Global Financial Stability Report

World Economic and Financial Surveys

Transition Challenges to Stability

October 2013

International Monetary Fund

Cataloging-in-Publication Data

Joint Bank-Fund Library

Global financial stability report – Washington, DC : International Monetary Fund, 2002– v. ; cm. – (World economic and financial surveys, 0258-7440)

Semiannual Some issues also have thematic titles. ISSN 1729-701X

1. Capital market — Development countries — Periodicals. 2. International finance — Periodicals. 3. Economic stabilization — Periodicals. I. International Monetary Fund. II. Series: World economic and financial surveys. HG4523.G563

ISBN 978-1-47552-497-0 (paper) 978-1-48431-915-4 (ePub) 978-1-48432-364-9 (Mobipocket) 978-1-48438-396-4 (PDF)

Please send orders to: International Monetary Fund, Publications Services P.O. Box 92780, Washington, DC 20090, U.S.A. Tel.: (202) 623-7430 Fax: (202) 623-7201 E-mail: [email protected] www.imfbookstore.org www.elibrary.imf.org

©International Monetary Fund. Not for Redistribution Contents

Assumptions and Conventions viii Preface ix executive summary xi Chapter 1 Making the transition to stability 1 Financial Stability Overview 1 Challenges Related to Accommodative Monetary Policies Will Test Markets and Policymakers 4 Box 1.1. Mortgage Real Estate Investment Trusts: Business Model Risks 12 Box 1.2. First-Time Issuers: New Opportunities and Emerging Risks 17 The Euro Area Banking, Corporate, and Sovereign Nexus 31 Global Banking Challenges: Profitability, Asset Quality, and Leverage 41 Box 1.3. Financial Regulatory Reform Update 47 Box 1.4. Recent Financial Sector Assessment Program Mission Findings 49 Making the Transition to Stability 50 Annex 1.1. Exploring the Factors Driving Bank Interest Rates on Corporate Loans 53 Annex 1.2. Euro Area Corporate Debt Overhang and Implications for Bank Asset Quality 57 Box 1.5. The GFSR Analysis of CorporateCredit Quality versus Bank Stress Tests 61 References 62 Chapter 2 Assessing Policies to Revive Credit Markets 63 Summary 63 Introduction 63 Recent Developments in Credit Markets 63 What Policies Have Been Implemented to Support Credit? 67 Box 2.1. Policies to Diversify Credit Options for Small and Medium Enterprises in Europe 70 Are Current Policies on Target? 71 Box 2.2. Challenges in the Structural Estimation of Credit Supply and Demand 76 Box 2.3. The Effect of the LiquidityC risis on Mortgage Lending 82 Designing Effective Policies for Reviving Credit Markets 83 Box 2.4. Policy Measures to Finance Small and Medium Enterprises during Crises: The Case ofKorea 84 Box 2.5. Lessons from the Nordic Banking Crises 87 Annex 2.1. Previous Findings in the Literature on Credit Constraints 90 Annex 2.2. Determinants of Bank Lending Standards 98 Annex 2.3. A Model of Bank Lending 98 References 101 Chapter 3 Changes in Bank Funding Patterns and Financial stability Risks 105 Summary 105 Introduction 106 Bank Funding Structures: Determinants and Recent Developments 107 Box 3.1. Typology of Bank Funding 108

International Monetary Fund | October 2013 iii

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 3.2. What the Crisis Taught Us about Bank Funding 113 Are Bank Funding Structures Relevant to Financial Stability? 114 Box 3.3. Changes in Cross-Border Bank Funding Sources 117 Crisis and the Impact of Regulatory Reforms on the Pricing of Bank Liabilities 119 Box 3.4. Bank Funding in Emerging Market Economies and the Impact of Regulatory Reforms 121 Box 3.5. Investor Base for Bail-in Debt and Bank Bond Ratings 126 Summary and Policy Recommendations 131 Annex 3.1. Data Description and Additional Empirical Results 134 Annex 3.2. Regulatory Developments AffectingBank Funding 139 Annex 3.3. Bank Bond Pricing Model 143 References 147 Glossary 149 Annex: IMF executive Board Discussion summary 157 statistical Appendix 159 Figures 1. Major Net Exporters and Importers of Capital in 2012 161 2. Sovereign Credit Default Swap Spreads 162 3. Selected Credit Default Swap Spreads 163 4. Selected Spreads 164 5. Implied Volatility Indices 165 6. United States: Corporate Bond Market 166 7. Euro Area: Corporate Bond Market 167 8. United States: Commercial Paper Market 168 Tables 1. Capital Market Size: Selected Indicators, 2012 169 2. MSCI Equity Market Indices 170 3. Emerging Markets Bond Index: EMBI Global Sovereign Yield Spreads 172 4. Emerging Market Private External Financing: Total Bonds, Equities, and Loans 174 5. Emerging Market Private External Financing: Bonds 177 6. Emerging Market Private External Financing: Equities 179 7. Emerging Market Private External Financing: Loans 181 8. Equity Valuation Measures: Dividend-Yield Ratios 184 9. Equity Valuation Measures: Price/Earnings Ratios 185 10. Emerging Markets: Mutual Funds 186 tables 1.1. Market-Implied Interest Rate Pricing versus Historical Cycles 6 1.2. Bond Portfolio Interest Rate Sensitivities 9 1.3. Structure of the Japanese Government Bond Market 28 1.4. Foreign Assets Held by Japanese Investor Groups, end-2012 28 1.5. Japan Scenarios: Complete, Incomplete, and Disorderly 29 1.6. European Union Bank Deleveraging 46 1.7. Policy Recommendations 51 1.8. Determinants of Bank Interest Rates on New Small Loans 56 1.9. Amadeus Database, 2011 57 1.10. Mapping of Corporate Vulnerability Indicators to Probabilities of Default 60 1.11. Comparison of GFSR Analysis with Oliver Wyman’s Stress Tests for Spain 61 1.3.1. Comparison of Bank Regulations across Jurisdictions 47 2.1. Identifying Countries with Weak Credit Growth, BIS Data 65 2.2. Identifying Countries with Weak Credit, Other Data Sources 66

iv International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution Contents

2.3. Credit Policies Implemented since 2007 72 2.4. Determinants of Credit Growth 73 2.5. Structural Determinants of the Supply and Demand of Bank Lending to Firms in Selected Countries 77 2.6. Firm-Level Regressions of Changes in Debt-to-Asset Ratio for Manufacturing Firms 78 2.7. Previous Findings in the Literature 91 2.8. Euro Area: Determinants of Bank Lending Standards 97 2.9. United States: Determinants of Bank Lending Standards 97 3.1. Characteristics of Four G-SIBs in Simulation Exercise 131 3.2. Country and Bank Coverage Statistics 135 3.3. List of Variables Used in the Panel Data Analysis 136 3.4. Illustration of Creditor Hierarchy and Loss Sharing under Alternative Resolution Tools 141 3.5. Cross-Country Comparison of Covered Deposits, end-2010 148

Figures 1.1. Global Financial Stability Map 1 1.2. Global Financial Stability Map: Assessment of Risks and Conditions 2 1.3. Market Volatility Shock 3 1.4. Global Financial Conditions 3 1.5. Market Dashboard 4 1.6. U.S. NonfinancialF irms’ Credit Fundamentals 5 1.7. Decomposition of Term Premium: Change From September 2008 to March 2013 7 1.8. Simulated Shock to 10-Year U.S. Treasury Term Premium 7 1.9. U.S. Mutual Fund Cumulative Flows 8 1.10. Global Bond Portfolio Duration 8 1.11. Nongovernment Bond Inventories, Total Trading Volumes, and Outstanding Bonds 9 1.12. Fire Sale “Risk Spiral” 10 1.13. Estimated Average Change in Mortgage Real Estate Investment Trust Portfolio Value for Parallel Interest Rate Shifts 11 1.14. Leveraged Mortgage Real Estate Investment Trusts Are More Vulnerable to Interest Rate Increases 11 1.15. Above-Trend Bond Flows from Advanced to Emerging Market Economies 15 1.16. Net Portfolio Flows into Fixed-Income Equity Markets by Country, 2009–12 16 1.17. Impact of Portfolio Flows on Local Currency Bond Yields 16 1.18. Allocation of Major Bond Funds to Emerging Markets 16 1.19. Duration of Emerging Market Fixed-Income Indices 20 1.20. Share of Nonresident Holdings of Local Currency Government Debt and Market Liquidity 20 1.21. Composition of the Holders of Local Currency Government Debt 20 1.22. Net New Issuance of Emerging Market Bonds 20 1.23. Credit Ratings of Emerging Market Corporate Bond Issues 21 1.24. Corporate Rating Changes in Emerging Markets 21 1.25. Nonfinancial CorporateB alance Sheet Metrics 22 1.26. Sharp Increase in Corporate Debt Defaults 22 1.27. China: Corporate Sector Fundamentals 23 1.28. External and Domestic Vulnerabilities 23 1.29. China: Credit Developments 23 1.30. Recent Stress in Emerging Markets 24 1.31. May 2013 Sell-Off ofE merging Market Bonds versus the Lehman Brothers Episode 25 1.32. Estimated Impact on Bond Yields from a Reversal of Capital Flows and Other Factors 25 1.33. Japanese Flows into Currency Overlay Funds 28 1.34. Japanese Banks’ Sensitivity to an Interest Rate Shock 30 1.35. Value-at-Risk in the Disorderly Scenario 30 1.36. Projected Flows to Selected Emerging Markets in the Disorderly Scenario 31 1.37. Bank-Corporate-Sovereign Nexus 32

International Monetary Fund | October 2013 v

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

1.38. Stressed Euro Area Economy Bank Credit 33 1.39. Bank Buffers andI nterest Rates on Corporate Loans 33 1.40. Individual Bank Buffers and Lending inStr essed Economies, 2013:Q1 33 1.41. Euro Area Sovereign Spreads, April–August 2013 33 1.42. Leverage Ratios 34 1.43. Leverage Ratios by Firm Size, 2011 34 1.44. Share of Debt at Firms with Various Debt-to-Assets Ratios, 2011 34 1.45. Share of Debt at Firms with Various Interest Coverage Ratios, 2011 34 1.46. Nonperforming Corporate Loans 35 1.47. Corporate Expected Default Frequencies and Bank Loan Interest Rates 35 1.48. Listed Firms: Changes in Debt, Capital Expenditures, and Dividends 36 1.49. Factors AffectingB ank Interest Rates on Corporate Loans 36 1.50. Interest Rates on Small Bank Loans and Model-Based Factor Decomposition 37 1.51. Corporate Debt Overhang 38 1.52. Distribution of Estimated Corporate Sector Probabilities of Default 39 1.53. Potential Losses on Corporate Loans and Banking System Buffers 40 1.54. Large Bank Tier 1 Ratios 43 1.55. Large Bank Tangible Leverage Ratios, 2012:Q4 43 1.56. Large Bank Tier 1 Capital and Tangible Leverage Ratios 44 1.57. Bank Profitability Comparison 44 1.58. Bank Profitability andMar ket Valuation of Assets 45 1.59. Large Bank Capitalization 46 1.60. Normalization of Monetary Policy: Smooth or Turbulent? 51 1.61. France: Deviations from Cointegrating Equilibrium 54 1.62. Spain and Italy: Deviations from the Cointegrating Equilibrium 55 1.63. Leverage, Profitability, and Debt at Risk 58 1.64. Bank Lending Rates to Small and Medium Enterprises 58 1.65. Projected Corporate Debt Overhang in Italy, Portugal, and Spain 59 1.66. Probabilities of Default in the Corporate Sector 60 1.1.1. Example of the Real Estate Investment Trust Maturity Transformation Process 12 1.1.2. Holdings of Agency Mortgage-Backed Securities 13 1.1.3. Real Estate Investment Trust Dependence on Short-Term Financing 13 1.1.4. REITs’ Agency MBS Holdings versus GSEs’ MBS Investment Portfolio Holdings 13 1.2.1. International Bond Debut Issuance 17 1.2.2. Investor Base by Region and Type 17 1.2.3. Performance of Frontier Markets during Emerging Market Bond Sell-Off 18 1.2.4. Size of Selected Frontier Market International Bond Issuance 18 2.1. Real Credit Growth 64 2.2. Perceived Obstacles in Access to Finance 66 2.3. Interest Rate Spread between Loans to SMEs and to Larger Firms 67 2.4. Corporate and Household Debt Outstanding 68 2.5. Stock Price Index 68 2.6. Real House Price Index 68 2.7. Relative Number of Credit Supply and Demand Policies Currently in Place 69 2.8. Decomposing Credit Growth: Corporate Loans 74 2.9. Decomposing Credit Growth: Mortgage Loans 75 2.10. Decomposition of Change in Debt-to-Asset Ratios for Firms 79 2.11. Real Total Credit Growth, by Borrowing Sector 79 2.12. Decomposing Lending Standards: Corporate Loans 95 2.13. Decomposing Lending Standards: Mortgage Loans 96 2.14. Effects of a Tightening of Lending Supply and a Drop in Lending Demand 98 2.15. Fitted Supply and Demand Curves for Bank Loans to Firms 99 2.3.1. U.S. Banks’ Core Deposits-to-Assets Ratio 82

vi International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution Contents

2.4.1. Outstanding Balance and Growth of SME Loans 84 2.4.2. Financial Support Programs for SMEs 84 2.4.3. Outstanding Balance and Growth of Credit Guarantees for SME Loans 85 2.4.4. Aggregate Credit Ceiling Loans 85 2.4.5. Number and Growth of Bankrupt Enterprises 86 2.4.6 Growth in Number of Employees 86 2.5.1. Real House Price Index in the Nordic Countries 87 2.5.2. Lending Growth by Banks 87 3.1. Banks’ Liability Structures across Major Economies and Regions 107 3.2. Banks’ Long-Term Wholesale Funding across Major Economies and Regions: Outstanding as of July 31, 2013 109 3.3. Evolution of Bank Funding Structures, Global and Systemically Important Banks 110 3.4. Evolution of Bank Funding Structures, Advanced and Emerging Market Economies 110 3.5. Determinants of Bank Funding 112 3.6. Wholesale Bank Funding 114 3.7. Regulatory Capital Ratios across Major Economies and Regions 115 3.8. Asset Encumbrance: December 2007 and June 2013 115 3.9. Share of Retained Bank-Covered Bonds in Europe 116 3.10. Contribution of Funding Characteristics to Bank Distress 116 3.11. Evolution of Bank Funding Characteristics 119 3.12. Priority of Claims of Bank Liabilities 120 3.13. Bank Bond Yield to Maturity at Issuance: Selected Advanced Economies 127 3.14. Debt Pricing under Depositor Preference and Asset Encumbrance 129 3.15. Debt Pricing under Bail-in Power 130 3.16. Simulation Results for SpecificB anks 131 3.17. Equity Value for Original Shareholders under Bail-in Regime Converting Debt to Equity 132 3.18. Contribution of SpecificV ariables to Bank Distress in Probit Models 138 3.19. Basel III Minimum Capital Requirements and Buffers 140 3.20. Pricing of Senior and Subordinated Debt and Equity 144 3.21. Pricing of Liabilities with Depositor Preference and Asset Encumbrance 145 3.22. Pricing of Liabilities under Bail-in Power 146 3.1.1. Breakdown of Bank Liabilities 108 3.3.1. Euro Area: Foreign Holding of Bank Debt Securities 117 3.3.2 Non-Euro Area: Foreign Holding of Bank Debt Securities 117 3.3.3. U.S. Money Market Fund Exposure to European and Other Banks 118

International Monetary Fund | October 2013 vii

©International Monetary Fund. Not for Redistribution AssuMPtIons AnD ConventIons

The following conventions are used throughout the Global Financial Stability Report (GFSR): . . . to indicate that data are not available or not applicable; – between years or months (for example, 2012–13 or January–June) to indicate the years or months covered, including the beginning and ending years or months; / between years or months (for example, 2012/13) to indicate a fiscal or financial year. “Billion” means a thousand million. “Trillion” means a thousand billion. “Basis points” refer to hundredths of 1 percentage point (for example, 25 basis points are equivalent to ¼ of 1 percentage point). If no source is listed on tables and figures, data are based on IMF staff estimates or calculations. Minor discrepancies between sums of constituent figures and totals shown reflect rounding. As used in this report, the terms “country” and “economy” do not in all cases refer to a territorial entity that is a state as understood by international law and practice. As used here, the term also covers some territorial entities that are not states but for which statistical data are maintained on a separate and independent basis.

Further Information and Data This version of the GFSR is available in full through the IMF eLibrary (www.elibrary.imf. org) and the IMF website (www.imf.org).

The data and analysis appearing in the GFSR are compiled by the IMF staff at the time of publication. Every effort is made to ensure, but not guarantee, their timeliness, accuracy, and completeness. When errors are discovered, there is a concerted effort to correct them as appropriate and feasible. Corrections and revisions made after publication are incorporated into the electronic editions available from the IMF eLibrary (www.elibrary.imf.org) and on the IMF website (www.imf.org). All substantive changes are listed in detail in the online tables of contents.

For details on the terms and conditions for usage of the contents of this publication, please refer to the IMF Copy- right and Usage website, www.imf.org/external/terms.htm.

viii International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution PReFACe

The Global Financial Stability Report (GFSR) assesses key risks facing the global financial system. In normal times, the report seeks to play a role in preventing crises by highlighting policies that may mitigate systemic risks, thereby contributing to global financial stability and the sustained economic growth of the IMF’s member countries. The global financial system is now undergoing a series of transitions along the path to financial stability that have led to an uptick in liquidity and emerging market risks. For policymakers, a major challenge is to respond to the increase in market volatility associated with expectations for an eventual withdrawal from unconventional monetary policy in advanced economies. The current report analyzes risks to financial stability associated with the buildup of pockets of leverage in advanced and emerging economies during the extended period of monetary accommodation, and assesses policies to minimize those risks during the transition. The report also examines the potential reasons behind the weak developments in private credit since 2008 in some advanced economies and offers a framework for determining the types of policies that may be best suited to address this issue. Lastly, the report looks at bank funding structures, identifying how they have changed over time and how they affect financial stability. The report notes that the recent reform efforts have potentially different implications for funding structures leading to some tension among them. The analysis in this report has been coordinated by the Monetary and Capital Markets (MCM) Department under the general direction of José Viñals, Financial Counsellor and Director. The project has been directed by Jan Brockmeijer and Peter Dattels, both Deputy Directors; Laura Kodres, Assistant Director; and Matthew Jones, Divi- sion Chief. It has benefited from comments and suggestions from the senior staff in the MCM department. Individual contributors to the report are Ruchir Agarwal, Sergei Antoshin, Nicolas Arregui, Serkan Arslanalp, Jorge A. Chan-Lau, Yingyuan Chen, Julian Chow, Nehad Chowdhury, Jihad Dagher, Reinout De Bock, Marc Dobler, Johannes Ehrentraud, Martin Edmonds, Xiangming Fang, Brenda González-Hermosillo, David Grigorian, Tryggvi Gudmundsson, Sanjay Hazarika, Anna Ilyina, Heedon Kang, William Kerry, Yitae Kim, Koralai Kirabaeva, Fred- eric Lambert, Samar Maziad, Rebecca McCaughrin, Peter Lindner, André Meier, Paul Mills, S. Erik Oppers, Nada Oulidi, Hiroko Oura, Evan Papageorgiou, Vladimir Pillonca, Alvaro Piris Chavarri, Jean Portier, Jaume Puig, Narayan Suryakumar, Takahiro Tsuda, Kenichi Ueda, Nico Valckx, Chris Walker, and Mamoru Yanase. Mustafa Jamal, Oksana Khadarina, and Yoon Sook Kim provided analytical support. Juan Rigat, Adriana Rota, and Ramanjeet Singh were responsible for word processing. Eugenio Cerutti, Ali Sharifkhani, and Hui Tong provided database and program- ming support. Additional input was received from Nicolas Blancher, Ana Carvajal, Karl Driessen, Jennifer Elliott, Michaela Erbenova, Ellen Gaston, Anastasia Guscina, Javier Hamann, Eija Holttinen, Bradley Jones, Emanuel Kopp, Fabiana Melo, Mala Nag, Oana Nedelescu, Guilherme Pedras, Gabriel Presciuttini, Amadou Sy, Hideyuki Tanimoto, Constant Verkoren, and Jianping Zhou. Linda Griffin Kean and Joe Procopio from the Communications Department edited the manuscript and coordinated production of the publication with assistance from Lucy Scott Morales and Linda Long. This particular issue draws in part on a series of discussions with banks, clearing organizations, securities firms, asset management companies, hedge funds, standards setters, financial consultants, pension funds, central banks, national treasuries, and academic researchers. The report reflects information available up to September 26, 2013.The report benefited from comments and- sug gestions from staff in other IMF departments, as well as from Executive Directors following their discussion of the Global Financial Stability Report on September 23, 2013. However, the analysis and policy considerations are those of the contributing staff and should not be attributed to the Executive Directors, their national authorities, or the IMF.

International Monetary Fund | October 2013 ix

©International Monetary Fund. Not for Redistribution This page intentionally left blank

©International Monetary Fund. Not for Redistribution eXeCutIve suMMARY

he global financial system is undergoing a foreign investors have crowded into local markets and series of transitions along the path toward may withdraw. Emerging market fundamentals have greater financial stability. The United States weakened in recent years, after a protracted interval of may soon move to less accommodative mon- credit expansion and rising corporate leverage. Manag- Tetary policies and higher long-term interest rates as ing the risks of the transition to a more balanced and its recovery gains ground. After a prolonged period of sustainable financial sector, while maintaining robust strong portfolio inflows, emerging markets are facing growth and financial stability, will be a key undertak- a transition to more volatile external conditions and ing confronting policymakers. higher risk premiums. Some need to address financial As central banks elsewhere consider strategies for and macroeconomic vulnerabilities and bolster resil- eventual exit from unconventional monetary poli- ience, as they shift to a regime in which financial sector cies, Japan is scaling up monetary stimulus under the growth is more balanced and sustainable. Japan is mov- Abenomics framework, aiming to pull the economy ing toward the new “Abenomics” policy regime marked out of its deflationary rut. Successful implementation by more vigorous monetary easing, coupled with fiscal of a complete policy package that features fiscal and and structural reforms. The euro area is moving toward structural reforms would reinforce domestic financial a more robust and safer financial sector, including a stability, while likely boosting capital outflows. But sub- stronger monetary union with a common framework stantial risks to financial stability could accompany the for risk mitigation, while strengthening financial program if planned fiscal and structural reforms are not systems and reducing excessive debt levels. Finally, the fully implemented. Failure to enact these reforms could global banking system is phasing in stronger regulatory lead to a return of deflation and increased bank holdings standards. Chapter 1 examines the challenges and risks of government debt, further increasing the already-high of each of these transitions. sovereign-bank nexus. In a more disorderly scenario, The primary challenge resulting from these changes with higher inflation and elevated risk premiums, the relates to managing the side effects and eventual risks to both domestic and global financial stability withdrawal of accommodative monetary policy in the could be greater still, including rapid rises in bond yields United States. Such a transition, including the benefits and volatility, and sharp increases in outflows. of a strong U.S. economy, should help limit financial In the euro area, reforms implemented at the stability risks associated with an extended period of national level and important steps taken toward low interest rates. Yet managing a smooth transition improving the architecture of the monetary union could prove challenging, as investors adjust portfolios have helped reduce funding pressures on banks and for a new regime with higher interest rates and greater sovereigns. However, in the stressed economies of Italy, volatility. The analysis in Chapter 1 highlights the risk Portugal, and Spain, heavy corporate sector debt loads that long-term interest rates could rise more sharply and financial fragmentation remain challenging. Even than currently anticipated. Structural reductions in if financial fragmentation is reversed over the medium market liquidity and leveraged positions in short-term term, this report estimates that a persistent debt over- funding markets and the shadow banking system (for hang would remain, amounting to almost one-fifth of instance, in the mortgage real estate investment trust the combined corporate debt of Italy, Portugal, and sector) could amplify these rate increases and spill over Spain. Assuming no further improvement in economic to global markets. and financial conditions which would correspond to a Financial stability challenges are also prevalent in more adverse outcome than the cyclical improvement many emerging market economies. Bond markets built into the October 2013 World Economic Outlook are now more sensitive to changes in accommoda- baseline scenario, some banks in these economies tive monetary policies in advanced economies because might need to further increase provisioning to address

International Monetary Fund | October 2013 xi

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

the potential deterioration in asset quality of corporate • Containing the risks to China’s financial system is as loan books. This could absorb a large portion of future important as it is challenging. Broad credit growth bank profits. Recent efforts to assess asset quality and needs to be reined in to contain financial stability boost provisions and capital have helped to increase the risks and promote the rebalancing of China’s econ- loss-absorption capacity of banks, but further efforts omy away from credit-fueled capital and property to cleanse bank balance sheets and to move to full investment. It is important that prudential oversight banking union are vital. These steps should be comple- of shadow banking activity be tightened, that incen- mented by a comprehensive assessment and strategy to tives for regulatory arbitrage be removed through address the debt overhang in nonfinancial companies. continued financial liberalization (for example, of A number of policy actions can help promote an deposit rates), and that the widespread belief in orderly move toward greater financial stability: implicit guarantees and bailouts of risky corporate • Stronger growth in the United States is setting loans and saving products be counteracted. Unless the stage for monetary normalization. Achieving credit losses are taken by lenders and savers, the state a smooth transition requires policies that man- faces large and unpredictable fiscal costs. age the effects of increased volatility and portfolio • Japan’s bold policies need to be completed, with the adjustments, while addressing structural liquidity authorities following through on fiscal and structural weaknesses and systemic vulnerabilities. A clear and policy commitments, to avoid downside risks. These well-timed communication strategy by central bank policies are needed to contain a potential sharp rise officials is critical. Compared with previous tighten- in government bond risk premiums if sovereign debt ing cycles, the authorities have a bigger toolkit at their dynamics do not improve. To help mitigate stability disposal. Yet in the event of adverse shocks, contin- risks, market structures also need to be made more gency backstops may be needed to address the risk resilient (such as through the modification of circuit of fire sales in some market segments and to manage breakers in derivatives markets) and the risk profile orderly unwinding or liquidation. Increased oversight of regional banks addressed. would help reduce related risks of excessive leverage in • In the euro area, further progress in reducing debt the shadow banking system and, in particular, in the overhangs and bolstering bank balance sheets needs larger mortgage real estate investment trusts. to go hand in hand with a strengthened euro area • For emerging market economies, the principal trans- financial architecture and the completion of the mission channel of external pressures is more likely to banking union agenda. Investors’ faith in euro area be via liquidity strains in bond and foreign exchange bank balance sheets needs to be fully restored and markets, rather than through bank funding chan- credit flows to viable enterprises strengthened: a nels. In addition, the response of foreign investors first step, as planned, is to conduct a thorough, to changing expectations for U.S. monetary policy realistic, and transparent balance sheet assessment. will continue to affect local markets. In the event of Credible capital backstops to meet any identified significant capital outflows, some countries may need shortfalls need to be put in place and communicated to focus on ensuring orderly market functioning, in advance of the publication of results from the using their policy buffers wisely. Keeping emerging exercise. The corporate debt overhang should be market economies resilient requires increased focus on addressed using a more comprehensive approach, domestic vulnerabilities as relative growth prospects including corporate debt cleanups, improvements moderate, U.S. nominal rates rise, and capital flows to corporate bankruptcy frameworks, and active recede. Policymakers should carefully monitor and facilitation of nonbank sources of credit. Further contain the rapid growth of corporate leverage. monetary support by the European Central Bank Local bank regulators need to guard against foreign and credit support to viable firms by the European currency funding mismatches affecting bank balance Investment Bank are crucial to providing time for sheets, including through foreign currency borrowing the repair of private balance sheets. by companies. In addition, establishing sufficient buf- • Global bank capitalization remains diverse, because fers and addressing macroeconomic imbalances will institutions are at different stages of balance sheet likely prove to be worthwhile steps for cushioning repair and operate in different economic and regula- against increased volatility and risk premiums. tory environments. The key tasks are to improve

xii International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ExEcutIvE suMMary

the credibility, transparency, and strength of balance writing standards and the risk of “evergreening” existing sheets, while avoiding undue pressures on banks loans. Mitigation of these risks may not be necessary or from uncoordinated national regulatory initiatives appropriate while the economic recovery is still weak and uncertainty. Further efforts are needed to assess because it could run counter to the objectives of the the way in which market developments and regula- credit policies (which are often designed to increase tory initiatives affecting dealer-bank business models risk taking); still, policymakers will need to continually may influence the cost and provision of market weigh the near-term benefits against the longer-run costs liquidity. At a minimum, increased surveillance of of policies aimed at boosting credit. and vigilance over the effects of trading liquidity Chapter 3 explores how bank funding structures pressures will be needed as financial markets move affect financial stability and whether regulatory reform to a regime with higher interest rates and volatility. initiatives are likely to make them more stable, diversi- If these policy challenges are properly managed, and fied, and resilient. The chapter finds that healthy if reforms are implemented as promised, the transition banks rely more on equity and less on debt—especially toward greater financial stability should prove smooth short-term wholesale funding that contributed to and provide a more robust platform for financial sector the global financial crisis—and use deposits as their activity and economic growth. But a failure to imple- primary funding source. Various reforms are rightly ment the reforms necessary to address the many policy promoting many of these desirable attributes, but there challenges highlighted above could trigger profound could be potential trade-offs among them. On the one spillovers across regions and potentially derail the hand, there are pressures to use more secured fund- smooth transition to greater stability. ing—which increases asset encumbrance—as well as Chapter 2 looks at efforts by policymakers to revive deposits, to reduce banks’ vulnerability to turbulence weak credit growth, which has been seen by many as a in wholesale funding markets. On the other hand, primary reason behind the slow economic recovery. The bail-in power and depositor preference give better pro- chapter catalogues the policies implemented by vari- tection to taxpayers and depositors at the expense of ous countries and offers a framework for assessing their unsecured wholesale debt holders. A numerical analysis effectiveness. It argues that policies are most effective if illustrates the impact on the cost of unsecured debt they target the constraints that underlie the weakness in as the proportion of newly protected creditors rises. credit. Using several analytical tools, the chapter finds Under current conditions and depositor protections that the constraints in credit markets differ by country (and especially for well-capitalized banks) the increase and evolve over time, requiring a careful country-by- would be modest; however, if depositor protections country assessment. Better data on new lending would were to be expanded substantially, the impact could also help identify constraints. In many cases, demand- be quite large. Careful implementation of the reforms and supply-oriented policies are complementary, but can moderate tensions: Basel III and over-the-counter their relative magnitude and sequencing can be impor- derivative reforms should be implemented as planned, tant. Moreover, policymakers should be cognizant of but policymakers should monitor the increased the fiscal costs and implications for financial stability demand for collateral and ensure that enough unen- of credit-supporting policies. The main risks center on cumbered assets are available to permit the meaningful increased credit risk, including a relaxation of under- bail-in of unsecured senior creditors.

International Monetary Fund | October 2013 xiii

©International Monetary Fund. Not for Redistribution This page intentionally left blank

©International Monetary Fund. Not for Redistribution chapter 1 Making the tranSitiOn tO Stability

Financial stability risks are in transition. Although prospects Financial Stability Overview for U.S. growth are solidifying, market and liquidity risks The Global Financial Stability Map indicates that risks have risen. Expectations of reduced monetary accom- are in transition (Figures 1.1 and 1.2). modation in the United States may cause further global Macroeconomic risks remain unchanged overall, with market adjustments and expose areas of financial excess global activity expected to strengthen moderately, though and systemic vulnerability. Emerging markets face tighter with forecast risks remaining to the downside, as discussed financial conditions as they cope with weaker economic out- in the October 2013 World Economic Outlook (WEO). looks and rising domestic vulnerabilities. In the euro area, European recovery has been tepid, and growth in an increas- further progress has been made toward banking union, but ing number of emerging market economies is slowing. At the outlook remains clouded by the unfinished business of the same time, the U.S. recovery is gaining ground, which restoring bank health and credit transmission and reduc- is positive for global growth, but is also leading markets to ing the corporate debt overhang. Japan’s bold policies hold price in an earlier tightening of U.S. financial conditions. hope for reinvigorating growth and ending corrosive debt Thus, the process of normalization of global asset allocations deflation dynamics, but implementation challenges are large has begun, pushing up interest rates and risk premiums as and halfway policies would pose serious downside risks. markets shift away from a regime of suppressed market volatility and very favorable liquidity conditions. These changes Prepared by Peter Dattels and Matthew Jones (team leaders), Sergei Antoshin, Yingyuan Chen, Julian Chow, Nehad Chow- are creating a host of new challenges for financial stability, dhury, Reinout De Bock, Martin Edmonds, Xiangming Fang, leading to higher market and liquidity risks. Sanjay Hazarika, Anna Ilyina, William Kerry, Koralai Kirabaeva, Developments since late May 2013 have brought about Rebecca McCaughrin, André Meier, Paul Mills, Nada Oulidi, Evan Papageorgiou, Vladimir Pillonca, Jean Portier, Jaume Puig, Narayan a “mini stress test” in the form of a global volatility shock, Suryakumar, and Chris Walker. uncovering some important channels of potential financial

Figure 1.1. Global Financial Stability Map

Risks Emerging market risks Credit risks

Oct. 2013 GFSR April 2013 GFSR

Market and Macroeconomic risks liquidity risks

Away from center signiﬁes higher risks, easier monetary and ﬁnancial conditions, or higher risk appetite.

Monetary and ﬁnancial Risk appetite Conditions

Source: IMF staff estimates.

International Monetary Fund | October 2013 1

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: Transition Challenges to Stability

Figure 1.2. Global Financial Stability Map: Assessment of Risks and Conditions (Notch changes since the April 2013 GFSR)

Macroeconomic risks remain unchanged, but global activity has become Emerging market risks have increased as the result of weaker more uneven and is projected to expand only modestly in 2014. growth prospects and rising domestic and external vulnerabilities. 1. 2. 2 2 More risk

1 1 More risk

0 0 Unchanged

–1 Less risk –1 Less risk –2 –2 Overall (9) Sovereign credit Inﬂation Economic activity Overall (6) Fundamentals Inﬂation (1)Corporate Liquidity (2) (1) variability (1) (7) (2) sector (1)

Market and liquidity risks have increased as markets adjust to prospects of Risk appetite has contracted, resulting in reversals of capital ﬂows to reduced monetary accommodation with implications for asset prices. emerging markets. 3. 4. 4 4 Higher risk appetite 3 3 2

2 1 0 More risk 1 –1 –2 0 Lower risk appetite –3 –4 –1 Less risk –5 –2 –6 Overall (6) Liquidity and Volatility (2) Market Equity Overall (4) Institutional Investor Relative asset Emerging funding (1) positioning valuations (1) allocations (1) surveys returns (1) markets (1) (2) (1)

Monetary and ﬁnancial conditions remain broadly accommodative, as lending Credit risks are broadly unchanged, reﬂecting the uneven progress in conditions have improved, but emerging market risk premiums have risen. balance sheet repair and pressures on euro area banks. 2 5. 6. 2

More risk 1 Easier 1

0 0 Unchanged Unchanged

–1 –1 Tighter Less risk

–2 –2 Overall (6) Monetary Financial Financial Lending QE and CB Overall (8)Banking Banking Household Corporate policy conditions in conditions conditions balance sector, Europe sector, non- sector (2) sector (3) conditions advanced in (1) sheet (1) Europe (2) (3) markets emerging expansion (0.65) markets (1) (0.35)

Source: IMF staff estimates. Note: Changes in risks and conditions are based on a range of indicators, complemented with IMF staff estimates (see Annex 1.1 in the April 2010 GFSR and Dattels and others, 2010, for a description of the methodology underlying the Global Financial Stability Map). Overall notch changes are the simple average of notch changes in individual indicators. The number next to each legend indicates the number of individual indicators within each subcategory of risks and conditions. For lending standards, positive values represent a slower pace of tightening or faster easing. CB = central bank; QE = quantitative easing.

2 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 Making the Transition to Stability

fragility. A substantial increase in volatility occurred, espe- (Figure 1.3). Market conditions have subsequently calmed, cially through the interest rate channel, as monetary policy but transition challenges remain. At the time of writing, a expectations reset and strongly affected emerging markets political standoff in the United States has led to a shutdown of its federal government. The analysis in this report Figure 1.3. Market Volatility Shock assumes that the shutdown is short, discretionary public (Index) spending is approved and executed as assumed in the forecast, and the debt ceiling––which may be reached by United Emerging Euro area Japan States markets mid-October––is raised promptly. There is uncertainty on 400 all three accounts. While the damage to the U.S. economy from a short shutdown is likely to be limited, a longer 350 shutdown could be quite harmful. And, even more impor- 300 tantly, a failure to promptly raise the debt ceiling, leading

2007 = 100) to a U.S. selective default, could seriously damage the

250 y 1, global economy and financial system. Although monetary and financial conditions overall remain accommodative, 200 risk premiums in emerging markets have risen, tightening 150 financial conditions in those markets (Figure 1.4). Against

olatility index (Januar this backdrop, emerging market risks have increased because V 100 of weaker growth prospects coupled with less accommoda-

50 tive external conditions and more worries about domestic and external vulnerabilities. Risk appetite has fallen, resulting 0 y y y y in some outflows from emerging market funds. Bond Bond Bond Bond Equity Equity Equity Equity Credit risks remain broadly unchanged, reflecting Currenc Currenc Currenc Currenc insufficient balance sheet repair and slow progress in August 8, 2013 addressing the lingering risks that materialized as a result January 2, 2013 of the crisis. The subdued outlook in Europe and chal- Postcrisis average (April 1, 2009–May 22, 2013) lenges in bank asset quality and capital continue to keep Sources: Bloomberg, L.P.; and IMF staff estimates. credit risks elevated, and this has been compounded Note: The historical volatilities are computed using a rolling 60-day standard deviation of by the problems posed by debt-burdened companies, index returns, which are then indexed with January 1, 2007, as the reference point. further undermining the prospects of a recovery. This chapter examines prospects for and risks to global Figure 1.4. Global Financial Conditions financial stability. The next section asks whether the (Index; normalized, two-week moving averages) prospect of tighter financial conditions in the United Emerging markets 1.0 States will result in a smooth normalization of financial Europe markets and portfolio allocations, or whether markets will Japan May 22 become turbulent and financial stability risks will arise. United States 0.5 How will emerging markets be affected by changes in advanced economy monetary policies and asset allocations? 0.0 Do domestic risks in emerging markets themselves pose a threat? Will Japan’s bold policies be successful, and what are –0.5 the downside risks if policy commitments are not met? Tighter ﬁnancial conditions The task of addressing legacy risks from the global finan-

–1.0 cial crisis remains unfinished. The third section assesses these risks by focusing on the remaining challenges in the euro area. The analysis suggests that addressing the debt –1.5 overhang in the nonfinancial corporate sector is critical. If it is not addressed, bank health cannot be restored and the –2.0 sovereign-banking-corporate nexus will remain unbroken. Jan 2013 Feb 13 Mar 13 Apr 13 May 13 Jun 13 Jul 13 Aug 13 Sep 13 The fourth section examines developments in systemi- Source: Goldman Sachs Financial Conditions Index. cally important banks and the progress they have made in

International Monetary Fund | October 2013 3

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.5. Market Dashboard (Percentiles over the past three years)

September 26, 2013 May 22, 2013 Average as of September 26, 2013 Average as of May 22, 2013

100

0 S&P 500 U.S. (VIX) U.S. CDX IG (5y) CDX HY (5y) CDX EM (5y) Europe (V2X) iTraxx IG (5y) iTraxx Euro Stoxx 50 GBI-EM (yield) MSCI EM Equities USDJPY basis (2y) EMBI global (yield) EURUSD basis (2y) Euro Stoxx Financials USDJPY ATM vol (6m) ATM USDJPY USDBRL ATM vol (6m) ATM USDBRL EURUSD ATM vol (6m) ATM EURUSD EU 1y10y swaption vol Fwd libor-fwd OIS (3m) Fwd libor-fwd U.S. 1y10y swaption vol U.S. CRB Commodities Index Equity U.S. Treasuries (10y yield) Treasuries U.S. Japan 1y10y swaption vol

Credit asset swap spread (2y) U.S. volatility Fwd euribor-fwd EONIA (3m) Fwd euribor-fwd Equities Foreign exchange Bonds Funding and liquidity Interest rate volatility premiums volatility Sources: Bloomberg, L.P.; and IMF staff estimates. Note: Each marker corresponds to the percentile of the level of the asset in relation to its three-year history of levels. Fifty corresponds to the median during the period, zero corresponds to the level consistent with the highest risk aversion during the period, and 100 corresponds to the level consistent with the lowest risk aversion during the period. 10y = 10 years; 5y = 5 years; 2y = 2 years; 6m = 6 months; V2X = Dow Jones EURO STOXX 50 Volatility Index; VIX = Chicago Board Options Exchange Market Volatility Index; EM = emerging market; OIS = overnight index swap.

strengthening their balance sheets. The fifth section tackles Brothers collapse in September 2008, while evidence of key policies that can safeguard financial stability. slowing growth mounted. Markets came to question both the upside and the downside risks of Japan’s bold set of challenges related to accommodative quantitative and qualitative monetary easing policies, Monetary policies Will test Markets and reflected in rising market volatility observed in April and policymakers May 2013. Against this backdrop, this section explores the transition challenges from an end to accommodative Before the market correction that began in May 2013, monetary policies and describes how markets and policy- prices of many assets had risen to multi-year highs, makers could be tested. underpinned by three key expectations. First, quantitative easing in the United States was expected to be protracted. Second, U.S. economic prospects were expected to catch the United States: Uncertainties in Making the up to the buoyancy in markets. Third, low yields were transition to a new regime expected to persist alongside low volatility and rising asset Stronger growth in the United States is setting the stage prices. Starting in May, markets were rattled by shifts in for a start toward monetary normalization. From a the perceived regime (Figure 1.5). The Federal Reserve financial stability standpoint, such a transition should signaled that improvements in the U.S. economy could help limit risks associated with a prolonged period prompt a tapering of its asset purchase program before of low interest rates. Yet managing a smooth transi- the end of the year. Emerging markets faced sustained tion could prove challenging, with a key risk being the capital outflows for the first time since the Lehman potential for long-term interest rates to overshoot. A

4 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.6. U.S. Nonﬁnancial Firms’ Credit Fundamentals

Leverage has risen meaningfully as debt levels have grown and The trend has been broad based, with leverage rising among both EBITDA gains have slowed. low- and high-quality credit.

1. Leverage: Ratio of Investment-Grade and High-Yield 2. Share of Firms Exhibiting an Increase in Leverage 70 Gross Debt to EBITDA 130 1997:Q1 60 2000:Q4 120 2007:Q4 50 Current cycle (2011:Q4) 40 110 30 100

100 = start of cycle 20 Percent, year over Percent, 90 10

80 0 0 4 8 12 16 20 Dec. Dec. 99 Dec. Dec. 03 Dec. 05 Dec. 07 Dec. 09 Dec. 11 Quarters from start of leverage cycle 1997 2001

Liquidity conditions are deteriorating... …while underwriting standards continue to weaken.

20 3. Liquidity: Cash-to-Debt Ratio 300 4. High-Yield Covenant-Lite Loans 40 18 High yield Covenant-lite loan issuance (left scale) 35 Investment grade 250 Share of institutional leveraged loans (right scale) 16 High-yield long-term average 30 14 Investment grade long-term average 200 12 25 10 150 20 Percent

8 Percent 15 6 100 Billions of U.S. dollars Billions of U.S. 10 4 50 2 5 0 0 0 1997 99 2001 03 05 07 09 11 13 1997 99 2001 03 05 07 09 11 13 (annualized) Reﬁnancing risk is not an immediate concern because of low rates …but defaults are still on track to rise owing to past excesses and a and liability management… turn in the credit cycle.

300 5. High-Yield Debt Maturity Proﬁle 6. U.S. High-Yield Default Rate 16 Leveraged loans 14 250 High yield 12 200 10

150 8 Percent 6 100 Billions of U.S. dollars Billions of U.S. Baseline 4 50 2

0 0 2013 14 15 16 17 18 19 20 21 22 23 1997:Q4 2000:Q4 03:Q4 06:Q4 09:Q4 12:Q4 15:Q4

Sources: Deutsche Bank; Federal Reserve; Moody’s; Morgan Stanley; S&P Leveraged Commentary and Data; Thomson Reuters; and IMF staff estimates. Note: EBITDA = earnings before interest, taxes, depreciation, and amortization.

International Monetary Fund | October 2013 5

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

decline in structural market liquidity, coupled with Table 1.1. Market-Implied Interest Rate Pricing versus leveraged funding and mortgage structures, could Historical Cycles amplify market movements and lead to systemic finan- Start of Federal Total Hike Average Hike cial strains in the United States and across the globe. Reserve Rate Cycle Length Total Hikes in First Year per Month Hiking Cycle (months) (basis points) (basis points) (basis points) Jul-1958 19 342 246 18 The Federal Reserve has indicated that if the economic Jan-1962 59 425 150 7 Jun-1967 27 625 200 23 recovery continues as expected, it would be appropriate to Dec-1971 31 1,031 250 33 Dec-1976 49 1,783 236 36 begin scaling back its asset purchase program as a first step Apr-1983 17 288 188 17 toward phasing out monetary stimulus.1 Gradually making Dec-1986 30 394 94 13 Jan-1994 14 300 250 21 the transition to a higher interest rate regime should be Mar-2004 28 425 175 15 Median 28 425 200 18 positive for financial stability, because risks associated with Mar-2015 49 381 73 6 low rates and the accumulation of financial excesses will be Sources: Bloomberg, L.P.; Deutsche Bank; and IMF staff estimates. curtailed. This is especially critical given that some of these Note: March 2015 figures are projections. risks have continued to build, including the deterioration in corporate credit conditions (Figure 1.6), yield-seeking behavior among pension funds and insurers (see the April financial conditions and increased portfolio losses, poten- 2013 GFSR), and an extension in portfolio duration.2 tially aggravated by reduced market liquidity and forced Ideally, the normalization of interest rates and volatil- asset sales (particularly where leverage and maturity mis- ity would be orderly and unfold as follows: short-term matches are sizable), with spillover implications for broader interest rate expectations rise along a smooth, gentle global financial conditions.3 These developments could lead path, consistent with current market expectations; the to a bumpier transition and strain financial stability. term premium compression unwinds gradually; the portfolio adjustment response occurs smoothly, and Containing long-term rates and market volatility will credit valuations reprice modestly; pockets of balance be a key challenge. sheet leverage are unwound at a gradual pace, with limited knock-on effects; market liquidity is sufficient to Following the turbulence in May and June 2013, financial accommodate these adjustments; and all of these devel- markets shifted forward their expectations about the start opments occur in the context of an economy gathering of the tightening cycle in response to an anticipated scaling strength. The current WEO projections assume that the back in Federal Reserve asset purchases. Then at its Septem- latest tightening in financial conditions was largely a ber meeting, the Federal Reserve surprised markets by one-time event and that the actual tapering of purchases deciding to delay the start of its tapering process. Neverthe- will further tighten conditions only modestly. less, interest rate futures markets are still pricing in only a But a less-benign scenario is a distinct risk. The failure very gradual, modest tightening relative to the historical of any one or all of the elements outlined here could lead trend (Table 1.1). Although the actual path could ulti- to a more abrupt, sustained move in long-term interest mately prove to be sharper and swifter, the Federal Reserve rates and excess market volatility as prior accommodation has a number of tools to guide short-term rates. is reversed (IMF, 2013c). The shift in short-term interest In contrast, controlling long-term rates is more difficult. rate expectations and term premiums could be sharper and Various factors influence term premiums and long-term the cycle more frontloaded, leading to a rapid tightening in rates that are collectively more difficult for central banks to contain. To assess the potential trajectory of long-term 1 The Federal Reserve surprised markets in mid-September by rates, a term premium model is estimated based on changes voting not to scale back asset purchases at that time, but suggested that if the economy continued to recover as it expected, it would, in macroeconomic fundamentals, macroeconomic volatility, at subsequent meetings, assess incoming information to determine financial market volatility, market expectations about the when to moderate the pace of asset purchases. See IMF U.S. Article future interest rate path, and the size and persistence of the IV Consultation Report (IMF, 2013c). 2Both high-yield and investment-grade firms continue to relever Federal Reserve’s asset purchase program. as debt levels have risen and earnings growth has slowed. The leverage distribution has worsened, suggesting that the cycle is moving toward a later, less-healthy stage. Meanwhile, free cash flow and over- 3Box 1.1 in the October 2013 WEO finds that the external conse- all cash balances are diminishing, issuance quality has deteriorated, quences of an eventual tightening of U.S. monetary policy are more there is a more persistent willingness to accept weaker covenants, and damaging the faster the pace of the adjustment and the weaker the credit conditions have weakened further. external policy framework.

6 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

The model reveals a substantial and statistically signifi- Figure 1.7. Decomposition of Term Premium: cant effect of quantitative easing policies on long-term rates. Change from September 2008 to March 2013 (Basis points) The decline in the term premium accounts for roughly half of the compression in 10-year nominal Treasury bond 100 yields since late 2008, when quantitative easing policies were first announced. Decomposing the term premium 50 Residual Short-rate further into its individual components shows that market expectations Macro fundamentals expectations about the Federal Reserve’s balance sheet (for 0 Market example, the various asset purchase announcements and volatility/interest forward guidance), the reduction in market volatility, and –50 rate uncertainty lower interest rate uncertainty account for almost the entire decline in the term premium (Figure 1.7).4 –100 Expectations about the Federal Future shocks to market volatility and uncertainty about Change in term Reserve's balance asset purchases and forward guidance could have a pro- –150 premium sheet nounced impact on the term premium and thus on long- term rates. Figure 1.8 presents two simulation exercises –200

based on different assumptions about volatility and the Determinants of 10-year yields Determinants of term premium Federal Reserve’s balance sheet evolution (IMF, 2013d): –250 • The baseline scenario assumes a return to trend in financial market volatility from depressed levels Sources: Bloomberg, L.P.; and IMF staff estimates. and an exit process that is consistent with current Federal Reserve guidance. Under this scenario, the compression in term premiums gradually eases and

returns to its precrisis level by 2020. • The adverse scenario reflects the effects of increased bond market volatility and market expectations that could result from a sharper, frontloaded tapering of Figure 1.8. Simulated Shock to 10-Year Treasury quantitative easing. This scenario results in a similarly Term Premium (Percentage points) sized adjustment (100 basis points) in long-term rates as the baseline case, but the adjustment is abrupt. 0.6 The rise in long-term rates that took place during the Adverse scenario 0.4 May-June episode mostly reflected an increase in term premiums rather than short-rate expectations. That trajec- Baseline scenario 0.2 tory (represented by the blue dot in Figure 1.8) so far + 100 basis points lies above the baseline scenario, but overall term premi- 0.0 ums are still at extraordinarily low levels. If the adverse After May-June (2013) mini stress test scenario materializes, the Federal Reserve would likely –0.2 seek to temper such a shock through communication and by fine-tuning policies (for example, adjusting its asset –0.4 Start of simulation purchase schedule), but its effectiveness may be limited by –0.6 persistent financial stability risks and difficulty in offset- 2013 14 15 16 17 18 19 20 ting sudden, large portfolio shifts and managing volatil- Sources: Bloomberg, L.P.; and IMF staff estimates. ity shocks. Although long-term rates under the adverse scenario eventually converge with rates under the baseline scenario, the frontloaded nature of the shock would have pervasive effects on financial markets.

4To capture variations in the market’s expectation of the size and persistence of the asset purchase program, a measure is constructed following Chung and others (2011). In particular, the measure estimates a present discounted value of the current and expected future securities holdings in excess of its historical normal level as a ratio to potential GDP.

International Monetary Fund | October 2013 7

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: Transition Challenges to Stability

Figure 1.9. U.S. Mutual Fund Cumulative Flows Overextended fixed-income allocations and duration (Trillions of U.S. dollars) risk are likely to magnify losses. 3.5 Corporate and Yankee bonds To illustrate how such a shock would affect financial MMunicipal debt AgencyA debt and GSE-backed securities 3.0 markets, an instantaneous hike of the same magnitude TreasuryT and open market paper TrendT (for ﬁxed income) is applied to major bond portfolios. Recall that as EquitiesE $1.3 2.5 part of the yield-seeking behavior under quantitative trillion easing, there was a broad-based shift into fixed-income 2.0 assets and an extension in portfolio duration well above the historical norm (Figures 1.9 and 1.10). This 1.5 increase in duration significantly raises the sensitivity of portfolios to rising interest rates: a 100 basis point 1.0 increase in interest rates from current levels generates higher aggregate losses on global bond portfolios (5.6 0.5 percent or $2.3 trillion) than a similarly sized increase has generated on prevailing portfolios during previous 0.0 historical tightening episodes (Table 1.2).5 This is the 2000 01 02 03 04 05 06 07 08 09 10 11 12 13 case for global, U.S., and emerging market bond port- Sources: Federal Reserve; and IMF staff estimates. folios. Of course, the impact of such losses depends Note: GSE = government-sponsored enterprise. on the nature of the underlying shock, distribution, time frame, and other conditions. A normalization in response to improved economic conditions and broadly distributed losses would likely be more easily absorbed, whereas losses concentrated in entities with Figure 1.10. Global Bond Portfolio Duration (Benchmarks; years) large unhedged positions or asset-liability mismatches would increase instability. Advanced economies Emerging market local currency Emerging market hard currency United States 8 Structural reductions in market liquidity could amplify these effects, leading to an overshooting of interest rates.

7 It is important to stress that a more probable outcome would be a smooth portfolio rebalancing out of longer-duration, fixed-income assets on the back of a 6 gradual rise in interest rates and repricing of credit risk. However, overshooting may occur as a result of any

5 number of unanticipated events. For instance, some fund managers may seek to adjust portfolios ahead of future monetary policy tightening to avoid crystallizing 4 losses, thereby exacerbating market volatility. Recent changes in structural market liquidity could also magnify an increase in long-term rates as financial 3 6 2000 02 04 06 08 10 12 conditions normalize. Securities dealers’ inventories of fixed-income instruments have declined since 2007 Sources: Barclays Capital; and IMF staff estimates.

5For instance, during the last three tightening episodes in 1994–95, 1999–2000, and 2004–06, an instantaneous 100 basis point increase would have resulted in an average 4.8 percent loss on U.S. bond portfolios prevailing at the time. 6Liquidity risk premiums—defined as the ability to trade in large size without having a significant impact on market prices—are not directly captured in this term premium model.

8 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Table 1.2. Bond Portfolio Interest Rate Sensitivities Emerging Market Hard Emerging Market Local Global Bond Aggregate U.S. Bond Aggregate Currency Currency1 Duration (years) Average for Last Three Tightening Cycles2 5.0 4.8 4.0 . . . July 2013 6.2 5.5 5.9 4.9 Total Market Value (billions of U.S. dollars) Average for Last Three Tightening Cycles2 13,319 5,833 209 . . . July 2013 41,541 16,065 1,225 1,634 Impact from 100 Basis Point Increase (billions of U.S. dollars) Average for Last Three Tightening Cycles2 –664 –281 3 . . . July 2013 –2,325 –876 –68 –76 Impact from 100 Basis Point Increase (percent) Average for Last Three Tightening Cycles2 –4.9 –4.8 3.2 . . . July 2013 –5.6 –5.5 –5.5 –4.6 Sources: Barclays Capital; Bloomberg, L.P.; and IMF staff estimates. 1Data are unavailable before July 2008. 2Cycles include 1994–95, 1999–2000, and 2004–06.

Figure 1.11. Nongovernment Bond Inventories, and the willingness to make markets and intermediate Total Trading Volumes, and Outstanding Bonds liquidity more pronounced as dealers adjust their value- at-risk frameworks. Outstanding bonds 500 1,200 (outer right scale) 50 Higher interest rates may also reveal weak links in the Trading volumes 1,000 400 40 (inner right scale) shadow banking system, exacerbating liquidity and 800 300 30 market strains. Inventories 600 200 (left scale) 20 400 Repo and other forms of short-term wholesale funding Billions of U.S. dollars Billions of U.S. dollars Billions of U.S. Trillions of U.S. dollars of U.S. Trillions 100 200 10 markets in the United States have been a potential source of systemic stress ever since the crisis.9 A deep, 0 0 0 2000 02 04 06 08 10 12 well-functioning repo market is critical to ensuring sufficient market liquidity in the underlying collateral Sources: Federal Reserve; Securities Industry and Financial Markets Association; and IMF staff estimates. because repo is the primary market used by market Note: Average daily volumes include municipal securities, treasuries, agencies, asset- and mortgage-backed securities, corporate debt, and federal agency securities. participants for financing positions. Some progress has been made in reducing financial stability risks surrounding repo markets.10 In par- owing to efforts to reduce market leverage and to a shift ticular, the Financial Stability Board has made policy in funding and trading models. The decline has been recommendations to mitigate the risk of fire sales of accompanied by lower trading volumes even though the collateral securities by limiting the buildup of excessive outstanding stock of fixed-income tradable instruments leverage and reducing procyclicality. These recommen- has expanded (Figure 1.11). Leaner inventories and tight dations include minimum haircuts, regulation of cash nongovernment repo financing has led securities dealers collateral reinvestment, requirements on rehypotheca- to migrate toward more frequently traded issues, result- tion, and the introduction of central counterparties ing in a bifurcation between large, more recently issued (which also helps to mitigate contagion effects arising bonds and smaller, seasoned credits. Other changes since from over-the-counter derivatives markets) (FSB, the crisis have also affected market liquidity, including 2013). Shadow banking liabilities have continued to shifts in the investor base (for example, a shift from decline, repo concentration risks have eased, collateral more active, leveraged investors to unleveraged, buy- 7 and-hold investors), risk appetite, and trading behavior. ficient to fill in the gaps left by retrenching broker-dealer intermedia- Although the postcrisis system has yet to be tested, this tion capacity. 9 shift potentially reduces dealers’ ability to act as shock See Begalle and others (2013) and Dudley (2013). See Chapter 8 3 of this report for a discussion of recent changes in bank funding absorbers during market stress. In a higher-volatility structures. environment, inventories are likely to be even lower 10These efforts include a reduction in excessive reliance on intraday credit, improvement in risk management policies, bolstered capi- 7See Box 2.6 in the October 2012 GFSR. tal and liquidity buffers for large banks, diversified funding sources 8Some nonbank entities have emerged as agents using their own for large financial institutions, and strengthened liquidity require- portfolios to match buyers and sellers, but this has not been suf- ments and concentration limits for money market mutual funds.

International Monetary Fund | October 2013 9

©International Monetary Fund. Not for Redistribution global FInanCIal stabIlIty rePort: transItIon Challenges to stabIlIty

quality has improved, and the volume of intraday Figure 1.12. Fire Sale “Risk Spiral” credit has decreased. However, short-term secured funding markets are Interest Rate Shock Higher rates lead to wider still exposed to potential runs that a rising-rate, higher- MBS spreads. volatility environment may reveal, owing to the follow- MBS Spread-Widening and Rising Losses ing vulnerabilities: Wider MBS spreads reduce • Asset fire sales: Asset sales drive book value, equity, and Fire sales may result either from a MBS spreads wider, Increased Volatility and assets available for repo. reinforcing the borrower default that leads to a liquidation of collat- Counterparty Credit Risk rise in rates. Increased volatility and eral in a volatile market or in response to preemptive counterparty risk concerns lead to tighter funding asset sales triggered by the mere risk of default. Higher Haircuts and conditions. • Flight-prone investor base: Lenders may cease rolling Reduced Funding Reduced funding, inability to over repo funding with limited notice.11 raise equity, and market Forced Asset Sales demands to reduce leverage • Contagion risks: Forced liquidations or the inability lead REITs to sell assets. to unwind illiquid assets could lead to greater pressure on other traditionally more liquid securities and Source: IMF staff estimates. market participants. Note: MBS = mortgage-backed security; REIT = real estate investment trust. Entities in the shadow banking system that use repo markets as a source of funding for longer-term, less-liquid assets are vulnerable to these risks. One example of such entities is mortgage real estate investment trusts (mRE- to sell MBSs because higher rates and wider MBS spreads ITs). Although their sheer size does not signal systemic were leading to declining portfolio values, reduced equity importance as a sector (assets total about $500 billion), cushions, and higher margins. To sustain the level of mREITs have grown signifi cantly in recent years and now borrowing relative to their net worth, the largest mREITs have a more important role in mortgage-backed security unwound $30 billion of MBS over the course of a single (MBS) markets (see Box 1.1). Furthermore, the mREIT week. To put that fi gure into context, a daily liquida- business model layers on other risks that could amplify tion of more than $4 billion by any MBS investor under market dislocations in a rising-rate environment. Specifi - normal market conditions adversely aff ects MBS prices cally, mREITs are leveraged, exposed to volatility shocks (Begalle and others, 2013). Th ese large sales weighed on (as a result of the prepayment option embedded in their overall MBS valuations and fueled an increase in primary MBS holdings), and highly dependent on short-term repo mortgage rates. Further interest rate increases could lead funding to fi nance their long-term assets. Th e combina- to a more destabilizing unwinding of positions (Figure tion of these risks increases their vulnerability to a fi re sale 1.13), with higher leverage magnifying losses (Figure event (Figure 1.12) in which higher interest rates pressure 1.14). An instantaneous interest rate shock of 50 basis mortgage rates and MBS spreads to widen and volatil- points or more would lead to portfolio value declines ity to increase, leading repo lenders to raise margins or among the top mREITs large enough to generate at least 12 reduce funding. Th is in turn induces mREITs to unwind temporary dislocations in the MBS market. their holdings in a declining market, thereby triggering a Such a scenario of rapid mREIT deleveraging has more disorderly adjustment in MBS valuations and exac- important spillover implications. Consistent selling erbating broader market discontinuities as MBS investors pressure could negatively aff ect MBS valuations and rebalance the hedges they use to manage the interest rate thus weigh on the balance sheets of other MBS inves- exposure of their portfolios. tors (for example, commercial banks, government- A version of this scenario played out during the market sponsored enterprises, the Federal Reserve). Sizable correction in May-June 2013. Many mREITs were forced disruptions in secondary mortgage markets against a backdrop of rising mortgage rates could also have macroeconomic implications, jeopardizing the still- 11Money market mutual funds, for instance, are important cash providers in the repo market but have limited ability to deter or slow an exit by investors. Reforms made in 2010—as well as the 12Th is assumes that declines in mREIT portfolio values lead to U.S. Securities and Exchange Commission’s proposal to require forced asset sales of a similar size over a compressed time frame, prime funds to adopt fl oating share prices or impose liquidity fees owing to reduced funding availability, an inability to raise equity, or restrictions on withdrawals—have signifi cantly reduced the risk of and market pressure to reduce leverage, all of which further magnify investor fl ight. But the system has yet to be tested. valuation declines.

10 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.13. Estimated Average Change in Mortgage Figure 1.14. Leveraged Mortgage Real Estate Real Estate Investment Trust Portfolio Value for Parallel Investment Trusts Are More Vulnerable to Interest Interest Rate Shifts Rate Increases

5 30

0 20

–5 10

–10 0 Percent –15 –10 point shock

–20 –20

–25 –30 Percent change in net interest income for a +100 basis Percent

–30 –40 –200 –100 –75 –50 –25 25 50 75 100 200 4 5 6 7 8 9 10 11 12 Yield curve shift (basis points) Leverage ratio (assets/equity) as of 2013:Q2

Sources: Company ﬁlings; and IMF staff estimates. Sources: Bloomberg, L.P.; and IMF staff estimates. Note: Includes impact of interest rate hedges.

fragile housing recovery. For instance, rising mort- risks to substantially affect the underlying collateral gage rates and widening MBS spreads have already credit risk for a protracted period. However, given led to a significant pullback in mortgage refinancing that the repo funding of the two largest mREITs is activity.13 Given the importance of MBS collateral in comparable to Lehman Brothers’ precrisis repo book, repo markets, a large enough shock to MBS valua- at the very least the mREITs point to a microcosm of tions, combined with a weakening in risk sentiment, fragilities in the shadow banking system that deserve could also induce repo lenders to pull back funding closer monitoring.15 or raise rates more broadly (or both), with negative consequences for other leveraged short-term bor- Policymakers can take a number of actions to help rowers.14 Securities dealers are currently net borrow- ensure a smooth transition. ers using MBS repo (their borrowing exceeds their lending by about $185 billion), increasing the risk Achieving a smooth transition requires policies that that repo lines would likely be cut fairly quickly to manage the effects of increased volatility and destabiliz- leveraged investors in the event of a deterioration ing portfolio adjustments and that address structural in MBS valuations. Disruptions to secured funding liquidity weaknesses and systemic vulnerabilities in the markets that occurred during the global financial shadow banking system. This is a major policy chal- crisis, following the deterioration in credit quality lenge that requires a number of actions, as outlined in of structured finance markets, are an apt reminder the following. of the ripple effects. Granted, agency MBS markets • A clear and well-timed communication strategy by are deeper, more liquid, and less risky, and mREIT central bank officials is critical. Compared with balance sheets are too small to allow counterparty previous tightening cycles, the authorities have a broader toolkit at their disposal and have made 13The 115 basis point uptick in mortgage rates since May has progress in developing a more refined communica- been accompanied by a 52 percent decline in overall mortgage applications during the same period, mostly reflecting reduced refinancing 15The two largest mREITs currently have repo liabilities of about activity. $100 billion to $125 billion each (one-third of which is less than 30 14MBS collateral represents nearly 40 percent of repo-funded days in maturity), as compared with Lehman’s repo book of $150 transactions. billion in September 2008.

International Monetary Fund | October 2013 11

©International Monetary Fund. Not for Redistribution global FInanCIal stabIlIty rePort: transItIon Challenges to stabIlIty

box 1.1. Mortgage real estate investment trusts: business Model risks

This box discusses the main institutional weaknesses that business model layers on other risks that could amplify expose mortgage real estate investment trusts to risk along a market dislocations: number of dimensions. • Funding and liquidity risk: Although mREITs have always relied to a certain extent on short-term Real estate investment trusts (REITs) own, and in secured financing, that share mushroomed during most cases operate, income-producing real estate. A the financial crisis when the cost advantage between subset of these companies, mortgage REITs (mREITs), the secured and unsecured market expanded and are involved in lending money to owners of real estate the availability of long-term financing dried up and buying (mostly agency-backed) mortgage-backed (Figure 1.1.3). securities (MBSs).1 The mREITs engage in leveraged • Refinancing and rollover risk: Because debt maturi- maturity transformation by relying on short-term repo ties are short, considerable refinancing and rollover funding—some of which is channeled indirectly from risks also arise. Unlike European banks—which money market mutual funds via securities dealer inter- when faced with a pullback in repo funding by U.S. mediaries—to finance their long-term MBSs (Figure money market funds in mid-2011 turned to cross- 1.1.1). currency basis swap markets and European Central Although mREITs are not large holders of MBSs Bank long-term refinancing operations as a substi- on a relative basis (Figure 1.1.2), they have grown in tute—mREITs have limited funding alternatives. importance since the global financial crisis, and their Furthermore, because the bulk of mREIT earnings are required to be paid out to investors, minimal This box was prepared by Rebecca McCaughrin. 1 cash flow can be retained for other purposes, result- Agency mREITs represent roughly 85 percent of the REIT 2 sector. Another smaller subset, credit REITs, typically securitize ing in slim liquidity buffers. pools of loans and sell the senior tranche, while retaining the 2To maintain their advantageous tax status, REITs are required subordinate first-loss (credit) tranche. to pay a large share of their taxable income as dividends.

Figure 1.1.1. Example of the Real Estate Investment Trust Maturity Transformation Process

1. A broker-dealer executes a 2. The broker-dealer uses the cash short-term, collateralized reverse to execute a bilateral repo with a repo with a liquidity-rich entity, REIT with a longer maturity and typically through a triparty clearing higher haircut, at a higher repo rate bank for a small fee (owing to the (owing to the longer tenor and conservative nature of the higher counterparty risk), earning a transaction). spread on the difference in rates of the two legs.

Cash Cash MMMF short-term REIT short-term cash cash investor borrower Terms of transaction: Tri-party Intermediated Terms of transaction: * Overnight clearing by securities * Term * 20 basis point repo bank dealers * 50 basis point repo rate rate * 2 percent haircut Mortgage- Mortgage- * 5 percent haircut backed backed securities securities

3. The REIT then invests the short-term cash obtained from the repo in long-dated MBS, earning a spread between the two rates.

Sources: Company statements; Fitch Ratings; and IMF staff. Note: MBS = mortgage-backed security; MMMF = money market mutual fund; REIT = real estate investment trust. Transaction terms relate to intermediating securities dealers.

12 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

box 1.1 (continued)

Figure 1.1.2. Holdings of Agency Figure 1.1.4. REITs’ Agency MBS Holdings Mortgage-Backed Securities versus GSEs’ MBS Investment Portfolio Other GSEs Holdings 9% 4% Federal Reserve 14% REIT holdings of MBS Pension funds GSE holdings of MBS 5% 1,200

Insurers REITs 5% 5% 1,000

Foreigners 800 16% Depository institutions 600 26% Mutual funds 18% dollars Billions of U.S. Total: $7.6 trillion 400

Sources: Federal Reserve; and IMF staff estimates. 200 Note: Total may differ from 100 percent due to rounding. GSE = government-sponsored enterprise; REIT = real estate investment trust. 0 90 93 96 99 05 08 11 87 1984 2002

Sources: Federal Reserve; and IMF staff estimates. Note: GSE = government-sponsored enterprise; MBS = mortgage-backed securities; REIT = real estate investment trust. Figure 1.1.3. Real Estate Investment Trust Dependence on Short-Term Funding

Funding by repo and federal funds (left scale) • Maturity mismatch risk: Some REITs have sought to Repo as a share of total liabilities (right scale) increase the maturity of their repo-related financ- 350 45 ing, diversify their repo counterparties, and shift

40 into other (more costly) sources, but most mREITs 300 are still highly dependent on short-term funding to 35 finance long-term assets.3 This maturity transfor- 250 mation risk is akin to the funding problems that 30 emerged during 2008 in the asset-backed commer- 200 25 cial paper market. • Convexity risk: All mREITs are exposed to inter-

150 20 Percent est rate and convexity risk. Given the prepayment

Billions of U.S. dollars Billions of U.S. options embedded in MBSs, the effective duration 15 100 of MBSs increases as interest rates rise, because 10 higher rates reduce mortgage refinancing activity 50 and slow the rate of prepayments. Generally, mRE- 5 ITs hedge the interest rate risk of their mortgage portfolios through Treasury bills, interest rate swaps, 0 0

95 98 04 07 10 swaptions, and other MBSs, but only partly. In 1992 2001 13:Q1 addition to a worsening in the duration mismatch, rising rates result in higher valuation losses on MBS Sources: Federal Reserve; and IMF staff estimates. holdings. Given current convexity risk, the average

3Among the largest mREITs, about 90 percent of assets are used as collateral in repos, which leaves limited unencumbered assets.

International Monetary Fund | October 2013 13

©International Monetary Fund. Not for Redistribution global FInanCIal stabIlIty rePort: transItIon Challenges to stabIlIty

box 1.1 (concluded)

mREIT MBS portfolio value would decline by • Wrong-way risk: Because mREITs pledge collat- roughly 10 percent in the event of a 100 basis point eral on the asset side of the balance sheet to fund parallel interest rate shock. themselves, they may be simultaneously exposed to • Concentration and correlation risk: Most mREITs pressure to make payments to investors and pressure hold fixed-rate agency MBSs, private-label MBSs, on the value of assets pledged for financing. and commercial MBSs, and so are sensitive to • Market risk: Increased capital market volatility tends shocks to mortgage and property markets.4 (By con- to reduce access to sources for refinancing and trast, the other large investors in MBSs, as shown capital. in Figure 1.1.2, have more diversified portfolios.) These risks are interrelated. Higher interest rates Their assets have expanded significantly since exacerbate convexity-related risks, which in turn raise the crisis, to the point that mREITs now hold a lenders’ concerns about the underlying collateral, larger stock of agency MBSs than the government- aggravate short-term funding conditions, and reinforce sponsored entities do in their investment portfolios the maturity transformation risk. Collateral and coun- (Figure 1.1.4). Furthermore, these risks are concen- terparty correlation risk also raise investors’ concerns trated in two large institutions. about the strength of future earnings and dividends, in turn increasing the cost of capital. Figure 1.12 in 4Regulatory guidelines require mREITs to hold a minimum of the main text illustrates how the presence of these risks 75 percent of agency MBSs. could lead to a fire sale event.

tion policy. But unexpected large increases in long- and susceptibility to short-term funding pressure— term rates, as the May-June episode suggests, cannot would help reduce the risk of a cascading failure be ruled out. of counterparties. A review of repo haircuts and • In the event of adverse shocks, contingency backstops margins would be desirable to limit the degree of need to be in place that reduce the likelihood of leverage and procyclicality inherent in these markets. cascading forced asset sales. Although a number of Greater disclosure by repo market participants and steps have been taken to mitigate the risks pres- mREITs would also help markets more accurately ent in short-term wholesale funding markets, other assess the risks to which these entities are exposed. options should be considered to address the risk In addition, the authorities could consider chang- of fire sales. Establishing incentives that lengthen ing the exemption status for certain mREITs, or if the maturity of repo contracts for borrowers in the warranted, designate the largest mREITs as systemi- shadow banking system may help reduce the initial cally important entities, subjecting them to greater buildup of maturity and liquidity transformation supervisory oversight. risk. In a severe crisis scenario, a mechanism (such as • Finally, further efforts are needed to assess how mar- a resolution authority) that can manage an orderly ket developments and regulatory initiatives affecting and appropriately timed unwinding or liquidation of dealer-bank business models may affect the cost repo collateral may be warranted.16 and provision of market liquidity. At a minimum, • Policies also need to be focused on structural vulner- increased surveillance of and vigilance over the effects of abilities. In particular, increased oversight of shadow trading liquidity pressures will be needed as financial banking entities (including repo market participants markets make the transition to a regime with higher and the larger mREITs)—given such entities’ inher- interest rates and volatility. In the longer term, secu- ent vulnerability to prepayment and interest rate risk rities and market regulators need to ensure that fund managers in illiquid and opaque underlying markets 17 16Such a facility would allow a repo cash lender to sell its collateral are mindful of the risks of liquidity drying up. to a well-capitalized liquidation agent with the ability to manage an orderly sale of the underlying collateral instead of liquidating the collateral received from a failing counterparty in a stressed market. See 17See the recommendations by the International Organization of Acharya and Öncü (2013). Securities Commissions in OICV-IOSCO (2012).

14 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

emerging Markets: riding the ebbing tide Figure 1.15. Above-Trend Bond Flows from Advanced of capital Flows to Emerging Market Economies (Percent of advanced economies' GDP) Accommodative monetary policies in advanced economies have encouraged foreign inflows into emerging 6 market bond markets squarely above their long-term trend. This raises the question of whether monetary 5 Long-term trend $470 billion policy normalization in the United States will result in further turbulence in emerging markets. Although 4 emerging market economies in general now have more buffers than in previous episodes of market volatil- 3 ity, events since May point to new financial stability concerns. The sensitivity of emerging market yields $370 billion to changes in external conditions has increased as 2 Cumulative flows foreigners have crowded into local markets, duration has lengthened, and market liquidity has dimin- 1 Cumulative flows, ished. Emerging market fundamentals have recently net of valuation effects weakened against the backdrop of weakening mac- 0 roeconomic positions and rising financial leverage. 2002 03 04 05 06 07 08 09 10 11 12E 13F Sources: IMF Consolidated Portfolio Investment Survey (CPIS); JP Morgan; and IMF Low growth, low rates, and unconventional monetary staff calculations. Note: E = estimate; F = forecast. The long-term trends were extrapolated from the policies in advanced economies have boosted inflows 2002–07 period to remove the effects of the global financial crisis and unconventional to the bond markets of emerging market economies. monetary policies. Data for 2012–13 were calculated from the trend of 2009–11 and estimates. Advanced economies = Bermuda, Canada, Cayman Islands, France, Germany, Hong Kong SAR, Italy, Japan, Jersey, Luxembourg, Netherlands, Singapore, Foreign portfolio investment in emerging market Switzerland, United Kingdom, and United States. bonds has been on an increasing long-term path since 2002, reflecting higher growth differentials and a expectations than on the effects of unconventional structural increase of allocations into emerging market monetary policies in advanced economies. assets. But since the pullback during the 2008 global Countries receiving relatively higher bond inflows financial crisis, cumulative bond inflows have risen generally experienced greater yield compression, with by an estimated $1.1 trillion through 2013, or $0.9 10-year bond yields in Indonesia, Mexico, and the trillion excluding portfolio and currency effects. These Philippines declining by more than 300 basis points cumulative flows represent 5.5 percent of advanced from their long-term average levels through mid-May economy nominal GDP (or 4.7 percent in net terms), 2013 (Figure 1.17). As discussed in the April 2013 and puts the 2013 forecast squarely above its long-term GFSR, external factors accounted for about two-thirds structural trend by an estimated $470 billion (or $370 of the local currency yield compression since 2008, billion in net terms; Figure 1.15).18 with domestic improvements explaining the smaller Foreign inflows into bonds have averaged more than share. These conditions have also enabled low-income 2 percentage points of recipient-country GDP a year countries to issue hard currency debt (Box 1.2). during the previous four years, mainly into higher- yielding, more liquid markets (Figure 1.16). Equity Foreign investors have crowded into local emerging portfolio flows have been less consistent than fixed- markets but market liquidity has deteriorated, making income flows since 2009, albeit of the same order of an exit more difficult. magnitude, and they are more dependent on growth

Yield-sensitive (so-called crossover) investors have 18The 2012 estimate and 2013 forecast of the cumulative fixed- income portfolio flows are extrapolated from the linear trend of much larger positions in emerging markets today than the previous three years, taking into consideration the outflows in in 2009. A trend that started out with mostly dedi- 2013:Q2–2013:Q3 and assuming continuing outflows in 2013:Q4. cated emerging market funds now includes “global They are conservative estimates of the portfolio flow increases when compared with more high frequency portfolio allocation surveys, or total return bond funds” and other crossover inves- the increase in the market capitalization of major bond indices. tors attracted by yield and an improvement in credit

International Monetary Fund | October 2013 15

©International Monetary Fund. Not for Redistribution global FInanCIal stabIlIty rePort: transItIon Challenges to stabIlIty

Figure 1.16. Net Portfolio Flows into Fixed-Income Figure 1.17. Impact of Portfolio Flows on Local Currency Equity Markets by Country, 2009–12 Bond Yields (Annual average; percent of GDP) 0 Fixed income Equities China India Greater yield 7 Chile changes 6 Thailand –100 Russia 5 Romania South Africa Hong Kong SAR 4 Israel Peru –200 Brazil 3 Colombia Poland Hungary 2 Turkey Mexico –300 1 Change in yield (basis points) 0 –400 Indonesia –1 Larger capital Philippines inﬂows Israel China India Brazil Peru Chile Russia Turkey PolandMexico Hungary RomaniaColombia Thailand Indonesia Philippines –500 South Africa 0 1 2 3 4 Hong Kong SAR Net portfolio inﬂow into debt instruments, 2009–12 Sources: Bloomberg, L.P.; CEIC Data; Haver Analytics; IMF, International Financial (annual average; percent of GDP) Statistics database; and IMF staff calculations. Note: Data for some countries are not necessarily until 2012 but are instead the Sources: Bloomberg, L.P.; CEIC Data; Haver Analytics; IMF International Financial Statistics latest data available. database; and IMF staff calculations. Note: The vertical axis shows the deviation of the mid-May 2013 10-year local government bond yield from its average since 2005.

f undamentals (Figure 1.18). At the same time, the benign external environment and search for yield facilitated a lengthening of maturities. Although this is sup- Figure 1.18. Allocation of Major Bond Funds to portive of government debt liability management, the Emerging Markets (Percent) increased duration of bond issues poses greater risks to investors from a rise in interest rates (Figure 1.19). 14 At the same time that foreign investors have 12 crowded into fixed-income assets, liquidity in several emerging market economy bond markets has declined 10 considerably in recent years (Figure 1.20). Offshore banks have scaled back their market-making activi- 8 ties, increasing reliance on local players for liquidity. Reduced turnover in secondary markets during the last 6 year is particularly evident in Hungary, Indonesia, and Malaysia, where foreign investor holdings now amount 4 to more than 20 times (75 for Indonesia) the average

daily trading volume (see Figure 1.20). In turn, dur- 2 ing periods of reduced liquidity, the increased foreign

exchange hedging activity by foreign institutional 0 investors can weaken local currencies, despite relatively June 2001 June 03 June 05 June 07 June 09 June 11 June 13 few outflows from domestic assets. This effect has Source: PIMCO. occurred in many countries since May 2013 on expec- Note: Allocation of PIMCO total return fund to emerging market ﬁxed income. tations of reduced U.S. monetary accommodation. Furthermore, the domestic investor base in many countries may be unwilling or unable to increase its holdings of fixed-income assets to provide adequate buffers against volatility during protracted sell-offs, as analysis in the Octo-

16 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

box 1.2. First-time issuers: new Opportunities and emerging risks

Hard currency bond issuance by first-time issuers has risen in recent years.1 Although these issuers do not currently Figure 1.2.1. International Bond Debut appear to pose systemic risks to the global financial system, Issuance in some instances these developments represent a significant (Millions of U.S. dollars) rise in external indebtedness, and may heighten stability Africa Eastern Europe Latin America 2 risks within particular countries. Such countries should Middle East Asia issue external debt in the context of a comprehensive medium-term debt management strategy and concurrently 4,000 deepen local markets to reduce dependence on volatile foreign capital. Debut issuers performed less poorly than their 3,500 more liquid emerging market counterparts in the ongoing 3,000 sell-off, but they have not been tested by a more prolonged period of repricing and therefore merit ongoing monitoring. 2,500

During the past 10 years, 23 emerging market econo- 2,000 mies and low-income countries have issued bonds inter- 1,500 nationally for the first time or have reentered the market after a long hiatus (Figure 1.2.1).3 The issuers are diverse, 1,000 both geographically and in terms of income levels, but 500 generally have a sub-investment-grade (BB) rating. The recent spike in issuance can be explained by 0 demand and supply factors. The search for yield and 2004 0506 07 08 09 10 11 12 13 demand for portfolio diversification have resulted in Source: Dealogic. demand-driven easy financing conditions, despite an ambiguous improvement in fundamentals.4 Further- more, rising financing needs, coupled with reduced access to concessional financing, relatively undeveloped Figure 1.2.2. Investor Base by Region domestic markets, and a favorable interest rate envi- and Type ronment, have made international bonds an attractive (Percent) financing alternative. 100 Pensions and Despite many similarities in the investor bases of insurers debut issuers and frequent issuers, notable differ- Others Banks

Own Others ences are apparent. In recent years, investors in global Banks 80 Banks Europe The authors of this box are Nehad Chowdhury, Anastasia Gus- Europe cina, Guilherme Pedras, and Gabriel Presciuttini. 60 1Most of these issuers would be considered frontier markets by bond investors, but for the purpose of this box, the term Europe “first-time” or “debut” is used. For the purpose of this study, we classified as first-time issuers only countries that have issued for 40 the first time since 2004, in amounts of at least $200 million. Fund managers

2 Fund managers The sum of issuance since 2004 ($14 billion) represents less Fund managers than 3 percent of the market capitalization of emerging market United States bonds. The market capitalization of JP Morgan’s EMBIG was 20 United States

$579 billion at end-April 2013. United States 3The 23 economies are Albania, Angola, Belarus, Bolivia, Ecuador, Gabon, Georgia, Ghana, Honduras, Jordan, Mongolia, 0 Montenegro, Namibia, Nigeria, Pakistan, Paraguay, Rwanda, Asia Africa Latin Asia AfricaLatin Senegal, Seychelles, Sri Lanka, Tanzania, Vietnam, and Zambia. America America 4Real GDP growth in the year of issuance was higher than the Source: Based on lead-managers information. average of the previous three years. However, current accounts Note: Weighted average for deals by Bolivia, Guatemala, deteriorated in the year of issuance compared with historical Honduras, Mongolia, Nigeria, Paraguay, and Zambia. averages, indicating borrowers’ need for hard currency.

International Monetary Fund | October 2013 17

©International Monetary Fund. Not for Redistribution global FInanCIal stabIlIty rePort: transItIon Challenges to stabIlIty

box 1.2 (continued)

Figure 1.2.3. Performance of Frontier Markets during Emerging Market Bond Sell-Off (Yield change in basis points May 1– June 25, 2013) 300

250

200

Median = 150 basis points 150 Median = 123 basis points

100

–50 Gabon Ghana Bolivia Jordan Nigeria Albania Zambia Belarus Rwanda Vietnam Ecuador Namibia Tanzania Mongolia Paraguay Honduras Montenegro EMBI Global –100

BBB BB B NR Sources: Bloomberg, L.P.; and IMF staff estimates. Note: Yellow bars represent issuances before 2008, and blue bars represent issuances thereafter. Average remaining maturity of the “ﬁrst wave” is 2.9 years, whereas for the “second wave” it is 7.6 years. EMBI = Emerging Markets Bond Index; NR = not rated.

investment-grade credit have crossed over (and are Figure 1.2.4. Size of Selected Frontier therefore referred to as crossover investors) to purchase Market International Bond Issuance investment-grade and relatively liquid emerging market debt (that of Brazil, Mexico, Russia, and others), Percent of GDP Percent of debt stock (right scale) but have not purchased the mostly lower credit quality 25 35 debt of debut issuers, and neither have hedge funds. In contrast, the investor base for debut issuers is still 30 dominated by dedicated, real money investors (Figure 20 1.2.2). 25 First-time issuers typically access markets at spreads notably wide of the Emerging Markets Bond Index 15 (EMBI). The higher spreads reflect their weaker credit 20 profiles, poorer secondary market liquidity, poorer 15 transparency, and lack of capital market financing 10 track record.5 Although debut issuers have not sold off more dra- 10 matically than the higher credit quality issuers during 5 the current sell-off (Figure 1.2.3), how they will fare 5 in a more prolonged period of repricing remains to be seen. On average, debut issuers were able to withstand 0 0 the shock on par with the more liquid issuers because investors across the board, particularly cross-over Gabon 2007 Ghana 2007 Bolivia 2012 Jordan 2010 Albania 2010 Zambia 2012 Georgia 2008 Rwanda 2013 Namibia 2011 Tanzania 2013 Tanzania

5 Mongolia 2012 Paraguay 2013 Paraguay According to IMF staff estimates, first-time issuers are Honduras 2013 Seychelles 2006 borrowing at a spread over EMBI that can only partially be Montenegro 2010 explained by ratings, macroeconomic and institutional character- Source: Dealogic. istics, and fiscal variables.

18 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

box 1.2 (concluded)

investors and hedge funds, first sold the most-liquid cially if accompanied by depreciating exchange rates.6 assets. The relative illiquidity of debut issuers’ bonds Policymakers should tap international markets only protected them from a more dramatic sell-off in the in the context of a comprehensive medium-term debt initial stage. It remains to be seen what would happen management strategy that makes the trade-off between in a more sustained sell-off. costs and risks explicit, and at the same time should Debut issuers should adopt policies that mitigate deepen local markets to reduce dependence on volatile risks associated with external debt. Some countries foreign capital. have issued bonds in large amounts compared with the size of their economies (Figure 1.2.4) or without 6Exposure to exchange rate depreciation is the most prominent a clearly defined use of the proceeds. The unwinding risk, given that many countries’ already-significant exposures to of unconventional monetary policies and increases in currency risk in their portfolios has further increased with the interest rates may pose refinancing challenges, espe- issuance of Eurobonds.

ber 2012 GFSR explained. Accordingly, asset prices may be domestic firms. Faced with underlying weakness in more vulnerable in Hungary, Indonesia, Mexico, and, to a demand and excess capacity across many industries, lesser extent, South Africa, as coverage of foreign investor corporate earnings have been falling (Figure 1.27, outflows by local investors is limited (Figure 1.21).19 panel 1). This development, along with the rise in corporate leverage in the past few years, explains why Corporate sector vulnerabilities are on the rise as the interest coverage ratios have progressively weakened leverage cycle advances. (Figure 1.27, panel 2; see also Box 1.1 of the April 2013 GFSR). Sustained pressure on financial posi- Corporate sector borrowing has surged since the crisis tions in the corporate sector would undoubtedly hit began, facilitated by foreign investors (Figure 1.22). banks’ loan portfolios, putting at risk the still-intact While in general highly rated firms typically raise pattern of strikingly low reported nonperforming the most capital, so far in 2013 the credit quality of loan ratios. new issues has deteriorated (Figure 1.23). Indeed, improvements in the overall credit profile of emerg- Financial vulnerabilities are rising because macro- ing market companies have peaked and are showing economic fundamentals have recently weakened. signs of deterioration as credit downgrades rise (Figure 1.24). Corporate leverage is also on the rise: net debt The external positions of emerging markets have deterio- to common equity increased to more than 60 percent rated since 2007, partly because of economic weakness for Latin American companies in 2012, and it remains in advanced economies, with the exception of those elevated for Asian companies (Figure 1.25, panel 1). eastern European countries that were previously running This trend, together with some slowdown in corporate exceptionally high deficits. This change in external posi- earnings, has caused interest coverage ratios among tions has arguably supported global rebalancing, but has Asian corporates to dip to a multiple of three times left some economies (especially Asian) that traditionally in 2012, down from a multiple of almost five times have large current account surpluses in a weaker external in 2010 (Figure 1.25, panel 2). In 2012, corporate position. Against the backdrop of weak global growth defaults reached their highest level since the global since 2009, many emerging markets pursued coun- financial crisis with 20 credit events amounting to tercyclical policies that expanded domestic credit. The $22 billion (Figure 1.26). long period of rapid credit expansion and easy access These trends are also evident in China, where to funding has given rise to greater domestic financial slower economic growth has begun to put pressure on vulnerabilities. For example, countries in the shaded areas of Figure 1.28 are faced with increased external 19 In Poland, the size of the nonbank financial sector may decline and domestic vulnerabilities at a time when many are relative to the nonresident holdings of local currency bonds following plans to absorb the government bond holdings of the pillar II also finding themselves with shrinking fiscal space (see pension fund assets into general government debt. the October 2013 Fiscal Monitor).

International Monetary Fund | October 2013 19

©International Monetary Fund. Not for Redistribution global FInanCIal stabIlIty rePort: transItIon Challenges to stabIlIty

Figure 1.19. Duration of Emerging Market Fixed-Income Figure 1.20. Share of Nonresident Holdings of Local Indices Currency Government Debt and Market Liquidity (Years; 30-day moving average) Share of total (current; percent) Share of average daily volume (current; days) Local sovereign Share of total Share of average daily volume Hard currency corporate (7-year average; percent) (October 2012; days) 80 5.4 Hard currency sovereign (right scale) 7.3 70 7.1 5.2 60 6.9 5.0 6.7 50

4.8 6.5 40

6.3 4.6 30

6.1 20 4.4 5.9 10 4.2 5.7 0 4.0 5.5 Brazil 2009 10 11 12 13 Turkey Poland Mexico Colombia Thailand Romania MalaysiaIndonesia Hungary South Africa Source: Bank of America Merrill Lynch. Sources: Asian Development Bank; national authorities; and IMF staff calculations.

Figure 1.21. Composition of the Holders of Local Currency Figure 1.22. Net New Issuance of Emerging Market Bonds Government Debt (Billions of U.S. dollars) (Percent) 300 Sovereign Corporate Foreigners Banks Nonbank ﬁnancial institutions Others 100 250

80 200

60 150

100 40

50 20

0 2008 09 10 11 12 13 YTD 0 Sources: Bond Radar; and Morgan Stanley. rkey Brazil Tu Poland Mexico Note: Data available through August 2013. YTD = year to date. Colombia Thailand Romania Malaysia Indonesia Hungary South Africa

Sources: Asian Development Bank; national authorities; and IMF staff calculations.

20 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.23. Credit Ratings of Emerging Market Figure 1.24. Corporate Rating Changes in Emerging Markets Corporate Bond Issues (Number of changes) BBB+ Upgrades Downgrades 160

140 Median (weighted by issuance size)

120 BBB 100

Weighted average 80

60 BBB– Investment-grade credit rating threshold 40

BB+ 0 2009 10 11 12 13 2000 01 02 03 04 05 06 07 08 09 10 11 12 13 YTD YTD Sources: Bond Radar; and IMF staff calculations. Source: Fitch Ratings. Note: Data available through July 2013. YTD = year to date. Note: Data available through 2013:Q1. YTD = year to date.

Rapid credit growth in the shadow banking system in • Lack of market disclosure: The new credit instruments China remains a key vulnerability. lack the central element of market-based intermediation, that is, effective market discipline. The Credit creation in China reaccelerated in early 2013, possibility of default is crucial to inducing proper as broad credit expanded by more than 22 percent pricing of credit risk. Yet China’s financial system (year over year). This level was well below the peak features a pervasive perception that alternative saving rates of credit growth in 2009–10 but further extends vehicles, including wealth management products, are the sharp rise in China’s credit-to-GDP ratio to almost effectively guaranteed by issuers. A history of bail- 180 percent of GDP (Figure 1.29). It also heightens outs has created similar moral hazard in the market worries that the rapid credit expansion may foreshadow for corporate bonds. a marked worsening of asset quality. Rapid disinterme- • Ties with the traditional banking system remain too diation has pushed the share of bank loans in total new close for comfort: Although financial innovation credit down to just above 55 percent in the first half of superficially reduces their role, China’s banks remain the year. This trend has helped diversify the financial deeply involved in many new forms of credit inter- system and introduce more market-based lending and mediation, although without the safeguards of capi- investment products, but the surge in nonstandard tal requirements, provisioning, or detailed disclosure. instruments—exemplified by the doubling of trust loans For example, some trust companies rely on banks to in less than 12 months—also carries considerable risks: both refer borrowers and provide funding. • Lack of oversight: Many of the new funding channels are subject to lighter regulation and supervi- As the United States approaches exit from sion. Trust companies have faced little regulatory unconventional monetary policies, emerging market constraint in ramping up their exposure to two vulnerabilities have come to the fore. sectors that are largely excluded from access to new bank loans: local government financing vehicles Since Federal Reserve Chairman Ben Bernanke’s and the property sector. Both of these sectors have testimony to Congress on May 22, emerging market been important drivers of recent economic activ- assets have come under pressure. Initially, the sell-off ity, but face serious questions about their financial was strong in most countries, reflecting the first two sustainability. key vulnerabilities: (1) yields and risk premiums had

International Monetary Fund | October 2013 21

©International Monetary Fund. Not for Redistribution global FInanCIal stabIlIty rePort: transItIon Challenges to stabIlIty

Figure 1.25. Nonﬁnancial Corporate Figure 1.26. Sharp Increase in Corporate Debt Defaults Balance Sheet Metrics (Billions of U.S. dollars)

600 90 Asia CEEMEA Latin America (left scale) (right scale) 1. Ratio of Net Debt to Equity

65 500 75

Strong 60 400 increase 60 in 2012 defaults 300 45 55

200 30 50 Percent

100 15

0 0 08 09 10 11 12 08 09 10 11 12 08 09 10 11 12 08 09 10 11 12 2007 2007 2007 2007 40 United States Europe Other advanced Emerging market economies economies 2004 05 06 07 08 09 10 11 12 Source: Standard & Poor's. Note: Other advanced economies = Australia, Canada, Japan, and New Zealand. 2. Ratio of EBIT to Interest Expense 5.5 become overly compressed and are likely to be repriced 5.0 further as monetary conditions normalize; and (2) the sensitivity of emerging market yields to changes in 4.5 external conditions and foreign flows has increased,

4.0 owing to crowded positions in local markets, lengthen-

Times ing duration, and reduced market liquidity. After June,

3.5 the sell-off became more concentrated along country fundamentals, highlighting the third key vulnerabil- 3.0 ity, (3) slowing growth and rising domestic financial vulnerabilities. 2.5 Currencies and bonds in Brazil, India, Indonesia, South Africa, and Turkey came under intense weaken- 2.0 ing pressure since May as their current account deficits 2004 05 06 07 08 09 10 11 12 persist, inflation remains elevated, and monetary policy Sources: Bloomberg, L.P.; and IMF staff calculations. room seems limited in the face of decelerating growth Note: CEEMEA = central and eastern Europe, Middle East, and Africa; EBIT = earnings (Figure 1.30). The perception of good fundamentals before interest and taxes. Computed as the median of all available ﬁrm data. Firms with negative net debt were excluded. and prudent approaches to macroeconomic and fiscal policies, together with robust financial systems, have contributed to resilience. For example, Chile, Mexico, and Poland fared relatively better with their local and hard currency bond spreads over U.S. Treasur- ies remaining within their long-term range. (See also Box 2.2 in the May 2013 Western Hemisphere Regional Economic Outlook about the role of exchange rates in capital outflows.) The pattern of volatility in emerging markets continues to be driven by expectations of monetary policy in the United States. Following the Federal Reserve’s

22 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.27. China: Corporate Sector Fundamentals Figure 1.28. External and Domestic Vulnerabilties

1. Proﬁtability of Listed Nonﬁnancial Companies CEEMEA Asia Latin America (Percent) 10 Share of loss-making companies (based on negative RoA) Malaysia Median return on assets (right scale) 8 20 7 6 18 Philippines Russia 6 4 16 China 2 5 14 Hungary Thailand Mexico 0 12 Indonesia 4 Chile –2 10 Brazil

Colombia 2010–12 (percent of GDP) 3 South Africa India –4 8 Poland Average current account balance, Romania –6 6 2 Ukraine Turkey

4 Greater external and –8 domestic vulnerabilities 1 2 –10 –8 –6 –4 –2 0 2 4 6 8 10 12 14 16 0 0 Real credit growth in excess of GDP growth, average, 2010–12 (percent) 2007 08 09 10 11 12 13:Q1 Sources: IMF, International Financial Statistics and World Economic Outlook databases. Note: CEEMEA = central and eastern Europe, Middle East, and Africa. 2. Median EBIT/Interest Expense by Firm Leverage

2007 2012

12 Least leveraged Most leveraged Figure 1.29. China: Credit Developments companies companies Stock of broad credit (percent of GDP; left scale) 10 Share of new credit from sources other than bank loans (percent of ﬂow; right scale) 200 60

8 50 180 6 40 160 4

EBIT/interest expense (percent) 30

2 140 20

0 0 1,010 120 Rolling window of 200 companies ranked by debt/asset ratio 10

Sources: WIND; and IMF staff calculations. Note: EBIT = earnings before interest and taxes; ﬁrm leverage = total liabilities/total 100 0 assets; RoA = return on assets. Top panel is computed for a balanced panel of 2,146 Dec. 2003 Dec. 05 Dec. 07 Dec. 09 Dec. 11 companies. Data for 2013:Q1 for the RoA are annualized, but may somewhat overstate the deterioration in performance, as a result of seasonal effects. Bottom panel is computed for a balanced panel of 1,210 nonﬁnancial companies. Sources: CEIC data; Haver Analytics; and IMF staff calculations. Note: Broad credit comprises bank loans, entrusted loans, trust loans, acceptance bills, and corporate bonds.

International Monetary Fund | October 2013 23

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.30. Recent Stress in Emerging Markets

Yields rose the most in the economies that had the greatest declines. Countries with macroeconomic weaknesses, such as high inﬂation…

CEEMEA Asia Latin America

1. Change in 10–Year Local Currency Government Bond Yields 2. Inﬂation and Exchange Rates (Basis points) 400 2 ISR TUR CHN 0 350 ROM POL HUN –2 300 IDN COLCHL –4 BRA PER 250 –6 COL THA ZAF RUS 200 PHL MEX –8 PER ZAF HUN MEX –10 150 POL MYS TUR THA –12 May 22–Sept. 5, 2013 5, May 22–Sept. 100 IND BRA MYS RUS –14 CHN 50 ISR May 22–Sept. 5, 2013 (percent) PHL IDN IND –16 CHL ROM 0 –18 200 0 –200 –400 –600 –800 –1000 0 2 4 6 8 10 Foreign exchange change versus U.S. dollar, March 31, 2009–April 30, 2013 (reverse scale) Inﬂation rate (year over year, May–July average; percent)

…unfavorable growth and inﬂation dynamics… …and external imbalances, underwent the most pressure...

3. Growth, Inﬂation, and Bonds 4. Current Account and Exchange Rates 400 2 ISR CHN TUR 0 350 POL ROM HUN –2 300 IDN CHL –4 PER 250 BRA COL RUS –6 versus U.S. dollar, COL ZAF PHL 200 MEX THA –8 ZAF PER 150 HUN POL MYS –10 change (basis points) MEX TUR BRA –12 100 IND THA MYS RUS CHN –14 50 ISR 2013 (percent) 5, May 22–Sept. IND –16 ROM CHL PHL IDN Bond yield change, May 22–Sept. 5, 2013 5, May 22–Sept. Bond yield change, 0 –18 –6 –4 –2 0 2 4 6 –8 –6 –4 –2 0 2 4 6 Foreign exchange Growth-inﬂation differential (percent; latest quarter) Current account (percent of GDP, latest four-quarter average)

…with potential feedback to credit markets.

5. Exchange Rates and Credit Markets 200 IND 160 IDN TUR 120

80 BRA ZAF MYS PHL PER RUS CHN MEX 40 COL HUN ROM THA Credit default swap spread change, Credit default swap spread change, May 22–Sept. 5, 2013 (basis points) 5, May 22–Sept. CHL POL ISR 0 –18 –14 –10 –6 –2 2 Foreign exchange change, May 22–Sept. 5, 2013 (versus U.S. dollar; percent)

Sources: Bloomberg, L.P.; and national authorities. Note: CEEMEA = central and eastern Europe, Middle East, and Africa; BRA = Brazil; CHL = Chile; CHN = China; COL = Colombia; HUN = Hungary; IDN = Indonesia; IND = India; ISR = Israel; MEX = Mexico; MYS = Malaysia; PER = Peru; PHL = Philippines; POL = Poland; ROM = Romania; RUS = Russia; THA = Thailand; TUR = Turkey; ZAF = South Africa.

24 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 Making the Transition to Stability

decision in September to delay tapering of its asset Figure 1.31. May 2013 Sell-Off of Emerging Market purchasing program, emerging market bond yields and Bonds versus the Lehman Brothers Episode spreads over U.S. treasuries declined, and currencies Change in assets under management (index; left scale) reversed some of their earlier declines against the U.S. Change in emerging market yields (basis points; right scale, reversed) dollar. Primary issuance of corporate and sovereign August 27, 2008 = 0 May 22, 2013 = 0 bonds picked up significantly, and flows into emerging 110 –40 market debt funds restarted in late September.

100 0 What would happen if flows reversed more sharply in emerging markets? 90 40 These factors suggest that emerging markets may have become more vulnerable during the transition to a more challenging external financing environment. In 80 80 the 12 weeks following the May 22, 2013, reversal of risk sentiment, assets under management for emerging 70 120 market fixed-income funds fell 7.6 percent (or $19 billion). This pullback was much smaller compared with the one accompanying the systemic financial shock in 60 160 –2 0 2 4 6 8 –2 0 2 4 6 8 10 12 2008, when assets under management fell by 36 percent (or $26 billion) during the first round of the asset Weeks sell-off in September–October 2008 (Figure 1.31). Yet Sources: EPFR Global; JP Morgan; and IMF staff calculations. the impact on local currency bond yields was similar across the two episodes, which suggests that emerging markets are highly vulnerable to sudden outflows that Figure 1.32. Estimated Impact on Bond Yields from a Reversal would further strain liquidity conditions. of Capital Flows and Other Factors A pricing model is used to highlight a stress (Basis points) scenario in which 10-year bond yields are explained External Domestic Total by domestic and external variables. An external 300 shock consisting of a 30 percent reduction of current foreign holdings of local currency government debt, 250 an increase of 100 basis points in the U.S. treasury 200 note yield, and a 10-percentage point increase in the Chicago Board Options Exchange Market Volatility 150 Index (VIX), and domestic variables along the 100 October 2013 WEO forecasts for 2014 (for debt-to- GDP ratios, real GDP growth and fiscal balances), as 50 well as unchanged monetary policy rates would result 0 in substantial increases in government bond yields in several countries (Figure 1.32). Yields on 10-year –50 bonds in Indonesia, South Africa, and Turkey would –100 increase by more than 150 basis points, all mostly attributable to external factors, while most countries’ –150 bond yields would increase by more than the U.S. Chile Israel Brazil Turkey Russia Poland Mexico Hungary Thailand Malaysia Colombia

Treasury note yield change. Indonesia Philippines Domestic policies can counteract the rise in term South Africa premiums, such as in Colombia, Mexico, and the Source: IMF staff calculations. Note: The shock consists of a 100 basis points rise in U.S. Treasury yields, a 30 percent Philippines, or add to external woes, like in Indonesia, reduction in foreign holdings, and a 10 percentage point rise in the VIX. Values of domestic South Africa, and Turkey (red portions of the bars in variables are those used in the October 2013 WEO forecasts for 2014 (for debt-to-GDP ratios, real GDP growth, and ﬁscal balances), and the policy rate is assumed unchanged.

International Monetary Fund | October 2013 25

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.32).20 The simulation underscores the need Maintaining central bank credibility is paramount for emerging markets to rebuild resilience and address in times of increased risk aversion, so monetary policy vulnerabilities. More broadly, the ongoing rise in yields recommendations hinge on inflation expectations. and credit spreads and the depreciation of emerging Countries with well-anchored inflation and inflation market currencies could impose further refinancing expectations may have more room for policy easing and default risks on firms with inadequate debt-servic- or less tightening to withstand the cyclical growth ing buffers, although looser domestic monetary policy slowdown. The scope for easing may be very limited may offset some of the higher risk premiums. in countries with high inflation pressures, which may have to do more to anchor inflation expectations. What actions can emerging market countries take? Brazil, India, and Indonesia have tightened monetary conditions to address inflation pressures. The episodes of financial market turmoil in the second Policymakers should carefully monitor and contain and third quarters of 2013 underscore that some the rapid growth of corporate leverage. Also, local bank emerging market economies need to address macro- regulators need to guard against foreign currency fund- economic imbalances, enhance policy credibility, and ing mismatches building up directly on bank balance rebuild policy space to reduce vulnerabilities as finan- sheets, or indirectly through foreign currency borrow- cial conditions normalize. Emerging market econo- ing by firms. mies need to make a transition to a more balanced Containing the risks to China’s financial system is as and sustainable financial sector, while maintaining important as it is challenging. As elaborated in the IMF’s robust growth and financial stability. These actions will China 2013 Article IV Staff Report (IMF, 2013b), broad position them to effectively withstand future market credit growth needs to be reined in to contain financial turbulence. stability risks and promote the rebalancing of China’s In the event of significant capital outflows, and with economy away from credit-fueled investment. However, a elevated emerging market contagion risk, policymak- sudden credit squeeze could further decelerate economic ers can take various actions to mitigate potential activity and trigger serious asset quality problems. The damage. Depending on the extent of outflows and spike in interbank market rates in June 2013 illustrates liquidity pressures in market segments, some countries the risks from policies that are not clearly communicated. may need to act to ensure orderly market operations, Similarly, introducing default risk to the financial system such as using cash balances, reducing the supply of will be critical for sustainable market development, but long-term debt, and performing switching auctions steps in this direction need to be finely calibrated to avoid to temporarily reduce supply on the long-end of yield causing a full-blown run on new investment products. curves. Reversing macroprudential tightening measures Against this backdrop, it is important for the following and/or previous restrictions on capital inflows may also actions to be taken: help maintain orderly conditions. • Tighten prudential oversight, especially of shadow Exchange rates should be allowed to depreciate in banking activity, while removing incentives for response to changing fundamentals but policymakers regulatory arbitrage through continued financial need to guard against disorderly adjustment. Brazil’s liberalization (for example, of deposit interest rates); announcement of a transparent, but temporary, foreign • Enforce stronger disclosure practices for new finan- exchange intervention program to dampen the uncer- cial products, and counteract the current pattern of tainty around intraday currency volatility is a step in implicit guarantees and bail-outs; and that direction. In addition, emerging market econo- • Use on-budget fiscal stimulus toward boosting con- mies may benefit from establishing swap lines with sumption if economic growth starts falling signifi- major central banks to remove liquidity shortages in cantly short of the target. foreign exchange markets.

Japan’s bold policies 20 The size of the improvement of domestic policies in Poland may The firing of the monetary arrow of “Abenomics” by be overstated by the decline in the debt-to-GDP ratio owing to the transfer of the government bond holdings of the pillar II pension the Bank of Japan (BoJ) in April 2013 reverberated fund assets to general government debt. through domestic markets and the banking system,

26 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

boosting equities but increasing bond volatility.21 Full implementation of Abenomics would likely lead The weakening of the yen before and after the BoJ’s to an increase in capital outflows to both advanced and action reflected expectations for eventual outflows and emerging market economies (Lam, 2013). Japanese substantial spillovers to both emerging market and households and institutions already have substan- advanced economies. If the other two reform arrows tial holdings of foreign assets, totaling ¥542 trillion (fiscal and structural) are effectively deployed, and ($6.2 trillion) at the end of 2012, or 114 percent of efforts at pulling the economy out of deflation are GDP (Table 1.4). A return of Japanese flows to peak successful, major gains to financial stability could historical rates could have significantly positive effects occur. But if policy follow-through is inadequate, new for some of the receiving markets and could even com- risks to domestic and global stability could arise. pensate for net redemptions prompted by monetary tightening elsewhere. What would the success of Abenomics mean for Flows to emerging markets are likely to be led financial stability? by individual investors, who are already moving to increase their foreign currency exposures. The willing- Successful implementation of the full Abenomics ness of individuals to take on emerging market risk policy framework—consisting of the three arrows of has risen sharply in recent years, supported by the monetary stimulus, flexible fiscal policy, and structural development of new investment products. Among the reform—would have important benefits for stability. most popular are currency overlay funds (Figure 1.33), As projected in the “complete Abenomics package” which are structured products that consist of an out- scenario of the October 2013 WEO, effective deploy- right investment in an underlying asset such as domes- ment of all three arrows would raise inflation and infla- tic equities, compounded with a derivative exposure to tion expectations toward the BoJ’s target of 2 percent a high-yielding emerging market currency. Such funds and would increase domestic investment and credit have continued to receive inflows even during periods demand. Banks would continue to scale back their of yen strength, and now total more than ¥10 trillion, bond holdings,22 and the nominal 10-year Japanese up from only ¥1 trillion in 2009.24 Other emerging- government bond (JGB) rate would shift up toward 3 market-oriented investments include foreign currency percent. Capital outflows would accelerate, possibly to positions held by retail traders, and some broader historically high rates, prompted by a new search for investment funds that do not feature a specific overlay. yield and the scarcity of domestic government bonds. Under a complete Abenomics scenario, outflows to Under the scenario described here, the vulnerabil- developed markets would also increase, led by con- ity of domestic banks to bond market shocks would servative investors such as life insurance companies likely decline. BoJ purchases during the next two years and pension funds. Purchases of developed market should reduce the total amount of JGBs available to assets, largely investment-grade bonds, would take the market (the current market structure is shown in longer to develop, because these conservative institu- Table 1.3). Accordingly, if all aspects of Abenomics tions often have extensive approval processes for major are successfully implemented, the interest sensitivity of portfolio reallocations. Japanese purchases of some both regional and major banks would be expected to specific classes of assets, such as higher-grade euro area decline sharply as those institutions shift their portfo- government bonds, as well as other G7 bonds, could lios toward foreign asset purchases and more domestic be significant. Japanese banks have already stepped up lending to meet increased credit demand.23 acquisition of foreign assets (see Table 1.4), both loans and direct investment, in some cases filling in for deleveraging European banks. Major city banks have been 21“Abenomics” refers to a set of economic policies advocated by especially active on this front, acquiring retail banking Prime Minister Shinzo Abe. The “three arrows” is the symbolic name operations in developing Asia, Latin America, and the given to the three foundational pillars of the plan. 22Major city banks sold more than ¥15 trillion ($150 billion) in United States. These capital outflows improve financial government bonds, about 14 percent of their overall portfolio of JGBs, in April and May 2013, following the BoJ’s April 4 policy announcement, and bond market volatility increased sharply. 24As an overall gauge of the scale of these holdings, Japan’s current 23This analysis, the results of which are presented in Table 1.5, account surplus is projected to be ¥7 trillion (1.3 percent of GDP) is based on Arslanalp (2013). See also the October 2012 GFSR in 2013. The steady-state surplus is somewhat higher, at about 1.7 discussion. percent of GDP.

International Monetary Fund | October 2013 27

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 1.3. Structure of the Japanese Government Bond Market Stock, end-2012 Share of JGB Market Share of Own Assets (trillions of yen) (percent) (percent) Banks 299 38.1 18.2 City banks 102 13.0 22.0 Regional and Shinkin 43 5.5 16.2 Insurers and Pensions1 277 35.3 39.3 Investment Trusts and Households 24 3.1 50.9 Foreign 35 4.5 8.8 Other 57 7.3 . . . Bank of Japan 91 11.6 54.0 Total 783 100.0 . . . Source: Bank of Japan, Flow of Funds. 1Includes Government Pension Investment Fund. Note: JGB = Japanese government bond.

Table 1.4. Foreign Assets Held by Japanese Investor Figure 1.33. Japanese Flows into Currency Overlay Funds Groups, end-2012 (Billions of yen; monthly) (Trillions of yen) Domestic equity funds Total funds Foreign Assets Net Purchases in 2012 (Flows) 500 Banks 145 10.1 Insurers and Pensions 115 0.7 Households 8 2.0 Investment Trusts 57 –1.7 400 Nonfinancial Corporations 111 17.1 Government1 105 –0.5 Total 542 27.7 Source: Bank of Japan, Flow of Funds. 300 1Excludes Government Pension Investment Fund.

200 stability in Japan through portfolio diversification and 136 frequently in destination markets, where they may 100 compensate for net sales by other investors. 47 0 Incomplete implementation of Abenomics would pose risks to banks. –100 Dec. Mar. 11 Jun. 11 Sep. 11 Dec. 11 Mar. 12 Jun. 12 Sep. 12 Dec. 12 Mar. 13 Jun. 13 The promising start for Abenomics could still end in 2010 disappointment if support from fiscal and structural reforms is not forthcoming. In such a case, described Sources: Bank of Japan; and IMF staff estimates. in the October 2013 WEO as an “incomplete” scenario, initial success in raising inflation and inflation expectations could eventually be followed by a decline of inflation below the 2 percent target, and domestic such a scenario would be rising susceptibility to inter- credit demand could falter. Banks may return to their est rate shocks (Figure 1.34). Associated risks, such previous course of accumulating government bonds as simultaneous large sales of domestic bonds due to (Table 1.5A), equity prices could dip, and capital value-at-risk (VaR) “model herding,” could persist or outflows subside. even increase. The shift into an incomplete scenario would revive long-standing financial stability concerns about banks’ A “disorderly” scenario with high risk premiums would accumulation of government bonds. In this sce- pose numerous stability and spillover risks. nario, city banks are projected to initially scale back government bond holdings in response to the BoJ’s Failure to deliver on key components of the ambi- bond-buying program. But these reductions would tious reform agenda could also have a more pernicious eventually be reversed as banks absorb the extra bond downside. Market disillusionment could lead to fiscal issuance needed to sustain economic growth, while and inflation concerns, particularly if medium-term domestic loan demand stagnates. The consequence of fiscal adjustments are not completed and the structural

28 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Table 1.5. Japan Scenarios: Complete, Incomplete, and Disorderly A. Net JGB Purchases (trillions of yen) Complete Incomplete Disorderly Banks –55 –21 –60 Insurers and Pensions 10 8 8 Households and Investment Trusts 0 0 0 Foreigners 12 –3 9 Bank of Japan 100 100 100 Ministry of Finance –67 –84 –57 B. Medium-Term Outcomes (percent) Complete Incomplete Disorderly Average Inflation Rate, 2013-17 1.7 1.0 3.6 Average Growth Rate, 2013-17 1.4 0.9 0.9 Ten-Year JGB Rate in 2017 3.2 2.9 6.2 Equity Market Change to 2017 50 –10 –10 C. Flows to Emerging Markets (trillions of yen) Investor Group Complete Incomplete Disorderly Memo: Stock1 Toshin Emerging Market Portfolio 1.8 0.2 3.5 3.6 Toshin Emerging Market Overlay 2.0 1.3 4.0 5.5 Bank FDI 0.5 0.4 0.5 2.8 Bank Portfolio 0.4 0.4 0.4 1.8 Bank Loans 1.0 –0.1 1.0 19.1 Corporate FDI 1.3 1.3 2.6 13.0 Total 7.0 3.5 12.0 45.7 Source: IMF staff estimates. Note: For complete scenario, outflow in each is maximum historical, except Toshin overlay. For disorderly scenario, outflow is twice maximum historical for Toshin and corporates; maximum historical for banks. EU = European Union; FDI = foreign direct investment; JGB = Japanese government bond. 1Stock of foreign assets at end of 2012.

reform arrow is never fired. In this “disorderly” alterna- triggering a wave of selling, which would, in turn, tive to the incomplete scenario (Table 1.5B), calculated prompt further volatility spikes and price declines. using the same analytical framework as the other two Strains could develop in the banking system. In the scenarios, banks would continue to sell government disorderly scenario, banks would experience pres- bonds at a faster rate than in the complete scenario, sure from withdrawals as households scale back bank and capital outflows would accelerate to record rates, deposits (now 55 percent of their financial assets) in led by outflows from individual investors. Risks to favor of higher-yielding instruments, such as foreign financial stability would escalate sharply because infla- bonds. A lack of profitable lending opportunities at tion and risk premiums on government bonds would home would limit revenues, thus squeezing margins rise to levels well beyond those experienced in recent and shrinking capital buffers. Further pressure would decades. come from mark-to-market losses on remaining bond The chances of a large “VaR shock” could increase holdings, which would reduce the Tier 1 capital ratios sharply. Although measured VaR spiked during bouts of regional banks to 6 percent from 10 percent, and of bond market volatility in April and May 2013, those of major banks to 9 percent from 12 percent. few major banks appear to have hit their VaR limits Weak domestic conditions would likely accelerate during this period (Figure 1.35). In part, these limits outflows to both advanced and emerging markets. were not hit because other major assets such as equities With limited opportunities for funneling savings were registering gains even as bond prices dropped, so into the domestic stock market or domestic lending, that overall portfolio volatility did not rise as much as individuals, banks, and companies would be even it otherwise would have.25 However, in a disorderly more inclined to shift capital offshore. The lack of a scenario in which prices of most asset classes decline, recent history of substantial inflation in Japan makes this dampening effect might not come into play. Joint it difficult to project outflows in the disorderly case, declines in bond and equity prices could exacerbate but given the availability of numerous foreign invest- portfolio volatility, forcing up the measured VaR, and ment channels through an open financial account, a large increase could be possible at a rate well beyond 25 Rather, bond sales were precipitated by losses to capital, market that of the complete scenario (Table 1.5C). Based on uncertainty, and a desire to shrink exposures before VaR limits became binding. This stands in contrast to the VaR shock of 2003, recent flows, the net increase in exposure to emerg- when binding internal limits forced fire sales of bonds. ing market currencies could be considerably more

International Monetary Fund | October 2013 29

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.34. Japanese Banks’ Sensitivity to an Interest Rate Figure 1.35. Value-at-Risk in the Disorderly Scenario

Shock Daily portfolio VaR (left scale, reversed) (Percent of Tier 1 capital; response to a 100 basis point shock) 10-year JGB three-month annualized volatility (left scale) 35 Equity bond three-month correlation (right scale) Projected Scenarios 100 (2) ... raising selling (3) ...which in turn 1.0 Complete 30 pressures ... pushes up volatility Incomplete and VaR further. Disorderly 80 0.8 25 60 0.6

20 If the correlation 40 between bonds and 0.4 Regional banks equities rises, VaR 15 would be higher for the 20 same level of bond 0.2

(1) Higher bond volatility volatility. Correlation Basis points 10 0 increases VaR... 0.0 Major banks

5 –20 –0.2

0 –40 –0.4 2002 04 06 08 10 12 14 16 Fiscal Year –60 –0.6 Dec. Jan. Feb. March April May June July Scenario 2012 13 Sources: Bank of Japan; and IMF staff estimates. Source: IMF staff estimates. Note: JGB = Japanese government bond; VaR = value at risk.

than the ¥7 trillion ($70 billion) a year that represents Successful deployment of the three arrows of reform previous periods of heavy outflows (Table 1.5C and would support domestic financial stability, but Figure 1.36), even under the assumption that advanced incomplete implementation could bring new risks. economy assets would make up the bulk of new purchases.26 Popular targets for recent outflows have The success of the Japanese government’s economic included higher-yielding and more liquid currencies, revitalization efforts would yield dividends for domestic such as the Brazilian real, Mexican peso, Indonesian financial stability, notably by reducing interest rate risks rupiah, and Turkish lira. As projected in Figure 1.36, to the banking sector, improving portfolio diversifica- annual flows from Japan into these fixed-income tion, and dampening volatility. Beyond the broad policy markets could be significant, amounting to as much as framework of Abenomics, certain specific changes in 8 percent of the overall government bond market and market structure would help mitigate risks. Technical more than 30 percent of foreign holdings in the case of adjustments in derivatives markets, including widening Turkey. Such investments, particularly those employ- tolerance zones for the operation of circuit breakers, ing structured products, can be volatile, raising the could increase the usefulness of available hedging instru- prospect of increased volatility for currencies and asset ments. VaR models could be further adjusted to reduce markets in emerging markets. herding behavior. Regional banks should strengthen their capital bases to take better advantage of the BoJ’s increased purchases of JGBs and increase lending to households and corporates. On the external front, regulators need to be conscious of the potential for risky structured products, such as currency overlay funds, to 26The largest increase in outflows would be among individual investors, which is the group with the highest average share of for- generate sudden price movements, large losses on house- eign assets in emerging markets. hold balance sheets, and spillovers to other markets.

30 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.36. Projected Flows to Selected Emerging Bank (ECB) and credit support to viable firms by Markets in the Disorderly Scenario the European Investment Bank are crucial to pro-

Flows (billions of U.S. dollars) vide time for the repair of private balance sheets. Flows as share of market size (percent) Flows as share of ﬁnancial holdings in market (percent) The ECB’s Outright Monetary Transactions (OMT) framework has increased confidence that policymakers 35 will avoid tail risks. Initial progress has also been made 30 on banking union, including through the Single Super- visory Mechanism, political agreements on the Euro- 25 pean Stability Mechanism framework for direct bank recapitalization, and the Bank Recovery and Resolution 20 Directive. This progress has helped ease the severe mar-

15 ket pressures that had been weighing down on weaker sovereigns and banks, stabilize bank deposits, staunch 10 capital flight, and narrow Target2 imbalances.28

5 Weak banks have been reinforcing the problems of weak corporates. 0 Brazil Indonesia Mexico South Africa Turkey

Sources: Haver Analytics; and IMF staff estimates. Nonetheless, financial fragmentation within the euro area has persisted,29 reinforcing an adverse feedback loop between weak banks, corporates, and sovereigns in stressed economies and entrenching divergence in financial and eco- the euro area banking, corporate, and nomic conditions (Figure 1.37). As a result of this feedback Sovereign nexus loop, along with weak demand for credit, bank lending to Policy actions at the euro area and national levels stressed economies continues to contract, as discussed in have reinforced a collective commitment to the euro. more detail in Chapter 2 (Figure 1.38). Weak banks have This renewed commitment has helped ease the severe been exacerbating the problems of weak corporates because market stresses that had been weighing on sovereigns institutions with thin buffers have been tightening credit and banks. While funding conditions have improved, conditions for corporates by rationing credit and increas- financial fragmentation persists, allowing the adverse ing the interest rates on new loans (Figure 1.39). Evidence feedback loop between banks, corporates, and sovereigns from individual banks suggests that even within stressed to continue in stressed economies.27 While there has economies in the euro area, weaker banks are more likely to been progress on bank repair, weak banks have been cut back lending (Figure 1.40).30 reinforcing the problems of weak corporates, while Sovereign risks have abated, but sovereign spreads weak corporates have been exacerbating the pressures remain differentiated within the euro area (Figure on weak banks. As a result, interest rates on corporate 1.41). Furthermore, spreads widened somewhat dur- loans have remained elevated. Taking steps to reverse ing the recent period of market volatility, though in financial fragmentation will help reduce interest rates most cases they are now tighter than they were at the in stressed economies, but will not be sufficient to 28 resolve the corporate debt overhang. Therefore, it is Target2, the main payment system within the European Mon- etary Union, works through the individual national central banks essential that efforts to repair bank balance sheets and (NCBs) of each of the euro area countries. The settlement of cross- to move to full banking union be complemented by a border payment flows between euro area countries in Target2 results comprehensive assessment and strategy to address the in claims and liabilities for each NCB. The Target2 balance for an NCB is the net of these claims and liabilities. problem of debt overhang in the nonfinancial sector. 29Foreign claims of core euro area banks on stressed economy Further monetary support by the European Central sovereigns, banks, and the nonfinancial private sector are at 40, 38, and 26 percent of their June 2011 peaks, respectively. 30This is consistent with the Bank of Italy’s April 2013 Financial 27The term “stressed economies” generally includes Cyprus, Stability Report, which presents evidence that in 2012 the growth of Greece, Ireland, Italy, Portugal, Spain, and Slovenia, though in some lending to firms was positive for banks with stronger capital ratios parts of the section it may refer to a subset of these economies. and lower funding gaps.

International Monetary Fund | October 2013 31

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

time of the April 2013 GFSR. As discussed in previ- Figure 1.37. Bank-Corporate-Sovereign Nexus ous GFSRs, as well as IMF (2013a), divergence in sovereign spreads has raised funding costs for banks in stressed economies, putting further upward pressure on Increased sovereign lending rates. Second-tier and small banks in stressed contingent liabilities Sovereigns Weaker economic growth economies have been facing the greatest wholesale Highly indebted funding strains, and it is these banks that tend to be Higher funding Higher corporate the main providers of credit to small and medium costs bond yields enterprises (SMEs) (see ECB, 2013b, pp. 67–68). Banks Corporates Weak corporates have exacerbated the pressures at weak Thin buffers Debt overhang, banks. Weak proﬁts

At the same time, weak corporates have exacerbated the Higher nonperforming loans, Constrained lending, problems of weak banks. Corporate leverage increased Lender forbearance Higher interest rates on loans in stressed economies during the boom years, especially in Portugal and Spain, in contrast to the core euro area Source: IMF staff. (Figure 1.42).31 This is particularly the case for SMEs, which tend to have higher leverage than do larger firms interest rates in response to the increased riskiness of cor- (Figure 1.43). Overall, more than three-quarters of porate loans, starting the cycle again. Figure 1.47 shows corporate debt in Portugal and Spain and about half of that bank interest rates tend to be higher in economies in corporate debt in Italy is owed by companies with debt- which corporate risks are higher, as proxied by Moody’s 32 to-assets ratios at or above 40 percent (Figure 1.44). expected default frequencies of publicly traded firms. High to moderate leverage has interacted with weak Furthermore, greater debt-servicing difficulties at SMEs profitability to create debt-servicing difficulties for are reflected in higher interest rates on small bank loans. companies, particularly because sovereign and bank- Banks with weak balance sheets will be less able and ing stress along with other factors that contributed to willing to recognize losses and so will become more financial fragmentation have raised corporate funding likely to forbear on loans. Although some forbearance 33 costs in stressed economies. Overall, almost 50 percent may help ease pressures on individual borrowers, wide- of debt in Portugal, 40 percent of debt in Spain, and 30 spread forbearance poses the risk that banks will devote percent of debt in Italy is owed by firms with an interest scarce resources to unhealthy corporates, crowding out 34 coverage ratio of less than 1 (Figure 1.45). These firms lending to healthier and more productive firms. would be unable to service their debts in the medium In addition, firms facing higher debt-servicing costs— term unless they make adjustments such as reducing caused by high leverage and remaining fragmentation— debt, operating costs, or capital expenditures. have been forced to adjust their businesses, as discussed These debt-servicing pressures—along with a weak in the April 2013 GFSR. In 2012, dividend payments economic environment—have led to an increase in were reduced sharply by Spanish and Italian companies, nonperforming loans, worsening the quality of the assets and large international firms have been selling foreign on bank balance sheets (Figure 1.46). Banks have raised assets.35 In addition, publicly traded firms in Portugal and Spain reduced capital expenditures by over 15 per- 31ECB (2013a) also discusses the accumulation of corporate debt cent (Figure 1.48). Although deleveraging is needed, in the euro area. IMF (2013a) also looks at constraints to growth excessive cutbacks in capital expenditure—especially and credit posed by the negative feedback loop between high private debt and the weak financial sector. amid remaining fragmentation—may further undermine 32A debt-to-assets ratio of 30 percent usually corresponds to a Ba economic growth prospects. credit rating, and a 35 percent debt-to-assets ratio usually corresponds to a B credit rating. 35The need to preserve or obtain investment-grade credit ratings to 33See also Chapter 2. maintain or gain access to capital markets is a critical driver of dele- 34Interest coverage ratio (ICR) is defined as earnings before inter- veraging efforts by large companies in stressed euro area economies. est and taxes (EBIT) divided by interest expense. Interest revenues This is especially the case because rating agencies have tightened or financial revenues are included in the calculation of earnings (and requirements for the ratio of debt to EBITDA (earnings before inter- thus partly offset interest expense). est, taxes, depreciation, and amortization) during the euro area crisis.

32 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.38. Stressed Euro Area Economy Bank Credit Figure 1.39. Bank Buffers and Interest Rates on Corporate (Percent change, cumulative since September 2011) Loans

October 2012 GFSR 7 scenario projections CY 2 GR 6 0 PT 5 Complete –2

–4 4 ES IE IT –6 Baseline 3 –8 NL FR DE AT 2 –10 BE

–12 1 Interest rate on nonﬁnancial corporate loans (percent) Weak Actual –14 0 0 1 2 3 4 5 6 7 –16 Bank buffers (capital and reserves to nonperforming loans) Sep. Dec. March June Sep. Dec. March June Sep. Dec. 2011 2012 2013 Sources: European Central Bank (ECB); ECB Consolidated Banking Data; IMF Financial Soundness Indicators; and IMF staff estimates. Sources: Haver Analytics; and IMF staff estimates. Note: Differences in deﬁnitions of nonperforming loans complicate comparisons across Note: Euro area lending by banks located in Ireland, Italy, Portugal, and Spain, adjusted for economies. Italian nonperforming loans have been adjusted to make them more comparable securitizations. with other economies, following Barisitz (2013). German nonperforming loans are estimated using ECB Consolidated Banking Data. Figure shows data as of July 2013 or latest data available. AT = Austria; BE = Belgium; CY = Cyprus; DE = Germany; ES = Spain; FR = France; GR = Greece; IE = Ireland; IT = Italy; NL = Netherlands; PT = Portugal.

Figure 1.40. Individual Bank Buffers and Lending in Stressed Economies, 2013:Q1

3 Figure 1.41. Euro Area Sovereign Spreads, April–August 2013 (Five-year spreads to German bunds; basis points)

700 2

600 1

500 0 400

–1 300 Lending growth (percent; year over year)

–2 200

–3 100 1st 2nd 3rd 4th 5th Bank buffers (quintiles) 0 AMJJAAMJJAAMJJAAMJJA AM J JA AMJJA Sources: SNL Financial; and IMF staff calculations. Note: The figure shows average lending growth for the individual banks in each quintile. The Portugal Spain Italy Ireland Belgium France figure is based on consolidated data for a sample of almost 70 banks headquartered in Cyprus, Greece, Ireland, Italy, Portugal, Slovenia, and Spain. Bank buffers are defined as the ratio of core Tier 1 capital and loan loss reserves to impaired loans. The figure uses 2013:Q1 Sources: Bloomberg, L.P.; and IMF staff calculations. or latest available data. Note: AMJJA = April, May, June, July, August. Dotted line represents April 2013 level.

International Monetary Fund | October 2013 33

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.42. Leverage Ratios Figure 1.43. Leverage Ratios by Firm Size, 2011 (Debt to EBITDA) (Debt to EBITDA)

Germany France Portugal Italy Spain Large Medium Small 6.5 8

6.0 7 5.5 6 5.0 5 4.5

4.0 4

3.5 3

3.0 2 2.5 1 2.0

1.5 0 2001 02 03 04 05 06 07 08 09 10 11 Germany France Italy Spain Portugal

Sources: Amadeus database; and IMF staff estimates. Sources: Amadeus database; and IMF staff estimates. Note: EBITDA = earnings before interest, taxes, depreciation, and amortization. Note: EBITDA = earnings before interest, taxes, depreciation, and amortization.

Figure 1.44. Share of Debt at Firms with Various Debt-to- Figure 1.45. Share of Debt at Firms with Various Interest Assets Ratios, 2011 Coverage Ratios, 2011 (Percent of total debt) (Percent of total debt)

Above 40 percent 35–40 percent 30–35 percent Below 30 percent Below 1 1–2 2–3 Above 3 100 100

90 90

80 80

70 70

60 60

50 50

40 40

30 30

20 20

10 10

0 0 Germany France Italy Spain Portugal Germany France Italy Spain Portugal

Sources: Amadeus database; and IMF staff estimates. Sources: Amadeus database; and IMF staff estimates.

34 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.46. Nonperforming Corporate Loans Figure 1.47. Corporate Expected Default Frequencies and

Portugal (left scale) Italy (left scale) Bank Loan Interest Rates (Percent; July 2013) Spain (left scale) Total ratio (right scale) 300 14 Small bank loan (left scale) Small bank loan (left scale) Large bank loan (left scale) Large bank loan (left scale) 12 Expected default frequency (right scale) 250 7 12

10 200 6 10 8 5 150 8 6 Percent Billions of euros 4 100 6 4 3 50 2 4 2

0 0 2 2006 07 08 09 10 11 12 13 1

Sources: National central banks; and IMF staff calculations. 0 0 Note: Differences in deﬁnitions complicate the comparison of nonperforming loans across BE AT FI FR DE NL IT ES PT GR economies. Data for Italy show bad loans only. Core euro area Stressed euro area countries

How has the bank-corporate-sovereign nexus affected Sources: Bloomberg, L.P.; European Central Bank; Moody’s Credit Edge; and IMF staff calculations. interest rates on corporate loans? Note: Small loans have a value of €1 million or less. Expected default frequencies (one year) are the average of the 25th, 50th, and 75th percentiles; the sample comprises publicly traded ﬁrms. AT = Austria; BE = Belgium; DE = Germany; ES = Spain; FR = France; GR = Greece; FI In general, banks should price loans so that the inter- = Finland; IT = Italy; NL = Netherlands; PT = Portugal. est rate is greater than the sum of their funding costs, required return on equity backing the loan, and a credit margin. In stressed economies of the euro area, these three These findings are broadly consistent with recent studies, components of interest rates have been affected by several including the ones on Portugal.37 factors: (1) higher sovereign risk, (2) bank balance sheet Conversely, the pass-through of the ECB’s easy mone- health, (3) corporate riskiness, and (4) the economic and tary policy stance has provided some downward pressure policy environment, as illustrated in Figure 1.49. on bank lending rates. Yet monetary policy has been The importance of these factors is assessed econo- insufficient to offset other pressures that have driven up metrically using monthly data over 2003–13 for France, interest rates on loans to SMEs. In addition, deep and Italy, and Spain for interest rates on small loans, many prolonged recessions in Italy and Spain have depressed of which are extended to SMEs.36 The results suggest, as the demand for loans from nonfinancial corporates. expected, that sovereign stress and banking system weaknesses have been the key driving forces behind higher 37 interest rates on small loans in Italy and Spain, particu- IMF (2013a) concludes that “funding costs, credit risk, and leverage have become important determinants of lending rates since the onset of larly from mid-2011 onward (Figure 1.50). Corporate the crisis, particularly for stressed countries, and that these factors appear credit risk is also a significant factor in higher lending to be more relevant for small loans, typically associated with SMEs.” A rates in Italy and Spain (see Annex 1.1 for details). recent study by the Bank of Portugal (Santos, 2013) using data on new loans to nonfinancial firms found that the firm-level z-score indicator (which captures the firm’s credit risk) and bank deposit rates are significantly and positively related to the level of interest rates (after controlling 36The analysis is based on a vector error correction model, which for several loan-, firm-, and bank-specific characteristics). Furthermore, includes money market rates, sovereign stress, and banking and IMF (2013e) identifies sovereign debt crisis and bank funding pressures business cycle variables as endogenous variables that determine equi- as the key determinants of the higher lending rates in Portugal, together librium lending rates, as well as a number of exogenous variables, with weak domestic conditions and profitability in the context of over- including corporate credit risk. (See Annex 1.1 for details.) leveraged private sector balance sheets.

International Monetary Fund | October 2013 35

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.48. Listed Firms: Changes in Debt, Capital Figure 1.49. Factors Affecting Bank Interest Rates on Expenditures, and Dividends Corporate Loans (Percent; end-2011 to end-2012) Worsening economic Capital expenditures Debt (left scale) Dividends (right scale) environment lowers (left scale) ﬁrm income and 5 10 raises ﬁrms’ riskiness. Corporate riskiness Credit margin

0 Deteriorating economy worsens Bank Economic bank asset quality. balance –10 environment sheet Interest rate on –5 and health Cost of policy equity corporate bank loan uncertainty ( E × re ) Sovereign –10 risk Economic outlook –30 Bank affects sovereign funding yields and policy rate. Policy cost –15 rate ( D × rd )

Source: IMF staff. –20 –50 Note: E = proportion of equity backing the loan; r = required rate of return on equity; D = Germany France Italy Portugal Spain e proportion of wholesale funding and deposits backing the loan; rd = marginal bank funding cost. This simpliﬁed loan pricing equation abstracts from tax rates and other expenses or Sources: Worldscope; and IMF staff estimates. beneﬁts related to the loan.

The factor decompositions in Figure 1.50 suggest that estimates are only indicative. That said, if lending rates were sovereign and banking stresses have played an important to decline to the levels consistent with their precrisis spreads role in keeping the lending rates elevated in Italy and over 7-year swap rates (see Figure 1.50), they would be Spain.38 Spanish bank stress had been simmering since about 150 and 200 basis points lower in Italy and Spain, early 2010, a longer period than in Italy, where the respectively. sovereign crisis did not escalate until mid-2011. Figure 1.50 also shows that the contributions of sovereign and Can the corporate debt overhang be resolved by banking stress have declined since the establishment of removing financial fragmentation? the ECB’s OMT framework and because of the reform progress at the national level.39 In contrast, in France, To assess the scale of the current corporate debt sovereign and banking stress have played virtually no overhang—measured as the share of corporate sector role in determining interest rates on corporate loans debt with an interest coverage ratio (ICR) of less than with lending rates driven primarily by monetary policy. 1—a detailed data set covering more than 3 million This framework can be used to estimate the impact of individual companies is used (see Annex 1.2 for more financial fragmentation—the contribution of sovereign details).40 The current debt overhang is estimated to and banking stress—on bank lending rates. If the influence of sovereign and bank stress (the red bars in Figure 1.50) 40ICRs for 2013 are estimated based on a regression of corporate is removed, the current interest rate on new small loans profitability (EBIT over assets) on GDP growth estimates and actual would be about 100 basis points lower in Italy and 160 interest rates on corporate loans (see Annex 1.2). In the case of Portu- basis points lower in Spain. As with any model, these point gal, the estimated ICRs are adjusted using actual 2012 data by sector/ size provided by the Bank of Portugal. Debt is assumed to be constant at 2011 levels throughout the projection period of the exercise. This 38See also Chapter 2. A high degree of interdependence between assumption may overstate the extent of debt overhang estimated for sovereign and banking risks means that any separation of their 2013 in the three economies. The available data for Spain show a respective contributions is bound to be imprecise and dependent on significant decline in corporate debt levels in 2012. However, data on the specific way in which these risks are measured (see Annex 1.1 for reduction in assets are not available, and these are necessary to estimate details). the effect on profitability and, consequently, the debt overhang. In 39In the case of Spain, progress on the restructuring of the bank- addition, price effects of asset sales have to be taken into account, as ing sector has been an important factor in the improvement of price discounts that are likely to be incurred by SMEs and firms under financial conditions. deleveraging pressures would hurt profitability.

36 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.50. Interest Rates on Small Bank Loans and Model-Based Factor Decomposition

Italian lending rates on small loans have remained elevated despite a …reﬂecting elevated banking stress and residual sovereign signiﬁcant decline in swap rates… pressures.

1. Lending Rate and Swap Rate 2. Model-Based Decomposition of Factor Contributions Explaining the Lending (Percent) Rate’s Deviation from Its Mean (Percentage points) Actual lending rate Seven-year swap rate Monetary policy Business cycle Sovereign and bank stress 6 Italy Italy 3

5 2

4 1

3 0

2 –1

1 –2

0 –3 2009 10 11 12 13 2009 10 11 12 13

In Spain, the spread between the interest rate on small loans and the swap …predominantly due to persistent banking strains… rate has also widened well beyond its historical norm… 3. Lending Rate and Swap Rate 4. Model-Based Decomposition of Factor Contributions Explaining the Lending (Percent) Rate’s Deviation from Its Mean 6 Spain (Percentage points) Spain 3

5 2

4 1

3 0

2 –1

1 –2

0 –3 2009 10 11 12 13 2009 10 11 12 13

…while in France, it remained constant. In France, the transmission from monetary policy is the dominant factor keeping lending rates low.

5. Lending Rate and Swap Rate 6. Model-Based Decomposition of Factor Contributions Explaining the Lending (Percent) Rate’s Deviation from Its Mean 6 France (Percentage points) France 3

5 2

4 1

3 0

2 –1

1 –2

0 –3 2009 10 11 12 13 2009 10 11 12 13

Source: IMF staff estimates.

International Monetary Fund | October 2013 37

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

be large, amounting to between 45 and 55 percent in Figure 1.51. Corporate Debt Overhang the stressed economies of the euro area in 2013 (Figure (Debt of ﬁrms with an interest coverage ratio < 1 as a percentage of total debt)

1.51). To gauge the scale of the debt overhang on a 2013 2018, chronic phase forward-looking basis, two scenarios are used: 2018, reversal of fragmentation 2013 average for the core • Chronic-phase scenario. This scenario assumes that "Persistent" debt overhang bank lending rates rise further as stalled delivery 60 on policy commitments leads to persistent financial fragmentation and as credit margins increase, 50 following a deterioration in the economic outlook under the October 2013 WEO alternative baseline 40 scenario.

• Reversal-of-fragmentation scenario. This scenario 30 assumes that sovereign and banking risks abate as

further progress is made toward banking and fiscal 20 union, leading to a decline in corporate funding

costs (in line with the results shown in Figure 1.50). 10 Growth in stressed economies recovers along the

lines of the October 2013 WEO baseline scenario, 0 which assumes an improvement in competitive- Italy Spain Portugal ness on the back of continued implementation of Sources: Amadeus database; and IMF staff estimates. national reforms. Under the chronic-phase scenario, the size of the debt overhang remains broadly unchanged from cur- The systemic nature of the debt overhang in Italy, rent high levels, and corporates fail to escape the debt Portugal, and Spain is further underscored by the fact overhang trap even in the medium term, further sup- that corporate sector strains are not limited to just pressing prospects for economic recovery (see Figure the sectors that experienced credit booms (construc- 1.51). Under the reversal-of-fragmentation scenario, tion and real estate in Spain and Portugal). Estimated the debt overhang is reduced substantially as corporate probabilities of default (PDs) suggest that stresses are profitability benefits from economic recovery under- also high in the cyclical and manufacturing sectors in pinned by structural reforms and favorable financing the stressed economies (Figure 1.52).42 In addition, conditions. However, even when economic growth strains at SMEs are greater relative to those at large picks up and financial fragmentation is reversed, a corporates in Italy, Portugal, and Spain, and also in sizable portion of firms in stressed economies remains France, because large corporates benefit from stronger financially vulnerable. Hence, a more comprehensive fundamentals and financing conditions. approach to address this “persistent” debt overhang, amounting to almost one-fifth of total corporate debt What are the implications of the corporate debt in these three countries, will be required to support overhang for banks? the flow of credit to healthier companies needed for sustained economic recovery.41 This GFSR examines the corporate exposures of banks in Portugal, Spain, and Italy, as these are systemically 41The “persistent” debt overhang is the share of debt in the important economies where the corporate debt over- stressed economies owed by financially vulnerable firms (those hang is sizable and where firm-level data are sufficiently with an ICR of less than one) under the reversal-of-fragmentation comprehensive to carry out this type of exercise. scenario, in excess of the equivalent share of debt in the core euro area economies. The core euro area is chosen as a benchmark because This analysis provides an illustration of the potential the debt-at-risk levels in the core have been relatively stable before magnitude of corporate risks for banking systems, and throughout the crisis (see Annex 1.2) and under the reversal- thus making the strong case that the ECB’s upcoming of-fragmentation scenario, corporates in the stressed economies bank balance sheet assessment should, among other and the core euro area will face similar financial conditions. The cross-country differences in the industrial structure per se should not lead to divergent levels of debt-at-risk across countries with similar economic and financial conditions. 42See Annex 1.2 for details.

38 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.52. Distribution of Estimated Corporate Sector Probabilities of Default (2011, over the next two years; based on interest coverage ratios of nonﬁnancial ﬁrms)

Large Small and medium enterprises Construction Manufacturing Cyclicals Construction Manufacturing Cyclicals Germany 1 1 1 1 1 2 France 1 2 1 3 4 3 Italy 2 3 1 3 3 3 Spain 2 2 2 4 4 4 Portugal 3 2 4 4 4 4

Source: IMF staff estimates. Note: The numbers indicate quartiles for the distributions of probabilities of default (PDs) across countries, sectors, and firm sizes. Segment-specific PDs are weighted averages of firm-specific PDs. Manufacturing includes manufacturing, utilities, and information technology. Cyclicals include wholesale and retail trade and all services.

things, focus on corporate exposures.43 It is important it to capture uncertainties about collateral valuations to note, however, that to properly assess potential bank and recoveries.45 Because the LGD assumptions are losses, a detailed bank-by-bank asset quality review and exogenous and the same for all countries, they may not stress test is required, which is a different and a more capture some country-specific circumstances, including precise exercise than the one presented in this report. ongoing bank restructuring processes. The forthcoming bank balance sheet assessment and Assuming no further improvement in economic and stress tests provide a golden opportunity to carry out a financial conditions—which would correspond to a more comprehensive and transparent evaluation across euro adverse outcome than the cyclical improvement built into area banks that could help restore investor confidence the October 2013 WEO baseline—some banks in the in the quality of their balance sheets. stressed economies could face sizable potential losses on The analysis in this report aims to assess the impact their corporate exposures. Figure 1.53 presents estimates of corporate strains on banks in the stressed economies of potential losses over the next two years for the bank- from the corporate sector balance sheet perspective. ing systems in Portugal, Spain, and Italy and compares It maps corporate vulnerability indicators (such as them with banks’ estimated total loss-absorption capac- ICRs) into historical default rates to estimate firm-level ity, which includes current provisions for corporate probabilities of default (PDs). The country-level PDs loans, future pre-provision earnings, and capital buffers are then calculated as weighted averages of the firm- (green bars in Figure 1.53).46 level PDs.44 Finally, the bank losses by country are Based on this indicative exercise for the more adverse estimated as the product of the country-level PDs, an outcome and under the 45 percent LGD assumption, the assumed loss given default (LGD) rate, and the stock Spanish banking system could face an estimated €104 bil- of corporate loans in the banking system. The poten- lion of gross losses on corporate exposures, but this is fully tial losses for banks operating in Italy, Portugal, and covered by existing provisions. Following several asset Spain are estimated for 2014–15 based on projected quality reviews and stress tests, Spanish banks have signifi- corporate sector vulnerability indicators as of 2013 cantly increased provisions, especially on construction and (Figure 1.53). A range of potential losses is estimated real estate exposures. In the case of Italy, the estimated using a standard Basel LGD of 45 percent as the gross losses on corporate exposures could amount to €125 mid-point and a 10 percentage point variation around billion, which exceeds existing provisions by €53 billion. As Figure 1.53 illustrates, these estimated net losses (€53 43The upcoming balance sheet review by the ECB will cover a billion) are covered by operating profits without erod- wider range of assets, including those in other euro area countries and stemming from other types of exposures. ing existing capital buffers, under the 45 percent LGD 44In the absence of more precise bank-level information on corporate loan portfolios, ICRs are mapped into PDs by (1) assigning 45In the case of Spain, the stress test carried out by Oliver Wyman implied credit ratings to companies in the sample based on average has reduced uncertainty about collateral valuations. ICRs by credit rating for companies rated by Moody’s and (2) 46Spain’s operating profits include domestic operating profits and assigning PDs to each implied rating based on historical default rates foreign net profits (after provisions and taxes abroad), while provi- of companies rated by Moody’s (see Annex 1.2 for more details). sions refer to business in Spain only.

International Monetary Fund | October 2013 39

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: Transition Challenges to Stability

Figure 1.53. Potential Losses on Corporate Loans and Banking System Buffers (For the next two years, based on projected corporate vulnerability indicators as of 2013; billions of euros)

Losses Buffers Losses in excess of provisions (net) Tier 1 capital Losses covered by provisions Operating proﬁts (2010–12 average x 2) Provisions on corporate loans 2. Italy 3. Portugal 1. Spain 60 350 300 300 250 50

250 200 40 200 150 30 150 100 20 100 50 50 10 0 0 0 35 45 55 35 45 55 35 45 55 percent percent percent Buffers percent percent percent Buffers percent percent percent Buffers LGD LGD LGD LGD LGD LGD LGD LGD LGD Losses Losses Losses

Sources: Amadeus database; Financial Soundness Indicators database; national central banks; and IMF staff estimates. Note: LGD = loss given default. Spanish potential losses are adjusted for transfers to SAREB, the Spanish asset management company that took over the bad debt of restructured banks during the global financial crisis. Spanish operating profits include domestic operating profits and foreign net profits (after provisions and taxes abroad). Spanish provisions include about €4 billion of dynamic provisions. assumption. For Portugal, the estimated gross losses on in operating costs and reductions in dividends, will corporate exposures could be €20 billion, or €8 billon in also help improve profitability and/or boost capital. excess of existing provisions. As Figure 1.53 illustrates, However, as mentioned previously, provisioning and/or these estimated net losses (€8 billion) could be covered by capital needs can only be ascertained precisely through operating profits without eroding existing capital buffers, a bank-by-bank asset quality review that looks into under the 45 percent LGD assumption.47 individual bank loan portfolios and takes into account Despite recent efforts to assess asset quality and boost provisions and capital held by each bank. provisions, this analysis suggests that some banks in Specifically, the analysis in this report differs from the the stressed economies might need to further increase standard bank solvency stress tests in several important provisioning to address the potential deterioration of respects: (1) it considers gross corporate sector exposures of asset quality on their corporate loan books, which could a banking system, including both performing and nonper- absorb a large portion of future bank profits.48 Recently forming loans (and hence, both expected and unexpected increased capital provides additional loss-absorption losses), whereas bank stress tests tend to focus on perform- capacity, if needed.49 Further measures, such as cuts ing loans (unexpected losses); (2) it relies on PDs derived from the detailed firm-level data on corporate sector vul- 47 Buffers on domestic corporate exposures may be overestimated nerabilities using the same methodology for all economies, because provisions (including generic provisions), operating profits, and Tier 1 capital data are available only on a consolidated basis whereas solvency stress tests typically use country-specific at the system level. Also, some of the losses may be borne by the PDs based on national historical default rates and models, household sector, as some SMEs may be able to draw on their own- as well as country-specific LGDs; and (3) it does not rely ers’ personal wealth. 48The central bank of Portugal has conducted three in-depth asset on bank-specific data and is not suitable for assessing bank quality reviews with support from external consultants, including a capital needs (see Annex 1.2 for more details). Hence, the detailed review of construction and real estate exposures (39 percent main goal of this exercise is to illustrate the potential scale of the corporate sector), as well as a recent review of large exposures and collateral valuations (49 percent of total assets). Both reviews of the asset quality issues in banks’ corporate exposures for identified some shortfalls that were subsequently addressed. Similarly, the forthcoming bank balance-sheet review to focus on. the central bank of Italy has evaluated provisioning in selected banks (see Box 1.4 for more details). 49In some cases, banks are also able to provision against future ing to banks’ medium-term funding and capital plans. In general, losses. Core Tier 1 ratios of several banks in countries with IMF implementation of forward-looking provisioning rules is, however, programs are comfortably above the hurdle rates set under the being undermined by the stalled attempts to adopt forward-looking baseline and stress test scenarios on a forward-looking basis, accord- impairment loss recognition in accounting standards.

40 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

What needs to be done to address bank weaknesses and • Expediting improvements to corporate bankruptcy the bank-sovereign negative feedback loop? frameworks in stressed economies to allow for swifter court processes, to provide clarity of collat- Investors’ faith in euro area banks’ balance sheets must eral ownership and the exercising of rights over secu- be fully restored: rity, and to encourage out-of-court debt resolutions • A first step will be to conduct a comprehensive and and write-offs, as recently done in Portugal.50 rigorous bank balance sheet assessment and stress • Taking a more comprehensive approach to corporate test, with involvement of independent, third-party debt cleanup. Where warranted, measures could auditors, as planned by the ECB. include establishing a special asset management • For the exercise to be credible, the sources of addi- company to restructure corporate loans or provid- tional capital should be identified ahead of time, if ing incentives to banks to aggressively provision for shortfalls are found and private funds are insufficient. nonperforming loans through tax or capital rules. These funds need to be sufficiently large to accom- Provisioned loans could then be written down or modate the limited ability of some sovereigns to take sold at a discount to specialist third parties. on more debt. Adequate backstops are also important • Actively facilitating nonbank sources of corporate to avoid putting pressure on banks to scale back their credit. Steps could be taken to emulate France in balance sheets ahead of the assessment. developing a domestic corporate medium-term note • Determination to resolve nonviable institutions market that has maintained positive net supply to will be critical to restoring the financial system to domestic companies in recent years (e.g., through long-term health and to improving credit supply, the maintenance of a domestic SME credit register especially to SMEs. by the central bank). Similarly, life insurers and pen- The banking union must be completed: sion funds could be encouraged to hold longer-term • Completion entails expediting reforms already under corporate loans or bonds if the authorities were to way, such as implementing the legislation for the Single give them capital or regulatory relief for mitigating Supervisory Mechanism and reaching final agreement reinvestment risk.51 on the Bank Resolution and Recovery Directive. Further monetary support by the ECB is crucial to • The process should also involve the introduction of provide time for the repair of private balance sheets. a strong Single Resolution Mechanism that ensures Additional unconventional measures—including ensur- the swift restructuring or winding-up of banks while ing term funding for weak but solvent banks, or target- limiting the overall cost to taxpayers and establishing ing credit-easing measures to SMEs—would be in line clear rules for investors. The euro area bank resolution with the recent strengthening of the ECB’s collateral process, as proposed in the draft Bank Resolution framework and would help reduce fragmentation and and Recovery Directive, will help weaken the bank- prevent a more severe contraction in credit, while sovereign link. However, in the current environment, further conventional easing through lower policy rates the limited scope and “negative leverage” entailed in would support demand across the euro area. At the European Stability Mechanism direct bank recapital- same time, recent initiatives by the European Invest- ization places the burden of raising capital firmly back ment Bank and the European Commission to increase on bank shareholders and creditors or on the sover- lending to SMEs could complement these efforts. eign (even if financed upfront by European Stability Mechanism loans), or on both, and thus may not provide sufficient backstop should substantial capital global banking challenges: profitability, asset shortfalls be found in economies with weak sovereign Quality, and leverage balance sheets. Global bank capitalization remains divergent because institutions are at different stages of balance sheet repair and operate in different economic and regula- What needs to be done to address the corporate sector tory environments. Asset quality and profitability debt overhang?

50See Chapter 2. Measures to deal with the overhang should include the 51Solvency II proposals currently provide limited capital benefits following: for holding longer-maturity assets against long-term liabilities.

International Monetary Fund | October 2013 41

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

pressures at some euro area banks have reduced their Based on these estimates, banks from advanced econo- ability to increase capital levels through retained mies tend to have slightly higher fully loaded Basel III earnings. Some institutions may, therefore, need to Tier 1 ratios (more than 10 percent, on average) than further cut back their balance sheets or raise capital do banks headquartered in emerging market economies to meet higher capital standards. The way in which (over 9 percent, on average). this adjustment will take place has implications for In addition to risk-weighted capital ratios, investors the financial system and the real economy and has to are increasingly using unweighted leverage ratios to be monitored. The key tasks are to improve credibility, assess bank capitalization. This is partly in anticipa- transparency, and the strength of balance sheets, while tion of new rules: the Basel Committee on Banking avoiding undue pressures on banks from uncoordi- Supervision has finalized its leverage ratio proposal, nated national regulatory initiatives and uncertainty. and the United States has proposed new leverage standards.55 But it also reflects lingering concerns Bank capitalization remains divergent. about the consistency of approaches used by banks in different jurisdictions for calculating risk-weights, an Bank capital ratios—for this section’s sample of institu- issue that is being examined by the Basel Committee tions from jurisdictions with systemically important and by the European Banking Authority.56 Because the financial sectors—remain diverse.52 Tier 1 capital ratios data on netting and off-balance-sheet positions, which reported at end-2012 ranged from 5 to 21 percent, are needed to calculate the Basel III leverage ratio, are with the asset-weighted average standing just under 13 not published by all banks, investors often use tangible percent (Figure 1.54). Although these ratios are above leverage ratios—such as the ratio of tangible equity to the current regulatory minimum, full implementation tangible assets—to gauge the relative strength of banks of the Basel III standards will raise both the quantity (Figure 1.55). and the quality of capital that banks have to hold to For some banks, these simple tangible leverage ratios meet these standards.53 and Tier 1 ratios appear to give conflicting signals As Basel III capital standards became effective in about the strength of bank balance sheets. This tension 2013, many banks began reporting their capital ratios is illustrated in Figure 1.56, which shows a number on a Basel III basis.54 Based on the latest available of banks in either the bottom-right or top-left quad- information and IMF staff estimates for sample banks, rants of the figure; these quadrants are where the two fully loaded Basel III Tier 1 capital ratios are more ratios give different signals about bank balance sheet than 2 percentage points lower than Tier 1 ratios strength. reported at end-2012, on average (see Figure 1.54). This apparent conflict reflects, in part, differences in business models and regulatory environments. The 52The analysis in this section is based on a sample of 113 large “universal banking” model, which tends to be used banks headquartered in jurisdictions with systemically impor- more in Europe, will naturally lead to a larger balance tant financial sectors (see IMF, 2010), plus two European banks headquartered in other countries that are considered systemically sheet when compared with a bank with the originate- important for the region. Large banks in the following economies are to-distribute model, more commonly used in North included: advanced Asia-Pacific (Australia, Hong Kong SAR, Japan, America. The conflicting signals also highlight the Korea, Singapore); emerging Asia (China and India); emerging Europe (Russia and Turkey); euro area (Austria, Belgium, France, importance of restoring investor confidence in the Germany, Ireland, Italy, Netherlands, Spain); Latin America (Brazil accuracy and consistency of bank risk weights. This and Mexico); North America (Canada and United States); and other also suggests that risk-weighted capital ratios should advanced Europe (Denmark, Norway, Sweden, Switzerland, United Kingdom). be supplemented by leverage ratios, as proposed in the 53See Box 1.3 for a comparison of regulatory requirements in Basel III framework. selected jurisdictions. 54As of June 2013, 38 percent of sample banks had published their fully loaded Basel III Tier 1 capital ratios and another 17 percent of sample banks had published their core Tier 1 ratios. The 55The Basel III leverage ratio began parallel run with the Basel II September 2013 Basel III Monitoring Report, which uses detailed leverage ratio in January 2013 (see Box 1.3). information that is not always publicly available, found that Basel III 56Details of the Basel Committee’s Regulatory Consistency Tier 1 ratios for a group of large internationally active banks were Assessment Program can be found at http://www.bis.org/publ/ around 3 percentage points lower than current Tier 1 ratios, based bcbs216.htm; the European Banking Authority’s work on this issue on December 2012 data. The report is available at http://www.bis. is available at http://www.eba.europa.eu/risk-analysis-and-data/ org/publ/bcbs262.htm. review-of-consistency-of-risk-weighted-assets.

42 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.54. Large Bank Tier 1 Ratios Figure 1.55. Large Bank Tangible Leverage Ratios, 2012:Q4 (Percent) (Percent)

Tier 1 ratios (as reported at end-2012) Basel III Tier 1 ratios (reported or estimated) 14 22 12 20

18 10 16

14 8

12 6 10

8 4 6

4 2

2 0 0 Emerging Latin Emerging Advanced North Other Euro Other North Euro Advanced Emerging Latin Emerging Europe America Asia Asia-Paciﬁc America advanced area advanced America area Asia-Paciﬁc Europe America Asia Europe Europe

Sources: Bloomberg, L.P.; company reports; Federal Deposit Insurance Corporation; and IMF Sources: Bloomberg, L.P.; company reports; and IMF staff estimates. staff estimates. Note: The fully loaded Basel III Tier 1 ratios are either values reported by banks (June 2013 or Note: The tangible leverage ratio is the ratio of adjusted tangible equity to adjusted tangible latest available) or are estimated by adjusting the Basel III core Tier 1 ratios (or the Tier 1 assets. The adjustment is made by subtracting goodwill, other intangibles, and deferred tax assets. For U.S. banks, these numbers also include adjustments for accounting differences in ratios reported according to current regulatory standards) by the average gap between the derivatives netting, in line with the methodology used in Hoenig (2013). However, some ratios reported by banks in the same region. The horizontal lines show the asset-weighted differences in accounting deﬁnitions may remain. The horizontal lines show the asset-weighted average ratios for the banks in each region. See footnote in text for makeup of regions. average ratio for the banks in each region. See footnote in text for makeup of regions.

Asset quality pressures at some banks are affecting their provisions. The latter reflects deteriorating asset quality profitability. from the weak cyclical positions of these economies, exacerbated by the corporate debt overhang in stressed Bank profitability is now generally lower than it was economies of the euro area. Some euro area banks— before the onset of the global financial crisis, but this including Dutch, Irish, as well as Spanish banks—face is likely the result of some unwinding of unsustainable challenges from their exposures to household debt. levels of pre-crisis profitability. In emerging market Recent IMF Financial Sector Assessment Program economies, large banks are able to generate higher (FSAP) assessments of a number of European econo- profits from their assets (return on assets of about 1.4 mies also found that continuing deterioration of credit percent) than are large banks in advanced economies quality weighs heavily on banks’ already-thin profit- (return on assets of about 0.4 percent), on average ability (see Box 1.4). (Figure 1.57). Revenues, especially net interest income, Concerns about bank asset quality are further com- are significantly higher for banks in emerging Europe pounded by uncertainty about the extent and nature and Latin America than for banks in advanced econo- of lender forbearance. Although the ECB’s upcoming mies, although loan loss provisions and expenses tend euro area asset quality review should help resolve some to be larger as well. of these concerns, some supervisors are acting preemp- Among advanced economy banks, European institu- tively. The Italian central bank recently carried out a tions—and euro area banks, in particular—currently review of asset quality; the Bank of Spain is conducting have the weakest profitability. Euro area banks have an assessment of restructured loan classification; the faced the combined pressures of increased funding Dutch central bank is reviewing commercial real estate costs, falling operating incomes, and rising loan loss lending; and U.K. authorities completed their asset

International Monetary Fund | October 2013 43

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.56. Large Bank Tier 1 Capital and Tangible quality review in June 2013 by publishing bank-by- Leverage Ratios bank capital shortfalls.57 (Percent) The link between weak profitability and asset quality Euro area Other advanced Europe is reflected in market valuations of institutions. Figure Advanced Asia-Paciﬁc North America Emerging Europe Emerging Asia 1.58 shows that market capitalization as a percentage Latin America of assets—a market indicator of the effect of asset qual- 16 Higher leverage ratio, Higher leverage ity on bank capital—tends to be lower for banks with lower Tier 1 ratio and Tier 1 ratios 14 weak profitability.

12 Asset quality and earnings pressures will affect some banks’ ability to increase their capitalization. 10 Weak profitability makes it more difficult for banks to 8 raise their capitalization organically through retained 6 earnings. This effect can be illustrated through a Tangible leverage ratio Tangible forward-looking exercise that projects bank capital- 4 ization in 2018 using analysts’ forecasts of bank net income, assuming that balance sheets are unchanged. 2 Lower leverage Lower leverage ratio, The objective of this exercise is to see how many and Tier 1 ratios higher Tier 1 ratio 0 institutions will likely not be able to reach these targets 0 5 10 15 20 25 through retained earnings alone and therefore would Estimated Basel III Tier 1 ratio

Sources: Bloomberg, L.P.; and IMF staff calculations. Note: See Figure 1.54 for details on the estimated Basel III Tier 1 ratio and Figure 1.55 for details on the tangible leverage ratio. The dotted lines show asset-weighted average ratios. 57For more information on the U.K. exercise, please see www. See footnote in text for makeup of regions. bankofengland.co.uk/publications/Pages/news/2013/081.aspx.

Figure 1.57. Bank Proﬁtability Comparison (Percent of total assets)

Trading income Net interest income Other net income Fees and commissions Loan loss provisions Total net income 10

–2

–4

–6

–8 2006 08 10 12 2006 08 10 12 2006 08 10 12 2006 08 10 12 2006 08 10 12 2006 08 10 12 2006 08 10 12 Emerging Latin Emerging Advanced North Other advanced Euro Europe America Asia Asia-Paciﬁc America Europe area

Sources: Bloomberg, L.P.; Federal Deposit Insurance Corporation; and IMF staff calculations. Note: Shows the unweighted average for large banks in each region. Total assets of U.S. banks have been adjusted for netting of derivatives. See footnote in text for the countries in each region.

44 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

have to make further adjustments, that is, shrink their Figure 1.58. Bank Proﬁtability and Market Valuation of Assets balance sheets, reduce risk-weighted assets, or raise Euro area Other advanced Europe capital. Projected bank capital levels are tested against North America Advanced Asia-Paciﬁc two targets: an 11 percent target for the Basel III Tier Emerging Europe Latin America Emerging Asia 1 capital ratios and a 4 percent target for tangible 2 leverage ratios. Although these two targets are not minimum regulatory requirements, they represent ratios that institutions may seek to reach given regula- 1 tory and market expectations.58 This projection exercise reveals that most banks in 0 the sample already have, or should have, an estimated Basel III capital ratio of 11 percent (a tangible leverage ratio of 4 percent) by 2018 (Figure 1.59). However, –1 around 4 percent of banks may not be able to meet these targets organically through retained earnings. Net income (percent of assets, 2012) Net income (percent of assets, Most of these institutions are in the euro area. –2

European banks have been deleveraging in response to –3 market and regulatory concerns about capital levels, 0 5 10 15 20 25 and may continue to do so. Market capitalization (percent of tangible assets, August 2013)

Sources: Bloomberg, L.P.; and IMF staff estimates. Banks that are unable to meet capital ratio targets Note: Assets of U.S. banks have been adjusted for netting of derivatives. See footnote in text organically through retained earnings will need to for the countries in each region. either raise fresh equity in markets or cut back balance sheets. Indeed, a combination of market and regula- 60 tory concerns about bank capitalization has already and by $2.1 trillion on a net basis (Table 1.6). These led to an increase in capital levels at EU banks.59 At cutbacks in assets are currently running at a similar the same time, large EU banks have continued to pace to the baseline scenario in the October 2012 shrink their balance sheets, in aggregate. Over the GFSR. About 40 percent of the reduction by the banks period 2011:Q3–2013:Q2, large EU banks reduced in the EU as a whole was through a cutback in loans, their assets by a total of $2.5 trillion on a gross basis— with the remainder through scaling back noncore which includes only those banks that cut back assets— exposures and sales of some parts of their businesses. Banks have been reducing their risk-weighted assets at a faster speed and have already cut back risk-weighted assets more than was envisaged in the 58Because the Basel III standards have not been universally adopted, October 2012 GFSR baseline scenario (see Table 1.6). identifying a common benchmark that banks across more than 20 As discussed in the April 2013 GFSR, banks have been jurisdictions may strive to achieve is not straightforward. Some regula- concentrating on derisking their balance sheets by tors may actually set more ambitious and/or different targets for their banks than the Basel III minimum requirements described in Box reducing capital-intensive businesses, holding greater 1.3. For example, (1) the U.K. Prudential Regulatory Authority has proportions of assets with lower risk weights (such asked banks to meet a Basel III 7 percent common equity Tier 1 ratio as government bonds), and optimizing risk-weight by end-2013, ahead of the Basel III timetable, after implementing additional deductions from capital for potential losses and expected models. The capital ratio projection exercise previ- conduct-related costs, as well as using higher risk weights for certain ously discussed suggests that some banks will need to exposures; (2) the United Kingdom’s 3 percent leverage ratio has continue raising equity or cutting back balance sheets similarly been set in more tightly defined terms than in Basel III; and as they endeavor to repair and strengthen their balance (3) the United States has proposed its own leverage ratio minimum of 4 percent. Furthermore, some banks may seek to have capital ratios sheets. that are above regulatory minimums and so other institutions could be under pressure to catch up with their peers. 59EBA (2012) provides the results of their capital exercise, which resulted in an increase in capital levels at the banks included in the 60Adjustment is also occurring on the liabilities side of the balance exercise. sheet, although generally more slowly (see Chapter 3).

International Monetary Fund | October 2013 45

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.59. Large Bank Capitalization deleveraging will need to be monitored to ensure that it occurs in an orderly manner and does not create 1. Fully Loaded Basel III Tier 1 Ratios in 2. Tangible Leverage Ratios in Relation Relation to an 11 Percent Benchmark to a 4 Percent Benchmark adverse spillovers to the financial system and the real May not meet by 2018 May not meet by 2018 economy. In particular, it is important for the upcom- without further adjustments without further adjustments ing balance sheet review in the euro area to encourage Should 4% Already 4% meet banks to adjust in a “healthy” manner (for example met by 2018 27% through disposal of nonperforming assets and by raising capital) to avoid putting undue pressure on the real 44% 52% economy. 70% Credibility and transparency of balance sheets need Should meet Already to be shored up. Finalizing work on risk weights, har- by 2018 met monizing definitions of key financial indicators (such as nonperforming loans) used in different jurisdictions, Sources: Bloomberg, L.P.; company reports; and IMF staff estimates. completing accounting convergence, and introduc- Note: For the categories “Should meet by 2018” and “May not meet by 2018 without further adjustments,” the test is to allow banks to reach the target by retaining all of their net income ing forward-looking provisioning will all help in that (but without reducing their risk-weighted assets or raising new equity), where future average regard. Restoring investor faith in risk weights will also annual income is based on consensus analysts’ forecasts. See Figures 1.54 and 1.55 for details of the estimated Basel III and tangible leverage ratios. Totals may not equal 100 help ensure that risk-weighted capital ratios remain the percent due to rounding. main capital benchmark, with leverage ratios having a supplementary backstop function, as envisaged in the Basel III framework. The transition to a stronger banking system requires Finally, regulatory uncertainty and unintended further policy effort. consequences from multiple uncoordinated national regulatory initiatives should be minimized. National Banks in advanced economies continue to face profit- structural measures for banks (such as the Volcker, ability and asset quality pressures against a weak eco- Vickers, and Liikanen proposals, as well as others) are nomic backdrop. These pressures keep banks focused another potential challenge, if implemented differently on rationalizing their business models and balance across jurisdictions, and could have unintended conse- sheets. However, bank balance sheet repair has yet to quences on markets.61 be completed. Although European banks have made significant progress on derisking and deleveraging their balance sheets, more needs to be done to improve 61As discussed in the April 2013 GFSR and Viñals and others earnings prospects and investor perceptions. Further (2013).

Table 1.6. European Union Bank Deleveraging Change in Balance Sheet Actual Change October 2012 GFSR Scenarios 2011:Q3–2013:Q2 2011:Q3–2013:Q4 Progress against GFSR Baseline (trillions of U.S. dollars) (trillions of U.S. dollars) (percent) Assuming Smooth Gross Net Complete Baseline Weak Gross/Baseline Adjustment Tangible Assets (minus derivatives and cash) –2.5 –2.1 –2.3 –2.8 –4.5 88 78 Risk-Weighted Assets –1.3 –1.2 –0.8 –1.0 –1.9 126 78 Sources: SNL Financial; and IMF staff estimates. Note: For a sample of 58 large European Union banks (see the April 2012 GFSR for a description of the sample). Gross shows the results for banks in the sample that cut back balance sheets. Net shows the change for all banks in the sample. The figures are rounded to the nearest 0.1 trillion.

46 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 Making the Transition to Stability

Box 1.3. Financial Regulatory Reform Update

Although progress in global regulatory reform has implementing effective domestic and cross-border been achieved over the past six months, there are a resolution regimes; facilitating implementation of over- number of areas where further coordinated efforts the-counter (OTC) derivatives reforms through further are needed. While many of the reform initiatives are cross-border coordination; and enhancing monitoring under way, gaps remain. Focus on timely and consis- of shadow banking. tent implementation of agreed measures will remain a Progress on Basel III continues with 25 of the 27 high priority. Priorities include strengthening pruden- Basel Committee on Banking Supervision member tial supervision through such measures as securing jurisdictions having issued the final set of Basel III resources and independence of supervisors, restoring capital regulations.1 Two main jurisdictions—the confidence in bank balance sheets, developing and United States and the EU—published their final Basel III regulations in the first week of July 2013 The authors of this box are Ana Carvajal, Marc Dobler, Ellen Gaston, Eija Holttinen, Fabiana Melo, Mala Nag, Oana Nede- 1For details, see the August 2013 BCBS progress report on lescu, Nobuyasu Sugimoto, and Mamoru Yanase. Basel III implementation (www.bis.org/publ/bcbs260.pdf).

Table 1.3.1. Comparison of Bank Regulations across Jurisdictions Regulation Basel Minimum Standard United States European Union Capital Quality of Capital Common equity to compose CET1, Common equity to compose CET1, Common equity to compose CET1, conservation and countercyclical conservation and countercyclical conservation and countercyclical buffers, and G-SIB surcharge buffers; no G-SIB surcharge buffers, and G-SIB surcharge Full compliance by 2018 (separately treated) Full compliance by 2018 Full compliance by 2018 Quantity of Capital CET1 4.5% CET1 4.5% CET1 4.5% Conservation buffer 2.5% Conservation buffer 2.5% Conservation buffer 2.5% Countercyclical buffer 2.5% Countercyclical buffer 2.5% Countercyclical buffer 2.5% G-SIB Buffer Surcharge 1.0–3.5% Not part of U.S. Basel III Surcharge 1.0–3.5% Leverage Ratio BCBS has set minimum requirement U.S. has revised its existing leverage EU is expected to adopt leverage at 3% for leverage ratio to ratio to require 4% (minimum) ratio within Basel III proposed complement risk-based capital for all banking organizations. framework. CRR/CRDIV includes ratio. Supplementary ratio (BCBS the calculation and reporting format) was adopted at 3% of a leverage ratio but does (minimum) for internationally not yet establish it as a pillar 1 active banking organizations. requirement. Enhanced supplementary ratio has been proposed for bank holding companies (with over $700 bn in assets or $10 trn in assets under custody) at 5%. Further, insured depository subsidiaries of these firms will have to meet 6% leverage ratio to be well capitalized under the prompt corrective action regime. Liquidity Liquidity Supervision U.S. Dodd-Frank Act, Section 165, The EU plans to adopt LCR and Net requires banks with assets of Stable Funding Ratio. more than $50 billion to hold LCR implementation phased in liquidity buffers of highly liquid beginning in January 2015 at 60%, assets; this is broadly consistent with full compliance by 2019. with the objective of Basel III EU member states are to carry out liquidity ratios. supervision and monitor reporting of LCR compliance progress. Liquidity Coverage Ratio BCBS has identified the list of eligible No proposals. The EU has outlined outflows and Level 1 and Level 2 assets to inflows in Capital Requirements constitute High Quality Liquid Regulation. Further refinements Assets. BCBS has proposed to come from EBA on regulatory phase-in period starting in January standards and to be adopted by 2015 and lasting through 2019. the European Commission. Net Stable Funding Ratio BCBS intends to review NSFR. The No proposals but expected at later EU plans to adopt NSFR once the objective is to ensure that banks date. BCBS has finalized it. maintain stable asset-liability profiles over a one-year horizon. Source: IMF staff. Note: BCBS = Basel Committee on Banking Supervision; CET1 = common equity Tier 1; EU = European Union; G-SIB = global systemically important bank; LCR = Liquidity Coverage Ratio; NSFR = Net Stable Funding Ratio. U.S. leverage ratio is defined as Tier 1 capital over on-balance-sheet assets, whereas the U.S. supplementary leverage ratio is defined as Tier 1 capital over total leverage exposure, which includes both on-balance-sheet and certain off-balance-sheet exposures.

International Monetary Fund | October 2013 47

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

box 1.3. (continued)

(Table 1.3.1). The BCBS is assessing the quality of traditional prudential regulatory and bank resolution implementation of its members through “Level 2” tools to enhance financial stability. Nevertheless, given assessments of its Regulatory Consistency Assessment their potentially significant costs, which can permeate the Program (RCAP).2 global economy, the implications of these measures for The BCBS is assessing the consistency of regulatory other jurisdictions should be weighed in. outcomes of its capital standards (“Level 3”). Pre- Efforts are pending to develop effective domestic and liminary findings focusing on the application of risk cross-border resolution regimes, and implementation weights by advanced approaches in the banking and remains challenging. Many countries are in the process trading books indicate discrepancies due to national of upgrading their legislation to reflect the Key Attributes supervisory action and variations in accepted model- of Effective Resolution Regimes for Financial Institutions ing practices. The findings from this analysis will feed (Key Attributes).4 An assessment methodology to evaluate into further policy recommendations and guidance to country compliance has been published and pilot assess- harmonize risk-weighting approaches. A fundamen- ments are being planned. Implementation of the Key tal review is under way regarding the standardized Attributes will require capacity-building and resources, approaches to regulatory capital for market, credit, and as well as strengthened and more systematic coopera- operational risks. tion among relevant authorities both within and across The first of two liquidity standards—the Liquid- borders. The Financial Stability Board (FSB) is leading ity Coverage Ratio—was agreed on in January 2013. efforts to offer more specific guidance on operational- With implementation scheduled to start in January izing recovery and resolution plans and on the resolution 2015, the final standards include a broadened defini- of financial market infrastructure and insurers. tion of High Quality Liquid Assets and a phase-in The International Association of Insurance Supervi- period. Discussions are ongoing regarding design and sors (IAIS) has agreed on a methodology for identifying calibration of the second liquidity standard—the Net globally systemically important insurers (G-SIIs) and Stable Funding ratio. on policy measures for G-SIIs focused on shielding In June 2013, the BCBS issued a consultative docu- traditional insurance activities from designated non- ment on the revised Basel III leverage ratio framework traditional and non-insurance (NTNI) activities. Based and disclosure requirements.3 The numerator of the on the assessment methodology, the FSB and national leverage ratio is Tier 1 capital of the risk-based capital authorities, in consultation with the IAIS, identified an framework and the denominator is the sum of bal- initial list of G-SIIs in July 2013. The policy measures ance sheet exposures, derivatives exposures, securities that will apply to G-SIIs include the development and financing transaction exposures, and other off-balance- implementation of systemic risk management plans, sheet exposures. The minimum requirement in the recovery and resolution planning requirements under transition period is 3 percent. Adjustments to the the Key Attributes, enhanced group-wide supervision, definition and calibration of the leverage ratio will be and higher loss absorbency capital requirements. The made by 2017 based on the results of the parallel run IAIS is also developing a straightforward group-wide consultations, with a view to migrating to a Pillar 1 capital requirement that will serve as a foundation for treatment on January 1, 2018. higher loss absorbency requirements. “Structural measures” that would impose business The International Accounting Standards Board model restrictions on banks are still under discussion. (IASB) and the U.S. Financial Accounting Standards The so-called Volcker Rule has not yet been implemented Board (FASB) are continuing to work on the conver- in the United States, but the recommendations from the gence of financial reporting standards, but progress Vickers report in the United Kingdom have become part has been slow. All four convergence projects (Finan- of U.K. banking law, and a draft German banking law cial Instruments, Revenue Recognition, Insurance setting some restrictions is also in progress. The French Contracts, and Leases), which started after the global legislature passed its version of structural regulation in the financial crisis, are at various stages of discussion. Con- summer of 2013. Appropriately designed and judiciously vergence between the two proposed models for asset implemented, these policies can work in tandem with impairment loss recognition remains challenging. International standard setting on OTC derivatives 2Details of the Regulatory Consistency Assessment Program reforms is almost complete but implementation chal- can be found at www.bis.org/publ/bcbs216.htm. 3See www.bis.org/press/p130626.htm. 4See www.financialstabilityboard.org/publications/r_111104cc.htm.

48 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

box 1.3. (concluded) lenges remain. Important policy developments include by the major OTC derivatives regulators could pave the the September 2013 publication of the BCBS-IOSCO way for much needed progress. final report on margin requirements for non-centrally Data constraints remain a key challenge for proper cleared derivatives. In June 2013, the BCBS proposed monitoring and supervision of shadow banking at the final capital requirements for banks’ exposures to global level. The FSB will address data constraints by central counterparties (CCPs) to replace the current developing standards for data collection on securities interim rules. New policy work streams have also been financing markets and information-sharing processes launched, focusing on recovery and resolution of Finan- for shadow banking entities in 2014. In the policy cial Market Infrastructures and conducting a feasibility arena, some progress has been made with the adop- study on aggregating OTC derivatives data reported to tion by the IOSCO of principles for money market trade repositories. While most of the larger jurisdictions funds and with the proposals by BCBS to limit large are finalizing their OTC derivatives frameworks, key exposures to shadow banking entities and to introduce implementation issues remain outstanding, in particular risk-sensitive capital requirements on equity invest- in relation to the treatment of cross-border activities. ment. In addition, the FSB has published documents The recent set of understandings between the United setting out (1) an overall approach to address financial States and the EU on the establishment of a mutual stability concerns associated with shadow banking reliance framework to regulate the cross-border activities entities and (2) a policy framework for addressing of swap dealers and the broader understandings reached shadow banking risks in securities lending and repos.

box 1.4. recent Financial Sector assessment program Mission Findings

Recent IMF Financial Sector Assessment Program address concerns about risks in the southeastern and (FSAP) missions to a number of European countries central European region and to repay government determined that financial sectors have largely sta- capital. Some banks should also further strengthen bilized since the peak of the global financial crisis, their foreign currency funding structures. but challenges remain as continuing deterioration of The FSAP found that, despite effective bank supervi- credit quality weighs heavily on banks’ already-thin sion practices, some governance improvements should profitability.1 Substantial amounts of euro area public be pursued in both the financial market authority and debt on banks’ and insurers’ balance sheets still bear the industry, and certain supervisory powers could be considerable risks. Central recommendations com- enhanced. A special bank resolution regime is needed mon to all of these FSAPs include the strengthening of in Austria to provide a wide range of tools and powers capital buffers, further cleanup of balance sheets, and to resolve failing banks in an orderly and least-cost derisking of activities. manner. The existing fragmented system of deposit During the recent crisis, the Austrian financial system guarantee schemes should be replaced with a unified, benefited from limited exposures to sovereign and prefunded, and publicly administered scheme. other market risks and relatively favorable domestic The Belgian financial sector has become smaller, less macroeconomic conditions. Stress test results indicate complex, and less leveraged. Its ongoing transforma- that, under adverse medium-term scenarios, virtually tion, however, involves significant downside risks from all Austrian banks, including all internationally active low profitability and weak macroeconomic prospects. institutions, would meet regulatory capital require- Structurally high costs for banks are compounded by ments (taking into account Basel III implementation). increased competition, diminished earning capac- However, stronger capital buffers appear desirable to ity, and the impact of regulatory reforms. The links between banks, insurers, and the Belgian sovereign The authors of this box are Javier Hamann and Emanuel have intensified against the backdrop of large public Kopp. debt. The government’s limited fiscal capacity makes 1 FSAPs assess the stability of the financial system as a whole it important to guard against inaction and supervisory and not that of individual institutions. They are intended to help countries identify key sources of systemic risk in the financial forbearance. sector and implement policies to enhance its resilience to shocks A prolonged period of low interest rates would cre- and contagion. ate vulnerabilities for banks and life insurers, while a

International Monetary Fund | October 2013 49

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

box 1.4 (concluded)

downturn in housing prices would further exacerbate dispose of impaired assets, and strengthening capital bank capital pressures. Stress tests revealed that initial and funding plans, where needed, can make important capital levels are solid in aggregate, but several banks contributions. Some of these steps have already been would experience significant deterioration of profit- initiated by the Bank of Italy. ability under stress, inducing solvency pressures. The The financial sector in Poland emerged unscathed FSAP mission recommended strengthening banks’ from the crisis. Banks have been profitable and hold capital buffers. Insurers meet the requirements of the relatively high levels of core Tier 1 capital. Vulnerabili- current solvency regime, but vulnerabilities are appar- ties lie in euro area interconnectedness and exposure ent, which means that supervisors must remain vigi- to foreign exchange risk. Stress tests suggest, how- lant and contingency plans need to be put in place ever, that these vulnerabilities are unlikely to become under the new recovery and resolution framework. systemic. Although the new regulatory structure is function- The FSAP stressed that persistent nonperforming ing well, more intensive and intrusive supervision is loans and the cyclical deterioration in credit quality needed. Compliance with international standards for need to be addressed. Furthermore, tax disincentives, regulation and supervision of banks and insurers is income accrual practices, and obstacles to out-of-court generally high, but national resolution and deposit restructurings need to be removed, and improve- insurance frameworks need to be strengthened, and ments in restructuring, accounting practices, and the positive changes to supervisory practices need to be insolvency framework would be helpful. To prevent sustained. a further rise in nonperforming loans, care should be The Italian financial sector has shown resilience in taken with ongoing regulatory revisions, including the the face of a severe and prolonged recession. Con- removal of uniform debt-to-income thresholds, tight- tinuing weaknesses in the real economy and the link ening of loan-to-value ratios, and currency matching between the financial sector and the sovereign remain of income and borrowing. key risks. If these risks materialize, the impact on Poland was found to be broadly compliant with core banks could be significant, albeit cushioned by their principles in the regulation and supervision of banks, own capital buffers and the availability of European insurance companies, and deposit insurance schemes. Central Bank liquidity. However, the supervisor needs greater powers, inde- The FSAP concluded that targeted financial sector pendence, and resources, and legislation to introduce action should be taken to shore up the defenses of a systemic risk board needs to be accelerated. Rebal- Italian banks. Increasing provisions, improving bank ancing the financial system toward capital market efficiency and profitability, developing a market to development is also important.

Making the transition to Stability moving toward a stronger monetary union with a common framework for risk mitigation while strength- The global financial system is undergoing a series of ening financial systems and reducing excessive debt transitions along the path toward greater financial levels. Finally, the global banking system is phasing stability. The United States may soon move to less in stronger regulatory standards. A number of policy accommodative monetary policies and higher sus- actions can help promote an orderly passage to greater tained long-term interest rates as its recovery gains financial stability, as summarized in Table 1.7. ground. After a prolonged period of strong portfolio inflows, emerging markets are facing a transi- The shift from prolonged periods of monetary accom- tion to more volatile external conditions and higher modation poses challenges. risk premiums. Some need to address financial and macroeconomic vulnerabilities and bolster resilience Experience suggests that transitions from monetary as they progress to a regime in which financial sec- accommodation can give rise to financial stability risks. tor growth is more balanced and sustainable. Japan As Figure 1.60 illustrates, during the period of Great is moving toward the new Abenomics policy regime Moderation, benign monetary and financial conditions marked by more vigorous monetary easing coupled drove investors to adopt similar investment strategies, with fiscal and structural reforms. The euro area is leading to a rise in correlation of asset prices and a

50 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Table 1.7. Policy Recommendations Reducing the market impact of • Carefully communicate the Federal Reserve’s quantitative easing asset-purchasing intentions to minimize interest rate volatility. monetary policy transition • Increase oversight of mutual fund, mortgage real estate investment trust, and exchange-traded fund liquidity terms for investors and management practices. • Develop a contingency leverage unwinding facility in the United States to act as a circuit-breaker in markets that heavily use repo funding. • In Japan, deliver on structural reforms and medium-term fiscal consolidation in addition to monetary stimulus to contain fiscal risk premiums in government bond yields. • Monitor Japanese regional bank exposures to interest rate risks. • Pursue reforms in Japanese bond and derivatives markets to manage rate volatility. Tackling emerging market • Address underlying macroeconomic vulnerabilities through credible fiscal or regulatory reforms. vulnerabilities • Prepare for and manage the reversal of capital inflows by ensuring orderly market operations and establishing swap lines with major central banks. • Restore policy buffers where needed, including through tighter monetary policy if inflation or currency vulnerabilities warrant. • Focus surveillance on domestic bank exposure to vulnerable corporates, especially liquidity and currency mismatches. • In China, rein in total credit growth, notably via the shadow banking system, by gradually liberalizing deposit rates and addressing moral hazard concerns. • Enhance supervision and disclosure in the Chinese nonbank financial system, including insurers and trust funds. Addressing legacy balance sheet • Restore investor confidence in euro area bank balance sheets with a credible balance sheet assessment and stress test, with clearly issues identified capital backstops. • Address euro area financial fragmentation through speedy implementation of the Single Supervisory Mechanism and the Single Resolution Mechanism with a commitment to cross-border deposit insurance. • Resolve the corporate debt overhang in stressed economies through a more systematic approach, including improved insolvency and debt workout arrangements, while fostering nonbank sources of corporate credit. • Provide time to repair private balance sheets through further European Central Bank monetary support and European Investment Bank credit support to viable firms. Improving regulation and market • Continue progress on strengthening regulatory frameworks and monitor progress toward achieving goals of higher capital standards. liquidity • Minimize regulatory uncertainty and unintended consequences on markets from national structural measures for banks (e.g., Volcker, Vickers, and Liikanen proposals). • Finish work on risk weights, complete accounting convergence, and introduce forward-looking provisioning to improve the credibility and transparency of bank balance sheets. • Assess the impact of regulatory and transaction tax proposals on market liquidity and rebalance where necessary, while clarifying issues that have increased uncertainty surrounding market liquidity and funding providers. • Increase focus on the implications of lower market liquidity and higher volatility through enhanced stress testing of bank’s mark-to- market books and repo-funded nonbank intermediaries.

decline in volatility. Arguably, those strategies resulted Figure 1.60. Normalization of Monetary Policy: Smooth or in excesses that led to the global financial crisis. In its Turbulent? wake, crisis measures and monetary accommodation (Standard deviations from mean) 1.5 have suppressed volatility, while the sensitivity of asset Low volatility, High volatility, Low rates, prices to central bank monetary policy remains high. high common factor unconventional monetary policyhigh common factor

Policymakers and markets need to prepare for struc- 1.0 turally higher market volatility because the probable July 2010–April 2013 withdrawal of the Federal Reserve’s quantitative easing Aug. 2007–June 2009 0.5 stimulus and tighter regulatory constraints on financial June 2006–July 2007 Global Financial Crisis intermediaries mean that market liquidity is likely to be further curtailed. Indeed, the rise in global rates and 0.0 Great Moderation volatility since May 2013—prompted first by uncer- Jan. 1999–June 2002

tainty over Bank of Japan policy implementation and –0.5 then by concerns about the Federal Reserve tapering U.S. recession and recovery Common component in returns its quantitative easing—precipitated a volatility spike June 2004–May 2006 –1.0 in global bond markets, prompting turbulence in a Low volatility, July 2002–May 2004 High volatility, number of important emerging markets. low common factor low common factor Achieving a smooth transition requires policies that –1.5 –1.5 –1.0 –0.5 0.0 0.5 1.0 1.5 carefully manage portfolio adjustments while addressing Volatility structural liquidity weaknesses and systemic vulnerabili- Source: IMF staff estimates. ties. Policymakers can take a number of actions to reduce Note: Common factor refers to the principal component that explains the largest possible the impact of elevated market volatility. These include variation in asset returns, based on the historical daily data (1994–2013) for advanced and clarity of communication about the parameters for the emerging market economies. withdrawal of monetary stimulus, and regulatory scrutiny

International Monetary Fund | October 2013 51

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

of the liquidity offered to investors in funds exposed to ers of recent capital flows to emerging markets are illiquid assets, especially when repo-funded, to mitigate weakening as relative growth prospects moderate, spikes in asset correlations and volatility. Indeed, authori- U.S. nominal rates rise, and volatility picks up. These ties may need to develop contingency backstops to reduce inflows have been intermediated primarily through the likelihood of cascading forced asset sales. sovereign and corporate bond markets, rather than through domestic banks engaged in cross-currency The transition to higher rates and volatility puts a credit intermediation. Therefore, the principal trans- premium on addressing legacy balance sheet problems. mission channel of volatility is likely to be through liquidity strains on sovereigns and leveraged cor- The rise in nominal global rates and volatility will make the porates with immediate borrowing and refinancing refinancing of stretched corporate and bank balance sheets needs, rather than through bank funding channels. more costly and difficult. The analysis of the euro area Consequently, emerging market investors are likely corporate debt overhang in this GFSR shows that unless to focus more on country-specific factors and insti- steps are taken to break the feedback loop between weak tutional robustness in evaluating risk-return trade- banks and corporates, a long period of weak asset quality offs, with the increasing likelihood that the portfolio and a drag on economic activity are probable risks. Hence, capital inflows of recent years will be partly reversed, further progress in reducing debt overhangs and strength- at least in the near term. ening bank balance sheets remains urgent, especially in the In the event of significant capital outflows, some stressed economies of the euro area. To succeed, investors’ countries may need to focus on ensuring orderly faith in euro area bank balance sheets must be restored market functioning, using their policy buffers wisely. (through the planned asset quality review and resulting Keeping emerging market economies resilient calls for recapitalization, if necessary) and banking union completed an increased focus on domestic vulnerabilities. Policy- to fully reverse financial fragmentation. Otherwise, the euro makers should carefully monitor and contain the rapid area risks entering a lengthy, chronic phase of low growth growth of corporate leverage. Local bank regulators and balance sheet strains. need to guard against foreign currency funding mismatches building up directly on bank balance sheets, Keeping emerging markets resilient calls for an or indirectly through unhedged foreign currency bor- increased focus on addressing domestic vulnerabilities. rowing by corporates.

Emerging markets are now encountering a less benign external environment. The fundamental driv-

52 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 Making the Transition to Stability

Annex 1.1. Exploring the Factors Driving Bank the sovereign bond yield spreads, but is not biased by Interest Rates on Corporate Loans episodes of flight to quality that tend to drive down Objectives and Analytical Approach German yields and exert an upward bias on sovereign spreads measured against German bunds. This exercise aims to explain the dynamics of bank inter- •• Bank health, as proxied by the banking system price- est rates on corporate loans in the euro area economies to-book ratio (pbkt). A healthier banking system in relation to their fundamental determinants. The start- will have a higher average price-to-book ratio, which ing point, building on previous research, is the notion captures the perceived health and expected future that the interest rate on corporate loans is a function profitability of banks, enabling them to borrow and of the monetary policy stance, which influences banks’ lend more cheaply. The higher price-to-book ratio funding costs via money market rates; the business cycle, outperformed alternative measures of bank balance which affects the demand for loans and asset qual- sheet strength (such as bank equity prices and credit ity; and stress in the banking sector, which determines default swap spreads) in diagnostic tests. banks’ ability to finance themselves, borrow, and extend •• The state of the business cycle, as captured by the credit (see also ECB, 2009; IMF, 2013a). The analysis industrial production index (ipt). When the level of also includes sovereign stress, given the importance of output declines, economic uncertainty rises, profits feedback effects between sovereign and bank stresses, come under pressure, and demand for corporate and a measure of corporate credit risk. loans typically falls. Consistent with other studies The main building block of this analysis is the (ECB, 2009) and bank lending surveys, one would (cointegrating) equilibrium that links the long-term expect weaker loan demand from firms and house- 62 dynamics of the following five variables: holds to put downward pressure on bank lending nfc •• The lending rate on new corporate loans (rt ) of less rates, especially during the deep and extended reces- than €1 million in France, Italy, and Spain; many sions seen in Italy and Spain. of these loans are extended to small and medium Furthermore, a number of exogenous variables are enterprises (SMEs). included, notably a corporate credit risk variable based •• The monetary policy stance, as captured by the on Moody’s KMV expected rates of default. This 7swp seven-year swap rate (rt ). The seven-year swap variable was added exogenously because its time-series rate was found to significantly outperform shorter- properties did not make it amenable to inclusion in dated maturities and other money market rates. For the cointegrating vector. Other variables included example, a recent study by the ECB concluded that exogenously are contemporaneous changes in euro “through its influence on expectations on future overnight index average (EONIA) rates, and changes in monetary policy actions, changes in monetary economic policy uncertainty (see Bloom, 2009). policy stance will often also have a strong impact on By exploiting the vector error correction model’s longer-term market rates, such as long-term govern- (VECM) long-term cointegrating relationship, the ment bond yields and swap rates, by moving the analysis determines the “equilibrium” levels of lending yield curve” (ECB, 2009, p. 97). rates under the current state of financial fragmentation. •• Sovereign stress, as proxied by the deviations of asset Subsequently, by setting banking and sovereign stresses swap spreads (10-year sovereign bond yields minus to zero in the cointegrating vector, a hypothetical swap rate of the same maturity) from their time-vary- shadow rate is constructed that captures the notion of aswp 63 ing trend (devt ). This spread was used as a proxy no fragmentation. The construction of this latter no- for sovereign credit risk because it behaves similarly to fragmentation proxy is what differentiates this analysis from previous studies. The cointegrating relationship The author of this annex is Vladimir Pillonca. can be expressed in terms of the key variable of inter- 62 Cointegration tests were performed using the Johansen method- est, the corporate lending rate r nfc:64 ology (see Johansen, 2009). t 63 The fixed-rate arm of an interest rate swap captures a highly r nfc = ψ + b r 7swp + b dev aswp + b pbk + b ip + ξ . liquid risk-free rate needed to compute bond spreads, as an alterna- t 1 1 t 2 t 3 t 4 t t tive to German bund yields. The time-varying trend was estimated (1.1.1) with a Christiano-Fitzgerald asymmetric bandpass (Christiano and Fitzgerald, 1999), which allowed the extraction of a signal of sovereign stress that was not overly collinear with the other variables 64As is standard practice, the coefficient on the variable of interest in the system. is normalized to unity.

International Monetary Fund | October 2013 53

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.61. France: Deviations from Cointegrating Equilibrium

The French corporate lending rate has not deviated sharply from its cointegrating The deviations from equilibrium tend to be corrected over time. equilibrium.

1. French Corporate Lending Rate 2. Deviations from Cointegrating Equilibrium 7 200 (Percent) (Basis points)

6 150

100 5

50 4 0 3 Cointegrating equilibrium –50 Lending rate on new corporate loans (< €1 million) 2 –100

1 –150

0 –200 Mar. Jun. 04 Sep. 05 Dec. 06 Mar. 08 Jun. 09 Sep. 10 Dec. 11 Mar. 13 Jan. Apr. 06 Jul. 07 Oct. 08 Jan. 10 Apr. 11 Jul. 12 Oct. 2003 2005 12

Source: IMF staff estimates.

The beta coefficientsb ( 1, b2, b3, b4) define the coin- variables dynamically adjust toward their cointegrating 66 tegrating relationship, and ψ1 is a constant. These five equilibrium. Although these deviations reflect rela- variables are individually nonstationary, but are jointly tively small shocks for France, they point toward much stationary, and thus share a common stochastic trend. larger and more persistent shocks for Italy and Spain.

The ξt term can be thought of as a deviation from equilibrium, the expected value of which is zero.65 The economic interpretation is that these variables share a Data and estimation common equilibrium driven by a small set of factors; The models for France, Italy, and Spain were esti- econometrically, this is a “state from which there is no mated using monthly data for 2003–13 (about 120 endogenous tendency to deviate” (see Amisano and observations).67 The estimation was carried out in two Giannini, 1997). steps. In the first step, the cointegrating relationships Most of the time, however, the cointegrating vector were estimated following the Johansen methodology.68 will not be exactly in equilibrium. Figure 1.61 shows In the second step, the error correction terms from the that the actual corporate lending rate on small loans in estimated cointegrating relationships were constructed France has been fairly close to equilibrium. to enable the estimation of a vector autoregression Since 2007, there have been a large number of in first differences (with the error correction terms as shocks to sovereign, banking, and monetary variables; regressors). The final specification was obtained by therefore, the deviations from equilibrium have been starting out with a large number of variables proxying large and persistent. Figure 1.62 shows that the actual the key determinants of bank lending rates (the mon- interest rates on small loans in Italy and Spain are currently more than 100 basis points higher than what 66The difference between the current lending rate and the no-frag- their cointegrating equilibrium relationship would sug- mentation proxy reflects these deviations in addition to the steady- gest. These can be interpreted as short-term deviations state contributions of sovereign and banking stresses (estimated at from equilibrium that are corrected over time as the 100 basis points for Italy and 160 basis points for Spain). 67Lending rates on new small loans are from the ECB; swap, money market rates, and sovereign yields are from Bloomberg, L.P.; 65This measure captures the short-term deviation of the actual price-to-book ratios and other equity variables are from MSCI; and lending rate from the equilibrium lending rate, as characterized by industrial production data are from national statistical offices. the full sample (2003–13) parameter estimates of the cointegrating 68Unrestricted rank and maximum eigenvalue cointegration tests vector, conditional on the current level of the endogenous variables. were performed (see Johansen, 2009).

54 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Figure 1.62. Spain and Italy: Deviations from the Cointegrating Equilibrium

Corporate lending rates for both Italy and Spain are above their cointegrating equilibrium …signaling that lending rates are too high relative to their fundamental determinants. levels…

1. Spain: Deviations from Cointegrating Equilibrium 2. Italy: Deviations from Cointegrating Equilibrium 200 200 (Basis points) (Basis points)

150 150

100 100

50 50

0 0

–50 –50

–100 –100

–150 –150

–200 –200 Jan. Apr. 06 Jul. 07 Oct. 08 Jan. 10 Apr. 11 Jul. 12 Oct. Jan. Apr. 06 Jul. 07 Oct. 08 Jan. 10 Apr. 11 Jul. 12 Oct. 2005 12 2005 12

Source: IMF staff estimates.

etary policy, sovereign stress, bank health, and business • The third factor is bank health, as captured by the cycle), then narrowing them down to the “best” prox- price-to-book ratio of the banking system.70 A ies using general-to-specific modeling and extensive healthier banking system will have a higher price- diagnostic testing.69 to-book ratio, which, in turn, enables banks to lend more cheaply. Negative and statistically significant coefficients for Italy and Spain confirm these Model estimates dynamics are at play. In contrast, the bank health Table 1.8 shows the coefficients of the cointegrating coefficient for France is not statistically significant, vector for each country estimated for 2003–13 (the reflecting considerably lower banking and sovereign same model was also estimated for the crisis period, pressures. but the results are not shown because the sample • The fourth factor is the state of the business cycle, period is short and volatile). as captured by the industrial production index. As The key findings follow: found in other studies, the coefficient results indi- • The first factor, the seven-year swap rate, captures cate that weaker loan demand from firms has put the pass-through of monetary policy to lending downward pressure on lending rates. This parameter rates. It is highly statistically significant and has the is not significant for France, highlighting the lack of expected sign. In Italy and France, a 100 basis point sensitivity of lending rates to the state of the busi- policy rate cut translates into a decline of 57 basis ness cycle, especially compared with Italy and Spain. points in the corporate lending rate and a decline of • Finally, the sensitivity of lending rates to the corpo- 40 basis points in Spain. rate credit risk factor and the statistical significance • The second factor is sovereign stress. This factor is of this coefficient have increased during the crisis in significant for Spain and Italy, but not for France. all three countries, although the size of the coefficients are significantly larger for Italy and Spain. 69Akaike Information Criteria (AIC), Schwarz-Bayes Criteria (SBC), recursive stability tests, and analysis of residual behavior, 70Alternative measures of bank health, such as bank equity prices among others. and credit default swap spreads, produced similar results.

International Monetary Fund | October 2013 55

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 1.8. Determinants of Bank Interest Rates on New Small Loans (Loans of <€1 million) Italy France Spain Full Sample Full Sample Full Sample 2003–13 2003–13 2003–13 Endogenous Cointegrating Factors 1. Monetary Policy Stance 0.5689*** 0.5669*** 0.3965*** 2. Sovereign Stress 0.0074*** 0.004 0.0042*** 3. Bank Health –1.0061*** 0.490 –2.5011*** 4. Business Cycle 6.4442*** –0.074 5.6004*** Exogenous Factor 5. Corporate Credit Risk 0.0117** 0.0004*** 0.0460** R-squared 0.79 0.58 0.69 Source: IMF staff. 1. Seven-year swap rate. 2. Deviation of asset swap spreads (10-year sovereign bond yields minus swap rate of the same maturity) from their trend. The trend is time-varying and is estimated with a Christiano-Fitzgerald (1999) asymmetric band pass. 3. Bank price-to-book ratio, log. 4. Industrial production, log. 5. The rate of change of the difference between the 90th percentile and the mean of the corporate sector expected default frequency distribution, at the country level. ***, ** and * denote significance at the 1 percent, 5 percent, and 10 percent level, respectively.

56 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

annex 1.2. euro area corporate Debt Overhang Table 1.9. Amadeus Database, 2011 and implications for bank asset Quality Number of Firms Total Assets (thousands) Billions of Euros Percent of Total1 Objectives and analytical approach France 866 3,398 43 Germany 145 3,389 66 The challenges posed by the debt overhang for large Italy 1,035 3,194 100 Portugal 352 361 52 publicly traded firms in stressed euro area economies Spain 818 2,199 67 71 were analyzed in the April 2013 GFSR. In this Sources: Amadeus; national central banks; and IMF staff estimates. GFSR, the analysis of debt overhang is extended to 1Percent of financial and nonfinancial assets of the entire corporate sector, based on central bank flows of funds data; and IMF staff estimates. the broader corporate sector, particularly to the small and medium enterprise (SME) segment. Because percent of available coverage from public and official smaller firms in stressed euro area economies tend to sources.73 Coverage of the SME segment is especially have higher leverage and lower profitability than larger good in Italy, Portugal, and Spain. Although coverage firms, and also face tighter financing constraints and of the SME segment is considerably smaller in Ger- fewer deleveraging options, the focus is on firms’ debt- many, Amadeus still captures two-thirds of corporate servicing capacity. The capacity to service debt can be sector assets. gauged by looking at a firm’s interest coverage ratio (ICR).72 The size of the debt overhang in the broader corporate sector is defined as the share of total debt leverage, profitability, and Debt-at-risk outstanding owed by firms with ICRs of less than 1; Debt-at-risk in stressed euro area economies has this concept is often referred to as debt-at-risk. An ICR increased since 2001 and tends to be larger in the of less than 1 means that a firm is unable to service its SME sector (Figure 1.63, panels 1 and 2). SMEs have debt without making some adjustments, such as reduc- higher debt-at-risk because of a combination of high ing operating costs, or drawing down its cash reserves, leverage and weak profitability: or even borrowing more. The analysis of corporate • Leverage—as measured by the debt-to-EBITDA debt overhang concludes by drawing the implications ratio—increased sharply in stressed euro area for bank asset quality. economies and is now much higher than in the core, especially in Portugal and Spain, and among SMEs Data (Figure 1.63, panels 3 and 4). • These firms entered the crisis with weak profitabil- The analysis is based on firm-level annual data from ity (Figure 1.63, panel 5). In contrast to the core the Bureau van Dijk’s Amadeus database. The sample economies, in stressed economies, SMEs tend to includes more than 3 million nonfinancial firms, both have much weaker profitability than large firms have publicly traded and private, from France, Germany, (panel 6). Italy, Portugal, and Spain (see Table 1.9). In these Higher lending rates caused by financial fragmenta- economies, Amadeus’s coverage approaches 100 tion in the euro area have contributed to the higher debt-at-risk among corporates and SMEs in stressed The authors of this annex are Sergei Antoshin, Xiangming Fang, and Jaume Puig. euro area economies (Figure 1.64). 71The analysis in the April 2013 GFSR focused on debt repayment capacity. The debt overhang was defined as debt owed by firms that are unable to generate sufficient cash flows to repay debt (i.e., analysis of corporate Debt Overhang to reduce debt to sustainable levels in the medium term). The main conclusion was that the deleveraging required to bring the stock The “Chronic-Phase” and “Reversal-of- of debt down to sustainable levels could be a significant drag on Fragmentation” Scenarios growth. 72The interest coverage ratio (ICR) is defined as earnings before To assess debt-at-risk on a forward-looking basis, ICRs interest and taxes (EBIT) divided by interest expense. Interest are forecast under a “chronic-phase” scenario and a revenues or financial revenues are included in the calculation of earn- “reversal-of-fragmentation” scenario. ings (and thus partly offset interest expense). Given that the focus of the analysis is on firms’ medium-term prospects, the concept of EBIT—rather than EBITDA—is used because it allows the analysis to assess whether a firm is economically viable. In some cases, rating 73Variations in coverage across countries reflect mostly the agencies and analysts may use EBITDA when the focus is on a firm’s stringency of filing requirements at local registries and associated short-term cash position. penalties for failure to comply.

International Monetary Fund | October 2013 57

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 1.63. Leverage, Profitability, and Debt at Risk Figure 1.64. Bank Lending Rates to Small and Medium Enterprises Germany France Italy Large Medium Small (Percent) Spain Portugal 2. Debt at Risk by Firm Size, based 1. Debt at Risk based on ICR below 1 on ICR below 1, 2011 Portugal Spain Italy Three-year euro swap rates (spot and forward) 50 (Percent of total debt) (Percent of total debt by firm size) 60 45 10 50 40 9 35 40 30 8 25 30 20 7 20 15 6 10 10 5 5 0 0 2002 03 04 05 06 07 08 09 10 11 Germany France ItalySpain Portugal 4 3. Leverage Ratios 4. Leverage Ratios by Firm Size, 2011 3 6.5 (Debt to EBITDA) (Debt to EBITDA) 8 6.0 7 2 5.5 6 5.0 1 Scenarios 4.5 5 0 4.0 4 Jan. 2003 Jan. 05 Jan. 07 Jan. 09 Jan. 11 Jan. 13 Jan. 15 Jan. 17 3.5 3 3.0 Sources: Haver Analytics; and IMF staff estimates. 2 2.5 2.0 1 1.5 0 2001 02 03 04 05 06 07 08 09 10 11 Germany France Italy Spain Portugal Earnings before interest and taxes (EBIT) are projected using GDP growth forecasts. Time-series regressions 5. Profitability 6. Profitability by Firm Size, 2011 13 (EBITDA to assets; percent) (EBITDA to assets; percent) 18 specific to country, sector, and firm size are estimated, as

12 16 are country-specific panel regressions, where corporate 14 profitability (EBIT over assets), is regressed on GDP 11 12 growth. GDP growth projections under the October 10 10 2013 World Economic Outlook baseline and alternative 9 8 scenarios are used in the reversal-of-fragmentation and 6 8 chronic-phase scenarios, respectively. 4 Interest rates on corporate debt are also projected 7 2 under the chronic-phase and reversal-of-fragmentation 6 0 2001 02 03 04 05 06 07 08 09 10 11 Germany France Italy Portugal Spain scenarios. The symmetric shocks are calibrated based on the econometric exercise presented in Annex 1.1.74

Sources: Amadeus database; and IMF staff estimates. This is broadly consistent with a return of SME lending Note: EBITDA = earnings before interest, taxes, depreciation, and amortization. spreads over swaps to precrisis levels under the reversal- ICR = interest coverage ratio. of-fragmentation scenario (see Figure 1.64). The shock for large companies is assumed to be half that for SMEs, also in line with a return to precrisis lending spreads.

74The exercise described in Annex 1.1 finds that removing fragmentation would result in a reduction of lending rates to small and medium enterprises (SMEs) of about 100 basis points in Italy and 160 basis points in Spain. We assume that the effect on lending rates to SMEs in Portugal would be about 200 basis points. The reduction in lending rates under the reversal-of-fragmentation scenario is assumed to be phased in during 2014–16 as gradual progress is made toward banking and fiscal union. A symmetric shock is assumed under the chronic-phase scenario.

58 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

“Persistent” Debt Overhang Figure 1.65. Projected Corporate Debt Overhang in Italy, Portugal, and Spain The debt overhang declines significantly as growth (Based on ICR < 1; share of total debt) recovers and financing costs decline under the reversal- of-fragmentation scenario.75 Sensitivity analysis shows 0.55 that the debt overhang declines by about 5 percentage 0.50 points, on average, if fragmentation is reduced by 100 basis points or growth improves by 3 percentage points. 0.45

The reversal-of-fragmentation scenario provides a 0.40 basis for assessing the size of the “persistent” corporate debt overhang. This persistent debt overhang is defined 0.35 as the share of debt in stressed euro area economies 0.30 that is owed by firms with an ICR of less than 1 Italy, Portugal, Spain: Chronic-phase scenario under the reversal-of-fragmentation scenario, in excess Italy, Portugal, Spain: Reversal of fragmentation scenario 0.25 Core of the equivalent share in the core. Firms in stressed 0.20 economies and in the core are expected to face similar financial conditions under the reversal-of-fragmenta- 0.15 tion scenario, but even under these benign financing 0.10 conditions, and the assumed recovery in profitability 2011 12 13 14 15 16 17 18 in line with the projected economic recovery, a sizable persistent debt overhang of almost one-fifth of total Sources: Amadeus database; and IMF staff estimates. corporate debt remains in stressed economies (indi- Note: ICR = interest coverage ratio. cated by the bracket in Figure 1.65). assessing implications for bank asset Quality entire banking system. In addition, the GFSR analysis Finally, this GFSR illustrates the implications of corpo- has the advantages of using a consistent approach across rate sector stresses for bank asset quality by estimating firms and countries, and providing an up-to-date assess- potential bank losses on corporate exposures (assuming ment of corporate sector stress and its implications for no improvement in corporate fundamentals over the banks (see Box 1.5 for more details). next two years) and comparing them with bank buffers Assuming that corporate fundamentals remain to gauge the extent to which these asset quality prob- unchanged, the potential losses during 2014–15 arising lems might not have yet been dealt with. from the corporate exposures of the banking system are Compared to the standard bank solvency stress tests, assessed as follows: the GFSR analysis provides a complementary (yet, less • ICRs as of 2013 are extrapolated using the latest data precise) perspective on the problem of corporate stress available, with estimates of EBIT based on the 2011 and its implications for bank asset quality. While stan- firm-level data from Amadeus and October 2013 World dard bank solvency stress tests typically rely on granular Economic Outlook GDP growth and the estimates of information on the individual bank exposures to dif- interest expense based on actual lending rates.76 ferent types of borrowers, the GFSR analysis considers • The firm-level ICRs are mapped into the prob- aggregate banking system exposures, and hence cannot abilities of default (PDs) by (1) assigning implied yield any insights about individual banks. On the other credit ratings to companies in the sample based on hand, the GFSR analysis uses very detailed nonfinancial average ICRs by credit rating for companies rated firm-level data to assess the extent of potential credit by Moody’s, and (2) assigning PDs over the next quality deterioration on corporate exposures of the two years to each implied rating based on historical 75The analysis assumes that balance sheets remain static in the forecast period. Aggregate data for 2012 show that corporate debt 76The EBIT projections use the same empirical relationships between declined in Spain, and credit data suggests that the decline in debt is profitability and GDP growth as the ones discussed in the section on greater in weaker companies. However, the lack of data on the asset “Analysis of Corporate Debt Overhang” in this Annex. In the case of side and on the effect of asset sales on the income statement prevents Portugal, the estimated ICRs are adjusted using actual 2012 data (avail- this study from taking deleveraging into account. able to date) by sector/size that were provided by the Bank of Portugal.

International Monetary Fund | October 2013 59

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

default rates of companies rated by Moody’s. Aggre- Figure 1.66. Probabilities of Default in the Corporate Sector gate PDs on corporate debt owed to banks are esti- (Percent as of 2011; over the next two years) mated at the country level as the average of PDs of Based on ICRs Based on leverage, ICRs, and proﬁtability individual firms weighted by the share of each firm’s 40 debt in aggregate country debt.77 This mapping of corporate credit scores into implied ratings and PDs is a standard approach used by rating agencies and 35 banks. The estimation of PDs is robust to the use of alternative corporate vulnerability indicators (other than ICRs), such as profitability and leverage ratios (Figure 1.66), and to the use of historical default 30 rates from other rating agencies (Table 1.10). Gener- ally, PDs based on ICRs and on Moody’s historical default rates tend to be lower than those based on 25 other vulnerability indicators and rating agencies. • Loss rates at the country level are obtained by multi- plying estimated aggregate PDs by loss given default (LGD) ratios. A range of 10 percentage points 20 Italy Spain Portugal around the standard Basel LGD ratio of 45 percent is used to estimate a range of potential loss rates (to reflect uncertainties about collateral valuations). Sources: Amadeus database; and IMF staff estimates. Note: ICR = interest coverage ratio. Potential bank losses from corporate exposures at the aggregate country level are obtained by applying these aggregate loss rates to the stock of loans The key parameters used in the GFSR analysis, such as extended to nonfinancial corporates by monetary PDs and LGD ratios, appear to be broadly in line with financial institutions in each country.78 those used in available stress testing exercises that consider • The estimated potential losses are related to existing the entire stock of loans. For example, using the same buffers, including provisions on corporate loans, approach as described previously to estimate three-year operating profits, and Tier 1 capital79 (see Figure PDs at the end of 2011 yields an estimated aggregate PD 1.53 in the main text of the chapter). for Spain that falls within the range of the parameters used in the Oliver Wyman stress tests published in 2012 77Fifty percent of debt of large corporates and all debt of SMEs is (Table 1.11); the same is true for the LGD assumptions. assumed to be owed to banks. 78For Spain, potential losses on bank loans are adjusted for the loans transferred to SAREB (Spain’s asset management company) in available only on a consolidated basis at the system level. Provisions December 2012 and February 2013. on corporate loans are estimated by applying the share of corporate 79Buffers on domestic corporate exposures may be overestimated loans in nonperforming loans to the stock of total provisions, includ- because provisions, operating profits, and core Tier 1 capital data are ing general provisions.

Table 1.10. Mapping of Corporate Vulnerability Indicators to Probabilities of Default Corporate Vulnerability Indicators1,2 Cumulative Default Rates3 Moody’s Standard & Poor’s Fitch ICR Profitability Leverage Implied Rating Year 1 Year 2 Year 1 Year 2 Year 1 Year 2 27.0 21.1 0.6 Aaa/AAA 0.0 0.0 0.0 0.0 0.0 0.0 14.7 13.5 1.5 Aa/AA 0.0 0.1 0.0 0.0 0.0 0.0 9.3 12.0 2.0 A/A 0.1 0.2 0.1 0.2 0.1 0.2 5.2 9.9 2.6 Baa/BBB 0.2 0.5 0.2 0.6 0.2 0.7 3.4 9.3 3.2 Ba/BB 1.1 3.1 0.9 3.0 1.1 2.8 1.6 7.3 4.8 B/B 4.1 9.6 4.5 10.0 2.0 4.8 0.5 3.2 7.6 Caa-C/CCC-C 16.4 27.9 26.8 36.0 24.9 31.9 Sources: Fitch; Moody’s; Standard and Poor’s; and IMF staff estimates. 1ICR is defined as EBIT/interest expense; profitability is defined as EBIT/average assets; leverage is defined as Debt/EBITDA. 2The probabilities of default are extrapolated beyond those corresponding to the implied rating C for firms with weaker vulnerability indicators. 3Based on 1970–2012 for Moody’s, 1981–2011 for S&P, and 1990–2012 for Fitch. Note: EBITDA = earnings before interest, taxes, depreciation, and amortization; ICR = interest coverage ratio.

60 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 1 MakIng the transItIon to stabIlIty

Table 1.11. Comparison of the GFSR Analysis with Oliver Wyman’s Stress Tests for Spain PD PD LGD LGD Baseline Adverse Baseline Adverse Oliver Wyman, as of 2011 (for 2012–14) Real Estate Developers 0.61 0.88 0.39 0.47 Large Corporates 0.09 0.17 0.47 0.49 Small and Medium Enterprises 0.21 0.35 0.40 0.42 Total Corporate Sector 0.29 0.45 0.42 0.46 GFSR, as of 2011 (for 2012–14) 0.37 0.45 Sources: Bank of Spain; IMF staff estimates. Note: LGD = loss given default; PD = probability of default.

box 1.5. the gFSr analysis of corporate credit Quality versus bank Stress tests

The methodological approach used in this GFSR Probabilities of Default to assess potential losses on corporate exposures of the • In a standard bank solvency stress test, PD is typically banking systems can be compared with standard stress defined as the one-year probability that a performing tests that are carried out in the context of Financial loan becomes nonperforming (actual default rates from Sector Assessment Programs, by looking at the main the central credit registry provided by central banks are elements of the analysis: commonly used; forward-looking PDs are also often tied to specific macroeconomic assumptions). Exposures • In the GFSR analysis, the PDs are estimated at the firm • Standard bank solvency stress tests focus mainly on level (not at the loan level) and are obtained by map- additional losses on performing loans and, in some ping current corporate vulnerability indicators into PDs cases, capture the impact on existing nonperforming through implied credit ratings for individual companies. loans (NPLs) through, for instance, adjusting loss given default (LGD) rates in the stress scenario. The Loss Given Default Rates analysis is based on granular, bank-level data on loan • The LGD rate used in many standard stress tests are exposures. In some cases, the adequacy of provisions typically provided by supervisory authorities, who against the existing stock of NPLs is assessed as well. may use different methodologies to estimate aggre- • The GFSR analysis considers the entire stock of gate LGDs (e.g., coverage ratios, LGDs estimated loans, sidestepping the issue of banks’ classification of from collateral valuation models, and so forth). exposures as performing or nonperforming and any • The GFSR analysis uses the Basel LGD ratio of cross-country differences in NPL definitions. The 45 percent (and a range of ±10 percentage points analysis considers aggregate corporate loan exposures around the 45 percent level to reflect uncertainties of all banks operating in a given country. about collateral valuation).

International Monetary Fund | October 2013 61

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY references Financial Stability Oversight Council (FSOC), 2013, Annual Report (Washington). Acharya, Viral V., and T. Öncü, 2013, “A Proposal for the Reso- Goretti, Manuela, and Marcos Souto, 2013, “Macro-Financial lution of Systemically Important Assets and Liabilities: The Implications of Corporate (De)Leveraging in the Euro Area Case of the Repo Market,” International Journal of Central Periphery,” IMF Working Paper No. 13/154 (Washington: Banking, Vol. 9, No. S1, pp. 293–394. International Monetary Fund). Amisano, Gianni, and Carlo Giannini, 1997, Topics in Structural Hoenig, Thomas M., 2013, “Tangible Equity Capital Ratios for VAR Econometrics (New York: Springer). Global Systemically Important Banks,” Federal Deposit Insur- Arslanalp, Serkan, 2013, “Banking Sector Risks under the ance Corporation (Kansas City). Government’s New Policies,” in Japan: Selected Issues, IMF International Monetary Fund (IMF), 2010, Integrating Stability Country Report No. 13/254 (Washington: International Assessments under the Financial Sector Assessment Program into Monetary Fund). Article IV Surveillance: Background Material (Washington). Banca d’Italia, 2013, Financial Stability Report (Rome). ———, 2013a, Euro Area Policies: 2013 Article IV Consulta- Barisitz, Stephan, 2013, “Nonperforming Loans in Western tion: Selected Issues, IMF Country Report No. 13/231 Europe—A Selective Comparison of Countries and National (Washington). Definitions,” Oesterreichische Nationalbank Focus on European ———, 2013b, People’s Republic of China: 2013 Article IV Con- Economic Integration Q1/13 (Vienna). sultation, IMF Country Report No. 13/211 (Washington). Begalle, Brian, Antoine Martin, James McAndrews, and Susan ———, 2013c, United States: Article IV Consultation, IMF McLaughlin, 2013, “The Risk of Fire Sales in the Tri-Party Country Report No. 13/236 (Washington). Repo Market,” Federal Reserve Bank of New York Staff ———, 2013d, United States: Article IV Consultation: Selected Report No. 616 (New York). Issues, IMF Country Report No. 13/237 (Washington). Bloom, Nicholas, 2009, “The Effect of Uncertainty Shocks,” ———, 2013e, Portugal: Seventh Review Under the Extended Econometrica, Vol. 77, No. 3, pp. 623–85. Arrangement and Request for Modification of End-June Christiano, Lawrence J., and Terry J. Fitzgerald, 1999, “The Performance Criteria, IMF Country Report No. 13/160 Band Pass Filter,” NBER Working Paper No. 7257 (Cam- (Washington). bridge, Massachusetts: National Bureau of Economic International Organization of Securities Commissions (OICV- Research). IOSCO), 2012, “Principles on Suspensions of Redemptions Chung, Hess, Jean-Philippe Laforte, David Reifschneider, and in Collective Investment Schemes: Final Report,” FR02/12 John C. Williams, 2011, “Have We Underestimated the Like- (Madrid: Technical Committee of the International Organiza- lihood and Severity of Zero Lower Bound Events?” Federal tion of Securities Commissions). Reserve Bank of San Francisco Working Paper No. 2011–01. Johansen, Søren, 2009, Likelihood-Based Inference in Cointegrated Dattels, Peter, Rebecca McCaughrin, Ken Miyajima, and Jaume Vector Autoregressive Models (New York: Oxford University Puig, 2010, “Can You Map Global Financial Stability?” Press). IMF Working Paper No. 10/145 (Washington: International Lam, Raphael, 2013, “Japanese Financial Institutions Expanding Monetary Fund). Abroad: Opportunities and Risks,” in Japan: Selected Issues, Dudley, William, 2013, “Fixing Wholesale Funding to Build IMF Country Report No. 13/254 (Washington: International a More Stable Financial System,” remarks at the New York Monetary Fund). Bankers Association Annual Meeting and Economic Forum, Oliver Wyman, 2012, “Bank of Spain Stress Testing Exercise,” New York, February 1. June 21 (Madrid). European Banking Authority, 2012, Final Report on the Santos, Carlos, 2013, “Bank Interest Rates on New Loans to Implementation of Capital Plans Following the EBA’s 2011 Non-Financial Corporations—A Look at a New Set of Micro Recommendation on the Creation of Temporary Capital Buffers Data,” Banco De Portugal Financial Stability Report May 2013 to Restore Market Confidence (London). (Lisbon). European Central Bank (ECB), 2009, Monthly Bulletin (August). Vinãls, José, Ceyla Pazarbasioglu, Jay Surti, Aditya Narain, ———, 2010, Monthly Bulletin (May). Michaela Erbenova, and Julian Chow, 2013, “Creating a Safer ———, 2013a, Corporate Finance and Economic Activity in the Financial System: Will the Volcker, Vickers, and Liikanen Euro Area: Structural Issues Report 2013, Occasional Paper No. Structural Measures Help?” IMF Staff Discussion Note No. 151 (Frankfurt). 13/4 (Washington: International Monetary Fund). ———, 2013b, Financial Stability Review (Frankfurt, May). Zoli, Edda, 2013, “Italian Sovereign Spreads: Their Determi- Financial Stability Board (FSB), 2013, “Strengthening Oversight nants and Pass-Through to Bank Funding Costs and Lending and Regulation of Shadow Banking: A Policy Framework for Conditions,” IMF Working Paper 13/84 (Washington: Inter- Addressing Shadow Banking Risks in Securities Lending and national Monetary Fund). Repos” (Basel).

62 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution chapter 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

SUMMARY

olicymakers in economies hit hard by the global financial crisis have been concerned about weak growth in credit, considered a main factor in the slow economic recovery. Many countries with near-zero or negative credit growth for a number of years sense that the strategy of very accommodative macroeconomic policies has been insufficient in reviving credit activity. Authorities have therefore implemented a host of Ppolicies to target credit creation (which are documented in an appendix to the chapter).1 Effectively targeting these policies requires identifying the factors that underlie the weakness in credit. In credit markets, these factors center around the buildup of excessive debt in households and firms, reducing their credit demand, as well as excessive leverage (or a shortage of capital) in banks, restricting their ability or willingness to provide additional loans. The government could also usefully alleviate a shortage of collateral (perhaps resulting from large declines in asset values), which could constrain credit activity. To address such a technically challenging exercise, this chapter takes a stepwise approach. The first step is an attempt to identify the constraints to credit through the use of lending surveys—trying to disentangle whether banks are unwilling to lend (on the supply side) or whether firms or households are reluctant to borrow (on the demand side). This distinction helps narrow down the set of policies to consider, which differ depending on the side of the market that faces the major constraint. A more challenging second step—which is hampered by the lack of sufficient data for many countries—is to identify the individual factors that are constraining credit, specifically what makes banks unwilling to lend or households and firms reluctant to borrow. Using this approach for several countries that have sufficient data, the analysis finds that the constraints in credit markets differ by country and evolve over time. This reinforces the importance of a careful country-by-country assessment and the need for better data on new lending. In many cases, demand- and supply-oriented policies will be complementary, but their relative magnitude and sequencing will be important. For example, relieving excessive debt in firms will help only if the banking sector is adequately capitalized. Policymakers should also recognize the limits of credit policies and not attempt to do too much. Because many policies will take time to have an impact, assessment of their effectiveness and the need for additional measures should not be rushed. When credit policies work well to support credit growth and an economic recovery, financial stability is enhanced, but policymakers should also be cognizant of longer-term potential risks to financial stability. The main risks center on increased credit risk, including a relaxation of underwriting standards and the risk of “evergreening” existing loans. Mitigation of these risks may not be necessary or appropriate while the economic recovery is still weak, as it could run counter to the objectives of the credit policies (which are often designed to increase risk taking); still, policymakers will need to continually weigh the near-term benefits against the longer-run costs of policies aimed to boost credit.

1Appendix 2.1 is available online on the GFSR page at both www.imf.org and http://elibrary.imf.org.

International Monetary Fund | October 2013 63

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Introduction Figure 2.1. Real Credit Growth (Percent; year over year)

This chapter examines possible reasons behind the Advanced economies Emerging market economies Global weakness in private credit in many countries since 2008, and it offers a framework for assessing the 1. Total Credit 25 various policies that have been implemented to revive 20 credit markets. These policies were put in place in the wake of a sharp decline in lending growth in most 15 advanced economies and some emerging markets 10 (Figure 2.1). Total credit to the private sector showed sluggish growth, while credit extended by domestic 5 banks declined for advanced economies. 0 Policymakers want to support credit markets because –5 the decline in lending is seen to be a primary factor 2000 02 04 06 08 10 12 in the slow recovery. Well-functioning credit markets make major contributions to growth and macroeco- 2. Bank Credit 25 nomic stability, and restarting credit plays an impor- 20 tant role in economic recovery after a downturn. Recent studies show that creditless recoveries are typi- 15 cally slower than those with more robust credit growth, 10 at least for the first few years, especially after recessions 5 that feature large declines in asset prices, a characteris- tic of this financial crisis.2 0 Credit-supporting policies are most effective if they –5 target the constraints that underlie the weakness in 2000 02 04 06 08 10 12 credit. Policymakers are sensing that the exceptionally accommodative macroeconomic policies imple- Sources: Bank for International Settlements (BIS); and IMF staff estimates. mented since the crisis have been insufficient and that Note: Unweighted average of real credit growth rates across countries. Total credit includes private sector borrowing (loans and debt instruments) from domestic banks and all other additional measures targeting credit creation could sources (“other credit”), such as other domestic nonbanks and foreign lenders (see BIS, 2013). further underpin the recovery. To target such policies Advanced economies include Australia, Austria, Belgium, Canada (not included in panel 2), Czech Republic, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Korea, effectively, policymakers must determine the factors Luxembourg (from 2004:Q1), Netherlands, Norway, Portugal, Singapore, Spain, Sweden, that constrain lending activity. This chapter provides a Switzerland, United Kingdom, and United States; emerging market economies include Argentina, Brazil, China, Hungary, Indonesia, India, Malaysia, Mexico, Poland, Russia, South Africa, Thailand, 3 framework for this purpose. and Turkey. Global consists of advanced and emerging market economies identiﬁed above. In the past, a clear case for government intervention emerged only when there were market failures or exter- sidered. In addition to exacerbating the current crisis, nalities, but this crisis showed that such developments these amplifying tendencies appear also to be present in in credit markets can be prevalent, amplifying upturns upswings, as the current crisis was in part precipitated and downturns. This is leading to some rethinking that by excessive credit creation during the preceding boom. the role of government policies, particularly macropru- Therefore, policymakers need also to mitigate exces- dential policies, may be larger than previously con- sive credit creation during economic upswings, which would lower the risk of similar future crises, and thus in The authors of this chapter are S. Erik Oppers (team leader), Nicolas Arregui, Johannes Ehrentraud, Frederic Lambert, and turn obviate the need for credit-supporting policies. Kenichi Ueda. Research support was provided by Yoon Sook Kim. Although well-designed credit policies can support Fabian Valencia shared data and methodology. credit intermediation and a more robust economic 2 The importance of credit in supporting economic recovery has recovery, the choice of policies should also take into been discussed at length in the literature. See Table 2.7 for a summary of these studies, under the heading “Creditless Recovery.” account direct or indirect fiscal costs and unintended 3Focusing on these potential constraints to credit (rather than consequences for financial stability. Although many simply its weakness) could also prevent policymakers from doing too policies have been implemented in a range of coun- much. In some cases, it may be that an expansion of credit is not desirable; deleveraging by firms or households may in fact be impor- tries, which helped to keep financial instability from tant to pave the way for more sustainable economic growth. worsening and the supply of credit from slipping

64 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

even further, there is not always a clearly favorable Table 2.1. Identifying Countries with Weak Credit Growth, BIS Data cost-benefit nexus. In particular, policymakers should Total Bank Credit Total Credit Total Credit to be mindful of possible consequences for financial to Private to Private Credit to Nonfinancial stability in the medium term, especially if new credit Sector Sector Households Corporations is extended without adequate attention to the risks Advanced Economies Australia involved (including if credit is extended by nonbanks). Austria Weak Weak Belgium Weak In addition, these policies may have fiscal costs, and Canada . . . policymakers should make sure that initiatives are as Czech Republic cost-effective as possible. Denmark Weak Weak Weak Weak Finland In connection with recent efforts to revive credit mar- France Germany Weak Weak Weak Weak kets, the chapter addresses the following questions: Greece Weak Weak Weak Weak • Which countries have seen weak credit growth Ireland Weak Weak Weak recently, and what are the potential causes? Italy Weak Weak Weak Weak Japan Weak Weak Weak • What policies have been put in place in various Korea countries to support credit? Luxembourg Weak • Have the policies targeted the constraints that Netherlands Weak Weak Weak Norway Weak underlie the weakness in credit? Portugal Weak Weak Weak Weak Singapore • What, if anything, can policymakers do to make Spain Weak Weak Weak Weak credit policies more effective? Sweden The analysis confirms that constraints in credit markets Switzerland United Kingdom Weak Weak Weak Weak differ by country, and policies to support credit should United States Weak Weak Weak be based on a country-specific analysis of the constraints Emerging Market Economies Argentina ...... that government policy may alleviate. As expected, higher Brazil ...... bank funding costs and lower bank capital have reduced China Hungary Weak Weak Weak Weak the ability of banks to supply loans, and high debt levels India in firms and households (along with lower GDP growth Indonesia Malaysia ...... forecasts) have lowered credit demand (and affected credit Mexico supply). These factors are present to different degrees in Poland different countries. Policymakers should be mindful of Russia ...... South Africa interactions with other policies, including regulatory mea- Thailand sures, direct and contingent costs to the government, and Turkey Sources: Bank for International Settlements (BIS); De Nederlandsche Bank; Instituto Nacional potential longer-term financial stability implications. If de Estadistica y Censos (INDEC); IMF, World Economic Outlook; Banca d’Italia; and IMF staff appropriate, prudential measures to mitigate such stability estimates. Note: Weak credit is identified if the average year-over-year credit growth (deflated by con- risks should be put in place. sumer price index inflation; official wage index inflation for Argentina) is negative over a two- year window (2011:Q1–2012:Q4). Growth rates are computed using stocks in local currency and not adjusted for exchange rate variations. Cells are blank if this criterion is not met. Cells with “. . .” indicate that the data are not available, except for bank credit in Canada, which is Recent Developments in Credit Markets ignored because of a break in the series. Total credit includes private sector borrowing (loans and debt instruments) from domestic banks and from all other sources (“other credit”), such Where has Credit Growth Been Weak? as domestic nonbanks and foreign lenders (see BIS, 2013). To find where credit growth has been weak, a simple rule can be applied. A transparent operational rule separate determination is made for particular segments used in the literature defines weak credit growth as of credit markets when disaggregated data are available. negative average real credit growth over a certain Many advanced economies have experienced weak

period.4 To identify where credit is currently still weak bank credit growth (Table 2.1), including the United several years into the crisis, this rule is applied to a Kingdom and the United States, as have many euro number of countries, using data from the Bank for area countries (including Austria, Belgium, Germany, 5 International Settlements (BIS) and other sources. A Greece, Ireland, Italy, Portugal, and Spain). Interestingly, 5The selection of countries is mostly unchanged if only the last 4For instance, Abiad, Dell’Ariccia, and Li (2011) and Sugawara year of credit is considered. The Netherlands would join the group and Zalduendo (2013) use negative average credit growth over recov- of countries with weak bank credit growth, and the United States ery periods to identify creditless recoveries. and Luxembourg would drop from the list. Austria, Belgium,

International Monetary Fund | October 2013 65

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 2.2. Identifying Countries with Weak Credit Figure 2.2. Perceived Obstacles in Access to Finance Growth, Other Data Sources (Percent of respondents) Bank Credit to Private Sector 1. Firms Seeing Access to Finance as Most Pressing Problem 30 Albania Belarus Weak Bosnia and Herzegovina 25 Bulgaria Weak Croatia Weak 20 Estonia Weak Iceland Weak 15 Kosovo Latvia Weak 10 Lithuania Weak FYR Macedonia 5 Moldova Montenegro Weak 0 Romania Large SMEs Large SMEs Large SMEs Large SMEs Large SMEs Serbia Spain Italy France Germany Euro area Slovak Republic Slovenia Weak Ukraine 2. Reasons Not to Apply for a Bank Loan Other reasons Did not apply because of possible rejection Sources: European Central Bank; IMF, International Financial Statistics and World Economic Outlook; Haver Analytics; and IMF staff estimates. 100 Note: Weak credit is identified if the average year-over-year credit growth (deflated by consumer price index inflation) is negative over a two-year window 80 (2011:Q1–2012:Q4). Growth rates are computed using stocks in local currency and not adjusted for exchange rate variations. Column is blank if this criterion is not met. 60 Ireland and the United States show weak credit growth 40 (from all sources) to households but not to nonfinancial 6,7 corporations. In addition, data from non-BIS sources 20 indicate that many countries in central, eastern, and 0 southeastern Europe, including Bulgaria, Croatia, Slove- Large SMEs Large SMEs Large SMEs Large SMEs Large SMEs nia, and the Baltic countries, have also recently seen weak Spain Italy France Germany Euro area bank credit growth (Table 2.2). 3. Success Rate on Applications for Bank Loans Survey data indicate particular challenges faced by Don't know Applied but rejected small and medium enterprises (SMEs) as they attempt Applied but refused Applied and got a portion to access credit. The most recent European Central because cost too high Applied and got most or all 100 Bank (ECB) Survey on the Access to Finance of SMEs in the euro area (SAFE) (ECB, 2013) shows that SMEs 80 tend to report access to finance as their most pressing problem more often than do large companies (Figure 60 2.2). Also, their loan applications were less success- 40 ful than those of large corporations. In addition, the survey showed that SMEs were discouraged more often 20 than larger firms from applying for a loan because of 0 the anticipation of rejection. A reluctance to apply Large SMEs Large SMEs Large SMEs Large SMEs Large SMEs may also be a result of the higher lending rates they Spain Italy France Germany Euro area

Luxembourg, and Norway had mildly negative bank credit growth Source: European Central Bank (2013). and actually had positive average real credit growth if other sources Note: SMEs = small and medium enterprises. The distinction between large corporations and of credit (in addition to banks) are included. SMEs is available only for the countries shown. 6Ireland showed negative real growth of credit to nonfinancial corporations in the last quarter of 2012. 7Alternative definitions of weak credit growth could be based on either real credit or a ratio of credit to GDP significantly below trend. Most of the countries selected with this chapter’s basic rule are also selected by at least one of these additional criteria. These definitions are the converse of methodologies in the literature that identify credit booms, including Borio and Lowe (2002); Mendoza and Terrones (2008); Borio and Drehmann (2009); and Drehmann, Borio, and Tsatsaronis (2011).

66 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Figure 2.3. Interest Rate Spread between Loans to SMEs and may choose not to take out loans, but rather focus to Larger Firms on paying off their loans. Banks may also find highly (Basis points) indebted borrowers less creditworthy. Debt overhang Euro area Germany Spain in banks can also affect credit supply: highly leveraged France Italy 300 banks may have difficulty obtaining funding and thus lack the liquidity to make additional loans.

In most credit cycles, government intervention to mitigate the factors constraining credit is generally not 200 necessary and may ultimately spur too much credit activity, but when various amplification mechanisms are at play, such as in the current cycle, government intervention has a clearer role. In the past, the difficulties mentioned previously could be overcome by the 100 private sector, but they may persist in times of crisis, amplifying the downturn. For example, in the current crisis, declining asset prices restricted credit, worsening the recession, which led to further downward pressure 0 on asset prices. In such situations, the government can 2007 08 09 10 11 12 13 implement various policies (detailed below) to ease

Sources: European Central Bank; and IMF staff estimates. credit constraints and break the downward spiral. Note: SMEs = small and medium enterprises. Spread is calculated as the difference between This chapter investigates the role of these factors in the lending rate for loans of less than €1 million and loans greater than €1 million. detail, but on the face of it, evidence is growing that they have contributed to the weakness in credit in recent face relative to other corporations (see Chapter 1 and years. Indebtedness of households and firms rose mark- Figure 2.3). edly in the run-up to the crisis, potentially contributing to a problem of debt overhang for borrowers in What Factors May Be Constraining Credit? some countries (Figure 2.4). Also, the major asset price declines seen globally in 2008 and 2009 depressed the Theoretically, credit markets suffer from potential diffi- value of large classes of collateral (Figures 2.5 and 2.6). culties that may be amplified in recessions (Annex 2.1). A later section investigates the extent to which these Some major factors that may constrain credit include developments played a role in recent years (and perhaps the following: still do) in restricting credit demand and supply. • Collateral constraints: To secure a loan, a borrower must often post collateral (an asset), because there is an information asymmetry: the lender does not know What Policies have Been Implemented to the borrower’s repayment behavior. A drop in the Support Credit? value of collateral as a result of asset price declines (in Policymakers have sought to boost economic activity real estate or stock markets, for example) shrinks the by implementing policies to support credit growth. loan that can be obtained with that collateral, tight- Appendix 2.1 provides an inventory of the policies ening credit demand as well as supply—indeed, the adopted in the major economies that have experienced amount of collateral required by banks may also rise weakness in private credit growth.8 The goal of these if bankers forecast further declines in its value. Lower collateral prices also lower the amounts banks will 8 lend to each other in interbank markets, restricting This appendix is only available online at www.imf.org/External/ Pubs/FT/GFSR/2013/02/index.htm. This inventory includes the bank funding and again tightening credit supply. group of countries covered in Tables 2.1 and 2.2, most European • Debt overhang: Excessively indebted firms may not countries (except, notably, the financial centers Luxembourg and pursue otherwise profitable business opportunities Switzerland), along with Japan, the United States, and some G20 countries that showed a marked deceleration of credit growth even and may strive to bring down their leverage, lowering though the simple rule in this analysis did not identify them as hav- credit demand. Similarly, highly indebted households ing weak credit (Australia, India, Korea, and South Africa).

International Monetary Fund | October 2013 67

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 2.4. Corporate and Household Debt Outstanding Figure 2.5. Stock Price Index (Percent of GDP) (2005 = 100)

Global Advanced economies Emerging market economies France Italy Spain United States Japan 200 United Kingdom Nonﬁnancial Corporations1 220 1. 2. 220 200 200 150 180 180 160 160 140 140 100 120 120 100 100 80 80 50 60 60 40 40 20 20 2000 02 04 06 08 10 12 2000 02 04 06 08 10 12 0 2000 02 04 06 08 10 12 Households2 120 3. 4. 120 Source: Morgan Stanley Capital International. Note: Global comprises advanced and emerging market economies. 100 100

80 80

60 60 Figure 2.6. Real House Price Index 40 40 (2005 = 100)

20 20 1. France Italy Spain 0 0 140 2000 02 04 06 08 10 12 2000 02 04 06 08 10 12 120

100 Source: Haver Analytics. Note: Seasonally adjusted GDP. 80 1Corporate debt includes securities other than shares (excluding financial derivatives for the United Kingdom), loans, and other accounts payable on a nonconsolidated basis. Consolidated 60 debt levels are significantly lower for some countries, especially those in which intercompany loans represent a large share of nonfinancial corporate debt. This calls for caution when doing 40 cross-country comparisons. 2Including nonprofit institutions serving households. 20 0 2000 02 04 06 08 10 12

2. United Kingdom United States Japan policies includes addressing the restrictions mentioned 140 in the previous section (mainly by alleviating debt 120 overhang) and easing various other constraints to free up the supply of credit. 100 Policies aimed at alleviating balance sheet problems 80 include the following: 60

• Corporate debt restructuring: To ease the debt overhang in 40 the corporate sector, which has depressed loan demand, 20 many governments have taken a leading role in corporate debt restructuring through state-owned banks and 0 2000 02 04 06 08 10 12 through asset management companies that took over the assets of distressed banks. In some countries, corporate Sources: Organization for Economic Cooperation and Development; and IMF, International bankruptcy rules were modified and speedier out-of- Financial Statistics. court resolution programs were introduced. Note: Deﬂated by consumer price inﬂation.

68 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

• Household debt restructuring: Applying strategies Figure 2.7. Relative Number of Credit Supply and Demand similar to those used in corporate debt restructur- Policies Currently in Place ing, some governments have sought to ease house- Bosnia and Herzegovina, Czech Republic, South Africa, Turkey hold debt overhang by implementing household Austria, France, Germany debt restructuring programs, most importantly for Finland, Slovak Republic, Sweden “underwater” mortgages (that is, the loan balance Bulgaria, Norway, Poland 0.4 is higher than the home value). In some countries, personal bankruptcy rules were modified, and out- Moldova Latvia Ireland of-court resolution programs were implemented. 0.3 Portugal • Bank restructuring: In the recent past, many govern- Estonia ments have recapitalized banks (both directly and United States through incentives for private investors), imple- Spain Slovenia 0.2 mented programs to purchase distressed bank assets, Greece Serbia 9 and provided guarantees for existing bank assets. Italy Albania Credit demand policy index Many countries increased the coverage of deposit Iceland Romania India Ukraine insurance to avoid deposit drains, which threatened Korea 0.1

to force banks to shrink their loan books. Lithuania CroatiaHungary United KingdomJapan Montenegro Denmark Macedonia Other policies fall into several broad categories: Australia • Monetary policies: Central banks have expanded their Russia Belgium Netherlands 0.0 monetary policy toolkits to enhance the demand 0 0.1 0.2 0.3 0.4 0.5 0.6 and supply of credit in addition to using tradi- Credit supply policy index tional tools such as changes in the policy rate. For Source: IMF staff estimates. example, the ECB’s “fixed-rate full allotment” policy Note: The indices are computed by dividing the number of policy measures currently in place to (in which banks’ bids for liquidity from the central support the supply of or demand for credit in each country by the total number of possible measures in the list of all policy measures in Appendix Table 2.1 (excluding “stress test,” bank are fully satisfied), as well as its long-term “coverage enhancement of deposit insurance,” “other policies to enhance credit supply,” and (three-year) refinancing operations, were aimed in “other policies to mitigate debt overhang”). EU-wide ﬁscal programs (e.g., through the European Investment Bank and the European Bank for Reconstruction and Development) are counted with part at supporting credit. Many central banks have half weights for the European Union member countries that do not have national ﬁscal programs. eased collateral constraints for banks, in part by accepting a wide range of private assets. Some have adopted policies of direct credit easing through Some countries have implicitly or explicitly allowed purchases of corporate bonds, mortgage bonds, and forbearance on recognition of nonperforming loans. other private sector assets. A few central banks have • Capital market measures: To promote the diversification engaged in indirect credit easing by making available of financing options for firms, several governments special lending facilities to promote bank lending. have made efforts to lower barriers to corporate bond • Fiscal programs: Many national treasuries have sought issuance for SMEs and to promote securitization mar- to promote expansion of corporate and mortgage kets for SME loans and household debt (Box 2.1). loans through direct extension of loans and through Most countries have relied on a variety of policies to subsidies or guarantee programs for new loans. These support both credit demand and credit supply, recog- programs have often been implemented through nizing that these are often complementary. Figure 2.7 state-owned or state-sponsored institutions. and Table 2.3 list the various credit-supporting policies • Financial regulations: Prudential regulators have implemented in 42 countries. The policies are limited instituted measures designed to ease bank balance to those directly targeting credit market constraints and sheet restrictions that have made banks unwilling do not include more general fiscal and monetary policies or unable to extend new loans. In some countries (including quantitative easing—that is, direct purchases (particularly in the European Union), regulators of government bonds) that have also underpinned credit have relaxed capital requirements for loans to SMEs. activity. In addition, the indices in Figure 2.7 refer only to the number of different measures currently in place; they do not account for the size of the programs or their 9See further discussions on restructuring programs in Landier and Ueda (2009) for banks, Laeven and Laryea (2009) for households, effectiveness. Despite this somewhat narrow scope, the and Laryea (2010) for firms. data yield the following main conclusions:

International Monetary Fund | October 2013 69

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 2.1. Policies to Diversify Credit Options for Small and Medium Enterprises in Europe

This box explores options for diversifying credit creation stock of bank loans to SMEs, which is estimated for small and medium enterprises (SMEs), which have to be approximately €1.5 trillion. traditionally been constrained in their credit channels. o Addressing the asymmetric treatment of securitized Options for access to credit are much more restricted assets vis-à-vis other assets with similar risk char- for SMEs than for larger firms. Larger companies have acteristics: Currently, securitized assets are often benefited from historically low costs of funding and treated less favorably by investors and central ample liquidity through a variety of credit channels. banks. For example, the haircut imposed by the Conversely, SMEs have virtually no access to bond ECB on asset-backed securities is 16 percent, markets and continue to face higher interest rates and much more than on other assets of similar risk— restricted access to bank credit. Although the availability such as covered bonds with a similar rating—that and conditions of external financing appear to have are also accepted in liquidity facilities and direct improved in the last year or so—including for bank purchases. Aside from the differences in the legal loans, bank overdrafts, and trade credit—these improve- frameworks governing securitized assets and ments have been less obvious for SMEs than for larger covered bonds, there are important inconsisten- companies. In a recent survey by the European Central cies in capital charges that provide incentives for Bank, for example, “access to finance” was the second covered bond issuance and bank cross-holdings most important concern mentioned by SMEs, on aver- of covered bonds, at the expense of securitiza- age, throughout the euro area, although the magnitude tions with the same credit rating and duration of the concern differed by country—38 percent of risk (Jones and others, forthcoming). SMEs in Greece reported this as their biggest concern, o Introducing government guarantees for SME 25 percent in Spain, and 24 percent in Ireland, while securitizations (covering credit and sovereign risk): only 8 percent of SMEs in Germany and Austria viewed Guarantees could encourage private investment access to finance as a primary issue (ECB, 2013). in these securities by offsetting some of the infor- SMEs were also hit harder by the crisis. There is mational asymmetries and SME credit risk, espe- evidence (Iyer and others, 2013) that the magnitude of cially from investors that can only buy securities the reduction in credit supply was significantly higher with certain minimum credit ratings. The effect for firms that (1) are smaller (as measured by both on lender incentives and the fiscal cost of these total assets and number of employees); (2) are younger guarantees should be appropriately recognized (as measured by the age of incorporation); and (3) (see the main text). have weaker banking relationships (as measured by the o Including SME loans in the collateral pool for cov- volume of their bank credit before the crisis). Regu- ered bonds: Currently, only mortgage, municipal, lation may also play a role. Some studies (OECD, ship, and aircraft loans are eligible collateral for 2012; Angelkort and Stuwe, 2011) suggest that Basel covered bond issuance; extending eligibility to III implementation could lead banks to reduce their SME loans will improve their attractiveness. lending to SMEs. This problem is likely to be larger in o Improving risk evaluation for SME securities by countries with bank-based financial systems and less- regulating and standardizing information disclo- developed financial markets. sure: More uniform information disclosure would Improving the availability of credit to the corporate reduce investors’ uncertainty about the quality of sector in general, and SMEs in particular, is essential SME securities and thus would tend to reduce to supporting the economic recovery. The following SMEs’ cost of bond and commercial paper policy measures may help achieve this goal. issuance. • Advancing the securitization agenda, including by: • Encouraging development of factoring of SME receiv- o Developing primary and secondary markets for ables: By facilitating the sale of account receivables, securitization of SME loans: Of the total euro area SMEs can finance working capital. If this form securitized bond market of €1 trillion at the end of financing is underdeveloped, then better credit of 2012, only some €140 billion was backed by information and quality of credit bureau data will SME loans. This contrasts with the much larger improve assessment of borrowers’ ability to pay. • Encouraging companies to lend to each other: Larger The authors of this box are David Grigorian, Peter Lindner, companies could provide financing to their smaller and Samar Maziad. suppliers (for example, via faster payment cycles).

70 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Box 2.1 (continued)

• Paving the way (including through appropriate regula- • Facilitating establishment of “direct lending” funds tion) for market-based credit guarantee programs and targeting SMEs that have difficulty getting other the development of small-bond markets: Government- types of financing: These funds could include direct backed partial credit guarantee and mutual guarantee financing by distressed-debt firms, private equity programs (similar to microfinance) could support firms, venture capital firms, hedge funds, and busi- expanded credit to SMEs (Honohan, 2010; Columba, ness development corporations. Gambacorta, and Mistrulli, 2010). Italy’s introduction of fiscal incentives for the issuance of minibonds by The relative effectiveness of these policies in unlisted firms in 2012 provides an example. providing credit to SMEs and their attendant costs • Tax incentives for banks that expand credit to SMEs: would need to be evaluated on a country-by-country These incentives could take the form of lower tax basis. The authorities should ensure that these rates on earnings from SME lending. However, any measures are sufficiently targeted to address the root tax subsidies should be carefully designed so as not to causes of lack of credit to SMEs. They must also encourage excessive risk taking by banks or weaken minimize moral hazard and financial stability risk loan underwriting standards, or create opportunities by ensuring adequate risk management practices are for tax avoidance, which will be very hard to reverse in place and requiring banks to hold a portion of later. Also in this case, the effect on lender incen- securitized SME-backed assets on their balance sheets tives and the fiscal cost of these guarantees should be to be sure they have a sufficient financial interest in appropriately and transparently recognized. monitoring the loans.

• Figure 2.7 suggests that some countries have cho- This chapter takes a stepwise approach to identify- sen to target only one side of the market, usually ing underlying constraints affecting credit markets. As focusing more on policies to boost credit supply. a first step to target policies, it proposes to distinguish However, countries that have not used targeted between demand and supply constraints, which can be demand-side policies—including the core euro area useful to narrow the policy options that may be effec- and the Nordic countries—have still relied to a con- tive. Moreover, if the sensitivity of supply or demand to siderable extent on more general fiscal and monetary interest rates can be determined, policymakers may be policies to support credit demand. able to discern which policies are likely to be most effec- • Emerging market economies in central and eastern tive in increasing credit volume. In a more challenging Europe have implemented relatively fewer policies second step, the chapter attempts to identify the specific to support credit, perhaps because some have less factors that may constrain credit demand or supply. In monetary and fiscal policy room. Some institutions countries for which sufficient data are available for this (including the European Investment Bank and the second step, results from such an analysis could further European Bank for Reconstruction and Develop- narrow the set of credit-supporting policies that are ment) are providing support for credit supply poli- likely to be most effective. Last, the chapter uses other cies in several of these countries. information gleaned from country-specific sources to add to the overall assessment. The analytical results should be interpreted with Are Current Policies on Target? caution. The factors that determine credit supply and Given limited policy resources, policymakers should demand are technically difficult to identify. The analy- target the constraints on the demand or the supply sis is further complicated by a lack of appropriate data, of credit that can be effectively addressed by govern- even in the advanced economies considered here. Still, ment intervention. To facilitate the usefulness, timing, this exercise provides a useful framework for assess- and sequencing of the various policies, it is helpful to ing the appropriate targeting of policies and offers a identify the factors that underlie credit demand and tentative and preliminary assessment of their effective- credit supply. Depending on how these factors influ- ness for countries where sufficient data were available. ence lending activity, one or more could be the target Further refinement of this framework would be useful, of government policies. and would greatly be facilitated by the availability of

International Monetary Fund | October 2013 71

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 2.3. Credit Policies Implemented since 2007 Enhancing Credit Supply Supporting Credit Demand Fiscal Supportive Monetary Programs on Financial Capital Market Bank Corporate Debt Household Debt Policy1 Credit Regulation2 Measures Restructuring3 Restructuring Restructuring Euro Area Austria Y Y Belgium Y Y Y Y Estonia Y Y Y Finland Y France Y Y Y Germany Y Y Y Greece Y Y Y Y Y Ireland Y Y Y Y Y Italy Y Y Y Y Y Y Y Netherlands Y Y Y Y Portugal Y Y Y Y Y Slovak Republic Y Slovenia Y Y Y Y Y Y Spain Y Y Y Y Y Y Other Advanced Europe Denmark Y Y Y Iceland Y Y Y Y Norway Y Y Sweden Y United Kingdom Y Y Y Y Y Non-European Countries Australia Y India Y Y Y Y Y Y Japan Y Y Y Y Y Y Korea Y Y Y Y Y Y Y South Africa United States Y Y Y Y Y Y Non-Euro-Area Central, Eastern, and Southeastern Europe Albania Y Y Y Bosnia and Herzegovina Y Bulgaria Y Croatia Y Y Y Y Czech Republic Hungary Y Y Y Latvia Y Y Y Lithuania Y Y FYR Macedonia Y Y Y Moldova Y Y Y Montenegro Y Poland Y Romania Y Y Y Russia Y Y Y Y Serbia Y Y Y Y Y Turkey Ukraine Y Y Y Y Source: IMF staff. Note: This table lists the various types of policies countries have implemented since 2007, based on Appendix Table 2.1, without consideration of the scope, duration, or effectiveness of those policies. “Stress test” and “coverage enhancement of deposit insurance” are excluded from the policies supporting credit demand. EU-wide fiscal programs (e.g., through the European Investment Bank and the European Bank for Reconstruction and Development) are not included although they are available for firms in the EU member countries (and in some non-EU European countries). 1Monetary policy measures that may ease constraints to credit supply, such as direct and indirect credit easing as well as widening of collateral eligibility for private sector assets (see also Appendix Table 2.1). 2Measures include a reduction in risk weights for small and medium enterprise loans when calculating banks’ capital adequacy ratios, forbearance of nonperforming loans, and countercyclical macroprudential regulations. In the United Kingdom, the authorities have recently relaxed liquidity requirements for banks. 3This category includes ad hoc public assistance to banks that may not have been initiated to counter undercapitalization (in or out of crisis situations) but were intended to improve credit supply. For India, the “Y” includes an ongoing government contribution to the equity capital of banks that is a consequence of the partial government ownership of banks, for which the relevant statute does not allow their ownership stake to go below 51 percent. Such contributions are a regular feature of the Indian banking system.

72 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

expanded and more detailed data (beyond the imper- Table 2.4. Determinants of Credit Growth fect proxies that are used in this analysis) that could Euro Area Euro Area United States Corporate Mortgage Commercial and more clearly identify the constraints to credit demand Loans Loans Industrial Loans and supply. Credit Growth (t – 1) 0.511*** 0.331** 0.628*** (0.134) (0.138) (0.112) ΣDemand Index (t – i) 0.030** 0.014** 0.009 (0.013) (0.007) (0.125) Disentangling Credit Supply and Demand ΣPure Supply Index (t – i) –0.040** –0.052** –0.126** (0.011) (0.021) (0.062) Data from bank lending surveys can help distinguish Source: IMF staff estimates. Note: Regressions include a lag of the dependent variable and four lags of the between demand and supply factors that underlie credit demand indicator and the “pure” supply indicator (see Annex 2.2) as well as developments. Identifying supply and demand shocks seasonal dummies. For the euro area, Arellano and Bond (1991) regressions with robust standard errors are in parentheses. The euro area estimation covers typically requires an exogenous source of demand 2003:Q1–2013:Q1 and includes Austria, France, Germany, Italy, Luxembourg, the and supply variation (Ashcraft, 2005), an exogenous Netherlands, Portugal, and Spain. For the United States, an ordinary least squares regression is estimated for the period 1999:Q1–2013:Q1. ** and *** denote instrument (Peek and Rosengren, 2000), or matched significance at the 5 and 1 percent levels, respectively. borrower-bank data (Jiménez, Ongena, Peydró, and Saurina, 2012). In the absence of such data, the analysis banks reporting in the survey that they observed an here relies on answers to bank lending surveys con- increase in demand for loans minus the fraction that ducted by central banks in the euro area and the United observed a decrease. Supply factors are proxied by a States.10 For these surveys, bank loan officers are asked measure of lending standards from which the influence for their views about the various factors affecting credit of factors that are not related to bank balance sheets is demand and credit supply using questions on credit statistically removed. These factors should be removed demand conditions and changes in lending standards. because lending standards reported in surveys may not Although the survey responses are qualitative (for reflect “pure” shifts in credit supply but instead may example, credit is assessed as having “tightened consider- respond to changes in factors such as borrowers’ credit ably or somewhat,” “eased considerably or somewhat,” worthiness, the economic outlook, and uncertainty, or “no change”), they can be assigned a numerical value which also affect loan demand conditions. After cleans- to obtain a quantitative index. The approach in this ing the raw data to arrive at a better measure of “pure” chapter assumes that the responses from loan officers in supply factors, credit growth can be decomposed into the bank lending surveys are good proxies for unob- demand and supply influences. These influences are served demand and supply.11 computed using the estimated coefficients from a The approach determines how much credit growth regression of credit growth on the demand index and 12 can be attributed to demand or supply factors (Annex the adjusted lending standards (Table 2.4). 2.2). Demand factors are proxied by the fraction of The results of this decomposition show that both demand and supply factors are important in explaining credit developments in both the euro area and the 10 In the euro area, the ECB conducts the quarterly Bank Lending United States but that their relative influence varies Survey (www.ecb.europa.eu/stats/money/surveys/lend/html/index. en.html), and in the United States, the Federal Reserve conducts over time. the quarterly Senior Loan Officer Opinion Survey on Bank Lending • Corporate credit (Figure 2.8): Demand factors had a Practices (www.federalreserve.gov/boarddocs/snloansurvey). Data negative effect in late 2009 in Austria, France, the series that are long enough for this analysis are available for Austria, France, Germany, Italy, Luxembourg, the Netherlands, Portugal, Netherlands, and Spain. Most countries saw deterio- Spain, and the United States. The surveys include questions such as, rating demand conditions in the most recent period, “How has the demand for loans changed at your bank over the past including Germany, where demand conditions had three months?” and “How have your bank’s credit standards changed been relatively favorable since the start of the crisis. over the past three months?” 11Although this analysis provides useful insight, it still suffers Supply factors have had a negative effect throughout from potential bias. For example, reporting bias is a concern: the period in most countries (with particularly strong surveyed banks may try to please their supervisors and fail to report negative effects in Portugal), but eased in most euro true credit supply conditions. Despite this problem, an emerging literature makes use of survey data to shed light on the determinants of credit growth, and there is evidence that it contains useful 12Unfortunately, the reasons provided in the survey as explana- information. For example, Lown and Morgan (2006) and De Bondt tions for changes in demand do not allow for a straightforward and others (2010) show that the surveys have predictive power for classification between supply and demand factors as is the case for output and credit growth in the United States and in the euro area, the supply questions and hence cannot be used to perform the same respectively. technique to “cleanse” the data as done for the supply side.

International Monetary Fund | October 2013 73

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 2.8. Decomposing Credit Growth: Corporate Loans area countries in the first half of 2012, likely as a

Demand component Supply component result of the long-term refinancing operations of the 1. Austria 2. France ECB. More recently, demand constraints appear to 0.5 2 outweigh supply constraints in France. 1 0.0 • Mortgage credit13 (Figure 2.9): The negative effect of 0 –0.5 demand factors in 2009 and 2010 on mortgage credit –1 –1.0 in a number of countries was more moderate than on –2 –1.5 corporate loans, and demand recovered in 2011 and –3 2012 before turning down again more recently (except –2.0 –4 in Austria and Germany). Most countries saw a double- –2.5 –5 2008 09 10 11 12 13 2008 09 10 11 12 13 dip in supply constraints, with a temporary relaxation 3. Germany 4. Italy around 2010. However, most recently (and in contrast 2 2 to developments for corporate loans), supply constraints

1 for mortgage loans eased in 2013 in a number of coun- 1 tries, most markedly in France, Italy, and Portugal. 0

0 –1 Identifying Factors Constraining Credit This section offers a more detailed set of tools to identify –1 –2 2008 09 10 11 12 13 2008 09 10 11 12 13 the factors constraining credit by estimating the under- 5. Netherlands 6. Portugal lying determinants of credit demand and credit supply. 1 2 Two approaches are employed: (1) an estimation of the 1 0 country-specific structural determinants of bank credit 0 –1 supply and demand; and (2) a firm-level panel estima- –1 tion of factors that affect manufacturing firms’ borrow- –2 –2 ing. Both approaches focus on credit to firms.

–3 –3 Evidence from a structural model of bank lending –4 –4 2008 09 10 11 12 13 2008 09 10 11 12 13 This approach estimates supply and demand equations 7. Spain 8. United States for aggregate bank lending for major countries that 1 2 14 1 have had weak credit growth. The exercise has exten- 0 0 sive data requirements and presents challenging econo- –1 metric issues (Box 2.2). As a result, reliable results were –1 –2 obtained only for corporate loans in France, Japan, –3 Spain, and the United Kingdom.15 –2 –4 Because shifts in demand and supply cannot be –5 –3 –6 observed directly, the analysis uses “shifters” that are 2008 09 10 11 12 13 2008 09 10 11 12 13 meant to affect only one, but not the other, side of the market, thus allowing demand and supply to be Sources: European Central Bank, Bank Lending Survey; Federal Reserve, Senior Loan Ofﬁcer Survey; and IMF staff calculations. 13 Note: Demand and supply components are constructed using the estimates in Table 2.4. The The analysis of mortgage lending does not include the United demand component is the ﬁtted values constructed recursively using the lags for the demand States because of a break in the mortgage lending standards series in index and setting the "pure" supply index to zero. The supply component is constructed 2007 and because the Senior Loan Officer Survey does not include analogously. questions regarding the reasons for tightening or easing lending standards for mortgages. 14See Annex 2.3 for details of the model’s design. 15France and Japan were included in the estimation, although bank credit growth to the private sector (nonfinancial corporations and households alike) was not identified as weak according to Table 2.1. Still, bank credit in Japan was identified as weak until the third quarter of 2012, and bank credit to nonfinancial firms in France (ECB data) declined in the last quarters of 2012. In addition, both countries implemented credit-supporting policies.

74 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Figure 2.9. Decomposing Credit Growth: Mortgage Loans identified separately. This econometric technique is commonly used but is difficult to implement because it Demand component Supply component 1. Austria 2. France requires accurately identifying variables associated with 1 2 either demand or supply, but not with both. The vari- 1 ables chosen that affect only supply (thereby tracing 0 0 out and identifying the demand curve) include the cost of bank funding and basic balance sheet variables (the –1 16 –1 bank’s capital-to-asset ratio). On the demand side, –2 the variables include the rate of capacity utilization and 17 –2 –3 a proxy for the availability of market financing. 2008 09 10 11 12 13 2008 09 10 11 12 13 The supply and demand equations include several 3. Germany 4. Italy variables to capture more directly some of the market 1 2 constraints previously discussed. In particular, the 1 nonfinancial firms’ debt-to-equity ratio aims to capture the effect of debt overhang on credit demand (and 0 0 serves as an indicator of credit risk from the viewpoint of banks on the supply side). Although the growth of –1 the stock market index is correlated with the value of

–1 –2 firms’ collateral (a supply-side constraint), it may also 2008 09 10 11 12 13 2008 09 10 11 12 13 increase firms’ preference for equity financing (affect- 5. Netherlands 6. Portugal 1 2 ing credit demand). The presumed relationships and reasons for choosing the specific variables are discussed 1 0 in Annex 2.3. 0 –1 The estimated supply and demand equations for –1 bank credit are well identified overall. For all coun- –2 –2 tries, one or more of the demand and supply shifters –3 –3 is significant in the regression, identifying the demand

–4 –4 and supply equations for these countries (Table 2.5). 2008 09 10 11 12 13 2008 09 10 11 12 13 On the supply side, lower funding costs (proxied by 7. Spain 2 deposit rates) tend to increase the supply of bank loans. The amount of capital a bank holds relative to 1 its total assets yields a counterintuitive negative sign in France and Spain. These results should probably not 0 be given too much weight, because they may reflect an inaccurate proxy for bank capital, a scaling down of –1 lending by banks that are building up their capital buf- 18 –2 fers, or ongoing major bank restructuring in Spain. 2008 09 10 11 12 13 Additional results (see below) show a positive relationship between bank capital and lending by banks. On Sources: European Central Bank, Bank Lending Survey; and IMF staff calculations. Note: Demand and supply components are constructed using the estimates in Table 2.4. The the demand side, in most cases, capacity utilization has demand component is the ﬁtted values constructed recursively using the lags for the demand index and setting the "pure" supply index to zero. The supply component is constructed the expected positive effect on firms’ demand for loans, analogously. 16Unfortunately, a better proxy—regulatory capital—is not available. 17Although finding one shifter each for the supply and demand side is theoretically enough to identify the model empirically, the potential endogeneity of some shifters complicates proper identification. 18Despite the increase in system-level capitalization (including injection of public capital), lending continues to contract, which may reflect in part the deleveraging requirements imposed on banks that receive government aid.

International Monetary Fund | October 2013 75

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 2.2. Challenges in the Structural Estimation of Credit Supply and Demand

This box draws attention to some limitations related • Most variables in the analysis are more or less jointly to the estimation of a structural model of supply and determined. For instance, future GDP (and there- demand for bank lending, and discusses attempts to fore GDP forecasts) depend on the amount of credit overcome them. granted by banks today. To alleviate the resulting endogeneity, most regressors are lagged by one period. Data measurement issues • Potential endogeneity is a major challenge for find- Measurement issues affect both the dependent and the ing variables that can separately identify credit sup- explanatory variables and constrain the estimation of ply and demand (which the chapter calls “shifters”). the determinants of credit supply and demand. A number of criteria were used to decide whether • Because of a lack of data on new bank loans gross the model was properly identified: (1) at least one of repayments, the analysis uses as the dependent of the shifters in each equation is statistically signifi- variable net transaction flows or the changes in the cant at the 5 percent level, and the shifters on each stock of bank loans. This underestimates the actual side are jointly significant; and (2) the coefficients volume of new loans, because repayments will offset on the lending rates in both the supply and demand some new loans. equations are of the expected sign, so that the • Among the explanatory variables, bank-specific resulting supply curve has a positive slope and the variables, such as the capital-to-asset ratio, are demand curve has a downward slope. A Hausman derived from monetary and financial statistics test based on the comparison of the two-stage and usually provided by central banks. They do not cor- three-stage least squares estimators was further used respond to regulatory ratios and may not accurately to verify the exogeneity of shifters. capture the constraints weighing on banks’ ability to lend. Many variables were considered in the supply Potential structural breaks equation as alternatives or in addition to the capital With the exception of the United Kingdom, the ratio of banks, in particular the price-to-book ratio, sample period considered in the analysis covers both changes in the level of capital, the deposit-to-total- the precrisis and crisis periods, raising the question liabilities ratio (to capture the extent of funding of whether the relationships in the estimation have constraints), the ratio of nonperforming loans changed over time and are robust to changes in the to total loans (as a proxy for the quality of bank sample period. For example, assets), the stock market index for the financial sec- • Restricting the sample to the period before or after tor, and banks’ z-score. Few came out as statistically 2008 prevents a proper identification of the model significant to allow for a proper identification of the in most cases because of the resulting large reduc- demand curve. One reason for this lack of signifi- tion in the number of observations. The estima- cance could be heterogeneity of the banking sector, tion therefore assumes that the coefficients do not with weaker banks behaving very differently from change over the full sample period. Alternative stronger ones, masked by the averages. specifications (not reported) allowed some coefficients to change before and after September 2008 Identification challenges by including a dummy variable for the period after Endogeneity issues complicate the proper identification September 2008 and interaction terms between of the supply and demand equations. For example, that dummy and some variables in the model, such as the lending rate or the capital ratio of banks. In most cases, the coefficients on the interaction terms The author of this box is Frederic Lambert. were not statistically significant.

76 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Table 2.5. Structural Determinants of the Supply and Demand of Bank Lending to Firms in Selected Countries Expected Signs France Spain United Kingdom Japan Supply Equation Lending Rate + 2,082.0*** 5,962.4*** 7,296.1*** 2,957.2 GDP Forecasts + 462.5 1,993.3*** 2,534.1** 106.8 Standard Deviation of GDP Forecasts – –5,879.6 3,300.1 6,752.2 496.9 Inflation – 666.5 541.8 –587.7 511.8* Growth of Stock Market Index + –5,121.1 –1,753.6 –9,427.0 –3,309.6 Lagged NFCs’ Debt-to-Equity Ratio – –176.4*** –41.9 240.8* –3.9 Lagged NFCs’ Profitability + –444.4 –1,979.9*** 1,242.7 2,621.3** Corporate Spread (investment grade) – n.a. n.a. n.a. 68.1*** Constant 38,351.8*** 80,127.5*** –87,380.5** –12,031.7** Supply Shifters Deposit Rate – –16,850.2** –28,978.5*** –11,077.6** –6,314.8** Lagged Banks’ Capital Ratio + –2,183.3** –923.1** 642.9 604.1 Bank CDS Spread – n.a. n.a. 2.8 n.a. F Statistics for Supply Shifters 4.780 23.348 6.147 4.371 P Value 0.092 0.000 0.105 0.112 Demand Equation Lending Rate – –2,009.0 –2,012.1*** –228.1 –1,573.2 GDP Forecasts + 1,318.3 3,009.8*** 1,026.1 152.7 Standard Deviation of GDP Forecasts – –3,405.0 6,501.2* 8,024.9 514.1 Inflation + 1,613.5* 1,042.9** –2,251.7 491.2* Growth of Stock Market Index – –5,312.6 799.5 –11,785.1 –3,307.7* Lagged NFCs’ Debt-to-Equity Ratio – –207.0*** –48.4 195.6 –5.7 Lagged NFCs’ Profitability – –150.5 –805.8*** 475.1 975.2 Corporate Spread (investment grade) + n.a. n.a. n.a. 37.7*** Constant 19,447.3 30,449.0* –94,991.7** –7,645.0* Demand Shifters Lagged Capacity Utilization + 319.4* 233.4 866.5** 34.4* Market Financing (average over past year) – –1,539.3** –13,084.5*** –103.2 279.3** F Statistics for Demand Shifters 4.482 27.784 6.258 5.590 P Value 0.106 0.000 0.044 0.061 Number of Observations 122 122 53 117 Sample Period 2003:M2–2013:M3 2003:M2–2013:M3 2008:M8–2012:M12 2003:M5–2013:M1 Source: IMF staff estimates. Note: CDS = credit default swap; NFC = nonfinancial corporation; M = month; n.a. = not applicable. *, **, and *** denote significance at the 10 percent, 5 percent, and 1 percent levels, respectively. The dependent variable is the net flow of bank loans to NFCs. NFCs’ profitability is computed as the ratio of NFCs’ gross operating surplus to gross value added. NFCs’ market financing is the average ratio of NFCs’ debt in the form of securities to total debt over the past year.

while the availability of market financing has the oppo- debt-to-asset ratio corresponds to net borrowing; there- site effect, as expected. This analysis provides no strong fore, the determinants of the changes in the corporate evidence that firms’ high current debt or low profit- debt-to-asset ratio can shed light on the factors that ability is holding back the demand for credit, except constrain corporate credit. maybe in France and Spain.19 Similarly, in contrast to The analysis uses annual data for 1991–2012 to ongoing discussions in some policy circles, the disper- conduct firm-level panel regressions to explain changes sion of growth forecasts (a measure of uncertainty in the debt-to-asset ratio for the manufacturing sectors about future growth) does not appear to play a large in France, Italy, Japan, Spain, the United Kingdom, role for either the supply of or demand for bank loans and the United States.20 Explanatory variables are the in this analysis. following: • The firm’s own debt-to-asset ratio, to capture debt- Evidence from firm-level data overhang effects that would constrain the willingness Additional evidence on specific factors that constrain or ability of firms to take on additional debt. It also credit emerges from data on firm indebtedness. These reflects the riskiness of firms, which would make data allow for a richer analysis that takes into account banks less willing to lend to them; the different characteristics of individual firms. Fairly comprehensive firm-level data are available from corpo- 20Firm-level balance sheet data are from the IMF Research rate balance sheets of exchange-listed firms that show Department’s Corporate Vulnerability Utility, based on Thomson Reuters data. House price data are from the Organization for Eco- total debt as a share of total assets. The change in the nomic Cooperation and Development and national sources. Credit includes bank credit and other forms of credit. All explanatory 19However, the results from the firm-level regressions show stron- variables are lagged by one period to mitigate possible simultaneity ger results for firms’ current debt levels. problems.

International Monetary Fund | October 2013 77

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 2.6. Firm-Level Regressions of Changes in Debt-to-Asset Ratio for Manufacturing Firms France Italy Spain United Kingdom Japan United States Return on Assets (%) –0.058 –0.083** –0.113** 0.018 –0.057*** –0.020*** Debt-to-Asset Ratio (%) –0.357*** –0.303*** –0.313*** –0.395*** –0.234*** –0.371*** Average Banking Sector Liability-to-Asset Ratio (%) 0.031 –0.294*** –0.765*** 0.019 0.213*** –0.558*** Real Household Consumption Growth Rate (%) 0.314*** 0.167* 0.120 0.264*** –0.256*** 0.212*** House Price Index (2010 = 100) 0.001 0.004 0.072*** 0.016* 0.037*** –0.002 Observations 4,613 1,621 961 7,819 30,581 33,358 Number of Firms 393 146 74 693 1,929 2,739 F Statistic P Value 0.00 0.00 0.00 0.00 0.00 0.00 R Squared 0.17 0.15 0.17 0.20 0.12 0.18 Sources: IMF, International Financial Statistics and Research Department, Corporate Vulnerability Utility, based on Thomson Reuters data; national sources; Organization for Economic Cooperation and Development; and IMF staff estimates. Note: Firm-level panel estimation is conducted with firm-fixed effects for each country using 1991–2012 data for the manufacturing sector. The dependent variable is the change in the debt-to-asset ratio (%). The manufacturing sector is defined as Division D of the Standard Industrial Classification (SIC), and the banking sector is defined as SIC 2-digit codes 60 (banks) and 61 (credit institutions) as well as four-digit code 6712 (bank holding companies). The coverage of firms is incomplete in 2012. All the explanatory variables are lagged by one period. *, **, and *** denote significance at the 10 percent, 5 percent, and 1 percent levels, respectively, based on robust standard errors clustered at the firm level.

• The firm’s return on assets, to capture the ability of more debt. Finally, higher consumption growth is sup- firms to fund investment projects internally as well portive of credit growth in most countries, except in as their creditworthiness;21 Spain and Japan. • The average liability-to-asset ratio of the banking Figure 2.10 shows the importance of each factor in sector in each country, to capture banks’ balance explaining recent deviations of corporate credit growth sheet constraints to making additional loans (a from each country’s average during 1992–2013. Credit higher ratio implies a more leveraged bank); has been restricted by bank capital in Spain (and in • Real household consumption growth, to capture Italy most recently) and by debt overhang in Italy consumer demand, a major driver of economic and Spain. Tepid consumer demand has slowed credit growth; and growth in France and Italy and also in the United • Real house prices, as a proxy for the value of loan Kingdom and the United States at the beginning of collateral.22 the crisis. Low real estate prices have been an impor- The regression results show that the factors con- tant factor constraining credit in Japan. straining corporate credit growth vary by country, but higher corporate debt levels, lower bank capital, and collateral constraints can play a role (Table 2.6).23 Cor- Are Credit Policies on Target? Some Examples porate debt levels matter for credit to manufacturing The results from the analyses in the previous sec- firms in all countries investigated: firms with higher tions can be used to evaluate whether specific policies debt levels (an indication of possible debt overhang) implemented in countries with weak credit growth tend to take on less additional debt. Credit to firms in are effectively targeting the specific factors that con- Italy, Spain, and the United States is also affected by strain credit growth (Figure 2.11). The analysis using the liability-to-asset ratio in banks: higher ratios (cor- bank lending surveys provides a first indication of responding to higher leverage and lower bank equity) the relative importance of supply and demand fac- are associated with lower debt in firms, suggesting that tors. The structural model and the firm-level analysis weaker banks lend less to firms. In Japan, Spain, and identify the specific factors that may constrain credit the United Kingdom, the results suggest that higher and how their influence has changed over time. The collateral values make it easier for firms to take on estimated demand and supply equations shed light on the potential effectiveness of specific policies on credit 21A drawback of this approach is that it does not distinguish volume, which depends on the relative sensitivity of between supply and demand. Here it is assumed that low profit- demand and supply to changes in the lending rate. For ability means firms would demand more external financing through example, if credit demand appears relatively insensitive loans. However, persistent low profitability may also cause banks to see the firm as less creditworthy, restricting supply. This latter effect to changes in the interest rate (its coefficient is close to is, however, partially absorbed by firm-fixed effects. zero or not significantly different from zero), govern- 22 The land price index is used for Japan. ment measures aiming to increase the supply of loans 23The sample includes only exchange-listed firms, which may bias downward the role of some constraints for firms with less easy access would lower the lending rate but would likely not lead to finance, such as SMEs. to a substantial increase in the demand. If the objective

78 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Figure 2.10. Decomposition of Change in Debt-to-Asset Figure 2.11. Real Total Credit Growth, by Borrowing Sector Ratios for Firms (Percent; year over year) France Italy Spain United States Japan Real estate price Bank capital Indebtedness United Kingdom Proﬁt Consumer demand Nonﬁnancial Corporations 1. 2. 1. France 2. Italy 25 25 1.0 2.0 1.5 20 20 0.5 1.0 15 15 0.5 10 10 0.0 0.0 –0.5 5 5 –1.0 0 0 –0.5 –1.5 –2.0 –5 –5 –1.0 –2.5 –10 –10 2009 10 11 12 13 2009 10 11 12 13 –15 –15 3. Japan 4. Spain 2004 06 08 10 12 2004 06 08 10 12 3.0 2.0 2.0 Households 1.5 1.0 3. 4. 1.0 20 20 0.0 0.5 –1.0 15 15 0.0 –2.0 –0.5 10 10 –1.0 –3.0 –1.5 –4.0 5 5 –2.0 –5.0 2009 10 11 12 13 2009 10 11 12 13 0 0 5. United Kingdom 6. United States 2.0 2.0 –5 –5 1.5 1.5 –10 –10 1.0 1.0 2004 06 08 10 12 2004 06 08 10 12 0.5 0.5 0.0 0.0 –0.5 Sources: Bank for International Settlements; and IMF staff estimates. Note: Total credit includes private sector borrowing (loans and debt instruments) from –1.0 –0.5 domestic banks and all other sources (“other credit”), such as other domestic nonbanks and –1.5 –1.0 foreign lenders (see BIS, 2013). –2.0 –1.5 2009 10 11 12 13 2009 10 11 12 13

Sources: IMF, International Financial Statistics, and Research Department, Corporate Vulnerability Utility, based on Thomson Reuters data; national sources; Organization for Economic Cooperation and Development; and IMF staff estimates. Note: The components add up to the deviation of the predicted change in the debt-to-asset ratio in each year from the average change in the debt-to-asset ratio over the period 1992–2013. A positive (negative) value means that the factor contributes to a positive (negative) change in the debt-to-asset ratio. Light colors indicate insigniﬁcant factors.

of policy is to increase the volume of lending, measures so some of the assessment is based on the previous that address demand-side frictions—corporate debt analyses of others (including from within the IMF and overhang, for example—would be more effective. outside). Clearly, the empirical work would benefit A preliminary assessment of policies for the major from further refinement, including by using more countries follows. This assessment is preliminary detailed data that could more effectively identify the because policies take some time to make an impact, constraints to credit, but it was not available for the and a number of policies have been implemented only research in this chapter. For a more explicit analysis of relatively recently. In addition, as indicated previously, funding costs in several European countries and their the technical analysis contains various weaknesses, potential effect on lending, see Chapter 1.

International Monetary Fund | October 2013 79

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

France also benefited from the ECB’s policies supporting credit For France, the results from the bank lending survey, supply. Bank lending survey results point to a large role the firm-level analysis, and the credit model show for bank balance sheet constraints in the tightening of a substantial negative effect from demand factors. lending standards at the beginning of 2012 and again Supply factors appear to play a lesser role, perhaps in more recently. The firm-level analysis confirms that low part because of the extensive supply-oriented poli- bank capital has played an important role most recently. cies that were implemented. The French government It also shows that debt overhang in firms may also helped ease credit supply by setting up state-sponsored play a role in restricting credit to firms. Other authors agencies to undertake refinancing operations and have confirmed the importance of bank capitalization, recapitalize banks. As a euro area member, France including Del Giovane, Eramo, and Nobili (2011), also benefited from the ECB’s efforts to support the who use confidential bank-level data in their analysis. supply of credit (including the widening of col- Albertazzi and Marchetti (2010) present evidence, based lateral eligibility). The firm-level analysis identifies on bank-firm matched data, that low bank capitaliza- weak consumption growth as a major factor in weak tion and scarce liquidity dampened lending following credit. This relationship likely reflects the strong role the collapse of Lehman Brothers. Also, Zoli (2013) that household consumption has played in sustaining finds that funding costs of banks with lower capital growth in the precrisis period, and the adverse impact ratios are more sensitive to changes in sovereign spreads. of uncertainty and rising unemployment on consump- These various analyses would suggest that measures tion in the latter period. By contrast, debt overhang that encourage banks to increase their capital would be in households does not appear to be an impediment useful. In particular, further provisioning and write-offs to consumption and credit growth, as discussed in the could be encouraged by increasing tax deductibility of 2013 IMF Article IV Staff Report for France (IMF, loan loss provisions and by expediting judicial process of 2013c). Therefore, further policy actions, if needed, corporate and household debt restructuring. could usefully focus on creating conditions for stronger Spain growth and employment, rather than on boosting credit directly. Debt overhang in banks, firms, and households is the key factor constraining credit volume in Spain. The Italy bank lending survey shows that Spain saw a substantial The Italian government has adopted a wide range of tightening of credit supply in 2009. The firm-level policies, particularly to ease the corporate debt over- analysis suggests that this tightening was in part due to hang and help households adjust during a period of constraints on bank capitalization. The decomposition large fiscal consolidation, but the most important factor of interest rates in Chapter 1 (see Figure 1.50) also restraining credit currently appears to be the capital suggests that the financial position of banks and sov- position of banks. On the demand side, corporate and ereign stress have contributed to higher interest rates personal bankruptcy laws were amended to speed up (and therefore lower loan volumes). Corporate debt restructuring procedures. A temporary moratorium on overhang also played a role, restricting credit demand. debt-service payments was implemented for both corpo- Jiménez and others (2012) underline the importance of rate and household debt, although this action may have supply constraints for Spain using bank-firm matched created other distortions because banks did not have to loan-level data and provide evidence that banks’ capital classify these loans as nonperforming. To address supply and liquidity ratios matter for their ability to extend constraints, the Italian government provided guarantees loans to firms. To ease these constraints, the govern- for corporate and mortgage loans and launched an ini- ment has helped guide a major restructuring of the tiative to promote the development of a corporate bond banking sector, including through a significant recapi- market. Some measures were taken in 2009 to support talization program (see IMF, 2013e and 2013f). Also, the recapitalization of the banking sector and one bank Spanish state- sponsored institutions have been provid- received additional support this year.24 Finally, Italy has ing direct loans to firms and guarantees for corporate

24While direct capital injections were not undertaken to a large government. These bonds are used as regulatory capital with special extent, the Italian government encouraged the issuance of spe- terms that allow banks to forgo the payment of interest if they are cial bank bonds (Tremonti bonds), which were purchased by the unprofitable.

80 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

loans. In addition, the government has been taking funding problems (through the Credit Guarantee steps to promote SME bond and equity financing and Scheme and Asset Protection Scheme). The Bank to address debt overhang in firms and households, of England and the Treasury jointly implemented a including through resolution programs for heavily Funding for Lending Scheme in mid-2012 (expanded indebted households and amendments to bankruptcy in April 2013) to lower funding costs and to provide rules. In view of the analysis in this chapter, further incentives for new lending. Although these measures useful steps to ease credit constraints could include appear to have helped ease funding conditions and additional reforms to ensure efficient and timely some lending rates, it is less clear that credit vol- resolution of corporate and household debt (see IMF, umes have increased as a result. This in part reflects 2013g), as well as reforms to further ease bank funding still-ongoing deleveraging by major banks with weak costs, such as additional steps toward a full banking asset quality or insufficient capital buffers. However, union (see the discussion in Chapter 1). preliminary econometric results in this chapter suggest that the demand for additional loans is relatively Japan insensitive to changes in lending rates. If this were The firm-level analysis suggests that declining collateral to be confirmed through additional, more detailed values have been a particular constraint to credit interme- analysis (including over a longer time period), then diation in Japan. Policies in Japan since 2008 have largely policies that support credit demand may be more focused on credit support measures to SMEs, including effective in boosting credit volumes.25 public credit guarantees and credit subsidies and direct credit provision by public financial institutions. Many United States of these measures had already been put in place in the The constraints that the U.S. corporate loan market early 2000s when Japan experienced a slowdown and a witnessed in the early stages of the crisis appear to banking crisis. As noted in Japan’s 2012 Financial Sector have dissipated. The analysis of lending surveys shows Assessment Program Update (IMF, 2012b), although that the United States saw a substantial tightening these credit policies have largely sheltered incumbent of corporate lending standards as a result of credit firms from a tightening of financing conditions and have supply constraints and the weaker economic outlook prevented widespread SME bankruptcies, reliance on in 2008 and 2009. Both supply and demand factors public credit guarantees in SME lending tends to weaken have improved since then, and total credit growth to banks’ incentives to undertake rigorous credit assessments nonfinancial corporations has turned positive. The and reduces incentives for restructuring, and entails fiscal improving housing market may improve access to costs that may begin to outweigh benefits. In addition to finance for SMEs given that they often use housing as the measures specifically geared toward SMEs, the Bank collateral (IMF, 2013i). Large purchases of mortgage- of Japan also established several lending facilities at low backed securities by the Federal Reserve, combined interest rates to encourage bank lending and lending with mortgage securitization through government- toward growth sectors. Further measures would be useful, sponsored enterprises, have helped alleviate supply- including (1) phasing out the full-value credit guarantees; side constraints in the mortgage market (Box 2.3). (2) increasing the availability of risk capital for start-ups However, the still-negative growth rate of credit to through asset-based lending; and (3) implementing a households (driven by housing debt) may call for structural reform of lending practices based on fixed-asset further measures. Some demand-side policies have collateral. been implemented: to ease household debt overhang, loan modification programs were introduced in 2009, United Kingdom and subsidies and tax incentives were provided to The U.K. authorities adopted a number of measures to boost credit, but their effectiveness has yet to be 25 demonstrated. This could be due to the relatively Credit supply and demand equations for the United Kingdom were estimated for the post-2008 period only. Empirical analysis by short period during which they have been in place. Aiyar, Calomiris, and Wieladek (2012) on the precrisis period with The Bank of England widened collateral eligibility confidential bank-by-bank data finds that the lending behavior of and purchased limited amounts of corporate bonds banks was sensitive to changes in capital requirements. The 2013 IMF Article IV Staff Report for the United Kingdom (IMF, 2013h) and commercial paper. The Treasury provided tem- also suggests the need for strengthening banks’ balance sheets and porary guarantees for bank assets to mitigate banks’ capital buffers as a prerequisite for a durable credit recovery.

International Monetary Fund | October 2013 81

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 2.3. The Effect of the Liquidity Crisis on Mortgage Lending

This box examines the credit supply impact resulting from the exposure of U.S. banks to market liquidity risk through Figure 2.3.1. U.S. Banks’ Core wholesale funding, based on Dagher and Kazimov (2012). Deposits-to-Assets Ratio

In the two decades leading up to the global financial Small banks 0.8 crisis, U.S. banks reduced their reliance on traditional retail deposits, as shown by a drop in their average ratio of core deposits to assets (Figure 2.3.1).1 Banks 0.7 Medium banks have increased their flexibility by moving away from traditional deposits and into market (or “wholesale”) 0.6 funding, but they are now more vulnerable to swings Large banks in market funding, as became apparent when wholesale funding liquidity dried up in the third quarter of 0.5 2007. The empirical literature on this topic provides 1986 90 94 98 2002 1006 evidence that banks that relied more on short-term wholesale funding reduced their lending more during Sources: Federal Financial Institutions Examination Council, the crisis than other banks. However, this literature Consolidated Reports of Condition and Income; and IMF staff has relied only on aggregate data, which makes the estimates. Note: Computed as the ratio of demand deposits plus fully insured task of disentangling demand and supply effects very time deposits to total assets. Small, medium, and large banks are challenging. designated according to total assets for lower third, middle third, Dagher and Kazimov (2012) make use of loan-level and top third, respectively. data on mortgage applications available through the Home Mortgage Disclosure Act, combined with bank 2 financial data from the Reports of Condition and less during the crisis. The analysis also shows that the Income collected by the Federal Deposit Insurance relative reduction in credit by wholesale-funded banks Corporation. The data allow for an analysis of banks’ was more severe for so-called jumbo loans, which decisions to reject loan applications while controlling cannot be sold to government-sponsored enterprises for a host of applicant, bank, and geographical charac- (GSEs). This suggests that the reduction in lending teristics. Bank characteristics include the ratio of core was likely associated with liquidity challenges in banks. deposits to assets, size, liquidity, leverage, and banks’ Indeed, the regressions indicate that banks that relied reliance on securitization. By focusing on a homoge- more on securitization through GSEs continued to neous category of credit and studying bank decisions lend more because such securitization offered a stable rather than the volume of credit, this approach greatly source of liquidity for mortgage financing for banks. reduces the potential for demand factors to confound Therefore, the results indirectly suggest that the the supply effects. The regression compares the effect Federal Reserve’s purchases of mortgage-backed securi- of bank characteristics on the decision to reject a loan ties, to the extent that they contributed to improving in 2008 with the crisis year (2007) and with the pre- the liquidity of mortgage loans, helped ease supply crisis years 2005 and 2006. constraints in mortgage lending. The results show that banks with a higher reliance on core deposits in 2007 increased their rejection rate

The author of this box is Jihad Dagher. 1The core deposit ratio is a commonly used measure of the extent to which banks rely on traditional insured deposits as 2Specifically, a 1 standard deviation (14 percentage point) a source of funding. It is computed as the ratio of transaction increase in the core-deposit-to-asset ratio is associated with a 3.7 deposits plus fully insured time deposits to total assets. percentage point relative decrease in the rejection rate.

82 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

encourage banks to restructure debt instead of pursu- negatively affect bank balance sheets. Hence, to restart ing foreclosure. credit, the restructuring of this debt must go hand in hand with more general repair of banks’ balance Other countries sheets. Sometimes credit policies can be reinforcing. For Data limitations and econometric challenges prevented example, policies to boost aggregate demand may be a similar analysis in this chapter for other countries, expected to boost the demand for credit, but the result- but the general analytical framework can be used else- ing improved economic outlook may also strengthen where. The use of better data (including supervisory banks’ balance sheets and relax credit supply constraints. data connecting individual banks to borrowers) could Sequencing is also important: policies to ease credit reveal the factors underlying weak credit developments supply constraints may be appropriate initially, but on a country-by-country basis and pinpoint the poli- once they take hold, credit demand may become the cies that would most effectively revive credit activity.26 constraining factor and additional policy measures may In most cases, measures to stimulate loan demand and be necessary to boost credit demand. Finally, policymak- loan supply will both work; however, their respective ers should attempt to determine whether constraints effectiveness will depend on the relative sensitivity of are temporary or require a more permanent form of credit demand and supply to changes in interest rates intervention. Most obviously, emergency measures and on the other factors that underlie these curves. implemented in times of crisis to counter acute market distortions may not be warranted during more tranquil times and should be only temporary. Designing Effective Policies for Reviving Credit Credit policies can usefully underpin financial stabil- Markets ity by preventing a deeper downturn than otherwise Appropriate policies to boost credit activity differ by and by sustaining an economic recovery, but as with country. The analysis shows that the causes of slow the use of unconventional monetary policy, policymak- credit growth differ by country, even for countries that ers should also recognize the limitations of credit poli- are closely linked (as in the euro area), and may be cies. Most policies will be effective only to the extent connected to specific factors that affect the demand that they can target underlying constraints to credit for credit (the profitability of firms, their capacity demand or supply. Ill-targeted measures may have utilization, or debt overhang), or to “pure” credit sup- adverse or conflicting effects. For example, the direct ply factors (the cost of funds for banks or the level of provision of credit by government-sponsored institu- bank capital), or to both (GDP growth or economic tions can lead to a suboptimal allocation of capital uncertainty). The set of policies that are likely to be and significant credit risk if loans are awarded on a effective will differ too and should be identified using a noncommercial basis. Also, for countries in which the thorough analysis of the underlying constraints in the deleveraging process in banks is seen as an essential ele- particular country. Such policies may also target sectors ment for bringing the financial sector back to health, that face particular credit challenges, such as SMEs (see policymakers may need to accept a period of slower Box 2.4 for policies in Korea). In that context, it may credit growth or a decline in credit. Finally, because be particularly helpful to promote diversification away policies take time to have an impact, there should be from bank credit to increase the options for finance no rush to judgment as to their effectiveness and the (see Box 2.1). Evidence from previous crises also indi- need for additional measures. cates that swift and comprehensive policy action leads The potential effectiveness of policies in the near to better outcomes (as in the Nordic countries in the term should be balanced with potential risks to early 1990s; see Box 2.5). financial stability in the longer run. If multiple policies In many cases, demand- and supply-oriented policies to enhance credit would be effective, relatively more are complementary, but the relative magnitude and effort should be placed on those policies likely to have sequencing of those policies is important. For example, the least detrimental effect on medium-term financial the restructuring of household and corporate debt may stability. Risks fall into several broad categories: • Credit risk: Policymakers should keep in mind that 26Such data are typically confidential and were not available for some policies, while potentially effective in sup- the analysis in this chapter. porting credit, may provide adverse incentives that

International Monetary Fund | October 2013 83

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 2.4. Policy Measures to Finance Small and Medium Enterprises during Crises: The Case of Korea

This box demonstrates how Korean authorities responded Figure 2.4.2. Financial Support Programs to crisis-related shocks forcefully and promptly to contain for SMEs a possible credit crunch for small and medium enterprises

(SMEs). Treasury investment SMBA Direct loan and loan (Small and Medium SMEs have been important contributors to eco- Business Administration) Financing nomic output, employment, and balanced regional Indirect Central Contribution KoFC loan development in Korea. SMEs represented 99.9 percent gov’t (Korea Finance Corp.) Onlending of the total number of firms and 86.9 percent of the guarantee) (fund+ Contribution total labor force in 2011. They contributed 48 percent KODIT (Korea Credit Guarantee Fund) Credit to GDP in 2011 and 69.8 percent of new job creation KOTEC guaranteeCredit guarantee during 2008–10. More than half of SMEs are located (Korea Technology loan Bank Credit Guarantee Fund) outside the Seoul metropolitan area, contributing to of Korea regional economic development. Aggregate credit ceiling loan Small and medium enterprises Financial institution An economic crisis often constrains financial access KOREG for SMEs, but lending to SMEs continued to grow (Korean Federation of Credit Indirect Credit Guarantee Foundations) guarantee loan during economic crises in Korea (Figure 2.4.1).1 Local gov’t Financial crises have a negative impact on SMEs’ Subsidies for interest rates gap profitability and creditworthiness in many coun-

Sources: Yi (2012); and IMF staff modifications. Figure 2.4.1. Outstanding Balance and Growth of SME Loans tries. Financial intermediaries typically tighten credit 80 500 conditions, thus worsening SMEs’ access to finance Asian Dot-com financial bubble (OECD, 2013). In contrast, SME loans in Korea crisis burst recorded positive growth in the year following crises.2 60 400 During the Asian crisis, the Korean authorities Outstanding Global responded with a host of financial support programs balance financial

won for SMEs (Figure 2.4.2). First, the authorities ramped 40 (right scale) 300

crisis n up existing credit guarantee programs by more than 90 percent on an annual basis (Figure 2.4.3), through the of Korea Korea Credit Guarantee Fund (KODIT), the Korea year-over-year change

20 200 Technology Credit Guarantee Fund (KOTEC), and Trillions

Percent, the Korean Federation of Credit Guarantee Founda- 3 0 100 tions (KOREG). Second, the Bank of Korea raised its aggregate credit ceiling and decreased preferential Growth rate (left scale) interest rates on loans by commercial banks to SMEs –20 0 to provide an additional incentive for SME lending 1997 99 2001 03 05 07 09 11

2 Sources: Financial Supervisory Services; and IMF staff calculations. Bank financing remains the most important source of Note: SME = small and medium enterprise. external financing for SMEs (83.3 percent) in Korea, followed by public lending (10.6 percent). Equity and bond financing accounted for 1.1 percent and 3.2 percent, respectively, in 2011. 3The funds facilitate loans by extending credit guarantees to The authors of this box are Heedon Kang and Yitae Kim. SMEs that lack tangible collateral but have good growth poten- 1Korea was affected by the 1997–98 Asian financial crisis, the tial. Three agencies support different types of SMEs: the KODIT bursting of the dot-com bubble in 2001, the credit card crisis in provides guarantees mostly for non-information-technology- 2003, and the global financial crisis. The credit card crisis related oriented start-ups and exporting SMEs; the KOTEC focuses mainly to household financial conditions, but the other three on information-technology-oriented SMEs; and the KOREG crises significantly affected the business environment for SMEs. supports regional SMEs.

84 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Box 2.4 (continued)

Figure 2.4.3. Outstanding Balance and Figure 2.4.4. Aggregate Credit Ceiling Growth of Credit Guarantees for SME Loans Loans

100 80 Outstanding balance (left scale) Ceiling (left scale) Asian Lending rate (right scale) Dot-com Global ﬁnancial ﬁnancial crisis bubble 14 6 burst crisis 80

60 12 Outstanding balance 5 60 (right scale) 10 4 40 40 8 3 Trillions of Korean won 20 Percent 6

Percent, year-over-year change Growth rate 20

(left scale) of KoreanTrillions won 2 0 4 Dot-com bubble 1 2 Asian bust –20 0 Global ﬁnancial ﬁnancial 1997 99 2001 03 05 07 09 11 crisis crisis 0 0 Sources: Korea Technology Credit Guarantee Fund (KOTEC); Korea 97199599 2001 03 05 07 09 11 13 Credit Guarantee Fund (KODIT); Korean Federation of Credit Guarantee Foundations (KOREG); and IMF staff calculations. Note: SME = small and medium enterprise. Source: Bank of Korea.

(Figure 2.4.4).4 Third, the Small and Medium Business helped stem job losses. Although SMEs were finan- Administration increased its policy lending to SMEs cially stressed and many went bankrupt at the outset by more than 60 percent. of the Asian financial crisis, the number of bankrupt- A successful experience during the Asian crisis led cies started to fall dramatically in 1999 (Figure 2.4.5); the authorities to repeat prompt policy responses in during later crises, these policies successfully prevented later crises.5 The quick recovery in Korea after the the bankruptcy of solvent SMEs with temporary Asian crisis is generally attributed in large part to liquidity shortages. Job losses also reversed quickly in accommodating macroeconomic policies, a favor- 1999 and did not occur during other crises (Figure able external environment, and a recovery in exports 2.4.6).6 Empirical studies show that supportive pro- supported by sharp depreciation of the Korean won. grams had strong profit-enhancing effects, especially However, specific policies to support SMEs also for innovative start-up SMEs, whose market access is contributed, and so the authorities were quick to limited despite their higher growth potential (Kang implement similar policy measures when the dot-com and Jeong, 2006; Kim, 2005).7 bubble burst in 2001 and when the global financial Although such policy measures can be seen as effec- crisis erupted in 2008. tive in easing access to finance for SMEs, they can The policy measures were instrumental in the pre- give rise to unintended consequences, such as missed vention of many disorderly SME bankruptcies, which opportunities for restructuring and high fiscal costs. SME financing support programs can undermine

4Aggregate credit ceiling loans (ACCLs) are extended by the Bank of Korea to commercial banks based on their SME loan 6Bankruptcy data disaggregated by enterprise size are not performance, up to a ceiling set by the Monetary Policy Com- available. mittee. The lending rates on ACCLs are kept lower than the 7The Bank of Korea enhanced its support for commercial policy rate to encourage banks to lend to SMEs. bank loans to innovative start-up SMEs by increasing the ACCL 5The Korea Finance Corporation was established in October ceiling by 3 trillion won and lowering preferential interest rates 2009; one of its purposes is to assist SMEs by supplying funds to from 1.25 percent to 0.5 percent. The Korea New Exchange financial institutions for onlending. (KONEX), a new stock market for SMEs, opened July 1, 2013.

International Monetary Fund | October 2013 85

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 2.4 (concluded)

Figure 2.4.5. Number and Growth of Figure 2.4.6. Growth in Number of Bankrupt Enterprises Employees (Percent; year-over-year change) Growth in number of bankruptcies (left scale) Number of bankrupt enterprises (right scale) 15

300 4,000 Asian Dot-com Global Asian financial bubble financial financial crisis burst crisis 250 crisis 3,500 10 Dot-com bubble 200 burst Global 3,000 financial crisis 5 150 2,500

100 2,000

Number 0

50 1,500 Percent, year-over-year change year-over-year Percent, 0 1,000 –5

–50 500

–10 –100 0 97199599 2001 03 05 07 09 11 9819962000 02 04 06 08 10 12

Sources: Bank of Korea; and IMF staff calculations. Sources: Bank of Korea; and IMF staff calculations.

creative destruction of nonviable SMEs. Despite the suggesting that political economy considerations may authorities’ strong commitment to reducing the pro- have played a role, which has resulted in a buildup in grams’ scale, in the wake of the Asian financial crisis government contingent liabilities. Nevertheless, the there has been an underlying upward trend. This trend policies so far have aided credit provision to SMEs and is particularly strong in the credit guarantee program, supported the Korean economy.

raise financial stability risks, most importantly incentives because they may lead banks to relax their by affecting credit risk in banks. For example, an screening and monitoring of borrowers. In addition attempt to encourage lending to SMEs by changing to increasing risks in banks, these incentive effects prudential rules (such as reducing prudential risk may lead to a misallocation of capital. weights) could jeopardize financial stability if the • Liquidity risk: Central bank provision of ample resulting risk weights do not appropriately account liquidity to banks, in part to encourage credit for the risks embedded in those exposures. Some extension, may weaken liquidity management and policies have tolerated or encouraged forbearance on discourage repair of private bank funding markets, loan payments by distressed firms, which could lead leaving banks overly reliant on central bank funding. to the practice of “evergreening,” whereby banks • Market risk: Authorities have directly intervened delay or fail to recognize loans as nonperforming.27 in credit markets to lower interest rates and ease Government guarantees of loans also affect lender financing conditions.28 Although appropriate for boosting growth in the current environment, when 27For risks associated with recent unconventional monetary policies (including the possibility of evergreening), see Chapter 3 of the 28As an additional risk, low interest rates tend to reduce interest April 2013 GFSR. margins, lowering bank profitability.

86 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Box 2.5. Lessons from the Nordic Banking Crises

This box discusses the policy responses of the Nordic authorities to the financial crises of the late 1980s and early Figure 2.5.2. Lending Growth by Banks 1990s, noting the importance of taking decisive action to (Percent) avert a lengthy recovery of credit growth. Finland Norway Sweden

60 Banking crises struck Norway in 1988 and Finland and Sweden in 1991. Although the episodes varied, each 50 was precipitated by significant financial liberalization and procyclical macroeconomic policies, which triggered rapid 40 credit growth, asset price inflation, and elevated private 30 sector indebtedness (Figures 2.5.1 and 2.5.2). Correc- tions to real estate prices, rising bankruptcies, and credit 20 losses followed various external shocks (for example, oil price declines, the collapse of the Soviet Union, and the 10 1 European Exchange Rate Mechanism crisis). 0 Sufficient macroprudential measures were absent in the run-up to the crises. This was in contrast to –10

–20 Figure 2.5.1. Real House Price Index in the Nordic Countries –30 (Quarterly index; historical average = 100) 1985 87 89 91 93 95 97 99

Finland Norway Sweden Source: Organization for Economic Cooperation and Development. 160

140 Denmark, which successfully avoided a crisis. While financial liberalization also began earlier, Danish banks 120 were better capitalized, in part due to favorable tax treatment of provisions and stricter requirements. Inadequate regulation of large exposures also allowed 100 substantial risks to accumulate in the other Nordic financial systems. 80 Once the crisis hit, responses varied: • In Norway, an independent fund was established to provide capital when losses threatened to deplete 60 capital at two of the four largest banks. The government eventually took ownership of both, alongside 40 the largest bank. 1985 87 89 91 93 95 97 99 • In Finland, following the takeover of the failed central savings bank, Skopbank, a fund was established Source: IMF staff calculations. to inject capital into the banking system together Note: Historical average refers to the average price computed with blanket guarantees. over the period 1980 to 2012. • In Sweden, one of the two largest banks that failed to meet regulatory capital requirements, Nordban- The author of this box is Ruchir Agarwal. ken, was merged with another bankrupt bank and 1 Average loan loss provisions over 1990–93 came to 3.4 subsequently broken up into a “bad” and “good” percent of total loans for Finland, 2.7 percent of total loans for Norway, and 4.8 percent of total loans for Sweden. See Drees bank. Government capital was injected into the and Pazarbasioglu (1998) for a comprehensive treatment of the failed banks and to fund the “bad” bank. Blanket Nordic banking crisis. guarantees were also issued.

International Monetary Fund | October 2013 87

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 2.5 (continued)

Conditional government support and government Decisive policy actions with little political uncertainty takeover were a critical part of the resolution. The Nordic were crucial. While lending contracted in the region, a governments protected taxpayers by wiping out most of serious credit crunch was avoided. Credit recovered by the incumbent shareholders and forcing banks to write the mid-1990s due to sound institutions that enabled down losses before injecting funds. In Finland and Sweden, orderly restructuring and strong governments with the “bad” assets were transferred to asset management compa- trust of the public to act in their best interest. nies that operated independently and with limited regulatory constraints, while the “good” banks focused on core banking tasks, facilitating credit within the system. Unlike the Finnish and Swedish governments, the Norwegian government did not extend its role as “owner of last resort” by guaranteeing bank liabilities and setting up a “bad bank” to deal with nonperforming loans. Since then, each government has maintained a portion of bank ownership.2

2Nordbanken eventually grew through regional mergers into the pan-Nordic bank, Nordea, in which the Swedish government’s addition, Solidium Oy, set up initially to manage Skopbank’s stake was 13 percent until July 2013, when it was reduced to industrial holdings and still fully owned by the Finnish govern- 7.1 percent. The Norwegian government maintained a stake of ment, retains a 3 percent share in Nordea through its holdings of 34 percent in Norwegian bank DNB as of December 2012. In the Sampo group.

central banks exit from their intervention, inter- robust liquidity and capital requirements. However, est rates will eventually rise. If such a rise is more some credit-enhancing policies are in fact designed to abrupt than expected (as in the adverse scenario in increase risk taking by lenders—for example, changing Chapter 1), banks may face substantial capital losses risk weights for loans to certain sectors. Offsetting pru- on holdings of fixed-rate securities. In addition, dential measures would undo the effects the policy is interest rate increases could lead to losses in the loan trying to achieve. Other policies also serve to enhance book as banks pass on their higher cost of funds to financial stability, either directly—for example, by borrowers (through, say, variable-rate loans), who improving the financial position of banks—or indi- may struggle to make higher loan payments. rectly—for example, by improving confidence—so that • Risk of moral hazard: Government financial support the extreme downside risks that were present in the cri- carries the chance that financial institutions will take sis are ameliorated. Still, in some cases, there could be more risks than they otherwise would, anticipating tension between supporting credit and raising financial that the government will again intervene and bail stability risks. If, in such circumstances, the authori- them out if they face trouble. Policy design should ties choose to promote credit, then it would suggest take into account such “moral hazard” and build in that increased credit risk in banks is accepted as part incentives for beneficiaries of government interven- of the cost of credit-supporting policies. Nevertheless, tion to behave prudently so as not to jeopardize policymakers need to continually weigh the near-term public funds. Recent efforts to introduce such incen- benefits against the longer-run costs of policies aimed tives are ongoing (FSB, 2011; IMF, 2012c). at boosting credit. Mitigation of these risks may not be necessary or Credit-enhancing policies raise similar issues of a appropriate while the economy is still weak, as it could possible trade-off between objectives in the context of run counter to the objectives of the credit policies; still, the broader agenda for financial reform. This impor- policymakers will need to remain cognizant of these tant and ambitious policy agenda includes more robust potential risks. In principle, the appropriate supervi- capital and liquidity standards for banks under Basel sory response to increased risks is to put prudential III, enhanced monitoring for shadow banks and other measures in place for mitigation, including enhanced nonbank financial intermediaries, and implementa- credit risk management, adequate loss provisions, and tion of macroprudential frameworks. The goals of this

88 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

broader policy agenda are to improve the quality and Better data are crucial for improving the analysis of quantity of capital, foster better liquidity manage- factors underlying weak credit. The investigation in ment and more accurate asset valuation, and develop this chapter was hampered significantly by a dearth of and implement more effective macroprudential tools. appropriate data, even for the major advanced econo- Overall, these measures should make banks stronger mies. Policymakers should aim to expand the scope and thus help sustain their role in credit markets of available data, in particular information that would in the medium term. Still, in the short term, some allow for identification of factors that may constrain regulatory changes may restrain bank lending; for loan demand and loan supply. For example, access to example, enhanced capital requirements may make it disaggregated loan data with information on borrow- more difficult for banks to increase lending. Therefore, ers and lenders would facilitate the examination of putting offsetting measures in place until these short- shifts in the supply of credit by effectively controlling term constraints are eased may be useful; for example, for demand, as that data would allow matching of data authorities may wish to urge banks to raise capital so from borrowers applying for loans at multiple banks. that enhanced capital requirements do not lead to less Data from credit registries could be useful in this regard. lending by banks. In addition, more extensive use of lending surveys with In addition to financial stability risks, the potential better-directed questions would allow for improved fiscal costs of policies should be considered.29 Some analysis. These recommendations are important also measures may raise credit activity but may impose a for policymakers in emerging markets, who could then substantial fiscal cost, including in the form of con- apply the framework developed in this chapter. tingent liabilities. Costs can include potential losses In sum, measures to stimulate private credit should on assets purchased by the central bank, loan losses in be designed with care. Policies to boost lending in the state-sponsored institutions engaged in direct lending short term can be beneficial, but can also carry costs to firms and households, and the carrying cost (inter- and potential risks to future financial stability if poorly est) on funds used to recapitalize banks, among others. designed or targeted. For prudent policymaking in this Contingent liabilities could include expanded deposit area, authorities should (1) identify the constraints to insurance and loan guarantees given by the public loan demand or supply that can be addressed with gov- sector. Some policies, such as adjustments in basic ernment intervention; (2) align the policies with the regulation or legal changes, do not incur substantial identified constraints; (3) be mindful of interactions direct fiscal costs. with other policies, including regulatory measures; (4) keep in mind direct and contingent costs of these policies to the government; (5) assess potential longer-term financial stability implications of such policies; and (6) 29See IMF (2010) for estimates of the fiscal costs associated with financial sector support measures during the 2008 crisis for G20 if warranted, establish appropriate prudential measures countries. to mitigate such stability risks.

International Monetary Fund | October 2013 89

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Annex 2.1. Previous Findings in the Literature households face tightened collateral constraints, they on Credit Constraints may increase precautionary saving (by lowering consumption). Although more saving eases credit supply Economic theory suggests that financial intermedia- constraints, lower consumption dampens credit tion suffers from potential intrinsic difficulties in the demand. These mechanisms may slow economic efficient allocation of scarce credit. Two important recovery (Guerrieri and Lorenzoni, 2011). difficulties involve (1) a maturity mismatch between • Debt overhang: Debt overhang can affect credit long-term borrowers and short-term creditors, and (2) demand and credit supply. Highly indebted firms an informational asymmetry between creditors and may not pursue otherwise profitable business borrowers. Informational asymmetries occur when a opportunities (Myers, 1977), thus lowering credit borrower’s misbehavior is not observed (moral hazard); demand. Similarly, more highly indebted households when borrowers’ risk types are not observed (adverse may choose not to take out loans, even though selection); or when information can be obtained but doing so could increase their overall current and with some costs (costly state verification). The litera- future well-being. Thus, an economy-wide debt ture has shown that, despite these market failures, overhang can slow growth and deflate asset prices efficient allocation of credit can still be achieved, and (Adrian and Shin, 2013), negatively affecting col- permanent government intervention is not necessary lateral values (and thus further constraining credit (Townsend, 1979; Prescott and Townsend, 1984a, creation). Debt overhang can also affect credit sup- 1984b; Bisin and Gottardi, 2006; Allen and Gale, ply when the overhang is in banks: highly leveraged 2004).30 banks may have difficulty obtaining funding (for However, in recessions, these market failures may example, in the interbank markets) and thus lack amplify credit contractions. The financial amplification the liquidity to make additional loans. mechanisms and their key factors described below have • Relationship banking: Informational asymmetry can been confirmed empirically for past major recessions. ease when banks and their borrowers have on going Preliminary evidence also suggests that these mecha- business relationships, which allow banks to know nisms are at work in the current recession (see Table 2.7, their customers and keep borrowers from mis- under the heading Identifying Amplification Frictions). behaving in order to obtain loans in the future • Collateral constraints: Requiring collateral (an asset) (Townsend, 1982; Sharpe, 1990; Rajan, 1992). from a borrower to secure a loan is appropriate However, in a severe recession, many of those behavior by a lender to help mitigate informational relationships may disappear because of the actual (or asymmetry. Using collateral to obtain a loan eases potential) bankruptcies of banks and firms. Banks the borrower’s liquidity constraint (a form of matu- respond by raising the risk premium they charge on rity mismatch), because liquidity is obtained from loans, in essence tightening the supply of credit. a less liquid asset. A drop in the value of collateral During normal times, the government’s role in miti- as a result of asset price declines (in stock or bond gating intrinsic market failures is limited. The govern- markets, for example) shrinks the loan that can be ment cannot acquire better information on borrowers obtained with that collateral, tightening credit supply. or change maturity preferences. Still, structural policies A similar mechanism affects interbank markets: lower can be pursued to increase information flows (for collateral prices would lower the amount banks will example, by instituting or improving a credit registry lend to each other in interbank markets, restricting or enhancing accounting standards and public disclo- bank funding and again tightening credit supply. sures) or to promote alternatives to bank credit, such On a macroeconomic level, this may further lower as a corporate bond market or securitization. asset prices (Kiyotaki and Moore, 1997; Gertler and But when market failures amplify severe downturns, Karadi, 2012; Geanakoplos, 2010). Moreover, when government intervention has a clearer role. In such The author of this annex is Kenichi Ueda. situations, the government can use its credit rating, 30Exceptions are government intervention through deposit insur- generally higher than that of the private sector, to ease ance and microprudential regulation. The former prevents bank runs credit constraints. For example, a central bank could that may result from maturity mismatches and the latter prevents excessive risk taking by banks, including as a result of deposit lend directly to firms (Gertler and Karadi, 2012), insurance. thus taking over the financial intermediation role. It

90 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS (Continued) finance. Creditless recovery is slower but only during the occurred in 2009–10. first two years. amounts. importance of stochastic collateral Finds stochastic collateral constraint). constraint to explain actual data. recovery during recession with credit crunch because households increase precautionary savings after loss of borrowing capacity. when applying for loans. The ratio of loan amount to collateral value has when applying for loans. more loan applications Yet Higher margins and fees are paid. declined. are rejected. lending. so labor market mismatch could worsen. Household movements indeed so labor market mismatch could worsen. but declined within out-of-county a migration county, was not affected suggesting effects on the labor market were small. much, to take advantage of improved competitiveness, but they suffer from to take advantage of improved competitiveness, Similar exporters collateral constraints and banking sector distress. confirming with higher foreign ownership have much larger investment, importance of bank liquidity/capital channel. their branches. well for firms that do not rely on banks but trade credit. banks reallocated loans away from riskier smaller firms, low-capital loans. “evergreen” banks seem to to evergreen loans, e.g., to circumvent the capital ratio requirement with e.g., to evergreen loans, Increase in the number of zombies depressesregulatory forbearance. investment and employment growth and lowers productivity. Key Findings Quick recovery is often observed without credit. Creditless recovery is slower, in particular after asset price bust. Creditless recovery is slower, Creditless recovery is slower, especially for industries relying on external Creditless recovery is slower, About half 25 percent of all episodes. Creditless recovery is not rare, collateral (land) value matters for investment (0.08 elasticity) and loan Firm’s Estimates key parameter values of DSGE model with financial frictions (i.e., Households’ collateral constraints can explain low interest rates and slow Compared with precrisis, in 2007–09, more SMEs are asked for collateral in 2007–09, Compared with precrisis, nonperforming loans) affects new Banks’ exposure to real estate (i.e., Banks’ capital ratio and liquidity ratio matter for loan provision to firms. Inefficient lock-in effect may Underwater arise: households cannot move, Innovative small firms find it harder to access finance than other firms. exporters should demand credit and banking) crisis, In a twin (currency State-level real activities are affected by distressed Japanese banks through Firms that had relationshipFirms with western European banks suffered more. With subsidized loans, unviable firms (zombies) survive. Banks have incentives unviable firms (zombies) survive. With subsidized loans, loan level loan level loan level Japanese bank-level data exposition Data Country-level data Country-level data Industry-level data Country-level data Bank-firm matched micro data, Flow of funds County-level data SME firm-level data with bond ratingsFirms Bank-firm matched micro data, Corporate debt overhang is confirmed. Bank-firm matched micro data, County-level data SME firm-level data Firm-level data U.S. SMEs U.S. firm-level data Bank relationship is important for quantity of firm-level credit. U.S. state-level U.S. activity data and U.S. county-level U.S. activity dataBank-firm matched micro data County-level real activities are affected by sudden bankruptcy of banks. Moral hazard is not a problem in bank-firm relationship but explains data Bank-firm matched micro data Bank-firm matched micro data Low bank capitalizationWhile larger and low-capitalscarce liquidity matter. Firm-level Firm-level regression/theoretical for creditless recovery) regression stochastic general equilibrium general equilibrium theory regression equilibrium model with moral hazard regression regression Methodology Descriptive charts Regression of duration of recession Panel regression Panel Probit regression (takes value of 1 panel Natural experiment, dynamic Structural estimation, dynamic stochastic Calibration, regression Panel Structural estimation, investment Structural estimation, panel Natural experiment, regression Panel Panel regression Panel Panel regression Panel regression/crisis event study Panel Panel regression Panel Natural experiment Structural estimation, general Structural estimation, Natural cross-section experiment, Natural cross-section experiment, regression Panel post Lehman 1980–2004 1960–2010 1964–2004 1965–2011 1994–98 1984–2010 2000–10 2001–09 1992–95 1994–98 2002–10 2007–09 2007–12 1990–2005 1988–89 1990s 1988, 19921988, Natural experiment 2000–06 2008–09 Six months 1993–2002 ©International Monetary Fund. Not for Redistribution ©International Monetary countries Country/Region Year Japan United Kingdom United States Japan Spain Six Latin American United States United States United States Spain Italy Japan Sanchez (2010) Collateral Constraints Gan (2007a) Jermann and Quadrini (2012)Guerrieri and Lorenzoni (2011) United States (2012) Fraser United States Debt Overhang Hennessy (2004) Gan (2007b) Jiménez and others (2012) Donovan and Schnure (2011) United States Lee, Sameen, and Martin (2013) Sameen, Lee, United Kingdom Kalemli-Ozcan, Kamil, Villegas- and Kamil, Kalemli-Ozcan, Relationship Banking and Rajan (1994) Petersen Peek and Rosengren (2000) Peek Ashcraft (2005) Karaivanov and others (2010) and van Horen (2013) Peydró, Ongena, Eastern Europe Albertazzi and Marchetti (2010) and Kashyap (2008) Hoshi, Caballero, Category/Paper Creditless Recovery (2006)Talvi and Izquierdo, Calvo, 31 emerging markets Claessens, Kose, and Terrones (2012)Terrones and Kose, Claessens, 44 countries Abiad, Dell’Ariccia, and Li (2011) Dell’Ariccia, Abiad, 48 countries Sugawara and Zalduendo (2013)Identifying Amplification Frictions 96 countries Table 2.7. Previous Findings in the Literature 2.7. Table

International Monetary Fund | October 2013 91 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY (Continued) security market. credit line drawdowns increased. Given bank funding strains, credit line Given bank funding strains, credit line drawdowns increased. drawdowns further constrained banks from making new loans. may provide misleading picture of true financing needs. firms in general but increased credit for first-time borrowers. a particularly large role in 2009. In 2010 and 2011, however, the however, In 2010 and 2011, a particularly large role in 2009. large factor was banks’ own weakened fundamentals and their more conservative way of responding to these fundamentals. growth. lending growth with one unit increase in demand factor in the survey), lending growth with one unit increase in demand factor the survey), although supply factors play a role in mortgage lending and pockets of SME lending. However, according to the INE Investment Survey, credit conditions have according to the INE Investment Survey, However, tightened significantly in some segments. appears stabilized. Still, conditions vary widely across countries. Still, appears stabilized. Banks with lower capital ratios and higher nonperforming loans show corporate loan rates are affected by sovereign In turn, more sensitivity. with risks, 30 to 40 percent Both pass-through. demand and supply factors from bank lending surveys explain quarterly credit growth significantly. existence of many unviable SMEs, partly as the result of credit support existence of many unviable SMEs, policies. to be important in explaining actual loan amount and real output. to be important in explaining actual loan amount and real output. the This excess time-varyingbond premium. premium explains most of GDP growth. by the Bank Lending Survey. The credit supply factor is stronger in the by the Bank Lending Survey. 1 crisis period in quarterly growth rate of bank lending to firms (2 bps vs. Demand factor is about 1 bp. bp beforehand). supply shocks. Loan supply shocks are found to be significant. The loan Loan supply shocks are found to be significant. supply shocks. to zero by 2010:Q2. but is close supply shock was quite large in 2008:Q4, Heterogeneity among countries has also been reduced. Key Findings Trade credit is widely used in the world. Trade Bank credit lines are used well during episodes of sudden malfunctioning New loans dwindled about 80 percent from the peak in 2008:Q4, but New loans dwindled about 80 percent from the peak in 2008:Q4, Total available funds seem sufficient to cover investment, but aggregate data available funds seem sufficient to cover investment, Total SMEs’ use of trade credit increased after onset crisis. Securitization of real estate loans did not affect credit for nonreal-estate German firms became less reliant on bank lending. Abnormally low credit supply and high risk premium were observed. In the credit slowdown during 2008–11, macroeconomic conditions played In the credit slowdown during 2008–11, High levels of sovereign, corporate, and household debt are detrimental to corporate, High levels of sovereign, Weak lending is mostly demand driven (3 to 4 bps increase in quarterly Weak Firms’ rapid deleveraging is mainly voluntary, driving deleveraging by banks. driving deleveraging by banks. rapid deleveragingFirms’ is mainly voluntary, Overall, SME access to finance in 2011 and early 2012 was tight but Overall, News on sovereign crisis affects Italian banks’ CDS and bond spreads. News on sovereign crisis affects Italian banks’ CDS and bond spreads. Bank lending standards (U.S. Senior Loan Officer Opinion Survey) are found Bank lending standards (U.S. Constructs a better credit spread index based on corporate bond spread: Constructs a better credit spread index based on corporate bond spread: Bank loan growth is explained by both supply and demand factors covered Uses sign restrictions to identify aggregate supply and demand loan loan level survey survey survey Data Transaction-level data Transaction-level Flow of funds Various micro data Various Flow of funds SME firm-level data Bank-firm matched micro data, Country-level data Country-level data Bank-level data Country-level data Country-level data/bank lending Country-level data Country-level data Country-level data Various country Various level dataA reason appears to be the especially for SMEs. Credit growth is low, Country-level data/bank lending Country-level data Country-level data/bank lending Country-level data Methodology OLS Descriptive charts Descriptive charts Descriptive charts Demand/supply disequilibrium MLE Panel regression Panel Descriptive charts Descriptive charts Panel regression Panel Aggregate country panel regression Panel regression Panel Descriptive charts Descriptive charts/statistics VAR/system of equations VAR/system Descriptive charts VAR VAR Panel regression Panel VAR 2005 2001–08 2000–08 1952–2012 1994–2008 1999–2009 1991–2010 2008–13 2001–11 1980–2009 2002–12 2007–12 2007–12 2006–12 2000–12 1968–84 1990–2000 1973–2010 2003–09 2003–10 ©International Monetary Fund. Not for Redistribution ©International Monetary (continued) Country/Region Year United States United States Spain Spain Germany Europe 21 CESEE countries 18 OECD countries Ireland Portugal OECD countries Italy Japan United States United States Euro area Euro area and Udell (2012) (2012) Category/Paper Alternative Credit Sources and Rajan (2012) Laeven, Klapper, 500 Global Fortune Chari, Christiano, and Kehoe (2008) Christiano, Chari, United States Ivashina and Scharfstein (2010) Chari (forthcoming) Carbó-Valverde, Rodríguez-Fernández, Rodríguez-Fernández, Carbó-Valverde, Jiménez and others (2011) Deutsche Bundesbank (2012) Credit Supply and Demand GFSR (2011–13) IMF, IMF (2013b) IMF (2013a) IMF (2012a) IMF (2013d) OECD (2013) Zoli (2013) Lam and Shin (2012) Lown and Morgan (2006) Gilchrist and Zakrajsek (2012) Hempell and Sørensen (2010) Hristov, Hülsewig, and Wollmershäuser Wollmershäuser and Hülsewig, Hristov, Table 2.7. Previous Findings in the Literature 2.7. Table

92 International Monetary Fund | October 2013 ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Monetary transmission is affected through the credit channel only in The transmission stems from impaired bank balance distressed countries. sheets for 2008–09 (only) and from weak credit demand 2008–11. was effective implying monetary policy The former effect is also smaller, for the former but not latter. supply plays a negligible role. factor lowered loan amounts 2.3 to 3.1 of percent, which one-quarter is attributed to banks’ weak capital positions and the other to increased perception of borrowers’ credit risk. period. At the peak between 2009:Q3 and 2010:Q1, credit supply factors At the peak between 2009:Q3 and 2010:Q1, period. in bank lending surveys explain 35 to 40 percent of bank lending, equivalent to about a 0.5 percent dampening of quarterly loan growth. lending survey. For business loans, each factor contributes about half of business loans, For lending survey. the loan growth deviation from the sample mean during crisis period. but supply factors were stronger, from 2008:Q2 to 2009:Q2, However, demand factors became stronger since. that is unexplained by macro and bank-level Credit factors. supply effects become larger than when using only the bank lending survey. recently. However, demand remained mixed, likely due to a lack of demand remained mixed, However, recently. the amount of new secured As for households, confidence among firms. credit increased significantly recently. shocks. The bank supply shocks explain 40 percent of aggregate loan shocks. and investment fluctuations. Key Findings Uses bank lending survey outcomes as credit supply and demand shocks. Uses bank lending survey outcomes as credit supply and demand shocks. Loan amounts are mostly demand driven (about 90 percent) and credit Both credit supply and demand play a role. For 2007–09, the credit supply 2007–09, For Both credit supply and demand play a role. Bank lending has been both supply and demand driven, even in the crisis Bank lending has been both supply and demand driven, Bank lending is explained by both supply and demand factors in the bank Constructs a cleaner credit supply measure: a lending survey component credit supply measure: Constructs a cleaner The overall availability of credit to the corporate sector has increased Loan movements are decomposed into bank, firm, industry, and common industry, firm, Loan movements are decomposed into bank, survey survey data data survey bank lending survey survey Data Country-level data/bank lending Country-level data/bank lending Bank lending survey/bank-level Bank lending survey/bank-level Country-level data/bank lending Country-level/loan level data/ Country-level/bank lending Bank-firm matched micro data Methodology VAR Bayesian model averaging Panel regression Panel Panel regression Panel VAR/panel regression VAR/panel Descriptive charts Decomposition of loan movements 2002–11 2002–11 2002–09 2003–10 2003–10 OLS 1992–2011 2007–13 1990–2010 ©International Monetary Fund. Not for Redistribution ©International Monetary (concluded) Country/Region Year Austria Italy Germany France United States United Kingdom Japan Category/Paper Ciccarelli, Maddaloni, and Peydró (2013) and Peydró Maddaloni, Ciccarelli, Euro area Beer and Waschiczek (2012) Waschiczek Beer and Del Giovane, Eramo, and Nobili (2011) Eramo, Del Giovane, Blaes (2011) Lacroix and Montornès (2010) Bassett and others (2012) Bank of England (2013) Amiti and Weinstein (2013) Weinstein Amiti and Source: IMF staff. Source: bp Note: = basis point; CDS = credit default swap; CESEE = and eastern, central, southeastern Europe; DSGE = dynamic stochastic general equilibrium; INE = Instituto Nacional de Estatística; MLE = maximum likelihood = vector autoregression. estimation; OECD = OrganizationVAR for Economic Cooperation and Development; OLS = ordinary least squares; SMEs = small and medium enterprises; Table 2.7. Previous Findings in the Literature 2.7. Table

International Monetary Fund | October 2013 93 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

can also loosen collateral rules to ease the liquidity appears that existing bank credit lines were used more constraints that result from declines in collateral values. intensively in the United States, although perhaps Treasuries can use their superior credit status simi- by crowding out new loans (Chari, Christiano, and larly, for example, by extending subsidized loans via Kehoe, 2008; Ivashina and Scharfstein, 2010). In state-sponsored institutions. In addition, governments another example, credit-constrained SMEs in Spain can remedy debt overhang by facilitating debt restruc- increased their use of trade credit (Carbó-Valverde, turing—for example, through bank recapitalization, Rodríguez-Fernández, and Udell, 2012). purchases of nonperforming assets, or reforms of laws Previous studies have also looked at credit market related to bankruptcy. These government interventions developments in various countries (see Table 2.7, under also help preserve relationships between banks and the heading Credit Supply and Demand). Some studies clients, easing another potential market failure. have found that credit supply appeared to constrain The market itself may also find ways to ease credit credit growth in many countries, in particular during constraints. In some countries, credit from alternative late 2008 and 2009 (Hempell and Sørensen, 2010; Del sources has likely mitigated increased market fric- Giovane, Eramo, and Nobili, 2011). Others also found tion during the recent recession (see Table 2.7, under low credit demand from 2008 to date in a number of the heading Alternative Credit Sources). For example, (mostly advanced) economies (Ciccarelli, Maddaloni, when the money and corporate bond markets did not and Peydró, 2013). function well after the collapse of Lehman Brothers, it

94 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Annex 2.2. Determinants of Bank Lending Figure 2.12. Decomposing Lending Standards: Corporate Standards Loans European Central Bank and Federal Reserve survey Lending standards Economic outlook factors Balance sheet factors 1. Austria 2. France results indicate that lending standards for corporate 60 120 and mortgage loans tightened considerably in late 50 100 2008 for most countries (Figures 2.12 and 2.13). 40 80 30 60 Conditions eased during 2010, but during the past 20 40 two years some European countries experienced a sec- 10 20 ond round of tightening in lending standards. In the 0 0 United States, corporate lending standards have not –10 –20 seen further strains since 2008–09. –20 –40 2003 05 07 09 11 13 2003 05 07 09 11 13 The surveys ask loan officers for the reasons behind 4. Italy tightened lending standards, which allows the con- 40 3. Germany 60 struction of a variable that reflects mostly supply 30 50 40 constraints. Responses on the tightness of lending con- 20 30 ditions may not necessarily reflect “pure” constraints 10 on the supply of credit, such as bank liquidity and 20 0 capital. The responses could also reflect effects on the 10 standards from changes in borrowers’ creditworthiness, –10 0 –20 –10 the economic outlook, economic uncertainty, and the 2003 05 07 09 11 13 2003 05 07 09 11 13 like. Aside from potentially affecting the willingness of 5. Netherlands 6. Portugal banks to make loans, these factors are also related to 60 100 loan demand conditions. The influence of these factors 40 80 can be statistically removed from the lending standards 60 20 variable (following Valencia, 2012) to obtain a measure 40 of lending standards that more closely reflects the 0 20 ability of banks to supply credit—that is, connected to –20 0 bank balance sheet constraints. –40 –20 To find the determinants of bank lending standards, 2003 05 07 09 11 13 2003 05 07 09 11 13 a regression is run with the overall credit standards 7. Spain 8. United States index as a dependent variable and the reasons for tight- 60 100 ening as explanatory variables. The results for the euro 80 40 area are shown in Table 2.8.31 The sample includes 60 40 Austria, France, Germany, Italy, Luxembourg, the 20 Netherlands, Portugal, and Spain. Regressions are also 20 run in which the real GDP forecast and stock market 0 0 volatility are included instead of answers related to the –20 –20 –40 economic environment, as more direct proxies for the 2003 05 07 09 11 13 2003 05 07 09 11 13 latter. This specification corresponds to the second and fifth columns in Table 2.8, for corporate and mortgage Sources: European Central Bank, Bank Lending Survey; Federal Reserve, Senior Loan Officer loans, respectively.32 Balance sheet constraints (capital Survey; and IMF staff calculations. Note: Y-axes have different scales. For European countries, lending standards correspond to enterprises and are measured as weighted net percentages. For the United States, lending standards correspond to commercial and industrial loans to large and middle-market firms The author of this annex is Nicolas Arregui. and are measured as unweighted net percentages. Economic outlook and balance sheet 31The specifications for corporate and mortgage loans differ factors are constructed using the first specification in Table 2.8 (Table 2.9 for the United because the available options included in the surveys to justify the States). Economic outlook factors are the fitted values constructed using the responses to tightening or easing in lending standards for corporate and mortgage general economic activity and industry and firm outlook (general economic activity for the United States) and setting all other coefficients to zero. Analogously, balance sheet factors are loans differ. the fitted values constructed using the responses to capital and liquidity position and access 32We also include a specification augmented with the expected to market financing (capital position for the United States). behavior of demand taken from the survey because banks may change lending standards based on an expected change in demand behavior. The variable is not significant.

International Monetary Fund | October 2013 95

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 2.13. Decomposing Lending Standards: Mortgage and liquidity position, access to market financing for Loans corporate credit, and cost of funds for mortgage loans)

Lending standards Economic outlook factors Balance sheet factors are significant. Competition from other banks turns 1. Austria 2. France out to be significant for both types of credit. The gen- 30 40 eral outlook and housing prospects are also significant. 20 30 Table 2.9 shows the results for the United States. The 20 10 capital position and economic outlook are significant 10 0 in this case. 0 –10 Using the coefficients from the first stage, measures –10 of lending standards are constructed in which the –20 –20 influence of non-balance-sheet factors is removed. Fit- –30 –30 2003 05 07 09 11 13 2003 05 07 09 11 13 ted values of the dependent variables are constructed

3. Germany 4. Italy using the coefficients on the balance sheet factors: 30 60 capital position, market financing, liquidity (for corpo- 50 20 rate loans), and the cost of funds (for mortgage loans), 40 while all other coefficients are set to zero. The capital 10 30 position is used for the United States. 20 0 Figures 2.12 and 2.13 show the resulting decom- 10 –10 position of lending standards for corporate loans and 0 mortgage loans, respectively, into demand and supply –20 –10 2003 05 07 09 11 13 2003 05 07 09 11 13 factors for major countries for which long data series

5. Netherlands 6. Portugal are available (with different y-axis scales, as appropri- 60 100 ate). In general, the figures show that lending standards 40 80 are, in fact, affected to a considerable extent by the 60 20 economic outlook, which also affects loan demand. 40 The supply factors related to bank balance sheet 0 20 constraints come into play in specific periods during

–20 0 the crisis and its aftermath. For example, for corporate

–40 –20 loans, supply factors restricted lending standards at 2003 05 07 09 11 13 2003 05 07 09 11 13 the start of the financial crisis in France, Germany, 7. Spain and the United States and also came into play in early 60 2012 in France and Italy as financial strains increased 33 40 in the euro area. For mortgage loans, balance sheet constraints also restricted lending standards at the 20 beginning of the crisis in most European countries shown and again in 2012 in Austria, France, Italy, and 0 Portugal.

–20 The next step is to determine how credit growth is 2003 05 07 09 11 13 affected by the demand and supply effects measured by the adjusted survey responses. Credit growth is Sources: European Central Bank, Bank Lending Survey; and IMF staff calculations. assumed to depend partly on past credit growth (to Note: Y-axes have different scales. Lending standards correspond to mortgage loans and are measured as weighted net percentages. The results for France are weighted by the share of capture momentum or “persistence” effects) and partly the outstanding loans issued by each bank in the French Bank Lending Survey sample in the total outstanding loans issued by all the banks in the sample. Economic outlook and balance on loan demand and supply conditions as measured sheet factors are constructed using the first specification in Table 2.8. Economic outlook factors are the fitted values constructed using the responses to general economic activity and setting all other coefficients to zero. Analogously, balance sheet factors are the fitted values constructed using the responses to cost of funds. 33The analysis does not show supply factors playing a significant role in recent years for Spain. Because the survey shows only changes in lending standards, it may be that the level is already quite tight. Alternatively, this may be the result of reporting bias (with banks adjusting their survey responses to downplay funding strains).

96 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Table 2.8. Euro Area: Determinants of Bank Lending Standards Dependent Variable: Overall Lending Standards, 2003:Q1–13:Q2 Corporate Loans Residential Mortgage Loans (1) (2) (3) (4) (5) (6) Capital Position 0.112 0.308*** 0.112 Cost of Funds 0.384*** 0.679*** 0.363*** (0.085) (0.062) (0.084) (0.087) (0.097) (0.099) Access to Market Financing 0.317* 0.436*** 0.317* Competition from Other Banks 0.234** 0.217 0.230** (0.141) (0.092) (0.143) (0.089) (0.126) (0.093) Liquidity Position 0.243** 0.175 0.243** Competition from Nonbanks –0.231 –0.261 –0.237 (0.093) (0.102) (0.090) (0.177) (0.243) (0.163) Competition from Other Banks 0.179*** 0.271** 0.179*** General Economic Activity 0.197*** 0.193*** (0.034) (0.095) (0.038) (0.037) (0.036) Competition from Nonbanks –0.256 –0.357 –0.256 Housing Market Prospects 0.274** 0.260** (0.252) (0.338) (0.247) (0.106) (0.095) Competition from Market Financing 0.557* 0.775 0.557* (0.263) (0.425) (0.252) General Economic Activity 0.125* 0.125* (0.062) (0.062) Industry or Firm Outlook 0.128* 0.128 (0.061) (0.068) Collateral Risk 0.338 0.338 (0.230) (0.231) Stock Market Volatility 0.521*** 0.374** (0.131) (0.134) Expected Real GDP Growth 1.663** 1.336 (0.542) (1.748) Expected Behavior of Demand 0.001 Expected Behavior of Demand –0.033 (0.035) (0.041) Observations 336 287 336 336 287 336 R Squared 0.767 0.710 0.767 0.617 0.540 0.619 Number of Countries 8 7 8 8 7 8 Source: IMF staff estimates. Note: Variables measured as weighted net percentages (share of banks that report a significant or moderate tightening, mutiplied by 1 and 0.5, respectively, minus the share of banks that report a significant or moderate easing, mutiplied by 1 and 0.5, respectively). Sample includes Austria, France, Germany, Italy, Luxembourg, the Nether- lands, Portugal, and Spain. Fixed effects regressions with robust standard errors are in parentheses. ***, **, and * indicate significance at the 1 percent, 5 percent, and 10 percent levels, respectively.

Table 2.9. United States: Determinants of Bank by the decomposition of the lending standards variable Lending Standards from the surveys.34 Formally, the regression Dependent Variable: Overall Lending Standards, 1999:Q1–2013:Q2 United States Credit growtht = a + bCredit growtht–1 + giDemand Commercial and Industrial Loans factorst–i + diSupply factorst–i + et (2.1) Capital Position 0.601** (0.270) is estimated using quarterly data for the period Economic Outlook 0.290*** (0.085) 2003:Q1–2013:Q1 for European countries and Liquidity in Secondary Market 0.049 (0.161) 1999:Q1–2013:Q1 for the United States. The sub- Competition from Other Banks 0.039 script i indicates lags of the variables. Several lags could (0.031) Tolerance for Risk 0.036 be included, adding more terms to the equation. e is a (0.093) random error term. Observations 58 R Squared 0.899 The coefficients found in the regressions, shown in Source: IMF staff estimates. Table 2.4 in the main text for the euro area and the Note: Variables are measured as unweighted net percentages (share of banks United States, can be used to calculate how much of reporting a significant or moderate tightening minus the share of banks reporting a significant or moderate easing). Ordinary least squares regressions with the recent evolution in corporate and mortgage credit robust standard errors are in parentheses. ***, **, and * indicate significance at the 1 percent, 5 percent, and 10 percent levels, respectively. growth can be explained by demand and supply factors (see Figures 2.8 and 2.9 in the main text). The demand component is the fitted values constructed recursively using the lags for the demand index and setting the “pure” supply index to zero. The supply component is constructed analogously.

34Demand factors are measured by the net fraction of banks that report in the survey that they observe an increase in demand for loans.

International Monetary Fund | October 2013 97

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Annex 2.3. A Model of Bank Lending Figure 2.14. Effects of a Tightening of Lending Supply and a Drop in Lending Demand A simple model of credit markets consists of two equations: a supply equation for new loans and a demand 35 1. Tightening of Bank Lending Supply 2. Drop in the Demand for Bank Loans equation. Both the supply of and demand for bank Lending Lending loans are functions of the lending rate and other vari- rate Demand rate Demand ables. In the familiar price-quantity plot (Figure 2.14), Supply Supply the supply curve slopes upward and the demand curve r2 slopes downward: banks will supply more loans if the r1 r1 r2 interest rate is higher, and borrowers will demand fewer loans if the rate is higher. The lending interest rate adjusts to clear the market—that is, to equalize Q2 Q1 Volume of Q2 Q1 Volume of demand and supply.36 The magnitude of the reduc- new loans new loans tion in the equilibrium quantity of new bank loans associated with an increase in lending rates depends on Source: IMF staff illustration. the sensitivity (or elasticity) of both credit demand and supply to interest rates. banks (proxied by the deposit rate and by banks’ credit Changes in other determinants of the volume of default swap spreads)37 is a shifter—presumably it does loans will shift these curves. For example, if banks’ not affect the demand for loans by borrowers. The funding costs rise, they will tend to supply fewer banks’ capital-to-total-assets ratio (banking regulations loans at an unchanged interest rate, so the supply impose certain capital requirements on banks, affecting curve will shift left. If the determinants of demand their ability to lend) is another supply shifter.38 do not change, then the equilibrium interest rate will Potential demand shifters are also included in the rise and the volume of loans will fall. Similarly, if the model. The rate of capacity utilization affects firms’ demand for loans contracts (as a result of a reduction decisions to invest and consequently their demand in economic activity, for instance), then the demand for credit. The availability of other sources of financ- curve will shift downward. In the new equilibrium, the ing, especially market financing, will also determine lending rate will fall, as will the volume of loans. firms’ demand for bank loans, to the extent that debt The shifts in the demand and supply curves cannot be issuance and bank loans are substitutes from the firm’s observed directly, but if underlying factors can be found point of view.39 that shift one and not the other, the supply and demand Other variables affecting both the supply of and equations can be traced out—or “identified”—sepa- demand for bank lending are included in both equa- rately. Those variables are referred to as “shifters” because tions. Table 2.5 in the main text includes a column they move one or the other curve, as in Figure 2.14. Finding shifters is an econometric challenge owing to 37 the many variables that affect both curves, and if both Credit default swap spreads affect the cost of wholesale funding for banks, but are available only for a few banks in each country curves shift simultaneously, neither one is identified. The (which may not necessarily be representative of that country’s entire proper identification of the model is further complicated banking sector) and have been available only for the past few years. by the potential endogeneity of shifters. These data were used only when the resulting sample reduction did not prevent a proper identification of the model. There are several potential shifters for the supply 38The results for Japan, Spain, and the United Kingdom are robust curve. As suggested earlier, the cost of funding for to using the bank price-to-book ratio instead of the capital-to-asset ratio. However, this variable, which is more volatile than the ratio based on accounting data and reflects the condition of listed banks The author of this annex is Frederic Lambert. only, does not allow for proper identification of the model in the 35Theoretically, repayments of previously granted loans should case of France. not be deducted from new loans. However, because data on gross 39The availability of other financing is proxied by the average flows of bank loans are not available, the empirical analysis uses net outstanding debt securities issued by nonfinancial firms as a share of transaction flows or changes in stocks as a proxy for new loans. total nonfinancial corporate debt. It is computed over the previous 36Market failures, such as maturity mismatches and informational four quarters to limit the endogeneity bias that may result from asymmetries, will add certain surcharges (or premiums) to the risk- firms’ recourse to capital market financing in response to a contrac- free short-term interest rate (for example, a term premium and a risk tion in the supply of bank loans, while still capturing recent progress premium). Equilibrium interest rates contain such premiums. in the development of corporate bond markets.

98 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

with the expected influence (sign) of each variable on Figure 2.15. Fitted Supply and Demand Curves for Bank either the supply or demand, or both. Loans to Firms

• GDP forecasts are expected to be positively related Demand before September 2008 Supply before September 2008 to both loan supply (higher future output imply- Demand after September 2008 Supply after September 2008 ing a greater ability of borrowers to repay) and loan 6 1. France 2. Spain 6 demand (higher expected output encouraging firms to borrow to invest). 5 5 • An increase in economic uncertainty (represented by the standard deviation of the GDP forecast) has the 4 4 r new loans to NFCs loans new r r new loans to NFCs loans new r fo opposite effect. Inflation is expected to negatively fo g rate rate g affect the supply of loans and positively affect demand 3 3 rate g ndin because it reduces the real value of debt over time. ndin Le Le • Growth in the stock market index (covering finan- 2 2 cial and nonfinancial firms) is used as a proxy for –10,000 –5,000 0 5,000 10,000 –10,000 0 10,000 20,000 Bank loans to NFCs Bank loans to NFCs changes in the value of collateral that firms can use (net monthly ﬂow, millions of euros) (net monthly ﬂow, millions of euros) to secure loans; higher collateral value should imply 7 a higher willingness of banks to lend. In addition, 3. United Kingdom 4. Japan 2.0

higher stock values make it easier for banks to raise 6 new capital for lending. It also makes it easier for 1.7 firms to raise new capital for investment without 5 having to borrow. The variable should thus be 1.4 4 positively associated with the supply of loans but 1.1 negatively with the demand for loans. 3 Lending rate for new loans to NFCs • The debt-to-equity ratio and profitability of firms, Lending rate for new loans to NFCs along with corporate spreads, are used to capture 2 0.8 –5,000 0 5,000 10,000 15,000 20,000 –1,000 –500 0 500 1,000 1,500 2,000 the quality of the pool of borrowers: higher debt Bank loans to NFCs Bank loans to NFCs to equity and higher corporate spreads should be (net monthly flow, millions of pounds) (net monthly flow, billions of yen) associated with reduced lending from banks, while higher firm profitability should increase credit sup- Source: IMF staff. Note: NFC = nonfinancial corporation. The plots show the fitted supply and demand curves ply. Higher debt may also reduce the demand for before and after the collapse of Lehman Brothers in September 2008, using the coefficients additional loans (the debt overhang effect discussed estimated over the full sample period from Table 2.5 and assuming that the explanatory variables equal their means over the two separate periods. Light shades of red and blue earlier), whereas higher profitability increases the indicate that the slope is not statistically significant. amount of resources available for self-financing, thus limiting the need for bank lending. Higher GDP forecasts and changes in the stock market index corporate spreads indicate a higher market funding (which reflect markets’ expectations about the future) cost, which should lead firms to prefer bank credit, are likely affected by the ability of firms to get funding thereby raising bank credit demand. to finance their activities. The system of two equations is estimated on coun- Because finding appropriate demand and supply try-level data by three-stage least squares. The sample shifters at a monthly or quarterly frequency is a period varies depending on the country. The longest challenge, data availability restricted the sample of period covers a little more than 10 years, from Febru- countries significantly. For some countries, conceptu- ary 2003 to March 2013. All variables are monthly ally appropriate demand shifters could be identified, except those relating to debt of nonfinancial corpora- but adequately long time series of sufficient frequency tions, profitability, and capacity utilization, which are could not be found. Highlighting the technical chal- quarterly and are linearly interpolated. The lending lenge of identification, even in some cases in which rate is “instrumented” by all other variables in the data were available, the shifters were not significant system. The potential endogeneity of other regressors in the regressions or other econometric problems is dealt with by lagging some of the variables by one emerged. In the end, results were obtained for France, period. Yet endogeneity issues remain. For example, Japan, Spain, and the United Kingdom.

International Monetary Fund | October 2013 99

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

The plots of the estimated demand and supply with error and should be viewed as purely indicative of curves as functions of the lending rate show how the the direction of movement.41 curves shifted after September 2008 (Figure 2.15). The • The demand curves shift downward in France, Japan, plots are constructed using the coefficients estimated and Spain, indicating that the decline in lending was over the full sample period and the means of the due in large part to a drop in lending demand. For explanatory variables over the two separate periods, as the United Kingdom, data availability restricted the is typically assumed for fitted relationships.40 Because estimation to the postcrisis period. of a shorter sample period for the United Kingdom, • The supply curve also shifts left in Spain and, to a the supply and demand curves are plotted only for much lesser extent, in France, suggesting that part of the period following the Lehman Brothers bank- the decline in lending in those countries reflects less ruptcy (October 2008–December 2012). Because of willingness or ability of banks to lend. This result the way the curves are constructed, the shifts reflect broadly confirms the analysis of the survey data. only changes in the average value of the explanatory The rightward shift of the supply curve in Japan can variables before and after the crisis and not changes be interpreted as reflecting improvement in the Japa- in the relationships between the variables. As with all nese banking sector after 2008 over the earlier part econometric estimations, these curves are estimated of the sample period (which reflects the aftermath of the Japanese banking crisis from the late 1990s through the early 2000s), along with the effect of

40The analysis assumes that the slopes of both the supply and credit support policies and the exceptional monetary demand curves have remained the same over the full sample period policy measures announced since 2008. (the elasticity of supply and demand to interest rates has not changed over time). The results of an alternative specification (not 41In some cases, the coefficient on the lending rate is not signifi- reported) allowing the elasticity to change before and after Septem- cant, so the slope of the curve is particularly uncertain. These curves ber 2008 did not contradict this assumption. are shown with lighter shades in Figure 2.15.

100 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

References ing Survey,” Discussion Paper Series 1: Economic Studies No. 31/2011 (Frankfurt: Deutsche Bank). Abiad, Abdul, Giovanni Dell’Ariccia, and Bin Li, 2011, “Credit- Borio, Claudio, and Mathias Drehmann, 2009, “Assessing the less Recoveries,” IMF Working Paper No. 11/58 (Washing- Risk of Banking Crises—Revisited,” BIS Quarterly Review ton: International Monetary Fund). (March), pp. 29–46. Adrian, Tobias, and Hyun Song Shin, 2013, “Procyclical Lever- Borio, Claudio, and Philip Lowe, 2002, “Assessing the Risk age and Value-At-Risk,” NBER Working Paper No. 18943 of Banking Crises,” BIS Quarterly Review (December), pp. (Cambridge, Massachusetts: National Bureau of Economic 43–54. Research). Caballero, Ricardo J., Takeo Hoshi, and Anil K. Kashyap, 2008, Aiyar, Shekhar, Charles Calomiris, and Tomasz Wieladek, “Zombie Lending and Depressed Restructuring in Japan,” 2012, “Does Macro-Pru Leak? Evidence from a UK Policy American Economic Review, Vol. 98, No. 5, pp. 1943–77. Experiment,” NBER Working Paper No. 17822 (Cam- Calvo, Guillermo, Alejandro Izquierdo, and Ernesto Talvi, 2006, bridge, Massachusetts: National Bureau of Economic “Phoenix Miracles in Emerging Markets: Recovering without Research). Credit from Systemic Financial Crises,” NBER Working Albertazzi, Ugo, and Domenico J. Marchetti, 2010, “Credit Paper No. 12101 (Cambridge, Massachusetts: National Supply, Flight to Quality and Evergreening: An Analysis of Bureau of Economic Research). Bank-Firm Relationships after Lehman,” Economic Working Carbó-Valverde, Santiago, Francisco Rodríguez-Fernández, and Paper No. 756 (Rome: Banca d’Italia). Gregory Udell, 2012, “Trade Credit, the Financial Crisis, Allen, Franklin, and Douglas Gale, 2004, “Financial Inter- and Firm Access to Finance,” Funcas Working Paper No. mediaries and Markets,” Econometrica, Vol. 72, No. 4, pp. 683/2012 (Madrid). 1023–61. Chari, Varadarajan, forthcoming, “A Macroeconomist’s Wish Amiti, Mary, and David Weinstein, 2013, “How Much Do List of Financial Data,” in Risk Topography: Systemic Risk and Bank Shocks Affect Investment? Evidence from Matched Macro Modeling (Cambridge, Massachusetts: National Bureau Bank-Firm Loan Data,” NBER Working Paper No. 18890 of Economic Research). (Cambridge, Massachusetts: National Bureau of Economic ———, Lawrence Christiano, and Patrick J. Kehoe, 2008, Research). “Facts and Myths about the Financial Crisis of 2008,” Federal Angelkort, Asmus, and Alexander Stuwe, 2011, “Basel III Reserve Bank of Minneapolis Working Paper No. 666. and SME Financing” (Zentrale Aufgaben, Germany: Ciccarelli, Matteo, Angela Maddaloni, and José-Luis Peydró, Friedrich-Ebert-Stiftung). 2013, “Heterogeneous Transmission Mechanism: Monetary Arellano, Manuel, and Stephen Bond, 1991, “Some Tests of Policy and Financial Fragility in the Euro Area,” European Specification for Panel Data: Monte Carlo Evidence and an Central Bank Working Paper No. 1527 (Frankfurt). Application to Employment Equations,” Review of Economic Claessens, Stijn, M. Ayhan Kose, and Marco E. Terrones, 2012, Studies, Vol. 58, No. 2, pp. 277–97. “How Do Business and Financial Cycles Interact?” Journal of Ashcraft, Adam, 2005, “Are Banks Really Special? New Evidence International Economics, Vol. 87, pp. 178–90. from the FDIC-Induced Failure of Healthy Banks,” American Columba, Francesco, Leonardo Gambacorta, and Paolo Emilio Economic Review, Vol. 95, No. 5, pp. 1713–32. Mistrulli, 2010, “Mutual Guarantee Institutions and Small Bank for International Settlements, 2013, Long Series on Credit Business Finance,” Journal of Financial Stability, Vol. 6, No. to Private Non-Financial Sectors (Basel). 1, pp. 45–54. Bank of England, 2013, Trends in Lending (London, April). Dagher, Jihad, and Kazim Kazimov, 2012, “Banks’ Liability Bassett, William F., Mary Beth Chosak, John C. Driscoll, and Structure and Mortgage Lending during the Financial Crisis,” Egon Zakrajesek, 2012, “Changes in Bank Lending Standards IMF Working Paper No. 12/155 (Washington: International and the Macroeconomy,” Federal Reserve Board Finance and Monetary Fund). Economics Discussion Paper No. 2012–24 (Washington: De Bondt, Gabe, Angela Maddaloni, José-Luis Peydró, and Board of Governors of the Federal Reserve System). Silvia Scopel, 2010, “The Euro Area Bank Lending Survey Beer, C., and W. Waschiczek, 2012, “Analyzing Corporate Loan Matters: Empirical Evidence for Credit and Output Growth,” Growth in Austria Using Bank Lending Survey Data: Con- European Central Bank Working Paper No. 1160 (Frankfurt). ceptual Issues and Some Empirical Evidence,” Monetary Policy Del Giovane, Paolo, Ginette Eramo, and Andrea Nobili, 2011, and the Economy, Vol. 2012, No. 2, pp. 61–80. “Disentangling Demand and Supply in Credit Developments: Bisin, Alberto, and Piero Gottardi, 2006, “Efficient Competi- A Survey-Based Analysis for Italy,” Journal of Banking and tive Equilibria with Adverse Selection,” Journal of Political Finance, Vol. 35, pp. 2719–32. Economy, Vol. 114, No. 3, pp. 485–516. Deutsche Bundesbank, 2012, “Long-Term Developments in Blaes, Barno, 2011, “Bank-Related Loan Supply Factors During Corporate Financing in Germany—Evidence Based on the the Crisis: An Analysis Based on the German Bank Lend- Financial Accounts,” Monthly Report, pp. 13–27.

International Monetary Fund | October 2013 101

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Donovan, Colleen, and Calvin Schnure, 2011, “Locked in the International Monetary Fund (IMF), 2010, Financial Sector House: Do Underwater Mortgages Reduce Labor Market Taxation: the IMF’s Report to the G-20 and Background Mate- Mobility?” (Washington: Freddie Mac and National Associa- rial, Report for the G-20 (Washington). tion of Real Estate Investment Trusts). ———, 2012a, Ireland—Selected Issues, IMF Country Report Drees, Burkhard, and Ceyla Pazarbasioglu, 1998, The Nordic Bank- No. 12/265 (Washington). ing Crisis: Pitfalls in Financial Liberalization, IMF Occasional ———, 2012b, Japan: Financial Sector Stability Assessment Paper No. 161 (Washington: International Monetary Fund). Update, IMF Country Report No. 12/210 (Washington). Drehmann, Mathias, Claudio Borio, and Kostas Tsatsaronis, ———, 2012c, “The Key Attributes of Effective Resolution 2011, “Anchoring Countercyclical Capital Buffers: The Role Regimes for Financial Institutions: Progress to Date and Next of Credit Aggregates,” BIS Working Paper No. 355 (Basel: Steps” (Washington). Bank for International Settlements). ———, 2013a, Euro Area Policies: Selected Issues Paper, IMF European Central Bank (ECB), 2013, “Survey on the Access to Country Report No. 13/232 (Washington). Finance of Small and Medium-Sized Enterprises in the Euro ———, 2013b, “Financing Future Growth: The Evolving Role Area” (Frankfurt). of Banking Systems in CESEE,” Central, Eastern and South- Financial Stability Board (FSB), 2011, “Key Attributes of Effec- eastern Europe: Regional Economic Issues (Washington). tive Resolution Regimes for Financial Institutions” (Basel). ———, 2013c, France: Staff Report for the 2013 Article IV Con- Fraser, Stuart, 2012, “The Impact of the Financial Crisis on sultation, IMF Country Report No. 13/251 (Washington). Bank Lending to SMEs” (London: Economic and Social ———, 2013d, Portugal, Selected Issues, IMF Country Report Research Council). No. 13/19 (Washington). Gan, Jie, 2007a, “Collateral, Debt Capacity, and Corporate ———, 2013e, Spain: Financial Sector Reform – Third Progress Investment: Evidence from a Natural Experiment,” Journal of Report, IMF Country Report No. 13/205 (Washington). Financial Economics, Vol. 85, pp. 709–34. ———, 2013f, Spain: Staff Report for the 2013 Article IV Con- ———, 2007b, “The Real Effects of Asset Market Bubbles: sultation, IMF Country Report No. 13/244 (Washington). Loan and Firm Level Evidence of a Lending Channel,” ———, 2013g, Spain, Selected Issues, IMF Country Report No. Review of Financial Studies, Vol. 20, pp. 1941–73. 13/245 (Washington). Geanakoplos, John, 2010, “The Leverage Cycle,” in NBER Mac- ———, 2013h, United Kingdom: Staff Report for the 2013 Article IV roeconomics Annual, Vol. 24, ed. by Daron Acemoglu, Ken- Consultation, IMF Country Report No. 13/210 (Washington). neth Rogoff, and Michael Woodford (Chicago: University of ———, 2013i, United States: Selected Issues, IMF Country Chicago Press and National Bureau of Economic Research), Report No. 13/237 (Washington). pp. 1–65. Ivashina, Victoria, and David Scharfstein, 2010, “Bank Lending Gertler, Mark, and Peter Karadi, 2012, “QE 1 vs. 2 vs. 3… : during the Financial Crisis of 2008,” Journal of Financial A Framework for Analyzing Large Scale Asset Purchases as a Economics, Vol. 97, No. 3, pp. 319–38. Monetary Policy Tool,” International Journal of Central Bank- Iyer, Rajkamal, Samuel Da-Rocha-Lopes, Jose-Luis Peydró, and ing, Vol. 9. No. S1, pp. 5–53. Antoinette Schoar, 2013, “Interbank Liquidity Crunch and Gilchrist, Simon, and Egon Zakrajsek, 2012, “Credit Spreads the Firm Credit Crunch: Evidence from the 2007–2009 and Business Cycle Fluctuations,” American Economic Review, Crisis,” Barcelona GSE Working Paper Series No. 687. Vol. 102, No. 4, pp. 1692–720. Jermann, Urban, and Vincenzo Quadrini, 2012, “Macroeco- Guerrieri, Veronica, and Guido Lorenzoni, 2011, “Credit Crises, nomic Effects of Financial Shocks,” American Economic Precautionary Savings, and the Liquidity Trap,” NBER Work- Review, Vol. 102, No. 1, pp. 238–71. ing Paper No. 17583 (Cambridge, Massachusetts: National Jiménez, Gabriel, Atif R. Mian, José-Luis Peydró, and Jesús Bureau of Economic Research). Saurina, 2011, “Local Versus Aggregate Lending Channels: Hempell, Hannah Sabine, and Christoffer Kok Sørensen, 2010, The Effects of Securitization on Corporate Credit Supply in “The Impact of Supply Constraints on Bank Lending in the Spain,” Proceedings (May), Federal Reserve Bank of Chicago, Euro Area—Crisis Induced Crunching?” ECB Working Paper pp. 210–20. No. 1262 (Frankfurt: European Central Bank). Jiménez, Gabriel, Steven Ongena, José-Luis Peydró, and Jesús Hennessy, Christopher, 2004, “Tobin’s Q, Debt Overhang and Saurina, 2012, ”Credit Supply and Monetary Policy: Identify- Investment,” Journal of Finance, Vol. 59, No. 4, pp. 1717–42. ing the Bank Balance-Sheet Channel with Loan Applica- Honohan, Patrick, 2010, “Partial Credit Guarantees: Principles and tions,” American Economic Review, Vol. 102, No. 5, pp. Practice,” Journal of Financial Stability, Vol. 6, No. 1, pp. 1–9. 2301–326. Hristov, Nikolay, Oliver Hülsewig, and Timo Wollmershäuser, Jones, Bradley, Peter Lindner, Miguel Segoviano, and Takahiro 2012, “Loan Supply Shocks during the Financial Crisis: Evi- Tsuda, forthcoming, “Securitization: Lessons Learned and the dence for the Euro Area,” Journal of International Money and Road Ahead,” IMF Working Paper (Washington: Interna- Finance, Vol. 31, pp. 569–92. tional Monetary Fund).

102 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChAPTER 2 ASSESSING POLICIES TO REVIVE CREDIT MARKETS

Kalemli-Ozcan, Sebnem, Herman Kamil, and Carolina Villegas- Ongena, Steven, José-Luis Peydró, and Neeltje van Horen, 2013, Sanchez, 2010, “What Hinders Investment in the Aftermath “Shocks Abroad, Pain at Home? Bank-Firm Level Evidence of Financial Crises: Insolvent Firms or Illiquid Banks?” on Financial Contagion during the Recent Financial Crisis,” NBER Working Paper No. 16528 (Cambridge, Massachu- Center for Economic Research Discussion Paper No. 2013- setts: National Bureau of Economic Research). 040 (Tilburg, Netherlands: Tilburg University). Kang, Jong-ku, and Hyung-Kwon Jeong, 2006, “Effectiveness Organization for Economic Cooperation and Development of Policy Measures for Lending to SMEs,” Financial System (OECD), 2012, “Financing SMEs and Entrepreneurs: An Review [Korean], Vol. 250 (Seoul: Bank of Korea). OECD Scoreboard” (Paris). Karaivanov, Alexander, Sonia Ruano, Jesus Saurina, and Robert ———, 2013, “Financing SMEs and Entrepreneurs: An OECD Townsend, 2010, “No Bank, One Bank, Several Banks: Does Scoreboard” (Paris). It Matter for Investment?” Working Paper No. 1003 (Madrid: Peek, Joe, and Erick S. Rosengren, 2000, “Collateral Damage: Effects Bank of Spain). of the Japanese Bank Crisis on Real Activity in the United States,” Kim, Hyeon-Wook, 2005, “The Profitability Improving Effects American Economic Review, Vol. 90, No. 1, pp. 30–45. of Korean SME Policy Lending Programs,” KDI Journal of Petersen, Mitchell, and Raghuram Rajan, 1994, “The Benefits of Economic Policy, Vol. 27, No. 2, pp. 45–88. Lending Relationship: Evidence from Small Business Data,” Kiyotaki, Nobuhiro, and John Moore, 1997, “Credit Cycles,” Journal of Finance, Vol. 49, No. 1, pp. 3–37. Journal of Political Economy, Vol. 105, No. 5, pp. 211–48. Prescott, Edward C., and Robert M. Townsend, 1984a, “General Klapper, Leora, Luc Laeven, and Raghuram Rajan, 2012, “Trade Competitive Analysis in an Economy with Private Informa- Credit Contracts,” Review of Financial Studies, Vol. 25, No. 3, tion,” International Economic Review, Vol. 25, No. 1, pp. pp. 838–67. 1–20. Lacroix, Renaud, and Jérémi Montornès, 2010, “Analysis of the ———, 1984b, “Pareto Optima and Competitive Equilibria Scope of the Results of the Bank Lending Survey in Rela- with Adverse Selection and Moral Hazard,” Econometrica, Vol. tion to Credit Data,” Bank of France, Quarterly Selection of 52, No. 1, pp. 21–45. Articles, No. 16, pp. 33–51. Rajan, Raghuram G., 1992, “Insiders and Outsiders: The Choice Laeven, Luc, and Thomas Laryea, 2009, “Principles of House- between Informed and Arm’s-Length Debt,” Journal of hold Debt Restructuring,” IMF Staff Position Note No. Finance, Vol. 47, No. 4, pp. 1367–400. 09/15 (Washington: International Monetary Fund). Sharpe, Steven A., 1990, “Asymmetric Information, Bank Lend- Lam, Raphael, and Jongsoon Shin, 2012, “What Role Can ing and Implicit Contracts: A Stylized Model of Customer Financial Policies Play in Revitalizing SMEs in Japan?” IMF Relationships,” Journal of Finance, Vol. 45, No. 4, pp. Working Paper No. 12/291 (Washington: International 1069–87. Monetary Fund). Sugawara, Naotaka, and Juan Zalduendo, 2013, “Credit-Less Landier, Augustin, and Kenichi Ueda, 2009, “The Economics of Recoveries, Neither a Rare nor an Insurmountable Chal- Bank Restructuring: Understanding the Options,” IMF Staff lenge,” World Bank Policy Research Working Paper No. 6459 Position Note No. 09/12 (Washington: International Monetary (Washington). Fund). Townsend, Robert M., 1979, “Optimal Contracts and Com- Laryea, Thomas, 2010, “Approaches to Corporate Debt Restruc- petitive Markets with Costly State Verification,” Journal of turing in the Wake of Financial Crises,” IMF Staff Position Economic Theory, Vol. 21, No. 2, pp. 265–93. Note No. 10/02 (Washington: International Monetary Fund). ———, 1982, “Optimal Multiperiod Contracts and the Gain Lee, Neil, Hiba Sameen, and Lloyd Martin, 2013, “Credit and from Enduring Relationships under Private Information,” the Crisis: Access to Finance for Innovative Small Firms Journal of Political Economy, Vol. 90, No. 6, pp. 1166–86. since the Recession” (Lancaster, United Kingdom: Work Valencia, F., 2012, “Credit Supply Shocks in the Euro Area” Foundation). (unpublished; Washington: International Monetary Fund). Lown, Cara, and Donald P. Morgan, 2006, “The Credit Cycle Yi, Jong-Goo, 2012, “Towards a Better Framework for Supply and the Business Cycle: New Findings Using the Loan Officer of Funds,” in Asian Financial Markets: Lessons from Korea,” Opinion Survey,” Journal of Money, Credit and Banking, Vol. presentation prepared for the International Conference on 38, No. 6, pp. 1576–97. “Asian Market Integration and Financial Innovation,” Tokyo, Mendoza, Enrique G., and Marco E. Terrones, 2008, “An Anat- February 10. omy of Credit Booms: Evidence From Macro Aggregates and Zoli, Edda, 2013, “Italian Sovereign Spreads: Their Determi- Micro Data,” NBER Working Paper No. 14049 (Cambridge, nants and Pass-through to Bank Funding Costs and Lending Massachusetts: National Bureau of Economic Research). Conditions,” IMF Working Paper No. 12/84 (Washington: Myers, Stewart, 1977, “Determinants of Corporate Borrowing,” International Monetary Fund). Journal of Financial Economics, Vol. 5, pp. 147–75.

International Monetary Fund | October 2013 103

©International Monetary Fund. Not for Redistribution This page intentionally left blank

©International Monetary Fund. Not for Redistribution ChAngeS in BAnk fUnding pAtteRnS And chapter 3 finAnCiAl StABilitY RiSkS

SUMMARY

unding structures matter for financial stability. In particular, overreliance by some banks on certain types of wholesale funding—especially by U.S. and European banks—contributed to the global financial crisis. Most banks have recently made their funding structures more resilient by raising their capital adequacy ratios and reducing their dependence on short-term wholesale funding. However, some distressed banks Fremain vulnerable because their equity capital levels are inadequate and they are highly dependent on central bank funds. This chapter examines how bank funding structures have changed over time—especially in the run-up to the crisis—and how these structures affect financial stability. The analysis considers banks in a number of advanced and emerging market economies and includes systemically important banks. The analysis shows that healthy banks rely more on equity and less on debt (especially short-term debt) and have more diversified funding structures with lower loan-to-deposit ratios. Adequate capital buffers reduce a bank’s probability of default and support financial stability. Therefore, Basel III capital regulations that aim to raise the quantity and quality of capital should continue to be a mainstay of the reform efforts. Basel III liquidity regulations will also play a role by reducing banks’ overreliance on short-term wholesale funding, which has proven detrimental to financial stability. Current reform efforts are aimed at reducing financial instability, but there can be tension among some key regulatory reforms that affect bank funding structures. As this chapter shows, such tension can arise, on the one hand, from pressures to use more secured funding (thereby raising levels of asset encumbrance) as well as deposits and, on the other hand, from bank-resolution initiatives (including the introduction of bail-in powers and the prospects for additional depositor preference) that are designed to reduce the burden on taxpayers while also protecting depositors. A numerical example examines funding costs under various scenarios. The analysis suggests that the effects may not be large under current conditions but that they depend importantly on the share of protected creditors and the size of equity buffers. Careful implementation of the reform efforts can help mitigate potential trade-offs so as to ensure that the financial stability benefits are realized. In particular, Basel III and over-the-counter (OTC) derivatives reforms should be implemented as planned. However, policymakers will want to monitor the increased demand for collateral (including from new liquidity standards and OTC derivatives reforms) to ensure that there are enough unencumbered assets to meaningfully attract senior unsecured creditors. Going forward, limits on asset encumbrance or minimum proportions of bail-in debt relative to assets may be required so that a sufficiently large proportion of unsecured debt is preserved to absorb losses when bank capital is exhausted as an important protection against future use of taxpayer funds. The introduction of such changes, however, will need to be mindful of funding market conditions to ensure that they are not introduced during periods of funding difficulties.

International Monetary Fund | October 2013 105

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY introduction with deposits. Also, banks that face higher currency volatility, stronger regulatory frameworks, and stricter The global financial crisis revealed the risks to financial disclosure requirements rely less on wholesale funding. stability arising from banks’ reliance on certain types Banks’ funding structures affect their stability, and of wholesale funding.1 Before the financial crisis, many although most banks have improved their funding U.S. and European banks relied on wholesale debt structures since the crisis began, distressed banks remain funding to expand asset growth. Since the crisis, these vulnerable. More equity and less debt (in particular less private market funds have diminished in size, whereas short-term debt), lower loan-to-deposit ratios, and more collateralized borrowing, including covered bond issu- diversified funding structures improve banks’ stability. ance and central bank funding, has risen (especially in Since the crisis began, banks around the world have Europe). Counterparty risk has prompted the grow- raised their capital adequacy ratios, reduced wholesale ing use of secured funding, pushing up the share of funding, and in some cases raised more deposits, all of assets pledged as collateral for liabilities (termed “asset which have improved their stability. However, distressed encumbrance”). At the same time, new regulations banks’ funding structures have not similarly improved, are being proposed or implemented that aim to make and they continue to be vulnerable. financial systems safer (including Basel III capital and There are potential tensions among some regula- liquidity regulations and over-the-counter [OTC] tory reforms, including regulations designed to increase derivatives reforms) and to improve bank resolution resilience to short-term liquidity shocks, measures to mechanisms (for example, bail-in powers and deposi- improve crisis management, and proposals to facilitate tor preference). This chapter examines funding market bank resolutions without the use of taxpayer support. developments and the implications of the reform Increasing banks’ equity capital, as intended by Basel III efforts for bank funding structures and their costs. capital regulations, reduces the cost of any type of debt by Against this backdrop, this chapter examines the increasing loss-absorbing buffers before any debt hold- following questions: ers face losses, and Basel III liquidity regulations should • What determines bank funding structures, and how help maintain liquidity buffers. Both measures should have they changed? improve financial stability. However, continuing weakness • How did funding structures relate to banks’ stabil- in bank funding markets (particularly in Europe), OTC ity in the run-up to the crisis? Have bank funding derivatives reforms, and some aspects of Basel III liquidity structures changed so as to improve financial stabil- regulation may encumber more assets, thereby increas- ity since the crisis began? ing unsecured bondholders’ potential losses. Unsecured • How will key regulatory initiatives affect bank bondholders may also face larger losses if (1) a country funding structures? What are the potential tensions introduces new depositor preferences for bank closures (in among the initiatives, if any? which case some or all retail depositors will be paid ahead • Considering the outcomes of various reforms, how of other unsecured creditors), and (2) the bondholders are will funding costs likely develop? bailed in when a bank is restructured (that is, they assume The analysis shows that banks have diverse, but more of the losses than do creditors that cannot be bailed slow-to-change, funding patterns. Larger banks in in). When the risk of losses rises (including from the fear advanced economies, excluding Japan, rely more on of being bailed in), the costs of unsecured debt also rise wholesale funding, whereas those in Japan and most because this class of investors will require higher returns emerging market economies fund themselves primarily (holding all else constant). To the extent that the possibility of bail-in removes the too-big-to-fail perception for The authors of this chapter are Brenda González-Hermosillo and Hiroko Oura (team leaders), along with Jorge Chan-Lau, Tryggvi systemically important institutions, some of the implicit Gudmundsson, and Nico Valckx. Other contributors include Serkan funding subsidy that they have received may be removed, Arslanalp, Marc Dobler, Alvaro Piris Chavarri, Lev Ratnovski, Taka- potentially raising overall funding costs to more appropri- hiro Tsuda, and Mamoru Yanase. Research support was provided by ate levels. However, some banks may find it difficult to Oksana Khadarina. 1See Chapter 2 of the October 2010 Global Financial Stability issue enough senior unsecured debt to ensure this market Report (GFSR) for developments in bank funding markets during discipline role, and if holders of this class of debt are less the global financial crisis. Berkmen and others (2012) and Chapter 4 tolerant when bank distress is imminent, then financial of the October 2012 GFSR show that banks that funded themselves with nondeposit liabilities fared worse during the financial crisis, and instability may ensue. Despite this proviso, overall, the their countries experienced weaker growth outcomes. introduction of bail-in powers alongside greater transpar-

106 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

ency is likely to make funding costs better reflect the risks • Regulatory reforms can affect bank funding structures of banking and hence enhance financial stability. both positively and negatively, so, these reforms need to A numerical exercise shows that some funding be calibrated carefully. Policymakers must be particu- structure configurations (including equity) and the larly watchful to ensure that the reforms—including order of creditor seniority can substantially alter the OTC derivatives reforms—do not encourage banks to cost of debt—perhaps in unanticipated ways. Capital- issue or hold certain types of securities that excessively ization and the riskiness of bank assets have a quanti- encumber assets. Incentives arising from regulations tatively large effect on the cost of bank debt. The share that may lead to the overuse of secured funding can be of preferred deposits and liabilities exempted from contained by introducing a maximum proportion of being bailed in (including secured borrowing) is also encumbered assets. To reap the benefits of the resolu- an important driver of the cost of unsecured (bail-in) tion reforms, policymakers will need to ensure that debt, which rises disproportionally more than increases the amount of bail-in debt is sufficient to induce these in these other components in the funding structure. debt holders to exercise market discipline and thereby There are two key policy messages from the analysis: to encourage safer banking. Hence, a minimum bail-in • Funding structures matter for financial stability because requirement may be necessary. a healthy funding structure lowers the probability that a bank will fall into distress. Adequate capital buffers Bank funding Structures: determinants and reduce the probability of default and, all else equal, Recent developments improve the chances that depositors and debt holders are repaid their funds. Hence, Basel III capital regula- What determines Bank funding Structures? tions aimed at raising the quantity and quality of capi- The empirical analysis shows that banks have very tal should continue to be the mainstay of the reform diverse funding structures and that, in general, these effort. The Basel III liquidity regulations will also play change only gradually. Modern banks use various forms a role by reducing the use of short-term wholesale of funding instruments other than deposits (Box 3.1). funding—a component of funding that the analysis These vary substantially across banks and regions (Fig- shows to be detrimental to financial stability. ures 3.1 and 3.2). Advanced economies, except Japan,

Figure 3.1. Banks’ Liability Structures across Major Economies and Regions (Percent)

Other Total equity Subordinated debt Senior debt Bank deposits Customer deposits

100

0 08 12 08 12 08 12 08 12 08 12 08 12

2005 United States 2005 Japan 2005 Euro area 2005 Other AE 2005 EM CEE 2005 EM Asia

Sources: Bank of Japan; SNL Financial; and IMF staff estimates. Note: "Other" includes ﬁnancial and accounting liabilities, such as derivatives liabilities, insurance liabilities, noncurrent liabilities, accounts payable and accrued expenses, deferred taxes and tax liabilities, and other provisions. Japan data for 2005 omitted due to data limitations. AE = advanced economies; CEE = central and eastern Europe; EM = emerging market economies. For region coverage, refer to Table 3.2.

International Monetary Fund | October 2013 107

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 3.1. typology of Bank funding

Bank funding sources can be distinguished by inves- times of distress. For example, the cost of interbank tor type, instrument type, and priority (Figure 3.1.1). loans (for example, the London interbank offered rate) Customer deposits are the main funding source for rose dramatically, and the issuance of CP and CDs banks that have traditional deposit-taking and loan- dropped sharply following the Lehman Brothers failure. making business models. • Short-term secured funds include repurchase agreements • Deposits payable at par and “on demand” carry the (repos), swaps, and asset-backed commercial paper. most liquidity risk because of their maturity mis- These were considered safe before the crisis, but suffered match with longer-term loans, and they could be a run in its early stages. Reuse of collateral (rehypotheca- subject to runs. However, in practice, retail deposits tion) also contributed to increasing the interconnected- are relatively stable, particularly if covered by a cred- ness among financial institutions that were using repos. ible deposit guarantee scheme. • Long-term funds include bonds and various forms of • Other types of deposits can be less stable, including securitization (including covered bonds and private- uninsured deposits, foreign currency deposits, depos- label mortgage-backed securities). These instruments its collected though Internet banking, and those are less likely to cause immediate funding difficulties. collected from nonresidents, corporations, money Capital, as defined by Basel III, absorbs incurred losses market funds, and high-net-worth individuals.1 before any other creditors (see BCBS, 2010a for details). • Regulatory capital includes common equity and Wholesale funds are often used for investments in certain types of subordinated debt. The highest qual- financial assets, including those used in the bank’s ity (that with the highest loss-absorbing capacity) is proprietary trading. known as common equity Tier 1 (CET1) capital, • Assets secured as collateral (and thus “encumbered”) which is mostly in the form of common equity. Cer- are designated for paying secured creditors first. tain types of subordinated debt, which are paid after Senior unsecured wholesale funds may rank equal other debt holders, also qualify as additional Tier 1 or to depositors or below depositors in countries with Tier 2 capital, including contingent convertible debt depositor preference. (CoCos), preferred shares, and perpetual bonds.2 • Short-term unsecured funds include some interbank loans, commercial paper (CP), and wholesale certificates 2Preferred shares are senior to common equity and usually carry no of deposit (CDs). These funds can be volatile during voting rights, but receive dividends before common equity. CoCos are bonds that would be converted into common equity when the regula- 1Uninsured deposits include those eligible for a deposit guar- tory capital ratio reaches a prespecified threshold. See Pazarbasioglu antee scheme, but exclude covered deposits (for example, retail and others (2011) on the economic rationale for introducing CoCos. deposits exceeding the maximum coverage) and ineligible deposits. See Barclays (2013) for a list of existing CoCos and their structures.

Figure 3.1.1. Breakdown of Bank Liabilities By Investor Type By Instrument By Priority Secured debt Stable deposits, including insured deposits Customer deposits ST: repo, swap, LT: covered Less stable deposits, including uninsured, FX, ABCP bonds, MBS Internet, HNW individual deposits

Unsecured: Senior unsecured debt interbank deposits, CP, CD Short term (ST) Secured: repo (including CB), swap, ABCP Wholesale funding Unsecured: Deposits

senior unsecured bonds interbank ST:

Long term (LT) CD CP, deposits,

Secured: unsecured bonds LT: covered bonds, ABS, MBS

Subordinated debt, including preferred shares, CoCo, Junior debt Regulatory capital perpetual bonds (retail/wholesale) Common equity Equity

Source: IMF staff. Note: ABCP = asset-backed commercial paper; ABS = asset-backed securities; CB = central bank; CD = certiﬁcate of deposit; CoCo = contingent convertible; CP = commercial paper; FX = foreign exchange; HNW = high net worth; LT = long term; MBS = mortgage- backed security; ST = short term. The example presented here is for an economy without deposit preference.

108 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Figure 3.2. Banks’ Long-Term Wholesale Funding across Major Economies and Regions: Outstanding as of July 31, 2013

1. Structure of Bank Debt 2. Structure of Secured Bank Debt (billions of U.S. dollars and percent of GDP) (percent of total)

Secured Tier 2 Tier 3 MBS High-yield corporate Medium-term notes Senior unsecured Tier 1Total, % GDP; right scale Covered bonds Investment-grade corporate ABS 4,000 30 100

90 3,500 25 80 3,000 70 20 2,500 60

2,000 15 50

40 1,500 10 30 1,000 20 5 500 10

0 0 0 area area Euro Euro Japan Japan States States United United EM CEE EM CEE EM Asia EM Asia EM Other AE Other AE Other EM other EM other

Sources: Dealogic; IMF, World Economic Outlook database; and IMF staff calculations. Note: ABS = asset-backed securities; AE = advanced economies; CEE = central and eastern Europe; EM = emerging market economies; MBS = mortgage-backed securities.

typically rely more on wholesale funding; Japan, by con- indication of the need for wholesale funding) are stud- trast, has an ample retail deposit base. Even for whole- ied for 751 banks, applying a dynamic panel regression sale funding, there is significant variation among banks, with bank-specific fixed effects for a large set of coun- with a few (the 90th percentile) using a preponderance tries (see Annex 3.1 for details).3 The roles of bank- of noncore funding (debt as a proportion of equity and specific factors are examined along with country-level deposits—Figures 3.3 and 3.4). Banks in emerging mar- macroeconomic, financial market, and regulatory and ket economies also fund themselves primarily with retail institutional factors. The sample is also split between deposits and are much more homogeneous in their use advanced economies and emerging market economies of various funding instruments than advanced economy and across specific periods. Systemically important banks. European banks are the largest issuers of bank banks are considered separately.4 bonds, especially covered bonds, both in absolute terms In line with earlier studies, the empirical evidence and relative to GDP.2 Despite some movements in non- suggests that bank funding is affected mainly by bank- core versus core funding instruments, on average, bank specific factors and to a lesser extent by macrofinancial funding structures change only gradually over time. To better understand how banks choose their fund- 3 ing structures and thus how they can be made more Some studies look at different specifications of funding, express- ing total liabilities or deposit and nondeposit liabilities as shares of resilient, we examine the factors influencing these banks’ market value (that is, more as indicators or components of structures between 1990 and 2012. The composition market leverage). However, this approach neglects the role of equity of the liability structure (equity, nondeposit liabilities, as a separate funding instrument. In addition, using market value restricts the analysis to listed banks. and deposits) as well as the loan-to-deposit ratio (an 4The subsample comprises 27 global and 84 domestic systemically important banks (global systemically important banks—G-SIBs— 2U.S. banks in the SNL Financial sample include only deposit- and domestic systemically important banks—D-SIBs—respectively). taking institutions, thus excluding broker dealers and various shadow The G-SIBs are those chosen by the Financial Stability Board banks. See Chapter 1 of this report for more information on shadow (2012b), and the selection of D-SIBs is based on whether a bank’s banks. total assets are close to or exceed 20 percent of GDP.

International Monetary Fund | October 2013 109

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.3. Evolution of Bank Funding Structures, Global and Figure 3.4. Evolution of Bank Funding Structures, Systemically Important Banks Advanced and Emerging Market Economies (Percent) (Percent) All Banks Systemically Important Banks Advanced Economy Banks Emerging Market Economy Banks 1. Equity-to-Asset Ratio 2. Equity-to-Asset Ratio 1. Equity-to-Asset Ratio 2. Equity-to-Asset Ratio 20 15 25 12 90th percentile 90th percentile 20 15 10 10 75th percentile 8 75th percentile 15 10 6 Median Median 10 5 25th percentile 25th percentile 4 5 5 10th percentile 2 10th percentile 0 0 0 0 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10 3. Debt-to-Asset Ratio 4. Debt-to-Asset Ratio 3. Debt-to-Asset Ratio 4. Debt-to-Asset Ratio 80 80 80 80

60 60 60 60

40 40 40 40

20 20 20 20

0 0 0 0 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10

5. Deposit-to-Asset Ratio 6. Deposit-to-Asset Ratio 5. Deposit-to-Asset Ratio 6. Deposit-to-Asset Ratio 100 100 100 100

80 80 80 80

60 60 60 60

40 40 40 40

20 20 20 20

0 0 0 0 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10 7. Loan-to-Deposit Ratio 8. Loan-to-Deposit Ratio 7. Loan-to-Deposit Ratio 8. Loan-to-Deposit Ratio 200 250 200 250

200 200 150 150 150 150 100 100 100 100 50 50 50 50

0 0 0 0 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10

9. Noncore-to-Core Funding Ratio 10. Noncore-to-Core Funding Ratio 9. Noncore-to-Core Funding Ratio 10. Noncore-to-Core Funding Ratio 200 300 200 150

250 150 150 200 100

100 150 100

100 50 50 50 50

0 0 0 0 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10 1990 95 2000 05 10

Sources: Bloomberg, L.P.; and IMF staff estimates. Sources: Bloomberg, L.P.; and IMF staff estimates. Note: Figures show the median (black line), interquartile range (red dashed lines), and upper and lower Note: Figures show the median (black line), interquartile range (red dashed lines), and upper and decile (blue solid lines) of the distribution of the share of equity, debt, and deposits as percentages of lower decile (blue solid lines) of the distribution of the share of equity, debt, and deposits as total assets and the loan-to-deposit and noncore-to-core funding ratios (in percent). The latter ratio is percentages of total assets and the loan-to-deposit and noncore-to-core funding ratios deﬁned as debt to equity and deposits. (in percent). The latter ratio is deﬁned as debt to equity and deposits.

110 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

and market variables.5 Institutional factors also seem to Bank funding before and after the global play a role. The key results are illustrated graphically in financial Crisis Figure 3.5:6 Focusing on developments just before the crisis, • Bank-level factors matter most, but regulation also banks, especially in Europe, relied largely on low- plays a role. Bank-specific fixed effects and past cost wholesale funding to expand investments (Box funding structure choices dominate the results. 3.2). U.S. banks rapidly increased interbank loans In contrast to previous studies, this analysis finds (unsecured debt and secured repos; Figure 3.6) and that proxies for the general regulatory environment issued securitized products, albeit from a lower base (including nonfinancial factors such as the “rule than their European counterparts. Japanese banks, of law”) influence bank funding structures.7,8 On however, needed little wholesale funding given their average, over all countries and the entire sample ample deposit base. Emerging market economy banks, period, countries with higher-quality regulations are especially those in central Europe, saw some erosion of associated with banks that have more deposit and their customer deposits in favor of interbank deposits less debt funding. Banks in advanced economies but maintained higher capital ratios (see Figures 3.1 with higher disclosure requirements (holding all else and 3.7). constant) tend to have higher deposit-to-asset ratios The global financial crisis caused substantial stress and lower loan-to-deposit ratios. in wholesale funding markets, forcing banks to adjust • Capital structures are generally highly persistent, but their funding models. In particular, many banks had to the speed of adjustment varies across time and coun- rely on central bank funding to survive systemic liquid- tries. Capital structures appear to be changing, but ity shortages. For banks that had relied on dollar-based only slowly. Equity funding tends to adjust faster funding, currency swap lines were provided by the than debt and deposit funding. However, since Federal Reserve to relieve U.S. dollar liquidity short- 2007, banks have adjusted at a faster and more ages abroad.9 Banks in all regions recapitalized, often similar pace across all types of funding. with government support (see Figure 3.7). Financial • Asset size plays an important role. Large banks gener- fragmentation and bank deleveraging have also affected ally take on more debt (perhaps because investors cross-border bank funding patterns. In particular, there are more familiar with them) and fund using less was a significant decline in foreigners’ investments in equity and deposits. bank-issued debt securities located in the stressed euro • More traditional, safer banks depend less on wholesale area countries of Ireland, Italy, Portugal, and Spain, funding. Banks with more securities and tangible while banks in core euro area countries generally expe- assets and those that pay dividends rely less on rienced the opposite. Changes appear to be smaller in wholesale funding (that is, have lower loan-to- non-euro-area advanced economies (Box 3.3). deposit ratios). Some diverging regional trends are noteworthy: • In Europe, for many banks there continues to be 5Existing studies show that a firm’s size and profitability, whether limited access to private short-term wholesale and it pays dividends, its cash flow volatility (as a measure of risk), and interbank markets. As a substitute, banks have its “tangibility” matter for bank funding. Tangibility for financial firms (such as banks) refers to the value of securities, cash and funds become more reliant on European Central Bank due from banks, fixed assets, and other tangible assets. (ECB) funding and on covered bond issuance, 6 See Gudmundsson and Valckx (forthcoming) for further regional which increases asset encumbrance (Figure 3.8), analysis. 7Based on the World Bank’s Doing Business Indicators of regula- especially during periods of stress (Figure 3.9, tory quality, effectiveness of governance, rule of law, and voice and panel 1). Notably, about 30 percent of covered accountability, two principal components are derived that reflect the bonds issued by European banks are retained by the level of regulation and disclosure. This interpretation is based on correlations and signs with other legal, regulatory, and institutional issuers for potential use as collateral for ECB facili- characteristics. ties (Figure 3.9). 8This conclusion was based on the large impact of bank fixed • U.S. banks have reduced their reliance on secured (for R2 effects on the explanatory power of the model (measured by ) and example, private-label mortgage-backed securities) on the difference in speed of adjustment (1 minus the coefficient of the lagged dependent variable) in a specification with and without and unsecured funding, replacing it with deposits and fixed effects, similar to Gropp and Heider (2010). Unlike Gropp and Heider (2010), in this study regulatory factors appear to help explain 9 See Chapter 3 of the April 2013 GFSR on central bank liquidity the variation in funding structures. support since the crisis.

International Monetary Fund | October 2013 111

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.5. Determinants of Bank Funding (Relative sizes of factors; percentage points)

1. All Banks, 1990–2012 2. All Banks, 2007–12

LDV LDV Size Size Proﬁtability Proﬁtability Total assets growth GDP growth Dividend Spread Collateral FX volatility GDP growth CA-to-GDP ratio FX volatility CA-to-GDP ratio Crisis Crisis Equity Equity Stock market cap. Stock market cap. Bond market cap. Debt Debt Savings Government debt Deposits Deposits Regulation Regulation –3 –2 –1 10 2 3 –2 –1 0 1 2 3

3. Advanced Economy Banks, 1990–2012 4. Emerging Market Economy Banks, 1990–2012

LDV LDV Size Size Total assets growth Equity Dividend Proﬁtability Collateral Debt Dividend GDP growth Deposits Spread Collateral CA-to-GDP ratio GDP growth Crisis Stock market cap. Crisis Equity Bond market cap. Stock market cap. Savings Debt Savings Government debt Deposits Disclosure Disclosure

–3 –2 –1 10 2 3 –2 –1 0 1 2 3

5. Systemically Important Banks, 1990–2012 6. Loan-to-Deposit Ratios, 1990–2012

LDV LDV Size Size Total assets growth Total assets growth GDP growth Dividend Spread Collateral FX volatility NII share CA-to-GDP ratio Equity GDP growth Stock market cap. FX volatility Debt Bond market cap. CA-to-GDP ratio Deposits All banks Savings Stock market cap. Government debt AE Savings Regulation EM Regulation Disclosure Disclosure

–3 –2 –1 10 2 3 –6 –4 –2 420 6 8 10

Sources: Bloomberg, L.P.; and IMF staff estimates. Note: CA = current account; cap. = capitalization; FX = foreign exchange; LDV = lagged-dependent variable; NII share = net interest income in percent of operating income. Regulation and disclosure are the first and second principal component scores, derived from four World Bank indicators of regulatory and institutional quality. See Table 3.3 for further details on factors and their definitions. Figures show the economic relevance of bank characteristics and macrofinancial and regulatory factors on bank funding through equity, deposits, and debt (as a percent of total assets), and on loan-to-deposit ratios (panel 6) based on panel estimations for all banks, advanced economy banks, emerging market economy banks (from developing Asia and central and eastern Europe), and global and domestic systemically important banks. Economic relevance is computed as coefficients multiplied by 1 standard deviation of each variable (averaged across banks). Variables shown are chosen using the general-to-specific selection method, which starts with a general regression model and narrows it down to a model with only significant variables. See Annex 3.1 for further details on data and estimation results.

112 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Box 3.2. What the Crisis taught Us about Bank funding

This box summarizes the leading current research on bank During the crisis, banks hoarded liquidity because funding sources and capital structures, focusing on their of perceived credit and liquidity risks (includ- role for financial stability. The literature demonstrates that ing their own inability to monitor risks) (Heider, bank wholesale funding does not provide sufficient market Hoerova, and Holthausen, 2009; Farhi and Tirole, discipline and is unstable during crises. 2012). • Wholesale funding created complex interactions Since the 1990s, banks have increasingly used between bank assets and liabilities, such that a fall wholesale funding—repurchase agreements (repos), in asset values could compromise banks’ ability to brokered deposits, interbank loans, and commercial obtain funds. Hence, a funding freeze could lead to paper—to supplement retail deposits (Feldman and asset fire sales to generate liquidity. As an alterna- Schmidt, 2001). The precrisis literature generally sug- tive, banks may be encouraged to securitize assets, gested that this trend was advantageous. Unlike retail but may continue to hold them on the balance depositors, the providers of wholesale funding were sheet—instead of selling off the new securities—to thought to be “sophisticated,” that is, able to monitor pledge them in repos for an additional source of and discipline risky banks (Calomiris and Kahn, 1991; funding (Acharya, Gale, and Yorulmazer, 2011; Rochet and Tirole, 1996; Flannery, 1998; Calomiris, Brunnermeier and Pedersen, 2009; Shin, 2009a). 1999) because they were not protected by (explicit) • At a macroeconomic level, variations in the value of deposit insurance schemes. collateral and margin requested, and other funding Yet the crisis revealed wholesale funding to be a market conditions, became a major determinant of major source of instability. In particular: bank leverage and banks’ ability to extend credit • Banks attracted wholesale funds at short maturities (Geanakoplos, 2009; Adrian and Shin, 2010), creat- because they are cheaper than at longer maturities. ing larger boom and bust cycles. Wholesale providers of funding did not adequately • Many empirical studies show that the reliance on monitor banks because they knew they could with- wholesale funding was a major source of bank vul- draw at a hint of negative news by not rolling over nerability during the crisis (Huang and Ratnovski, their funding. During the crisis, collective withdrawals 2009; Shin, 2009b; Demirgüç-Kunt and Huizinga, triggered generalized funding disruptions (Huang and 2010; Goldsmith-Pinkham and Yorulmazer, 2010; Ratnovski, 2011; Brunnermeier and Oehmke, 2013). Bologna, 2011; Vazquez and Federico, 2012). • Banks attracted wholesale funding on a secured In sum, the literature suggests that bank wholesale basis—against the collateral of securitized debt and funding has become an inherent feature of the modern other assets for repo transactions. Sudden concerns financial system. It can be explained as a response to about the quality of collateral led to a freeze of repo financial innovation and a buildup of excess savings in funding markets (“a run on repo,” as described by some countries’ corporate sectors (so-called cash pools) Gorton and Metrick, 2012). as well as by increases in official reserves of many • Wholesale funding made the financial system emerging market economies. However, the literature (not just the banking system) more intercon- highlights that wholesale funding is associated with nected because both bank and nonbank financial some problematic properties, specifically a lack of suf- institutions provided liquidity to each other. The ficient market discipline and instability in crises. An interbank market proved to be particularly fragile. important conclusion is that any regulation designed to counter potential downside risks to wholesale fund- The author of this box is Lev Ratnovski. ing will need to account for potential trade-offs.

International Monetary Fund | October 2013 113

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

equity (see Figure 3.6). The share of net repo funding Figure 3.6. Wholesale Bank Funding for U.S. banks declined from about 8 percent of total 1 liabilities in 2008 to 2 percent in 2013. 1. Cumulative Percentages 70 • In Asian and central and eastern European emerg- Repo Interbank Debt

ing market economies and in Japan, banks have 60 slightly increased wholesale funding since the crisis

began while expanding their balance sheets, but 50 these funding categories remain proportionately less

than in Europe or the United States. While Japanese 40 banks primarily rely on deposits at home, they are

increasingly relying on wholesale funding abroad. 30

Are Bank funding Structures Relevant to 20 financial Stability? 10 Can funding structures that are likely to improve financial stability be empirically identified? The 0 relationship between bank funding characteristics and 07 09 11 08 10 12 07 09 11 08 10 12 07 09 11 2005 2006 2005 2006 2005 bank distress is examined for a broad group of coun- Euro area United States Other AE2 EM Asia and CEE tries from 1990 through 2012 (see Annex 3.1). The Japan characteristics included in the analysis are the stability of the structure (amount of short-term debt subject 2. Secured Senior Debt in Percent of Total Senior Debt to rollover risk), diversity (concentration of banks’ 70 funding via debt, equity, and deposits), asset-liability Euro area AE United States EM CEE EM Other AE2 EM Asia mismatches (loan-to-deposit funding gap), and leverage 60 (debt and equity relative to total assets), in line with Le Leslé (2012). Three separate variables are used to check 50 the sensitivity of the funding structures to various definitions of bank distress: a balance sheet measure of 40 risk (low z-scores),10 an asset-price-based indicator (low price-to-book ratio), and bank equity analysts’ rating 30 (buy or sell) recommendations. As expected, funding characteristics matter for 20 bank distress (Figure 3.10). The results support the view that overall banking-sector stability requires that 10 funding structures be stable, diversified, and involve less leverage. Limiting the mismatch between loans 0 2005 06 07 08 09 10 11 12 and deposits, which reduces the need for wholesale funding, is also important—a finding that is in line with the literature on this topic (see Box 3.2).11 More Sources: Dealogic; SNL Financial; and IMF staff estimates. Note: AE = advanced economies; CEE = central and eastern Europe; EM = emerging market specifically: economies. 1 • Better capitalization (a higher equity-to-asset ratio) Debt, interbank liabilities, and repurchase agreements (repos) as cumulative percent of wholesale funding plus customer deposits. contributes to bank stability for both advanced 2Other AE excludes European Union, Norway, and the United States.

10The z-score is defined as the equity-to-asset ratio plus return on assets (ROA), divided by the standard deviation of ROA. It is a measure of the risk-adjusted ROA, and the higher the z-score, the more resilient the bank. Chapter 3 of the April 2013 GFSR also found positive results using z-scores as a measure of bank-level risk. 11See Annex 3.1 for additional results and the economic magni- tudes of the effects.

114 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Figure 3.7. Regulatory Capital Ratios across Major Economies and Regions (Percent of risk-weighted assets)

Common equity Tier 1 Additional Tier 1 Tier 2 Tier 3 20

0 08 12 08 12 08 12 08 12 08 12 08 12

2005 United States 2005 Japan 2005 Euro area 2005 Other AE 2005 EM CEE 2005 EM Asia

Sources: SNL Financial; and IMF staff estimates. Note: Japan data for 2005 omitted due to data limitations. AE = advanced economies; CEE = central and eastern Europe; EM = emerging market economies.

Figure 3.8. Asset Encumbrance: December 2007 and June 2013 (Percent of total bank assets)

Central bank funding Covered bonds and ABS Repurchase agreements 40

0 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 07 13 2007 ITA IRL FIN BEL ESP LUX SVK PRT FRA AUT CZK SVN USA CHE DEU NLD GRC GBR CAN HUN DNK NOR SWE

Sources: European Central Bank; European Covered Bond Council; Haver Analytics; and IMF staff estimates. Note: ABS = asset-backed securities; AUT = Austria; BEL = Belgium; CAN = Canada; CHE = Switzerland; CZK = Czech Republic; DEU = Germany; DNK = Denmark; ESP = Spain; FIN = Finland; FRA = France; GBR = United Kingdom; GRC = Greece; HUN = Hungary; IRL = Ireland; ITA = Italy; LUX = Luxembourg; NLD = Netherlands; NOR = Norway; PRT = Portugal; SVK = Slovak Republic; SVN = Slovenia; SWE = Sweden; USA = United States.

International Monetary Fund | October 2013 115

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.9. Share of Retained Bank-Covered Bonds in Europe

1. Covered Bond Issuance by European Banks 2. Share of Self-Funded Covered Bond Issuance (billions of U.S. dollars) (percent of total covered debt issuance, 2005±1 3) 500 120 Sold to third party 450 Self-funded 100 400

350 80 300

250 60

200 40 150

100 20 50

0 0 2005 06 07 08 09 10 11 12 13 ry Italy ance Spain rtugal Fr Ireland Cyprus Austria Greece Finland Norway Sweden Belgium Po Hunga Germany Denmark Switzerland Netherlands Czech Republic United Kingdom Sources: Dealogic; and IMF staff estimates. Note: The sample includes public and private sector banks, but excludes agency bonds. For 2013, the data are annualized using data through the end of July.

Figure 3.10. Contribution of Funding Characteristics to Bank Distress (Relative size of factors; percentage points)

Z-score Price-to-book ratio Analysts’ ratings 6

–2

–4

–6 All All All All All AE AE AE AE AE EM EM EM EM EM G/DSIB G/DSIB G/DSIB G/DSIB G/DSIB 2007–12 2007–12 2007–12 2007–12 2007–12

Loan-to-deposit ratio Short-term debt Debt ratio Equity ratio Diversity

Sources: Bloomberg, L.P.; and IMF staff estimates. Note: AE = advanced economies; EM = emerging market economies; G/DSIB = global and domestic systemically important banks. Figure shows the economic significance of bank funding characteristics, evaluated at 1 standard deviation away from the variable’s mean, on the probability of distress specified under alternative distress models and samples. Bank distress is a dummy variable, defined either as a z-score below 3, price-to-book ratio below 0.5, or average analyst ratings of 2.5 or lower. G/DSIB is a subsample consisting of systemically important banks: G-SIBs are from Financial Stability Board (2012b), and D-SIBs are banks whose assets account for close to or exceed 20 percent of GDP. Lighter-shaded bars denote nonsignificant effects. See Annex 3.1 for further details. The emerging market economy sample contains banks from developing Asia and central and eastern Europe.

116 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Box 3.3. Changes in Cross-Border Bank funding Sources

Since the global financial crisis began, financial fragmenta- ing. In 2004, foreign holdings of bank debt securi- tion and bank deleveraging have affected cross-border bank ties accounted for 40 percent of the total for France, funding patterns. In particular, foreigners’ investments in Germany, and Spain, whereas holdings for Italy debt securities of banks located in stressed euro area countries were about 10 percent (Figure 3.3.1). The financial have declined significantly; banks in core euro area countries fragmentation and bank deleveraging in some stressed have generally experienced the opposite. Changes appear to be euro area countries has led to a decline of foreign smaller in non-euro-area advanced economies.1 holdings for Italy and Spain. This declining path has been associated with steady increases of foreign hold- In the euro area, foreign investors can be differenti- ings for France and Germany. This divergent trend has ated between core and stressed economies, reflecting eased since the European Central Bank’s announce- financial segmentation and ongoing bank deleverag- ment of Outright Monetary Transactions in September 2012, which has helped mitigate tail risks. The authors of this box are Serkan Arslanalp and Takahiro Despite the high variation in foreign holding pat- Tsuda. terns across countries outside the euro area, the foreign 1The estimation methodology is based on Arslanalp and Tsuda (2012). Total debt securities issued by banks are from the Bank investor base for bank debt securities has been quite for International Settlements (BIS) Debt Securities database, stable (Figure 3.3.2). For instance, Korea has had very and foreign holdings of those securities are from the IMF-World low foreign holdings (about 10 percent) relative to Bank Quarterly External Debt Statistics. The BIS debt securities the total size of bank debt securities, whereas more statistics include debt securities issued by all financial corpora- than 50 percent of Sweden’s bank debt securities have tions, not just depository corporations. The foreign share of bank debt may, therefore, be understated in countries in which been held by foreigners. Yet in both countries, changes nonbank financial corporations issue a large amount of debt. over time have been small, with a modest increase of Both databases are based on the residency principle in relation foreign holdings in recent years. to the issuer and holder of debt. The analysis covers selected A similar picture emerges from public disclosures advanced economies for which long-term data are available. The of large U.S. money market funds. Before the global Fitch sample includes the 10 largest U.S. prime money market funds with total exposure of $654 billion as of the end of April financial crisis, U.S. money market fund allocations 2013, representing 46 percent of total U.S. prime money market to European banks represented about half their total fund assets. exposure to banks, based on Fitch Ratings’ sample of

Figure 3.3.1. Euro Area: Foreign Holding of Figure 3.3.2. Non-Euro Area: Foreign Holding Bank Debt Securities of Bank Debt Securities (Percent of total) (Percent of total) 60 70 Germany Spain Australia Sweden France Italy United Kingdom Korea 60 50

50 40

40 30 30

20 20

10 10

0 0 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 Q3 :Q3 2004:Q1 2005:Q1 2006:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 2011:Q1 2012:Q1 2004:Q1 2005:Q1 2006:Q1 2007:Q1 2008:Q1 2009:Q1 2010:Q1 2011:Q1 2012:Q1

Sources: Bank for International Settlements; IMF/World Bank, Quarterly External Debt Statistics; and IMF staff estimates.

International Monetary Fund | October 2013 117

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 3.3. (continued)

Figure 3.3.3. U.S. Money Market Fund Exposure to European and Other Banks (Percent of total)

Europe United States Australia Canada 60 United Kingdom Japan Germany Stressed euro 16 area economies 14 50 12 40 10

30 8

6 20 4 10 2

0 0 2006:H2 2007:H1 2007:H2 2008:H1 2008:H2 2009:H1 2009:H2 2010:H1 2010:H2 2011:H1 2011:H2 2012:H1 2012:H2 2006:H2 2007:H1 2007:H2 2008:H1 2008:H2 2009:H1 2009:H2 2010:H1 2010:H2 2011:H1 2011:H2 2012:H1 2012:H2 2013:Q1 2013:Q1

Sources: Fitch Ratings; and IMF staff estimates. Note: Stressed euro area economies include Ireland, Italy, Portugal, and Spain. H = half year; Q = quarter.

U.S. money market funds (Figure 3.3.3). This share rebounded recently. Meanwhile, U.S. money market declined rapidly starting in 2010, as U.S. money funds continue to increase allocations to Australian, market funds stopped funding banks in Ireland, Italy, Canadian, and Japanese banks, which combined Portugal, as well as Spain, and reduced their alloca- represent about one-third of their total exposure to tion to core euro area banks, although the latter have banks.

economy and emerging market economy banks, for emerging market economy banks and for sys- except for the case in which distress is measured by temically important banks.13 the price-to-book ratio. For systemically important • Higher reliance on wholesale funding (a higher loan- banks, the effect of better capitalization is much to-deposit ratio), is linked to higher bank distress in smaller, possibly reflecting their too-big-to-fail status both advanced economies (under all distress mea- during the sample period.12 sures) and emerging market economies (using the • Debt, in particular short-term debt, harms bank balance sheet distress measure) during the sample stability. Higher reliance on short-term debt is asso- period.14 However, especially in the absence of cred- ciated with an increase in bank distress. Higher debt

ratios are also correlated with an increase in bank 13For the full period and for advanced economy banks, the distress, especially in the recent period (2007–12), results for the analysts’ ratings-based measure associate lower distress probabilities with higher debt-to-asset ratios, which likely reflects the (eventually unsustainable) buildup of leverage before the global financial crisis. However, analysts assign lower distress probabilities to systemically important banks with lower debt ratios. 12In related research, Bertay, Demirgüç-Kunt, and Huizinga 14A similar result was found in a country-based panel framework (forthcoming) find that systemically important banks are less profit- for emerging market economies. No threshold effects, in which able and do not have lower risk. Ueda and Weder di Mauro (2012) other interest-bearing liabilities above a certain level were associated find that credit ratings of systemically important banks imply a with banking crises, were found in this study. See Chapter 4 of the structural subsidy. October 2012 GFSR. Gudmundsson and Valckx (forthcoming) also

118 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Figure 3.11. Evolution of Bank Funding Characteristics (Ratios)

Nondistressed Distressed 1.8

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0 2007–10 2011–12 2007–10 2011–12 2007–10 2011–12 2007–10 2011–12 2007–10 2011–12 2007–10 2011–12 2007–10 2011–12 1999–2006 1999–2006 1999–2006 1999–2006 1999–2006 1999–2006 1999–2006 Noncore Loan-to- Short term Diversity Deposits Debt Equity funding deposit ratio

Sources: Bloomberg, L.P.; and IMF staff estimates. Note: Distressed banks are those with z-scores that fall in the lowest 10 percent of the distribution (z-scores below 3). Noncore funding is the ratio of debt to customer deposits and equity. Short term is short-term debt plus repurchase agreements (repos) as a percent of total bank debt and repos. Diversity is a Herfindahl index of funding concentration (lower ratio indicating more funding diversity, defined as the squared shares of deposits, debt, and equity). Deposits, debt, and equity are expressed relative to total bank assets. Deposits are total customer deposits. Debt is defined as nondeposit bank liabilities. Equity is total common equity (at book value).

ible deposit insurance programs, panic deposit runs fallen, while the loan-to-deposit ratio has remained could be destabilizing. broadly stable. • Higher concentration in funding sources is associ- • Distressed banks have made some improvements to ated with a higher level of bank distress in some their funding structures, but most components have cases, which suggests that banks need to seek a bal- changed for the worse. On the positive side, their anced funding mix. use of short-term debt and repos has fallen close to Since the crisis began, most banks have altered their the levels for nondistressed banks (perhaps because funding structures to make themselves less vulnerable of an inability to roll over short-term debt), and to financial instability, but distressed banks (those with their funding mix has become more diversified than low z-scores) are still subject to unfavorable funding for nondistressed banks. However, their loan-to- market developments (Figure 3.11).15 deposit ratios have increased as a result of reduced • Nondistressed banks have improved their funding access to deposits, and debt financing (including structures by slightly increasing their capitalization recourse to central bank funding through repos) has ratios (equity-to-asset ratios) and lowering their increased, pushing up their leverage and reducing debt ratios. Also, their funding sources have become equity-to-asset ratios considerably. slightly more diversified, and reliance on short-term debt and repos (relative to total borrowings) has Crisis and the impact of Regulatory Reforms on the pricing of Bank liabilities analyze the importance of core and noncore funding ratios as in The crisis has prompted various regulatory reform pro- Hahm, Shin, and Shin (2012). 15Distressed banks are defined as those with z-scores below 3 posals, some of which are aimed at directly changing (those in the lowest 10 percent of the distribution). bank funding structures and loss-sharing rules across

International Monetary Fund | October 2013 119

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.12. Priority of Claims of Bank Liabilities brance (elevated, in part, as a result of the ongoing crisis) magnify the expected losses that senior unse- Without depositor preference With depositor preference or bail-in or bail-in cured debt holders will suffer in the event of default, which can increase their costs. The overall effect on High Secured debt Secured debt funding costs is not easily surmised, as not only will

Senior unsecured debt the rates associated with each liability type change, but Deposits the amounts used of each type will also change. Most likely, systemically important banks that had been able

Deposits Other debt to fund themselves at a lower overall cost than other Other (bail-in) senior banks as a result of their implicit too-big-to-fail status

unsecured debt Senior unsecured debt will see their cost of funding rise. For other banks, the effects will depend on a combination of factors, as highlighted in the following examples (Box 3.4). Subordinated debt Subordinated debt A numerical analysis based on the option-like Equity Equity features of bank funding structures can help shed light Low on the possible repricing effects of some key aspects of these reforms. This approach, which builds on Merton Source: IMF staff. (1974), allows a holistic analysis by linking the price of debt to the overall composition of funding (including the equity buffer) and to the risks on the bank’s bal- various funding instruments. Basel III capital regula- ance sheet. At the same time, not all regulatory reforms tions should raise banks’ loss-absorbing equity buffers, can be placed into this framework. For example, and the accompanying liquidity regulations should Basel III liquidity regulation can make a bank safer strengthen funding structures against liquidity shocks. by changing its asset structure, rather than its liability OTC derivatives reforms, by requiring collateral to be structure, and by reducing its asset-liability maturity set aside in bilateral trades and at centralized counter- mismatch, which is the main source of liquidity risk. parties, will enhance the safety of these markets but Before discussing the numerical exercise, this section will encumber more assets. Proposals to strengthen first reviews selected aspects of regulatory reforms and resolution frameworks (such as introducing depositor asset encumbrance, providing a sense of their likely preference and providing bail-in powers) may increase effects on funding structures (for more details, see losses for some bank creditors in an effort to protect Annex 3.2). small depositors and limit the burden on taxpayers in the event of resolution (Figure 3.12 and Annex 3.2).16 Policymakers need to be aware of the complex Basel iii Capital Regulations: More and higher-Quality interactions of these reforms—while acknowledging Capital the legacy effects of the crisis—on bank funding struc- The Basel III capital regulations promote higher levels tures and costs. In particular, some changes to funding of minimum equity capital and improve its qual- structures (including more equity) combined with ity, making any debt safer and cheaper.17 Although reallocation of losses upon bank failure among differ- the minimum total capital requirement is set at the ent debt holders can produce disproportionate changes same level as in Basel II—8 percent of risk-weighted of funding costs that are not easily anticipated. On the assets—the quality of capital in Basel III is higher, one hand, having more equity (a larger loss-absorbing requiring 4.5 percent of risk-weighted assets to be of buffer) makes all debt safer and cheaper. On the other higher-quality capital (common equity Tier 1 [CET1]). hand, bail-in powers and the introduction of deposi- In addition, Basel III sets considerably more stringent tor preference—which are being actively discussed in criteria for what qualifies as CET1, additional Tier 1 Europe—combined with high levels of asset encum- capital, and Tier 2 capital. The Basel III framework

16Bail-in powers are generally designed to ensure that shareholders and debtors internalize the cost of bank failure rather than being 17The Basel Committee on Banking Supervision (BCBS) issued bailed out by taxpayers. See Le Leslé (2012) for the broad impact of the details of its global regulatory capital standards in 2010 (BCBS, these regulatory initiatives on European banks. 2010a). They are expected to be phased in by 2019.

120 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Box 3.4. Bank funding in emerging Market economies and the impact of Regulatory Reforms

This box summarizes funding structures of emerging market • In general, the already-higher capitalization and economy banks and provides some indications of how greater reliance on deposits should support emerg- regulatory reforms will affect these banks. ing market economy banks’ transition to Basel III. However, there are variations across jurisdic- In general, emerging market economy banks have tions. For instance, banks in Mexico tend to rely safer funding structures than advanced economy much more on repurchase agreements and other banks. Emerging market economy banks are better wholesale funding sources than do their Asian peers capitalized, rely more on deposits and less on debt, (CGFS, 2013), which could mean lower liquid- and have lower funding gaps (loan-to-deposit ratios), ity ratios. While some jurisdictions have voiced all of which are desirable features for a more resilient concern about their limited supply of government bank (Figures 3.4 and 3.10). Even larger banks do not securities, which is a major component of high- appear to rely excessively on wholesale versus deposit quality liquid assets in satisfaction of the liquidity funding (Figure 3.5). Asset encumbrance appears to coverage ratio, many emerging market economy be limited as well: most of their medium-term debt is banks have an even higher share of government either senior unsecured or capital-qualifying debt (Fig- securities on their balance sheet than do their ures 3.2 and 3.6). In some economies, funding from advanced economy bank peers, including banks a foreign parent bank—a type of wholesale funding— from Saudi Arabia or financial centers such as could be a relevant source of bank funding, although Hong Kong SAR and Singapore (see Chapter 3 of in some cases subsidiaries provide funds to parents. the April 2012 Global Financial Stability Report, Current bank funding structures in emerging although liquidity of these securities could be less market economies appear to be less affected by regula- than those in advanced economies. tory reforms, although some cross-border effects pose • One area of uncertainty faced by emerging market concerns. Emerging market banks seem to be better economy banks is how any funding they receive positioned to satisfy Basel III requirements, on aver- from their parent banks in advanced economies will age, than their advanced economy peers. Potential ten- be treated. In principle, liquidity regulations are sions arising in advanced economies (among liquidity applied at group levels, covering both parent and regulations, asset encumbrance, depositor preference, subsidiary, and it is up to host supervisors to decide and bail-in power) appear less stark as well. However, whether to additionally apply the regulation on a interactions among home (advanced economies) and solo basis to foreign bank subsidiaries in their juris- host (emerging market economies) jurisdictions will diction, which should help to ensure that liquidity require enhanced cooperation and communication buffers are sufficient for the local bank. However, so as to reduce potential cross-border tensions from there could be a direct impact on their funding if reforms that aim to strengthen resolution framework these banks are borrowing substantially from their and lower financial stability risks. parent and their parents need to adjust their own • Basel III capital and liquidity requirements are operations to cope with new regulatory require- expected to be implemented on the same schedule ments, including by deleveraging and by increasing for all Basel Committee on Banking Supervision local high-quality liquid assets and deposits. member jurisdictions, including those in emerging • The Financial Stability Board (FSB) is encourag- market economies such as Argentina, Brazil, China, ing the G20, including the major emerging market India, Indonesia, Mexico, Russia, Saudi Arabia, economies, to adopt the legal reforms necessary to South Africa, and Turkey, and those in the Euro- fully meet the Key Attributes of Effective Resolution pean Union. Some other emerging market econo- Regimes for Financial Institutions (FSB, 2011) by mies in Latin America and Asia have also indicated the end of 2015. Emerging market economy banks’ that they will implement the new regulations. high share of deposit funding, combined with However, in other countries, it could take much bail-in powers and deposit preference (if adopted) longer before they adopt Basel III. could potentially push up their cost of issuing unsecured debt. However, low asset encumbrance and relatively high equity capital buffers should The authors of this box are Marc Dobler, Hiroko Oura, and help to mitigate any adverse impact on overall fund- Mamoru Yanase. ing costs.

International Monetary Fund | October 2013 121

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 3.4. (continued)

• Key concerns of policymakers in emerging mar- capacity is held and which jurisdiction is permit- ket economies are (1) how the reforms for global ted to trigger a bail-in. The FSB is encouraging systemically important banks would affect their enhanced cooperation and communication between scale of operations and intermediation costs in host home and host authorities, including with host jurisdictions (particularly when the host bank- authorities who have not been invited to participate ing systems are largely foreign-owned); and (2) in the crisis management groups that have been set whether benefits and costs of the reforms would up for each global systemically important bank to be spread unevenly across home and host jurisdic- address these risks. tions, depending on where additional loss-absorbing also adopts a non-risk-sensitive simple leverage ratio Basel iii liquidity Regulations: longer and More Stable that serves as a backstop to the risk-based measures by funding constraining the buildup of leverage in the banking The systemic liquidity shocks during the global finan- system. Furthermore, Basel III adds various buffers,18 cial crisis promoted globally agreed-upon quantitative which will eventually raise the effective total capital liquidity regulations for the first time. The regulations ratio to between 10.5 and 15.5 percent, depend- are formulated as the liquidity coverage ratio to improve ing on the applicability of the extra buffers, mostly resilience to short-term liquidity shocks by encouraging in CET1.19 Global systemically important banks are banks to hold high-quality liquid assets for such events, subject to surcharges, given their critical relevance for and the net stable funding ratio (NSFR) requiring long- financial stability. With no change in assets, higher term assets to be financed by stable funding (BCBS, capital buffers should reduce the probability of default, 2010b). These regulations aim to reduce liquidity risks reducing the costs of debt regardless of the remaining arising from maturity mismatches and short-term fund- funding structure.20 ing sources and to provide a stronger incentive for banks Basel III also raises the loss-absorbing capacity of debt to shift their funding mixes to include more insured that qualifies as additional Tier 1 and Tier 2 capital, deposits (from individuals and small and medium better protecting senior debt. In particular, the relevant enterprises) and more longer-term funding (secured authority should have discretion to write off or convert or unsecured), which have been shown to be relatively these other instruments to common equity if the bank is more resilient during the recent crisis. judged to be nonviable.21 The objective is to give better Most banks are on track to satisfy the liquidity incentives for investors to limit banks’ risk taking and requirements, implying little additional need to modify to increase the private sector contribution in resolving liability structures. The latest Quantitative Impact failed banks while reducing fiscal costs. Study (QIS) for liquidity coverage ratios (BCBS, 2012) showed that banks in the BCBS member jurisdictions already had a greater than 90 percent liquidity coverage 18These include (1) a conservation buffer (additional 2.5 percent of risk-weighted assets with CET1) that triggers supervisory limits ratio, on average, at the end of 2011, compared with on a bank’s payouts (for example, dividends) when banks fall into the 100 percent requirement to be achieved by 2019, the buffer range; (2) a countercyclical buffer (an additional zero to although European banks lag somewhat.22 With the 2.5 percent of risk-weighted assets with CET1) that is added when supervisors judge that credit growth is leading to an unacceptable 2013 revision to the rule (BCBS, 2013a), the aver- buildup in systemic risk; and (3) additional charges on G-SIBs (cur- age liquidity coverage ratio for those banks is likely to rently 1 to 2.5 percent of risk-weighted assets with CET1) to ensure exceed 100 percent. The latest QIS (as of June 2012) they have higher loss-absorbing capacity to reflect risks that they pose to the financial system. suggests that the average net stable funding ratio had 19Some view this minimum capital requirement to be insuffi- already reached the required 100 percent level (BCBS, ciently large (Admati and Hellwig, 2013). 20See also the section in this chapter on “Are Bank Funding Struc- 22Central bank funding is less likely to affect the liquidity cover- tures Relevant to Financial Stability?” age ratio because it reduces both the unencumbered high-quality 21For instance, this would occur if minimum capital require- liquid assets (that is, the numerator of the ratio) and, because of the ments are breached and recapitalizing through private markets is not stability of central bank funding, the amount of funds that can be feasible. lost within 30 days (that is, the denominator).

122 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

2013b), although the rule is currently under review by • Regulatory factors: In contrast, regulatory changes the BCBS and has yet to be finalized. could lead to more permanent changes in asset encumbrance.24 o Some aspects of the Basel III liquidity regula- the impact of the Crisis and Various Reforms on Asset tions could encourage covered bond issuance and encumbrance increase asset encumbrance. For instance, covered The more assets are used as collateral (termed “encum- bonds qualify as a part of high-quality liquid assets, bered”) to mitigate counterparty risks, the less likely which would improve the liquidity coverage ratio it is that unsecured creditors will receive what they if a bank holds covered bonds as assets.25 The ratio are due in the event of a resolution, thus raising their for an issuing bank would improve if long-term costs. Encumbered assets are used to back up repay- covered bonds replace shorter-term wholesale ments owed to secured debt holders or the settlement funding. Issuing covered bonds can also improve of losses on derivatives contracts (see Table 3.4 in the net stable funding ratio by raising the available Annex 3.2 for an illustration). Collateral is useful for amount of long-term stable funding. mitigating counterparty risks, and secured funding o OTC derivatives reforms will lower counterparty (including central bank funding) could be the only credit risks at the expense of higher encumbrance. available source of market funding during a systemic The reforms will encourage participants to place liquidity crisis. However, higher asset encumbrance collateral either with derivatives counterparties reduces the amount that debt holders without col- (including dealer banks) or with a formal central lateral will receive if the bank becomes insolvent, and counterparty, both of which will receive preferen- therefore those debt holders will require higher yields tial treatment in the event of resolution. Because to hold this debt. At the same time, other liability activity in this market is dominated by banks, it is holders will be better protected (including those expected that the collateral requirements could be holding secured debt), and their required returns will quite large, encumbering more assets. likely be lower. The overall effect on funding costs will depend on the amounts of various types of funding instruments, the relative funding costs, and the under- the impact of Bank Resolution Reforms lying riskiness of the banks’ assets (both encumbered Two elements of the current bank resolution reform and unencumbered), leading to an ambiguous overall proposals could especially affect bank funding patterns effect on funding costs. and costs. These are: (1) depositor preference in liqui- Asset encumbrance can be driven by both transient dation, when bank operations are discontinued; and and permanent factors. (2) the bail-in of creditors in resolution, when bank • Transient factors (including crises): Periods of financial operations are maintained but, possibly, restructured. distress can be accompanied by systemic liquid- Depositor preference gives depositors legal seniority ity shortages resulting from the declines in private over other senior unsecured creditors when a bank is short-term wholesale funding that occur when closed, providing better protection for (small) deposi- participants withdraw due to elevated counterparty tors at the expense of bondholders (see Figure 3.12 and credit risk. During such times, central banks provide Table 3.4 in Annex 3.2). This preference contrasts with liquidity to banks against collateral, leading to corporate liquidation systems in which all unsecured higher asset encumbrance (see Figure 3.8).23 More- creditors are ranked equally (that is, pari passu), unless over, weaker banks may only be able to tap private contracts state otherwise. Depositor preference can markets if they offer secured debt. These increases contribute to financial stability by enhancing depositor should dissipate once financial conditions normalize. confidence and reducing contingent liabilities of the

23Gorton and Metrick (2012) indicate jumps in the reductions 24For a discussion of covered bonds and the degree to which they (“haircuts”) assigned in the U.S. private repo market. Covered alter bank incentives, see Jones and others (forthcoming). bonds are typically issued with collateral whose value exceeds that of 25Banks are becoming major investors of covered bonds issued bonds, and this excess is measured by overcollateralization. Rating by other banks, in part motivated by their preferential treatment agencies often set the level of overcollateralization that is necessary to in the liquidity coverage ratio framework. However, at the end of maintain a certain rating. These levels vary significantly across bonds, 2011, covered bonds amounted to less than 3 percent of high-quality from less than 10 percent to more than 100 percent. liquid assets.

International Monetary Fund | October 2013 123

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

deposit guarantee scheme. Many countries, including bank bondholders fully bear the risks they assume.28 the United States, already have some form of explicit This action should remove bank bondholders’ expecta- depositor preference, and many provide implicit tions that their investments in systemically important preferences during a systemic crisis. The international banks will be bailed out by taxpayers. Bank bondhold- proposal—the Financial Stability Board’s Key Attributes ers are typically institutional investors who are assumed of Effective Resolution Regimes for Financial Institu- to have the capacity to make more informed choices tions (FSB, 2011)—does not require countries to adopt and absorb losses more easily than retail depositors. depositor preference. However, its use may facilitate Therefore, the Financial Stability Board’s proposal the use of other resolution tools, such as bail-in.26 A (FSB, 2011) excludes insured depositors (and secured number of countries, including the European Union as debt holders) from bail-in, although some countries a whole, are actively considering depositor preference. may exclude additional liabilities, such as short-term Depositor preference can be “tiered” so that insured debt and interbank funding. These exclusions would deposits are covered first, with the deposit guarantee increase losses for bail-in debt holders beyond what scheme stepping in to assume the rights of the depositor would have applied when they were ranked equally in liquidation proceedings (called subrogation), and then with all other senior creditors.29 Hence, to attract other deposits that are eligible for the deposit guarantee bondholders for bail-in debt, their yields would need scheme coverage (but that exceed the insurance limit) to rise to reflect the increased prospect of losses (Figure are covered before payouts are made to senior unsecured 3.12 and Table 3.4). creditors. This tiered structure offers the greatest finan- For bail-in powers to effectively provide more loss- cial protection for the state (or the deposit guarantee absorbing capacity, banks would need to maintain a scheme), but also would concentrate potential losses on certain amount of bail-in debt, leading to proposals for a smaller group of creditors. some quantitative targets. The 2012 European Com- Statutory bail-in aims to hold bank bondholders mission’s proposal (EC, 2012) suggests 10 percent of accountable for the risks they assume by removing total liabilities (including regulatory capital) as the the implicit too-big-to-fail subsidy for systemically target. In the United Kingdom, the Vickers report important institutions and by imposing larger losses on (ICB, 2011) proposes loss-absorbing capacity between them than on smaller retail creditors. Statutory bail-in 7 and 10 percent of risk-weighted assets (in addition grants authorities the power to write down debt or to equity amounting to 10 percent of risk-weighted convert debt to equity when a bank is near failure so assets). This level was set to ensure that banks would that these bailed-in debts absorb losses should capital have enough loss-absorbing capacity to cover losses be exhausted (see Zhou and others, 2012).27 These comparable to those that have materialized in the most powers become available when a bank is no longer recent bank failures.30,31 viable but before it becomes insolvent, ensuring that 28The point at which a resolution authority decides a bank is not viable should be somewhere between breaching the regulatory minimum capital requirement and becoming insolvent, and it should be the same as for other bank resolution tools. The Basel III capital 26For example, if a resolution authority decides to restructure regulations already incorporate such bail-in characteristics with and revive a bank, forcing general debt holders to forgo some value capital-qualifying debt instruments. (that is, bail-in) while protecting insured depositors, the general debt 29However, in the past, resolution authorities have protected some holders can potentially sue the resolution authority, claiming they depositors without legal rights. For instance, a failing bank’s deposits would have been better off if the bank had been liquidated. Intro- and some corresponding assets may be transferred to other banks. ducing depositor preference for insured depositors would align the Therefore, the current yield for senior unsecured debt should already recovery amount for general debt holders more closely to the bail-in reflect such differential treatment to some degree. amount, preventing such a lawsuit. 30During 2007–10, the Anglo Irish Bank suffered a loss of 39 27Statutory bail-in power and bail-in debt should not be confused percent of risk-weighted assets, though all other banks saw losses of with contingent convertible capital instruments (CoCos), despite less than 16 percent. their similarities. CoCos are new bank capital instruments that have 31This emphasis on “large enough” loss-absorbing capacity con- contractual clauses indicating they are written off or converted to trasts with some of the traditional views that emphasize the role that equity when contractually set criteria are met, such as a decline in even a small amount of debt (for example, subordinated debt) can the CET ratio to, say, 7 percent (a level that could be set above regu- play in motivating such creditors to monitor and discipline banks’ latory minimums). In contrast, statutory bail-in powers give legal activities (Calomiris, 1999; Calomiris and Kahn, 1991). In contrast, rights to a country authority to give a haircut to general debt (such Admati and Hellwig (2013) challenge the disciplining role of bank as senior unsecured debt or uninsured deposits, unless explicitly debt and propose that banks should have a higher amount of equity exempt) or convert it to equity when a bank is deemed not viable. capital (15 percent of unweighted total assets) to absorb losses.

124 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

potential Challenges posed by the Regulatory Reforms higher asset encumbrance is difficult when banks face systemic funding difficulties. Attempts to introduce bail- Strengthening resolution regimes will increase the cost in rules or limits to asset encumbrance in the middle of of senior unsecured bonds but will also require a suf- a systemic crisis could exacerbate instability. Moveover, ficient amount of bail-in debt to provide potential loss limits on asset encumbrance may also make it more absorption. difficult to achieve the goal of making OTC derivatives • Introducing depositor preference would increase safer. On the other hand, without such limits, a bank unsecured creditor losses in the event of a bank fail- may have too few assets to be shared among unsecured ure by reducing their seniority rank. This provides creditors (including uninsured depositors and the better protection for retail depositors and small deposit guarantee scheme) when they face resolution. and medium enterprises. Bail-in powers could also Basel III liquidity regulations and the altera- impose higher losses on unsecured creditors, increas- tions in resolution regimes may push bank funding ing the cost of this bail-in debt. The largest effect on structures in different directions and will likely drive funding costs will likely be on systemically impor- some intermediation into the shadow banking arena tant banks because it will lessen the implicit subsidy (see Chapter 1). The liquidity regulations encourage they have received from their too-big-to-fail status. (insured) deposit funding that is likely to be protected Some researchers (for example, Ueda and Weder di by depositor preference and from being bailed in, and Mauro, 2012) estimate that the implicit subsidy is hence may reduce the proportion of bail-in debt.33 worth between 80 and 100 basis points. The exact Banks also may rely on long-term secured debt to cost impact of several configurations is explored reduce maturity mismatches, encumbering more assets. quantitatively in the following section. The higher Although the latest Quantitative Impact Study indi- cost could drive banks to increase insured deposits cates banks are broadly on track to meet the liquidity and secured funding. It also raises the question of ratios, European banks—the main issuers of covered whether traditional investors in bank debt will pur- bonds—have tended to lag. And in general, banks’ chase bail-in debt in the future (see Box 3.5). ability to acquire funding may become more difficult, • The growing use of deposits in some jurisdic- leaving room for other nonbank institutions (shadow tions and the likelihood that uninsured deposits banks) to collect savings and intermediate credit. will either be formally preferred or given de facto preferential treatment in a resolution (for instance, via public guarantees to contain a deposit run) may implications of Regulatory Reforms on Bank funding reduce the effectiveness of bail-in powers, without Costs: A numerical exercise commensurate efforts to ensure that sufficient bail- There have been many attempts to assess the cost in debt is issued. implications of bail-in powers, depositor preference, Some reforms encourage asset encumbrance, even and asset encumbrance, but few of these fully incorpo- though this may be detrimental to resolution processes. rate the changes in the overall funding structure of a Excessive asset encumbrance reduces bail-in debt and bank. So far, the difference between the yield spreads makes resolution less effective. When too many assets of secured and senior unsecured bank debt has been are encumbered, unsecured creditors (including the relatively small compared with the spread against deposit guarantee scheme) will incur higher losses in subordinated debt (Figure 3.13). Various market order to honor secured debt contracts and collateral payments. The full extent of asset encumbrance, including central bank funding during a crisis, short-term repos, corresponding to its allowance of covered bond issuance). The and covered bonds with overcollateralization, is hard to Netherlands, Norway, and the United Kingdom took a case-by-case approach that set threshold values for covered bond issues for indi- gauge with current reporting systems. Therefore, some vidual institutions (Houben and Slingenberg, 2013). The European countries are improving the reporting of asset encum- Banking Authority issued a consultation paper (2013) on strengthen- brance or setting limits on encumbrance, for example, ing reporting and transparency of asset encumbrance. 33 by limiting the combined value of assets that can be As an extreme example, suppose a bank funds itself with 90 percent insured deposits and 10 percent equity. This liability structure 32 used to secure covered bonds. However, avoiding would be desirable from the perspective of the Basel III liquidity requirements but inconsistent with the desire to have bail-in debt. 32For example, Australia, Canada, and Singapore set limits on Of course, enough equity capital would supplant the need for bail-in asset encumbrance (with Australia’s introduction in October 2011 debt.

International Monetary Fund | October 2013 125

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Box 3.5. investor Base for Bail-in debt and Bank Bond Ratings

Three types of bail-in bonds are potentially available— off or converted to equity (the trigger point) when the senior, subordinated, and contingent convertible debt resolution process is introduced (for example, when (or CoCos)—with different investor bases. The degree to the corresponding capital ratio is between zero and which traditional buyers of senior bank debt are willing to the minimum requirement). By contrast, the trig- purchase bail-in debt will largely depend on whether the ger for CoCos is usually set at higher levels (closer to issuing banks are able to maintain stand-alone investment the minimum capital requirement), which would, all grade status. CoCos would likely attract investors with else equal, result in a higher probability of default, higher risk tolerance because of their higher trigger points, making these securities more attractive to investors compared with senior and subordinated debt. New regula- with a higher tolerance for credit risk, such as hedge tions and accounting standards may also play a role. funds or high-yield investment funds. Given the more limited investor base, development of CoCos may Traditionally, the main buyers of senior bank debt be constrained. Total assets under management for have been insurers and pension funds, as well as some event-driven credit arbitrage hedge funds are only $16 mutual funds devoted to investment-grade fixed- billion, although the hedge fund industry has been income instruments and sovereign wealth funds that growing rapidly, with year-over-year growth of 10 per- have a moderate appetite for credit risk. Event-driven cent as of the end of 2012. Barclays (2013) estimates credit arbitrage hedge funds have also participated that the European CoCo market currently stands at in this market, but they are more prominent in the only about €19 billion, but if interest from investors subordinated bank debt market. expands, then this could rise to as much as much as Investor demand for senior debt critically depends €400 billion, which is equivalent to the size of the on whether the issuing banks maintain investment- existing European bank subordinated debt market. grade ratings. According to a recent investor survey Some investors may be constrained by regulations by JPMorgan (Henriques, Bowe, and Finsterbusch, even though the current low interest rate environment 2013), 34 percent of European bank debt investors say would otherwise make them likely candidate buy- they would reduce their investment in senior unse- ers for bail-in debt. Insurance companies are a good cured debt if it became a bail-in instrument, while example—two opposite factors influence their appetite 63 percent of them would maintain it as is. At the (CGFS, 2011). The negative factor includes prospec- same time, survey participants indicated that the most tive changes to international regulatory and accounting important factor determining their decision would standards, which can reduce demand for riskier bonds. be whether the debt would still carry investor grade New mark-to-market rules in international account- ratings. Recent guidelines provided by rating agencies ing standards are expected to increase the volatility suggest that only issuers with high stand-alone ratings of insurance companies’ financial statements, making would have investment grade senior bail-in debt. If riskier assets with higher price variation less attrac- that is the case, the investor base for senior debt may tive. The Solvency II Directive in Europe, currently shrink. Currently, more than 90 percent of the senior scheduled to be phased in beginning in 2014, will also unsecured debt issued by banks is investment grade. require assets to be marked to market and more capital CoCos would likely attract investors with higher to be held against equity-like instruments, structured risk tolerance because of their higher trigger points, products, and long-term or low-rated corporate and compared with senior and subordinated debt. The bank bonds. However, the current low interest rate payoff structures of senior and subordinated debt are environment lowers insurers’ profits (because many similar in the sense that the value of debt is written of them continue to offer high guaranteed returns or generous defined-benefit-type products), encourag- The authors of this box are Serkan Arslanalp and Takahiro ing them to search for higher-yielding assets, creating Tsuda. potential demand for the riskier bail-in debt.

126 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Figure 3.13. Bank Bond Yield to Maturity at Issuance: Selected Advanced Economies (Percent)

Secured Senior unsecured Subordinated

06 07 08 09 10 11 12 13

06 07 08 09 10 11 12 13 06 07 08 09 10 11 12 13 06 07 08 09 10 11 12 13 06 07 08 09 10 11 12 13

2005

2005 2005 2005 2005 France and Germany Ireland, Italy, Portugal, Euro area United States Switzerland Japan and Spain

Sources: Dealogic; and IMF staff estimates. Note: Weighted average yield using deal volume, through May 2013. Excludes deals without volume or yield-to-maturity data, perpetual bonds, and those in foreign currencies. Sample includes both private and public banks but excludes agency bonds.

estimates indicate the yield of senior unsecured debt liability structure, as in Figure 3.12, bondholders face could increase by 100 to 300 basis points under bail-in losses on their debt when the total losses of the bank powers. The current spread between existing CoCos exceed the sum of all the claims with lower priority and senior debt (about 500 basis points) is viewed by (that is, subordinated debt defaults when the losses are some as a good approximation, although CoCos are larger than the amount of equity). Therefore, changes part of subordinated debt, which would continue to in the ranking of priority or in the size of each type of be ranked below senior bail-in debt (Le Leslé, 2012). debt affect the cost to other bondholders. For con- Moreover, the estimates typically fail to account for the venience, all types of instruments (including secured positive influence of the larger equity buffers that will debt) that are ranked above other creditors are labeled be required under Basel III. “preferred creditors” in this exercise. Because the resolution framework reforms are Depositor Preference and Asset Encumbrance currently being actively debated in Europe, the yield Depositor preference and asset encumbrance can be spreads on each type of debt are calculated for a hypo- analyzed using a similar pricing model, despite their thetical bank that has characteristics broadly similar conceptual and legal differences. Both secured debt and to those of large European banks.35 In particular, the preferred deposits have priority over other unsecured proportions of equity and subordinated debt to total bondholders (see Annex 3.3).34 Based on a stylized assets are about 5 percent (see Figure 3.1) and 2 percent, respectively (using only balance sheet assets, not risk- 34To be exact, there are clear differences between having priority weighted assets). To see the sensitivity of bank funding claims based on depositor preference and on asset encumbrance. costs vis-à-vis bank capital levels, we also examine the Depositor preference provides legal seniority to deposits over other unsecured creditors. Secured debt holders have priority claim only up to the value of the collateral assets. If collateral value falls short 35Based on the average capital structure for Royal Bank of Scot- of the face value of secured debt, then the creditors rank equally to land, Commerzbank, DnB NOR, Société Générale, Lloyds, Barclays, other general debt holders for the shortfall amount. See Chan-Lau HSBC, Banco Bilbao Vizcaya Argentaria, Intesa Sanpaolo, Nordea and Oura (forthcoming) for a fuller analysis of asset encumbrance. Bank AB, Danske Bank, Crédit Agricole S.A., and BNP Paribas.

International Monetary Fund | October 2013 127

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

yields when a bank hypothetically maintains two higher of senior unsecured debt changes (along the line) for equity-to-total asset ratios: 10 percent, at the highest end different shares of preferred creditors: of possible capital requirements across countries, and • As bank capitalization (the equity-to-asset ratio) 15 percent, an even higher level.36 The other liabilities declines and asset volatility increases, spreads rise are assumed to be funded either by deposits or senior disproportionately, indicating much higher funding unsecured debt. For large European banks, secured debt costs for riskier and less-capitalized banks. If a bank and deposits average about 17 percent and 48 percent, maintains recent levels of safety (5 percent asset vola- respectively (Street, Ineke, and McGrath, 2012).37 With tility) and is exceptionally well capitalized (15 percent 24 to 70 percent of the deposits insured in Europe (see equity-to-asset ratio), even subordinated debt can be Annex 3.2), the exercise assumes that preferred credi- issued at a fairly low cost (below 200 basis points over tors account for 30 to 50 percent of total liabilities.38 the risk-free rate), and the senior unsecured debt yield The analysis measures bank riskiness using total asset rises fewer than 50 basis points regardless of the pro- volatility and considers two levels: 5 percent, close to the portion of preferred creditors (Figure 3.14, panel 6). current average for global systemically important banks; • The exercise shows that the share of preferred credi- and 10 percent, the worst case during the global finan- tors has a major influence on the spreads of senior cial crisis.39 All debts are assumed to have zero coupons unsecured debt (see Figure 3.14). with a maturity of five years. The five-year risk-free rate (1) Asset encumbrance alone appears less likely to is set at 3 percent. increase the cost of senior unsecured bonds to The spreads across different funding instruments unbearable levels for European banks (Figure depend mostly on the underlying health and riskiness 3.14, panel 4). The share of secured debt, at of bank assets and the share of preferred creditors. an aggregate level, is about 25 percent even Figure 3.14 shows the calculated spreads for all types for Greece (see Figure 3.8). At these levels, the of debt over the risk-free rate for several underlying increase in the senior unsecured debt spread situations: (1) alternative proportions of preferred (along the “senior unsecured” line in Figure creditors (horizontal axis); (2) equity buffers; and (3) 3.14) is less than 30 basis points. The spread of different levels of asset volatility. The figure also shows senior unsecured debt over secure debt (pre- the yield of senior unsecured debt when it is ranked ferred creditors’ yield) is about 55 to 75 basis equally with preferred creditors (labeled as pari passu points, comparable to the actual differences for yields) for comparison. Introducing depositor prefer- most European banks (see Figure 3.13). ence changes the seniority structure and raises senior (2) However, the senior unsecured debt yield spread unsecured debt yields from the pari passu levels to the rises more appreciably with depositor preference. “senior unsecured“ line in Figure 3.14. If preferred The spread would rise relative to the “senior creditors represent secured bondholders, then the cost unsecured pari passu” line and would depend on the share of preferred deposits, which can be much 36A level of 10 percent equity to total assets roughly corresponds larger than secured debt. For European banks to the CET1 requirement with maximum possible buffers and a 0.7 (Figure 3.14, panel 4), the increase is about 30 percent ratio between risk-weighted assets and total assets (comparable to 50 basis points when preferential treatment is to the levels in the United States and emerging market economies). It is worth noting that U.S. banks had an equity-to-total-asset ratio of more limited to insured deposits (dark orange section in than 10 percent for the decades before World War II (Miles, Yang, and Figure 3.14) on top of secured debt. But it could Marcheggiano, 2012). Although it is not universally endorsed by econo- range from 50 to 120 basis points when deposits mists, Admati and Hellwig (2013) propose a 15 percent ratio. 37Assuming repos and short sales are net with reverse repos. that receive preferential treatment rise from 50 to, 40 38These are very rough estimates, applying a range of national aggre- say, 65 percent of assets (light orange section). gate estimates for the share of insured deposits to the average share of The actual increases critically depend on the size of deposits in total liabilities among the 13 large European banks. Much larger variations across individual banks could be present. 39The 10 percent corresponds to the highest observation across time and across banks using total asset volatility as calculated by 40Depositor preference should also reduce the cost of deposits Moody’s KMV for global systemically important banks (as defined from the senior unsecured pari passu debt levels to preferred credi- by the Financial Stability Board) from January 2005 through June tors levels. However, banks might already enjoy low deposit funding 2013. The median (across time and banks) is about 4 percent, and costs thanks to a deposit guarantee scheme. In that case, higher the average for May 2013 is 4.2 percent. seniority benefits the deposit guarantee scheme but not the banks.

128 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Figure 3.14. Debt Pricing under Depositor Preference and Asset Encumbrance (Spread over risk-free rate; basis points)

Preferred creditors Senior unsecured Subordinated Senior unsecured, pari passu

Asset Volatility = 10 Percent; Ratio of Subordinated Debt to Total Assets = 2 Percent

1. Equity/Total Assets = 5 Percent 2. Equity/Total Assets = 10 Percent 3. Equity/Total Assets = 15 Percent 2,500 2,500 2,500

2,000 2,000 2,000

1,500 1,500 1,500

1,000 1,000 1,000 Including all deposits Including Insured deposits + secured Including all deposits Including 500 500 all deposits Including 500 Insured deposits + secured Insured deposits + secured

0 0 0 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 Preferred creditors in percent of assets Preferred creditors in percent of assets Preferred creditors in percent of assets

Asset Volatility = 5 Percent; Ratio of Subordinated Debt to Total Assets = 2 Percent

4. Equity/Total Assets = 5 Percent, European Banks 5. Equity/Total Assets = 10 Percent 6. Equity/Total Assets = 15 Percent 1,200 1,200 1,200

1,000 1,000 1,000

800 800 800

600 600 600

400 400 400 Including all deposits Including Including all deposits Including Including all deposits Including

200 200 Insured deposits + secured 200 Insured deposits + secured Insured deposits + secured

0 0 0 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 Preferred creditors in percent of assets Preferred creditors in percent of assets Preferred creditors in percent of assets

Source: IMF staff estimates. Notes: “Preferred creditors” may include secured debt, preferred deposits, or both. The lines for "senior unsecured" show yield for holders of senior unsecured debt when they have lower priority than "preferred creditors." The lines for "senior unsecured, pari passu" show yields for senior unsecured debt (and for preferred creditors) when it is ranked the same as for preferred creditors. Assumptions: Equity-to-total-asset ratio for European banks is about 5 percent (Figure 3.1), and the subordinated debt ratio is about 2 to 3 percent. A 10 percent equity-to-asset ratio roughly corresponds to the Basel III CET1 requirement, with maximum possible buffers and 70 percent risk-weighted-assets-to-total-asset ratio (e.g., U.S. and emerging market bank levels). The 15 percent corresponds to the proposal by Admati and Hellwig (2013). The asset volatility assumption is based on the estimate by Moody’s KMV for global systemically important banks (January 2005–June 2013), with 10 (4) percent as the highest (median) across time and banks. The average for May 2013 is 4.2 percent. For large European banks, secured debt (assessing repos on a net basis) and deposits average 17 percent and 48 percent of the assets, respectively, and 24 to 70 percent of the deposits are insured (Table 3.5).

preferred deposits and the other parameters in the emerging Asia (see Figure 3.1). Senior unsecured model. debt is likely to remain a distinct asset class from • Depositor preference or asset encumbrance subordinated debt. increases the cost of senior unsecured debt but not to the levels of subordinated debt. The spreads Bail-in Powers for senior unsecured debt are well below those for The pricing effects of introducing bail-in powers subordinated debt even when the share of preferred depend on the conditions for initiating a bail-in and creditors is as high as 70 percent—the current the liabilities excluded from being bailed in. This share of total deposits for banks in Japan and section assumes that the bail-in debt is converted to

International Monetary Fund | October 2013 129

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

equity when the equity-to-asset ratio falls to 5 percent Figure 3.15. Debt Pricing under Bail-in Power and original equity holders are diluted.41,42 It is further (Spread over risk-free rate; basis points) assumed that banks have three types of liabilities: (1) a The liability side has three instruments: deposits—exempt from bail-in; senior unsecured liability that is exempted from being bailed in, labeled bail-in debt—converted to equity when the capital-to-asset ratio declines to 5 percent; and capital. as “preferred creditors;”43 (2) bail-in senior unsecured debt; and (3) capital (equity and capital-qualifying European Banks: Asset Volatility = 5 Percent, Capital-to-Asset Ratio = 7 Percent Preferred creditors Senior unsecured, benchmark subordinated debt combined). Capital buffers of 7 Subordinated, no bail-in Senior unsecured, pari passu percent, 12 percent, and 17 percent are considered.44 Senior unsecured, bail-in Other assumptions are the same as in the depositor 1,200 preference and asset encumbrance cases. The simple existence of bail-in powers would have 1,000 a relatively small impact on bail-in bond yield spreads (Figure 3.15): 800 • The effect of converting bail-in liabilities to equity is small. The “benchmark” yield spread shows the yield spread of senior unsecured debt that is junior 600 to preferred creditors. The difference between the Including all deposits Including 400

benchmark and bail-in debt yield spreads represents Insured deposits + secured the effects of conversion to equity. For European

banks, the difference is small when the exemption is 200 limited to insured deposits and secured debt (dark orange section). When all deposits are exempted 0 (the share of preferred creditors is about 65 percent), 0 10 20 30 40 50 60 70 80 90 100 bail-in debt costs about 50 basis points more than Preferred creditors in percent of assets the benchmark yields (difference between red dashed and red solid lines). Source: IMF staff estimates. Note: The “senior unsecured, benchmark” line is the same as the “senior unsecured” line in • However, the share of exempt liabilities (namely, Figure 3.14 and represents the yield when senior unsecured debt is junior to “preferred creditors,” but not subject to bail-in. When “senior unsecured, benchmark” bonds are de facto preferred creditors) plays a large role similar to that junior to “preferred creditors” their yield is already higher than that of “preferred creditors” of the depositor preference and asset encumbrance and the yield when the two are ranked equally (“senior unsecured, pari passu”). In addition, applying bail-in power and converting them to equity when the bank becomes unviable will cases. The benchmark yield spreads themselves raise their yield from the “senior unsecured, benchmark” line to the “senior unsecured, are already 120 basis points higher than the yield bail-in” line yield. The equity buffer in this ﬁgure corresponds to the sum of equity and subordinated debt in Figure 3.14. spreads when senior unsecured debt is ranked Assumptions: The capital-to-total-asset ratio for European banks is about 7 percent (equity plus subordinated debt). The asset volatility assumption is based on the estimate by Moody’s equally to preferred creditors (pari passu yields), KMV for global systemically important banks (January 2005–June 2013), with 10 (4) percent because seniority is given to preferred creditors. as the highest (median) across time and banks. The average for May 2013 is 4.2 percent. For large European banks, secured debt (assessing repos on a net basis) and deposits average 17 percent and 48 percent of the assets, respectively, and 24 to 70 percent of the deposits are Bank-Specific Estimates insured (Table 3.5). The simulation is applied to four global systemically important banks with distinctive capital structures and outcomes. These represent an investment bank, a risks to gauge whether the model produces realistic global retail bank, a stressed European bank, and a U.S. retail bank (Table 3.1 and Figure 3.16). 41This is a fairly high trigger point: the equity-to-asset ratio for • Senior unsecured debt: The difference between the European banks is a little higher than 5 percent (see Figure 3.1). 42In practice, there will be uncertainty as to exactly when authori- simulated yields and the actual yields is consistent ties will exercise their bail-in power. This uncertainty is excluded across the four banks. For example, the yields are from this illustrative exercise. much higher for the stressed European bank than 43 As discussed in Annex 3.2, in reality, some deposits may be for the U.S. retail bank. Across banks, the actual considered to be bail-in instruments, while some types of senior unsecured debt (for example, short-term debt) may be exempted. yields are close to those of bail-in debt, indicating 44For simplicity, the subordinated debt and equity funding in the that changes in resolution frameworks may already previous exercises are combined into capital. Therefore, the capital be priced into current yields (although the current levels of 7 percent, 12 percent, and 17 percent are considered, respectively combining 5 percent, 10 percent, and 15 percent of yields could also reflect heightened sovereign risk of equity with 2 percent of subordinated debt. the countries in which they are headquartered).

130 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Table 3.1. Characteristics of Four G-SIBs in Simulation Exercise Investment Bank Global Retail Bank Stressed European Bank U.S. Retail Bank Percent of Total Capital Structure Secured Debt 16.0 6.5 22.2 2.5 Deposits 44.7 67.0 49.2 70.5 Senior Unsecured Debt 30.5 18.1 19.1 14.8 Subordinated Debt 3.7 2.5 3.0 1.2 Equity 5.1 5.9 6.4 11.1 Percent Asset Volatility 4.5 3.8 5.1 3.9 Sources: Company annual reports; Moody’s KMV; Street and others (2012); and IMF staff estimates. Note: G-SIBs = global systematically important banks.

Figure 3.16. Simulation Results for Speciﬁc Banks • Subordinated debt: The large equity buffer and low (Yield to maturity for ﬁve-year debt; spreads over risk-free rate; basis points) risk of the U.S. retail bank keep its simulated subor- 1. Secured Debt 2. Deposits dinated debt yield at low levels, which is in line with Actual General DP Pari passu Bail-in the actual and at much lower levels than for other Pari passu Bail-in General DP 150 50 banks (Figure 3.16, panel 4). However, for other banks, market yields are much lower than simulated 100 40 yields, which could reflect in part a too-big-to-fail subsidy. 50 30 • Secured debt and deposits: Depositor preference and 0 20 bail-in powers can provide strong protections to depositors, reducing the deposit rate to the near- –50 10 risk-free rate, even without deposit insurance. Simu- lated secured debt yields are also near risk-free rates, –100 0 Investment Global Stressed U.S. retail Investment Global Stressed U.S. retail although they are not close to the actual yields, bank retail European bank1 bank retail European bank bank bank bank bank perhaps owing to specific characteristics of the debt that are not well captured in the model.45 3. Senior Unsecured Debt 4. Subordinated Debt Actual General DP Actual Simulation 350 Pari passu Bail-in 900 Funding Structure and Incentives to Make a Bank Safer

300 800 Although difficult to determine for banks as a whole, 700 250 banks’ total funding costs may decline if the reforms 600 are calibrated to provide shareholders with incentives 200 500 to prefer safer asset portfolios. For instance, bail-in 150 400 powers that ensure that shareholders are heavily diluted 300 100 200 when a bank becomes unviable could be particularly 50 100 effective for reducing the risk-increasing behavior that 0 0 shareholders normally exhibit in a limited liability Investment Global Stressed U.S. retail Investment Global Stressed U.S. retail bank retail European bank bank retail European bank setting. With bail-in powers, the gains from pursuing bank bank bank bank higher asset volatility for the original shareholders may be offset by the costs that would come from equity Sources: Bloomberg, L.P.; and IMF staff estimates. Note: DP = depositor preference. The pari passu case assumes that all deposits and senior dilution. When the cost is sufficiently large, the origi- unsecured debt are ranked equally. The general DP case assumes that all depositors nal shareholders would prefer a safer portfolio with low (irrespective of sector or insurance coverage) will rank above senior unsecured debt. The bail-in case also assumes that all deposits are exempt from bail-in. Simulated deposit rates exclude asset volatility (Figure 3.17). effects from deposit insurance. Risk-free rates are proxied by ﬁve-year government bond yields for Germany (for euro-denominated debt) and the United States (for dollar-denominated debt). Actual yield to maturity corresponds to August 2013 and is computed as the weighted average of bonds issued in the currency most used by the bank and including all maturities. Summary and policy Recommendations 1No actual data are available. The analysis confirms the relevance of bank funding structures for financial stability. Banks have diverse

45These characteristics would include the maturity, collateral, and extent of overcollateralization.

International Monetary Fund | October 2013 131

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.17. Equity Value for Original Shareholders under encumbrance) and insured deposits and, on the other Bail-in Regime Converting Debt to Equity hand, ensuring that some debt holders bear more losses

Historical asset volatility range for G-SIBs in a resolution through reforms to resolution regimes 30 (bail-in debt and the prospects for additional depositor preference). Altogether, these elements of reform raise the cost of unsecured bail-in debt, in particular. For systemically important banks, the reforms will likely 25 increase the overall cost of funding, particularly by reducing the too-big-to-fail subsidy enjoyed by these financial institutions. Weaker institutions may also 20 experience a larger impact, particularly if they have

Dilution 95% inadequate amounts of capital. For other banks, the overall impact is ambiguous and will depend on the Dilution 90% relative costs and amounts of different funding sources, Dilution 85% 15 Value of equity for original shareholders Value the level of equity capital, and the underlying riskiness Dilution 80% of their assets. Dilution 75% A numerical examination of these potential trade- 10 offs shows that the simulated price impact on unse- 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Asset volatility in percent cured senior debt spreads is relatively small under present conditions, including in euro area countries.

Source: IMF staff estimates. But the share of preferred deposits and the level of Note: G-SIBs = global systematically important banks. Ten percent dilution means bail-in debt asset encumbrance are important drivers of the cost of holders receive 10 percent of (new) equity after they are converted to shareholders, whereas 90 percent of the new equity is given to original shareholders. Original shareholders’ claim is diluted bail-in debt, which rises disproportionately more than by 10 percent in this case. For simplicity, this exercise assumes two types of liabilities: (bail-in) when the share of these other liabilities increase in the debt and equity. It is assumed that $100 is initial asset value, $90 is face value of total debt, and 3 percent is the risk-free rate. Debt is converted to equity when asset value declines to $92. funding structure. For weaker banks, the increased risk to unsecured bondholders may leave traditional investors unwilling to hold this debt and may make it funding patterns that change only slowly. The empiri- difficult to issue enough of it to maintain its market- cal results suggest that countries in which banks were discipline role. In this event, these institutions would overly reliant on short-term wholesale funding (primar- need to raise more equity capital and perhaps restruc- ily larger banks in many advanced economies) were ture their operations and alter their funding structures. more likely to experience financial instability. They However, these potential trade-offs can be managed also suggest that banks with more stable, diversified so as to ensure that the financial stability benefits of funding structures and those that carry less leverage are the reforms can be realized and hence the current set less likely to experience distress. Since the start of the of reforms should move forward in a deliberate man- global financial crisis, some improvements have been ner, paying close attention to their potential interac- made, with most banks lowering their overexposure to tions. The following policy recommendations for short-term wholesale funding, but the funding struc- capital and liquidity rules, asset encumbrance, bail-in tures of some banks, particularly those in distress, have powers, and depositor preference will help. not improved similarly, and they remain vulnerable. Overall, the reform agenda aims to make financial systems safer by improving the shock resistance of bank funding structures and by forcing bank creditors new Basel Capital and liquidity Regulations to assume their contractual obligations. However, the • First and foremost, equity capital plays a quanti- reforms to bank capital and liquidity, to OTC deriva- tatively significant role in reducing the probability tives, and to bank resolution will likely have differ- of bank failures and in lowering the cost of any ent and perhaps unintended consequences for some type of debt. Capital requirement reforms that raise institutions. Specifically, there is a trade-off between, the amount of common equity should be imple- on the one hand, pressuring banks to increase their mented without delay because more equity supports use of more secured funding (raising levels of asset economic growth and mitigates the effects of other

132 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

reforms that may increase the cost of bail-in debt.46 Bail-in powers and depositor preference The positive effects on the cost of debt are dispro- • When the proportion of preferred creditors is too portionately large if a bank builds its capital buffer large, a bank may find it difficult to preserve a beyond Basel III requirements. sufficiently large proportion of unsecured debt to • As noted in previous issues of the GFSR, supervisors absorb losses if capital is exhausted. In such cases, should continue to implement Basel III liquidity minimum bail-in debt requirements can be used. standards as agreed. The new global liquidity stan- By the same token, depositor preference regimes can dards for the liquidity capital ratio and net stable usefully signal to depositors the likelihood that they funding ratio are designed to discourage short-term will receive their deposits in case of bank distress wholesale funding, and they are unlikely to result in and thereby prevent runs and support financial rapid, large-scale changes in banks’ funding struc- stability. To the extent that a deposit guaran- tures, in part because many banks already satisfy the tee scheme is already in place, a tiered depositor requirements and their implementation is gradual. preference structure is desirable—one that prefers Indeed, the standards for the net stable funding insured deposits (and the deposit guarantee scheme ratio have yet to be completed, and early agreement that substitutes for such depositors when liquida- would help lessen uncertainties surrounding its final tion takes place, that is, through subrogation) over contours. uninsured deposits and that prefers both over other senior unsecured creditors, as this will help to lower Concentration in funding and Asset encumbrance contingent claims on the government. • To the extent that bail-in powers and depositor • Although the reforms should continue to be imple- preference reduce demand for debt issued by banks mented on the current timelines, regulators and regarded as systemically important, market disci- policymakers need to monitor the effects of all the pline is enhanced because these banks no longer policies that increase demand for collateral (including receive a funding advantage. Traditional long-term the introduction of the liquidity capital ratio and net buyers of senior bank debt—insurers and pension stable funding ratio as well as reforms to OTC deriva- funds—appear to be willing to purchase bail-in debt tives) and weigh the resulting asset encumbrance if the issuing banks are able to maintain stand-alone against the resilience to liquidity risk and lower coun- investment-grade ratings and carry sufficient equity terparty risks. Limits on encumbrance, for example, capital buffers. If the debt turns out to be too risky on covered bonds, may be one way of ensuring a for traditional holders even at higher yields, a differ- diversified funding structure and the benefits from ent investor base may develop. Regardless, it will be other reforms. However, consideration would need important to ensure that all investors are fully aware to be given to different business models and country of the risks they assume by means of appropriate circumstances. In particular, the introduction of limits disclosures of the terms under which they could be on encumbrance during a period of funding stress bailed in. This calls for greater clarity around the may be counterproductive, limiting the ability of statutory criteria used by resolution authorities for banks to obtain necessary funding through the use of, putting a bank into resolution and for applying the for instance, covered bonds. bail-in tool, among others. Hence, appropriately • Market discipline and appropriate risk-pricing balanced with other reforms, bail-in powers and mechanisms for bank debt can be enhanced by depositor preference can be effective ways to limit requiring banks to provide regular, standardized government bailouts and enhance financial stability. public disclosure of their liability structures and asset • The timing of any introduction of depositor encumbrance. Appropriately priced liabilities are preference or bail-in powers should be carefully important for ensuring that good risk- and burden- considered, taking into account the specific fund- sharing arrangements exist across all stakeholders. ing structures of banks in each country and their vulnerability to systemic funding shocks. If systemic financial stress is low, depositor preference or bail-in 46See Chapter 4 of the October 2012 GFSR for an estimate of the powers could usefully be introduced sooner rather positive implications of higher capital buffers on output growth. than later, so as to be in place in advance of bank

International Monetary Fund | October 2013 133

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

failures. However, in countries in which balance earlier for U.S. banks), while the Bloomberg cover- sheet repair and the restructuring of distressed age starts as early as 1990 globally. The analysis ends institutions are still under way, the introduction with data for 2012. The empirical analysis also uses of these measures could inadvertently increase the IMF and World Bank macrofinancial time series and likelihood of failures. For example, the recent shift governance indicators. The definitions of the variables in nonresident holdings of bank debt securities in and data sources are in Table 3.3. the euro area suggests that the risks may be getting more localized and concentrated in some countries. At the same time, distressed banks rely increasingly determinants of Bank funding patterns on secured funding. Hence, the ongoing risk of To answer the question of what drives bank funding a recurrent systemic liquidity crisis highlights the structures, the following panel regression model is urgency of first dealing with distressed banks, before estimated: introducing depositor preference or bail-in powers. Z = α BANK + β MACFIN + γ REG Discussions within the EU appear to be focusing ijt ijt–1 jt–1 jt + δ Z + Fixed effects + ε , (3.1) on dates far enough in the future to reduce the risk ijt–1 ijt that the introduction will be destabilizing, but only in which Z denotes a source of funding (bank equity, if balance sheet repair and restructuring are accom- debt, or deposits expressed as a fraction of total assets) plished first. or the loan-to-deposit ratio. BANK is a vector of bank- Overall, some bank funding structures are more specific factors, MACFIN is a vector of macro-financial closely associated with financial stability than others. factors, REG is a vector of institutional and governance Many banks in emerging market economies already indicators, and ε is the model’s residual for bank i in have safer structures than do their advanced economy country j in year t.47 The coefficients (or coefficient counterparts, and some of the reforms discussed above vectors) to be estimated are α, β, γ, and δ. Separate may not be necessary in those economies. Regardless ordinary-least-squares panel regressions are estimated of the funding structure, however, any type of bank for each source of funding, with and without cross- debt is safer and less costly when there is adequate section and time-fixed effects, using robust standard equity capital in place. Therefore, policymakers in errors. Models are estimated in levels because funding both advanced and emerging market economies must variables do not contain unit roots by construction continue to pay close attention to this component, so (funding structure shares are bound between zero that these other reforms can achieve their intended and 1), but include a lagged dependent variable to objectives. account for slow adjustment toward a preferred funding structure.48,49 The general-to-specific approach is applied to arrive at the final specification for each Annex 3.1. data description and Additional funding source. empirical Results The empirical evidence here indicates that bank This annex describes the data sources, contains technical funding structures are affected mainly by bank-specific background, and provides key results from the empirical factors, but also by macrofinancial and market vari- analysis in this chapter. ables as well as by the regulatory environment. The data Sources and Coverage 47See Table 3.3 and Gudmundsson and Valckx (forthcoming) for The analysis is based on detailed bank-level balance a more detailed description of the explanatory variables and expected sheet and market statistics from listed and unlisted signs. banks in advanced and emerging market economies. 48To attenuate potential endogeneity, explanatory variables are Table 3.2 reports country and bank coverage statistics. lagged one period. 49Gropp and Heider (2010); Octavia and Brown (2010); Brewer, Primary data sources are SNL Financial and Dealogic Kaufman, and Wall (2008); Demirgüç-Kunt and Huizinga (2010); for the stylized developments in funding patterns. Rauh and Sufi (2010); Lemmon, Roberts, and Zender (2008); Bloomberg, L.P., is used for the empirical analysis. and Antoniou, Guney, and Paudyal (2008) also analyzed bank and nonfinancial companies’ funding structures, using similar firm- SNL Financial’s annual data coverage starts in 2005 or specific variables but different country samples or time periods. See 2007 for banks outside the United States (somewhat Gudmundsson and Valckx (forthcoming) for a detailed review.

134 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Table 3.2. Country and Bank Coverage Statistics Bloomberg, L.P. Sample SNL Financial Sample Advanced Economies Emerging Market Economies Advanced Economies Emerging Market Economies All Euro Area Other All Asia CEE AE Euro Area Other EM Asia CEE Australia 6 Austria Australia Argentina 6 China Croatia Australia 16 Austria Australia Bulgaria 8 China Bulgaria Austria 7 Belgium Canada China 16 India Poland Austria 16 Belgium Canada China 77 India Croatia Belgium 3 Cyprus Denmark Croatia 5 Indonesia Russia Belgium 11 Cyprus Czech Republic Croatia 14 Indonesia Hungary Canada 7 Finland Hong Kong SAR India 30 Malaysia Turkey Canada 11 Finland Denmark Hungary 8 Malaysia Lithuania Cyprus 3 France Japan Indonesia 29 Philippines Cyprus 4 France Hong Kong SAR India 38 Pakistan Poland Denmark 10 Germany Norway Malaysia 9 Thailand Czech Republic 6 Germany Iceland Indonesia 25 Philippines Romania Finland 2 Greece Singapore Philippines 14 Denmark 40 Greece Japan Lithuania 5 Thailand Russia France 7 Ireland Sweden Poland 13 Latin America Finland 5 Ireland Korea Malaysia 20 Vietnam Turkey Germany 6 Italy Switzerland Russia 36 Argentina France 34 Italy New Zealand Pakistan 4 Ukraine Greece 13 Malta United Kingdom Thailand 11 Germany 75 Luxembourg Norway Philippines 8 Hong Kong SAR 7 Netherlands United States Turkey 14 Greece 12 Malta Sweden Poland 16 Ireland 3 Portugal 183 Hong Kong SAR 20 Netherlands Switzerland Romania 9 Italy 16 Slovenia Iceland 3 Portugal Taiwan Province Russia 33 Japan 52 Spain Ireland 13 Slovak Republic of China Thailand 14 Malta 3 Italy 56 Slovenia United Kingdom Turkey 17 Netherlands 3 Japan 42 Spain United States Ukraine 11 Norway 3 Korea 14 Vietnam 5 Portugal 6 Luxembourg 14 312 Singapore 3 Malta 4 Slovenia 5 Netherlands 14 Spain 10 New Zealand 7 Sweden 4 Norway 27 Switzerland 9 Portugal 7 United Kingdom 9 Singapore 4 United States 46 Slovak Republic 6 243 Slovenia 3 Spain 55 Sweden 6 Switzerland 37 Taiwan Province of China 21 United Kingdom 30 United States 75 688 Source: IMF staff. Note: AE = advanced economies; CEE = central and eastern Europe; EM = emerging market economies. Number of banks effectively used in the computations and estimations is indicated after the country’s name. Banks are stand-alone legal entities (subsidiaries) within the country in question. SNL Financial data cover both listed and nonlisted banks (top 100 by assets for the United States) which are either operating or acquired/defunct companies from North America, Europe, and the Asia-Pacific region. The Bloomberg sample contains listed and nonlisted banks from western and eastern Europe, developed and developing Asia, and North and Latin America, retrieved using Bloomberg’s EQS and PSCR functions.

main results are as follows, focusing on statistically and than for advanced economy banks (–0.3 percent- economically50 significant variables: age point) and systemically important banks (–0.5 • Partial adjustment to preferred funding levels: A 1 percentage point). standard deviation shock to banks’ funding sources • Profitability and securities and tangible assets: Banks can shift funding sources by between 1 and 3 that pay dividends and those with higher profit- percentage points (5 to 10 percentage points for ability have lower equity ratios (–0.3 percentage the loan-to-deposit ratio). The impact is larger point for a 1 standard deviation shock to the return since 2007 compared with the precrisis period and on assets). Safer banks, with more securities and for emerging market economy banks relative to tangible assets to total assets, tend to have lower advanced economy banks. Also, comparable shocks debt ratios (–1 percentage point for every 1 standard to equity funding result in smaller adjustments than deviation increase in “tangibility”) and lower loan- do debt and deposit shocks. to-deposit ratios (about a 3 percentage point impact • Size: Larger banks have less equity and more debt from a 1 standard deviation shock). (and higher loan-to-deposit ratios). The reduction • Growth and currency volatility: Banks in countries in equity ratios is proportionately larger for emerg- with (1 standard deviation) higher GDP growth ing market economy banks (–1.6 percentage points) experience about 2 percentage points less debt and higher deposit and equity-to-asset ratios (and 4 percentage point lower loan-to-deposit ratios). 50Economic significance is gauged by 1 standard deviation shocks to the explanatory variables, which makes their effects comparable. Higher currency volatility reduces debt reliance (and

International Monetary Fund | October 2013 135

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 3.3. List of Variables Used in the Panel Data Analysis Variable Definition Data Source Dependent Variables Equity/Assets Total equity divided by total assets Bloomberg, L.P. Deposits/Assets Customer deposits divided by total assets Bloomberg, L.P. Debt/Assets Nondeposit liabilities divided by total assets Bloomberg, L.P. Explanatory Variables Bank-Specific Variables Size Log of total assets Bloomberg, L.P. Profitability Pretax income divided by total assets Bloomberg, L.P. Growth Assets Annual growth in total assets Bloomberg, L.P. Dividend Payer Dummy that equals 1 if bank pays dividend Bloomberg, L.P. Collateral Securities + interbank assets + fixed assets divided by total assets Bloomberg, L.P. Business Model I Share of net interest income to interest and noninterest income Bloomberg, L.P. Business Model II Loans to total assets Bloomberg, L.P. Asset Quality Loan loss provisions to loans Bloomberg, L.P. Macroeconomic and Financial Market Variables GDP Growth Annual growth rate of real GDP WEO Inflation Annual change in the consumer price index WEO Interest Spread Long-term bond yield minus short-term interest rate WEO, WB Stock Return Annual change in the country’s main stock market index Bloomberg, L.P., WB Bond Market Capitalization Outstanding volume of nonfinancial corporate bonds to GDP BIS, WB Stock Market Capitalization Outstanding volume of stock market capitalization to GDP WB Household Saving Ratio Household savings to disposable income WEO, WB Government Debt General government gross debt to GDP WEO Openness I Current account surplus or deficit, percent of GDP WEO Openness II Exports plus imports to GDP WEO Openness III External positions of reporting banks vis-à-vis individual countries’ banks (difference BIS Locational Banking Statistics between all sectors and nonbanks) relative to GDP Foreign Exchange Volatility Standard deviation of monthly currency rate return against SDR IFS Stock Market Volatility 260-day standard deviation of daily stock returns Bloomberg, L.P. Banking Crisis Dummy Dummy variable that equals 1 if the country experiences a systemic banking crisis for the Laeven and Valencia (2012) duration of the crisis Regulatory and Institutional Variables Government Effectiveness1 Perception of the quality of public services, the quality of the civil service and the degree WB of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies Regulatory Quality1 The ability of the government to formulate and implement sound policies and regulations WB that permit and promote private sector development Rule of Law1 The extent to which agents have confidence in and abide by the rules of society, and WB in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence Accountability and Voice1 Perception of the extent to which a country’s citizens are able to participate in selecting their WB government, as well as freedom of expression, freedom of association, and a free media Legal Origin A dummy variable that identifies the legal origin of the company law or commercial code La Porta and others (2000) of each country. The five origins are English, French, German, Nordic, and socialist. Bank Funding Structure Variables Senior Debt Principal amounts outstanding on loans, notes payable, bonds, securities sold under SNL Financial repurchase agreements (repos), mortgage-backed bonds, short-term borrowing, mortgage notes and other notes payable, capitalized lease obligations, and other debt instruments not classified as subordinated debt Subordinated Debt Debt in which the creditor’s claims to the bank’s assets are subordinated to those of SNL Financial other creditors Total Equity Includes par value, paid-in capital, retained earnings, and other adjustments to equity. SNL Financial Minority interest may be included per relevant accounting standards. Wholesale Funding Ratio Interbank borrowing, repo debt, and senior and subordinated debt relative to total debt SNL Financial and customer deposits Secured Funding Ratio Secured senior debt relative to both secured and unsecured senior debt outstanding, DCM Analytics aggregated by country Core Tier 1 Capital Core common capital as defined by regulatory guidelines SNL Financial Additional Tier 1 Capital Tier 1–eligible hybrid capital securities, reserves, and allowances; minority interests; and SNL Financial other Tier 2 capital adjustments as defined by the bank’s domestic central bank/regulator Tier 2 Capital Tier 2–eligible hybrid capital securities, reserves, and allowances; minority interests; and SNL Financial other Tier 2 capital adjustments Tier 3 Capital Eligible subordinated debt and other capital adjustments SNL Financial Source: IMF staff. Note: BIS = Bank for International Settlements; IFS = IMF, International Financial Statistics Database; SDR = special drawing right; WB = World Bank; WEO = IMF, World Economic Outlook Database. 1Governance indicators are in units of a standard normal distribution, with mean zero, standard deviation of 1, and ranging from approximately –2.5 to 2.5, with higher values corresponding to better governance. Data are taken from the World Bank Doing Business Database.

136 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

lowers loan-to-deposit ratios) by about 1 percent various 0–1 dummy variables are constructed to char- (2 percent) in the full sample and for systemically acterize banking distress: important banks. • Balance sheet distress: Bank z-scores below 3, which • Savings and deposits: Banks in countries with (1 stan- corresponds to the lowest decile of the panel series’ dard deviation) higher household savings rates enjoy distributions, are used as an indicator of potential (between 0.5 and 0.8 percentage point) higher deposit capital shortfall. financing and have lower loan-to-deposit ratios. • Bank equity price distress: Price-to-book ratios below • Regulatory factors: Systematically important banks in 0.5, which comprises the lowest 7.5 percent of the countries with high-quality regulatory environments banks, are used. Stock returns falling by 60 to 90 (“Regulation”) have more than 1 percent higher percent during a given year are also considered and deposits. Banks in countries with stronger disclosure yield broadly similar results. have marginally higher equity and deposit ratios and • Analysts’ ratings: Bank equity analysts’ ratings (on a lower loan-to-deposit ratios. scale of 1 to 5, with 1 a strong sell and 5 a strong buy) below 2.5, which corresponds to the 10 percent left tail, are used. Bank funding patterns and financial Stability The exercise uses five different characteristics of bank In line with recent studies, we examine whether bank funding: funding structures have an impact on financial stability • Loan-to-deposit ratio: roughly corresponds to the when combined with other bank characteristics and wholesale funding ratio because it measures the macrofinancial factors.51 Using the Bloomberg panel deposit funding gap to be filled by debt (or equity); data set described previously, we estimate a (panel/ • Funding concentration: a Herfindahl index of bank pooled time series) probit model: funding structure (sum of squared percentages of debt, deposits, and equity), with higher values indi- P{Distress | X , Z } = F(β X + β Z ), (3.2) ijt ijt–1 jt–1 ij ijt–1 j jt–1 cating less diverse funding; in which P{} is the probability that bank i from • Short-term debt funding: the share of debt expiring country j will be in distress at time t, conditional on within the year, as a share of total bank debt; bank-specific and country-level characteristics Xijt and • Banks’ debt-to-asset ratios: the ratio of debt liabilities Zjt. F() is the standard normal distribution function to total assets; and that transforms a linear combination of the explana- • Banks’ equity-to-asset ratios: the ratio of total equity tory variables into the [0,1] interval. The estimations to total assets. use lagged explanatory variables to reduce endogene- The assumption is that higher loan-to-deposit ratios, ity concerns and report robust standard errors. The less diverse funding sources, higher reliance on short- general-to-specific approach is applied to arrive at the term debt funding, and higher leverage will increase final probit specifications. banks’ probabilities of distress. Given that the data do not directly provide bank Other bank-specific factors and general macroeco- status characteristics (default versus going concern), nomic conditions are controlled for. These include size, asset growth, the loan loss provision ratio, 51Vazquez and Federico (2012) find that European and U.S. banks real GDP growth, inflation, the interest rate term with higher net stable funding ratios (NSFRs) and equity-to-asset spread, as well as the broad stock market return and ratios before the 2008–09 crisis had lower crisis failure probabilities. Demirgüç-Kunt and Huizinga (2010) showed that wholesale funding volatility. and banks with higher noninterest income experience higher average The results suggest that, in addition to bank fund- fragility, for banks from 101 countries. Demirgüç-Kunt, Detragia- ing, some other bank characteristics, as well as the che, and Merrouche (2010), for 313 banks from 12 countries, and macrofinancial and broad regulatory environment, Beltratti and Stulz (2009), for 98 large banks from 20 countries, find that better capitalized banks (and large, more deposit-financed significantly affect banks’ distress probabilities (Figure banks) saw smaller stock price declines during the crisis, whereas 3.18). Focusing on 1 standard deviation shocks away wholesale funding increased bank fragility. Bologna (2011) and from the mean, the impact on distress probabilities are Berger and Bouwman (2013) find that U.S. banks with less stable deposit funding were more likely to fail, controlling for nonperform- as follows: ing loans and capital ratios. Huang and Ratnovski (2009) find that • Size: Bigger advanced economy banks and larger deposit funding contributed to the stability of banks in Canada and systemically important banks seem 3.6 to 4.5 percent 72 other large Organization for Economic Cooperation and Devel- opment country banks during the crisis. more likely to be under stress (under the price-to-book

International Monetary Fund | October 2013 137

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.18. Contribution of Speciﬁc Variables to Bank Distress in Probit Models (Relative size; percentage points)

Z-score Price-to-book ratio Analysts’ ratings 1. All Banks (1990–2012) 2. Systemically Important Banks

Size Size Total assets growth Total assets growth LLP ratio LLP ratio NII share NII share Loan-to-deposit ratio Loan-to-deposit ratio Short-term debt Short-term debt Debt ratio Debt ratio Equity ratio Equity ratio Concentration GDP growth GDP growth Inﬂation Inﬂation Yield spread Yield spread Stock return Stock return Stock volatility Stock volatility Regulation Regulation Disclosure Disclosure –6 –4 0–2 2 4 6 –4 –2 0 2 4 6 3. All Banks (1990–2006) 4. All Banks (2007–12)

Size Size Total assets growth LLP ratio LLP ratio NII share NII share Loan-to-deposit Loan-to-deposit ratio Short-term debt Short-term debt Debt ratio Debt ratio Equity ratio Equity ratio Concentration Concentration GDP growth GDP growth Inﬂation Inﬂation Yield spread Yield spread Stock return Stock return Stock volatility Stock volatility Regulation Regulation Disclosure –6 –4 0–2 2 4 6 –6 –4 –2 0 2 4 6

5. Advanced Economy Banks 6. Emerging Market Economy Banks

Size Size LLP ratio Total assets growth NII share NII share Loan-to-deposit ratio Loan-to-deposit ratio Short-term debt Short-term debt Debt ratio Debt ratio Equity ratio Equity ratio Concentration GDP growth GDP growth Inﬂation Inﬂation Yield spread Yield spread Stock return Stock return Stock volatility Stock volatility Regulation Regulation Disclosure Disclosure –4 –2 0 2 4 6 –10 –5 0 5 10 15

Sources: Bloomberg, L.P.; and IMF staff estimates. Note: LLP = loan loss provisions; NII share = net interest income in percent of operating income. Regulation and disclosure are the first and second principal component scores, derived from the four World Bank indicators of regulatory and institutional quality. See Table 3.3 for details on factors and their definitions. Figure shows the economic significance of bank and country characteristics, evaluated at the variable’s mean plus 1 standard deviation on the probability of distress specified under alternative distress models and samples. Bank distress is a dummy variable, defined either as a z-score below 3, price-to-book ratio below 0.5, or average analyst ratings of 2.5 or lower. Different probit estimations are performed for the full 1990–2012 sample (all banks), the 2007–12 period, advanced economy banks, and emerging market economy banks. The emerging market economy sample contains banks from developing Asia and central and eastern Europe.

138 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

measure), whereas large emerging market economy Annex 3.2. Regulatory developments Affecting banks seem 2 to 4 percent less likely to be under stress. Bank funding • Asset growth: More rapidly growing emerging market This annex summarizes the details of Basel III capital and economy banks seem less likely to be under stress, liquidity regulation and proposals for strengthening resolu- by up to 8 percent. tion framework for financial institutions. • Asset quality: Banks with higher loan loss ratios have up to 4.5 percent higher distress probabilities with the z-score measure (but not using the price-to-book Basel iii Capital Regulation or analysts’ ratings distress measures). • Retail versus wholesale focus: Banks with more tradi- Basel III capital regulations require more and better tional business (higher net interest income to total capital than do Basel II regulations. The majority of income) experience slightly lower distress overall, the minimum capital requirement should be of the especially using the analysts’ ratings-based distress highest quality (common equity). Various buffers are measure. However, chances of distress for more tra- added for macroprudential purposes or to account ditional advanced economy banks and systemically for the systemic relevance of some institutions (Figure important banks increase by 1 to 1.5 percent using 3.19). Basel III also requires more capital to bet- the z-score measure. In the same vein, banks with a ter cover risks from securitization, the trading book more wholesale orientation experience substantially (including proprietary trading), and banks’ exposures higher distress probabilities, of 4 percent or more, to derivative counterparties, other financial institu- with some measures. tions, and central counterparties (namely, counterparty • Funding structure, debt, and equity: Increases in risks). A non-risk-based leverage ratio will be added short-term debt, or in overall debt ratios for emerg- to minimum requirements in 2018 and could stem a ing market economy banks and systemically impor- buildup in leverage caused by off-balance-sheet expo- tant banks, raise banks’ distress probabilities by 1 to sures and repo transactions. 4 percent. Higher equity buffers, however, uniformly lower distress probabilities across all measures, by up to 5.5 percent. Basel iii liquidity Regulations • GDP growth, yield spreads, and inflation: Higher The Basel III liquidity regulation includes two quanti- growth results mostly in 0.5 to 2.5 percent lower tative ratios: the liquidity capital ratio (LCR) and the banking sector distress. Similarly, higher yield net stable funding ratio (NSFR). The LCR assesses spreads reduce the likelihood of distress using the shorter-term (30-day) vulnerability to liquidity shocks, z-score and analysts’ ratings measures by 1.0 to and the NSFR aims to reduce maturity mismatches 2.5 percent (but using the price-to-book measure, over one year. Specifically: higher yield spreads raise distress). Banks in higher- • The LCR is defined as the stock of high-quality inflation countries are more likely to be in distress liquid assets as a proportion of the bank’s net cash according to the z-score measures (+5 percent dis- outflows over a 30-day time period. Banks will be tress for emerging market economy banks), whereas required to maintain a 100 percent LCR when the the price-to-book and analyst ratings measures phase-in period ends in 2019. The size of the net indicate the reverse, probably reflecting the possibil- outflow is based on assumed withdrawal rates for ity of hedging against inflation with stocks (up to 4 short-term liabilities, according to their stability percent lower distress). (for example, withdrawal rates are lower for insured • Stock return and volatility: Higher market returns retail deposits than for deposits from corporations and lower volatility are beneficial to banking stabil- and nonresidents) and the potential drawdown of ity and are significant across the various specifica- contingency facilities. Having more long-term debt tions, with effects on distress probabilities broadly (maturities greater than 30 days) is positive for the between 1.0 and 2.5 percent. LCR, because its associated outflow within 30 days • Regulatory quality and disclosure: Stronger and better- is zero. quality regulatory environments, as well as countries • The NSFR is defined as a bank’s available stable with higher disclosure requirements, reduce banking funding (ASF) divided by its required stable funding distress probabilities by between 1 and 5 percent. (RSF) and must be greater than 100 percent. Each

International Monetary Fund | October 2013 139

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.19. Basel III Minimum Capital Requirements and Buffers (Percent of risk-weighted assets)

Minimum Buffers G-SIB charge: 2 percent T2 G-SIBs must have higher loss-absorbency capacity to reﬂect the greater risks that they pose to the 1.5 percent AT1 ﬁnancial system, ranging by institution from 1 to 2.5 1–2.5 percent CET1 percent. G-SIB charge Countercyclical buffer: 0–2.5 percent CET1 A buffer is added if supervisors judge that credit countercyclical buffer growth is leading to unacceptable systemic risk 2 percent T2 2.5 percent CET1 buildup. 1.5 percent AT1 conservation buffer Conservation buffer: 10.5%–15.5% total capital 8.5%–13.5% T1

7%–12% CET1 Constraint on a bank's payout (e.g., dividends) is 4.5 percent CET1 imposed when banks fall within the buffer range. 4.5 percent CET1 6% T1 7% CET1

8% total capital minimum

Source: IMF staff based on BCBS (2010a). Note: AT1 = additional Tier 1; CET1 = common equity Tier 1; G-SIB = global systemically important bank; T2 = Tier 2. The G-SIB surcharge, in principle, could be as high as 3.5 percent, but currently no SIBs are charged more than 2.5 percent.

asset category is assigned an RSF “factor,” which is however, some countries already had bail-in powers or lower for liquid assets and higher for illiquid assets. depositor preference (for example, the United States). Similarly, ASF factors are assigned to each liability Depositor preference provides seniority to some deposi- category, and the factors are higher for more stable tors over other senior unsecured debt holders at liquidation liabilities (for example, capital, long-term debt (Table 3.4). Liquidation is a form of resolution in which with a maturity of more than one year, and insured the bank’s assets are sold and the values recovered are used deposits) and lower for less stable funding (for to pay creditors in the order of priority. Without depositor example, short-term wholesale funding). preference, insured depositors hold the same seniority as other senior unsecured debt holders; therefore, their recovery ratios (without considering payouts from the deposit Reform Agenda for Resolution frameworks guarantee scheme) at the time of a bank failure are the Despite an agreement on the broad initiatives for same (examples [A] and [E] in Table 3.4). In a liquidation strengthening resolution frameworks, there is not yet full with depositor preference, and when asset recoveries are agreement on some specific aspects, including the scope insufficient to repay all senior creditors, depositors are paid of bail-in, depositor preference, and minimum holdings before senior unsecured debt holders, and their recovery of bail-in debt. At the global level, the Financial Stability ratios are higher (examples [B] and [F] in Table 3.4). The Board’s Key Attributes of Effective Resolution Regimes for formal introduction of depositor preference with bail-in Financial Institutions were agreed to by the G20 in 2011 powers would help to limit legal challenges and claims and cover both bail-in powers for authorities and protec- for compensation in cases of resolution, even if the bank tion of insured depositors. Many countries are making is not liquidated, making bail-in powers more effective. progress in implementing them, and a date of the end of There are two main forms of depositor preference. 2015 has been set.52 Even before the global initiatives, The specifics of existing and proposed forms vary across countries, suggesting that the share of preferred deposits 52At the same time, separate proposals in individual countries in total liabilities varies substantially (Table 3.5).53 or regions have emerged, including the Dodd-Frank Act (Dodd- Frank [2012]) in the United States, which has provisions for bank resolution, and the EU’s Recovery and Resolution Directive. The EU Banking Sector, 2012) also discuss providing bail-in powers to agreement of the European Commission on the directive would, if authorities and raising the loss-absorbing capacity of banks. enacted, introduce depositor preference and phase in bail-in powers 53Several countries have some forms of depositor preference in the European Union. In addition to the legislative proposals, legislation in place, including Argentina, Australia, Austria, Belgium, recommendations by high-level committees and expert groups such China, Germany, Greece, Hong Kong SAR, Italy, Latvia, Norway, as the U.K.’s Vickers’ report (ICB, 2011) and the EU’s Liikanen Portugal, Romania, Russia, Singapore, Spain, Switzerland, the report (High-Level Expert Group on Reforming the Structure of the United Kingdom, and the United States.

140 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks 4 . . 0% 0% 100% equity 46% + $7 $0 $0 . debt deposits 70–40–7 $7 equity $0 equity unsecured $23 senior $40 insured $0 sub. debt $0 sub. Continue $70 [D]:Bail-in, with exclusion of insured deposits with exclusion [D]:Bail-in, Remaining value of asset . . 100% 100% 100% 100% 3 . . . $30 debt deposits $8 equity unsecured $50 senior $40 insured [C]:Bail-out $2 sub. debt $2 sub. Continue $70 Bail-out Remaining subsidy $30 value of asset 2 . . 0% 0% 60% 100% [B]:Liquidation Without Asset Encumbrance $0 $0 . . debt (+DGS and insured DP) 70–40 deposits $0 equity unsecured $30 senior $40 insured $0 sub. debt $0 sub. Discontinued 1 . 0% 0% 78% 78% 100% (78% without DGS) ©International Monetary Fund. Not for Redistribution ©International Monetary $9 $0 debt deposits unsecured $39 senior $40 insured [A]:Liquidation (+DGS, no DP) [A]:Liquidation (+DGS, (including $9 (including Discontinued DGS payment) 70x(40/[40+50]) 70x(50/[40+50]) 40–70x(40/[40+50]) debt deposits $8 equity $0 equity unsecured $50 senior $40 insured $2 sub. debt$2 sub. debt $0 sub. Balance sheet $70 assets Assets Liabilities Liabilities Recovery Liabilities Recovery Assets Liabilities Recovery Assets Liabilities Recovery Remaining value of asset $30 losses on Bank operations net DGS cost, Taxpayer cost, excluding DGS excluding cost, Taxpayer Recovery value calculations DGS/insured deposits without DGS Senior unsecured debt DGS net payment to depositors Table 3.4. Illustration of Creditor Hierarchy and Loss Sharing under Alternative Resolution Tools 3.4. Table

International Monetary Fund | October 2013 141 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY . . 0% 0% 100% 100% equity 30% + $7 4 $0 $0 . deposit $7 equity $0 equity unsecured $13 senior $40 insured $10 secured $0 sub. debt $0 sub. 70–10–40–7 secured debt Continue . . $70 Remaining [H]:Bail-in, with exclusion of insured deposits and with exclusion [H]:Bail-in, value of asset . . 100% 100% 100% 100% 100% 3 . . . . $30 debt deposits $8 equity unsecured $40 senior $40 insured [G]:Bail-out $10 secured $2 sub. debt $2 sub. Continue (concluded) $70 Bail out Remaining subsidy $30 value of asset 5 2 . . 0% 0% 50% 100% 100% With Asset Encumbrance [F]:Liquidation $0 $0 debt (+DGS and insured DP) deposits $0 equity unsecured 70–10–40 $20 senior $40 insured $10 secured $0 sub. debt $0 sub. Discontinued 1 . 0% 0% 75% 75% 100% 100% (75% without DGS) ©International Monetary Fund. Not for Redistribution ©International Monetary $0 debt $10 [40+40]) [40+40]) [40+40]) deposits unsecured $30 senior [E]:Liquidation (+DGS, no DP) [E]:Liquidation (+DGS, $40 insured (70–10)x(40/ (70–10)x(50/ Discontinued (including $10 (including DGS payment) 40–(70–10)x(40/ debt deposits $8 equity $0 equity unsecured $40 senior $40 insured $10 secured $10 secured $2 sub. debt$2 sub. debt $0 sub. Balance sheet $70 assets Assets Liabilities Liabilities Recovery Liabilities Recovery Assets Liabilities Recovery Assets Liabilities Recovery Remaining value of asset $30 losses on Bank operations net DGS cost, Taxpayer cost, excluding DGS excluding cost, Taxpayer Recovery value calculations DGS/insured deposits without DGS Senior unsecured debt DGS net payment to depositors Liquidation with DGS: Insured-depositor claims are paid in full by the DGS and the DGS “steps into” the rights of the insured depositors in liquidation proceedings. DGS (taking the place of depositors) ranks equally (pari passu) with other senior the rights of insured depositors in liquidation proceedings. “steps into” are paid in full by the DGS and Insured-depositor claims Liquidation with DGS: ranks above senior under DP, DGS (taking the place of depositors), the rights of insured depositors in liquidation proceedings. “steps into” are paid in full by the DGS and Insured-depositor claims Liquidation with DGS and DP: The government could instead inject capital (perhaps after writing off the government is assumed to make good on losses caused by too-big-to-fail concerns through an asset-protection scheme without charge. In the example, Bail-out: (This does not necessarily imply giving DP Attributes. Key as per the FSB’s Insured deposits are assumed to be exempt from bail-in, under the bail-in power given to the countryThe bank is rehabilitated authority. with debt restructuring, Bail-in: If the remaining value of assets is secured debt is assumed to be fully collateralized. simplicity, For on the collateral Secured debt holders haveassets to (insured) deposits and senior unsecured are exempt from bail-in. senior claim Table 3.4. Illustration of Creditor Hierarchy and Loss Sharing under Alternative Resolution Tools Illustration of Creditor Hierarchy and Loss Sharing under Alternative Resolution Tools 3.4. Table 1 unsecured creditors. 2 unsecured creditors in the proceedings. 3 and secured debt holders. depositors, while protecting the payments to senior debt holders, subordinated debt and equity) to absorb the losses add fresh equity buffers, 4 in liquidation—a country with a bail-in resolution regime may have a as liquidation in framework cases with [A] DP, and Losses [E]). are first absorbed by equity and subordinate debt holders andThe conversion values are set to achieve a 10 percent capital-to-asset then ratio in this example. by senior unsecured are partly written down and partly converted into new equity. claims debt whose holders, 5 below $50 in cases [F] and losses [H], are first absorbed by depositors (through the Secured DGS). debt holders experience losses only when the asset value falls below $10. Source: IMF staff. Source: uninsured deposits. Senior unsecured debt includes rounding errors. Loss and recovery amounts include DGS = depositor guarantee scheme; DP preference. Note:

142 International Monetary Fund | October 2013 ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Table 3.5. Cross-Country Comparison of Covered Deposits, end-2010 Total Domestic Deposit Base Covered Deposits Eligible Deposits (billions of U.S. dollars) (percent of total) (percent of total) Cyprus1 128 24 . . . Greece1 169 63 . . . France 1,742 67 92 Germany 3,395 . . . 40 Italy 2,050 31 45 Netherlands 1,202 48 59 Spain 1,963 47 65 Switzerland 1,481 24 73 United States 7,888 79 100 Source: Financial Stability Board (2012a). 1IMF staff estimates.

• Insured depositor preference provides preferential treat- make senior unsecured debt holders worse off than in ment for insured deposits and ranks all other senior liquidation, in which they would be treated equally unsecured creditors, including uninsured deposits, (cases [A] and [E] in Table 3.4). This outcome could equally. lead to a lawsuit. Formal introduction of depositor • General depositor preference gives preference to all preference simultaneously with bail-in power would deposits of a deposit-taking institution, including to align the recovery for debt holders in liquidation and balances higher than the deposit insurance limit over restructuring, limiting legal challenges and claims for senior unsecured creditors. compensation. • Tiered depositor preference prefers insured deposits (and the deposit guarantee scheme through subrogation) over uninsured deposits and prefers both over Annex 3.3. Bank Bond pricing Model senior unsecured creditors. Merton-Style Bond pricing framework for Senior and In contrast to depositor preference, bail-in power is Subordinated debt and equity applied when a bank failure is resolved while keeping The price of a bond that may default depends on the bank operational (see Table 3.4). Junior stakehold- the value of a bank’s assets relative to the face value ers (subordinated debt and shareholders) are the first to of the bond and its seniority rank. Consider a bank lose their stakes, and if these amounts are not sufficient that issues only three types of liabilities, senior and to restore viability, senior debt holders are then bailed subordinated debt as well as equity (Figure 3.20). The in at the discretion of the resolution authority (Table total liabilities of the bank, excluding equity, are $95. 3.4, examples [D] and [H]). Secured debt holders and If the asset value is greater than $95, both senior and some depositors may be exempt from bail-in; therefore, subordinated debt holders (creditors) recover the full their recovery amounts are assumed to be higher than face value of debt and the rest goes to shareholders (for those of the senior unsecured debt holders. example, if the asset value is $110, shareholders receive Several aspects of bail-in power for bank rehabilita- $15). But if the value of assets declines below $95, tion need to be established in advance. The first is the the bank defaults. The recovery after bank failure for scope of bail-in debt as discussed in the main text. debt holders depends on their seniority and the capital The second is establishing when this power would be structure. For instance, if the asset value becomes $50, exercised. The power should be applied when a bank senior debt holders receive $50, while subordinated becomes unviable, which could be any time after a debt creditors and shareholders recover $0. If the asset bank breaches a regulatory capital ratio but before value is between $93 and $95, senior creditors receive it becomes insolvent, and therefore requires further $93, shareholders recover $0, and the rest goes to specificity in legislation. The third is creditor seniority subordinated creditors. order. Bail-in powers could impose losses on credi- The contingent nature of the liabilities suggests tors in a different order and of a different magnitude that a standard option pricing formula can be used to than losses in liquidation. For instance, bailing in valuate these liabilities. The value of equity is the same senior unsecured debt holders while exempting insured as the value of buying a call option—that is, the right depositors (cases [D] and [H] in Table 3.4) could to buy the asset at a strike price of $95—on the bank’s

International Monetary Fund | October 2013 143

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.20. Pricing of Senior and Subordinated Debt and Equity

Balance sheet Senior debt $93 Asset $100 Subordinated debt $2 Equity $5

Balance sheet identity Value of total liability (debt+equity) Value of senior debt Value of subordinated debt Value of equity 100 93

2 5 100 93 93 95 95 100 Value of assets Value of assets Value of assets Value of assets

Source: IMF staff. Note: The balance sheet identity implies that the total value of assets should be equal to the total value of liabilities, which is the vertical sum of the values of senior debt, subordinated debt, and equity. total assets (see Figure 3.20). The value of senior debt Bond pricing with depositor preference or Asset over the risk-free asset is represented by the value of encumbrance selling a put option—the right to sell the asset at the Both depositor preference and asset encumbrance strike price of $93—on the bank’s assets. (that is, the use of secured debt) in effect create an Once the value of equity and senior debt is calcu- additional type of liability that is ranked above other lated, the balance sheet identity determines the value senior debt for pricing purposes. The senior debt in of subordinated debt that sits between senior debt Figure 3.20 is now split into a “preferred creditor” sta- and equity. It is calculated as the difference between tus that is senior to all other debt and to both senior total assets and the sum of the equity and senior debt debt and subordinated debt in Figure 3.21. “Preferred values, because balance sheet identity implies that the creditor” debt in this exercise includes every type of value of all types of debt and equity will sum to the debt that will have preferential ranking over other value of total assets. In other words, a liability with senior debt as a result of depositor preference or asset a seniority ranking between senior debt and equity encumbrance.55 can be modeled as a combination of purchasing a call The changes in seniority ranking affect the value option with a strike price of $93 and selling a call of liabilities relative to the case without asset encum- option with a strike price of $95 as shown in Figure brance or depositor preference (see Figure 3.12 for an 3.20. (Options strategists call this a “vertical spread.”) illustration of the creditor hierarchy). The values of This figure represents the potential payoffs to subordi- equity and subordinate debt remain the same, because nated debt holders at maturity of the option. their seniority ranking is unaffected. Figure 3.21 shows The chapter’s analysis adopts all the assumptions stated by Merton (1974), including that the asset value follows a geometric Brownian motion. The asset tive examples rather than precise estimates; the qualitative analysis, value changes at any given future date are distributed however, is robust. 55To be precise, secured debt holders have seniority only up to 54 normally. Default is assumed to occur only at matu- the value of their collateral assets. However, central bank repurchase rity—that is, the options are “European.” agreements (mostly short term, with haircuts on the collateral assets) and covered bonds (overcollateralization, which implies the collateral is greater than that needed to ensure payments) are structured such that they are very likely to recover full value of the debt. See Chan- 54More complex and realistic processes, including jump-diffusions Lau and Oura (forthcoming) for a fuller analysis of asset encum- or distributions with fatter tails, can be accommodated within brance in the situation in which secured creditors have less than full this framework. The numerical results should be taken as illustra- seniority over other creditors. The quantitative impact is small.

144 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

Figure 3.21. Pricing of Liabilities with Depositor Preference and Asset Encumbrance

Balance sheet Value of preferred creditors Value of subordinated debt Preferred credit $40 40 Senior debt 2 Asset $100 $53 Subordinated 40 93 95 debt $2 Value of assets Value of assets Equity $5

Balance sheet identity Value of total liability (debt+equity) Value of senior debt Value of equity 100

53 5 100 40 93 95 100 Value of assets Value of assets Value of assets

Source: IMF staff. Note: The balance sheet identity implies that the total value of assets should be equal to the total value of liabilities, which is the vertical sum of the values of preferred credit, senior debt, subordinated debt, and equity. how different combinations of puts and call options to-asset ratio calculated using market values falls below on the asset value of the bank can be used to price the a prespecified level, set at 5 percent in this exercise. liabilities. The recoveries for bail-in debt and equity depend on Preferred creditors face losses only when the asset whether the event is triggered (Figure 3.22). Their value declines to less than $40. In contrast, senior debt values can be expressed as a combination of two barrier faces losses when the asset value declines to less than options that have closed-form solutions: a down-and- $93, which is when the equity and subordinated debt out call option that assigns recovery values provided buffers are used up. Moreover, the recovery value of the bail-in is not triggered and a down-and-in call senior debt is also lower than it would be if it were option for when bail-in is applied.56 When bail-in is ranked equally with deposits, given that $40 of the triggered, senior debt holders and existing shareholders assets’ value is reserved to first pay off preferred credi- are assumed to receive new equity in proportion to the tors. The probability of default of senior debt remains market value of their respective claims at the time of the same because the bank defaults whenever the asset bail-in. value is $93 but its recovery value declines, which is reflected in higher yields relative to the case in which senior debt ranks equally with deposits.

56Barrier options are options whose payoffs depend on the strike Bail-in debt price and an additional event. A down-and-out (down-and-in) option ceases to exist (becomes activated) if the value of the underly- When bail-in powers are exercised, all bail-in debt is ing asset falls below a prespecified value, or barrier value, at some assumed to be converted to equity when the equity- point during the life of an option.

International Monetary Fund | October 2013 145

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 3.22. Pricing of Liabilities under Bail-in Power

Balance sheet No bail-in triggered Bail-in triggered (down-and-out call option) (down-and-in call option) Preferred Value of preferred creditors Value of preferred creditors credit $40 Asset $100 Senior debt $53 40 40 Equity $7 40 40 Value of assets Value of assets Value of senior debt Value of senior debt Balance sheet identity Value of total liability X (debt+equity) 53

100 40 93 40 100 Value of assets Value of assets Value of equity Value of equity

100 Value of assets 7 Y 93 100 40 100 Value of assets Value of assets

Source: IMF staff. Note: X and Y depend on the extent of dilution for existing shareholders when bail-in power is applied. In this exercise, senior debt holders and existing shareholders are assumed to receive new equity in proportion to the market value of their respective claims. Suppose SenD* and E* represent the market value of senior debt and equity, respectively, when bail-in kicks in. Senior debt holders receive SenD*/(SenD*+E*) percent of new equity and the rest goes to existing shareholders. The balance sheet identity implies that the total value of assets should be equal to the total value of liabilities, which is the sum of the values of preferred credit, senior debt, subordinated debt, and equity.

146 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution ChApteR 3 Changes In Bank FundIng patterns and FInanCIal staBIlIty rIsks

References Brewer, Elijah, George G. Kaufman, and Larry D. Wall, 2008, “Bank Capital Ratios Across Countries: Why Do They Vary?” Acharya, Viral V., Douglas Gale, and Tanju Yorulmazer, 2011, Journal of Financial Services Research, Vol. 34, pp. 177–201. “Rollover Risk and Market Freezes,” Journal of Finance, Vol. Brunnermeier, Marcus, and Martin Oehmke, 2013, “The Maturity 66, No. 4, pp. 1171–209. Rat Race,” Journal of Finance, Vol. 68, No. 2, pp. 483–521. The Bankers’ New Admati, Anat, and Martin Hellwig, 2013, ———, and Lasse Heje Pedersen, 2009, “Market Liquidity and Clothes: What’s Wrong with Banking and What to Do About It, Funding Liquidity,” Review of Financial Studies, Vol. 22, No. (Princeton, New Jersey: Princeton University Press). 6, pp. 2201–38. Adrian, Tobias, and Hyun Song Shin, 2010, “Liquidity and Calomiris, Charles, 1999, “Building an Incentive-Compatible Leverage,” Journal of Financial Intermediation, Vol. 19, No. 3, Safety Net,” Journal of Banking and Finance, Vol. 23, pp. pp. 418–37. 1499–519. Antoniou, Antonios, Yilmaz Guney, and Krishna Paudyal, 2008, ———, and Charles Kahn, 1991, “The Role of Demandable “The Determinants of Capital Structure: Capital Market-Ori- Debt in Structuring Optimal Banking Arrangements,” Ameri- ented versus Bank-Oriented Institutions,” Journal of Financial can Economic Review, Vol. 81, No. 3, pp. 497–513. and Quantitative Analysis, Vol. 43, No. 1, pp. 59–92. Chan-Lau, Jorge, and Hiroko Oura, forthcoming, “Repricing Arslanalp, Serkan, and Takahiro Tsuda, 2012, “Tracking Global Bank Liabilities: An Option Price Approach,” IMF Working Demand for Advanced Economy Sovereign Debt,” IMF Working Paper (Washington: International Monetary Fund). Paper No. 12/284 (Washington: International Monetary Fund). Committee on the Global Financial System (CGFS), 2011, Barclays Capital, 2013, “CoCos—A Structural Revolution,” “Fixed Income Strategies of Insurance Companies and Pen- Credit Research, April. sion Funds,” CGFS Paper No. 44. Basel Committee on Banking and Supervision (BCBS), 2010a, ———, 2013, “Asset Encumbrance, Financial Reform and the “Basel III: A Global Regulatory Framework for More Resil- Demand for Collateral Assets,” CGFS Paper No. 49. ient Banks and Banking Systems” (Basel; revised June 2011). Demirgüç-Kunt, Asli, and Harry Huizinga, 2010, “Bank Activity ———, 2010b, “Basel III: International Framework for Liquid- and Funding Strategies: The Impact on Risk and Returns,” ity Risk Measurement, Standards and Monitoring” (Basel, Journal of Financial Economics, Vol. 98, No. 3, pp. 626–50. December). Demirgüç-Kunt, Asli, Enrica Detragiache, and Ouarda Mer- ———, 2012, “Results of the Basel III Monitoring Exercise as rouche, 2010, “Bank Capital: Lessons from the Financial of 31 December 2011” (Basel, September). Crisis,” IMF Working Paper No. 10/286 (Washington: Inter- ———, 2013a, “Basel III: The Liquidity Coverage Ratio and national Monetary Fund). Liquidity Risk Monitoring Tools” (Basel, January). Dodd-Frank Wall Street Reform and Consumer Protection Act, ———, 2013b, “Results of the Basel III Monitoring Exercise as 2010, H.R. 4174 (111th Congress), July 21. of 30 June 2012” (Basel, March). European Banking Authority (EBA), 2013, “Consultation Paper ———, 2013c, “Global Systemically Important Banks: Updated Draft Implementing Technical Standards on Asset Encum- Assessment Methodology and the Higher Loss Absorbency brance Reporting under Article 95a of the Draft Capital Requirement” (Basel, July). Requirements Regulation,” EBA/CP/2013/05. Beltratti, Andrea, and René Stulz, 2009, “Why Did Some Banks European Commission (EC), 2012, “Impact Assessment Accom- Perform Better During the Credit Crisis? A Cross-Country panying the Draft Directive Establishing a Framework for the Study of the Impact of Governance and Regulation,” NBER Recovery and Resolution of Credit Institutions and Invest- Working Paper No. 15180 (Cambridge, Massachusetts: ment Firms,” SWD(2012) 166 final, June 6 (Brussels). National Bureau of Economic Research). Farhi, Emmanuel, and Jean Tirole, 2012, “Bubbly Liquidity,” Berger, Allan N., and Christa H. S. Bouwman, 2013, “How Review of Economic Studies, Vol. 79, No. 2, pp. 678–706. Does Capital Affect Bank Performance During Financial Cri- Feldman, Ron, and Jason Schmidt, 2001, “Increased Use of ses?” Journal of Financial Economics, Vol. 109, pp. 146–76. Uninsured Deposits: Implications for Market Discipline,” Berkmen, Pelin, Gaston Gelos, Robert Rennhack, and James Federal Reserve Bank Minneapolis Fed Gazette, March 18–19. Walsh, 2012, “The Global Financial Crisis: Explaining Financial Stability Board (FSB), 2011, “Key Attributes of Effec- Cross-Country Differences in the Output Impact,” Journal of tive Resolution Regimes for Financial Institutions” (Basel). International Money and Finance, Vol. 31, pp. 42–59. ———, 2012a, “Thematic Review on Deposit Insurance Systems” Bertay, Ata Can, Asli Demirgüç-Kunt, and Harry Huizinga, (Basel). forthcoming, “Do We Need Big Banks? Evidence on Perfor- ———, 2012b, “Update of Group of Global Systemically mance, Strategy and Market Discipline,” Journal of Financial Important Banks (G-SIBs)” (Basel). Intermediation. Flannery, Mark, 1998, “Using Market Information in Prudential Bologna, Pierluigi, 2011, “Is There a Role for Funding in Bank Supervision: A Review of the U.S. Empirical Evidence,” Explaining Recent U.S. Banks’ Failures?” IMF Working Paper Journal of Money, Credit and Banking, Vol. 30, No. 3:1, pp. No. 11/180 (Washington: International Monetary Fund). 273–305.

International Monetary Fund | October 2013 147

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Geanakoplos, John, 2009, “The Leverage Cycle,” in NBER Governance,” Journal of Financial Economics, Vol. 58, No. Macroeconomics Annual, Vol. 24, ed. by Daron Acemoglu, 1–2, pp. 3–27. Kenneth Rogoff, and Michael Woodford (Chicago: University Le Leslé, Vanessa, 2012, “Bank Debt in Europe: Are Funding of Chicago Press). Models Broken?” IMF Working Paper No. 12/299 (Washing- Goldsmith-Pinkham, Paul, and Tanju Yorulmazer, 2010, ton: International Monetary Fund). “Liquidity, Bank Runs, and Bailouts: Spillover Effects During Lemmon, Michael L., Michael R. Roberts, and Jaime F. Zender, the Northern Rock Episode,” Journal of Financial Services 2008, “Back to the Beginning: Persistence and the Cross- Research, Vol. 37, No. 2–3, pp. 83–98. Section of Corporate Capital Structure,” Journal of Finance, Gorton, Gary, and Andrew Metrick, 2012, “Securitized Banking Vol. 63, No. 4, pp. 1575–608. and the Run on Repo,” Journal of Financial Economics, Vol. Merton, Robert, 1974, “On the Pricing of Corporate Debt: The 104, No. 3, pp. 425–51. Risk Structure of Interest Rates,” Journal of Finance, Vol. 29, Gropp, Reint, and Florian Heider, 2010, “The Determinants No. 2, pp. 449–70. of Bank Capital Structure,” Review of Finance, Vol. 14, pp. Miles, David, Jin Yang, and Gilberto Marcheggiano, 2012, 587–622. “Optimal Bank Capital,” Economic Journal, Vol. 123, pp. Gudmundsson, Tryggvi, and Nico Valckx, forthcoming, “What 1–37. Determines Bank Funding Structures?” IMF Working Paper Octavia, Monica, and Rayna Brown, 2010, “Determinants of (Washington: International Monetary Fund). Bank Capital Structure in Developing Countries: Regulatory Hahm, Joon-Ho, Hyun Song Shin, and Kwanho Shin, 2012, Capital Requirement versus the Standard Determinants of “Non-Core Bank Liabilities and Financial Vulnerability,” Capital Structure,” Journal of Emerging Markets, Vol. 15, No. NBER Working Paper No. 18428 (Cambridge, Massachu- 1, pp. 50–62. setts: National Bureau of Economic Research). Pazarbasioglu, Ceyla, Jianping Zhou, Vanessa Le Leslé, and Heider, Florian, Marie Hoerova, and Cornelia Holthausen, Michael Moore, 2011, “Contingent Capital: Economic Ratio- 2009, “Liquidity Hoarding and Interbank Market Spreads: nal and Design Feature,” IMF Staff Discussion Note No. The Role of Counterparty Risk,” ECB Working Paper No. 11/01 (Washington: International Monetary Fund). 1126 (Frankfurt: European Central Bank). Rauh, Joshua D., and Amir Sufi, 2010, “Capital Structure and Henriques, Roberto, Alan Bowe, and Azel Finsterbusch, 2013, Debt Structure,” Review of Financial Studies, Vol. 23, No. 12, “European Bank Bail-In Survey Results,” J.P. Morgan Europe pp. 4242–80. Credit Research, April 24. Rochet, Jean-Charles, and Jean Tirole, 1996, “Interbank Lending High-level Expert Group on Reforming the Structure of the and Systemic Risk,” Journal of Money, Credit and Banking, EU Banking Sector, 2012, “Final Report (Liikanen Report),” Vol. 28, No. 4, pp. 733–62. chaired by Erkki Liikanen, October 2. Shin, Hyun Song, 2009a, “Reflections on Northern Rock: The Houben, Aerdt, and Jan Willem Slingenberg, 2013, “Collateral Bank Run that Heralded the Global Financial Crisis,” Journal Scarcity and Asset Encumbrance: Implications for the Euro- of Economic Perspectives, Vol. 23, No. 1, pp. 101–20. pean Financial System,” Banque de France Financial Stability ———, 2009b, “Securitization and Financial Stability,” Eco- Review, No. 17, April. nomic Journal, Vol. 119, No. 536, pp. 309–32. Huang, Rocco, and Lev Ratnovski, 2009, “Why are Canadian Street, Lee, Jackie Ineke, and Sean McGrath, 2012, “European Banks More Resilient?” IMF Working Paper No. 09/152 Banks: Depositor Preference Is the Real Threat—Not Encum- (Washington: International Monetary Fund). brance,” Morgan Stanley Research, May 24. ———, 2011, “The Dark Side of Bank Wholesale Funding,” Ueda, Kenichi, and Beatrice Weder di Mauro, 2012, “Quanti- Journal of Financial Intermediation, Vol. 20, No. 2, pp. fying Structural Subsidy Values for Systemically Important 248–63. Financial Institutions” IMF Working Paper No. 12/128 Independent Commission on Banking (ICB), 2011, “Final (Washington: International Monetary Fund). Report Recommendations (Vicker’s Report),” chaired by Sir Vazquez, Francisco, and Pablo Federico, 2012, “Bank Funding John Vickers, September. Structures and Risk: Evidence from the Global Financial Jones, Bradley, Peter Lindner, Miguel Segoviano, and Takahiro Crisis,” IMF Working Paper No. 12/29 (Washington: Inter- Tsuda (forthcoming), “Securitization: Lessons Learned and national Monetary Fund). the Road Ahead,” IMF Working Paper (Washington: Interna- Zhou, Jianping, Virginia Rutledge, Wouter Bossu, Marc Dobler, tional Monetary Fund). Nadege Jassaud, and Michael Moore, 2012, “From Bail- Laeven, Luc, and Fabián Valencia, 2012, “Systemic Banking Cri- Out to Bail-In: Mandatory Debt Restructuring of Systemic ses Database: An Update,” IMF Working Paper No. 12/163 Financial Institutions,” IMF Staff Discussion Note 12/03 (Washington: International Monetary Fund). (Washington: International Monetary Fund). La Porta, Rafael, Florencio López-de-Silanes, Andrei Shleifer, and Robert Vishny, 2000, “Investor Protection and Corporate

148 International Monetary Fund | October 2013

Abenomics Refers to the set of policy measures Bank insolvency A bank becomes insolvent when its introduced by Japanese Prime Minister Shinzo Abe after equity value falls below zero—namely, when the value the December 2012 elections to boost the domestic of its assets falls below the value of its debt. Before economy. The set of policies encompasses three “arrows”: becoming insolvent, a bank breaches minimum regulatory monetary stimulus, fiscal flexibility, and structural reforms. capital requirements, which usually leads to a series of actions from the supervisor, including, but not limited Accommodative policies Central bank policies designed to, intensified supervisory monitoring, restrictions to stimulate economic growth by making borrowing less on dividend payouts, and instructions to take specific expensive. managerial actions. Additional Tier 1 capital The sum of (1) instruments Bank resolution There are two broad forms of bank issued by banks that meet the criteria for inclusion resolution. One is a liquidation—namely a “gone concern” in Additional Tier 1 capital (and are not included resolution, under which a bank ceases to operate and its in Common Equity Tier 1); (2) stock surplus (share assets are distributed among creditors according to their premium—the value of paid-in capital that exceeds the seniority. The other is a “going concern” resolution, under shares’ nominal value) resulting from the issuance of which some parts of the bank’s operations continue, instruments included in Additional Tier 1 capital; (3) typically with some financial and operational restructuring. instruments issued by consolidated subsidiaries of the The latter could include purchase and assumption (P&A): a bank and held by third parties that meet the criteria for healthy bank purchases the assets and assumes the liabilities inclusion in Additional Tier 1 capital and are not included of an unhealthy bank, and a bridge bank, authorized to hold in Common Equity Tier 1; and (4) applicable regulatory the assets and liabilities of an unhealthy bank, continues adjustments. See Common Equity Tier 1 and Tier 1 capital. the bank’s operations until it is solvent and is acquired by Asset encumbrance An asset is considered encumbered another entity or until it is liquidated. if it has been pledged or may be required to secure or Banking union A European Commission policy collateralize a transaction from which it cannot be freely response to the global financial crisis: establishment of withdrawn. The asset may be pledged to reduce the credit a single supervisory-regulatory framework, harmonized risk of the underlying transaction (for example, a “credit national resolution regimes for credit institutions, and enhancement”). harmonized standards across euro area national deposit Bail in A statutory power to restructure the liabilities insurance programs. of a distressed financial institution by writing down, or Basel Committee on Banking Supervision (BCBS) converting to equity, its unsecured debt. A committee of banking supervisory authorities that Bail-in debt Also frequently called bailin-able debt refers provides a forum for regular cooperation on banking to any liabilities that can be “bailed-in,” by being written supervisory matters. The committee develops guidelines off, written down, or converted into equity through and supervisory standards in various areas, including the the application of statutory bail-in powers in a bank international standards on capital adequacy, the Core resolution. Principles for Effective Banking Supervision, and the Concordat on cross-border banking supervision. Balance-sheet constraints In the context of Chapter 2, constraints related to the capital or liquidity position of Basel III A comprehensive set of reform measures banks, to their access to market finance or, more generally, introduced as a result of the global financial crisis to to their cost of funds, all of which can make lending by improve the banking sector’s ability to absorb financial banks more difficult or expensive. and economic shocks, enhance banks’ risk management

International Monetary Fund | October 2013 149

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

and governance, and increase transparency and disclosure. Credit policies Policies implemented to promote credit These measures revise the existing definition of regulatory creation. capital under the Basel Accord, enhance capital adequacy Credit registry A database, often maintained by central standards, and introduce minimum liquidity adequacy banks or bank supervisors, with detailed information on standards for banks. loans granted by financial intermediaries. In particular, a CDS spread A credit default swap (CDS) is a credit credit registry usually contains detailed information about derivative whose payout is triggered by a “credit event,” both the borrower and the lender. often a default. The “spread” of a CDS is the annual Credit risk The risk that a party to a financial contract amount (the “premium”) the protection buyer must will incur a financial loss because a counterparty is unable pay the protection seller over the length of the contract, or unwilling to meet its obligations. expressed as a percent of the notional amount. Creditless recovery A situation in which economic Collateral Assets pledged or posted to a counterparty to recovery after a downturn is not associated with secure an outstanding exposure, derivative contract, or loan. corresponding growth in credit. Common Equity Tier 1 (CET1) The sum of (1) Currency overlay funds Structured products that common shares issued by a bank that meet the regulatory feature an outright investment in an underlying asset, criteria for classification as common shares (or the such as a domestic stock, compounded with an “overlay” equivalent for non-joint-stock companies); (2) stock exposure to a (possibly unrelated) currency. surplus (share premium—the value of paid-in capital that exceeds the shares’ nominal value) resulting from the Debt overhang A situation in which excessively issuance of instruments included in CET1; (3) retained indebted borrowers do not act as they would if they had earnings; (4) accumulated other comprehensive income less debt outstanding. For example, corporations might and other disclosed reserves; (5) common shares issued by not pursue otherwise profitable business opportunities, or consolidated subsidiaries of the bank and held by third highly indebted households may choose not to invest or parties (that is, minority interest) that meet the criteria for consume, but rather pay off their loans. inclusion in CET1 capital; and (6) applicable regulatory adjustments. See Tier 1 capital. Debt restructuring A change in the terms of a borrower’s outstanding debt, often to the benefit of the Contingent convertible bonds or CoCos Bonds with borrower. principal and coupon payments that are automatically Deleveraging converted into equity or written down, in accordance with The reduction of the leverage ratio—the their contractual terms, when a predetermined trigger percent of debt on a financial institution’s balance sheet. event occurs. Deposit preference Preference given to depositors in the creditor hierarchy that gives them a preferential claim over Contingent cost Cost that may or may not materialize, the assets of a failed deposit-taking institution compared depending on the outcome of a future event. with other senior unsecured creditors. Corporate spread Difference between the yield on a Direct credit easing Direct purchases (or sales) by corporate bond and the yield on a government bond of the the central bank in specific credit market segments with same maturity. impaired functioning. Countercyclical Movement of an economic or financial Duration A measure of the sensitivity of bond prices to quantity that is opposite to the economic cycle. For example, interest rate fluctuations, based on the bond’s weighted countercyclical capital buffers are built up during an economic average cash flows. upturn so that they can be drawn down in a downturn. Earnings before interest, taxes, depreciation, and Credit cycle The expansion and contraction of credit amortization (EBITDA) A measure of a company’s over time. operating cash flow obtained by looking at earnings before Credit guarantee A promise to repay the lender if the the deduction of interest expenses, taxes, depreciation, and borrower defaults. amortization. This measure is used to compare companies’

150 International Monetary Fund | October 2013

profitability without the accounting and financing effects bids are fully satisfied (against adequate collateral) at a of various asset and capital structures. This measure may predetermined price. be of particular interest to creditors because it represents a Flexible Credit Line (FCL) An IMF credit line for the company’s income available for interest payments. purpose of crisis prevention and mitigation for countries Endogeneity In a statistical model, endogeneity may with very strong economic fundamentals and policy track arise when an independent variable (regressor) is correlated records. with the error term, which makes it difficult to identify Forbearance A temporary postponement of loan causal relationships. Endogeneity may be caused, for payments granted by a lender or creditor. Forbearance example, by omitted variables or simultaneity. gives the borrower time to make up overdue payments on European option An option that may be exercised only a loan. on its expiration date. Foreclosure A lender’s seizure of pledged or mortgaged European Stability Mechanism (ESM) An assets, such as a house, usually with the intention of selling international organization that assists members of the them to recover part or all of the amount due from the euro area in financial difficulty to safeguard the financial borrower. stability of the euro area. The ESM may raise funds, for Funding cost Cost at which banks can obtain funds (in example, by issuing bonds and other debt instruments and the form of equity or debt). entering into arrangements with euro area members. Funding liquidity risk The risk that increases in assets Evergreening Additional loans by banks to stressed cannot be funded or obligations met as they come due borrowers to enable them to repay existing loans or without incurring unacceptable losses. Funding liquidity interest. This practice can prevent loans from becoming risk sometimes refers to the risk that solvent counterparties nonperforming, but it further increases a bank’s exposure might have difficulty borrowing in the very short term to to a troubled borrower. meet required liability payments.

Exogenous variable In an econometric model, an Global systemically important bank (G-SIB) Large explanatory variable is exogenous if it is not correlated banking institution with global operations with a potential with the error term. impact on the financial system. The Financial Stability Board (FSB) has tentatively identified 29 global banks as Externality Cost or benefit arising from an economic G-SIBs. These banks have been provisionally earmarked activity that affects not only those engaged in the activity for additional loss absorbency, or capital surcharges, but also those not engaged in the specific activity. ranging from 1 percent to 2.5 percent of the ratio of Financial fragmentation A broad retrenchment in Common Equity Tier 1 capital to risk-weighted assets. cross-border flows and assets so that private capital is Government-sponsored enterprise (GSE) A financial invested and held more along national lines. In a currency institution that provides credit or credit insurance to specific union such as the euro area, fragmentation can lead to a groups or areas of the economy, such as farmers or housing. breakdown in monetary policy transmission across the In the United States, such enterprises are federally chartered region’s banking and credit markets. and maintain legal and/or financial ties to the government. Fire sale A panic condition in which many holders of Great Moderation Period beginning in the 1980s of an asset or class of assets attempt a market sale, thereby substantially reduced macroeconomic volatility in the driving down the price to extremely low levels. A fire sale United States. may also denote a seller’s acceptance of a low price for Identification assets when faced with bankruptcy or other impending In an econometric model, a parameter distress. is said to be identified if it can be consistently estimated from the observed data. Fitted values Values predicted by a model that has been Indirect credit easing The provision of long-term funds fitted to a set of data. to banks by the central bank (instead of through regular Fixed-rate full allotment Under fixed-rate full weekly operations) specifically so that banks can expand allotment liquidity provisions, central bank counterparties’ lending to firms and households.

International Monetary Fund | October 2013 151

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Information asymmetry Situation in which one party cash outflows over a 30-day time period.” Two types of in a transaction has more or better information than the liquid assets are included, both of which should have high other. This imbalance in information can potentially affect credit quality and low market risk: Level 1 assets should the nature of the transaction and lead to a market failure. be unlikely to suffer large price changes during periods See Market failure. of distress; Level 2 assets are more likely to suffer price changes and be subject to a haircut and a limit on their Instrumental variable (instrument) Alternative variable quantity in the overall liquidity requirement. used in an econometric analysis whose original variable represents either cause or effect. Ideally, an instrumental Loan covenants Provisions in a loan agreement binding variable is highly correlated with the original variable so on the borrower or lender. that it behaves like the original variable, but it should have Loan loss provision Losses (noncash charge to earnings) little correlation with the dependent variable to eliminate that a bank expects as a result of uncollectible or troubled effects of dependent variable movements. loans and that is used to create a loan loss reserve. Insured deposits Deposits insured by deposit guarantee Examples include transfers to bad debt reserves (Japan) programs. Some types of deposits, such as retail deposits, and amortization of loans (Japan). are eligible for the insurance but are not insured because Long-term refinancing operation (LTRO) Open the amount of the deposit exceeds the maximum insurance market operations conducted by the European Central coverage. Bank to provide long-term liquidity to the banking Interest coverage ratio (ICR) Earnings before interest, system. taxes, depreciation, and amortization (EBITDA) divided Loss-absorbing buffers (or capacity) Bank liabilities by the interest expense. It measures firms’ ability to service that can be used to absorb losses from assets to maintain their debt. the bank’s viability. Equity and capital-qualifying debt, Jumbo loan In the U.S. mortgage market, a mortgage recognized under bank capital rules, are important loan that exceeds a certain legally determined limit and components. Additional debt instruments could also be can therefore not be sold by banks and other lenders to used to absorb losses without going through a liquidation government-sponsored enterprises. process, including under statutory bail-in powers.

Lending standards Internal guidelines or criteria that Macroprudential policies Policies to maintain the reflect the conditions under which a bank will grant a safety and soundness of the financial system as a whole (for loan. These include various nonprice lending terms in example, countercyclical capital buffers). a typical bank business loan or line of credit, such as Mark-to-market valuation The act of recording the collateral, covenants, and loan limits. price or value of a security, portfolio, or account to reflect its current market value rather than its book value. Leverage The proportion of debt to equity (also assets to equity) often expressed as a multiplier, such as 20X, or Market failure Occurs when free markets fail to allocate the capital-to-asset ratio in banking, expressed as a percent. resources efficiently. Market failures are often associated Leverage can be built up by borrowing relative to a fixed with asymmetric information (when buyers and sellers amount of capital (on-balance-sheet leverage) or through do not operate with the same set of information), non- off-balance-sheet transactions that increase the future competitive markets (such as monopolies), externalities exposure of the bank relative to its loss-absorbing capacity. (see externality), or public goods (when the traded good See Loss-absorbing buffers (or capacity). cannot be excluded from others’ use).

Leverage ratio A bank’s leverage ratio typically refers Market liquidity Ability to trade an asset’s large to Tier 1 capital as a ratio of adjusted assets. Assets are nominal value without significantly altering its market adjusted for intangible assets not included in Tier 1 price. capital. Microprudential policies Supervisory and regulatory Liquidity coverage ratio (LCR) A liquidity standard policies aimed at maintaining the safety and soundness introduced by Basel III. It is defined as the stock of high- of individual financial institutions. Examples are capital quality liquid assets as a proportion of the bank’s “net and liquidity requirements, banks’ recovery and resolution

152 International Monetary Fund | October 2013

plans, restrictions on executive compensation, limits on international banking regulatory agency. They facilitate dividend distributions, etc. bank-related financial services, such as investment, risk pooling, contractual savings, and market brokering, and Model herding Tendency for financial sector players to can include money market mutual funds, investment act together, often as the result of using similar common banks, finance companies, insurance firms, pension funds, financial models. hedge funds, currency exchanges, and microfinance Money market mutual fund (MMMF) An open-ended organizations. mutual fund that invests in short-term money market Nonperforming loan (NPL) A loan for which the securities, such as U.S. Treasury bills and commercial contractual payments are delinquent, usually defined as paper. being overdue for more than a certain number of days (for Moral hazard A situation in which an agent (an example, more than 30, 60, or 90 days). The NPL ratio is individual or institution) will act less carefully than the amount of nonperforming loans as a percent of gross otherwise because the consequences of a bad outcome will loans. be largely shifted to another party. Often such behavior is Originate-to-distribute model present because the other party cannot observe the actions. A banking model, For example, a financial institution may take excessive popular in North America, whereby banks tend to risks if it believes that governments will support them distribute loans, such as mortgages, credit card credits, and during a crisis and that governments cannot observe the corporate loans, that they originate to other investors. risky behavior ex ante to prevent it. Overcollateralization When issuing covered bonds, Mortgage-backed security (MBS) A security, backed issuers usually pledge collateral so that the total value of by pooled mortgages on real estate assets, that derives the collateral exceeds the borrowed amount. The extent its cash flows from principal and interest payments on of overcollateralization varies significantly across bonds, those mortgages. An MBS can be backed by residential ranging from a few percent to well over 100 percent. mortgages (RMBS) or mortgages on commercial A rating agency will often require a certain degree of properties (CMBS). A private-label MBS is typically a overcollateralization for the bond to attain a high rating structured credit product. RMBSs that are issued by a (for example, AAA). government-sponsored enterprise are not structured (that Quantitative easing (QE) Direct purchases of is, do not have a tiered or tranched payments structures government bonds by the central bank, usually when with payment priorities to the different holders). the official policy interest rate is at or near the zero lower Mortgage real estate investment trusts (mREIT) bound. Investment vehicles designed for borrowing at short- Quantitative and qualitative monetary easing term rates and investing in long-term mortgage-related (QQME) Policies introduced by the Bank of Japan that securities. involve significantly increasing its holdings of government Net stable funding ratio (NSFR) Introduced by Basel bonds and other assets through extending the maturity III to provide a more sustainable maturity structure of of Japanese government bond purchases. The aim is to assets and liabilities. The NSFR stipulates that the ratio achieve a consumer price index stability target of 2 percent of a bank’s available stable sources of funding to its year over year as soon as possible. required stable funding be greater than 100 percent. Each Pari passu When creditors rank equally in the creditor asset category (including off-balance-sheet contingent hierarchy for repayment of their debt from the obligor’s assets. instruments) is assigned a factor to reflect its potential liquidity characteristics. The NSFR aims to limit Perpetual bonds Perpetual bonds, also known as perp overreliance on short-term wholesale funding during bonds, are those with no maturity date. times of buoyant market liquidity and to encourage better Pillar 1 (of Basel II) One of the three mutually assessment of liquidity risk across all on- and off-balance- supporting pillars that form the Basel II accord. Pillar 1 sheet items. sets a minimum capital requirement for all internationally Nonbanks Financial institutions that do not have full active banks that covers credit risk, operational risk, and banking licenses or are not supervised by a national or market risk.

International Monetary Fund | October 2013 153

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Pillar 3 (of Basel II) One of the three mutually cash in return and pledges the legal title of a security as supporting pillars that form the Basel II accord. Pillar 3 collateral. provides disclosure requirements for information regarding Return on assets (RoA) The amount an investor would regulatory capital ratios. earn from a firm as a proportion of the total assets. Usually Preferred creditor An individual or organization that calculated as: (Net income before preferred dividends plus has repayment priority if the debtor declares bankruptcy. ((interest expense on debt-interest capitalized) multiplied Preferred share A preferred share (or stock) is an equity by (1 minus tax rate))) divided by last year’s total assets security that has features not possessed by common stock, multiplied by 100. including properties of both an equity and bonds and is Risk premium The extra expected return on an asset that generally considered a hybrid instrument. It usually has investors demand in exchange for accepting its higher risk. no voting rights, but may carry dividends and may have Secondary market The financial market in which priority over common stock in the payment of dividends, previously issued financial instruments, such as stock and may receive cash flows upon liquidation. or bonds, are bought and sold. The existence of liquid Price-to-book ratio Used to compare a firm’s stock secondary markets can encourage people to buy in the market value to its book value. It is calculated by dividing primary market, as they know they are likely to be able to the current closing price of the stock by the firm’s recent- sell easily should they wish to. See Primary market. quarter accounting book value per share. Secured creditor Any creditor or lender that takes Primary market The financial market that deals with collateral for the extension of credit, loan, or bond issuance. the issuance of new financial instruments, such as stock and bonds. See also Secondary market. Secured funding Funding secured by certain collateral, including repos, asset-backed securities, mortgage-backed Probability of default (PD) Likelihood of default over a securities, and covered bonds. At liquidation, secured debt given time horizon. holders have priority claim up to the value of the pledged Prudential measures These comprise micro- and collateral over general creditors, including depositors. macroprudential policy measures. Securitization The creation of securities from a reference Regulatory forbearance A situation in which bank portfolio of preexisting assets or future receivables that regulators or supervisors allow banks to avoid adhering to are placed under the legal control of investors through a established regulations. To temporarily help borrowers, specially created intermediary: a “special purpose vehicle” regulators or supervisors may allow banks to avoid (SPV) or “special purpose entity” (SPE). In the case of recognizing nonperforming loans on their balance sheets “synthetic” securitizations, the securities are created from a or discourage banks from seizing collateral underlying portfolio of derivative instruments. their loans. Senior creditor A creditor who receives higher priority Relationship banking A situation in which a bank for the repayment of a debt instrument from the obligor’s attempts to cultivate a long-term relationship with their assets, for example, compared to subordinated or junior borrowers. Typically, a bank will attempt to accumulate creditors. soft (proprietary) information in existing customers in Shadow banks Nonbank financial intermediaries that addition to hard (quantifiable, verifiable) information to provide services similar to traditional commercial banks, assess a borrower’s creditworthiness. Various products, but are not regulated or supervised like a bank. These can such as long-term contracts, can help to establish a include hedge funds, money market funds, and structured borrower’s long-term commitment to the bank. investment vehicles (SIVs), depending on their investment Repurchase (repo) transaction A sale of securities and funding strategies. coupled with an agreement to repurchase the securities at Shifter In the context of Chapter 2, a variable that an agreed price at a future date. This transaction occurs shifts either the credit supply curve or the demand curve, between a cash borrower (or securities lender)—typically a without affecting the other. fixed-income securities broker-dealer—and the cash lender (or securities borrower), such as a money market mutual Single supervisory mechanism (SSM) A common fund or a custodial bank. The securities lender receives banking supervision framework under the aegis of the

154 International Monetary Fund | October 2013

European Central Bank for the euro area banks, as proposed Underwriting The process that a financial institution, such by the European Commission in September 2012. as a bank or insurer, uses to assess the eligibility of a customer to receive a financial product, such as credit or insurance. Small and medium enterprises (SMEs) In Europe, these firms are classified based on the number of employees and Universal banking model Banking system, popular in balance sheet turnover (according to EU law). Europe, whereby banks often provide a range of financial products and services, including both investment and Stop-loss sales Sales orders that are executed when a commercial banking, transaction banking, asset gathering, security falls to a prespecified price. and retail banking.

Stress test A process that evaluates an institution’s ability Unsecured creditor Any creditor or lender that lends to financially withstand adverse macroeconomic and money without obtaining prespecified assets as collateral. financial situations. Value-at-risk (VaR) An estimate of the loss, over a given horizon, that is statistically unlikely to be exceeded Subordinated debt (or junior debt) This debt at a given probability level, usually based on the historical instrument receives lower seniority than general debt returns, covariances, and volatilities of a portfolio of assets. in the event that a company falls into liquidation or bankruptcy, and it receives payments only after all senior Vertical (bull) spread An options strategy involving debt holders are paid but before equity holders receive any buying a call and selling a put option on the same money. underlying security with the same expiration date but at different strike prices. In Chapter 3, value of senior debt Swaptions Interest rate instruments that allow mirrors a strategy in which the call is purchased at a lower investors to take a view on future interest rate volatility, strike price than the put is sold. using options to trigger underlying interest rate swap Vienna Initiative The European Bank Coordination agreements. A 10-year by 10-year swaption allows an “Vienna” Initiative (EBCI) was launched in January investor to buy/sell a 10-year option on an underlying 2009 to provide a framework for coordinating the interest rate swaps contract with a 10-year maturity. crisis management and resolution regime that involved Tangible assets (TA) Total assets less intangible assets large cross-border banking groups in emerging (such as goodwill and deferred tax assets). Europe. The European Bank for Reconstruction and Development, the IMF, the European Commission, Tangible leverage ratio A measure of financial and other international financial institutions initiated a strength using the ratio of a bank’s total liabilities to its process to address possible collective actions for dealing shareholder’s equity less goodwill and tangible assets. It is with financial instability. In a series of meetings, the not a regulatory requirement. international financial institutions and policymakers from home and host countries’ banks met with Term premium The premium in terms of yield that commercial banks active in emerging Europe to discuss an investor expects to receive for buying longer-dated what measures might be needed to reaffirm their securities compared to the yield received if short-term presence in the region in general and, more specifically, securities were to be reinvested as they come due until the in countries that were receiving balance of payments maturity of the longer-dated securities. support from the international financial institutions. Tier 1 capital Under Basel III, Tier 1 capital (or going VIX Chicago Board Options Exchange Volatility Index concern capital) comprises Common Equity Tier 1 capital that measures market expectations of financial volatility and Additional Tier 1 capital. See Common Equity Tier 1 over the next 30 days. The VIX is constructed from S&P capital and Additional Tier 1 capital. 500 option prices. Tier 1 capital ratio This is the ratio of a bank’s Tier 1 Wholesale funding Bank funding instruments typically capital to its total risk-weighted assets (RWA). Under Basel issued in money and capital markets, including interbank III, banks in member countries are required to meet the deposits, commercial paper (CPs), certificates of deposit minimum Tier 1 capital ratio requirement of 6 percent (CDs), repurchase agreements (repos), swaps, and and Common Equity Tier 1 capital ratio of 4.5 percent by various kinds of bonds. These are typically purchased by January 1, 2015. See Tier 1 capital. institutional investors, including other banks.

International Monetary Fund | October 2013 155

The following remarks were made by the Acting Chair at the conclusion of the Executive Board’s discussion of the World Economic Outlook, Global Financial Stability Report, and Fiscal Monitor on September 23, 2013.

xecutive Directors broadly shared the staff’s conditions. However, lingering supply side bottlenecks assessment of the state of the global economy in infrastructure, labor markets, and regulatory and and financial markets, risks, and key policy financial systems could have lowered potential output recommendations. They observed, in particular, for many of these economies. Growth in low-income Ethat global growth remains subdued and that uncertainty countries remains robust, supported by enhanced and downside risks dominate the outlook. The recovery policy frameworks, although less favorable commodity in the United States and Japan has gained ground and prices and external financing may weaken their fiscal the euro area is pulling out of recession, while growth in positions. many emerging market economies has slowed. Directors Directors expressed concern that multiple vulner- underscored that policymakers in all economies have a abilities may have raised the risk of a prolonged period shared responsibility to sustain balanced growth while of lower global growth. They noted that important continuing to build resilience. legacy risks are still present in advanced economies. Directors stressed that changing growth dynamics, These include unfinished financial sector reforms in combined with the anticipation of the start of the nor- the euro area, impaired monetary policy transmission malization of U.S. monetary policy, pose new policy and corporate debt overhang in some of its economies, challenges, particularly in emerging market economies. and high levels of government debt and related fiscal Many of these countries have recently experienced and financial risks in many other advanced economies, increased capital outflows, currency depreciation, lower including Japan and the United States. equity prices, and higher sovereign risk premiums. In Directors noted that downside risks to growth addition, external financial conditions have generally in emerging market economies have become more tightened and the fiscal space has narrowed, while risks prominent, reflecting risks of further asset repricing of interest rate and exchange rate overshooting have in anticipation of the normalization of U.S. monetary increased. In this regard, Directors took note of the policy as well as rising domestic vulnerabilities in some U.S. Federal Reserve’s guidance that monetary policy countries. Fiscal vulnerabilities are increasing as policy normalization will occur in the context of stronger buffers are used, potential output declines, and public U.S. growth and employment that, in turn, should be contingent liabilities build up. Nevertheless, Direc- beneficial for global growth. tors noted that, in general, these economies are in a Directors noted that global growth is expected stronger position now than in the past to withstand to improve modestly in the near term. Activity in the looming turbulence, with improved fundamentals advanced economies is accelerating as fiscal consolida- and policy frameworks, more flexible exchange rates, tion eases and monetary conditions remain accom- and higher international reserve buffers. modative. In the euro area, policy actions have reduced Directors underscored the need for credible policy tail risks and stabilized financial markets, but growth actions to forestall downside risks and address old chal- remains fragile, given persistently high unemployment, lenges decisively. In the euro area, priorities continue financial fragmentation, and weak credit developments. to be—building on recent progress—bank balance Growth in emerging market economies, which contin- sheet repair, a comprehensive assessment of, and ues to account for the bulk of global growth, remains measures to reduce, corporate debt overhang in some driven by solid consumption and, in a historical per- countries, and completion of a full-fledged banking spective, still supportive fiscal, monetary, and financial union, with an effective common backstop. Directors

International Monetary Fund | October 2013 157

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

underscored that, while the fiscal adjustment path is toolkit could be useful, anchored in credible monetary currently appropriate for the euro area as a whole, policy frameworks. Directors emphasized the impor- the speed and composition of fiscal consolidation in tance of prudential oversight and regulation to contain each country would need to take into account cyclical any further buildup of foreign currency mismatches considerations, debt levels, and financing conditions. and risks stemming from shadow banking activities in In Japan and the United States, Directors empha- key emerging markets. sized the importance of carefully pacing fiscal adjust- Directors noted that, in emerging market econo- ment and placing government debt on a sustainable mies where inflation is low and expectations are firmly track, anchored in a medium-term plan that includes anchored, monetary policy should be used as the first durable tax and entitlement reforms. Promptly lift- line of defense if downside risks materialize. They ing the debt ceiling is also a priority in the United stressed the need to rebuild fiscal buffers, unless growth States. More generally, Directors agreed that there is deteriorates significantly. In countries with high debt, scope for broader tax reforms to improve efficiency fiscal consolidation remains a high priority, taking and fairness, and for strengthening cooperation on advantage of still favorable cyclical conditions. Further international taxation. In most advanced economies, a structural reforms are also essential to boost potential sustained focus on structural reforms over the medium growth, including improving infrastructure, productiv- term remains crucial to reduce rigidities in labor and ity, and the investment climate. Low-income countries product markets, enhance competitiveness, and boost need to step up revenue mobilization, including from potential output. natural resources, to rebuild their fiscal buffers and Directors agreed that monetary conditions need to support higher priority public spending. stay accommodative in major advanced economies. In Directors concurred that a further narrowing of the United States, it is important that monetary policy global imbalances would help maintain more sustain- respond gradually to changing prospects for growth, able and stable global growth. They observed that the inflation, and financial stability, accompanied by clear, recent exchange rate depreciations have facilitated well-timed communication about the policy direction some rebalancing in many deficit emerging market and strategy. Directors also emphasized the need to economies. However, further efforts are needed to address structural liquidity weaknesses and vulnerabili- increase national saving and boost productivity and ties in the shadow banking system, which would help competitiveness in many countries, including Brazil, reduce financial market volatility during the transition India, Russia, and South Africa. For the United States, to higher interest rates. gradual progress on fiscal deficit reduction through a Directors noted that policy priorities and options comprehensive plan for medium-term consolidation differ across emerging market economies, depending would help support global rebalancing. In surplus on the degree of economic slack, the nature of vulner- countries, priorities include measures to promote more abilities, and available policy buffers. They pointed consumption-based growth in China and structural to the role of exchange rates as a shock absorber and reforms and medium-term fiscal consolidation in the need to guard against excessive volatility, while Japan. Directors were of the view that further exter- macroprudential measures should be used to mitigate nal rebalancing within the euro area requires deeper financial stability risks. A few Directors took the view structural reforms, including sustained efforts to raise that exploring policy options beyond the traditional investment in Germany.

158 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT STATISTICAL APPENDIX, OCTOBER 2013

his presentation complements the main Please note that effective with the April 2011 issue, text of the Global Financial Stability Report the IMF’s Statistics Department assumed responsibil- (GFSR) with data on financial develop- ity for compiling the Financial Soundness Indicators ments in regions and countries as well as in tables, and they are no longer part of this appendix. Tselected sectors. However, these tables will continue to be linked to Unless otherwise noted, the data reflect information the GFSR Statistical Appendix on the IMF’s public available up to July 31, 2013. The data come for the website. most part from sources outside the IMF. Although the Effective with the April 2013 issue, the database and IMF endeavors to use the highest quality data available, filtering criteria for the external private financing Tables it cannot be responsible for the accuracy of information 4, 5, 6, and 7 were changed. Consequently, there was a obtained from independent sources. significant break in the data reported in previous issues.

The following symbols and conventions have been used in this appendix: . . . to indicate that data are not available; — to indicate that the figure is zero, or less than half the final digit shown, or the item does not exist; – between years and months (for example, 2008–09 or January–June) to indicate the years or months covered, including the beginning and ending years or months; / between years (for example, 2008/09) to indicate a fiscal or financial year. “Billion” means a thousand million; “trillion” means a thousand billion. “Basis points” refers to hundredths of 1 percentage point (for example, 25 basis points is equivalent to ¼ of 1 percentage point). “n.a.” means not applicable. Minor discrepancies between constituent figures and totals are due to rounding. As used in this volume, the term “country” does not in all cases refer to a territorial entity that is a state as understood by international law and practice. As used here, the term also covers some territorial entities that are not states but for which statistical data are maintained on a separate and independent basis.

International Monetary Fund | October 2013 159

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table of Contents

Figures 1. Major Net Exporters and Importers of Capital, 2012 ...... 161 2. Sovereign Credit Default Swap Spreads ...... 162 3. Selected Credit Default Swap Spreads ...... 163 4. Selected Spreads ...... 164 5. Implied Volatility Indices ...... 165 6. U.S. Corporate Bond Market ...... 166 7. Euro Area Corporate Bond Market ...... 167 8. U.S. Commercial Paper Market ...... 168

Tables 1. Capital Market Size: Selected Indicators, 2012 ...... 169 2. Morgan Stanley Capital International: Equity Market Indices ...... 170 3. Emerging Markets Bond Index: Global Sovereign Yield Spreads ...... 172 4. Emerging Market Private External Financing: Total Bonds, Equities, and Loans ...... 174 5. Emerging Market Private External Financing: Bonds ...... 177 6. Emerging Market Private External Financing: Equities ...... 179 7. Emerging Market Private External Financing: Loans ...... 181 8. Equity Valuation Measures: Dividend-Yield Ratios ...... 184 9. Equity Valuation Measures: Price/Earnings Ratios ...... 185 10. Emerging Markets: Mutual Funds ...... 186

160 International Monetary Fund | October 2013

Figure 1. Major Net Exporters and Importers of Capital, 2012

1. Economies That Export Capital1

Other economies2 Germany 14.9% 21.0%

Taiwan Province of China China 12.1% 3.1%

Singapore 3.2%

Japan 3.8% Saudi Arabia Qatar 3.9% 10.3% United Arab Emirates Kuwait 4.2% Switzerland 5.0% 4.4% Norway 4.4% Netherlands 4.9% Russia 4.7%

2. Economies That Import Capital3

Other economies4 31.8% United States 35.2%

Brazil 4.3%

Australia 4.5%

France 4.6% Canada 5.0%

India United Kingdom 7.0% 7.5%

Source: IMF, World Economic Outlook database as of Sept. 24, 2013. 1As measured by economies’ current account surplus (assuming errors and omissions are part of the capital and fi nancial accounts). 2“Other economies” includes all economies with shares of total surplus less than 3.1 percent. 3As measured by economies’ current account defi cit (assuming errors and omissions are part of the capital and fi nancial accounts). 4“Other economies” includes all economies with shares of total defi cit less than 4.3 percent.

International Monetary Fund | October 2013 161

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 2. Sovereign Credit Default Swap Spreads (Five-year tenors; basis points)

200 Japan France 700 180 United States Germany 600 160 United Kingdom Italy 140 Spain 500

120 400 100 80 300 60 200 40 100 20 0 0 2008 09 10 11 12 13 2008 09 10 11 12 13

1,800 Greece 30,000 1,400 1,600 (right scale) 25,000 China 1,200 1,400 Ireland Indonesia 1,000 1,200 Portugal 20,000 Korea 1,000 Philippines 800 15,000 Thailand 800 600 600 10,000 400 400 5,000 200 200 0 0 0 2008 09 10 11 12 13 2008 09 10 11 12 13

1,200 600 5,000 Hungary Argentina (right scale) Poland 1,000 500 Brazil Russia 4,000 Chile 800 South Africa 400 Turkey Colombia 3,000 600 300 Mexico 2,000 400 200

1,000 200 100

0 0 0 2008 09 10 11 12 13 2008 09 10 11 12 13

Source: Bloomberg, L.P.

162 International Monetary Fund | October 2013

Figure 3. Selected Credit Default Swap Spreads (Five-year tenors; basis points)

1. Sovereigns by Region 1,200 Western Europe Asia Paciﬁc 1,000 Emerging markets 800 CEEMEA 600

400

200

0 2008 09 10 11 12 13

2. Corporates by Credit Quality 1,600 European investment grade European crossover North American investment grade 1,200 North American crossover

800

400

0 2008 09 10 11 12 13

3. Banks by Region 700 Euro area 600 United Kingdom United States 500

400

300

200

100

0 2008 09 10 11 12 13

Sources: Bloomberg L.P.; and Datastream. Note: CEEMEA = Central and Eastern Europe, Middle East, and Africa.

International Monetary Fund | October 2013 163

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Figure 4. Selected Spreads (Basis points; monthly data)

1 1. Repo Spread 160

120

–40 1998 00 02 04 06 08 10 12

2. Commercial Paper Spread2 280

240

200

160

120

–40 1998 00 02 04 06 08 10 12

3 3. Swap Spreads 160

120

80 United States

Euro area

0 Japan –40 1998 00 02 04 06 08 10 12 Sources: Bloomberg, L.P.; and Bank of America Merrill Lynch. 1Spread between yields on three-month U.S. Treasury repo and on three-month U.S. Treasury bill. 2Spread between yields on 90-day investment-grade (fi nancial and nonfi nancial) commercial paper and on three-month U.S. Treasury bill. 3Spread over 10-year government bond.

164 International Monetary Fund | October 2013

Figure 5. Implied Volatility Indices

90 300

80 250 70

60 200 MOVE (basis points; right scale) 50 VIX (percent; leftscale) 150 40

30 100

20 50 10 G7currencies (percent; leftscale) 0 0 1998 2000 02 04 06 08 10 12 Source: Bloomberg, L.P. Note: G7 currencies = VXY index from JPMorgan Chase & Co. and denotes G7 foreign exchange volatility. MOVE = Bank of America Merrill Lynch Option Volatility Estimate index and denotes one-month Treasury options volatility. VIX = Chicago Board Options Exchange volatility index on the Standard & Poor’s 500 and denotes equity volatility.

International Monetary Fund | October 2013 165

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

FigureFigure 6.6. U.S.U.S. CorporateCorporate BondBond MarketMarket

1. Investment Grade 80 Bank of America Merrill Lynch 700 investment-grade option-adjusted spread (basis points; right scale) 70 600

60 Gross issuance 500 (billions of U.S. dollars; 50 left scale) 400 40 300 30

200 20

10 100

0 0 1998 2000 02 04 06 08 10 12

2. High Yield 35 2,500

Bank of America Merrill Lynch 30 high-yield option-adjusted spread (basis points; right scale) 2,000 Gross issuance 25 (billions of U.S. dollars; left scale) 1,500 20

15 1,000

10 500 5

0 0 1998 2000 02 04 06 08 10 12

Sources: Board of Governors of the Federal Reserve System; and Bank of America Merrill Lynch. Sources: Board of Governors of the Federal Reserve System; and Bank of America Merrill Lynch.

166 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution FigureFigure 7.7. EuroEuro AreaArea CorporateCorporate BondBond MarketMarket

1. Investment Grade 200 Bank of America Merrill Lynch 500 investment-grade option-adjusted spread 180 (basis points; right scale) 450

160 400

140 350

120 300

100 Gross issuance 250 (billions of U.S. dollars; 80 left scale) 200

60 150

40 100

20 50

0 0 1998 99 2000 01 02 03 04 05 06 07 08 09 10 11 12 13

2. High Yield 3.5 2,500 STATISTICAL APPENdIx

3.0 Bank of America Merrill Lynch 2,000 high-yield option-adjusted spread 2.5 (basis points; right scale)

1,500 2.0

1.5 Gross issuance (billions of U.S. dollars; 1,000 left scale) 1.0

500 0.5

0.0 0 1998 99 2000 01 02 03 04 05 06 07 08 09 10 11 12 13

Sources: Dealogic; and Bank of America Merrill Lynch. Sources: Dealogic; and Bank of America Merrill Lynch.

International Monetary Fund | October 2013 167

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

FigureFigure 8.8. U.S.U.S. CommercialCommercial PaperPaper MarketMarket

1. Discount Rate Spread1 600 (Basis points; weekly data)

500

400

300

200

100

0 2001 02 03 04 05 06 07 08 09 10 11 12 13

2. Amount Outstanding 300 (Billions of U.S. dollars; monthly data) 1,400 Asset-backed (right scale) 1,200 250 Nonﬁnancial (left scale) 1,000 200

800 150 600

100 Financial (right scale) 400

50 200

0 0 2001 02 03 04 05 06 07 08 09 10 11 12 13

Source: Board of Governors of the Federal Reserve System. 1Source:Difference Board between of Governors 30-day ofA2/P2 the Federaland AA Reservenonfinancial System. commercial paper. 1Difference between 30-day A2/P2 and AA nonﬁnancial commercial paper.

168 International Monetary Fund | October 2013

Table 1. Capital Market Size: Selected Indicators, 2012 (Billions of U.S. dollars, unless noted otherwise) Total Bonds, Bonds, Equities, Reserves Stock Market Total Debt Bank Equities, and and Bank Assets5 GDP Minus Gold2 Capitalization Securities3 Assets4 Bank Assets5 (percent of GDP) World 72,216.4 11,403.5 52,494.9 99,134.2 116,956.1 268,585.2 371.9 European Union1 15,514.8 498.0 9,732.2 29,456.8 43,028.6 82,217.5 529.9 Euro area 12,199.5 332.5 5,492.1 21,983.1 30,314.4 57,789.6 473.7 North America 18,066.0 207.5 18,883.2 37,255.5 18,858.9 74,997.6 415.1 Canada 1,821.4 68.4 2,027.6 2,100.5 3,800.9 7,929.0 435.3 United States 16,244.6 139.1 16,855.6 35,155.0 15,058.0 67,068.6 412.9 Japan 5,960.3 1,227.1 3,638.6 14,592.4 14,166.1 32,397.2 543.6 Memorandum EU countries Austria 394.9 12.2 112.3 641.7 1,295.7 2,049.7 519.1 Belgium 483.9 18.6 297.8 736.2 1,180.4 2,214.4 457.6 Denmark 314.9 86.1 243.8 920.2 1,188.2 2,352.1 747.0 Finland 247.6 8.5 163.0 281.8 575.2 1,019.9 411.8 France 2,613.9 54.2 1,662.7 4,530.0 9,426.6 15,619.4 597.5 Germany 3,429.5 67.4 1,567.1 4,355.2 4,967.6 10,889.9 317.5 Greece 249.2 1.3 44.0 237.5 443.0 724.5 290.7 Ireland 210.9 1.4 107.2 1,322.1 1,151.4 2,580.7 1,223.9 Italy 2,014.1 50.5 509.7 3,895.1 3,143.8 7,548.6 374.8 Luxembourg 57.1 0.9 70.3 729.7 969.8 1,769.9 3,097.4 Netherlands 770.9 22.1 298.2 2,328.3 2,567.7 5,194.1 673.8 Portugal 212.4 2.2 70.9 395.2 656.4 1,122.6 528.4 Spain 1,323.5 35.5 567.9 2,423.8 3,750.0 6,741.7 509.4 Sweden 523.8 45.5 580.6 775.3 808.0 2,164.0 413.1 United Kingdom 2,476.7 88.6 3,415.7 5,778.2 10,717.9 19,911.9 804.0 Newly industrialized Asian economies6 2,143.5 1,302.7 5,943.6 2,317.7 4,844.1 13,105.3 611.4 Emerging market economies7 26,975.0 7,384.2 11,196.3 10,870.7 28,599.0 50,666.0 187.8 of which: Asia 12,358.5 4,187.3 5,852.7 5,530.2 19,836.6 31,219.5 252.6 Latin America and the 5,629.0 798.2 2,475.6 3,589.6 3,774.0 9,839.1 174.8 Caribbean Middle East and North Africa 3,201.5 1,278.2 895.3 217.6 1,705.1 2,818.0 88.0 Sub-Saharan Africa 1,272.0 200.7 606.1 259.8 600.7 1,466.6 115.3 Europe 4,514.0 919.8 1,366.6 1,273.6 2,682.6 5,322.7 117.9 Sources: World Federation of Exchanges; Bank for International Settlements (BIS); IMF, International Financial Statistics (IFS) and World Economic Outlook databases as of September 24, 2013; ©2003 Bureau van Dijk Electronic Publishing-Bankscope; Board of Governors of the Federal Reserve System, Flow of Funds; and Bloomberg, L.P. 1This aggregate includes euro area countries, Denmark, Sweden, and the United Kingdom. 2Data are from IFS. For euro area, the data also include the total reserves minus gold holdings of the European Central Bank. 3Data are from BIS as of September 24, 2013. The data include total debt securities, all issuers, amounts outstanding by residence of issuer. BIS compilation methodology changed in December 2012. For the new data definition and classificaiton, refer to “Enhancements to the BIS debt securities statistics” publication. 4Total assets of domestic commercial banks, including foreign bank’s subsidiaries operated domestically. For Austria, the data are from Austrian National Bank. For Ireland, the data are from Central Bank of Ireland. For Luxembourg, the data are from Commission de Surveillance du Secteur Financier. It comprises the assets of commercial, savings, and private banks. For Portugal, the data are from Bank of Portugal. For the United States, the data are from the Flow of Funds. It comprises the assets of private depository institutions. 5Sum of the stock market capitalization, debt securities, and bank assets. 6Hong Kong SAR, Korea, Singapore, and Taiwan Province of China. 7This aggregate comprises the group of emerging and developing economies defined in the World Economic Outlook.

International Monetary Fund | October 2013 169

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 2. Morgan Stanley Capital International: Equity Market Indices 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Period-on-period percent change

Global 31.5 10.4 –9.4 13.4 6.2 2.5 6.0 –1.2 Emerging Markets Index1 74.5 16.4 –20.4 15.1 7.0 5.2 –1.9 –9.1

Latin America 98.1 12.1 –21.9 5.4 4.3 3.4 0.5 –16.5 Brazil 121.3 3.8 –24.9 –3.5 4.3 2.2 –1.3 –18.4 Chile 81.4 41.8 –22.1 5.6 1.1 –1.0 4.1 –15.5 Colombia 76.5 40.8 –7.1 31.6 2.0 12.3 –6.8 –15.1 Mexico 53.1 26.0 –13.5 27.1 6.5 5.4 5.9 –11.7 Peru 69.3 49.2 –23.9 15.5 2.7 5.2 –2.7 –28.4 Asia 70.3 16.6 –19.1 18.1 7.7 5.8 –1.6 –6.3 China 58.9 2.6 –20.4 18.7 3.9 12.8 –4.4 –9.1 India 91.5 14.7 –26.3 27.9 8.6 4.2 –3.5 2.6 Indonesia 90.3 25.8 4.7 8.8 9.2 1.4 14.1 –6.1 Korea 56.6 22.1 –11.5 11.7 6.6 0.9 –0.3 –7.7 Malaysia 46.2 19.3 –0.2 6.8 0.5 3.1 –0.1 7.0 Pakistan 89.8 21.4 –12.9 33.5 8.8 5.7 2.8 14.3 Philippines 55.8 23.5 –3.1 34.7 3.0 9.6 17.1 –4.0 Taiwan Province of China 70.7 7.9 –20.3 8.8 5.8 0.6 2.8 1.8 Thailand 63.0 36.4 –1.2 26.9 6.6 5.2 4.7 –4.2 Europe, Middle East, 63.5 20.9 –22.6 17.7 7.6 5.5 –5.8 –9.8 and Africa Czech Republic1 13.9 –5.9 –6.8 –6.1 6.0 –6.6 –9.5 –12.2 Egypt 32.1 15.9 –46.8 52.5 23.4 –6.9 –4.7 –8.6 Hungary 71.2 –1.6 –23.7 8.1 7.8 –1.8 0.2 4.5 Morocco –10.5 17.2 –16.5 –17.6 –5.1 –0.6 –0.9 –7.1 Poland 32.6 16.3 –21.7 19.0 4.0 8.0 –7.0 –3.5 Russia 100.3 17.2 –20.9 9.6 9.0 2.1 –3.2 –11.1 South Africa 22.2 17.4 0.9 20.6 6.5 8.4 –2.4 –0.4 Turkey 86.5 21.5 –22.4 51.7 7.4 17.6 9.7 –11.4 Sectors Energy 82.1 7.5 –20.1 2.5 10.3 0.7 –6.0 –13.9 Materials 74.8 14.7 –23.0 6.4 2.5 5.1 –9.3 –14.1 Industrials 56.3 27.1 –30.6 14.9 4.2 5.0 –1.5 –9.6 Consumer discretionary 113.0 29.5 –10.4 14.6 9.9 3.8 –2.6 –3.6 Consumer staple 66.7 27.6 –1.4 23.0 6.6 7.2 1.6 –4.5 Health care 40.1 25.7 –23.2 31.6 10.8 4.7 2.5 –1.0 Financials 76.6 14.5 –25.6 22.0 6.7 9.7 1.5 –11.3 Information technology 104.7 13.9 –17.1 26.3 9.2 5.4 0.8 –4.0 Telecommunications 21.8 10.9 –8.0 9.6 6.1 0.4 –5.5 –1.0 Utilities 51.2 4.9 –16.4 2.4 –1.2 0.6 1.6 –11.8

170 International Monetary Fund | October 2013

Table 2. (concluded) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Period-on-period percent change

Developed Markets 27.0 9.6 –7.6 13.2 6.1 2.1 7.2 –0.1 Australia 30.8 –3.5 –14.9 14.9 7.2 6.0 7.3 –2.9 Austria 34.1 14.8 –35.7 20.7 5.3 15.5 –2.4 –5.3 Belgium 49.4 4.6 –9.6 34.0 8.3 3.6 11.7 –7.4 Canada 29.7 12.0 –12.2 4.3 6.0 1.3 2.4 –4.5 Denmark 31.0 39.0 –14.3 28.1 9.9 0.9 5.6 –5.3 Finland 7.2 7.1 –34.2 10.0 10.6 12.7 1.8 –1.6 France 27.6 –6.7 –19.3 17.7 6.8 10.6 0.4 0.8 Germany 21.3 6.0 –20.1 27.2 13.9 8.5 –0.1 0.5 Greece 22.6 –46.4 –63.6 –0.8 –1.4 28.1 14.0 –12.8 Hong Kong SAR 55.2 19.7 –18.4 24.4 11.8 5.2 3.1 –5.9 Ireland 9.9 –19.7 11.4 3.8 –2.1 2.9 11.4 –3.7 Israel 51.3 2.2 –29.8 –7.0 5.3 –4.4 6.6 –5.3 Italy 22.6 –17.6 –25.8 8.6 6.8 9.1 –9.8 –1.4 Japan 4.4 13.4 –16.2 5.8 –1.8 5.7 10.7 4.2 Netherlands 37.9 –0.6 –14.4 17.2 8.7 9.2 2.2 1.3 New Zealand 43.0 3.2 1.1 23.0 12.8 4.3 8.0 –10.4 Norway 82.5 7.4 –12.8 13.7 13.3 0.5 0.5 –8.5 Portugal 35.4 –14.6 –25.7 –0.7 12.8 10.1 –0.3 –3.6 Singapore 67.3 18.4 –21.0 26.4 9.3 2.7 2.8 –7.6 Spain 36.5 –25.4 –16.9 –3.3 9.7 8.4 –6.4 –1.5 Sweden 60.2 31.3 –17.8 18.7 10.6 5.0 8.6 –7.7 Switzerland 22.9 9.8 –9.1 17.3 7.7 8.0 10.4 –1.7 United Kingdom 22.3 8.5 –5.4 5.9 2.9 2.8 8.6 –3.2 United States 24.2 13.2 –0.1 13.5 5.8 –0.9 10.1 2.2 Source: Morgan Stanley Capital International (MSCI). Note: Price indices are in local currency terms. 1The country and regional classifications used in this table follow the conventions of MSCI, and do not necessarily conform to IMF country classifications or regional groupings.

International Monetary Fund | October 2013 171

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 3. Emerging Markets Bond Index: Global Sovereign Yield Spreads (Basis points) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

End-of-period spread levels

EMBI Global 294 289 426 266 308 266 307 353 Latin America 355 357 468 326 369 326 358 424 Argentina 660 507 925 991 897 991 1,307 1,199 Belize 1,177 617 1,391 2,245 2,399 2,245 789 872 Brazil 189 189 225 140 162 140 190 243 Chile 95 115 172 116 143 116 153 180 Colombia 198 172 191 112 132 112 147 193 Dominican Republic 405 322 597 343 418 343 385 401 Ecuador 769 913 846 826 743 826 700 665 El Salvador 326 302 478 396 426 396 350 436 Jamaica 719 427 637 711 662 711 680 623 Mexico 192 173 222 155 172 155 182 223 Panama 166 162 201 129 148 129 169 218 Peru 165 165 216 114 125 114 147 201 Uruguay 238 188 213 127 139 127 173 235 Venezuela 1,041 1,114 1,258 786 956 786 797 976 Europe 226 231 440 208 265 208 270 300 Bulgaria 179 195 340 94 124 94 99 114 Belarus … 623 1,164 695 831 695 630 747 Croatia 195 298 602 311 349 311 386 361 Georgia 467 504 471 357 389 357 363 402 Hungary 186 345 605 345 383 345 429 352 Kazakhstan 393 324 453 207 247 207 270 316 Lithuania 332 267 447 149 209 149 182 211 Poland 124 151 310 108 144 108 145 157 Romania … … … 235 362 235 283 270 Russia 203 224 364 157 208 157 210 234 Serbia 333 418 601 391 523 391 416 440 Turkey 197 177 385 177 236 177 229 255 Ukraine 989 461 940 632 722 632 631 782 Middle East 335 284 439 426 442 426 450 450 Iraq 447 314 603 465 504 465 494 576 Jordan … … 500 436 405 436 446 348 Lebanon 287 270 384 412 420 412 437 419 Africa 211 329 452 264 278 264 312 381 Côte d'Ivoire … 1,154 1,192 473 563 473 483 573 Egypt –3 221 607 453 444 453 672 764 Gabon 390 258 422 252 265 252 237 313 Ghana 462 363 534 397 374 397 412 525 Nigeria … … 435 261 345 261 288 369 South Africa 149 145 261 163 176 163 217 268 Asia 206 175 271 165 187 165 201 239 China 64 126 278 146 171 146 153 194 Indonesia 230 183 274 179 204 179 226 275 Malaysia 136 117 178 98 131 98 128 163 Philippines 206 163 242 121 144 121 153 172 Sri Lanka 382 290 461 342 352 342 379 436 Vietnam 314 323 510 304 352 304 280 358

172 International Monetary Fund | October 2013

Table 3. (concluded) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Period-on-period spread level changes

EMBI Global –430 –6 138 –161 –66 –42 42 46 Latin America –391 2 111 –142 –59 –43 32 66 Argentina –1,044 –153 418 66 –191 94 316 –108 Belize –613 –560 774 854 708 –154 –1,456 83 Brazil –240 0 36 –85 –46 –22 50 53 Chile –248 20 57 –56 –24 –27 37 27 Colombia –300 –26 19 –79 –26 –20 35 46 Dominican Republic –1,200 –83 275 –254 –70 –75 42 16 Ecuador –3,962 144 –67 –20 –149 83 –126 –35 El Salvador –528 –24 176 –82 –54 –30 –46 86 Jamaica –466 –292 210 74 22 49 –31 –57 Mexico –242 –19 49 –67 –30 –17 27 41 Panama –373 –4 39 –72 –39 –19 40 49 Peru –344 0 51 –102 –49 –11 33 54 Uruguay –447 –50 25 –86 –58 –12 46 62 Venezuela –823 73 144 –472 –173 –170 11 179 Europe –514 5 209 –232 –88 –57 62 30 Bulgaria –495 16 145 –246 –168 –30 5 15 Belarus … … 541 –469 –98 –136 –65 117 Croatia … 103 304 –291 –202 –38 75 –25 Georgia –1,434 37 –33 –114 –44 –32 6 39 Hungary –318 159 260 –260 –152 –38 84 –77 Kazakhstan –910 –69 129 –246 –107 –40 63 46 Lithuania … –65 180 –298 –109 –60 33 29 Poland –190 27 159 –202 –71 –36 37 12 Romania ………… –101 –127 48 –13 Russia –602 21 140 –207 –86 –51 53 24 Serbia –891 85 183 –210 –30 –132 25 24 Turkey –337 –20 208 –208 –68 –59 52 26 Ukraine –1,782 –528 479 –308 –175 –90 –1 151 Middle East –599 –51 155 –13 –32 –16 24 0 Iraq –835 –133 289 –138 –142 –39 29 82 Jordan … … … –64 –9 31 10 –98 Lebanon –507 –17 114 28 0 –8 25 –18 Africa –559 118 123 –188 –68 –14 48 69 Côte d'Ivoire … … 38 –719 –140 –90 10 90 Egypt –388 224 386 –154 –87 9 219 92 Gabon –796 –132 164 –170 –65 –13 –15 76 Ghana –1,023 –99 171 –137 –130 23 15 113 Nigeria … … … –174 –70 –84 27 81 South Africa –413 –4 116 –98 –37 –13 54 51 Asia –390 –31 96 –106 –63 –22 36 38 China –164 62 152 –132 –51 –25 7 41 Indonesia –532 –47 91 –95 –69 –25 47 49 Malaysia –234 –19 61 –80 –43 –33 30 35 Philippines –340 –43 79 –121 –62 –23 32 19 Sri Lanka –1,484 –92 171 –119 –108 –10 37 57 Vietnam –433 9 187 –206 –73 –48 –24 78

Source: JPMorgan Chase & Co. Note: EMBI = emerging market bond index. The country and regional classifications used in this table follow the conventions of JPMorgan, and do not necessarily conform to IMF country classifications or regional groupings.

International Monetary Fund | October 2013 173

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 4. Emerging Market Private External Financing: Total Bonds, Equities, and Loans (Millions of U.S. dollars) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Total 665,710.5 870,762.4 748,194.8 774,263.5 191,324.0 212,277.1 238,771.6 236,487.3 Sub-Saharan Africa 35,821.7 36,549.1 39,820.8 49,491.4 18,882.5 11,932.4 5,880.8 7,650.8 Angola 2,495.2 4,391.2 3,110.7 2,847.9 231.3 2,446.1 87.1 … Benin 120.0 … 11.0 … … … … … Botswana 2,465.0 869.9 255.0 79.7 … 79.7 … … Burkina Faso 75.3 10.9 … … … … … 111.9 Burundi … 15.0 69.5 157.6 … … 150.0 … Cameroon 125.0 … 239.6 492.1 142.6 47.2 … 91.5 Cape Verde … 78.7 10.0 … … … … … Chad … … 14.7 … … … … … Congo … 250.0 … … … … … … Congo, Democratic Republic of the 300.0 12.2 169.9 100.0 … … … … Côte d'Ivoire … 97.0 930.9 152.6 152.6 … … 163.0 Djibouti … … 1.2 … … … … … Equatorial Guinea … … 390.0 600.0 … 600.0 … … Eritrea 446.0 … … … … … … … Ethiopia 590.7 1,342.8 1,694.3 1,497.8 … 1,263.3 … … Gabon 68.7 500.2 197.5 493.8 58.0 85.8 350.0 … Ghana 1,775.4 2,113.2 6,052.9 7,210.1 1,601.0 1,529.6 666.3 45.0 Guinea … … 34.8 198.9 … … … … Guinea-Bissau … … 60.4 … … … … … Kenya 539.5 703.4 660.8 1,510.4 22.5 1,050.7 101.0 … Liberia 24.5 1,902.5 11.0 24.9 … … … 130.0 Madagascar … 78.8 … … … … 75.2 … Malawi … … 39.0 … … … … … Mali 10.5 … 68.5 … … … … … Mauritania 1,108.0 … 64.4 … … … … … Mauritius 49.0 … 14.0 240.0 … … 240.0 270.0 Mozambique 188.0 164.9 206.9 84.5 8.2 12.0 64.3 … Namibia 196.5 … 536.2 23.2 … 23.2 … 60.6 Niger … 20.0 … 15.0 … … … … Nigeria 7,757.8 3,919.5 4,372.3 3,171.6 1,363.0 1,174.5 846.3 3,778.9 Rwanda 70.0 14.0 284.9 13.6 … … … 392.9 Senegal 413.2 348.1 515.4 … … … … … Seychelles 21.2 … … … … … … … Sierra Leone … 44.4 217.6 95.3 52.0 43.3 … … South Africa 15,486.4 16,352.2 17,641.0 16,426.9 6,065.5 3,075.7 2,498.7 2,607.1 South Sudan … … … 8,000.0 8,000.0 … … … Sudan … 89.3 … 2,000.0 … … … … Tanzania 440.0 398.0 1,078.4 331.4 123.0 … 742.0 … Togo 566.9 … … 52.4 … 52.4 … … Uganda 319.0 2,242.5 25.0 225.8 … 65.8 60.0 … Zambia 90.0 533.0 576.2 3,065.9 1,062.8 3.1 … … Zimbabwe 80.0 57.5 267.0 380.0 … 380.0 … … Central and Eastern Europe 97,264.6 95,287.8 124,007.1 91,350.2 24,148.4 30,275.6 27,569.1 13,076.6 Albania 116.8 405.3 145.9 168.4 104.8 … … … Bosnia and Herzegovina 400.6 70.5 92.0 6.2 6.2 … … … Bulgaria 1,043.9 360.0 281.2 2,268.2 1,611.9 94.5 265.0 98.3 Croatia 5,656.5 2,602.0 4,376.9 4,213.0 1,022.8 567.7 1,976.2 … Hungary 7,793.3 5,390.4 11,289.2 2,646.5 1,777.3 628.7 3,370.8 739.8 Latvia 4,565.4 374.5 528.8 2,233.5 … 1,238.9 … … Lithuania 3,929.3 3,023.6 1,666.4 2,442.4 189.2 135.3 953.1 38.9 Macedonia 460.8 113.6 357.3 61.5 … 61.5 … … Montenegro 42.1 369.2 287.0 15.9 … … … … Poland 21,206.8 42,385.8 49,518.6 20,789.3 4,992.2 5,892.4 7,218.4 1,173.8 Romania 29,133.6 3,241.7 13,469.9 6,181.9 1,326.1 2,029.6 1,793.1 443.3 Serbia 2,424.9 577.0 2,757.9 3,218.4 1,041.8 743.5 1,476.0 … Turkey 20,490.8 36,374.1 39,235.9 47,105.1 12,076.1 18,883.6 10,516.6 10,582.4

174 International Monetary Fund | October 2013

Table 4. (continued) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Commonwealth of Independent States 122,043.2 106,405.1 119,504.0 138,877.6 34,155.8 32,526.4 55,947.5 57,329.8 Armenia 150.2 105.6 143.6 31.0 … 30.0 1.0 … Azerbaijan 1,065.3 3,615.1 1,417.0 1,372.7 27.0 320.0 1,523.2 … Belarus 3,353.1 1,837.7 5,102.9 11,008.0 … … 1,008.0 … Georgia1 980.5 298.8 878.7 1,135.9 17.5 … 127.2 … Kazakhstan 16,979.7 5,979.6 3,751.6 8,707.2 1,650.1 3,257.8 1,576.6 5,311.9 Kyrgyzstan 116.7 5.8 3.0 … … … … … Moldova 138.9 23.2 21.7 5.0 … 5.0 … … Mongolia1 4.4 1,228.6 271.7 3,676.6 778.0 1,500.0 52.2 … Russian Federation 92,218.4 83,897.8 92,327.4 100,048.3 24,946.9 23,113.7 46,529.4 49,478.6 Tajikistan 148.2 10.5 8.0 … … … … … Turkmenistan 4,036.7 500.2 4,225.0 … … … … … Ukraine 2,823.0 8,491.1 11,328.6 9,938.9 4,157.4 4,300.0 5,130.0 2,539.3 Uzbekistan 28.2 411.2 25.0 2,954.0 2,579.0 … … … Developing Asia 156,812.3 248,908.7 198,236.5 224,300.0 53,177.2 61,000.5 69,043.4 73,613.5 Bangladesh 176.9 197.0 228.0 1,686.6 855.0 150.0 481.6 46.1 Bhutan … 47.4 … … … … … … Brunei Darussalam … … … 353.5 … 169.9 183.6 … Cambodia 462.7 65.0 591.0 155.8 … 51.4 156.0 … China 61,106.4 110,851.3 77,541.7 91,490.5 21,819.0 24,124.2 27,504.3 38,335.6 Cook Islands 4.4 … … … … … … … Fiji … … 250.2 … … … … … India 32,656.2 68,426.0 51,928.0 47,200.3 11,068.4 15,276.5 17,902.3 12,920.3 Indonesia 21,825.2 24,685.5 28,711.4 29,765.3 5,328.9 6,674.4 6,258.9 11,723.1 Laos 146.0 1,143.2 120.0 241.5 241.5 … … 275.0 Malaysia 12,466.0 12,677.4 11,206.0 19,753.0 5,249.2 3,715.9 4,931.8 2,443.4 Maldives … … 2.0 16.0 … … … 115.0 Marshall Islands … 660.0 1,946.5 497.9 … 162.0 215.9 … Myanmar … 2,400.0 … … … … … … Pakistan 742.7 516.2 1,270.6 1,499.2 333.2 … 1.5 … Papua New Guinea 14,078.5 … 980.3 222.0 … 57.0 … … Philippines 8,767.4 12,556.5 8,168.1 9,491.4 2,451.6 1,797.6 4,383.0 4,797.9 Sri Lanka 560.0 1,310.8 1,791.6 2,076.9 1,270.9 44.1 199.4 666.9 Thailand 1,642.8 7,975.3 6,674.4 16,259.8 4,220.6 7,034.6 6,096.7 2,190.2 Vietnam 2,176.9 5,397.1 6,826.5 3,590.3 339.0 1,742.8 728.4 100.0

International Monetary Fund | October 2013 175

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 4. (concluded) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Middle East and North Africa 78,193.9 110,589.4 74,247.8 76,504.9 16,166.1 25,917.3 19,399.4 29,184.1 Afghanistan 65.0 … … … … … … … Algeria 33.6 12.9 … … … … … … Bahrain 3,314.1 5,666.3 2,513.0 3,439.7 570.4 749.3 169.0 695.7 Egypt 3,863.0 16,541.0 8,788.1 2,978.2 240.0 110.8 335.8 3,196.4 Iran 78.7 … … 214.9 … 214.9 … … Iraq … 991.0 831.8 … … … … … Jordan 938.0 1,479.0 1,860.0 277.0 … … 270.0 608.0 Kuwait 2,568.0 3,962.2 2,246.0 1,223.5 150.0 820.3 150.0 428.4 Lebanon 1,118.0 2,103.5 2,687.4 2,324.3 … 1,320.9 … 1,095.6 Libya … … 40.0 … … … … … Morocco 618.5 3,764.8 318.0 3,509.3 884.5 2,534.8 … 748.9 Oman 1,065.6 3,604.9 2,310.5 994.2 251.2 656.0 499.2 6.3 Qatar 21,179.1 14,025.9 13,900.9 16,379.6 6,443.9 5,441.8 1,692.5 2,145.5 Saudi Arabia 2,178.4 17,766.7 9,627.0 12,576.6 3,326.2 3,693.9 2,917.2 3,499.3 Syria 796.9 … … … … … … … Tunisia 973.3 930.2 997.7 1,586.7 485.0 303.4 9.2 155.0 United Arab Emirates 39,295.0 39,320.9 27,563.7 30,876.0 3,690.0 10,071.3 13,356.4 16,605.1 West Bank and Gaza 85.0 50.3 … 125.0 125.0 … … … Yemen 23.7 369.8 563.7 … … … … … Latin America and the Caribbean 175,574.8 273,022.3 192,378.6 193,739.5 44,793.9 50,625.0 60,931.5 55,632.5 Argentina 3,715.2 5,143.8 10,142.3 2,697.6 485.0 241.1 455.9 460.0 Barbados 450.0 403.3 … 340.0 340.0 … … 400.0 Belize 2,500.0 … … … … … … … Bolivia 280.0 253.0 200.0 500.0 … 500.0 … … Brazil 88,841.8 154,757.3 77,356.4 72,747.0 13,377.4 16,288.3 15,703.2 28,174.1 Chile 5,388.8 12,541.6 17,444.7 22,714.9 6,457.1 5,370.1 12,731.8 4,128.2 Colombia 8,959.1 5,519.3 16,500.1 11,885.1 3,832.5 2,242.1 4,033.8 3,687.5 Costa Rica 853.5 31.0 479.0 1,322.4 … 999.9 … 1,491.8 Dominican Republic 556.7 2,475.6 1,209.0 900.0 … 550.0 447.4 1,000.0 Ecuador 430.0 22.0 36.0 … … … … … El Salvador 1,485.0 644.1 653.5 1,099.9 … 799.9 300.0 … Guatemala 46.0 604.0 333.2 1,429.9 40.0 500.0 690.8 109.2 Guyana 24.8 … … … … … … … Haiti 149.3 … … … … … … … Honduras … … … … … … 500.0 … Jamaica 1,804.1 1,833.5 1,568.4 1,770.5 1,500.0 … 1,300.0 0.5 Mexico 39,934.7 45,669.4 38,804.1 59,989.8 14,978.4 19,955.7 20,254.0 9,914.0 Nicaragua 50.4 185.0 … … … … … … Panama 3,924.6 1,478.2 3,138.2 1,906.7 1,247.8 2.0 … 1,258.0 Paraguay 234.0 … 100.0 651.0 … 651.0 500.0 … Peru 5,735.1 9,333.9 4,841.5 10,717.5 2,033.4 1,692.1 3,572.4 4,923.4 Trinidad and Tobago 843.3 93.5 182.5 27.2 … … 27.2 … Uruguay 1,195.5 … 3,323.8 720.0 … 720.0 … … Venezuela 8,061.8 31,807.7 15,635.8 2,320.0 502.4 113.0 414.9 … Source: Dealogic. Note: For inclusion criteria, please see notes for Tables 5, 6, and 7. 1Georgia and Mongolia, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarities in economic structure.

176 International Monetary Fund | October 2013

Table 5. Emerging Market Private External Financing: Bonds (Millions of U.S. dollars) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Total 165,931.9 253,731.1 251,312.0 362,620.6 91,012.9 105,837.7 112,972.2 110,726.8 Sub-Saharan Africa 3,709.2 4,673.1 8,113.1 8,306.3 3,924.9 1,204.4 1,565.7 900.0 Botswana … … … 79.7 … 79.7 … … Namibia … … 490.6 … … … … … Nigeria … … 986.0 350.0 350.0 … 575.0 298.4 Rwanda … … … … … … … 392.9 Senegal 196.1 … 487.9 … … … … … South Africa 3,513.1 4,673.1 6,148.7 7,140.8 2,839.1 1,124.7 390.7 208.7 Tanzania … … … … … … 600.0 … Zambia … … … 735.8 735.8 … … … Central and Eastern Europe 24,188.9 30,879.2 31,853.1 53,149.2 12,560.8 19,470.4 15,529.7 5,476.1 Albania … 405.3 … … … … … … Bulgaria … … … 1,343.3 1,343.3 … … … Croatia 3,098.8 1,238.8 2,748.4 3,104.0 718.6 500.0 1,485.8 … Hungary 3,545.9 3,518.1 8,752.3 1,763.5 1,134.8 628.7 3,236.0 259.8 Latvia … … 490.8 2,233.5 … 1,238.9 … … Lithuania 2,380.9 2,710.1 1,495.7 2,214.4 189.2 … 860.4 … Macedonia 243.2 … … … … … … … Montenegro … 252.8 252.5 … … … … … Poland 11,116.0 11,512.9 7,773.0 16,283.2 3,541.3 5,002.5 1,950.2 … Romania … 1,418.4 2,106.1 5,182.5 1,007.8 1,921.9 1,484.4 … Serbia … … 982.6 1,785.3 1,041.8 743.5 1,476.0 … Turkey 3,804.0 9,822.9 7,251.6 19,239.6 3,584.0 9,434.9 5,036.8 5,216.3 Commonwealth of Independent States 13,773.1 42,733.5 31,293.4 62,013.5 17,842.3 21,635.6 22,282.2 22,258.3 Azerbaijan … … 125.0 500.0 … … 1,000.0 … Belarus … 1,327.3 800.0 … … … … … Georgia1 … 248.8 491.2 996.1 … … … … Kazakhstan 2,299.2 4,840.5 1,072.9 3,242.7 875.9 2,366.7 422.4 4,080.9 Mongolia1 … 174.0 … 2,979.0 299.0 1,500.0 … … Russian Federation 11,473.9 30,869.5 22,924.6 48,852.9 13,024.5 15,968.9 16,952.7 15,661.6 Ukraine … 5,273.4 5,879.8 5,442.8 3,642.8 1,800.0 3,907.1 2,515.8 Developing Asia 23,037.7 45,968.7 58,630.7 83,008.5 18,366.6 19,811.4 34,079.7 37,795.9 China 2,233.9 18,058.6 30,654.6 40,404.0 6,352.9 11,636.3 15,267.3 23,823.6 Fiji … … 250.0 … … … … … India 2,140.6 9,045.8 9,307.0 10,435.2 5,066.8 2,758.9 6,427.4 4,840.6 Indonesia 7,840.6 5,794.1 6,363.9 12,336.4 362.2 2,985.7 2,508.8 6,886.4 Malaysia 5,007.2 2,638.5 4,170.8 8,929.0 2,386.9 89.9 4,217.1 700.0 Philippines 5,315.5 6,400.0 4,175.6 3,769.5 … 1,250.2 2,208.9 1,045.3 Sri Lanka 500.0 1,000.0 1,000.0 1,500.0 1,000.0 … … 500.0 Thailand … 2,046.0 2,622.3 5,387.1 3,198.0 1,090.2 3,450.1 … Vietnam … 985.8 86.6 247.5 … … … …

International Monetary Fund | October 2013 177

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 5. (concluded) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Middle East and North Africa 35,480.0 32,505.4 26,666.5 40,670.8 7,508.1 15,775.9 12,090.9 11,993.6 Bahrain 750.0 2,460.5 1,050.0 2,343.6 96.3 749.3 … 695.7 Egypt 295.6 2,095.3 500.0 … … … 40.0 3,196.4 Jordan … 741.6 … … … … … … Kuwait 493.6 989.3 446.7 923.5 … 820.3 … 161.7 Lebanon 944.7 1,925.0 2,687.4 2,278.3 … 1,320.9 … 1,095.6 Morocco … 1,340.1 … 1,479.6 … 1,479.6 … 748.9 Oman … 320.0 … … … … 496.5 … Qatar 15,284.0 8,743.5 5,087.7 10,508.8 5,068.9 3,453.9 1,324.9 1,495.5 Saudi Arabia 990.0 2,586.4 … 3,800.0 … 500.0 2,000.0 2,147.6 Tunisia … … … 1,288.4 485.0 303.4 … … United Arab Emirates 16,722.2 11,303.8 16,894.7 18,048.7 1,858.0 7,148.6 8,229.5 2,452.3

Latin America and the Caribbean 65,743.1 96,971.4 94,755.2 115,472.3 30,810.2 27,940.1 27,424.0 32,303.0 Argentina 506.7 4,129.9 2,552.8 1,111.2 … … 298.0 200.0 Barbados 450.0 403.3 … 250.0 250.0 … … 400.0 Bolivia … … … 500.0 … 500.0 … … Brazil 25,427.7 40,513.3 38,988.5 51,539.5 10,650.7 10,134.0 8,491.3 16,740.1 Chile 1,976.9 7,522.3 5,795.8 9,631.6 3,350.6 4,195.1 3,097.5 2,887.2 Colombia 5,922.2 1,939.8 6,374.3 7,342.5 3,732.5 … 3,289.8 1,089.1 Costa Rica … … 250.0 1,262.4 … 999.9 … 1,491.8 Dominican Republic … 750.0 777.6 550.0 … 550.0 297.4 1,000.0 El Salvador 800.0 444.1 653.5 799.9 … 799.9 … … Guatemala … … … 1,389.9 … 500.0 690.8 109.2 Honduras … … … … … … 500.0 … Jamaica 1,042.3 1,083.3 695.2 1,750.0 1,500.0 … 1,300.0 … Mexico 17,039.4 27,412.7 20,537.9 31,424.4 8,776.0 8,986.2 6,121.2 5,100.7 Panama 1,324.2 … 1,045.8 797.8 797.8 … … 750.0 Paraguay … … 100.0 500.0 … 500.0 500.0 … Peru 2,910.4 6,476.3 2,394.7 6,123.1 1,752.7 275.1 2,837.9 2,534.9 Trinidad and Tobago 843.3 … 175.0 … … … … … Uruguay 500.0 … 1,969.8 500.0 … 500.0 … … Venezuela 7,000.0 6,296.5 12,444.2 … … … … … Source: Dealogic. Note: Search criteria by deal nationality and filtered by international tranche, and excludes money market and short-term bonds, and supranationals. Deal inclusion conforms to the vendor’s criteria for external public and private sector syndicated gross issuance, generally excluding bilateral deals. 1Georgia and Mongolia are not members of the Commonwealth of Independent States, but are included in this group for reasons of geography and similarities in economic structure.

178 International Monetary Fund | October 2013

Table 6. Emerging Market Private External Financing: Equities (Millions of U.S. dollars) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Total 125,005.1 246,593.3 100,115.0 116,598.4 34,602.8 37,791.0 39,563.9 32,037.2 Sub-Saharan Africa 2,364.7 3,599.5 2,690.5 3,377.2 348.0 1,689.9 828.5 916.7 Botswana … 44.9 … … … … … … Burkina Faso 64.7 … … … … … … … Ghana … 45.5 … … … … … … Kenya … … 37.9 30.2 21.5 8.7 … … Madagascar … 78.8 … … … … 75.2 … Mauritius … … 14.0 … … … … … Niger … 20.0 … … … … … … Nigeria 64.5 140.3 … 219.9 … … 189.5 393.2 Rwanda … … 90.9 … … … … … South Africa 1,668.6 3,266.5 2,452.7 3,055.8 326.5 1,609.9 563.8 523.4 Tanzania … … 23.7 … … … … … Togo 566.9 … … 2.4 … 2.4 … … Uganda … 3.5 … 65.8 … 65.8 … … Zambia … … 71.2 3.1 … 3.1 … … Central and Eastern Europe 3,836.3 10,482.8 5,980.9 6,986.3 1,019.2 3,455.3 4,641.8 986.4 Bulgaria 6.6 … 18.4 1.6 … 1.6 … … Hungary … … 14.7 … … … 134.7 … Latvia 3.0 … … … … … … … Lithuania … 209.7 … 30.2 … 30.2 … 38.9 Poland 3,702.5 8,827.6 4,865.3 2,911.2 1,019.2 889.9 4,507.1 333.8 Romania … … … 76.3 … 26.1 … 167.8 Turkey 124.1 1,445.6 1,082.5 3,967.0 … 2,507.6 … 445.9 Commonwealth of Independent States 8,024.0 9,663.7 11,517.4 10,083.2 6,382.7 3,192.6 1,728.7 4,518.1 Armenia 2.4 … 11.6 … … … … … Georgia1 … … … … … … 74.9 … Kazakhstan 435.1 309.2 1.3 593.6 … 593.6 … … Kyrgyzstan … 5.8 … … … … … … Mongolia1 3.4 683.5 … 81.6 … … … … Russian Federation 7,377.5 8,005.0 11,137.0 9,400.6 6,375.2 2,599.0 1,637.6 4,499.6 Ukraine 205.7 660.1 367.5 7.4 7.4 … 16.3 18.5 Developing Asia 78,452.7 120,325.9 48,827.4 65,877.7 18,201.3 20,542.3 18,979.1 13,951.0 Bangladesh 70.5 … 86.0 … … … … … Cambodia … … … 155.8 … 51.4 156.0 … China 53,008.0 74,859.7 31,816.5 30,993.5 10,643.2 6,980.9 9,011.0 5,685.2 Fiji … … … … … … … … India 17,480.3 26,200.8 8,409.5 14,476.6 1,627.1 5,899.5 5,393.0 1,959.8 Indonesia 1,803.2 8,066.6 3,259.4 3,581.8 1,217.3 984.5 1,837.0 1,351.8 Laos … 111.2 … 241.5 241.5 … … … Malaysia 5,195.2 6,917.5 2,634.2 7,310.2 2,460.3 1,998.4 507.5 1,161.4 Maldives … … … 16.0 … … … … Pakistan … … … … … … 1.5 … Philippines 466.9 1,783.3 1,047.1 2,721.3 1,123.4 427.4 1,019.1 2,131.6 Sri Lanka … 5.6 … … … … … … Thailand 410.3 2,379.0 1,514.7 6,380.9 888.6 4,200.2 1,054.1 1,661.2 Vietnam 18.3 2.1 60.0 … … … … …

International Monetary Fund | October 2013 179

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 6. (concluded) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Middle East and North Africa 2,358.4 4,466.7 414.7 5,296.8 2,457.6 672.0 370.0 531.3 Bahrain 300.0 1,585.4 … … … … … … Egypt 114.2 1,095.3 … … … … … … Iran 78.7 … … … … … … … Iraq … … 8.5 … … … … … Kuwait 91.5 … … … … … … … Morocco … 20.8 13.0 … … … … … Oman … 474.8 63.9 357.2 251.2 106.0 2.7 6.3 Qatar 952.2 137.5 … 2,073.6 … 188.0 260.9 … Saudi Arabia 639.9 720.8 105.6 2,384.8 2,206.4 143.9 97.2 28.0 Syria 37.0 … … … … … … … Tunisia 19.8 175.7 … … … … 9.2 155.0 United Arab Emirates 125.1 206.2 223.7 481.3 … 234.2 … 341.9 West Bank and Gaza … 50.3 … … … … … … Latin America and the Caribbean 29,969.0 98,054.6 30,684.1 24,977.3 6,194.2 8,238.9 13,015.8 11,133.8 Argentina 319.4 109.7 4,978.0 60.2 … … 127.9 … Brazil 26,123.3 94,356.7 14,339.4 8,650.6 1,079.9 2,958.9 3,114.2 8,239.0 Chile 92.5 1,309.7 5,252.9 4,319.3 715.1 492.9 5,226.1 … Colombia 921.6 295.5 5,307.2 2,461.5 … 1,129.5 … 879.4 Jamaica … … … … … … … … Mexico 2,278.2 1,692.7 765.3 8,693.6 4,399.3 3,197.7 4,518.3 2,015.0 Panama … 103.0 41.3 … … … … … Peru 234.1 187.4 … 792.2 … 460.0 29.5 … Source: Dealogic. Note: Search criteria by issuer nationality and filters by initial and follow-on offerings, and international tranche. Deal inclusion conforms to the vendor’s criteria for external public and private sector syndicated gross issuance, generally excluding bilateral deals. 1Gerogia and Mongolia are not members of the Commonwealth of Independent States, but are included in this group for reasons of geography and similarities in economic structure.

180 International Monetary Fund | October 2013

Table 7. Emerging Market Private External Financing: Loans (Millions of U.S. dollars) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Total 374,773.5 370,438.0 396,767.8 295,044.5 65,708.3 68,648.4 86,235.5 93,723.4 Sub-Saharan Africa 29,747.8 28,276.5 29,017.3 37,807.9 14,609.7 9,038.0 3,486.6 5,834.2 Angola 2,495.2 4,391.2 3,110.7 2,847.9 231.3 2,446.1 87.1 … Benin 120.0 … 11.0 … … … … … Botswana 2,465.0 825.0 255.0 … … … … … Burkina Faso 10.6 10.9 … … … … … 111.9 Burundi … 15.0 69.5 157.6 … … 150.0 … Cameroon 125.0 … 239.6 492.1 142.6 47.2 … 91.5 Cape Verde … 78.7 10.0 … … … … … Chad … … 14.7 … … … … … Congo … 250.0 … … … … … … Congo, Democratic Republic 300.0 12.2 169.9 100.0 … … … … of the Côte d'Ivoire … 97.0 930.9 152.6 152.6 … … 163.0 Djibouti … … 1.2 … … … … … Equatorial Guinea … … 390.0 600.0 … 600.0 … … Eritrea 446.0 … … … … … … … Ethiopia 590.7 1,342.8 1,694.3 1,497.8 … 1,263.3 … … Gabon 68.7 500.2 197.5 493.8 58.0 85.8 350.0 … Ghana 1,775.4 2,067.7 6,052.9 7,210.1 1,601.0 1,529.6 666.3 45.0 Guinea … … 34.8 198.9 … … … … Guinea-Bissau … … 60.4 … … … … … Kenya 539.5 703.4 622.8 1,480.2 … 1,042.0 101.0 … Liberia 24.5 1,902.5 11.0 24.9 … … … 130.0 Malawi … … 39.0 … … … … … Mali 10.5 … 68.5 … … … … … Mauritania 1,108.0 … 64.4 … … … … … Mauritius 49.0 … … 240.0 … … 240.0 270.0 Mozambique 188.0 164.9 206.9 84.5 8.2 12.0 64.3 … Namibia 196.5 … 45.6 23.2 … 23.2 … 60.6 Niger … … … 15.0 … … … … Nigeria 7,693.3 3,779.2 3,386.3 2,601.8 1,013.0 1,174.5 81.8 3,087.2 Rwanda 70.0 14.0 194.0 13.6 … … … … Senegal 217.2 348.1 27.5 … … … … … Seychelles 21.2 … … … … … … … Sierra Leone … 44.4 217.6 95.3 52.0 43.3 … … South Africa 10,304.7 8,412.5 9,039.6 6,230.3 2,900.0 341.1 1,544.2 1,875.0 South Sudan … … … 8,000.0 8,000.0 … … … Sudan … 89.3 … 2,000.0 … … … … Tanzania 440.0 398.0 1,054.7 331.4 123.0 … 142.0 … Togo … … … 50.0 … 50.0 … … Uganda 319.0 2,239.0 25.0 160.0 … … 60.0 … Zambia 90.0 533.0 505.0 2,327.0 327.0 … … … Zimbabwe 80.0 57.5 267.0 380.0 … 380.0 … … Central and Eastern Europe 69,239.5 53,925.8 86,173.2 31,214.7 10,568.5 7,349.9 7,397.7 6,614.1 Albania 116.8 … 145.9 168.4 104.8 … … … Bosnia and Herzegovina 400.6 70.5 92.0 6.2 6.2 … … … Bulgaria 1,037.3 360.0 262.8 923.4 268.6 92.9 265.0 98.3 Croatia 2,557.6 1,363.2 1,628.5 1,108.9 304.2 67.7 490.4 … Hungary 4,247.4 1,872.3 2,522.3 883.0 642.5 … … 480.0 Latvia 4,562.4 374.5 38.0 … … … … … Lithuania 1,548.5 103.8 170.8 197.7 … 105.1 92.7 … Macedonia 217.5 113.6 357.3 61.5 … 61.5 … … Montenegro 42.1 116.5 34.5 15.9 … … … … Poland 6,388.2 22,045.3 36,880.3 1,594.9 431.8 … 761.2 840.1 Romania 29,133.6 1,823.3 11,363.8 923.1 318.3 81.6 308.6 275.5 Serbia 2,424.9 577.0 1,775.3 1,433.1 … … … … Turkey 16,562.6 25,105.7 30,901.8 23,898.4 8,492.0 6,941.1 5,479.8 4,920.3

International Monetary Fund | October 2013 181

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 7. (continued) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Commonwealth of Independent States 100,246.1 54,008.0 76,693.2 66,780.9 9,930.9 7,698.3 31,936.5 30,553.5 Armenia 147.8 105.6 132.0 31.0 … 30.0 … … Azerbaijan 1,065.3 3,615.1 1,292.0 872.7 27.0 320.0 523.2 … Belarus 3,353.1 510.5 4,302.9 11,008.0 … … 1,008.0 … Georgia1 980.5 50.0 387.5 139.8 17.5 … 52.3 … Kazakhstan 14,245.4 829.9 2,677.4 4,871.0 774.1 297.5 1,154.2 1,231.0 Kyrgyzstan 116.7 … 3.0 … … … … … Moldova 138.9 23.2 21.7 5.0 … 5.0 … … Mongolia1 … 371.0 271.7 616.0 479.0 … 52.2 … Russian Federation 73,367.1 45,023.3 58,265.9 41,794.8 5,547.1 4,545.8 27,939.1 29,317.5 Tajikistan 148.2 10.5 8.0 … … … … … Turkmenistan 4,036.7 500.2 4,225.0 … … … … … Ukraine 2,617.3 2,557.6 5,081.2 4,488.7 507.2 2,500.0 1,206.6 5.0 Uzbekistan 28.2 411.2 25.0 2,954.0 2,579.0 … … … Developing Asia 55,321.9 82,614.2 90,778.4 75,413.9 16,609.3 20,646.8 15,984.7 21,866.6 Bangladesh 106.4 197.0 142.0 1,686.6 855.0 150.0 481.6 46.1 Bhutan … 47.4 … … … … … … Brunei Darussalam … … … 353.5 … 169.9 183.6 … Cambodia 462.7 65.0 591.0 … … … … … China 5,864.6 17,933.0 15,070.7 20,093.1 4,823.0 5,507.0 3,226.0 8,826.9 Cook Islands 4.4 … … … … … … … India 13,035.3 33,179.4 34,211.5 22,288.5 4,374.6 6,618.1 6,081.9 6,119.9 Indonesia 12,181.4 10,824.8 19,088.2 13,847.1 3,749.5 2,704.2 1,913.1 3,484.8 Laos 146.0 1,032.0 120.0 … … … … 275.0 Malaysia 2,263.7 3,121.4 4,401.1 3,513.8 402.0 1,627.6 207.1 582.0 Maldives … … 2.0 … … … … 115.0 Marshall Islands … 660.0 1,946.5 497.9 … 162.0 215.9 … Myanmar … 2,400.0 … … … … … … Pakistan 742.7 516.2 1,270.6 1,499.2 333.2 … … … Papua New Guinea 14,078.5 … 980.3 222.0 … 57.0 … … Philippines 2,985.1 4,373.2 2,945.4 3,000.6 1,328.2 120.0 1,155.0 1,621.0 Sri Lanka 60.0 305.2 791.6 576.9 270.9 44.1 199.4 166.9 Thailand 1,232.5 3,550.4 2,537.5 4,491.8 134.1 1,744.2 1,592.5 529.0 Vietnam 2,158.7 4,409.3 6,680.0 3,342.8 339.0 1,742.8 728.4 100.0 Middle East and North Africa 40,355.6 73,617.4 47,166.6 30,537.2 6,200.4 9,469.4 6,938.4 16,659.2 Afghanistan 65.0 … … … … … … … Algeria 33.6 12.9 … … … … … … Bahrain 2,264.1 1,620.4 1,463.0 1,096.1 474.1 … 169.0 … Egypt 3,453.2 13,350.4 8,288.1 2,978.2 240.0 110.8 295.8 … Iran … … … 214.9 … 214.9 … … Iraq … 991.0 823.3 … … … … … Jordan 938.0 737.4 1,860.0 277.0 … … 270.0 608.0 Kuwait 1,982.8 2,972.9 1,799.3 300.0 150.0 … 150.0 266.7 Lebanon 173.3 178.5 … 46.0 … … … … Libya … … 40.0 … … … … … Morocco 618.5 2,403.9 305.0 2,029.7 884.5 1,055.2 … … Oman 1,065.6 2,810.1 2,246.6 637.0 … 550.0 … … Qatar 4,942.8 5,145.0 8,813.3 3,797.2 1,375.0 1,800.0 106.7 650.0 Saudi Arabia 548.5 14,459.6 9,521.4 6,391.8 1,119.8 3,050.0 820.0 1,323.7 Syria 759.9 … … … … … … … Tunisia 953.5 754.5 997.7 298.3 … … … … United Arab Emirates 22,447.8 27,811.0 10,445.2 12,346.0 1,832.0 2,688.5 5,126.9 13,810.9 West Bank and Gaza 85.0 … … 125.0 125.0 … … … Yemen 23.7 369.8 563.7 … … … … …

182 International Monetary Fund | October 2013

Table 7. (concluded) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Latin America and the Caribbean 79,862.7 77,996.3 66,939.2 53,289.9 7,789.5 14,446.0 20,491.7 12,195.8 Argentina 2,889.1 904.3 2,611.6 1,526.2 485.0 241.1 30.0 260.0 Barbados … … … 90.0 90.0 … … … Belize 2,500.0 … … … … … … … Bolivia 280.0 253.0 200.0 … … … … … Brazil 37,290.8 19,887.3 24,028.5 12,556.9 1,646.8 3,195.5 4,097.8 3,195.0 Chile 3,319.5 3,709.5 6,396.0 8,764.0 2,391.5 682.1 4,408.2 1,241.0 Colombia 2,115.3 3,284.0 4,818.6 2,081.2 100.0 1,112.6 744.0 1,719.0 Costa Rica 853.5 31.0 229.0 60.0 … … … … Dominican Republic 556.7 1,725.6 431.4 350.0 … … 150.0 … Ecuador 430.0 22.0 36.0 … … … … … El Salvador 685.0 200.0 … 300.0 … … 300.0 … Guatemala 46.0 604.0 333.2 40.0 40.0 … … … Guyana 24.8 … … … … … … … Haiti 149.3 … … … … … … … Jamaica 761.8 750.2 873.2 20.5 … … … … Mexico 20,617.1 16,564.1 17,500.9 19,871.8 1,803.1 7,771.8 9,614.5 2,798.3 Nicaragua 50.4 185.0 … … … … … … Panama 2,600.4 1,375.1 2,051.1 1,108.9 450.0 2.0 … 508.0 Paraguay 234.0 … … 151.0 … 151.0 … … Peru 2,590.6 2,670.3 2,446.8 3,802.1 280.7 956.9 705.0 2,388.5 Trinidad and Tobago … 93.5 7.5 27.2 … … 27.2 … Uruguay 695.5 … 1,354.0 220.0 … 220.0 … … Venezuela 1,061.8 25,511.2 3,191.6 2,320.0 502.4 113.0 414.9 … Source: Dealogic. Note: Search criteria by deal nationality and filters by hard currency. Deal inclusion conforms to the vendor's criteria for external public and private sector syndicated gross issuance, generally excluding bilateral deals. 1Georgia and Mongolia, which are not members of the Commonwealth of Independent States, are included in this group for reasons of geography and similarities in economic structure.

International Monetary Fund | October 2013 183

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 8. Equity Valuation Measures: Dividend-Yield Ratios 2012 2013 10-year 2008 2009 2010 2011 2012 Q3 Q4 Q1 Q2 average

Emerging Markets 4.1 2.0 2.1 3.0 2.7 2.9 2.7 2.8 2.9 2.6 Asia 4.2 1.7 2.0 2.8 2.3 2.5 2.3 2.4 2.5 2.4 Europe/Middle East/Africa 4.3 2.2 2.1 3.2 3.5 3.6 3.5 3.5 3.7 2.7 Latin America 4.0 2.7 2.3 3.3 3.2 3.5 3.2 3.2 3.5 2.9 Argentina 2.7 1.1 1.9 8.4 5.6 16.6 5.6 5.4 5.8 3.2 Brazil 4.6 2.9 2.7 4.1 4.1 4.6 4.1 4.1 4.7 3.5 Chile 2.6 1.6 1.4 2.3 2.1 2.2 2.1 2.1 1.9 2.1 China 3.1 1.9 2.2 3.2 2.9 3.2 2.9 3.1 3.5 2.4 Colombia 2.4 2.8 2.1 2.6 2.9 4.1 2.9 3.0 3.2 2.7 Egypt 6.3 4.8 3.5 5.3 3.5 3.2 3.5 3.1 3.3 3.5 Hungary 4.6 1.3 1.6 2.4 3.5 3.1 3.5 3.5 3.6 2.3 India 1.8 0.9 0.9 1.5 1.3 1.3 1.3 1.4 1.4 1.3 Indonesia 5.4 1.9 2.2 2.4 2.4 2.5 2.4 2.2 2.4 2.9 Jordan 3.4 3.1 2.5 3.1 4.3 4.4 4.3 4.7 4.8 2.7 Malaysia 4.1 2.4 2.3 2.8 2.9 2.9 2.9 2.9 2.9 2.7 Mexico 2.8 2.4 1.6 1.2 1.4 1.3 1.4 1.5 1.5 1.8 Morocco 3.2 4.9 4.3 5.5 4.7 4.7 4.7 4.8 4.5 3.9 Pakistan 12.5 6.4 5.6 8.3 7.1 7.5 7.1 7.6 7.0 6.6 Philippines 4.4 2.2 2.4 2.7 2.0 2.2 2.0 1.6 1.8 2.3 Poland 5.9 3.0 2.5 5.4 5.6 6.0 5.6 5.9 6.1 3.6 Russia 3.5 1.4 1.5 2.4 3.6 3.6 3.6 3.8 3.9 2.1 South Africa 4.5 2.7 2.3 3.2 3.2 3.2 3.2 3.2 3.2 3.0 Sri Lanka 9.8 1.6 1.2 2.2 2.2 2.2 2.2 2.1 2.1 2.6 Thailand 6.5 2.9 2.6 3.3 2.8 3.0 2.8 2.8 2.9 3.4 Turkey 5.8 2.1 2.2 3.3 2.2 2.5 2.2 1.8 2.3 2.8 Source: Morgan Stanley Capital International (MSCI). Note: The country and regional classifications used in this table follow the conventions of MSCI, and do not necessarily conform to IMF country classifications or regional groupings.

184 International Monetary Fund | October 2013

Table 9. Equity Valuation Measures: Price/Earnings Ratios 2012 2013 10-year 2008 2009 2010 2011 2012 Q3 Q4 Q1 Q2 average

Emerging Markets 8.5 20.6 14.6 10.8 12.7 12.5 12.7 12.5 11.8 13.9 Asia 9.4 24.3 15.2 11.4 13.2 13.3 13.2 13.1 12.1 14.9 Europe/Middle East/Africa 6.7 16.2 12.1 8.2 9.2 8.9 9.2 8.8 8.5 12.5 Latin America 9.0 18.3 15.9 11.8 16.4 15.3 16.4 16.1 15.8 13.8 Argentina 3.7 8.0 8.8 5.2 3.3 3.0 3.3 3.8 3.9 18.4 Brazil 7.9 17.0 13.8 9.8 14.3 13.0 14.3 13.8 13.5 12.1 Chile 13.3 18.7 21.4 17.2 23.2 20.1 23.2 23.9 25.6 21.7 China 10.3 21.1 14.6 9.4 11.3 10.0 11.3 10.7 9.2 15.1 Colombia 13.4 25.1 23.5 17.2 19.3 17.7 19.3 19.3 18.4 19.3 Egypt 7.1 13.9 17.4 10.3 13.8 16.7 13.8 13.6 17.2 15.8 Hungary 3.7 14.2 12.2 8.7 13.7 15.4 13.7 10.4 12.6 11.5 India 10.5 21.8 22.4 14.4 16.3 16.2 16.3 15.4 15.4 18.6 Indonesia 8.7 16.4 19.0 15.2 16.2 16.4 16.2 18.7 17.7 14.9 Jordan 14.4 15.9 21.3 16.9 11.6 11.0 11.6 12.3 13.0 22.7 Malaysia 10.2 20.3 18.1 16.9 14.8 15.8 14.8 15.5 16.6 16.3 Mexico 12.3 22.7 23.9 21.8 21.5 23.8 21.5 21.2 20.9 17.7 Morocco 26.0 14.3 17.5 14.0 12.3 12.6 12.3 12.5 11.6 20.6 Pakistan 3.8 10.1 9.1 6.2 7.6 7.0 7.6 7.8 8.9 9.6 Philippines 11.7 19.1 17.5 15.8 19.9 18.5 19.9 22.4 20.6 16.9 Poland 7.3 19.3 14.1 8.0 8.7 8.2 8.7 9.6 10.2 13.7 Russia 3.4 15.6 8.3 4.9 5.6 5.5 5.6 5.1 4.9 10.2 South Africa 10.7 16.6 18.9 16.4 15.9 14.8 15.9 16.4 16.4 15.3 Sri Lanka 7.1 77.7 20.5 13.2 14.1 14.7 14.1 15.2 14.6 19.2 Thailand 7.1 19.3 14.8 11.1 15.9 16.0 15.9 15.1 15.0 12.9 Turkey 5.3 12.6 10.8 9.2 12.0 11.1 12.0 12.5 10.5 11.1 Source: Morgan Stanley Capital International (MSCI). Note: The country and regional classifications used in this table follow the conventions of MSCI, and do not necessarily conform to IMF country classifications or regional groupings.

International Monetary Fund | October 2013 185

©International Monetary Fund. Not for Redistribution GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY

Table 10. Emerging Markets: Mutual Funds Net Flows (Billions of U.S. dollars) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Bonds 9.7 53.3 15.9 58.8 16.4 18.9 21.2 –9.2 Global 9.5 46.5 13.6 47.1 16.0 16.9 18.2 –8.0 Asia 0.3 6.6 2.6 0.8 0.4 1.6 2.9 –0.6 Europe/Middle East/Africa –0.5 –0.2 –1.0 –0.4 –0.1 0.2 0.3 0.6 Latin America 0.4 0.4 0.8 0.7 0.1 0.1 –0.3 –1.2 Equities 83.0 95.7 –46.2 52.3 5.0 33.4 29.8 –31.9 Global 44.2 63.6 –4.6 33.8 6.9 19.5 23.3 –10.7 Asia 26.1 22.1 –23.7 –6.2 –2.9 14.5 8.8 –14.5 Europe/Middle East/Africa 1.5 7.3 –7.0 –1.7 0.3 –1.1 –1.5 –2.6 Latin America 11.1 2.6 –10.9 –1.5 0.8 0.5 –0.7 –4.0

Net Asset Values (Billions of U.S. dollars) 2012 2013 2009 2010 2011 2012 Q3 Q4 Q1 Q2

Bonds 87.8 162.0 183.3 301.8 265.9 301.8 349.0 375.7 Global 76.2 141.9 157.3 264.7 233.8 264.7 291.7 316.8 Asia 7.0 14.5 20.0 28.7 24.4 28.7 32.5 34.1 Europe/Middle East/Africa 3.0 3.2 3.0 4.5 4.0 4.5 5.3 6.0 Latin America 1.5 2.4 2.9 4.0 3.8 4.0 19.6 18.7 Equities 702.1 950.2 774.1 1,016.9 930.2 1,016.9 1,057.0 1,068.9 Global 334.5 476.8 416.4 562.4 513.2 562.4 582.9 581.6 Asia 257.3 329.7 262.6 343.9 309.1 343.9 361.4 373.7 Europe/Middle East/Africa 42.7 62.6 40.1 52.2 52.7 52.2 53.1 58.5 Latin America 67.6 81.1 55.0 58.4 55.2 58.4 59.6 55.1 Source: EPFR Global. Note: Flows data derive from both tradtional and alternative funds domiciled globally with $19 trillion in assets. The country and regional classifications used in this table follow the conventions of Emerging Portfolio Fund Research and individual fund managers, and do not necessarily conform to IMF country classifications or regional groupings.

186 International Monetary Fund | October 2013

©International Monetary Fund. Not for Redistribution INTERNATIONAL MONETARY FUND Expand your global expertise. Visit the IMF Bookstore. World Economic Outlook Energy Subsidy Reform: Lessons The World Economic Outlook, and Implications packed with country-specific Energy subsidies have wide-ranging facts, figures, and global economic consequences. This book projections, presents the provides the most comprehensive outlook for growth, inflation, estimates of energy subsidies trade, and other economic currently available for 176 countries. developments in a clear, It also presents an analysis of “how practical format. to do” energy subsidy reform, drawing on insights from 22 country Fiscal Monitor case studies undertaken by the IMF The Fiscal Monitor, published twice a year, surveys staff and analyses carriedo ut by other institutions. and analyzes developments in public finance, updates medium-term fiscal projections, and assesses policies to put public financeso n a sustainable footing. The Monitor’s projections are based on the same database used for the World Economic Outlook.

China’s Economy in Transition: From External to Internal Rebalancing This volume looks at various aspects of the rebalancing under way in China. It focuses on three themes: the reasons for the decline in China’s current account and signs of growing domestic imbalances; the implications for trading partners; and policy lessons to achieve Order Now: Visit the IMF Bookstore today and browse a stable, sustainable, and more inclusive transformation to find the latest reports, publications, and data. of China’s growth model. IMF BOOKSTORE imfbookstore.org/go13

Bookstore ad for GFSR_285c.indd 1 10/2/13 2:18 PM This page intentionally left blank

©International Monetary Fund. Not for Redistribution Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing − − − − − − ECB EU Policy Netherlands From (mo. Yr.) (mo. Oct 2008 Y pre-2007 Y pre-2007 Y pre-2007 Y pre-2007 Until (mo. Yr.) (mo. Jan 2010 ongoing Y − − − ECB Germany EU Policy From (mo. Yr.) (mo. Oct 2008 Y Jan 2009 Y Until (mo. Yr.) (mo. ongoing ongoing Oct 2009 − − − − ECB France EU Policy From (mo. Yr.) (mo. Oct 2013 Oct 2008 Jan 2013 Y Y Y Until (mo. Yr.) (mo. ongoing ongoing − − − ECB Belgium EU Policy From (mo. Yr.) (mo. Aug 2012 May 2009 Y Until (mo. Yr.) (mo. ongoing Oct 2012 ongoing ongoing Y ongoing ongoing ongoing − − − − − − − − ECB Euro Area / EU Wide From Jul 2009 (mo. Yr.) (mo. Oct 2008 Jan 2007 Jan 2007 Jan 2007 Jun 2013 Jun 2013 ©International Monetary Fund. Not for Redistribution ©International Monetary Y Y Y Y Y Y Y assets bank liquidity operations liquidity operations corporates households Widening of collateral eligibility to include private Widening of collateralsector eligibility to include Allow nonbank financial institutions to access central Allow nonfinancial corporations to access central bank Purchase of corporate bonds Purchase of corporate stocks, ETFs Purchase of corporate stocks, other corporate short-term assets MMF, Purchase of CP, other real-estate-related assets REIT, Purchase of MBS, Guarantees on asset prices Special lending facilities to promote bank Special lending facilities to promote bank Direct provision Subsidies and tax programs Guarantees Direct provision Subsidies and tax programs Guarantees capital adequacy ratio capital adequacy seizure With conditions to expand bank lending Monetary policy operationMonetary policy Direct credit easing Indirect credit easing Corporate loans and funding Mortgage loans Reduction of risk weights for SME loans when calculating Forbearance on recognizing nonperforming loans/collateral Forbearance macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate bonds Create securitization markets for SME loans Create securitization markets for household debt Recapitalization program Monetary policies, direct involvement with nonbanks Monetary policies, Fiscal programs by governments and state-ownedFiscal institutions Financial sector regulations Financial Capital markets Other policies to enhance credit supply Bank restructuring programs Euro Area Enhancing Credit Supply Mitigating Debt Overhang Appendix Table 2.1. Policies Implemented to Support Credit Markets Policies 2.1. Appendix Table

International Monetary Fund | October 2013 1 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY Until (mo. Yr.) (mo. − EU Policy EU Policy Netherlands From (mo. Yr.) (mo. Y Until (mo. Yr.) (mo. − Germany EU Policy EU Policy From (mo. Yr.) (mo. Y Until (mo. Yr.) (mo. − France EU Policy EU Policy From (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing ongoing − − − Belgium EU Policy EU Policy From (mo. Yr.) (mo. Oct 2008 Dec 2012 Y Y Y 2 Until (mo. Yr.) (mo. ongoing ongoing − − − Euro Area / EU Wide From (mo. Yr.) (mo. Sep 2009 Dec 2011 ©International Monetary Fund. Not for Redistribution ©International Monetary Y Y Y institutions companies restructure loans procedures institutions companies restructure loans With capital ratio requirement higher than Basel III Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Ad hoc public assistance Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in corporate-bankruptcy-related Improvements in accounting standards for SMEs Changes in securities and other related laws workout plan Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Centralized arbitration scheme Moratorium on debt service write-down of loans Forced procedures Legal changes in personal-bankruptcy-related workout plan banking depress credit flows Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Stress tests Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Basel Accord Ring-fencing and subsidiary requirements for cross-border Other policies to increase regulatory barriers to potentially Other policies to contain banking sector vulnerability Corporate debt restructuring Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum required by Euro Area, continued Euro Area, New Regulatory Barriers Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

2 International Monetary Fund | October 2013 Appendix TAble 2.1. Policies imPlemented to suPPort credit markets Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing − − − − − − − − − − − − ECB Slovenia EU Policy EU Policy From (mo. Yr.) (mo. Sep 2012 Sep 2012 Sep 2012 May 2013 May 2013 May 2011 Y Y Y pre-2007 Y pre-2007 Y pre-2007 Y pre-2007 Y Y Y Y Y pre-2007 Y Y Y Until (mo. Yr.) (mo. ongoing − − ECB EU Policy EU Policy Slovak Republic From (mo. Yr.) (mo. July 2009 Y Until (mo. Yr.) (mo. − ECB Finland EU Policy EU Policy From (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing − − ECB Estonia EU Policy EU Policy From (mo. Yr.) (mo. Mar 2011 Until (mo. Yr.) (mo. ongoing Dec 2010 ongoing Y − − − − ECB Austria EU Policy EU Policy From (mo. Yr.) (mo. Oct 2008 Oct 2008 Oct 2008 ©International Monetary Fund. Not for Redistribution ©International Monetary Y Y Y Y Y institutions companies restructure loans Direct provision Subsidies and tax programs Guarantees Direct provision Subsidies and tax programs Guarantees capital adequacy ratio capital adequacy seizure With conditions to expand bank lending With capital ratio requirement higher than Basel III Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Ad hoc public assistance Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Corporate loans and funding Mortgage loans Reduction of risk weights for SME loans when calculating Forbearance on recognizing nonperforming loans/collateral Forbearance Countercyclical macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate bonds Create securitization markets for SME loans Create securitization markets for household debt Recapitalization program Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Stress tests Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Monetary policies, direct involvement with nonbanks Monetary policies, Fiscal programs by governments and state-ownedFiscal institutions Financial sector regulations Financial Capital markets Other policies to enhance credit supply Bank restructuring programs Other policies to contain banking sector vulnerability Corporate debt restructuring Euro Area, continued Euro Area, Enhancing Credit Supply Mitigating Debt Overhang Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

International Monetary Fund | October 2013 3 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY Until (mo. Yr.) (mo. ongoing ongoing − − − Slovenia EU Policy From (mo. Yr.) (mo. May 2012 May 2012 Y Y Until (mo. Yr.) (mo. ongoing − − Slovak Republic EU Policy Jan 2012 Until (mo. Yr.) (mo. Y Until (mo. Yr.) (mo. − Finland EU Policy Until (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing − − − − − − Estonia EU Policy Apr 2011 Apr 2011 Apr 2011 Apr 2011 Jun 2011 Until (mo. Yr.) (mo. Y Y Y Y Y Until (mo. Yr.) (mo. ongoing ongoing − − − Austria From (mo. Yr.) (mo. Jan 2013 Jan 2013 ©International Monetary Fund. Not for Redistribution ©International Monetary Y Y procedures institutions companies restructure loans Legal changes in corporate-bankruptcy-related Improvements in accounting standards for SMEs Changes in securities and other related laws workout plan Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Centralized arbitration scheme Moratorium on debt service write-down of loans Forced procedures Legal changes in personal-bankruptcy-related workout plan Coordination of creditors (and debtors) to reach orderly Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Basel Accord banking depress credit flows Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum required by Ring-fencing and subsidiary requirements for cross-border Other policies to increase regulatory barriers to potentially Euro Area, concluded Euro Area, New Regulatory Barriers Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

4 International Monetary Fund | October 2013 Appendix TAble 2.1. Policies imPlemented to suPPort credit markets Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing ongoing ongoing − − − − − − − − ECB Spain EU Policy EU Policy From Jul 2012 Jul 2011 (mo. Yr.) (mo. Jul 2012 Jun 2013 Mar 2010 Y pre-2007 Y pre-2007 Y Y Y Y Y Y Until (mo. Yr.) (mo. Dec 2013 Jun 2013 − − − ECB Portugal EU Policy EU Policy From (mo. Yr.) (mo. Jan 2008 Jun 2012 Y Y Y Y Until (mo. Yr.) (mo. ongoing ongoing Y ongoing ongoing ongoing Dec 2009 ongoing − − − − − − − − ECB Italy EU Policy EU Policy From (mo. Yr.) (mo. Feb 2009 Feb Jan 2000 Jun 2012 Aug 2011 Aug 2009 May 2009 May 2011 Y Y Y Y Y Until (mo. Yr.) (mo. ongoing Y ongoing Y ongoing ongoing Oct 2010 ongoing ongoing ongoing Y − − − − − − − − − ECB Ireland EU Policy EU Policy From (mo. Yr.) (mo. Jan 2009 . 2011 Nov 2009 Sep 2008 Sep 2008 May 2011 Y pre-2007 Y pre-2007 Y Y Y Y Y Y Y Until (mo. Yr.) (mo. ongoing Y Jun 2013 Apr 2012 − − − − ECB Greece EU Policy EU Policy From (mo. Yr.) (mo. Apr 2012 Dec 2012 ©International Monetary Fund. Not for Redistribution ©International Monetary Y pre-2007 Y Y institutions companies restructure loans Direct provision Subsidies and tax programs Guarantees Direct provision Subsidies and tax programs Guarantees ratio capital adequacy seizure With capital ratio requirement higher than Basel III With conditions to expand bank lending Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Ad hoc public assistance Corporate loans and funding Mortgage loans Reduction of risk weights for SME loans when calculating Forbearance on recognizing nonperforming loans/collateral Forbearance macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate bonds Create securitization markets for SME loans Create securitization markets for household debt Recapitalization program Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Stress tests Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal liabilities Monetary policies, direct involvement with nonbanks Monetary policies, Fiscal programs by governments and state-ownedFiscal institutions Financial sector regulations Financial Capital markets Other policies to enhance credit supply Bank restructuring programs Other policies to contain banking sector vulnerability Corporate debt restructuring Euro Area Stressed Countries Enhancing Credit Supply Mitigating Debt Overhang Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

International Monetary Fund | October 2013 5 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing − − − − − Spain EU Policy From (mo. Yr.) (mo. Jan 2012 Jan 2012 Nov 2012 May 2013 Y Y Y Until (mo. Yr.) (mo. ongoing ongoing Y ongoing ongoing ongoing − − − − − − Portugal EU Policy From (mo. Yr.) (mo. Jan 2012 Sep 2011 Sep 2011 Nov 2012 May 2012 Y Y Y Until (mo. Yr.) (mo. ongoing Y ongoing Y Mar 2013 ongoing − − − − − Italy EU Policy From (mo. Yr.) (mo. Feb 2010 Feb Aug 2009 Y Y Until (mo. Yr.) (mo. ongoing Y pre-2007 ongoing ongoing ongoing Jul 2013 ongoing Y pre-2007 ongoing − − − − − − − − Ireland EU Policy From Jul 2013 (mo. Yr.) (mo. Feb 2009 Feb Jan 2013 Jan 2013 Jan 2013 Dec 2012 Mar 2013 Y Y Y Y Until (mo. Yr.) (mo. . 2013 ongoing Y ongoing ongoing Y ongoing Y − − − − − − Greece From (mo. Yr.) (mo. . 2010 . 2010 . Aug 2013 Aug 2013 ©International Monetary Fund. Not for Redistribution ©International Monetary Y Y Y Y Y Y procedures institutions companies restructure loans Centralized arbitration scheme Moratorium on debt service Forced write-down of loans Forced Legal changes in corporate-bankruptcy-related Improvements in accounting standards for SMEs Changes in securities and other related laws workout plan Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Centralized arbitration scheme Moratorium on debt service Forced write-down of loans Forced Legal changes in personal-bankruptcy-related procedures Legal changes in personal-bankruptcy-related workout plan Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Basel Accord banking depress credit flows Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum required by Ring-fencing and subsidiary requirements for cross-border Other policies to increase regulatory barriers to potentially Euro Area Stressed Countries, concluded Euro Area Stressed Countries, New Regulatory Barriers Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

6 International Monetary Fund | October 2013 Appendix TAble 2.1. Policies imPlemented to suPPort credit markets Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing − − − − − − − − − − − − EU Policy United Kingdom From Jul 2012 Jul 2012 (mo. Yr.) (mo. Apr 2012 Jan 2009 Jan 2009 Jan 2009 Jan 2009 Sep 2007 Mar 2012 Y Y Y Y Y Y Y pre-2007 Y Y Y Y pre-2007 Y Until (mo. Yr.) (mo. ongoing − − Sweden EU Policy From (mo. Yr.) (mo. Y pre-2007 Until (mo. Yr.) (mo. Oct 2009 Nov 2009 ongoing ongoing ongoing − − − − − − Norway From (mo. Yr.) (mo. Oct 2008 Nov 2008 Mar 2009 Y Y Y Y pre-2007 Y pre-2007 Until (mo. Yr.) (mo. ongoing − − Iceland From (mo. Yr.) (mo. Y pre-2007 Until (mo. Yr.) (mo. ongoing ongoing − − − Denmark EU Policy Peg to euro Peg From (mo. Yr.) (mo. Aug 2011 Y Y pre-2007 ©International Monetary Fund. Not for Redistribution ©International Monetary 3 4 assets liquidity operations liquidity operations corporates households Widening of collateral eligibility to include private Widening of collateralsector eligibility to include Allow nonbank financial institutions to access central bank Allow nonfinancial corporations to access central bank Purchase of corporate bonds Purchase of corporate stocks, ETFs Purchase of corporate stocks, other corporate short-term assets MMF, Purchase of CP, Purchase of MBS, REIT, other real-estate-related assets REIT, Purchase of MBS, Guarantees on asset prices Special lending facilities to promote bank Special lending facilities to promote bank Direct provision ratio capital adequacy seizure Subsidies and tax programs Guarantees Direct provision Subsidies and tax programs Guarantees Monetary policy operationMonetary policy Direct credit easing Indirect credit easing Corporate loans and funding Reduction of risk weights for SME loans when calculating on recognizing nonperforming loans/collateral Forbearance macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate bonds Create securitization markets for SME loans Create securitization markets for household debt Mortgage loans Monetary policies, direct involvement with nonbanks Monetary policies, Fiscal programs by governments and state-ownedFiscal institutions Capital markets Other policies to enhance credit supply Financial sector regulations Financial Non-Euro Area Advanced Europe Non-Euro Area Advanced Enhancing Credit Supply Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

International Monetary Fund | October 2013 7 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY Until (mo. Yr.) (mo. Nov 2012 ongoing − − − EU Policy EU Policy United Kingdom From (mo. Yr.) (mo. Oct 2008 Jan 2009 Y Y Y Until (mo. Yr.) (mo. ongoing ongoing − − − Sweden EU Policy From (mo. Yr.) (mo. Oct 2010 Jan 2013 Y Y Until (mo. Yr.) (mo. − Norway EU Policy EU Policy From (mo. Yr.) (mo. Until (mo. Yr.) (mo. Dec 2010 Aug 2012 ongoing ongoing Jul 2009 ongoing Jun 2011 Dec 2012 Apr 2009 ongoing Dec 2012 Dec 2012 ongoing − − − − − − − − − − − − − − Iceland From Jul 2009 (mo. Yr.) (mo. Oct 2008 Oct 2009 Oct 2008 Oct 2009 Oct 2008 Oct 2009 Jan 2011 Jun 2010 Dec 2010 Aug 2012 Mar 2009 Y Y Y pre-2007 Y Y Y Y Y Y Y Y Y Until (mo. Yr.) (mo. ongoing Y Dec 2010 ongoing Y − − − − Denmark EU Policy EU Policy From (mo. Yr.) (mo. Oct 2008 Oct 2008 Oct 2008 Y Y Y ©International Monetary Fund. Not for Redistribution ©International Monetary institutions restructure loans institutions restructure loans With conditions to expand bank lending With capital ratio requirement higher than Basel III Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management companies Subsidies and tax programs to encourage banks to Ad hoc public assistance Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in corporate-bankruptcy-related procedures Improvements in accounting standards for SMEs Changes in securities and other related laws workout plan Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management companies Subsidies and tax programs to encourage banks to Centralized arbitration scheme Moratorium on debt service write-down of loans Forced procedures Legal changes in personal-bankruptcy-related workout plan Recapitalization program Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Stress tests Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Basel Accord banking depress credit flows Bank restructuring programs Other policies to contain banking sector vulnerability Corporate debt restructuring Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum required by Ring-fencing and subsidiary requirements for cross-border Other policies to increase regulatory barriers to potentially Non-Euro Area Advanced Europe, concluded Europe, Non-Euro Area Advanced Mitigating Debt Overhang New Regulatory Barriers Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

8 International Monetary Fund | October 2013 Appendix TAble 2.1. Policies imPlemented to suPPort credit markets Until (mo. Yr.) (mo. ongoing Jun 2010 Jun 2010 Feb 2010 Feb ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing − − − − − − − − − − − − − United States From (mo. Yr.) (mo. Nov 2008 Sep 2008 Sep 2009 Sep 2009 Sep 2009 Sep 2008 Sep 2008 Mar 2008 Mar 2008 Y pre-2007 Y Y Y Y Y Y Y Y Y Y pre-2007 Y pre-2007 Until (mo. Yr.) (mo. − South Africa From (mo. Yr.) (mo. Until (mo. Yr.) (mo. Nov 2009 Jul 2009 ongoing ongoing ongoing Mar 2009 ongoing ongoing ongoing − − − − − − − − − − Korea From (mo. Yr.) (mo. Oct 2008 Oct 2008 Oct 2008 Nov 2008 Nov 2008 Dec 2008 Dec 2008 Y Y pre-2007 Until ongoingongoing Y Y ongoing Y ongoing ongoingongoing Y ongoing ongoing (mo. Yr.) (mo. ongoingongoing Y ongoing Y Mar 2013 ongoing ongoing Y pre-2007 − − − − − − − − − − − − − − − Japan From (mo. Yr.) (mo. Oct 2010 Oct 2010 Oct 2012 Jan 2009 Dec 2008 Dec 2009 May 2010 Y pre-2007 Y pre-2007 Y Y Y Y Y pre-2007 Y pre-2007 Y Y Y pre-2007 6 Until (mo. Yr.) (mo. Mar 2010 ongoing Y pre-2007 ongoing ongoing Y ongoing ongoing Y pre-2007 ongoing − − − − − − − − India From … (mo. Yr.) (mo. Sep 2008 Sep 2008 Sep 2008 Y Y pre-2007 Y Y Y pre-2007 Y Y pre-2007 Until (mo. Yr.) (mo. ongoing − − ©International Monetary Fund. Not for Redistribution ©International Monetary Australia From (mo. Yr.) (mo. Oct 2007 Y 5 private sector assets central bank liquidity operations central bank liquidity operations term assets related assets lending to corporates lending to households Widening of collateral eligibility to include Widening of collateral eligibility to include Allow nonbank financial institutions to access Allow nonfinancial corporations to access Purchase of corporate bonds ETFs Purchase of corporate stocks, Purchase of CP, MMF, other corporate short- MMF, Purchase of CP, other real-estate- REIT, Purchase of MBS, Guarantees on asset prices Special lending facilities to promote bank Direct provision Subsidies and tax programs Guarantees Special lending facilities to promote bank Direct provision Subsidies and tax programs Guarantees calculating ratio capital adequacy loans/collateral seizure bonds nonbanks operationMonetary policy Direct credit easing Indirect credit easing institutions Corporate loans and funding Mortgage loans Reduction of risk weights for SME loans when on recognizing nonperforming Forbearance Countercyclical macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate Create securitization markets for SME loans Create securitization markets for household debt Monetary policies, direct involvement with Monetary policies, Fiscal programs by governments and state-ownedFiscal Financial sector regulations Financial Capital markets Other policies to enhance credit supply Enhancing Credit Supply Other Areas Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

International Monetary Fund | October 2013 9 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY Until (mo. Yr.) (mo. ongoing ongoing Dec 2012 Jun 2009 ongoing Dec 2012 ongoing ongoing ongoing ongoing − − − − − − − − − − − United States From (mo. Yr.) (mo. Jul 2008 Oct 2008 Oct 2008 Oct 2009 Oct 2008 Oct 2008 Jul 2008 Jul 2008 May 2009 May 2009 Y Y Y Y Y Y Y Y Y Y Y Until (mo. Yr.) (mo. ongoing Y − − South Africa From (mo. Yr.) (mo. Y May 2008 Until ongoing ongoing (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing − − − − − − − − Korea From (mo. Yr.) (mo. Feb 2009 Feb Mar 2009 Mar 2009 Mar 2013 Nov 2008 Mar 2009 Y Y Y Y Y Y pre-2007 Until (mo. Yr.) (mo. ongoing Y ongoing ongoing Y Mar 2013 − − − − − Japan From (mo. Yr.) (mo. Oct 2009 Oct 2009 Dec 2009 Y Y Y Y Until (mo. Yr.) (mo. … ongoing Y pre-2007 ongoing Y ongoing … … − − − − − − − India From … (mo. Yr.) (mo. … … Sep 2008 Y Y pre-2007 Y pre-2007 Y Y Y Until (mo. Yr.) (mo. ongoing Y − − ©International Monetary Fund. Not for Redistribution ©International Monetary Australia From (mo. Yr.) (mo. pre-2007 Y

Basel III state-owned institutions management companies banks to restructure loans related procedures state-owned institutions management companies banks to restructure loans procedures With conditions to expand bank lending With capital ratio requirement higher than liabilities Restructuring of loans provided or owned by Restructuring of loans using asset Subsidies and tax programs to encourage Ad hoc public assistance involvement) Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in corporate-bankruptcy- Improvements in accounting standards for SMEs Changes in securities and other related laws orderly workout plan liabilities Restructuring of loans provided or owned by Restructuring of loans using asset Subsidies and tax programs to encourage Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in personal-bankruptcy-related orderly workout plan Recapitalization program Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Stress tests Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal Legal approach (without direct fiscal Coordination of creditors (and debtors) to reach Government-led scheme with contingent fiscal Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach Accord required by the Basel border banking potentially depress credit flows Bank restructuring programs Other policies to contain banking sector vulnerability Corporate debt restructuring Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum Ring-fencing and subsidiary requirements for cross- Other policies to increase regulatory barriers to Other Areas, concluded Other Areas, Mitigating Debt Overhang New Regulatory Barriers Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

10 International Monetary Fund | October 2013 Appendix TAble 2.1. Policies imPlemented to suPPort credit markets Until (mo. Yr.) (mo. − Czech Republic From (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing ongoing Jul 2012 ongoing − − − − − Croatia Oct 2008 Feb 2010 Feb From Jan 2010 Mar 2011 (mo. Yr.) (mo. Y Y Y Y Until (mo. Yr.) (mo. ongoing ongoing Y − − − Bulgaria Peg to euro Peg Feb 2009 Feb From (mo. Yr.) (mo. Y Y pre-2007 Until (mo. Yr.) (mo. − Peg to euro Peg Bosnia and Herzegovina From (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing − − Albania Apr 2013 From (mo. Yr.) (mo. ©International Monetary Fund. Not for Redistribution ©International Monetary Y assets bank liquidity operations liquidity operations corporates households Widening of collateral eligibility to include private Widening of collateralsector eligibility to include Allow nonbank financial institutions to access central Allow nonfinancial corporations to access central bank Purchase of corporate bonds ETFs Purchase of corporate stocks, other corporate short-term assets MMF, Purchase of CP, other real-estate-related assets REIT, Purchase of MBS, Guarantees on asset prices Special lending facilities to promote bank Special lending facilities to promote bank Direct provision Subsidies and tax programs Guarantees Direct provision Subsidies and tax programs Guarantees ratio capital adequacy seizure Monetary policy operationMonetary policy Direct credit easing Indirect credit easing Corporate loans and funding Mortgage loans Reduction of risk weights for SME loans when calculating on recognizing nonperforming loans/collateral Forbearance macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate bonds Create securitization markets for SME loans Create securitization markets for household debt Monetary policies, direct involvement with nonbanks Monetary policies, Fiscal programs by governments and state-ownedFiscal institutions sector regulations Financial Capital markets Other policies to enhance credit supply Non-Euro-Area Central, Eastern, and Southeastern European Countries Eastern, Non-Euro-Area Central, Enhancing Credit Supply Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

International Monetary Fund | October 2013 11 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY Until (mo. Yr.) (mo. ongoing − − Czech Republic From (mo. Yr.) (mo. Feb 2010 Feb

Until (mo. Yr.) (mo. ongoing Y ongoing ongoing Oct 2008 − − − − − Croatia From (mo. Yr.) (mo. Oct 2012 Apr 2010 Jan 2006 Y Y Y Y Until (mo. Yr.) (mo. ongoing Y pre-2007 ongoing Y − − − Bulgaria From (mo. Yr.) (mo. Y pre-2007 Until (mo. Yr.) (mo. ongoing Y pre-2007 ongoing − − − From … Bosnia and Herzegovina (mo. Yr.) (mo. Dec 2008 Y Until (mo. Yr.) (mo. ongoing Y ongoing Y ongoing ongoing − − − − − Albania From (mo. Yr.) (mo. Apr 2013 Apr 2013 Mar 2009 ©International Monetary Fund. Not for Redistribution ©International Monetary Y pre-2007 Y Y Y Y institutions companies restructure loans procedures institutions companies restructure loans With conditions to expand bank lending With capital ratio requirement higher than Basel III Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Ad hoc public assistance Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in corporate-bankruptcy-related Improvements in accounting standards for SMEs Changes in securities and other related laws workout plan Restructuring of loans provided or owned by state-owned Restructuring of loans using asset management Subsidies and tax programs to encourage banks to Centralized arbitration scheme Moratorium on debt service write-down of loans Forced procedures Legal changes in personal-bankruptcy-related workout plan Recapitalization program Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Stress tests Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Government-led scheme with contingent fiscal liabilities Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach orderly Basel Accord banking depress credit flows Bank restructuring programs Other policies to contain banking sector vulnerability Corporate debt restructuring Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum required by Ring-fencing and subsidiary requirements for cross-border Other policies to increase regulatory barriers to potentially Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table continued and Southeastern European Countries, Eastern, Non-Euro-Area Central, Mitigating Debt Overhang New Regulatory Barriers

12 International Monetary Fund | October 2013 Appendix TAble 2.1. Policies imPlemented to suPPort credit markets 1 Until ongoing (mo. Yr.) (mo. − − Montenegro From (mo. Yr.) (mo. Unilateral euro adoption Y Y Apr 2010 Y Until Apr 2011 Apr 2011 (mo. Yr.) (mo. − − − Moldova From (mo. Yr.) (mo. Y May 2009 Y May 2009 Until ongoing ongoing ongoing ongoing ongoing ongoing (mo. Yr.) (mo. − − − − − − − FYR Macedonia From (mo. Yr.) (mo. Y Jul 2011 Y Jul 2011 Y Feb 2013 Y Feb Y Dec 2012 Y Dec 2012 Until Dec 2008 Y Jan 2012 (mo. Yr.) (mo. − − Lithuania Peg to euro Peg From (mo. Yr.) (mo. Y pre-2007 Until (mo. Yr.) (mo. − Latvia Peg to euro Peg From (mo. Yr.) (mo. Until ongoing ongoing ongoing Aug 2013 ongoing (mo. Yr.) (mo. − − − − − − ©International Monetary Fund. Not for Redistribution ©International Monetary Hungary From (mo. Yr.) (mo. Y pre-2007 Y pre-2007 Y Aug 2012 Y Y Jun 2013 Y Jun 2013 lending to households central bank liquidity operations central bank liquidity operations term assets related assets lending to corporates private sector assets bonds Guarantees calculating ratio capital adequacy loans/collateral seizure Direct provision Subsidies and tax programs Guarantees Direct provision Subsidies and tax programs Special lending facilities to promote bank Allow nonbank financial institutions to access Allow nonfinancial corporations to access Purchase of corporate bonds ETFs Purchase of corporate stocks, other corporate short- MMF, Purchase of CP, other real-estate- REIT, Purchase of MBS, Guarantees on asset prices Special lending facilities to promote bank Widening of collateral eligibility to include Widening of collateral eligibility to include Countercyclical macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate Create securitization markets for SME loans Create securitization markets for household debt Reduction of risk weights for SME loans when on recognizing nonperforming Forbearance Mortgage loans institutions Corporate loans and funding Direct credit easing Indirect credit easing Monetary policy operationMonetary policy nonbanks Capital markets Other policies to enhance credit supply Financial sector regulations Financial Fiscal programs by governments and state-ownedFiscal Monetary policies, direct involvement with Monetary policies, Non-Euro-Area Central, Eastern, and Southeastern European Countries, continued and Southeastern European Countries, Eastern, Non-Euro-Area Central, Enhancing Credit Supply Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

International Monetary Fund | October 2013 13 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY Until (mo. Yr.) (mo. ongoing ongoing − − − Montenegro From Oct 2008 (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing Y pre-2007 ongoing Y ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing − − − − − − − − − − − Moldova From Jul 2012 Jul 2012 Jul 2012 Jul 2012 Jul 2012 Jul 2012 Jan 2010 Dec 2011 Dec 2011 Dec 2011 (mo. Yr.) (mo. Y Y Y Y Y Y Y Y Until (mo. Yr.) (mo. ongoing Y ongoing Y − − − FYR Macedonia From May 2011 (mo. Yr.) (mo. Y pre-2007 Until (mo. Yr.) (mo. ongoing Y ongoing ongoing − − − − Lithuania 10 2008 From Mar 2013 (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing Y pre-2007 ongoing Y ongoing ongoing ongoing ongoing ongoing Y ongoing ongoing − − − − − − − − − − Latvia From Jul 2009 Jun 2011 Nov 2010 Nov 2010 Nov 2010 Aug 2009 Dec 2010 Aug 2009 Aug 2009 (mo. Yr.) (mo. Y Y Y Y Y Y Y Y Y Y Y Until (mo. Yr.) (mo. ongoing May 2013 − − − ©International Monetary Fund. Not for Redistribution ©International Monetary Hungary From Jun 2012 Sep 2011 Aug 2012 (mo. Yr.) (mo. Y Y Y state-owned institutions management companies banks to restructure loans related procedures state-owned institutions management companies banks to restructure loans procedures With conditions to expand bank lending With capital ratio requirement higher than Basel III liabilities Restructuring of loans provided or owned by Restructuring of loans using asset Subsidies and tax programs to encourage Ad hoc public assistance involvement) Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in corporate-bankruptcy- Improvements in accounting standards for SMEs Changes in securities and other related laws orderly workout plan liabilities Restructuring of loans provided or owned by Restructuring of loans using asset Subsidies and tax programs to encourage Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in personal-bankruptcy-related orderly workout plan Recapitalization program Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Stress tests Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal Legal approach (without direct fiscal Coordination of creditors (and debtors) to reach Government-led scheme with contingent fiscal Legal approach (without direct fiscal involvement) Coordination of creditors (and debtors) to reach required by the Basel Accord required by the Basel cross-border banking potentially depress credit flows Bank restructuring programs Other policies to contain banking sector vulnerability Corporate debt restructuring Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum Ring-fencing and subsidiary requirements for Other policies to increase regulatory barriers to Non-Euro-Area Central, Eastern, and Southeastern European Countries, continued and Southeastern European Countries, Eastern, Non-Euro-Area Central, Mitigating Debt Overhang New Regulatory Barriers Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

14 International Monetary Fund | October 2013 Appendix TAble 2.1. Policies imPlemented to suPPort credit markets Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing ongoing ongoing ongoing − − − − − − − − − Ukraine From (mo. Yr.) (mo. Jan 2012 Jan 2009 Jun 2013 Y Y pre-2007 Y pre-2007 Y Y pre-2007 Y pre-2007 Y Y Until (mo. Yr.) (mo. ongoing Y pre-2007 − − Turkey From (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing ongoing Nov 2012 ongoing ongoing ongoing ongoing ongoing ongoing Y pre-2007 − − − − − − − − − − Serbia From (mo. Yr.) (mo. Apr 2012 Jan 2010 Sep 2012 Dec 2011 Dec 2007 Y Y pre-2007 Y Y pre-2007 Y Y Y Until ongoing Y (mo. Yr.) (mo. Dec 2010 ongoing Y pre-2007 Jun 2010 Y pre-2007 ongoing ongoing Y − − − − − − − Russia From (mo. Yr.) (mo. Oct 2008 Oct 2008 Dec 2008 Y pre-2007 Y Y pre-2007 Y Y Y Until (mo. Yr.) (mo. ongoing ongoing ongoing Y pre-2007 ongoing − − − − − Romania From (mo. Yr.) (mo. Jun 2009 Nov 2008 Y Y pre-2007 Y Until (mo. Yr.) (mo. ongoing Y pre-2007 ongoing ongoing − − − − ©International Monetary Fund. Not for Redistribution ©International Monetary Poland From (mo. Yr.) (mo. Jul 2013 Mar 2013 Y Y pre-2007 Y private sector assets central bank liquidity operations central bank liquidity operations term assets related assets lending to corporates lending to households Widening of collateral eligibility to include Widening of collateral eligibility to include Allow nonbank financial institutions to access Allow nonfinancial corporations to access Purchase of corporate bonds ETFs Purchase of corporate stocks, other corporate short- MMF, Purchase of CP, other real-estate- REIT, Purchase of MBS, Guarantees on asset prices Special lending facilities to promote bank Special lending facilities to promote bank Direct provision Subsidies and tax programs Guarantees Direct provision Subsidies and tax programs Guarantees calculating ratio capital adequacy loans/collateral seizure With conditions to expand bank lending With capital ratio requirement higher than Basel III Monetary policy operationMonetary policy Direct credit easing Indirect credit easing institutions Corporate loans and funding Mortgage loans Reduction of risk weights for SME loans when on recognizing nonperforming Forbearance Stress tests Countercyclical macroprudential regulations Countercyclical Lower barriers for SMEs to issue corporate bonds Create securitization markets for SME loans Create securitization markets for household debt Recapitalization program Asset purchase scheme Guarantees for bank asset values Ad hoc public assistance Monetary policies, direct involvement with nonbanks Monetary policies, Fiscal programs by governments and state-ownedFiscal Financial sector regulations Financial Other policies to contain banking sector vulnerability Capital markets Other policies to enhance credit supply Bank restructuring programs Enhancing Credit Supply Non-Euro-Area Central, Eastern, and Southeastern European Countries, continued and Southeastern European Countries, Eastern, Non-Euro-Area Central, Mitigating Debt Overhang Appendix Table 2.1. Policies Implemented to Support Credit Markets (continued) Policies 2.1. Appendix Table

International Monetary Fund | October 2013 15 GLOBAL FINANCIAL STABILITY REPORT: TRANSITION ChALLENGES TO STABILITY - Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing ongoing − − − − − − − Ukraine From (mo. Yr.) (mo. Jan 2013 Aug 2012 Y Y pre-2007 Y Y pre-2007 Y pre-2007 Y pre-2007 Until (mo. Yr.) (mo. − Turkey From (mo. Yr.) (mo. Until (mo. Yr.) (mo. ongoing ongoing ongoing ongoing ongoing − − − − − − Serbia From (mo. Yr.) (mo. Jun 2010 Dec 2009 Dec 2011 May 2011

Y Y Y Y pre-2007 Until (mo. Yr.) (mo. ongoing Y ongoing ongoing − − − − Russia From (mo. Yr.) (mo. Oct 2008 Mar 2013 Y pre-2007 Until (mo. Yr.) (mo. ongoing Y ongoing ongoing ongoing ongoing Y − − − − − − Romania From (mo. Yr.) (mo. Oct 2008 Apr 2013 Jun 2012 Sep 2010 Mar 2009 Y Y Y Y Y Until (mo. Yr.) (mo. ongoing Y ongoing Y − − − ©International Monetary Fund. Not for Redistribution ©International Monetary Poland From (mo. Yr.) (mo. Jun 2013 Jun 2012 Y Y state-owned institutions management companies banks to restructure loans related procedures SMEs state-owned institutions management companies banks to restructure loans procedures liabilities Restructuring of loans provided or owned by Restructuring of loans using asset Subsidies and tax programs to encourage Ad hoc public assistance involvement) Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in corporate-bankruptcy- Improvements in accounting standards for Changes in securities and other related laws orderly workout plan liabilities Restructuring of loans provided or owned by Restructuring of loans using asset Subsidies and tax programs to encourage involvement) Centralized arbitration scheme Moratorium on debt service write-down of loans Forced Legal changes in personal-bankruptcy-related orderly workout plan Coverage enhancement of deposit insurance Government-led scheme with contingent fiscal Legal approach (without direct fiscal Coordination of creditors (and debtors) to reach Government-led scheme with contingent fiscal Legal approach (without direct fiscal Coordination of creditors (and debtors) to reach Accord required by the Basel cross-border banking potentially depress credit flows Corporate debt restructuring Household debt restructuring Other policies to mitigate debt overhang Higher capital requirement than the minimum Ring-fencing and subsidiary requirements for Other policies to increase regulatory barriers to Unilateral euro adoption refers to the situation in FYR Montenegro, which uses the euro as its currency without being a member of the euro zone. which uses the euro as its currency Unilateral euro adoption refers to the situation in FYR Montenegro, EBA recapitalization exercise. authorities have recently relaxed liquidity requirements for banks. U.K. The Breedon Review made concrete proposals that have been partially implemented. work in progress. the United Kingdom, For ensure thefinancialTo Bank stability, of Japan purchased corporate stocks from financial institutions from February 2009 to AprilA similar 2010. program existed between November 2002 and September 2004. for which the relevant statute an ongoing government contribution to the equity capital does not allow of banks that their ownership includes is a consequence of the partial government ownership stake to go below of banks, 51 “Y” the India, For Non-Euro-Area Central, Eastern, and Southeastern European Countries, concluded and Southeastern European Countries, Eastern, Non-Euro-Area Central, New Regulatory Barriers Appendix Table 2.1. Policies Implemented to Support Credit Markets (concluded) Policies 2.1. Appendix Table 1 2 3 4 5 6 Such contributions are a regular feature of the Indian banking system. percent. Source: IMF staff. Source: CP = commercial paper; ECB = European Central Bank; ETF exchange-traded fund; EU Union; MBS mortgage-backed security; MMF = money market fund; REIT real estate investment trust; SME = small and medium Note: indicateswas imple that such a policy “Y” Authority. such as the European Investment Bank and Banking measures that have refers to policy been taken by the ECB and other European institutions, Wide” Area/EU enterprises.”Euro mented is “ongoing” in used the when country. the policy is still where effective: it is entered for a legal introduction change, or amendment of a law is not meant to indicates“…” be temporary. insufficient information.

16 International Monetary Fund | October 2013