Global Institutional Consulting

Chief Investment Office Global Debt: MAY 2017 Emmanuel D. Hatzakis Challenges and Opportunities Director Chief Investment Office The world is deep in a flood tide of debt. Do we care, and what can we do about it? Debt is inextricably linked to economic growth. Whether in the form of government, business or household credit, debt can fuel economic growth, and economic growth can increase the propensity to borrow. And when growth is too weak, credit can be used to stimulate it. The cumulative effect of credit use by the various sectors of an economy is a rise in its overall debt. While economic prosperity is often credit-financed, higher levels of debt appear connected to lower future economic growth. Almost all debt accumulations have been hard to reverse, most have involved tough policy choices, and some notable ones have ended badly. In this whitepaper, we discuss the state of global debt, how it came to be, where it may lead, and what we believe portfolio implications to be.

The state of global debt More than eight years since the 2008 global financial crisis started, the world seems to be drowning in debt. Global economic growth, meanwhile, remains anemic, after years of sluggishness. Some economists attribute it to high levels of debt1, reasoning that they have been preventing economies from realizing their full potential because governments, businesses and households have been devoting significant resources to debt servicing — resources that otherwise would have been available for productive activities. Global debt was on a steep rise before the crisis, and continued growing after it. The McKinsey Global Institute estimates that, by mid-2015, the world’s debt stood at $204 trillion — $68 trillion (or 50%) higher than it had been in 2007, immediately before the crisis, and $121 trillion (or 145%) higher than in 2000. Governments, businesses and households all contributed to the increase. (See Exhibit 1.) What is more, global debt has been growing faster than the world’s economy. As of mid- 2015, it stood at 294% of global gross domestic product (GDP), up 25 percentage points (pp) since the end of 2007 and 48 pp since the end of 20002. In many countries, debt has increased to levels not normally seen during peacetime3. In advanced economies, government debt has had several spikes associated with wartime borrowing or economic crises over more than two centuries, and is now at its highest levels since World War II4 (see Exhibit 2). Businesses and households have never been as indebted before5.

Bank of America Merrill Lynch makes available products and services offered by Merrill Lynch Pierce Fenner & Smith (MLPF&S) and other subsidiaries of Bank of America Corporation (BofA Corp.). Investment products: Are Not FDIC Insured Are Not Bank Guaranteed May Lose Value

MLPF&S is a registered broker-dealer, a registered investment adviser and Member SIPC. Please see important disclosure © 2017 Bank of America Corporation. All rights reserved. information on the last page. Global Debt: Challenges and Opportunities | 2

Exhibit 1: Global Debt Outstanding, by Type Exhibit 3: China’s Outstanding Debt, by Type

Household Business-Non-Financial Government Business-Financial Household Business-Non-Financial Government Business - Financial 250 300 290%

$204 39pp 200 250 $40 + $121 + $68 200 150 $136 132pp $57 158% $32 150 20pp 100 121% $83 8pp $37 100 72pp $18 $58 51pp Debt as % of GDP 50 $25 $31 83pp 50 $21 42pp $36 $48 68pp $ Trillion, Constant Exchange Rates $19 23pp 24pp 0 0 7pp 2000 2007 Q2 2015 2000 2007 Q2 2015 TOTAL DEBT AS % OF GDP TOTAL DEBT, $ TRILLION, CONSTANT EXCHANGE RATES 246% 269% 294% $1.9 $6.9 $30.0

Source: McKinsey Global Institute and Chief Investment Office. Source: McKinsey Global Institute and Chief Investment Office. Note: Totals may be off due to rounding.

Exhibit 2: Global Government Debt Outstanding — How this came to be Advanced versus Emerging Economies Debt has characteristics that make it attractive both to Advanced Economies Emerging Economies borrowers, as a source of funds, and to lenders, as an 140 investment. Borrowers may like its favorable tax treatment, 120 and the limited surrender of control rights it offers when 100 compared to equity; lenders often perceive it as less

80 risky than equity, and its contractually specified stream

60 of cash flows may serve their needs with reasonable certainty. A confluence of factors contributed to the global 40 Debt as % of GDP debt accumulation7: 20

0 • In many economies, robust real GDP growth and booming real estate markets and construction led to 1800 1809 1818 1827 1836 1845 1854 1863 1872 1881 1890 1899 1908 1917 1926 1935 1944 1953 1962 1971 1980 1989 1998 2007 fast increases in credit. Source: Reinhart, Reinhart and Rogoff (2012). Data through 2012 for advanced economies and through 2011 for emerging economies. • The liberalization of the financial sector over the last three decades opened up major financial markets, and innovative ways to engineer debt instruments with risk profiles Emerging economies alone contributed 47% to the growth in perceived to be suitable for specific classes of lenders global debt since 2007. China, whose government launched an fueled a self-reinforcing increase in credit availability. ambitious stimulus program in late 2008, in response to the global financial crisis, played a leading role6, with its debt more • A savings glut, mainly in emerging economies, and than quadrupling since 2007 (see Exhibit 3). China’s debt is declining real interest rates across the world made mostly domestic, held by banks, and has been used primarily to credit cheaper and more available. finance real estate and other investments. Continued on next page. Global Debt: Challenges and Opportunities | 3

Debt accumulations are historically associated with slower Continued from previous page. future GDP growth, systemic instability, and economic and 9 • Credit use increased as both lenders and borrowers financial-system vulnerability . The faster the debt build-up, felt more comfortable in a relatively stable the worse the ensuing contraction has tended to be. In the macroeconomic environment. years leading up to the global financial crisis, major advanced economies saw household debt spiral upward to unsustainable • The 2008 global financial crisis caused governments to levels along with housing prices, until real estate collapsed and increase their borrowing steeply to offset tax revenue domestic spending slowed, with more-leveraged households declines and stimulate their economies. experiencing larger spending slowdowns10. Faster private- sector deleveraging and quicker balance-sheet recognition of Debt levels can increase in both good economic times and losses have resulted in more robust recoveries11. bad. When growth is too weak, governments may attempt to The economic slowdown following debt accumulation is well stimulate economic growth by investing or spending borrowed understood by those familiar with the aftermath of the lending funds, as well as adopt policies that encourage businesses booms in the U.S. in the 2000s and Japan in the 1980s, and and households to invest or spend using credit. Several prior research findings corroborate it12. Exhibit 4 outlines an illustrative periods of debt accumulation were followed by severe economic example. When domestic demand is fueled entirely by credit, $1 contractions8, and what we are experiencing now may have of borrowing generates $1 of spending, and the stock of debt similarities to past cycles. The current secular debt cycle is now increases by $1, all else being equal. To keep demand at the same more than five decades long, with deleveraging pauses in the level through credit, debt needs to keep increasing. When the 1980s and 1990s. It intensified in the last few years, raising flow of credit stops, demand goes back to what it was before some critical questions: borrowing, and the stock of debt remains at a high plateau. To • Should we be concerned? decrease debt, we need to reverse the credit flow, which causes • Does this pose any risks? If it does, is there a way out? a contraction in demand. What is more, weak demand leads • Do we need to deleverage at any cost, or should we rather to disinflation or even deflation, which, in turn, triggers easier focus on learning to manage the debt and its growth trends? monetary policy, currency depreciation, and increased vulnerability of the banking system to crises. • Could it be that it does not matter as much as we fear? What if debt has become a structural element of the modern global Since debt overhangs have often happened during or after deep economy, with its deep financial markets, sophisticated risk recessions, wars, or other crises, it is not clear whether anemic or management tools, and steadily declining interest rates to negative growth can all be blamed on high debt. Also, the impact historically low levels that make debt servicing as easy as it on growth depends on each economy. Emerging economies, has ever been? as well as countries that lack inclusive political and economic • Finally, how should investor portfolios be positioned in this institutions, are considered to be more adversely affected13. environment? Some deleveraging has been taking place since the crisis14, but it has been uneven across economies. Global Debt: Challenges and Opportunities | 4

Exhibit 4: Debt would rise perpetually in an economy Exhibit 5: Corporate bond issuance and S&P 500 relying solely on credit to sustain demand (depicted by buybacks have been increasing at similar paces dotted lines). If credit stopped flowing, demand would S&P 500 buybacks Bond issuance stagnate while the stock of debt stabilized, and it would contract if the credit flow reversed in order to deleverage. 3,000

2,500 PERIOD 1 PERIOD 2 PERIOD 3 PERIOD 4 2,000

+$1 -$1 1,500 USD Billions $0 CREDIT 1,000 -$1 500

0 +$1 2009 2010 2011 2012 2013 2014 2015

$0 DEBT STOCK Source: Strategas Research Partners and Chief Investment Office. Past performance is no guarantee of future results. Please see Appendix for index definition.

+$1 -$1 This trend may be indicative of a shift in the capital structure $0 DEMAND of firms toward less equity and more debt. The rebalancing to -$1 a lower cost-of-capital equilibrium in the current environment $0 borrowed. $1 borrowed. $0 borrowed. $1 paid back. of low or negative interest rates could be reflective of capital $0 debt build-up. $1 debt build-up. $1 debt build-up Debt stock falls $1. $0 demand. Demand expands unchanged. Demand contracts market efficiency and deepening. Share buybacks have proven by $1. $0 demand. by $1. rewarding to investors for several years, as Exhibit 6 shows.

Source: Chief Investment Office. Exhibit 6: Share buybacks have proven rewarding to investors for several years Business debt A study by the McKinsey Global Institute found that business S&P 500 Buyback Index S&P 500 Equal Weight Total Return* Total Return Index* deleveraging has been taking place in the developed world, with 400 debt-to-GDP ratios falling in the U.S. and a few other countries 350 and stabilizing elsewhere. The financial sector deleveraged 300 significantly in the years since the financial crisis, mainly as a 250 200 result of new regulations and stricter capital standards. While this 150 has made the sector more stable, it has also restricted availability 100 of credit to households and businesses15. The decline in corporate 50 1/2/2009 1/2/2010 1/2/2011 1/2/2012 1/2/2013 1/2/2014 1/2/2015 1/2/2016 bank lending has been somewhat offset by an increase in non- bank credit. A significant part of the latter includes the issuance Source: Bloomberg and Chief Investment Office. *Bloomberg index tickers: SPBUYUT, SPXEWTR. Please see Appendix for index definitions. of corporate bonds by larger companies, while share buybacks Past performance is no guarantee of future results. have been increasing at a pace similar to that of bond issuance in the U.S., according to Strategas Research Partners (see Exhibit 5), Other forms of non-bank credit include transparent forms of as well as in Japan and, to a lesser degree, Europe. securitization, as well as lending by hedge funds, pension funds, insurers, leasing and government programs, and the still small but rapidly growing peer-to-peer lending platforms. All of these can be Global Debt: Challenges and Opportunities | 5

vital sources of funding to smaller entities. Non-bank lending can has payments linked to the local house price index, and can serve an important role in promoting economic growth by providing adjust downward if house prices fall, while the amortization needed credit to the private sector16. schedule remains the same, resulting in a principal reduction to the homeowner. In turn, the lender would receive a portion of the Among proposals to better manage the growth in corporate debt capital gains upon sale of the house, if it appreciates22. is reducing or phasing out the deductibility of interest, coupled with reductions in marginal rates and other potential adjustments17. Bankruptcy and foreclosure can be lengthy and costly. In their place, several economists have advocated debt reduction through Household debt renegotiation, mortgage principal reduction, refinancing and, for Household spending and residential investment were credit- certain types of household loans, such as unsecured credit card financed to a large degree before the global financial crisis. debt, even forgiveness. Some point to historical evidence Since it broke out, households in the U.S. and several other of voluntary debt forgiveness in ancient Mesopotamia, advanced economies, such as Ireland, the U.K. and Spain, have and the United States in the early 19th century and during 18 been deleveraging . This is being done mainly through a rise the Great Depression23, and during biblical Jubilee years in precautionary savings and paying down debt to weather the (Leviticus 25). Broader adoption of non-recourse loans outside economic uncertainty. Delinquency write-offs and reduced credit the United States, combined with sound macro-prudential availability for new mortgage, auto and credit card loans also policies to prevent lending excesses, could speed up bad-debt played a significant role. In the emerging world, household debt resolution, with lesser adverse impact to the economy24. has been rising at different rates across countries, with China leading the trend, and is still at modest levels relative to income. Sovereign debt Countries have typically borrowed to finance several Real estate prices and household debt are highly correlated endeavors, including wars, investments in productivity- in several economies because the larger the mortgages for enhancing projects (e.g., infrastructure), and fiscal policies which home buyers are approved, the more they bid up house that stimulate economic growth through increased consumer prices. The link between mortgage debt and real estate prices spending. The level of public debt a country can sustain is makes household debt hard to reduce19. determined by its ability to continue servicing it through Central banks can be effective in managing mortgage credit organic economic growth or access to more borrowing. When trends and house prices through the appropriate mix of the stock of debt rises above a “tipping point” that impairs interest rate and macro-prudential policies20. These have been either or both of these sources, it ought to be reduced25. In the found to meaningfully reduce real household credit growth and absence of default or restructuring, debt build-ups can take a house price growth21. long time, or be impossible, to eliminate. Deleveraging after Among proposals to improve the stability of the home mortgage major financial crises typically takes longer and is more painful market is requiring or encouraging homeowners to purchase than after wars and, in some cases, debt remains high or gets insurance that makes mortgage payments after a job loss even higher. In the current cycle, deleveraging started slowly in or other adverse event. Tax incentives, such as deductibility the world’s major economies. of mortgage interest, could be reconsidered, since they lead Debt-reduction episodes have involved “belt-tightening,” households to take larger mortgages to maximize tax benefits. elevated inflation, financial repression, stronger economic They mainly benefit high-income households, and contributed to growth, and default or restructuring26. Fiscal consolidation at the unsustainable home price increases that caused the systemic the right pace and speed27 has played a key role in reversing instability during the 2008 global financial crisis. public debt build-ups, as Sweden and Finland demonstrated Innovative proposals, mainly by think tanks and academics, in the 1990s. After World War II, the U.S. and other countries include flexible mortgage contracts. By automatically adjusting reduced their public debt through higher inflation and financial payments, a “continuous workout mortgage” would protect repression28. France restructured hers after the Revolution and homeowners during recessions, unemployment, or other triggers Napoleonic Wars, and so did several European countries after impairing their ability to pay. A “shared responsibility mortgage” World War II. A number of emerging countries restructured Global Debt: Challenges and Opportunities | 6

theirs under the Brady Plan in the 1980-90s29. Post-Brady Plan, and understood better, especially with regard to their potential large restructurings include those in Argentina (2005), Greece for giving rise to moral hazard issues and risk of loss of central (2012), and Russia (1998)30. Britain, which twice in the last bank independence32. three centuries ran enormous public debts, associated mostly In the conventional sense, deleveraging without default, with wartime spending, deleveraged through primary budget with as fast a return to robust economic growth as possible, surpluses and robust economic growth (see Exhibit 7)31. is considered a successful outcome after a debt build-up. But, can economic growth be achieved during deleveraging? Exhibit 7: Public debt in Britain, 1700–2010 Economic Growth During Deleveraging 250 Achieving economic growth while deleveraging is hard but

200 possible, as Sweden and Finland demonstrated during the 1990s33. The two Nordic countries quickly recapitalized or 150 nationalized banks and set up institutions to dispose of bad loans, thus ensuring stability in the banking system. They cut 100 government spending at a pace that put debt under control without threatening long-term fiscal sustainability. They adopted % of National Income 50 meaningful structural reforms, including joining the European Union, which attracted foreign investment and helped exports, 0 1700 1750 1810 1850 1880 1910 1920 1950 1970 1990 2010 and enacted measures to boost productivity and growth in sectors such as retail and banking. Their exports rose by almost Source: T. Piketty: Capital in the 21st Century (see piketty.pse.ens.fr/capital21c). 10% annually in mid-decade, through a focus on a few export- oriented sectors and companies, and currency depreciations. Central banks around the world recognize the state of global debt Private investment revived and helped growth during the late as the new normal and have been broadening their mandates, stages of deleveraging. Finally, a stabilized housing market using new tools or resurrecting old ones to manage it effectively. and a rebound in construction helped the economy return to Financial repression, which helps governments ease the burden of normal conditions. Ireland’s recent success could be even more debt service through artificially low or even negative interest rates, remarkable: The country grew while deleveraging in the arguably borrowing from captive domestic sources (e.g., pension funds), more challenging, post-2008, economic environment and within cross-border capital controls, and tighter connection with banks, the confines of the European common currency. is making a comeback. When certain issuers’ heavy leverage is perceived as a systemic risk, central banks bail them out, as the The Conference Board proposed two policy levers as key to Federal Reserve did with several financial firms in 2007­ – 09. The a successful sovereign debt resolution without the need for European Central Bank and other institutions did the same with harsh austerity measures or stimulus programs that could add the eurozone’s periphery earlier in this decade, but the required to debt accumulation: productivity-driven GDP growth, and austerity exceeded the natural speed limit of fiscal consolidation constant per-capita government spending over medium-term and caused a severe economic contraction in certain countries. planning horizons. The fiscal surplus that would emerge could eventually reduce debt-to-GDP ratios34. Some have proposed that central banks engage in fiscal expansion by “printing money” — i.e., growing the monetary Economies can grow while deleveraging even with weak or no base or the total bank reserves and currency in circulation, credit growth, especially after banking crises and credit booms35. and investing it in infrastructure and other projects, or even Examples include the U.S. and Japan after the Great Depression dropping it to the population from “helicopters.” This approach and the U.K. in the 25 years following World War II. The U.S. would allow governments to continue spending without the experienced real GDP growth even though debt-to-income ratios need to accumulate more debt or raise new taxes. These were declining since the trough of the 2008 global financial crisis. practices, termed “monetary finance,” are being discussed In these and similar cases, the flow of credit remained positive, more and more. We feel that they need to be researched more even though the stock of debt was shrinking36. Global Debt: Challenges and Opportunities | 7

The role of default or restructuring in achieving Defaults, restructurings and market access successful outcomes Potential investors duly avoid the country’s public and private Sometimes, defaults and/or restructurings are inevitable, and assets for as long as any uncertainty remains that its troubles decisive action is generally the fastest way to deleverage and are over and there is almost no chance that a relapse will force return to growth37. Such episodes can be painful in the short run, them to assume any future losses. The caution is not limited forcing investors to take significant losses38. However, if executed to the public sector; it also encompasses the private sector. In in a careful and orderly manner, a restructuring can lead to fact, businesses in a defaulted, or otherwise financially troubled, sustainable long-term economic growth and a rise in asset prices, domicile are hurt more than the sovereign borrower, as their as historical evidence shows (see Exhibit 8). access to capital markets is severely restricted for as long as their home country’s financial situation remains unresolved41. Liquidity crunch versus solvency crisis A country can get heavily in debt for various reasons, and may not Investors will eagerly come only after all past losses have been have cause for concern. If it temporarily lacks, but expects to soon assigned. The safest way to achieve this is to take decisive have, the funds it needs for debt service, interim financing or at action to make public debt sustainable in the present. For most an extension of bond maturities can relieve the liquidity example, after the Brady Plan restructurings finally eliminated crunch. But if it both lacks liquidity and, given meager economic debt overhang in countries that had defaulted on commercial growth prospects, is not expected to be able to generate the bank loans in the 1980s, investors responded with a wave of income and fiscal surpluses needed to service its debt in the capital inflows to the relieved sovereigns, whose stock markets future, more decisive actions may be in order. Such actions rallied, economies benefitted from new investment and grew, include debt rescheduling (i.e., lengthening of maturities, possibly and capital market access was restored42. involving lower interest rates), debt reduction (i.e., a “haircut” in There is no sovereign bankruptcy court, and “gunboat diplomacy” the nominal value of debt) or, in the most undesirable case, a seems to be a relic of history. In their absence, what motivates disorderly default, whereby the country fails to make principal or countries to pay their debts? The common assumption that they interest payments on schedule. try to preserve their reputation as responsible borrowers is not A liquidity crunch should not be mistaken for a solvency crisis, supported by recent evidence43. In reality, governments typically or vice versa, and the key to resolution is an accurate case-by- tend to avoid default to contain political costs to incumbent case diagnosis39. governments44. A timely and orderly default and restructuring It is hard to diagnose certain cases ex ante. What initially may allows a country to receive rating upgrades and borrow again 45 seem just a liquidity crunch could be more serious and deteriorate soon . This is because global interest rates and liquidity into a solvency crisis, and may require that it be treated as such. conditions, decreased investor risk aversion during lending booms, In early 2010, few had foreseen the severe downward spiral that yield-seeking attitudes, and portfolio manager strategies and the Greek economy would enter as austerity measures were incentives are stronger determinants of sovereign capital flows 46 implemented. These measures significantly impaired the country’s and market access . ability to service its debt, and will continue doing so in the absence Yet, while default frequency does not seem to affect future 40 of an effective resolution . In a similar manner, as the 2008 global market access, it appears that the magnitude of the most financial crisis was unfolding, the U.S. financial institutions that recent default likely plays a role: Restructurings involving failed, or nearly did so and had to be rescued, were those that deeper “haircuts” are associated with significantly higher had not been aggressive enough in raising adequate capital to subsequent bond yield spreads and longer periods of capital remain solvent through the worst-case scenario that materialized. market exclusion47. As a special case, Argentina, whose long Even if a solvency crisis is diagnosed, a variety of factors involving absence from global markets was caused by holdout creditor multiple stakeholders and political constituents may impede litigation, reached a settlement with those creditors in April administration of the necessary remedies. 2016 and is again reissuing external sovereign debt. In line with research evidence, it is being well received by global investors, and it will be interesting to analyze its performance in the medium to long term48. Global Debt: Challenges and Opportunities | 8

Exhibit 8: After sovereign debt restructurings, equity markets in seven of ten countries strongly outperformed global equities, and three underperformed them

We identified ten sovereign debt restructuring cases over the last three decades and examined the performance of a hypothetical investment in each country’s equity market made at the end of the month during which the restructuring was completed and held for two yearsa. By the end of the two-year period, on the basis of price returns in U.S. dollars, seven of the ten investments would have outperformed the MSCI All Country World Index (ACWI) by an average of 92%, and three would have underperformed it by an average of 38%. On the aggregate, the ten investments would have outperformed the ACWI by an average of 56%. Eight of the ten would also have had positive absolute price returns. For comparison, we included Argentina’s 2016 sovereign debt resolution.

Average of Seven Outperformers Average of Three Underperformers Average of all 10 Restructuring Cases Argentina (2016)

125% 92% 100%

ACWI (USD, % ) 75% 56% 50%

25% 15% 0%

-25% -38%

Relative Daily Price Return vs. MSC I -50% 0 3 6 9 12 15 18 21 24 Months Since Sovereign Debt Restructuring

Sovereign Debt Restructuring Caseb Equity Market Price Returns (2-year post-restructuring, end-of-period, USD, %) Country Completion Date Performance Period Absolutec Relative to MSCI ACWId Mexico March 1990 03/31/1990 – 03/31/1992 295.4% 292.6% Argentina January 2005 01/31/2005 – 01/31/2007 161.5% 128.1% Philippines April 1992 04/30/1992 – 04/30/1994 134.5% 109.3% Average: Russia November 1998 11/30/1998 – 11/30/2000 83.4% 74.1% 92% Greece March 2012 03/31/2012 – 03/31/2014 50.6% 27.3% Average: Poland October 1994 10/31/1994 – 10/31/1996 28.3% 8.5% 56% Brazil April 1994 04/30/1994 – 04/30/1996 28.4% 3.8% Argentina April 1993 04/30/1993 – 04/30/1995 8.9% -9.9% Average: Jordan December 1993 12/31/1993 – 12/31/1995 -3.9% -24.2% -38% Peru March 1997 03/31/1997 – 03/31/1999 -39.8% -79.3% Argentinae April 2016 4/30/2016 – 4/30/2017 (1 year) 27.7% 14.9%

Source: Bloomberg and Chief Investment Office. The graph shows average daily price returns of the MSCI equity market index over the seven countries that outperformed and over the three that underperformed the MSCI ACWI during the two-year performance period after a restructuring. It also gives the average daily price return of the MSCI equity market index for Argentina the year after the country’s sovereign debt resolution. Past performance is no guarantee of future results. Please see Appendix for index definitions. a. Each country’s MSCI equity market index was used to represent its equity market. As a requirement for inclusion in our sample, MSCI must have been offering a continuous daily series of the country’s equity market index at the time of, and for the two years following, the end of the month during which each restructuring was reportedly completed. We chose a single index provider — rather than a mix of index providers — for the benefit of a uniform equity market index calculation methodology. As a result, several restructuring cases for which MSCI country equity market index data were not available for the aforementioned time period were excluded from the sample. b. All cases except Argentina 2005, Greece 2012, and Russia 1998 were Brady Plan restructurings of distressed sovereign debt. c. Based on the MSCI equity market index for each country during the two-year performance period after a restructuring, as shown. d. Calculated as the difference between absolute price returns of each MSCI equity market index and the price return of the MSCI ACWI in the same performance period. e. Argentina did not restructure its sovereign debt in April 2016, but reached a settlement with its holdout creditors from prior restructurings. That resolution enabled it to access the global capital markets again. Global Debt: Challenges and Opportunities | 9

The outlook for debt and growth Securities with such characteristics have started to be used Government debt crises might not have happened, or their by governments, or have been envisioned and recommended effects might have been significantly attenuated, had better as viable means of public funding. They present new risk sharing been practiced. Today’s sovereign lenders and opportunities to attenuate or prevent insolvencies: borrowers may draw some lessons from the Spanish Empire • GDP-linked warrants, whose payments are triggered when a th of the 16 century (see feature box). country’s economy meets pre-established growth targets, have Today, sovereign borrowers could make their debt service been offered as part of bond exchange packages in sovereign contingent on their GDP growth rates. After World War II, debt restructurings, most notably by Argentina in 2005 and John Maynard Keynes negotiated such a clause in a U.S. loan Greece in 2012. Economic statistics as reported by the issuing to the U.K.50 Countries could issue securities with equity-like country will have to be used for valuing these securities, and 52 features — e.g., bonds with coupons linked to the country’s hence their reliability must be a factor in investment decisions . economic growth51 and, hence, the country’s ability to service • Equity-like shares representing claims on a country’s GDP its debt. Government debt crises could be avoided through use have been proposed53 or already issued54. Such securities have of such securities. During an expansion, higher coupons would even been envisioned for the United States55. allow investors to share in the country’s prosperity. Conversely, • Commodity-linked bonds can sidestep GDP-related issues a contraction would not necessarily cause a debt crisis, since because they derive their value from the price of a commodity lower coupons could protect the country’s ability to service that a country produces and sells. In certain countries, a single its debt without raising tax rates or cutting spending, both of commodity accounts for a major share of GDP, such as copper which could deepen the contraction. Lower coupons could also in Chile, crude oil in Russia, Nigeria, Saudi Arabia and Venezuela, prevent a default or restructuring. and potentially natural gas in Cyprus in the future56.

The King of Sovereign Default King Philip II (1527-1598) of Spain ruled for most of the second half of the 16th century over an empire spread across three continents, one of the largest and richest in history. His wealth and power did not prevent him from defaulting four times. Yet, his serial defaults on sovereign debt comparable to today’s levels relative to the size of his empire’s economy did not cause systemic problems or severe economic downturns.

King Philip’s bankers pooled risk through loan syndication, whereby many investors of small and medium-sized wealth participated in the lending. They also shared risk and reward with him by agreeing to debt service contingent on the evolving financial conditions of the crown. King Philip issued mostly short-term debt under flexible clauses— for example, making debt service contingent on the size and arrival date of the annual silver deliveries from the Americas. Lending to him was profitable, with average annual returns of 15% or more, even after factoring in his insolvencies49. Image Source: iStock. Global Debt: Challenges and Opportunities | 10

Proposed automatic restructuring mechanisms would trigger delinquencies and foreclosures if and when the economic lower coupons, longer maturities, and conversion to equity-like cycle turns. instruments upon a pre-defined triggering event — for example, The benefits of deleveraging are most directly observable in when the debt-to-GDP ratio reaches a certain threshold57. cases where there has been quick and decisive resolution: We ought to point out a critical difference between securities Equity markets of countries that have just restructured their that investors will want to buy based on their superior risk- debt are very likely to rally in the short to medium term return characteristics, and securities that investors are forced (see Exhibit 8 again). There are multiple reasons for this. to receive as part of a sovereign debt restructuring. When Countries whose economies were stifled under heavy debt designing, structuring and offering such securities, political, service are all of a sudden relieved from the burden and have social, economic, moral hazard, financial engineering, and a at their disposal resources to pursue pro-growth activities. multitude of other public funding issues ought to be better A portion of their resources comes from renewed access to understood and addressed, and incentives for all stakeholders debt markets, and this is also true for their corporations. The must be aligned as best as possible58. credit cycle affects equity returns, and attempts to time it may be informed by historical experience and other factors. Portfolio implications Equity market underperformance has followed periods of While credit fuels economic growth, there comes a point rapid credit build-ups, especially in emerging economies, so where rising global debt leads to muted future economic this is a valuable metric to track and use as a guide for asset growth. This implies low expected returns and higher volatility allocations in portfolios. for assets. Investors would be well advised to practice sound risk management and take the long-term view with portfolio For fixed income investors, sovereign spreads of countries and positioning, since research evidence and experience indicate corporate spreads of firms that have low leverage are viewed these trends play out over very long horizons. That said, as relatively safer and therefore enjoy tight spreads. Spreads there can be cyclical moves within the secular trends of of countries that recently deleveraged are narrowing, those debt and growth that may present tactical positioning of countries still deleveraging are likely to narrow, and so are opportunities — for example, in response to central bank those of corporates. This points to strong balance sheets and actions or sovereign debt resolutions. high quality in both sectors.

Equities or traded debt of entities with low leverage or in Debt accumulation, deleveraging and credit flow trends impact the process of deleveraging are attractive. This holds true demand for commodities as a function of economic growth, for all types of debt — government, business and household. and by commodity type depending on each economy’s specific For governments, a lighter debt load to service means more characteristics. financial resources available to productively deploy elsewhere. Real estate is driven to a large extent by the credit cycle, Less indebted businesses have more funds available to invest interest rates, and macro-prudential policies. Prices are positively for growth — through capital expenditures, acquisitions, or correlated with credit flows and inversely correlated with interest research and development — or return money to shareholders rates and tightening lending standards. Investors with a long- through higher dividends or share buybacks. Finally, households term horizon could take advantage of opportunities presenting with stronger balance sheets can fuel economic growth and themselves as the credit cycle unfolds, together with asset values corporate profits through more robust consumption, and fewer and prices in each geography.

Acknowledgments: We are grateful to Susan Lund at the McKinsey Global Institute and Roy Butta at Strategas Research Partners for making analyses available for our use. We would also like to thank Ricardo Penfold of Goldman Sachs Asset Management, Ralf Preusser and Athanasios Vamvakidis of BofA Merrill Lynch Global Research, and Karin Kimbrough, Niladri Mukherjee and Rodrigo Serrano of the Chief Investment Office for providing thoughtful feedback. Global Debt: Challenges and Opportunities | 11

Endnotes 19 House prices are also driven by factors such as land scarcity, regulatory rules and urbanization. Monitoring mortgage debt trends, especially in major urban 1 See, for example, Lo and Rogoff, 2015. The authors include debt overhang among centers with high real estate prices, is therefore one of the most important the possible explanations for sluggish growth. Others they include are a secular policy matters (McKinsey Global Institute, 2015). deficiency in aggregate demand, slowing innovation, adverse demographics, lingering policy uncertainty, post-crisis political fractionalization, insufficient fiscal 20 Macro-prudential policies could help control total leverage in an economy by stimulus, and excessive financial regulation. We argue that it is too soon to tell, limiting credit growth. They may include limitations on loan-to-value (LTV) and and add underinvestment in productive resources, lower labor productivity, and debt-service-to-income (DSTI) ratios, restrictions on interest-only loans and (in the case of the U.S.) weak demand from abroad as potential contributing other types of mortgages, bank capital requirements, and other underwriting factors to the slower growth. rules and banking regulations (see Kuttner and Shim, 2013). 2 McKinsey Global Institute analysis, March 2016. 21 See, for example, Jarocinski and Smets, 2008, and Iacoviello and Neri, 2010. 3 The term “peacetime” is used loosely, since several countries have been involved 22 Shiller et al., 2011, discuss continuous workout mortgages, and Mian and in conflicts during the global debt build-up. For example, the wars in Afghanistan Sufi, 2014, shared responsibility mortgages. Any misalignment of interests and Iraq may have contributed significantly to the U.S. stock of debt. between homeowners and lenders in contract design, and the fact that their 4 See Reinhart et al., 2012. risk-sharing features may disqualify them as debt instruments, making their 5 See Reinhart and Rogoff, 2011. mortgage interest ineligible for tax deduction, could prevent widespread 6 See McKinsey Global Institute, 2015. adoption of shared responsibility mortgages. 7 See Buiter and Rahbari, 2012. 23 Mian and Sufi, 2014. “Debt Forgiveness in History,” houseofdebt.org, May 18 8 See Reinhart and Rogoff, 2009. (accessed on November 27, 2015). 9 Results are based on analyzing extensive datasets assembled by researchers (e.g., 24 Non-recourse loans, broadly used in the United States, give creditors rights Barro and Ursua, 2008, Devries et al., 2011, Reinhart et al., 2012, Jorda et al., 2013, only to a defaulted loan’s collateral and do not allow claims on the borrower’s 2015, McKinsey Global Institute, 2012, Medas et al., 2013, and Schularick and future income or other assets. This helps write off debts quickly and prevents Taylor, 2012). severe disruptions to households’ economic life. On the other hand, recourse 10 Mian and Sufi, 2014, present compelling evidence of a link between rising loans, common in many parts of the world, allow creditors to seize assets household debt levels and a subsequent spending slowdown affecting all other than a defaulted loan’s collateral, or lay claim on the borrower’s future households, and not only the most leveraged ones, even though the latter income. As borrowers slash spending to commit more funds to debt service in experienced larger cuts in spending. See also Dynan et al., 2012. Interestingly, their efforts to avoid default at all costs, recessions can deepen and lengthen balance-sheet adjustments were limited even for households most affected (see, for example, McKinsey Global Institute, 2015). by the crisis, as reported by Cooper, 2013. 25 See, for example, Ostry et al. 2015. 11 See Chen et al, 2015. 26 For a detailed discussion, see McKinsey Global Institute, 2012. 12 Reinhart and Rogoff, 2010, found that after government debt has exceeded 27 See Alesina and Ardagna, 2010. 90% of GDP, median growth rates have been one pp lower, and mean rates 28 even lower. Kumar and Woo, 2010, found that, for every 10 pp that the starting A good analysis can be found in Reinhart and Sbrancia, 2011. debt-to-GDP ratio is higher, the subsequent real per-capita GDP growth can 29 Most Brady Plan deals took place in Latin American countries — i.e., Argentina, be 0.2 pp lower, with emerging economies more severely impacted than Bolivia, Brazil, Costa Rica, Dominican Republic, Ecuador, Mexico, Panama, Peru, advanced economies. Uruguay and Venezuela. Others took place in Bulgaria, Ivory Coast, Jordan, 13 See Acemoglu, D., and J. Robinson, 2012. Nigeria, the Philippines, Poland and Vietnam. Morocco was going to, but eventually did not, participate (see Das et al. 2012). 14 See McKinsey Global Institute, 2015. 30 For a complete list of sovereign debt restructurings from 1950 to 2010, see 15 The riskiest forms of shadow banking, including repos, CDOs, CDSs, and off- Das et al., 2012. balance-sheet securitization vehicles, have been in a material decline since 31 2007. The unwinding of a global financial system that had become excessively For a discussion on France’s restructuring of Revolutionary and Napoleonic complex, opaque and ultimately unstable and vulnerable to shocks is a debts, and Britain’s debt reduction through fiscal prudence and economic welcome development (see McKinsey Global Institute, 2015). growth, see Piketty, 2014. 32 16 The growth of non-bank lending ought to be carefully managed, as hidden For details, see Turner, 2015. Bernanke, 2016, emphasizes implementation risks and leverage may be building and could cause systemic instability issues. A related discussion can be found in Montier, 2016. until detected. A sound regulatory framework, with appropriate reporting 33 See McKinsey Global Institute, 2010, 2012. requirements and monitoring mechanisms, can reduce its role as a source of 34 See Sexauer and van Ark, 2010. systemic risk. 35 Resources on creditless recoveries include Abiad et al., 2011, Bijsterbosch 17 Tax code changes are always challenging, and this could prove a thorny policy and Dahlhaus. 2011, Takats and Upper, 2013, and Constancio 2014. Some matter, but would have the additional benefit of better capital structure and researchers call recoveries without a rebound in credit “Phoenix Miracles”; see, allocation decisions in firms, by increasing the relative attractiveness of equity for example, Calvo et al., 2006, and Biggs et al., 2010. versus debt, and labor versus capital goods investments. 36 See Dalio, 2015. 18 Notable exceptions include the Netherlands, Denmark and Norway, where 37 household debt as a percentage of household income has been rapidly Levy-Yeyati and Panizza, 2011, have demonstrated that the costs of delaying a increasing. needed default, in terms of output losses, are more significant than the default itself, which often marks the end of economic malaise and the beginning of Global Debt: Challenges and Opportunities | 12

strong economic recovery for the defaulting sovereign. The timing of default or 48 Despite its debt restructurings of 2005 and 2010, after its 2001 default, significant debt restructuring hence is a crucial policy matter, and policymakers’ Argentina has not issued debt internationally given pending litigation with efforts to postpone a default that has been widely anticipated and priced in by the holdout creditors in U.S. courts. Not issuing bonds abroad is a high cost to market may be misguided (see Zettelmeyer et al., 2013). This is demonstrated pay for default, but not always one imposed by unwilling foreign creditors; by the cases of Greece and Puerto Rico, both of which are constrained by their in Argentina’s case, it is being forced by a minority of holdout bondholders. inability to print money, and by the political complexities within the union that The distinction may seem subtle but, given increased concern now with each is part of. the serious threat holdouts pose to sovereign debt restructuring deals, it is 38 The “haircuts” in some recent restructurings ranged from 13% (Uruguay in warranted, and significant effort is being undertaken on appropriate Collective 2003) to 73% (Argentina in 2005). See Sturzenegger and Zettelmeyer, 2008. Action Clauses (CAC) to include in future bond offerings to mitigate this threat and facilitate a needed restructuring (see Gulati and Buchheit, 2011). 39 See, for example, International Monetary Fund, 2013-2015, and Das et 49 al., 2012. Voth, 2014, defines sovereign debt sustainability as a country’s See Drelichman and Voth, 2014. expected value of future primary fiscal surpluses. Net debt, which excludes 50 See Griffith-Joneset al., 2011. borrowing by central banks and other government agencies, enables a better 51 See Shiller, 1993, and Borensztein et al., 2004. assessment of sustainability. A common error when treating a solvency crisis 52 as a liquidity problem is trying to make an entity’s debt sustainable at a future See Shiller, 2004, Griffith-Jones and Sharma, 2006, and Griffith-Jones and date. The assumption is that the entity’s economy will have recovered enough Hertova, 2012. Although attempts have been made to rigorously value these over some period to make its debt sustainable and serviceable then, because it instruments (e.g., Miyajima, 2006, Ruban et al., 2008) and market experience will be a certain lower percentage of GDP than it is at present. This assumption with the Argentine warrants has been quite rewarding for investors, a number is particularly misguided in that it supposes that we know exactly what will of issues ought to be better understood before GDP-linked warrants become happen on the path to debt sustainability five to ten years from the present. widely accepted. An entity issuing GDP-linked securities signals its willingness to share its future expected economic prosperity with investors, to a certain 40 In three articles published after the Greek sovereign debt restructuring degree, even if the securities have embedded redemption features or are in 2012, Barrionuevo et al. argue that Greece’s solvency crisis was initially somehow callable. Moral hazard around an entity’s GDP and inflation figures misdiagnosed as a liquidity crisis, and treated through temporary financing must be accounted for. These may be reported in a manner that lowers warrant that proved ineffective and became a permanent burden to the country and a payments resulting in losses to investors; or they could be manipulated potential future write-off to its rescuers. to increase payments in certain years to create, or revive, interest in the 41 See Arteta and Hale, 2008, and Das et al., 2010. securities, to the detriment of the entity’s taxpayers. In each case, economic 42 See Arslanalp and Henry, 2005. Success proved unsustainable for some Brady statistics can be revised to their true values at a later time. Therefore, issuing Plan countries (see Chuhan and Sturzenegger, 2005). Arslanalp and Henry GDP-linked warrants must be accompanied by robust procedures, controls and 2006 contend that for poorer sovereigns, debt forgiveness entails a superficial audits around the preparation and reporting of economic statistics, and their boost if unaccompanied by policies that promote significant reforms to weak structure should allow coupon payment adjustments upon revisions in the local economic institutions partially responsible for these countries’ malaise. figures used to calculate them. Adverse selection could also play a role. Entities that believe or know themselves to have weak economic prospects may prefer 43 Eaton and Gersovitz, 1981 present analysis on countries’ credit access that to issue GDP-linked warrants structured in a manner that makes them look supports the common assumption that countries avoid default to steer clear attractive but secures a future stream of low coupon payments; conversely, from reputational damage that could severely restrict new credit or make those with strong confidence in their prospects would choose to structure it more costly. Yet, beyond disputable centuries-old experience (see Tomz, GDP-linked securities so their future payment stream resembles that of fixed- 2007), contemporary evidence for this assumption is weak (see Panizza et al., coupon debt. Hence, moral hazard around an entity’s GDP, and a few other 2009, and Datz, 2014). Sovereign debtors’ market access was not determined issues, ought to be better understood and accounted for before GDP-linked by their default records during the debt crises of the 1930s and the 1950s warrants become more broadly accepted. credit boom (see Lindert and Morton, 1989). 53 Griffith-Jones and Shiller, 2006, contend that “GDP-linked bonds” would create 44 Borensztein and Panizza, 2009 contend that the main motivation to avoid markets for countries’ economies. In contrast, they contend, stocks are claims default is limiting political costs rather than preserving reputation, preventing on net corporate profits that can constitute as little as 10% of GDP. Similar international trade exclusion, or limiting costs to the domestic economy market signaling and information issues (moral hazard and adverse selection) through the financial system. could exist for these securities as with GDP-linked warrants. If properly 45 Gelos et al., 2011, did not find market access to be a function of a country’s constructed, portfolios of such securities from several countries could offer frequency of defaults, and, when resolved quickly, past defaults did not investors true global diversification. prevent sovereigns from obtaining fresh funds in financial markets. 54 The “New Singapore Shares,” issued in 2001, paid a minimum dividend of 46 Time horizons tend to be longer-term for investors and shorter-term for 3%, plus the country’s real annual GDP growth rate, with no claw-back in years managers, since performance must be reported monthly, quarterly and when it would be negative. annually. Compensation through performance fees also shortens the 55 Kamstra and Shiller, 2009, proposed that the U.S. Treasury issue a new security investment time horizons of these managers. Therefore, when liquidity is with a coupon tied to the current GDP of the U.S. Because the coupon of plentiful, industry incentives point toward assets and economies that provide this security might, per their proposal, equal one-trillionth of the country’s relatively higher yields and growth prospects, regardless of a (recent) history GDP, they suggest the security be called a “trill.” Shiller, 2013, demonstrates of default (see Datz, 2009). through examples how this might work. A U.S. trill held by an investor during 47 See Cruces and Trebesch, 2011. Results perhaps reflect market caution toward 2011 would have paid the holder a coupon of USD 14.50, since the U.S. GDP a sovereign based on the degree to which it got itself in trouble in the past. was about USD 14.5 trillion in 2011. A Canadian trill would have paid CAD 1.81, Global Debt: Challenges and Opportunities | 13

with the country’s GDP at CAD 1.81 trillion. The security’s coupon would be by companies, which are issued as bonds and can later convert to equity. If determined in a similar manner for each trill-issuing country. And it could the borrowing decision-making process is deliberately manipulated and a effectively protect its debt servicing ability: A Greek trill would have paid a 2008 sovereign’s debt pushed past the restructuring trigger thresholds by a little coupon of EUR 0.23; the same trill’s coupon for 2013 would be estimated at more borrowing, moral hazard can lead to adverse selection. EUR 0.18, about 21% lower, consistent with the magnitude of the country’s 58 Also, there is a difference between countercyclical mechanisms that ease economic contraction during the period. Had the country issued trills several the burden of debt repayment in economic downturns as opposed to a more years ago, it might not have needed the bailouts, debt restructuring, and specific group of assets based on automatic/mechanical approaches such draconian programs of austerity. as those proposed in the case of CoCos. The latter could be subject to moral 56 GDP in each of these countries is highly correlated with that commodity’s hazard, which can undermine the integrity of the process. A better approach price, which is transparently set in global markets and can be assumed to is to pursue a careful case-by-case debt analysis and negotiation. The renewed be free of the risk of sovereign manipulation of GDP and inflation statistics. effort in official and legal circles to design a sovereign debt restructuring However, moral hazard could remain a problem, especially in the case of a mechanism in light of the increased threat of holdout creditors and Greece’s country that has the power to manipulate a commodity’s price because it lingering difficulties is welcome (see Committee on International Economic controls a sizable share of its global production and supply. Price manipulation Policy and Reform, 2013), but a universally effective construct may be hardly can be particularly severe when practiced by groups of countries producing feasible. Political, legal and financial strategies are adaptive. A framework that the same commodity, as in the case of crude oil. allows flexibility to accommodate the idiosyncrasies of each case and takes 57 See, for example, Steinhauser, 2013, and Mody, 2013. This would work in a advantage of innovative financial products may not be ideal, but is certainly manner similar to the case of contingent convertible bonds (CoCos) issued promising for all parties involved. Global Debt: Challenges and Opportunities | 14

Alphabetical list of references Cooper, D. 2013. “Changes in U.S. Household Balance Sheet Behavior after the Housing Bust and Great Recession: Evidence from Panel Data,” Federal Reserve Abiad, A., G. Dell’ Ariccia, and B. Li. 2011. “Creditless Recoveries,” IMF Working Bank of Boston Public Policy Discussion Paper No. 13-6, September 6. Paper 11/58, March. Cruces, J.J., and C. Trebesch. 2011. “Sovereign defaults: the price of haircuts,” Acemoglu, D., and J. Robinson. 2012. Why Nations Fail, Crown Business. CESINFO Working Paper 3604. Alesina, A. and S. Ardagna. 2010. “Large Changes in Fiscal Policy: Taxes versus Dalio, R. 2015. “How the Economic Machine Works: Leveragings and Spending,” Tax Policy and the Economy, Vol. 24, ed. Jeffrey R. Brown, National Deleveragings,” Bridgewater Associates. Bureau of Economic Research. Das, U.S., M.G. Papaioannou, and C. Trebesch. 2010. “Sovereign Default Risk and Arslanalp, S., and P.B. Henry. 2005. “Is Debt Relief Efficient?”Journal of Finance, Private Sector Access to Capital in Emerging Markets,” IMF Working Paper 10/10, 60(2):1017-1051. January. Arslanalp, S., and P.B. Henry. 2006. “Debt Relief,” Journal of Economic Perspectives, Das, U.S., M.G. Papaioannou, and C. Trebesch. 2012. “Sovereign Debt 20(1):207-220. Restructurings 1950-2010: Literature Survey, Data, and Stylized Facts,” IMF Arteta, C., and G. Hale. 2008. “Sovereign Debt Crises and Credit to the Private Working Paper 12/203, August. Sector,” Journal of International Economics, 74(1):53-69. Datz, G. 2009. “What life after default? Time horizons and the outcome of the Barrionuevo, J.M, N. Georgikopoulos, E.D. Hatzakis, and G. Nielsen. 2012. “Greece’s Argentine debt restructuring deal,” Review of International Political Economy, Ignorant Savior,” Financial Times Deutschland, June 25 (in German; original title: 16(3):456-484. Griechenlands ignorante Retter). Datz, G. 2014. “Sovereign Debt Default.” International Studies Encyclopedia, Barrionuevo, J.M, N. Georgikopoulos, E.D. Hatzakis, and G. Nielsen. 2012. “Putting International Studies Association Compendium Project. Greece Back on Track,” The Banker, July 16. Devries, P., J. Guajardo, D. Leigh, and A. Pescatori. 2011. “A New Action-based Barrionuevo, J.M, N. Georgikopoulos, E.D. Hatzakis, and G. Nielsen. 2012. “How to Dataset of Fiscal Consolidation,” IMF Working Paper 11/128, June. Pay Greece’s Creditors and Rebuild its Economy,” The New York Times/International Drelichman, M., and H.-J. Voth. 2014. “Lending to the borrower from hell: Debt, Herald Tribune, November 6. taxes, and default in the age of Philip II,” Princeton University Press, Princeton, NJ. Barro R., and J. Ursua. 2008. “Macroeconomic Crises since 1870,” NBER Working Dynan, K., A. Mian, and K.M. Pence. 2012. “Is a Household Debt Overhang Paper 13940, April. Holding Back Consumption?” Brookings Papers on Economic Activity. Brookings BCA. 2015. “The Impotence of Monetary Policy,” Research Report, September 2. Institution Press, 299–362, Spring. Bernanke, B. 2016. “What tools does the Fed have left? Part 3: Helicopter money,” Eaton, J., and M. Gersovitz. 1981. “Debt with Potential Repudiation: Theoretical Brookings Institution, April 11. (http://www.brookings.edu/blogs/ben-bernanke/ and Empirical Analysis,” Review of Economic Studies, 48(2):289-309. posts/2016/04/11-helicopter-money). Gelos, R.G., R. Sahay, and G. Sandleris. 2011. “Sovereign Borrowing by Developing Biggs, M., T. Mayer, and A. Pick. 2010. “Credit and Economic Recovery: Countries: What Determines Market Access?” Journal of International Economics, Demystifying Phoenix Miracles,” Working Paper, SSRN ID: 1595980. 83(2):243-254. Bijsterbosch, M., and T. Dahlhaus. 2011. “Determinants of credit-less recoveries,” Griffith-Jones, S., R. Akerlof, and M. Miller. 2011. “Growth bonds a win win for ECB Working Paper No. 1358, June. troubled Eurozone,” Financial Times, December 9. Borensztein, E., P. Mauro, M. Ottaviani, and S. Claessens. 2004. “The Case for GDP- Griffith-Jones, S. and D. Hertova. 2012. “Growth-Linked Securities.” In F. Bergstein indexed Bonds,” Economic Policy, 19(38):165-216. and R. Henning (eds.), Global Economics in Extraordinary Times: Essays in Borensztein, E., and U. Panizza. 2009. “The Costs of Sovereign Default,” IMF Staff Honor of John Williamson. Washington, DC: Peterson Institute for International Papers, 56(4):683-741. Economics. Buiter, W. and E. Rahbari. 2012. “Debt of Nations: Mr. Micawber’s Vindication: Griffith-Jones, S., and K. Sharma. 2006. “GDP-Indexed Bonds: Making it Happen,” Causes and Consequences of Excessive Debt,” Citi Global Perspectives and DESA Working Paper No. 21, April. Solutions, November 20. Griffith-Jones, S., and R. Shiller. 2006. “A bond that insures against instability,” Calvo, G.A., A. Izquierdo, and E. Talvi. 2006. “Sudden Stops and Phoenix Miracles in Financial Times, July 10. Emerging Markets,” American Economic Review, 96(2):405-410, 2006. Gulati, M., and L. Buccheit. 2011. “Drafting a Model Collective Action Clause for Chen, S., M. Kim, M. Otte, K. Wiseman, and A. Zdzienicka. 2015. “Private Sector Eurozone Sovereign Bonds,” Capital Markets Law Journal, 6:317-325. Deleveraging and Growth Following Busts,” IMF Working Paper 15/35, February. Iacoviello, M., and S. Neri. 2010. “Housing Market Spillovers: Evidence from an Chuhan, P., and F. Sturzenegger. 2005. “Default Episodes in the 1980s and Estimated DSGE Model,” American Economic Journal: Macroeconomics, 2(2): 125- 1990s: What Have We Learned?” in Managing Economic Volatility and Crises: 64, April. A Practitioner’s Guide, eds. Aizenman, J., and B. Pinto, 471-519. Cambridge, MA: International Monetary Fund. 2013. “Sovereign Debt Restructuring—Recent Cambridge University Press. Developments and Implications for the Fund’s Legal and Policy Framework,” April Committee on International Economic Policy and Reform. 2013. “Revisiting 26. Sovereign Bankruptcy,” Brookings Institution, October. International Monetary Fund. 2014. “The Fund’s Lending Framework and Constancio, V. 2014. “Growing out of the crisis – is fixing finance enough?” ECB Sovereign Debt — Preliminary Considerations,” May 22. speech given on April 10 (https://www.ecb.europa.eu/press/key/date/2014/html/ International Monetary Fund. 2015. “The Fund’s Lending Framework and sp140410_1.en.html). Sovereign Debt – Further Considerations,” April 29. Global Debt: Challenges and Opportunities | 15

Jarocinski, M. and F.R. Smets. 2008. “House Prices and the Stance of Monetary Piketty, T. 2014. Capital in the 21st Century, Harvard University Press. Policy,” Federal Reserve Bank of St. Louis Review, 90(4):339-65, July/August. Reinhart, C.M., V. Reinhart, and K.S. Rogoff. 2012. “Public debt overhangs: Jorda, O., M. Schularick, and A.M. Taylor. 2013. “When Credit Bites Back,” Journal of advanced economy episodes since 1800,” Journal of Economic Perspectives, Money, Credit and Banking, 45(2):3-28, December (supplement). 26(3):69-86, Summer. Jorda, O., M. Schularick, and A.M. Taylor. 2015. “Leveraged Bubbles,” NBER Working Reinhart, C.M., and K.S. Rogoff. 2009.This Time Is Different: Eight Centuries of Paper 21486, August. Financial Folly. Princeton University Press. Kamstra, M., and R.J. Shiller. 2009. “The case for Trills: giving the people and their Reinhart, C.M., and K.S. Rogoff. 2010. “Growth in a Time of Debt,”American pension funds a stake in the wealth of the nation,” Cowles Foundation Discussion Economic Review: Papers and Proceedings, 100:573-578, May. Paper No. 1717, Yale University (http://cowles.econ.yale.edu). Reinhart, C.M., and K.S. Rogoff. 2011. “A decade of debt,” NBER Working Paper Kumar, M.S., and J. Woo. 2010. “Public Debt and Growth,” IMF Working Paper 16827. 10/174. Reinhart, C.M., and M.B. Sbrancia. 2011. “The Liquidation of Government Debt,” Kuttner, K.N., and I. Shim. 2013. “Can Non-Interest Rate Policies Stabilize Housing NBER Working Paper 16893, March. Markets? Evidence from a Panel of 57 Economies,” NBER Working Paper 19723, Ruban, O.A., S-H. Poon, and K.N. Vonatsos. 2008. “GDP Linked Bonds: Contract December. Design and Pricing,” Available at SSRN: http://ssrn.com/abstract=966436 or Levy-Yeyati, E., and U. Panizza. 2011. “The Elusive Cost of Sovereign Defaults,” http://dx.doi.org/10.2139/ssrn.966436. Journal of Development Economics, 94(1):95-105. Schularick, M. and A. Taylor. 2012. “Credit booms gone bust: Monetary policy, Lindert, P.H., and P.J. Morton. 1989. “How Sovereign Debt Has Worked,” leverage cycles, and financial crises, 1870-2008,”American Economic Review In Developing Country Debt and Economic Performance, Volume 1: The 102(2):1029-61. International Financial System, ed. J. Sachs, 39-106. Chicago: University of Chicago Sexauer, S., and B. van Ark. 2010. “Escaping the Sovereign-Debt Crisis: Press. Productivity-Driven Growth and Moderate Spending May Offer a Way Out,” The Lo, S., and K. Rogoff. 2015. “Secular Stagnation, debt overhang and other Conference Board Executive Action Series, December. rationales for sluggish growth, six years on,” BIS Working Paper No. 482, Bank for Shiller, R.J. 1993. Macro Markets: Creating Institutions for Managing Society’s Largest International Settlements, January. Economic Risks. Oxford. Clarendon Press. McKinsey Global Institute. 2010. Debt and Deleveraging: The Global Credit Bubble Shiller, R.J. 2004. “Radical Financial Innovation,” Cowles Foundation Discussion , January. and its Economic Consequences Paper No. 1461, Yale University (http://cowles.econ.yale.edu). McKinsey Global Institute. 2012. Debt and Deleveraging: Uneven Progress on the Shiller, R.J. 2013. “Capitalism and Financial Innovation,” Financial Analysts Journal, January. Path to Growth, 69(1):21-25, Jan-Feb. McKinsey Global Institute. 2015. , February. Debt and (Not Much) Deleveraging Shiller, R.J., R.M. Wojakowski, M.S. Ebrahim, and M.B. Shackleton. 2011. Medas, P., J. Reynaud, T. Mahedy, I. Krznar, and P. Rabanal. 2013. IMF Country Report “Continuous Workout Mortgages,” NBER Working Paper 17007, May. 13/41. January. Steinhauser, G. 2013. “Can debt cuts be automated?” The Wall Street Journal, Mian, A., and A. Sufi. 2014.House of Debt: How they (and you) caused the Great August 13. , University of Chicago Recession, and how we can prevent it from happening again Sturzenegger, F., and J. Zettelmeyer. 2008. “Haircuts: Estimating Investor Losses Press. in Sovereign Debt Restructurings,” Journal of International Money and Finance, Mian, A., and A. Sufi. 2014. “Debt Forgiveness in History,” houseofdebt.org, May 27(5):780-805. 18 (accessed on November 27, 2015). Takats, E., and C. Upper. 2013. “Credit and growth after financial crises,” BIS Miyajima, K. 2006. “How to Evaluate GDP-Linked Warrants: Price and Repayment Working Paper No 416, July. Capacity,” IMF Working Paper 06/85, March. Tomz, M. 2007. Reputation and International Cooperation: Sovereign Debt across Mody, A. 2013. “Sovereign debt and its restructuring in the Euro area,” Bruegel Three Centuries. Princeton: Princeton University Press. working paper, August. Turner, A. 2015. “The Case for Monetary Finance – An Essentially Political Issue,” Montier, J. 2016. “Market Macro Myths: Debts, Deficits, and Delusions,” GMO 16th Jacques Polak Annual Research Conference, International Monetary Fund, White Paper, January. November 5-6. Ostry, J.D., A.R. Ghosh, and R. Espinoza. 2015. “When Should Public Debt Be Voth, J. 2014. “Fear, Folly and Financial Crises: Some Policy Lessons from History,” Reduced?” IMF Staff Discussion Note 15/10, June. UBS Center Public Paper #2, April. Panizza, U., F. Sturzenegger, and J. Zettelmeyer. 2009. “The Economics and Law of Zettelmeyer, J., C. Trebesch, and M. Gulati. 2013. “The Greek Debt Restructuring: Sovereign Debt and Default,” Journal of Economic Literature, 47(3):651-698. An Autopsy,” Peterson Institute for International Economics Working Paper 13-8, August. Global Debt: Challenges and Opportunities | 16

Appendix Index Definitionsa

Standard and Poor’s (S&P) 500 Index: A market capitalization- MSCI Jordan Equity Market Indexb: Measures the performance of weighted free float-adjusted index of 500 large-capitalization the Jordanian market. It was developed with a base value of 100 as of U.S. stocks. It captures approximately 80% of available market December 31, 1987. capitalization. The index was developed with a base level of 10 for b the 1941-43 base period. MSCI Mexico Equity Market Index : Measures the performance of the large- and mid-cap segments of the Mexican market. It was S&P 500 Equal Weight Index: The equal-weight version of the developed with a base value of 100 as of December 31, 1987. With 27 S&P 500. constituents, the index covers about 85% of the free float-adjusted market capitalization in Mexico. S&P 500 Buyback Index: An equal weighted index designed to measure the performance of the top 100 stocks with the highest MSCI Peru Equity Market Indexb: Measures the performance of buyback ratios in the S&P 500. the large- and mid-cap segments of the Peruvian market. It was developed with a base value of 100 as of December 31, 1992. With b MSCI All Country World Index (ACWI) : Captures large- and mid- three constituents, the index covers approximately 85% of the capitalization representation across 23 Developed Market (DM) and Peruvian equity universe. 23 Emerging Market (EM) countriesc. It was developed with a base value of 100 as of December 31, 1987. With 2,482 constituents, MSCI Philippines Equity Market Indexb: Measures the performance the index covers approximately 85% of the global investable equity of the large- and mid-cap segments of the Philippines market. It was opportunity set. developed with a base value of 100 as of December 31, 1987. With 22 constituents, the index covers about 85% of the Philippines equity b MSCI Argentina Equity Market Index : Measures the performance universe. of the large- and mid-cap segments of the Argentine market. It was developed with a base value of 100 as of December 31, 1987. MSCI Poland Equity Market Indexb: Measures the performance With nine constituents, the index covers about 85% of the Argentine of the large- and mid-cap segments of the Polish market. It was equity universe. developed with a base value of 100 as of December 31, 1992. With 23 constituents, the index covers about 85% of the Polish equity universe. MSCI Brazil Equity Market Indexb: Measures the performance of the large- and mid-cap segments of the Brazilian market. It MSCI Russia Equity Market Indexb: Measures the performance was developed with a base value of 100 as of December 31, 1987. of the large- and mid-cap segments of the Russian market. It was With 60 constituents, the index covers about 85% of the Brazilian developed with a base value of 100 as of December 30, 1994. With 20 equity universe. constituents, the index covers about 85% of the free float-adjusted market capitalization in Russia. MSCI Greece Equity Market Indexb: Measures the performance of the large- and mid-cap segments of the Greek market. It was developed with a base value of 100 as of December 31, 1998. With ten constituents, the index covers about 85% of the Greek equity universe.

a. Index definitions in this Appendix are based on information available at the index providers’ websites and on Bloomberg as of March 31, 2016. b. The MSCI ACWI and all MSCI Country indices defined in this Appendix are free-float market capitalization-weighted. c. DM countries include: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Hong Kong, Ireland, Israel, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Singapore, Spain, Sweden, Switzerland, the U.K. and the U.S. EM countries include: Brazil, Chile, China, Colombia, Czech Republic, Egypt, Greece, Hungary, India, Indonesia, Korea, Malaysia, Mexico, Peru, Philippines, Poland, Russia, Qatar, South Africa, Taiwan, Thailand, Turkey and United Arab Emirates. Global Debt: Challenges and Opportunities | 17

CHIEF INVESTMENT OFFICE Christopher Hyzy Chief Investment Officer Bank of America Global Wealth & Investment Management

Mary Ann Bartels Karin Kimbrough Niladri Mukherjee Head of Merrill Lynch Wealth Head of Investment Strategy Director of Portfolio Strategy, Management Portfolio Strategy Merrill Lynch Wealth Management Private Banking & Investment Group (PBIG) and International Nicholas Giorgi Tony Golden Emmanuel D. Hatzakis Marci McGregor Rodrigo C. Serrano John Veit Vice President Director Director Director Vice President Director

The Global Wealth & Investment Management Chief Investment Office (GWIM CIO) provides industry-leading investment solutions, portfolio construction advice and wealth management guidance. This material was prepared by the GWIM CIO and is not a publication of BofA Merrill Lynch Global Research. The views expressed are those of the GWIM CIO only and are subject to change. This information should not be construed as investment advice. It is presented for information purposes only and is not intended to be either a specific offer by any Merrill Lynch entity to sell or provide, or a specific invitation for a consumer to apply for, any particular retail financial product or service that may be available. Past performance is no guarantee of future results. This information and any discussion should not be construed as a personalized and individual client recommendation, which should be based on each client’s investment objectives, risk tolerance, liquidity needs and financial situation. This information and any discussion also is not intended as a specific offer by Merrill Lynch, its affiliates, or any related entity to sell or provide, or a specific invitation for a consumer to apply for, any particular retail financial product or service. Investments and opinions are subject to change due to market conditions and the opinions and guidance may not be profitable or realized. Any information presented in connection with BofA Merrill Lynch Global Research is general in nature and is not intended to provide personal investment advice. The information does not take into account the specific investment objectives, financial situation and particular needs of any specific person who may receive it. Investors should understand that statements regarding future prospects may not be realized. No investment program is risk-free and a systematic investing plan does not ensure a profit or protect against a loss in declining markets. Any investment plan should be subject to periodic review for changes in your individual circumstances, including changes in market conditions and your financial ability to continue purchases. The investments discussed have varying degrees of risk. Some of the risks involved with equities include the possibility that the value of the stocks may fluctuate in response to events specific to the companies or markets, as well as economic, political or social events in the U.S. or abroad. Bonds are subject to interest rate, inflation and credit risks. Investments in foreign securities involve special risks, including foreign currency risk and the possibility of substantial volatility due to adverse political, economic or other developments. These risks are magnified for investments made in emerging markets. Trading in commodities is speculative and can be extremely volatile. Market prices of the commodities may fluctuate rapidly based on numerous factors, including changes in supply and demand relationships; weather; agriculture; precious metals; trade; fiscal, monetary and exchange control programs; domestic and foreign political and economic events and policies; disease; technological developments; and changes in interest rates. Alternative investments are intended for qualified and suitable investors only. Alternative investments involve limited access to the investment and may include, among other factors, the risks on investing in derivatives, using leverage, and engaging in short sales, practice which can magnify potential losses or gains. Alternative investments are speculative and involve a high degree of risk. Neither Bank of America nor any of its affiliates provide legal, tax or accounting advice. You should consult your legal and/or tax advisors before making any financial decisions. Bank of America Merrill Lynch is a brand name used by several Bank of America Corporation (BofA Corp.) businesses, including, but not limited to, Global Institutional Consulting, which offers products and services for the benefit of institutional and ultra high net worth clients. Global Institutional Consulting (GIC) is part of the Global Wealth and Retirement Services business of BofA Corp. Institutional Investments & Philanthropic Solutions (Philanthropic Solutions) is part of U.S. Trust, Bank of America Corporation (U.S. Trust). U.S. Trust operates through Bank of America, N.A., member FDIC, and other subsidiaries of BofA Corp. For both GIC and Philanthropic Solutions, banking and fiduciary activities are performed by wholly owned banking affiliates of BofA Corp., including Bank of America, N.A. Both brokerage and investment advisory services are provided by Merrill Lynch, Pierce, Fenner & Smith Incorporated, a registered broker-dealer, Member Securities Investor Protection Corporation (SIPC), www.sipc.org, and registered investment adviser; the nature and degree of advice and assistance provided, the fees charged, and clients’ rights and MLPF&S’s obligations will differ depending upon the products and services actually provided. © 2017 Bank of America Corporation. All rights reserved. AR8D7PNV | 00-63-1598NSB