COMMUNITY BANKING

Economic Capital and the Assessment of Capital Adequacy

by Robert L. Burns hat will Basel II mean for community banks? This question can’t be answered without first understand- W ing economic capital. The FDIC recently produced an excellent analysis of economic capital and Basel II, which The RMA Journal is pleased to reprint. It is a must read for all community banks. he assessment of capital adequacy is one of ulations have not been the driving force behind the the most critical aspects of bank supervision. development of these models as such methodologies TIn completing this assessment, examiners have been in use for more than 10 years at some of focus on a comparison of a bank’s available capital the nation’s largest banks. Economic capital has also protection with its capital needs based on the bank’s become a useful and sometimes necessary tool for overall risk profile. other insured institutions. Several regional banks and Bank management must likewise continuously some community banks have developed or are evaluate capital adequacy in relation to risk. In exploring implementation of economic capital mod- recent years, many banks have adopted advanced els with more banks likely to do so in the future. modeling techniques intended to improve their abili- This article provides an introduction to the concept ty to quantify and manage risks. These modeling of economic capital, describes the relationship techniques frequently incorporate the internal alloca- between economic capital and the revised Basel tion of “economic capital” considered necessary to framework, and discusses examiner review of eco- support risks associated with individual lines of busi- nomic capital models as a part of the supervisory ness, portfolios, or transactions within the bank. As a assessment of capital adequacy. result, economic capital models can provide valuable additional information that bankers and examiners Economic Capital can use in their overall assessment of a bank’s capital Economic capital is a measure of risk, not of cap- adequacy. ital held. As such, it is distinct from familiar account- As will be discussed later, economic capital mod- ing and regulatory capital measures. The output of els or similar risk and capital adequacy assessment economic capital models also differs from many other processes are important to banks adopting the measures of capital adequacy. Model results are revised Basel framework. But revisions to capital reg- expressed as a dollar level of capital necessary to

Robert L. Burns, CFA, CPA, is senior examiner, Large Financial Institutions, Federal Deposit Insurance Corporation, Washington, D.C. Reprinted with permission from the FDIC.

54 The RMA Journal April 2005 Economic Capital and the Assessment of Capital Adequacy adequately support specific risks assumed. Whereas tribution and the expected loss. It is sometimes most traditional measures of capital adequacy relate referred to as “unexpected loss at the confidence existing capital levels to assets or some form of level.” adjusted assets, economic capital relates capital to The confidence level is established by bank risks, regardless of the existence of assets. Economic management and can be viewed as the risk of insol- capital is based on a probabilistic assessment of vency during a defined time period at which man- potential future losses and is therefore a potentially agement has chosen to operate. The higher the con- more forward-looking measure of capital adequacy fidence level selected, the lower the probability of than traditional accounting measures. The develop- insolvency. For example, if management establishes ment and implementation of a well-functioning eco- a 99.97% confidence level, that means it is accepting nomic capital model can make bank management a 3 in 10,000 probability of the bank becoming insol- better equipped to anticipate potential problems. vent during the next 12 months. Many banks using Conceptually, economic capital can be expressed economic capital models have selected a confidence as protection against unexpected future losses at a level between 99.96% and 99.98%, equivalent to the selected confidence level. This relationship is pre- insolvency rate expected for an AA or Aa credit rat- sented graphically in Figure 1. ing. Expected loss is the anticipated average loss The primary value of economic capital, and the over a defined period of time. Expected losses repre- reason that banks have already adopted such sent a cost of doing business and are generally methodologies, is its application to decision making expected to be absorbed by operating income. In the and risk management. Specifically, the use of such case of loan losses, for example, the expected loss models can: should be priced into the yield and an appropriate • Contribute to a more comprehensive pricing sys- charge included in the allowance for loan and lease tem that covers expected losses. losses. • Assist in the evaluation of the adequacy of capi- Unexpected loss is the potential for actual loss to tal in relation to the bank’s overall risk profile. exceed the expected loss and is a measure of the • Develop risk-adjusted performance measures uncertainty inherent in the loss estimate.1 It is this that provide for better evaluation of returns and possibility for unexpected losses to occur that neces- the volatility of returns.2 sitates the holding of capital protection. • Enhance risk management efforts by providing a Economic capital is typically defined as the dif- common currency for risk. ference between some given percentile of a loss dis- The following example illustrates how each of these potential uses could be applied at a bank.3 This Figure 1 Economic Capital example describes only cred- it risk quantification and its Expected Loss Confidence Level translation to economic capi- tal for commercial lending activities. Obviously, risks are evident in activities other than commercial lending, Economic Capital and commercial lending itself involves numerous risks in addition to credit

Frequency of Loss risk.4 Banks that use eco- nomic capital models gener- ally identify and quantify all types of risk across all lines of business throughout the Amount of Loss (increasing to the right) bank. 55 Economic Capital and the Assessment of Capital Adequacy

Economic Capital Allocation for Table 1 Commercial Credit Risk Example Obligor Grades and Associated Default Probabilities At its most fundamental level, Internal 1 2 3 4 5 6 7 8 9 10 credit risk is associated with loan Loan Grade losses resulting from the occur- Average rence of default and the subse- Probability 0.03% 0.06% 0.10% 0.25% 0.50% 1.00% 2.50% 8.00% 22.00% 100.00% of Default quent failure to collect in full the Mapping to AA A BBB+ BBB BB+ BB B+ B CCC D balances owed at the time of External Ratings default.5 Expected credit losses associated with default can therefore be determined to loans are often associated with factors such as loan from parameters associated with the likelihood of a type, collateral type, collateral values, guarantees, or loan defaulting, or an estimate of the probability of credit protection such as credit default swaps.6 default (PD) during a defined time period, and the Pricing Implications: A credit facility that is the severity of loss expected to be experienced in the same in all other respects may be priced differently event of a default, or an estimate of loss given based on its expected loss.7 Table 2 shows expected default (LGD). Naturally, this ratio would be applied losses for three different borrowers with the same to a measure of estimated exposure at default (EAD) loan structure and collateral support resulting in a to convert loss expectations to dollar amounts. The 40% loss severity in the event of default. The resulting formula is: higher-risk credit grade has five times the expected loss of the lower-risk credit grade. Expected losses ($) = PD(%) * LGD(%) * EAD($). If a bank made middle-market loans that fell into the three grade bands shown in Table 2, but priced most of these loans with an implicit loss PD and LGD parameter estimates are drawn expectation from the bank’s historical performance or from a of 50 basis mapping of internal portfolio risk assessments to Table 2 Expected Loss points, the external information sources for PD and LGD Loan Grade PD * LGD = Loss bank is parameters. This requires that banks have in place 5 0.50% 40% 20 basis points overcharging processes that enable them to periodically assess 6 1.00% 40% 40 basis points stronger bor- credit risk exposures to individual borrowers and 7 2.50% 40% 100 basis points rowers and counterparties with robust internal credit-rating sys- undercharg- tems that reflect implicit, if not explicit, assessments ing weaker borrowers. One potential result is that the of loss probability. Definitions of credit grades bank could end up with stronger borrowers exiting should be sufficiently detailed and descriptive to the bank and find its loan pool progressively weaker clearly delineate risk level between grades and and portfolio returns inadequate for losses experi- should be applied consistently across all business enced. lines. Although such a highly quantitative process may For example, a bank could have a 10-grade cred- appear somewhat foreign to many bankers, a form of it rating system with associated one-year probabili- probability of default estimates is considered in the ties of default drawn from their historical default use of consumer FICO scores or banks’ own internal experience within each grade, as shown in Table 1. loan scorecards. Furthermore, many banks, including In this example, the historical default rate experi- many community banks, are already relating this enced for loans internally graded as a “6” has been type of analysis to their allowance for loan and lease 1%, which is approximately equivalent to the long- loss determination. term default frequency associated with an S&P cred- it rating of BB. Capital Adequacy: The allocation of economic Estimates for loss severity in the event of default capital to support credit risk begins with similar could likewise be constructed. LGD grades assigned inputs to derive expected losses, but considers other

56 The RMA Journal April 2005 Economic Capital and the Assessment of Capital Adequacy factors to determine Table 3 unexpected losses, such Example Economic Capital Allocations ($) for $100 1-Year Maturity Commmercial Loan as credit concentrations Facility Grades and default correlations and Associated Obligor Grades Associated Default Probabilities among borrowers. Loss Given 1 2 3 4 5 6 7 8 9 Because borrower Default defaults are not perfectly 0.03% 0.06% 0.10% 0.25% 0.50% 1.00% 2.5% 8.00% 22.0% correlated, the default A 10% $0.13 0.23 0.33 0.62 0.93 1.30 1.84 2.84 4.05 risk of a credit portfolio is B 20% 0.27 0.46 0.66 1.23 1.85 2.61 3.67 5.69 8.10 less than the sum of the C 30% 0.40 0.69 1.00 1.85 2.78 3.91 5.51 8.53 12.14 risks contained in the D 40% 0.54 0.91 1.33 2.46 3.71 5.21 7.35 11.38 16.19 underlying loans. Economic capital credit E 50% 0.67 1.14 1.66 3.08 4.64 6.51 9.18 14.22 20.24 risk modeling therefore F 60% 0.81 1.37 1.99 3.70 5.56 7.82 11.02 17.06 24.29 measures the incremental G 70% 0.94 1.60 2.32 4.31 6.49 9.12 12.86 19.91 28.33 risk that a transaction H 80% 1.08 1.83 2.66 4.93 7.42 10.42 14.69 22.75 32.38 adds to a portfolio rather than the absolute level of I 90% 1.21 2.06 2.99 5.55 8.35 11.72 16.53 25.60 36.43 risk associated with an cism; i.e., both credits would be “pass” credits. individual transaction. Complex models are required However, the economic capital allocations show a to derive this measure of portfolio loss volatility and considerable difference in the inherent risk between translate that into an associated economic capital these loans. A $100 one-year maturity commercial charge. loan that is graded a 6 would receive a $5.21 credit Table 3 shows an example of credit risk econom- economic capital allocation compared with a $3.71 ic capital allocations (credit risk only) determined allocation for a similar loan graded 5, an increase of using the PD and LGD parameters previously dis- approximately 40% in estimated risk. cussed and a model translation of those parameters into a credit risk capital charge.8 The bank’s obligor Risk-Adjusted Performance Measures: grades and associated PDs are shown at the top of Economic capital is also used to evaluate risk-adjust- this table. The bank’s facility grades and associated ed performance; without some quantification of risk loss severity estimates are shown on the left-hand associated with an activity, it is not possible to meas- side of the table. The associated capital charges rep- ure performance on a risk-adjusted basis. Several resent the dollar amount of capital needed to support techniques have been developed with two such a $100 one-year maturity commercial loan based on approaches that incorporate economic capital alloca- parameter inputs (such as the PD estimate) and tions demonstrated below: model assumptions (such as default correlations). • Risk Adjusted Return on Capital (RAROC), a Credit economic capital allocations for a non- percentage measure of performance = Economic defaulted $100 one-year maturity commercial loan Net Income / Economic Capital Allocation using this model would range from as low as 13 cents • Economic Profit, or Shareholder Value Added to as high as $36.43. Everyone intuitively expects (SVA), a dollar measure of performance = increased risk to be associated with lower-quality Economic Net Income - (Economic Capital graded loans or loans with higher loss severity, but Allocations * Hurdle Rate)9 the allocation of economic capital estimates the level Assume that a bank is considering the perform- of risk associated with a particular grade band and ance of two loan portfolios: Portfolios X and Y, with differentiates risk among bands. Portfolio X assumed to be higher risk and producing For example, commercial loans graded as a 5 or a a higher return relative to Portfolio Y (see Table 4). 6 with an LGD of 40% in the table above would not Using internal grading parameters and economic cap- likely be subject to regulatory classification or criti- ital modeling for credit risk, management can

57 Economic Capital and the Assessment of Capital Adequacy strengthen its evaluation of the risk/return trade-off that produce lower risk. Economic capital should be of the two portfolios.10 viewed as a tool to enhance risk identification and Please note this example considers only credit selection. Decisions resulting in the acceptance of risk. Bank management would incorporate assess- higher credit risk can be expected to occur when ments of other risks in determining risk-adjusted supported by transaction level returns that compen- performance. sate for higher risk or increased portfolio diversifica- tion benefits. Table 4 Example Risk-Adjusted Performance Measures Risk Management: The implications for loan Portfolio X Portfolio Y pricing, capital analysis, and risk-adjusted perform- Portfolio Balances $100,000,000 $100,000,000 ance measures relate directly to risk management, but economic capital, as a common currency of risk, Net Income before $1,400,000 $1,100,000 Losses* can provide additional potential applications to the Loan Parameters: risk management process. For example, some banks use credit economic capital allocations in place of or - PD 0.50% 0.25% in addition to more traditional credit hold limits - LGD 50% 40% based on notional exposures that may not fully cap- - EL (in bps) 25 10 ture factors such as potential loss severity, default Expected Losses $250,000 $100,000 correlations with the rest of the credit portfolio, or 12 Income after maturity effects on default probability. $1,150,000 $1,000,000 Expected Losses The preceding example focused on credit risk. Economic Capital But similar assessment and quantification efforts can $4,640,000 $2,460,000 (credit only)** help banks identify, monitor, and manage other risks in other lines of business as well. RAROC 24.8% 40.7% The effectiveness of a bank’s risk management Economic Profit $686,000 $754,000 (10% hurdle rate) practices is an important consideration in the super- * Net income before losses = loan interest + fees + soft dollars - funding costs visory evaluation of an institution and directly influ- - operating costs. ences the regulatory assessment of capital adequacy. ** Determined from the exonomic capital charges shown in Table 3. Strong risk management practices can compensate in part for higher levels of inherent risk in a bank’s Initially, bank management may have been business activities. Recognizing this relationship, the inclined to select Portfolio X, based on simple return revised Basel capital framework promotes the adop- characteristics, as shown below. On a risk-adjusted tion of stronger risk management practices through- basis, however, Portfolio Y is the preferred alterna- out the banking industry by incorporating industry tive. Although Portfolio X produces higher expected advances in risk modeling and management into reg- book and economic net income, the volatility of ulatory capital requirements. Portfolio X’s return (i.e., risk) is not adequately com- pensated for in comparison to Portfolio Y. Portfolio Y generates a higher RAROC and results in a greater Economic Capital and Basel II The revised Basel framework seeks to create economic profit on a significantly lower economic more risk-sensitive regulatory capital requirements in capital allocation.11 order to address concerns that the regulatory capital Income after Expected Losses $1,150,000 $1,100,000 measures established by the 1988 Basel Accord do Flat Capital Charge (e.g., 8%) $8,000,000 $8,000,000 not adequately differentiate risk, and to reduce regu- latory capital arbitrage activities that have eroded the Return on Equity 14.4% 12.5% relevance of current risk-based capital measures at some institutions. Many industry participants and Although the decision reached in this example observers have associated economic capital with the resulted in lower overall credit risk, economic capital calculation of minimum regulatory capital require- models are not designed to always favor strategies

58 The RMA Journal April 2005 Economic Capital and the Assessment of Capital Adequacy ments under the first pillar of the revised framework substantially with economic conditions.14 and the supervisory review process under the second More fundamentally, the risks captured under pillar. As discussed below, however, economic capital regulatory and economic capital differ. The regulato- and regulatory capital under the revised framework ry capital charge captures only credit, market, and are not synonymous. operational risk. Furthermore, the regulatory capital calculation does not fully address certain aspects of The First Pillar: Minimum Capital Requirements these risks, such as credit concentration risk. As pre- The calculation of minimum regulatory capital viously discussed, economic capital models generally under the revised framework relies heavily on certain address all risks arising from the bank’s business inputs from the bank’s assessment of its individual activities. risk profile. For example, the calculation of the capi- Economic capital also typically incorporates a tal charge for credit risk considers the distribution of diversification benefit that is not considered in the a bank’s specific credit exposures among internally regulatory capital calculation. This diversification assigned PD and LGD grades. The translation of benefit is a top-line measure of how changes in the that risk profile into a capital charge, however, is con- risk associated with each business activity occur in sistent for all institutions. The required inputs follow relation to changes in risk in all other activities. specific slotting criteria and are applied against regu- Figure 2 provides a graphic example of some of latory risk weight curves that are the same for all the potential differences between regulatory capital institutions.13 This need to Figure 2 ensure consistency necessarily creates Economic Comparison of Minimum Regulatory Capital with Economic Capital differences between a bank’s Capital ($ billions) internal capital 35 Regulatory Model Bank Model allocations and the Model Drivers: minimum regula- Liquidity Risk funding sources and 30 4 uses stress scenerio tory capital charge. analysis Potential dif- Business Risk Measure of potential ferences also exist 25 4 earnings volatility in the inputs used. Interest Rate Risk Economic Value of Model Drivers: 5 Equity (EVE) results For example, in its 20 10-day Value-at Risk plus specific risk Value-at-Risk over a economic capital Market Risk charges Market Risk liquidation period model, a bank 4 5 plus stress scenerio 15 Frequency and sever- analysis ity loss distributions may use a long- Frequency and sever- and other factors Capital Operational Risk term estimate of $25 ity loss distributions Capital Operational Risk 5 and other factors LGD that covers 10 $21 5 PD and LGD bands, EAD, and some PD, LGD, EAD, and all economic maturity data as maturity as inputs, inputs, regulatory observed correlations cycles, but for reg- 5 risk curves used to Credit Risk used; credit consider- Credit Risk capture correlations; 10 ations considered; ulatory capital 12 credit losses related credit losses related to changes in eco- purposes, the to default nomic value 0 LGD estimate Diversification Vector analysis of should reflect eco- Benefit risk correlations -7 nomic downturn -5 Other Model Differences: - Different confidence leves used conditions for - Variations in input data exposures where -10 loss severities are Minimum Regulatory Capital Economic Capital expected to vary

59 Economic Capital and the Assessment of Capital Adequacy under the revised Basel framework and economic sory requirement for economic capital methodologies capital at a hypothetical bank. In this example, total to be employed in this process, many large institu- economic capital allocations are higher than the regu- tions appear likely to use their economic capital latory minimum capital charge. While this typically models to demonstrate capital adequacy in relation may be expected to be the case, in some instances a to risk under Pillar 2. bank could reasonably have lower economic capital allocations than regulatory capital requirements Supervisory Review of Economic Capital depending on the specific risk characteristics of the Regulators expect certain large or complex banks bank and the significance of the diversification bene- to perform appropriate risk quantification and capital fit. analysis regardless of whether the bank is subject to As demonstrated by the above discussion, a bank the revised Basel framework. This is particularly is not required to have a fully functional economic important at banks where more traditional capital capital model to develop the necessary inputs for the adequacy measures may not adequately capture the calculation of the minimum regulatory capital charge. inherent risk of their business activities, such as at These inputs generally can be determined inde- banks heavily engaged in securitization activities. An pendent of any comprehensive risk measurement economic capital model is one tool available for such and management process. However, the second pillar analysis. of the revised framework creates a more direct link At banks where economic capital models are to a bank’s own risk and capital adequacy assess- used, considerable supervisory effort is focused on ments. the process. Examiners consider both the adequacy of economic capital processes and the results of such The Second Pillar: Supervisory Review Process processes in their supervisory evaluation of the bank. The second pillar establishes a regulatory expec- Furthermore, as discussed later in this article, exam- tation for the evaluation of how well banks assess iners may find it beneficial to modify certain tradi- their own capital needs. The second pillar does not tional examination procedures to more fully evaluate explicitly require banks to adopt economic capital risk management practices associated with the eco- models. It does, however, establish an expectation nomic capital process and other risk modeling tech- for banks to perform a comprehensive assessment of niques. the risks they face and to relate capital adequacy to these risks.15 Process Review Furthermore, the bank’s own capital analysis is When properly used, economic capital models expected to encompass all risks, not only those risks can improve risk management and the evaluation of captured by the minimum regulatory capital calcula- capital adequacy. However, these models can suffer tion. The revised Basel framework describes three from data limitations, erroneous assumptions, inabili- areas not addressed in the minimum capital calcula- ty to sufficiently quantify risks, and potential misuse tion that should be specifically considered under the or misunderstanding of model outputs. Examiner second pillar: assessment of the appropriateness of a bank’s capital • Risks that are not fully captured under the first adequacy analysis, potentially including economic pillar, such as credit concentration risk. capital methodologies, can be a consideration in the • Risks that are not considered under the first pil- supervisory evaluation and rating of bank manage- lar, such as interest rate risk. ment. Institutions found to have material weakness- • Factors external to the bank, such as economic es in their methodologies may be directed to conditions.16 strengthen risk measurement and management capa- The supervisory qualification and ongoing vali- bilities. dation of a bank’s compliance with regulations The supervisory approach used to evaluate a implementing the revised framework will necessarily bank’s economic capital process will necessarily vary incorporate review of a bank’s risk quantification based on the complexity of the institution and the efforts and capital analysis. While there is no supervi- extent of use of the economic capital process by

60 The RMA Journal April 2005 Economic Capital and the Assessment of Capital Adequacy bank management. Examination guidance on eco- • The quality of management information sys- nomic capital models is limited. Federal Reserve tems, including the timeliness of incorporation Board Supervisory Letter SR 99-18 and the second of changes in the bank’s risk profile. pillar of the revised Basel framework do not specifi- • Internal or external audit program review of eco- cally address economic capital methodologies, but nomic capital methodologies. both documents describe the supervisory review of a • The corporate governance structure as it relates bank’s capital analysis process. Many of the princi- to risk management and economic capital. ples discussed in these documents are included in Determine the extent to which the economic the general review concepts examiners may want to capital process is used in decision making, such as in consider that are discussed below. setting risk limits or evaluating performance. Economic capital processes that are in place but not Evaluate the adequacy of board and manage- integrated with the institution’s risk management ment oversight concerning economic capital. procedures generally are ineffective. Management is responsible for understanding the nature and level of risks undertaken in the bank’s Results Review activities and how these risks fit within the overall The results of economic capital models can pro- business strategy of the bank. To evaluate this over- vide examiners another tool in the supervisory evalu- sight, examiners could review: ation of capital adequacy, enabling examiners to • Specific board approval of risk tolerances and compare tangible capital levels (capital available to associated capital levels. support risk) with economic capital levels (the bank’s • Periodic economic capital reports provided to the own measure of its risk). As has always been the board and senior management. Such reports case, an institution found to hold inadequate capital should be sufficient to allow the board and man- in relation to risk, regardless of the institution’s com- agement to evaluate risk exposures, determine pliance with minimum regulatory capital require- that the bank holds sufficient capital relative to ments, is expected to take appropriate actions to identified risk, and incorporate capital needs into reduce risk or increase capital. the strategic planning process. Banks generally operate with a capital cushion Determine that economic capital methodologies above the level of risk measured by the economic appropriately incorporate all material risks. At a mini- capital model, recognizing the imprecision inherent mum, this should include assessments of credit, mar- in such estimation and the need for the bank to be ket, operational, liquidity, and business risks. To responsive to potential changes in conditions. make this determination, examiners could review: Several factors can be considered in determining the • A mapping of data inputs to material exposures, appropriate cushion, including: ensuring accuracy and completeness. • The robustness of the bank’s economic capital • Documentation supporting the appropriateness methodologies. of specific risk quantification techniques. • The quality of data inputs, assumptions, and • Analysis supporting the reasonableness and parameters. validity of stress tests and scenarios used. • Volatility of the business model. • Analysis and testing of model sensitivity to key • The composition of capital,17. assumptions and data inputs used. • External factors, such as business cycle effects • Model validation work, including, where appro- and the macroeconomic environment. priate, the evaluation of developmental evi- dence, process verification, benchmarking, and Incorporation into the Overall Supervisory back-testing. Process Evaluate the control environment. Controls The development and implementation of risk should be in place to ensure the integrity of data models such as economic capital often represent a inputs and the overall management process. In eval- significant change in a bank’s overall risk manage- uating such controls, examiners could consider: ment philosophy and practices. Likewise, the overall

61 Economic Capital and the Assessment of Capital Adequacy supervisory process for banks default-only measure of credit risk discussed in the lower than their existing business line RAROC. For example, should be viewed as potential approach- example, a manager might choose to reject an oth- adopting economic capital models es rather than as the only or best alternative. erwise desirable investment with a 20% RAROC if can be affected as examination his line of business had an average RAROC of 4 Examples are interest rate risk and operational 25%. focus may shift more to process risk associated with underwriting and servicing of evaluation. Transactional testing loans. 11 Note that Portfolio X and Portfolio Y, when con- sidering credit risk only, would be acceptable to would continue to figure promi- 5 The example describes a default-only perspective management as both generate positive economic nently in the examination func- to derive a loss distribution; i.e., loan defaults cre- profit assuming a hurdle rate of 10 percent. ate credit losses. Some banks have adopted a more tion, but the purpose of transac- robust perspective for credit loss that considers the 12 Intuitively, longer maturity loans to the same tional testing may be redirected to probability distribution of obligor grade migration borrower entail greater credit risk; i.e., the default and resulting changes in the economic value of the risk of a five year loan to a borrower, even a bor- validation. loan; i.e., a decline in the credit quality of a loan rower of strong credit quality, is significantly For example, the earlier dis- regardless of any default creates credit losses. greater than a six-month loan to the same borrow- er. However, traditional credit hold limits, such as cussion of commercial lending 6 Some banks consider guarantees and credit pro- notional exposures by loan grade, rarely capture credit risk highlights the need for tection as substitutes for the borrower and there- this maturity effect. Credit economic capital alloca- fore use guarantor or counterparty PDs in place of tions frequently adjust the one-year PD estimates examiners to focus on validating borrower PDs, while other banks retain the bor- for an obligor to reflect the differences in credit risk rower PD and consider guarantees and credit pro- resulting from facility maturity. the accuracy of the loan-grading tection in determining LGD. process at all grade bands rather 13 The regulatory risk-weight curves serve as a 7 Pricing models are considerably more complex proxy for default correlations, with the expected than concentrating their attention than the simplistic approach shown in this exam- default experience among weaker commercial bor- primarily on large or criticized ple. This discussion is merely intended to show that rowers (credits with higher PDs) assumed to be less expected losses are often built into the pricing of correlated with systemic risk (overall economic facilities. The classification of loans. conditions). individual loans becomes integrat- 8 The credit economic capital allocations shown in 14 Paragraph 468 of the revised Basel framework. ed with the evaluation of the the table were derived using the regulatory capital International Convergence of Capital bank’s internal loan-grading sys- calculation for corporate credit exposures under Measurement and Capital Standards, June 2004, the revised Basel framework. Refer to International Basel Committee on Banking Supervision. tem. Convergence of Capital Measurement and Capital Furthermore, economic capi- Standards, June 2004 text, Basel Committee on 15 Paragraph 732 of the revised Basel framework: Banking Supervision. As discussed later in this arti- "All material risks faced by the bank should be tal results can provide useful cle, the regulatory capital calculation under the addressed in the capital assessment process. While information for risk-scoping. revised framework differs in important ways from the Committee recognizes that not all risks can be economic capital methodologies, but is used for measured precisely, a process should be developed Examiners can incorporate the illustrative purposes in this example as a proxy for to estimate risk." International Convergence of bank’s risk quantification efforts an economic capital methodology to avoid disclos- Capital Measurement and Capital Standards, June ing information about proprietary models used by 2004, Basel Committee on Banking Supervision. and trends in economic capital any bank. The table includes nine obligor grades allocations as another tool to bet- and nine facility grades; the tenth borrower grade 16 Paragraph 724, International Convergence of previously discussed was for defaulted loans and is Capital Measurement and Capital Standards, June ter focus supervisory efforts on not shown as the methodology for estimating risk 2004, Basel Committee on Banking Supervision. in defaulted exposures varies considerably among areas of high or increasing risk. institutions. 17 This is particularly critical when considering the The use of economic capital capacity of various elements of capital to absorb 9 The hurdle rate can be viewed as the firm-wide losses under stress scenarios. and other risk modeling tech- cost of capital. Returns above the hurdle rate add niques is expected to continue to to shareholder value and those below, while per- evolve and expand to more indus- haps profitable, detract from shareholder value. try participants. Supervisory eval- 10 In many banks, risk-adjusted performance measures are built into the determination of com- uations of banks are also changing pensation for line of business managers and staff, to appropriately incorporate such directly influencing behavior at the business line level. Often, both a dollar level of risk-adjusted per- advances by the industry. ❒ formance, such as SVA, and a percentage measure, such as RAROC, are used. Footnotes Percentage measures of performance are often 1 Unexpected loss is often described as the volatil- used because dollar measures may not provide suf- ity of loss around the average over time. ficient information to distinguish between alterna- tive acceptable investments. For example, two port- 2 Risk-adjusted performance is typically measured folios could produce the same dollar measure of at the business unit level, but can also be used to risk-adjusted performance, but one could require evaluate how individual business unit returns con- substantially larger capital allocations. tribute to a bank’s overall profitability and risk pro- Dollar measures of performance may be used file. because managers might be inclined to reject an investment that would generate positive SVA if 3 Specific methodologies, such as the use of a that investment generated a RAROC that was

62 The RMA Journal April 2005