Each depositor insured to at least $250,000 per insured bank



Home > News & Events > Speeches & Testimony




Speeches & Testimony

Statement of Thomas J. Curry Director Federal Deposit Insurance Corporation
on Basel II: Capital Changes in The U.S. Banking System and
the Results of The Impact Study Before the Subcommittee
on Financial Institutions and Consumer Credit and Subcommittee
on Domestic and International Monetary Policy, Trade, and Technology
of The Committee on Financial Services U.S. House of Representatives
May 11, 2005
Room 2128, Rayburn House Office Building
   

Chairman Bachus, Chairman Pryce, Ranking Member Sanders, Ranking Member Maloney and members of the subcommittees, it is a pleasure to appear before you today to discuss the perspectives of the Federal Deposit Insurance Corporation regarding Basel II. Basel II is an effort to tie capital requirements more closely to risk and promote a disciplined approach to risk management at our largest banks. The FDIC supports these goals and the process of implementing a revised capital framework in the United States .

My testimony will focus on some concerns the FDIC has about the results of the recent quantitative impact study, QIS-4. I will also have some comments about the requirements for operational risk capital. The issues we discuss today may sound sweeping and fundamental, but we believe they can be resolved. Our intention is to work with our fellow regulators to address our concerns and to move forward expeditiously when this is done.

Background
In June 2004, the Basel Committee achieved an important milestone with the publication of “International Convergence of Capital Measurement and Capital Standards,” representing an informal agreement among the Committee members about the framework that would form the basis of national supervisors' efforts to implement the new approaches. When publishing the new framework, the Basel Committee recognized that individual countries must decide how to implement the new capital measurements and standards, given their own unique circumstances. The four federal banking agencies are in the process of drafting a proposed rule to implement Basel II's “Advanced Approaches” in the United States . The term “Advanced Approaches” refers to the Basel II approaches that rely most fully on banks' own risk estimates.

The FDIC brings a number of perspectives to the proposed rulemaking process. In addition to our role as primary federal supervisor of a number of institutions that have indicated an interest in opting in to the new framework, the FDIC's role as deposit insurer requires a keen interest in the risk profile of any bank adopting the new framework. In both our supervision and deposit insurance roles, we interact with the thousands of banks where capital will not be set by the Basel II standards, but will be affected, directly or indirectly, by the adoption of Basel II.

The work on the proposed rule, like all the agencies' work on Basel II, has been intensely collaborative, and characterized by vigorous give-and-take on many individual issues. In such a process, there is always a danger that the focus on the details can result in a loss of focus on the big picture. It is important from time to time to step back and take a fresh look at the totality of what we have created through years of negotiations. There have been a number of such opportunities during the development of Basel II. The 2004 Basel text was preceded by the Basel Committee's publication of three consultative papers, each of which received extensive comment from the banks that would be most affected. There were also various quantitative studies in which participating banks provided their own risk inputs to simulate the potential capital impact of the proposals. The comments received on each of the consultative papers and the insights gained from the quantitative studies resulted in significant changes to the framework over the years.

In the light of all the changes to the new framework, culminating in the Basel Committee's 2004 mid-year text, the U.S. agencies embarked on a fourth quantitative impact study, QIS-4. QIS-4 is a comprehensive effort completed by 26 large U.S. consolidated banking organizations during late 2004 and early 2005. The purpose of the impact study was to use these organizations' internal estimates of the key risk parameters driving capital requirements for credit risk and operational risk under the Basel II framework (not all banks provided estimates of exposure to operational risk). Each bank's risk parameters and exposures were fed into the Basel II formulas to estimate the minimum capital requirements that would result for each consolidated banking organization and each line of business under the new framework. The agencies have long envisioned that QIS-4 would serve as an important input to the proposed rulemaking process.

A summary of the results of QIS-4 is contained in an appendix to this testimony. It is important to note that these results are preliminary and that the agencies' review of QIS-4 is not complete. Nevertheless, in part because the QIS-4 results are consistent with previous FDIC analysis, we have formed some preliminary conclusions.

In the FDIC's view, QIS-4 shows excessive reductions in risk-based capital requirements. Capital requirements fell by more than 26 percent in more than half of the institutions in the study. This is without fully factoring in the benefits of credit risk hedging and guarantees that are likely to reduce capital requirements significantly more. For individual loan types at individual banks, over one third of the reductions in capital requirements were in the range of 50 to almost 100 percent. Numbers like this do not provide comfort that the Basel framework will require capital adequate for the risks of individual activities.

The FDIC also is concerned that the dispersion of results suggests there is a difficulty in applying the framework consistently across banks. Capital requirements in Basel II are very sensitive to inputs. Achieving consistency in Basel II depends on the idea that best practices, and best data, will lead to convergence in the capital treatment of similar loan portfolios across banks. At present, however, at least as indicated by QIS-4, there is little commonality in the approaches the various banks used to estimate their risk inputs.

The FDIC has communicated on many occasions about the continued need for a leverage ratio. As discussed at greater length later in this testimony, the QIS-4 results suggest that U.S. leverage requirements will be more important under Basel II than ever before. The FDIC can support moving forward with this new framework only because of the existence of the leverage-based component of U.S. capital regulation.

The FDIC also has a concern about the potential competitive effects of the new framework. If QIS-4 is representative of capital requirements going forward under Basel II, the competitive ramifications for community banks and large non-adopting banks could, in our view, be profound. If Basel II is implemented unchanged, mitigating these competitive effects would seem to require a substantial reduction in risk-based capital requirements for non-Basel banks.

The remainder of this testimony will contain a brief summary of how Basel II computes capital requirements and the conceptual change this approach represents. This is followed by a discussion of the QIS-4 results—our concern with those results—the role of the leverage ratio in U.S. capital regulation, competitive equity issues, and some observations on key implementation issues.

A new paradigm for capital regulation
To provide perspective, it is worthwhile to reflect on the overall thrust of the change that Basel II represents. The fundamental changes represented by Basel II provide an incentive to improve risk management practices, and elevate the role of banks' and supervisors' judgment in determining risk-based capital requirements. While this judgment is expected to be informed by analysis, the importance of judgment is nevertheless infinitely multiplied under Basel II's advanced approaches.

To calculate capital requirements for credit risk under the current capital standards ( Basel I), each exposure is slotted into one of a few simple categories, each with a predetermined capital requirement. Under Basel II, the same exposure could attract a capital charge of anywhere from essentially zero to many multiples of the current charge, depending on specific risk inputs for that exposure estimated by the bank and approved by the supervisor. The capital requirements under the proposed Basel II will be much more risk-sensitive, and much more subjective, than at present.

Basel II computes minimum capital requirements for credit risk using a set of formulas that are, at least in comparison to other risk-modeling approaches, relatively simple. The inputs to these formulas are probability of default (PD), loss given default (LGD), exposure at default (EAD) and, for wholesale exposures, maturity (M). While each bank supplies its own inputs, the regulators have developed the formulas themselves through the Basel Committee on Banking Supervision. The regulators have six separate formulas for: wholesale credit exposures, small business wholesale exposures, high volatility commercial real estate, revolving exposures (mostly credit cards), residential mortgages, and other retail loans. The capital requirement is computed for each exposure or pool of exposures based on the inputs the bank provides and added across the entire bank to get the total capital requirement for credit risk. The capital requirement is always eight percent of risk-weighted assets, just as it is now, but the risk weighted asset number now becomes the byproduct of the aforementioned calculations and assumptions.

One of the outputs of the calculations above is a number called the “expected loss,” interpreted as the amount of credit losses a bank would expect over a one year period given the assumptions it made about the PDs and LGDs for its exposures. The total expected loss for the bank is compared to its allowance for loan and lease losses and other reserves. If the expected loss exceeds these reserves, the difference is deducted from capital. If reserves exceed expected loss, the excess up to a limit is added to capital. These adjustments can add quite significantly to the capital requirements for certain kinds of retail credit, especially credit cards.

A capital requirement also exists for operational risk. Operational risk is the risk of loss associated with human error, failed systems or external events. As discussed later in this testimony, banks will develop historical databases of operational risk losses under the “Advanced Measurement Approach” (AMA). Banks will use these databases to attempt to estimate an amount of operational loss that is highly unlikely to be exceeded. At least in theory, a bank will estimate an amount of capital sufficient to absorb operational loss in 99.9 percent of all scenarios. This estimated loss, after subtraction of certain offsets such as permissible reserves, is—with supervisory approval—the bank's operational risk capital requirement. Banks outside the U.S. are permitted to use simpler approaches to calculate capital for operational risk.

Although this describes the relatively simple part of the Basel II framework, there also is an extremely detailed, complex, formula-driven and internal model-driven infrastructure surrounding the calculation of capital for exposures to securitization, repurchase agreements, equity investments, and a host of other exposure issues that are beyond the scope of this testimony.

The QIS-4 results
Data for the QIS-4 was collected in late 2004 and early 2005 and have been under review since February of this year. The preliminary review suggests that if the QIS-4 is representative of the risk inputs banking organizations use to calculate their capital requirements going forward, risk-based capital requirements in aggregate would decline by roughly 17 percent. An aggregate capital number is of some interest, but perhaps less important in terms of competitive implications and risks to the insurance funds than the results for individual banks. Half of the 26 banking companies participating in the QIS-4 reported capital reductions in excess of 26 percent, with a number of institutions reporting reductions in overall capital requirements in the range of 30 to 50 percent. Other banking companies reported increases in capital requirements of as much as 60 percent.

Capital requirements for specific business lines also showed striking results. Preliminary estimates show capital requirements for wholesale loans down 24 percent in aggregate, with the outcomes ranging from an increase of more than 50 percent to a decrease of almost 75 percent. Capital for high volatility commercial real estate was down 33 percent in aggregate. The wholesale lending category, in particular, is an area where banks appear to have substantial latitude to take advantage of the benefits of guarantees and credit risk hedging in order to further reduce their capital requirements.

Preliminary results show that capital requirements for retail loans were down 26 percent in aggregate, with half the banks showing a reduction of 50 percent or more. Within that retail category, capital requirements for mortgage loans showed a decline of 62 percent in aggregate, with half the banks showing declines in excess of 73 percent. Capital requirements for home equity lines of credit were down 74 percent, with half the banks showing declines in excess of 79 percent. Capital requirements for credit cards were up substantially on average, but ranged from over a 100 percent increase to a decrease of 90 percent.

A number of observations about these results are in order. First, at this time the framework does not appear to comport with the stated expectation of the Basel Committee that overall capital should remain about the same throughout the system, with perhaps only modest reductions. While some have accused the framework of excessive conservatism, the QIS-4 results suggest that Basel II in its current form would bring substantial reductions in risk-based capital requirements. In terms of their average direction and magnitude, the FDIC does not see these QIS-4 results as surprising. An FDIC paper published in December 2003 suggested that when reasonable PDs and LGDs, estimated based on twenty years of U.S. bank charge-off history, are entered into the underlying Basel II credit risk formulas, those formulas can be expected to deliver substantial reductions in risk-based capital requirements.

The QIS-4 results also illustrate that under the advanced approaches, there is potential for substantial dispersion in capital requirements in ways that are not explainable by real differences in risk. Capital requirements under the advanced approaches depend heavily on the answers to questions that have no objectively best answer. For example, if a borrower defaults in the future, how much will the bank lose per dollar of the loan? One expert might guess 20 cents on the dollar and another might guess 30 cents on the dollar. While reasonable people might be hard-pressed to decide which expert is correct, the conclusion reached in this specific example would swing the Basel II capital requirement by 33 to 50 percent.

The agencies did, in fact, observe a wide range of practice in how banking organizations estimated their PDs, LGDs and exposures for QIS-4 purposes. This range of practice suggests that considerable practical challenges lie ahead in the supervision of Basel II's advanced approaches. In part, the challenge will be to achieve consistent application of Basel II across institutions. We may want to avoid a situation where a banking organization's Basel II risk-based capital requirement is, for all practical purposes, whatever capital level is acceptable to its regulator. To limit the potential unintended consequences of such a situation, implementing Basel II implies the need for an unprecedented degree of market transparency, interagency collaboration and information sharing. From the FDIC's perspective of assessing risks to the insurance funds, collaboration should include access by all bank regulators to information about the critical assumptions, models and data used to implement capital requirements based on banks' own estimates of risk.

Another important issue relates to the capital treatment of new or emerging lines of business where there is little or no relevant history of loss experience. The recent rapid growth of home equity lines of credit provides a useful current example. FDIC-insured institutions' holdings of this product have tripled in the past 5 years. This rapid growth, the unseasoned nature of the portfolio, and the agencies' belief that home equity lenders' underwriting standards have loosened considerably in recent years all suggest that a certain amount of supervisory and regulatory caution is appropriate. The recent loss experience, however, is favorable, and this is what drives the 80 percent capital reduction for this activity reported in the QIS-4.

As long as banks are growing and innovating, there will always be new and rapidly growing lines of business with little relevant loss history. The example of home equity lending suggests to us that Basel II has not solved the problem of finding the “right” level of capital for such emerging activities, and that further thought is needed about the appropriate prudential approaches in this area. For example, to what extent will the significant reductions in capital requirements for these activities result in a de facto expansion of the federal safety net? One of the classic antidotes to the moral hazard problems associated with deposit insurance is for regulators to require an adequate amount of private equity capital to be at risk. In this respect, the QIS-4 results for mortgages and home equity loans suggest the need for a hard look at how this part of the Basel II framework should be implemented.

The Basel Committee envisioned that calibration issues could be handled by means of a simple multiplier. Specifically, if at some future date the Committee decides that the overall capital required under the new framework is inconsistent with its objectives, either because of being too high or too low, the Committee has reserved the option of proposing that each jurisdiction multiply the capital requirements by a single number, thereby bringing overall capital more in line with the Committee's objectives. Given the wide dispersion and extreme outcomes of the QIS-4, it appears at this time that the need for adjustments to the framework for U.S. implementation purposes could go beyond a simple multiplier adjustment. Serious thought needs to be given to finding ways to achieve results under Basel II that are less extreme and more consistently applicable across banks.

The accuracy of Basel II and the role of the leverage ratio
The Basel II capital accord reflects the significant input of the world's largest banks and has been described by some as a codification of current best practices in risk measurement with a dose of conservatism. Given all this, a natural question that many U.S. bankers have asked is why the U.S. regulators would not place exclusive reliance on the results of the Basel II formulas. These bankers have asked why the U.S. leverage ratio requirements would not be jettisoned or phased out over time as part of the implementation of Basel II. Indeed, they ask, if leverage ratio requirements are retained, why bother with Basel II?

Clearly, a robust and appropriate set of risk-based capital requirements is an important part of our overall regulatory capital system. An equally important role is played by the leverage ratio to ensure that regardless of the risk-based capital model used by a Basel II bank, there will be a base level of capital available in the event of a crisis. Basel II, with its reliance on internal methods and models, does not provide us with that same degree of comfort because there can be little certainty that the Basel II formulas produce an adequate level of capital. For many reasons, we believe that the leverage ratio will continue to serve as a relevant and reliable indicator of bank solvency to be used in conjunction with the Basel II risk-based measures.

First, the Basel II minimum capital measure is not comprehensive. For example, capital is not required for interest rate risk associated with loans held to maturity, or for liquidity risk. These are material risks. The elimination of the leverage ratio would send the signal that these are secondary risks of little importance to the regulatory community.

Second, Basel II is only as good as the inputs entered into the formulas. Analytical mishaps or faulty assumptions that prove to be overly optimistic could have a disastrous effect on the solvency of an institution, as well as the financial system.

Third, no matter how the data used to drive the capital calculation is sliced, we cannot lose sight of the fact that the past ten years have been some of the best years in banking. It is difficult to expect this data—collected during good economic times—will be sufficient to generate capital requirements robust enough to withstand extreme losses under adverse conditions. While the past can be a useful guide to the future, reliance on historical losses as the risk profile of the business line increases could leave a bank unprepared to absorb unexpected losses.

It should also be noted that the Basel II formulas include assumptions with recognized limitations. In Basel II, LGDs are assumed not to increase during a recession. This amounts to assuming that extreme loss scenarios will be less extreme than they might actually be. In Basel II, capital requirements are literally zero for an exposure with a zero reported LGD, whereas economic theory suggests if the bank is being compensated for holding the exposure, this compensation is in return for assuming some risk. In Basel II, credit losses are assumed to have a normal distribution, whereas there is widespread consensus that historical credit losses display a much greater frequency of extreme outcomes than would be predicted by a normal distribution.

While all of these factors bias the capital requirement downward and work against its ability to serve as a buffer against unexpected losses, Basel II does contain other elements that work against these downward biases. The most important is probably that the total capital required for all exposures is the sum of the capital for the individual exposures. This approach, by design, does not allow the bank to benefit from the fact that not all its portfolios are likely to experience the thousand year flood at the same time. Another important factor driving Basel II is the extent that defaults are assumed likely to occur together. The greater this correlation among defaults, the higher the Basel capital requirement. There appears to be a consensus that Basel II is relatively conservative in its correlation assumptions.

The net effect of these multiple offsetting assumptions is that we cannot have confidence that the capital requirement coming out of a Basel II formula is “the right number,” even if reasonable PDs and LGDs were used as inputs. Nevertheless, the hope is that the capital requirements coming out of Basel II will encourage over time a disciplined approach to risk measurement, provide a relative measure of risk among asset types and lessen the incentives for banks to structure their activities in ways that are driven primarily by flaws in regulatory capital.

Quite apart from the specifics of the Basel II models, there is a more fundamental issue that some have raised about the future role of the leverage ratio within the overall structure of U.S. bank capital regulation. The leverage ratio is a simple, clear-cut minimum amount of capital banks need to hold as a percentage of their assets. As indicated earlier, some observers are now suggesting that the arrival of Basel II will, sooner or later, make the leverage requirement obsolete. A closely related idea is that the appropriate benchmarks for capital regulation are banks' own estimates of their capital needs. According to this view, regulators should always set capital requirements less than what “best practice banks” estimate is optimal for their own needs. Requiring more capital than this, it is said, distorts the otherwise optimal function of the marketplace.

This conflicts with a large body of longstanding academic literature, and with the principles reflecting that view as embodied in the FDIC Improvement Act. The existence of a substantial federal safety net underlying banking, including but not limited to deposit insurance, means that the marketplace left to its own devices is likely to establish capital levels for banking organizations that are too low. This reflects the so-called “moral hazard problem” associated with safety nets. Because some creditors, and most notably insured depositors, are insulated from risk, those creditors do not demand any compensation for an increase in the bank's risk profile. The bank, consequently, takes on more risk than it otherwise would. As a result, the argument suggests that a bank enjoying a measure of federal safety net support will tend to hold less capital to support a given risk profile than if it did not enjoy the safety net support.

This is a standard rationale for bank capital regulation and the fundamental basis for the FDIC Improvement Act's Prompt Corrective Action requirements to maintain bank capital at prudent levels. Indeed, if the market could be relied upon to maintain acceptable levels of bank capital, there would seem to be no need for regulatory capital standards. In that case, bank supervision should be sufficient to address any outlier institutions with below-market capital positions.

To suggest that a bank's estimate of its own capital needs is an optimal number ignores both the moral hazard problem and the systemic implications of a large bank failure. There are, in short, legitimate and compelling public policy reasons for bank regulators to require more capital than what a bank estimates for its own needs.

Despite the intensive effort on Basel II development, the framework continues to produce outcomes with which supervisors are not comfortable. The QIS-4 results support the notion that no matter how refined the risk-based capital framework, there will always be a need for straightforward capital minimums. Phasing out the leverage ratio would suggest a willingness to contemplate a significant expansion of the federal safety net, and a significant increase in risk to the financial system. As stated earlier, the FDIC is able to support moving forward with Basel II primarily because of the continued existence of a set of straightforward leverage requirements.

Competitive effects
Absent a substantial reduction in capital requirements for non-Basel II banks, implementing risk-based capital requirements along the lines depicted in the QIS-4 results could have profound competitive implications and could significantly harm the community banking sector in the U.S. , as well as large non-adopters. In our market economy, assets and lending will migrate to where it is most economical to house them. Today, risk-based capital requirements for identical assets are identical across banks so that there is no systematic regulatory capital economy achieved by moving an asset from a small bank to a large bank. Basel II would appear to create significant differences between the capital requirements of small and large banks for many activities. Owners of small banks will receive sub-par returns on their investments in capital-disadvantaged assets compared to the returns that owners of large banks could earn on the same assets. As a result, market forces would likely drive those assets over time away from smaller banks, toward the Basel II adopting banks.

Certainly many factors other than regulatory capital have influenced the distribution of lending activity between small and large banks over time. In fact, since identical assets have received identical capital requirements across banks, one could argue that regulatory capital played no role in affecting market shares. If Basel II results in significant differences in the risk-based capital requirements of small and large banks, it is likely to alter the existing equilibrium.

The FDIC believes it is important to address the potential competitive implications of Basel II. In part, this could be achieved by revisions to the general capital requirements for all U.S. banks. From the standpoint of competitive equity, such revisions ideally would produce like capital requirements for like assets, regardless of whether the bank holding the asset is a Basel II bank or a non-Basel II bank. The magnitude of capital reductions suggested by the QIS-4, however, is likely to raise other issues. As yet, neither bank supervisors nor the FDIC in its role as insurer have had serious discussions about reducing risk-based capital requirements for all U.S. banks in a way that would broadly match the reductions suggested by QIS-4. It is likely, however, that such a discussion would raise issues not only about competitive equity, but about the safety and soundness implications of such a substantial reduction in capital requirements. Moreover, should large financial institutions employ their excess capital through acquisitions of non-Basel II institutions, then this framework would result in a shift in the industry toward greater consolidation, and concentration of exposure and risks. Finding ways to moderate Basel II's potential for extreme results might make these competitive equity considerations easier to address.

Potential revisions to capital standards for non-Basel banks
The U.S. agencies are preparing a proposal outlining potential changes to risk-based capital regulations for all U.S. banks. The agencies will be soliciting comments on ways to achieve greater risk sensitivity in capital in a way that does not create undue burden for insured institutions and is consistent with safety-and-soundness objectives.

These proposals will likely focus on a number of ideas, such as the creation of additional risk buckets for various lending categories, expanded recognition of collateral, and enhancements to the current rules in a few specialty areas. Expanded risk buckets would allow for lower capital requirements for less risky assets and higher requirements for more risky assets. It is anticipated that comments by banks and thrifts will contribute significantly to the agencies' discussions of the factors that should be considered in assigning assets to specific risk buckets. Moreover, unlike Basel II which has separate charges for credit risk and operational risk, the agencies do not envision a capital charge for operational risk to be applied to non-Basel II banks in the U.S.

Current plans are to publish the new capital proposals for all U.S. banks simultaneously with the Basel II proposed rule. These proposals could be compared side-by-side to determine the likely competitive implications of the overall package of proposed changes to U.S. capital regulation.

Implementation issues for insured depository institutions
Global banking organizations have expressed a concern about the practicality of implementing Basel II if the supervisors of individual banks around the world all insist on a bank by bank implementation of the new framework. In the extreme scenario, every bank could be required to maintain its own historical loss databases for credit and operational risk, a separate credit rating system, separate methodologies for determining PDs and LGDs, separate internal audit of the results, and so on. Such a situation would represent an inefficient allocation of resources and, from the perspective of an organization with many subsidiaries, would be unworkable.

At the same time, we must consider the perspective of supervisors of individual banks or other entities such as broker dealers. The supervisors, and for that matter boards of directors and senior management, have significant legal and statutory mandates to ensure the safe and sound operation of the entities under their jurisdiction and governance. In a Basel II world, these supervisors are going to need capital requirements that make sense for their individual supervised entities. Clearly, there is a tension between the responsibilities of individual supervisors and the cost advantages of organization-wide approaches to the implementation of Basel II.

In managing this tension, the principle of absolute accountability of the management and directors of FDIC-insured institutions for the governance of their institutions needs to be preserved. The FDIC's considerable potential deposit insurance liability and, conversely, its ability to recover over time the costs of that liability are specifically attached to insured institutions. It is certainly true that there are large financial conglomerates that in the normal course of events manage on a business line basis rather than a legal entity basis. History has demonstrated repeatedly, however, that in severe, solvency-threatening conditions, organizations move to defend themselves along legal boundaries. The legal location of risks and capital matter very much in these scenarios. That is why the only relevant measures of risk and of regulatory capital adequacy for an insured bank are measures of the bank's risk and the bank's capital. While there may be synergies in data and models that can be realized to streamline the implementation of Basel II in a conglomerate, such models and the data must capture the unique risks present in insured banks that may not be captured if such analysis was performed on a consolidated basis.

The practical questions surrounding these issues are currently being debated vigorously with respect to the proposed capital requirements for operational risk. The Basel II text contains a provision that, with supervisory approval, non-significant subsidiaries of organizations adopting the advanced approaches need not adopt their own advanced measurement approach (AMA) for operational risk. Instead, these non-significant subsidiaries could rely on an “allocation” of the AMA computed by the larger organization. Moreover, again with supervisory approval, that allocation could reflect “diversification benefits.” Diversification benefits in this context essentially refers to the idea that not every legal entity within an organization is likely to have high operational losses at the same time. Consequently, each entity's capital requirement should be lowered relative to the stand-alone amount it would have to hold if it computed its own AMA, reflecting, in effect, an averaging out of potential losses across multiple entities. The use of capital allocation to determine a non-significant subsidiary bank's operational risk capital requirement, and the possible use of diversification benefits to reduce those capital requirements, are collectively known as the hybrid approach to the AMA. Under a hybrid approach, significant banks would compute a stand-alone AMA while non-significant banks need not do so.

The concept of “capital allocation” depends on the idea that the relevant measure of operational risk is at the holding company level and not at the bank level. The idea of diversification benefits goes farther and assumes, in effect, that capital in any legal entity within a holding company structure is equally available to each of the legal entities. Both of these ideas are inconsistent with the fundamental principle that the relevant measures of risk and regulatory capital for an insured bank are those measures that refer to that bank's risk and that bank's capital. Compromising this principle would tend to erode the accountability of the insured bank, contribute to a de facto extension of the federal safety net to non-bank entities, and increase risks to the deposit insurance funds.

The AMA is highly complex and extremely expensive to implement. Moreover, given the wide range of variation in how banks are estimating risk inputs for the comparatively mature discipline of credit risk measurement, it also must be acknowledged that estimates of capital requirements for operational risk capital will be, for the foreseeable future, of unknown usefulness. Rather than compromise the principle of bank-centric capital calculation because the unique features of the AMA force such compromises upon us, we would be more inclined to offer significant flexibility to any bank that is computing a capital requirement for operational risk, including reliance on data and analysis developed outside the insured bank, or allowing banks to use approaches simpler than the AMA.

In short, the tensions between the important principle of stand-alone bank capital calculation, and the costs and burdens of the AMA, can be resolved. There is room for substantial additional thinking in this area, and the FDIC is committed to working with our fellow regulators to arrive at a sensible solution that does not impose excessive burdens on U.S. banks or banking organizations.

Conclusion
The agencies stand at an important crossroads in the development of U.S. capital regulation. The considerations outlined in this testimony suggest to us that achieving an implementation of Basel II that will represent positive change for the U.S. financial system is contingent on several important factors. In brief, they are:

  • Preserving a set of straightforward minimum capital requirements to complement the more risk-sensitive, but also more subjective, approaches of Basel II;
  • Maintaining competitive equity; and
  • Finding ways to achieve results under Basel II that are less extreme and more consistently applicable across banks.

The FDIC, like the other banking agencies, will proceed with the implementation of Basel II in an appropriately deliberative manner and with full consideration of the comments of all interested persons.

Appendix

    QIS-4 Preliminary Change in Effective Minimum Capital Requirements of Participating Institutions: Basel I to Basel II

    Chart: QIS-4 Preliminary Change in Effective Minimum Capital Requirements of Participating Institutions: Basel I to Basel IId
*This is the change in the amount of Tier 1 capital and Tier 2 elements other than reserves needed to meet the minimum capital requirement.

Note:
This is preliminary data as of May 5, 2005 for the twenty-six participating QIS-4 institutions, and caution should be used in drawing any inferences from the aggregate data at this stage. The U.S. banking agencies plan additional work to determine whether these results reflect differences in risk, reveal limitations of QIS4, identify variations in the stages of bank implementation efforts (particularly related to data availability), and/or suggest the need for adjustments to the Basel II Framework.

QIS-4 Preliminary Change in Minimum Capital Requirements of Participating Institutions: Basel I to Basel II

Portfolio

% Change in Portfolio MRC

Median % Change in Port. MRC

Share of Basel I MRC

Share of Basel II MRC

Wholesale Credit

(25%)

(24%)

44.3%

38.8%

     Corporate, Bank, Sovereign

(22%)

(30%)

33.9%

30.7%

     Small Business

(26%)

(27%)

4.6%

4.0%

     High Volatility CRE

(33%)

(23%)

1.8%

1.4%

     Incoming Producing RE

(41%)

(52%)

4.0%

2.7%

Retail Credit

(26%)

(50%)

30.5%

26.3%

     Home Equity (HELOC)

(74%)

(79%)

6.1%

1.8%

     Residential Mortgage

(62%)

(73%)

11.1%

4.9%

     Credit Card (QRE)

66%

63%

6.1%

11.7%

     Other Consumer

(7%)

(35%)

6.0%

6.5%

     Retail Business Exposures

(6%)

(29%)

1.2%

1.3%

Equity

11%

(9%)

1.3%

1.6%

Other assets

(12%)

(3%)

10.1%

10.4%

Securitization

(20%)

(40%)

7.9%

7.7%

Operational Risk

 

 

0.0%

9.0%

Trading Book

0%

0%

5.2%

6.0%

Portfolio Total

(14%)

(24%)

100.0%

100.0%

Change in Effective MRC*

(17%)

(26%)

 

 

*This is the change in the amount of Tier 1 capital and Tier 2 elements other than reserves needed to meet the minimum capital requirement.

Note:
This is preliminary data as of May 5, 2005 for the twenty-six participating QIS-4 institutions, and caution should be used in drawing any inferences from the aggregate data at this stage. The U.S. banking agencies plan additional work to determine whether these results reflect differences in risk, reveal limitations of QIS4, identify variations in the stages of bank implementation efforts (particularly related to data availability), and/or suggest the need for adjustments to the Basel II Framework.


Last Updated 05/11/2005 communications@fdic.gov