July 04, 2006
Addressing Challenges Raised by Basel II Implementation
Governor Susan Schmidt Bies
At the Risk Capital 2006 Forum, Paris, France
Good morning. And if you will indulge me on this U.S. national holiday, I would like to wish you a happy Fourth of July. I am honored to be addressing such an esteemed collection of professionals in the field of risk management. As you know, the Federal Reserve is strongly committed to the continuing evolution of risk measurement and management at U.S. banking organizations. Thus, we want you to be successful at your craft, and meetings such as this will facilitate the communication of emerging best practice ideas across financial institutions.
Today I will provide an update on progress with respect to Basel II implementation in the United States, and describe areas where further advances in risk management practices are most needed to ensure the success of this new risk-based capital standard. Naturally, given the international composition of this audience, I will also offer some thoughts on cross-border implementation issues associated with Basel II, including so-called home-host issues.
Moving to Basel II
By now most of you are aware that on March 30 of this year the Federal Reserve Board approved a draft of the U.S. notice of proposed rulemaking (NPR) on the Basel II capital framework. I imagine that many of you have already read the draft’s 400-plus pages. Once all of the U.S. banking agencies have completed their individual review and approval processes, the NPR will be issued in the Federal Register, meaning that it will be officially out for comment. It appears that those approval processes are going well and there have not been any unanticipated issues so far. Indeed, the recent completion of the Office of Management and Budget’s (OMB) review of the draft NPR brings us one step closer to issuing the NPR for comment.
Since the release of the draft NPR by the Federal Reserve, the U.S. banking agencies have met with a number of industry groups, comprising both U.S. and foreign banks, to discuss the document. From our perspective, these discussions have been informative and helpful, and we believe that some of the clarifications we have provided have been useful to bankers. I would like to make one further clarification, on a procedural matter that may be a bit confusing: Although we welcome comments on the draft NPR at this time, bankers should not expect that those comments will be incorporated in the NPR that is released in the Federal Register. That stage will come later in our rulemaking process. A proposed 120-day comment period will be provided once the final NPR has been released, and all comments given now and during that period will be considered as the final rule is written.
Before discussing some of the specifics in the draft NPR, I would like to review the Federal Reserve’s reasons for pursuing Basel II.
Rationale for Moving to Basel II
The current Basel I capital framework, adopted nearly twenty years ago, has served us well. But it has become increasingly inadequate for large, internationally active banks that are offering ever more complex and sophisticated products and services. We need a better capital framework for these particular banks, and we believe that Basel II is such a framework.
One of the major improvements in Basel II is the closer linkage between capital requirements and the way banks manage their actual risk. The current Basel I measures have very limited risk-sensitivity and do not provide bankers, supervisors, or the marketplace with meaningful measures of risk at large complex organizations. Under Basel I, a bank’s capital requirement does not adequately reflect gradations in asset quality and does not change over time to reflect deterioration in asset quality. Further, there is no explicit capital requirement to account for the operational risk embedded in many of the services from which the largest institutions generate a good portion of their revenues.
In addition to strengthening the linkage between “minimum regulatory capital,” as calculated in Pillar 1, and the way banks manage their “actual” capital, Basel II should make the financial system safer by encouraging continuing improvement in risk-measurement and risk-management practices at the largest banks. Pillar 1 of Basel II is based on many of the economic capital practices of the most sophisticated banks and therefore brings minimum regulatory capital requirements closer to the capital generated by banks’ internal models. By providing a consistent framework within which the largest banks calculate minimum regulatory capital requirements, supervisors will more readily be able to identify those portfolios and banks whose capital is not commensurate with their inherent risk levels. Working within this consistent framework and engaging in ongoing and regular dialogue with supervisors will in turn help inform management about how its proprietary risk measurement and management models compare with the range of current practices and where enhancements are needed. We have already seen some progress in risk measurement and management at many institutions in the United States and around the globe as a result of preparations for Basel II. Admittedly, banks have told us that some of the costs for Basel II would have been incurred anyway. But if anything, Basel II has accelerated the pace of this change.
Basel II can also provide supervisors with a more conceptually consistent and more transparent framework for evaluating systemic risk in the banking system, particularly through credit cycles. Thus it improves on Basel I, which requires banks to hold the same level of capital for a given portfolio, no matter how the portfolio’s inherent risk may change over time. Further, as bankers gain experience with the advanced approaches under Basel II, they will have better information on how their risk taking may vary through various cycles. Therefore, Basel II establishes a more coherent relationship between how supervisors assess regulatory capital and how they supervise banks, enabling examiners to better evaluate whether banks are holding prudent levels of capital, given their risk profiles.
The reasons I’ve just given for pursuing Basel II also provide justification for the recent Basel revisions to the 1996 Market Risk Amendment (MRA). Since adoption of the MRA, banks’ trading activities have become more sophisticated and have given rise to a wider range of risks that are not easily captured in their existing value-at-risk (VaR) models. For example, more products related to credit risk, such as credit default swaps and tranches of collateralized debt obligations, are now included in the trading book. These products can give rise to default risks that are not captured well in methodologies required by the current rule specifying a ten-day holding period and a 99 percent confidence interval, which can create potential arbitrage opportunities between the banking book and the trading book. The U.S. agencies are in the final stages of drafting the NPR to revise Market Risk capital requirements, which will be issued for comments. In the United States we would continue to have banks with significant trading book activity hold additional capital for the risks inherent in that line of business, whether they remain Basel I banks or move to Basel II.
Bridging the Gap Between Regulatory Capital Requirements and Internal Bank Practice
With Basel II, U.S. supervisors are attempting to use the internal risk-measurement and risk-management information produced by large complex institutions to manage their own risks in such a way as to augment the risk sensitivity and overall meaningfulness of minimum regulatory capital measures. Basel II, by tying Pillar 1 minimum regulatory capital calculations to bank-generated inputs, offers greater transparency about the practices that stand behind the inputs provided by banks and the way those inputs are calculated. In fact, through the Basel II framework, supervisors will be able to compare results across banks and provide better information to bankers about how their models and methodologies agree with sound practice and where they may be an outlier.
Of course, we understand that the extent to which banks’ internal inputs can be used in minimum regulatory capital requirements is limited, for a variety of reasons. Today’s banks have highly customized models for running their businesses, which of course is entirely appropriate. But as supervisors, we need to ensure adequacy and enforceability of our minimum regulatory capital requirements while maintaining some consistency across banks. In a broader sense, we are working to protect the safety and soundness of our financial system. Naturally, as we seek to develop a common framework that will work for large complex banks globally, we recognize an inherent tension between our regulatory requirements and internal bank practice. We are working to strike the right balance to achieve our goals without making Basel II purely a compliance exercise and creating undue burden.
Need for Strong Capital
Basel II is intended to improve regulatory capital requirements, especially for large complex organizations, through greater risk sensitivity of regulatory capital and improved linkage to banks’ actual capital risk management. That is why the U.S. agencies have chosen to adopt only the most advanced options for credit risk and operational risk minimum regulatory capital calculations in the United States, and to limit the requirement of Basel II to only a small number of banking institutions that fit the definition of large, complex, and internationally active.
It is important to recognize that Basel II is a complete capital framework consisting of three pillars. While much of the focus to date has been on Pillar 1 and the calculation of minimum regulatory capital, the importance of Pillar 2, which provides for supervisory review and oversight of an institution’s overall capital adequacy, should not be overlooked. Pillar 2 is intended to ensure that banks have adequate capital to support all the risks to their business, and to further encourage them to develop and use better techniques to monitor and manage the risks. It addresses some kinds of risk that are not captured in Pillar 1, such as credit concentration, interest rate, and liquidity risks. And it provides supervisors with the opportunity to assess compliance with the minimum standards and disclosure requirements of the advanced approaches in Pillar 1.
Some key principles of the supervisory review process are discussed in the framework document. Time does not permit a full discussion of each, but I do want to at least mention them today. First, banks should have a process for assessing their overall capital adequacy in relation to their risk profile and a strategy for maintaining their capital levels. Second, supervisors should review and evaluate banks’ internal capital adequacy assessments and strategies, as well as their ability to monitor and ensure their compliance with regulatory capital ratios--and should take appropriate supervisory action if they are not satisfied with the results of this process. Third, supervisors should expect banks to operate above the minimum regulatory capital ratios and should have the ability to require banks to hold capital in excess of the minimum. Finally, supervisors should seek to intervene at an early stage to prevent capital from falling below the minimum levels indicated by the risk characteristics of a particular bank--and should require rapid remedial action if capital is not maintained or restored.
Many of the actions implied by the Pillar 2 principles are already part of the supervisory process in the United States. The agencies plan to provide some information about our expectations for Pillar 2, beyond what is included in the existing U.S. supervisory process, in forthcoming supervisory guidance.
Let me assure you that we at the Federal Reserve would not be pursuing Basel II if we thought that it would in any way undermine the strong capital base that U.S. institutions now enjoy. As a central bank and a supervisor of banks, bank holding companies, and financial holding companies, the Federal Reserve is committed to ensuring that the Basel II framework delivers a strong and risk-sensitive base of capital for our largest and most complex banking institutions. That is why we supported moving ahead with the draft NPR, which was modified to address concerns identified in the fourth quantitative impact study, known as QIS4, and includes additional safeguards to ensure strong capital levels during the transition to Basel II. We will remain vigilant in monitoring and assessing Basel II’s impact on individual and aggregate minimum regulatory capital levels on an ongoing basis. As an extra degree of precaution, the U.S. banking agencies also decided to delay for a year the start of the parallel-run period.
Starting with the parallel run, and both during and after the transition period, the U.S. agencies will rely on ongoing, detailed analyses to evaluate the results of Basel II, so as to ensure prudent levels of capital. Importantly, Basel II represents a new way of thinking about regulatory capital; it is complex, reflecting the complexity of risk measurement and management for the largest, most complex banking institutions. Banking institutions and their supervisors will need to have ongoing dialogue and work diligently to make sure Basel II is working as we expect it to. But we believe it is a powerful approach to making regulatory capital more risk sensitive. To be quite clear, the Federal Reserve believes that strong capital is critical to the health of our banking system, and we believe that Basel II will help us continue to ensure that U.S. banks maintain capital levels that serve as an appropriate cushion against their risk taking.
Some Aspects of the U.S. Proposals
As you know, the draft U.S. Basel II NPR is based on the 2004 framework issued by the Basel Committee and adheres to the main elements of that framework. But the U.S. agencies have exercised national discretion and have tailored the Basel II framework to fit the U.S. banking system and U.S. financial environment--just as their counterparts in other countries have tailored the framework to their situations. For example, as I have just mentioned, the U.S. agencies continue to propose that we implement only the advanced approaches of Basel II, namely the advanced internal-ratings-based approach (AIRB) for credit risk and the advanced measurement approaches (AMA) for operational risk.
Also, the draft NPR proposes a more gradual implementation timetable and a more rigorous set of transition safeguards than those set forth in the 2004 Basel II framework. For instance, the U.S. agencies are proposing three years of transition floors below which a bank’s minimum required capital under Basel II would not be permitted to fall. The first transition period would have a floor of 95 percent relative to the general risk-based capital rules, the second period 90 percent, and the third period 85 percent. Implementation of a more gradual transition timetable is justified in part by a recognition that banks need more time to prepare--and that we as supervisors need more time to analyze transition information and ensure that there are no unintended consequences.
As you are aware, the QIS4 exercise identified some areas requiring further clarification by regulators and additional work by bankers on risk models and databases. One of the key areas in the NPR influenced by these results pertains to banks’ estimates of loss given default (LGD). Many QIS4 participants reported difficulty computing LGDs (which must reflect downturn conditions), in part because their data histories were not long enough to capture weaker parts of the economic cycle. To address this problem, the agencies have proposed a supervisory mapping function that can be used in the interim by those institutions unable to estimate appropriate downturn LGDs. The mapping function allows an institution to “stress” its expected LGDs, generating an input to the capital calculation that conforms to the Basel II requirements and hence produces a more appropriate capital requirement. The Federal Reserve supported the introduction of this supervisory mapping function, as an important component of U.S. Basel II implementation, in order to address a specific challenge articulated by the industry. Banks would be able to shift from using the mapping function to using their own internal estimates of LGDs when their own estimates become reliable.
About a month ago, the Basel Committee released the results of QIS5, which was conducted by a number of countries but not the United States, since we had already conducted QIS4. As you know, the QIS4 results from U.S. institutions are not completely comparable with the QIS5 results from institutions in other countries. Nonetheless, some similarities are worth noting. First, the aggregate declines in minimum regulatory capital of banks using the advanced approaches were similar in the two exercises, particularly considering that QIS4 did not include the 1.06 capital multiplier and QIS5 did. For both exercises the aggregate declines reflected relatively good economic times--which, as we know all too well, do not last forever. Second, both studies pointed to dispersion of changes in minimum required capital. The dispersion among banks in QIS5 was attributed largely to a combination of differences in portfolio characteristics and differences and uncertainties in estimation methodologies. The U.S. agencies highlighted similar issues in their public release on QIS4 and have stated their intention to monitor issues related to dispersion very closely in the future, because we want to ensure that the Basel II framework does indeed accurately produce similar capital for similar risk. The Basel Committee’s release on QIS5 also stated that methodologies and systems for LGD calculation are still being developed and that, as a result, some of the effects of downturns may have been underestimated. The U.S. agencies' release on QIS4 stated that U.S. banks also faced some challenges in estimating downturn LGDs.
Basel I Modifications
At this point I would like to say just a few words about ongoing efforts to revise the existing Basel I regulatory capital rules for non-Basel II institutions. We expect only one or two dozen banks to move to the U.S. version of Basel II in the near term, meaning that the vast majority of U.S. banks will continue to operate under Basel I, which will be amended through a separate rulemaking process. The Basel I framework has already been amended more than twenty-five times in response to changes in banking products and the banking environment and as a result of a better understanding of the risks of individual products and services. The U.S. agencies believe that now is another appropriate time to propose modifications to Basel I rules. The agencies have issued an advance notice of proposed rulemaking discussing possible changes to increase the risk sensitivity of U.S. Basel I rules and to mitigate any competitive distortions that might be created by introducing Basel II. We are now in the process of reviewing comments on the advance notice and working on a notice of proposed rulemaking. We are mindful that within the current structure of the Basel I rules, amendments to those rules should not be too complex or too burdensome for the large number of banks to which the revised rules will apply.
With regard to both the Basel II proposals and the proposed Basel I amendments, we understand the need for full transparency. For that reason, we expect to have overlapping comment periods for the Basel II NPR and the NPR for the proposed Basel I amendments. In fact, we want all interested parties to compare, contrast, and comment on the two proposals in overlapping timeframes. Accordingly, our proposals could change as a result of comments received or new information gathered.
Cross-Border Implementation of Basel II
As I noted earlier, each country must implement Basel II as appropriate for its particular jurisdiction. To that end, the U.S. banking agencies are acting to ensure that the Basel II framework is implemented in the United States in a prudential manner. We recognize that the adoption of differing approaches to Basel II by various countries may create challenges for banking organizations that operate in multiple jurisdictions. It is good to remember that cross-border banking has always raised specific challenges that supervisors from various countries have worked hard to address. Let me assure all bankers here that supervisors are aware that the process of changing to new national versions of Basel II has heightened concerns about home-host issues. The Federal Reserve and other U.S. agencies have, for many years, worked with international counterparts to limit the difficulty and burden that have arisen as foreign banks have entered U.S. markets and as U.S. banks have established operations in other jurisdictions.
The United States is working to complete its national-standard-setting process, as we recognize that the lack of a final rule creates uncertainty, for both banks and foreign supervisors, about exactly what will be required. At the same time, we have been working with our colleagues on the Basel Accord Implementation Group (AIG) for the past few years to identify issues arising from differences in national standards of the Basel II framework. All of the supervisory bodies participating in the AIG effort are committed to making the transition to Basel II successful.
We have heard from some bankers that they are concerned about home-host issues. Many of the issues are institution specific, and all the U.S. banking agencies encourage regular meetings between bankers and supervisors to identify specific concerns. These meetings are also an opportunity for bankers to make supervisors aware of their individual bank’s implementation plans and progress, and for supervisors to make bankers aware of current supervisory expectations with respect to those banks. Indeed, the meetings we have had to date to discuss the draft NPR have provided us with useful information about how our proposals are being interpreted and perceived.
We continue to ask for details about concerns bankers may have about cross-border implementation; it is a great help, as we work on our proposals, to hear your specific issues. We were pleased that at a few recent meetings, bankers did, in fact, provide some very specific information about how possible differences in implementation could generate extra burden and affect the way in which the bankers conduct their business. Suggestions for changes to our proposals that are specific, detailed, and supported by facts will allow us, in drafting the final rule, to more clearly identify alternative solutions to the issues raised.
Of course, all Basel-member countries have their own rollout timelines and their own ways of addressing matters that are left to national discretion under the Accord, which is entirely appropriate. So while the United States appears to be a bit of an outlier with respect to its implementation timetable, other countries are grappling with their own challenges, too. As you know, a number of other countries plan to implement the full set of Basel II approaches for credit risk (standardized, foundation IRB, and advanced IRB) and operational risk (basic indicator, standardized, and the advanced measurement approaches), with all but the advanced approaches expected to be implemented next year in many Basel countries. This underscores the importance of regular dialogues between home and host supervisors and banks. I hope that bankers pay special attention to the Basel Committee’s June 2006 paper “Home-Host Information Sharing for Effective Basel II Implementation,” as it provides guidance on how cooperative efforts can successfully realize the objectives of Basel II.
The Federal Reserve, as I have indicated, believes that Basel II is a worthwhile endeavor despite the challenges facing us. The current regulatory capital framework, based on Basel I, is not adequate for the largest, most complex U.S. banking organizations. We applaud the risk-management improvements made by institutions so far, but additional work remains to be done before banks can meet our expectations--not just for Pillar 1, but also for Pillars 2 and 3.
Implementing Basel II is a serious undertaking, with many moving parts both domestically and internationally. The U.S. banking agencies are cognizant of the challenges associated with Basel II, including the challenge of ensuring that its effects are those that we intend. We are also aware of the difficulties that could arise in implementing Basel II on a cross-border basis. So far, we have been pleased with our talks with the industry about our implementation plans and how those plans interact with the plans of other countries. We understand that there are questions to be answered and issues to be addressed, and we hope that bankers--and other interested parties--will provide detailed feedback on the U.S. NPR so that we can assess the entire spectrum of comments. While some observers may be critical of the slower pace of the U.S. agencies, we believe that our deliberate pace is necessary to ensure that the effects of Basel II are indeed understood and that all comments are taken into consideration.