Moving Forward into the 21st Century It is a pleasure to be here, when--once again--the industry has enjoyed another year of strong performance. That is not to say bank supervision and regulation today is without its challenges, or that there are few risks to U.S. banks. Perhaps to the contrary, as a result of market dynamics, both domestically and abroad, we can expect to see some loan losses and earnings pressures. Some weaknesses have already surfaced. For the most part, though, we continue to be spared the crisis events that can be so disruptive. This "lull" in domestic financial problems has come at an opportune time, as U.S. and world financial systems adjust to the profound changes of the last decade. The number of insured commercial banks continues to decline--down 4 percent last year and by roughly one-third since the decade began. Meanwhile, the scope and pace of financial innovation continues to expand, making many transactions increasingly difficult to manage and more opaque. Risk management and measurement techniques throughout the industry have become much more quantified with greater integration of information systems and financial theory. Large banks, in particular, have also become far more diversified, now that they can expand nationwide. These changes and the industry's transition, in general, should be viewed as a continuous and natural response by banks to evolving market needs and new technologies. That the industry, itself, is changing is not a problem; banks must adapt to survive. The fact that the industry, rather than the legislative or regulatory process, is leading the change is also appropriate; the private sector should almost always show the way. The result, though, must be managed with care. During the 1990s, banking organizations have increased tremendously in size as a result of the consolidation process, and the complexity of many bank activities has grown as well. These developments have crucial implications for bank supervisors, including those pertaining to systemic risks. In many respects, they have also made bank supervision more difficult. We have not yet achieved "financial modernization" in terms of legislation, but we certainly have a far different banking and financial industry than existed a decade ago. Undoubtedly, more change is on the horizon, as distinctions among financial institutions continue to erode. That fact simply underscores the need for Congress to modify U.S. banking laws and permit the regulatory environment to catch up with market events. Meanwhile, bank supervisors and regulators should remain focused on their principal tasks. First, to ensure that the banking system remains sufficiently safe and sound, posing little risk to the federal safety net and adequately protected against systemic risk. Second, to ensure that the industry continues to provide the American public with a full range of competitively priced banking services and conforms to legislative standards of competitiveness. Perhaps more than before, achieving these goals requires us to adapt our practices to changing circumstances within the banking industry and to take full advantage of the technologies that exist. Clearly, as the industry has changed, so has bank supervision. As banks expanded nationwide, state and federal supervisors worked together, producing the interstate supervisory protocol that provides a more seamless oversight process for state chartered banks. As banks grew larger and more complex, we focused more on risk management practices and controls and less on a bank's condition at a point in time. We also became more risk focused in our overall supervisory approach, emphasizing those activities that presented the greatest risks. As financial innovation and capital arbitrage took hold, we also became more aware of the need to update regulatory capital standards and to make greater use of market discipline. We are pursuing our objectives both domestically among ourselves and abroad through the Basel Committee on Bank Supervision, under the auspices of the Bank for International Settlements in Basel, Switzerland. We are designing a way forward, building upon the "three pillars" approach outlined in a consultative document released today by the Basel Supervisors Committee, a subject I will return to in a moment. This approach encompasses (1) a strong, risk-sensitive regulatory capital standard; (2) an active supervisory program; and (3) improved bank disclosures that allow the marketplace to evaluate an institution's risk posture and to reward or discipline it appropriately. In my remarks today, I would like to address many of these and other points, with particular emphasis on the supervisory process and how we at the Federal Reserve are adapting to change. At the outset, I would emphasize that bank supervision is, by its nature, a dynamic process. Our practices must constantly improve or they will become quickly outdated. Supervisors must also be flexible, both in their application of supervisory techniques to banks and in their expectations regarding what practices individual banks should follow. Perhaps more so than any other, the U.S. banking system is highly diverse, with its thousands of small community banks and a small number of increasingly large, highly complex, internationally active institutions accounting for a growing share of total bank assets. Neither a single supervisory approach, nor a single risk management technique will work for all. That need for flexibility and adaptation has been well served by our dual banking system and by the ability of individual states and state chartered banks to innovate. Large and Complex Banking Organizations One aspect of supervision that has become more crucial to our oversight process relates to systemic risk and to the activities of our largest banking organizations. A decade ago, for example, the 20 largest U.S. banking organizations held 68 percent of the assets of the 50 largest bank holding companies; now its 82 percent. Then, the 20 largest holding companies held 37 percent of all U.S. commercial bank assets; now that figure has risen to 64 percent. Those figures conceal, of course, the dramatic increase in the complexity of their activities represented by securitizations and derivative products. The notional value of derivative and futures contracts held by U.S. banks now exceeds $33 trillion, nearly five times the level at the end of 1990. Securitizations by U.S. banks, at $270 billion, have grown as fast and are expanding beyond consumer-based loans, such as credit card and auto loans, to commercial credits. Virtually all of these securitization and derivative activities are concentrated among the largest banks. While notional values and amounts securitized say almost nothing about the level of underlying risk to individual banks, they speak strongly to the increased volume and complexity of large bank activities and of the somewhat hidden risks they face. For these organizations, balance sheets and traditional lending have much different meanings from a decade ago. Last year, the Federal Reserve responded to this trend by sharpening its supervisory focus on a smaller number of large complex banking organizations, both domestic and foreign. We now give increased attention to roughly twenty U.S.-owned and another ten foreign-owned banks. Although they are generally the largest institutions we supervise, they warrant the greater attention not only because of their size, but also because of their on-and off-balance sheet activities, their broad range of products and services, their more complex domestic and international oversight structure, and their role in payment and settlement systems. We refer to them as LCBOs, for "large, complex banking organizations." In supervising these institutions we recognize that each is unique and complex and that it is particularly necessary for our analysts, examiners, and supervisors to understand sound practices within the industry and to compare activities and risk management techniques among institutions. Accordingly, we are taking a "portfolio" approach, whereby we evaluate practices across institutions where we find similar business lines, characteristics, and risk profiles. This approach fosters more informed and consistent supervision among institutions and provides supervisory staff with greater opportunities to identify and promote sound practices. It also accommodates more readily the development and coordination of staff expertise throughout the Federal Reserve System. The Federal Reserve's supervisory approach toward LCBOs requires ongoing monitoring, including a formal re-evaluation of an institution's risk profile and a quarterly update of our supervisory plan. This periodic assessment is based, in part, on internal management reports, internal and external audit reports, and publicly available information. Since these organizations typically conduct a broad range of regulated activities, supervisory staff must also frequently communicate and coordinate their own activities with those of other bank and nonbank regulators. Management of this oversight process rests with a senior staff member designated as CPC, or "central point of contact." That individual, in turn, coordinates virtually all interaction between the Federal Reserve and the institution, and directs an identified team of examination and supervisory staff having specialized skills tailored to the unique profile of the institution. This structure, combined with the ability of the CPC to draw upon additional staff throughout the Federal Reserve System, should promote greater understanding of an institution's business and risk management process, while reducing our level of intrusion. Indeed, a necessary aspect of our supervisory review is maintaining a steady flow of relevant information about an institution's exposures and risk management system in order to reduce the time-consuming and burdensome discovery process often associated with traditional examination and oversight techniques. Periodic review of management reports should not only enhance our knowledge of specific exposures and events, but also provide insights into a bank's control process and about what information management deems important. In some cases, it may be most convenient to us and to the bank if we have direct access, on-line, to management information. Indeed, that is the case now for a couple of our largest institutions, particularly with respect to the internal audit process. Effective supervision of an LCBO requires a supervisory plan that is tailored to the institution's current risk profile and organizational and operational structure and that considers the activities of other supervisors--highlighting, once again, the need for communication and coordination. The plan should address the major risks (e.g., credit risk, market risk, and so forth) and should employ follow-up actions ranging from off-site analysis and meetings with management, to targeted or full-scope examinations. CPCs should also structure the plan to achieve the proper balance of review of risk management practices and transaction testing, the latter relying typically on statistically sound sampling techniques. Information sharing and coordination with other supervisors are key elements of the program and are essential to successful supervision of these large institutions. For this purpose, the Federal Reserve will continue to enhance its base of information technology and extend its resources to other supervisors. Many of you are already aware of an information system we are developing called the Banking Organization National Desktop, or "BOND." When introduced next year, that system should provide supervisors with both public and confidential information about an institution in a highly user-friendly way. The system should prove particularly helpful in monitoring and evaluating conditions at the largest institutions. For the system to be useful, though, it needs to be used--and to be fed the information people want. These requirements, in turn, require a high degree of security, so that individuals can take comfort that information they put into the system is not misused or misdirected. This aspect of the system has been given great importance and should actually strengthen the level of security surrounding confidential information, while also disseminating necessary information. In supervising LCBOs, we not only expect more of ourselves, we also have higher standards for the institutions. A fundamental tenet of supervision is that the nature of a bank's risk management process must be consistent with the level of underlying risk. More sophistication is necessary as transaction volume and complexity rise. Credit Risk and Capital Our higher expectations in the level of management skills and sophistication at larger banks will also become more apparent in the years ahead in terms of capital standards. Much has been said recently in supervisory statements and industry publications about the need to revise the 1988 Basel Capital Accord and to improve, more generally, the credit risk management of banks. Credit risk has always been the dominant risk in banking, yet it remains crudely measured. This lack of quantitatively rigorous risk measurement within the industry explains why we developed the current Accord as we did. The trouble is that measuring credit risk is hard. Experienced bankers and examiners can usually distinguish a good loan from a bad one, but quantifying the level of risk on a portfolio basis and bank-wide is quite a different matter. Much attention has been devoted to the exercise within the industry and among bank supervisors, but no solution is at hand. Best practice banks and early research at the Federal Reserve suggest that significant strides are being made in credit risk management, but the industry and regulators still have a long way to go. It is--and should be--the highest priority for the industry and the supervisors. Last year, as you may recall, the United States and the other countries represented on the Basel Supervisors Committee adopted new capital requirements for trading activities that are based on a bank's internal measure of "value at risk." That regulatory amendment represented an important shift in regulatory thinking and a greater willingness by the regulatory community to build on risk management practices of banks. With market risk, though, the basic elements of the "value-at-risk" measure were relatively well established, although most institutions still needed to strengthen certain aspects of their calculations and management processes. In that exercise, the necessary data for identifying current trading positions and measuring the historical volatility of their market values were also generally available. The mark-to-market process and short horizon of daily trading also helped greatly in evaluating the effectiveness and overall "accuracy" of the market risk models. None of these crucial elements exists today for measuring credit risk, and industry practice has not yet converged around a particular measure of credit risk, or even a conceptual definition of credit loss. Some models, for example, identify a loss only when a borrower defaults, largely reflecting the view that the bank will hold the asset until it matures. If the model forecasts a default during the relevant time horizon, it then calculates an expected loss, or "loss rate, given default." Other models take more of a mark-to-market approach, recognizing the gains or losses in the economic value of a loan portfolio resulting not only from defaults or expected defaults, but also from changes in the credit quality of a borrower or from different market and economic conditions. As you can sense, model structures and assumptions become crucial. Moreover, the fundamental input--a borrower's credit risk rating--can be highly subjective and is largely determined internally within the bank. Some borrowers have public debt ratings from recognized rating agencies, but most do not. Even a public rating needs to be translated into the rating schedule of each bank. This lack of credit risk data is a serious weakness, with even large banks lacking enough historical default experience for a given borrower type to determine appropriate capital charges without substantial judgmental input. The subjective and variable quality of risk ratings, lack of historical data, and the long time horizon before answers are known about a portfolio's underlying strength make validating credit risk models a difficult task. If more risk-sensitive models are to be used for regulatory capital standards, these differences become more important because they can have material effects on competition and on the safety and soundness of banks, both domestically and abroad. Moreover, unlike trading activities, where the related capital requirements represent a small part of the total, credit risk counts. We need to get this measure right for obvious reasons. "Getting it right" means also providing the proper risk management incentives to banks. Far more needs to be done in measuring and managing credit risk than has been done so far by U.S. and foreign banks. As I noted, much progress has been made in recent years, make no mistake. But much more progress is necessary before most large banks, themselves, can gain a solid grasp on their risk exposures for risk management purposes, let alone before supervisors will be able to substantially revise the Capital Accord. In recent months Federal Reserve staff visited a number of large money center banks to understand better what role credit risk models perform now in senior management's internal assessments of the institution's capital adequacy. While, again, progress is being made, the results were somewhat disappointing. Nearly all institutions indicated that in their own internal reviews, they focused largely on factors such as their targeted external credit rating and their regulatory risk-based capital ratios relative to those of their primary competitors. If these figures were in line, they generally viewed their capital as adequate. Although the targeted and actual ratios were significantly above regulatory minimums, these responses were disappointing, indeed. It should be noted that a key ingredient in rating agency evaluations of bank capital is the risk-based capital ratio. While we regulators are flattered by the use of our capital standard for internal and marketplace analysis, we must emphasize that the well-known shortcomings of the standard make it an inappropriate tool for many internal and market purposes. We expect institutions to be ahead of regulators in this analysis, not the other way around. Where models are available, generally pertaining to commercial credits, they are used principally in setting concentration and exposure limits, pricing, and evaluating performance on a risk/return basis. Important uses, for sure, but in no case did management indicate their risk measures offered a significant input to evaluating the institution's overall capital adequacy. This assessment is not intended to be pessimistic. I believe significant progress can be made if sufficient attention and resources are devoted to the effort, and if the industry is given the right incentives to make it work. Supervisors can provide some of the incentives--both the carrot of an improved capital standard and a better risk management process, and the stick that management will be judged, in part, by its ability to quantify risk. In large part, virtue can also be its own reward. Banks that effectively measure and manage risk will make and price credit better. New Capital Proposal As I mentioned earlier, the Basel Committee on Banking Supervision has now released its long awaited consultative report on revisions to the 1988 Basel Capital Accord. For the largest institutions, the Accord has increasingly been weakened by the changes that have occurred in financial markets. Most importantly banks here and abroad have been engaging in capital arbitrage techniques designed to move their higher quality, lower-risk assets to securities markets, sometimes reducing their capital charges on these assets more than proportional to the retained risk positions. In addition, the remaining higher-credit risk assets have the same regulatory capital charges as the lower-risk assets that have been securitized, changing the meaning of the resultant capital ratio. For these and other reasons, the 1988 Accord has become increasingly undermined and the risk-weighted capital ratios have become more difficult to interpret. Modifying the Accord is an incredibly complex and difficult procedure, not only because it must be negotiated among 12 nations and affect the policies of many more, but also because the issues are so difficult. As I noted, the underlying approach has three equally important legs, all of which reflect efforts to respond to the evolving changes in financial markets. The first leg is modification of the capital Accord per se, especially for the large complex banking organizations, the most important of which will be to change the risk-weighting scheme on portfolio assets. The framework calls for moving from a four-weight scheme--with most of the assets at one weight regardless of risk--to multiple and more sensitive risk weights and also includes steps to curtail loopholes dealing with securitization transactions. The risk weights, in turn, would be based perhaps on one or a combination of techniques: external ratings, internal management risk ratings, and/or bank-specific formal risk models. Please note that in each of these, the process is leveraging off the market's risk evaluation, including what the bank management applies for its own purposes. Consistency and improvement in bank risk management is thus a prerequisite to improved, and more rational, capital regulation. The second leg is increased market discipline. Market discipline, of course, can occur only to the extent that the banks make information available to creditors and counterparties that have the ability to respond to that information. Thus, the consultative document contemplates more transparency about bank risk-taking and controls so that creditors and counterparties can decide more rationally about their required compensation for the risk of dealing with that bank. Of course, the objective is to create the incentives for more rational and efficient risk taking by the bank. The final leg is supervisory review of the capital adequacy of the banking organization. The purpose is twofold: first, to ensure that a bank's capital position is consistent with its overall risk profile and strategy and, second, to encourage early supervisory intervention. The purpose of this review is to provide supervisory comfort that each bank's internal process for assessing its capital adequacy, and that each bank's actual capital levels, are consistent with the scale and complexity of its risk taking activities. In some cases, these reviews may well result in requiring individual banks to hold more capital than the minimum regulatory standard. As we think about capital standards for the years ahead, it seems appropriate to consider a more bifurcated approach: one standard for large, complex institutions; another for most other banks. That direction seems especially necessary if we do pursue a more sophisticated, risk sensitive measure of credit risk to capture developments and techniques at the larger and more complex banking organizations. My sense is that greater complexity would be unnecessary for community banks, where a simpler, less burdensome approach may be quite satisfactory for supervisory purposes for most banks. Nevertheless, all banks should take to heart the message the Federal Reserve and other supervisors are sending about the need for stronger practices for evaluating credit risk. Loan Loss Reserves On the topic of capital, I would like to say a few words about loan loss reserves and the interaction of the Federal Reserve and other federal banking agencies with the Securities and Exchange Commission. In recent months, as you know, the Commission has devoted increased attention to practices of large U.S. banks in setting their level of loan loss reserves. The issue surfaced last fall, when SunTrust was required to reduce its reserves and revise previous financial statements. Since then, several other institutions have been asked to explain their reserve practices to staff of the SEC. As supervisor of these holding companies, the Federal Reserve has been actively involved in this matter from the outset and has urged the Commission to work with us, with the institutions, and with the other banking agencies toward a satisfactory resolution. Obviously, this means reconciling different perspectives on this issue. For example, in light of increased volatility and banking risks in recent years, the banking industry has appropriately maintained robust reserving practices and levels. From a safety and soundness perspective, the Federal Reserve and other banking regulators have expected institutions to maintain strong loan loss reserves that are conservatively measured. In carrying out its responsibilities, the SEC has emphasized the need for financial statements and reported earnings to be transparent and, therefore, for allowances to be adequate but not excessive. Enhanced transparency has also been a critical objective of bank regulators, both domestically and internationally. Last week, press reports characterized the Fed's position as being different from that of other federal banking agencies. That is not true. The main point of contention between the banking agencies and the SEC appears to be whether the recent guidance issued by the Financial Accounting Standards Board (FASB) represents a mandate to reduce reserves. The Federal Reserve has worked with the SEC to issue guidance emphasizing that the FASB guidance does not mandate any material change and that bank management should feel free to maintain reserves at the high end of a reasonable range. The Federal Reserve's policy guidance provides background information that is intended to assist institutions and their auditors in understanding the SEC announcement and the FASB article in the broader context of other accounting initiatives and discussions between the SEC and the Federal Reserve on allowance accounting matters. Moreover, our policy letter sends a clear message that the Federal Reserve wants banks to maintain prudent reserving practices and not to over-react as a result of a narrow interpretation of the FASB guidance. The other banking agencies appear less sanguine about the intent of the FASB guidance and have registered protests on Capitol Hill. The Federal Reserve will continue to work with the SEC and the accounting profession in the months to come in providing further information regarding appropriate documentation and other matters. I understand Richard Spillenkothen, the Federal Reserve's Director of Banking Supervision and Regulation, will also speak to this topic in his comments at lunch. Disclosure and Market Discipline Although we disagree with the need for banking organizations to revise previous financial statements, battling the SEC on many of these issues seems not the proper course. They have an obligation to enforce sound reporting and disclosure practices as best they can, and our financial markets have been well served in the process. The U.S. banking industry has its obligations, too, to manage its risks and to tell its story to bank supervisors, the SEC, and the general public. If for no other reason than the fact that banks today are so large and complex and have the potential to present such widespread risk, these largest institutions, in particular, should be held to high performance and compliance standards. As bank supervisors, we should welcome the market's help to identify and assess banking risks and to minimize the risk of moral hazard. One approach the Federal Reserve is exploring would enhance the role of investors in bank or bank holding company subordinated debt. Unlike shareholders, who benefit from any gains from excessive risk, subordinated debt holders have only downside risk. As a result, their incentives are similar to those of supervisors and the bank insurance fund: they lose if the bank defaults but they don't participate in outsized gains. A difficulty, however, in creating a greater role for subordinated debt is determining how to provide investors with adequate and timely information about a bank's risks and with sufficient leverage to affect management decisions. From the supervisor's perspective, another difficulty is separating market "noise" in changing yield spreads from meaningful signals they may provide. We will be collecting and analyzing these data in the months ahead and will be evaluating their potential usefulness, both as a tool for supervision and as a market mechanism for providing feedback to banking organizations. Whether or not the exercise proves fruitful, it points in the right direction--providing incentives for greater market discipline and for sound management of banking risks. Conclusion In closing, I would remind you that we are beginning to see slippage in important indicators of industry strength. Though still low by historical standards, the volume of nonperforming assets increased last year for the first time since 1991, with the deterioration concentrated within commercial and industrial loans. Delinquencies in agricultural loans have also risen, as a result of extremely weak markets for many farm products. Continued weakness in much of this sector could begin to weigh on some community banks. The next stress-point for any particular bank may come from poor credit quality, from structural and competitive pressures within the industry, or from many other sources. Fortunately, the U.S. commercial banking system has demonstrated a great deal of strength and resiliency in dealing with challenges of the past, and it still seems as strong and as well positioned overall now to handle stress as it has been in many years. I have no doubt that the U.S. banking system will continue to grow and that it will remain central to the nation's financial system. To do that, though, requires that we all to adapt to changing times and that banks manage risk carefully in both good times and bad.

Return to top 1999 Speeches