Remarks by Julie L. Williams First Senior Deputy Comptroller and Chief Counsel Office of the Comptroller of the Currency Before the Robert Morris Associates Lending and Credit Risk Management Conference New Orleans, Louisiana October 5, 1999 It is a real honor and genuine pleasure to again have the opportunity to address an RMA conference, and to talk again about credit risk developments in the banking industry. As you probably are aware, the OCC has just released its latest annual credit underwriting survey. The 1999 survey results show some positive trends from previous years' survey findings, as underwriting standards for commercial loans appear to have strengthened in several significant categories. I am happy to be able to report these improvements, but they are not cause for complacency. The aggregate data from our underwriting survey does not tell the whole story. Beneath the statistics of the survey, there are some troubling trends that require bankers' attention. Easing continues in some important loan markets. And today's improvements cannot mitigate the overhang of credit risk created by weaker standards used in previous years. To illustrate this concern, I'll be talking about the findings of our credit-focused exam work, including our "Ugly Loan" review. This work provides important perspectives and depth to understanding the credit underwriting picture we see today. Finally, I want to talk about what you, as bankers, can and should do to manage and mitigate your institutions' credit risk. First, the results of the OCC's 1999 credit underwriting survey reflect some positive developments. For the first time in five years, the survey indicates that underwriting standards for commercial loans have strengthened, with examiners reporting that 25 percent of the surveyed banks tightened standards, compared to just 4 percent in 1998. Examiners also reported that only 13 percent of surveyed banks eased standards, compared to 44 percent in 1998. Overall, firming up was evident in syndicated/national and international lending, areas that appear to be most affected by the credit events in the fall of 1998, and in agricultural lending, which reflects low commodity prices and weakened export demand. While we're encouraged to see strengthened underwriting standards in these areas, easing of standards persisted for some commercial products -- commercial real estate and middle market lending. This is of concern because these two categories represent a large portion of bank loan portfolios. Further, a large percentage of banks reported either a mixture of easing and tightening, or no change from last year. While this indicates standards did not deteriorate, we must keep in mind that "no change" also means no improvement. Increased pricing was the most prevalent method used to tighten standards (56 percent), followed by adjusting maximum credit availability (33 percent), increasing collateral requirements (31 percent), and strengthening loan covenants (28 percent). For those banks strengthening - standards, examiners reported economic outlook and a change in risk appetite as the primary reasons. The story has parallels on the retail side. Strengthening continued for the third successive year with slightly more banks tightening standards at 27 percent, than easing (22 percent). Products where tightening was most pronounced included consumer leasing, indirect consumer loans, credit cards, and other direct consumer loans. This seems appropriate as these products have been particularly vulnerable to rising consumer delinquency and charge-off rates in recent years. But, among the retail products, the trend toward easing standards for home equity loans continued for the third successive year. By definition, a high LTV home equity product represents an easing of underwriting standards, and poses greater risk of default and risk of loss in the event of default. Those banks that strengthened underwriting standards for retail products most commonly used scorecards or pricing and, for real estate-related products, enhanced collateral requirements. Most banks tightening standards did so because of a change in risk appetite or market strategy. Perhaps the most worrisome finding in the 1999 survey is the continued increase in credit risk reported for all commercial and retail products. For the fourth successive year, examiners report increased inherent portfolio risk, and they expect risk to continue to increase over the next twelve months. Overall, the modest changes in risk tolerance and the additional emphasis on risk-based returns for commercial credit highlighted by this year's survey are certainly trends that we would all like to see continue. But, we all recognize that competitive pressures of the marketplace will continue to exert pressure, and that continued strengthening of underwriting standards in this economic environment is very challenging. Moreover, positive news this year about underwriting standards does not erase the increased risks embedded in bank portfolios caused by four successive years of easing commercial loan underwriting standards. Nor, in the face of this year's results, can we ignore that banks shoulder increased portfolio credit risk not only as a legacy of weaker underwriting standards employed in recent years, but also as a result of the more limited credit risk selection options available to them. As the better quality loans continue to migrate to the capital markets, bankers are left to compete with other lenders for less desirable, riskier paper, which increases the risk of default. In response to all these developments, the OCC has been implementing a series of modulated supervisory responses. These actions commenced with early warnings and advisories. Last year, when I spoke at the RMA conference in Chicago, I announced a series of calibrated actions designed to further focus examiner and banker attention on signs of deteriorating credit standards, and to highlight the need for bankers to implement risk evaluation and management systems commensurate with the increasing credit risks they held. We instructed our examiners to identify loans with specific types of structural weaknesses in reports of examination and factor those weaknesses into their judgments about the quality of individual credits, portfolios, and the quality of portfolio risk management. Our subsequent findings give rise to concerns. Particularly at our largest banks, we found significant numbers of loans containing various structural weaknesses. The most frequently cited weaknesses include: * Non-existent, weak or waived covenants; * Indefinite or overly liberal payment terms; * Inadequate financial analysis; * Insufficient collateral support; * Elevated leverage ratios; and * Repayment dependent on highly optimistic or undemonstrated cash flows. Loans with such structural weaknesses were listed in the reports of examination or other supervisory communications with management and/or boards of directors, to ensure that they understood the nature and extent of the weaknesses disclosed and their impact on our judgments on portfolio quality and management. Simultaneously, we formed a team of our best credit experts and instructed them to review loans across the population of our larger banks in order to better identify examples of the types of structural weaknesses that we were seeing. Unfortunately, the team had no trouble finding example after example. Based on this review -- which came to be nicknamed the "Ugly Loan Project" -- we developed and delivered a focused training program designed to advance the credit risk evaluation and classification skills of our examiners and clarify our expectations about how structural weaknesses should be incorporated in their judgments of credit risk in individual loans and portfolios. We also communicated these efforts to the CEOs of our largest banks. These efforts coincided with the interagency Shared National Credit Review, which recently concluded. As you all probably know, the Shared National Credit Review is jointly conducted by the OCC, FRB and FDIC each Spring. That review covers all commercial loans that are shared among three or more banks and exceed $20 million. This year's program covered approximately 30 percent of all commercial loans in the U.S. banking system. The results of this year's SNC review have shown a marked increase in classified and special mention credits compared to 1998 -- an increase of roughly 70 percent in the dollar value of adversely rated loans. Fifty-nine percent of the adversely rated loans had been rated "pass" last year. Of particular concern, 14 percent of the adversely rated loans were to new borrowers. In other words, banks are booking new loans that are weak at their inception. These results reflect the increasing portfolio risk building up in the system as a result of weakened underwriting and risk selection standards. Thus, while our most recent credit underwriting survey indicates that banks are firming up their underwriting standards, our exam work also tells us that banks have heightened embedded credit risk as a result of previous years' underwriting practices. What should bankers be doing about this? Let me offer some suggestions and cautions. We all know that risk-taking is inherent in the business of banking and that exceptions are an expected component of the underwriting process. We expect that you will underwrite a certain number, hopefully not too large, of loans with exceptions, but we also expect that you will have good controls over and intelligence on changes in the underwriting standards applied in your banks. This requires a system that routinely identifies the aggregate level of loans that do not meet your own underwriting criteria, segmented by portfolio and type of exception. Over he years, many bankers have resisted the need to identify and track exceptions to their risk selection and underwriting policies; however, recent studies indicate that those that do routinely track exceptions and analyze their implications for portfolio risk are among the better performing banks. The need for such information is precisely why OCC initiated examination procedures to identify loans with structural weaknesses last year. Indeed, a number of the banks we supervise were appreciative of the detailed work performed by our examiners to identify "ugly" loans. Because examiners identified, reported and discussed the specific underwriting weaknesses they found as well as their risk implications, bank management and boards were better informed and could more readily determine whether weaknesses represented isolated events or common occurrences. Bankers found examiner's feedback regarding specific underwriting weaknesses to be much more valuable than general expressions of concern about deteriorating underwriting standards. As a result, we will continue to perform these procedures in every safety and soundness examination, and we will continue to urge bankers to adopt the exception reporting processes. Another critical component of sound portfolio risk management hinges on the integrity of banks' own internal risk ratings. Many banks are making substantial progress, setting higher internal standards for accuracy and using credit models to back-test and cross-check the integrity of their internal ratings. We applaud these advances, but our recent examination activity, including the 1999 SNC review, has made clear that most banks have room to improve the accuracy and timeliness of their problem loan identification. Even the most sophisticated banks can improve the level of precision and reliability of distinctions among various risk levels of "pass" rated credits. This becomes particularly imperative for the additional reason that today consideration is being given to using those internal risk ratings to set components of banks' capital requirements. Unfortunately, in too many banks today, risk management advancements and MIS to provide timely risk analysis information to senior management and the board of directors are not being pursued with the same vigor as new business and revenue generation opportunities are being pursued. In practical terms, this is akin to ignoring termites in the foundation of your house while you're busy adding a sunroom to the second floor. We also are concerned with the continuing relative decline in the ALLL. The average ratio of the ALLL to loans for all national banks is at its lowest level since 1986. That ratio, which was maintained above 2 percent for all years between 1987 and 1995, dropped to 1.82 percent at the end of the second quarter. That is why the OCC and the other bank regulatory agencies were extremely concerned earlier this year that certain actions and pronouncements by the Securities and Exchange Commission last year and earlier this year would have the unintended effect of prompting unwarranted reductions in banks' reserve levels. We applaud SEC Chairman Levitt's statements clarifying the SEC's positions in this area, and we continue to strongly believe that given the higher risk inherent in bank portfolios today, it would be more appropriate for banks to be increasing their reserve coverage, rather than decreasing it. One particular credit issue that caught our attention during our efforts of the past year has been the evolution of leveraged finance. Earlier this year, the OCC issued guidance to bankers advising them of the risks in leveraged lending activities and the need for more intensive risk management of these types of credits. This guidance was in keeping with our measured responses to loans with structural weaknesses, but also reflected our growing concerns with the developing trend toward higher leverage in the lending markets overall. In that guidance, we expressed a number of concerns about the risk characteristics involved with today's leveraged lending activity. As a result of our more recent exam work, however, our concerns have particularly escalated about the increasing reliance on so-called "enterprise values" to justify collateral shortfalls common to many leveraged financing packages. The premise behind banks' reliance on "enterprise values" as a secondary source for repayment of a loan is that the borrower's business can be sold or refinanced at an amount that can be estimated based on a projected earnings capacity. But, the interagency Shared National Credit Review revealed that too often documentation and analytical support for enterprise values are inadequate. Let me give you some examples of enterprise value lending situations that concern us: The first example is a loan made to finance an acquisition. The loan was "collateralized" by all assets of the company; however, the bank group did not value the assets, but instead relied on the estimated enterprise value. Optimistic earnings and cash flow projections never materialized. Market prices for the company's product declined and now the company is closing plants and attempting, thus far unsuccessfully, to sell assets. The acquisition price now appears inflated as does the "enterprise value" used to underwrite the loan. The potential for substantial loss to the bank group is considered very high. Lending against valuations based on multiples of stabilized cash flow valuations can be dangerous. In another example, the borrower is a highly specialized manufacturer who sought funds for working capital and growth. Accounts receivable, inventory and fixed assets only covered 60 percent of total bank debt. The balance of the loan was supported by a stabilized cash flow enterprise valuation that provided a modest 10 percent cushion. The borrower's product has met reduced demand and increased competition. Cash flow has been severely reduced -- it was inadequate to service debt in 1998 and is negative YTD 1999. A final example is a consumer goods manufacturer that required re-capitalization. Although the bank group took a lien on fixed assets, their value was considered minimal. To justify the loan, he bank group relied on an enterprise value based on anticipated refinance or sale of the borrower. Unfortunately, product demand has weakened, the borrower has lost a number of key customers, and operating margins have shrunk,. Cash flow is no longer adequate to service debt. The likelihood of a refinancing or sale of the company that will rescue the banks from the loan is bleak. Thus, the risk of loss is very high. Stepping back from these particulars, this type of transaction raises several fundamental concerns: Our first concern arises from the interdependence between the primary and secondary sources of repayment when "enterprise value" is the secondary repayment source. Enterprise value is often a calculated multiple of the borrower's historical or projected cash flow. As long as cash flow is sufficient to meet the repayment terms of the loan, there's no problem; but if cash flow does not materialize as projected, both the primary and secondary sources of repayment of the loan are at risk. The same events that endanger the first source of repayment -- cash flow -- also will endanger the second source of repayment -- the enterprise or going concern value of the business. Our second concern lies with loan officers' and loan review officers' apparent level of comfort with "enterprise value" as a substitute for real collateral. "Enterprise value" is a volatile, disappearing intangible, inherently prone to vanish when it is most needed. It is not the equivalent of hard assets and equity, properly assigned and hypothecated, valued under normal and stressed conditions, with reasonable margins at or above the loan amount. There is a reason why market professionals refer to the enterprise value component of the support for a loan as the "airball." Our third concern in this area is with the increasingly equity-like appearance of many leveraged loan structures and whether banks are appropriately recognizing and mitigating the risks inherent in these loans. In addition to relying on enterprise values, many such loans also evidence weakened structures that diminish a bank's controls, rights and collateral protection, extended maturities that defer meaningful repayment, and reliance on refinancing or re-capitalization of the borrower as a primary source of repayment. As we said in the leveraged lending advisory we issued earlier this year, bankers need to be able to identify the amount of leveraged loans in their portfolios, and should also be able to identify those loans where the financing gap -- the "airball" -- between the value of assigned collateral and the amount financed relies on enterprise value. Bankers should carefully analyze the assumptions used to support underlying enterprise values and be realistic in challenging those assumptions under normal and stressed conditions. Where identified financing gaps are large relative to the overall portfolio, the bank's allowance, or its capital, bankers should be developing plans to reduce these gaps in an orderly manner and to augment the allowance or capital as necessary. Bankers must not assume that economic conditions and the capital markets will always be conducive to a refinancing or re-capitalization needed to close the financing gap when a loan comes due. I particularly hope that some bankers are not complacent because they implicitly assume that their "section 20" securities affiliate will be available to arrange a market transaction that will extract the bank from the loan. Where a bank has such an affiliate, it must not only guard against that assumption infiltrating its credit decisions, but also ensure that its conflict of interest policies are up to date and vigorously enforced. With the prospect of "financial modernization" legislation that will enhance the ability of banks and securities firms to be affiliated, it becomes essential that bankers assure that their credit decisions are not being shaded by the prospect that they will be rescued from questionable loans by their securities affiliate. Bankers that will be successful -- in good times and bad -- will be the bankers who recognize risk earliest and move decisively to control it and reduce it where warranted. This can only be done with good intelligence -- meaningful and accurate MIS, reliable and timely risk ratings, effective and objective internal loan review and, most importantly, credit risk management that is adept at assessing the data, the trends, and relevant market intelligence. Bankers should be asking themselves: "Am I comfortable that my bank's culture and credit controls, including my credit risk rating system, would prevent the problems bankers suffered in the late 1980's and early 1990's?" And bank CEOs, in particular, should be asking: "Are my independent risk identification and control functions telling me what I want to hear, or what I need to hear? Downsizing to improve efficiency ratios and efforts to achieve merger-related synergies must not reduce the effectiveness of your credit risk control processes. Given the higher risk inherent in bank portfolios today, bankers need to be more skilled at quantifying and mitigating the credit risk they undertake to be pursued with the same enthusiasm as new business and revenue generation activities. And, risk management functions need the independence and empowerment to candid about risk in the portfolio. You in this room, our industry's credit risk managers, face all these challenges. I urge you to take them up with vigor. Your efforts will be key to the industry's stability and success as we enter the next century.