Remarks by Douglas E. Harris Senior Deputy Comptroller before the International Swaps and Derivatives Association 11th Annual General Meeting

Thank you. First of all, I would like to say that I am very honored to have the opportunity for the second time in as many years to address ISDA's Annual General Meeting. The work that is done by my group at the Office of the Comptroller of the Currency (OCC) most directly affects the banks that are ISDA members. But it is also the case that the issues that we address, and the policy positions that we develop, have an affect on ISDA's non-bank members. Therefore, the opportunity to come before you and explain our positions and new initiatives and to get your feedback is invaluable.

My topic this morning is OCC's revised Examiner Guidance for Financial Derivatives. We are still developing that Guidance and at this point I cannot give you a firm timetable for its release, although sometime in the second quarter is likely. Nonetheless, I would like to talk to you about some of the modifications and additions that we are considering at this time to our existing Guidance.

Before I discuss the guidance, however, I would first like to talk to you about what I consider to be one of the more important documents that we have released over the last two years: a paper entitled "Trading Activities at Commercial Banks", which was released in December, 1995. That paper was prepared by the staffs of the OCC, the Fed and the FDIC, with input from the SEC and the CFTC.

Many of you will remember that in April of 1994, the House Banking Committee held a hearing on hedge funds' trading activities. At the hearing, George Soros, the principal manager of The Quantum Group of Funds, very deftly turned some of the concerns and criticisms directed at hedge funds toward commercial banks. He claimed that banks were engaged in the same sort of activities as hedge funds under the name "proprietary trading," and that they did so with the support of federal deposit insurance.

Then in May, 1994, the U.S. General Accounting Office (GAO) released a report to Congress entitled "Financial Derivatives: Actions Needed to Protect the Financial System." Included in one of the GAO's Recommendations to Congress was the following statement:

"Gaps and weaknesses in OTC derivatives regulation clearly demonstrate that the existing regulatory structure has not kept pace with the dramatic and rapid changes in the domestic and global financial markets that have occurred over the past several years. Banking, securities, futures and insurance are no longer separate and distinct industries that can be well regulated by the existing patchwork quilt of federal and state agencies. Many issues need to be debated and decided, including the appropriate uses of federally insured deposits and the extent to which they should be used to finance large-scale proprietary trading in derivatives or other financial instruments."

In the following months, several members of Congress, including Senators Dorgan and Riegle, introduced bills, some of which would have forced either all derivatives activities or all proprietary trading out of commercial banks and into separate subsidiaries or affiliates. One of the things that should have been obvious to many of you was that there was considerable confusion, especially in Washington, between proprietary trading and derivatives trading, with many unable to distinguish between the two.

In a July 18, 1994, response to Congressman Dingell regarding the GAO's recommendation, then Treasury Secretary Bentsen said:

"The Treasury, which includes the [OCC] and the [OTS], and the Federal Reserve do not perceive the risks associated with proprietary trading, if properly managed, to be inherently greater than those associated with other banking activities ... Nevertheless, the central role that banks play in the economy, and the fact that federally insured institutions are engaging in these activities, raise public policy issues. As a result, federal banking regulators are devoting further attention to this area."

During the summer of 1994, the three federal banking agencies undertook a study to address many of the questions raised about banks' proprietary trading activities: "What is proprietary trading? How large is it? How widespread is it? Who engages in it? What are the risks? How are they managed and controlled? Are current risk management processes sufficient?" Underlying these questions were concerns that revenues from trading activities can be more volatile than those from other banking activities; that trading activities may produce sudden and sizeable losses that could threaten the solvency of the bank; and that, therefore, proprietary trading may be an inappropriate activity for a federally insured depository institution.

The paper we produced is largely an educational document full of interesting information (much of it in the appendices), much of which had not previously been compiled in this form. Though our original intent was to study and report on proprietary trading activities, the paper is also an overview of commercial bank trading activities generally. It discusses trading instruments and trading methods; the growth and involvement of banks in those activities; and the risks and benefits posed by trading; how those risks are managed by the banks; and how those risks are supervised by the regulators.

One of the first issues we address is "what is proprietary trading?" and "how do you distinguish it or can you distinguish it from other types of trading activities?" We found that trading occurs along a continuum. Though we were able to make distinctions between each of "matched trading", "market-making," "positioning" and "proprietary trading," in practice, the lines between them became blurred. In fact, only a handful of the major trading banks formally segregate their proprietary trading activities.

What about the risks that arise from trading activities generally and proprietary trading, in particular? Not surprisingly, we found that these are the same risks that banks face in other bank activities and products — credit risk, price risk, liquidity risk, transaction risk, compliance risk, strategic risk and reputation risk. Though the significance and scope of these risks may vary from bank to bank, it is the significance of price (or market) risk that most clearly distinguishes trading from other bank activities and proprietary trading from other types of trading. And it is market movements in either price or volatility that are the primary source of revenue in proprietary trading.

On average, the size of bank trading operations relative to bank capital is small. Most of the major trading banks now maintain their daily earnings-at-risk for all trading activities at less that 3% of Tier I capital, although there are a couple of major exceptions. As I mentioned earlier, very few institutions formally segregate their proprietary trading operations. Those that do typically have sublimits for proprietary activities of less than 1% of Tier I capital. A bank's aggregate trading limit generally reflects a bank's overall appetite for risk. It forms an upper bound for aggregate exposure. However, low loss experiences and on-site examinations findings by our resident examiners at all the major national bank dealers indicate that such aggregate trading limits are rarely exceeded.

So, has this market risk been realized in the form of sudden and sizeable losses? Historical data indicate that trading revenues have been less volatile than is generally perceived. We found that, despite some short-term volatility from quarter to quarter, the major dealer banks have had successful long-term trading results over several years. Trading revenues were consistently positive for the seven major dealer banks over the 44 calendar quarters from June 30, 1984, to June 30, 1995, except for six instances of losses. One bank accounted for four of those instances of losses. And, even in the quarters in which losses were reported, the size of the losses were minimal compared to the quarterly revenue of the bank.

We also found that trading revenues tend to exhibit cyclical components — components which appear to be consistent with business and interest rate cycles. (This is also the case with earnings from securities.)

A finding that raises some concern for us supervisors is that trading revenues also appear to be highly correlated across the major dealer banks — when trading revenues are down for one of the major trading banks, trading is likely to be down for most of the others. The possibility that a large number of financial institutions would simultaneously experience declines in revenue certainly has systemic implications. It also raises the specter that any one bank which has other dealer banks as significant counterparties could face a sudden and significant broad-based decline in counterparty credit quality.

We also found that banks are generally measuring, monitoring and controlling well the risks arising out of trading activities — and that includes price risk. They use a number of complimentary methods to accomplish this. This is consistent with the information that we at the OCC have compiled over the last couple of years with respect to national bank compliance with BC 277, our guidance for risk management of financial derivatives.

With respect to risk measurement, individual banks tended to employ different market and mathematical assumptions based on their experience and theoretical perspectives. Many of the largest dealer banks have already moved to a value-at-risk system for measuring price risk. They often supplement these systems with notional or par limits. Virtually all banks further supplement these risk measurement and control mechanisms with loss control limits or management action triggers.

Overall, we found that banks' trading activities provide substantial advantages to banks, their customers and to the markets. There is no doubt that the primary reason banks trade is to generate revenues. They have done this quite successfully. For the 11 largest dealer banks, average trading revenue has grown from 5.85% of total revenue in 1989 to 12.80% in 1994. However, in the process, they have also diversified their revenue sources.

At the same time, the growth in trading activities has allowed banks to diversify their products and services. This allows them to strengthen existing institutional and "high" net-worth customer relations and to be competitive in attracting new customers. Of course, the customer also has the ability to shop around for the product that best meets the customer's risk management or investment needs at the best possible price. That is taking place as these customer needs have been growing and evolving.

As you all know, traders provide a significant benefit to their banks by obtaining and providing first-hand knowledge of the current market levels, magnitudes and directions of movements in interest and exchange rates, as well as commodity and equity prices. This information is essential for making accurate pricing decisions.

Traders are also often the first to hear and spread market rumors. These rumors can often be vital to an institution in monitoring and controlling its various risks. After the Barings episode last year, the OCC went to the largest national bank dealers and inquired about any problems with their risk management systems that episode had revealed. We also inquired as to whether they planned to make any changes in those systems. In response, one theme that was repeated was that many traders had heard rumors of the size of Barings positions on the exchange-traded markets and of potential problems with those positions well before Barings went under. However, this information had not been communicated to senior management and the corporate office. As a result, some banks were unable to take appropriate defensive measures to limit any exposure. (As it turned out, no U.S. bank was significantly affected in the long-term by Barings' demise.) These banks said that they will attempt to further encourage, if not formalize, the flow of gossip, rumor, and information from the trading floor to senior management.

It is also clear that some banks have benefited from the new and evolving techniques used to manage risk in trading, in general, and derivatives, in particular. The benefit has been in applying these techniques to other bank products and activities. For example, some banks have extended portfolio-based market risk management techniques to the management of the credit risk in their loan portfolios.

The size and diversity of commercial bank trading activities also contribute significantly to the overall depth and liquidity of the cash and derivatives markets. And there is anecdotal evidence to suggest that derivatives-related trading allows markets to adjust more rapidly to changing economic conditions.

The staffs of the Fed, the FDIC and the OCC caution readers that different results may occur in different economic environments. We also caution that a successful history is no guarantee that a sizeable and sudden loss may not occur in the future. But we can conclude that:

• the market risk from trading activities has resulted in less volatility to dealer banks' earnings than has credit risk from lending activities
• trading activities provide an important source of revenue (and a diversified source of revenue) for the banks involved in those activities;
• trading activities allow banks to enhance customer relationships and benefit bank customers and the markets; and
• the risks from trading activities appear to be manageable as long as senior management and the board are appropriately informed about, and committed to addressing, risk-related issues that arise from those activities.

It is clear, however, from the incidents at both Daiwa and Barings that all bets are off in the case of fraud, just as with any other banking or commercial activity.

I would now like to turn for a few moments to the OCC's forthcoming revisions to our existing Examiner Guidance on Financial Derivatives. That original guidance was released in October, 1994. We have decided to revise the guidance at this time in order to:

• consolidate our previous guidance on trading activities;
• conform the guidance with the OCC's new Supervision-by-Risk program, and specifically to conform to risk definitions used in that program; and finally
• to expand and clarify the existing guidance based on events over the past year and a half and comments from our field examiners who use this guidance in their examinations of national banks' derivatives activities.

Though no final decisions have been made at this time, I would like to mention a few of the issues we are considering addressing and I would be happy to have your feedback.

First, we are aware of increasing pressure on banks to deal with undisclosed (or blind) counterparties. I have been a little confused as to why they are called "blind" counterparties. The counterparties know exactly who they are dealing with; it is the banks that are in the dark. The OCC is concerned about the credit, legal and reputation risks that are inherent in this activity and is considering guidance that would:

• urge banks to obtain legal opinions on the enforceability of any written "customer" agreements;
• warn banks of the possible conflict between dealing with undisclosed counterparties and complying with local money laundering regulations; and
• discourage banks generally from dealing with unnamed counterparties unless certain types of controls are established. Controls might include (i) restricting transactions to an approved list of counterparties; (ii) limiting the size of transactions with unnamed counterparties individually and in the aggregate; (iii) limiting transactions to very liquid contracts.

We are also considering additional guidance on the importance of stress-testing derivatives portfolios on a regular basis, and conducting stress scenarios that produce the greatest losses or exposure. We know that many banks use a large market move in their stress scenarios. However, this movement may not expose the portfolio's greatest vulnerabilities. The more sophisticated and better managed banks will identify the environment that produces the most undesirable results and estimate the probability of their occurrence. These institutions will also provide the results of stress-testing, along with major assumptions, to senior management and the board of directors on a periodic basis.

Over the last year and a half, the issue of risk models is one that has attracted considerable attention. We are considering additional guidance on the importance of validating, backtesting and recalibrating risk measurement models on a regular basis. Included in that discussion would be a discussion of the limitation of value-at-risk models.

Finally, we are considering expanding our discussion of revaluations. Accurate market values are the key to the production of meaningful reports regarding risk levels, profitability and market trends. A bank can take different approaches based upon the liquidity and complexity of the contract. We may suggest that banks establish policies which (i) specify required valuation adjustments, (ii) require documentation of the rationale supporting certain valuations, (iii) require a periodic review of assumptions, and (iv) provide for proper accounting treatment.

As you can see from the issues that we are considering addressing, we continue to learn from you. For the most part, our issues are your issues. We continue to look for the best practices at the best banks, and to impart that information to others either in the form of guidance or an advisory. And we are committed to modifying our guidance as necessary in order to keep it current. That means eliminating any elements of our guidance that become stale, and addressing new issues, such as unnamed counterparties, as they become more visible and of greater concern. And, as always, we welcome your feedback and discussion with you of these issues.

Thank you.