Dealers and end-users should use derivatives in a manner consistent with the overall risk management and capital policies approved by their boards of directors. These policies should be reviewed as business and market circumstances change. Policies governing derivatives use should be clearly defined, including the purposes for which these transactions are to be undertaken. Senior management should approve procedures and controls to implement these policies, and management at all levels should enforce them.

Derivatives activities merit senior management attention because they can generate significant benefits or costs for any firm. A firm's policies for derivatives should be an integral part of its overall policies for risk taking and management, either in its underlying business (if it is an end-user) or in its other lines of business (if it is a dealer). Periodic reviews will help ensure that these policies reflect changing circumstances and innovations.

Recommendation 2: Marking to Market

Dealers should mark their derivatives positions to market, on at least a daily basis, for risk management purposes.

Marking to market is the only valuation technique that correctly reflects the current value of derivatives cash flows to be managed and provides information about market risk and appropriate hedging actions. Lower-of-cost-or-market accounting, and accruals accounting, are not appropriate for risk management.

The Survey of Industry Practice shows that the practice of marking to market daily is widespread among dealers, reflecting the importance of the information it provides to risk managers. Intraday or even real time valuation can help greatly, especially in managing the market risk of some option portfolios.

Recommendation 3: Market Valuation Methods

Derivatives portfolios of dealers should be valued based on mid-market levels less specific adjustments, or on appropriate bid or offer levels. Mid-market valuation adjustments should allow for expected future costs such as unearned credit spread, close-out costs, investing and funding costs, and administrative costs.

Marking to mid-market less adjustments specifically defines and quantifies adjustments that are implicitly assumed in the bid or offer method. Using the mid-market valuation method without adjustment would overstate the value of a portfolio by not deferring income to meet future costs and to provide a credit spread.

Two adjustments to mid-market are necessary even for a perfectly matched portfolio: the "unearned credit spread adjustment" to reflect the credit risk in the portfolio; and the "administrative costs adjustment" for costs that will be incurred to administer the portfolio. The unearned credit spread adjustment represents amounts set aside to cover expected credit losses and to provide compensation for credit exposure. Expected credit losses should be based upon expected exposure to counterparties (taking into account netting arrangements), expected default experience, and overall portfolio diversification. The unearned credit spread should preferably be adjusted dynamically as these factors change. It can be calculated on a transaction basis, on a portfolio basis, or across all activities with a given client.

Two additional adjustments are necessary for portfolios that are not perfectly matched: the "close-out costs adjustment" which factors in the cost of eliminating their market risk; and the "investing and funding costs adjustment" relating to the cost of funding and investing cash flow mismatches at rates different from the LIBOR rate which models typically assume.

The Survey reveals a wide range of practice concerning the mark-to-market method and the use of adjustments to mid-market value. The most commonly used adjustments are for credit and administrative costs.

Recommendation 4: Identifying Revenue Sources

Dealers should measure the components of revenue regularly and in sufficient detail to understand the sources of risk.

By identifying and isolating individual sources of revenue, dealers develop a more refined understanding of the risks and returns of derivatives activities. Components of revenue generally include origination revenue, credit spread revenue, if applicable, and other trading revenue. It is useful, though complex, to split other trading revenue among components of market risk.

The Survey of Industry Practice indicates that few dealers identify individual sources of revenue. This should become a more common practice.

Recommendation 5: Measuring Market Risk

Dealers should use a consistent measure to calculate daily the market risk of their derivatives positions and compare it to market risk limits.

• Market risk is best measured as "value at risk" using probability analysis based upon a common confidence interval (e.g., two standard deviations) and time horizon (e.g., a one-day exposure).
• Components of market risk that should be considered across the term structure include: absolute price or rate change (delta); convexity (gamma); volatility (vega); time decay (theta); basis or correlation; and discount rate (rho).

Reducing market risks across derivatives to a single common denominator makes aggregation, comparison, and risk control easier. "Value at risk" is the expected loss from an adverse market movement with a specified probability over a particular period of time. For example, with 97.5% probability (that is, a "confidence interval" of 97.5%), corresponding to calculations using about two standard deviations, it can be determined that any change in portfolio value over one day resulting from an adverse market movement will not exceed a specific amount. Conversely, there is a 2.5% probability of experiencing an adverse change in excess of the calculated amount.

Value at risk should encompass changes in all major market risk components listed in the recommendation. The difficulty in applying the technique of value at risk increases with the complexity of the risks being managed. For comparability, value at risk should be calculated to a common confidence interval and time horizon.

For most portfolios without options, once the expected loss is known for events with a given probability, the loss for a more likely or less likely scenario can easily be deduced. Therefore, for such portfolios, the choice of confidence interval is of no great significance. For option-based portfolios, however, this does not hold true. In their case, it would also be useful to calculate the loss from more and less likely scenarios.

A time horizon of one day is consistent with Recommendation 2 for daily marking to market, which allows management to know and decide daily any change of the risk profile.

Once a method of risk measurement is in place, market risk limits must be decided based on factors such as: management tolerance for low probability extreme losses versus higher probability modest losses; capital resources; market liquidity; expected profitability; trader experience; and business strategy.

The Survey suggests that most dealers know and consider some or all of the components of market risk. However, the use of one consistent measure of market risk, such as value at risk, is more prevalent among large dealers.

Recommendation 6: Stress Simulations

Dealers should regularly perform simulations to determine how their portfolios would perform under stress conditions.

Simulations of improbable market environments are important in risk analysis because many assumptions that are valid for normal markets may no longer hold true in abnormal markets.

These simulations should reflect both historical events and future possibilities. Stress scenarios should include not only abnormally large market swings but also periods of prolonged inactivity. The tests should consider the effect of price changes on the mid-market value of the portfolio, as well as changes in the assumptions about the adjustments to mid-market (such as the impact that decreased liquidity would have on close-out costs). Dealers should evaluate the results of stress tests and develop contingency plans accordingly.

The Survey indicates that some stress testing is being conducted, mainly by large dealers, and that broader usage is planned.

Recommendation 7: Investing and Funding Forecasts

Dealers should periodically forecast the cash investing and funding requirements arising from their derivatives portfolios.

The frequency and precision of forecasts should be determined by the size and nature of mismatches. A detailed forecast should determine surpluses and funding needs, by currency, over time. It also should examine the potential impact of contractual unwind provisions or other credit provisions that produce cash or collateral receipts or payments.

The Survey indicates that at present, half of responding dealers are conducting forecasts of cash investing and funding requirements. This type of forecast should become a more common practice.

Recommendation 8: Independent Market Risk Management

Dealers should have a market risk management function, with clear independence and authority, to ensure that the following responsibilities are carried out:

• The development of risk limit policies and the monitoring of transactions and positions for adherence to these policies. (See Recommendation 5.)
• The design of stress scenarios to measure the impact of market conditions, however improbable, that might cause market gaps, volatility swings, or disruptions of major relationships, or might reduce liquidity in the face of unfavorable market linkages, concentrated market making, or credit exhaustion. (See Recommendation 6.)
• The design of revenue reports quantifying the contribution of various risk components, and of market risk measures such as value at risk. (See Recommendations 4 and 5.)
• The monitoring of variance between the actual volatility of portfolio value and that predicted by the measure of market risk.
• The review and approval of pricing models and valuation systems used by front- and back-office personnel, and the development of reconciliation procedures if different systems are used.

The growth of activities in derivatives and other financial instruments has led many firms to establish market (and credit) risk management functions to assist senior management in establishing consistent policies and procedures applicable to various activities. Market risk management is typically headed by a board level or near board level executive.

The market risk management function acts as a catalyst for the development of sound market risk management systems, models, and procedures. Its review of trading performance typically answers the question: Are results consistent with those suggested by analysis of value at risk? The risk management function is rarely involved in actual risk-taking decisions.

According to the Survey, a large majority of dealers already have such a function in place and over 50% of those that do not plan to establish one in the near future.

Recommendation 9: Practices by End-Users

As appropriate to the nature, size, and complexity of their derivatives activities, end-users should adopt the same valuation and market risk management practices that are recommended for dealers. Specifically, they should consider: regularly marking to market their derivatives transactions for risk management purposes; periodically forecasting the cash investing and funding requirements arising from their derivatives transactions; and establishing a clearly independent and authoritative function to design and assure adherence to prudent risk limits.

While many end-users do not expect significant change in the combined value of their derivatives positions and the underlying positions, others do. Derivatives are customer-specific transactions, often designed to offset precisely the market risk of an end-user's business position (e.g., buying a commodity as a raw material). End-users should establish the performance assessment and control procedures that are appropriate for their derivatives activities.

Less than half of those end-users surveyed currently mark their derivatives hedges to market for risk management purposes. About half plan to do so.

Credit Risk Measurement and Management

Recommendation 10: Measuring Credit Exposure

Dealers and end-users should measure credit exposure on derivatives in two ways:

• Current exposure, which is the replacement cost of derivatives transactions, that is, their market value.
• Potential exposure, which is an estimate of the future replacement cost of derivatives transactions. It should be calculated using probability analysis based upon broad confidence intervals (e.g., two standard deviations) over the remaining terms of the transactions.

To assess credit risk, a dealer or end-user should ask two questions. If a counterparty was to default today, what would it cost to replace the derivatives transaction? If a counterparty defaults in the future, what is a reasonable estimate of the future replacement cost?

Current exposure is an accurate measure of credit risk that addresses the first question. It simply evaluates the replacement cost of outstanding derivatives commitments. The result can be positive or negative. It is an important measure of credit risk as it represents the actual risk to a counterparty at any point in time. The regular calculation of current exposure is a broadly accepted practice today.

Potential exposure is more difficult to assess, and the methods used to determine it vary. The most rigorous methods use either simulation analysis or option valuation models. The analysis generally involves a statistical modeling of the effects on the value of the derivatives of movement in the prices of the underlying variables (such as interest rates, exchange rates, equity prices, or commodity prices). These techniques are often used to generate two measures of potential exposure: expected exposure; and maximum or "worst case" exposure.

Dealers and end-users that cannot justify the simulation and statistical systems needed to perform such potential exposure calculations should use tables of factors developed under the same principles. The factors used should differentiate appropriately by type and maturity of transaction and be adjusted periodically for changes in market conditions.

The Survey shows that dealers use several different methods for calculating credit exposures. These include: the BIS original and current exposure methods, used by one-third of all dealers; methods based on worst-case scenarios applied to each transaction, used by about a quarter of dealers and expected to become the most common in the future; and methods that rely upon tables of factors, used by almost 40% of dealers. End-users tend to rely on simpler methods primarily based on notional amounts.

Recommendation 11: Aggregating Credit Exposures

Credit exposures on derivatives, and all other credit exposures to a counterparty, should be aggregated taking into consideration enforceable netting arrangements. Credit exposures should be calculated regularly and compared to credit limits.

In calculating the current credit exposure for a portfolio of transactions with a counterparty, the first question is whether netting applies. If it does, the current exposure is simply the sum of positive and negative exposures on transactions in the portfolio.

The calculation of potential exposure is more complicated. Simply summing the potential exposures of all transactions will in most cases dramatically overstate the actual exposure, even if netting does not apply. This is because a straight summation fails to take into account transactions in the portfolio that offset each other or that have peak potential exposures at different times. The most accurate calculation of potential exposure simulates the entire portfolio. Although portfolio-level simulation is not commonly used by dealers at present, they should pursue it more widely to avoid overstating aggregate exposure.

Credit exposures should be calculated regularly. In particular, dealers should monitor current exposures daily; they can generally measure potential exposures less frequently. End-users with derivative portfolios should also periodically assess credit exposures. For them, the appropriate frequency will depend upon how material their credit exposures are.

Credit exposures should also be regularly compared to credit limits, and systems should be in place to monitor when limits are approached or exceeded, so that management can take appropriate actions.

By aggregating credit exposures on derivatives as described above, participants will have a consistent basis for comparison with other credit exposures including those resulting from on-balance-sheet activity. This would permit a more effective evaluation of the adequacy of credit reserves relative to overall credit exposure.

The Survey suggests that most dealers monitor gross credit use against limits. Aggregating current and potential exposures by counterparty on a net basis is not common among dealers, although some who do not net at present plan to in the future. Frequent monitoring of credit exposure is widespread among dealers, with three-quarters of respondents doing it either intraday or overnight. The majority of end-users monitor credit exposures at least once a month.

Recommendation 12: Independent Credit Risk Management

Dealers and end-users should have a credit risk management function with clear independence and authority, and with analytical capabilities in derivatives, responsible for:

• Approving credit exposure measurement standards.
• Setting credit limits and monitoring their use.
• Reviewing credits and concentrations of credit risk.
• Reviewing and monitoring risk reduction arrangements.

For dealers, credit exposures should be monitored by an independent credit risk management group. According to the Survey, most dealers and some end-users have such a group. For end-users, this role may not necessarily be performed by a separate group; however, the credit risk should be managed independently from dealing personnel. This separation of responsibility is intended to prevent conflicts of interest and to ensure that credit exposure is assessed objectively. The credit risk management function should approve exposure management standards, and should establish credit limits for counterparties consistent with these standards. Specifically, it should conduct an internal credit review before engaging in transactions with a counterparty, and should guide the use of documentation and credit support tools. Credit limits and guidelines should ensure that only those potential counterparties that meet the appropriate credit standards, with or without credit support, become actual counterparties.

The credit risk management function should continually review the creditworthiness of counterparties and their credit limits.