An introductory overview should provide some concise information on the credit risk policy at the trading activities level and of its integration into the overall risk policy of the firm. This should be followed by a short analysis of the major economic and firm specific events, as well as important decisions, that affected the economic performance and credit risk exposures of the client and proprietary trading books. The total revenue generated by each book should be presented by rating class, geographic region and industrial sectors for the important market risk factors and instruments traded. The credit limits authorisation policy should be presented, mentioning in particular whether the firm has guidelines regarding the interaction with the market risks trading limits set for the various instruments traded.

The following presents the presentation format for evaluating the quality of credit risk management for the trading book, including the tables and the quantitative and qualitative information.

a) General description of trading position notional amounts

This should include the notional amounts of the aggregate trading positions. These are broken into major asset and maturity profiles as in Table 2, by instrument type (swaps, futures, forwards and options) and by their OTC versus exchange traded specifics (or by internal rating system characteristics, though this would make cross firm comparisons difficult). Ideally, these notional amount analyses should distinguish between the proprietary and client-based trading activities.

Table 2

b) Current and potential credit risk exposures

In Tables 3 and 4, the current exposures are expressed in two ways: gross and net of netting agreements, collateral and other guarantees. They are categorised by official rating categories and by country to identify concentration risk. In addition, additional information regarding OTC derivatives past-due amounts and credit losses experienced during the period should be added. This was recently suggested by the joint IOSCO and BIS 'Framework for supervisory information about derivatives and trading activities' issued in September 1998 (their suggested presentation of the table is reproduced in the Appendix).

Table 3

Table 4

Given credit exposure changes over time, firms should also disclose their potential credit exposures using BIS add-ons or the firm's internal credit risk model. If the latter is used it should be explicitly mentioned below the table. To refine the trading book positions credit risk exposure analysis, the following information should then complement those tables:

• changes in credit exposure during the year,
• underlying assumptions of the internal credit risk model type (firm value, structural or rating changes based internal add-on model), input data source to the model (KMV expected default probabilities, internally computed rating migration tables, S&P or Moody's rating migrations, etc.) as well as information regarding the exposures, the recovery rates distribution, the joint default probabilities computation and the treatment of non-linear payoff credit sensitive claims,
• method of computation used to integrate OTC and credit derivatives default risk within the model ,
• whether the credit risk model used also accounts for the market risk sensitivities of the positions ,
• backtesting results of the internal credit risk model and stress testing scenarios and outcomes.

Finally, given the importance of integrating market and credit risks management, the firm should provide relevant information on methods and models used, for some or all of the trading books and especially for credit derivatives. The modelling approach, the integrated VAR model used and its backtesting procedures should then be discussed along the same lines as in section IV 2 together with a description of the method used to compute the correlations between the market and the credit risk factors.

c) Concentration risks

The traditional way to distinguish between OECD and non-OECD country banks, governments, central banks and agencies is out-of-date. Noting the recent credit risk crunches in Asia and Russia, and the potential systemic risk of LTC hedge fund collapse, we see that financial firms need a more refined partitioning method to understand the multiplier effects stemming from holding concentrated positions in tight or disrupted markets. Thus, concentration risk should be assessed at the rating, geographic region, currency and industrial sector levels. Along the lines suggested in Tables 5 and 6 below, the firm should present the concentration level by the industrial sector and by the major currencies of the trading positions, based on their gross replacement values.

Table 5

Table 6

Concentration risk should also be identified and examined for major individual counterparties. For instance the current gross individual and cumulative credit exposures to the 10 largest counterparties and the their fraction of the economic capital should be presented. As illustrated in Table 7 below, this could be refined by reproducing the same table for the main geographical regions, industrial sectors and rating classes.

Table 7

d) Other relevant information

The marking-to-market, or any other valuation methods used for the derivative and cash instruments traded, have to be explicitly stated for each major risk category. The proportion of marked-to-model cash and derivative instruments should be explicitly mentioned for both OTC as well as exchange-traded products for the aggregate trading portfolio and for each major risk category.