The analysis starts with a qualitative description of the credit risk policy and whether it applies to all or part of its non-trading activities, as well as its coverage across products and markets. The firm should explicitly mention what type of rating system (internal versus external) it uses and whether or not the internal rating system is based on an internal credit risk model. In the latter case, the main features and assumptions of the model should be stated, as discussed above in section IV.3 for the credit risk management disclosure for trading activities. Furthermore, the management of counterparty risk should be discussed in terms of frequency and procedures for rating revisions, special procedures for non-performing loans evaluation and concentration risk management. In particular, the firm should clearly define its on- and off-balance sheet items (see Table 2 in the Appendix). Further the firm should mention if it relies on an internal model to manage and thus diversify the credit risk exposure of its overall loan portfolio and to what extent it also uses it for its off-balance sheet positions. The credit losses total allowances and provisions for unexpected losses policy should be described, in particular if the firm calculates its allowances for expected losses and provisions for unexpected credit losses based on an internal credit risk model.

The performance of its loan portfolio and off-balance sheet positions credit risk management should be briefly reviewed in light of the domestic and international performance of the firm as well as the major economic and political events during the year. (Whereas the framework is discussed in detail for the credit risk disclosure of the loan portfolio, it is explicitly intended to apply to all on- and off-balance sheet items, as listed in Table 2 of the Appendix, according to their economic importance at the level of each firm.)

b) Counter party data analysis

More specifically, this section should include the following information in table format.

• Distribution of the loans (in % terms) according to internal/external ratings. If internal ratings are used then a description of the main features of the rating system should be included in the previous qualitative discussion and, if possible, a correspondence to an official rating system should be provided for the sake of comparison.
• Distribution of the loans (in % terms) according to geographic region (following the same geographical regions as Table 4)
• Analysis of the loan portfolio and of its performance according to geographic region and industrial sector, along the lines proposed in Table 9.
• Analysis of the total allowances for the credit risk exposures of the entire firm and of its coverage ratio for loans, along the lines presented in Table 10.

If the institution uses an elaborate credit risk model to compute its expected and unexpected credit losses allowances, the analysis in Table 10 can be refined along the lines discussed above. A discussion of the statistical robustness of the model-based allowances (similarly to a backtesting procedure) should be added.

Given a recent initiative by the BIS and domestic supervisors, one would expect that internationally active financial firms would rapidly conform to the following presentation format for their loan portfolio management. Ideally, these above requirements would extend to off-balance sheet positions as well (such as letters of credit as well as performance and other traded guarantees). However, the current state of financial theory does not enable the extension of the existing credit risk internal models to these instruments.

Table 9

Table 10

c) Concentration risk

For similar reasons to those in section IV.3, concentration risk is crucial at the level of the loan and other on- and off-balance sheet positions of a financial firm. Concentration risk is amplified by lack of liquidity in the short run and insolvency in the long run. Both can quickly magnify risks for cumulated large exposures to a given sector, geographic region or a large counterparty. Nevertheless the client-oriented nature of the loan and trade financing business makes it difficult for many medium sized institutions to avoid concentration risk. As presented in Table 11, concentration risk should be quantified at the industry level, since it is well known that the risk of a growing economic sector is less a declining sector. Concentration by currency should be presented, given the wide range of currency volatilities (as in Table 6). Finally, the ratings-based and individual-counterparty-risk decompositions of the loan portfolio exposures, provided in Tables 12 and 13 respectively, complete the identification and measurement of the concentration risk for the loan portfolio (and by analogy, for other on- and off-balance sheet positions that are listed in Table 2 of the Appendix).

Table 11

Table 12 Pie Chart of Loan Portfolio Decomposition by Ratings* *The rating source (internal or external) and computational methodology (credit risk model or accounting ratios based) should be mentioned and the pie chart decomposition should be based on the gross replacement cost of the exposures.

Table 13

d) Cross-border risk analysis

Currently cross-border risk analysis is performed primarily by U.S. banks, although the trend may quickly follow in Europe (see for instance Barclays' 1997 annual report). We feel that a minimum set of information, as summarised in Table 14, should be included in the annual report to present the additional political and economic risks of its foreign outstanding commitments, even if these are exchange listed products. (This is based on the March 1997 recommendations on cross-border risk by the Federal Financial Institutions Examination Council (FFIEC).)

Table 14

The information provided in Table 14 should be supplemented with a risk assessment of each country (or economic region) based, for instance, on an internal model which accounts for economic and political stability and for sovereign borrower reputation.