RDP 9409: Default Risk and Derivatives: An Empirical Analysis of Bilateral Netting 8. Conclusion

There are strong theoretical arguments in favour of a more sophisticated approach to setting a credit risk capital charge when obligations are netted. However, empirical support for these arguments does not appear to be as strong. The work with Australian banks' portfolios provides some evidence to suggest that a number of approaches to the calculation of the add-on provides a more efficient coverage of potential exposure than the current Basle formula, namely:

• the current Basle add-on scaled by the absolute ratio of net to gross market value;
• the short/long approach;
• a method which allows the add-ons for contracts within given time bands to be netted; and
• a smooth, rather than stepped, function of term to maturity.

The approach to the calculation of a total capital charge which appears most robust to interest rate movements over the cycle is the scenario approach. While several approaches are some improvement over the proposed method, no single formulation clearly outranks all others.

The better performance of these alternative approaches to capturing potential exposure is not sufficient to outweigh the dominance of current exposure in the determination of the total capital charge. So long as the coverage of potential exposure afforded by the add-ons remains low relative to that provided for current exposure, and that is one conclusion of this work, then the capital charge will continue to be dominated by current exposure. That suggests that the case for a more complex formulation of the capital charge is not warranted.

The conclusions of the analysis are, nonetheless, tentative since the work covers only portfolios comprising simple interest rate swaps and forward rate agreements. Exchange rate and other, more complex, derivatives have not been considered. The add-on factors for exchange rate contracts are considerably higher than those for interest rate contracts. Hence, the add-ons for a portfolio made up of both interest rate and other contracts could represent a higher proportion of the total capital charge than for a portfolio of interest rate contracts alone. This remains an empirical question since current exposure on these contracts could also be higher. Whether the add-ons represent a higher or lower proportion of the total capital charge for more complex portfolios, will depend upon the composition of the portfolio and on market conditions.