How Important is the Modeling of Interest Rate and Credit Spread Risk in Standard and Non-Standard Credit Portfolio Models?

Abstract

In this comprehensive simulation study the effect on the economic capital of a credit portfolio which results from integrating interest rate and credit spread risk into portfolio models with different specifications is analyzed. By using forward rates, most credit portfolio models currently employed in the banking industry treat these risk factors as known in advance. In contrast to previous studies, in this paper transition risk, credit spread risk, interest rate risk and - for one specification - also recovery rate risk are assumed to be correlated risks. The simulations show that generally the error made when neglecting the stochastic nature of interest rates or credit spread risk is significant, especially for high quality credit portfolios with low correlations between the obligors' asset returns. Considering inhomogeneities in the portfolio composition as they appear in real-world credit portfolios or different distributional assumptions for the credit risk factors might reduce the effect of an underestimation of economic capital, but it is still relevant.