Pricing Tranched Credit Products with Generalized Multifactor Models

The market for tranched credit products (CDOs, Itraxx tranches) is one of the fastest growing segments in the credit derivatives industry. However, some assumptions underlying the standard Gaussian onefactor pricing model (homogeneity, single factor, Normality), which is the pricing standard widely used in the industry, are probably too restrictive. In this paper we generalize the standard model by means of a two by two model (two factors and two asset classes). We assume two driving factors (business cycle and industry) with independent t-Student distributions, respectively, and we allow the model to distinguish among portfolio assets classes. In order to illustrate the estimation of the parameters of the model, an empirical application with Moody’s data is also included and potential relationships between default rates and macroeconomic variables are analyzed.