Correlation Calibration with Stochastic Recovery

Abstract

In this chapter, we expand the base correlation framework by enriching it with Stochastic Recovery modelling as a way to address the model limitations observed in a distressed credit environment. We introduce the general class of Conditional-Functional recovery models, which specify the recovery rate as a function of the common conditioning factor of the Gaussian copula. Then, we review some of the most popular ones, such as: the Conditional Discrete model of Krekel (2008), the Conditional Gaussian of Andersen and Sidenius (2005) and the Conditional Mark-Down of Amraoui and Hitier (2008). We also look at stochastic recovery from an aggregate portfolio perspective and present a top-down specification of the problem. By establishing the equivalence between these two approaches, we show that the latter can provide a useful tool for analyzing the structure of various stochastic recovery model assumptions.