Joint Default and Recovery Risk Estimation: An Application to CDS Data
by Jens Henrik Eggert Christensen of the Federal Reserve Bank of San Francisco
January 22, 2007
Abstract: Motivated by the results in Christensen (2006b) this paper tests the separation of default and recovery risk in a reduced-form affine credit risk model using CDS data for Ford Motor Co. The results indicate that default and recovery risk can in fact be jointly identified from cross-sections of CDS rates in this case. Furthermore, it is investigated whether changes in the risk premium specification has any impact on the separation. To that end, the essentially affine and the extended affine risk premium specifications are analyzed. The results show that the Feller conditions required in the extended affine risk premium specification are binding restrictions which significantly change the outcome of the CDS rate decomposition. For that reason the essentially affine specification is preferred for credit risk modeling purposes.