Extension of Spot Recovery Model for Gaussian Copula

by Hui Li of AIG

October 17, 2009

Abstract: Heightened systematic risk in the credit crisis has created challenges to CDO pricing and risk management. One important focus has been on the modeling of stochastic recovery. Different approaches within the Gaussian Copula framework have been proposed, but a consistent model was lacking until the recent paper of Bennani and Maetz [6] which shifted the modeling from period recovery to spot recovery. In this paper, we generalize their model to an arbitrary spot recovery distribution setup and extend the deterministic dependency on systematic factor to a random one. Besides, an extra parameter is introduced to control the correlation between default and recovery rate and the correlation between the recovery rates.

JEL Classification: G32, G13.

Keywords: CDO, Gaussian Copula, Stochastic Recovery, Spot Recovery Model.

Download paper (192K PDF) 20 pages

Related reading: Extensions to the Gaussian Copula: Random recovery and random factor loadings