Benzschawel, Terry, Adoito Haroon and Tuohua Wu, "A Model for Recovery Value in Default", Journal of Fixed Income, Vol. 21, No. 2, (Fall 2011), pp. 15-29.
Abstract: The amount recovered in default is as important as default probability for estimating expected losses on risky assets and for determining fair credit spreads, yet recovery value has been much less well studied than default. Recent research has increased substantially information about recovery value in default. From those studies, principles have emerged regarding the influence of collateral, credit cycle, seniority, industry sector, credit quality, and geography on loss given default. This article presents a decision-tree model that embeds known determinants of recovery value. Simulations of historical default rates and correlated defaults using the model generate distributions of recovery values that replicate well the statistical properties of recovery values reported in the literature.