modelling lgd using survival analysis

Loss given default (lgd) is one of the key parameters needed in order to estimate expected and unexpected credit losses necessary for credit pricing as well as for calculation of the regulatory basel ii requirement (bcbs, 2006). while the credit rating and probability of default (pd) techniques have been well developed in recent decades, lgd has attracted little attention before 2000s.in this paper, we compare linear regression and survival analysis models for modelling recovery rates and recovery amounts, in order to predict the  lgd for unsecured consumer loans or credit cards.