The Importance of Simultaneous Jumps in Default Correlation

2 Abstract Correlated defaults have been an important area of research in credit risk analysis with the advent of a basket of credit derivatives. Even the simple credit derivatives should be considered a basket of two default risks since the bankruptcy risk of the derivative issuer is also a factor. Considering jumps in the asset value helps to model the surprise risk of default in a group of firms. Simultaneous jumps in the asset values of companies can explain the default correlation. The multivariate jump diffusion model is used for modeling the asset value in the structural approach to credit risk modeling. GMM implemented on the moments generated by empirical characteristic function is the method used for estimation of the parameters. The principal component method is used for reducing the hassle of moment conditions in the characteristic function estimation of the model. At the end, the empirical result of joint default credit risk of a basket of two firms, Ford and General Motors, are shown using two models: one without jump and the other one with the simultaneous jump. Model selection criterion proves that the model with jump is a better model. The model without simultaneous jump underestimates the joint default probability of two firms.