Sensitivity Analysis of Portfolio Credit Derivatives by Conditional Monte Carlo Simulation

Abstract

We study sensitivity analysis of portfolio credit derivatives, including basket default swaps and collateralized debt obligations. An unbiased estimator is derived using conditional Monte Carlo for sensitivities with respect to systemic parameters (parameters that influence some or all the entities). Copula-based methods are used to model the joint distribution of the default times. Simulation experiments demonstrate the advantages of the proposed derivative estimator over other methods.

Keywords: simulation, stochastic gradient estimation, credit derivative, conditional Monte Carlo, copula model