Abstract
This chapter provides an overview of modeling and computational issues associated with portfolio credit risk. We consider the problem of calculating the loss distribution in a portfolio of assets exposed to credit risk, such as corporate bonds or bank loans. We also discuss the pricing of portfolio credit derivatives, such as basket default swaps and collateralized debt obligations. A portfolio view of credit risk requires capturing dependence between the assets in the portfolio; we discuss models of dependence and associated computational techniques. A standard modeling framework takes the assets to be conditionally independent given a set of underlying factors, and this is convenient for computational purposes. We discuss a recursive convolution technique, transform inversion, saddlepoint approximation, and importance sampling for Monte Carlo simulation.