A saddle for complex credit portfolio models

Quantifying credit portfolio losses under realistic assumptions - including correlated defaults, stochastic loss given default (LGD), and dependencies between a counterparties' probability of default (PD) and LGD - is usually based on Monte Carlo simulation and hence computationally expensive. Analytic approximations have the advantage of being computationally less expensive and enabling the user to perform additional calculations, such as calculating risk contributions, risk-adjusted pricing