Applying importance sampling for estimating coherent credit risk contributions
提出基于重要性抽样的蒙特卡洛方法，用于估计信用组合中每个借款人的个体风险贡献，并与标准方法对比，证明了渐近最优性。
A Monte Carlo simulation method based on importance sampling is applied to the problem of determining individual risk contributions of the obligors in a credit portfolio. The effectiveness of the method is benchmarked against standard Monte Carlo techniques and the asymptotic optimality of the method is proved. The risk measure adopted is expected shortfall, a particualr coherent risk measure. The concept of a coherent risk spectrum is discussed on the basis of some numerical examples.