Subsampling and other considerations for efficient risk estimation in large portfolios

Computing risk measures of a financial portfolio comprising thousands derivatives is challenging problem both because it involves nested expectation requiring multiple evaluations the loss for different scenarios and evaluating expensive cost increases with size. We apply multilevel Monte Carlo simulation adaptive inner sampling to this discuss several practical considerations. In particular, we subsampling strategy whose computational complexity does not increase size portfolio. also control variates that significantly improve efficiency in our setting.