Monte Carlo risk management

Abstract

In this paper we propose a Monte Carlo based approach to Risk Management. Our approach applies to any system subject to random uncorrelated losses under very general conditions. Our methodology consists of the following steps: model the distribution of losses and the distribution of time intervals between losses via an analysis of historical data; perform a Monte Carlo simulation of a finite period of time in the future; use the Monte Carlo data to estimate the 99.9% VaR.