A Probabilistic Simulation Based VaR Computation and Sensitivity Analysis Method

Abstract

This paper presents a new method to compute VaR (value at risk) and perform corresponding variance based sensitivity analysis. VaR has a long history of being applied in stock price prediction and investment portfolio analysis. Traditional method, however, is mainly analytical, and has certain limitations. The VaR simulation, on the other hand, provides more realistic analysis but is very slow which affects the applications. This study proposes a new VaR computation method based on a probabilistic simulation technique called Simulation As You Operate (SAYO). It is always helpful to know the most influential factors in an investment, and thus a sensitivity analysis method based on SAYO is also introduced to enhance investment analysis.