Value at Risk with Filtered Historical Simulation

Abstract

In this paper we study the properties of estimates of the Value at Risk (VaR) using the historical simulation method. Historical simulation (HS) method is widely used method in many large financial institutions as a non-parametric approach for computing VaR. This paper theoretically and empirically examines the filtered historical simulation (FHS) method for computing VaR that combines non-parametric and parametric approach. We use the parametric dynamic models of return volatility such as GARCH, A-GARCH. We compare FHS VaR with VaR obtained using historical simulation and parametric VaR.