A comparison of heavy-tailed VaR estimates and Filtered Historical Simulation: Evidence from emerging markets

Abstract

The purpose of this paper is to investigate relative performance of Value at Risk (VaR) models in forecasting risk of emerging markets financial series. Among well-known Value-at-Risk modelling techniques, we based our estimates on GARCH, Extereme Value Theory (EVT) and Filtered Historical Simulation (FHS). Risk estimates are then exposed to the backtesting models inorder to evaluating comparative accuracy of the VaR models. The results indicate that EVT-based VaR estimates provides us with more accurate estimation of VaR specially in higher quantiles.