A Simple Estimate of VAR under Garch Modelling

This paper calculates Value at Risk under under a GARCH framework for returns, without assuming a given probability distribution for errors of GARCH process. The procedure is given and its accuracy is checked. The bootstrap method is proposed to study the finite sample properties of estimate. Using five examples, we show that our approach works well. The good properties of our approach are proved. A methodology is proposed for selecting the time horizon. Comparisons with JPMorgan Riskmetrics outcomes are performed. Finally, a real data set is considered.