The performance of GARCH (1,1) models in forecasting volatility of financial markets.

Abstract

The purpose of this study is to examine the forecasting performance of the GARCH (1, 1) model in an attempt to answer to the findings of several forecasts competitions that present the GARCH models as poor forecast predictors. We compare the forecast accuracy of the GARCH (1, 1) model with that of a homoskedastic one, by using as statistical criterion the mean squared prediction error. We utilize bilateral daily data for the dollar versus the currencies of other ten countries and bilateral daily data of stock market returns for ten financial markets. We compare the out of sample performance realization of the squared of the daily change in an exchange of stock return rate with the value predicted by a model of the conditional variance