The Volatility and Density Prediction Performance of Alternative GARCH Models

Abstract

This study compares the volatility and density prediction performance of alternative GARCH models with different conditional distribution specifications. The conditional residuals are specified as normal, skewed-t or compound Poisson (jump) distribution based upon a non-linear and asymmetric GARCH (NGARCH) model framework. The empirical results for the S&P 500 and FTSE 100 index returns suggest that the jump model outperforms all other models in terms of both volatility forecasting and density prediction. Nevertheless, the superiority of the non-normal models is not always significant and diminished during the sample period on those occasions when volatility experiences an obvious structural change.

Keywords: GARCH, Model fitting, Volatility forecasting, Density prediction, Jumps

JEL Classification: G10, C10, C30