Empirical application of normal mixture GARCH and value-at-risk estimation

Abstract

Normal mixture (NM) GARCH model can capture time variation in both conditional skewness and kurtosis. In this paper, we present the general framework of Normal mixture GARCH (1,1). An empirical application is presented using Malaysia weekly stock market returns. This paper provides evidence that, for modeling stock market returns, two-component Normal mixture GARCH (1,1) model perform better than Normal, symmetric and skewed Student’s t-GARCH models. This model can quantify the volatility corresponding to stable and crash market circumstances. We also consider Value-at-Risk (VaR) estimation for Normal mixture GARCH model.