Bayesian Analysis of a Triple-Threshold GARCH Model with Application in Chinese Stock Market

junjun zhu,shiyu xie

2010-01-01

Abstract:This paper constructs a triple-threshold GARCH model, which incorporates threshold regime determination, asymmetric response of conditional mean to past returns, and asymmetric reaction of conditional volatility to past innovations. In this model, threshold variable for regime switching is defined as moving average of past returns. Parameter estimation is performed using Griddy-Gibbs sampling method. Empirical analysis in Chinese stock market shows that 12-days-average return Plays an important role in defining different regimes, and Chinese stock market indicates evidences of asymmetric reaction in both conditional mean and conditional volatility. In VaR assessment, the triple-threshold GARCH model shows more accurate forecast performance than GARCH and GJR models.