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

摘要：

　　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.

关键词： threshold Griddy-Gibbs sampling MCMC method GARCH

作者: Zhu Junjun Xie Shiyu

作者单位: School of Economics, Fudan University

会议类型: 国际会议

会议名称: The 4th Conference on Chinas Economic Operation Risk Management(2010·Shanghai)(第四届中国立信风险管理论坛)

会议地点: 上海

会议语种:英文

页码: 326-333

在线出版日期: 2010-10-14（万方平台首次上网日期，不代表论文的发表时间）