基于MRS Copula模型的沪港股市相依关系研究

Dependence between Shanghai and Hong Kong Stock Market: An Empirical Analysis Based on MRS Copula Model

本文引入ARFIMA-FIAPARCH-Skewed t模型刻画上证与恒生指数收益率的典型事实特征,并进一步结合EVT极值理论建立边缘分布;在此基础上,为了准确刻画沪港股市间相依关系的结构突变特征,本文构建了三类马尔可夫机制转换Copula(MRS Copula)模型,并与普通时变Copula模型进行比较,选取拟合效果相对较好的Copula模型,探讨沪港股市间相依关系变化。实证结果表明:相较于普通时变Copula模型,MRS Copula模型能够更为有效地刻画沪港股市间的尾部相关性,更进一步地,MRS SJC Copula模型的拟合效果相对较好;2008年次贷危机、沪港通和深港通的开通等重要事件均会对沪港两市之间的相依状态产生影响。

In this paper,the ARFIMA-FIAPARCH-Skewed t model is used to characterize the typical facts of the Shanghai Composite Index and Hang Seng Index,which establishes the marginal distribution in combination with EVT theory;on this basis,in order to accurately characterize the structure switching of the dependence between Shanghai and Hong Kong stock market,three Markov regime switching Copula (MRS Copula)models are eonstrueted and compared with the eommon time-varying Copula model.Finally,the Copula function with relatively good fitting effect is selected to research the changes of the dependence between Shanghai and Hong Kong stock market.The empirical results show that compared with common time-varying Copula model,the MRS Copula model can more effectively characterize the tail correlation between Shanghai and Hong Kong stock market.Further,the fitting effect of MRS SJC Copula is relatively good.In addition,subprime crisis in 2008,the Shanghai-Hong Kong Stock Connect and Shenzhen-Hong Kong Stock Connect will affect the state of dependence between Shanghai and Hong Kong stock market.