中国和其它金砖国家股票市场动态相关性研究——基于动态时变Copula模型分析

Analysis of the Dynamic Dependence between China and other BRICs Countries——Based on Dynamic Time-varying Copula Model

该文选取2007年至2014年金砖五国的股票指数日收盘价数据,运用AR(1)-FIGARCH(1,d,1)-偏t模型对各自的边缘分布进行拟合,然后基于13种不同的动态时变Copula模型对中国和其它金砖国家股票市场之间的相关结构进行拟合估计,从中选出T-DCC-Copula模型作为最优的模型。实证结果表明:中国与其他金砖国家经济体股票市场之间动态相关性在整体上并没有因为金砖政治体而显著增强,在局部上却都具有"峰会效应",该效应在中印、中南以及中巴股市之间持续时间较短,而在中俄股市之间持续时间较长,这说明与其他三个金砖国家相比,中俄之间的贸易往来更为活跃。

This paper selects daily closing price data of stock index from stock markets of the BRICs coun- tries from 2007 to 2014, use the AR （ 1 ） -FIGARCH （ 1, d, 1 ）-partial t models to fit their respective marginal distribution, and thenfit to estimate the correlation structure betweenstock markets of China and other BRICs countries based on the 13different dynamic time-varying Copula model, from which we selects the T-DCC-Copu- la model as the optimal model. Based on this model, the empirical results show that： the dynamic correlations between China and other economies＇ stock markets are not significantly enhanced in the whole because of this BRICs political system, but there has a ＂Summit＂ effect in the local, the effeetbetween China and India, Chi- na and South Africa, and also China and Brazil stock markets is of shorter duration, while the effect between China and Russia＇ s stock market is for a long time, and this shows that compared with the other three coun- tries, the trade between China and Russia is more active. The results of this study can provide some reference for investors and financial regulators in portfolio investment and risk management.