论文部分内容阅读
近年来,Copula方法在金融市场相关性的研究越来越广泛。由于金融时间序列具有时变非对称相关的特性,本文引入时变相关非对称Copula函数(GJC-Copula)应用非参数核密度估计刻画序列概率分布,再采用极大似然估计方法估计尾部相关参数,研究了沪深股市之间的相关性,结果表明股市在下跌时相关系数更大,并通过加入虚拟变量方式分析了股权分置改革后相关系数的变化,结果表明05年股权分置改革后沪深股市上尾和下尾相关系数均有一定程度下降。
In recent years, Copula’s research on the correlation of financial markets has become more and more widespread. Due to the time-varying and asymmetric correlation of financial time series, this paper introduces the non-parametric kernel density estimation of GJC-Copula to describe the sequence probability distribution, and then uses the maximum likelihood estimation method to estimate the tail-related parameters , The correlation between the Shanghai and Shenzhen stock markets was studied. The result shows that the stock market has a larger correlation coefficient when it falls, and the change of the correlation coefficient after the non-tradable share reform has been analyzed by adding the dummy variables. The results show that after the reform of the share structure in 2005 Shanghai and Shenzhen stock market at the end of the tail and the correlation coefficient has declined to some extent.