基于Copula-EVT模型的沪深股市尾部相关性分析

A Copula-EVT Model for Measure Tail Dependence between SHI and SZI

考虑到实际金融收益变量具有尖峰厚尾性,将边缘分布选取为极值分布和Laplace分布组合,结合Copula技术来对上海综合指数和深圳成分指数的尾部相关性进行研究.其中Copula函数选择具有下尾相关性的Clayton Copula和上尾相关的Gumbel Coula来度量尾部相关性,量化后的指标可以很好预测股票市场的发展.

In this paper, as for financial income variable owns character of high kurtosis, thick tail and extreme phenomenon, we combine extreme value distribution with Laplace distribution as margin distribution, and make use of Copula technique to measure the tail dependence between Shanghai compositive index and Shenzhen componential index. We choose Clayton Copula with lower tail de- pendence and Gumbel Copula with upper dependence to measure dependence, and the tail quantified dependence could forest the change in the stock market in the future.