基于机制转换混合Copula模型的我国股市间极值相依性

Extreme dependence in Chinese stock markets based on regime-switching mixed Copula

通过构建机制转换混合Copula模型,考察了沪、深股市与港、台股市间的尾部相依特征,研究发现:它们间的尾部相依性呈非对称动态过程,在低风险状态下,右尾部相依性普遍高于左尾部相依性;而在高风险状态下,左尾部相依性普遍高于右尾部相依性.相对于沪、深股市与台股而言,沪、深股市与港股间的尾部相依性更强,其尾部相依性对外来冲击更敏感;相对于沪市与港、台股市而言,深市与港、台股市间的尾部对外在冲击更敏感.在此次次贷危机中,沪、深股市与港、台股市间极值风险显著增加,并呈现明显的金融感染.而且各尾部相依性的两个结构变化几乎同时发生并分别对应于危机的第一、二阶段,这表明:沪、深股市与港、台股市间的风险呈系统性特征,而危机的第三阶段对沪、深股市与港、台股市间的影响较有限.

Based on regime switching mixed Copula model, this paper finds the tail dependence between Shanghai, Shenzhen and Hongkong, Taiwan stock markets is a dynamic asymmetric process. In the state of low risk, right tail dependence is higher than left tail dependence; in the state of high risk, the situation is opposite. The tail dependence between Shanghai, Shenzhen and Hongkong stock markets is much stronger and more sensitive to exterior shocks than that between Shanghai, Shenzhen and Taiwan stock markets. The tail dependence between Shenzhen and Hongkong, Taiwan stock markets is more sensitive to exterior shocks than that between Shanghai and Hongkong, Taiwan stock markets. Tail dependence shows Chinese stock market risk increases significantly in the crisis and financial contagion happens. Furthermore, the tail dependence shows two big simultaneous structure changes correspond to the first and second stages of the crisis, which means the stock market risk is systematic. However, the third stage of the crisis almost has no influence on Chinese stock markets.