基于多种Copula函数的沪新股市尾部相关性比较分析  

Comparison on Tail Dependence Between SCI and SSTI Based on Copulas

作　　者：王晓芳[1] 谢金静[1] 

机构地区：[1]西安交通大学经济与金融学院,陕西西安710061

出　　处：《统计与信息论坛》2009年第6期27-32,共6页Journal of Statistics and Information

基　　金：国家社会科学基金项目<防止资产价格过快上涨和抑制资产泡沫问题研究---基于货币政策调控的理论与实证>(08BJY153)

摘　　要：投资组合理论要求在尽可能多的市场上进行投资组合,因而研究两个股票市场之间的尾部相关关系对全球投资组合和风险管理具有重要的意义。采用基于秩的极大似然估计法,分别采用静态和动态Copula函数对比研究上证综指和新加坡海峡时报指数收益序列的尾部相关关系,结果发现时变的SJC Copula函数较其他3种静态Copula函数的拟合效果好;两市股指收益序列的尾部相关关系不对称,存在显著的下尾相关,而上尾相关不明显,而且上证综指和海峡时报指数收益序列的下尾相关系数是不断变化的,并呈现出日益上升的趋势。According to port-folio theory, port-folio should be made in more markets, hence it is important for global port-folio and risk management to research the tail dependence between two stock markets. 3 stationary Copula and time - varying SJC Copula functions and MLE based on ranks are used in tail dependence structure research between Shanghai composite index and Singapore＇ s Straits limes index. We find evidence that tail dependence between two markets is asymmetry. The lower tail dependence is stronger than the upper tail dependence and time - varying SJC Copula function is better than other 4 stationary Copula functions. By applying SJC Copula model, we find that tail dependence coefficient values are experiencing dynamic changes and Showing a rising trend.

关 键 词：上证综指 新加坡海峡时报指数 尾部相关性 COPULA函数 

分 类 号：F222.3[经济管理—国民经济]