基于双参数Copula沪港股市的相关性分析  

Dependence Analysis of SZI & HSI Based on the Two-Parameter Copula

作　　者：王玥[1] 程希骏[1] 马利军[2] 

机构地区：[1]中国科学技术大学管理学院,安徽合肥230051 [2]深圳大学管理学院,广东深圳518060

出　　处：《数学的实践与认识》2011年第17期1-7,共7页Mathematics in Practice and Theory

基　　金：中国科学院知识创新工程重要方向项目(KJXC3-SYW-S02)

摘　　要：通过双参数Copula分析上证指数和恒生指数的尾部相关性,并与单参数Copula及混合Copula进行比较分析,参数估计使用半参数估计法,结果表明:与单参数Clayton Copula、Gumbel-Hougaard Copula以及由两者组成的混合Copula相比,双参数BB1 Copula对数据具有更好的拟合效果;且通过分析发现两股市的上尾相关性大于下尾相关性.The tail dependence of SZI and HSI analysis results are compared with that of using is studied by two-parameter copulas, and the one-parameter copulas and mix copula which is composed of two one-parameter copulas, and the method of semiparametric estimation is used to estimate parameters. The results show that the BB1 copula has the best Goodness- of-fit tests among the copulas which we selected; and we also found that the two stock markets have larger upper tail dependence than lower tail dependence.

关 键 词：双参数Copula 半参数估计法 尾部相关性 

分 类 号：F224[经济管理—国民经济] F832.51