基于Spearman的rho的Copula参数模型的选择  

Choice of the Copula Parameter Model Based on Spearman's Rho

作　　者：王沁[1] 王璐[1] 袁代林[1] 

机构地区：[1]西南交通大学数学学院,四川成都610031

出　　处：《数学的实践与认识》2011年第15期145-150,共6页Mathematics in Practice and Theory

基　　金：2009年教育部人文社会科学研究项目基金(09YJCZH104);中央高校基本科研业务费专项资金(SWJTU09CX075;SWJTU09ZT37);2010年教育部人文社会科学研究基金(10YJCZH157)

摘　　要：从Spearman的rho与Kendall的tau的关系入手,讨论了一类二元Copula参数模型的选择问题.由于这类二元Copula参数模型的Spearman的rho与Kendall的tau存在某种函数关系,模型选择问题转化为了曲线拟合检验问题.对于正态Copula、Frank-Copula,FGM-Copula、B11-Copula等这类Copula参数模型,说明了两种情况下进行模型选择的方法,并对中国股市的上证指数与深证综指作了实证分析,结果表明两者存在着较强的正相关性,相关性模型选取B11-Copula参数模型最合适.Based on Spearman＇s rho and Kendall＇s tau relationship, the method of choice Copula parameter model is put forward firstly There are lots of binary Copula parameter models whose Spearman＇s rho and Kendall＇s tau exist a type of function relationship, so the method of choice Copula parameter model is transferred to curve-fitting testing problem. Two methods of choice Copula model are given for several of the type Copula models such as the normal Copula, Frank-Copula, FGM-Copula, B11-Copula, etc At last, the empirical analyses of dependent mechanism between Shanghai and Shenzhen stock index have been done based on the choice method of Copula model. The results showed that there existed a strong positive correlation between Shanghai and Shenzhen stock index, and the Bll-Copula model is the most appropriate model which describes the characteristic.

关 键 词：Copula参数模型 Spearman的rho Kendall的tau 曲线拟合 蒙特卡洛 模拟 

分 类 号：O242.1[理学—计算数学]