基于变结构Copula模型的相依关系分析  

Dependent Analysis Based on Variable Structure Copula Model

作　　者：王沁[1] 

机构地区：[1]西南交通大学数学学院,四川成都610031

出　　处：《数理统计与管理》2012年第2期341-347,共7页Journal of Applied Statistics and Management

基　　金：2009教育部人文社会科学研究项目基金资助(编号09YJCZH104);西南交通大学"希望之星"资助;中央高校基本科研业务费专项资金资助(编号SWJTU09CX075;SWJTU09ZT37)

摘　　要：本文首次从Spearman的rho出发,提出了一种检验Copula模型是否存在变结构特征的方法,并通过蒙特卡洛模拟验证了这种方法的有效性。在此基础上,利用分阶段建模技术和Spearman的rho对沪深股市的相依关系是否存在变结构进行了分析。结果证实,利用Spearman的rho能捕捉序列之间相依关系是否存在变结构的信息,具有一定的优越性。In the paper, based on Spearman＇s rho, a method by which a Copula model with the characteristics of variable structure can be diagnosed is put forward at the first time. Though Monte Carlo Simulation （MC） technology, the validity of the diagnostic method is verified. By using of the staged modeling technique and Spearman＇s rho, it is analyzed whether or not the dependency structure between Shanghai and Shenzhen stock markets possesses the characteristics of variable structure. The results confirm that the dependent relationship of variables possessed the characteristics of variable structure can be captured by using Spearman＇s rho. The diagnostic method using Spearman＇s rho has a certain degree of superiority.

关 键 词：Spearman的rho 变结构Copula模型 蒙特卡洛模拟 沪深股市 

分 类 号：O211.3[理学—概率论与数理统计]