中美主要金融市场相关结构及风险传导路径研究--基于Copula理论与方法  

Study on the Correlation Structure and the Transmission Path of Risks between the Financial Markets in China and America — Based on Copula Theory and Method

作　　者：黄在鑫[1] 覃正 

机构地区：[1]上海财经大学信息管理与工程学院 [2]南方科技大学

出　　处：《国际金融研究》2012年第5期74-82,共9页Studies of International Finance

基　　金：国家自然科学基金项目“应对国际金融风险的危机-机遇模式研究”(70971083);211项目基金(211-5-1);教育部2009年博士点基金(20090078110001)资助

摘　　要：金融危机背景下的股市表现出更加复杂的动荡性,本文在传统GARCH模型的基础上引入了风险值对收益率的影响因素,运用GARCH-M模型来刻画股票收益率序列边缘分布,通过构建GARCH-M-t边缘分布过滤模型获取收益率残差序列,最后采用Copula函数对边缘分布拟合后的残差序列建模构建出Copula-GARCH-M-t相关结构模型。经过参数估计及多种Copula函数的拟合优度检验,最终成功刻画出中美金融市场五大证券交易中心股票收益率之间的相关结构模型。通过秩相关系数、尾部相关系数等相关性度量工具对中美两国金融市场的相关性进行分析,最后通过对不同股票市场之间的尾部相关性分析确定两国金融市场之间风险传导路径。The financial markets appear to be more unstable against the background of the financial risks. This paper uses GARCH-M model to fit the marginal distribution of the benefit rate instead of the traditional GARCH model which does not consider the risk factor. Through the GARHC-M-t model, we gain the residuals series, based on which we construct the Copula-GARCH-M-t model. After fitting the constructed Copula-GARCH-M-t model, we found that the model with the GARCH-M marginal distributions fits the residual series quite well. After estimating the parameters of the Copula functions, we choose the most accurate copula model for these series, and finally got the model of the correlation structure of the five benefits rate. We use the analysis tools of rank correlation coefficient and correlation coefficient of tail to analyze the correlation between the financial markets of China and America, finally found the way that the risk is infected between the two countries.

关 键 词：COPULA GARCH—M相关性度量 金融风险 关联模型 

分 类 号：F831[经济管理—金融学]