c-D-Copula模型构建及其在金融风险传染中的应用
Linkage of International Stock Market and Financial Infection Analysis Based on c-D-Copula Model
金融危机的频繁发生, 使得国际金融传染的研究成为了一个重要的课题. 文章在动态Copula模型的基础上, 对其参数的动态演变方程进行推广, 借鉴c-DCC模型的思想假定其截距项存在结构突变, 构建了c-D-Copula模型. 为了避免人为选择结构性变化时间, 对c-D-Copula模型进行变点检测, 由数据驱动自动选择断点. 为了进行比较分析, 同时给出了动态演变方程的参数全部变化的变结构全模型, 并给出了相应的假设检验方法. 文章选择美国标准普尔500指数和全球主要5个国家(地区)的股票指数的对数收益率进行实证研究, 通过检验得出大多数情况下用c-D-Copula模型比变结构全模型更合适对数据进行描述. 通过这5个国家或地区与美国之间联动性的动态变化, 对美国次贷危机金融传染的存在性和变化过程进行了研究, 变点时刻和下尾部相依系数的实证结果表明所分析的国家和地区都受到了美国次贷危机的传染.
The frequent occurrence of financial crisis makes the research on the financial contagion become an important subject. Based on the dynamic Copula model, we promote the dynamic evolution equation of the parameters, which refers to c-DCC model and assume that there is a structural break in the intercept term to establish c-D-Copula model. In order to avoid the artificial choice of structural breaks, change point test is applied on the c-D-Copula model to change the point of detection, and the data is driven by the data automatically to select the change point. In order to carry on the comparative analysis, we also analyze the full-model of which all parameters of the dynamic evolution equation have structural change, and the corresponding hypothesis test method is also given. This paper analyzes the American Standard \& Poor's 500 index and the stock index of the global five major countries (regions). Through the test, in most cases, c-D-Copula model is more appropriate to describe the data than thefull-model. Through changes of the linkage between the five countries or regions and the United States, we analyze the existence and the process of changes of the financial crisis of U.S. subprime mortgage crisis. The empirical results of the change points and the lower tail dependency coefficient show that the analyzed countries and the region are subject to the U.S. subprime mortgage crisis.
/
〈 | 〉 |