Copula理论在我国上市公司债券与股票间相关结构分析中的应用

发布时间：2018-04-23 20:14

本文选题：联动效应 + Copula　；参考：《华中科技大学》2012年硕士论文

【摘要】：金融风险间的联动效应是近年来学术界和实务界关注的热门课题，在此背景下，本文重点探讨中国金融市场中信用风险与市场风险间的联动效应，以考察我国金融市场的风险配置效率。考虑到两者之间可能存在的非线性相关关系，本文应用研究非线性关系强有力的统计技术Copula函数来研究两者之间的相关结构，本文的研究对我国的金融市场参与者具有一定的参考价值。本文首先介绍了Copula函数的概念和性质，并重点研究了几种重要的相关性测度，接着对在本文中有重要应用的三种Archimedean Copula函数进行了详细介绍，并研究了它们在金融市场相关性分析中的特点。最后，本文介绍了Copula函数的参数估计以及模型选择理论。在实证研究部分，本文利用我国47家发行有公司债的上市公司2010年1月1日至2011年4月1日的交易数据，结合Copula函数相关理论，以债券利差和股票收益波动率分别作为信用风险和市场风险的代理变量，考察了我国上市公司债券利差与股票收益波动率间的相关结构，试图找到我国金融市场中不同风险之间会产生联动效应的证据，但实证分析发现，，两者之间并不存在确定的相关关系，本文对此做出了解释并给出了相应的政策建议；研究还发现，信用评级低的公司两者之间更倾向于正相关，表明这类公司更易产生风险联动效应；最后，在模型选择部分，研究发现，Gumbel Copula是对现实数据拟合能力最好的模型。
[Abstract]:In recent years, the linkage effect of financial risk is a hot topic in academic and practical circles. Under this background, this paper focuses on the linkage effect between credit risk and market risk in Chinese financial market. In order to investigate the risk allocation efficiency of our financial market. Considering the possible nonlinear correlation between the two, this paper applies the powerful statistical technique Copula function to study the correlation structure between the two. The research of this paper has certain reference value to the financial market participant of our country. In this paper, the concept and properties of Copula function are introduced, and several important correlation measures are studied. Then, three kinds of Archimedean Copula functions which have important applications in this paper are introduced in detail. And studied their characteristics in the financial market correlation analysis. Finally, the parameter estimation and model selection theory of Copula function are introduced. In the part of empirical research, this paper uses the transaction data of 47 listed companies with corporate bonds issued in our country from January 1, 2010 to April 1, 2011, and combines with the relevant theory of Copula function. Taking bond spread and stock return volatility as proxy variables of credit risk and market risk respectively, this paper investigates the correlation structure between bond interest margin and stock return volatility of listed companies in China. This paper tries to find evidence that there is linkage effect between different risks in China's financial market, but the empirical analysis shows that there is no definite correlation between them. It is also found that the companies with low credit rating tend to be positively correlated, indicating that they are more prone to risk linkage. Finally, in the selection of models, it is found that Gumbel Copula is the best model to fit the real data.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.5;F224

【参考文献】