基于非参数时变Copula模型的美国次贷危机传染分析

发布时间：2018-01-19 13:16

本文关键词： 金融危机传染 Copula 尾部相依局部极大似然估计　出处：《中国科学技术大学》2015年硕士论文　论文类型：学位论文

【摘要】：全球经济金融一体化和自由化的持续深入带来了金融创新的飞速发展,大规模的金融危机开始日益频繁地发生,而显著的传染性是这些危机最主要的特征。在以往的研究中,不同种类的模型被开发出来用于检验金融危机传染是否存在,但是这些模型普遍存在某些方面的缺陷：一些模型忽略了市场之间非线性以及非对称的复杂相依结构；另外一些模型只能够对金融危机传染进行静态分析,而不能够动态的研究金融危机的发展过程,而这往往是投资者更关心的问题,因此如何解决以上两个问题成为了本文的研究焦点。本文最主要的贡献是将非参数时变Copula方法引入到金融危机传染的研究中,并根据模型的性质以及优点来分析金融危机传染发展过程。文章首先详细地介绍了金融危机传染领域的国内外研究现状,其中包括静态模型和动态模型；接着系统地给出了Copula理论的具体内容,其中包括Copula的定义以及性质,常用的一些Copula函数,Copula函数的参数估计方法及尾部系数的估计方法。紧接着,本文利用非参Copula方法检验了金融危机传染,提出了一种非对称、非线性的金融危机传染方法,并且引入了非参数时变Copula模型,实证结果表明,Copula模型在研究金融危机传染的发展历程中是一种非常有效的模型。最后,文章将非参数时变Copula的金融危机传染模型应用到美国次贷危机传染的分析中,结果表明此方法可以较好的研究危机传染,一方面该模型弥补了之前一些研究中不足的地方,因为通过尾部相关系数能够刻画金融市场指数之间非线性、非对称的相依结构,而且可以得到动态的尾部相依系数,这是本文最重要的贡献,它有效地解决了现有研究中难以精确地度量金融危机传染问题,丰富了现有的金融危机传染研究体系。
[Abstract]:The continuous deepening of global economic and financial integration and liberalization has brought the rapid development of financial innovation, large-scale financial crisis began to occur more and more frequently. In previous studies, different types of models have been developed to test the existence of contagion in financial crises. But there are some defects in these models: some models ignore the nonlinear and asymmetric complex dependent structures between markets; Other models can only be static analysis of financial crisis contagion, but can not dynamically study the development process of financial crisis, which is often more concerned by investors. Therefore, how to solve the above two problems has become the focus of this paper. The main contribution of this paper is to introduce the nonparametric time-varying Copula method into the study of financial crisis contagion. And according to the nature and advantages of the model to analyze the financial crisis contagion development process. Firstly, this paper introduces the domestic and foreign research status of financial crisis contagion field in detail, including static model and dynamic model; Then the specific contents of Copula theory are given systematically, including the definition and properties of Copula, and some commonly used Copula functions. The parameter estimation method of Copula function and the method of tail coefficient estimation. Then, this paper uses the non-parametric Copula method to test the financial crisis contagion, proposes an asymmetric, nonlinear financial crisis contagion method, and introduces a non-parametric time-varying Copula model. The empirical results show that Copula model is a very effective model in the study of financial crisis contagion. Finally, the non-parametric time-varying Copula financial crisis contagion model is applied to the analysis of sub-prime crisis contagion in the United States. The results show that this method can better study crisis contagion. On the one hand, the model makes up for the shortcomings of some previous studies, because the tail correlation coefficient can describe the nonlinear and asymmetric dependent structure of the financial market index. And the dynamic tail dependency coefficient can be obtained, which is the most important contribution of this paper. It effectively solves the problem of financial crisis contagion that is difficult to measure accurately in the existing research. Has enriched the existing financial crisis contagion research system.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F224;F837.12

【参考文献】