文档介绍：华中科技大学
硕士学位论文
基于Copula-GARCH模型的偏股型基金的风险分析
姓名:张艳
申请学位级别:硕士
专业:数量经济学
指导教师:周少甫
20090501
华中科技大学硕士学位论文

摘要
现今随着金融市场不断发展,在金融市场的分析中相关性的研究占有越来越重
要的地位。如今金融市场越来越关注组合的风险,投资者的需求已经不能满足于单
一资产的风险度量,而精确的金融资产间相关结构度量直接决定着 VaR 的度量风险
的准确性。基于线性相关系数的分析方法基本已经不能准确反映金融市场的相关关
系,金融市场和金融资产间的相关性形式越来越复杂多样,它们大多都呈现出非线
性和非对称性等相关模式,这一难题正好由 Copula 函数来解决。Sklar 认为用 Copula
函数可以描述随机变量间的相关性,而各变量的边缘分布由特定模型来确定。
本文主要研究 Copula 函数理论及其在金融市场的数量分析中的应用。在对
Copula 函数理论进行了系统的阐述之后,本文用二元 Clayton Copula 函数与 GARCH
模型相结合构成 Copula-GARCH 模型,由于该模型能够描述金融市场间的非线性相
关性,因而可用于投资组合的风险度量。实证部分以中信红利精选[288002]2009 年
第一季度的前十只权重股票为研究对象,然后结合 Copula-GARCH 模型和 Monte
Carlo 模拟法来计算出风险价值 VaR。结果证实本文中提出的模型和方法是可行的、
有效的。

关键词: Copula 函数 VaR Copula-GARCH 模型
I
华中科技大学硕士学位论文

Abstract
With the development and improvement of financial markets, correlation analysis is
playing an increasingly important role in quantitative analysis. Since the way of risk
measuring of single assets can no longer meet the demand of investment, more attention
has been focused on the risk of portfolio, and the accuracy of VaR measurement largely
depends on that of correlation structure. The plicated correlation between
financial markets and financial assets shows non-linear or asymmetry model, while
analysis based on linear coefficient of correlation is obsolete as to accurately reflect the
correlation of financial market. Copula function theory just provides the right therapy.
Sklar's Copula theory holds that the correlation of random variables can be decided by the
Copula function, and the demographic characteristics of the variables can be determined
by the marginal distribution.
In this paper, the Copula Function Theory and its application in the mathematic
analysis of financial markets are mainly analyzed. The Copula function theory is
elaborated, and the Copula-GARCH model is structured through bination of
binary Clayton Copula function and GARCH model. The Copula-GARCH model can be
used for i