基于Copula的金融资产风险度量研究

发布时间：2018-03-09 09:56

本文选题：风险度量分析　切入点：随机波动模型　出处：《湖南大学》2013年硕士论文　论文类型：学位论文

【摘要】：基于Copula的金融资产投资组合风险测度的应用研究是金融风险管理的前沿性课题.随着经济全球化的加快，国内的金融市场也变得日益复杂和多样化，金融市场风险度量的相关模式呈现出非正态、非对称和尾部相关等特性，Copula函数相较于以往基于正态分布假设的分析方法，对金融风险进行相关性分析有其特有的优势.本文将Copula理论结合描述边缘分布的随机波动模型，对中国两大能源股票板块的相关性进行分析，构建t-Copula-SV-t模型，然后结合风险度量理论计算风险值来观察股票的风险. 本文在理论上，首先介绍了随机波动模型以及Copula理论在国内外金融领域中的研究现状，并简要介绍了贝叶斯计量工具的应用.然后分章节对Copula函数、随机波动模型的相关内容作了详细的阐述，包括模型的介绍、分类、参数估计以及检验等方面.接着介绍了风险值的概念以及不同条件下的计算方法，为后面的风险度量做好理论准备. 在实证方面，本文选取上证能源股票和深证能源股票作为研究样本，首先分析数据的基本统计特征，然后运用随机波动模型对变量建立边缘分布，并通过检验比较并选择出最适合的SV-t模型；接着以选择的SV-t模型为边缘分布，建立不同类型的Copula模型，通过Copula密度函数图及理论评价均可看出t-Copula模型是几类模型中拟合效果最好的，在Copula模型的基础上计算组合的风险值. 最后本文对Copula理论在中国金融市场相关性研究以及风险度量分析中的应用进行总结，并提出了有待我们进一步研究的问题和展望.
[Abstract]:The research on the risk measurement of financial asset portfolio based on Copula is a leading subject of financial risk management. With the acceleration of economic globalization, the domestic financial market is becoming more and more complex and diversified. The related patterns of financial market risk measurement show the characteristics of non-normal, asymmetric and tail correlation. Compared with the previous analysis methods based on the normal distribution hypothesis, the Copula function is analyzed. This paper combines the Copula theory with the stochastic volatility model to describe the marginal distribution, analyzes the correlation between the two energy stocks in China, and constructs the t-Copula-SV-t model. Then combining the risk measurement theory to calculate the risk value to observe the stock risk. In theory, this paper first introduces the stochastic volatility model and the research status of Copula theory in the field of finance at home and abroad, and briefly introduces the application of Bayesian econometric tools. The related contents of the stochastic volatility model are described in detail, including the introduction, classification, parameter estimation and test of the model. Then, the concept of the risk value and the calculation methods under different conditions are introduced. Prepare for the following risk measurement. In the empirical aspect, this paper selects the energy stocks of Shanghai Stock Exchange and Shenzhen Stock Exchange as the research samples, analyzes the basic statistical characteristics of the data, and then uses the stochastic volatility model to establish the marginal distribution of the variables. The most suitable SV-t model is compared and selected by testing, and then different types of Copula models are established with the selected SV-t model as the edge distribution. Through the Copula density function diagram and the theoretical evaluation, it can be seen that t-Copula model is the best fit among several kinds of models. The risk value of combination is calculated on the basis of Copula model. Finally, this paper summarizes the application of Copula theory in the study of the correlation of Chinese financial market and the risk measurement analysis, and puts forward some problems and prospects for our further study.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224

【参考文献】