金融资产间的相关性分析研究

发布时间：2018-01-09 08:13

本文关键词：金融资产间的相关性分析研究　出处：《温州大学》2014年硕士论文　论文类型：学位论文

【摘要】：相依性研究是金融领域中的一个重要问题，投资组合、资产定价和风险管理等问题都与相依性研究有关。随着金融市场的快速发展和经济全球化的不断深入，金融数据表现出的分布越来越不规则，金融资产间的相关性越来越复杂。作为统计学中相关性分析的有力工具，Copula函数不仅能够较好刻画随机变量间的相关性，还能够将相关关系和数据的分布分开来研究，这为研究金融资产间的相关性提供了方便，也为人们理解资产之间的相关性提供了理论基础。鉴于非对称Laplace分布具有尖峰厚尾及有偏等特性，能够较好地刻画金融数据的分布。本文以Copula函数和非对称Laplace分布为理论基础，，尝试地做了如下几方面的工作：（1）对于非对称Laplace分布是否能作为金融资产收益的密度估计，做了实证分析，通过图形拟合检验，得出非对称Laplace分布能作为资产收益分布的密度估计，并以此作为后文Copula理论建模的边缘分布。（2）探讨了混合Copula函数与成分Copula函数间相关性测度的关系；选用Gumbel、Clayton和Frank Copula函数的线性组合构建了混合Copula函数，用实证分析的方法，研究了混合Copula函数在描述相关性方面的灵活性和准确性，发现混合Copula函数更能准确灵活地描述我国沪深股市间的相关性。（3）金融资产间的相关性分析是金融领域研究中的一个重要问题，以上证指数和恒生指数为例，本文选取了Gumbel、Clayton、Frank和C Copula函数，对金融资产间的相关性进行了分析，分析发现Clayton Copula函数最适合描述上证指数和恒生指数间的相关性。最后，总结和分析了本文的研究工作，并给出了下一步研究的课题。
[Abstract]:Dependency research is an important issue in the field of finance, investment portfolio, asset pricing and risk management problems are related to dependency research. With the rapid development of financial market and the deepening of economic globalization, the distribution of financial data showed more and more irregular, and the correlation between financial assets is more complicated. A powerful tool for the analysis of statistical correlation, the Copula function can not only describe the correlation between random variables, can also be studied separately and related distribution data, the correlation study between financial assets to provide a convenient, but also provides the theoretical basis for people to understand the correlation between assets. In view of the asymmetric Laplace distribution with leptokurtic and partial characteristics, can better describe the distribution of financial data. The theory based on Copula function and asymmetric Laplace distribution On the basis of this, we have tried to do the following work:
(1) for asymmetric Laplace distribution can serve as a financial asset return distribution estimation, do the empirical analysis by graph fitting test, the asymmetric Laplace distribution as asset return distribution and density estimation, which is the Copula theory modeling of marginal distribution.
(2) to investigate the relationship between the correlation measure of mixed Copula function and composition of Copula function; using Gumbel, Clayton and Frank linear combination of Copula functions to build a hybrid Copula function, use the methods of empirical analysis, study the mixed Copula function in describing the correlation of flexibility and accuracy, found that the mixed Copula function more accurately and flexibly the relationship between China's Shanghai and Shenzhen stock market.
(3) the correlation analysis between financial assets is an important issue in the financial field, with the Shanghai Composite Index and the Hang Seng Index as an example, this paper selects Gumbel, Clayton, Frank and C Copula function, the correlation between financial assets are analyzed. The analysis showed that the Clayton Copula function is most suitable to describe the correlation between the Shanghai index and the Hang Seng Index.
Finally, the research work of this paper is summarized and analyzed, and the next research topic is given.

【学位授予单位】：温州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F830.9

【参考文献】