基于Copula函数股票相关性分析系统的设计与实现

发布时间：2017-12-31 12:31

本文关键词：基于Copula函数股票相关性分析系统的设计与实现　出处：《太原科技大学》2012年硕士论文　论文类型：学位论文

【摘要】：随着近年来金融市场的发展，金融风险的管理已经变得尤为突出，所以全面、准确的刻画金融市场之间的相关结构成为金融市场管理的重点。由于Copula函数导出的一致性和相关性测度可以反映变量之间非线性的相关关系，因此Copula函数作为一种灵活的非线性相关性研究工具已经在金融风险的建模中得到了广泛的应用。基于Copula函数的广泛应用，本文以Copula函数为分析资产收益率相关性的依据，使用MATLAB与VC混合编程的方法，设计并实现了基于Copula函数的股票相关性分析系统。本文的主要工作如下：首先，论文系统的回顾了Copula函数的发展历史背景和理论基础，简要的介绍了基于Copula函数的几种相关性测度指标以及用于构建Copula模型边缘分布函数的金融时间序列模型。然后，论文介绍了该系统的详细设计过程。简单的介绍了本文的总体研究思路，系统开发过程中所采用的数据库，重点介绍了Copula函数在MATLAB软件中的算法设计过程，将算法过程供VC调用来解决本系统中用到的计算问题。再次，论文重点介绍了该系统主要功能模块的设计与实现。该系统从两个角度来设计系统功能的，一方面该系统可以使用二维Copula函数计算两支股票的常相关系数也能计算两支股票的动态相关系数，另一方面，既能使用二维Copula函数计算两支股票之间的相关性，也能使用多维Copula函数计算多支股票之间的相关性。该系统能根据股票收益率自动选择较优的金融时间序列模型，并且可以根据不同Copula函数的特点，自动选择与股票间相关性拟合较好的Copula函数，从而定量地计算股票间相关性系数。最后，论文使用该系统以分析两支股票的相关性为例，简单介绍了该系统的使用过程。该系统操作简单、方便，可以为风险投资决策者提供很好的决策支持。
[Abstract]:With the development of the financial market in recent years, the financial risk management has become particularly prominent, so overall, the correlation structure between the accurate characterization of the financial market has become the focus of financial market management. Because of the relationship between Copula function to derive the consistency and correlation measure can reflect the nonlinear variable, so the Copula function as a flexible non linearcorrelation tools have been widely used in the modeling of financial risk. Applications based on the Copula function, the Copula function for the analysis of asset return correlation based method using mixed programming of MATLAB and VC, the design and implementation of the system of stock correlation analysis based on Copula function. The main work is as follows:
First, the paper systematically reviews the development history and theoretical basis of Copula function, and briefly introduces several correlation measure indexes based on Copula function and the financial time series model for building Copula model edge distribution function.
Then, the paper introduces the detailed design process of the system. Briefly introduces the general idea of this paper, the system development process in the database, introduces the algorithm design process of Copula function in MATLAB software, the algorithm for the VC to solve the problem of calculation used in this system.
Again, this paper mainly introduces the design and Realization of the main functional modules of the system. The system from two aspects to the design of system function, on the one hand, the system can often use two-dimensional Copula function to calculate the correlation coefficient of the two stocks two stocks also calculate the dynamic correlation coefficient, on the other hand, can be used two dimensional Copula function to calculate the correlation between the two stocks, can also use the Copula function to calculate the multidimensional correlation between multiple stocks. The system can automatically select according to the stock returns of financial time series model is better, and can according to the different characteristics of Copula function, Copula function and automatic selection of good correlation between the stock fitting. In order to quantitatively calculate the coefficient of correlation between stocks.
Finally, the paper uses the system to analyze the correlation between two stocks as an example, and briefly introduces the application process of the system. The system is simple and convenient, and it can provide good decision support for venture capital decision-makers.

【学位授予单位】：太原科技大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F224;F830.91

【参考文献】