尾部相关系数与资产选择研究

发布时间：2018-03-11 06:01

本文选题：Copula　切入点：尾部相关性　出处：《厦门大学》2014年硕士论文　论文类型：学位论文

【摘要】：相关性分析是一个重要的金融研究课题,被广泛运用于风险管理、资产定价和资产配置等领域。Copula函数与传统的相关分析工具相比,具有两个明显的优势：第一,它对金融市场的线性和非线性、对称和非对称的相关关系都能较好地描述；第二,它可分析变量间的相关结构。由Copula导出的下尾相关系数,可以度量股票下跌的风险,能为风险厌恶的投资者提供资产选择的有效信息。因此,本文研究尾部相关系数的资产选择问题有一定的理论和现实意义。本文首先介绍了Copula定义和类型、尾部相关系数的概念,以及基于极值模型下尾相关系数的计算方法。接着,本文借助Rietz的理论模型,分析了以下尾相关系作为资产选择指标的理论依据。在实证部分,本文借助二元阈值模型,以沪深300指数为大盘,计算其160支成分股与大盘、5000个资产组合与大盘的尾部相关系数,对尾部相关系数的稳定性、相关性和与股票市场表现的关系做了实证分析。结果表明,下尾相关系数是一个稳定性较高的且与其他风险指标相关度不高(除了Beta系数)；下尾相关系数低的股票的市场表现优于下尾相关系数高的股票。这具体表现在下尾相关系数低的股票,在熊市时期能抵御来自大盘下跌的风险；无论在任何时期,它们都具有较小的方差和较高的夏普比率；长期来说它们有较高累计收益率。最后,本文还是考察以上尾相关系数作为选股指标的效果,发现由于上尾相关系数与下尾相关系数之间有很高的相关性,上尾相关系数与下尾部相关系数选股效果相似。
[Abstract]:Correlation analysis is an important financial research topic, which is widely used in risk management, asset pricing and asset allocation. Copula function has two obvious advantages compared with traditional related analysis tools: first, It can well describe the linear and nonlinear, symmetric and asymmetric correlation of financial markets. Secondly, it can analyze the correlation structure between variables. The lower tail correlation coefficient derived by Copula can measure the risk of stock falling. It can provide effective information of asset selection for risk-averse investors. Therefore, it is of theoretical and practical significance to study the problem of asset selection with tail correlation coefficient in this paper. This paper first introduces the definition and type of Copula, the concept of tail correlation coefficient, and the calculation method of tail correlation coefficient based on extreme value model. The following tail phase relation is analyzed as the theoretical basis of asset selection index. In the empirical part, with the help of the dual threshold model and taking the CSI 300 index as the market, this paper calculates the tail correlation coefficient of 160 constituent stocks and the market, and 5000 asset combinations and the stock market, and the stability of the tail correlation coefficient to the tail correlation coefficient. The correlation and the relationship with the performance of the stock market are analyzed empirically. The lower tail correlation coefficient is a stable stock with low correlation with other risk indicators (except Beta coefficient); the market performance of stocks with lower tail correlation coefficient is better than that of lower tail correlation coefficient. Stocks with low tail correlation coefficient, In bear market times they can resist the risk of falling market; at any time they have smaller variance and higher Sharp ratio; in the long run they have higher cumulative yield. Finally, this paper investigates the effect of the above tail correlation coefficient as a stock selection index. It is found that the upper tail correlation coefficient is similar to the lower tail correlation coefficient because of the high correlation between the upper tail correlation coefficient and the lower tail correlation coefficient.
【学位授予单位】：厦门大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F224;F830.91

【参考文献】