沪深股市收益率及其相关性的实证分析

发布时间：2018-03-27 11:13

本文选题：沪深股市　切入点：收益率分布　出处：《天津财经大学》2013年硕士论文

【摘要】：股指的波动特征无外乎单个股指的波动特征与多个股指间的波动特征,关于单个股指波动特征的研究,即单个股指收益率分布的研究,本文基于非参数核密度估计方法,充分利用样本数据本身特征来拟合股指收益率密度曲线,有效降低了参数模型中因既定假设分布所引起的拟合误差；关于多个股指间波动特征的研究,即不同股指间波动相关性的研究,本文使用连接函数与核密度估计相结合,即在不假设边缘分布的条件下,利用半参数方法求解连接函数中的参数,有效提高了拟合精度。在以上研究方法体系之上,本文对中国沪深股市收益率及其相关性进行了不同层次与时间段的实证分析,得出相关结论并提出政策建议。股指收益率方面,将大盘、板块及个股三者收益率分布情况联系起来分析,得出结论：核密度估计可以灵活准确拟合股指收益率曲线；宏观、中观、微观三层面的收益率分布情况均为非正态,存在尖峰厚尾；宏观、中观收益率分布有更为突出的左侧厚尾,说明更多投资者是风险厌恶型,对股市中利空消息表现出更为强烈的协同效应；微观个股表现出左右厚尾情况均存在的复杂现象,说明上市公司的信誉和行业背景影响投资者风险偏好。股指相关性方面,本文完成了1997年1月至2012年7月沪深两市大盘指数相关性的静态分析,并对2008年9月金融危机前后沪深两市相关性特征进行了对比分析,得出结论：上尾更为紧凑的Gumbel Copula函数基本能够捕捉沪深两市相关结构的特征,即当一个市场指数出现高收益率时,另一个市场出现高收益率指数的可能性也很大,出现低收益率时虽然也有类似效应,但不如高收益率时的范围区间广。由对比分析可知,相关程度方面,金融危机后两市的相关依存度较危机前明显增强,说明沪深两市在整体上熊市相关依存度更强；相关结构方面,危机前后沪深两市的相关结构并没有发生非常明显的变化,说明由美国次贷危机引发的全球金融危机对我国股市的冲击并没有改变沪深股市整体相关结构。
[Abstract]:The volatility characteristic of the index finger is not related to the fluctuation characteristic of single index finger and the fluctuation characteristic between multiple index fingers , namely the study of single index fluctuation characteristic , that is , the study of single index yield distribution . Based on the non - parameter kernel density estimation method , the paper uses the characteristic of sample data itself to fit the curve of the yield density curve of stock index , effectively reduces the fitting error caused by the established hypothesis distribution in the parameter model ;
In this paper , we study the relationship between the fluctuation characteristics of multiple index fingers , that is , the correlation between different index fingers . In this paper , the connection function is combined with kernel density estimation , that is , under the condition that the edge distribution is not assumed , the parameters in the connection function are solved by using the semi - parametric method , and the fitting accuracy is effectively improved .

Based on the above research methods , this paper analyzes the yield and its correlation of Shanghai - Shenzhen stock market at different levels and time periods , and draws relevant conclusions and puts forward policy recommendations .

Based on the analysis of the yield distribution of stock index , plate and stock , it is concluded that nuclear density estimation can fit the yield curve of stock index flexibly and accurately .
Macro - , meso - and micro - level yield distributions are non - normal , and there is a peak - thick tail .
Macro - and middle - view yield distributions have a more prominent left - hand thick tail , which shows that more investors are risk - averse and show a stronger synergy in the stock market .
The micro - stock shows the complex phenomenon in the case of left and right thick tail , which indicates that the reputation and the background of the listed company affect the investor ' s risk appetite .

Based on the analysis of the correlation between Shanghai and Shenzhen in Shanghai and Shenzhen from January 1997 to July 2012 , the paper concludes that the more compact Gumbel Copula function can capture the characteristics of the related structures in Shanghai and Shenzhen .
There is no obvious change in the structure of Shanghai and Shenzhen in Shanghai and Shenzhen before and after the crisis , which shows that the impact of the global financial crisis caused by the U.S . subprime crisis on China ' s stock market has not changed the overall related structure of Shanghai and Shenzhen stock market .

【学位授予单位】：天津财经大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224

【参考文献】