时变copula函数及其在股指相关性研究中的应用
发布时间：2018-10-25 09:02
【摘要】：本文研究了copula函数的构造，给出了水平(垂直)截面为有理函数截面时的copula构造方法，证明了不存在水平(垂直)截面为二次有理函数的阿基米德copula函数。并使用多种copula函数及时变copula函数，，研究我国股票市场中不同市场行情下上证指数、中小扳指、沪深300指数的相依性结构及尾部相关性。实证结果表明：指数间的下尾相关程度普遍大于上尾相关程度。在牛市行情下，symmetrised-Joe Clayton copula函数能够较好的拟合三个指数之间的相依结构；而在熊市以及震荡市的行情下，t-copula函数则能够更好的描述三个指数之间的相依性结构。使用时变的symmetrised-Joe Clayton copula函数描述在不同的市场行情下股票市场三指数间的尾部相关性，结果表明：上证指数和沪深300指数均具有稳定的尾部相关性，在牛市行情下，上证指数与中小板指数的上尾相关系数具有较为明显的波动，在熊市行情下，上证指数与中小板指数的下尾相关系数具有较为明显的波动，而在振荡盘整行情下，上证指数与中小板指、中小扳指与沪深300指数的上尾相关系数与下尾相关系数都具有明显的波动。
[Abstract]:In this paper, we study the construction of copula function, give the copula construction method when the horizontal (vertical) section is a rational function section, and prove that there is no Archimedes copula function where the horizontal (vertical) section is a quadratic rational function. Using various copula functions to change the copula function in time, this paper studies the dependence structure and tail correlation of Shanghai Stock Exchange Index, small and medium index, Shanghai and Shenzhen 300 index under different market prices in our country. The empirical results show that the degree of lower tail correlation is higher than that of upper tail correlation. In the bull market, the symmetrised-Joe Clayton copula function can fit the dependence structure of the three indices better, while in the bear market and the shock market, the t-copula function can better describe the dependency structure between the three indices. The time-varying symmetrised-Joe Clayton copula function is used to describe the tail correlation between the three indices in different market prices. The results show that both the Shanghai Stock Exchange Index and the Shanghai and Shenzhen 300 Index have stable tail correlation. The upper tail correlation coefficient of Shanghai stock index and small and medium-sized board index has obvious fluctuation. In the bear market, the lower tail correlation coefficient of Shanghai stock index and small and medium-sized board index has obvious fluctuation, but under the oscillatory consolidation market, The upper tail correlation coefficient and lower tail correlation coefficient of Shanghai Stock Exchange index and small board index, middle and small index and Shanghai and Shenzhen 300 index have obvious fluctuation.
【学位授予单位】：暨南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F830.91;O174
本文编号：2293238
[Abstract]:In this paper, we study the construction of copula function, give the copula construction method when the horizontal (vertical) section is a rational function section, and prove that there is no Archimedes copula function where the horizontal (vertical) section is a quadratic rational function. Using various copula functions to change the copula function in time, this paper studies the dependence structure and tail correlation of Shanghai Stock Exchange Index, small and medium index, Shanghai and Shenzhen 300 index under different market prices in our country. The empirical results show that the degree of lower tail correlation is higher than that of upper tail correlation. In the bull market, the symmetrised-Joe Clayton copula function can fit the dependence structure of the three indices better, while in the bear market and the shock market, the t-copula function can better describe the dependency structure between the three indices. The time-varying symmetrised-Joe Clayton copula function is used to describe the tail correlation between the three indices in different market prices. The results show that both the Shanghai Stock Exchange Index and the Shanghai and Shenzhen 300 Index have stable tail correlation. The upper tail correlation coefficient of Shanghai stock index and small and medium-sized board index has obvious fluctuation. In the bear market, the lower tail correlation coefficient of Shanghai stock index and small and medium-sized board index has obvious fluctuation, but under the oscillatory consolidation market, The upper tail correlation coefficient and lower tail correlation coefficient of Shanghai Stock Exchange index and small board index, middle and small index and Shanghai and Shenzhen 300 index have obvious fluctuation.
【学位授予单位】：暨南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F830.91;O174
【参考文献】
相关期刊论文 前5条
1 韦艳华,张世英,孟利锋;Copula理论在金融上的应用[J];西北农林科技大学学报(社会科学版);2003年05期
2 孙炳X,史道济;二元Cauchy分布尾部的相关性[J];天津大学学报;2000年04期
3 王沁;王璐;程世娟;;基于时变Copula模型的沪深股市相依分析[J];统计与决策;2010年19期
4 朱世武;基于Copula函数度量违约相关性[J];统计研究;2005年04期
5 刘曦;王沁;易文德;江松明;;基于核密度估计的时变copula函数在铜期货套期保值的研究[J];时代金融;2014年06期
相关博士学位论文 前1条
1 胡铮洋;证券市场风险度量—时变Copula和极值Copula的应用研究[D];吉林大学;2009年
本文编号：2293238
本文链接：https://www.wllwen.com/jingjilunwen/touziyanjiulunwen/2293238.html
