基于尾部变结构Copula模型的股市波动溢出效应研究
本文关键词:基于尾部变结构Copula模型的股市波动溢出效应研究 出处:《浙江工商大学》2013年硕士论文 论文类型:学位论文
更多相关文章: Copula 尾部变结构 相依结构 波动溢出效应
【摘要】:随着中国经济的快速发展和市场的逐步开放,中国股市受到世界主要股市的传染和影响日益加深。近年来,由于金融市场波动的日益剧烈与金融危机的频发,各国金融市场之间的波动溢出效应研究越来越受到国内外学者的密切关注。以往传统的相关关系度量方法只能度量变量间的线性相关关系,越来越不能适应人们对于相关性度量的需要,因此,本文引入可以度量非线性相关关系的度量工具一Copula函数来建立非线性相关性模型。 以往相关研究较多的是运用单一Copula函数,并不能全面地描述金融变量间的相依结构。但是,相对单一Copula函数,混合Copula函数可以包含不同类型的Copula,能够更加灵活地度量变量之间的相依性。基于此,本文通过线性组合方式构建一个具有尾部变结构的混合Copula模型。该混合Copula函数不仅能度量变量之间的上尾和下尾相关关系,还能同时度量变量间的相依程度和相依模式。同时,本文另一创新之处是考虑了混合Copula模型权重参数的时变性,而Copula函数参数是不变的,权重参数的时变性能够更灵活地捕捉到变量之间相依模式的动态性。 本文在数据上选取2000年1月至2013年1月上证综指、香港恒生指数、美国标普500指数和日经225指数,研究中国股市与这些股市之间相依性的变化以及是否存在风险溢出效应的特征。首先使用AR(1)-GARCH(1,1)-t模型来模拟指数收益率序列的边际分布,然后使用静态formal Copula, Gumbel Copula,Clayton Copula, SJC Copula函数和本文构建的具有尾部变结构特征的时变混合Copula函数来对中国大陆股市与国际主要股市间的风险传染及相依特征进行建模,并对实证结果进行分析。实证结果表明,十多年来,在中国资本市场逐步对外开放进程中,上证指数与国际主要指数之间的联动性并不强,上证综指对国际主要股市的影响力还较弱。不过,虽然中国股市与其他四个股市间的尾部相依程度较低,但是当市场在面对极端情况时,例如2008年的次贷危机,仍然有同时发生大起或大落的可能,这在一定程度上反映了大的金融危机爆发时对全球股市的影响,特别是由美国引起的金融危机。
[Abstract]:With the rapid development of China gradually open economy and market, the stock market has been the world's major stock markets Chinese infectious and influence is growing. In recent years, due to the frequent fluctuations in the financial market is becoming increasingly fierce and the financial crisis, the volatility spillover effect between financial markets research pay more and more attention of scholars at home and abroad. The past relationship the traditional measurement method can only measure the linear relationship between the variables, increasingly unable to meet the needs of the people for, so we introduce the correlation metric, a measure tool can measure the nonlinear relationship between the Copula function to establish the nonlinear correlation model.
The more research is the use of a single Copula function, and can not fully describe the dependence structure of financial variables. However, relative to a single Copula function, mixed Copula function can contain different types of Copula, can be more flexible to measure the dependency between variables. Based on this, this paper constructs a hybrid Copula model with variable tail structure by linear combination method. The mixed Copula function can not only measure the variables between the upper and lower tail correlation, but also measure the dependence between variables and dependent mode. At the same time, this is another innovation of the modified hybrid Copula model weight parameters, and the parameters of Copula function is constant. Time-varying weight parameters can be more flexible to capture the dynamic dependencies between variables of the model.
In this paper, from January 2000 to January 2013 the Shanghai Composite Index in the data selection, the Hongkong Hang Seng Index, the S & P 500 and the Nikkei 225 index, the stock market and the characteristics between China stock dependent changes and the existence of Risk Spillover effect. The first use of AR (1) -GARCH (1,1) -t model to simulate the marginal distribution of index returns sequence, and then use the static formal Copula, Gumbel Copula, Clayton Copula, SJC Copula function and the tail has the characteristics of variable structure and time-varying mixed Copula function on the mainland stock market and the model dependent characteristics and risk contagion between Chinese main international stock markets, and analyzed the results of empirical analysis. The empirical results show that, more than 10 in China years, the capital market gradually in the process of opening up, the linkage between the Shanghai index and international index is not strong, the Shanghai Composite Index of major international The influence of the stock market is still relatively weak. However, although the stock market and the other four Chinese tail stock market's dependence degree is low, but when the market in the face of extreme conditions, such as the 2008 subprime crisis, there are still large or occur at the same time big potential, impact on the global stock market to a certain extent reflects the outbreak of the financial crisis, especially the United States caused by the financial crisis.
【学位授予单位】:浙江工商大学
【学位级别】:硕士
【学位授予年份】:2013
【分类号】:F832.51;F224
【参考文献】
相关期刊论文 前10条
1 田光;张瑞锋;;基于Copula的股票市场波动溢出分析[J];财经理论与实践;2011年06期
2 韦艳华,张世英;金融市场的相关性分析——Copula-GARCH模型及其应用[J];系统工程;2004年04期
3 韦艳华;齐树天;;亚洲新兴市场金融危机传染问题研究——基于Copula理论的检验方法[J];国际金融研究;2008年09期
4 黄在鑫;覃正;;中美主要金融市场相关结构及风险传导路径研究——基于Copula理论与方法[J];国际金融研究;2012年05期
5 刘平;杜晓蓉;;对金融危机风险传染效应的比较研究——基于静态与动态copula函数的分析[J];经济经纬;2011年03期
6 赵留彦,王一鸣;A、B股之间的信息流动与波动溢出[J];金融研究;2003年10期
7 孙清;蔡则祥;;基于COPULA函数的信贷资产风险估计模型[J];南京社会科学;2007年07期
8 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
9 李勇;李传乐;;风险传染效应在牛市、熊市中的异化现象——来自A+H双重上市公司的证据[J];金融发展研究;2008年11期
10 孙彬;杨朝军;于静;;基于copula函数的国际证券市场传染效应实证分析[J];上海交通大学学报;2009年04期
,本文编号:1387602
本文链接:https://www.wllwen.com/guanlilunwen/zhqtouz/1387602.html