摘要
文章通过选取2004年1月1日到2009年6月30日中国、香港、日本、英国和澳大利亚五个股票市场日收盘价的道琼斯数据,采用三状态Markov机制转换模型研究这些股市间相依性结构的变化。通过对目标股市结构变化的研究,可以描述并预测股市的波动性,从而指导风险管理。实证分析表明,在有机制转换条件下,澳大利亚与英国、日本股市间的相依性比无机制转换条件下均有所下降,而中国与香港股市间相依性却大幅上升。同时,本文采用总体拟合效果法来选取合适的copula函数并运用基于copula理论的相关系数法进行对比研究,发现次贷危机后各股市间的尾部相依性出现不同的变化,市场收益率呈现下降趋势,波动性均有所增加。其中,澳大利亚与英国股市间的尾部相依性最强,而中国股市与其他股市之间的相依性较弱,说明受到影响的程度较小。
The paper has selected the Dow Jones Indexes from January 1st, 2004, to June 30th, 2009, in the five stock markets of China, Hong Kong, Japan, Britain and Australia, and approached the changes of dependence structure between these target stock markets by employing three states Markov switching model. This kind of work might describe and predict the volatility in stock markets and is helpful to guide risk management. It is found that, compared with no Markov switching model, the dependences of stock markets have decreased in Australia, Britain and Japan on the condition of Markov switching model. However, it is worth noting that the correlation between Chinese stock market and Japanese has presentedan increase in a vast scale. Meanwhile, the paper has analyzed the dependence structure by selecting proper copulas with the method of overall fitting and the method of correlation based on copula theory is used for comparative analysis in this paper. It indicates that different degree of dependence change appears in these stock markets through empirical analysis after subprime crisis happened and market returns have declined obviously being accompanied to volatility increase. The highest dependence is exhibited between the stock markets in Australia and Britain, while the markets between China and the other four markets have displayed low dependences comparatively.
出处
《数理统计与管理》
CSSCI
北大核心
2013年第2期343-358,共16页
Journal of Applied Statistics and Management
基金
国家自然科学基金项目(70861003
70825005
71171168)
教育部社科研究基金规划项目(09YJA790092
10YJA790200)
中国博士后科学基金项目(20110490877)
中国博士后科学基金特别资助项目(2012T50726)
中央高校基本科研业务费专项资金资助(112010004005080001)资助
