主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2015, Vol. 23 ›› Issue (4): 20-29.doi: 10.16381/j.cnki.issn1003-207x.2015.04.003

• 论文 • 上一篇    下一篇

典型事实、混合Copula函数与金融市场相依结构研究

林宇1, 陈王2, 王一鸣3, 黄迅1   

  1. 1. 成都理工大学商学院, 四川 成都 610059;
    2. 西南交通大学经济管理学院, 四川 成都 610031;
    3. 北京大学经济学院, 北京 100871
  • 收稿日期:2013-03-27 修回日期:2014-01-28 出版日期:2015-04-20 发布日期:2015-04-24
  • 作者简介:林宇(1973-), 男(汉族), 四川仪陇人, 成都理工大学商学院副教授, 博士, 研究方向:金融风险管理、金融市场与公司理财.
  • 基金资助:

    国家自然科学基金资助项目(71171025);国家社科基金资助项目(12BGL024);成都理工大学金融与投资创新团队(KYTD201303);四川软科学计划项目(2014ZR0093);四川矿产资源研究中心重点项目(SCKCZY2012-2D002)

Study on Dependency Structure of Financial Markets Based on the Stylized Facts and Mixed Copula Function

LIN Yu1, CHEN Wang2, WANG Yi-ming3, HUANG Xun1   

  1. 1. Business School, Chengdu University of Technology Chengdu 610059, China;
    2. School of Economics and Management, Southwest Jiaotong University Chengdu 610031, China;
    3. School of Economics, Peking University Beijing 100871, China
  • Received:2013-03-27 Revised:2014-01-28 Online:2015-04-20 Published:2015-04-24

摘要: 运用ARFIMA-FIAPARCH-skst模型对沪深300指数和香港恒生指数建立收益-波动模型, 然后结合估计的参数对模型进行修正以确立最终模型, 排除金融市场典型事实对相依关系的影响, 进而运用由Clayton、Frank和Gumbel组成的混合copula模型对相依结构进行建模。研究结果表明:内地市场和香港市场均未观察到显著的杠杆效应;由Clayton、Frank和Gumbel组成的混合Copula模型能够准确地描述两个市场之间的相依结构, 且两个市场下尾相依关系要强于上尾的相依关系, 通过动态混合copula也验证了这一明显的非对称关系。

关键词: 金融市场, 典型事实, 混合Copula, 相依结构

Abstract: The ARFIMA-FIAPARCH-skst model is applied to establish return-volatility model to CSI300 and HIS. Then the parameter estimated is combined to revise model to confine the final model and get rid of the effect of dependence relation from stylized facts in financial market. And then the mixed copula model made of Clayton, Frank and Gumbel is applied to establish a model of dependence structure. The result of the research indicates that evident of leverage effects found by the existing research hasn't been observed by local market and HongKong market. Moreover, the mixed copula model made of Clayton, Frank and Gumbel can describe the dependence structure between two markets accurately and the dependence relation of lower tail of two markets is stronger than the dependence relation of upper tail. Besides, the time varying mixed-copula also indicates that there is an obvious asymmetric dependence relationship.

Key words: financial market, stylized facts, mixed copula, dependence structure

中图分类号: