欢迎访问中国科学院大学学报,今天是

中国科学院大学学报 ›› 2010, Vol. 27 ›› Issue (4): 440-447.DOI: 10.7523/j.issn.2095-6134.2010.4.002

• 论文 • 上一篇    下一篇

基于pair copula-GARCH模型的多资产组合VaR分析

黄恩喜, 程希骏   

  1. 中国科学技术大学统计与金融系,合肥 230026
  • 收稿日期:2009-11-03 修回日期:2010-03-05 发布日期:2010-07-15
  • 通讯作者: 黄恩喜
  • 基金资助:

    中国科学院知识创新工程重要方向项目(KJCX3-SYW-S02)资助 

Analysis of portfolio VaR by pair copula-GARCH

HUANG En-Xi, CHENG Xi-Jun   

  1. Department of Statistic and Finance, University of Science and Technology of China, Hefei 230026, China
  • Received:2009-11-03 Revised:2010-03-05 Published:2010-07-15

摘要:

提出了多资产组合风险分析的pair copula-GARCH 模型.相比于以往的copula-GARCH模型,它能够更好描述投资组合中两两资产间的尾部相关性的差异,从而更好地度量多资产组合间的相依结构.在此基础上,还探讨了pair copula-GARCH 模型下的多资产线性组合的VaR的计算方法.最后给出模型的实证分析.

关键词: pair copula, GARCH, Monte Carlo, VaR

Abstract:

We propose a pair copula-GARCH model to analyze the risk of Multi-asset portfolio. Compared to other Copula-GARCH models,our method can capture the tail dependence of different pairs of risk factors in a portfolio so that it can describe the dependence structure of the risk factors better. By using this model combined with Monte Carlo techniques, the numerical analysis of portfolio VaR is fully discussed in this paper. Finally, the empirical portfolio risk analysis is given.

Key words: pair copula, GARCH, Monte Carlo, VaR

中图分类号: