An Empirical Analysis of Quantile Regression Based Risk Measurement in the Chinese Stock Markets

摘要：

　　This paper introduces the concepts and methods of the Value at Risk (VaR) and quantile regression,and uses the lag yields as explanatory variables to establish the conditional quantile regression model,and the R software to dynamic estimate the VaR in the Chinese stock markets during 1996-2010.Empirical research more than a decade of the impact of lag yields to VaR of Chinas Shanghai and Shenzhen stock markets,and the non-uniformity of each trading day within a week.And compare the results with the GARCH (2,1) model by the Kupiec likelihood ratio test.The results show that the quantile regression model has many good properties,applicable to VaR estimation of financial time series data with heavy tail and that is an effective semi-parametric risk measurement method.

关键词： Quantile regression VaR GARCH model Kupiec likelihood ratio test

作者: Chen Hua Kang Yixue

作者单位: Mathematics,Collage of Science China University of Mining and Technology Xuzhou,China SunYueqi Honors Collage China University of Mining and Technology Xuzhou,China

会议类型: 国际会议

会议名称: 2012 IEEE 5th International Conference on Management Engineering & Technology of Statistics (第五届(2012)管理工程与统计技术国际学术研讨会(ICMETS2012))

会议地点: 青岛

会议语种:英文

页码: 30-34

在线出版日期: 2012-07-20（万方平台首次上网日期，不代表论文的发表时间）