MARKOV-GARCH模型对我国证券市场在险价值的度量  

The Measure of VaR of the Chinese Stock Exchanges Based on a New MARKOV-GARCH Model

作　　者：陈燕武[1] 吴承业[1] 

机构地区：[1]华侨大学商学院,福建泉州362021

出　　处：《华侨大学学报（自然科学版）》2009年第5期580-584,共5页Journal of Huaqiao University(Natural Science)

基　　金：高等学校博士点学科专项科研基金(20050385001);福建省高等学校新世纪优秀人才支持计划项目(07FJRC07);华侨大学科研基金资助项目(06BS211)

摘　　要：提出一种度量我国证劵市场在险价值的MARKOV-GARCH(马尔可夫-广义自回归条件异方差)模型.即通过MARKOV的性质,提示市场的跳跃规律,结合GARCH模型得出市场的波动率,进一步通过波动率度量市场的在险价值.实证结果表明,在确保成功率的前提下,其度量的在险价值的置信区间比一般GARCH模型所度量的置信区间更小,而且稳定性也较其他模型强.该模型在不放大所度量的在险价值区间的前提下,却取得较高的成功率.A Markov-generalized autoregressive conditional heteroscedasticity （MARKOV-GARCH） model has been put forth in this paper to measure the value at risk （VaR） of our stock exchange market, namely, it reveals the jumping law of the market based on the MARKOV theory and gets the fluctuation ratio of market combined with the GARCH model and, with witch to have the further measure the VaR of the market. The conclusion shows that, under the insured fitness ratio, the new MARKOV-GARCH model has a narrower confidence interval than that by general GARCH models and also has better stability than other modals. The MARKOV-GARCH model has got a relatively higher fitness ratio on the premise of the modal without expanding the interval of VaR.

关 键 词：MARKOV-GARCH模型 在险价值 拟合优度 中国证券市场 

分 类 号：O211.62[理学—概率论与数理统计] F832.5[理学—数学]