基于小波已实现波动的动态风险价值研究  

Research for Dynamic Value at Risk Based on Wavelet Realized Volatility

作　　者：伍习丽[1] 彭选华[2] 

机构地区：[1]重庆三峡学院数学与统计学院非线性科学与系统结构重点实验室,重庆404100 [2]西南政法大学经济学院,重庆401120

出　　处：《重庆师范大学学报（自然科学版）》2017年第3期73-78,共6页Journal of Chongqing Normal University:Natural Science

基　　金：重庆市教委科技计划项目(No.KJ130107);重庆市自然科学基金项目(No.cstc2012jjA00023);重庆三峡学院青年项目(No.14QN22)

摘　　要：【目的】对股票市场的VaR动态风险价值进行研究。【方法】采用小波多分辨技术将高频已实现波动率分解为近似信号和细节信号,建立MRA-RV-ARFIMA GARCH-VaR类模型,分别在1~2d、2~4d、4~8d和8~16d的尺度下进行动态风险价值度量。【结果】实证表明该模型能很好地捕捉到市场的信息,对风险预测效果较好。【结论】经过多分辨分解后的信号能有效地捕捉到不同时间尺度上的波动信息,近似信号能很好的反应波动的变化趋势,资产波动对短期交易反应敏感,不同时间尺度拟合的VaR比低频GARCH类模型效果更好。[Purposes] It does research for dynamic value at risk of stock market. [Methods]It uses the wavelet multi-resolution technique to decompose high frequency realized volatility into approximation and details signals, establish MRA-RV-ARFIMA- GARCH-VaR model to measure dynamic value at risk under 1-2day, 2-4day, 4-8day and 8-16day scales. EFindings3The empirical analysis shows that the model can primely catch the imformation of finacial market. It can also effectually forecast risk. [Conclusions]Multi-resolution decomposition of the signal can effectively capture the volatility information on different time scales. The approximate signal can be a very good fluctuation response for change trend. The asset volatility is sensitive to short-term trading reaction. VaR on different time scales is better than the low-frequency GARCH class models.

关 键 词：小波多分辨分析 已实现波动率RV 风险价值VAR ARMA GARCH 

分 类 号：O29[理学—应用数学] F830.91[理学—数学]