已实现GARCH-GED模型的研究及应用  

Research and Application of Realized GARCH-GED Model

作　　者：魏正元 余徳英 李素平 WEI Zhengyuan;YU Deying;LI Suping(College of Science, Chongqing University of Technology, Chongqing 400054, Chin)

机构地区：[1]重庆理工大学理学院,重庆400054

出　　处：《重庆理工大学学报（自然科学）》2018年第3期273-278,共6页Journal of Chongqing University of Technology：Natural Science

基　　金：国家统计局统计科研重点项目(2014LZ25);重庆市教委科学技术研究项目(2014CJ39);重庆市教委科技项目(KJ1709207)

摘　　要：针对高频金融数据分布普遍存在尖峰厚尾的现象,将经典的已实现GARCH模型的残差分布拓展到服从广义误差分布的形式,并将杠杆函数的幂次放松为待估参数,建立了新的已实现GARCH-GED模型。基于上证50指数5 min频率的高频数据的实证研究结果表明:相对于正态分布假设下的已实现GARCH模型,已实现GARCH-GED模型更能刻画波动率的杠杆效应,在一定程度上提升了对尾部风险度量的精度。For the universal existence of high-frequency financial data＇s sharp peak and fat tail,this paper constructed a new classical realized GARCH-GED model with the residual distribution of realized GARCH model has been extended to generalized error distribution and the power of leverage function is relaxed to estimated parameter. The empirical results based on the 5-minute highfrequency data of SSE 50 index show that compared with the realized GARCH model based on normalized residual,The realized GARCH-GED model can describe the leverage effect of the price of stock and improve the accuracy of tail risk measurement in a certain extent.

关 键 词：高频金融数据 厚尾分布 已实现GARCH-GED模型 VaR返回测试 

分 类 号：O21[理学—概率论与数理统计]