基于高频数据的GARCH模型拟极大指数似然估计  

Quasi Maximum Exponential Likelihood Estimation of GARCH Model Based on High Frequency Data

作　　者：李莉丽 张兴发 邓春亮[2] 李元 LI Lili;ZHANG Xingfa;DENG Chunliang;LI Yuan(School of Economics and Statistics,Guangzhou University,Guangzhou 510006,China;School of Mathematics,Jiaying University,Meizhou 514015,China)

机构地区：[1]广州大学经济与统计学院,广州510006 [2]嘉应学院数学学院,梅州514015

出　　处：《应用数学学报》2022年第5期652-664,共13页Acta Mathematicae Applicatae Sinica

基　　金：国家自然科学基金(11731015);广东省教育厅青年创新人才项目(2018KQNCX241);广州大学校内科研基金(YG2020029)资助项目。

摘　　要：基于拟极大指数似然方法,本文研究了利用日内高频数据来估计日频GARCH模型的参数.已有的相关研究均假定GARCH方程中常数项是给定的,限制了方法的广泛应用.本文基于常规的GARCH(1,1)模型框架,针对模型全部参数给出了两步估计方法,讨论了估计量的理论性质.针对不同的波动率替代,文章也给出了最优波动率替代选择标准.模拟研究和实证研究表明所提估计方法有很好的表现和一定的应用价值.Based on quasi maximum exponential likelihood approach,this paper studies the daily GARCH model estimation by introducing the intraday high frequency data.Existent results assume the constant term in GARCH equation is given,which restricts the extensive applications of the method.In this paper,based on the framework of GARCH(1,1)model,two-step estimation method for all model parameters,together with the according asymptotic properties,are provided.A criterion is also given to choose the optimal volatility proxy.Simulation and empirical studies show that the estimators perform well and have potential applications in practice.

关 键 词：GARCH模型 高频数据 拟极大指数似然估计 

分 类 号：O212.1[理学—概率论与数理统计]