基于LHAR-RV-V模型的中国股市波动性研究  

Research on China's stock market fluctuations based on LHAR-RV-V model

作　　者：文凤华[1,2] 刘晓群 唐海如[1] 杨晓光[1,3,2] 

机构地区：[1]长沙理工大学经济与管理学院,长沙410114 [2]湖南省金融工程与金融管理研究中心,长沙410114 [3]中国科学院数学与系统科学研究院管理决策与信息系统重点实验室,北京100190

出　　处：《管理科学学报》2012年第6期59-67,共9页Journal of Management Sciences in China

基　　金：国家自然科学基金资助项目(70971013;71171024);国家973计划资助项目(2010CB731405);湖南省杰出青年基金资助项目(09JJ1010)

摘　　要：在已实现波动率异质自回归模型(HAR-RV模型)的基础上,基于市场微观结构的理论,同时考虑市场波动的杠杆效应和量价关系,构造了已实现波动率及交易量之长记忆异质自回归模型(LHAR-RV-V模型).利用该模型对沪深300指数的等时1min高频数据进行实证分析,实证结果表明该模型能够较好地捕捉到我国股票市场波动的长记忆性和杠杆效应,且杠杆效应具有一定的持续性.此外,过去不同周期交易量的加入不仅能够更为细微的反映量价之间的关系,而且在一定程度上改善了模型的预测能力.On the basis of heterogeneous autoregressive realized volatility model （ HAR-RV model）, combining the market microstructure theory, the paper proposes a new dynamic volatility model which considers both the leverage effect and price-volume relation and which is called Leverage Heterogeneous Autoregressive Realized Volatility with Volume model（ LHAR-RV-V model）. The proposed model is applied to empirical analysis with the 1 minute high frequency data of Shenzhen 300 index. The empirical results show that the model can well capture the long memory and the leverage effect characteristics, and the leverage effect has certain sustainabili- ty. In addition, the past different cycle volume joined in the model can not only reflect the relationship be- tween the volume and price in a more sophisticated way, but also improve the predictive power of the model to some degree.

关 键 词：异质市场假说 LHAR—RV-V模型 交易量 杠杆效应 

分 类 号：F830.91[经济管理—金融学]