股市波动长期成分与宏观基本面的非线性格兰杰因果检验  

The Nonlinear Granger Causality Test between Macroeconomic Fundamental and Stock Market long Run Volatility

作　　者：尚玉皇[1] 郑挺国[2] SHANG Yu-huang;ZHENG Ting-guo(Institute of Chinese Financial Studies,Southwestern University of Finance and Economics,Sichuan Chengdu 611130,China;School of Economics,Xiamen University,Fujian Xiamen 361005,China)

机构地区：[1]西南财经大学中国金融研究中心,四川成都611130 [2]厦门大学经济学院,福建厦门361005

出　　处：《数理统计与管理》2018年第6期1102-1113,共12页Journal of Applied Statistics and Management

基　　金：国家自然科学基金青年项目(71701165);国家自然科学基金面上项目(71371160);教育部人文社会科学研究青年基金项目(16YJC790084);长江学者奖励计划青年学者项目(Q2016131);西南财经大学一流学科项目;研究阐释党的十九大精神国家社科基金专项(18VSJ073);中央高校基本科研业务费之重大基础理论项目(JBK171124)

摘　　要：揭示股市波动与宏观基本面相互影响关系为市场参与主体行为决策提供了重要的参考依据。但该作用机制受到波动率自身噪音及经济周期的影响。为此,本文设定混频随机波动率模型以过滤波动率短期噪音成分,识别长期稳定成分,进而利用非线性格兰杰因果检验（MS-VAR）模型讨论经济周期对股市波动的非对称性影响。研究表明：首先,混频随机波动率模型能够稳健识别出波动率中的确定成分即长期成分;其次,非线性格兰杰因果检验结果表明宏观经济与股市长期波动成分之间的波动溢出具有显著的非对称特征,而该非对称特征主要由经济周期因素驱动;最后,在经济周期的衰退时期,宏观基本面与股市长期波动成分的正向溢出效应更加明显,而在经济扩张周期内,两者之间不存在显著的格兰杰因果关系。Investigating the relation between stock volatility and macro fundamental is important for market participators. However, this relation suffers from business cycle and volatility noise. This paper setups a SV-MIDAS model to filter the noise component and identify volatility component. Then we analyze the asymmetric effect of business cycle to volatility with MS-VAR model. We find that： first, the SV-MIDAS models can identify the volatility component robustly. Second, the nonlinear Granger causality test shows that the asymmetric figures of volatility spillover between macroeconomic and stock market is affected by business cycle. Third, the positive volatility spillover is occurred in economic recession, but not in economic boom period.

关 键 词：波动溢出 成分分解 区制转移 混频数据抽样 格兰杰因果 

分 类 号：F830[经济管理—金融学] O212[理学—概率论与数理统计]