基于MS-HAR-TVP模型的原油价格对行业指数波动溢出与预测研究  

Study on the Spillover Impact and Forecasting of Crude Oil Prices on Industry Index Fluctuations Based on MS-HAR-TVP Modeling

作　　者：张虎[1,2] 徐泽宇 高子桓 ZHANG Hu;XU Ze-yu;GAO Zi-huan(School of Statistics and Mathematics,Zhongnan University of Economics and Law,Wuhan 430073,China;Big Data Institute,Zhongnan University of Economics and Law,Wuhan 430073,China)

机构地区：[1]中南财经政法大学统计与数学学院,湖北武汉430073 [2]中南财经政法大学大数据研究院,湖北武汉430073

出　　处：《数理统计与管理》2024年第6期1121-1140,共20页Journal of Applied Statistics and Management

基　　金：国家社会科学基金重大项目(23&ZD057);中央高校基本科研业务费专项资金资助项目(2722024EJ002,2722024EJ026)。

摘　　要：本文研究了高频条件下原油市场对国内股票市场的波动溢出效应,提出了一种基于HAR模型的改进方法,即MS-HAR-TVP模型,该模型通过TVP模型将原油价格对国内金融市场的影响分解为趋势和跳跃波动的溢出效应,并结合马尔可夫转换机制捕捉波动率的状态变化。本文选取中证煤炭指数和中证新能源指数作为国内能源类股票市场的代表,采用滚动窗口法和MCS检验法评估模型的预测性能,并与其他常用模型进行比较。实证结果表明:(1)拆解后的高频波动溢出具有明显的波动聚集和不对称性,且趋势和跳跃溢出对未来波动的预测能力有显著提升。(2)原油短期趋势与新能源指数趋势相反,但与煤炭指数短期趋势相同,说明原油价格对不同能源类股票市场的影响存在差异性;(3)结合原油波动溢出后的MS-HAR-TVP模型和MS-HAR-TVPJ/TCJ模型在高波动时期的样本内外预测精度显著高于其他模型,本文提出的模型能够更好地刻画和预测国内能源类股票市场的波动特征。In the face of the uncertainty caused by commodity price fluctuations on the financial market,this paper further explores the volatility spillover effect of the crude oil market on the domestic stock market under high-frequency conditions.Decomposing the high-frequency volatility with HAR model,the trend and jump volatility spillover caused by crude oil prices on China financial market are imported by the TVP-VAR model,and then combined with the Markov Regime-Switching to propose the MSHAR-TVP model.An empirical study is conducted on the different impact of crude oil on Coal and Sola energy index,and the model results are evaluated by adopting a rolling window period and the MCS test method.The empirical research shows that the high-frequency volatility spillover after decomposition still has obvious volatility clustering and asymmetry.The trend and jump spillover have a higher ability to predict future volatility.The short-term trend of crude oil is opposite to that of the new energy index but the same as that of the coal index.The MCS test verifies that combining crude oil volatility spillover can significantly improve the in-sample and out-of-sample predictive accuracy of domestic energy index during high volatility periods.Both industry indices have higher out-of-sample predictive accuracy in the trend model(MS-HAR-TVP)and the jump model(MS-HAR-TVP-J/TCJ).still has obvious volatility clustering and asymmetry.The trend and jump spillover have a higher ability to predict future volatility.The short-term trend of crude oil is opposite to that of the new energy index but the same as that of the coal index.The MCS test verifies that combining crude oil volatility spillover can significantly improve the in-sample and out-of-sample predictive accuracy of domestic energy index during high volatility periods.Both industry indices have higher out-of-sample predictive accuracy in the trend model(MS-HAR-TVP)and the jump model(MS-HAR-TVP-J/TCJ).

关 键 词：波动率预测 波动溢出 TVP-VAR MS-HAR 

分 类 号：F832[经济管理—金融学] F713.35[理学—概率论与数理统计] O212[理学—数学]