基于消息模型的股票市场波动与相关消息因素的关系研究  

A Test of the News Model of Shanghai A Share Market

作　　者：王鹏 吕永健[3] 

机构地区：[1]金融安全协同创新中心,四川成都610074 [2]西南财经大学中国金融研究中心,四川成都610074 [3]西南财经大学金融学院,四川成都610074

出　　处：《数理统计与管理》2013年第6期1124-1131,共8页Journal of Applied Statistics and Management

基　　金：国家自然科学基金(71101119);西南财经大学和四川省教育厅创新团队建设项目(JBK130401)

摘　　要："消息"(News),对于验证市场有效性和研究股票市场的波动特征,具有重要价值.本文基于消息模型(News Model),对上证A股综合指数及其相关影响因素进行了实证分析.通过对样本数据分别利用ARMA(Autoregression Moving Average)、VAR(Vector Autoregressive)、HPF(Hodrick Prescott Fliter)模型提取消息部分,进行回归分析;再利用估计ARDL(Autoregressive Distributed Lag)形式的消息模型,检验相关因素对股市是否存在滞后效应.实证结果显示,一方面,与对众多国外市场的研究保持一致,论文对于中国股票市场的研究同样证实,消息对股市是具有影响力的,另一方面,我们还得出以下结论:(1)财政收入消息对股市存在显著的影响;(2)财政支出消息对股市存在显著的正向影响;(3)消息对股市影响的滞后效应不明显.On the basis of news model, this paper tests five price determination that may have a effect on Shanghai A share market over the period January 1996 to December 2010. First, we use ARMA models, VAR models and HPF models to extract the news components, and then estimate news model with news components. Second, we make nonnested model selection test through MacKinnon J test. At last, through the use of Autoregressive Distributed Lag version of former news model to test if the news components have lagged effect on stock price. The empirical results tell us that news also have effect on China stock market, and we also find that： （1） Government revenue news has a effect on Shanghai A share, which is consistent with Imad Mossa （2010）＇s conclusion. （2） Government spending news has significant effect on Shanghai A share. （3） There＇s no significant signs that news components have the lagged effect on Shanghai A share.

关 键 词：消息模型 ARMA VAR ItPF 财政支出 

分 类 号：O212[理学—概率论与数理统计]