有信息冲击、无信息冲击与波动率非对称性  

News-Driven,No-News-Driven Innovations and Volatility Asymmetry

作　　者：谭地军[1] 田益祥[1] 黄文光[1,2] 

机构地区：[1]电子科技大学管理学院,四川成都610054 [2]Cardiff Busincss School, UK

出　　处：《管理工程学报》2009年第2期92-98,共7页Journal of Industrial Engineering and Engineering Management

基　　金：教育部新世纪优秀人才支持计划资助项目(教技函[2005]35号);电子科技大学中青年学术带头人+创新团队支持计划;教育部优秀青年教师计划资助项目(教人司[2003]355号)

摘　　要：本文将金融市场的冲击(innovations)分成有信息冲击和无信息冲击,构建了一种新的非对称GARCH类模型,区间均值GARCH模型(sectional mean GARCH,SM-GARCH)。在中国股票市场的实证分析中发现:SM-GARCH模型能够更好地捕捉波动率特征和预测波动率;条件均值预测误差用于描述新到来的冲击对条件方差的影响,所提供的信息是有限的,在无信息冲击的情况下可以忽略它在波动率预测中的作用。This paper classifies the innovations of financial markets into news-driven and no-news-driven innovations, and proposes a new type of asymmetric GARCH model, sectional mean GARCH （SM-GARCH）. In the empirical study of the Chinese stock market, it has been found that the SM-GARCH behaves better than the classical GARCH model in capturing the volatility characteristics as well as forecasting volatility. Moreover, the forecast errors of the conditional mean, which is usually seen as the new innovations to the market, provide little information about the future volatility. The no-news-driven innovations, especially, can be ignored in volatility forecasting.

关 键 词：信息 区间均值 GARCH模型 知情交易 流动性交易 

分 类 号：F830.9[经济管理—金融学]