向量GARCH模型的非线性协同持续  

The Nonlinear Common Persistence of Multivariate GARCH Model

作　　者：刘丹红[1] 徐正国 张世英[2] 

机构地区：[1]天津大学理学院,天津300072 [2]天津大学管理学院,天津300072

出　　处：《系统工程》2004年第6期33-38,共6页Systems Engineering

基　　金：国家自然科学基金资助项目(70171001);南开大学-天津大学刘徽应用中心资助项目

摘　　要：讨论波动持续性的函义,介绍向量GARCH模型的持续性及协同持续的定义,来研究上海和深圳股市,发现股市表现很强的持续性,但不存在线性的协同持续。提出非线性协同持续的概念,并提出用小波神经网络逼近非线性协同持续函数,同时证明沪深股市存在非线性协同持续关系。This paper discusses the signification of persistence in volatility, introduces the definition of the multivariate (GARCH) persistence and the common persistence to study Shanghai and Shenzhen stock markets and find these markets (show) strong persistence, but no linear common persistence. So, this paper gives out the definition of the nonlinear common persistence. We make use of the wavelet neural network to approach the nonlinear common (persistence.) The results also (testify) the nonlinear common persistence of two markets at the same time.

关 键 词：持续性 协同持续 非线性协同持续 小波神经网络 

分 类 号：F830[经济管理—金融学]