高频金融时间序列的协同持续关系研究  

Research on the relationship of co-persistence based on high-frequency financial time series

作　　者：唐勇[1] 张世英[1] 张瑞锋[1] 

机构地区：[1]天津大学管理学院,天津300072

出　　处：《系统工程学报》2006年第5期455-462,共8页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(70471050)

摘　　要：对向量高频时间序列的“已实现”协方差阵提出相应的模型并建立了“已实现”向量自回归模型.应用Bollerslev和Engle提出的持续和协同持续概念,讨论了“已实现”向量自回归模型存在线性协同持续的充要条件和寻找这种线性协同持续向量的方法,在此基础上进行了实证分析,表明沪深两股市之间不存在线性协同持续关系.最后指出协同持续概念在动态组合投资、风险规避策略中的意义和作用.The corresponding model of realized covariance matrix of in this paper vector high-frequency financial time series is brought forward and the realized vector autoregressive model is set up is this paper. The necessary and sufficient condition of the existence of linear co-persistence in this model and the method of seeking the co-persistence discussed by applying the concepts of persistence ＆ co-persistence proposed by Bollerslev and Engle. Under these bases, it is indicated that the co-persistence does not exist in Shanghai and Shenzhen stock markets through empirical analysis. Finally, the meaning and action in dynamic portfolio and risk avoiding are pointed out.

关 键 词：高频金融时间序列 协同持续 动态投资组合 

分 类 号：F830[经济管理—金融学]