中国股市收益非线性可预测性的实证检验  

Empirical Study on Nonlinear Predictability of Stock Returns in China

作　　者：黄元元[1] 曾峻 王传彬[1] HUANG Yuan-yuan;ZENG Jun;WANG Chuan-bin(School of Economics,Management and Humanities,Jiangsu Architectural Vocationaland Technical College,Xuzhou Jiangsu 221116;Hunan University of Finance and Economics,Changsha Hunan 410205)

机构地区：[1]江苏建筑职业技术学院经济管理与人文学院,江苏徐州221116 [2]湖南财政经济学院会计学院,湖南长沙410205

出　　处：《湖南财政经济学院学报》2018年第2期14-20,共7页Journal of Hunan University of Finance and Economics

基　　金：江苏省社会科学基金项目"江苏开发区创新驱动中的转型升级与环境规制问题研究"(项目编号:15EYB005)

摘　　要：以有效市场假说和理性人假设为前提的传统线性范式的资本市场理论,越来越受到金融学界的普遍质疑,导致非线性金融建模的迅速发展。采用ARCH-LM、BDS、PBT等非线性检验方法,对我国上证指数、深证成指和深证综指三大股指的收益率时间序列进行检验,发现非线性的可预测性是我国股市收益率序列相关的主要形式,而导致整个时间序列的非线性可预测的却往往只是其中的某几个时间段,在这几段时间范围内信息不能充分反映在股价里。The traditional linear capital market theory,which is based on the efficient market hypothesis and rational person hypothesis,has been widely questioned by the financial circles,which leads to the rapid development of nonlinear financial modeling.This paper employs the nonlinear test method of ARCH LM,BDS and PBT to examine the stock return time series in China.The result shows that:the nonlinear dependency structures is the main form of sequence correlation of Chinese stock returns,however,the cause of the whole time series correlation is only a few certain brief periods that information is not fully reflected into stock price.

关 键 词：非线性相关 可预测性 BDS检验 PBT检验 

分 类 号：F832.5[经济管理—金融学]