中国股票市场关于股票收益的两个假设的检验  

Test of Two Hypotheses on Stock Returns in China's Stock Market

作　　者：蒋正华[1] 王建伟[1] 芮萌[1] 陈工孟[1] 

机构地区：[1]西安交通大学管理学院,陕西西安710061

出　　处：《财经理论与实践》2003年第6期49-53,共5页The Theory and Practice of Finance and Economics

摘　　要：使用方差比方法检验中国股票市场上股票收益服从随机游走的假设 ,检验结果拒绝了该假设。而使用Geweke和Porter -Hudak (1983)提出的分数差分检验 ,检验结果却支持股票收益存在长期记忆的假设。长期记忆性的存在使得有可能通过建立非线性经济计量模型以改进价格预测效果。We use variance-ratio methodology to test the hypothesis that stock return follows a random walk. But the results of variance-ratio tests reject the random walk hypothesis in China. In order to test the long memory in stock return in Chinese stock market, we employ the fractional differencing test devised by Geweke and Porter-Hudak (1983). We find evidence of long memory of returns. The discovery of long memory suggests possibilities of constructing nonlinear econometric models to improve price forecasting performance.

关 键 词：中国 股票市场 股票收益 随机游走 长期记忆性 

分 类 号：F832.51[经济管理—金融学]