作　　者： ; ;

机构地区： 山东经济学院信息管理学院

出　　处： 《管理科学学报》 2008年第6期77-83,94,共8页

摘　　要： 为了摆脱证券市场弱式有效性传统检验范式的理论困境,提出用价格时间序列转化为数位序列后的条件可预测性进行检验的新途径,并定义条件熵为这种研究范式下市场有效性的度量.新方法不仅简单明确,且给出了有效性程度的量化度量.实证分析了上证综合指数的条件可预测性特征及其历史变迁趋势,并用香港恒生指数进行对比,取得了几个重要的检验成果,并阐述了理论和政策意义. The traditional paradigm of testing for weak form of Efficient Market Hypothesis （EMH） has a theoretical obstruction, so in this paper we suggested a new approach to test EMH by testing the conditional predictability of the bit serial that is converted from the original time serial, and defined conditional entropy as the measure of market efficiency. The new approach is simple, clear and advantageous to traditional paradigm for the possibility to quantitatively measure market efficiency. We empirically analyzed the features and tendency of conditional predictability of Shanghai Composite Index and drew a comparison with Hong Kong HangSeng Index. We gained several significant conclusions from the results of the tests.

关 键 词： 有效市场假说 可预测性 条件熵 上证综合指数

领　　域： [自然科学总论] [经济管理]