摘要
本文从预期收益期望的统计学计算原理出发,研究发现在CAPM经验检验时,交易前后预期收益的残差分布将发生变动,这一变化可使得交易后预期收益的期望会增加、方差却减少。该结论也可同时解释许多当前有关CAPM有效性不足的研究成果:无风险收益可作为预期收益的影响因子,市场风险溢价者具有时变性,经验检验设计存在偏差。这种较强的解释力,不仅表明本研究方法的有效性,也将为我们进一步研究和提高CAPM及其经验检验的有效性提供一种新的方法和途径。
In this paper,according to the statistic principle for calculating the expectation,we found that in the empirical test process for the CAPM,the residual distribution of the estimative earnings would vary after the exchange,which would make the expectation of the estimative earning increase,while the variance reduce. These findings could explain many research achievements about the CAPM shortages.For example,risk - free income should be a factor to forecast the asset earnings,the risk premium to the market would vary according to the time,and bias still remains in design of empirical test.All this,has demonstrated better effectiveness of the methods in this paper.It will also provide a new means for us to study CAPM further and raise its effectiveness.
出处
《贵州大学学报（社会科学版）》
2010年第6期24-29,共6页
Journal of Guizhou University(Social Sciences)
作者简介
罗兵（1966-），男，贵州财经学院教授，武汉理工大学在读博士。
唐先勇（1973-），男，贵州财经学院会计学院副教授、博士、中国注册会计师。
