本文建立了股票非预期收益三因素定价模型,三因素包括:当期非预期会计收益与期初股票价格之比(反映价值相关性),分析师盈余预测修正变量(反映会计收益增长预期),以及市场非预期收益(URM)变量.在此基础上,本文建立了多变量回归模型,并采用2002年1月至2011年3月间中国股票市场的有关交易数据、机构收益预测数据和财务数据,来检验理论模型和实证模型的预测,发现:1)三因素模型框架可以精确地解释股票非预期收益,模型的截距项接近于零,调整R^2为52%;2)非预期市场收益(URM)吸收了账面市值比对股票非预期收益的解释能力;3)存在证券分析师偏向乐观的盈余预测,并且,基于证券分析师盈余的当期盈余预测修正变量对股票非预期收益具有显著的解释能力.
A new three-variables model has been established to explain unexpected individuM stock returns. Three-variables are inclusion in current earnings, often used in studies to examine the value relevance of accounting earnings beginning with Ball and Brown[11, incorporating earning growth with the revisions in forecasts of current-period earnings based on analyst predicting and unexpected market return reflecting market sentiment. First, I find my three-variables is the exact pricing model to unexpected individual stock returns in China stock markets. Second, the navel measure of unexpected market return absorbs the apparent roles of book to market ratio on individual expected stock returns with the sign of coefficient of book-to market ratio becoming to be negative after I control for unexpected market returns. Thirdly, the optimism bias in analyst forecasts remains unchanged during the sample period. The measure of current revision of earning forecasts used reflecting earnings growth has explanatory power for individual stock returns.