新三因素模型在A股市场的实证检验doc.doc

摘 要 资产定价理论是金融学领域研究的核心。传统资产定价理论,从CAPM到Fama-French三因素模型，将生产因素作为外生变量，从消费者效用最大化角度来研究资产定价模型,其在不断发展过程中面临实证的挑战。而新三因素模型基于托宾q理论，从生产者价值最大化的角度着手，将股票收益率与公司特征联系起来，认为市场因子、投资因子和ROA因子能够很好解释股票横截面收益率，特别是价值溢价和短期收益惯性现象。 本文以我国A股市场的股票作为研究对象，对新三因素模型在A股市场的适用性进行了实证检验。首先采用Fama-Macbeth两步回归方法和分组计算加权平均收益率方法对投资效应和ROA效应进行检验，并且分别考虑不同规模水平下以及股权分置改革前后投资效应和ROA效应的显著性；其次构造投资因子和ROA因子，采用Fama-French时间序列因素回归法检验新三因素模型的适用性，重点考察其对价值溢价和短期收益惯性的解释能力，并将其与CAPM、Fama-French三因素模型进行比较。 实证结果表明A股市场存在较为显著的投资效应和ROA效应，小规模公司投资效应的显著性强，股权分置改革前投资效应显著性较差，股权分置改革后投资效应较为显著。新三因素模型能够较好地解释A股市场所存在的价值溢价和短期收益惯性的现象。与国外发达国家的资本市场相比，我国资本市场起步较晚，发展尚未成熟，我国股市的投资效应还是较弱，这在一定程度上也制约了新三因素模型对价值溢价现象的解释能力。 关键词： 新三因素模型；价值溢价；短期收益惯性 Abstract Asset pricing is one of the hot points in financial empirical researches. The traditional asset pricing models, from CAPM to Fama-French three-factor model，tend to look at the demand side of the economy taking production as exogenously determined. Although continuous development, they face more empirical challenges. The alternative three-factor model, based on Tobin’s q theory, motivates factors by exploiting a direct link between stock returns and characteristics from the production side. The alternative three-factor model, including market factor, investment factor, and ROA factor, can explain cross-section returns very well, especially value premium and momentum. Based on China A stock market , the article studies the empirical performance of the alternative three-factor model. First, I test the investment effect and ROA effect using Fama-Macbeth regression method and sorting method. I also consider the significance of the investment effect before and after the reform of non-tradable shares. Secondly, I construct the investment factor and ROA factor, study the empirical performance of the alternative three-factor model in explaining value premium and momentum using the calendar-time factor regression method. The empirical results show that there exists significant investment effect and ROA effect in A stock mark