本文在行为金融学相关理论的基础上，围绕噪音交易者模型（DSSW模型），通过理论演绎归纳出5个有关投资者情绪假说。并利用伍燕然老师构造的投资者情绪指数，使用GARCH-M模型分别验证了（1）投资者情绪变化对沪深两市整体当期收益的影响；（2）投资者情绪变化对大盘股、中盘股与小盘股，高市净率股、中市净率股与低市净率股，高价股、中价股与低价股，周期股与非周期股的当期收益影响的不同。实证结果表明，投资者情绪变化显著地正向影响沪深两市的超额收益，并能显著地正向修正沪深两市的收益波动，通过风险奖励进一步影响收益；小盘股票、成长性股票（高市净率股）、高价股票和周期性股票对投资者情绪变化更敏感。

On the basis of the theory of behavioral finance, this paper conclude five investor sentiment hypothesis around the DSSW (1990) model. Using the investor sentiment indexconducted by Wu Yanran, this paper employ a generalized autoregressive conditional het-eroscedasticity-in-mean (GARCH-M) specification to test (1) How the changes in investorsentiment influence the return and volatility of Shanghai and Shenzhen stock market. (2) How the changes in investor sentiment affect differently on large-cap stocks, medium-capstocks and small-cap stocks; high-book-value stocks, medium-book-value stocks and low- book-value stocks, high-priced stocks, medium-priced stocks and low-priced stocks；cycli-cal stocks and non-cyclical stocks. The empirical results show that changes in investor s-entiment have significant positive effect on the excess return of the Shanghai and Shenz-hen market, and can correct the volatility of return of Shanghai and Shenzhen positively.Small-cap stocks, growth stocks (high-book-value stocks), high-priced stocks and cyclical stocks are more sensitive to the changes in investor sentiment.