Capital Assets Pricing Model with Fuzzy Return

During the last decades several models have provided for predicting stock returns. The Capital Asset Pricing Model (CAPM), has attracted much attention, but also raised to it a lot of criticism. One of the criticisms is mentioned to this model is ignoring the intra-period volatility, structural breaks and duration of the calculations. Therefore, this study by using fuzzy regression has been tried to intra-period volatility to be included in this model. Then the stability of beta and accurately of the return prediction on the Capital Asset Pricing Model by using of Fuzzy logic system (FLS), Ordinary Least Squares regression model (OLS) and Generalized Method of Moments (GMM) are compared. The population of research is all companies listed in Tehran Stock Exchange and the research sample is consisted of 31 companies that in the period of 2011-2014, have been with the highest turnover. The results of research shows that the prediction error of stock returns in fuzzy model is more than OLS and GMM model, but the prediction error of stock returns in both of ordinary least squares regression model and Generalized method of moments doesn’t have any significant difference with each other. So the classical beta still has offered the more stable beta than fuzzy beta and thus is offering a better prediction of stock returns.