Compared this three models, Fama-French model has highest adjusted R-square and lowest probability of F-statistic. Therefore, the most suitable model to capture the performance of mutual funds is Fama-French model. However, Fama-French model is not so good. Firstly, the β of ExRm and HML is insignificant, that is to say, the expect return of mutual funds is little related to market premium and stocks value. So it is meaningless to put ExRm and HML into the model. Secondly, the R-square of Fama-French model is low, maybe because the model omits some variables that highly explain the risk premium of mutual funds. …show more content…
DeMarzo. (2010). <Corporate Finance>, Pearson, 2nd edition
Bodie, Kane & Marcus. < Investments>. McGraw-Hill
Stock & Watson. < Introduction to Econometrics>, Pearson International Edition
Appendix:
Regress of CAPM model Dependent Variable: R | | | Method: Least Squares | | | Date: 12/06/12 Time: 13:24 | | | Sample: 1968M02 2005M07 | | | Included observations: 450 | | | | | | | | | | | | | Variable | Coefficient | Std. Error | t-Statistic | Prob. | | | | | | | | | | | C | -0.005027 | 0.000586 | -8.571419 | 0.0000 | EXRM | 0.011989 | 0.012712 | 0.943093 | 0.3461 | | | | | | | | | | | R-squared | 0.001981 | Mean dependent var | -0.004973 | Adjusted R-squared | -0.000246 | S.D. dependent var | 0.012379 | S.E. of regression | 0.012381 | Akaike info criterion | -5.940885 | Sum squared resid | 0.068673 | Schwarz criterion | -5.922622 | Log likelihood | 1338.699 | Hannan-Quinn criter. | -5.933687 | F-statistic | 0.889424 | Durbin-Watson stat | 2.143090 | Prob(F-statistic) | 0.346142 | | | | | | | | | | | | |