In this chapter, we saw how a multifactor model can be built and implemented. As a result of a principal component analysis, we identified five independent factors that explained asset returns, but they seemed to be insufficient, given that they explained only 30 percent of the variance. For illustration, we also reproduced the famous Fama-French model on real market data, where, apart from the market factor, two additional firm-specific factors (SMB and HML) were also used. We used built-in functions for principal component analysis and factor analysis, and we have shown how to use a general linear model for regression analysis.

We found that the three factors were significant. Hence, we can conclude that on a more recent sample, the Fama-French factors have explanatory power. We encourage you to develop and test new multifactor pricing formulas that work as the classical ones, or even better.