| Throughout history of the conditional asset pricing literature the goal has been to find the best possible model to explain what determines a firm's expected stock return. In Dickson (2015) the variables that prove to be best at explaining a firm's stock return is book-to-market, market capitalization, gross profitability, investment, short-term reversal, and momentum. The aim of this study is to further examine improvements in Dickson (2015) by changing the functional form and adding interaction terms between the variables. The chosen methodology is a version of the popular Fama-Macbeth regressions which are well documented in the literature to determine the added risk premium associated with firm characteristics. By allowing for the possibility of non-linear characteristics and interaction terms, this study shows that market capitalization follows a significant non-linear relationship with the average stock return and by adding the squared regressor to the model, the explanatory power and risk premium for market capitalization improves. The study further shows that including the interaction between investment and market capitalization improves the explanatory power of the investment variable and the market capitalization variable. |
