Industry Differences in the Explanatory Power of the Fama-French Three-Factor Model : Evidence from the Swedish Stock Market, 2010–2019

Sammanfattning

This study examines whether the explanatory power of the Fama-French three-factor model varies across sectors in the Swedish stock market during the time period of 2010–2019. Although the model has been widely employed in international asset pricing studies, there have been fewer empirical studies conducted to investigate if its performance differs across sectors in smaller developed markets such as Sweden. The empirical analysis is centered on firm-level panel data for publicly listed Swedish firms. The dependent variable, excess return, is regressed on the market risk premium, SMB, and HML based on separately estimated firm-level regressions, with a minimum criterion of 60 observations for each firm. The firm-level outcomes are then summarized at the sector level. Additionally, supplementary sector-level regressions are estimated, and a comparative analysis of the model performance is examined between the Fama-French model with CAPM. Model performance is measured using metrics such as adjusted R-squared, RMSE, MSE, and log-likelihood. A further robustness check is conducted to examine the sensitivity of the result to extreme return observations. The findings suggest that the explanatory power of the Fama-French three-factor model is generally weak at the firm-level. Adjusted metrics, such as mean and median values are close to zero for every sector included in the study, and in several sectors the mean adjusted is negative. Nonetheless, the empirical findings suggest that model performance differs between sectors. The sectors Health Care and Consumer Discretionary demonstrate comparatively stronger model fit in the sector-level regressions, whereas Consumer Staples displays the weakest model fit of the sector sample. The comparative analysis with the benchmark CAPM offers mixed evidence, as the Fama-French three-factor model enhances model performance to a limited extent and is not uniformly observed across sectors.