Potential of fast and frugal trees in factor investing on the US equity market

Abstract This paper examines the predictive value of fast-and-frugal trees modelling embeding thefive factors suggested by model. We examine a fast-andfrugaltrees predictive model of returns upon stock traded on the S&P500 over the years2010 to 2024. We allow for three variations in the predictive context of the fast-and-frugaltrees used: Skewness in the coefficients of the Fama-French models OLS regressionestimates Analysts following the chosen firm/industry sector. Market value of the firm or sector examined.We find the FFT model is a worthy competitor to regression models in predictingstock returns for the S&P 500 in our sample period, especially in terms of theconsistent nature of the predictions across sectoral groups. But a logisticregression alternative outperforms both OLS and our fast-and-frugal-tree inpredicting S&P 500 returns in our sample period.