Alternative Value Premia

Abstract

The asset pricing literature has historically focused on the book-to-market (HML) factor as adriver of the value premium, but several scholars have noted that alternative fundamental-toprice ratios capture value premia comparable to the HML factor. Alternative value premia havein the literature received little attention; hence this study seeks to fill this gap in the literatureby analyzing the historical performance of alternative value factors starting from 1963 through2018, using CRSP and Compustat datasets. We test whether returns of long-short factors based on alternative value metrics are explainedby the five-, q- and six-factor models and investigate how characteristics of alternative valuefactors deviate from the traditional HML factor using multidimensional sorts of firms based ontheir fundamentals which allow for the detection of alpha generating subsections. FamaMacBeth regressions are then used to investigate if past changes in size drive value premia ona value-sorted portfolio level. Lastly, time-dependent factors are analyzed to examine if thevalue premia are driven by subperiods where the difference between cheap and expensivestocks are particularly large. Alternative value factors generally deliver higher returns as standalone strategies and thesehigher returns are often attributed to exposure to profitability risk factors. The cash-flow-fromoperations-to-enterprise-value factor generated alpha unaccounted for by the five- and q-factormodels which were, as for multiple other value strategies primarily driven by the short-leg andshowed a general tendency to intensify when controlling for gross profits to assets. Results areambiguous and lack enough statistical significance to conclude to what extent, past changesare a driver of alternative value metrics. Timing value strategies based on the relative valuespread did not prove to be an effective strategy, suggesting value strategies also generatespositive returns in the subperiods where the value spread is small. While factor subsectionsdrive many factor premia, we call for future research to examine how these subsections interactas they hold the potential to improve the existing asset pricing models.