Size and Value Matter But Not the Way You Thought

We propose a simple but fundamental methodological change to Fama and French (1993) factor construction procedure. Consistent with Lambert and Hübner (2013) sequential sorting procedure to classify stocks, our methodology controls ex ante for pricing errors produced by multifactor models. Our size and value factors deliver less specification errors when used to price portfolios, especially regarding low size and high B/M stocks. Furthermore, this alternative framework generates much stronger “turn-of-the-year” size and “through-the-year” book-to-market effects than conventionally documented. The factors also display a slight competitive advantage on the taxonomy of low turnover market anomalies defined by Novy-Marx and Velikov (2015).