Factoring Characteristics into Returns: A Clinical Approach to Fama-French Portfolio Decomposition

This paper thoroughly analyzes competing construction methods for factoring characteristics into
returns. We show the importance of ensuring a proper diversification of the factor's portfolio constituents for producing relevant and unbiased risk factors or benchmark portfolios. This is an important issue to be
solved for asset pricing and performance models defined as a function of characteristics. As a practical
case, the paper works on the design of size and value spread portfolios à la Fama-French. This quasi-
clinical investigation examines three methodological choices that have an impact on portfolio
diversification: the (in)dependence and the (a)symmetry of the stock sorting procedure, and the sorting
breakpoints. A sequential and symmetric sort of stocks into long and short portfolios conditioned on
control variables produces unbiased factors. Our results are stronger when whole firm samples are used to define breakpoints and are also robust to the inclusion of a third dimension in the multiple sorting.