Who Cares about Purity of Factor Indexes? A Comment on “Evaluating the Efficiency of ‘Smart Beta’ Indexes”

Abstract

A recent article by Hunstad and Dekhayser (JII, Summer 2015) introduced a novel measure of factor “purity”—the factor efficiency ratio—and concluded that the indexes analyzed “were generally unable to provide desired factor exposures without taking on substantial unintended exposures” and that indexes are not “pure” in their delivery of intended factor exposures. This note points out several questions regarding the relevance of factor efficiency ratios and similar assessments of purity of factor indexes.