From Academic Factor Portfolios to Investable Factor Products: Return Differences, Mapping Losses, Implementation Frictions, and a Conditional Framework

Abstract

The academic factor premium comes from the “laboratory measurement” of an idealized long-short combination. For investors to actually gain factor exposure, they must first map academic signals into portfolioable data, and then execute the data through the product vehicle. This article organizes its overview along the three-layer filtering logic of “existence–persistence–availability” and summarizes a two-layer explanatory framework of combination mapping loss (short selling constraints, benchmark constraints, factor definition changes) and implementation delivery loss (transaction costs, rates, strategy capacity). The review found that: (1) the shrinkage system exists systematically, and existing major studies generally show a shrinkage ranking of momentum > value > market ≈ size (mainly based on US active mutual funds samples); (2) the shrinkage estimates of the same factor vary greatly in different studies, due to differences in three systematic methodologies: sample period, cost model and factor definition; (3) the impact of turnover rate and short selling contribution on the shrinkage magnitude has strong cross-documentary evidence, while the role of execution capability and constraint strength mainly comes from inferences from comparisons between different studies. Based on this, this paper proposes an inductive framework based on factor type × product carrier conditionalization. The contribution of this paper is not to provide a universally applicable reduction ratio, but to establish a conditional analytical framework that distinguishes between three levels of evidence: “original conclusions, cross-documentary consensus, and inferences from this paper,” in order to explain the discrepancies between studies, and to identify, in particular, the empirical gap in the measurement of factor capture rate in Smart Beta ETFs.