Equity Multi-Factor Approaches: Sum of Factors vs. Multi-Factor Ranking

Abstract

There is a plethora of literature on the persistence and rationale of individual equity factor premia, which investors are often advised to extract through forming multi-factor portfolios – however, not so much attention is paid to the nuances of forming such portfolios and their implications.

There are generally two approaches for forming equity multi-factor portfolios:

- Sum of Factors (SoF) – where separate respective factor portfolios are first formed which are then subsequently combined.

- Multi-Factor Ranking (MFR) – where individual stocks are selected based on a specified multi-factor ranking model.

In this paper, we empirically demonstrate the implications of the two approaches when allocating across four equity factors: Value, Momentum, Low Volatility and Quality. We do so in the context of two topical implementation considerations:

- Portfolio size: Concentrated versus diversified portfolios.

- Factor exposure: Ex-ante versus ex-post factor exposure.

In summary when comparing the two approaches, we find:

- MFR portfolios have historically outperformed SoF portfolios irrespective of portfolio size.

- However, this has been historically driven by the persistently lower CAPM Beta of the MFR portfolios, which has not been by design and which is accentuated in concentrated portfolios.

- This lower CAPM Beta is from an implicit low volatility factor bias.

- Such a bias comes with performance risk implications that investors should be mindful of.

Keywords: smart beta; factors; portfolio construction

JEL Classification: G10; G11