Fama–French five-factor model

Fama–French five-factor model
Name	Fama–French five-factor model
Authors	Eugene F. Fama; Kenneth R. French
Introduced	2015
Field	Asset pricing
Components	Market risk, Size, Value, Profitability, Investment
Related	Capital Asset Pricing Model, Three-factor model, Carhart four-factor model

Fama–French five-factor model The Fama–French five-factor model is an empirical asset pricing model developed to explain cross-sectional variation in stock returns by augmenting the earlier Capital Asset Pricing Model and Three-factor model with additional firm characteristics. Introduced by Eugene F. Fama and Kenneth R. French in 2015, the model adds profitability and investment factors to the existing market, size, and value factors to improve explanatory power across portfolios. The model has been central in debates among practitioners at firms such as BlackRock, Vanguard Group, and JP Morgan Chase and academics at institutions like the University of Chicago and Dartmouth College.

Overview

The five-factor model builds on the literature that includes the Capital Asset Pricing Model, the Three-factor model, and subsequent tests by researchers at Harvard University and MIT. The model was presented in papers circulated through outlets associated with National Bureau of Economic Research, discussed at conferences hosted by the American Finance Association and cited by researchers at Columbia Business School and London School of Economics. Its development reflects ongoing empirical work in asset pricing involving datasets compiled by firms such as CRSP and Compustat, and motivated critiques from scholars affiliated with Princeton University and Stanford University.

Model Specification and Factors

The specification regresses excess returns on five risk factors: a market factor derived from a broad index such as the S&P 500; a size factor distinguishing small-cap and large-cap firms similar to portfolios used by Renaissance Technologies and Goldman Sachs; a value factor based on book-to-market ratios as in studies by Harry Markowitz-influenced literature; plus profitability and investment factors measuring operating profitability and asset growth rooted in research by Michael Jensen and Fischer Black. The factors are constructed using portfolio sorts and time-series of returns à la procedures implemented by Ken French’s data library and replicated in empirical work at New York University and University of Pennsylvania.

Formally, the model extends the three-factor regression by adding two additional factors: robust minus weak profitability and conservative minus aggressive investment. Estimation often employs ordinary least squares with adjustments for heteroskedasticity and autocorrelation as recommended in methodology texts from Econometrica and by econometricians at University of California, Berkeley institutions.

Empirical Evidence and Performance

Empirical tests conducted by teams at Wharton School, London Business School, and INSEAD find that the five-factor model explains cross-sectional returns for many sets of portfolios better than the Three-factor model, particularly when assessing anomalies cataloged by researchers at Columbia University and University of Michigan. Other studies by scholars at Yale University and Northwestern University highlight cases where value-related return patterns diminish once profitability and investment are accounted for, echoing findings published in journals such as the Journal of Finance and Review of Financial Studies.

However, replication efforts by researchers at University of California, Los Angeles and critics from University of Cambridge show persistent anomalies—such as momentum—that the five-factor model does not capture, aligning with results in the Carhart four-factor model literature and empirical work by practitioners at AQR Capital Management. Large-sample analyses using databases maintained by CRSP and Compustat reveal time-varying explanatory power and mixed out-of-sample performance noted by analysts at Moody's and Standard & Poor's.

Extensions and Criticisms

Scholars at Massachusetts Institute of Technology and Princeton University have proposed extensions incorporating momentum, liquidity, and tail risk—ideas also explored by teams at Two Sigma Investments and Citadel LLC. Critics from Columbia Business School and Oxford University argue that factor definitions can be data-mined and that the model may not hold across international markets studied by researchers at IMF and World Bank datasets. Methodological critiques draw on work by econometricians at University of Chicago and Stanford University concerning measurement error and factor redundancy, while applied researchers at Deutsche Bank and Barclays assess implementation frictions in real-world portfolio strategies.

Notable academic responses include proposals to integrate macroeconomic variables as in research from Federal Reserve Board economists and to reconcile the model with consumption-based approaches advanced by scholars at Brown University and Duke University.

Applications in Asset Pricing and Portfolio Management

In practice, asset managers at BlackRock, Vanguard Group, State Street Corporation, and hedge funds including AQR Capital Management and Bridgewater Associates use factor exposures from the five-factor model for performance attribution, risk management, and smart-beta product design. Academics at Columbia Business School and practitioners at Goldman Sachs apply the model in performance evaluation, portfolio construction, and stress-testing frameworks that interact with regulatory reporting overseen by entities such as the Securities and Exchange Commission.

The model informs academic curricula at Harvard Business School and London Business School and underpins empirical modules used in executive education by CFA Institute. Empirical implementations leverage historical return series from CRSP and accounting measures from Compustat to form tradable factor portfolios employed by institutional investors at CalPERS and Norwegian Government Pension Fund Global.

Category:Financial models