Fama & French (1993) - Common Risk Factors in Stock and Bond Returns
Key Insights
- Fama and French extend the CAPM by adding size (SMB) and value (HML) factors to explain cross-sectional variation in stock returns, demonstrating that the three-factor model explains patterns that the CAPM cannot.
Edit on GitHub — registry.json
Overview
Fama and French (1993) introduce the three-factor model, extending the CAPM with SMB (size) and HML (value) factors. It is one of the most influential asset pricing models, explaining cross-sectional patterns that the CAPM cannot.
The CAPM's Empirical Failure
By the early 1990s the CAPM was in retreat: small stocks had outperformed large, and cheap (high book-to-market) stocks had outperformed expensive ones, with neither effect explained by beta. The failures were systematic, not noise, which meant either the market was broadly mispriced or the asset pricing model was missing priced risk factors.
The Construction
The paper builds the factors by sorting stocks into portfolios on market equity (ME, size) and book-to-market equity (BE/ME). SMB (small minus big) is the return spread between small and large firms; HML (high minus low) is the spread between high and low book-to-market firms. Used alongside the market factor, the three-factor model regresses every stock's returns on market, size, and value exposure.
The Results
The three-factor model absorbs most of the CAPM's anomalies: the size and value effects disappear into their factor loadings, and portfolios that differ wildly in average returns now differ wildly in factor exposure instead. The paper also extends the framework to bonds with term and default factors, establishing the multi-factor methodology as the default way to describe the cross-section of returns.
Why It Matters
The paper created the factor industry: style boxes, smart beta, and performance attribution all speak its language. The enduring debate is interpretation — Fama and French read HML and SMB as risk factors, while behavioralists read them as mispricing patterns — but both sides use the same empirical technology. Every later extension, including the five-factor model, is a refinement of this paper's method.
Key Takeaways
- SMB and HML absorb the anomalies the CAPM could not price — size and value became factors, not puzzles.
- The 2x3 portfolio construction is the industry template for defining and benchmarking factors.
- The risk-versus-mispricing debate is about interpretation, not about the model's empirical power.