Definition
Fama-French factors are traded or constructed return series used to describe systematic variation in asset returns. The familiar three-factor model combines a market excess return with size and value factors; later versions add factors such as profitability and investment. A factor exposure is an estimated relation in a specified sample, not a promise of a future premium.
How it is calculated
Asset excess return = alpha + market beta times market excess return + size loading times SMB + value loading times HML + residual.
| Symbol | Meaning | Unit |
|---|---|---|
| R | asset or market return | return per period |
| R_f | risk-free return | return per period |
| alpha | intercept | return per period |
| beta | market loading | dimensionless |
| s | size-factor loading | dimensionless |
| h | value-factor loading | dimensionless |
| SMB | small-minus-big factor return | return per period |
| HML | high-minus-low value factor return | return per period |
| epsilon | regression residual notation | return per period |
Interpretation
Factors are labeled return streams that act like explanatory coordinates for a portfolio. The loadings say how strongly an asset moved with each coordinate in the sample. They do not by themselves establish why the factor exists or whether it will be rewarded later.
Assumptions and conventions
-
The factor definitions and portfolio construction are fixed before estimation.
-
The linear relation is an approximation over the selected assets and sample.
All returns and alpha use the same period; loadings are dimensionless.
Limitations and common misreadings
-
A factor loading is not the same thing as a causal mechanism.
-
The factor label does not make an estimated alpha economically real.
References
- Eugene F. Fama, Kenneth R. French. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics. Sections 2–3. doi:10.1016/0304-405X(93)90023-5. Source

