Skip to content
All library documents

Interpreting Fama–French Three-Factor Fit for Industry Portfolios

Article Quant Q&A · Author: Gustav
Read the original

Summary

The document raises a model-fit question about two-stage Fama–MacBeth regressions using ten industry-sorted portfolios as dependent variables and the Fama–French three factors as explanatory returns. It reports that the value-weighted portfolios have an R-squared of 31% and adjusted R-squared of 11%, while equal-weighted portfolios have higher reported values of 52% and 44%, respectively. The question is whether the lower values for value-weighted portfolios are expected in this setting.

No answer or explanation is included, so the document does not establish whether either level of fit is typical or diagnose the difference between weighting schemes. It also does not provide sample period, estimation details, standard errors, or portfolio construction specifics. The figures are therefore observations prompting interpretation, not evidence for a general conclusion about the explanatory power of the three-factor model. Readers would need additional model and data context to assess the fit.

Key ideas

  • The question concerns two-stage Fama–MacBeth regressions on industry-sorted portfolios with the Fama–French three factors.
  • The document reports lower R-squared measures for value-weighted portfolios than for equal-weighted portfolios.
  • It asks whether the lower fit is expected but includes no response or diagnosis.
  • The reported fit statistics alone do not explain the difference between weighting schemes.

Tags

EquitiesFactor investingStatisticsPortfolio construction
Full text
# R^2 and adj. R^2 Fama French 3 factor with industry portfolios


# R^2 and adj. R^2 Fama French 3 factor with industry portfolios












I'm running two-stage Fama Macbeth regressions with 10 portfolios sorted on industry on the LHS and Fama French 3-factors (excess market returns, SMB, HML) created by using the same stocks on the RHS.

The "issue" that I have is rather low values for R2(31%) and adj R2(11%) for value weighted portfolios. If I equally weight the portfolios R2 is 52% and adj. R2 44%. Are low R2 values to be expected for industry-sorted portfolios for FF3 models?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.