Adding an ESG Variable to a Fama–French Three-Factor Regression

Summary

The document addresses how to represent sustainability in a return comparison using the Fama–French three-factor model. It considers two related but distinct questions: estimating the return difference between a portfolio of high-scoring ESG firms and a peer portfolio of lower-scoring firms, or estimating how ESG ratings relate to returns across individual constituents. For the portfolio comparison, the suggested explanatory variable is an indicator that identifies membership in the ESG portfolio. For constituent-level analysis, the suggested variable is the ESG score or rating itself.

These choices correspond to different research questions: a binary indicator captures a group distinction, whereas a score preserves variation in ESG ratings. The response is a concise modeling suggestion and provides no dataset, regression specification details, empirical results, or guidance on controls, portfolio construction, or causal interpretation. Researchers would need to define the sample and variable timing carefully before drawing conclusions about performance or sustainability effects.

Key ideas

• A binary indicator can identify membership in the high-ESG portfolio in a return comparison.
• An ESG score or rating can serve as an explanatory variable for constituent-level analysis.
• The choice between an indicator and a score depends on the research question.
• The document offers a variable construction idea but no empirical results or full regression design.

Tags

# How to create an additional ESG factor into a Fama and French 3 Factor Model


# How to create an additional ESG factor into a Fama and French 3 Factor Model












I'm writing my thesis on the relationship between financial performance and sustainability. For that I'm comparing the returns of 2 portfolios, the first having really good ESG scoring firms, while the second portfolio consists of similar firms to the first (same peer group), but showing poorer ESG scores. I'm then using the Fama and French 3 Factor model, while integrating a new ESG factor in the regression. However I'm having some trouble building that variable. Can someone help ?

## Answer by Andreas (score 4, accepted)

https://quant.stackexchange.com/a/53065

If you just want to compare the portfolios you could work with a dummy variable (A column in your data which is "1" for the ESG firms and "0" for the non-ESG firms). The regression output will then state the impact of such a variable.

Alternatively, if you intend to measure the impact of an ESG-Score/Rating on the individual portfolio constituents, you could add such a Score as an explanatory variable.

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.