This is a summary of “Toward ESG Alpha: Analyzing ESG Exposures through a Factor Lens” by Ananth Madhavan, Aleksander Sobczyk, and Andrew Ang, published in the First Quarter 2021 issue of the Financial Analysts Journal.
Rather than trying to assess the likely success of idiosyncratic ESG stock selection, observing factor weightings in ESG funds could provide better insight into future performance.
What’s the Investment Issue?
Robust evidence indicates that style factors are linked to long-term risk premiums. But to what extent are factors linked to environmental, social, and governance (ESG) performance?
Factors such as value, quality, momentum, size, and minimum volatility have decades-long series. So if links can be discerned between the ESG scores of the underlying companies in investment funds and the funds’ factor weightings, key long-term drivers of ESG performance would be revealed.
To put it another way, when ESG scores are correlated with factors with a history of outperformance, it seems reasonable that the fund could deliver high excess returns too. The reverse should also be true.
How Do the Authors Tackle the Issue?
To test their theory, the authors use the MSCI ESG scoring system, which covers 1,312 active US equity mutual funds and more than 600,000 equity and fixed-income securities globally sorted into three ESG categories: sustainable impact, values alignment, and risks.
The authors use bottom-up analysis, assessing each individual stock holding rather than regressions of portfolio returns. They argue that bottom-up analysis confers greater statistical power in testing how ESG scores affect alpha.
The chosen factors, proxied by relevant MSCI indexes, are value, size, quality, momentum, large-cap multifactor, small-cap multifactor, and minimum volatility.
The authors’ approach breaks down a fund’s alpha (return in excess of the fund’s benchmark) into (1) returns from static factor exposures, for instance, the long-term overweighting of quality stocks; (2) factor timing, where managers use their judgment to regularly adjust factor exposures; and (3) individual security selection.
Additionally, the article separates the impact of ESG into “Factor ESG” and “Idiosyncratic ESG.” As the label implies, Factor ESG is the component of fund ESG scores that is related to style factors. Idiosyncratic ESG, in contrast, reflects stock selection that is not correlated with factors.
What Are the Findings?
In terms of how funds’ factor weightings compare with ESG scores, the authors find that funds with high environmental scores tend to load up with quality and momentum factor stocks. Some 75% of “E” scores can be explained by style factors, but factors explain only 25% of “S” scores and a mere 14% of “G” scores. Funds with high E scores also overweight low volatility and larger companies.
The authors also discover a strong positive link between fund alpha and Factor ESG scores. They find that Factor ESG components are correlated with fund alpha, and the relationship is statistically strong. There is, however, little relationship between Idiosyncratic ESG and alpha; some evidence even suggests that this relationship is negative.
There is weak evidence of correlation between active returns and ESG scores. However, funds in the top decile of active returns do have significantly lower carbon intensity scores—85.1 carbon emissions per dollar of sales, compared with 164.2 for the other nine deciles.
What Are the Implications for Investors and Investment Managers?
Investors need to be aware that ESG portfolio construction may lead to factor tilts that differ from the market as a whole. If these exposures are not wanted in the portfolio, investors will need to adjust exposures accordingly.
ESG exposure—especially to funds that have high environmental scores associated with larger quality and momentum factor weightings—is shown to be rewarded. But the link between high ESG ratings and high returns is found only in ESG components that are correlated with certain factors. Idiosyncratic ESG security selection, which is unrelated to factor weightings, results in insignificant excess return premiums.