Scientific Beta
The successive growth of ESG and climate investing has led practitioners to promote strategies that aim to fulfil both higher ESG scores and lower carbon emissions, without ever considering the potential trade-off between these two dimensions. We quantify this very trade-off by measuring the carbon intensity increase investors face when they add ESG score objectives to a low carbon intensity objective in global equity portfolios.
The successive growth of ESG and climate investing has led practitioners to promote strategies that aim to fulfil both higher ESG scores and lower carbon emissions, without ever considering the potential trade-off between these two dimensions.
We quantify this very trade-off by measuring the carbon intensity increase investors face when they add ESG score objectives to a low carbon intensity objective in global equity portfolios. We account for heterogeneity in ESG preferences by relying on 25 ESG theme scores from three major ESG rating providers, and building portfolios based on numerous combinations of ESG objectives and carbon reduction. By comparing the greenness of portfolios built to have both higher ESG scores and lower carbon intensity to that of portfolios solely built to reduce carbon intensity, we are able to compute the incremental impact of the inclusion of ESG scores on carbon intensity reduction, which we call green dilution.
We show green dilution is pervasive, regardless of which ESG scores are targeted as objectives, substantial, with an average of 92% across our portfolios, and robust across several alternative specifications. A 92% green dilution means that 92% of the carbon intensity reduction investors could have reached by solely weighting stocks to minimise carbon intensity is lost when adding ESG scores as a partial weight determinant. Only 8% of the carbon reduction objective survived the inclusion of ESG scores in portfolio weighting schemes.
Adding a single ESG score in portfolio construction, so that stock weights are equally determined by carbon intensity and the ESG score in question, leads to a green dilution of 65% on average. Mixing ESG scores one might expect to be green, scores belonging to the environmental pillar, with carbon intensity also leads to a substantial deterioration in green performance. Mixing scores from the social or governance pillars with carbon intensity routinely results in portfolios than are less green than the cap-weighted index: on average, social and governance scores more than completely reversed the carbon reduction objective.
Green dilution has a simple explanation. The cross-sectional rank correlation1 between ESG scores and carbon intensity is close to zero. The two objectives are unrelated and are therefore hard for investors to simultaneously achieve. This low correlation explains why one should not mix ESG and carbon scores in portfolio weighting schemes. A more sensible alternative is to separate the two objectives, by first screening out stocks with low ESG scores, and then weighting the remaining stocks by the investor’s key objective, carbon intensity in our case. Since both dimensions are unrelated, screening out stocks by ESG scores does not affect the carbon intensity distribution of the stock universe. ESG exclusions thus result in a neutral impact on portfolio carbon intensity, with a green dilution close to zero.
Overall, we provide clear evidence against the quantitative mixing of ESG and carbon scores in equity portfolio weighting schemes, which comes at great carbon cost for green investors. Conversely, we provide evidence in favour of the exclusionary approach to ESG objectives, to best accommodate multiple non-financial and unrelated objectives.
1The cross-sectional absolute correlation is even lower due to outliers.