Environmental, social and governance (ESG) scores are becoming an increasingly important tool for asset managers to design and implement ESG investment strategies. However, there are drawbacks in using headline ESG scores that limit their usefulness. ESG scores amalgamate a broad range of fundamentally different factors, which creates ambiguity. Weak scores in one pillar can offset strong scores in another pillar.
We demonstrate an investment strategy based on deconstructing ESG scores. The strategy focuses on specific underlying ESG categories such as emissions reduction and human rights. To implement our investment strategy, we exclude firms with the lowest scores in certain ESG categories of interest and implement a best-in-class investment strategy.
This approach helps investors overcome the "aggregated confusion" inherent in ESG scores. Moreover, it enables investors to better track the sustainability performance trajectory of their portfolio against their stated sustainable investment objectives.
We find that simple exclusions enable substantial improvements to the headline ESG score of the portfolio. Here, the portfolio's financial performance only suffers a marginal impact relative to a broad stock market benchmark. However, the exclusion results in regional and sectoral biases compared to the benchmark.
To counter this, we adopt a best-in-class strategy that excludes firms with the lowest category scores and reinvests the proceeds in firms with the highest scores. This approach helps reduce the tracking error of the portfolio, and slightly improve its risk adjusted performance while still yielding a large gain in the headline ESG score.
Environmental, Social, and Governance (ESG) scores are becoming an increasingly important tool for asset managers to design and implement ESG investment strategies. They amalgamate a broad range of fundamentally different factors, creating ambiguity for investors as to the signals of higher or lower ESG scores. We explore the feasibility and performance of more targeted investment strategies based on specific categories by deconstructing ESG scores into their granular components. First, we investigate the characteristics of the various categories underlying ESG scores. Not all types of ESG categories lend themselves to more targeted strategies, which is related to both limits to ESG data disclosure and the fundamental challenge of translating qualitative characteristics into quantitative measures. Second, we consider an investment scheme based on the exclusion of firms with the lowest scores in each category of interest. In most cases, this targeted strategy still allows investors to substantially improve the portfolio headline ESG score, with only a marginal impact on financial performance relative to a broad stock market benchmark. The exclusion results in regional and sectoral biases relative to the benchmark, which may be undesirable for some investors. We then implement a "best-in-class" strategy, based on excluding firms with the lowest category scores and reinvesting the proceeds in firms with the highest scores maintaining the same regional and sectoral composition. This approach reduces the tracking error of the portfolio and slightly improves its risk-adjusted performance while still yielding a large gain in the headline ESG score.