There is growing interest by investors to “do the right thing” by using their influence to pressure companies to improve their approach to environmental, social and governance (ESG) issues. But a major challenge is how to minimize the potential costs imposed by ESG constraints on portfolios and overcome the persistent sparsity of ESG data resulting from companies’ non-reporting.
In this study, we propose a quantitative approach to integrating ESG into portfolios that is expected to deliver comparable performance to non-ESG portfolios and is capable of classifying companies based on ESG even when they do not disclose sufficient data. The approach is particularly suitable for quantitative portfolios with large numbers of positions and many small exposures. In such portfolios, one can generally identify companies with bad ESG metrics and swap them out for companies with similar expected future returns and better ESG scores. This allows the manager to efficiently tilt the entire portfolio towards better ESG companies without the need to employ detailed ESG analysis of individual firms.