Portfolio Choice with ESG Disagreement: Customizing Sustainability through Direct Indexing

There is strong demand for sustainable investing in direct indexing strategies. We examine implications of disagreement about environmental, social, and governance (ESG) ratings for portfolio choice by maximizing ESG scores subject to a tracking error constraint. Varying the ESG score we optimize on results in portfolios with substantial differences. Correlations between active weights of the ESG-optimized portfolios are even lower than correlations between ESG scores. Optimal portfolios have positive (negative) active weights in stocks with high (low) ESG scores, as expected, but in both cases a small market capitalization or high specific risk pulls the active weight toward zero. To attenuate ESG disagreement, we propose an optimal portfolio that maximizes the average ESG score across vendors and explicitly manages ESG disagreement by penalizing stocks with high ESG uncertainty. Increasing ESG uncertainty aversion thus means investing less in stocks with high ESG disagreement. Our solution is well suited for direct indexing clients wanting to express their sustainability beliefs.