Optimization of Portfolios Using ESG Factors

Abstract

This study presents a thorough analysis of the U.S. stock market, examining how ESG metrics can be practically integrated into portfolio construction and investment strategies. Building upon the foundational work of Fama-French and Carhart, we extend their traditional factor models by incorporating granular ESG metrics retrieved from MSCI’s ESG database. Our methodology involves constructing decile portfolios from the lowest-performing decile (bottom 10%) to the highest-performing decile (top 10%) sorting stocks monthly based on their individual ESG dimension scores and simulating dynamic investment strategies. Leveraging regression modelling loops and sorting algorithms in Python we systematically generate 300 unique combinations of ESG dimensions portfolio strategies with 4.500 factor model risk indicators to identify optimal and profitable trading strategies.Our findings reveal that traditional risk factors are neither uniform nor linear in their interaction with ESG scores but instead exhibit nuanced and multidimensional behavior across the ESG spectrum. Once we accounted for market, size, value, profitability, investment, and momentum effects, neither the aggregate ESG score nor the individual Environmental, Social, and Governance pillars generated any statistically significant abnormal returns, aside from a few isolated results in specific deciles and models.In contrast, when examining the individual ESG dimensions, we uncover several targeted portfolio strategies capable of delivering robust excess returns beyond traditional factor explanations and outpacing the S&P 500 on a risk-adjusted basis. Intriguingly, our analysis identifies the most significant investment opportunities not at the extremes of ESG performance but rather in the overlooked segments prior to the top or above the bottom. These "ESG followers" and "ESG laggards" appear consistently mispriced, highlighting a critical blind spot in conventional ESG investing strategies. Examining this behavior, our evidence indicates that the significant abnormal return portfolios were disproportionately associated with ESG dimensions exhibiting heightened investor scrutiny.