The Black‐Box of ESG Scores From Rating Agencies: Do They Genuinely Reflect Sustainability Practices, or Are They Disproportionately Shaped by Financial Performance?

This study examines the environmental, social and governance (ESG) scoring methodologies used by Bloomberg and S&P Global through the lens of Data Envelopment Analysis (DEA). It addresses a notable gap in the literature by identifying the underlying factors that shape ESG scores and providing practical insights for companies seeking to understand or improve their sustainability ratings. Our comparative analysis reveals clear differences between the two rating agencies. While Bloomberg's raw ESG scores are generally higher than those of S&P Global, the DEA‐normalised results tell a different story. Bloomberg applies stricter internal benchmarks, resulting in lower efficiency scores. In contrast, S&P's lower raw scores convert into higher DEA efficiencies, suggesting a more lenient, peer‐based benchmarking approach that tends to cluster firms near the top regardless of their absolute ESG performance. A particularly striking finding is that 99% of ESG scores from both agencies correlate with net income, highlighting a strong connection between financial performance and ESG ratings. Our regression analysis supports this, showing that firms with better financial outcomes tend to receive higher ESG scores. However, we also find that companies with growing cash reserves—often indicative of reinvestment and expansion—may be penalised, receiving lower ESG scores. This suggests a potential bias against firms prioritising long‐term growth over immediate returns. This study lays the groundwork for future research aimed at refining ESG datasets and expanding the scope of analysis.