ESG performance, analysts’ forecast accuracy and board gender diversity: international evidence

Purpose The purpose of this study is to investigate whether environmental, social, and governance (ESG) ratings are associated with analysts’ forecast accuracy and whether board gender diversity moderates the relationship between ESG ratings and analysts’ forecast accuracy. Design/methodology/approach The sample comprises 23,498 firm-year observations across 36 countries over the period 2011–2022, with data obtained from I/B/E/S, Thomson Reuters and Refinitiv. Generalized Least Squares with panel-specific heteroskedasticity is used as the baseline estimator, with model choice supported by a Wald test. Robustness checks include fixed-effects models with clustered standard errors, propensity score matching and a Heckman selection correction. Findings ESG ratings are associated with lower analyst forecast errors when board gender diversity remains below approximately 18.5% (or 20.7% for executive-level diversity). Beyond these levels, the relationship reverses, with higher diversity associated with reduced ESG informativeness. The interaction term (+0.0097***) reflects this pattern: ESG is associated with improved accuracy on less diverse boards, has limited net effect near the threshold and is associated with reduced accuracy on more diverse boards. This non-linearity is not captured under a standard linear specification and becomes apparent only through marginal effects analysis. Practical implications For firms, the findings suggest that board diversification efforts may be more effective when undertaken prior to large-scale ESG commitments. Analysts may benefit from considering board composition when interpreting ESG ratings. Regulators may wish to consider coordinating diversity-related requirements with ESG disclosure mandates rather than addressing them independently. Originality/value This study contributes evidence of a threshold-based relationship between board diversity and ESG informativeness using cross-country data spanning 36 countries. The identified threshold of approximately 18.5% may serve as a reference point for discussions on sustainability governance standards across different national contexts.