Female Equity Analysts and Corporate Environmental and Social Performance

This paper examines the impact of female analyst coverage on firms’ environmental and social (E&S) performance. Exploiting broker closures as a quasi-exogenous shock to analyst coverage, we find that firms experiencing an exogenous decline in female analyst coverage subsequently show a significantly larger drop in E&S scores than those experiencing an equivalent decline in male analyst coverage. To explore the underlying mechanisms, we develop novel machine-learning models to analyze more than 2.4 million analyst reports and 120,000 earnings call transcripts. Our analysis shows that, compared with their male counterparts, female analysts are more likely to address E&S issues, particularly those involving regulatory compliance, stakeholders, and the environment, in both research reports and earnings conference calls. They also display distinct cognitive and linguistic patterns when discussing E&S issues. Furthermore, female analysts are more likely to issue lower stock recommendations and target prices (lower stock recommendations) following negative E&S discussions in their reports (E&S incidents) than male analysts. Finally, investors respond more strongly to female analysts’ negative tones when discussing E&S issues. Overall, our findings suggest that gender diversity among analysts plays a significant role in shaping corporate E&S practices and provide new insights into the origins of gender differences in skills within the equity analyst profession.

This paper was accepted by Will Cong, finance.

Funding: We acknowledge financial support from the Social Sciences and Humanities Research Council of Canada [Grant 435-2022-0285] and the Sauder Exploratory Research Grants Program. K. Li acknowledges financial support from the Canada Research Chair in Corporate Governance. T. Zhang acknowledges data support from the Whitcomb Center for Research in Financial Services at Rutgers University.

Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2024.06429.