The Effect of ESG Disclosure on Firm Performance: Evidence from Pharmaceutical, Chemical and Textile Industries of Bangladesh

Abstract

This research explores the impact of the practice of Environmental, Social, and Governance (ESG) disclosure on the financial performance of listed companies in the sector of Pharma, Chemical and Textile in Bangladesh. ESG disclosure is further measured using a custom unweighted binary scorecard with 33 indicators across the dimensions of environmental, social, and governance (ESG) based on a balanced sample of 105 firm-year observations drawn from 21 listed companies in the Dhaka Stock Exchange (DSE) for the 2021–2025 period. Financial performance is measured using Return on Assets (ROA) and control variables are firm size, leverage and revenue growth. ESG disclosure practices are well dispersed and differ significantly across the sample firms (mean score of 17.09/33), whilst the profitability too is well dispersed. The correlation and regression analyses indicate a positive relationship between ESG disclosure and ROA that is statistically significant, even when allowing for heteroskedasticity-consistent standard errors, meaning that companies with high ESG disclosure scores are more likely to have high ROAs. Leverage is also negatively and significantly associated with ROA, which reflects the higher financial risk of the debt-heavy capital structure while the firm size is positively associated with ROA. But once robust standard errors are used, revenue growth is not statistically significant. Such findings align with the Stakeholder and Legitimacy Theory which indicates that ESG disclosure is a value relevant signal for stakeholders even in an emerging market with weak enforcement like Bangladesh. The study also enriches the scarce empirical evidence on ESG performance relationship in environmental sensitive industries in Bangladesh and provides implications for industry stakeholders, regulators, investors and industry bodies that aim to have robust ESG reporting. Limitations such as sample size, survivorship effect and the binary nature of the disclosure scorecard are recognized and suggestions for future research employing richer panel methods and disclosure-quality measures are offered.