Essays in Sustainable Finance and ESG modeling

Environmental, social, and governance (ESG) information has become increasingly important to investors, corporations, academics, and policymakers. As financial markets increasingly rely on non-financial information to assess corporate sustainability and risk, this dissertation examines how ESG measures are constructed, interpreted, and incorporated into financial analysis. Using econometric and empirical finance methods, the three chapters investigate the relationship between corporate governance and ESG performance, the role of missing data in ESG measurement, and the implications of incomplete emissions information for asset pricing. The first chapter examines whether board diversity and network centrality are associated with firms’ environmental and social performance, finding signifi cant relationships between corporate governance structures and ESG outcomes. The second chapter evaluates alternative approaches to missing-value imputation in ESG data and finds that machine-learning methods, particularly K-nearest-neighbors imputation, can improve the accuracy of emissions measures and materially affect resulting ESG scores. The third chapter extends this analysis to financial markets by examining whether missing emissions information is associated with equity returns. Using international firm-level data, the analysis develops emissions and missing-data factors within a Fama–French-style framework and finds evidence that incomplete emissions information is associated with distinct return patterns. Taken together, the chapters demonstrate that the economic interpretation of ESG information depends not only on firms’ underlying sustainability characteristics but also on corporate governance, data construction, and disclosure quality. The dissertation highlights the importance of accounting for these factors when using ESG information in financial research and investment analysis.