Understanding Corporate Emissions Data: Implications for Research Design

Abstract

This study investigates the characteristics and limitations of corporate greenhouse gas (GHG) emissions data used in environmental and financial research. Using data from S&P Global Trucost, LSEG, and Sustainalytics, the researchers examine differences between disclosed and vendor-estimated emissions across publicly listed companies. The findings reveal substantial inconsistencies in estimated emissions, strong associations with company sales and exchange rates, and limited availability of disclosure-linked data. The research demonstrates how these factors can influence financial analysis and provides methodological recommendations to improve the reliability of emissions-based research.

Introduction

Corporate emissions data have become increasingly important for assessing climate-related risks, environmental performance, and sustainable investment opportunities. However, researchers and investors often rely on emissions estimates from data providers because many companies do not publicly disclose their greenhouse gas emissions.

The study examines whether estimated emissions accurately reflect differences in corporate environmental performance and how variations in data sources, measurement methods, and sample composition affect research outcomes. It highlights the importance of distinguishing between company-disclosed and vendor-estimated information when evaluating emissions and financial performance.

Research Objectives and Methodology

The research aims to identify the empirical characteristics of corporate emissions data, assess differences between disclosed and estimated emissions, and determine how measurement choices influence financial research findings.

The researchers analyzed Scope 1 and Scope 2 emissions data from S&P Global Trucost between 2002 and 2022, supplemented by information from LSEG and Sustainalytics. The analysis included 154,366 emissions observations matched with Compustat financial data and compared emissions classifications, cross-provider consistency, company characteristics, and exchange-rate sensitivity. Regression analyses further examined relationships between emissions, firm valuation, cost of equity, and institutional ownership.

Implications and Applications

The findings highlight significant implications for researchers, investors, companies, and policymakers using corporate emissions data. Researchers should distinguish disclosed emissions from vendor estimates, account for company size and exchange-rate effects, and conduct sensitivity analyses across alternative data sources and sample structures. Using emissions intensity rather than absolute emissions can also improve the consistency of empirical results.

For investors and policymakers, the study emphasizes the importance of evaluating data quality before comparing corporate environmental performance or making investment and regulatory decisions. Greater transparency in estimation methodologies, improved corporate emissions disclosure, and careful interpretation of vendor-generated data can support more reliable climate risk assessments and sustainable finance practices.

Conclusion

The research concludes that corporate emissions data differ substantially in coverage, measurement characteristics, and comparability. Vendor-estimated emissions provide important information where corporate disclosures are unavailable, but their strong associations with sales, exchange rates, and industry-level estimation methods can influence empirical findings. Disclosure-linked emissions data remain limited, covering only 5.3% of Compustat firm-year observations in the study.

The authors recommend stronger methodological controls, greater transparency in emissions estimation, and improved corporate disclosure practices. These measures can strengthen environmental research and support more informed investment decisions and climate-related policymaking.

DOI: 10.1016/j.jbankfin.2026.107822

Original Research Paper Link: https://doi.org/10.1016/j.jbankfin.2026.107822