Mandatory Environmental, Social and Governance Reporting, Capital Allocation and the Green Behaviour of Firms

Achieving net-zero targets by 2050 necessitates active private sector involvement in reducing greenhouse gas emissions and advancing technological innovations. The UK and the EU have emerged as leaders in corporate sustainability legislation, implementing mandatory ESG (Environmental, Social, and Governance) reporting for listed or large firms. These regulations require companies to disclose their environmental impact, climate-related risks and opportunities, and management strategies.

Mandatory ESG disclosure enhances corporate accountability and reduces information asymmetries among firms, investors, governments, and the public, and its impact is further amplified when paired with environmental regulations and innovation incentives.

While existing literature has primarily focused on factors influencing the quantity and quality of ESG reporting, as well as its effects on corporate financial performance and investment efficiency, this project aims to delve deeper. Specifically, it will examine the impact of mandatory ESG reporting on firms' access to capital, capital allocation, and green behaviour, including environmental performance and green innovation. The study will address three key research questions:

1. How does ESG reporting affect firms' access to capital and subsequent capital allocation, particularly for those with lower climate-related risks or stronger existing ESG performance? 2. To what extent does enhanced ESG reporting influence firms' short- and medium-term environmental performance?

3. Does stricter ESG reporting regulation drive firms to invest in green innovation, and how is this influenced by the presence of innovation incentives?

The research will fill a critical gap in the existing literature by providing insights to help firms and governments understand the effects and challenges of implementing and expanding ESG reporting requirements, alongside incentives for green innovation.

To address these research questions, the study will employ statistical analysis and econometric modelling of panel data. Additionally, machine learning techniques - such as textual analysis, decision trees, random forests, and neural networks - will be utilized for prediction, classification, and pattern recognition, particularly when handling variables with diverse distributions and complex functional forms.