We are investigating whether a the subset of Environmental, Social and Governance indicators defined by the Sustainable Finance Disclosure Regulation (SFDR) is a good base to predict company ESG (Environmental, Social, and Governance) performance for a subset of 2,900 companies, mainly based in the United-States.
These 2,900 companies, our reference universe, will correspond to all authorized stocks within the Natixis Investment Managers Challenge, 2021 edition. It comprises over 2,000 companies for which we retrieved ESG data for year 2020.
As a proxy for ESG performance, we consider ESG grades given by three data providers : Thompson-Reuters & MSCI, both of which are used by Asset Managers, and CSRHub, which is open access.
Our goal is to build our own robust understanding of ESG performance:
- Compare notation agency notations and evidence bias as per paper [XXX]
- Test the robustness of SFDR indicators as an ESG performance predictor
- Build our own transparent ESG Scoring methodology.
Our aim is to predict and compare scorings from three ESG data providers on our subset of 2,000 companies, evidence notation biases as per [quote paper],
We start off by using data from Refinitiv-Eikon, a well-known notation agency in the financial industry.