Journal of
Italian
Banking
Association

year 100
 

Archive » April 2024 » Investment decisions and Esg ratings: a Machine Learning approach with Balance sheet and Systemic risk indicators

Investment decisions and Esg ratings: a Machine Learning approach with Balance sheet and Systemic risk indicators

Rosella Castellano, Annalisa Ferrari, Federico Cini
April 2024 - n. 4
Keywords: Machine Learning, rischi Esg, rischio di performance, strategie di investimento
Jel codes: G3, G11, G32, C6, C8

For more than ten years, the concept of sustainability has been identified as a key factor in investment strategies. Empirical evidence attributes better risk-return profiles to sustainable investments, and therefore investors consider Esg ratings to be essential for profitable decisions. The classic portfolio selection metrics have, therefore, been enriched by these new Esg measures, both individually and as a whole. The sustainable investor has a medium-long term time horizon that favors the contamination of classic metrics by Esg ones. The aim of the paper is to verify whether a selected set of balance sheet variables and a dynamic measure of systemic risk make it possible to identify a company's Esg rating class. We use companies in the EuroStoxx 600 index from 2016-2021, and applying a Machine Learning (Random Forest) model, we estimate the Esg rating class with an accuracy unprecedented in literature. This agile and thrifty model provides information to the sustainable investor for strategic choices and paves the way for estimating the Esg rating also for unlisted companies and Smes.

Interested in this paper?
Buy the issue