The pertinence of incorporating ESG ratings to make investment decisions: a quantitative analysis using machine learning

ABSTRACT

Global sustainability being the major goal ahead, socially conscious investors are concerned about non-financial dimensions of investments like impact on environment (E), social relations (S), and corporate governance (G). This research aims to answer whether including ESG data points is conducive to profitable investments while promoting sustainability. Web-scraped a unique dataset of ESG and key financial data of 1400+ companies from 34 stock markets internationally. Quantitative analysis is performed on this data with the aim of determining whether the qualitative aspect of sustainable investments is tantamount to financial parameters. Better ESG scores indicate better financial performance. Return on equity was 14% greater for top 10% ESG companies than bottom 10%. Prediction accuracy of ML models like linear, random forest regression increased when training data included both ESG and financial data. The research concludes with a propitious relationship between ESG data and financial growth parameters which are worth probing further.

KEYWORDS:

• ESG
• machine learning
• sustainable
• finance
• investment
• stock market

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data used for the findings of this study is not readily available anywhere. It was collected using web-scraping. The data used is organized and scraped to suit the needs of this study specifically. The data was derived from the following resources available in the public domain: (1) https://in.finance.yahoo.com/ and (2) https://www.sustainalytics.com/esg-ratings. The data is available on request.