AI for ESG Investing

Course Introduction

AI for ESG Investing is a two-day online workshop that explores how investment managers and analysts can use a large language model (LLM) such as ChatGPT to compute ESG ratings on listed companies. Delegates will learn about the capabilities and limitations of ChatGPT and how it can be used to compute ESG ratings of companies that result in similar outcomes as expensive rating products. The course also examines how to use corporate sustainability data to calculate and compare ESG ratings and how to analyse ESG sentiment using big data sources.

Who should attend: fund managers, hedge funds, ESG and equity analysts, sustainability professionals.

Course Details
Course Outline


Large Language Models (LLMs) and Generative Pre-trained Transformers (GPT)

  • Introduction to LLMs, GPT-3, ChatGPT, and other large language models
  • Typical Applications of GPT-3 and other language models
  • How GPT-3 and other language models work
  • Utilizing ChatGPT on the web and through the API

Building Applications with GPT

  • Overview of Prompt engineering
  • Building applications such as text generation, summarization, etc.
  • Few-shot learning with GPT
  • Introduction to embeddings.
  • Overview of the OpenAI embeddings API and its usage

Risks Associated with LLMs

  • Understanding main risks with LLMs, such as, hallucinations, bias, consent and security
  • Methods for reducing the risks of hallucinations, such as, retrieval augmentation, prompt engineering, and self-reflection
  • Methods to detect and address hallucinations, including human feedback and model-based approaches

Individual Company ESG Ratings using Generative AI

  • Introduction to how ESG ratings are calculated
  • Main sources of divergence are in ESG ratings across different benchmarks provided agency
  • Selecting Relevant ESG Factors: Identifying key ESG factors for inclusion in the computation such as climate and DEI data
  • Company data collection and preprocessing using NLP from sustainability reports & news articles, etc. .
  • Derive ESG ratings for listed companies based on selected factors

Establishing Benchmarks for Validating ESG ratings

  • Establish the main sources of divergence in ESG ratings
  • Establish ESG’s industry benchmark
  • Analyzing and addressing any divergences or discrepancies in the results
  • Visualizing the scores across time and sectors
  • Benefits and potential of using generative AI for ESG ratings
  • Importance of continuous improvement, validation, and refinement in ESG analysis using generative AI
  • Guidelines for utilizing additional information sources, such as corporate filings, earnings calls, corporate announcements, and press releases, to enhance ESG computation

Deploying GPT and Other Language Models in Production

  • Best practices for deploying GPT in production
  • Overview of alternative generative models such as Cohere, LLaMA, Alpaca, etc.