Ujjwal Sharma (ABS, UvA) - "Greenwashing detection using multimodal deep learning"

Modern ESG corporate reporting often falls short in capturing corporate alignment with established sustainability metrics. Furthermore, traditional ESG rating systems struggle to process and distill the wealth of information reported by companies, resulting in an increased reliance on low-quality metrics that can potentially mislead investors—commonly known as greenwashing. This project aims to quantify the gap between predicted and official rankings as an indicator of greenwashing. In doing so, it seeks to develop detection models using AI techniques to automatically identify instances of greenwashing. To achieve this, we will leverage recent advancements in text and vision transformers to create models that accurately assess sustainability-related messaging based on a corporation's category. This approach enhances the explainability and interpretability of existing models, allowing us to map textual and visual information sources to sustainability axes. Ultimately, the project will yield insights that simplify the rating process by introducing a 'greenwashing score' to evaluate companies based on their commitment to sustainability.

Ramon de Punder (ASE, UvA) - "Quantifying the impact of ESG risk on insurance and financial portfolios"

Upcoming legislation highlights the need to quantify the effect of non-financial risk on insurance and financial portfolios. The scope is not restricted to climate risk but extends to the broader category of ESG risk, crucially including biodiversity risk. These developments are making it crucially important for the insurance and pensions industry to be able to stress test their insurance and financial portfolios for ESG risk. This project aims to quantify the impact of ESG risk on insurance and financial portfolios by developing Integrated Assessment Models (IAMs) that can serve as a basis for nature-related ESG risk stress testing in the insurance industry and by analysing the policy implications of nature-related ESG risk for risk management, portfolio choice, pricing, and product development to assess which products an insurer must offer towards the future as part of the climate transition.

Information and registration:

When? Monday, March 18, 2024 between 13:00-14:00 (walk-in with lunch from 12:30 on; coffee and cake afterwards between 14:00-15:00)

Where? REC E0.03

Please contact us at asf-feb@uva.nl if you want to join this workshop.