Combining AI and Domain Expertise to Assess Corporate Climate Transition Disclosures

43 Pages Posted: 14 May 2024

See all articles by Chiara Colesanti Senni

Chiara Colesanti Senni

University of Zurich - Department of Finance

Tobias Schimanski

University of Zurich

Julia Bingler

University of Oxford

Jingwei Ni

ETH Zurich

Markus Leippold

University of Zurich; Swiss Finance Institute

Date Written: May 13, 2024

Abstract

Companies need sound planning to reduce their emissions and deal with the transition to a more sustainable economy. The disclosure of such plans is key for effective capital allocation and risk management. Transition and sustainability disclosures are a compass for market participants to guide their actions and strategies toward the net-zero target. If companies plan their transition appropriately, the negative implications of physical and transition risks for micro- and macro-financial stability can be reduced. Many frameworks have been suggested to assess transition plans’ ambition, credibility, and feasibility. However, the lack of one clear reference framework paves the way for inconsistencies in transition plans and the risk of greenwashing. We propose a set of 64 common ground indicators from 28 different transition plan disclosure frameworks to comprehensively assess transition plans and develop a novel natural language processing (NLP)–based tool to automate the assessment of companies’ disclosures. This can help investors and financial supervisors assess transition risks while supporting companies’ disclosure efforts. Applying the tool to 143 reports from the carbon-intensive CA100+ companies, we find that companies tend to disclose more indicators related to target setting (talk) but fewer indicators related to the concrete implementation of strategies (walk). Our results demonstrate that machine learning can be used to generate a positive impact on the transition towards a more sustainable economy by identifying the elements of transition plans that require further scrutiny and/or effort. Our work will be a starting point for further leveraging new technologies in sustainable finance. For example, the assessment of the plans could be used by financial regulators in their supervisory practices or to investigate whether the risk of greenwashing is reflected in stock returns.

Keywords: Climate disclosure, RAG system, transition strategies, human evaluation, CA100+

Suggested Citation

Colesanti Senni, Chiara and Schimanski, Tobias and Bingler, Julia and Ni, Jingwei and Leippold, Markus, Combining AI and Domain Expertise to Assess Corporate Climate Transition Disclosures (May 13, 2024). Available at SSRN: https://ssrn.com/abstract=4826207 or http://dx.doi.org/10.2139/ssrn.4826207

Chiara Colesanti Senni

University of Zurich - Department of Finance ( email )

Plattenstr 32
Zurich, 8032
Switzerland

Tobias Schimanski (Contact Author)

University of Zurich ( email )

Schönberggasse 1
Zürich, 8001
Switzerland

Julia Bingler

University of Oxford ( email )

Mansfield Road
Oxford, Oxfordshire OX1 4AU
United Kingdom

Jingwei Ni

ETH Zurich ( email )

Markus Leippold

University of Zurich ( email )

Rämistrasse 71
Zürich, CH-8006
Switzerland

Swiss Finance Institute ( email )

c/o University of Geneva
40, Bd du Pont-d'Arve
CH-1211 Geneva 4
Switzerland

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
483
Abstract Views
1,115
Rank
111,920
PlumX Metrics