95% of researchers rate our articles as excellent or good
Learn more about the work of our research integrity team to safeguard the quality of each article we publish.
Find out more
BRIEF RESEARCH REPORT article
Front. Artif. Intell.
Sec. AI in Finance
Volume 8 - 2025 | doi: 10.3389/frai.2025.1578190
This article is part of the Research Topic Applications of AI and Machine Learning in Finance and Economics View all 8 articles
Explainable Machine Learning to predict the Cost of Capital
Provisionally accepted- 1 University of Pavia, Pavia, Italy
- 2 Department of Economics and Management, University of Pavia, Pavia, Italy
- 3 CAMRisk - Centre for the Analysis and Measurement of Global Risks, Pavia, Italy
- 4 Birkbeck, University of London, London, United Kingdom
This study investigates the impact of financial and non-financial factors on a firm's ex-ante cost of capital, which is the reflection of investors' perception on a firm's riskiness. Departing from previous literature, we apply the XGBoost algorithm and two explainable Artificial Intelligence methods, namely the Shapley value approach and Lorenz Model Selection to a sample of more than 1,400 listed companies worldwide. Results confirm the relevance of key financial indicators such as firm size, ROE, firm portfolio risk, but also individuate firm's non-financial features and country's institutional quality as relevant predictors for the cost of capital. These results suggest the importance of non-financial indicators and country institutional quality on the firm's ex-ante cost of equity that expresses investors' risk perception. Our findings pave the way for future investigations on the impact of ESG and country factors in predicting the cost of capital.
Keywords: Explainable AI, Shapley values, XGBoost models, Cost of capital, non financial disclosure
Received: 17 Feb 2025; Accepted: 17 Mar 2025.
Copyright: © 2025 Bußmann, Giudici, Tanda and Yu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence:
Alessandra Tanda, Department of Economics and Management, University of Pavia, Pavia, Italy
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.