Environmental-Anomaly Index for Early Detection of Green-Money-Laundering Risks in Climate-Finance Portfolios
American Journal of Machine Learning · 2026
DOI: 10.71465/ajml3538 ↗Kurz gefasst
Fachartikel in der Zeitschrift American Journal of Machine Learning, 2026. Verfasst von Mats Johansen und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
This study proposes an Environmental-Anomaly Index (EAI) to identify inconsistencies between environmental performance and financial flows in climate-finance projects. The index is computed from three components: satellite-observed environmental deviations, inconsistencies in ESG reporting, and discrepancies between risk ratings and capital-allocation patterns. A dataset of 2,036 green-finance projects from 14 institutions between 2016–2022 was used for evaluation. Incorporating EAI features increased model AUC from 0.79 to 0.88 and precision at 60% recall from 0.43 to 0.64. Projects in the t op-risk decile contained 71.3% of known problematic cases. Back-testing showed that applying EAI-based screening could have reduced misallocated capital by an estimated 28.1%. The results demonstrate the potential of environmental anomaly monitoring for early detection of green-money-laundering risks.
Zitieren
APA
Johansen, M., Kristoffersen, H., & Solberg, I. (2026). Environmental-Anomaly Index for Early Detection of Green-Money-Laundering Risks in Climate-Finance Portfolios. American Journal of Machine Learning. https://doi.org/10.71465/ajml3538
Vancouver
Johansen M, Kristoffersen H, Solberg I. Environmental-Anomaly Index for Early Detection of Green-Money-Laundering Risks in Climate-Finance Portfolios. American Journal of Machine Learning. 2026. doi:10.71465/ajml3538
BibTeX
@article{Johansen2026,
author = {Johansen, Mats and Kristoffersen, Hanne and Solberg, Ingrid},
title = {Environmental-Anomaly Index for Early Detection of Green-Money-Laundering Risks in Climate-Finance Portfolios},
journal = {American Journal of Machine Learning},
year = {2026},
doi = {10.71465/ajml3538}
}
EUVIMED