FinTech and European Bank Performance: A Machine-Learning Analysis of Greenwashing and Sustainability Channels
Year: 2026
Type: article
Abstract: The parallel diffusion of financial technology and the tightening of European sustainable-finance regulation have reshaped banking, yet how digitalisation affects financial performance remains unclear. Using a panel of 90 European banks over 2014–2025 (1080 bank-year observations) and a Fintech index built from annual-report text through a Dynamic Factor Model, we examine whether, and through which channels, Fintech adoption is associated with bank performance, focusing on two mechanisms usually treated as controls: greenwashing and substantive ESG practices. We combine causal mediation, a sensitivity analysis to hidden confounding, a doubly robust Double Machine Learning estimator, unconditional-quantile (RIF) regressions, Local Projections, a Causal Forest, and optimal policy learning. Four findings emerge. First, Fintech’s direct effect raises long-term soundness while depressing short-term profitability, though its total effect on profitability is insignificant. Second, both channels mediate the relationship in the baseline specification, but only the substantive-ESG channel is robust to alternative measurement: the greenwashing channel, significant in the baseline, does not survive a performance-free re-measurement of the disclosure–performance gap, so we read these as associations rather than established causal mechanisms. Third, effects are heterogeneous across the performance distribution and over time, with profitability turning positive from the second year. Fourth, the benefits concentrate among large, well-capitalised, high-ESG, low-greenwashing banks, a profile that an exploratory policy rule can target. These associations are specific to Europe’s sustainable-finance architecture; because the design is not experimental, they are interpreted as conditional on the identifying assumptions, and their external validity remains open.
Source: FinTech
Authors: Mejda Tebessi, Heni Boubaker
Institution: University of Sousse
Language: English
Subfield: Management Information Systems
Domain: Social Sciences
Open Access status: gold
Estimated APC paid: $1,207
OpenAlex ID: W7236655259
