Digital synergies boost bank returns, but only for the digitally ready

One of the study’s most pressing takeaways is that standard regulatory or investment strategies are insufficient to foster equitable digital transformation across the EU’s banking landscape. As digital banking platforms become more dominant, the risk of digital exclusion and widening profitability gaps looms larger.

The banking sector across the European Union (EU) is undergoing a profound digital transformation, marked by the rapid adoption of mechatronic systems and artificial intelligence (AI). A newly published study titled “Decoding Digital Synergies: How Mechatronic Systems and Artificial Intelligence Shape Banking Performance Through Quantile-Driven Method of Moments” in Applied Sciences (2025) examines this convergence using a quantile-driven econometric model. The findings uncover that automation’s impact on profitability is not uniform, and it varies substantially depending on a bank’s performance level, operational scale, and regional regulatory context.

Using Method of Moments Quantile Regression (MMQR), the researchers assess how physical infrastructure (like ATMs) and software-based AI tools (like chatbots and robo-advisors) affect banking performance across different levels of return on equity (ROE). The study spans EU banks from 2017 to 2022 and highlights the need for tailored digital strategies and policy reforms to ensure that technological advances do not deepen the existing performance divide.

How do mechatronic systems and AI technologies influence bank profitability?

The study differentiates between two key classes of banking automation technologies: hardware-centric mechatronic systems, exemplified by ATMs, and software-driven tools powered by artificial intelligence, such as chatbots and robo-advisors. Each plays a distinct role in operational efficiency and customer service delivery.

The researchers find that mechatronic systems yield the greatest efficiency gains in lower-performing banks. Institutions in the lower ROE quantiles benefit from automation through basic cost reduction and operational streamlining. ATMs, for example, are particularly impactful in enabling these banks to deliver 24/7 service, reduce teller labor costs, and widen financial access in underserved regions. The MMQR model shows a statistically significant and positive correlation between ATM density and ROE in the 0.25 and 0.50 quantiles.

Conversely, higher-performing banks, those in the upper ROE quantiles, gain more from AI-driven solutions. These institutions leverage digital interfaces to optimize customer engagement, tailor financial recommendations, and drive strategic decision-making through analytics. The model demonstrates that variables representing digital inclusion (like e-banking adoption) and AI intensity (measured through DESI indicators and R&D spending) have a stronger impact on profitability in the 0.75 and 0.90 quantiles.

Why is a Quantile-based analysis crucial for understanding banking automation?

Traditional econometric models typically focus on average effects, which may obscure divergent impacts across banks with different financial profiles. This study departs from that approach by applying the Method of Moments Quantile Regression (MMQR), allowing researchers to estimate the effects of automation at different points in the ROE distribution.

The MMQR model incorporates higher-order moments, MOMENT1 and MOMENT2, to account for linear and nonlinear variations in profitability across quantiles. This enables a nuanced analysis of how digitalization influences bank performance, taking into account heterogeneity in bank size, regulatory exposure, and technology adoption levels.

For instance, the study finds that economic growth and human capital investment are consistently positive drivers across the spectrum, but the magnitude and significance vary. In higher quantiles, innovation yields greater returns due to scale advantages and better absorption capacity. In lower quantiles, however, regulatory fragmentation and weak infrastructure diminish the returns from digital investment.

Moreover, the study validates these insights through robustness tests, including cross-sectional dependency checks, unit root assessments, and cointegration analysis, reinforcing the reliability of the quantile-based findings. This approach reveals that automation and digital transformation in banking should not be treated as a one-size-fits-all intervention.

What are the broader implications for policymakers and banking leaders?

One of the study’s most pressing takeaways is that standard regulatory or investment strategies are insufficient to foster equitable digital transformation across the EU’s banking landscape. As digital banking platforms become more dominant, the risk of digital exclusion and widening profitability gaps looms larger.

The authors call for customized policy interventions that reflect the institutional capabilities and digital readiness of each member state. Specifically, banks in lagging regions would benefit from increased investment in broadband infrastructure, financial technology incubation, and workforce upskilling programs. Simultaneously, high-performing banks must manage the risks associated with excessive digital scaling, such as cybersecurity threats and regulatory compliance gaps.

Additionally, the study emphasizes the need for regulatory harmonization. Fragmented oversight structures across the EU have led to inconsistencies in technology adoption, compliance costs, and innovation speed. A unified regulatory framework, focused on transparency, ethical AI use, and inclusive access, would support balanced growth and prevent monopolistic advantages for banks with early digital dominance.

The authors urge continued research on the causal mechanisms between ROE and automation, supported by advanced modeling techniques and longitudinal datasets. As financial technologies continue to evolve, the MMQR model offers a blueprint for evaluating digital transformation outcomes in a dynamic, performance-sensitive manner.