FinTech Adoption and Bank Sustainability in Nigeria: Evidence from Disaggregated Channels and Mediating Effects of Operating Efficiency and Financial Inclusion

This study examines the effect of FinTech adoption on the sustainability of deposit money banks in Nigeria over the period 2010–2024. It focuses on three major FinTech channels which include mobile banking, point-of-sale (POS) transactions, and automated teller machines (ATMs, and evaluates how their effects on bank sustainability are mediated by operating efficiency and financial inclusion. Using secondary data from the Central Bank of Nigeria, the Nigeria Inter-Bank Settlement System, and bank annual reports, the study employs the Autoregressive Distributed Lag (ARDL) model to estimate short-run and long-run relationships, while Structural Equation Modelling (SEM) is used to assess mediation effects. The findings show that FinTech adoption has heterogeneous effects on bank sustainability. Mobile banking has a significant negative effect on return on equity, suggesting that adoption costs, service disruptions, and technology-related risks may weaken short-run performance. In contrast, POS and ATM channels exert positive effects on financial performance, indicating their contribution to revenue generation and service efficiency.

The mediation analysis further shows that operating efficiency is a strong positive transmission channel between FinTech adoption and bank sustainability, whereas financial inclusion has a weaker negative mediating effect, reflecting the short-run costs of expanding access to underserved populations. The ARDL bounds test confirms the existence of a long-run equilibrium relationship between FinTech adoption and bank sustainability in Nigeria. The study concludes that FinTech adoption can enhance bank sustainability when supported by efficient operations, sound risk management, and balanced regulation. It contributes to the literature by integrating disaggregated FinTech indicators with mediation analysis in the Nigerian banking context.