Digital Reach, Delayed Returns: What Fintech Adoption Means for Asian Banks

Fintech is delivering one of banking's most important promises across Asia: wider access to deposits and credit. Still, the same digital transformation is not producing an equally clear payoff for the banks financing it.

In "Fintech Adoption, Financial Inclusion, and Bank Performance: Evidence from the Asian Banking Industry," published in the International Journal of Financial Studies, Helal Uddin and Munim Kumar Barai analyse 173 banks in 14 economies between 2016 and 2024. Their findings reveal a fintech paradox: stronger financial inclusion across every market group, but uneven and sometimes negative effects on profitability.

The research suggests that fintech may first expand the reach and functioning of the financial system before strengthening the balance sheets of the institutions funding the transition.

A continent-wide test of digital banking

Much of the existing evidence on fintech has come from single-country studies or research focused separately on financial inclusion and bank performance. Uddin and Barai bring the two questions together: does fintech help banks mobilise more deposits and provide more loans, and does that expansion make them more profitable?

Their sample spans four groups. Japan, South Korea, Singapore and Hong Kong represent advanced economies. China, India, Indonesia and Malaysia are classified as large economies. The Philippines, Thailand and Vietnam form the emerging-economy group, while Bangladesh, Pakistan and Sri Lanka are treated as frontier economies.

The study uses 1,557 bank-year observations drawn from Bloomberg. Financial inclusion is measured through total deposits and total loans, expressed in logarithmic form, while performance is assessed through return on assets and return on equity.

Rather than relying on national indicators of digital development, the authors construct a bank-level fintech adoption index. It captures eight publicly observable capabilities: mobile banking, online banking, remote account opening, digital loan applications, digital wallets or contactless payments, real-time payments, AI-enabled customer assistance and personalised online financial analytics.

Banks receive scores for no, partial or full adoption. This approach recognises that digital transformation is rarely binary. A bank may offer a mobile application, for example, while still requiring customers to visit a branch for verification or service activation.

The authors employ fixed-effects models with Driscoll-Kraay standard errors, designed to account for common regional shocks, heteroskedasticity and serial correlation. They also use an instrumental-variable analysis based on lagged fintech adoption to test whether stronger banks simply adopt more technology, rather than technology producing stronger outcomes.

This methodological breadth is important because Asian banking systems do not operate in isolation. Interest-rate shifts, economic downturns, pandemics and financial-market disruptions can affect banks across several countries at once. A model that ignores those shared pressures could attribute too much to fintech itself.

Access is rising faster than profits

The clearest result is that fintech adoption is positively and significantly associated with both deposits and loans in every country group. For deposits, the largest estimated fintech coefficient appears in emerging economies, followed by large, advanced and frontier economies. For lending, the strongest coefficient is found in advanced economies, followed by emerging, large and frontier markets.

The geographic pattern is revealing. In emerging economies, digital channels may be especially effective at mobilising deposits because they reduce the need for extensive branch networks and make formal banking more convenient. Customers can open or operate accounts remotely, transfer funds digitally and engage with banks at lower cost.

Advanced economies, by contrast, may gain more through the lending channel. Banks in highly digital markets can use established data systems, automated processing and online application infrastructure to improve credit delivery. Their advantage may not be bringing basic banking to first-time users, but making mature financial intermediation faster and more efficient.

Frontier economies record smaller coefficients than the other groups, but the effects remain statistically significant. That matters because these are precisely the markets where physical access, infrastructure constraints and high service costs can exclude customers. Even incomplete digital adoption can therefore expand formal financial activity.

The profitability picture is far less straightforward. In advanced economies, fintech adoption is negatively and significantly associated with both return on assets and return on equity. In large, emerging and frontier economies, the estimated relationships are mostly positive but not statistically significant in the main subgroup models.

The study interprets the advanced-economy result as evidence that technology expenditure and operating costs may initially outweigh financial gains. Mature markets may also expose banks to stronger digital competition, compressing margins even as services improve.

The robustness analysis adds another layer. In the full-sample instrumental-variable model, fintech remains strongly associated with deposits and loans, while its relationship with return on equity becomes positive and statistically significant. Return on assets remains statistically insignificant. The broader conclusion, therefore, is not that fintech necessarily damages profitability, but that the outcome depends on timing, market maturity and model specification.

The crucial insight is that financial reach and financial return move on different timelines. A bank can attract more deposits, process more loans and expand customer participation without immediately generating higher profits.

The inclusion dividend comes with blind spots

The study makes a valuable contribution by showing that fintech's most consistent effect lies in expanding banking activity rather than guaranteeing commercial returns. Yet its measures also leave important questions unanswered.

Deposits and loans indicate active use of banking services, making them more informative than account ownership alone. But they do not reveal who benefits. Rising deposit volumes could reflect larger balances held by existing affluent customers rather than the entry of previously unbanked households. Loan growth could be driven by corporations and established borrowers rather than microenterprises, rural communities or low-income consumers.

The research thus demonstrates financial deepening more clearly than distributional inclusion. The sample is also restricted to listed banks. Such institutions tend to be larger, better capitalised and more capable of financing digital systems than rural banks, cooperatives, microfinance institutions or small community lenders. The findings may not fully reflect the experience of institutions serving harder-to-reach populations.

The fintech index is transparent and practical, but it assigns equal weight to all eight components. In development terms, those capabilities may not be equally important. Online account opening and instant payments could materially reduce access barriers, while chatbots or spending analytics may mainly improve convenience for existing customers.

The study also cannot independently verify whether digital services are affordable, reliable or widely used. A bank may technically offer online lending, but high data costs, weak connectivity, difficult interfaces or complex documentation could still restrict access.

Finally, the authors acknowledge that they do not directly test why profitability declines in advanced economies. They suggest high investment costs and competition as possible explanations, but bank-level technology spending and revenue data are not included. The Bloomberg data are also unavailable for public replication because of licensing restrictions.

These limitations do not undermine the central findings, but they reinforce the need to distinguish digital availability from meaningful inclusion.

Policy must reward reach without ignoring risk

For governments and regulators, the study offers a strong case for supporting fintech adoption while resisting simplistic digitalisation targets. In emerging and frontier economies, policy should focus on infrastructure that lowers the cost of participation: interoperable payment systems, reliable digital identification, proportionate know-your-customer requirements, affordable connectivity and remote account-opening frameworks. Shared infrastructure may be especially important where individual banks cannot justify large investments in low-income or sparsely populated markets.

Regulators should also measure outcomes, not merely the number of digital products launched. Useful indicators would include active account usage, transaction costs, loan approval patterns, rural participation, gender gaps, service reliability and customer complaints.

The profitability results demand equal attention. If digital investment weakens margins in the short term, banks may reduce services, charge higher fees, consolidate operations or take additional risks to protect returns. Supervisors should therefore monitor whether cost pressures are affecting lending standards, cybersecurity spending, consumer protection or financial stability.

Bank managers, meanwhile, may need to judge fintech through a wider performance framework. Return on assets and return on equity remain important, but they may not capture early-stage gains in customer reach, deposit growth, processing speed, service quality and operating resilience. Digital investments should be evaluated over longer periods and linked to clear institutional objectives rather than pursued as a response to market fashion.

The research also carries implications for development agencies and investors. Fintech projects should not be assessed only by the number of accounts or platforms created. Their value depends on whether people can use them safely, affordably and repeatedly, and whether the institutions providing them can sustain the service.

The broader lesson is that fintech is not automatically profitable, inclusive or transformative. Its most credible contribution emerges when technology is embedded in supportive infrastructure, effective regulation and viable banking models.