Trust Bank’s announcement that it recorded its first profitable month in March 2026, just over three years after launch, deserves attention. Not because a profitable month in itself proves that digital banking has become economically sustainable, but because it offers a useful point from which to ask a harder business question, what actually makes these economics work?

Digital banking globally has shown that customer acquisition and innovation are easier than building durable economics. Institutions have demonstrated that they can launch digital propositions, attract users and build services that customers are willing to adopt. Far fewer have shown that these models can translate into sustainable profitability.

Even markets that appeared structurally attractive have struggled. Hong Kong’s eight virtual banks, now more commonly described as digital banks, launched with considerable ambition in a market with proximity to the wider Greater Bay Area. Most have yet to demonstrate broad profitability after several years of operation. Malaysia’s five digital banks are still at an earlier stage of proving their business models. Australia’s neo-bank experiment had been more sobering, with several entrants either exited, acquired or restructured after struggling to establish viable standalone retail economics.

Against that backdrop, Singapore becomes a particularly useful test case. It is highly banked, digitally advanced and intensely competitive. The incumbent institutions are not weak legacy banks vulnerable to digital disruption. They are already technologically mature, operationally sophisticated and deeply embedded in customers’ financial lives. Any new entrant therefore has to do more than create a digital proposition. It has to become relevant in customers’ everyday financial behaviour and build economics that can endure.

Dwaipayan Sadhu, CEO of Trust Bank, described the bank’s March profitability milestone not as a sudden breakthrough, but as the outcome of disciplined execution over time. “It has been a story of very steady progress,” he said. He rejected the idea that profitability emerged because of a single initiative or product. “It’s not one single thing,” he said. The drivers, he explained, included stronger customer engagement, product diversification, increasing operational efficiency and notably, strong credit performance.

The disclosed figures support that view. In March 2026, Trust reported that risk-adjusted revenue, defined as revenue net of credit losses, rose 67% year on year while monthly costs fell 9%, resulting in positive monthly profitability. In financial year 2025, the bank reported revenue of SGD 135 million ($101 million) against costs of SGD 146 million ($109 million), narrowing its pre-tax loss to SGD 54 million ($40 million) from SGD 93 million ($69 million) a year earlier. The more useful question is what changed underneath.

Building economic value through behavioural depth

Sadhu’s explanation began with customer behaviour. By the end of 2025, Trust had accumulated SGD 4.0 billion ($3.0 billion) in deposits, SGD 1.2 billion ($900 million) in loans and advances, and had disbursed more than SGD 900 million ($675 million) in loans during the year. Around 170,000 customers were using Trust as a primary banking relationship, while approximately one-third of deposit balances came from salary-crediting customers.

For Sadhu, however, scale was only part of the story. “Long-term value comes from regular usage,” he said. Trust’s credit card users transacted around 25 times a month, which Sadhu described as evidence that the bank had become part of customers’ daily spending behaviour rather than functioning as a secondary wallet or occasional spending tool.

He pointed to other indicators of engagement quality. Card activation rates stood at 85%, which he said was among the highest in the market. Trust had reached 7% market share in unsecured lending and cards, while accounting for 22% of Visa foreign exchange spending. Around 70% of new customers came through referrals from existing users, reinforcing his argument that the proposition was generating advocacy rather than simply promotional acquisition.

That matters because one of digital banking’s recurring economic challenges has been acquisition cost. Models dependent on constant incentive-led customer acquisition can become structurally expensive. Referral-led growth suggests a proposition that customers find sufficiently useful to recommend.

Sadhu also pointed to product design choices intended to drive habitual use, zero foreign exchange fees for travel spending, digital supplementary cards, card lock and unlock functionality, savings pots and fractional investing capabilities. These are not individually transformative features. Collectively, they are designed to create practical, repeat reasons for customers to return.

Diversification with institutional credit discipline

Sadhu repeatedly returned to one distinction between Trust and some digital banking peers, product breadth.

Trust did not launch as a single-product institution. It began with a dual-function debit and credit card and personal family accident insurance, before expanding into unsecured lending, investments and a broader range of insurance offerings. Sadhu argued that this was economically important because narrow digital models are structurally exposed.

“A bank which only has deposits will be really exposed to a dropping interest rate environment,” he said. “The fact that we have a strong credit card book, the fact that we have a strong lending book and increasingly we have a growing investment proposition is helping us to diversify our income stream.”

Trust’s broader product model now includes 50,000 TrustInvest accounts and more than 75,000 insurance policies sold. But the more important point is that profitability was not framed as dependence on a single revenue engine.

Credit performance was clearly part of the March outcome. Sadhu explicitly cited strong credit performance as one contributor to profitability. But he was equally clear that Trust’s lending model was not built from an immature starting point.

This is where Standard Chartered’s influence becomes strategically important.

Trust’s lending discipline, underwriting standards, approval frameworks and lifecycle risk management were built on established institutional banking practices rather than being developed from scratch. That distinction matters when evaluating digital banking economics because rapid lending growth can improve top-line economics while creating future provisioning pressure.

MariBank’s trajectory illustrates the point. Public disclosures indicate stronger income growth, but also SGD 133.4 million ($100 million) in credit allowances, suggesting heavier credit drag from a more lending-intensive model.

Sadhu’s point was not that lending is unimportant. It was that profitable lending requires discipline.

AI as operating leverage with governance

The most striking operational feature of Trust’s recent performance was cost discipline.

In a conventional banking model, costs typically rise as customer numbers, transactions and product complexity increase. Trust reported the opposite. Revenue grew 39% in 2025 while costs fell 7%. March’s 9% cost reduction extended that pattern.
Sadhu attributed this to automation and artificial intelligence (AI), but not in the simplistic sense of labour substitution.

“Digital-only where possible,” he said, describing the bank’s operating philosophy. Combined with greenfield infrastructure, this meant Trust avoided legacy technology constraints that often slow incumbent institutions.

Sadhu described Trust’s operating mindset as “AI-first”. The bank’s generative artificial intelligence (Gen AI)-enabled customer service chatbot now handles nearly 50% of customer chats end to end. He contrasted this with conventional scripted chatbot models, arguing that the Trust model allows natural conversation and smart escalation to human agents where necessary.

But he also emphasised governance discipline. Trust’s customer-facing AI operates within a controlled knowledge environment, with clear escalation and oversight mechanisms rather than unrestricted external access. That reflects institutional banking discipline as much as technological experimentation.

The complaint-management use case is perhaps more instructive. Trust uses AI-driven sentiment analysis to identify deteriorating customer interactions before they become formal complaints. Sadhu said this reduced complaints by 40% within three months.

He was careful not to reduce AI to efficiency alone. “We are still scratching the surface,” he said. His longer-term view was that AI would increasingly support hyper-personalised engagement, financial insights and advisory interaction.

Is Trust proving digital banking economics, or something more specific?

The harder analytical question is whether Trust is proving digital banking economics generally or demonstrating something more specific. Singapore’s digital banking market is already evolving into distinct economic models.

GXS has built meaningful scale, with public disclosures showing SGD 3.4 billion ($2.55 billion) in assets, SGD 2.3 billion ($1.73 billion) in deposits and SGD 1.03 billion ($773 million) in loans, alongside sharply improving operating metrics, but it remains loss-making.

MariBank’s economics look different again, with stronger income momentum accompanied by materially heavier credit provisioning.

ANEXT Bank and Green Link Digital Bank are less directly comparable because they operate SME and wholesale models rather than mass retail consumer propositions, though they remain part of the broader Singapore digital banking experiment.

Trust’s structural advantages therefore matter. Sadhu was explicit about the role of the FairPrice and NTUC ecosystem. The advantage was not simply distribution reach, but practical frequency and everyday relevance. Customers already had reasons to engage regularly.

But ecosystem access alone does not explain Trust’s trajectory. The Standard Chartered relationship matters just as much. “Running a bank is very different from running a fintech,” Sadhu said. That observation goes beyond governance rhetoric. It points to lending discipline, risk management, regulatory execution and institutional operating maturity.

That sharpens the central question. Trust may be showing that digital banking economics can be made to work. But it may also be showing that these economics become materially stronger when combined with incumbent-grade institutional discipline.

The next test is broader relevance

A profitable month is an important milestone. It is not yet proof of structural sustainability. Trust’s economics clearly look stronger than they did two years ago. Customer engagement is deeper. Product participation is broader. AI appears to be delivering measurable operating leverage. Credit performance has held up.

But important questions remain unresolved. Monthly profitability is not sustained profitability. Exact revenue composition remains undisclosed. It remains unclear how much of the model’s economics depend on macro conditions, ecosystem advantages or sponsor support.

For Sadhu, the next phase is less about celebrating a profitability milestone than becoming more relevant across a broader set of customer financial needs.

That includes digital wealth, particularly for emerging affluent customers where traditional wealth propositions have often focused disproportionately on affluent and private banking clients, leaving a meaningful segment under-served. It may also include hybrid models where digital engagement is complemented by human advisory support, particularly in areas such as insurance where complexity still matters.

The broader lesson may therefore be more nuanced than a simple digital banking success story. If digital banking economics in mature markets are becoming sustainable, the evidence increasingly suggests that digital capability alone is not enough. Product breadth, behavioural engagement, credit discipline, operational execution, ecosystem relevance and institutional governance may matter just as much.