SINGAPORE, 17 AUG 2026 — MariBank, the Singapore digital bank owned by Sea Limited, intends to build a regional digital banking group run from Singapore, starting with the Philippines. It lost S$55.6 million in 2025, and it has about three years left of the window Singapore's digital banks were given to show a path to profit.
Exporting a model before it has proven itself at home is a defensible strategy. It is also the point that deserves the most scrutiny: the model has not yet worked in the market it was built for.
The position
Chief executive Natalia Goh has described the plan as building a regional digital banking group headquartered in Singapore, with the city-state as the hub for innovation, talent and strategy, and the Philippines as the first market.
The route into the Philippines was an acquisition rather than an application. MariBank acquired SeaBank Philippines, a rural bank, in a transaction effective April 2025. The entity was upgraded when Bangko Sentral ng Pilipinas issued its certificate of authority on 8 July this year; it began operating under the digital banking licence on 18 July.
What the losses mean, and what they do not
A loss of S$55.6 million against S$51.3 million the year before is a bank still buying customers, and that is what a digital bank is supposed to be doing at this stage.
Context matters more than the absolute figure. GXS Bank, the Grab and Singtel venture, lost S$208 million in 2025 — roughly four times as much. Trust Bank, backed by Standard Chartered and FairPrice, recorded its first profitable month in March and is the only Singapore digital bank to have got there.
So MariBank is neither the outlier nor the leader. It is in the middle of a cohort that has collectively demonstrated how expensive it is to acquire deposit customers in a market where three incumbent banks are competent, trusted and already hold the accounts.
The five-year framing is the constraint that makes the expansion timing legible. Singapore's digital banks were licensed on the understanding that they would show a viable path to profitability within five years of launch. MariBank began operating in 2023, which leaves about three. Regional expansion is one of the few things that can change the shape of that arithmetic inside the time available, because it changes the size of the addressable market rather than the cost of serving the current one.
The Philippines is a different problem, not a bigger one
The strategy Goh describes is not to ship the Singapore product across. It is to take the technology, the product knowledge and the operating model and localise them — lower ticket sizes, altered features, and, notably, cash.
That last item is the one to watch. MariBank is piloting cash-in and cash-out partnerships with retail outlets in the Philippines, which is a category of work it does not have to do in Singapore at all. Goh's own framing is that the Philippines is moving toward cashlessness but still has real need for cash.
This is the honest difficulty in every ASEAN regional banking plan. The markets share a region and very little else. A digital bank built for a population with universal bank accounts, high smartphone penetration, instant domestic transfers and near-total card acceptance is not obviously transferable to a market where a large share of transactions begin and end in notes. Solving that is a physical distribution problem — agents, retail partners, reconciliation — and it is a different business from the one Singapore taught them.
What travels is the technology and the cost base. What does not travel is the assumption about how money moves.
Why Singapore as the hub is the interesting choice
The logic for keeping the group headquarters in Singapore, while the growth is elsewhere, is clear.
Singapore supplies regulatory credibility, access to talent and a Monetary Authority licence that carries weight with counterparties across the region. It is an expensive place to run operations and a cheap place to be trusted from. Several regional financial groups already use this split structure: strategy and product in Singapore, customer growth in markets with more headroom.
It also means the Singapore business does not have to become profitable on its own for the group to work, which changes what the five-year window is actually measuring. A regulator assessing a path to profitability for a Singapore-licensed bank that is deliberately running its growth elsewhere is assessing something more complicated than a single entity's accounts.
The comparison worth making
MariBank is not alone in this move. GXS is expanding into Malaysia and the Philippines on similar reasoning, and we reported earlier this year on Grab taking majority control of Superbank in Indonesia.
Singapore's digital banks seem to have concluded, at roughly the same time, that the domestic market cannot carry them to profitability. That is a meaningful collective verdict on a licensing experiment that is only a few years old.
The Philippines is receiving several of them at once. Tonik reached profitability as a standalone digital bank there, which proves the market can support one, and the central bank's digital payments push is expanding the addressable base. Arriving with an acquired rural bank and a parent company's balance sheet is a stronger entry than most, but the market already has competition from those who got there first.
What we could not establish
MariBank's deposit base, customer numbers and loan book in Singapore, none of which appear in the reporting we could read. Without those, the loss figure cannot be assessed against the size of the business it is buying, which is the only way to tell an expensive customer acquisition from a failing one.
Several other points are unestablished: SeaBank Philippines' customer numbers and deposits; what the upgraded licence permits that the rural bank licence did not; how much capital the Philippine entity has; whether other markets have been identified; what the group structure means for capital treatment; and how MAS will assess the five-year profitability path for a bank whose growth is deliberately offshore.
What to watch
The next set of accounts is the first checkpoint. A loss that widens again while the Philippine business is still being built would be expected; one that widens without deposit or customer growth to show for it would not.
The next question is whether the cash-in and cash-out pilots become a permanent network. Retail agent networks are expensive, operationally heavy and the single largest determinant of whether a digital bank works in a partly cash economy. Scaling that is the real test of whether the Singapore model can be localised or merely translated.
Finally, watch what MAS says as the five-year marks approach for this cohort. Three banks with very different results and one shared deadline will force a conversation about what the licences were for, and that conversation will shape whether any regulator in this region issues another round.
