What is the difference between a digital bank and a traditional bank?
A digital bank delivers banking primarily through an app or website, with little or no branch network. A traditional bank combines established banking operations with branches and digital channels. Both can accept deposits and make loans within their licences. The difference is mainly distribution, operating structure and customer focus; an attractive app alone does not make a company a bank.
Singapore also distinguishes licence categories. GXS and MariBank are digital full banks; Trust is a full bank operating with a digital model. ANEXT and Green Link Digital Bank serve the wholesale banking market. Their permitted activities and target customers are not interchangeable. Established banks such as DBS, OCBC and UOB also compete through digital services.
How do their business and revenue models work?
The core banking model combines funding, lending, investments and services. Net interest income is interest earned on loans and other interest-bearing assets minus interest paid on deposits and other funding. Fees add revenue from activities such as payments, foreign exchange, wealth distribution and lending services. Product mix and licence scope determine which sources matter.
| Dimension | Digital bank | Traditional bank |
|---|---|---|
| Customer acquisition | App onboarding, referrals and ecosystem partners; promotions can be expensive. | Branches, relationship managers, existing customers and digital channels. |
| Revenue opportunity | Interest margins, card income and fees; newer banks often start with fewer products. | The same core sources, often across broader retail, business and wealth franchises. |
| Cost structure | Fewer premises; substantial technology, compliance, security, support and incentive costs remain. | Branch, staffing and legacy-system costs alongside digital investment. |
| Cost structure at scale | Automation can lower service cost per active customer as usage grows. | Larger existing revenue can absorb fixed costs, but branches and legacy systems require investment. |
Revenue is the money earned; profit is what remains
A bank earns interest from loans and investments, and collects fees for some services. It must then pay interest to savers, salaries, technology bills, customer support, compliance costs and tax. It also records an allowance for loans that customers may not repay. Only what is left is profit. More accounts or deposits do not automatically mean more profit.
For a simplified, made-up example, suppose a bank earns S$5 a year from S$100 of interest-earning assets and pays S$3 for the funding. The S$2 difference must still cover its other costs and possible loan losses. If funding costs rise to S$4, only S$1 remains before those costs. An efficient app helps, but cannot fix a business that repeatedly pays out more than it earns.
Shareholders must also provide capital: their own money that can absorb losses. Regulators require banks to maintain this cushion. It is not simply cash locked in a drawer, but it limits how far a bank can expand and how much it can return to owners. Profitability, available cash and capital strength are related, but they are different tests.
Singapore market size: can five digital banks build sustainable businesses?
Market capitalisation means the stock market’s value of a listed company. It is a poor tool for this local comparison: DBS, OCBC and UOB are listed groups with overseas businesses, while GXS, MariBank and Trust have no separately traded public shares. Their parents’ valuations cannot stand in for the banks. To understand competitive scale in Singapore, compare deposits, lending and the available customer base instead.
AMRO’s Singapore 2025 report puts the five digital banks’ combined end-2024 share at 0.45% of commercial-bank deposits, 0.19% of loans and 0.30% of assets. These are historical system-wide shares, not September 2026 figures or shares of the domestic SME market. Singapore’s international banking activity makes the denominator much broader than local shops and households.
| Measure | Five digital banks combined | Plain-language meaning |
|---|---|---|
| Deposits | 0.45% | About 45 cents in every S$100 deposited across the system. |
| Loans | 0.19% | About 19 cents in every S$100 lent. |
| Assets | 0.30% | About 30 cents in every S$100 of banking assets. |
The sector is still young in balance-sheet scale, even where an individual bank has attracted many users. The five are GXS, MariBank, Trust, ANEXT and Green Link Digital Bank. They compete with one another in overlapping segments, as well as with DBS, OCBC, UOB and foreign banks. They do not all have the same licence or chase the same customers: the two wholesale banks focus on business banking.
The local business pool is finite. SingStat counted 358,300 enterprises in 2024, excluding the public sector; 77% had annual revenue of S$1 million or less. Its infographic was updated in March 2026. This count covers enterprises with revenue or employment, not every registered entity, and it is not a count of businesses seeking a new bank or loan.
EISOL’s assessment: this creates a credible risk of too many similar offers chasing too few profitable relationships. A small merchant may open several free accounts but keep little money in each and never borrow. Each provider still pays to onboard, monitor and support that merchant. If enough customers behave this way, sign-up growth will not cover the bills. That makes a sustainable profit-and-loss result harder to reach.
Five players alone do not prove oversupply. Specialised lending, underserved customers and cross-border business can create room to grow. The test is whether each bank earns enough from retained customers after funding costs, losses and operating expenses—not whether it can keep announcing new accounts.
The customer bargain: higher savings interest and lower business fees
Savers can benefit from attractive deposit rates, while businesses save on account and transfer fees. Those benefits are real. For a bank using both offers to win customers, however, the economics are squeezed from two sides: higher deposit interest increases its costs, and lower business fees reduce one source of income. Neither is, by itself, a new source of revenue.
For example, MariBank’s business-account page advertises no account or fall-below fees and free local transfers; its overseas transfer-fee promotion is stated to run until 31 December 2026. These terms were checked on 16 September 2026. Free services can be sustainable if lending, foreign exchange or other products pay enough to support them. A headline fee waiver alone does not reveal the profitability of the full customer relationship.
The danger is a prolonged price war. Rivals may match rates, waive more fees and increase rewards to prevent customers leaving. If customers mainly collect incentives and move on, the expected future earnings never arrive. The bank may then cut benefits, charge fees, reduce costs or raise fresh shareholder money. If losses keep eroding capital and further funding is unavailable, a sale, merger or regulated wind-down becomes possible. This is a risk scenario, not a prediction that any named bank is running out of cash.
0% MDR: a useful offer, with costs somewhere in the system
MDR, or merchant discount rate, is the fee a merchant pays for accepting a payment, usually expressed as a percentage. EPOS360 BlueTap advertises 0% PayNow MDR and requires an ANEXT Bank account. Its page separately describes Antom card processing and an EPOS360 subscription. This is a bank-linked payment offer, not evidence that ANEXT provides free acceptance for every card or bears every cost. The published terms were checked on 16 September 2026.
Commercially, zero pricing can help win market share, but payment processing, fraud controls and support still cost money. Other income or an explicit subsidy must cover the gap. If the plan depends on merchants buying other services and they do not, higher volume can magnify the shortfall. Later fee increases can then drive merchants away, leaving acquisition costs unrecovered. This is the risk of aggressive acquisition pricing; the advertisement alone does not establish below-cost pricing or reckless management. UOB also advertises a 0% PayNow MDR campaign, so this tactic is not unique to digital banks.
Chocolate Finance: a warning about promotions and access to money
On 10 March 2025, Chocolate Finance paused instant withdrawals amid a surge in requests, and temporarily paused card transactions. This followed the removal of AXS payments from its miles offer. Its founder told CNA that bill-payment volumes had made the arrangement unsustainable and that the change had been poorly communicated. It is a relevant Singapore warning about incentives, customer confidence and withdrawal expectations, rather than a new September 2026 incident.
Chocolate Finance is a fund-management platform, not a licensed digital bank; its investment balances are not SDIC-insured bank deposits. By 21 March 2025, the company said it had paid all requests received from 10 March through the 18 March cut-off, including invested capital and earned returns. The incident should not be described as a bank collapse or evidence that customers lost their principal. The lesson is that a convenient app and attractive returns do not guarantee immediate access to money under stress.
Wirecard: what a provider failure can mean at the checkout
On 30 September 2020, MAS directed Wirecard’s Singapore entities to stop payment services and return customer funds by 14 October, following the parent company’s insolvency. MAS warned that merchants’ card acceptance and Wirecard-issued prepaid cards would be affected, and encouraged alternative arrangements. This was a payment-provider failure; it does not demonstrate that zero MDR caused a digital bank to fail.
For a bank, a run means many customers demanding their deposits back at once. If available cash cannot keep pace, withdrawals and payments may be disrupted even while loans still have value. For merchants, disruption at a bank or payment provider can interrupt takings, settlement, supplier payments or payroll. An orderly exit can also require account changes, terminal replacement and reconciliation work. Keeping an alternative payment route and accessible operating funds reduces reliance on one provider.
What are the inherent risks?
The main risks are the economics behind the offer and the continuity of access to money. They affect traditional banks too, but a young bank with limited recurring income has less room to absorb mistakes. A promotion, annual loss or small market share alone is not proof of imminent failure.
| Risk | What can go wrong | Possible customer effect |
|---|---|---|
| Profit and capital pressure | High deposit costs and low fees leave too little to cover bills and losses. | Reduced rewards, new fees, narrower services or a need to move providers. |
| Customers leave after incentives | Paid-for growth does not turn into lasting deposits or profitable borrowing. | Frequent changes to rates and product conditions. |
| Loan losses and rate changes | Borrowers fail to repay, or lending income falls faster than funding costs. | Tighter credit and fewer lending options. |
| Liquidity and confidence | Withdrawals arrive faster than assets can be turned into available cash. | Possible delays accessing money or completing payments under stress. |
| Technology, fraud and provider dependence | An outage, security incident or partner failure interrupts the service. | Checkout disruption, delayed settlement and recovery work. |
Deposit insurance depends on eligibility, not the app
In Singapore, eligible Singapore-dollar deposits are generally insured up to S$100,000 in aggregate per depositor per Deposit Insurance Scheme member. Opening several accounts at the same member does not multiply the ordinary limit. CPFIS and CPFRS monies have separate aggregation rules. Foreign-currency deposits, structured deposits and investment products are excluded. Check the institution’s scheme membership and register of insured deposits.
What does current competition in Singapore show?
Recent disclosures show different stages of development. Profitability measures need a period and an entity: one profitable month is not a profitable year, and a regional group is not the same as its Singapore bank. The following observations were available by 15 September 2026.
Trust: a monthly milestone, with annual losses still in the record
The Business Times reported that Trust made a profit in March 2026 but remained loss-making for Q1. Its FY2025 loss was about S$54 million, versus S$93 million in FY2024. This demonstrates progress without establishing a full-year profit for 2026.
MariBank Singapore: higher income does not guarantee lower losses
MariBank Singapore’s FY2025 loss widened to S$55.6 million from S$51.3 million despite higher income, with increased credit and other loss allowances. These are Singapore-bank figures; the wider group, including the Philippines, reported a different result.
GXS: competition is expanding beyond savings accounts
GXS announced a credit card developed with Grab and Singtel on 17 August 2026. The product expansion illustrates competition for spending relationships as well as deposits. EISOL’s interpretation: customer activity and additional revenue sources increasingly matter alongside account sign-ups.
GLDB: evidence that outcomes can differ
Linklogis, an investor in Green Link Digital Bank, reported an operating gain from the bank in its FY2025 results announcement. Together with Trust’s monthly profit milestone, this is a reason to assess each business individually. Pricing pressure is real, but failure is not the only possible outcome.
The central challenge is turning customer acquisition into durable, risk-adjusted earnings. Attractive savings rates raise funding costs; rewards raise acquisition costs; lending creates income but also potential losses. Meanwhile, established banks can respond with their own digital products. Winning an account opening is easier than becoming the account used for salary, supplier payments and repeat borrowing.
What could happen when competition intensifies?
The following are scenarios, not forecasts for any named bank. Several could happen at the same time.
| Pressure | Possible response | Effect on customers |
|---|---|---|
| Rate and reward competition | Higher promotional spending initially; later, reduced rewards, tighter eligibility or repricing to protect margins. | Short-term offers can improve, then become less generous. |
| Insufficient scale | Specialise in a viable customer segment, add partners or seek further shareholder capital. | More specialised services, with a narrower target market. |
| Persistent weak economics | Possible acquisitions, mergers, product exits or an orderly wind-down, subject to regulatory requirements. | Account or service migration may be required; an exit does not automatically mean deposit losses. |
Survivability depends on capital and liquidity headroom, credit quality, customer retention after promotions, operating efficiency and the ability to earn sustainable returns. Shareholder support can buy time, but an assumed parent guarantee is not a substitute for the bank’s own financial position. Losses alone do not prove imminent failure; rapid customer growth alone does not prove long-term viability.
Payment acquirer vs digital bank: different economics
A payment acquirer enables a merchant to accept payments and handles acquiring and settlement obligations. A card issuer serves the cardholder; Visa and Mastercard operate card networks. A gateway carries payment information, while a processor handles transaction processing. One provider may combine several roles, but payment processing volume is not the same as banking assets or deposits.
| Question | Digital bank | Payment acquirer |
|---|---|---|
| Core service | Accounts, deposits, credit and other banking services within its licence. | Merchant acceptance, transaction handling and settlement. |
| Revenue engine | Interest income and fees, less funding and service costs. | Processing or acquiring fees; potentially FX, software and other service fees. |
| Key exposure | Borrower losses, funding liquidity and interest-rate mismatch. | Fraud, chargebacks, merchant failure, settlement exposure and outages. |
The merchant service charge is not all acquirer profit. Card costs can include interchange paid to the issuer, scheme fees and the provider’s own charge. Profit depends on the retained revenue after network, processing, risk and operating costs. A merchant failure can also leave the acquirer exposed to refunds or chargebacks after funds have already been paid out.
A Singapore Major Payment Institution licence is not a banking licence. For example, Stripe’s Singapore disclosure lists merchant acquisition among its licensed payment activities and describes safeguarding relevant customer money in trust accounts. Safeguarding is a different legal arrangement from an insured bank deposit; a payment balance should not automatically be treated as one. Some banking groups also acquire payments, so check the contracting entity and service.
Examples of payment acquirers operating in Singapore
Visa’s Singapore acquirer directory lists Adyen Singapore, Airwallex (Singapore), Stripe, Global Payments Singapore and WorldPay, alongside bank acquirers including DBS, OCBC and UOB. These are examples, not a ranking or a complete market list. Acceptance channels, card coverage and commercial terms vary by provider and merchant contract.
What this means for a Singapore business
Choose the bank for treasury, credit and account needs, and the acquiring arrangement for payment acceptance, settlement and support. Then connect them operationally. A successful checkout is only the first step: the sale, payment reference, fees, refund and eventual bank deposit must reconcile.
EISOL helps businesses design and integrate POS, payments and accounting workflows. The practical objective is traceable transaction records, clear exception handling and less manual reconciliation. Technology creates value when a business can see what was sold, what was collected and what actually reached its bank account.