On 24 September 2026, Bangladesh Bank issued Letters of Intent to five proposed digital banks: bKash Digital Bank, Nova Digital Bank, Boost Digital Bank, DK Digital Bank and Kori Digital Bank. Each must meet regulatory conditions, including Tk300 crore in paid-up capital, before receiving a final licence. They will operate without branches and focus on smaller borrowers.
Since then, I have been asked two questions repeatedly: Why do we need digital banks when every bank already has an app? And how does a digital bank actually work?
Both deserve a better answer than “it is a bank on your phone”. If that is all these licences produce, Tk1,500 crore of sponsor capital will have built five more apps.
A digital bank is still a bank. It takes deposits, lends, manages liquidity and manages risk. The ledger does not change. What changes is everything around it: lower cost to serve, distribution through smartphones and ecosystems, and products that can improve in weeks rather than quarters.
But the most important difference is not technology. It is context.
A customer does not wake up thinking, “I need a banking product.” A shopkeeper needs inventory. A parent needs to pay school fees. A farmer needs money before harvest. Banking should sit underneath those moments.
That is where digital banking becomes more than digitised banking.
Take a simple example. You buy groceries for Tk1,233. The bank rounds it up to Tk1,240 and moves Tk7 into savings. By evening, you have saved without making a separate decision. Your salary arrives on the first; your largest bills fall later. With your consent, the bank helps you set money aside early and warns you before a cash-flow gap opens.
But data must be earned, not assumed. A digital bank must explain what it knows, why it uses it and how customers remain in control.
The real test is lending
China’s MYbank pioneered the “310” model: three minutes to apply, one second to approve and no human intervention. By the end of 2023, it had served more than 53 million small and micro enterprises; more than 72 per cent of its new borrowers that year were taking a bank business loan for the first time.
Instead of asking only, “What collateral do you own?” it asks, “What does your economic activity say about your ability to repay?”
That question matters enormously here. CMSMEs contribute around a quarter of GDP and employ more than 34 million people, yet face a financing gap estimated at over US$73 billion. Meanwhile, mobile financial services move more than Tk5,000 crore a day.
A tea-stall owner in Rangpur who has received digital payments for three years may carry a richer credit signal than her lack of collateral suggests.
Nor is the business case simply “no branches, lower costs”. The larger prize is deeper relationships: payments, savings, working capital and insurance built around the same economic activity.
By March 2026, Nubank reported more than 135 million customers and a 29 per cent return on equity. Its monthly average cost to serve an active customer was about $1.
Technology alone will not win
Should bankers then make way for engineers? Technology accounted for 36 per cent of MYbank’s operating expenses.
Yet digital banks rarely fail for want of a feature. They fail on funding, underwriting and economics; Australia’s Xinja and Volt both handed back their licences.
Engineers must understand credit, and compliance must sit at the design table, not arrive after launch.
Singapore’s DBS rebuilt itself around technology and data while keeping what incumbents have: a balance sheet, deposits and trust.
The smartest model may be banks, fintechs, MFS providers and telecoms connecting their strengths. Customers will not care who owns the infrastructure, only whether the service is fast, affordable, safe and useful.
Bangladesh already has the ingredients: MFS, QR payments, agent networks, fintechs and millions of small businesses.
A digital bank could fund a merchant inside an e-commerce platform, finance a purchase at checkout or lend to a farmer based on the economics of a crop cycle. None of this needs to begin with “Open your banking app”.
That is embedded finance. And that is the real opportunity.
Three risks stand in the way
Deposit-first thinking can raise funding costs without widening inclusion.
Reckless digital credit can create over-indebtedness; speed is not good banking.
Weak data ethics can also erode trust. A digital bank cannot say, “We know everything about you.” It should say, “With your permission, we understand enough to serve you better.”
Five years from now, these banks should not be judged by downloads or deposits alone, but by how many first-time borrowers they bring into formal credit, how many people begin saving regularly, how efficiently they serve customers, and whether customers trust them with their money and data.
If successful, Bangladesh will move from banking as a destination to banking as infrastructure.
We will not simply have more banks.
We will have changed what banking means.
Writer is the he author of “From Cash to Code: How Digital Finance Can Transform Bangladesh’s Invisible Economy and Visible Work, Invisible People.”
Views expressed are solely those of the author
