Bangladesh’s Fintech Story Is Told at the Wrong Altitude

When foreign investors ask about Bangladesh’s fintech sector, the conversation begins well, and it should. Mobile financial services now reach a population the banks never did. Close to half of adults hold an MFS account, nearly double the share who hold a bank account, and bKash alone moves more transactions in an average day than every bank card in the country combined, through an agent network of more than 300,000 outlets. The reach has produced real inclusion: the formally banked share of adults has risen from around 31 percent to roughly 50 percent in a few years, and the FY2026-27 budget named digital finance as a pillar of reform (BBS, 2023; Bangladesh Bank, 2024-25). This is a genuine achievement, and the people who built it deserve the credit they get.

One number, though, rarely shares the conversation. MFS transaction value grew 28.4 percent in 2024 and only 7.8 percent in 2025 (Bangladesh Bank, via Future Startup and The Daily Star, 2025-26). The easy explanation is the political transition after August 2024, and it probably accounts for much of the fall; with a stable government in place, growth may recover. But that a single shock could slow an entire sector by two-thirds is worth reading as a stress test, and what it exposes is the real story. Bangladesh built extraordinary reach at the point of transaction and very little underneath it.

The contrast with India makes the point. India’s Unified Payments Interface did not arrive as a product. It arrived on a decade of groundwork: a national identity system, a drive that opened hundreds of millions of basic accounts, and only then the payment rails on top. Identity first, then accounts, then rails. The order is not a technicality. Identity and verified accounts are what let a rail carry more than money, which is how India layered credit, savings, and lending onto the same pipes. Bangladesh inverted the sequence. Distribution came first, and from the private sector rather than the state, so the country now has a dominant wallet sitting on layers that were never laid, and a rail that moves money well but struggles to carry anything else.

Interoperability is the clearest example. At the merchant counter, Bangladesh has solved it: one Bangla QR code lets a shop accept payment from more than fifty banks and providers. But at the account-to-account layer the system stalls. Binimoy, the central bank’s transfer platform, requires both parties to register an alias before money can move, so a tool meant to connect wallets has instead become a second one that few providers bother to use (TIB, 2025). The consequence shows in the data. Around 80 percent of e-commerce is still settled cash on delivery, and Bangladesh Bank’s target of 75 percent of retail transactions cashless by July 2027 has no visible mechanism strong enough yet to make it credible.

Underneath the plumbing sits the layer nobody names. Cash reliance is not inertia. It is a priced response to observed risk. Bangladeshis have watched trust break at every level of the system: the 2016 theft of USD 81 million from the central bank’s own account at the New York Federal Reserve, the 2021 collapse of Evaly owing customers more than Tk 500 crore, the 2023 exposure of national ID data for over fifty million citizens, and the steady drip of SIM-swap and OTP fraud at the consumer’s screen. Someone who has watched all of that is behaving rationally when she pays the delivery rider in cash.

For an investor, one more problem compounds the rest. Registered-account and transaction totals are easy to obtain; active users, dormancy, and segment-level demand are not. Registered accounts and merchant sign-ups are distribution metrics. They are not demand metrics. Asked to underwrite a market it cannot measure, capital does what it always does with unmeasurable risk. It demands a return no early-stage firm can promise, or it waits.

None of this means the market is unready for sophisticated products, and the common claim that it is deserves correcting. bKash, through a City Bank partnership, has disbursed Tk 2,300 crore in collateral-free nano loans to nearly a million borrowers, at ticket sizes no manual underwriting process could serve economically (Bangladesh Bank, via TBS, 2025). It works for one reason: it sits directly on the wallet and the agent network, the one rail Bangladesh has built. That is the actual distinction, and it is not frontier versus foundational. It is whether a product rides that rail or needs one that does not yet exist. Standalone credit scoring struggles because there is no shared credit bureau for fintechs and no workflow to carry a score into a funded loan. Insurtech stays nascent, with penetration below one percent of GDP, because micro-insurance depends on a trust the rails do not yet carry. The products that work are riding the rail. The ones that stall are waiting on a layer no one has built.

The work ahead follows from that, and it is unfashionable precisely because none of it is a product. Interoperability has to become a mandate with a deadline rather than a platform providers may decline to join, starting with Binimoy’s alias requirement. Digital literacy and fraud infrastructure have to be funded as shared market development, not left to whichever operator is willing to subsidize its competitors. And the regulatory perimeter, drawn around banks and licensed operators, has to make room for the technology providers who now build the plumbing beneath them. None of this is only an investor’s concern. For Bangladesh Bank it is the distance between an announced cashless target and a met one; for the banks and founders it is where the next decade’s margins and fundable problems actually sit.

Bangladesh’s fintech story is not a bad one. It is a good one told at the wrong altitude. From high up, the subscriber and transaction numbers look like the profile of a mature market. Come down a few thousand feet and you find most of e-commerce still settled in cash, an interoperability platform almost nobody uses, and growth that fell by two-thirds in a single year. Those who keep flying high will keep being disappointed. Those who come down will find that the unglamorous work is where the returns are.

References

1. Bangladesh Bank. (2024-25). Financial inclusion data; MFS transaction data; remittance inflow data; nano lending, BNPL, and credit card growth (via TBS, September 2025).
2. Bangladesh Bureau of Statistics. (2023). MFS and bank account penetration among population aged 15+.
3. Federal Reserve Bank of New York and Bangladesh Bank. (2016). SWIFT payment-instruction fraud incident (USD 81 million).
4. Future Startup. (2025, May), and The Daily Star. (2026, March). MFS transaction value growth, 2024-25.
5. National Budget Speech FY2026-27. (2026, June 11). Financial sector reform: digital banking and rural finance access.
6. Oxford Internet Institute (iLabour Project); Payoneer and BASIS. (2025-26). Bangladesh online freelance labour supply.
7. The Business Standard. (2021, September). Evaly collapse and advance-payment liabilities.
8. The Daily Star. (2023, July). Exposure of national ID data of more than 50 million citizens.
9. Transparency International Bangladesh. (2025, May). MFS Sector Report: Binimoy adoption and the alias requirement.
10. United States International Trade Administration. (n.d.). Bangladesh eCommerce Country Commercial Guide.
11. World Bank. (2025). Global Findex Database: Bangladesh.