Summary

  • VEON says its venture with Square Group received approval for a digital-bank licence. Bangladesh Bank's Version 2 framework first gives an eligible applicant a Letter of Intent; local reports quoting the regulator describe the September decisions as initial approvals, with final licences conditional on completion.
  • The remaining receipt is substantial: at least BDT 300 crore of unencumbered cash capital under lien, an approved public company and leadership, capital-maintenance commitments, Bangladesh-local cloud, separate Tier III primary and recovery sites, and an accredited pre-licence ICT audit.
  • Banglalink's Mukto Pay payment-service licence, VEON's 59 million financial-services users in three other countries and its US$250 million multi-sector Bangladesh investment are adjacent evidence. None is a deposit base, a bank-capital cheque or a final section 31 licence.

One announcement, two regulatory nouns

On 25 September, VEON said that a joint venture between VEON Digital Financial Group and Square Group had received Bangladesh Bank approval for a Digital Bank licence. The release described the decision as VEON's entry into full-scale digital banking in Bangladesh and said the service would integrate with Banglalink's mobile ecosystem. The headline is unequivocal. The operating details are not.

The release does not name the final licensed public company, disclose the partners' ownership percentages or split the equity cheque. It gives no launch date, no product list and no deposit, lending or profitability target. Square Group's own quotation says the benefits will come “once it goes into operation”. That is a useful boundary: an approved project and an operating bank are not yet the same asset.

Bangladesh Bank's Guidelines to Establish Digital Bank, Version 2, dated 20 August 2025 and still listed on the regulator's guidelines page, give the boundary a legal sequence. A digital bank must be a public limited company and must ultimately be licensed under section 31 of the Bank Company Act. But the evaluation stage first issues a Letter of Intent to an applicant deemed eligible. The LOI carries conditions and a completion clock. It automatically lapses if those conditions or the stipulated timetable are missed, and Bangladesh Bank retains the right not to grant the final licence.

Local reporting uses the narrower noun. bdnews24.com reported, citing Bangladesh Bank spokesman Arif Hossain Khan, that four applicants had received initial approval and would receive licences after meeting the conditions. The Daily Star likewise described Letters of Intent and identified the proposed VEON–Square entity as Nova Digital Bank. Those reports are not a substitute for a formal licence notice. They do show why the economically conservative reading is a conditional build mandate, not a live banking franchise.

This is a terminology mismatch, not evidence of deception. Corporate communications often name the intended destination; a prudential rulebook names the current stage. An investor, competitor or infrastructure supplier needs the latter because the value of the project now depends on which conditions become irreversible.

BDT 300 crore must become restricted cash

The first hard gate is not an addressable-market estimate. It is paid-up ordinary-share capital. The rulebook sets the minimum at BDT 300 crore. Sponsor contributions must be cash, free of encumbrance, deposited with a local scheduled bank, independently verified and held under lien in favour of Bangladesh Bank. Promises, group experience and a broad investment envelope cannot occupy that line.

The capital obligation also survives opening day. Substantial shareholders must enter a capital-maintenance agreement and can be required, jointly and severally, to inject more funds if regulatory capital falls below the required level. After the final licence, the bank must pursue an initial public offering within five years, for an amount no smaller than the original sponsor capital. Sponsor selection therefore creates a future financing architecture as well as an entry ticket.

The public release leaves the most useful denominator blank. It does not say how much VEON Digital Financial Group will contribute, how much Square Group will contribute, whether additional sponsors will participate or when the cash will be placed under lien. It also does not reconcile bank capital with VEON's separate US$250 million anchor-investment announcement. That anchor spans connectivity, digital infrastructure, financial services and artificial intelligence and is presented as a catalyst for a US$1 billion mobilisation ambition. It is not disclosed as a BDT 300 crore bank-capital allocation.

This distinction prevents an easy but invalid conversion. A group-level commitment across several sectors cannot be divided into bank equity without an allocation decision. A regulatory minimum does not prove the cheque has cleared. The next meaningful disclosure is therefore not another total for Bangladesh investment; it is a cap table and cash receipt for the licensed company.

Local cloud is a licensing condition, not a hosting preference

The second gate makes data locality physical. Bangladesh Bank requires at least a Tier III data centre and a disaster-recovery site located in different seismic zones. Cloud infrastructure must be located in Bangladesh. The applicant needs an ICT business-continuity plan, ICT risk management, compliant core-banking and security systems, and an infrastructure audit by an internationally accredited firm before the licence. External ICT audits continue annually after launch.

That design changes the economics of a supposedly branchless bank. Removing a branch estate does not remove fixed infrastructure; it relocates fixed cost into resilient sites, local cloud, security controls, payments connections, identity systems and audit. The local-cloud clause also limits the freedom to treat a regional platform as a frictionless copy. Group software may be reusable, but regulated data, failover, vendor contracts and operating accountability have to land inside Bangladesh's perimeter.

The corporate announcement says the venture will combine Banglalink's digital reach, VEON's technology and Square Group's domestic presence. It does not identify the core-banking supplier, cloud operator, primary site, recovery site, migration design or audit firm. Absence of those details is not proof that the work has not started. It means the announcement cannot yet price completion risk.

For vendors, the opportunity lies precisely in that gap. For sponsors, the gap is a sequence of expenditure before bank revenue. For the regulator, it is the point of control: the system must demonstrate resilience and local accountability before deposits and loans move from product plan to regulated balance sheet.

Mukto Pay is adjacent, not the bank

Banglalink already has a separate financial-services milestone. On 12 August, VEON announced that Banglalink had received a Payment Service Provider licence for Mukto Pay. It targeted a September commercial launch, subject to the remaining technical and regulatory requirements.

That licence matters. A payment service can build distribution, merchant acceptance, wallet activity and familiarity with regulated transaction systems. It can reduce the customer-acquisition friction facing a future bank. It does not make the two perimeters interchangeable.

A PSP moves payments under its own permissions. A section 31 bank may take deposits, extend credit and carry prudential balance-sheet risk after the corresponding licence and controls exist. If Mukto Pay launches first, its users, transactions and merchant network should be reported as PSP metrics. They should not be called Nova accounts, Nova deposits or proof that the bank has opened. The separation is not pedantry; it keeps consumer obligations, capital, liquidity and credit risk attached to the correct legal entity.

The same rule applies to Banglalink's telecom base. Distribution can be strategically valuable without becoming a banking denominator. A SIM relationship does not equal consent to a bank account. A digital-service user is not automatically a depositor. Cross-selling may create option value, but adoption must be measured after the product, permissions and legal perimeter exist.

Experience abroad is capability, not local demand

VEON's second-quarter results reported 59 million active users across its financial-services platforms in Pakistan, Kazakhstan and Uzbekistan. That scale supports a credible capability claim. The group has operated wallets and financial platforms in markets where telecom distribution can lower access costs.

It does not forecast Bangladesh. The 59 million users belong to different products, regulations, competitive sets and economies. They are not Nova customers waiting to be activated. Treating them as a transferable installed base would import revenue and adoption before the Bangladesh entity has a final name, licence or account.

The better use of the number is operational. It suggests VEON can bring product design, fraud controls, channel experience and engineering knowledge to the venture. Square Group can bring domestic commercial knowledge. Banglalink can provide a distribution and service interface. The value of those assets becomes measurable only when the sponsors disclose how they are separated, contracted and governed inside the bank.

Build the activation receipt

The licence story should now be reported as a ledger with no skipped lines.

First, publish the LOI date, its conditions and its deadline. Second, name the public limited company that will seek the section 31 licence and disclose its ownership. Third, show each sponsor's paid cash, the verification and the Bangladesh Bank lien. Fourth, identify the board, senior managers, fit-and-proper decisions and capital-maintenance allocation. Fifth, describe the local data-centre, disaster-recovery and cloud architecture without exposing security-sensitive detail. Sixth, obtain the accredited ICT audit and close the regulator's conditions. Seventh, publish the final section 31 licence.

Only after that sequence should operating metrics begin: first live account, first customer deposit, first payment, first loan, liquidity, capital ratios, arrears, credit losses, complaints and unit economics. Each is a different receipt. The phrase “full-scale digital banking” describes the destination; it should not collapse the route.

The September approval has real value. It narrows the competitive field, gives the sponsors authority to complete a regulated build and turns a strategic idea into a conditional option. The option is not worthless because it is conditional. Its value is simply a function of execution rather than headline grammar.

Evidence boundaries

No formal Bangladesh Bank notice granting this venture its final section 31 operating licence was located in the material reviewed for this article. The LOI description comes from the published Version 2 rules and local reports attributing the current decision to the regulator's spokesman. If Bangladesh Bank subsequently publishes a final licence, that new primary evidence should replace the present stage assessment.

The public record also does not disclose sponsor ownership, the equity split, LOI conditions, capital placement, infrastructure suppliers, audit completion, final launch date or bank economics. These are unknowns, not negative findings. The analysis does not infer failure, misconduct or delay from their absence.

Sources