Key points
- Banking acceptance is decided by the bank's risk appetite, not by the regulator
- A licence helps enormously and guarantees nothing
- The flow-of-funds narrative is what most account applications are refused on
- Card issuing needs an e-money footing plus a scheme relationship
- Plan banking during jurisdiction selection, not after approval
A licence you cannot bank is not an operating business. We work on bank and payment-institution account opening across the UK, the EU and Singapore, and on card issuing programmes including U Card in Singapore and Hong Kong. The determining factor is rarely the licence itself — it is whether the bank's own risk process can get comfortable with your specific flow of funds.
Founders arrive expecting the licence to be the hard part. For a significant number of them, it is not — the account is. This page is about the second problem, which is the one that decides whether a licensed business can actually trade.
Why is banking so hard for this sector?
Because the decision is not a regulatory one. A bank is not obliged to open an account for you; it is making a commercial judgement about how much work your business will generate for its financial-crime function, and how confident it is that it can explain your transactions to its own supervisor. Crypto flows are harder to explain than most, correspondent relationships are risk-sensitive, and one poorly-handled account can cost a bank far more than it earns from it.
A licence changes this materially. It gives the bank a supervised counterparty, an AML framework someone else has already reviewed, and a regulator to point at. It does not make the outcome certain, and any adviser who says otherwise is describing a decision they do not control. Which supervisor stands behind you also matters to the bank's own assessment — a permission from the Monetary Authority of Singapore or an SRO recognised by FINMA is diligenced faster than one from a regime the analyst has not seen before.
What decides the outcome?
What an account application actually turns on
- A flow-of-funds narrative the bank's analyst can follow without asking a single clarifying question
- Clear beneficial ownership, with the corporate structure documented up to natural persons
- The licence or registration itself, and evidence that its scope covers what you will actually do through the account
- Named counterparties: which exchanges, processors and corridors money will move to and from
- Source of the company's own funds, evidenced rather than asserted
- An AML framework the bank can read and recognise, ideally mapped to the standards it uses itself
- A named individual on your side who can answer follow-up questions quickly and consistently
In our experience the single biggest factor is the flow-of-funds narrative. An analyst has to be able to read one document and understand where money comes from, what happens to it, where it goes, and why each step exists. Applications fail when that narrative is vague, when it contradicts the technical architecture, or when it omits a corridor that later shows up in transaction monitoring.
The second is verifiable corporate information. Ownership that resolves cleanly to natural persons, entities that can be confirmed in a public register, and a structure without unexplained intermediate layers. This is the same argument made at greater length on verifiable entity proof, and it applies to you exactly as it applies to any adviser you are assessing.
Card issuing
Card programmes sit on top of two things: an e-money or equivalent footing, and a relationship with a scheme or an issuing partner. We work on programmes including U Card in Singapore and Hong Kong. The licensing groundwork underneath is on fintech licences — a card programme without the right permission behind it is not a shortcut, it is an unlicensed e-money operation.
The practical constraints are worth knowing early: schemes have their own onboarding standards independent of any regulator, programme managers have their own risk appetites, and the settlement mechanics determine how much working capital you need to hold. None of that appears in a licence application, and all of it determines whether the product works.
Sequencing
Banking belongs in the jurisdiction decision, not after it. The question "which regulator will approve this fastest" and the question "which permission will my bank accept" have different answers, and only the second one determines whether you can operate. We raise it at stage two of the engagement described on how we work, and it is one of the comparison factors on jurisdictions.
Where a business already holds a licence and cannot get banked, the work is diagnostic first: understanding why previous applications were refused, since the stated reason in a decline notice is rarely the operative one. Standards across this area derive from the FATF recommendations, which is why an AML framework mapped to them reads as familiar to a reviewing bank.
We introduce, prepare and support. We do not control a bank's credit or compliance committee, and we do not promise an account. Nothing here is legal advice; CryptoLicense is an advisory firm, not a regulator and not a law firm.
