Banking5 min read
Why banks close accounts, and what precedes it
Account closures feel arbitrary because the bank is frequently prohibited from explaining them. They are not arbitrary. There is a pattern, and most of it is visible months in advance.
The letter is two paragraphs long, gives sixty days' notice, cites a clause in the terms, and offers no reason. Calling the bank produces someone who genuinely does not know. This is experienced as arbitrary and it is not — it is a specific decision made in a specific place, under rules that also prevent anyone telling you about it.
What to do in the sixty days is a separate problem with a practical answer. This is about the months before, which is where the leverage actually is.
Two different events that look identical
- De-risking. The bank has decided that a whole *category* is not worth serving — a country, an industry, a payment pattern, a size of business. It is a portfolio decision, nothing to do with you specifically, and appealing it on your own merits does not work because your merits were never assessed.
- An exit decision on you. A review of your account concluded that the risk cannot be managed at an acceptable cost. Something specific triggered it. This one sometimes responds to information, if you supply it before the decision rather than after.
You can usually tell them apart by the timing. A de-risking exit arrives with no preceding contact, often alongside similar letters to businesses like yours. An account-specific exit is preceded by questions.
Why they cannot tell you
The commercial reality behind it is unromantic. Serving a customer whose profile requires enhanced due diligence, ongoing monitoring and periodic review costs a fixed amount in analyst time each year. Below a certain revenue, that arithmetic never works, and the bank's cheapest correct answer is to exit.
What actually triggers a review
| Trigger | Why it fires |
|---|---|
| Activity that does not match your stated business | You described a design agency; the account shows fifty small inbound payments a day. The mismatch, not the activity, is the problem |
| A step change in volume | Turnover that triples in a month re-opens a risk assessment made at a different scale |
| Payments to or from higher-risk jurisdictions | Screening lists are applied mechanically, and a legitimate customer in the wrong country still generates an alert |
| Crypto exposure | Inbound from an exchange, or outbound to one, is among the most reliable review triggers in retail and small-business banking |
| Round-number transfers between your own entities | Regular round sums with no invoice narrative read as layering, whatever their actual purpose |
| A change in beneficial ownership or address | Any change refreshes the file and can move you into a different risk band |
| Periodic review falling due | Nothing changed. The bank's own cycle came round, standards rose in the interim, and your file no longer meets the current one |
| An unanswered request for information | The single most avoidable cause. Non-response is treated as an unmanaged risk and closes the account by itself |
The signals that come first
An account-specific exit is rarely the first contact. Watch for these, and treat any of them as the beginning of a process rather than an administrative annoyance:
- A questionnaire about the nature of your business, particularly one you have already answered in the last two years.
- A request for source-of-funds documents on a specific payment.
- A single held or returned payment with a vague explanation.
- A request to re-verify beneficial owners outside the normal cycle.
- A relationship manager who stops being assigned to you, or an account quietly moved to a different service tier.
What reduces the risk
- Describe your business accurately at onboarding, including the awkward parts. A fact disclosed in week one is a risk the bank priced and accepted. The same fact discovered in month eighteen is a customer who concealed something.
- Keep the narrative on your payments. Invoice numbers and references on intercompany and supplier transfers do more to prevent a review than anything else on this list.
- Tell them before the change, not after. A new market, a new product line, a big contract that will triple volume — a two-line email in advance converts an alert into an expected event.
- Do not run a business through a personal account. It is the fastest route to a closure of both.
- Keep the entity's own house in order. Good standing, current filings, a clean register of members. A bank that cannot verify your company from public records treats that as your problem.
The structural answer
Everything above reduces the probability. None of it makes closure impossible, because the largest single cause is a category decision you cannot influence. So the position to be in is not *unclosable* — it is survivable.
- A second banking relationship, opened while you do not need it. Applying for a new account while holding a closure letter is materially harder: you have to explain the closure, and you cannot, because nobody told you why.
- Payroll runnable from either institution, tested at least once.
- A standing document pack — formation documents, ownership, contracts, statements, source-of-funds narrative — kept current, so a new application takes days rather than weeks. It is the same pack an account opening abroad asks for.
- Direct debits and customer payment details you can move. Knowing which counterparties need re-pointing, before you need to re-point them.
This is the same conclusion as processor freezes and deposit protection, reached from a third direction, which is a reasonable signal it is correct. The risk you are actually managing is loss of access, and the only real defence against it is redundancy you arranged in advance.
The second rail, before you need it
Accounts, processing and the document pack each application asks for — set up so no single provider can stop the business.
See the Business OS