Offshore company with a bank account" is two things, and the second one is failing

The phrase turns up in search boxes thousands of times a month, and it is really two separate purchases bolted together by wishful thinking.

The first, an offshore company, is a commodity. Somebody files a form, pays a registry fee, and a certificate arrives. It is cheap, it is fast, and dozens of agents will do it for you this week.

The second, a bank account for that company, is not a commodity at all. It is a credit and compliance decision made by an institution that is under no obligation to say yes, is measured on the cost of getting it wrong, and has watched its own correspondent banks exit relationships over exactly this category of customer.

The two are sold together. They are not delivered together, and the gap is where people lose a year.

Why the refusal is usually not about you

The most useful thing to understand is that most refusals are not a judgement on your business.

Banks increasingly manage this exposure at the level of the category rather than the applicant. A compliance function looks at a class of customer, weighs the cost of monitoring it against the revenue it produces, and exits the class. Your file may be immaculate. If the bank has decided it no longer serves companies incorporated in a particular jurisdiction, no amount of documentation in your application changes the answer, because nobody is reading your application on its merits.

This pattern has a name in the regulatory literature. De-risking is the practice of terminating or restricting relationships with whole categories of client, and the international standard-setters have been warning about its side effects for years, precisely because it pushes legitimate businesses out of the regulated system alongside the ones the rules were aimed at.

Underneath it sits the correspondent banking chain. Your bank does not move dollars or euros on its own; it relies on relationships with larger institutions that clear those currencies. When a correspondent decides a jurisdiction is more trouble than it is worth, the effect flows downhill to every bank that depends on it, and then to every customer of those banks. You experience this as a refusal, or as an account that worked fine for two years and then closed with two months' notice.

What actually gets asked

Strip away the paperwork and an onboarding file answers six questions. They are the same six almost everywhere, and how well you can answer them matters far more than which registry issued your certificate.

What does this company actually do? Not the objects clause. The business. Who the customers are, what they pay for, how the money arrives.

Who owns it, and can you prove it? A single ownership chain, registers that match the chart, and the ultimate beneficial owner identified. This is the point where corporate secrecy in the jurisdiction of incorporation stops helping and starts hurting, because the bank will require disclosure regardless of what the local law permits you to withhold. A structure sold on privacy is a structure that cannot be banked.

Is the activity licensed, where licensing applies? And if the licence sits in one entity and the account is for another, the relationship between them has to be documented rather than asserted.

Where do the incoming funds come from? This is source of funds, and it is the question that sinks most applications. A balance is not an answer. An explanation is not an answer. What a compliance officer needs is a chain of documents: contracts, invoices, settlement records, statements, tying the money in the account to activity that produced it.

Is tax being paid, and where? Returns filed, accounts audited where audit applies, a tax residency certificate available on request. A company that pays tax nowhere is a company whose profit has no explanation.

Who runs it day to day? Named directors who are reachable, and ideally resident where the company is. An office. Payroll. The bank is testing whether the company exists as an operating business or only as a certificate.

Why the offshore answer is structurally weak

Run an offshore company through those six questions and the pattern is obvious.

The activity is usually real. The ownership is often provable, if the client is willing. Where it falls apart is the last three. Tax is frequently paid nowhere, because that was the point. Substance is usually absent, because the company was never meant to have any. And the source-of-funds chain is thin, because the entity was inserted to hold profit rather than to earn it, so there are few contracts and fewer invoices explaining why the money is there.

Add a jurisdiction the correspondent chain has already decided against, and the application is refused before anyone reads it.

The cost nobody prices at the start

An offshore company is cheap to form and expensive to live with, and the expensive part shows up in places that never appear on the quote.

Payment providers increasingly want an EU or UK counterparty and will not onboard an entity without one. Partners run diligence and walk away from a structure they cannot follow. Lenders decline because the accounts are not audited. A buyer's advisers price the uncertainty into the offer, or carve it out of the consideration entirely. And when the account does close, replacing it takes months during which the business still has payroll to run.

None of that is a tax cost. It is the cost of a structure the financial system has decided not to serve.

What changes with an onshore entity

Now run the same six questions against an EU-resident company that genuinely operates.

It files audited accounts, so the tax and accounts questions answer themselves. It pays tax at a published rate, so profit has a provenance. It has an office, staff and directors where it is registered, so the substance question is a matter of record rather than argument. Its jurisdiction is not on anyone's list, so the correspondent chain has no view about it. And because the entity earns its income under contracts rather than receiving it as a transfer, the source-of-funds file assembles itself as a by-product of operating.

That is the trade. You give up secrecy, which was never worth much, and you give up a zero rate you probably could not bank anyway. What you get is a company that can hold an account, take card payments, borrow, and be sold.

The Cyprus example, and its conditions

Cyprus is the version of this that technology businesses tend to land on, because it pairs the onshore position with a genuine incentive rather than asking them to give up the economics entirely.

A Cyprus company is EU-resident, files audited accounts, and pays corporation tax at 15% following the reform in force from 1 January 2026. Cyprus levies no withholding tax on dividends to non-resident shareholders in standard jurisdictions. And where the company owns software it developed itself, the IP Box deducts 80% of qualifying profit and taxes the remaining 20% at 15%, which produces an effective rate of approximately 3% on the qualifying share.

Since 2026 Cyprus also charges defensive withholding on outbound dividends: 17% where the recipient sits in an EU non-cooperative jurisdiction and 5% where it sits in a low-taxed one. That is worth reading twice, because it tells you what kind of jurisdiction Cyprus is trying to be. It taxes opaque ownership chains at source rather than accommodating them. Putting an offshore holding company above your Cyprus company is no longer a neutral choice; it is a priced one.

Two conditions apply and neither is negotiable. Approximately 3% is a floor rather than a promise, tied to the OECD nexus approach: the qualifying share tracks the R&D expenditure behind the asset and who incurred it, so acquired IP and development recharged from related companies outside Cyprus both reduce it. And the company has to be genuinely operated, with development and decision making where the company is, because substance is exactly what the banking questions above are testing.

What a usable source-of-funds file contains

Since source of funds is the question that decides most applications, it is worth knowing what an answer looks like before you need one. Not a narrative: documents, in an order somebody else can follow.

Certificates of incorporation, registers and a structure chart for every entity in the chain, and the chart has to match the registers. Contracts and service agreements between group companies, so intercompany flows have a stated basis rather than appearing as unexplained transfers. Monthly invoices, with the settlements that paid them. Audited financial statements and filed tax returns. A tax residency certificate. Dividend resolutions and payment records for anything that reached a shareholder.

Two things about that list. First, none of it is exotic; it is what an operating company produces anyway. Second, it is very hard to assemble retrospectively, which is why the file is either a by-product of running the business properly or a project that takes months and still has gaps.

The people who ask for it are not only banks. Lenders want it before a mortgage or asset finance. A buyer's advisers work through the same documents in the same order during diligence. And a private bank onboarding you personally will ask where the money that funded the account came from, which routes straight back through the company.

If you are already offshore

Most people reading this are not choosing from scratch. They have an entity, it has been running for years, and the banking is deteriorating.

Migration is usually possible where the revenue can be evidenced. Historical income can often be rebuilt from platform data, payment provider statements and contracts into a file an institution will accept, and the ownership chain can be simplified so it no longer runs through a jurisdiction that triggers enhanced scrutiny.

What migration cannot do is manufacture provenance that does not exist. Funds that cannot be evidenced do not become evidenced by moving them, and any provider who suggests otherwise is describing a problem they intend to hand back to you at onboarding. Full customer due diligence applies at the new institution exactly as it would to a new applicant.

The realistic framing is that migration remediates documentation for legitimate revenue. That is a large and useful thing, and it is not the same as a fresh start.

Sequence matters more than jurisdiction

If you take one practical thing from this, take the order of operations.

The common sequence is to incorporate first, then look for banking, then discover the problem. The workable sequence is the reverse: establish what the banking will require, structure the entity so those answers exist, and only then form the company. Registering is the administrative part and it can wait a fortnight. Being bankable is the part that decides whether the structure was worth building.

If you are searching for an offshore company with a bank account, the honest translation of what you want is an entity that can hold money, prove where it came from, and survive somebody else's diligence. In 2026 an onshore EU company does that more reliably and, once the failed applications and the closed accounts are counted, usually for less.