Why the BVI still works for crypto startups— and when it does not.

Ask a dozen crypto founders why they incorporated in the British Virgin Islands and most will give an answer about tax. That answer is incomplete and it is the reason a meaningful share of BVI structures underperform. The jurisdiction's real advantage is legal infrastructure: a company law statute that has been tested in litigation for decades, a commercial court whose judgments are recognised internationally, and a professional services layer that has been serving institutional clients since long before digital assets existed.

For a crypto startup, that translates into things you can actually use. Investors have seen the documents before, so a BVI business company with a standard memorandum and articles does not slow a financing round. Exchanges have onboarded BVI issuers before, so their compliance teams know which registry extracts and register of members they need and how to verify them. Counterparties will sign a contract governed by BVI or English law without a fight. None of that is glamorous, and all of it saves months.

The second advantage is regulatory clarity of a particular kind. Rather than pretending digital assets did not exist or trying to supervise everything, the BVI defined a perimeter through its virtual asset service provider regime. Activities inside it require registration and supervision. Activities outside it — notably issuing your own token, holding treasury, or developing software — generally do not. That produces a workable answer to the question every founder and every bank asks: is this entity regulated, and if not, why not? A clear, defensible answer to that question is worth more than a vague hope that nobody asks.

The third advantage is speed and cost relative to the alternatives that offer comparable credibility. A BVI business company can be incorporated quickly, maintained at predictable annual cost, and does not require local directors, local office space or a minimum capital deposit. Compare that with an onshore European structure carrying employment obligations, statutory audits and substance requirements from the first day, and the difference in the first two years of a startup's life is material.

None of this makes the BVI right for every project. It is a poor fit for a business whose customers are consumers in a single regulated European market, or for a team that wants to run a licensed exchange serving retail users in the EU. This guide is about where it fits well, where it does not, and how to build the structure so it survives contact with banks, exchanges, investors and tax authorities.

The workhorse vehicle is the BVI business company, formed under the BVI Business Companies Act. It is a limited liability company with separate legal personality, shares, directors and members, and it behaves the way a company lawyer from any common law jurisdiction expects it to behave. That familiarity is the point.

A few features matter specifically for crypto projects. There is no minimum share capital, so a company can be capitalised with a nominal amount and funded through loans or subscriptions as needed. Shares can be issued in multiple classes with different rights, which accommodates the founder, investor and advisor arrangements that token projects tend to need. A single person can be both the sole director and sole shareholder, which suits an early-stage project without forcing an artificial board.

There is no requirement for a BVI-resident director and no requirement to hold meetings in the territory. That is a genuine flexibility, but it is also the feature that founders most often over-rely on. Where a company is managed and controlled has tax consequences in the countries where the directors actually sit, and the absence of a BVI residence requirement does not make those consequences disappear. Section seven covers this properly, because it is the single most common structural failure we see.

Accounting records must be kept and must be sufficient to show and explain the company's transactions. There is an annual financial return requirement to the registered agent. There is no general public filing of financial statements and no statutory audit requirement for an ordinary unregulated business company, which is often described as secrecy and is not: registers of directors are filed with the Registrar, beneficial ownership information is held and accessible to competent authorities, and the territory participates in international information exchange. The correct expectation is commercial confidentiality, not opacity from tax authorities.

Ongoing obligations are light but real: a registered agent and registered office in the BVI, annual government fees, maintenance of statutory registers, and economic substance reporting for the relevant activity categories. Missing these is how companies end up struck off the register at exactly the moment an exchange or a bank asks for a certificate of good standing.

The BVI operates a registration regime for virtual asset service providers, supervised by the Financial Services Commission. The regime targets defined activities carried on as a business: exchanging virtual assets for fiat or for other virtual assets, transferring virtual assets on behalf of others, safekeeping or administering virtual assets or the instruments that control them, and providing financial services related to the offer or sale of a virtual asset by an issuer.

The practical consequence is a bright-ish line. A project that issues its own token, holds its own treasury, and builds software that users interact with non-custodially is generally outside the regime. A project that holds customer assets, matches customer orders, or moves value on behalf of others is generally inside it. Between those poles sit the cases that need real analysis: a wallet whose key management arrangements could be characterised as custody, a bridge that takes control of assets in transit, a staking product where user assets are pooled, a launchpad that handles subscription funds.

Getting this analysis wrong in either direction is expensive. Operating inside the perimeter without registration is an enforcement matter and, more immediately, an instant disqualifier for banking and for institutional counterparties. Assuming you need registration when you do not costs a great deal of money and adds a supervisory relationship, capital and reporting obligations, and an approval timeline that you did not need to take on.

Where registration is required, the application is a real regulatory file rather than a form. It covers the business plan and activity description, ownership and control with fit-and-proper assessment of controllers, senior management including compliance and money laundering reporting appointments, anti-money-laundering and counter-terrorist-financing policies calibrated to the actual product, custody and key management arrangements, cyber security, financial projections and capital adequacy, outsourcing arrangements and a wind-down plan. Applications are prepared with qualified BVI counsel and a licensed agent; timelines run in months, not weeks, and the regulator asks questions.

Our consistent advice is to run the perimeter analysis before the product is finalised, because small design choices frequently decide the outcome. Whether the user or the protocol holds the key, whether funds route through a company-controlled account, whether the interface is the only route to the contract — these are engineering decisions with regulatory consequences, and they are far cheaper to make correctly than to unwind.

The BVI does not impose corporate income tax, capital gains tax or withholding tax on a business company's profits. That is the fact that draws founders, and it is true. What it does not mean is that the profits are untaxed, and the gap between those two statements is where most of the damage happens.

A BVI company with no local tax charge is still exposed to tax in every jurisdiction that has a claim on it. The three claims that matter are corporate residence, permanent establishment, and controlled foreign company rules. Corporate residence: many countries treat a company as resident, and therefore taxable on worldwide profits, where its central management and control or its effective management actually sits — which is where the directors make real decisions, not where the certificate was issued. Permanent establishment: even without residence, a fixed place of business or a dependent agent habitually concluding contracts in a country can create a taxable presence there. Controlled foreign company rules: many countries tax resident shareholders directly on the undistributed profits of low-taxed foreign companies they control, regardless of whether anything is distributed.

For a crypto startup this usually means the honest question is not what the BVI charges but where the team is. A BVI company whose two founders both live and work in a high-tax European country, make every decision there, and hold all the shares personally is very likely resident or attributed there. The company is not a shield; it is a filing obligation the founders may not know they have.

This is not an argument against the BVI. It is an argument for designing the structure and the founders' personal positions together. Where the founders are genuinely mobile or have already established residence in a jurisdiction with a favourable treatment of foreign company income, a BVI issuer can be extremely efficient and entirely defensible. Where they are not, the structure needs either genuine management substance in a suitable jurisdiction or a different shape altogether — and the honest version of that conversation should happen before incorporation.

Economic substance requirements add a further layer. The BVI applies substance rules to companies carrying on relevant activities, with holding companies subject to a reduced test and other categories requiring adequate local expenditure, employees and management. Pure token issuance and software development are not automatically relevant activities, but the classification needs to be assessed and reported annually rather than assumed. Everything in this section is general information and not tax advice; the specific analysis depends on facts we would need to see.

The BVI is one of the most commonly used jurisdictions for token issuance, and the reasons are practical. There is no bespoke prospectus regime that automatically catches a utility token, the company law is flexible enough to accommodate unusual capital and governance arrangements, and the professional services layer has issued enough legal opinions on token characterisation that counsel know how to write one that exchanges will accept.

The legal opinion is the deliverable that matters. Listing venues, custodians, market makers and banks will ask for a written opinion from recognised counsel addressing what the token is, whether it constitutes an investment or security under BVI law, whether the issuance or the ongoing operation falls within the VASP regime, and what restrictions apply to offers into other jurisdictions. An opinion is only as good as the facts it is built on, which is why the token's actual design — rights conferred, revenue linkage, redemption mechanics, marketing language — must be settled before the opinion is drafted rather than after.

A common structure pairs a BVI issuer with an operating company elsewhere. The issuer holds the token treasury and carries token-related obligations. The operating company employs the team, owns or licenses the technology, and provides development services to the issuer under a written agreement priced on arm's-length terms. This separates token liabilities from the operating business, allows each entity to sit where its activity fits, and produces a structure acquirers and investors recognise. It fails when the agreements are never signed or the pricing is arbitrary, which converts the arrangement into a transfer pricing exposure in the country where the team works.

Where a foundation-style vehicle with independent governance is required — typically for projects that need credible decentralisation or a non-owner-controlled treasury — the BVI is often paired with, or replaced by, a jurisdiction offering purpose trusts or foundation companies. That is a design question about who controls the treasury and how governance is exercised, not a tax question, and it should be answered on that basis.

One warning worth stating plainly: BVI incorporation does not exempt an offering from the securities laws of the countries where the token is actually offered. If the token is marketed to United States persons or to retail investors in regulated European markets, those regimes apply to the offer regardless of where the issuer sits. Distribution controls, geofencing and offering restrictions are part of the structure, not an afterthought.

A BVI company connected to digital assets is not an easy banking application, and any adviser who suggests otherwise is selling something. Institutions apply enhanced due diligence to both the jurisdiction and the sector, and the combination means the file is read carefully by people whose job is to find reasons to decline. That is the environment. Within it, well-prepared applications succeed regularly and poorly prepared ones fail regardless of the size of the deposit.

What an underwriter is actually assessing is whether they can explain your account to their own compliance committee and to their regulator in twelve months. That means they need a coherent answer to five questions: who ultimately owns and controls the company, what the business genuinely does and how it earns money, where the money comes in from and where it goes out to, what the source of the initial and ongoing funds is with documentation, and what controls exist to stop the account being used for something else.

The file that answers those questions well contains certified corporate documents and a current certificate of good standing, a register of members and directors, verified identification and proof of address for all controllers, a business description written for a compliance reader rather than an investor, projected transaction flows by counterparty type and corridor, documented source of wealth and source of funds, the legal opinion on the token or the regulatory perimeter analysis where relevant, and the AML framework the company actually operates. Where the company touches customer assets, the custody and safeguarding arrangements must be described precisely.

The architecture matters as much as the file. Most crypto companies should not run one account. A workable design separates the operating account for payroll and suppliers, a conversion relationship for digital-asset-to-fiat activity through a regulated venue or desk, and at least one contingency relationship opened before it is needed. Concentration in a single institution is the risk that actually materialises: a bank changes its sector appetite, the account closes with notice, and a company with no alternative cannot pay its staff. Building the second relationship while the first is healthy is the difference between an inconvenience and a crisis.

Expect the process to take weeks rather than days, expect follow-up questions, and expect at least one institution to decline for reasons that have nothing to do with your file. That is normal. Xavion works across a network of more than one hundred and twenty banking and payment institutions, and we position each application with the institutions whose stated appetite matches the flow. No adviser can guarantee an account — every institution decides independently, and we say so before an engagement starts.

The most common structural failure in BVI crypto structures is not exotic. It is a company incorporated in the BVI, directed entirely from a founder's apartment in a high-tax country, with a nominee director who signs whatever is put in front of them and a registered office that has never seen a board meeting. That arrangement is transparent to any competent tax authority and is increasingly transparent to banks and exchanges as well.

Substance means the company's decisions are genuinely made where the company says they are made. In practice, that requires directors who actually exercise judgement, board meetings that occur where they are minuted as occurring, minutes that record real deliberation rather than ratification, contracts that are negotiated and signed by the people the structure says are responsible, and books and records held in the same place. A director resident in the right jurisdiction who never exercises independent judgement provides no protection at all; the analysis follows where control actually sits.

For a startup with two founders in one country, honest substance in a third jurisdiction is often unaffordable. That is a legitimate finding, and the right response is to change the structure rather than to pretend. Options include placing the operating entity where the team genuinely is and accepting local tax on service revenue while the BVI entity holds only what it can defensibly hold, relocating the founders' personal residence if that is a life decision they were considering anyway, or appointing genuinely independent directors with real authority and paying for that authority to be exercised.

The BVI's own economic substance regime is separate from, and narrower than, this analysis. It applies to defined relevant activities — holding, financing and leasing, headquarters, distribution and service centres, intellectual property, fund management, banking, insurance and shipping — with a reduced test for pure equity holding companies. Assessment and annual reporting go through the registered agent. Satisfying BVI substance requirements does not resolve the corporate residence question in the countries where the directors live; those are two different tests and both need answering.

Read the trend line rather than the current rules. Information exchange between tax authorities has expanded every year for a decade, beneficial ownership registers are more accessible to competent authorities than they were, and crypto-specific reporting frameworks now bring exchange and custodian data into the same channels. A structure that only works if nobody looks is not a structure. Build the version that survives being examined.

The BVI is usually compared against the Cayman Islands, and for token projects the honest distinction is governance rather than tax. Cayman's foundation company is a well-understood ownerless vehicle, which is why projects that need a treasury genuinely outside founder control often choose it. Cayman also has deeper fund infrastructure, which matters if the project has an investment arm. The BVI is typically faster and cheaper to establish and maintain, and its business company is entirely adequate where an ownerless vehicle is not required. Many mature projects use both: a Cayman foundation for the protocol and a BVI company for a specific operating or issuance function.

Against the Gulf financial centres, the trade is credibility-through-supervision versus flexibility. A supervised entity in a Gulf financial free zone with a substantive local presence generally banks more easily and reads better to institutional counterparties, because there is a named regulator standing behind it. It also costs materially more, requires real local office space and staff, and takes longer to establish. For a project that has revenue and needs institutional relationships, that trade is often worth making. For a pre-launch project, it usually is not.

Against European authorisation, the comparison is about market access. A European authorisation buys the right to serve European customers and passporting across the bloc; that is a decisive advantage for a business whose users are in Europe and an expensive irrelevance for one whose users are not. Authorisation carries capital requirements, governance and reporting obligations, local staffing and a timeline measured in quarters. Choose it when the market you are serving requires it, not for its reputational halo.

Against Singapore, Hong Kong and similar hubs, the question is usually where the team is and where the investors are. Both offer strong legal systems and serious banking, both apply real substance expectations, and both have specific licensing regimes for the supervised activities. If the team is already there, the local entity is frequently the right operating company, potentially alongside a BVI or Cayman issuer.

And against doing nothing — running the project through a personal account or a company formed where the founder happens to live, with no analysis — the BVI wins comfortably. The most expensive structures we are asked to repair are not the wrong jurisdiction. They are the absence of any structure at all, discovered at the moment a bank freezes an account or an exchange requests corporate documents that do not exist.

Sequence decides cost. The same structure built in the right order costs a fraction of the same structure built in the wrong order, because migrations, restatements and retrospective agreements are far more expensive than getting it right first.

Start with the activity analysis. Write down precisely what the company will do, who holds assets at each step, who the counterparties are, and where the users are. That document drives the perimeter analysis, the token characterisation, the residence analysis and the banking narrative. Almost every later question is answered from it, and almost every expensive mistake comes from skipping it.

Second, resolve the founders' personal positions. Corporate residence, controlled foreign company exposure and the taxability of any token allocation to the founders depend on where those individuals are resident. Deciding the entity before the people is backwards, and it is the sequence that produces structures that have to be rebuilt in year two.

Third, form the entities and paper the relationships in the same pass: memorandum and articles fit for the actual capital and governance arrangement, directors appointed with a genuine understanding of their role, intercompany services and IP licence agreements drafted and signed, transfer pricing basis documented. Fourth, obtain the legal opinion or perimeter analysis before it is needed, because it is required by banks, exchanges and market makers and the timeline is measured in weeks.

Fifth, open banking before there is money to move. This is the step founders defer and the one that determines whether the launch is smooth. Sixth, complete any registration requirement if the perimeter analysis found one. Only then run the token event, listing or product launch. And from day one, maintain: annual fees, statutory registers, substance reporting, financial records, board minutes. Companies struck off for unpaid annual fees at the exact moment an exchange requests a certificate of good standing is a genuinely common and entirely avoidable failure.

The engagement opens with a scoping call and a written analysis rather than a formation order. We take your activity description, the founders' residence positions, where your users are, your funding history and your launch timetable, and we come back with the structure we would actually build, what it costs to establish and to maintain, the banking outcome we consider realistic, the timeline, and the specific risks we would not take on. Where the BVI is not the right answer, we say so and set out what is.

On formation, we handle the BVI business company end to end through licensed agents: name reservation, memorandum and articles drafted for your capital and governance arrangement rather than a template, director and shareholder appointments, registered agent and registered office, statutory registers, and a corporate pack — certified incorporation documents, certificate of good standing, registers, certified passport and address verification for controllers — assembled in the format banks and exchanges request. Where a group is required, we form the operating entity alongside it and draft the services agreement, IP licence and transfer pricing basis so the two entities relate to each other on paper as well as in the deck.

On the regulatory side, we run the VASP perimeter analysis against your real product design and, where the answer is uncertain, coordinate qualified BVI counsel for a written opinion or a token characterisation opinion in the form exchanges and custodians accept. Where registration is required, we manage the application file and the regulator's question rounds with the licensed agent. Where product design choices would move you across the perimeter line, we flag them while they are still cheap to change.

On banking, we position the application across a network of more than one hundred and twenty banking and payment institutions. That means writing the business description for a compliance reader, building the source-of-funds and source-of-wealth documentation, modelling projected flows by corridor and counterparty type, and matching the file to institutions whose stated appetite fits your activity — then building a layered architecture with an operating account, a conversion route and a contingency relationship rather than a single point of failure. We do not guarantee accounts; every institution decides independently.

After launch we stay on the maintenance that structures die from neglecting: annual government fees and registered agent renewals, statutory register upkeep, economic substance assessment and annual reporting, board minutes and management substance evidence, bank relationship reviews, and refreshed documentation when an exchange, custodian or counterparty asks for it. And where the project moves into listing, liquidity or institutional access, the same team runs that work against the structure we built. Send us the shape of what you are building and we will tell you plainly whether we can help.