The Canary Islands Special Zone attracts two very different kinds of enquiry. The first comes from groups that already have people, contracts and equipment somewhere, and are deciding where the next increment of activity should sit. The second comes from groups that have a tax charge and are looking for somewhere to book it. The regime was designed to serve the first and to defeat the second, and the distinction is not a matter of interpretation; it is written into the conditions of the authorisation. What follows is a sorting exercise rather than an argument in favour. The cost of discovering late that a business model does not fit is borne entirely by the group, and it is rarely limited to the tax at stake.
What the regime actually is
ZEC is a regional aid regime authorised by the European Commission. It sits inside Spain and inside the European Union, which means the entity is a Spanish resident company with ordinary reporting obligations, ordinary exposure to anti-abuse doctrine, and access to directives and treaties on the same conditions as any other Spanish company. It offers a reduced rate of corporate income tax, conditional on minimum investment, the creation of employment, effective activity carried on in the islands, and registration in the Registro Oficial de Entidades ZEC.
Read those conditions as a description of a business rather than as a checklist. Investment, employment and effective activity are operational facts. They are observable by a tax inspector, by a counterparty conducting diligence, and by an acquirer. A group that can satisfy them only through paperwork has not satisfied them. This is the single most useful thing a CFO can internalise before spending money on advice: the regime does not reward the structure, it rewards the activity, and it withdraws when the activity is absent.
Models that fit
Technology and digital services businesses with an engineering or product team resident in the islands fit well. The people are the productive asset, the asset is physically located where the rate applies, and the revenue is attributable to what those people build. The same logic applies to audiovisual production, where crews, studios, post-production and the associated intellectual property can be genuinely located, and where the local labour requirement is met by the nature of the work rather than in spite of it.
Logistics and trading operations with real physical presence fit for a different reason. Warehousing, handling, consolidation and onward distribution are inherently located activities. Where goods physically move through the islands and the local entity bears inventory risk and contracts in its own name, the profit attribution follows naturally. Research and development fits where the laboratory, the equipment and the researchers are there, not where a licence agreement says the intellectual property is. Business services with local payroll, whether shared-service functions, technical support or specialised professional teams, fit where headcount, supervision and decision-making are all in the same place.
The common feature is not sector. It is that in each case the enterprise would be materially different if the islands operation disappeared. That is the whole test, and it is worth stating plainly.
Models that do not fit
A pure holding company with no operating activity does not fit. It creates no employment of substance, requires no meaningful investment, and carries out no effective activity; it is a title-holding vehicle. There are jurisdictions and vehicles designed for that function, with mature regimes and settled administrative practice, and the Netherlands is one of them. The Dutch participation exemption is mandatory and symmetric, and it exists precisely to make holding activity coherent, subject to the minimum holding requirement and to the motive, subject-to-tax and asset tests. Attempting to place a holding function inside a regional aid regime designed for operating businesses is a category error before it is a tax risk.
Invoicing from the islands for services actually performed somewhere else does not fit. This is the most common failure pattern and the most expensive. The contract, the invoice and the bank account are in the islands; the engineers, the account managers and the decision-makers are in Madrid, Munich or Amsterdam. Transfer pricing analysis resolves this quickly and unfavourably, because the functional analysis follows people and risk, not paper. Article 8b of the Dutch corporate income tax act applies the arm’s length standard and its documentation requirement without threshold, and comparable rules apply in every jurisdiction on the other side of the invoice.
Structures resting on a single part-time director do not fit. A director who spends a limited part of the year on the islands and holds a nominal portfolio of directorships cannot supply the management substance the regime assumes. Nor can a service provider supplying the same individual to a long list of entities. The employment condition is not satisfied by a title, and the effective activity condition is not satisfied by attendance at board meetings.
The relocation test
There is a mental test that settles most cases in a single sentence. Imagine moving the entire islands team out tomorrow, to anywhere else the group operates. Would the business continue to function exactly as before?
If moving the team out of the islands would change nothing about how the group earns its money, the islands were never earning it.
If the answer is that nothing would change, the activity is not located there in any sense that matters, and the structure is decorative. If the answer is that delivery would stall, that clients would notice, that production would stop or that a capability would have to be rebuilt from scratch, the activity is real and the regime is doing what it was authorised to do. The test is useful because it is not a legal test at all. It is a business question, and management can answer it honestly without advice.
It also has a corollary worth applying at the design stage: a business model that requires the islands team to be small, cheap and interchangeable in order for the numbers to work is a model that will fail the test at the first examination.
Transfer pricing is where the answer is settled
Where a ZEC entity sits inside a wider group, the reduced rate does not determine the outcome; the allocation of profit does. Intercompany pricing must be at arm’s length and documented, and the documentation must describe functions, assets and risks as they actually are. Master File and Local File obligations arise from fifty million euro of consolidated turnover, country-by-country reporting from seven hundred and fifty million, and the Pillar Two minimum of fifteen per cent applies from the same seven hundred and fifty million threshold. Larger groups should therefore model the interaction between a reduced statutory rate and a global minimum before assuming that any benefit accrues at all.
The practical consequence is that a group cannot obtain a better result by asserting that the islands entity is an entrepreneur while operating it as a service provider. If it performs routine functions under instruction, it is remunerated as such, and the reduced rate applies to a modest margin. Groups that are disappointed by this were usually sold the rate rather than the model.
The Dutch layer above
Many international groups will hold a Spanish operating entity beneath a Dutch platform, and that combination is coherent when each layer does its own job. The Dutch entity performs the holding, financing and governance functions, subject to genuine substance requirements and to board decision-making that actually occurs in the Netherlands. Dutch corporate income tax applies at 25.8 per cent in the upper bracket, with a reduced rate in the first bracket. Dividend withholding tax is 15 per cent in general, with treaty reductions and intra-EU exemptions, all subject to anti-abuse conditions, and the conditional withholding tax on interest and royalties applies to payments to low-taxed or listed jurisdictions.
Two points deserve emphasis. First, Dutch ruling policy since July 2019 requires real economic nexus and declines to confirm arrangements whose decisive motive is tax saving. A group that would not survive that standard in the Netherlands is unlikely to survive the equivalent scrutiny in Spain. Second, where the structure is debt-funded, the ATAD earnings-stripping limitation on interest deduction applies as a percentage of fiscal EBITDA above a minimum threshold, and the parameters have changed over time. Reporting obligations under DAC6 may also arise for cross-border arrangements bearing the relevant hallmarks, with the obligation falling on the intermediary or, failing that, on the taxpayer.
A caution on personal regimes
The Spanish inbound expatriate regime is frequently raised in the same conversation and should be kept separate. It is optional, requires an absence of prior Spanish residence for the relevant period and a qualifying reason for the move, and taxes employment income at a fixed rate up to a threshold and at a higher rate above it. It is personal and it does not alter the taxation of companies. It can make relocating key people workable; it cannot make an unsuitable corporate model suitable, and conflating the two produces structures that satisfy neither set of conditions.
What this means in practice
The models that survive examination are the ones a group would have built anyway, placed where the reduced rate happens to apply. The ones that fail are the ones invented to capture the rate. That distinction is visible at the outset to anyone willing to apply the relocation test honestly, and it is far cheaper to apply before incorporation than after the first audit. Where the answer is that the model does not fit, the more sensible response is usually a conventional structure with sound governance and no regional aid element at all. None of the above is a recommendation to adopt any particular structure; each case turns on its own facts and requires advice on both sides of the border.
Montclare structures and operates Dutch and cross-border platforms for international groups. Our services are set out on our services page.
This article is informational and does not constitute tax, legal or investment advice. Each engagement is subject to scope and applicable regulation.