Pillar Two and the 15% Global Minimum Tax: What It Means for Crypto Groups Using Offshore Structures

Every crypto group with a Cayman, BVI, or UAE free-zone entity built its structure around one number: a 0% headline rate. That number was the load-bearing assumption behind the IP-holding company, the offshore licensing entity, and the place where mobile profit landed.

Pillar Two — the OECD's 15% global minimum tax — quietly rewrites that maths. If the group clears EUR 750m consolidated revenue, the zero does not stay a zero. It becomes a top-up tax payable somewhere — and the only open question is which jurisdiction collects it.

This guide sets out the GloBE Exposure Test — the three questions that decide whether your low-tax structure now triggers a top-up — and the structuring response for groups that are in scope. We name OECD rules, EU directives, and jurisdictions freely; we do not name banks or custodians, because this is a tax-structuring problem, not a banking one.

For two decades, the offshore crypto playbook was simple: park the intellectual property, the treasury, and the licensing income in a jurisdiction with a 0% corporate rate, run substance somewhere cheap, and let the headline rate do the work. The structure was legal, transparent, and — until recently — final.

Pillar Two breaks the finality. It does not ask what a jurisdiction's headline rate is. It asks what the group actually paid there, expressed as a jurisdictional effective tax rate, and tops the figure up to 15% wherever it falls short. A 0% Cayman entity inside a large group is now the most exposed part of the structure, not the cleverest.

The shift matters most for crypto because the sector concentrated high, mobile profit in low-substance entities — exactly the profile Pillar Two was designed to catch. The substance-based carve-out that shelters a factory or a workforce does very little for an IP-holding shell with three directors and a registered office.

Pillar Two is the second pillar of the OECD/G20 BEPS project. Its core mechanism is the Global Anti-Base Erosion rules — GloBE — set out in the OECD GloBE Model Rules¹^[1].

GloBE imposes a 15% minimum effective tax rate on the profits of large multinational groups, computed jurisdiction by jurisdiction. Where a group's profit in a jurisdiction is taxed below 15%, a top-up tax brings it up to the floor. The top-up is collected through three interlocking charging rules.

• IIR — Income Inclusion Rule: the parent jurisdiction taxes the low-taxed profit of its subsidiaries. The primary charging rule.

• UTPR — Undertaxed Profits Rule: a backstop that lets other jurisdictions deny deductions or make an equivalent adjustment when no IIR applies above.

• QDMTT — Qualified Domestic Minimum Top-up Tax: the low-tax jurisdiction itself charges the top-up domestically, keeping the revenue at home rather than ceding it to the parent country.

The EU transposed GloBE through the Minimum Tax Directive (Council Directive (EU) 2022/2523)²^[2], binding every EU member state. Many non-EU jurisdictions — including several offshore centres — have introduced their own QDMTT to collect the top-up locally rather than lose it abroad.

Whether your structure survives Pillar Two reduces to three sequential questions. Fail to clear the first and you are out of scope entirely. Clear it, and the second and third decide how much top-up is payable and where.

• Question 1 — Scope: does the group clear EUR 750m consolidated revenue? If not, Pillar Two does not apply and your 0% structure is untouched.

• Question 2 — Rate: is the jurisdictional GloBE effective tax rate below 15%? This is not the headline rate — it is a specific GloBE computation.

• Question 3 — Substance: after the substance-based income exclusion, is there any low-taxed excess profit left? For crypto groups, almost always yes.

Pillar Two applies only to a group whose consolidated revenue reaches EUR 750m in at least two of the four fiscal years preceding the tested year. The figure is taken from the ultimate parent entity's consolidated financial statements — not from any single subsidiary.

The threshold is deliberately aligned with country-by-country reporting under BEPS Action 13, so groups already filing a CbCR are usually the same population caught by GloBE. Below EUR 750m, a 0% offshore structure remains fully effective — scope is the first and best line of defence.

Once in scope, every entity in the consolidation is a constituent entity — including the Cayman IP-holder, the BVI licensing company, and the UAE free-zone operator. GloBE then groups constituent entities by jurisdiction and tests each jurisdictional blend against the 15% floor.

This jurisdictional blending matters. A group cannot dilute a 0% Cayman entity by pointing to tax paid in Germany or Ireland — each jurisdiction is tested separately. The low-taxed jurisdiction stands alone, which is precisely why offshore shells are exposed.

GloBE does not use the statutory headline rate. It computes a jurisdictional effective tax rate — the GloBE ETR — as covered taxes divided by GloBE income, aggregated across all constituent entities in that jurisdiction.

• Covered taxes: corporate income taxes and certain equivalents actually borne — not withholding the group can credit elsewhere, and not notional credits.

• GloBE income: financial-accounting profit, adjusted by a defined set of GloBE add-backs and exclusions, not local taxable profit.

The consequence for a 0% jurisdiction is blunt: covered taxes are zero, so the GloBE ETR is 0%, and the full 15% gap is exposed to top-up. There is no headline-rate engineering that changes a numerator of zero.

A 9% UAE free-zone entity is more nuanced. Qualifying free zone income taxed at 0% produces a 0% GloBE ETR on that slice; income outside the qualifying perimeter taxed at 9% still sits below 15%, leaving a 6-point top-up before any carve-out. Even the 9% regime does not reach the floor.

GloBE does not tax all low-taxed profit. It carves out a substance-based income exclusion — the SBIE — under the OECD substance-based income exclusion guidance³^[3]. The carve-out shields a return on real activity from the top-up.

The SBIE excludes a percentage of two things from the top-up base: eligible payroll costs and the carrying value of tangible assets in the jurisdiction. During the transition the carve-out percentages step down annually toward a long-run 5% of payroll and 5% of tangible assets.

• Payroll carve-out: a percentage of wages, salaries and other costs for employees performing activities in the jurisdiction. Contractors and remote staff elsewhere do not count.

• Tangible-asset carve-out: a percentage of the carrying value of property, plant and equipment located in the jurisdiction. IP and other intangibles are excluded.

Here is why crypto groups have little carve-out: their value sits in intangibles — code, brand, licences — and in a small mobile workforce, not in payroll-heavy operations or tangible plant. A Cayman IP-holder with no employees and no buildings carves out almost nothing. The exclusion is real but it barely touches the crypto profile.

Once a top-up is due, which jurisdiction collects it depends on the charging order set out in the OECD Administrative Guidance on the GloBE Rules⁴^[4]. The three mechanisms form a priority cascade: a QDMTT is applied first, then the IIR, with the UTPR as the final backstop.

The strategic point for offshore crypto groups: if your low-tax jurisdiction adopts a QDMTT, the top-up is collected there — and the IIR and UTPR fall away. The money is paid either way, but a QDMTT keeps it in a jurisdiction you control rather than handing it to a parent-country tax authority.

Pillar Two was engineered to catch exactly the structure most crypto groups already run: high-margin profit booked in a low-substance, low-tax entity. The features that made offshore attractive are the same features that maximise GloBE exposure.

• Profit is mobile: licensing fees, treasury yield and IP royalties can be booked anywhere, so groups booked them at 0%. GloBE now tops that 0% up to 15%.

• Substance is thin: the SBIE rewards payroll and tangible assets; crypto value is intangible, so the carve-out shields little of the profit.

• The ETR is genuinely zero: unlike a 12.5% jurisdiction that needs only a small top-up, a true 0% entity faces the full 15-point gap on its excess profit.

The verdict is uncomfortable but clear: the offshore IP-holding shell is the single most exposed entity in a large crypto group under Pillar Two. The structure was optimised for a world where headline rate was the only variable — and GloBE retired that world.

Walk the logic for a 0% Cayman IP-holder inside an in-scope group. The entity books substantial licensing profit and pays no local tax. Its covered taxes are zero, so its GloBE ETR is 0% — a full 15-point shortfall against the floor.

The SBIE then carves out a small slice — but with no employees and no tangible assets in Cayman, the carve-out is near zero. Almost the entire profit remains low-taxed excess profit subject to top-up.

Finally the charging cascade allocates the 15% top-up. If Cayman has adopted a QDMTT, Cayman collects it. If not, the parent jurisdiction's IIR collects it; failing that, the UTPR sweeps it up across the group. The 0% is paid as 15% somewhere — the only choice left is which authority receives it.

The same logic runs for a BVI licensing company and, with one wrinkle, for a UAE free-zone operator: the UAE's 9% headline still produces a 0% GloBE ETR on qualifying free zone income, so the free-zone exemption does not save it from a top-up once the group is in scope.

The table below shows the direction of travel for three common crypto archetypes. These are illustrative effective-rate outcomes, not a substitute for a modelled computation — but they show where the top-up bites and where it does not.

The pattern is consistent: the more real the substance and the higher the existing rate, the smaller the Pillar Two impact. An EEA operating company with a genuine workforce and 12.5–25% CIT is largely unaffected; a 0% offshore shell absorbs the full top-up.

If the group is in scope, the question is no longer whether to pay 15% but where it lands and how to avoid paying twice. There are four practical levers.

If a top-up is inevitable, prefer a jurisdiction with a Qualified Domestic Minimum Top-up Tax so the revenue stays at source and pre-empts a parent-country IIR. This is controlled top-up versus ceded top-up — the cash cost is the same, but the relationship and certainty are better.

Where you keep a low-tax entity, give it genuine payroll and tangible assets so the SBIE actually shields profit. Substance that was once optional is now a tax shield — but only real, local, demonstrable substance counts.

For many groups the cleaner answer is to move the IP onshore into a jurisdiction with a competitive but real CIT — often paired with a patent / IP box regime that is GloBE-compatible. A 12.5% effective rate with substance can beat a 0% rate topped up to 15% with friction.

Some offshore entities now add cost without saving tax: they trigger GloBE compliance and a top-up while delivering zero net benefit. For an in-scope group, a 0% shell with no substance is often worth collapsing — the headline saving is illusory once the top-up is priced in.

In-scope groups must file a standardised GloBE Information Return (Pillar Two)⁵^[5] — the GIR — reporting jurisdictional ETR computations, top-up calculations, and charging-rule allocations for every constituent entity.

The GIR is detailed: it captures covered taxes, GloBE income, SBIE inputs, and the top-up allocation across mechanisms. Groups generally have an extended deadline of up to 15 months after the fiscal year-end for the first return, with 18 months for the transition year.

The substantive rules sit within the broader OECD OECD BEPS project⁶^[6] framework, so groups already managing CbCR and transfer-pricing documentation should fold GloBE data collection into the same cycle rather than building a parallel process.

Yes — Pillar Two applies to any multinational group, crypto or not, that clears EUR 750m consolidated revenue. There is no crypto exemption. Smaller groups are out of scope, but large crypto groups with 0% offshore entities are squarely in the population the rules target.

The threshold is EUR 750m of consolidated group revenue in at least two of the four preceding fiscal years, measured from the ultimate parent's consolidated accounts. It mirrors the country-by-country reporting threshold under BEPS Action 13.

For a group below EUR 750m, yes — Pillar Two does not apply and the 0% rate stands. For an in-scope group, a 0% jurisdiction produces a 0% GloBE ETR and a full 15-point top-up on excess profit, collected by a QDMTT, IIR, or UTPR. The 0% is effectively paid as 15% somewhere.

A Qualified Domestic Minimum Top-up Tax is a top-up charged by the low-tax jurisdiction itself, bringing local profit to the 15% floor domestically. Because a QDMTT applies first in the charging cascade, it keeps the revenue at home and pre-empts a parent-country IIR. Many offshore centres have adopted one for exactly this reason.

No. Qualifying free zone income taxed at 0% gives a 0% GloBE ETR, and the 9% headline on other income still sits below 15%. For an in-scope group the free-zone exemption is ignored by GloBE and the profit is topped up like any other low-taxed jurisdiction.

• The CFC Map: how controlled-foreign-company rules attribute offshore profit home — the regime that overlaps with Pillar Two.

• Crypto Transfer Pricing: BEPS, Arm's Length: pricing intra-group flows defensibly, the data layer Pillar Two's ETR computation also relies on.

• Multi-Jurisdiction Crypto Structures: designing the cross-border entity map that GloBE now stress-tests jurisdiction by jurisdiction.

• Cyprus IP Box for Crypto IP: an onshore, GloBE-compatible home for crypto IP when the 0% offshore shell stops paying off.

Pillar Two did not ban the offshore structure — it priced it. For groups below EUR 750m nothing changes; for those above it, the 0% headline is now a 15% liability waiting for a collector.

The groups that win the next phase are the ones that run the GloBE exposure test early, choose where the top-up lands on their own terms, and retire the structures that no longer pay — before a tax authority models the number for them.