Few tax measures have generated as much preparatory work relative to the number of groups they actually bind as the global minimum tax. Pillar Two is now part of Dutch law, the compliance apparatus around it is substantial, and a market of advisers and software vendors has grown up to service it. It is also, for most internationally active mid-market groups, entirely irrelevant. Establishing which of those two positions applies is a short exercise, and worth doing precisely, because the cost of assuming wrongly runs in both directions.
What the rules are trying to achieve
Pillar Two is the outcome of the OECD and G20 Inclusive Framework’s work on a coordinated floor under corporate taxation, published as the global anti-base erosion model rules of December 2021. The European Union implemented it through Council Directive (EU) 2022/2523 of 14 December 2022, and the Netherlands transposed that directive into domestic law through the Minimum Tax Act 2024 (Wet minimumbelasting 2024). The three layers say much the same thing, but only the last of them imposes tax in the Netherlands, and it is the one against which a Dutch position has to be checked. The mechanics differ in detail across jurisdictions, but the objective is uniform: that the profits of a large multinational group bear tax at a minimum rate of 15 per cent in every jurisdiction where they arise.
The structural point is that Pillar Two is not an extension of the corporate income tax. It sits above the existing systems. Each jurisdiction continues to tax profit under its own rules; Pillar Two then measures the result, and if the outcome in a given jurisdiction falls short of the floor, a top-up is collected. The base is not taxable profit under national law but a separately defined figure derived from consolidated financial accounting, adjusted by a long list of prescribed items.
Who is actually in scope
The threshold is consolidated revenue of EUR 750 million or more in the consolidated financial statements of the ultimate parent entity, reached in at least two of the four financial years immediately preceding the year being tested. That condition carries as much weight as the figure: one year above the line does not bring a group into the regime, and one year below it does not take a group out. The test does not require the group to be international: a group whose constituent entities are all established in the Netherlands is caught on the same terms, and where its Dutch effective rate falls below the floor the top-up is collected as a domestic minimum top-up tax. Three features of the test are regularly misread.
The first is that it is a revenue test and not a profit test. A low-margin distributor can cross the threshold while barely profitable; a high-margin services business can be far more valuable and never approach it. The second is that it is tested at group level, which makes the size of the Dutch entity irrelevant. A small Dutch holding or finance company inside a large foreign-parented group is in scope; a large Dutch operating company inside a group that consolidates below the threshold is not. The third is that it looks backwards. Because the test refers to preceding financial years, a group does not decide to enter the regime. It discovers that it already has, on the basis of accounts that were closed some time ago.
Certain entities are excluded from the charging provisions, among them government entities, international organisations, non-profit organizations, pension funds, and investment funds and real estate investment vehicles that sit at the top of an ownership chain. Joint ventures and acquisitions are the two most common reasons a group crosses the line without having planned to, and the act sets out separately how the revenue threshold is applied where groups merge or demerge.
How the calculation works in essence
The unit of analysis is the jurisdiction, not the entity. For each jurisdiction in which the group has constituent entities, two aggregates are computed: a measure of income derived from the accounting result under the group’s consolidation standard, adjusted for items such as excluded dividends and excluded equity gains, and a measure of covered taxes, comprising current tax and certain deferred tax movements. Dividing the second by the first produces the effective tax rate for that jurisdiction.
If the effective rate is below 15 per cent, the shortfall is applied to the jurisdiction’s profit after a substance-based carve-out calculated by reference to payroll costs and the carrying value of tangible assets located there. The result is the top-up tax. The order in which that amount is collected then follows a defined sequence: a qualified domestic minimum top-up tax in the jurisdiction concerned takes priority, followed by an income inclusion rule applied at the level of the parent, with the undertaxed profits rule operating as a backstop where neither of the first two captures the amount. The Netherlands has legislated a domestic minimum top-up tax alongside the other two charges, which means that where a Dutch top-up arises it is collected in the Netherlands rather than abroad.
Two filters keep modest footprints out of the charge, and both are conditional. The de minimis exclusion can be elected for a jurisdiction where average qualifying revenue there is below EUR 10 million and average qualifying income is below EUR 1 million, each measured across the financial year and the two preceding ones. Transitional safe harbours, based largely on country-by-country reporting data, allow a group to switch off the full calculation for jurisdictions that plainly clear the floor, but they expire rather than settling into the system: the country-by-country version is available only for financial years that begin on or before 31 December 2026 and end before 1 July 2028, and the transitional safe harbour for the undertaxed profits rule ran out with financial years beginning after 31 December 2025. This is where most of the early workload sits, and where the quality of a group’s existing reporting data becomes immediately visible.
What it means for a group with a Dutch presence
The Dutch corporate income tax rate is 25.8 per cent in the upper bracket, with 19 per cent on the first EUR 200,000 of taxable profit. On headline rates alone, a Dutch jurisdictional effective rate below 15 per cent is unusual. But the Pillar Two effective rate is not the statutory rate, and the divergence between them is where the work lies. Loss carryforwards, deferred tax positions, non-deductible expenditure and the interaction between accounting and fiscal profit can all move the ratio. The participation exemption, which relieves qualifying dividends and capital gains, is broadly neutral in this context because the corresponding income is generally excluded from the Pillar Two income base as well as from the tax numerator; groups nonetheless need to demonstrate that treatment rather than assume it.
For an in-scope group, the practical Dutch consequences are administrative before they are financial: the filing of the top-up tax information return, or where that return is filed abroad and reaches the Dutch authorities through exchange of information, a notification saying which group entity files it and where, followed by the construction of a data set reporting covered taxes and deferred tax movements by jurisdiction on a basis reconcilable to the consolidated accounts. Most groups find that their finance systems were never designed to produce that view, and that rebuilding them takes longer than the calculation itself.
Why most mid-market groups should stop here
This is the part least often stated plainly. If a group consolidates well below EUR 750 million across the financial years the test looks back to, and has no realistic path to that figure, Pillar Two imposes no obligation on it. There is nothing to file, nothing to notify, no software to license and no restructuring to consider. Groups in that position are periodically sold readiness assessments for a regime that does not apply to them; the appropriate response is to run the threshold test against the consolidated accounts and record the conclusion.
The position changes for groups that are within sight of the threshold, whether organically or through a pending acquisition. Because entry depends on revenue already reported, the useful window for analysis is before the line is crossed, not after. What matters at that stage is not tax planning but readiness: knowing which entities and jurisdictions will be in the perimeter, whether the safe harbours are likely to be available for the years in which they still run, and whether the group’s reporting can produce the underlying figures at all. Those are systems questions, and they have long lead times.
The transfer pricing dimension
Pillar Two measures profit and tax by jurisdiction. Transfer pricing determines where the profit is recognized. The two are therefore mechanically linked: a change in intercompany pricing shifts the numerator and denominator of a jurisdictional effective rate, and can move a jurisdiction across the floor in either direction.
Pillar Two does not ask where a group is headquartered. It asks where the profit landed, and at what rate it was taxed. Transfer pricing answers the first half of that question, and it answers it whether or not the group has given the matter any thought.
Two consequences follow. The first is that transfer pricing documentation acquires an audience it did not previously have. The Dutch arm’s length principle and the associated documentation duty under article 8b of the corporate income tax act apply without a size threshold, with master file and local file obligations for groups that reached EUR 50 million of consolidated revenue in the immediately preceding financial year and country-by-country reporting for those that reached EUR 750 million in that same preceding year, a single-year test rather than the two-out-of-four test that governs Pillar Two itself. Since the safe harbours draw on country-by-country data, those figures now carry consequences beyond risk assessment, and inconsistencies between that report, the local files and the statutory accounts are harder to explain.
The second is that the allocation of profit needs to be defensible on its own terms. Jurisdictional outcomes driven by intangible ownership, intra-group financing or centralized service arrangements attract attention under both regimes at once. The Dutch approach to intangibles follows the development, enhancement, maintenance, protection and exploitation framework, so where royalty income sits must be supported by where the relevant functions are actually performed. Intra-group financing faces its own constraints, including the earnings stripping limitation on interest deduction, set at the higher of a percentage of fiscal EBITDA and a fixed minimum amount, and the conditional withholding tax, which reaches interest, royalties and dividends paid to a related party established in a low-taxed or listed jurisdiction, and also a related party elsewhere where it is entitled to the payment with the main purpose, or one of the main purposes, of avoiding that tax for another party and the arrangement is artificial. Since July 2019, the Dutch ruling practice has required genuine economic nexus and does not accommodate arrangements whose sole or decisive motive is tax saving.
A defensible sequence of steps
For a group approaching or within the regime, the order of work is reasonably settled. Confirm the threshold test against the consolidated statements for the relevant look-back period, and document the conclusion. Identify the ultimate parent, the filing position and the full list of constituent entities by jurisdiction. Test the available safe harbours against existing country-by-country data before committing to a full calculation. Establish whether the finance function can produce covered taxes and deferred tax movements per jurisdiction, and if it cannot, treat that as the project. Then review whether the transfer pricing policy and its documentation support the geographic profile of profit the group is about to report to multiple authorities in a single dataset.
Groups establishing or reorganizing a European platform have an advantage here, in that the perimeter and the reporting architecture can be designed together rather than reconciled afterwards; the considerations are set out in our note on building a Dutch holding structure. For everyone else the summary is short. Confirm the threshold. If you are outside it, say so in writing and move on. If you are inside it, or heading there, the work is data and documentation rather than structuring, and it starts earlier than most groups expect.
Montclare structures and operates Dutch and cross-border holding platforms for international groups. Our services are set out on our services page.
This article is informational and does not constitute tax advice. Each engagement is subject to scope and applicable regulation.