February 24, 2026 — The most consequential international tax reform of the past forty years entered its first complete fiscal year on January 1, 2024 in approximately thirty jurisdictions, principally the European Union Member States together with the United Kingdom, Norway, Australia, Japan, South Korea, and Canada. Twelve months later, the second wave of implementation arrived in countries that had previously deferred. The United Kingdom, Germany, France, Italy, Spain, the Netherlands, Ireland, Switzerland, Australia, Japan, South Korea, Canada, and others are now operating Income Inclusion Rules that compel ultimate parent entities to top up the effective tax rate of their international operations to fifteen percent. The Undertaxed Profits Rule, a backstop mechanism that allocates top-up tax to other group members when the parent jurisdiction does not apply Income Inclusion, is in force in several jurisdictions and approaches generalised application from January 2025. The Qualified Domestic Minimum Top-up Tax, a sovereign protective measure that allows source jurisdictions to collect their own top-up before another country claims it, has been adopted by virtually every jurisdiction with a corporate tax rate below fifteen percent — including, most consequentially, the United Arab Emirates, Switzerland, Hong Kong, Bermuda, Cayman Islands, and Bahrain.

The architectural assumptions on which forty years of multinational tax planning were built have been quietly invalidated. This article is not a primer on the rules, of which there are now several thousand pages of OECD guidance, country implementation legislation, and administrative interpretation. It is a field report on what tax functions inside multinational groups have actually changed, what they have not changed, what surprised them, and what the second-order consequences look like as the regime stabilises.

The fifteen percent question was always the wrong question

The architecture of low-tax jurisdictions did not collapse on January 1, 2024. The conventional pre-implementation narrative was that Pillar Two would render Bermuda, the Cayman Islands, the British Virgin Islands, and similar zero-rate jurisdictions instantly obsolete as holding company locations. That has not happened, and the reason it has not happened reveals more about the rule architecture than the original commentary did.

Consider a typical European-headquartered group with a Bermuda finance company. Pre-2024, the Bermuda entity earned interest income, paid no corporate tax, and distributed dividends upstream that were largely exempt under participation exemption regimes. Pillar Two should have imposed a fifteen percent top-up. In practice, Bermuda introduced a Qualified Domestic Minimum Top-up Tax effective January 1, 2025. The fifteen percent is now levied at source, in Bermuda, before any other jurisdiction can claim it. The cash leaves the group, but it leaves in Bermuda, not in the United Kingdom or Germany.

The structural consequence is that low-tax jurisdictions remain viable as holding locations because the QDMTT mechanism allows them to retain the additional revenue. What changed is the cost: a Bermuda finance company that paid zero tax on intercompany interest now pays fifteen percent on the same flows. The marginal effective tax rate on activities placed in QDMTT jurisdictions has converged on fifteen percent regardless of the headline rate.

The competitive implications are most significant for jurisdictions that fall between the QDMTT floor and the historical European range. Switzerland, where cantonal effective rates ranged from twelve percent to twenty-one percent, has implemented a federal QDMTT that levies the difference between the cantonal rate and fifteen percent for in-scope groups. Ireland, with its twelve and a half percent corporate rate, has implemented a domestic top-up that brings in-scope groups to fifteen percent. The Irish Revenue Commissioners issued detailed guidance through 2024 and 2025 on the interaction with the existing Knowledge Development Box and research and development credits, with the practical effect that the post-Pillar Two effective rate for in-scope multinationals using these regimes lands at approximately the same fifteen percent as a group in a higher-tax jurisdiction.

The fifteen percent is now the global floor, applied near-uniformly. The question is no longer whether the group can achieve a sub-fifteen percent rate. The question is where the additional revenue is collected, and that question is far from settled.

The architecture that disappeared

What did not survive the transition were structures predicated on stranding profit in low-tax jurisdictions without operational substance. The classic Bermuda intellectual property holding company — formed to receive royalties from operating subsidiaries, retain the income at zero percent, and accumulate reserves — has effectively disappeared as a forward-looking design. Groups that maintained such structures into 2023 have either migrated the IP to onshore locations during 2024 or absorbed the QDMTT cost as an ongoing compliance.

The migrations have been more complex than the commentary anticipated. Moving intellectual property between jurisdictions triggers exit charges in the originating country, which under most regimes are computed on the fair market value of the IP at the date of transfer. Pre-Pillar Two, the calculus on whether to migrate IP from Bermuda to the Netherlands or Switzerland depended on the comparison between the discounted value of future Bermuda zero-rate accumulation and the immediate exit charge plus future taxation in the destination. Post-Pillar Two, the Bermuda accumulation no longer escapes the fifteen percent QDMTT, so the comparison reduces to: pay fifteen percent at source in perpetuity, or pay the exit charge once and then operate under the destination jurisdiction's full corporate tax rate, often twenty-one to twenty-five percent.

The arithmetic has favoured migration in fewer cases than was widely predicted. Most groups that completed the analysis through 2024 chose to leave the Bermuda structure in place, accept the QDMTT, and avoid the exit charge entirely. The exception is where the IP holding company was inactive — generating no new IP, holding only legacy assets — in which case migration to a jurisdiction with a participation exemption on disposal proceeds offered a cleaner long-term position.

What has been more visibly affected is the family of pre-existing financing structures: the Luxembourg participating notes, the Belgian notional interest deductions, the Maltese refundable tax credit, the Cypriot notional interest deduction, the various preferential intellectual property regimes. Each of these mechanisms was constructed to deliver an effective tax rate well below the headline statutory rate. Each has been recalibrated. The Luxembourg authorities issued guidance in early 2024 confirming that participating notes remained in force but acknowledged that for in-scope groups the QDMTT would apply in addition. The Belgian notional interest deduction continues to operate but at an effective rate that, for in-scope groups, no longer drops below fifteen percent. The Maltese refundable credit, which historically delivered a five percent effective rate for foreign shareholders, is being analysed by the European Commission in connection with state aid considerations, with several Maltese law firms anticipating curtailment by 2027. The Knowledge Development Box in Ireland and analogous regimes in the Netherlands have been adjusted to operate within the Pillar Two framework but no longer deliver the headline incentive rates that drove pre-2024 placement decisions.

The aggregate effect is a flattening of the European tax landscape that took forty years of patient negotiation between jurisdictions to construct. The differentiation between Belgium, Luxembourg, the Netherlands, Ireland, Malta, and Cyprus that defined holding company strategy through the entirety of the post-1990 era has been compressed into a band perhaps two percentage points wide. Headline rates remain different. Effective rates for in-scope multinational groups have converged.

The substance pivot

If the rate differential has flattened, the substance differential has not. Pillar Two contains a Substance-Based Income Exclusion that reduces the top-up tax base by a percentage of payroll costs and tangible asset value located in the jurisdiction. The exclusion, set at five percent of qualifying payroll and five percent of qualifying tangible assets at full implementation but operating under transitional rates of nine point eight percent and seven point eight percent through 2032, creates a structural reward for substance that did not exist under prior regimes.

The consequence has been a measurable acceleration in the placement of operational substance in jurisdictions that had previously been used only for legal seat. Switzerland is the clearest case. Several major American technology and pharmaceutical groups that historically operated Swiss intellectual property holding entities with limited employee headcount have, through 2024 and 2025, expanded those operations materially. Zurich, Basel, and Zug together absorbed several thousand new tax-relevant headcount during the period from major US groups, the new positions concentrated in research and development, regional management, and customer-facing functions that increase qualifying payroll for Substance-Based Income Exclusion purposes.

The same pattern, at smaller scale, has appeared in Ireland, Singapore, and Luxembourg. Each jurisdiction has incentive to retain in-scope multinational groups whose tax revenue under the new regime depends partly on their continued substance presence. The competition for substance, as opposed to competition for legal seat, is the new field of multinational tax competition. It is being conducted not through statutory rate adjustments but through immigration policy, infrastructure investment, talent visa programmes, and operational support.

The United Arab Emirates implemented a federal corporate tax of nine percent in June 2023 and a Pillar Two-compliant Domestic Minimum Top-up Tax effective January 2025. Free zones that had previously offered zero percent rates were either brought within the corporate tax regime or, for qualifying entities, retained zero percent on qualifying income but with the QDMTT applying for in-scope groups. The substance angle is significant: the UAE's competitive proposition for multinational groups is now anchored not on rate but on the combination of fifteen percent effective tax (low by global standards), zero percent personal income tax, full repatriation, and an operational ecosystem that increasingly competes with Singapore and Switzerland for regional headquarters.

What the commentary missed

The implementation surprises fall into three broad categories.

The first is the operational complexity of the calculation itself. Pillar Two requires the determination, on a jurisdictional basis, of GloBE income — a measure derived from financial accounts but adjusted for several dozen items — and Covered Taxes — a measure of taxes paid plus or minus various adjustments. The ratio of Covered Taxes to GloBE income is the effective tax rate. Where this ratio falls below fifteen percent, top-up tax is owed. The calculation must be performed for each constituent entity, aggregated by jurisdiction, with elections, transitional safe harbours, and adjustments applied at various levels.

Tax functions inside multinational groups underestimated, almost universally, the data systems work this would require. Pre-2024, group consolidations were prepared in financial reporting frameworks that did not collect data at the granularity Pillar Two requires. The implementation work through 2023 and 2024 — the building of data pipelines from local statutory accounts to central calculation engines, the reconciliation of book and tax differences across entities, the construction of audit trails that survive regulatory review — has consumed budgets several multiples of what was originally allocated. Most large groups now operate dedicated Pillar Two technology stacks, often built on top of existing tax provisioning systems but with significant custom development.

The second surprise is the scope of administrative interpretation that has been required. The OECD has issued more than a dozen administrative guidance packages between 2022 and 2025, addressing issues that the original Model Rules did not anticipate. The qualification of Qualified Domestic Minimum Top-up Taxes, the treatment of partnerships and trusts, the calculation of substance-based income exclusion for shared payroll costs, the interaction with US GILTI, the treatment of deferred tax assets and liabilities — each has required administrative resolution. The pace of resolution has been faster than initial sceptics predicted, but it has placed multinational tax functions in continuous adaptation mode through the entire implementation period.

The third surprise is the incomplete alignment between the OECD framework and the United States' GILTI regime. The expectation in 2022 was that GILTI would be reformed to qualify as a Pillar Two-compliant Income Inclusion Rule. That reform has not happened. The Inflation Reduction Act of 2022 and subsequent Treasury guidance have not produced a formally aligned regime. The consequence is that US-headquartered multinationals operate under both GILTI (a US measure with its own calculation rules and rates) and Pillar Two (an OECD measure adopted by their non-US operations' jurisdictions). The two regimes overlap but are not identical, producing computational complexity that US multinationals had hoped would be eliminated.

The political question is whether the next administration in either Washington or the major European capitals will accept the framework as a permanent feature of international tax. The early 2025 indications suggest stability rather than rollback. The implementation cost has been incurred. The administrative infrastructure has been built. The revenue is being collected. None of these conditions favour reversal.

The structural adjustments that are working

Among the design responses that have emerged, three deserve specific mention because they appear in the practice of multiple groups across multiple sectors.

The first is the consolidation of holding entities. Pre-Pillar Two, large multinational groups commonly operated four to seven layers of holding entities, each chosen for a specific tax or regulatory advantage. Post-Pillar Two, the marginal benefit of additional holding layers has compressed because the consolidation of profit at any layer is subject to the same fifteen percent floor. Several major European groups have, through 2024 and 2025, undertaken simplification projects that reduce holding layers, eliminate inactive entities, and consolidate the legal structure to reduce Pillar Two compliance burden. The motivating factor is not tax saving — the rate is the same regardless of structure — but compliance cost.

The second is the migration of substance to align with where profit is reported. Pre-Pillar Two, transfer pricing operated on the principle that profit should follow value creation, which in practice often diverged from where operational substance was located. Post-Pillar Two, the Substance-Based Income Exclusion creates a direct mechanical reward for placing substance where profit is allocated. Groups that historically operated thin holding companies in Luxembourg, Switzerland, or the Netherlands have, in selected cases, expanded those operations to capture the SBIE. The investment is not always financially attractive on a pre-tax basis, but the post-tax economics shift in its favour under the new regime.

The third is the renegotiation of inter-company financing. Pillar Two treats deferred tax assets and liabilities in a manner that creates non-trivial interactions with the pre-existing intercompany debt architecture of large groups. In particular, the original ten-year transition rules for deferred tax accounts, and the recapture mechanism that applies if those accounts reverse, have caused groups to reconsider the structure of intercompany loans, the timing of repayments, and the recognition pattern of interest income. The renegotiation work is ongoing in many groups and is expected to continue through 2026 and 2027.

The M&A second-order effects

The effect of Pillar Two on cross-border merger and acquisition activity is now visible in the diligence packages of major transactions. Buyers operating in-scope groups now require, as a matter of standard process, a Pillar Two impact analysis as part of acquisition due diligence. The analysis covers the target's existing effective tax rate by jurisdiction, the deferred tax positions that will integrate into the buyer's group, the SBIE position of the target's operations, and the potential for transitional safe harbours to apply through 2026. Several transactions in 2025 priced in non-trivial Pillar Two adjustments to the post-acquisition effective tax rate, with consequent effects on transaction value.

A specific area of focus is the treatment of pre-acquisition deferred tax assets. Pillar Two contains rules that allow certain deferred tax assets to be recognised in the GloBE calculation, but with caps and timing constraints that do not align with financial reporting recognition. Acquired groups that carried significant deferred tax assets — typically from prior losses or from valuation step-ups — have produced unexpected Pillar Two outcomes for buyers who did not perform the analysis early. Several practitioners have observed that diligence procedures developed in 2024 are now standard in 2025, with most major transactions now including a Pillar Two workstream.

The second-order effect on private equity is more nuanced. Many private equity portfolio companies fall below the seven hundred and fifty million euro consolidated revenue threshold and are therefore out of scope. Where the threshold is exceeded — typically in larger buyout funds or in growth-stage technology investments — the Pillar Two analysis becomes part of the investment thesis. The rate floor reduces, in expected value terms, the after-tax return on investments structured through historically low-tax holding chains. The impact on returns has not been catastrophic, but it has been measurable, and is being absorbed into the underwriting models of major sponsors.

What comes next

The Pillar Two implementation has been consequential, but it is incomplete. Three forward-looking items merit attention.

First, the application of the Undertaxed Profits Rule from January 2025 onward will materially affect groups whose ultimate parent is in a jurisdiction that has not adopted Pillar Two. The largest such jurisdiction is the United States, which operates GILTI but has not adopted a Qualified Income Inclusion Rule. UPR allocates top-up tax to non-parent constituent entities in jurisdictions that have adopted UPR. For US-parented groups with European operations, this means that European subsidiaries may collect top-up tax that would otherwise have been collected by the parent jurisdiction. The mechanic creates a real revenue redistribution from Washington to European capitals on the income of US multinationals, and the political consequences are still being processed.

Second, the development of Pillar One — the OECD's parallel reform addressing the allocation of taxing rights over the largest hundred multinational groups — has lagged significantly behind Pillar Two. The Multilateral Convention to implement Amount A, opened for signature on 3 October 2024, has not attracted sufficient signatories to enter force, principally because of the absence of US support. The deadline for entry has been extended multiple times. Whether Pillar One will ultimately be implemented in its current form, modified, or abandoned remains genuinely uncertain. The current consensus among practitioners is that it is unlikely to enter force in its 2024 draft form, but some elements may persist in modified shape.

Third, the interaction between Pillar Two and the pre-existing network of bilateral tax treaties is producing edge cases that the framework's architects acknowledged but did not fully resolve. Treaty provisions on nondiscrimination, business profits, and dispute resolution interact in complex ways with the Pillar Two top-up. Several groups have already encountered situations where the Pillar Two calculation produces a result that, while compliant with the GloBE rules, is in tension with treaty principles. Resolution mechanisms exist but are slow.

The framework will continue to evolve. The first year produced enough surprises to suggest that the second, third, and fourth years will produce more. What will not change, in the foreseeable horizon, is the existence of a fifteen percent floor that applies, in some form, to every large multinational group regardless of the underlying jurisdiction's headline rate. The architecture of multinational tax planning is being rebuilt around this constant. The rebuilding is well underway, and the patterns of the new architecture are becoming visible.