May 22, 2025 — Withholding tax is the levy that source jurisdictions impose on cross-border payments — dividends, interest, royalties — made to non-residents. The historical bilateral tax treaty network reduced or eliminated withholding tax in many bilateral relationships, with the consequence that the routing of payments through the appropriate treaty jurisdiction became a significant element of cross-border tax planning. The post-MLI introduction of the Principal Purpose Test has constrained but not eliminated this planning. This article reads where withholding tax optimisation stands in 2025.

The historical architecture

The pre-MLI architecture allowed multinational groups to reduce withholding tax on cross-border payments by routing them through holding entities in jurisdictions with favourable bilateral treaty positions. A US technology group paying a dividend to its Cayman parent could reduce the US withholding tax (30 percent under domestic law, 0 percent under some treaties) by interposing a UK or Netherlands holding entity. A German group paying interest to its US parent could reduce withholding tax through a Luxembourg or Irish intermediate. The bilateral treaty network's variability produced numerous routing possibilities, and major multinationals built their corporate structures in part to capture them.

What the PPT changed

The PPT, addressed in detail in a separate article, denies treaty benefits where it is reasonable to conclude that obtaining the benefit was one of the principal purposes of any arrangement. Applied to withholding tax routing, the PPT has substantially constrained the use of intermediate holding entities established principally to access favourable bilateral treaties.

A holding entity in a treaty jurisdiction that exists principally to reduce withholding tax on payments routed through it is, in current administrative practice, exposed to PPT challenge. The defensive response has been the development of substance in such entities and the documentation of commercial rationale beyond tax benefit. Both responses have raised the cost of withholding tax routing structures.

What still works

The PPT has not eliminated all withholding tax planning. Several categories of structure continue to operate effectively post-MLI.

The first category is structures with genuine substance and clear commercial rationale. A holding entity that has been operational for many years, that performs genuine corporate functions, and that obtains incidental withholding tax benefits is generally not subject to PPT challenge. The substance and rationale must be demonstrable, but where they are, the treaty benefits remain available.

The second category is direct investment from jurisdictions with favourable bilateral treaty positions. A direct investment from a US parent to a Belgian operating subsidiary captures the US-Belgium bilateral treaty position without requiring an intermediate structure. The PPT does not affect direct investment in this configuration.

The third category is structures that fall within specific safe-harbour provisions in the relevant bilateral treaties. Some treaties contain provisions that protect specified categories of investor — listed companies, regulated investment funds, pension funds — from PPT challenge. Investments through these qualifying categories continue to capture treaty benefits.

The fourth category is structures using intermediate jurisdictions that provide non-treaty advantages. Luxembourg, the Netherlands, and Ireland continue to be used for European holding structures because of their substantive corporate law frameworks, their professional services ecosystems, and their treaty networks combined. The PPT may apply to limit specific treaty benefits but does not eliminate the broader value of these locations.

What does not work

Several categories of structure that operated effectively pre-MLI no longer work post-MLI.

The thin holding entity established principally for treaty access has been substantially eliminated. A Luxembourg or Dutch entity with no substantive presence, established to route payments and capture withholding tax reduction, is exposed to PPT challenge in essentially every case. The structure has disappeared from forward-looking practice.

The conduit structure layered through multiple treaty jurisdictions has also been substantially eliminated. The structure depends on each layer being viable as an independent treaty claimant, which the PPT has substantially undermined where the layers lack substance.

The corporate response

The corporate response has been a combination of substance investment in legitimate structures and structural simplification of legacy holding chains. The major multinational groups have, in the post-2018 period, invested in substance in their primary holding jurisdictions, eliminated unused entity layers, and consolidated their structures around fewer, more substantively legitimate, holding locations.

The simplification has produced operational benefits beyond the tax benefits. Reduced entity counts have lowered compliance costs, simplified governance, and reduced the operational burden of group operations. The post-MLI structures are simpler than the pre-MLI structures, and the simplification has been generally welcomed by the corporate functions that operate the structures.

The interaction with Pillar Two

The Pillar Two framework's 15 percent effective rate floor has changed the underlying economics of withholding tax planning. A multinational group whose effective tax rate would, but for Pillar Two, be reduced through withholding tax planning may find that Pillar Two captures back any reduction below the 15 percent floor. The combined effect is that the value of withholding tax planning is reduced for in-scope groups, with planning principally relevant for the rate differential between source-state withholding and the Pillar Two floor.

For out-of-scope groups (consolidated revenues below 750 million euros), Pillar Two does not apply and the historical economics of withholding tax planning continue to operate. The bifurcation has produced different planning approaches for in-scope and out-of-scope groups, with consequent complexity for advisers serving both populations.

The trajectory

The trajectory of withholding tax planning through the second half of the 2020s is one of continued constraint with selective continuation. The PPT will continue to constrain treaty shopping. Pillar Two will continue to reduce the marginal value of withholding tax reduction for large groups. Substance-based legitimate structures will continue to operate. The architecture is more constrained than at any prior point in the modern history of cross-border investment, but the constraints operate principally on artificial structures rather than on genuine cross-border activity.

For multinationals operating cross-border, the practical implication is that withholding tax planning is no longer a primary driver of structural choice. The structural choices that remain meaningful are those that operate on substance, on operational practicality, and on the broader business considerations that drive cross-border investment. The withholding tax position is a consequence of those choices rather than a determinant of them.