Luxembourg works for US money because it pairs a flexible holding and fund toolkit with a US treaty. But that treaty has a limitation-on-benefits clause, and the US side adds check-the-box, CFC and anti-hybrid rules. Getting it right means designing for two tax systems at once.
The European gateway for US capital
Luxembourg is the most used European country for US-sponsored investment into the continent. The reasons are practical: a mature fund industry, the flexible SOPARFI holding, fund regimes such as the RAIF and SIF, deep familiarity among US managers and their advisers, and a US tax treaty with a defined route for dividends, interest and gains.
For a US private equity, credit or real estate manager investing across Europe, one Luxembourg platform holds the whole portfolio in a recognised, bankable country, and keeps US investor reporting workable.
US fund or investors
Limited partners and the sponsor.
SOPARFI or fund
The holding and financing layer.
Portfolio companies
Germany, France, the Netherlands, Spain…
Four things a US link changes
Limitation on benefits
The US–Luxembourg treaty has an LOB clause. The Luxembourg company must fit a qualifying category (ownership, listing, active business or derivative benefits) to get reduced US withholding, whatever its business reasons.
Check-the-box
US classification elections let a Luxembourg entity be transparent or opaque for US purposes. Powerful for aligning both sides, and a common source of hybrid mismatches when handled carelessly.
Anti-hybrid rules
The EU's ATAD II and the US rules both bite when an instrument or entity is treated differently in each country. Mismatches that once created benefits are now neutralised.
CFC income and Subpart F
US shareholders face anti-deferral rules on income earned through foreign companies: Subpart F and GILTI, renamed net CFC tested income from 2026. The structure has to be modelled at the US shareholder's level too.
Which overlays apply to you?
Tick what applies. Each tick is one more rulebook to model before you set up.
How many US overlays does your structure trigger?
Complexity, not a verdict.
European logic vs US-connected logic
| Question | Europe-only structure | US-connected structure |
|---|---|---|
| Anti-abuse gate | PPT, via the MLI | PPT plus the US treaty's LOB |
| Entity classification | Local law | Plus a US check-the-box election |
| Hybrid risk | ATAD II | ATAD II plus US anti-hybrid rules |
| Investor-level tax | Local investor rules | Plus US CFC rules on US holders |
| Withholding relief | Directives and EU treaties | US treaty, subject to LOB |
| Design driver | Optimise one system | Reconcile two at once |
A Europe-only structure answers to one tax system. The moment US capital enters, it answers to two, and they do not always agree.
Design for both systems from day one
The recurring mistake is to design a clean European structure and treat the US side as an afterthought. The US overlay has to be modelled at the start: confirm the Luxembourg company qualifies under the treaty's LOB clause, choose classifications that avoid hybrid mismatches under both ATAD II and US rules, and test the structure against the US shareholder's CFC position.
Substance remains the foundation on both sides. A Luxembourg platform with real function and decision-making supports treaty access, beneficial ownership and the principal purpose test. Where a Dutch layer fits better, the same logic applies to a Dutch holding.
The default European hub for US fund and investment capital.
The US treaty adds a mechanical test on top of the PPT.
Check-the-box plus ATAD II can neutralise mismatches.
US CFC rules mean the holder's position matters too.