What I Would Test Before Relying on the GloBE Safe Harbours

Annex A: Where Each GloBE Safe Harbour Applies

This is the last note in my series on the OECD/G20 Global Anti-Base Erosion (GloBE) Model Rules. It deals with Annex A, which sets out the safe harbours under Pillar Two. A safe harbour does two things. It cuts the data a group has to gather, and it gives the group a way to show that no top-up tax is due in a jurisdiction. In this note I take them one at a time. For each one I set out what it is for, the tests it imposes, and the years it runs. I also say where I would want the underlying figures checked before a group relies on it.

Overview of Safe Harbours in GloBE Rules

A safe harbour is a shortcut written into the rules. Meet its tests, and the rules accept a result without a full calculation. The safe harbours in Annex A work in two ways. Some compare figures, such as a margin or an effective tax rate, against a set level. Others accept that a jurisdiction's own tax already does the work. Either way, the safe harbour has to be checked before it is relied on. That check is where groups come unstuck, because the relief is only as sound as the data behind it.

Transitional CbCR Safe Harbour

This is the safe harbour most groups will use first. It lets a group test a jurisdiction on the Country-by-Country Reporting (CbCR) data it already files. No new ledger is needed to run the test.

- Purpose and Application: the group takes the CbCR figures for a jurisdiction and runs the safe harbour tests on them. If a test is met, the group reports no top-up tax for that jurisdiction. The input is the data already in the CbCR file. The work sits in checking that those are the right figures to use, and that they tie back to the accounts.

- Duration and Conditions: this safe harbour is transitional. It runs for a fixed set of years from the start of the rules, and it has to be tested in each of them. A jurisdiction that passed in one year can fail in the next, so the test is run every year rather than once.

Permanent Safe Harbour

The permanent safe harbour does not end with the transitional years. It is the one a group will live with once that period closes. It is also the one that asks more of the group, because it is not a concession to help with the first filings.

- Eligibility and Benefits: the conditions here are stricter than the transitional tests, and that is the design. A permanent relief has to hold over many years, so it cannot rest on a rough cut of the data. A group that wants the benefit has to show it meets the conditions on its facts, and keep the working papers that show it.

Section 1. Simplified Calculations Safe Harbour Framework

Section 1 sets a simpler way to work out some of the figures the GloBE calculation needs. It does not change the tax result. It changes how much work goes into reaching it.

- Implementation Details: the framework names the figures that may be simplified and the cases in which the simplification is allowed. Outside those cases the full calculation stands. A group that uses the simplification should be able to point to the passage that allows it, and to the year in which it applies.

Section 2. Simplified Calculations for Non-Material Entities

Section 2 deals with entities in the group that are too small to justify a full calculation. Size is the first test. The share of group revenue an entity brings in is another. An entity can also fall here because of its tax profile rather than its size.

- Simplified Approach Benefits: effort follows materiality. A group spends its compliance time on the entities that move the effective tax rate, and takes the small ones on a simpler basis. I would test materiality each year, because an entity that was small in one year may not be small in the next.

QDMTT Safe Harbour

The Qualified Domestic Minimum Top-up Tax (QDMTT) safe harbour is the one that turns on a jurisdiction's own law. A QDMTT is a top-up tax that a jurisdiction applies to groups within its charge. Where that domestic tax meets the agreed standard, the safe harbour applies to the jurisdiction, and the group's GloBE position there is settled on that basis.

- Application and Scope: the test is on the domestic rule, not on the group's figures. It asks whether the jurisdiction's top-up tax is built to the agreed standard, and applied as such. If it is, the safe harbour covers the group's entities in that jurisdiction, and the group files on that basis. The group keeps the evidence that the standard is met, because that is where a review will start.

The agreed standard is not a single test. The administrative guidance breaks it into three parts: how the QDMTT accounts for its base, whether it reaches the same result a GloBE calculation would, and how the jurisdiction administers the tax. A domestic rule that meets the first two parts but not the third falls outside the safe harbour. Article 8.2 provides for the safe harbours and Annex A sets their terms, so the check starts with the jurisdiction's law and its published guidance, and only then moves to the group's figures.

Unlike the transitional CbCR safe harbour, this one has no end date, so the test does not stop when the transitional years close. The assessment is of the regime rather than of the group's return, but a group that relies on it should still hold the file showing that its own QDMTT computation agrees with the domestic return. Where both this safe harbour and the CbCR test are open for a jurisdiction, the working papers should say which one the group used, because the two rest on different figures and a later review will start from that choice.

Transitional UTPR Safe Harbour

This one is transitional as well, and it applies to the Under-Taxed Payment Rule (UTPR) rather than to the income inclusion rule. It softens the first years of the UTPR, while groups are still building the data the rule needs.

- Purpose and Transition: the aim is a workable start. A group gets time to pull together jurisdictional data and to test its systems before the UTPR applies in full. Like the other transitional relief, it has an end date. The planning question is what the group does in the year after it.

Conclusion

I would not treat these safe harbours as a set of boxes to tick at the year end. Each one rests on figures that come from somewhere else: a CbCR file, a domestic rule, a set of accounts. The relief is only as good as that source. So I would test it early, before the numbers are locked, and keep the working papers that show which test was used, on which data, for which year. That is also the answer I would give a tax authority that asked why a jurisdiction was left out of the calculation.

Read as a set, the safe harbours are the part of Pillar Two a group touches every year. They are also the part where preparation shows. A group that knows its margins and its effective tax rates by jurisdiction before the filing window opens will find the relief does what it says. A group that meets the figures late will do the full calculation instead.