
GILTI became NCTI and your provision model noticed
A renamed regime with a different deduction is not a labelling change. It moves an effective rate that several other calculations depend on.
- The 2025 US tax act replaced the previous international inclusion regime with net CFC tested income, with a reduced deduction that produces a different effective rate.
- Removing the qualified business asset investment exclusion changes which entities generate an inclusion and by how much.
- For global minimum tax purposes these amounts are treated as controlled foreign company taxes rather than as a domestic minimum tax in the source jurisdiction.
- The change reaches the rate reconciliation, so the explanation of your effective rate has to be rebuilt rather than relabelled.
- The data requirement moves toward entity level attributes that many provision models hold only in summary.
Provision models are among the most conservative artefacts in finance. They get carried forward, patched and reconciled, and the person who built the logic left three roles ago.
That is fine while the underlying regime is stable. It is a problem when a regime is renamed, recalculated and given a different deduction in the same act.
The rename is the visible part. The changes to the base and the deduction are the part that moves numbers.
What changed beneath the name
The 2025 act replaced the previous inclusion regime with net CFC tested income. The mechanics are recognisable, and the parameters are not the same.
The deduction available against the inclusion was reduced, which raises the effective rate on the inclusion relative to the prior position.
The exclusion based on qualified business asset investment was removed, which changes which entities produce an inclusion at all. Capital intensive foreign operations are affected most.
Each of these has a direct path into the provision, and none of them is handled by changing a label in a spreadsheet.

Four places the change surfaces
The inclusion calculation is the obvious one. These are the four that tend to be found later, alongside the project data question raised by section 174A and the project data you never kept.
The data problem underneath
Removing an asset based exclusion sounds like a simplification. In practice it changes which entity level attributes matter and therefore what data has to be collected.
Many provision models hold foreign operations at a summarised level because the previous calculation tolerated that. A recalculated base with a different deduction does not always tolerate it.
This is the same collection problem that runs through the global minimum tax work, described in Pillar Two data living in fifteen systems.
Teams that already built entity level collection for one purpose usually find it serves the other, which is an argument for doing it once properly.
One collection, several regimes. The entity level data is the reusable asset, not the model.
What to check in the model
A short review list that a preparer can work through before the next quarter rather than during it, in the same spirit as checking the fixed asset register before the rule bites.
| Model component | What to verify | Why it matters |
|---|---|---|
| Inclusion base | Which entities are in scope after the asset based exclusion was removed. | The population changes, which changes the answer before any rate is applied. |
| Deduction percentage | The rate used against the inclusion and the period it applies from. | A stale percentage produces a plausible number that is wrong. |
| Foreign tax treatment | How related foreign taxes are treated in the calculation and in the reconciliation. | This is where most restatements of the explanation originate. |
This is a general description of a complex regime with transitional aspects. Confirm the treatment for your facts with your tax advisers rather than relying on a summary.
A model that is patched every year eventually describes a regime that no longer exists.
ChatFinRebuilding the explanation, not just the number
Audit committees ask about the effective rate movement before they ask about the computation, so the explanation deserves as much attention as the calculation.
Restate the prior year on the new basis
A comparison on a consistent basis is what makes the movement explainable. Without it every discussion starts from scratch.
Split the movement into causes
Base change, deduction change and entity mix. Three components explain most of it and each has a different message.
Rebuild the reconciliation lines
The categories that explained the old regime may not exist. Adding a residual line is a signal that the work is unfinished.
Document the version of guidance used
Regimes with interim guidance move. Recording which version supported a position saves an argument later, as in provisional amounts when guidance lands mid quarter.

Two failure patterns in the first cycle
Both are common and both are avoidable with an hour of planning.
Relabelling
The model keeps its old logic with new field names. The numbers look reasonable and the parameters are stale, which is the hardest kind of error to find.
Rebuilding late
The work is deferred to year end, when there is no time to restate comparatives, so the explanation of the movement becomes an assertion.
Where ChatFin sits in provision work
We are not a tax engine and we do not calculate a provision. The value is in the collection and the traceability underneath it.
The technical positions remain with your advisers and your tax team, which is where they belong.
AI Cost Estimator for the CFO Office
Four inputs. A first year range, a timeline, and three delivery models compared. No login required.
Frequently Asked Questions
Is this only a naming change?
No. The deduction available against the inclusion was reduced and the asset based exclusion was removed, which changes both the population of entities that generate an inclusion and the effective rate on it. Treating it as a rename is the most common and most expensive first reading.
How does this interact with the global minimum tax?
Amounts arising under the regime are treated as controlled foreign company taxes for global minimum tax purposes rather than as a domestic minimum tax in the source jurisdiction, which affects how they are allocated. The interaction is technical and fact specific, so take it to your advisers.
What should we do about comparatives?
Restate the prior year on the new basis for internal purposes even where no restatement is required for reporting. Without a consistent comparison the movement in the effective rate cannot be explained in components, and the audit committee will ask for components.
Does this change our data collection?
Usually yes. Removing an asset based exclusion shifts which entity level attributes matter, and many models hold foreign operations in summary because the previous calculation allowed it. Entity level collection tends to serve several regimes at once, which is an argument for building it properly.
The parameters changed. The model should too.
Restating the prior year, splitting the movement into components and rebuilding the reconciliation lines is a week of work that prevents a difficult conversation at year end.
If you want a first year range for automated entity level collection into your provision process, the estimator will produce one.
Open the estimatorBook a Demo
See what this looks like in your ERP.
We will walk through the workflow behind this article on the system you already run, using your own data, and show what a first phase would actually cover.
This article is general information for finance and accounting teams. It is not legal, tax, accounting or investment advice. Dates and thresholds are given as published at the time of writing and several of these timelines have moved before, so confirm the current position with your advisers and with the relevant authority before acting.