Why a recurring UK permanent-establishment position may justify an Advance Pricing Agreement — and why bilateral certainty, critical assumptions and annual governance matter more than simply obtaining HMRC agreement

Angelo Chirulli FCA ADIT TEP
September 2026

In my previous article, I considered an international group that had moved beyond simply calculating the profit attributable to its UK permanent establishment.

The group had built a reproducible UK PE attribution model.

It had then mapped that result into the residence-state tax base, separated the UK subsidiary’s Corporation Tax from the overseas principal’s own UK tax, modelled the foreign-tax-relief position and identified the points at which a disagreement could require Mutual Agreement Procedure.

The double-tax position was no longer an assumption.

It was a controlled computation.

That should have been the end of the project.

Instead, the board asked a different question.

Are we going to have to prove all of this again next year?

The answer exposed the next stage in the group’s international tax governance.

A technically robust methodology does not necessarily provide certainty that two tax administrations will continue to accept that methodology.

For a material, recurring cross-border arrangement, there comes a point at which the question changes from:

Can we defend this position?

to:

Should we agree the methodology in advance?

That is where an Advance Pricing Agreement can become relevant.

A repeatable methodology was not the same as tax certainty

The group had already done much of the work that a sophisticated PE attribution required.

It had identified the functions performed in the United Kingdom.

It had determined which assets and economically significant risks were attributable to those activities.

It had established the dealings between the UK PE and the rest of the overseas enterprise.

It had separated those dealings from the actual transactions between the foreign principal and the UK subsidiary.

And it had created a tax balance sheet capable of reproducing the attribution from one accounting period to the next.

But none of that prevented HMRC from reaching a different view in a later enquiry.

Nor did it prevent the residence-state authority from accepting the existence of the UK PE while disagreeing with the amount of profit attributed to it.

The group could therefore produce the same answer every year and still face the same substantive controversy every year.

That distinction matters.

Consistency is an evidential strength. It is not an agreement with a tax authority.

HMRC describes an APA as a written agreement determining a method for resolving transfer-pricing issues in advance of a return being made. Where its terms are complied with, the taxpayer obtains assurance that HMRC will accept the agreed treatment for the period covered. HMRC’s published guidance expressly contemplates applications by a non-resident trading in the United Kingdom through a permanent establishment.

The group therefore had to consider whether its PE methodology had become sufficiently important, complex and recurring to justify moving from annual defence to prospective agreement.

The case for an APA was not that the tax calculation was difficult

Complexity alone was not enough.

The relevant question was whether there was genuine uncertainty about how the arm’s-length or separate-enterprise principles should apply to the particular facts.

HMRC indicates that the APA programme is intended principally for situations where the issues are genuinely complex rather than straightforward, including cases where there is real doubt about the application of the arm’s-length standard.

That distinction changed the group’s analysis.

A routine cost-plus arrangement supported by readily available comparables might be documented annually without requiring an APA.

The group’s position was different.

Its UK operations involved overlapping functions between employees of a UK subsidiary and activities performed for the overseas principal.

The attribution required distinguishing:

• the subsidiary’s own functions from functions relevant to the PE;
• risks contractually allocated outside the United Kingdom from risks actually controlled through UK activity;
• assets legally owned by the foreign enterprise from assets economically connected with the PE;
• dealings within the same legal enterprise from actual intercompany transactions;
• the subsidiary’s arm’s-length remuneration from the residual profit attributable to the PE; and
• the UK taxable measure from the amount recognised for relief purposes in the residence state.

None of those questions disappeared because the first year’s computation had been completed successfully.

If the operating model remained materially unchanged, the same technical questions would arise again.

That recurring uncertainty was the real reason to consider an APA.

A unilateral APA would have solved only the UK side

The first instinct was to seek certainty from HMRC.

That would have been useful.

But it would not necessarily have solved the group’s principal risk.

The concern was not merely whether HMRC would accept the PE attribution.

It was whether the United Kingdom and the residence state would accept a compatible outcome.

HMRC makes the distinction clearly. A domestic or unilateral agreement can establish the UK treatment, but it does not determine the treatment in the other jurisdiction and therefore does not ordinarily eliminate cross-border double-tax risk. Where a relevant treaty and MAP article permit it, a bilateral APA involves agreement between HMRC and the other tax administration.

That point was fundamental.

If HMRC agreed that £10 million of profit should be attributed to the UK PE but the residence-state authority considered that only £7 million properly belonged there, the group could still face double taxation on the difference.

A unilateral UK APA would provide UK certainty.

It would not compel the foreign authority to give relief on the same basis.

For the group, therefore, the economically relevant objective was not simply an APA.

It was bilateral certainty over the methodology and the resulting allocation of taxable profit.

The group first defined what actually needed to be agreed

An APA should not become an attempt to freeze an entire multinational group’s tax position.

The group therefore started with scope.

It identified the issues that created recurring uncertainty and separated them from matters that could continue to be dealt with through ordinary compliance.

The proposed architecture covered:

The UK PE attribution

The functions, assets and risks relevant to the permanent establishment and the methodology for determining the profit attributable to them.

The internal dealings

The basis on which economically significant dealings between the PE and other parts of the same enterprise would be recognised and measured.

The UK subsidiary

The remuneration of the separate UK legal entity for the functions it performed for the foreign principal.

The interaction between the two

Controls ensuring that the same UK activity was not rewarded once through the subsidiary’s transfer price and again through an inconsistent PE attribution.

The financial methodology

The allocation keys, segmentation rules, currency conventions, adjustment mechanisms and any agreed range or formula required to produce the annual result.

The covered period

The prospective years for which the methodology was expected to remain commercially appropriate.

HMRC’s formal APA guidance requires the applicant to explain the covered transactions or issues, functions, assets and risks, the wider value chain, the proposed methodology, financial data and projections, relevant comparables and the assumptions on which the methodology depends.

The exercise therefore forced the group to identify something that an ordinary year-end tax computation had not required with the same precision:

Which parts of the model were principles, and which parts were merely current-year facts?

That distinction became central to the APA design.

Critical assumptions were as important as the methodology

Management initially viewed an APA as an agreement over a formula.

That was incomplete.

A formula that produces an arm’s-length result today may cease to do so if the business changes.

The agreement therefore had to identify the assumptions that made the methodology reliable.

For this group, those assumptions included matters such as:

• where key commercial decisions were made;
• which personnel controlled economically significant risks;
• whether the UK subsidiary continued to perform the same functional profile;
• whether customer contracting processes remained materially unchanged;
• whether valuable intangibles continued to be developed, controlled and exploited in the assumed locations;
• whether the PE continued to use the same categories of assets;
• whether the contractual arrangements reflected the actual conduct;
• whether the relative scale of the UK and non-UK operations remained within expected parameters; and
• whether there had been a major acquisition, disposal or business restructuring.

HMRC’s APA framework refers to these as critical assumptions. They are intended to identify circumstances fundamental to the continuing reliability of the agreed methodology. A material change may require reconsideration, revision or potentially termination of the agreement rather than mechanical application of a methodology to facts for which it was never designed.

This changed the way the board viewed the APA.

The agreement would not permanently freeze the tax result.

It would define the conditions within which the agreed methodology remained reliable.

That was a much more useful form of certainty.

The APA therefore became an operating-model control

The tax team created a critical-assumptions register.

Each assumption was mapped to an operational owner.

Changes in contracting authority were linked to Legal.

Changes in senior personnel and reporting lines were linked to HR.

Changes in the location of decision-making were linked to the business leadership.

Changes in functions and cost centres were linked to Finance.

Acquisitions, disposals and restructurings were linked to Corporate Development.

Transfer-pricing and PE governance therefore stopped being an exclusively tax-department exercise.

The annual question was no longer simply:

Did we apply the calculation correctly?

It became:

Did the business that the methodology was designed for still exist?

That is the distinction between an APA as a tax document and an APA as a governance framework.

An APA still required annual compliance

The board also had to abandon another misconception.

Obtaining an APA would not eliminate annual work.

It would change the nature of that work.

HMRC ordinarily requires annual reporting demonstrating that the agreed methodology has been applied and that the critical assumptions continue to be satisfied. The annual report can include financial information, the outcome of the agreed methodology, relevant adjustments, changes in accounting treatment and a specific statement addressing compliance with the critical assumptions.

The group therefore designed its annual reporting pack before the APA application was made.

The pack contained:

1. the annual PE functional ledger;
2. the PE tax balance sheet;
3. the UK subsidiary transfer-pricing calculation;
4. the covered-dealings schedule;
5. the calculation applying the proposed APA methodology;
6. the reconciliation to statutory and management accounts;
7. the residence-state tax-base bridge;
8. the critical-assumptions certification;
9. a record of compensating or year-end adjustments; and
10. a change log covering personnel, contracts, assets, risk control and business reorganisations.

The objective was not to create another year-end compliance exercise.

It was to ensure that the data required to demonstrate compliance existed naturally within the group’s reporting process.

The group also considered the length of the agreement

A methodology can provide certainty only for as long as the commercial facts remain reasonably predictable.

HMRC’s published practice indicates that APAs typically operate for three to five years, with longer periods considered only exceptionally. The legislation also permits an agreement in appropriate circumstances to affect an accounting period that has already ended, allowing a degree of rollback where agreed.

That created two separate questions.

First, how far forward could the group reasonably expect the existing business model to remain stable?

Secondly, were there prior open periods for which applying the same agreed methodology would materially reduce uncertainty?

The group deliberately avoided requesting the longest theoretically available period.

Its commercial strategy anticipated significant growth in the UK and possible changes to customer contracting.

A shorter period with clearly defined critical assumptions provided more meaningful certainty than a longer agreement built on increasingly speculative forecasts.

Rollback was not simply an opportunity to reopen history

The possibility of applying an agreed methodology to earlier periods was attractive.

But the group treated rollback cautiously.

A methodology negotiated on the basis of the current operating model could not simply be imposed on earlier years if the historical facts were materially different.

The tax team therefore compared the historical functional profile with the proposed covered period.

It examined:

• personnel and decision-making;
• customer contracting;
• risk-control functions;
• ownership and use of assets;
• intercompany contractual arrangements;
• financing;
• changes in the subsidiary’s activities; and
• the scale and composition of UK operations.

Only periods sufficiently comparable to the agreed factual framework could sensibly support the same methodology.

Rollback was therefore treated as a factual comparability exercise, not as an automatic extension of the prospective APA.

The process itself required planning

The group also had to factor procedural timing into its decision.

The UK APA process begins with an expression of interest rather than immediately filing a full formal application. For bilateral or multilateral cases, the taxpayer should also approach the relevant foreign administrations.

If HMRC is prepared to consider the proposal, its current guidance generally envisages submission of the formal application within six months unless another period is agreed. In bilateral cases, information supplied to one administration is expected to be made available to the other.

This meant that the group could not treat the APA as something to start shortly before filing its Corporation Tax return.

The factual analysis, financial modelling and competent-authority strategy had to begin much earlier.

HMRC currently states a target of around 30 months from formal submission for completing the APA process, while bilateral or multilateral cases can take 36 months or longer depending on complexity and the working practices of the other administration.

Prospective certainty therefore required prospective planning.

The 2026 UK reforms did not remove the bilateral problem

The question had become more important following the UK reforms applying to chargeable periods beginning on or after 1 January 2026.

The UK’s domestic PE rules have been updated to align more closely with the latest international consensus, including changes to the dependent-agent PE definition and the allocation of profits. The reform package also introduced new mechanisms dealing with the interaction between transfer-pricing adjustments and UK permanent establishments.

Those reforms can improve coherence within the UK system.

They do not create bilateral agreement.

HMRC itself notes in its current Profit Diversion Compliance Facility guidance that although the domestic PE rules changed from 1 January 2026, the UK’s treaty network remains largely unchanged.

That distinction is easy to overlook.

Domestic alignment with OECD concepts may reduce certain interpretative differences.

But where two countries disagree over the application of the treaty to the same facts, the issue remains international.

No UK domestic reform can, by itself, oblige the other jurisdiction to accept HMRC’s attribution.

The new Advance Tax Certainty Service was a separate question

From 1 July 2026, HMRC also introduced the Advance Tax Certainty Service for qualifying major UK investment projects.

The service can provide binding advance certainty on the application of UK tax rules to a major investment project, from before the final investment decision through to the relevant tax return.

For groups considering significant new UK investment, that creates an additional route worth evaluating.

But it answers a different question from a bilateral APA.

A UK clearance or advance certainty mechanism can determine how the United Kingdom will apply its rules.

It does not, merely by doing so, bind another country’s tax administration to the same allocation of cross-border profit.

Where the commercial concern is bilateral double taxation, the relevant architecture must still consider the treaty relationship and competent-authority process.

The group therefore compared three alternatives

The board ultimately evaluated three approaches.

1. Continue with annual documentation

This was the lowest procedural burden.

It remained appropriate if the operating model was stable, the amounts were manageable and the methodology was unlikely to be contentious.

But it preserved enquiry and double-tax risk.

2. Seek unilateral UK certainty

This could substantially reduce uncertainty over the UK position.

It was potentially appropriate where the principal concern was HMRC’s treatment and the residence-state relief position was already clear or otherwise low risk.

But it could not guarantee corresponding treatment overseas.

3. Pursue bilateral agreement

This required the greatest initial investment in analysis, documentation and competent-authority engagement.

But for a high-value recurring arrangement where the same profit-allocation issue affected both countries, it most directly addressed the risk that mattered economically: two administrations applying different methodologies to the same profit.

The decision was therefore not driven simply by tax technical complexity.

It was driven by the expected value of certainty.

The group built an APA-readiness file before deciding whether to apply

Rather than moving immediately into a formal process, the tax team first created an APA-readiness file.

It contained:

• the legal and ownership structure;
• the relevant treaty provisions;
• the UK PE analysis;
• the PE functional and factual narrative;
• the UK subsidiary FAR analysis;
• the proposed covered issues;
• the economic methodology;
• internal and external comparables;
• historical financial outcomes;
• projected outcomes;
• the taxable-base bridge;
• the foreign-relief position;
• proposed critical assumptions;
• potential rollback periods;
• open enquiries and assessments;
• existing rulings or agreements in other jurisdictions; and
• a list of the questions on which advance agreement would materially reduce risk.

That file had an unexpected benefit.

Even before the group decided whether to proceed with an APA, it exposed weaknesses in the existing position.

Several assumptions that had previously appeared to be part of the tax methodology were actually unsupported operational assumptions.

One allocation key depended on a reporting system that Finance intended to replace.

One functional distinction depended heavily on two senior individuals whose roles were changing.

And one contractual description of risk did not fully reflect how decisions were being made in practice.

The APA-readiness exercise therefore improved the underlying tax position even before any tax authority became involved.

Certainty required the group to accept discipline as well as protection

This was ultimately the most important lesson.

An APA is sometimes viewed as protection from tax authority challenge.

That description is incomplete.

It also constrains the taxpayer.

The group must apply the agreed methodology.

It must monitor the assumptions on which that methodology depends.

It must identify relevant business changes.

It must produce the required information.

And where the operating model changes materially, it cannot simply continue relying on an agreement negotiated for different facts.

HMRC’s guidance expressly provides for revision in appropriate circumstances and requires renewal to be considered separately rather than assuming that an existing agreement will automatically continue.

The value of an APA therefore comes from exchanging some future flexibility for a defined degree of certainty.

For the right arrangement, that exchange can be commercially valuable.

For the wrong arrangement, it can create administrative cost without materially reducing risk.

The board’s question changed again

The group had originally asked:

How much profit belongs to the UK PE?

It then asked:

Will the residence state relieve the UK tax?

The next question was more strategic:

If the same material methodology will determine our tax position for the next several years, should we continue proving it after the event or seek to agree it before the event?

There is no universal answer.

But international groups with recurring permanent-establishment and transfer-pricing positions should be able to answer at least seven questions:

1. Is the methodology genuinely complex or uncertain enough to justify advance agreement?
2. Is the exposure sufficiently material and recurring to justify the process?
3. Is unilateral UK certainty sufficient, or is bilateral agreement required to address the real double-tax risk?
4. Are the business functions, assets and risk-control arrangements stable enough for an advance methodology to remain meaningful?
5. Which assumptions are fundamental to that methodology?
6. Can Finance and the business produce the data required to apply and monitor the agreement every year?
7. Are there earlier periods for which rollback may be appropriate, or open disputes that require MAP rather than prospective certainty?

The most important point is that an APA should not be considered only after several years of disputes.

Where the operating model is material, recurring and sufficiently stable, dispute prevention can be more valuable than dispute resolution.

A well-designed PE attribution tells the group what profit should be taxed in the United Kingdom.

A coordinated foreign-relief analysis tells the group whether that tax should produce double taxation.

An effective bilateral APA can go one stage further.

It can establish, in advance, the methodology that both administrations will use to answer the question.

That is a materially different level of tax certainty.

How Vectigalis AC Tax can assist

Vectigalis AC Tax advises international groups on permanent establishments, profit attribution, transfer pricing, treaty relief, Mutual Agreement Procedure and advance tax-certainty strategies.

We can assist with:

1. APA and BAPA readiness reviews, including assessment of whether advance agreement is proportionate to the risk;
2. UK PE attribution models integrating functions, assets, risks, dealings and tax-balance-sheet analysis;
3. coordination between PE attribution and UK subsidiary transfer pricing, including identification of overlapping or potentially inconsistent remuneration;
4. critical-assumptions frameworks and operating-model governance;
5. taxable-base bridges and bilateral relief analysis supporting competent-authority discussions;
6. APA expression-of-interest and application preparation, including economic analysis and supporting financial models;
7. rollback and open-period analysis; and
8. annual APA governance and monitoring frameworks designed to identify factual changes before they affect the agreed methodology.

For groups with a recurring UK PE or complex UK/international transfer-pricing position, the relevant question is not simply whether the current year’s return can be defended.

It is whether the same uncertainty is likely to arise again next year — and whether continuing to manage that uncertainty annually is commercially rational.

Angelo Chirulli FCA ADIT TEP

Vectigalis AC Tax | www.vectigalistax.co.uk | angelo@vectigalistax.co.uk

This article is for general information and does not constitute tax, legal or accounting advice. The availability and appropriateness of an APA, bilateral APA, MAP or other advance-clearance procedure depend on the specific facts, the relevant jurisdictions, the applicable treaty and the tax authorities involved.