Why registering a permanent establishment is only the beginning — and why a defensible attribution requires a functional ledger, a tax balance sheet and a reconciliation with the UK subsidiary

Angelo Chirulli ADIT FCA BFP IFA TEP CPA | September 2026

In my previous article, I considered an international group that had remediated a historic UK permanent establishment exposure and then faced a forward-looking choice.

It could attempt to remove the PE through genuine operational change. It could migrate the relevant business into its UK subsidiary. Or it could retain the existing commercial model and accept that the overseas principal would continue trading through an intentional UK PE.

The operational review showed that the UK executives were commercially central. Moving their authority overseas would either disrupt the business or require the group to build substantial new capability outside the United Kingdom. At the same time, the UK subsidiary was not intended to assume the full entrepreneurial role.

The group therefore chose the third route.

It accepted an intentional UK PE.

The tax team began the registration process. Responsibility for UK filings was allocated. The board assumed that the difficult decision had been made.

Finance then asked the question that exposed the next problem:

What profit should the PE report each year, and how will we calculate it without reconstructing the business after the year has ended?

The answer was not contained in the registration form.

It was not the profit already reported by the UK subsidiary.

It was not a fixed percentage of UK customer revenue.

And it could not safely be produced by adding a transfer-pricing adjustment to the accounts shortly before the Corporation Tax return was filed.

The group had accepted the existence of the PE. It had not yet built the system required to measure it.

Registration did not define the taxable perimeter

Registering a non-UK company for Corporation Tax acknowledges that the company is within the UK tax charge.

It does not determine which functions, assets, risks, income or expenses belong to its UK PE.

For a dependent-agent PE without a physical UK establishment, HMRC requires registration within three months of the date on which the non-UK company becomes liable to Corporation Tax. Where the company also has a physical UK presence, Companies House registration may be required through a different route.

Source: HMRC guidance — Register for Corporation Tax through a dependent-agent permanent establishment.

Those administrative steps are important. But they do not perform the attribution.

HMRC’s guidance states that the starting point should ordinarily include accounts for both the company as a whole and its UK operations. It also makes clear that the UK chargeable profits must be calculated by reference to a detailed understanding of how the UK activities fit within the non-resident company’s wider trade.

Source: HMRC International Manual — method of calculating PE profits.

That was precisely the group’s difficulty.

The overseas principal maintained global statutory accounts. The UK subsidiary maintained its own accounts. But nobody maintained a financial perimeter for the UK activities of the overseas principal.

There were two accounting entities.

There were potentially three relevant economic perimeters:

1. the overseas company as a whole;
2. the UK subsidiary; and
3. the UK PE forming part of the overseas company.

The third perimeter existed for tax purposes but was not visible in the accounting system.

The subsidiary and the PE could not be merged into one answer

The group’s first proposed shortcut was to use the UK subsidiary’s transfer-pricing result as the PE result.

The UK company received a cost-plus return for specified support activities. Finance suggested that the same figure could represent the total UK profit attributable to the group’s operations.

That approach confused two different questions.

The UK subsidiary is a separate legal enterprise. Its arm’s-length remuneration is determined under the associated-enterprise transfer-pricing rules.

The PE is part of the overseas company. Its profit is the profit that the hypothesised separate and independent UK enterprise might be expected to earn from the functions performed, assets used and risks assumed through the PE.

Those calculations may interact. They are not interchangeable.

HMRC expressly recognises that an arm’s-length commission or service fee earned by a dependent agent may not exhaust the profit attributable to the non-resident enterprise’s UK PE. The agent may earn its own return while additional profit remains attributable to the overseas company.

Source: HMRC International Manual — separate-enterprise principle.

In this case, the subsidiary performed administrative, technical-support and local marketing functions.

Some of its senior personnel also acted for the overseas principal. They led important negotiations, determined commercial parameters and controlled elements of customer and pricing risk.

The subsidiary needed to be rewarded for the services it provided.

The PE calculation then had to determine what additional profit, if any, the overseas company earned through the activities performed in the United Kingdom.

Increasing the subsidiary’s cost-plus percentage would not necessarily solve that second question. It might simply change the profit of the wrong taxpayer.

The attribution had to begin with the business, not the ledger

For chargeable periods beginning on or after 1 January 2026, the UK legislation aligns the attribution rules more expressly with Article 7 and the relevant OECD materials, including the 2010 Report on the Attribution of Profits to Permanent Establishments.

The legislation requires the PE to be treated as a separate and independent enterprise, taking account of the functions performed, assets used and risks assumed through the PE and through the other parts of the non-UK company.

Source: Finance Act 2026, Schedule 7 — permanent-establishment attribution.

The exercise therefore began with a functional and factual analysis.

For each material profit stream, the group mapped:

• who originated the customer opportunity;
• who determined the pricing range and other commercial parameters;
• who negotiated the material contractual provisions;
• who possessed the authority and capability to accept or reject the transaction;
• who controlled the relevant credit, delivery, product and market risks;
• where the assets and systems supporting those activities were used;
• which part of the enterprise possessed the rights and obligations arising from the transaction; and
• which activities were performed by the UK subsidiary on its own account rather than for the overseas principal.

This was not merely a list of employees.

Job titles did not show whether an individual acted for the subsidiary, the overseas company or both. Payroll location did not identify which part of the enterprise controlled a risk. An overseas signature did not establish that the underlying commercial decision had been made overseas.

The analysis had to follow conduct.

Revenue did not follow the customer’s address

The group’s second proposed shortcut was to attribute all UK customer revenue to the PE.

That was also rejected.

A British customer does not automatically make the entire customer margin a UK PE profit. The relevant product may have been developed overseas. Manufacturing or service delivery may have occurred outside the United Kingdom. The overseas principal may have deployed valuable intangibles, capital and specialist personnel.

Equally, the fact that the contract was booked by the overseas company did not keep the resulting profit outside the UK attribution exercise.

The correct approach was to examine the activities and risks producing the revenue.

The group therefore divided its income into material revenue streams rather than applying a single geographical percentage. It separately analysed new customer acquisition, contract renewals, key-account expansion, product-related income and support services.

For each stream, finance created a bridge from the consolidated customer result to:

• the return earned by the UK subsidiary;
• the profit attributable to the UK PE of the overseas principal; and
• the profit remaining with the other parts of the overseas enterprise.

The purpose was not to allocate every pound of revenue by employee location.

It was to identify the economically relevant contributions and then apply an appropriate transfer-pricing methodology to the hypothesised PE.

Depending on the facts, that could involve a comparable uncontrolled price, cost-plus, resale-price, transactional-net-margin or profit-split analysis. The method was the output of the functional analysis, not a substitute for it.

The PE required a tax balance sheet, not merely a profit-and-loss account

The initial finance model concentrated entirely on revenue and expenditure.

It omitted the assets, risks and capital required to support the UK operations.

That omission mattered.

A hypothesised independent enterprise performing the PE’s functions would ordinarily require assets and funding. It could not be assumed to operate indefinitely without appropriate equity or loan capital merely because the legal entity’s treasury function was situated overseas.

The group therefore constructed a tax balance sheet for the PE.

It considered tangible assets used in the UK operations, relevant rights and obligations, working capital, receivables associated with the attributed business and any intangible-related functions performed through the PE. It then assessed the amount and composition of capital that a comparable independent operation would reasonably require.

HMRC’s practical framework begins by attributing the relevant assets, determining the capital requirement, calculating the appropriate funding costs and making any resulting adjustment in the PE tax computation.

Source: HMRC International Manual — practical approach to PE capital attribution.

This did not require the group to transfer legal capital into a separate company.

The PE was not a separate company.

The exercise created the balance sheet that the PE would be expected to possess for attribution purposes if it were the hypothesised separate and independent enterprise required by Article 7.

Without that step, the PE profit could be distorted by too much debt, too little capital or funding costs that did not reflect the risks and assets attributed to the UK operations.

Internal dealings needed evidence, not invoices

The finance team next attempted to create intercompany invoices between the head office and the PE.

That language was misleading.

A PE and its head office are parts of the same legal person. An internal invoice does not create a legally enforceable transaction between them.

The authorised OECD approach can nevertheless recognise internal “dealings” where the functional and factual analysis identifies an economically significant event between different parts of the enterprise.

The distinction matters.

A dealing is not established merely because a spreadsheet labels an amount as an internal service fee, royalty or risk transfer. The group must be able to identify the underlying functions, assets, risks, rights and obligations that changed within the enterprise.

The group therefore maintained a dealings schedule recording:

• the nature of each recognised dealing;
• the parts of the enterprise involved;
• the business event giving rise to it;
• the people responsible for the relevant decisions;
• the supporting operational evidence;
• the pricing methodology; and
• the corresponding entries in the PE tax accounts.

No cash movement was required simply to prove that a dealing existed.

Conversely, a head-office allocation did not become a recognised dealing merely because it appeared in the management accounts.

The tax analysis followed the business event.

Costs needed to follow purpose and benefit

The group had historically allocated central expenditure using global revenue.

That produced superficially precise numbers but inconsistent outcomes.

Some costs related directly to the UK PE. Some supported the overseas company as a whole. Some related to shareholder activity. Others benefited the UK subsidiary rather than the PE.

The revised process therefore applied a hierarchy.

Directly traceable expenditure was allocated directly. Shared expenditure was apportioned using a driver connected to the purpose and benefit of the cost. Headcount was used where personnel drove the expenditure. Transaction volumes were used where processing activity was relevant. Revenue was retained only where revenue genuinely reflected the consumption of the service.

HMRC’s guidance accepts reasonable apportionment of shared and general administrative costs but expects the chosen basis to be applied consistently. It also distinguishes expenses incurred for the purposes of the UK operations from costs that merely appear in a central ledger.

Source: HMRC International Manual — allocation of expenses in a PE attribution.

Consistency did not mean that an allocation key could never change.

It meant that any change had to follow a change in the underlying business and be documented. A different answer could not be selected each year simply because it produced a more convenient taxable result.

The attribution model had to operate before year-end

The group’s historic calculations had been retrospective.

Tax reviewed emails, contracts and interviews after the period had closed. Finance then attempted to reproduce an economic perimeter that had never been recorded contemporaneously.

That was acceptable for remediation.

It was not a sustainable annual process.

The group therefore incorporated the PE model into its ordinary reporting cycle.

Customer and project records were tagged to the relevant profit stream. Personnel costs were divided by documented function rather than legal employer alone. Material contracts recorded the individuals responsible for origination, negotiation, approval and continuing risk control. Shared-cost drivers were captured during the year. Assets and working capital were reviewed quarterly.

The tax team also created trigger events requiring the functional analysis to be reopened. These included:

• a material change in the authority of a UK executive;
• the departure of a relevant overseas decision-maker;
• a new product or market led from the United Kingdom;
• a change in customer contracting arrangements;
• a material acquisition or disposal of assets;
• a change in the use of UK premises;
• a significant divergence between the subsidiary’s transfer-pricing policy and its actual conduct; or
• evidence that an overseas approval process had become routine ratification.

The year-end process then became a reconciliation exercise rather than a reconstruction exercise.

The four-way reconciliation was the principal control

The most important document was not the final attribution report.

It was the reconciliation connecting four different positions:

1. the statutory accounts and tax return of the overseas company;
2. the UK PE profit-and-loss account and tax balance sheet;
3. the UK subsidiary’s statutory accounts and transfer-pricing position; and
4. the residence-jurisdiction treatment, including double-tax relief and any relevant CFC implications.

A change to one element could affect the others.

If the UK subsidiary’s remuneration increased, the PE model had to determine whether that payment represented an expense attributable to the PE, another part of the overseas enterprise or both.

If more risk was attributed to the PE, the asset and capital analysis had to be reconsidered.

If the residence jurisdiction did not recognise the same PE profit, the group had to identify the resulting double taxation and the appropriate domestic or treaty-relief mechanism.

The 2026 reform package also introduced a mechanism under which a UK-resident company may claim relief where a transfer-pricing adjustment to a connected foreign company relates to its UK PE. That development makes coordination between the subsidiary and PE calculations more important, not less.

Source: HMRC policy paper — reform of transfer pricing, permanent establishments and Diverted Profits Tax.

The group could no longer prepare the subsidiary local file, the PE attribution and the foreign-relief computation as separate projects using different descriptions of the same people and risks.

They were different tax analyses of the same business.

The factual narrative had to reconcile.

The group replaced an estimate with a process

The first draft PE computation had treated the UK subsidiary’s cost-plus result as the total UK profit.

The second draft attributed all revenue from British customers to the PE.

Neither reflected the business.

The completed analysis distinguished the subsidiary’s own service activities from the functions performed in the United Kingdom for the overseas principal. It attributed the relevant assets, risks and capital to the PE, identified the recognised dealings with the remainder of the enterprise and selected the pricing method appropriate to each material profit stream.

The resulting PE profit was higher than the subsidiary’s cost-plus return.

It was lower than the full margin on UK customer revenue.

More importantly, it could be reproduced.

Finance could trace it to the underlying contracts, people, assets and decisions. Tax could reconcile it with the subsidiary’s remuneration and the overseas company’s accounts. The board could identify the operational changes that would require the model to be revisited.

The group had not eliminated judgement.

PE attribution will always require judgement.

It had eliminated the need to rediscover the business after every year-end.

The next question for international groups

Accepting an intentional UK PE is a legitimate operating-model decision.

But registration alone does not make the PE compliant.

A sustainable model requires the group to answer five connected questions:

What does the PE actually do?

Which assets and risks belong to those activities?

What dealings should be recognised with the remainder of the enterprise?

How should the resulting profit be priced?

Can the finance system reproduce that answer consistently and reconcile it with the UK subsidiary and the overseas tax position?

A PE profit that appears only in the Corporation Tax computation is usually fragile.

A defensible PE profit begins with the way the business records its functions, authority, assets and risk control throughout the year.

The objective is not simply to produce a number.

It is to build a number that the group can explain, evidence and repeat.

How Vectigalis AC Tax can assist

Vectigalis AC Tax advises UK and international groups on permanent establishments, transfer pricing, profit attribution, CFCs, corporate residence, treaty relief and cross-border operating-model design.

We can assist with:

• dependent-agent and fixed-place PE risk reviews;
• functional and factual analyses under the authorised OECD approach;
• PE profit-attribution models and tax balance sheets;
• mapping of assets, risks, capital and internal dealings;
• reconciliation of PE profit with UK subsidiary remuneration;
• historical remediation and double-tax-relief strategy; and
• implementation of practical PE governance and year-end control frameworks.

If your group has already identified a UK PE — or its UK personnel are increasingly negotiating contracts, controlling commercial risks or acting for an overseas principal — the appropriate time to review the attribution model is before the next year-end, not after the next HMRC enquiry.

Vectigalis AC Tax can undertake a focused PE attribution and governance review, identify the evidential and financial gaps, and design a model that aligns the legal structure, commercial conduct and taxable profit.

Angelo Chirulli FCA, ADIT, TEP
UK Chartered Accountant and Italian Dottore Commercialista
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 outcome depends on the specific facts, the applicable treaty, the jurisdictions involved and the way in which the operating model is implemented in practice.