Transfer Pricing in the UK as of 7 March 2024

UK statistics

The UK tax yield from transfer pricing (including from related actions and enquiries) rose marginally to GBP1,635 million in 2022–2023 (up from GBP1,482 million in 2021–2022 but still down from the high watermark of GBP2,162 million in 2020–2021). The 2022–2023 revenue stems from a lower number of settled enquiries (153, down from 175 in 2021–2022), indicating increasing HMRC focus on higher value actions.

Interestingly, only 15 Advance Pricing Agreements (APAs) were agreed with HMRC during the 2022–2023 tax year (down from 20 in 2021–2022, and the recent high point of 30 in 2018–2019). With APAs taking an average of over 45 months to agree (albeit down from over 58 months in 2022–2023) it is unsurprising that there has been some reluctance from taxpayers to pursue the process. The delays stem from HMRC’s approach in entering into bilateral or multilateral agreements with interested jurisdictions. In particular, a taxpayer’s business operations could undergo material changes in the period taken to reach agreement. In contrast, the average time to resolve mutual agreement procedure (MAP) cases is 28.4 months (up from 21.1 months in 2021–2022), which may have further encouraged taxpayers to seek forgiveness rather than consent.

However, recent figures indicate a change in this approach, with the number of APA applications increasing significantly in recent years (to 45 in 2022–2023, up from 40 in 2021–2022). With Pillar II proposals taking effect in many jurisdictions this year, this increase is understandable: Transfer pricing allocations will have a key impact in determining the effective tax rate in relevant jurisdictions, meaning that any subsequent challenge to transfer pricing will have a knock-on impact on Pillar II calculations. This heightens the risk of double taxation, as well as the related administrative burden (in having to resubmit returns, for example), increasing the benefits of achieving certainty via APAs.

In contrast to APAs, the number of Advance Thin Capitalisation Agreements (ATCAs) in force has fallen hugely, from 334 in 2017–2018 to only 30 in 2022–2023. This is primarily due to the introduction, in 2017, of the corporate interest restriction, an additional regime which broadly limits deductions for UK finance expenses to 30% of an adjusted EBITDA. In practice (although transfer pricing rules take priority under applicable law), the relatively formulaic operation of the corporate interest restriction means that it serves as the first line of defence against tax deductions, thereby reducing pressure on transfer pricing analysis in a finance context. Staffing levels within the relevant HMRC team have not fallen in this period, so, at first glance, it is perhaps surprising that this drop-off in ATCAs has not resulted in an appreciable reduction in the average time to agree APAs. However, this relative increase in HMRC manpower has been offset by APA processes becoming ever more complex as a result of increases in (i) the data available to HMRC (eg, as a result of country-by-country reporting (CbCR) rules), and (ii) the jurisdictional spread of taxpayer operations (as to which, see further below).

Unfortunately, the trend towards increasing complexity in the application of transfer pricing rules looks likely to continue.

Complexity arising from increased mobility

One of the key impacts of the recent pandemic, which is likely to impact transfer pricing for many years to come, has been the increased mobility of the workforce. Indeed, it is becoming increasingly necessary for multinational groups (MNGs) to offer flexibility/hybrid working of this kind to attract, and retain, key talent. This has led to an increased likelihood of (i) MNG functions being spread across an increased number of jurisdictions (including jurisdictions in which relevant MNGs have not had a historic presence and which may not have material experience in, or resources to devote to, negotiating and agreeing APA and MAP processes), and (ii) individuals holding key decision-making or risk functions carrying out such functions across a number of jurisdictions. Countries such as Croatia, Portugal, Brazil and Estonia (which have not historically served as locations where material MNG functions are carried out) have opened up specific visas for individuals working remotely, increasing the likelihood of these countries being drawn into disputes of this kind.

This flexibility in working arrangements is expected to materially complicate the transfer pricing exercise in the coming years, and to increase the information required to be maintained by MNGs to support positions taken.

• Historical means of allocating profits between low-risk functions, such as number of employees or floorspace, would need to be tracked on an ongoing basis, and/or may no longer serve as appropriate measures, respectively.
• For high risk/value functions, the roles and responsibilities of individuals would need to be reviewed with enhanced granularity (to consider whether the work carried out in a particular jurisdiction gives rise to a permanent establishment there and, if so, the appropriate level of profit attribution).
• While transfer pricing rules have always required transactions to be priced based on the reality on the ground, unless adequate limitations are put in place to restrict employees’ ability to carry out certain functions in particular jurisdictions, it will be increasingly difficult to predict and apply outcomes ahead of time, at the cost of MNGs’ certainty and ability to plan.
• Taxpayers that are party to APAs will need to be mindful that any flexibility proposed to be offered to employees does not breach the terms of any existing APAs.
• Timeframes in which APAs can be agreed and MAPs and other disputes resolved are likely to be extended, as it will be necessary for tax authorities to consider information at a previously unnecessary granular level of detail.

Such complexity seems likely to increase the risk of double taxation. In particular, jurisdictions to which profits were traditionally allocated may be reluctant to accept (and hence more likely to challenge) reductions in such allocations. In addition, given that OECD Pillar I proposals (if implemented) are expected to reduce the taxing rights of jurisdictions that have typically served as MNG headquarters (such as the UK, the US and the Netherlands), there is a risk that tax authorities in such jurisdictions may increasingly look to transfer pricing to stem expected revenue losses, becoming more aggressive in their approach thereto.

Somewhat helpfully, the Organization for Economic Co-operation and Development (the OECD) has identified tax complexities arising from the increase in remote-working as one of two key tax areas (together with climate change) on which it plans to focus in the near future. Such work will continue on from the helpful guidance published by the OECD in 2020 and 2021 in response to the pandemic (which covered employment tax, and residency and permanent establishment risk, as well as separate transfer pricing guidance). However, in recent years, the vast majority of OECD resources have been devoted to progressing Pillar I and Pillar II workstreams, and it is disappointing that no definitive timeline for, or scope of, the remote-working project has yet been published. It is hoped that:

• this workstream can be accelerated;
• the scope of the project addresses the full range of resulting tax complications, across personal, corporate (including transfer pricing) and indirect tax; and
• the need for timely progress is not at the expense of thoughtful consideration, and appropriate stakeholder input, as to how the burden of increased tax complications and compliance for taxpayers can best be mitigated.

Increase in data-keeping obligations and data available to tax authorities

HMRC has, in recent years, indicated a need to plug a perceived “information gap” in the context of transfer pricing. More detailed record-keeping requirements, and an increase in the data available to tax authorities, have followed, and are likely to be features of transfer pricing going forward. Indeed, if anything, the greater risk is that the information available to HMRC and other tax authorities may outweigh their resources to properly consider it.

Looking at the UK in particular, in 2023, transfer pricing record-keeping requirements were expanded.

• Broadly, UK taxpayers subject to CbCR (ie, those that are members of MNGs with global revenue of at least EUR750 million) are now required to maintain a local file and master file containing the information described in Annexes I and II to Chapter V of the 2022 OECD Transfer Pricing Guidelines (eg, a description of local management functions, a functional analysis of material related-party transactions, a description of the transfer pricing policy applied thereto and information supporting the taxpayer’s view that the policy is arm’s length). The files can be requested by HMRC at any time, must be provided within 30 days of request and taxpayers that fail to comply will be subject to a rebuttable presumption that errors are careless (such that tax geared penalties would apply). The changes are consistent with the information that HMRC would have expected taxpayers to maintain as part of their obligations to keep records supporting their corporation tax returns, so are not expected to materially increase transfer pricing obligations for in-scope taxpayers. However, interestingly, HMRC guidance advises large businesses that are not in scope to voluntarily prepare such data, suggesting a level of “mission creep” that may be of concern to relevant taxpayers.
• HMRC was also empowered to introduce legislation requiring in-scope taxpayers to prepare a short “Summary Audit Trail” (SAT). The SAT would set out the steps taken by the taxpayer in preparing the local file, to enable HMRC to undertake “high level quality assurance” on the data provided. Such legislation was delayed, pending a consultation that was due to take place in 2023 (but has not yet opened). Taxpayers have voiced concerns that:
1. the SAT may replicate information contained in the files themselves (thereby increasing compliance without any benefit to HMRC);
2. the requirement goes beyond international consensus; and
3. the requirement would, if genuinely necessary, be best implemented as part of a multilateral process to ensure standardisation.

In light of this, it is hoped that HMRC may decide not to pursue the proposal further.

In addition to an increase in legislative requirements, tax authorities, including HMRC, are likely to have access to an increase in publicly available information regarding taxpayers, due to a trend toward increasing tax transparency.

• EU public country-by-country requirements will take effect this year. Where MNGs subject to the rules have UK operations, the information published thereunder would in any event be available to HMRC under CbCR exchange of information rules. These EU disclosures will nevertheless constitute an additional source of information to HMRC, as they will provide comparative detail on non-UK operations. This additional information is likely to lead to enhanced scrutiny of benchmarking by UK taxpayers, and a better understanding by HMRC of how the UK fits into wider MNG operations. (The same is true of EU proposals to require large taxpayers to publicly publish their effective tax rate, if implemented.)
• MNGs are becoming increasingly focused on economic, social and governance (ESG) standards. Two of the most widely-used ESG standards, produced by the Global Reporting Initiative and the World Economic Forum, respectively, both include standards on tax transparency (proposing, for example, that signatories publish figures for their total tax paid, with varying degrees of specificity). While there has not yet been wide-spread take-up of voluntary tax disclosure standards, it is likely that there will be some movement in that direction in the coming years, and that MNG may come under increasing pressure to publicise information regarding their tax affairs.

This increase in publicly available information means that MNGs’ transfer pricing may be subject to scrutiny not only from tax authorities, but also from the wider public and the press (who may not have appropriate experience or context to interpret it). As such, transfer pricing is likely to represent an increasing reputational risk, and may be subject to increased attention from non-tax executives within MNGs. This creates enhanced risks of challenge, as it seems likely that, where MNGs’ tax positions are subject to public or press scrutiny, tax authorities will feel emboldened in pursuing enquiries and assessments.

Proposals for limited simplification

The increased complexity of MNG operations (as a result of globalisation, changes in supply chains, and the above-mentioned mobilisation) in recent years has increased the burden of transfer pricing compliance. Taxpayers have, as a result, called for simplification.

Such calls have been acknowledged in the UK by HMRC, and at an international level by the OECD, with varying degrees of success. Proposed changes contemplated by HMRC seem to be a welcome step toward a more pragmatic approach. However, the resulting benefits are likely to be outweighed by failures to reach international consensus on OECD-led proposals.

UK-specific proposals

In summer 2023, HMRC launched a consultation on potential changes to transfer pricing rules, with a general objective of simplifying the application of the rules (where possible). Having considered responses thereto, the government proposes to make some targeted changes which should be helpful to taxpayers, including the following.

1. UK–UK transactions

UK transfer pricing rules generally apply to UK–UK transactions. This has long been considered by taxpayers to introduce a disproportionate compliance burden (given the low risk of tax-loss to HMRC where UK taxpayers are on both sides of the related- party transaction). Most respondents to the consultation felt that the application of the rules in this context should be limited to scenarios where there is a UK tax advantage (eg, where one party is subject to a higher UK corporation tax rate under specific regimes, such as those applying to oil and gas companies). In response, the government has confirmed it will relax the obligation to apply transfer pricing between UK entities where the UK tax base is not disadvantaged. Respondents were split as to whether this would be best implemented via an express requirement for a UK tax advantage, or more prescriptive drafting (for example, expressly referencing a rate differential) for greater clarity. The government has not yet chosen a preferred approach, although has noted that it will consider whether an exhaustive and specific list of exceptions can be achieved without prejudicing its aim of simplification.

2. Participation condition

The consultation discussed the merits of changing the existing participation condition, which applies, broadly, where entities are under common control (by reference to shareholding, voting power or other powers conferred by governing documents). This was due to government concern that the existing definition of control does not adequately capture circumstances where excessive influence (eg, by major creditors) could impact provisions. Views were sought on a potential move to a more principle-based approach such as:

• the US approach in applying transfer pricing rules where taxpayers are “acting in concert”;
• the Norwegian approach, which requires a “community of interest” (a fact-dependent test that considers whether either party is dependent on, or under the influence of, the other); or
• the Swiss approach, which simply asks whether the tested transaction occurred only because of the relationship between the parties.

Most respondents considered that these alternatives would introduce subjectivity and decrease certainty, and favoured a prescriptive approach. While little detail has been provided, the government seems to have taken this feedback on board, noting that (i) it will address known problem-cases in a targeted and prescriptive manner (seeking to avoid material increases in the compliance burden), and (ii) where current rules produce uncertainty, amendments will be made.

3. Guarantees

Broadly, current UK law provides that in determining whether a financial transaction between related parties is arm’s length, account should be taken of all factors other than the effect of parent guarantees. There is currently doubt as to whether the exclusion extends to implicit support (by virtue of simply being part of an MNG). Following consultation, to better align with the most recent OECD guidance published in 2022, the government intends to amend the legislation to allow regard to be had to (i) implicit support (in line with guidance that the government intends to publish), and (ii) guarantees (within the scope of UK transfer pricing rules that reduce borrowing costs) when determining whether the terms of the debt (but not the amount) are arm’s length.

4. Interactions with market value rules

Currently, under UK rules for the taxation of (i) intangibles, there is a market value override which sits alongside the arm’s length rule (with taxpayers taxed in accordance with the former if it is higher), and (ii) loan relationships and derivative contracts, related-party transactions are required to be taxed in line with an “independent terms assumption” (which broadly refers to the terms that would have been entered into between knowledgeable and willing parties dealing at arm’s length). These various valuation premises were identified as increasing taxpayer compliance, and as potentially creating a different outcome to the outcome under applicable treaties. Following consultation, the government proposes to simplify related-party transactions by (i) only requiring the arm’s length provision to be considered for intangibles (thereby allowing related-party intangible transactions to benefit from APAs), and (ii) in the case of loan relationship and derivative transactions, “simplifying and clarifying” the rules (with further detailed information regarding the proposed changes not yet available).

5. Definition of permanent establishment

While not expressly a part of the transfer pricing rules, the consultation also addressed whether to expand the definition of “permanent establishment” to align with the current OECD guidance. Specifically, respondents were asked for feedback on government proposals to:

• expand the definition to cover “dependent agents that habitually play the principal role leading to the conclusion of contracts that are routinely concluded without modification by the enterprise”; and
• narrow the independent agent exclusion to remove any person “act[ing] exclusively or almost exclusively on behalf of one or more enterprises to which [they are] ... closely related” (changes, in each case, that were first introduced into OECD commentary in 2017 and against which the UK has reserved its position).

Respondents raised concerns that the proposed changes would lower the threshold for permanent establishments and increase uncertainty in tax treatment and the risk of double taxation. The proposals were considered to be particularly detrimental for the asset management industry in:

• potentially reducing the scope of activities (eg, discretionary asset management) that could be carried out in the UK without the possibility of a taxable presence; and
• limiting recourse to the independent agent exemption (given that most managers hold interests in funds they manage, and hence are likely to be “closely related”).

Respondents noted that, if implemented, there would, in particular, be an immediate detrimental effect on offshore fund structures which rely solely on domestic provisions to prevent the creation of a taxable presence for investors (especially where the structure did not qualify for the UK’s “investment management exemption” to prevent such taxable presence, and relied solely on the above definitions). More generally, respondents highlighted that the UK fund industry had been structured around the existing definitions, and that if the proposed changes were implemented, the resulting uncertainty in tax treatment would likely result in fund managers relocating to Luxembourg or Ireland and/or reduced investment into the UK. In light of these concerns, the government noted that it would consider further whether to implement the proposals, but gave assurances that it would, in any event:

• not amend the UK’s double tax treaties to mirror any such changes; and
• seek to prevent unintended consequences for offshore investors in UK-managed funds.

OECD

The OECD has attempted to address taxpayers’ desire for simplification with a proposal for a streamlined, formulaic approach to transfer pricing for baseline marketing and distribution functions.

Broadly, the formula (so called “Amount B”) would be applied on the basis of a “pricing matrix” which uses a specific return on sales as the net profit indicator. The matrix would provide for different pricing depending on (i) the applicable industry, and (ii) whether the taxpayer’s expenses and net operating assets, relative, in each case, to revenue, are high, medium or low. Where taxpayers’ priced in-scope related-party transactions are in line with the Amount B produced by the matrix, the provision would be deemed to be arm’s length.

While it is helpful that the OECD has sought to address taxpayers’ requests for simplicity, there are concerns that this dual-track approach may inadvertently increase the compliance burden, with taxpayers having to familiarise themselves with, apply, and police implementation of, a new second standard (in addition to the existing standards applied to other related-party transactions). In particular, such activities typically attract relatively simple and well-understood transfer pricing methodologies, and so the need for alternative standards is not necessarily clear.

Indeed, OECD proposals are likely to increase, rather than reduce, complexity (and the risk of double taxation) in light of recent announcements that:

• the implementation of Amount B rules will be optional for jurisdictions;
• where one relevant jurisdiction has decided not to implement the rules, it would not be bound by the application of the rules in other relevant jurisdictions (eg, for MAP purposes) and Amount B should not be used as the basis for MAP disputes; and
• jurisdictions opting into the proposals can decide whether to allow (on a permissive basis) or require (on a mandatory basis) taxpayers to apply the streamlined “Amount B” approach.

To prevent taxpayers bearing the burden and cost of fragmented approaches to adoption, it is hoped that either (i) international consensus can be reached, so that the rules will be adopted uniformly, or (ii) proposals are abandoned until such time as full consensus can be reached.

Conclusion

Unfortunately, transfer pricing trends generally appear to be moving against taxpayers’ interests.

Broadly:

• changes in working practices, and proposed OECD-led tax changes (ironically designed to increase simplicity), seem likely to increase complexity, the risk of double taxation and the timeframe for resolving disputes; and
• increases in data-keeping obligations, and in the information available to tax authorities, are likely to increase taxpayers’ compliance burdens and the likelihood of challenge.

While some recent UK developments buck the trend (such as proposed changes to simplify UK transfer pricing rules and recognition of the need for further deliberation before pursuing any changes to the domestic “permanent establishment” definition), they are unlikely to materially move the needle against an otherwise unfavourable outlook. Against this background, it would be prudent for MNGs to:

• build out their internal transfer pricing compliance resources/staffing;
• develop efficient and standardised methods of contemporaneous data collection and analysis (potentially with assistance from artificial intelligence); and
• explore the possibility of APAs to enhance certainty and minimise the risk of challenge.