Improving the OECD transfer pricing guidance on intragroup services

by Andrew Hickman, consultant and former OECD Transfer Pricing Unit Head

The OECD is considering revisions to the transfer pricing guidelines dealing with intragroup services and has invited public comment on this undertaking.

This consultation offers the transfer pricing community a valuable opportunity to share insights to help the OECD shape this critical guidance.

The following is an adaptation of my comments on this key initiative where I suggest that the OECD consider bold ways of significantly reducing the number of group companies to which low value centralised services are allocated, removing the benefits test, and providing insightful guidance on distinguishing and valuing high value services.

I encourage others to submit their comments by the 20 June deadline. The more views that the OECD can take into account at this scoping stage, the more relevant the resulting guidance is likely to prove.

Centralised corporate overheads

I use the term “centralised corporate overheads,” to cover low value-adding intragroup services as defined in the 2017 version of Chapter VII of the OECD transfer pricing guidelines. The omission of “services” perhaps better reflects the commercial motivation of the activity. 

The case for simplification of transfer pricing rules for centralised corporate overheads is compelling. The matter is regularly the subject of transfer pricing audits, a cause of double taxation, and consumes a disproportionate amount of compliance effort relative to its importance or value.

It is a concern for the robustness of the arm’s length principle that it seems to struggle to provide practical solutions for the commonest of commercial activities that seldom present valuation intricacies.

However, attempts by the OECD, EU, and some individual jurisdictions to simplify the charging of costs from the perspective of the services provider have not found widespread support.

One reason is that costs charged in the jurisdiction where the multinational enterprise receives the services can be significant relative to the cost-base and profits of the enterprise. Tax administrations located in such jurisdictions may have an incentive to challenge the charges and have little interest in adopting a simplification regime that applies global allocations.  

A global corporate overheads rule that allows group enterprises to deduct corporate overheads up to the group ratio of overheads to EBIT (similar in some respects to the general interest limitation rule) could be explored; however, this would require extensive buy-in from tax administrations around the world, with likely onus on developing countries, and the rule would continue to be based on the questionable assumption that group-wide allocations are required.  

I see no prospect of material simplification unless it includes a reduction in the number of group companies to which costs are allocated, but this would require bold and imaginative thinking and would involve compromises, perhaps mainly for developed countries.

A perception has been allowed to develop, I believe, that if costs of corporate overheads are not allocated to all group companies, some companies are undercharged while others are overcharged.  This perception should be challenged since material improvement to the Chapter VII rules could involve charging the costs of centralised corporate overheads to a very limited number of group companies.

A perception has been allowed to develop, I believe, that if costs of corporate overheads are not allocated to all group companies, some companies are undercharged while others are overcharged. This perception should be challenged since material improvement to the Chapter VII rules could involve charging the costs of centralised corporate overheads to a very limited number of group companies.

Such an approach is sometimes adopted by MNE groups that are structured through, say, regional principal companies, each with distribution affiliates that are remunerated by those principal companies through a fee determined by targeting an operating margin. Any centralised corporate overheads that could be allocated to those distribution affiliates would not affect their target operating margin, and those additional costs would be effectively borne by the regional principal companies who are the residual profit-takers.

For MNE groups structured in this way it can make little sense for costs to be allocated to all the distribution companies, simply for those costs to be effectively reimbursed by the regional principals. What makes sense is for the costs to be allocated just to the regional principals. 

Such an approach is not only more efficient and may reduce the scale of potential disputes, but it is also capable of defence under the arm’s length principle in that it is the entrepreneurial, value-creating activity of the residual profit-takers that ultimately benefits from the efficient centralising of corporate overheads.

It would be helpful if the revisions to the guidance explicitly endorsed such an approach, but it would require revisiting the concept of “beneficiaries” (which usually means the users of the service) as used in the current guidance as well as the benefits test (see below).  

However, not all MNE groups have supply chains and transfer pricing policies that are structured as described above, and it would be unfair to endorse an approach that is available to some without finding ways of enabling all MNE groups to limit the number of enterprises that are charged the costs of centralised corporate overheads.

I suggest that one way to cover more groups is to recognise that in all groups there are likely to be a small number of value centres that are effectively residual profit-takers even when they may not be so directly embedded in the supply chain as in the example of regional principal companies.

They may operate by exploiting significant intangibles or tangible assets, providing high-value services (see below), or controlling significant risks, as identified in the master file. Such value centres’ income and profits are ultimately affected by the costs that are incurred throughout the MNE group, and they have an interest in the efficient centralising of corporate overheads throughout the group to maximise their returns.  A case can be made under the arm’s length principle that they could solely be allocated corporate overheads.  

The revised guidance could consider how the arm’s length principle supports the allocation of centralised corporate overheads to a limited number of value centres, rather than to all group companies.

Such an approach endorses and makes more widely available the practical outcome that can already be achieved by MNE groups structured in certain ways. The identification of the types of value-creating enterprises in an MNE group qualifying for allocation of costs of centralised corporate overheads and the disclosure mechanism for MNEs applying this approach could be part of the scope of the revised guidance.

Objections may be made that such an approach means that the cost base of group enterprises may not be the same as an otherwise comparable independent enterprise. 

I am not sure that this is relevant to the arm’s length principle, though. What is relevant is the effect on the profits of a residual profit-taker, such as a licensor, were it to bear costs that relate primarily to the business of the licensees, but which also are anticipated to optimise efficiencies for the licensee in generating the income from which royalties are payable, under the hypothesis that the licensor and licensees were unrelated.

The answer to that question is complex, and it should be recalled that determining the profits of such an enterprise is already subject to imprecision under the arm’s length principle. But tax administrations have the option to turn aside this particular complexity for the sake of administrative simplicity and to lower the costs of dealing with disputes and double taxation.

This approach also has the advantage of putting the compromises of simplicity largely on those tax administrations best equipped to bear the burdens: developed countries that are likely to have relatively high anticipated taxable income from their residual profit-taking taxpayers.

The purest outcome, numerically at least, of limiting the number of group enterprises being charged for centralised corporate overheads would be one group company bearing all costs of centralised corporate overheads.

This is likely to be the parent company that ultimately earns the economic benefits of operating a unified business that are achieved in part through centralising corporate functions and which has an interest in overseeing that those functions are performed to standards it sets in an efficient, uniform, and adaptable manner.

However, some parent companies generate low levels of taxable profits in their own tax jurisdiction and some MNE groups have very high levels of centralised corporate overheads. Nevertheless, countries may see wider policy objectives being met by designing a corporate overheads transfer pricing regime which allows an enterprise to be exempt from current application of transfer pricing rules in relation to charging out centralised services and to deduct costs of all centralised corporate overheads.

Such a regime is likely to require a level of economic nexus, so that the activities are substantially conducted in that country (and costs are not simply recharged from other countries) and may involve limitations on the creation of losses or on the size of eligible MNE groups.  The design of such a regime and the disclosure of its adoption by an MNE group could be part of the scope of the revised guidance.

The benefits test

The benefits test in Chapter VII of the guidelines determines whether a transaction should be recognised as an intragroup service. (The test uses the term “activity” not “transaction” but it is an activity performed for one or more group members and therefore appears to convey the same meaning as “transaction” in the rest of the guidelines.)

The test is unique to services. When a transaction involving goods or intangibles is considered, we are not authorised to ignore it from the outset if there is a question about whether the recipient has enhanced or maintained its business position. Instead, the nature of the transaction and its economically relevant circumstances must be analysed under the Chapter I guidance (as supplemented for intangibles in Chapter VI).

It is probable that the Chapter I guidance on accurately delineating the actual transaction (including its contractual terms, the functions of the parties, and how those functions relate to the wider generation of value, the circumstances surrounding the transaction, the characteristics of the services provided, the economic circumstances of the parties) subsumes any purpose of the benefits test. Thus, incorporation of the Chapter I principles in the Chapter VII revision may lead to the conclusion that the benefits test is redundant.  

Removal of the benefits test would be a helpful outcome. It is difficult to answer the benefits test in principle: how do we know whether an independent enterprise would have been willing to pay for anything or to perform any activity itself, and to what do we turn when parties answer the question differently?

Removal of the benefits test would be a helpful outcome. It is difficult to answer the benefits test in principle: how do we know whether an independent enterprise would have been willing to pay for anything or to perform any activity itself, and to what do we turn when parties answer the question differently?

Second-guessing willingness or behaviour is not part of the arm’s length principle. It is particularly difficult to apply the benefits test to corporate overheads since they exist to prevent services being bought in from elsewhere or performed by individual companies.

Finally, the benefits test can cause problems when it is interpreted to encourage valuation of the transaction based on what it would have cost the enterprise to acquire the service from a local provider or perform the activity itself.

Chapter I principles requiring accurate delineation of the transaction with reference to the economic circumstances of the parties seem to provide a much better and comprehensive framework for analysing a services transaction than a test which relies on identifying what an independent enterprise might have been willing to do.

The benefits test has already been nudged to one side in the simplified approach in Chapter VII, and revisions to the chapter should consider whether it continues to serve any useful purpose.  

If the benefits test were to be retained, then care should be taken to align it with the general principles of Chapter I, including both the accurate delineation of the actual transaction and also non-recognition.

Distinguishing and valuing high value services  

The current version of Chapter VII provides limited or no guidance on how to identify and price high value intragroup services that can drive business performance.

Services can, in some cases, command arm’s length compensation that is relatively low and based on the costs of providing the service; whereas, in other cases, superficially similar services can command arm’s length compensation that is strikingly higher since it is based on results. 

For example, procurement or leadership team activities can, in practice, justify cost-based or performance-related compensation, depending on some critical distinguishing factors. There is a pressing need for guidance identifying those factors.

The guidance will probably need to cover how the services relate to the wider generation of value by the MNE group; any significant risks for the MNE group that are affected by the services; how such risks are controlled; and the existence of intangibles, including know-how, that can create scarcity and unique propositions.

Such guidance would neatly draw on the revisions already made to Chapters I and VI, in particular, and would likely relate to aspects of the recent work on profit splits.