Dear Members of Working Party No. 6,
I welcome the opportunity to comment on the discussion draft revising Chapter VII of the OECD Transfer Pricing Guidelines, released on 1 June 2026. My comments are confined to
- The treatment of duplicative services under paragraphs 7.32 to 7.34 and the proposed expansion of the duplication exceptions.
- Examples of shareholder activities under paragraph 7.26 e
- The treatment of duplicative services under paragraphs 7.32 to 7.34 and the proposed expansion of the duplication exceptions.
- Summary of Position
Multinational groups operate in an intensely competitive commercial environment in which duplicated cost is not tolerated as a matter of ordinary business discipline. Group finance functions actively eliminate redundant spend independent of any tax consideration. This commercial reality means that genuine, unexplained duplication of intra-group services is rare in practice.
In my experience, the far more common source of dispute is not real duplication but the mischaracterisation of complementary, additional, or centrally-mandated activities as duplicative by examining tax authorities who stop their analysis at superficial indicators such as a matching job title, a similarly worded service description, or an overlapping cost centre , without undertaking the fact-intensive, multi-factor review that paragraph 7.33 itself prescribes.
I therefore support the direction of the proposed revisions, and in particular the expansion of the duplication exceptions to cover regulatory requirements. However, I respectfully submit that the discussion draft, as currently framed, does not go far enough to correct the underlying audit practice problem. My recommendations are set out below.
- The Practical Problem: Asymmetric Application of the Case-by-Case Standard
Paragraph 7.33 correctly states that the determination of whether a service is merely duplicative “should be made on a case-by-case basis” and should focus on “the nature, goals, scope, customisation, duration of the activity and, when relevant, the reasons for apparent duplication.”
In practice, this multi-factor test is frequently applied asymmetrically: taxpayers are expected to substantiate each factor in detail to rebut a duplication challenge, while tax authorities routinely raise the challenge itself on the basis of superficial similarity alone, without first establishing which of the 7.33 factors actually supports a finding of duplication.
In other words, a test that is meant to be applied carefully, factor by factor, is instead being used as a quick, one-size-fits-all trigger for audit challenges. It also creates significant friction and cost for taxpayers who must reconstruct, often years after the fact, the scope and customisation distinctions between in-house and centrally-provided activities that were reasonably clear at the time the arrangement was entered into.
3. Recommendations
3.1 Require tax authorities to substantiate a duplication finding against the 7.33 factors before shifting the burden to the taxpayer
I recommend that the revised guidance state explicitly that a tax administration asserting duplication must first identify which of the nature, goals, scope, customisation, or duration factors in paragraph 7.33 supports that assertion, with reference to the actual functions performed, before the taxpayer bears the burden of demonstrating non-duplication.
This does not reverse the ultimate burden of proof under domestic law, but it would require the audit position itself to be reasoned rather than asserted, which would materially reduce the volume of duplication challenges raised on name-matching or cost-centre-matching alone.
3.2 Adopt a rebuttable presumption against duplication for centrally-driven, consistently-allocated functions
Given the commercial reality described in Section 1, I recommend the guidance adopt a rebuttable presumption that a function that is (1) provided centrally by the group, (2) allocated consistently to all entities, and (3) not separately re-performed in full scope at the local level, is not duplicative. The presumption would be rebuttable on specific evidence of true functional overlap, and would not apply to the recognised exceptions carve-outs already noted in the draft (temporary duplication during reorganisation, second-opinion risk mitigation, and regulatory requirements for local performance). This reflects how groups actually behave under competitive cost pressure and shifts the evidentiary starting point away from suspicion of duplication toward the fact pattern that is empirically more common.
3.3 Integrate the duplication inquiry into the benefit test rather than treating it as a free-standing ground for disallowance
The discussion draft usefully clarifies that the benefit test and the arm’s length charge are separate analyses, that a benefit need only be reasonably expected at inception, and that the test must be applied at each recipient entity’s level. I recommend that the question of duplication be formally absorbed into this benefit-test framework, i.e., whether a benefit was reasonably expected at inception given what the recipient already performs for itself, rather than remaining available to tax administrations as a separate, free-standing basis for disallowance that can be invoked without the same evidentiary discipline the benefit test now requires. This would also resolve the ambiguity, noted at paragraph 7.34 of the current text, between whether a service is duplicated and whether costs are charged out more than once; both questions would sit within a single, coherent inquiry.
3.4 Retain the case-by-case standard; avoid a bright-line anti-duplication rule
While the commercial case for discouraging duplicated cost is strong, I would not recommend resolving this issue through a bright-line rule that treats any apparent overlap as presumptively non-chargeable. Such a rule would be simpler for examiners to apply, but would remove the very case-by-case discipline that allows genuinely complementary, additional, or intentionally redundant arrangements e.g. getting a second opinion to reduce decision risk (an example from the consultation document), including group-wide compliance and risk-management functions that must be performed both locally and centrally, to be recognised on their merits. The correct target for reform is the rigour with which the existing standard is applied, not the standard itself.
4. Conclusion
The proposed expansion of the duplication exceptions to cover regulatory requirements is a welcome, if narrow, improvement. I encourage Working Party No. 6 to pair it with clearer expectations that tax administrations apply the paragraph 7.33 factors with the same rigour expected of taxpayers, a rebuttable presumption against duplication for centrally-driven functions, and integration of the duplication question into the benefit test. Together, these changes would reduce unproductive audit friction while preserving the fact-based, case-by-case standard that the current guidance rightly establishes.
- Shareholder activities under paragraph 7.26 e
- Summary of Position
I welcome the discussion draft’s move to treat shareholder activities, duplication, and incidental benefits as illustrations of situations that fail to satisfy the single benefit test, rather than as separate, free-standing tests in their own right. This is a conceptually cleaner approach: it recognises that these categories are not independent legal doctrines but different ways in which an activity may fail to demonstrate that an independent enterprise would have been willing to pay for it, or would have performed it in-house.
However, I have two specific concerns with how the shareholder-activities category is currently drafted:
- The catch-all for “ancillary activities to the corporate governance of the MNE as a whole” is too elastic and is likely to be over-used in audits
- The worked examples in Annex I illustrating shareholder activities are one-sided, which risks skewing the guidance toward a default assumption that central coordination is presumptively non-chargeable.
- The Practical Problem: An open-ended catch-all invites over use in audits
The discussion draft retains the existing list of shareholder-activity examples from the 2022 guidelines and invites stakeholder input on that list, including the catch-all phrase “ancillary activities to the corporate governance of the MNE as a whole.”
The difficulty with this phrase is that “corporate governance of the MNE as a whole” is not a bounded concept. Governance, in a modern multinational group, legitimately overlaps with a wide range of commercially valuable central functions such as treasury oversight, group risk policy-setting, strategic coordination, and cross-border legal or IT coordination among them. Because the catch-all is framed at the level of governance in the abstract, rather than by reference to the specific nature of the activity performed, almost any centrally-run function touching multiple entities can be characterised by a tax administrator as “ancillary to governance of the group as a whole,” regardless of whether an independent enterprise would in fact have been willing to pay for that same activity.
In other words, a catch-all that is meant to close a narrow gap left by the specific examples is, in practice, open-ended enough to pull genuinely chargeable, commercially-justified central functions back into the non-chargeable “shareholder activity” category simply by relabelling them as governance-related.
- Recommendations
3.1 Narrow the “ancillary activities to corporate governance” catch-all
I recommend that the catch-all be redrafted to refer to specific, identifiable activities — for example, activities relating solely to the legal structure of the parent as a shareholder (such as preparing consolidated financial statements for the parent’s own regulatory or listing obligations, or organising the parent’s own shareholder meetings) rather than the open-ended phrase “governance of the MNE as a whole.” A defined, closed-ended formulation would preserve the intended scope of the shareholder-activities category without inviting its extension to central functions that have independent commercial value to the group.
3.2 Add a counterbalancing example where central coordination clearly does qualify as a chargeable service
The current Annex I illustrations for shareholder activities (Examples 6 to 8) are all one-directional: they address activities performed by a parent as part of a potential divestiture of a subsidiary, expenses incurred by a parent in divesting a subsidiary, and expenses incurred by a parent in acquiring a target, and in each case conclude that the activity is a shareholder activity that does not satisfy the benefit test. Because all three worked examples point in the same direction, the draft risks leavingtax administrators and practitioners with the impression that any parent-level coordination or oversight function defaults toward non-chargeable shareholder activity.
I recommend that the Annex include at least one additional example illustrating central coordination or governance-adjacent activity that does satisfy the benefit test, for instance, central treasury risk oversight that actively manages the group’s funding, liquidity, or hedging exposures on behalf of group entities, as distinct from the passive monitoring of an investment that characterises a true shareholder activity. This would give the guidance the same balance recommended elsewhere in respect of duplication: a standard capable of distinguishing genuine shareholder activity from centrally-coordinated, commercially valuable services, rather than one that defaults to treating anything governance-adjacent as non-chargeable.
- Conclusion
The decision to treat shareholder activities, duplication, and incidental benefits as illustrations of a single benefit-test failure, rather than as separate tests, is a welcome simplification. I encourage Working Party No. 6 to pair this simplification with a narrower, more precisely defined catch-all for ancillary governance activities, and with a more balanced set of worked examples in Annex I that illustrate both when central coordination fails the benefit test and when it satisfies it. Together, these changes would reduce the risk that the shareholder-activities category is used in audits as a broad basis for disallowing centrally-provided services that a third party would, in fact, have been willing to pay for.
I thank you for the opportunity to contribute to this important work and please do not hesitate to contact me.
Yours faithfully
Rubeena Dina