Pillar One: Amount B

The October 2021 IF agreement, described the purpose of Amount B to simplify and streamline the application of the arm’s length principle to in-country baseline marketing and distribution activities. Amount B is a component of Pillar One of the OECD’s two-pillar approach that seeks to address the tax challenges arising from the digitalisation of the economy.

This description is misleading in two ways:

• Amount B is not aimed at (and to a large extent, appears not to cover) digital businesses.
• Amount B now appears to be on a separate track from the rest of Pillar One, and is slated to enter the OECD Guidelines and apply from January 2024, before the 2025 effective date projected for Amount A in the OECD’s recent Outcome Statement (PDF, 166KB).

Who will be impacted by Amount B? Amount B will primarily impact multinationals that sell tangible goods on a wholesale basis and operate through routine or limited risk distributors (a more common descriptor for baseline marketing and distribution entities, although the characterisation of a distributor as “routine” or “limited risk” for transfer pricing purposes does not necessarily mean it will qualify as baseline under Amount B, and vice versa).

This consultation document goes beyond theory and sets forth the actual margins that distributors would be expected to earn under Amount B. This means for the first time potentially affected groups have enough information to model the expected impact of Amount B, compare the expected results to their existing transfer pricing policies and consider whether it represents a positive simplification or risks diverging from the ALP.

The bulk of the consultation document is devoted to scope and pricing considerations, with shorter sections addressing an exception where an internal comparable uncontrolled price (“CUP”) is available, documentation, transitional issues, and tax certainty. These sections are covered in more detail in the rest of this report.

Amount B would apply to baseline wholesale distribution of goods selling to unrelated parties, whether carried out through buy-sell arrangements or through sales agent or commissionaire structures, which is an important clarification on the previous consultation. The scope of Amount B, unlike Amount A, is not limited to larger businesses.

The consultation document lays out two possible approaches to defining baseline distribution:

• Alternative A would place greater reliance on a distributor’s ratio of operating expenses to sales as a scoping criterion.
• Alternative B would require additional qualitative consideration of whether the distributor engages in non-baseline activities.

There is disagreement between IF members on which alternative better fulfills the aims of Amount B and comments are requested on which alternative would be preferable.

From a practical perspective, the difference between the alternatives may be less pronounced than it at first appears. Both Alternatives A and B ultimately rely on qualitative and quantitative criteria. Under both alternatives, a taxpayer or tax administration would need to establish the following to demonstrate that transaction falls within scope of Amount B (i.e., is a qualifying transaction):

1. The transaction can be reliably priced using a one-sided transfer pricing method (i.e., the transactional net margin method or “TNMM”), with the distributor, sales agent or commissionaire as the tested party;
2. Depending on which alternative is selected,
   1. No, or a more limited, Substance-based Income Exclusion;
   2. No, or a more limited, De Minimis Exclusion; and
3. The transaction does not involve the distribution of services or commodities; and
4. The tested party does not carry out non-distribution activities (e.g., manufacturing) that cannot reliably be priced and evaluated separately, subject to an “administrative guardrail” (discussed below).

In addition, Alternative B would add a further requirement that the tested party not make non-baseline contributions that are connected to the relevant transaction (or that are unconnected but cannot be reliably segmented and priced separately).

The requirement (under criterion 1) to evaluate whether a transaction can be reliably priced via the TNMM (including the requirement to accurately delineate the transaction) adds a significant qualitative element into the baseline definition under both alternatives. The additional requirement included in Alternative B would go beyond this criterion by recognising that in some cases the TNMM could be applied to supra-baseline activities. However, many of the most challenging transfer pricing questions – for instance, the performance of development, enhancement, maintenance, protection, and exploitation (“DEMPE”) functions and the sharing of economically significant risks – would be part of the qualitative review under both alternatives.

The jurisdictions that favor Alternative B seem to recognise that there may ultimately be little practical difference between the two, suggesting that if Alternative A is adopted, they will simply conduct more audits around the accurate delineation of the transaction and the appropriateness of a one-sided method. The significant qualitative component to the definition of baseline activities under both of the alternatives risks undermining the potential benefits of Amount B by creating material grounds for disputes.

The prior consultation document had suggested that multi-function entities (e.g., entities that perform manufacturing and distribution) would be excluded from scope. The latest document suggests these entities could be included in scope, where it is possible to reliably segment the baseline distribution from other activities. Many stakeholders will view this as a positive step forward. However, a proposed “administrative guardrail” would limit the scope of these entities ability to benefit from Amount B. While the precise contours of the guardrail remain subject to further work, the tentative proposal is that an entity would be ineligible for Amount B if more than 30% of its costs are allocated between the distribution and non-distribution businesses using allocation keys.

While retail activities are excluded from Amount B in principle, the consultation document provides that distributors for which retail sales do not exceed 20% of total net sales would be deemed to be engaged only in wholesale distribution. Recognising the mixed profile of many distributors, this allowance for retail distribution will result in a broader application of Amount B. 

The prior consultation document suggested that distributors that perform any valuable or material regulatory activity should be excluded from scope, leading many commentators to assume the objective was to exclude the distribution of pharmaceuticals. The latest document includes no such limitation, and indeed explicitly identifies pharmaceuticals as an industry the pricing matrix (more below) could apply to. Moreover, an example provided under Alternative B address the relevance of operating licenses that may be required for local regulatory purposes and clarifies that obtaining such licenses should not automatically disqualify a distributor as baseline. This example seems to make it more likely that distributors in industries where such licensure requirements are common, such as the pharmaceutical industry, could be in scope.

A significant open item on which public comments are requested is whether the scope of Amount B should include the wholesale distribution of digital goods. In the context of digital business models, it is not entirely clear what constitutes “wholesale” distribution. Moreover, in the digital space the distinction between digital goods (which are potentially in scope) and digital services (which are out of scope, along with other services) is not always clear, and applying an approach that provides disparate treatment for digital goods and digital services may be problematic. Many digital businesses are likely to be frustrated that they appear to be largely excluded from the scope of Amount B, and indeed the half-in-half-out approach that is currently proposed may be the worst of all worlds causing two separate transfer pricing systems to apply to integrated digital businesses and creating greater complexity and risk of dispute than exists today.

A more technical but important issue is the treatment of pass-through expenses when calculating the ratio of operating expenses to sales – a key issue for both scope and pricing. While a footnote indicates that in some cases such expenses ought to be excluded, more definite guidance on this point will be needed to ensure Amount B provides meaningful certainty.

Amount B would be based on application of the TNMM using an agreed benchmarking approach. The consultation document lays out four pricing models, with differing levels of detail.

The applicable model in a given case would be determined based on the relevant jurisdiction, but it is not yet clear which models would apply to which jurisdictions. All pricing models would use return on sales (“ROS”) as the profit level indicator, subject to a corroborative check, and potential adjustment, using the Berry ratio. The starting point for the benchmarking search criteria remains the same as those described in the December 2022 document; however the OECD has conducted further qualitative analysis around company websites and internet information.  The impact and materiality of this qualitative review is unknown. The pricing matrices (and hence benchmarking) would be subject to a five year update, absent a significant change in market conditions.

This raises significant concerns that the IF is relying on “secret” comparables to set the pricing matrix, i.e., the information is available and may be relied upon by tax authorities, but taxpayers have no ability to review the information. Though this may be partially driven by database license restrictions, it is important that the OECD find a mechanism through which to make this information publicly available to avoid this criticism.

For the generally applicable pricing model, the OECD sets forth a pricing matrix based on three industry classifications and five bands of operating asset and operating expense intensity (referenced as OAS and OES below). Of note, no pricing differences or adjustments are indicated for sales agent or commissionaire structures.  Depending on where in the matrix an entity falls, its Amount B ROS return would be anywhere from 1.5% to 5.5% (see table replicated below).

The returns in the matrix are subject to a 0.5% variance up or down, which would allow taxpayers to target a range rather than a single point and would thus greatly simplify operational transfer pricing considerations for Amount B. The delineation of the five intensity bands favours operating asset intensity as the predominant criterion, with operating expense intensity playing a lesser role.

The industry classifications are based on the OECD’s own classification of the products distributed by companies in its benchmarking search, which led to three groupings based on the OECD’s statistical analyses. These categories may be difficult to apply in practice. For instance, domestic vehicle distribution is categorised in one category, while another category (with higher Amount B returns) is used for domestic vehicles, industrial and agricultural vehicles, motorcycles, and vehicle parts. In reality, a single distributor may be responsible for distributing new and used domestic vehicles, commercial vehicles, and vehicle parts. The industry classifications also contain categories for “mixed goods” and “multiple products lines,” the application of which is unclear and has the potential to lead to dispute.

The IF is considering using a modified pricing matrix for jurisdictions with observed distribution results that differ materially from the global comparable set. The language of the consultation document suggests that a single modified pricing matrix would apply to all such jurisdictions. Annex C indicates that the jurisdictions eligible for the modified pricing matrix would be those where more than half the comparables in a jurisdiction have returns that fall outside the inter-quartile range for the applicable subset of the global dataset, although it is not clear why this would be appropriate. While the ROS returns for the modified pricing matrix have yet to be determined, it appears likely that they would represent increases over the results in the default pricing matrix.

For some jurisdictions, insufficient data is available to test whether the modified pricing matrix  should apply, but the OECD considers there is evidence that country risk may influence arm’s length distribution returns. Such jurisdictions would receive increased Amount B returns based on a formula that adds to the otherwise applicable return under the standard global pricing matrix an additional amount equal to the product of the tested party’s operating asset intensity percentage (i.e., net operating assets over net revenue)  and a net risk adjustment ranging from 0% (no adjustment) to 8.6% based on the jurisdiction’s sovereign credit rating. It is not clear what jurisdictions would be subject to this uplift.

Finally, the consultation document indicates that it will be possible for jurisdictions to substitute a “qualifying local dataset” for the global dataset that otherwise forms the basis for Amount B, and to apply an analogous pricing matrix-based approach on the basis of the local comparables. The local dataset and pricing matrix would require verification by the IF to ensure they meet the same standards as the global dataset, but it is unclear what level of verification would be needed (e.g., majority approval vs. unanimity). The local dataset rule appears to succeed the exemption for local comparables proposed in the December 2022 consultation, but still risks undermining the streamlining benefits of a uniform approach to Amount B and creating additional disputes.  Since the comparables used in the local dataset may not be disclosed to taxpayers, the use of this local pricing matrix raises concerns about the use of “secret” comparables. 

Whichever approach is applied, the ROS results would be subject to corroboration through a Berry ratio cap-and-collar approach. The ROS result determined based on the Amount B pricing approach would be converted into an implied Berry ratio result and compared to the cap (1.50) and collar (1.05); if the implied result fell outside this range, the distributor’s Amount B return would be adjusted to bring it to the nearest edge (i.e., either the cap or the collar). Including a corroboration mechanism in the final design of Amount B is important, particularly for low function distributors for whom the proposed ROS returns are unlikely to be consistent with the ALP.

There is an exception from the pricing matrix approach outlined above, where the CUP) method using internal data provides a better measure of an arm’s length result for a qualifying transaction. This limited exception is likely to be considered an improvement on the more amorphous “most appropriate method” exemption contemplated in the December 2022 consultation document, which would have introduced significant uncertainty into the application of Amount B.  It will be important to understand how broadly applicable to the tested transactions the internal CUP must be to create an exception.

Despite the progress that has been made to date, the consultation document defers consideration of the most critical question: whether Amount B would be implemented on a prescriptive basis or instead operate as a safe harbor. Many taxpayers will see benefit in a safe harbor approach, as this will enable them to elect to apply Amount B when it delivers outcomes that are consistent with the ALP. The uncertainty around this point will likely limit stakeholders’ ability to provide effective comments.

The consultation document indicates that documentation requirements for Amount B could rely on the existing local file/master file framework, while leaving jurisdictions the option to provide Amount B-specific documentation rules. While an affirmative embrasure of simplification of documentation requirements would have been desirable, this represents an improvement of the December 2022 consultation document’s proposal to subject Amount B to prescriptive documentation requirements above and beyond what is required in a local file.

Two items are worth noting: the document contemplates that jurisdictions could require taxpayers applying Amount B to commit to apply it for a minimum of three years,  and though there is no overarching requirement that a taxpayer with in-scope transactions enter into a written distribution contract, but individual jurisdictions may nonetheless require a written contract.

The consultation document takes a generally permissive view towards restructuring into or out of Amount B, noting the right of taxpayers to arrange their affairs as they see fit subject to the application of the restructuring guidance found in Chapter IX of the OECD Transfer Pricing Guidelines. This reflects a significant improvement over the December 2022 consultation document, which contemplated that rules might be needed to limit taxpayers’ ability to restructure into or out of Amount B.

The consultation document offers little in the arena of tax certainty. Amount B adjustments could be addressed through the mutual agreement procedure (“MAP”), if available under an applicable bilateral treaty, and existing MAP and advance pricing agreement (“APA”) resolutions would not be disturbed by the adoption of Amount B. However, the document does not provide for the ability of new bilateral or multilateral APAs to agree to apply results that diverge from Amount B to transactions that would otherwise be in scope.

The consultation marks the first time businesses are able to see the actual Amount B returns the IF proposes to apply to their baseline distribution activities, and to compare those returns against existing transfer pricing policies and observed comparable results. The longer consultation window provides opportunity for businesses to prepare the “empirical data and analysis” that the OECD requests for comments on the pricing framework.

Amount B has the potential to benefit both taxpayers and tax administrations. Progress has been made, but significant open items remain. Given the ambitious timeframe for finalising Amount B in January 2024, this consultation, which closes on September 1, is likely to be the last formal opportunity for business stakeholders to comment on the design of Amount B.

The consultation document specifically seeks comments on the use of Alternative A or B for scoping, the pricing framework, application to wholesale distribution of digital goods, country uplifts, and the criteria for using a local dataset in lieu of the global analysis. However, as noted above, there are other issues where comments may be helpful or necessary.