EBIT comments on the OECD Secretariat’s proposal for a “Unified Approach” under Pillar one

To: Tax Policy and Statistics Division, Centre for Tax Policy and Administration

Sent via upload TFDE@oecd.org

Dear David,

EBIT’s Members¹ are grateful for the opportunity to provide comments and input on the OECD’s public consultation running from 9 October – 12 November 2019 with regard to the OECD’s public consultation document (hereinafter: ‘the consultation document’) on the Secretariat’s proposal for a “Unified Approach” under Pillar one.

Comments

Below are a number of issues that EBIT believes are important for the Task Force on the Digital Economy as well as the Inclusive Framework (IF) on BEPS to consider in reaching a consensus solution on the so-called “unified approach”.

We note that an important consideration, that could also be seen as part of dispute prevention, will be the political commitment by countries to repeal (and not introduce new) unilateral measures which seek to increase their tax base (such as digital services taxes, diverted profits taxes, profit fragmentation rules etc.) and lead to increased risks of double taxation.

EBIT’s contribution follows the order of the questions for public comments raised at the end of the consultation document.

1. Scope

It is important that the scope of the unified approach is clearly defined and delineated in order to avoid any undesired spill over effects. In particular, the following would need to be defined clearly and narrowly:

• Who or what is exactly considered a ‘consumer’ / ‘consumer facing’ or a ‘user’ under the unified approach? In our view, under certain conditions also certain B2B transactions (which are not defined in the unified approach) might be considered to fall under the scope of Amount A but any such targeting would have to be clear and within principled objectives. Confusion on the scope must be avoided. For example, in the case of advertising on a webpage, the question is who is the consumer or user? What the proposed unified approach appears to capture is the target group of consumers / users to which the advertising is focused but that have no intervention in the actual B2B transaction (and are actually third parties). The actual consumers or users in the depicted case are the enterprises or persons using the digital channel to advertise (consumption of advertising space);
• What is considered to be a MNE group? In addition, differences in the composition of the MNE group may occur when using tax accounting or financial reporting. Further, countries’ legislations differ on what are considered to be associated enterprises (= members of an MNE group) and this may lead to controversy. A suggested approach could be to use the consolidation requirements of the ultimate parent entity or - when consolidation occurs at a regional level - of the regional parent entity.
• Tax accounting and financial (consolidation) accounting also produce different numbers as they serve different purposes. So, if a re-allocable super or residual profit were determined on the basis of the latter, the scope for double taxation is higher (for example under Amount A and Amount C or domestic tax rules: see below) where there is taxable income otherwise arising in a territory to which an allocation is made or where a credit is appropriate in a territory from which it is made.
• The scope of the unified approach should stop at the sale to the last associated enterprise which monetises the relationship with the consumer/user, usually via final sale to them. Sales performed by intermediaries (distributors, commissionaires or sales agents) independent from the MNE group should not be considered, as the MNE will have no control over the location of the sale, no ability to track an item through the value chain and no ability to obtain financial information from these unrelated parties.

2. New Nexus

Under the unified approach, the new nexus rules are intended to be largely based on sales. EBIT considers that sales can be a potential threshold but cannot be the only threshold.

Firstly, if the size of the group were to be considered part of the nexus, an alignment with the threshold for country-by-country reporting (currently set at consolidated group revenue of 750 million euro) has some administrative logic although there is a principled argument for a reduced amount involving only in-scope activities.

Secondly, a combined threshold in the market jurisdiction consisting of:

• A specific sales threshold;
• A country-specific adjustment factor that would increase or decrease the sales threshold to calibrate country-specific features. This country specific factor could be based for example on GDP, income per capita or other macro-economic factors. It would allow creating a level playing field between larger and smaller economies and between developed and developing economies.

The technical details regarding the treatment of losses are still to be announced. There may be an option to apply thresholds to group results measured on an average basis over, say, up to three or five years which may partly deal with this as well as inconsistency in businesses falling above or below cliff-edge thresholds due to variations in the business cycle.

Thirdly, EBIT welcomes the suggestion in the Secretariat proposal to carve out certain industries such as extractive industries and commodities, and, potentially, financial services.

Fourthly, we understand that - unfortunately - a significant incremental administrative burden will be generated through the new nexus rules. However, such administrative burden should be minimized by not requiring numerous additional registrations in all the countries where there is (a potential) new nexus. This could possibly be created through a “One Stop Shop” idea, leveraging on systems that are inter alia in use for existing VAT purposes in Europe. For example, the calculations defining Amount A could be sent to (and relevant tax paid to) one tax authority (in the parent jurisdiction) and that tax authority would audit it (perhaps allowing scrutiny from other tax authorities) and reallocate the tax revenues accordingly. This system shows some resemblance with the exchange of information system for country-by-country reporting. While tax authorities in other markets, likely including developing countries, may at first sight be reluctant to cede (primary) responsibility to the parent jurisdiction tax authority, there would be advantages to them in not having to set up new infrastructure of their own to collect tax from non-resident multinationals.

Finally, sales to unrelated distributors should not be included in the nexus calculations for the reasons set out in relation to scope.

3. Calculation of group profits for Amount A

Preliminary remark

EBIT considers the taxation based on Amount A as an additional tax, creating an additional burden for taxpayers within its scope. It seems that this additional tax has some hybrid features of different kinds of taxes (sales tax by using sales as allocation key, destination based tax, income tax). In any way certainty should be obtained that this newly created tax can be relieved (by preference automatically) with the states or states forfeiting their right to tax under the existing rules (e.g. by providing a deduction from taxable income in the surrender jurisdiction for the amount allocated to other jurisdictions under Amount A) and care being taken about interaction with Amounts B and C. The tax needs also be recognized in the treaties relieving double taxation, possibly through a MLI 2.0.

Other observations

The starting point for the calculation of group profits seems to be the accounting records and accounting profits. Although there may be some advantages to using such records (no additional burden, audited), it remains that these records are not compiled for taxation purposes as such, but with the aim of informing a large set of stakeholders. In principle, the accounting system to be used is the system in force in the jurisdiction of the ultimate parent, which must be accepted by the other countries in which a new nexus exists.

EBIT Members believe that consideration should be given to determining profit thresholds etc. segmentally, by business line and/or regionally, where considerable differences exist in practice. Bearing in mind that segmentation in consolidated financial statements is not standardised and that businesses often have their own unique methodologies, flexibility of approach would be necessary. Some potential principles might include the following (which are intended to co-exist):

• Enterprises should be permitted to use aggregate financials without segmentation if they prefer.
• Enterprises should be able to segment on their own in a manner which differs from their financial statement segmentation, provided they use the same segmentation globally, and potentially that they carry the same method forward with limited ability to change (i.e., a multi-year consistency requirement)
• Taxing authorities (whether in home or market jurisdictions) may not require an enterprise to segment in some way that they prefer – that is, enterprises must have the choice on whether and how to segment.

A related debate concerns which level of profit needs to be considered. Net sales as denominator and profit before tax as numerator may be most appropriate – but variations exist so definitions might still be necessary. Revenue is not suitable as denominator as it may be influenced by elements that have no bearing to the existence of the nexus. Profit before tax is considered appropriate as this represents the actual costs incurred in the business. Generally, some businesses look at earnings before interest and tax to take away the noise that interest can provide in results – reflecting those that take on more debt than others at different times.

The Secretariat proposal indicates that the profit under Amount A could not be duplicated under amount C. We would welcome a clarification during the public consultation as to how this would be achieved.

4. Determination of Amount A

Amount A is applicable to consumer-facing businesses. As already indicated in this contribution, it is not clear what is meant with consumer-facing businesses as ultimately, all goods produced or services rendered end up one way or another with a consumer or user. A ‘one size fits all’ consumer-facing approach is therefore going to be a real issue and we would welcome refinements based upon industry approaches and regional/ business line /approaches. EBIT is happy to provide three examples that deserve attention:

Example 1: An MNE with substantial sales in a limited number of countries, and relatively less sales in a substantial number of countries;

Example 2: An MNE situated in a relatively small economy where the value is created, with most of the sales generated outside that economy;

Example 3: An MNE active in certain markets that generate high profit margins versus low profit margin in other markets.

As indicated under Question 1, transactions with unrelated distributors should be out of scope. Acquiring the information from sales of unrelated distributors to local wholesalers or final consumers (situated in a different country than the unrelated party) will not be readily accessible for the MNE.

We note that the Secretariat suggests a 4-step approach:

1. determine the total profit to be used, either at group level or business line level;
2. exclude deemed routine profits to arrive at a deemed residual (through fixed percentages);
3. allocate a portion of the deemed residual profit on the basis of a formula to consumer- facing / user activities; and
4. allocate (c) among the different market jurisdictions on the basis of an agreed allocation key (formulary); Sales can be considered as appropriate allocation key.

In order for the system not to create an overly excessive burden, it is necessary that the IF agree on a uniform calculation and formulaic approach. We suggest that this is done base on consolidated income determined in accordance with the accounting rules of consolidated financial statements. This, however, may create a disconnect with the actual result generated in a market state through consumer-facing or user activities. In order to maintain as much the connection between the formulaic approach and real-life business as possible, the parameters used for the formula ought to be reviewed regularly (for example every 3 years).

Amount A should apply in a manner that is fair and equitable across industries, acknowledging losses and rewarding upfront investment costs (which have a higher risk profile, are incurred early in the lifecycle and often involve significant aborted spend). This is particularly important for industries unable to benchmark returns for high-risk investment and which have long product lifecycles (for example, where R&D costs are incurred many years in advance of income being generated). In a given year, this could create the illusion of residual profits, where the economic reality is that investment costs are still being recovered.

It is also important that the country of innovation be appropriately rewarded for the investment costs it has borne and the risk of innovation otherwise this would disincentivise investment in R&D with an adverse impact on global innovation.

Regarding the possible use of segmentation, should other than consolidated accounts be allowed for determining the determination and allocation of Amount A, EBIT points out that segmented reporting data in a set of (consolidated) IFRS accounts will generally be retrieved from management accounts rather than from statutory accounts, often based on a profit centre perspective. The profit centre view cannot be easily reconciled to a legal entity perspective and used for allocating profits to a specific country. Pragmatism and flexibility should govern the use of segmentation in order to lead to a reasonable profit allocation in such case. MNEs should maintain the freedom to use consolidated accounts in any case.

At this moment in time, it is still unclear how the formulaic approach under Amount A will turn out, as the different parameters are not yet known. In any way, the parameters and thresholds should be set high enough so as not to have an adverse effect and be set in line with economic reality.

We also recommend that the calculations for the profits to be allocated be based upon a share of group net income and not as a proportion of turnover. Regional profitability may not always be determined or may be difficult to determine and should be avoided of using it as a basis for calculating Amount A.

EBIT considers that Amount A should not be used as a basis for adjusting the customs base or any other indirect taxes.

5. Elimination of double taxation in relation to Amount A

Amount A should give entitlement to relief and therefore taxation under the new nexus rules should be considered to be covered under the treaties, possibly through the use of an MLI 2.0. The recognition through the insertion of amount A in the treaty is crucial.

As the Amount A is a formulaic approach, tax authorities should give automatic and full relief (e.g., by providing a deduction from taxable income or exemption in the surrender jurisdiction for Amount A allocated to other jurisdictions). This automatic and full relief could possibly be achieved when Amount A is determined and levied through the one-stop shop mentioned under Question 1.

The other existing mechanisms, including mandatory arbitration, should remain available and readily accessible to the taxpayer. As the approach is formulaic - and hence no extended discussions on the facts and circumstances of transactions should take place, the cases should be solved within a limited (and possibly mandatory) timeframe (6 months to one year).

EBIT considers that the system must remain neutral for the taxpayer. It should be possible that elimination of double tax is relieved at the level of the respective (sub) entrepreneurs and not only at the level of the parent company.

Although some opposition towards the use of mandatory arbitration can be expected, the IF should realise that:

• This is a new form of taxation;
• The result is based on a formulaic approach to which the IF has agreed;
• The taxpayer should not suffer from the disagreements between countries;
• No real factual elements need to be discussed;
• No one country has experience with Amount A;

Therefore, if there is no up-front agreement, opposition toward arbitration seems unfounded.

Important issues that need to be solved are, for example, relief in the case of loss-making or low profit generating groups, whether these groups should not be carved out and issues related to limiting the credit of Amount A tax to the corporate income tax paid in the residence state. According to EBIT, credit of Amount A tax should be allowed to the greatest extent possible.

In recognition that Amount A is designed to give the destination country a share in the residual profits of the MNE, relief should be given against the Amount A that is allocated to a destination country for any residual profits that are recognized in that destination country under the arm’s length transfer pricing rules (i.e. for residual profits that are included in Amounts B and C, or that are otherwise recognized in the destination country if Amount C is defined to be less than all other profits under arm’s length transfer pricing rules). For simplicity, residual profits that are recognized in the destination country could be defined as all profits above a fixed profit margin in that country (and that fixed profit margin could be set at the same percentage as the deemed non-routine profit margin in the calculation of Amount A globally). This should avoid double taxation of Amount A in a manner that is consistent with the policy intent. The formulaic nature would allow administration through the one-stop shop approach by providing the designated tax authority with information on sales per country and taxable profits per country prior to the Amount A reallocation.

6. Amount B

Amount B relates to routine or baseline marketing and distribution activities. It is not possible, however, to give an exhaustive definition of the activities or functions (and associated risks) that can be covered under routine or baseline marketing and distribution activities. No two businesses are alike and structures may vary significantly.

In our opinion routine or baseline marketing and distribution activities are activities that can relatively easily be benchmarked according to normal transfer pricing rules. We understand the call for simplicity and therefore a ‘safe harbour’ or ‘simplified approach’ could possibly be created for routine or baseline marketing and distribution activities. Such safe harbour or simplified approach should be accepted by all participants in the IF in order to guarantee certainty and avoidance of disputes.

Typically, to our experience, routine or baseline marketing and distribution activities as we understand from the intention under Amount B would be very close to a distribution function with low risk. As under Amount A, we would welcome refinements based upon industry approaches and regional approaches. We wonder, however, whether the analysis to be performed to select the appropriate return under B - and its timely updates - based upon industry and regional approaches - is not akin to a benchmarking analysis. In any way, the determination of the baseline return must be principled based in order not to become a wholly arbitrary allocation. Taxpayers should be given the possibility to indicate that they fall below the return or outside the range determined by the safe harbour or simplified approach. This is different from amount C that only provides in topping up the routine or baseline return under amount B.

In our opinion, the way in which the limited orientations on Amount B are drafted in the Secretariat proposal, other activities than routine or baseline marketing and distribution activities are not covered under an approach based on fixed returns.

EBIT interprets the unified approach in such a way that activities that go beyond routine or baseline activities indeed fall under the normal transfer pricing rules. Also other activities such as manufacturing, research and development, procurement, financial services, etc., would fall under the normal transfer pricing rules; i.e. no fixed margins would apply to other activities than routine or baseline distribution activities (see also our response to Question 7 in that regard).

7. Amount C / Dispute prevention and resolution

The new nexus and allocation rules under Pillar I need to be assorted with detailed guidelines and commentaries to avoid as much as possible different interpretations of the new rules. Unfortunately, as with any new rule or approach, different interpretations of the rules will arise unavoidably, leading to disputes and controversy. Hence it is necessary that the members of the IF agree on mandatory dispute resolution rules, or, even better, avoid disputes from arising.

More clarity is needed as to what activities would be rewarded under Amount C (which appears to apply the ALP in its current form to those activities). The consultation document refers to firstly marketing/distribution in the market jurisdiction going beyond the baseline level of functionality and other business activities in the jurisdiction unrelated to marketing and distribution. What constitutes ‘marketing/distribution’ and the meaning of the ‘baseline level of functionality’ need to be precisely defined and otherwise there is a clear risk that many countries will continue to assert rights under Amount C in relation to most MNEs and any benefits of simplicity using Amount B will be reduced or lost completely. Consideration might also be given as to whether there might be a level of marketing/distribution which falls below the baseline level, which would be rewarded under Amount C but not Amount B (or might altogether be scoped out by a local nexus threshold).

EBIT Members believe that some kind of mandatory form of relief should be made available; in particular with regard to Amount A and with possible effect on Amount C. As Amount A is based on a formulary approach and as such does not necessitate a facts and circumstances analysis, Amount A is particularly well placed for dispute avoidance and relief through automatic rules: fixed determination of which State shall benefit from amount A and which State shall give relief under C.

This can be done under:

• Automatic relief rules;
• Certainty up-front through multilateral safe harbours or simplified conventions;
• Simplified multilateral APAs or rulings (allowing for faster conclusion)
• Mandatory arbitration rules: we urge members of the IF to sign up for mandatory arbitration as much as possible.

EBIT sees mandatory arbitration as a ‘last resort’ mechanism only. It finds its place in the mechanisms with the aim to urge tax authorities to arrive at a solution within an acceptable period.

The current system of APA is lengthy (in particular for bilateral APAs, and even longer for multilateral APAs) and finalisation of the APA comes at times when the transactions have already been executed. In those cases, the use and merits of bilateral or multilateral APAs can be questioned.

Finally, the current dispute resolution mechanisms are mainly systems between tax authorities where the taxpayer has only a limited role. We believe that the taxpayer should be warranted to play a more prominent role in the mechanisms through, for example, attendance and active participation at meetings and deliberations, and submission of their points of views and potential solutions.

Of late, some new mechanisms have been introduced such as ICAP, simultaneous tax audits or joint tax audits. Although these mechanisms may have some merits in enhancing compliance, their effect on dispute prevention and resolution for the time being is rather limited. EBIT considers these tools rather to be audit tools than relief tools.

EBIT Members trust that the above comments are helpful and are taken into account.

Yours sincerely,

European Business Initiative on Taxation – November 2019