TEI comments on OECD attribution of profits draft: October 12, 2004.

On October 12, 2004, TEI President Judith P. Zelisko sent to following letter to Jeffrey Owens, Head of the Centre for Tax Policy and Administration for the Organisation for Economic Cooperation and Development. The Institute's letter was prepared by a working group headed by Vincent Alicandri of Hydro One Networks, Inc. It was coordinated through TEI's Canadian Income Tax Committee, whose chair is David V. Daubaras of General Electric Canada, and TEI's International Tax Committee, whose chair is John J. Herson, of Neenah Paper, Inc. A working group of the European Chapter members, chaired by Peter H Taylor of Du Pont de Nemours International, also contributed to the submission.

On behalf of Tax Executives Institute, I am pleased to respond to the request for comments on the OECD's revised Discussion Draft on the Attribution of Profits to a Permanent Establishment--Part I (General Considerations), which was published on 2 August 2004 (hereinafter referred to as the "Discussion Draft"). TEI is concerned that the proposed rules will significantly increase the risk of double taxation, as well as increase the number of disputes between taxpayers and taxing authorities and between the taxing authorities themselves.

Background

Founded in 1944 as a non-profit organization in the United States to serve the professional needs of business tax professionals, TEI now has an international scope with 53 chapters throughout North America and Europe. Our 5,400 members represent 2,800 of the largest companies in the United States, Canada, and Europe. Representing a cross-section of the business community; TEI is dedicated to the development and effective implementation of sound tax policy, to promoting the uniform and equitable enforcement of the tax laws, and to reducing the cost and burden of administration and compliance to the benefit of taxpayers and governments alike. As a professional association, TEI is firmly committed to maintaining tax systems that are administrable and with which taxpayers can comply.

The majority of TEI members work for multinational companies with substantial international operations and sales. Members of TEI are responsible for managing the tax affairs of their companies and must contend daily with the provisions of the various tax laws relating to the operation of business enterprises. Consequently, TEI members have a special interest in the OECD's revised Discussion Draft on attributing profits to a permanent establishment.

Article 7(1) of the OECD Model Tax Convention on Income and on Capital provides that the profits of an enterprise of a Contracting State shall be taxable only in that State unless the enterprise carries on business in other States through a permanent establishment (PE). In those circumstances, the profits of the enterprise in the other State may be taxed only to the extent they are attributable to the PE.

An earlier version of the Discussion Draft developed and tested a working hypothesis concerning the preferred approach for attributing profits to a PE. Two broad interpretations of Article 7 are currently used by OECD Member countries:

* The "relevant business activity" approach, which refers to the profit of the business activity in which the PE has some participation and provides that the attributed profits may not exceed the profits that the whole enterprise earns from the relevant business activity; and

* The "functionally separate entity" approach, which refers to the profits that the PE would have earned at arm's length as if it were a distinct and separate enterprise performing the same or similar functions under the same or similar conditions, determined by applying the arm's-length principle under Article 7(2).

The Discussion Draft avers that, if the latter approach were more explicitly supported in the Commentary, Member countries would generally prefer using the functionally separate entity approach because it is "simpler, more administrable, and more consistent with the understanding of the arm's-length principle" as applied under Article 9 (Associated Enterprises). Discussion Draft at [paragraph] 23. The Discussion Draft proposes to adopt this approach as the authorised OECD approach.

Executive Summary

TEI commends the OECD for its collaborative approach in seeking to provide clearer rules. As the Discussion Draft recognises, the domestic laws of OECD Member countries vary regarding the taxation of PEs and the lack of a common interpretation and consistent application of Article 7 may lead to double taxation. We appreciate the opportunity the OECD has provided the business community to comment on the Discussion Draft.

As a representative of many large, complex Multinational Enterprises (MNEs), TEI supports the development of appropriate rules to govern international taxation. As important, however, are effective rules. The Institute is concerned that the Discussion Draft may be interpreted to lower the threshold for determining source-country taxation of non-resident companies, which will impede the ability of MNEs to compete efficiently. Although the Discussion Draft disavows lowering that threshold, several statements in the draft may regrettably be construed to lend support to increased PE-based taxation.

In TEI's view, effective rules have two key elements:

* First, the rules must be embedded in the OECD Model Treaty to ensure that when taxpayers follow the administrative rules, the courts will respect the results. Regrettably, in some cases the Discussion Draft does not meet this test. The proposed rules are likely to spawn disagreements between taxpayers and taxing authorities and, indeed, between taxing authorities themselves.

* Second, where practicable, the rules should align with common business practice by respecting the taxpayer's books and records where (i) they are consistent with the functional and factual analyses; and (ii) the profit attributed to the PE is within the range of results of comparable separate and distinct enterprises. Insufficient attention has been paid to this issue. The Discussion Draft tends to focus on the difficulties experienced by taxing authorities rather than recognising that business practices may form a reasonable and economical basis for attributing profit to a PE. (1)

The concepts set forth in the Discussion Draft should apply only in respect of the attribution of profits of a PE under Article 7, and only after a PE has been established under Article 5. This should be clarified in the preamble.

Finally, TEI recommends that the effective date of the proposed rules be delayed until there is greater clarity concerning their application and a greater understanding of their effect on multinational companies and government revenues.

C-1: Form

The proposed authorised OECD approach hypothesises the PE as a distinct and separate enterprise engaged in the same or similar activities under the same or similar conditions, and it attributes profits to the PE using the OECD Transfer Pricing Guidelines. It applies these Guidelines by analogy, by adapting them where required to take into account factual differences between a PE and a legally separate enterprise. The Discussion Draft provides that the attribution of capital must be carried out in accordance with the arm's-length principle "to ensure that a fair and appropriate amount of profit is allocated to the PE." Discussion Draft at [paragraph] 62.

Thus, a two-step analysis is required: (i) a functional and factual analysis must be made to hypothesise the PE and the remainder of the enterprise (or a segment) as if they were associated enterprises, each undertaking functions, using assets, and assuming risks; and (ii) an analysis of the Transfer Pricing Guidelines must be made, applying the arm's-length principle to the hypothesised enterprises. Of particular importance under step (i) is the determination of the enterprise's key entrepreneurial risk-taking (KERT) functions and the extent to which the PE undertakes one or more of those functions. See Discussion Draft at [paragraphs] 54-56.

It is important that the structure of the final report be consistent with the objectives of fairness and administrability that underlie the consultative process taken by the OECD. To aid comprehension, TEI recommends that section C-1 be revised to describe the essential framework of the analysis, before jumping into its complexities.

Specifically, section C-1 should open with a paragraph similar to paragraph 69 of proposed Part II, which provides, as follows:

The assets and risks recorded in the accounts and books of the PE form a practical starting point for determining whether the economic ownership of assets has been assigned to the location where the key entrepreneurial risk-taking functions were performed. The accounts and books should be respected for tax purposes, provided they reflect an attribution of assets and risks that is consistent with the functional and factual analysis. There may, however, be cases where the accounts and records are inconsistent with the functional and factual analysis, for example, because material amounts of assets and risks may be booked in a location where none, or very few, of the key entrepreneurial risk taking functions related to their creation or subsequent management were performed. Respecting the booking location in such cases would not lead to an arm's length attribution of profit.

We also suggest that a statement be added recognising that, as in all transfer pricing cases, there is a range of acceptable results.

TEI understands that the principles described in section C-1 will apply only for the purposes of computing profit and are not relevant for other purposes, such as withholding tax. Paragraph 246 of the Discussion Draft expressly recognises this concept with respect to notional royalties by providing--

   The recognition of the notional
   royalty is relevant only
   to the attribution of profits to
   the PE and should not be understood
   to carry wider implications
   as regards withholding
   taxes, which are outside
   the scope of the Report.

TEI recommends that the OECD consider including a statement in its final report that the basic principles enunciated in section C-1 are relevant only to the attribution of profits to the PE and should not be used in respect of withholding taxes.

C-1: Basic Principles Used to Attribute Profits to a PE

The Discussion Draft's reliance on the OECD Transfer Pricing Guidelines may lead tax authorities to apply the Draft's principles outside the context of a PE. A basic principle underlying the authorised OECD approach is that capital follows risks and the economic owner will be attributed the capital necessary to support the associated risks. See Discussion Draft at [paragraph] 59. In the context of associated enterprises, however, the Guidelines clearly recognise that the provision of capital is a risk in itself that should be considered in determining the economic owner of an asset for transfer pricing purposes. For example, paragraph 7.41 of the Guidelines provides that the results of "broadly defined" research and development services provided by one member of a group to another member funding such services are economically owned by the enterprise funding the services. (2)

To avoid transfer pricing disputes, the final report should clarify that the concept that capital follows risks is not applicable in the context of transfer pricing between associated enterprises under Article 9.

C-2 (i), (ii), and (iii): KERT

Paragraph 81 of the Discussion Draft states that--

   The functional and factual
   analysis needs to be carried
   out in a thorough and detailed
   manner in order to establish
   the exact nature of the function
   being performed. This is
   because where the functional
   analysis has determined that
   the PE has performed the
   key entrepreneurial risk-taking
   [KERT] functions,
   the PE will be attributed the
   assets and risks associated
   with those functions. This
   in turn leads to the attribution
   to the PE of the income
   and expenses associated with
   those assets and risks.

Paragraph 86 then provides that--

   In determining the characteristics
   of the PE for taxation
   purposes, it is the economic
   (rather than legal) conditions
   that are most important because
   they are likely to have
   a greater effect on the economic
   relationships between
   the various parts of the single
   legal entity. Economic ownership
   of an asset, whether
   physical or intangible, is
   determined by a functional
   and factual analysis, and in
   particular rests upon performance
   of the key entrepreneurial
   risk-taking functions
   in relation to the asset.

Absent clarification, the Discussion Draft might be viewed as elevating the KERT concept to a level where it supplants the actual language of the OECD Model Treaty. While the KERT concept may be useful in determining the location of assets without a clear home, it must be applied in a manner consistent with the treaty and the facts. Some assets have an indisputable physical location. For example, an asset capable of creating a PE in Country A under Article 5 should be attributed to Country A, regardless of the location of the KERT functions. Of course, this does not mean that the KERT functions are not entitled to a fair return under principles analogous to Article 9. Paragraph 82 recognises this issue in the case of an e-commerce operation.

Similarly, where profit is booked in a manner "which it might be expected to make if it were a distinct and separate enterprise engaged in the same or similar activities under the same or similar conditions and dealing wholly independently with the enterprise of which it is a permanent establishment," (3) such treatment should be respected.

The OECD approach for allocating income to PEs depends upon the geographic location of the KERT functions, which are defined as "those which require active decision making with regard to the most important profit generators of the business." As a result--

* The determination of key profit generators or their functional analysis will be a significant factor in the ultimate income allocation. This functional analysis is fact-intensive and prone to varying economic and subjective interpretation, thereby leading to conflicts not only with the taxing authorities of the home country but with the host countries as well.

* Significant and burdensome documentation will be required in respect of the PEs and their attribution of assets, measurement of risks, and attribution of free capital and interest expense.

It is also important to recognise that KERT functions may change over time, for example, because of a reorganisation or centralisation of functions or a change in residence to a third country. MNEs must have flexibility to operate in the most efficient manner. The tax implications of KERT-related asset movements may be substantial. This is further compounded by the concept of KERT functions creating partial ownerships in different locations, as set out in paragraph 199.

Finally, when the KERT functions have been identified, whether any profits can be allocated to the PE should depend on where the important decisions related to those functions are made. In many cases, the KERT functions may reside in more than one country and decisions related to the KERT function may be made in both countries. The nature of the decision, however, is likely to be different. For example, the main decision-making power may be centralised in one country, but some limited authority may be delegated to the local operation. These cases may trigger lengthy and costly disputes with tax authorities, especially if the responsibility level of the local entity is reduced. The concept may also undermine the longstanding management and control test.

Consider an example where the local sales person has some authority to deviate from established price lists. The relatively minor level of authority to bind may encourage aggressive tax authorities to mistakenly assert that a dependent agent PE exists. If the determination of the price list itself and the level of authority exists outside the local sales country, the most important profit generators should be considered as residing outside the territory and hence no profit should be allocated to the dependent agent PE. The final report should be revised to reflect these common business practices.

C-2 (iv): Credit Worthiness

The Discussion Draft provides that the creditworthiness of the enterprise as a whole will be attributed to a PE. The OECD states that "it is an observable condition that permanent establishments generally enjoy the same creditworthiness as the enterprise of which they are a part." Discussion Draft at [paragraph] 91. TEI questions the validity of the propositions set forth in C-2(iv), and submits that the concept characterizing an attribute of the enterprise as a whole to a PE has not been consistently applied throughout the paper.

Paragraph 94 acknowledges that many factors contribute to the credit rating of an enterprise. The benefit associated with many of these factors, however, may not be attributable to any geographic location and in such cases it is reasonable to share the benefits, as long as the concomitant costs and capital savings are shared. Other factors clearly have a geographic location. For example, a common factor increasing ratings is a stable, sustainable source of income. This may well exist in only a few countries. It may be reasonable to assume that a capital allocation model based on economic capital already takes this factor into consideration. Where other models are used, however, a capital charge from such sources seems appropriate. Obviously, there are geographical or political reasons for negative effects on credit ratings as well.

Based upon a factual analysis, where a taxpayer is able to support that the creditworthiness of a PE is different from that of the entire enterprise, this method should be permitted as an authorised OECD approach.

Further, the rules for allocating expenses to a PE should specifically provide that expenses attributable to maintaining a credit rating should be shared, even though the Transfer Pricing Guidelines do not support recognition of charges for items in the nature of stewardship expenses. These expenses form the foundation for the credit rating. This view is consistent with Article 7(3), relating to the deduction for executive and general administrative expenses.

C-2 (v): Free Capital

Paragraph 98 of the Discussion Draft provides that, because interest expense is generally deductible, an appropriate amount of the enterprise's "free" capital must be attributed to a PE "in order to ensure an arm's length attribution of profits to the PE." Free capital is to be allocated to where the risk-taking functions are performed, and not necessarily to the location where the capital is actually deployed. This allocation is very subjective, will lead to limitations on interest deductibility, and will create administrative burdens and documentation requirements to substantiate taxpayer allocations to home and host countries.

With the development of global capital markets, countries have long abandoned regulated capital requirements for non-financial businesses to provide a "cushion" for risks inherent in the business. The ability of businesses to insure risks and relatively quickly obtain new capital ensures international expansion and capital efficiency.

Free capital should not be attributed to PEs on a notional or contingent basis, but only when it is actually used in the business.

C-3 (iv)(a): Capital Assets

This section focuses on determining the ownership of capital assets at the time of acquisition. Paragraph 198 provides that physical assets may be either owned or rented and "there are commercial pros and cons with either option." The starting point is to determine the nature of the risks related to the asset used in the PE, which will vary according to whether the asset is legally owned. Where the physical asset is acquired by the MNE as a whole but located in or used exclusively by the PE, an issue arises concerning which part of the enterprise is considered the economic owner.

Paragraph 199 provides that the economic ownership of an asset belongs with the part of the enterprise performing particular KERT functions in respect of the asset. The commentary is troubling. As paragraph 198 seemingly acknowledges, for many assets "a distinct and separate enterprise engaged in the same or similar activities under the same or similar conditions" may choose various ways to obtain the use of such assets. In the ordinary case, a functional analysis will not lead to a conclusion on the method (rental or ownership) that would be most appropriate for the PE.

TEI urges that the final report recognise that where the books reflect the legal ownership of assets by the PE, the books should be determinative of the method. Determining the "economic ownership" of the asset should not be necessary.

C-3 (iv)(v): Dependent Agent PEs

The Discussion Draft provides a detailed commentary on the treatment of PEs that are deemed to arise because of the activities of a dependent agent, providing that such dependent agent PEs should be subject to the same profit attribution rules as actual PEs. Although it does not rule out the possibility, the Discussion Draft confirms that the activities of a mere sales agent are unlikely to result in attributing profit in respect of marketing or trade intangibles. The draft specifically rejects the argument, however, that the payment of an arm's-length price to the dependent agent PE precludes the attribution of additional profits to that PE, noting that "this approach would not result in a fair division of taxing rights between host and home jurisdictions as it ignores assets and risks that relate to the activity being carried on in the source jurisdiction simply because those assets and risks legally belong to the non-resident enterprise." Discussion Draft at [paragraph] 275.

An example in the Discussion Draft suggests that a dependent agent PE may be required to recognise profits in addition to those already reported in the entity performing a sales function. (4) If the local entity performs KERT functions and therefore attracts to its location assets and risks under the methodology noted in the Discussion Draft--i.e., the local entity provides services both to and on behalf of the foreign company--additional profits may have to be allocated because the assets and risks associated with the KERT functions are allocated to the (dependent agent) PE. The assets and risks so allocated to the PE would then attract income in accordance with the Transfer Pricing Guidelines.

The problems associated with this interpretation can be illustrated by an example where the local entity is performing warehousing services (in addition to sales activities). If a dependent agent PE is recognised, the compensation for the inventory risk may be part of the profit of that PE. As a practical matter, whether an activity is considered a KERT function will be difficult to determine (e.g., the local warehousing function will not necessarily manage the inventory risk).

Consider also an arrangement where decision-making is highly centralized but some level of authority is left locally or the local function provides support for the decision-making at the foreign company. If a dependent agent PE is recognised, to what extent can the activities performed locally be considered KERT functions, thus attracting additional assets and risks resulting in taxable income? How is the profit determined? What is the effect if the sales function (causing the dependent agent PE to exist) is completely segregated (functionally and geographically) from the warehousing functions? As illustrated above, it is difficult to distinguish between the provision of services to the principal/foreign company for a fee and the provision of services on behalf of a principal/foreign company. These practical issues may outweigh the policy objective of allocating profits strictly based on a functional analysis.

It is also unclear which functions are covered by an arm's-length fee paid to the dependent agent and which are attributable to the dependent agent PE. Paragraph 271 provides that--

   On the one hand the dependent
   agent enterprise will be
   rewarded for the service it
   provides to the non-resident
   enterprise (taking into account
   its assets and its risks
   (if any)). On the other hand,
   the dependent agent PE will
   be attributed the assets and
   risks of the non-resident
   enterprise relating to the
   functions performed by the
   dependent agent enterprise
   on behalf of the non-resident,
   together with sufficient free
   capital to support those assets
   and risks.

If a PE is asserted, are the PE functions previously included and covered by the service fee shifted into the PE from the dependent agent enterprise? Or are the assets and risks related to any PE KERT functions recognised and thus related profit is attributable to the PE? It would appear that the former is intended. If this is correct, what are the implications for the dependent agent enterprise, i.e., has it been overpaid for the service provided to the non-resident enterprise or is the service fee part of the PE costs? We recommend that the OECD address these issues by providing examples to illustrate the principles of this requirement.

TEI regrets that the Discussion Draft may embolden taxing authorities to conclude that more income should be attributed to associated enterprises as PEs under Article 7 than warranted. This is particularly true because jurisdictions will likely differ on the identification of a given activity as involving a KERT function. Although the Discussion Draft faults the "single taxpayer" approach for inadequately allowing for the attribution of risks to a PE, the proffered solution goes too far in the other direction by potentially undermining the ability of associated enterprises to structure their relations in a legally and economically reasonable manner.

By failing to sustain an adjustment based on a dependent agent enterprise's transactions with the nonresident enterprise, the authorised OECD approach will permit taxing jurisdictions to sidestep the Article 9 hurdle and attribute assets and income under Article 7 to the same enterprise as a dependent agent PE. It will be necessary to assert only that the associated enterprise is engaging in a KERT function on behalf of the non-resident enterprise to successfully avoid the limitations of Article 9. Further, the existence of substantial KERT functions in an associated enterprise is largely inconsistent with the agent being "dependent."

Finally, TEI believes that where the dependent agent is an associated enterprise, Article 9, rather than Article 7, should govern the taxation of benefits in the host country. In this case, Articles 5(7) and 9 should be read in conjunction with Article 5(5), with Article 5(5) used only for unassociated dependent agents who are analogous to employees.

Some countries have approached this problem by negotiating specific treaty provisions. See, e.g., the protocol to Article 5 of the Austria-Germany Treaty. (5) This seems a more prudent approach.

Loss Enterprises

With respect to an MNE that incurs losses, the functionally separate entity approach espoused in paragraph 28 will lead to double taxation. For example, a host country may assert profits in its jurisdiction that the head office country may reject since the result will be an increase in losses in the head office jurisdiction. The consequence of a loss in one jurisdiction should be addressed in the final report.

Financing Costs

A governing principle in allocating profits to a PE should be that amounts treated as interest (other than interest that is capitalized) under the rules of section C-2(v) remain deductible to the PE, regardless whether the interest has been recorded in the books of account. For example, this situation may arise because of audit adjustments or reallocation of capital by taxing authorities.

Conclusion

The OECD should strive to adopt principles that are simple, easy to ado minister, and reflect sound tax policy, all with an objective of avoiding double taxation. The Discussion Draft is unfortunately inconsistent with these principles. Rather, the proposals significantly increase the risk of double taxation. In addition, they appear to seek a level of tax accuracy that imposes enormous documentation requirements on a PE, based upon notional transactions requiring functional analyses that are fact-intensive and prone to subjective interpretation. Taxpayers will find it difficult to determine how the proposed rules apply within the MNE.

In TEI's view, the proposals grant inappropriate power to taxing authorities to allocate profits to their jurisdiction, relying on transfer pricing principles that are subjective, difficult to substantiate, and prone to revision years after the fact. Basically, all cross-border transactions and operations will require functional analyses to substantiate income and expense allocations for tax and regulatory purposes in the future. This will result in--

* Increased transfer pricing data and justification (particularly with respect to taxpayers whose businesses are dependent upon significant intangible assets);

* Substantial administrative burden and cost to the taxpayer by requiring yet another transfer pricing study and contemporaneous documentation based upon notional transactions;

* Differing country approaches, particularly when the entity has a PE in various jurisdictions and each is using its own transfer-pricing methodology that is not limited by the actual profits earned by the entire entity. Consequently, it is possible that the entity will be taxed on amounts that in the aggregate exceed the net profits from the transactions. Furthermore, double taxation may still occur because of timing and permanent differences (6);

* A complex and impractical dependent agent concept, which contradicts the legal theory recognising the separate existence of different incorporated entities;

* Complex capital allocation and methodology; and

* Subjective functional analysis, which will lead to disagreements on transfer pricing methodologies (including profit-split issues, capital allocation, functional importance, and asset allocation). In the case of a MNE with complex functions and risks (such as unique intangibles), it is difficult to determine the appropriate arm's-length price because comparables are often hard to find. These difficulties will only be exacerbated by the hypothetical nature of the authorised approach.

For the above reasons, TEI recommends that the effective date of the proposed rules be delayed until there is greater clarity about their application and a greater understanding of their effect on MNEs and on government revenues. We would be pleased to meet with you to discuss our comments.

Tax Executives Institute appreciates this opportunity to present our views on the OECD's Discussion Draft on the attribution of profits to a permanent establishment. If you have any questions about the Institute's views, please contact TEI's acting General Counsel and Director of Tax Affairs, Mary L. Fahey, at 1.202.638.5601; [e-mail].

(1) See Discussion Draft at [paragraph] 63, acknowledging that there is a range of potential capital attributions within which it is possible to achieve the basic principles of the OECD approach. We are concerned that the variety of options may be used by host country tax authorities to reach a different result, even where the taxpayer's method is reasonable. This will create uncertainty, increase costs, and result in greater use of the Mutual Agreement Procedures, raising the potential of double taxation.

(2) The rules should also expressly exclude the attribution of profit to a PE in respect of stewardship functions conducted by the non-resident enterprise. In the wake of the U.S. Sarbanes-Oxley Act of 2002, publicly listed entities are incurring additional stewardship expenses to document internal controls and ensure that their financial position is fairly stated. Although these activities should qualify under Article 5(4) as preparatory or auxiliary in nature, the final report should clarify that the provision of executive, managerial, or administrative facilities or services for the non-resident enterprise will not attract any attribution of profit to the PE.

(3) See OECD Model Treaty, Art. 7(2).

(4) In this transaction, there are apparently three taxpayers: the local sales entity, the dependent agent PE, and the foreign company.

(5) "It is agreed that in the case of associated enterprises, neither of these enterprises shall be treated as an agency permanent establishment, provided that the respective functions--which would lead to the assumption of an agency permanent establishment in the absence of this agreement--are compensated by appropriate transfer prices, which include a remaining profit of an enterprise."

(6) See Discussion Draft, [paragraphs] 46-47.