2 The Development of the ALP
Richard S Collier, Joseph L Andrus
Richard Collier, Joseph L Andrus
- Formation of contract — Interpretation of contract — Payment of price
Much more importance is attached to the desirability of interfering as little as possible with existing business organization and of refraining from inflicting demands for information on foreign enterprises, which are unnecessarily onerous.
1963 OECD Model Commentary on Article 7(1)
2.01 After the dissolution of the League of Nations, it had been expected that the then newly formed United Nations would continue the work formerly carried on by the League of Nations Fiscal Committee.1 In 1946, the Economic and Social Council of the United Nations set up a Fiscal Commission to study and advise the Council in the field of public finance, particularly in (p. 51) its legal, administrative and technical aspects.2 However, the work of the United Nations Fiscal Committee proved short lived and the Committee was disbanded in 1954. The principal multilateral work on international tax matters passed to the Organisation for European Economic Co-operation (OEEC), which created its own Fiscal Committee in 1956 and from 1958 set about the task of creating a new model double tax treaty.3 The OEEC changed its name to the Organisation for Economic Co-operation and Development (OECD) in 1960, following the accession to membership of the US and Canada.4
2.02 Since its formation, a very significant part of the tax work of the OECD has been concerned with the development of model double tax treaties. The early work of the OECD led to the 1963 Draft Model Double Tax Treaty (referred to here as the ‘1963 Draft’). This was subject to further revisions from 1971, leading to the 1977 revised Model Double Tax Convention (the ‘1977 Model’). The 1977 Model was in turn revised in 1991, at which time the OECD adopted the concept of an ambulatory Model that could be regularly updated (the subsequent 1992 version of the model is referred to here as the ‘OECD Model’). There have been further, and more regular, revisions of the OECD Model from that time.5 Since 1963, the model treaties produced by the OECD have been accompanied by detailed Commentaries, which are intended to aid the interpretation of each Article.6 Comprehensive consideration of the work of the OECD on model double tax treaties is beyond the scope of this book except in so far as it concerns matters relating to the development and application of the ALP. On that topic, the OECD has been very active in expanding the thinking and guidance on transfer pricing and income attribution matters, both generally and in relation to specific technical issues and specific business sectors.
2.03 While the primary work on double tax treaties shifted to the OECD, important developments regarding transfer pricing and income attribution have come from two other sources. First, individual country developments both in the courts and through administrative and legislative pronouncements have furthered the evolution of the ALP. Developments in the US have been especially important. Second, the United Nations has played a role, particularly as it relates to the application of treaties and the ALP to developing countries which, until recently have not been fully engaged in the work of the OECD.
2.04 Given the interest of this book in charting the emergence and evolution of the ALP, the discussion in this chapter seeks to survey the more important developments from each of (p. 52) these sources, generally in chronological order. The discussion is inevitably incomplete, given the scale of developments in the period. It does, however, survey the most important steps in the evolution of the ALP from the time of its adoption to the beginning of the BEPS Project.
2.05 As will be seen, the ALP has evolved significantly since its original adoption in the 1920s and 1930s. In very broad terms, the stages of its evolution may be characterized as involving four distinct phases. The period prior to 1968 was a period in which the primacy of the ALP as an income allocation tool was established. During this period the OECD had an almost complacent belief in the problem-free operation of the rule, although at a national level problems began to be apparent. The period from the mid 1960s to the mid 1980s represented a sharp corrective to the position of complacency and it was in this period that the emphasis shifted to a highly detailed elaboration of transactional pricing, reflected first in the US 1968 regulations on section 482 of the Internal Revenue Code, and subsequently in the early transfer pricing reports of the OECD in 1979, 1984, and 1987. In the period from the mid 1980s, that detailed transactional pricing approach was taken further, and some of the perceived fundamental difficulties arising from that approach began to be addressed through litigation, administrative pronouncements, and legislation and through modification and elaboration of the OECD Guidelines. Notable in this period was the adoption and international accommodation of the US commensurate with income principle and the general acceptance of profits-based methods. Also notable in this period was a sharp focus on the problems associated with transfer pricing for intangibles. The period since 1995 has been characterized by the proliferation across the globe of highly detailed compliance regimes relating to transfer pricing and by the increasingly difficult issues—both practical and conceptual—that have emerged. All these developments are considered in the discussion below.
2.06 In the period following adoption of the ALP by the League of Nations, a process of elaboration and evolution that would characterize the next seventy years of development began. Initially the most influential developments took place in individual countries, none more important to the process than the US. In that country, important developments first took place in the courts. For purposes of the discussion here, those developments can be grouped into two broad categories. First were a series of judicial decisions that had the effect of establishing the role of the ALP as the exclusive standard to be applied in transfer pricing and income attribution matters. Second were a series of cases that affirmed that the ALP had a role to play in regulating transfers of income-producing assets via capital contribution by affirming that the transfer pricing rules apply to so-called non-recognition transfers.
1. Judicial Confirmation of the ALP as the Exclusive Income Allocation Standard
2.07 Beginning in about 1935 and continuing for nearly thirty years the US courts decided a number of cases that turned to some extent on the standard that should be applied in deciding whether an allocation of income between two associated enterprises was appropriate. The US statute differs to some extent from the treaty language adopted by the League of Nations embodying the ALP. The US statute grants the Internal Revenue Service the (p. 53) authority to make adjustments to the income of associated enterprises where necessary to prevent the evasion or avoidance of taxes by such entities, or ‘clearly to reflect the income’ of such enterprises.7
2.08 In 1935, the Internal Revenue Service (IRS) adopted regulations that indicated that the arm’s length principle was the ‘standard to be applied in every case’. However, the courts were not quick to accept that prices had to be arm’s length in the sense that they needed to be supported by prices of comparable uncontrolled transactions.8 Instead, when precise comparables were not readily available, the courts applied standards based on prices being ‘fair’ or ‘reasonable’ or ‘fairly arrived at’ and did not always demand a reference to comparable uncontrolled transactions. Over time, however, the courts migrated to a position that supported the statement in the regulations that the ALP was to be applied in every case, or at least in every case where reliable comparables were available.9 By the mid 1960s, the ALP had been established as the exclusive standard to be applied, though there was still a good deal of uncertainty about how to approach cases where there were no useful comparables.
2.09 This movement to confirm the pre-eminent position of the ALP was given encouragement by legislative developments. In the early 1960s concern arose in the US that companies were shifting income into tax havens and that the existing transfer pricing rules were not effective in preventing such actions. President Kennedy’s administration proposed a comprehensive overhaul of the rules on taxing international operations of US companies in the early 1960s. As part of its consideration of those proposals, the House of Representatives adopted legislation that would have applied a formulary apportionment approach as a default approach in situations where reliable comparables could not be identified.10 When the House bill was considered in the Senate, however, the business community was successful in convincing the legislators that the formulary apportionment backstop should not be adopted and that instead the problems with the ALP could be addressed through regulations. This legislative outcome led, after the passage of several years, to the publication of the 1968 transfer pricing regulations by the IRS, which are discussed in greater detail, below.
2. Cases on the Treatment of Capital
2.10 A separate line of US cases established very early that the transfer pricing rules had a role to play in situations where assets were transferred by way of capital contribution. In the Northwest Securities case,11 the taxpayer transferred appreciated securities to an affiliate by way of capital contribution immediately prior to the disposition of the securities. The affiliate transferee had losses that could be used to offset the gain on the disposition of (p. 54) the securities. When the transferee sold the securities, offsetting the gain on the sale with its unrelated losses, the IRS challenged the transaction arguing that the transfer pricing statute gave it authority to reallocate the gain on the disposition of the transferred securities back to the transferor, primarily on ‘clear reflection of income’ grounds. The court agreed with the IRS position without mentioning the ALP.
2.11 In a similar case, Central Cuba Sugar,12 the taxpayer transferred growing crops to a subsidiary by way of capital contribution immediately before the crops were harvested and sold. The court reallocated pre-transfer expenses of growing the crop to the transferee under the authority of the transfer pricing statute. Again, the notion was that the transfer pricing rules in the US had a sufficiently broad reach to demand that deductions incurred in planting and growing the crops be aligned with the income derived as a result of those deductions, notwithstanding the tax-free capital contribution of the growing crop. The decision relied heavily on the ‘clear reflection of income’ leg of the US statute.
2.12 These cases did not turn on the identification of comparables, or on any attempt to mimic the behaviour of unrelated parties. These early cases established that the US transfer pricing statute could be applied, at least in some circumstances, to challenge transfers of capital assets between affiliated companies that otherwise qualified for non-recognition treatment. The principles the cases established would, however, take on greater importance in the 1970s and 1980s when it became common for taxpayers to shift intangibles and other capital assets to affiliates in tax havens, particularly Puerto Rico, in order to shift income into those tax havens.
3. OECD 1963 Draft Model Treaty
2.13 The judicial and legislative developments in the US were consequential for the first serious consideration of the ALP at the OECD, encouraging a multilateral reaffirmation of the ALP as the exclusive transfer pricing standard. In 1963 the OECD released a draft of an updated model income tax treaty. The 1963 Draft was not a development that was focused on the ALP in particular, as the objective of the OECD at the time was to overhaul all of the model treaty provisions and supplement them with detailed commentaries. The 1963 OECD Draft Model Treaty did follow the growing consensus in the US and elsewhere that the ALP should be the sole transfer pricing standard and introduced the contemporary version of the ALP rule, entitled ‘Associated Enterprises’. This language has remained unaltered to date and is worded as follows:
Article 9, Associated Enterprises
Where a. an enterprise of a Contracting State participates directly or indirectly in the management, control or capital of an enterprise of the other Contracting State, or b. the same persons participate directly or indirectly in the management, control or capital of an enterprise of a Contracting State and an enterprise of the other Contracting State, and in either case conditions are made or imposed between the two enterprises in their commercial or financial relations which differ from those which would be made between independent enterprises, then any profits which would, but for those conditions, have accrued to one of the enterprises, but, by reason of those conditions, have not so accrued, may be included in the profits of that enterprise and taxed accordingly.
(p. 55) 2.14 The revised formulation of the ALP contained in the Draft corresponded in its essential elements to the version first introduced in Article 5 of the 1933 Model Treaty on the Allocation of Profits (which was reproduced in the drafts of the London and Mexico Model Tax Conventions). The basic requirement of the rule remained that the conditions of the commercial or financial relations between associated enterprises should be consistent with those found between independent enterprises. However, the drafting of the rule was modified in various respects.13 The relevant 1963 expression of what it is to be an ‘associated enterprise’ is captured in different terms to those used in 1933 but the approach adopted is broadly similar, in each case operating by reference to participation in management, capital, or control.14 As regards the scope of the rule, reference to the ‘commercial or financial relations’ of associated enterprises is common to both the 1933 and 1963 versions of the test. The 1963 version includes a reference to the relevant conditions being ‘made or imposed’ which did not feature in the earlier 1933 version and the earlier reference to profits having been ‘diverted’ has been replaced by the more neutral wording ‘but, by reason of those conditions, have not so accrued’. The mechanism for adjustment (e.g. price adjustment, deeming of income) is not included in Article 9 but is left to domestic law.
2.15 The new version of the rule was accompanied by a Commentary comprising a solitary (and very short) paragraph which records that the relevant tax authorities may ‘re-write’ the accounts of the enterprises ‘if as a result of the special relations between the enterprises the accounts do not show the true taxable profits arising in that country’. It is also recorded that ‘[i]t is evidently appropriate that rectification should be sanctioned in such circumstances’ and that ‘the Article seems to call for very little comment’.15 The impression given is that thinking on the ALP had developed little over the three decades since 1933 and that at this point there remained a ready confidence in the unproblematic operation of the rule. As with the previous version, the amended formulation of the ALP expresses the rule in very general and broad terms (e.g. the reference to ‘conditions made or imposed’ in relation to ‘commercial or financial relations’ which seems to take the potential scope of the provision beyond mere pricing arrangements or contractual terms).16
2.16 With the advent of the contemporary version of the ALP, it is relevant to consider the purpose of the rule when enacted in double tax treaties. Though probably most of the discussion relating to the development of the rule, including the rationale for it originally given by Mitchell B. Carroll and the comments referred to above from the 1963 Commentary (p. 56) are couched in terms of it being required to prevent or correct abuse, the relevant historical background (discussed in Chapter 1) makes it clear that the rule is essentially concerned with preventing economic double taxation (i.e. the situation where two different taxpayers are taxed in respect of the same income).17 This is achieved by ensuring domestic anti-avoidance rules that permit profit adjustments in cases involving related parties do not conflict with the requirements of the ALP by going beyond what is permissible under the ALP. In other words, it is intended that the uniform ‘arm’s length’ allocation standard of the ALP will prevent economic double taxation by the application of a uniform rule under treaties for the distribution of taxable profits among the states in which MNE groups do business. It is generally (but not universally) accepted that the treaty provision is not ‘self-executing’—that is, it does not, in itself, create a liability to tax and it will be necessary for a state to have in its domestic law transfer pricing provisions that operate to adjust the effect of any abnormal terms in arrangements between related parties.18 This point is significant because it underlines the need for a global consensus on the operation of the ALP. Without such a consensus, the ALP would fail in its primary function of avoiding economic double taxation. As will be seen, while there has been broad support for the principle of the arm’s-length test, there has not been a constant and uniform consensus on its application in practice.
2.17 The 1963 Draft also contains a materially expanded version of the rules dealing with the allocation of profits for permanent establishments. The basic rule, which seems clearer in its purpose and function than the Article 9 rule, continues to be that business profits of an enterprise of one contracting state are to be taxed in the other contracting state only if the enterprise has a permanent establishment in that other contracting state. If that is the case, tax may be levied in the other contracting state, but only on the profits that are ‘attributable’ to that state. The rule is expressed in Article 7(2) of the 1963 Draft as follows:
Article 7(2), Business Profits
Where an enterprise of a Contracting State carries on business in the other Contracting State through a permanent establishment situated therein, there shall in each Contracting State be attributed to that permanent establishment the profits 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.
2.18 This 1963 wording also retains the essential elements of the approach as first expressed in the 1933 Model Treaty on the Allocation of Profits (again, as reproduced in the drafts of the London and Mexico Model Tax Conventions). In the 1963 Draft, the principle is supplemented by various other provisions in the Article which: clarify the deductibility of expenses incurred for the purposes of the permanent establishment, whether in the same state or elsewhere (Article 7(3)); permit the use of customary apportionment methods to attribute profits provided the result is in conformity with the principles of the Article (p. 57) (Article 7(4)); provide that no profits are to be attributed by reason of the mere purchase of goods or merchandise for the enterprise (Article 7(5)); require a consistency of method for the attribution of profits in the absence of a good reason to the contrary (Article 7(6); and provide that if items of income are dealt with separately in other articles then the provisions of those articles are to be unaffected by Article 7 (Article 7(7)). The expanded Article 7, which remained largely unchanged until 2010, is accompanied by a detailed Commentary. The detailed ‘four step’ guidance on the approach to the calculation of taxable profits which had first appeared in Article 3 of the 1933 draft model on the allocation of profits (and which was also reproduced in the London and Mexico drafts) was not included in the text of Article 7 of the 1963 Draft. However, the 1963 Draft contained a much lengthier Commentary (compared to the 1933 draft and the London and Mexico drafts), dealing with the approach to the attribution of profits to permanent establishments and this Commentary includes points that previously were included in the 1933 guidance.19
The organisation of modern business is highly complex. In OECD countries, there are a considerable number of companies each of which is engaged in a wide diversity of activities and is carrying on business extensively in many countries. Current trends of political thought in Europe seem likely to make such companies even more common in future than they are at present.20
2.20 Apart from its restatement and modest reformulation of the basic transfer pricing rule in the 1963 Draft, OECD developments on the ALP remained at a relatively low ebb, a position that would endure until the mid to late 1970s, at which time concerns over the activities of multinational companies began to emerge.21
4. The 1968 US Regulations
2.21 In light of the influence they would have across the world, the release of the US 1968 regulations was, and remains, one of the most significant developments in transfer pricing since the inception of the ALP. As noted above, Congress had directed the preparation of these regulations as an alternative to adopting an approach that would have relied heavily on formulary allocations. Even though the regulations sought to apply a wide, substance-based approach,22 the primary effect of the detailed transfer pricing methodologies introduced (p. 58) by the regulations—at least in practice if not by design—was to shift the focus of the ALP from its broad-based and results-oriented principle to an approach that was rooted in a detailed transaction-by-transaction analysis. In the vernacular, the shift was from the wood to the trees.
2.22 The congressional concerns about transfer pricing abuse, primarily relating to the under-invoicing of outbound transfers, are evident throughout the regulations.23 Although there had been regulations on section 45 (renumbered as section 482 from 1954) since 1935,24 it had in practice been left to the courts to determine how to apply the provision in practice, leading to a wide variety of approaches. From 1968, it was the regulations that would provide the detailed starting point.25
2.23 The stated purpose of the 1968 regulations was to ensure taxpayers ‘clearly reflect’ income attributable to controlled transactions and to prevent the avoidance of taxes with respect to such transactions.26 Detailed methodologies were introduced for the first time by the regulations, with the focus being on related party transfers of tangible property. For tangible property, the regulations introduced three specific methods that were to be applied according to a stipulated hierarchy, being: ‘comparable uncontrolled price’; then ‘resale price’; and then ‘cost plus’. The comparable uncontrolled price (CUP) method requires a direct reference to prices in comparable transactions involving third parties (where there is such a CUP). The resale price method starts from the final selling price and then subtracts the cost to arrive at a profit mark-up equivalent to that earned in a comparable transaction. The cost-plus method starts from the cost of providing the goods or services and adds whatever uplift or profit is appropriate based on mark-ups in comparable transactions. If none of these three methods could be applied then other ‘fourth methods’ were possible under the regulations, and such methods could include a profit split method.27 In relation to transfers of intangible property, the regulations were less detailed but provided that the ‘comparable transaction’ method had to be used and if a comparable transaction was not found then twelve factors were applied to determine the arm’s-length price, starting with ‘prevailing rates in the industry’.
2.24 There were (and remain) two immediate difficulties with what was being proposed. These relate to unique products and the unique structures and arrangements that an MNE may (p. 59) have adopted. These difficulties arise directly from the heavy reliance on comparable transactions mandated by the regulations.
2.25 The unique products issue raises the question whether it will always prove possible to identify suitable third party comparables that can be used in the transfer pricing exercise. As noted above, the issue was not new, having been a central consideration in the cases establishing the primacy of the ALP. An early example of a situation lacking a reliable comparable would be the transfer of rights by IBM to the first programmable computer28 and a more contemporary example would involve the feasibility of finding reliable comparables for the transfer of the intellectual property rights in an Apple iPhone. The important point is that if the task of identifying suitable market comparables is problematic, then the arm’s length approach, based fundamentally on a comparison with the actions and arrangements of market, or unrelated parties, will in turn prove problematic. While the 1968 regulations recognized the problem, the partial solutions proposed—fourth methods and the twelve factors for determining an intangible price—would ultimately prove insufficiently robust to address adequately the variety and complexity of world-wide businesses operating in a rapidly globalizing economy.
2.26 The second issue stems from the very nature of MNEs. An MNE group, which typically comprises many different subsidiaries, may operate as a single economic unit. The risk and reward from the activity of that economic unit, including future profit or loss, remain within the economic unit (the group of companies) through the parent’s shareholding in the group subsidiary companies. Ultimately, all risks and rewards are concentrated in the parent company, the shares of which may be traded as a single asset. It has been argued by Ronald Coase that the MNE group exists because it internalizes and reduces transaction costs and so increases efficiency in dealing with overall costs: ‘The source of the gain from having a firm is that the operation of a market costs something and that, by forming an organization and allowing the allocation of resources administratively, these costs are saved.’29 Later thinking on the theory of the firm includes a more-recent focus on the value of information, leading to a knowledge-based rationale.30 These theories of the firm suggest that MNEs exist to deliver synergies that other market players cannot or do not deliver. These theories also suggest that much of the value of the MNE consists in the attributes of the group itself rather than being derived from the individual internal transactions that are entered into. That is, the MNE group of companies is not simply an amalgam of separate legal entities, each of which necessarily has a stand-alone and independent counterpart operating in the market, nor are all MNE transactions and arrangements necessarily mirrored in the market. What (p. 60) this means is that an economic slicing of MNE transactions into ‘market’ components will not necessarily allocate all of the group profits of an MNE to individual group members.31 The 1968 regulations, however, proceed on the basis that the value in the group can be quantified and captured, and divided between the individual legal entities that comprise the group, through a transactional pricing approach.
2.27 These issues were not tackled head on at the time of the 1968 regulations, though the issues were considered to some degree two decades later in the US ‘White Paper’, which is considered below. Nevertheless, the 1968 regulations proved to be the strongest formative influence on subsequent OECD thinking on transfer pricing.
2.28 Given the judicial developments that had by then taken place and the lack of concrete guidance or direction on the practical application of section 482, it seems wholly sensible for the transactional guidance comprised in the regulations to have been developed and released. The aim was clearly to provide practical methods that might be used by taxpayers on a day-to-day basis to deliver the required arm’s-length prices for related party transactions entered into. However, the main practical result of the methodologies developed by the regulations was to move the focus on the ALP and anchor it to the examination of actual transactions that had been undertaken by related parties using comparable pricing data as the benchmark. This meant that the focus inevitably shifted away from the wider question of what the parties would have done if they were not related at all and whether the allocation of income reflected clearly the contributions to profits of various members of the group. As will be seen, the effect of the regulations in this regard was all the more pervasive because of their influence on mainstream OECD thinking.32
2.29 After the publication of the definitive treaty language by the OECD and the detailed US 1968 regulations, there ensued a period of international elaboration of the ALP. The ideas of the 1968 regulations were exported to other countries and through the OECD the wealthiest countries began to standardize their approaches to applying the comparability-based rules of the ALP. As this occurred, countries gradually became aware of some of the problems of the ALP, notably problems caused by the lack of comparables, particularly in matters involving intangibles, and began early efforts deal with those problems.
1. OECD 1977 Model Treaty
2.30 Largely as a result of US pressure for a uniform global approach to the topic, the allocation of income was one of the most important topics discussed in the development of the 1977 Model.33 That Model made no changes to the 1963 expression of the ALP (which (p. 61) was retained in Article 9(1) of the 1977 Model) but added a new provision (which had also featured in earlier US regulations)34 dealing with ‘corresponding adjustments’ in a newly inserted Article 9(2), the text of which is as follows:
Article 9(2), Associated Enterprises
Where a Contracting State includes in the profits of an enterprise of that State—and taxes accordingly—profits on which an enterprise of the other Contracting State has been charged to tax in that other State and the profits so included are profits which would have accrued to the enterprise of the first-mentioned State if the conditions made between the two enterprises had been those which would have been made between independent enterprises, then that other State shall make an appropriate adjustment to the amount of the tax charged therein on those profits. In determining such adjustment, due regard shall be had to the other provisions of this Convention and the competent authorities of the Contracting States shall if necessary consult each other.
2.31 The 1977 Model retained without any alteration the one-paragraph Commentary from 1963 but added a number of additional paragraphs explaining the purpose of the new Article 9(2) provision, namely to avoid economic double taxation in the situation where an upward adjustment of profits is made in one state. It is explained that this purpose is achieved by the other contracting state making a corresponding downwards adjustment of the profits it taxes. However, reflecting concerns on the part of some states that they should not be at the mercy of heavy-handed transfer pricing adjustments imposed in other states, the Commentary emphasizes that any such adjustment is not automatic—rather, it is required only if the state concerned considers that the adjustment originally made in the first state is justified both in principle and as regards the amount.35 At a practical level, the terms of the new provision dealing with corresponding adjustments made the existence of a general consensus on the operation of the ALP (i.e. in relation to the circumstances in which it is appropriate to make transfer pricing adjustments) very important if the principle were to operate successfully to remove economic double taxation.
2.32 The Commentary also clarifies that the paragraph does not require the ‘secondary adjustments’ that would be needed to establish the situation exactly as it would have been if transactions which are adjusted on transfer pricing grounds had been originally entered into on arm’s-length terms (the nature of such secondary adjustments would depend on the specific facts of any case but typically might involve actual transfers of cash corresponding to any pricing adjustments made, the possible imposition of additional withholding tax depending upon the type of income adjustment concerned, etc.). However, the Commentary clarifies that nothing prevents such secondary adjustments from being made where they are permitted under the domestic laws of contracting states.
(p. 62) 2.33 With the addition of Article 9(2) in 1977, the OECD model treaty provisions were brought into the form in which they remain today, given that there have been no subsequent changes to the text of Article 9 in the OECD Model since 1977.
2.34 With regard to the rules on attributing profits to permanent establishments, the 1977 Model contained no material changes to the earlier 1963 version of Article 7 and only relatively minor changes to the accompanying Commentary.
2. The 1979 OECD Report on Transfer Pricing and Multinational Enterprises
2.35 The 1979 Report, Transfer Pricing and Multinational Enterprises,36 was the first major OECD contribution to address the detailed application of the ALP. The Report was intended to respond to the growing internationalization of economic activities and specifically the emergence of MNEs which it referred to as follows: ‘This increasingly common phenomenon of related companies operating in a group with some degree of centralised management, yet with the individual members of the group operating under different national laws, has given rise to important problems regarding the taxation of corporate profits.’ The primary concern expressed was that pricing used in relation to transactions within the group might ‘diverge considerably from the prices which would have been agreed upon between unrelated parties engaged in the same or similar transactions under the same or similar conditions in the open market’.37 Notwithstanding some of the high-profile concerns with the behaviour of MNEs that had by then already emerged, the overall tone of the Report is moderate, seeking to balance the interests of MNEs and tax authorities.38
2.36 The Report is largely based on the US 1968 regulations.39 The US was strongly in the forefront of campaigning for a single global standard to relieve the pressures of double taxation on US business. The discussion in the Report is overwhelmingly couched in transactional terms, referring to the pricing of transactions, in this respect mirroring the discussion in the earlier US regulations. The discussion concentrates on the pricing of sales of goods, transfers of technology and trademarks, the provision of services, and loans.
(p. 63) 2.37 The Report aims to set out the considerations to be taken into account, and the means available, for determining an arm’s-length price in widely varying circumstances involving related party transactions. As such, the Report aims to set out practical guidance to build on the ‘common concept and common language’ on transfer pricing provided by the 1963 Draft. The Report therefore discusses methods that might be used to determine arm’s-length prices and also deals with some of the more important transactions that take place between associated enterprises.
2.38 On transfer pricing methods, the Report largely follows the approach of the earlier US regulations (there was by this time widespread use by taxpayers of the transfer pricing methods first outlined in those US regulations). According to the OECD Report, the ideal test to assess conformity to arm’s-length pricing is direct reference to prices in comparable transactions involving third parties (where there is a CUP). However, it is recognized that such pricing information may not be available or be impractical to collect and so, following the US regulations, it also considers the cost-plus and resale price methods. The discussion of the methods is fairly detailed and includes comment on comparability of goods and markets, the impact of branding, sales volumes, and advertising costs, the impact of intangible property, the level of profit mark-ups, the allocation of overheads, and the allocation of fixed asset costs. As with the US regulations, the discussion is open to other methods being used but the discussion on such other possibilities is rather vague and negative.40 The Report also considers briefly profits-based methods and a return on capital approach, though the latter is noted as presenting difficulties and it is suggested that net yield expectations are too imprecise for routine use.41 It is also recognized that in many cases an arm’s-length price will not be precisely ascertainable and that in such circumstances it will be necessary to seek a reasonable approximation to it.42
2.39 The lengthiest section of the Report addresses intangible property, including patents and know-how, trademarks, research and development, and cost contribution arrangements involving intangible property. Intangible property transfers, particularly where they include cost-contribution arrangements,43 are already seen as presenting a number of particular difficulties for national tax authorities given the potential of such transactions to give rise to material shifting of profits within the MNE group.44 It is also noted that some arrangements adopted by groups (e.g. funding research and development activity through a cost contribution arrangement) would be ‘very unusual among unrelated enterprises’.45
2.40 The discussion notes that, whatever methods are used, problems of judgement and the evaluation of evidence will arise. However, it is confidently asserted that ‘[e]xperience shows that the difficulties can in general be satisfactorily dealt with and acceptable prices agreed’.46
(p. 64) 2.41 The Report also refers to the desirability of conducting a functional analysis to determine the functions that are carried out by the parties and in what capacity the functions are performed (e.g. whether as a principal with all the risk and reward or as an agent with limited risk). Various other issues, such as how to deal with set-offs, package deals, contracts, and documentation are all discussed.47 The Report also notes the desirability of contemporaneous documentation being prepared to evidence and support the transfer pricing adopted in any case.48 The discussion of service transactions introduces the notion of a ‘benefit’ test and distinguishes services that provide a benefit to the recipient from shareholder activities and costs.49
2.42 Throughout the Report, the bulk of the discussion concentrates on the task of ensuring related party pricing conforms to the ALP.50 This is reflected in the expression of the overall aim of the Report, namely to ‘set out the considerations to be taken into account, and the means available, for determining an arm’s length price in the widely varying circumstances which arise in practice in connection with transactions between associated enterprises’.51 This preponderant focus on transactional pricing reflects the influence of the US regulations. The approach leaves open the wider question of the degree to which the ALP has a role in recharacterizing a transaction entered into by related parties or determining whether it should be recognized at all. The Report comments on this point of uncertainty, though without providing a wholly clear position on the matter. Specifically, the Report acknowledges that the general approach it adopts is to ‘recognise the actual transactions as the starting point for the tax assessment and not, in other than exceptional cases, to disregard them or substitute other transactions for them’. However, it is then also noted: ‘The Report does, however, recognise that it may be important in considering, for example, what is ostensibly interest on a loan to decide whether it is an interest payment or, in reality, a dividend or other distribution of profit.’52 The Report is therefore indicating that ‘in exceptional cases’ (which are not clarified beyond the illustrative reference to ‘what is ostensibly interest on a loan’) a non-recognition or recharacterization of transactions may be justified under the ALP.53
2.43 The Report also touches on a related issue, namely the nature of some transactions entered into by MNEs. The Report refers to the ability of groups to enter into a greater variety of contracts and arrangements than are entered into by unrelated enterprises ‘because the normal conflict of interest which would exist between independent enterprises is often absent’. The discussion refers to cost contribution arrangements for research and development expenditure as an example of the type of arrangements that are very rarely encountered between unrelated parties. It also notes that contracts within an MNE ‘could be quite easily altered, suspended, extended or terminated according to the overall strategies of the (p. 65) MNE as a whole and such alterations may even be made retroactively’. It notes that there may be a difference between the legal character of a transaction and what it amounts to in economic terms.54 Though the Report identifies that transactions between members of an MNE group are not necessarily influenced by the market, these issues are not interpreted as raising profound questions about the suitability of the ALP nor as raising the fundamental concerns such issues would later be thought to raise. It concluded that tax authorities would have to determine ‘what is the underlying reality behind an arrangement in considering what the appropriate arm’s length price would be’.55 Nevertheless, the overall tone of the Report remains one of confidence in the ALP, ascribing to it an ability to accommodate any difficulties arising under the transactional approach.
2.44 The Report contains a relatively strong rebuttal of income allocation approaches not based on the ALP (largely directed at global formulary apportionment) on the basis that such approaches would be incompatible with Articles 7 and 9 of the OECD Model and also ‘would necessarily be arbitrary, tending to disregard market conditions as well as the particular circumstances of the individual enterprises and tend to ignore the management’s own allocation of resources’. The Report therefore suggests that formulary approaches could produce an allocation of profits which may bear no sound relationship to the economic facts and would inherently run the risk of allocating profits to entities which are in truth making losses.56 A key argument in the rejection of non-arm’s-length approaches is the impossibility of obtaining the full information necessary on the global activities of a MNE to enable the application of a global formulary methodology. In a comment that demonstrates how far thinking on the matter has now changed, the Report notes that the tax authorities in a country in which a subsidiary operates would not be able to obtain the relevant information ‘without imposing on the MNE itself a possibly intolerable administrative burden, or a similar burden on the tax authorities of the parent company’s country if they seek to get the information by way of exchange of information provisions under double taxation agreements’.57
2.45 The work of the United Nations on international tax lapsed in 1954 but later recommenced with the formation in 1968 of the Ad Hoc Group of Experts on Tax Treaties between Developed and Developing Countries.58 This revitalized United Nations involvement stemmed from the recognition that ‘the traditional tax conventions have not commended themselves to developing countries’, largely because it is the country of source that generally gives up revenue.59 Since developing countries tended not to be members of the OECD, their views were not given much consideration in OECD transfer pricing and tax treaty discussions.
2.46 The UN Group of Experts worked on preparing a UN Model Double Tax Convention, based on the OECD’s 1977 Model. This led to the adoption of a draft Model in late 1979. This was published in 1980 as the United Nations Model Double Taxation Convention between Developed and Developing Countries (‘UN Model’).60 Further editions of the UN Model have been worked on by the renamed (in 1980) Ad Hoc Group of Experts on International Cooperation in Tax Matters (and, from 2005, Committee of Experts on International Cooperation in Tax Matters) and published in 2001 and 2012. The UN Model is designed generally to expand source country taxing rights as compared with the approach in the OECD Model, the structure (and numbering) of which is largely retained. It is in some senses a successor to the League of Nations Mexico Model given the greater emphasis on source country taxing rights reflected in that version of the League of Nations model tax treaty.61
2.47 Though the position has become less certain from 2012 (see further below), the aim of the UN Model generally to expand source country taxing rights was initially of less relevance to the ALP given that the ALP is intended to apply symmetrically and uniformly to transactions between related parties, irrespective of the country in which they are located. Accordingly, the expression of the ALP in Article 9(1) of the UN Model (and of the ‘corresponding adjustments’ provision in Article 9(2)) is the same as in the OECD Model.62
2.48 The Commentary in the UN Model on Article 9 largely reproduces the OECD Model Commentary (including support for the application of Article 9 to deal with cases of thin capitalization).63 There are some more significant differences to the OECD Model (p. 67) approach in Article 7 of the UN Model, dealing with the attribution of profits to PEs. These include a limited force-of-attraction rule, which extends source country taxing rights to include sales and other business activities which, though not carried on in any PE, are similar to the activities of an existing PE of the company. There are also restrictions on deductions for certain payments to the relevant head office of the enterprise.64
2.49 The immediate relevance of the UN Model is twofold. First, the UN supported the ALP as the transfer pricing method of choice, even in developing countries. As will be seen below, the UN provided an outlet for what would ultimately grow to be at least a partially different perspective on the details of how the ALP should be applied. But its support of the ALP in general preserved the consensus forged through the League of Nations and OECD that the ALP was the transfer pricing approach of choice (in the authors’ view, the differences in perspective between ‘developed’ and ‘developing’ countries are limited to certain fairly circumscribed issues and are often exaggerated). Second, and more important in practice, the interest of developing countries in transfer pricing matters reflected in the work of the UN has underlined the significant challenges developing or emerging countries have in implementing and operating the ALP in practice. These challenges are discussed further in Chapter 4.
4. The 1984 OECD Report on Three Transfer Pricing Issues
2.50 The Three Taxation Issues report was produced as part of an effort by the OECD to continue the work on transfer pricing started with the 1979 Report.65 The Report addresses three specific issues, being (i) how to relieve the economic double taxation which would arise from one-sided transfer pricing adjustments, (ii) transfer pricing in the banking sector, and (iii) issues related to the allocation of management and service costs for tax purposes. As is discussed further in the next chapter, and though first addressed some thirty-five years ago, the first and third topics in practice remain highly problematic issues today.
2.51 The discussion on relieving double taxation is given over to the discussion of corresponding adjustments (following the introduction of the provisions in Article 9(2) of the 1977 Model) and reflects the importance of a consensus approach on the operation of the ALP. One of the more important issues addressed is the approach to be adopted if the tax authorities of different states have a different opinion on the need for a transfer pricing adjustment or its amount. This might arise where, for example, the amount of a cross-border royalty payment between affiliated companies is increased by a transfer pricing adjustment made by the tax authorities in the recipient state but the tax authorities in the state of the payer deny the need for any transfer pricing adjustment or dispute the amount of the adjustment made.66 It is noted that discussions between tax authorities (p. 68) would normally take place under the Model’s Article 25 ‘mutual agreement procedure’ (also often referred to as the ‘competent authority procedure’) as this is designed to facilitate the resolution by the relevant competent authorities of problems within the scope of bilateral tax treaties. However, as the discussion notes, an important limitation on the procedure is that competent authorities only have an obligation to negotiate; they are not required to reach an agreement. Material concerns on the part of taxpayers with the time-consuming and uncertain nature of the process are also noted.67 In a masterful understatement, it is noted that this makes the existing mechanism ‘a less than perfect instrument for resolving the problems that may arise in the implementation of double taxation agreements’.68 Nonetheless, the Report notes optimistically: ‘acceptable compromises have in practice nearly always been found’. Therefore, no major changes are suggested and in particular a process of mandatory mutual agreement (subject to arbitration) is not required, though it is noted the topic will be kept under review.69
2.52 The Report on banking, The Taxation of Multinational Banking Enterprises, is intended to supplement the 1979 Report, given that it did not address the position of banks and other financial traders. It is noted that banks conduct much of their international business through branches, and that while the transfer pricing guidance in the 1979 Report is relevant, special factors also apply to branch transactions.70 The Report includes a discussion of the general operation of Article 7 of the 1977 Model, the treatment of interest paid and received (which includes the recognition that intra-entity payments of interest for a bank should generally be recognized notwithstanding they are not normally taken account of in the case of other enterprises), the treatment of interest on bank capital, etc. As a practical matter, however, it is the (brief) discussion of the criteria by which loan assets are to be attributed to the different parts (i.e. branches) of the entity for the purposes of Article 7 of the Model that is arguably the most important feature of the Report. The discussion took place against the backdrop of concerns on the part of the tax authorities about ‘offshore booking’ of loans, i.e. the situation in which profitable loan assets were being originated in one jurisdiction but then booked in overseas tax haven locations. The Report proposes a functions-based approach to the allocation of such assets, namely identifying the relevant location where the loan assets are ‘substantially generated’ by the activities of the permanent establishment. This in turn is explained as depending on the extent to which the negotiation and conclusion of the transaction has been the work of the permanent establishment. The discussion sets out various activities that would normally be involved in the negotiation and conclusion of a transaction (e.g. obtaining the offer of new business, negotiating the terms of the loan with the borrower, deciding on the terms of the loan). The guidance is of note for two main reasons. First, it amounted to almost the only concrete guidance from the OECD on the topic of attribution until the OECD’s major project on the topic, which started in the late 1990s. In consequence, the 1984 guidance was in practice often applied by analogy to a broader range of situations than those for which it was intended. Second, it provided an early indication of the highly functions-driven approach that would later be taken up (p. 69) and developed in the lengthy OECD project on the attribution of profits to permanent establishments.
2.53 The final report, The Allocation of Central Management and Service Costs, is the shortest of the three. The Report identifies and discusses the wide range of costs and services that may need to be charged out within a group (e.g. for administrative and support services, marketing, and training), though noting that this will never include ‘shareholder costs’ of the parent. Cost allocation and apportionment methodologies are discussed and there is also a discussion of the risks of economic double taxation where the tax authorities of the parent (or service provider) require the allocation or charge-out of costs but these are not accepted (or only partially accepted) as deductible by the tax authorities in the recipient country. The Report warns that tax authorities should not substitute their own judgement for the commercial judgement of the entrepreneur but rather should simply apply the arm’s length principle.71 Further, as the Report notes, tax authorities have a highly variable approach to dealing with cost allocation issues and particularly in relation to what cost allocations are acceptable and whether a mark-up on costs should be included.72 Not surprisingly, given the issues identified, the discussion notes that to some extent problems relating to the symmetrical treatment of management expenses are always likely to arise, and this seems to be confirmed by the number of concerns expressed by MNEs that are referred to throughout the Report.73 The only practical remedy offered by the Report is use of the mutual agreement procedure, which has been discussed above.74
2.54 In 1987, the OECD released a report on the topic of thin capitalization.75 Whereas the two earlier OECD reports on transfer pricing had been concerned largely with the topic of price adjustments arising from the application of the ALP (including discussions on the pricing of interest on loans)76 this 1987 Report is chiefly concerned with the balance or ‘mix’ of debt and equity funding. As such, the focus of the discussion goes beyond the mere pricing of existing transactions and involves questions relating to the application of the ALP to recharacterize (or partially recharacterize) actual transactions entered into by the parties concerned.77
(p. 70) 2.55 The Report notes that non-tax factors will often determine an appropriate capital structure for a company and that in practice tax authorities have generally tended, in the absence of contrary indications, to regard the way a company is financed as primarily a matter for the judgement of the parties concerned. However, it is stated that in the case of a multinational group, the possibility of manipulation of its capital structure for tax purposes is a matter of concern. The various tax benefits from the use of debt finance are discussed and it is also noted that they can be combined with the use of tax-exempt dividend payments to give a high degree of tax planning flexibility to an MNC.78 It is acknowledged that it is ‘not at all clear what relationship between debt and equity should be taken as the norm in deciding in any particular instance whether a company’s debt is high in relation to its equity capital,’ though this does not deter the immediate observation that ‘a high debt/equity ratio may be an indication of an effort to achieve tax advantages by a disproportionate use of debt’.79 The Report contains a number of references to the absence of international norms or consensus on the issue of thin capitalization80 and this explains why the stated purpose of the Report is, broadly, simply to air the issues, though with particular emphasis on the impact of tax treaties and how to relieve unjustifiable juridical or economic double taxation.81 The Report thus represents the first OECD effort to grapple with some of the difficult transfer pricing issues that may be associated with the allocation of capital.
2.56 The Report notes the various types of country responses to dealing with problems from thin capitalization and these include what is described as the ‘arm’s length approach’ which is based on the size of the loan which would have been made in the arm’s length situation. However, the Report comments: ‘The main difficulty… is the absence of any clear guidelines as to what are the practices adopted by independent parties, and thus the difficulty of devising any consistent practice’.82 In something of an understatement, the Report concludes that there is no international consensus on the matter but in discussing the relevance of tax treaties, the Report identifies certain specific issues that concern Article 9.83 The most significant issue addressed is whether under Article 9 it is possible to deem the nature of a purported payment of interest to be something other than interest (i.e. in essence the question is whether Article 9 constitutes sufficient authority to permit the recharacterization and hence disallowance of a payment of interest on excessive debt on the basis that the interest payment would not have been made in a comparable situation involving independent parties dealing at arm’s length because, at arm’s length, there would not have been such excessive debt). The response given to this question is (p. 71) that Article 9 is generally relevant when countries are applying their domestic rules about thin capitalization and that its ambit goes beyond determining whether the rate of interest is at an arm’s length rate and includes determining whether a prima facie loan should be respected as such or regarded in whole or in part as some other kind of payment by reference to the arm’s-length test. The rationale for the response is that Article 9 allows adjustments to the accounts to include profits that would have accrued in an arm’s-length situation so that if profits have been reduced by interest payments which would not have been made in an arm’s-length situation, then a tax authority is acting in conformity with Article 9(1) in adjusting the profits to deal with cases of thin capitalization in order to deliver an arm’s-length result.84 The relationship between the provisions of Article 9 and domestic thin capitalization rules is not discussed in detail but the clear implication is that it will be necessary for a state to have in its domestic law transfer pricing provisions which are competent to adjust the effect of any abnormal loan terms by reference to the thin capitalization concept, leaving Article 9(1) to prevent the application of any such domestic rules beyond the arm’s-length standard, consistent with its purpose of preventing economic double taxation.85
2.57 The OECD view expressed modifies the hitherto highly transactional approach (as reflected in the two earlier reports) and is essentially a results-based approach to the application of the ALP: i.e. if the result of what is done leads to a non-arm’s-length result in terms of its impact on the respective profits of related parties, then the terms of the ALP justify the making of a transfer pricing adjustment, even if interest rates are comparable to those found in the marketplace. The Report also constitutes the high watermark for OECD consideration of arm’s-length levels of capital. While further internal discussions of the topic have been undertaken, there has not been sufficient agreement to provide further published guidance.
2.58 Given that in concept the OECD clearly viewed the ALP as relevant to cases of thin capitalization, the Report also addressed the chief practical barrier, namely by exploring what practical guidelines or standards were available to assist in the application of Article 9. The Report notes that it is difficult to provide precise guidelines but suggests the type of issues that would be relevant on a facts and circumstances approach. These include, among others, the uses to which the borrowed funds are put, the level of subordination of the debt concerned, the level of equity, and the perspective of an independent bank if asked to lend.86
2.59 The process of elaborating the ALP necessarily gave rise to consideration of the problems associated with its application. While some of those problems were apparent very early in the elaboration process, both judicial and administrative initiatives highlighted more difficult issues as time passed. Important developments occurred both in individual countries and at the OECD, with the interplay between US legal developments and the international deliberations in OECD publications being particularly illuminating. In large part, the issues being confronted involved two related topics: (i) the absence in many cases of useful comparables adequate to support application of the approved transactional methods, and (ii) the related problems surrounding the treatment of readily movable assets, especially intangibles. The interplay between transactional pricing and the treatment of capital movements became very important in the deliberation over how the ALP’s problems could be confronted. In at least one instance the challenges were deemed to require a legislative fix going beyond the ALP, rather than simply a more artful interpretation of the existing transfer pricing statute.
1. US Judicial and Legislative Developments Related to Economic Analysis and Intangibles
2.60 The courts in the US continued to struggle with how transfer pricing cases could be addressed when there were no good comparable transactions to guide pricing determinations. Litigants in these cases began turning to various forms of economic analysis to aid their cause, receiving a decidedly mixed judicial response. An early bellwether case involved the DuPont chemical company.87 In DuPont, the taxpayer had fairly plainly undercharged its Swiss marketing affiliate, DISA, for various products. Internal memoranda were identified that admitted as much. The company sought to justify its pricing by applying the resale price method from the 1968 regulations, but because its products were unique, comparables were difficult to identify and the court found that the data used in this effort did not satisfy the necessary standards of comparability.
2.61 In combating the taxpayer’s argument, the government presented testimony of two expert economists. Dr Charles Berry presented an analysis based on the ratio of gross income to total operating costs of arguably similar independent distribution companies (a measure ever-after referred to as the ‘Berry ratio’). Dr Irving Plotkin presented evidence of rates of return of a broad range of 1,100 companies. These expert presentations were intended primarily to demonstrate the unreasonableness of the profit levels achieved by DISA. The analyses presaged the much later development of the CPM and TNMM approaches (see below). The court noted these analyses but did not directly rely on them in reaching its decision. The approaches of Dr Berry and Dr Plotkin were very influential, however, in guiding the development of transfer pricing thinking in subsequent years. Both became regular expert witnesses in numerous cases over the ensuing years.
2.62 The other issue to confront the courts in this period related to the treatment of intangibles. Taxpayers adopted various strategies for transferring valuable intangibles to tax havens and (p. 73) then arguing, based on the intangible ownership rights of the low-tax transferee, that it should be entitled to earn substantial income. The most intense of these disputes involved so-called possessions corporations operating in Puerto Rico.
2.63 The US maintained a special tax incentive regime for companies that operated in US possessions, most notably Puerto Rico. US companies with most of their physical operations in Puerto Rico operated free of US tax in order to encourage employment in the possession, and could negotiate highly favourable tax incentive arrangements with Puerto Rico, as well. Pharmaceutical companies, and others with valuable patents, adopted structures under which valuable drug patents, trademarks or other intangibles were transferred to possessions companies qualifying for the tax incentives by means of capital contribution. The possessions subsidiaries then manufactured products in Puerto Rico and sold their products back to the US market at high margins that primarily reflected the value of the intangibles those subsidiaries had received by way of capital contribution.
2.64 The IRS challenged these arrangements in several cases involving pharmaceutical and other technology companies.88 The IRS argued that the possessions company affiliates were essentially contract manufacturers which should be compensated for their manufacturing services on a cost-plus basis, but that the benefit of the intangible assets they had received should be allocated back to the US parent companies. The argument was based on the contention that the parent companies had developed the intangibles and could not be expected to give those assets away in arm’s length dealings. The IRS had some initial success with these arguments in Lilly and Searle in the Tax Court, where the court rejected the contract manufacturer claim but essentially ruled that an independent enterprise would not have been willing to transfer these valuable intangibles, thereby depriving itself of cash flow needed to continue the transferor’s innovative research and development. The economic effects of the capital contributions of the intangibles were thus disregarded (at least in part) under transfer pricing principles.
2.65 On appeal in Lilly, however, the court ruled that the capital contribution of the intangibles (and the resulting intangible ownership of the possessions corporation subsidiary) had to be respected for transfer pricing purposes and could not be disregarded under the authority of the transfer pricing rules. This left the possessions subsidiary free to retain most of the financial benefit of the intangible ownership it had achieved by way of the contribution to its capital.89 Clearly, with this result, it became evident that the transfer of capital, a transfer that was tax-free under the relevant statutes, created an enormous hole in the ability of the transfer pricing statute to align income with income producing activity.
2.66 Sensing the possibility that the taxpayers in these cases could well prevail in their claim that the capital contributions could not be disregarded in applying the transfer pricing rules, Congress entered the fray even before the final court decisions were reached. In the early 1980s, with the court cases still pending, the adoption of special statutory provisions on possessions corporations limited the amount of income that could be attributed to such entities.90 The provisions effectively placed an arbitrary statutory limitation on the ability (p. 74) of taxpayers to claim a tax benefit for contributed intangibles. It refused to allow any tax benefit to a possessions corporation for trademarks and other ‘marketing intangibles’ (a term invented for the purpose which later took on a more generalized importance in transfer pricing discussion), and required the payment of a sizable cash royalty for patents and other trade intangibles transferred to a possessions company affiliate, even though the equity interest received for the transferred patents was by definition equivalent to the value of the patents transferred without any additional royalty being received.
2.67 It is noteworthy that, faced with a situation where the transfer pricing rules seem unlikely to be able to regulate the income shifting consequences of transfers of capital assets, Congress chose to respond with arbitrary limits on capital shifting. It did not respond by clarifying the ALP and how it might apply to regulate capital transactions, or by generally amending the US transfer pricing statute to make clear that it could address shifts of capital, or by giving the IRS the authority to disregard capital transfers that somehow led to a misalignment of economic activity and income. It simply found the transfer pricing statute to be inadequate to the task of regulating capital movements, and moved in a different direction to achieve its desired tax policy result.
2.68 The courts also struggled with shifts of intangibles in situations where direct capital contributions were not present. These cases essentially presented valuation problems, which also proved to be highly challenging and which for at least some observers confirmed that the transfer pricing rules might very well not be up to the task of regulating income shifting by means of transfers of intangibles.91
2. The Commensurate with Income Standard and Developments Leading to the 1994 US Regulations
2.69 While globally a broad consensus among states supporting the ALP had developed, the concerns in the US with the transactional approach reflected in the cases discussed above continued to grow. Ultimately, this would lead to a unilateral rewriting of the US transfer pricing regulations by the IRS. A Government Accountability Office (GAO) report in 1981 expressed concerns about the availability of meaningful comparables data (whether reliable data on comparable products for the CUP method or on gross profit margins for the cost-plus and resale price methods). This meant, the GAO said, that the ‘IRS can seldom find an arm’s length price on which to base adjustments but instead must construct a price’. The consequence noted was uncertainty for corporate taxpayers and a difficult and time-consuming enforcement process for both the IRS and taxpayers.92
2.70 At the same time, the US government harboured increased concerns about transfers of intangibles to tax havens, including the continuing Puerto Rico situation, but going far beyond that unique factual circumstance. These issues led to the beginning of a wider rethinking of the approach that had been adopted and this ultimately led to a much-expanded use (p. 75) of economic analysis and profits-based methods, as such methods avoided the problem of obtaining information on transactional pricing for comparables. Data on aggregate operating profits was much easier to obtain.
2.71 On the intangibles front, the Tax Reform Act of 1986 added a single sentence to the US transfer pricing statute designed to address the issue of intangible property valuation. Initially this provision was intended to further limit the benefits of shifting intangibles to Puerto Rico, but out of concern to maintain the argument that the provision would be consistent with the ALP, it was extended to cover all intangibles transfers. It stated: ‘In the case of any transfer (or license) of intangible property … the income with respect to such transfer or license shall be commensurate with the income attributable to the intangible’. The amendment, which came to be known as the ‘super royalty’ provision, was intended to look to future actual profit derived from the exploitation of a transferred intangible to determine the consideration at the date of the contract to transfer (or license) the intangible property. The intention of the measure was to allow upward adjustments to be made to the consideration received by the transferor of an intangible if the profits of the transferee turn out to be higher than those anticipated at the time of the transaction.
2.72 The 1986 Act also required the Treasury to conduct a broader study of the application of the ALP. This led in 1988 to the publication by the Treasury of A Study of Intercompany Pricing Under Section 482 of the Code (the ‘White Paper’).93 The White Paper explained that the general goal of the commensurate with income standard is ‘to ensure that each party earns the income or return from the intangible that an unrelated party would earn in an arm’s-length transfer of the intangible’.94 The White Paper considers the history and evolution of section 482, the regulations, and case law. It also outlines the problems experienced, including especially the difficulties of identifying and using relevant comparables, and proposes solutions to tackle the perceived weaknesses.
2.73 In broad terms, the White Paper aimed to restrict the use of transactional pricing methods to cases where an exact comparable could be found. Otherwise, it sought to use economic evaluation of industry-based profit measures to reward whatever functions were involved in any particular case. The paper discusses four possible methods for transfers of intangibles, two of which were based on a price approach (exact CUPs, which it acknowledged were rarely available, and inexact comparables requiring appropriate adjustments) and two profit-based methods, being the basic arm’s-length return method (BALRM), and a profit split method. The BALRM, which was central to the White Paper, was at the time a novel approach. It sought to identify the appropriate return for the transfer of an intangible based on applying industry-wide return rates to the relevant assets and functions of the parties concerned. Under the BALRM, first, the functions involved in the business affected by the intangible transfer would be identified; second, a value or return would be assigned to each function; third, the residual (which might be income solely due to the intangible) would be allocated to the parent (expected to be in the US), or the residual might be subject to (p. 76) a profit split if both parties concerned had developed intangibles. One by-product of the BALRM proposal (which continued as the proposal developed into the ‘CPM’ which is discussed below) was the impetus given to one-sided testing methods.95
2.74 The White Paper also discusses economic theories concerning section 482 and particularly whether the market-based arm’s-length approach is flawed as a matter of general principle given that it cannot be assumed that related parties conduct market-based transactions within the MNE group.96 The discussion concludes that the arm’s-length approach remains the best theoretical allocation method. This conclusion is reached on the basis that there are two types of arm’s-length transactions to consider, one in which the parties remain independent and another in which the two parties make an arm’s-length agreement to affiliate by merger, joint venture, etc. so that they are in a position similar to that of related companies. Put another way, the appropriate transfer pricing result is achieved if each related party is assigned the income that a corresponding unrelated party would earn if the latter were using the efficient cost structure of the MNE.97 However, there is no discussion of how the two different types of arm’s-length situations are to be identified nor by what means any resulting differences in profits are to be allocated to the parties.
2.75 The proposed BALRM method was heavily criticized on the basis it was: highly complex given the significant data required; too subjective in relation to the treatment of the residual; and inconsistent with the ALP given the required use of industry-wide data in setting a transfer price and the retrospective rewriting of agreements (even if they seemed arm’s length when entered into) due to the periodic reassessment of transfer prices, which reassessment could take account of new information arising in the interim. There were also concerns that it would be over-relied on by the IRS and have the effect of skewing profits to the US. The negative response led to the US Treasury extending the comment period on the paper indefinitely and to the subsequent protracted process for enacting regulations.
2.76 There then followed Proposed Regulations in 199298 and later the Temporary Regulations in 1993.99 The 1992 Proposed Regulations were concerned with intangibles only and were intended to implement the commensurate with income standard by replacing the BALRM with a comparable profit valuation method which would apply where there were no exact comparables. The 1992 Proposed Regulations contained two price-based methods as before, though with different names—the matching transaction method (MTM) corresponding to exact CUPs and the comparable adjustable transaction method (CATM) corresponding to inexact comparables. However, the application of CATM was required to fall within a ‘comparable profit interval’ (CPI) which was to be established by data (broadly, an index of acceptable profit ranges based on comparable functions) from similar but unrelated companies, or if not available, from the business sector concerned.100 The (p. 77) intention was that the CPI approach would replace the BALRM given the latter was by then accepted by the IRS to not be feasible.
2.77 The 1992 Proposed Regulations were perceived as being too complex and inflexible and the proposals gave way to further revisions in the 1993 Temporary Regulations. The 1993 Regulations were broader than the 1992 Regulations as they covered tangible as well as intangible property. The 1993 Regulations introduced the comparable profits method (CPM) as an alternative method for valuing transfers of tangible and intangible property.101 Like the CPI, the CPM is based on a comparison of the operating profit of the taxpayer with that of independent enterprises with similar types of transactions (or the sector as a whole) under comparable circumstances. The CPM starts by selecting the party to be tested, which is typically the party carrying out the simplest or most easily valued functions, and then selects comparable parties with similar ‘profit level indicators’ (financial ratios that measure profits, costs, resources, etc.) before determining the relevant arm’s-length range.
2.78 The 1993 Regulations are also noteworthy for introducing the ‘best method rule’ (which requires use of the method that gives the most accurate measure of an arm’s-length result based on the facts in any particular case) and for developing more detailed and flexible standards of comparability. Specifically, five comparability factors were introduced, being: functions, risks, contractual terms, economic conditions, and the nature of the property or services involved. The intent was to make it easier to compile a hypothetical comparable by adjustments to comparable components of third party transactions, easing the requirement for ‘exact’ comparables.
2.79 The 1994 Final Regulations clarify and refine the approach of the 1993 Temporary Regulations, but do not change the fundamentals. For tangible property, the regulations specify five possible methods—CUP, resale price method, cost-plus method, CPM, and profit split methods (either comparable profit split or residual profit split). For intangibles, the basic approach remained to first reward the functions of the parties on the basis of market comparables, with the residual profit then being allocated to the party who bore the costs of developing the intangibles, whether owned by that party or not.102 Three specific methods are available for intangibles—comparable uncontrolled transaction (CUT) (which is similar to the CUP method), CPM, and profit split. In the case of both tangible and intangible property, other methods may also be used if they can be justified in the circumstances. The result is that profit-based methods, including profit splits, were now accorded equal status with other (traditional) methods. The commensurate with income rule of section 482 is implemented by requiring periodic adjustments to the consideration for intangible property in relation to arrangements covering more than one year, though with certain exceptions.
(p. 78) 2.80 The overall result from this period of rethinking—and re-engineering—of the ALP is therefore something of a retreat by the US from the earlier comparables-based pricing check approach to the operation of section 482. By moving away from the strictly comparables-based approach and widening the methods that may be used by taxpayers, the intention of the Final Regulations was to avoid being hamstrung by the (as it had proved) more rigid approach of the 1968 Regulations and to be more aligned with what was seen as the proper purpose of section 482, namely ‘clearly to reflect the income’ of related party taxpayers.103 Economic analysis based on financial rates of return was assuming an increasingly important role in transfer pricing thinking while the practical role of comparable transactions declined.
2.81 The immediate issue raised by these unilateral US policy developments, particularly given the importance of a global consensus to protect against economic double taxation, was how they would be regarded by other states. Generally, the answer was not positively. Due to widespread concerns over the US developments on the part of a number of countries (some of which were in the process of moving to conformity with the then prevailing OECD approach), an OECD Task Force was established to provide to the US the collective view of the other OECD members. The Task Force released reports on both the 1992 and 1993 regulations.104 It had various concerns with what was being proposed, particularly the ‘ex post’ use of hindsight for purposes of the commensurate with income standard. This was considered contrary to the ALP because third parties would generally negotiate contracts on the basis of the situation prevailing at the time of the negotiations and such an ‘ex ante approach’ was already implicit throughout the existing OECD guidance.105 There were also general concerns on the proposed use of profit methods (particularly where these drew on sector profitability data) given that, for the OECD, profit methods were measures of last resort or to be used as backstop checks to other methods. Specific concerns were raised in connection with the CPM.106 The Task Force, which aimed to retain a uniformity of interpretation of the ALP in Article 9 of the OECD Model, emphasized that problems with the ALP cannot be resolved unilaterally and expressed concerns that the US approach would not be accepted by other countries. There were also concerns that the new US approach, together with the tougher stance on penalties and documentation that had been adopted,107 would unreasonably skew income to the US and therefore increase double taxation.108
2.82 The publication of the 1994 US Final Regulations created a material schism between the approach in the US and the prevailing OECD thinking on the ALP. One of the primary tasks of the 1995 Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations109 was to overcome those differences in order to ensure a uniform set of transfer pricing rules. As will be seen, some level of rapprochement was achieved by the OECD’s slightly softened stance on the use of profits-based methods, though this still left some significant differences between the US and OECD approaches and a number of OECD states remained strongly opposed to the use of profits-based methods.
2.83 The Guidelines are a revision and compilation of the three previous OECD reports that have been discussed above, namely Transfer Pricing and Multinational Enterprises (1979), Transfer Pricing and Multinational Enterprises, Three Taxation Issues (1984), and Thin Capitalisation (1987).110 However, the intention behind the Transfer Pricing Guidelines was not merely to revise and update the earlier documents but also to create guidance that could be periodically reviewed and revised on an ongoing basis. In the end, the Guidelines also borrowed heavily from the approaches contained in the new US regulations. It was also intended at the time of publication to supplement the guidance with additional chapters addressing other aspects of transfer pricing. This has subsequently happened, meaning that the Transfer Pricing Guidelines have been much expanded since their initial publication.111 Globally, the Guidelines now rank as the most influential guide to the operation of the ALP.112
2.84 The starting point for the 1995 Guidelines was a reference to the fact that the role of multinationals has increased dramatically over the twenty years leading to their publication. It is noted that the growth of MNEs presents increasingly complex issues, primarily the practical difficulty of determining the income and expenses that should be taken account of in each jurisdiction. The issues are said to be particularly difficult where the MNE group’s operations are highly integrated.113 The discussion refers to difficulties for the tax authorities of obtaining relevant data. There are also concerns that ‘the relationship among members of an MNE group may permit the group members to establish special (p. 80) conditions in their intra-group relations that differ from those that would have been established had the group members been acting as independent enterprises operating in open markets’.114 The text notes some clear concerns but conveys a confident message that the ALP will nonetheless eliminate the effect of such special conditions and secure the dual objective of determining the appropriate tax base in each jurisdiction and avoiding double taxation.115
2.85 A discussion of basic principles for applying the arm’s length principle comprises Chapter I of the Guidelines. This includes an explanation of why OECD Member countries have adopted the ALP and the discussion refers to the resulting ‘broad parity of tax treatment for MNEs and independent enterprises’ which puts MNEs and independent enterprises on a more equal footing for tax purposes.116 It is also stated that the ALP has been found to work in the vast majority of cases, though there is also recognition of the difficulties involved in cases where MNE groups deal in the integrated production of highly specialized goods, in unique intangibles and/or in the provision of specialized services. It is also noted that ‘transfer pricing is not an exact science but does require the exercise of judgement on the part of both the tax administration and the taxpayer’.117 This also leads to the recognition there is not always a single arm’s-length amount and a discussion of the ‘arm’s-length range’ (a range of possible pricing outcomes that would be acceptable from the application of the ALP, another concept borrowed from the US regulations).118
2.86 The discussion of transfer pricing methods is organized into two chapters—Chapter II deals with what are now referred to as ‘traditional’ methods and Chapter III deals with ‘other methods’. The traditional methods are identified as the CUP method, the resale price method and the cost-plus method. There is no formal order of ranking of these three methods. All three traditional methods were discussed in the earlier 1979 report (and all are included in the US regulations) and the discussion in Chapter II is therefore a revision of the earlier material.119 The preference of the Guidelines is where possible to apply the ALP on a transaction-by-transaction basis.120 To support use of these traditional methods, the discussion of comparability analysis (comparing the conditions in a controlled transaction with the conditions in transactions between independent enterprises) in the Guidelines is much expanded as compared with previous OECD guidance. The discussion refers to the need to take into account all options realistically available to the taxpayer concerned and also introduces and explains a five-factor approach for determining comparability, which is also derived from the US regulations.121
(p. 81) 2.87 In a major accommodation to the US approach, the 1995 Guidelines include two ‘transactional profit methods’. Use of these profits-based methods is discouraged, however, and it is stated that these should be considered only when traditional transactional methods cannot be reliably applied or applied at all, for example where transactions are very interrelated such that they cannot be evaluated on a separate basis.122 This contrasts with the position under the US regulations where there is no such relegation of profit-based methods and the choice of methods is governed by the best method rule alone. The discussion in the Guidelines confirms, however, that these profits-based methods are consistent with the ALP and presents a much-expanded consideration of profit-based methods, which were mentioned only very briefly in the earlier 1979 guidance.123
2.88 The two profits-based methods discussed are the profit split method and the transactional net margin method (TNMM). The profit split method first seeks to identify the relevant profit to be split and then splits that profit between the associated enterprises on an economically valid basis that approximates the division of profits that would have been anticipated and reflected in an agreement made at arm’s length. The profit split does not rely on the availability of closely comparable transactions, but is highly dependent on identifying the respective contributions of the parties, typically through a detailed functional analysis.124 A variant of the profit split approach is the residual profit split, which divides the profit in two stages—first, to give the parties a basic return appropriate for the type of transactions in which they are engaged and, second, an allocation of remaining or residual profits according to the contribution of the parties.
2.89 The alternative TNMM is intended to reflect the net profit margin of one party relative to an appropriate base (e.g. costs, sales, assets), which a taxpayer realizes from one or more controlled transactions. The TNMM is the OECD’s version of the US CPM, which has been discussed above, though it is not identical. In theory, the TNMM is concerned with the net profit margin from a transaction or aggregated transactions with a related party whereas the broader-based CPM may be applied on the basis of the profit results of the company as a whole or even industry sector returns. The TNMM therefore operates in a manner similar to the cost-plus and resale price methods, though the Guidelines express substantial concerns that the method may be (wrongly) applied without adequately taking into account the relevant differences between the associated enterprises and the independent enterprises being compared (e.g. due to the relevant net margin being skewed by high costs from inefficiencies or poor management).125 The discussion urges appreciable caution in using transactional profit methods and notes the limited experience of them in practice.126 At this stage, there was strong resistance by a number of (p. 82) OECD states to the use of such profit-based methods.127 Further, even though the 1995 Guidelines approve the limited use of profit-based methods as explained above, the US approach reflected in the ‘commensurate with income’ standard goes appreciably further by including future profits, whereas the OECD position remained strongly resistant to the use of hindsight.128 In the discussion of methods in the Guidelines, it is still left to the taxpayer to choose the most appropriate method (though traditional methods where applicable should prevail over transactional profit methods) and it is recognized that it is not possible to provide specific rules that will cover every case and that no one method is suitable in every possible situation.129 The OECD position was in contrast to the requirements under the US 1994 regulations where there was no such hierarchy between traditional and transactional profit methods—rather, the required method was the one that gave the most reliable result (the ‘best method’ rule). Also, for the OECD, the consideration of more than one method is not required, whereas under the US approach, at least at a practical level (to avoid penalties in the event of a disagreement with the IRS), it is.
2.90 It is notable that the Guidelines draw away from evidence of bargaining between related parties as a reliable factor in confirming the arm’s-length character of arrangements between related parties. A ‘negotiated price method’ had been used in practice by taxpayers but the Guidelines note that, while MNEs can act as autonomous entities in their relationship with each other, clear evidence of hard bargaining between related parties alone is not sufficient to establish that the dealings are at arm’s length.130
2.91 The Guidelines generally adopt a highly transaction-specific approach with a strong emphasis on identifying circumstances where price adjustments are required. However, one of the particularly important new sections of the guidance deals with the more fundamental ‘recognition’ of actual transactions undertaken. Echoing the position adopted in the earlier 1979 Report, the Guidelines clarify that the arrangements entered into by taxpayers should generally be accepted: ‘In other than exceptional cases, the tax administration should not disregard the actual transactions or substitute other transactions for them, since restructuring legitimate business transactions would be a wholly arbitrary exercise compounded by double taxation created where the other tax administration does not share the same views as to how the transaction should be structured.’131 However, the Guidelines go on to introduce two circumstances where it may be both ‘appropriate and legitimate’ for a tax administration to consider disregarding the structure adopted by a taxpayer in entering into a controlled transaction. First, where the economic substance of the transaction differs from its form, the tax administration may disregard the parties’ characterization of the transaction and recharacterize it in accordance with its substance. The second circumstance is where, while the form and substance of the transaction may be the same, the (p. 83) arrangements made, viewed in their totality, ‘differ from those which would have been adopted by independent enterprises behaving in a commercially rational manner and the actual structure practically impedes the tax administration from determining an appropriate transfer price’.132 The justification given for this approach to the recognition of transactions is that, in both sets of circumstances described above, the character of the transaction may derive from the relationship between the parties rather than being determined by normal commercial conditions and may also have been structured by the taxpayer to avoid or minimize tax. In such cases, the totality of the terms of the transaction would be the result of a condition that would not have been made if the parties had been engaged in arm’s-length dealings. Accordingly, Article 9 would allow an adjustment of conditions to reflect those which the parties would have attained had the transaction been structured in accordance with the economic and commercial reality of parties dealing at arm’s length. Despite the single justification given for the two circumstances, they are quite different in concept. The first example is essentially concerned with identifying the substance or reality of the transaction actually entered into (irrespective of the form of the transaction the parties may have agreed) and is therefore arguably closer to what has more recently been referred to as accurately ‘delineating’ the transaction.133 In the second case, however, there is no question of identifying the real transaction entered into. Rather, the intention is either to wholly disregard the transaction entered into or substitute for it a different form of transaction altogether.
2.92 The OECD’s limited approach to the non-recognition of transactions is presumably intended to minimize the discretion on the part of tax authorities, and thereby reduce material uncertainty on the tax treatment of transactions. The approach also avoids the inevitable practical difficulties that come with reconstituting actual transactions. As such, the approach is in line with the primary goal of avoiding or reducing economic double taxation through the development of uniform transfer pricing rules. However, a significant area of uncertainty has been how the OECD’s comments on the recognition of transactions can be reconciled with the normal treaty approach of leaving it to the domestic law of the parties to the treaty to deal with substance over form doctrines. The matter is discussed in Chapter 4.
2.93 Leaving that matter to one side, a perceived drawback of the OECD approach to recognition, especially in combination with the general transactional approach and the accommodation of MNE transactions that are not typically carried out by third parties,134(p. 84) has been the perception that the ALP has in consequence a limited ability to deal with tax-structured arrangements adopted by MNEs. This is a matter that is at the heart of the BEPS project and it will therefore be considered further in Chapter 6.
2.94 Given the concerns about the impact of the new US approach and the complexities of applying the ALP in practice, it is not surprising that the lengthiest chapter of the 1995 Guidelines is given over to the topic of administrative approaches to avoiding and resolving transfer pricing disputes, a topic which would grow in importance over the coming years.135 This material subsumes and expands the discussion in the 1984 report, Transfer Pricing, Corresponding Adjustments and the Mutual Agreement Procedure.136 The discussion also includes additional material on topics such as examination practices, the burden of proof in disputes, and penalties. There are also three entirely new and lengthy sections in the chapter dealing with simultaneous tax examinations, safe harbours, and advanced pricing arrangements. A brief discussion of arbitration is also included.137 Interest in the topic of simultaneous tax examinations reflected the increasing focus of the tax authorities on ensuring practical taxpayer compliance with transfer pricing rules.138 In a similar vein, the possibility of safe harbour regimes was raised as a possible means of mitigating the highly fact-intensive and judgement-based transfer pricing compliance process, though it would be some years before any attempt was made by the OECD to make full use of the idea.
2.95 The possibility of advanced pricing arrangements (APAs) is also mentioned. APAs were a relatively new concept at the time of the compilation of the 1995 Guidelines. An APA is an arrangement, initiated by the taxpayer, that determines an acceptable transfer pricing method in advance of any transactions being entered into and which applies for a fixed period of time by agreement with one or more tax authorities. The primary benefit of an APA is the certainty of transfer pricing method that is to be applied in any particular case. The first bilateral APAs were entered into by the US from 1991 and they are now common.
2.96 The Guidelines also include a chapter on documentation which includes a general listing of information that may be useful to support transfer pricing policies, including an outline of the business, the structure of the organization, financial information, and information on transactions. The OECD recommended approach is generally less stringent than under the US regulations.139 There are also chapters in the Guidelines dealing with intra-group services,140 and intangible property and cost contribution (p. 85) arrangements but these topics will be discussed further in the context of the BEPS changes in Chapters 6 and 7.141
2.97 The adoption of the US regulations in 1994 and the release of the 1995 Transfer Pricing Guidelines led to a flurry of actions by individual states in the latter half of the 1990s, with a large number of countries following suit with new legislation or new regulations strengthening existing measures or introducing new requirements. These included, for example, Latvia, South Africa, and Spain in 1995; Korea and Poland in 1996; Australia, Brazil, Canada, Chile, Czechoslovakia (now the Czech Republic), Mexico, New Zealand, the UK, and Vietnam in 1997; Argentina, China, and Denmark in 1998; and Russia and Zambia in 1999.142 These developments continued into the 2000s with similar legislative changes and/or the introduction or strengthening of supporting regulations on transfer pricing being made in a large number of countries, such as Azerbaijan, India, and Peru in 2001; Colombia and Israel in 2002; Tanzania and Belgium in 2004; Namibia, Croatia, and Egypt in 2005; Kenya, Bulgaria, Dominican Republic, and Estonia in 2006, and so on.143 In some of these cases (such as Azerbaijan, Colombia, Israel, and the Dominican Republic) transfer pricing requirements were introduced for the first time but for a number of countries the changes were intended to codify existing principles in the tax law or overhaul and expand existing statutory rules dealing with transfer pricing.144
2.98 It seems likely that many of these individual country actions were driven by concerns that, in the absence of having stricter rules, documentation requirements and penalty regimes, they would lose revenues to countries with much more onerous transfer pricing compliance obligations, such as the US. The result of these developments is that, within the space of approximately fifteen years, the number of countries with detailed regulations on transfer pricing had gone from six to at least thirty-seven by 2009 and in most cases detailed transfer pricing documentation requirements (some of which impose (p. 86) heavy reporting obligations and significant potential penalties on taxpayers) had been introduced.145
2.99 Two immediate consequences arose from the proliferation of this ‘arm’s race’ of new international transfer pricing obligations. First, the level of resources required to deal with transfer pricing issues and compliance, both within MNEs and their advisors as well as the tax authorities, shot up. Within MNEs, advisory firms, and tax authorities across the world, specialist transfer pricing teams, including economists and sector experts, were established. Second, the level of dispute (and the focus on efficient methods to resolve disputes), whether between taxpayers and tax authorities or between tax authorities and tax authorities, rose materially. Multilateral efforts to resolve disputes included the EU Arbitration Convention, which entered into force from 1995 and was intended to address the problems arising out of the non-binding nature of corresponding adjustments.146 In 2006, the OECD published a Manual on Effective Mutual Agreement Procedures (MEMAP) to give guidance on best practice in mutual agreement procedures and expedite unresolved disputes.147 At the same time, model treaty language was developed for implementing arbitration in MAP cases, including transfer pricing disputes, though effective arbitration arrangements remain more the exception than the rule.
2.100 OECD work on overhauling the technical approach to the attribution of profits to PEs started in the late 1990s.148 Differences in country practice had arisen in applying the much earlier work by the League of Nations on this topic. The overall objective of the new OECD project was to provide a clear and uniform approach for allocating business profits to PEs pursuant to Article 7 of the OECD Model.149 This was explained to be necessary because, while it was clear that the PE host country could, under Article 7, tax the ‘profits of an enterprise’ so far as attributable to the PE, the then existing rules and guidance did not adequately clarify how to arrive at the relevant measure of profits. The OECD noted that there were two main competing interpretations of the phrase ‘profits of an enterprise’ (p. 87) applied by states, and also further variations which meant there was no uniformity of approach,150 leading to the potential for double taxation and non-taxation.151
2.101 The OECD work on the topic carried on for more than a decade and involved a number of interim discussion drafts. This led to the lengthy 2008 Report on the Attribution of Profits to Permanent Establishments (the ‘2008 PE Report’), which was reissued in a slightly revised form in 2010 (the ‘2010 PE Report’), together with a completely new version of Article 7 (and accompanying Commentary) of the OECD Model.152 The work led to a single recommended ‘authorised OECD approach’ (AOA) to the attribution of profits to PEs, including a completely new approach to capital allocation. In broad terms, the project brought the taxation of PEs closer to that of subsidiaries, making the Transfer Pricing Guidelines much more relevant to the taxation of PEs. However, on key issues, especially the treatment of capital, important and consequential differences between the treatment of PEs and separate subsidiaries became very apparent. In the present context, the extensive OECD project is also notable for two other reasons: first, its impact on the relationship between the transfer pricing rules and the PE rules; and, second, its attempt to deal with the transfer pricing aspects (including in relation to capital) of global trading activities by investment banks.
2.102 The lengthy OECD 2008 and 2010 PE Reports each comprise four parts, dealing with general principles; banks; enterprises carrying on global trading of financial instruments; and insurance companies. The significant financial sector focus is due to the widespread use of branches in the financial sector.
2.103 The work on the topic was directed at identifying the preferred approach to attributing profits to a PE. This meant the work was not constrained by the original intent, the historical practice and interpretation, or even the language of Article 7 of the Model Treaty. The (p. 88) Reports adopt the ‘functionally separate enterprise’ approach and in the process explore how far the approach of treating a PE as a hypothetical and separate enterprise (to which the Transfer Pricing Guidelines could be applied) could be taken.153 The direction taken attempts to bring the approach to attributing profits to a PE closer to the position of a legally distinct and separate enterprise, though, as noted, there remain some material differences.
2.104 The position arrived at by the OECD is that, to apply the cardinal attribution principles of Article 7(2) of the OECD Model, a two-step approach is needed. Under the first step, the PE concerned is ‘hypothesized’ as a separate and independent enterprise. To do this, it is first necessary to carry out a full functional and factual analysis. The intention is that the output of that functional analysis will provide a mechanism for attributing risks, economic ownership of assets and capital to the hypothetically separate and independent PE and for recognizing and determining the nature of the ‘dealings’ (i.e. the intra-entity equivalents of separate enterprise transactions) between the hypothetically separate PE and other parts of the enterprise of which the PE is part. This is achieved by allocating the economic ownership of assets within the legal entity according to the location of the ‘significant people functions’ relating to the asset concerned. Risks are similarly allocated within the enterprise according to the location of the ‘significant people functions’ relevant to the assumption and management of risk. A different approach is taken to capital allocation. On the basis that a PE should have sufficient capital to support the functions it undertakes, the assets it economically owns, and the risks it assumes, capital is allocated as necessary to support the assets and risks of the PE. Thus, whereas assets and risks are allocated under the attribution test by reference to the geographical location of the relevant significant people functions, capital is allocated automatically to support those assets and risks.154 Lastly, under this first step to hypothesize the PE as a separate and independent enterprise, it is necessary to recognize (or not recognize) the ‘dealings’ between the PE and the rest of the enterprise of which it is part. The intention is that the functional analysis will identify whether a ‘real and identifiable event’ has occurred which needs to be recognized as a dealing of economic significance between the PE and another part of the enterprise. Generally, that threshold test will be met where there is real functional activity carried on in the PE that is consistent with the purported dealing. For example, if a PE claims to have entered into a dealing with its head office under which it has assumed the management of a risky asset, the relevant risk management people functions would need to be carried on by (or be actively supervised by) the PE.155
2.105 The second step is more familiar. Once, under step one, the assets, risk and capital have been allocated to the PE, and once the dealings have been identified, step two requires that the profits of the PE are determined based upon a comparability exercise, with arm’s-length pricing applied on the basis of the assets, risk and capital of the PE, taking into account the dealings to be recognized. This is achieved by applying the Transfer Pricing Guidelines by analogy, in other words comparing the dealings between the PE and the rest of the enterprise of which it is part with transactions between independent enterprises.156
(p. 89) 2.106 Notwithstanding the ‘functionally separate entity approach’ adopted for the attribution of profits to PEs, important differences remain in comparison with the strict separate entity approach of transfer pricing and Article 9 of the OECD Model. First, given a single legal entity is involved, it is obviously not possible to rely on intra-entity ‘contracts’ with legal consequences or validity. Instead, it is the location of the significant people functions or activity that is determinative as regards the location of assets and risks and which acts as the proxy for economic ownership of those assets and risks. The treatment of capital is another major area of difference. Under the AOA, with its ‘automatic’ attribution of capital to support the relevant assets and risks of the enterprise, it is not possible to separate capital from the business as a whole and assign it a reward (as it is for transfer pricing). There is also the point that the deemed separate enterprise approach has limitations—the credit rating of a PE, for example, is not derived from its stand-alone position but is based on the position of the enterprise as a whole.157
2.107 As noted, a new version of Article 7 of the OECD Model Treaty, together with a new Commentary, was released in 2010 to reflect the approach developed in the 2010 PE Report.158 The new Article is simplified by reducing it from the seven paragraphs of its previous version to only four paragraphs. However, the essentials of the Article remain. As before, Article 7(1) sets out the basic rule that profits of an enterprise in one state may only be taxed in the other if it carries on business through a PE and then only to the extent attributable to that PE. Article 7(2), the provision that provides the PE version of the ALP, is modified to take account of the concepts of the 2010 Report and defines the attributable profits of a PE as ‘the profits it might be expected to make, in particular in its dealings with other parts of the enterprise, if it were a separate and independent enterprise engaged in the same or similar activities under the same or similar conditions, taking into account the functions performed, assets used and risks assumed by the enterprise through the permanent establishment and through the other parts of the enterprise’. An entirely new provision is inserted as Article 7(3). This provides, broadly, a ‘corresponding adjustment’ mechanism for Article 7 similar to the one provided for transfer pricing purposes by Article 9(2) of the OECD Model. Specifically, the rule stipulates that if one treaty state adjusts the profits attributable to a PE by virtue of the rule in Article 7(2), the other state ‘shall, to the extent necessary to eliminate double taxation on these profits, make an appropriate adjustment to the amount of the tax charged on those profits’.159 Finally, Article 7(4) reproduces the wording of former Article 7(7) to the effect that where the profits of a PE include items of income dealt with separately in other articles of the treaty, the provisions of those other articles are not to be affected by the provisions of Article 7. The provisions of the earlier version of Article 7 dealing with the deduction of expenses incurred for the purposes of the PE, the possible use of a customary apportionment basis, the exemption for certain purchasing activity, and the requirement for consistency of the attribution method used (former Article 7(3), (4), (5), and (6)) have all been removed as unnecessary in the new version of the Article.
2.108 Progress since 2010 on adopting the new Article 7 and the AOA has not been rapid with the result that it has not been widely applied. One of the difficulties has been the need to amend (p. 90) existing tax treaties to give full effect to the new version of Article 7 and the AOA and this is inevitably a slow process. It has also been made more difficult by the decision of the UN not to adopt the AOA or the changes to Article 7 (see the further discussion in Chapter 3). The original intention of providing a clear and uniform approach to the interpretation of Article 7 has therefore not yet been fulfilled.
2.109 There are two respects in which the transfer pricing–PE relationship was materially changed by the AOA. The first relates to the technical shift to a transactional focus in the PE attribution rules brought about by the AOA. The second, relates to the increased use by tax administrators of the PE rules, in some cases as a preferred alternative to reliance on transfer pricing rules.
2.110 Dealing first with the shift to a transactional focus, this follows from the central role of ‘dealings’ introduced under the AOA. As discussed above, once the nature of the separate and independent PE has been established (‘hypothesized’) under step 1, it is then necessary to identify the relevant ‘dealings’ which are to be recognized as taking place between the PE and the rest of the enterprise of which it is part. This use of ‘dealings’ (intended as the Article 7 equivalent of ‘transactions’ for Article 9) was the mechanism by which the Transfer Pricing Guidelines could then be applied to the PE attribution exercise and it introduced a highly transactional approach for the purposes of the Article 7 attribution exercise. Formerly such a transactional approach would have been much less relevant to the analysis because of the greater latitude in applying the Article. This means that, under the AOA, it is now necessary to construct individual dealings (based on the relevant functional analysis and aiming to identify the relevant activities in the PE that are to be recognized), in arriving at the relevant intra-entity profit allocations. The AOA therefore brings a significant change to the approach to attributing profits to PEs. As a practical matter, though more disciplined, it is also a more onerous approach than previously applied. Further, this new focus on ‘dealings’ is arguably a step running counter to concerns created because of the highly transactional approach to transfer pricing under Article 9.160
2.111 The second aspect of the change in the transfer pricing–PE relationship concerns tax authority practice. Until relatively recently, and though there were some high-profile exceptions,161 disputes and problems relating to the PE rules have tended to be fairly infrequent, largely because the PE rules were typically regarded by the tax authorities as having a relatively restricted application in comparison with the transfer pricing rules. This was due to the fact that it was generally only financial sector entities (particularly banks) that operated commonly in branch form. As a practical matter, there was often also an assumption made that the PE attribution rules operated in broadly the same way as the transfer pricing rules, leading to a commonly held view that there would be no additional profits available to be taxed under the PE rules as compared with the result under the transfer pricing rules in the event that any PE had been created inadvertently by a taxpayer (as a result of fulfilling the threshold tests in Article 5 of the OECD Model). The consequences of this position in (p. 91) practice were that tax authorities would not generally seek to argue that PEs had been created where the transfer pricing policies adopted in any particular case seemed reasonable.
2.112 This situation began to change as the OECD work on the PE Report progressed, primarily because it became clearer that the revised two-step approach to computing the profits of a PE could give a materially different result to what may be termed the corresponding transfer pricing result.162 For example, if the operations of a local subsidiary company providing services in one jurisdiction caused the threshold PE test to be met (which might be the case where the dependent agent test of Article 5(5) of the OECD Model was fulfilled), then the profits attributable to the resultant PE created by that subsidiary could in some circumstances exceed the amount paid to that subsidiary for its services functions under the transfer pricing rules.163 Following the new AOA, it remained the case that the provision of ‘people functions’ (such as services) located in one territory would be rewarded in much the same way—and normally lead to the same level of reward—under the transfer pricing rules and the PE attribution rules. This meant there would generally be no difference to the return for the functions or activity according to whether that activity was carried on in a company or through a PE. However, the additional role in the PE attribution rules of the ‘significant people functions’ approach meant that in some cases there would be additional profits to be recognized in a PE that could never be recognized in a separate company under the transfer pricing rules. Those additional profits would follow from the use of the location of the ‘significant people functions’ as the attribution test for the location of assets (including intangibles), risks, and as the ultimate driver for the location of capital, as explained above.
2.113 Over time, tax authorities have begun to argue that in some cases additional profits over those required under the transfer pricing rules may be taxed in the event there is a PE. This has changed fundamentally the relationship between the transfer pricing and PE rules. Not surprisingly, it has become more common for tax authorities to challenge existing transfer pricing arrangements on the basis that the activities of the entity in the relevant jurisdiction cause a PE of a related party to be created where that related party is not otherwise present in that jurisdiction. There have also been concerns expressed that tax authorities are in some cases using the PE rules inappropriately to exert additional pressure on taxpayers with a view to achieving outcomes that might not otherwise be available under the transfer pricing rules.164 As will be seen in Chapter 6, the BEPS process has now materially increased the already existing concerns about the relationship between the transfer pricing rules and the PE rules.
2.114 Whereas the other three parts of the PE Report deal with attributing income to PEs, Part III of the PE Report considers first the transfer pricing issues arising where associated enterprises conduct global trading business, before then proceeding to consider such business in a PE context. The transfer pricing discussion in Part III is largely a continuation of (p. 92) the OECD’s work on global trading which was carried out in the period 1996–8.165 That early work sought to deal with a number of financial sector transfer pricing problems and concerns166 but is significant here for two specific reasons. It had by the time of that work become obvious that the traditional transfer pricing methods would struggle to accommodate global trading business carried on by investment banking groups due to the very high degree of integration of roles of the group entities involved and the very high number of individual related party transactions requiring analysis. It was also relevant that the structures used to conduct the business were often different to those of independent parties.167 These factors explain why the transfer pricing issues had by then already been addressed in a number of APAs involving financial sector groups and why the US IRS had developed guidance on its approach to the use of profit split methods and factor formulas to deal with difficult global trading cases in its Notice 94-40.168 The 1998 OECD paper initially refers to profit splits being a measure of last resort but the analysis soon proceeds more realistically to consider the factors that could be used when performing a contribution or residual profit analysis. Ultimately, the discussion recognizes that profit splits would more likely be justified across the industry sector as a whole, representing a significant acknowledgement of the scale of application of such profit-based methods.169 Such methods, particularly the residual profit split method, are now applied routinely in the sector in cases of global trading.
2.115 The work also represented an early effort to get to grips with some aspects of capital and risk. The importance of capital to the business was recognized170 and there was an early attempt to address the appropriate reward for a capital provider where capital was provided separately from the trading function.171 On risk, it was acknowledged that the Transfer Pricing Guidelines required account to be taken of risk, but that there was an open question as to exactly what that required. The paper considered whether it should be interpreted as risk exposure (i.e. the taking on of risk, being a function of available capital and the degree of risk being a measure of the contribution of that capital) or risk management. The issue was regarded as the most difficult of the questions arising in the project on global trading and was not satisfactorily resolved.172
2.116 By the time of the 2008/2010 PE Report and its comments on global trading carried on by associated enterprises, risk assumption and risk management were both regarded as highly important in the functional analysis.173 The paper also included a section on the role of capital and the required return to the capital provider, though this contemplated a broad range of possibilities, meaning that the guidance was set at a fairly generic level. (p. 93) The discussion seems to harbour some unarticulated concerns on the topic of capital, given that in passing it raised the question whether the arrangements involving capital should be disregarded or whether they might be construed as involving a dependent agent issue. However, its general conclusion was that there are numerous cases where capital will need rewarding and, in sharp contrast to current thinking post BEPS, it was also open to methods where the bulk of the return from a trading book is given over to a reward for capital, rather than as a reward to the trading function.174
2.117 Since 2009, UN work on taxation has been reinvigorated and has been focused especially on transfer pricing. This led to the publication in 2013 of the 495-page UN Practical Manual on Transfer Pricing for Developing Countries, which was produced by its Subcommittee on Transfer Pricing.175 Substantial additions and modifications to the Manual were approved in April 2017.
2.118 The Manual is largely consistent with the OECD approach as reflected in the OECD Transfer Pricing Guidelines, though it has led to some controversy as a result of the concern that it partially undermines the status of the OECD Guidelines as the global standard in the area of transfer pricing. For example, there have been concerns that observations in the UN Manual such as ‘the interpretation provided by the OECD Transfer Pricing Guidelines may not be fully consistent with the policy position of all developing countries’ actively encourages a departure from the prevailing OECD approach rather than simply describing the current state of affairs.176 There is also a chapter in the UN Manual that sets out a commentary on country practices in Brazil, China, India, and South Africa. Unlike the other chapters of the Manual, the chapter containing these individual country views (which were written by tax officials from the countries concerned) ‘does not reflect a consistent or consensus view of the Subcommittee’ and the individual country descriptions are in parts not necessarily consistent with OECD thinking as expressed in the Transfer Pricing Guidelines.177 The number of separate country chapters has increased in the revised 2017 version of the Manual.
2.119 Concern that the UN has become more circumspect about its previous support for the OECD Guidelines has also been fuelled by new commentary in the 2012 update to the UN Model. In the course of work on that 2012 Model update, Brazil, India, and China expressed reservations with the previous (2001) UN Model’s commitment to following the OECD Transfer Pricing Guidelines as internationally agreed principles. The 2012 UN (p. 94) Model now records that this commitment to the OECD standards stems from the views of the former Group of Experts and that these views ‘have not yet been considered fully’ by the (current) Committee of Experts and the language will be updated in the next version of the commentary.178
2.120 It is certainly wide of the mark to view countries as neatly segregated into two polarized camps of ‘developed country’ OECD Members and ‘developing country’ non-OECD Members. However, there clearly has been a growing level of unease or distrust on the part of a number of non-OECD Member states with the international approach to the allocation of taxing rights more generally, and this has also spilled over into the transfer pricing domain. The position has been compounded by the practical ‘capacity’ or resource challenges that a number of developing states have encountered in implementing the increasingly complex transfer pricing regime.179 The OECD has been seeking to counter this trend for some years and the BEPS project presented a major opportunity to extend the engagement of the OECD on tax matters with a very large number of non-OECD Members.
2.121 In 2010, the OECD updated the Transfer Pricing Guidelines and finalized a detailed report on business restructuring following a period of extensive consultation on the topic with business from 2005. The new business restructuring chapter of the Guidelines was divided into four parts dealing with: risk; the compensation for the restructuring; post-restructuring pricing; and the ‘recognition’ of the transaction.
2.122 The focus of the business restructuring project was on the internal restructuring that had by then become very common, i.e. the adoption by MNEs of structures involving a centralized principal in a low-tax state which would act as the primary risk taker for the group. The introduction of such structures would often involve the conversion of fully-fledged distributors in high-tax states into limited risk distributors or commissionaires for the foreign group risk taker (and/or the conversion of fully-fledged manufacturers into contract or toll manufacturers acting for the foreign group risk taker). Transfers of intangible property to a related party principal (an ‘IP company’) were also a common feature of these arrangements.180 These developments raise fundamental issues relating to the effectiveness of the ALP because risks and assets would typically be relocated from one company or companies to another company or companies, with significant international tax impacts, yet with the risks and assets remaining within the MNE group concerned.
2.123 Though appearing rather late in the day, given the scale of business restructuring that had by then been effected, the OECD work includes a careful rehearsing of all the relevant issues and arguments that might be explored by tax authorities in the context of a business restructuring.181 The discussion on risk expands significantly on the earlier explanation of (p. 95) ‘risks assumed’ in the Transfer Pricing Guidelines, with particular (and new) emphasis on the importance of identifying the party that controls the risk and the party that has the financial capacity to absorb the consequences of bearing risk, though at this stage the control of risk was a relevant, but not determinative, factor.182
2.124 There is much emphasis (especially in the discussion of the compensation for the restructuring itself) on the need to assess the ‘options realistically available’ to the taxpayer in any particular case. The guidance explains that independent enterprises would only enter into a transaction if it does not make them worse off than their next best option and therefore consideration of the other options realistically available is relevant to comparability analysis, to understand the respective positions of—and outcomes for—the parties.183 There is also increased emphasis that the relevant contractual terms should not be taken as definitive and that regard should be had to whether they are consistent with what independent parties would agree to, as well as taking account of the totality of the arrangements.184 These points are not new in comparison with the earlier versions of the Transfer Pricing Guidelines but the emphasis put on them is much increased. The approach suggested adds to the complexity—and subjectivity—of the required analysis, and reflects some attempt to move the ALP test to a broader assessment of whether the outcome of the arrangements adopted as a whole accords with what would result from normal commercial behaviour of independent enterprises.185
2.125 Not surprisingly, the discussion of transfer pricing methods reflects a preference for transaction-based methods, especially the CUP, but there is a perhaps more realistic acknowledgement that comparables will usually be imperfect or not available at all. It is also recognized that there may be situations where transfers of interrelated assets, risks, and functions may need to be valued on an aggregated basis and using valuation methods that are seen in acquisition deals.186
2.126 There is an extensive discussion of the ‘recognition’ of the actual transaction undertaken (i.e. the degree to which it should be respected for transfer pricing purposes) following the introduction in 1995 in the Transfer Pricing Guidelines of the two circumstances in which non-recognition is permitted. In a discussion that anticipates the emphasis in BEPS on the ‘delineation’ of the transaction, it is clarified that the starting point for the analysis is to properly identify and characterize the transaction under consideration.187 It is emphasized that tax authorities ‘should not normally interfere with business decisions (p. 96) of a taxpayer as to how to structure its business arrangements’. Therefore, transactions entered into by taxpayers should only exceptionally not be recognized with the result that transfer pricing adjustments will more usually arise not on the basis of the recognition doctrine but as a result of the appropriate comparability analysis in relation to the actual transaction undertaken.188 One practical reason for this caution is evident from the rather vague discussion of how to go about reconstructing a non-recognized transaction, namely the obvious difficulties that are involved in hypothesizing a ‘substitute’ transaction.189 Thus, while much of the discussion puts an emphasis on a more general ‘outcome-based’ approach, this is in the context of generally respecting the transaction actually entered into by the taxpayer.
2.127 The overall discussion of the topic of business restructurings brought out two major difficulties, one practical and one conceptual. The practical difficulty lay in the differences in views (between tax authorities and taxpayers) of the importance of the relevant commercial drivers within an MNE (e.g. to maximize synergies and economies of scale, to move to centralized global contracting and use of global IP companies, to improve the efficiency of the supply chain and to take advantage of internet-based technologies) as against the purely tax-driven goals of MNEs. While in principle, a tax-driven motive for a restructuring is of little relevance to whether it accords with the arm’s-length standard, in practice the matter is clearly relevant to the position taken by tax authorities.190
2.128 The (by now, familiar) conceptual difficulty lay in the tension between the fundamental approach of the ALP in treating group companies as separate and independent entities in circumstances where integrated business models and a single global organization and operating process were being implemented across the group. The discussion seeks broadly to accommodate, rather than target, this difficulty, responding to it in various ways. The assumption made is that if something is commercially rational from a group perspective, it can be ‘sliced’ or segmented by reference to the ALP to give an appropriate specific entity result.191 The discussion also draws a distinction between requiring related parties to actually act in the same way as unrelated parties (which is not a requirement of the Transfer Pricing Guidelines) as against requiring them to act in a ‘commercially rational’ manner, which includes considering alternative options.192 This distinction is behind the constant affirmations in the OECD discussion that the fact a related party transaction is not seen between independent parties does not necessarily make it a non-arm’s-length transaction.193 Where the transaction is not one seen between independent parties, the transfer pricing task is to ascertain what independent parties would have done under similar circumstances. This overall approach seems to stand the role of comparables on its head. Now, even though ‘group’ transactions may be different from those entered into by third parties, they are to be recognized in all but exceptional cases and appropriate adjustments or hypotheses can be made in the application of the ALP to find a price an independent (p. 97) party would have paid for something it would not have done in the first place.194 The OECD discussion of this difficulty itself seems to acknowledge dissatisfaction with the conclusion reached and specifically that the tensions are not fully answered by the discussion, meaning that it would need to be kept ‘in mind in an attempt to develop approaches that are realistic and reasonably pragmatic’.195
2.129 The OECD work on business restructuring is a precursor to BEPS in a number of ways. First, there were specific aspects of the topic, most notably the treatment of intangible property, that were carried over for further separate work which was ultimately addressed as part of the BEPS output. Second, a number of the themes raised in the business restructuring project—such as the treatment of risk, the proper identification and characterization of transactions, and the approach to the ‘recognition’ of transactions—would be revisited and reworked in the BEPS output. More fundamentally, the business restructuring project can be seen as the first major attempt by the OECD to consider the large-scale and materially tax-driven internal MNE restructuring which by then had become commonplace. The BEPS project would in due course seek to mount a much more fundamental set of challenges, both in relation to transfer pricing issues and on a much broader footing.
2.130 In addition to the new work on business restructurings, the 2010 update of the Guidelines included other notable changes. In particular, the increasing comfort with TNMM analysis governments had gained over the previous fifteen years led to a determination that the relative status given to profits methods could be increased. While TNMM is not quite made equivalent to the CUP and other traditional methods, it is made clear that profits methods are not simply a last resort. A most appropriate method analysis similar to that in the 1994 US regulations was incorporated in the Guidelines, and it was made clear that one-sided, profits-based approaches could well be the most appropriate approach in any given case.196
2.131 For at least the three decades following the end of the involvement of the League of Nations, there were few indications of any material change in global perspectives and thinking on the ALP. The general attitude to, and confidence in, the ALP in this period seems to be captured aptly by the comments in the 1963 OECD Commentary on the ALP, namely that the rule was ‘evidently appropriate’ and ‘seems to call for very little comment’.197
2.132 The earliest concerns with the ALP emerged from the US and from that country also came the early emphasis on a transactional-based approach to applying the ALP, which was subsequently taken up globally through the influence of the OECD. Subsequent US efforts to correct what were seen as weaknesses of this approach were not readily accepted globally. However, over time, the US-led use of profits-based methods has become increasingly accepted internationally and has led to a more flexible approach in applying the ALP, (p. 98) though this has not changed the general default to a transactional approach which takes as its starting point the actual transaction entered into by related parties.
2.133 By the time of the OECD’s BEPS project, the ALP had evolved in some very material ways from the concept that had been understood by the League of Nations in 1946. Most obviously, the earlier ready confidence in the ALP had subsided and given way to a number of significant concerns relating to its operation in practice (e.g. in relation to the complexity of its operation, the availability of suitable comparables, the different interpretations of what it requires) and there were also concerns with its application from a technical perspective (e.g. in relation to the technical purpose and scope of the provision, how it deals with group synergies, risk and capital, the importance and enhanced role of economic analysis). There remained a broad level of support from states for the ALP in principle, yet less consensus on its application and meaning. These concerns with aspects of transfer pricing had built from the mid 1970s through three ensuing decades and stemmed from the rise in importance of multinational companies and from the ease of conducting business across a group of companies on a unified global basis. These issues led, from the 1990s, to a material expansion in work on transfer pricing and measures directed against transfer pricing abuse—by both the supranationals, such as the OECD and United Nations, and domestic tax authorities.
2.134 The most acute practical and technical problems of the ALP that had emerged by the time of the BEPS project are described in Chapters 4 and 5. Before that, the next chapter will survey the operation in practice of the ALP on the eve of BEPS. What is very clear at this point, however, is that the ALP of 2010 was no longer the simple creation Mitchell B. Carroll had in mind.
1 The outgoing Fiscal Committee of the League of Nations had already suggested a programme of work for this new body which included revisiting the rules for the determination and allocation of taxable income in the case of business enterprises carrying on their activities in more than one country. See League of Nations Fiscal Committee, Report on the Work of the Tenth Session of the Committee, Geneva: League of Nations, 1946, C.37.M.37.1946.II.A, pp. 8–11.
2 The Fiscal Committee was created by resolution 2 (III) of 1 Oct. 1946 of the Economic and Social Council of the United Nations. See Department of Economic and Social Affairs, United Nations Model Double Taxation Convention between Developed and Developing Countries, New York, 2001, p. xvii.
4 The primary body of the OECD is its Council and its decisions are normally taken on a unanimous basis, according to Article 6(1) of the OECD Convention. In 1971, the OECD’s Fiscal Committee was renamed the ‘Committee on Fiscal Affairs’ (CFA). The CFA carries on much of its work through standing working parties, including Working Party 1 on Tax Conventions and Related Matters, Working Party 6 on The Taxation of Multinational Enterprises (addressing transfer pricing and profit attribution matters), and Working Party 8 on Tax Avoidance and Tax Evasion.
5 See Philip Baker QC, Double Taxation Conventions, London: Sweet & Maxwell, 2004, at A.02–A.06. The current version of the OECD Model is dated July 2014 and since 1992 there have been updates in 1995, 1997, 2000, 2003, 2005, 2008, and 2010.
8 See e.g. Frank v. International Canadian Corporation, 308 F.2d 520 (9th Cir. 1962); Seminole Flavor Co. v. Comm’r, 4 TC 1215 (1945); Ballantine Motor Co., Inc. v. Comm’r, 39 TC 348 (1962) aff’d, 321 F.2d 796 (4th Cir. 1963).
9 Oil Base, Inc. v. Comm’r, 362 F.2d 212 (9th Cir. 1966), aff’g 23 TCM 1838 (1964). For a more detailed consideration of these developments, see Reuven Avi Yonah, The Rise and Fall of Arm’s Length: A Study in the Evolution of U.S. International Taxation, University of Michigan Law School Scholarship Repository. http://repository.law.umich.edu/law.econ.archive/art73 [accessed 12 Apr. 2017].
13 The provision was originally contained in Article VII of the London and Mexico Drafts. In the period 1958–61, when the OEEC published a number of reports containing proposals for a new model, it was designated as Article XVI on associated enterprises, while Article XV was concerned with business profits. By the time of the 1963 Draft, the articles had been redesignated as Articles 9 and 7 respectively.
14 The term ‘associated enterprise’ is used only in the heading of the provision and the explanation of what this means gives only basic conditions for the application of the article. The detailed formulation of the rule (including the meaning of ‘management’, ‘control’, and ‘capital’) is left to countries to deal with in their domestic law. For an examination of the consequences of this approach as regards the term ‘associated enterprises’ (including the different approaches adopted by states), see Carmine Rotondaro, ‘The Notion of “Associated Enterprises”: Treaty Issues and Domestic Interpretations—An Overview’, ITPJ (Jan./Feb. 2000), 2.
16 For a discussion of the objective scope of Article 9(1) by reference to the words ‘commercial or financial relations’, see Jens Wittendorff, ‘The Object of Art. 9(1) of the OECD Model Convention: Commercial or Financial Relations’, International Transfer Pricing Journal (May/Jun. 2010), 200.
17 The notion of economic double taxation and juridical double taxation is explained further in Chapter 1, n. 9.
18 See further E. Reimer and A. Rust (eds.), Klaus Vogel on Double Tax Conventions, 4th edn, 2015, Wolters Kluwer, Article 9 at 11–15 (pp. 599–604). Unfortunately, and notwithstanding these summary comments, the ALP rule has led to considerable uncertainty and disagreement as to its scope and purpose. The matter is discussed further in Chapter 4.
19 The 1963 Commentary on Article 7 starts from the position that a complete solution to the problem of double taxation requires an agreed set of rules by reference to which the profits of a permanent establishment are to be calculated. The Commentary does not set out detailed criteria for attributing profits to a permanent establishment but notes that standard approaches have developed in a large number of European double tax treaties concluded since the Second Word War. Included in the Commentary is a discussion of some classes of intra-entity payments that cause special difficulties, including interest, royalties, and payments for ancillary services. See 1963 Draft, Commentary on Article 7, paras. 15–17.
21 It is suggested however that tax planning based on the use of intermediate group companies had become well advanced by the early 1950s—see Bret Wells and Cym H. Lowell, ‘Income Tax Treaty Policy in the 21st Century: Residence vs. Source’, Columbia Journal of Taxation 5 (2013), 32.
23 See Robert G. Clark, ‘Transfer Pricing, Section 482, and International Tax Conflict: Getting Harmonised Income Allocation Measures from Multinational Cacophony’, American University Law Review 42 (1993), 1167.
24 The regulations explained the purpose of section 45 (the predecessor provision to section 482 of the Code) as being ‘to place a controlled taxpayer on a tax parity with an uncontrolled taxpayer by determining, according to the standard of an uncontrolled taxpayer, the true net income from the property and business of a controlled taxpayer’. It was also explained that ‘the standard to be applied in every case is that of an uncontrolled taxpayer dealing at arm’s length with another uncontrolled taxpayer’ (Article 45–1(c) of Reg. 86 (1935), Revenue Act of 1934).
25 The US regulations have almost the same effect as statutory provisions due to the authority of the IRS to make adjustments based on section 482, coupled with the burden of proof resting on the taxpayer to demonstrate that any proposed IRS adjustments are ‘arbitrary and capricious’ in order to defeat the proposed adjustment.
27 The profit split method, which operates in a manner similar to a formulary approach, though with no preset factors, was not a standard recognized in the 1968 regulations but an approach that had been developed by the courts. See Treas. Reg. 1.482–2(d)(2)(iii)(a)–(m) (1968).
28 See Clark (n. 23), p. 1155, n. 35.
29 R. H. Coase, Industrial Organization: A Proposal for Research, reproduced in the Firm, the Market and the Law, Chicago: University of Chicago Press, 1988, at p. 63. See also ‘The Nature of the Firm’, reproduced in the same volume, at p. 33. Transaction costs are to be understood widely as including the costs of searching out people to contract with, negotiating with them and reaching a decision, drawing up the contract, and policing and enforcing its terms. These matters represent a ‘cost’ of using the price mechanism and firms will emerge to organize what would otherwise be market transactions whenever their costs are less than carrying out transactions through the market.
30 See generally, Oliver Hart, ‘An Economist’s Perspective on the Theory of the Firm’, Columbia Law Review 89(7) (Nov. 1989), 1757–74. It has also been suggested that tax avoidance, including the manipulation of transfer prices, is a principal reason for the existence of multinationals—A Small Open Economy? The Rule of Transfer Pricing and Income Shifting, US National Bureau of Economic Research Working Paper No. 4690 (1994).
32 As the discussion implies, the 1968 regulations had the effect of strengthening the separate enterprise approach but it has become increasingly questionable whether this is synonymous with delivering an arm’s-length result.
33 See further J. Wittendorf, Transfer Pricing and the Arm’s Length Principle in International Tax Law, the Netherlands: Kluwer, 2010, pp. 37–9; Martin Lehner, ‘Article 9 Associated Companies’, in M. Lang, T. Eckler, and G. Ressler (eds.), History of Tax Treaties, Berlin: Verlag Linde, 2011, pp. 398–9.
35 There were clearly some reservations on the provision, as indicated by the fact that, in the 1977 Model, Belgium, Finland, Germany, Italy, Japan, Portugal, and Switzerland reserved the right not to insert para. 2 in their conventions. The operation of Article 9(2) was subsequently discussed by the OECD in a 1984 report, Transfer Pricing, Corresponding Adjustments and the Mutual Agreement Procedure, which was contained in the volume, Transfer Pricing and Multinational Enterprises, Three Taxation Issues, Paris: OECD, 1984.
36 Transfer Pricing and Multinational Enterprises, Report of the OECD Committee on Fiscal Affairs, Paris: OECD, 1979. The Report has no legal force as such but its significance stems from being referred to in the 1992 revisions to the OECD Model as representing internationally agreed principles and providing valid guidelines for the interpretation of Article 9. Also, all OECD state ministries of finance adopted the 1979 report without reservation.
37 Transfer Pricing and Multinational Enterprises (n. 36), p. 7.
38 For example, significant publicity had been given to the case of Hoffmann-La Roche following the extensive 1973 report of the UK Monopolies and Mergers Commission (the forerunner to current UK Office of Fair Trading); Chlordiazepoxide and Diazepam: A Report on the Supply of Chlordiazepoxide and Diazepam, London: HMSO, 1973, which indicated blatant overcharging to a UK subsidiary for imports of Librium and Valium from another group entity. In consequence, the UK Monopolies and Mergers Commission ordered massively reduced prices to be charged and repayment of earlier overpricing of the product sold. There was also an additional UK tax liability. See further, Peter Muchlinski, Multinational Enterprises and the Law, 2nd edn, Oxford: Oxford International Law Library, 2007, pp. 275–6.
39 The US developments on transfer pricing were not merely domestic. The US recognized the need for an international consensus on the ALP and opened a dialogue with the OECD on the topic from 1965 leading to the establishment of an OECD working group on the topic. See Sol Picciotto, Is the International Tax System Fit for Purpose, Especially for Developing Countries? International Centre for Tax and Development, 2013, Working Paper 13, p. 17.
40 Transfer Pricing and Multinational Enterprises (n. 36), pp. 13–14 and 33–44.
41 Transfer Pricing and Multinational Enterprises (n. 36), pp. 73–4.
42 Transfer Pricing and Multinational Enterprises (n. 36), p. 33.
43 Cost-contribution arrangements involve an agreement between two or more members of a group for the sharing of costs and risks of developing intangible property in return for a specified interest in the property that may be produced. They are discussed at pages 57–8 of the Report and it is noted that no countries other than the US have specific rules relating to them.
44 Transfer Pricing and Multinational Enterprises (n. 36), pp. 48 and 59.
45 Transfer Pricing and Multinational Enterprises (n. 36), p. 56.
46 Transfer Pricing and Multinational Enterprises (n. 36), p. 14.
47 Transfer Pricing and Multinational Enterprises (n. 36), pp. 17–20.
48 Transfer Pricing and Multinational Enterprises (n. 36), p. 20.
49 Transfer Pricing and Multinational Enterprises (n. 36), p. 76. This is also derived from the 1968 US Regulations. See Reg. section 1.482–2 (b)(2).
50 See, for example, Transfer Pricing and Multinational Enterprises (n. 36), pp. 8–9 and 51.
51 Transfer Pricing and Multinational Enterprises (n. 36), p. 10.
52 Transfer Pricing and Multinational Enterprises (n. 36), p. 19.
53 The thinking on the matter was subsequently developed and led to the section in the 1995 Transfer Pricing Guidelines at paras. 1.36–1.41 on the ‘recognition of the actual transactions undertaken’, which is discussed below. See also the discussion on the scope of Article 9 in Chapter 4.
54 Transfer Pricing and Multinational Enterprises (n. 36), pp. 20, 26, 28.
55 Transfer Pricing and Multinational Enterprises (n. 36), p. 20.
56 Transfer Pricing and Multinational Enterprises (n. 36), p. 14. The 1979 Report was written in the middle of the UK/US disputes over state formulary apportionment. The 1975 renegotiation of the UK/US double tax treaty prohibited state imposition of taxation on UK companies under a global formulary apportionment approach. The US Senate refused to ratify this provision in 1978 under pressure from the states and concern about unequal treatment of US-based and UK-based companies. Several US Supreme Court cases addressed the constitutionality of state formulary apportionment, starting with the Asarco case in 1978 (Asarco Inc. v. Idaho State Tax Comm’n, 458 US 307 (1982)) and culminating with Container Corp. in 1983 (Container Corp. of America v. Franchise Tax Board, 103 S. Ct. 2933 (1983)). There was an intense US Treasury effort in the Container Corp. case to have state practices treated as unacceptable, though this effort ultimately failed. The UK Parliament made threats, Barclays filed a legal challenge against California directly posing the question of whether formulary apportionment as applied by the states to non-US-based companies was constitutional, and the Reagan administration sought to find a compromise, largely unsuccessfully. The comments on formulary taxation in the 1979 Report may be viewed in the context of the times and the opposition to formulary taxation of foreign corporations, which was a pressing international trade issue at the time. The US Treasury tended to favour the trading partner view and rejected the state view, which led them to be willing to use the OECD to try to further that anti-global formulary apportionment agenda. This explains the strong condemnation of formulary apportionment in the 1979 Report.
57 Transfer Pricing and Multinational Enterprises (n. 36), p. 15. A discussion repudiating formulary methods also features in the later 1995 Transfer Pricing Guidelines, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, Paris: OECD, 1995, ch. 3, paras. 3.58–3.74 (hereafter, 1995 OECD Guidelines).
58 See Michael Lennard, ‘The Purpose and Current State of the United Nations Tax Work’, Asia-Pacific Tax Bulletin (Jan./Feb. 2008), 23–4, for a summary of the early history of the United Nations, including the background to its involvement on international tax matters from 1967.
59 Fiscal Incentives for Private Investment in Developing Countries: Report of the OECD Fiscal Committee, Paris: OECD, 1965, paras. 163–5, as cited in United Nations Model Double Taxation Convention between Developed and Developing Countries, United Nations, May 1980, ST/ESA/102, p. 2.
60 United Nations Model Double Taxation Convention (n. 59).
61 The main differences between the OECD Model and the UN Model are discussed in Lennard (n. 58), pp. 25–7.
62 There is one difference however in that, as a result of an amendment made in 1999, the UN Model has a third paragraph to the effect that the ‘corresponding adjustment’ provisions of Article9(2) shall not apply where judicial, administrative, or other legal proceedings determine that one of the parties is liable to a penalty for fraud, gross negligence or wilful default. The Commentary recognizes this would have application in exceptional cases only.
63 The UN Commentary does, however, note that some developing countries wish to resist the automatic obligation to give effect to the ‘corresponding adjustments’ clause and so alter the terms of their obligation from ‘shall make an appropriate adjustment’ to ‘may make an appropriate adjustment’. United Nations Model Double Taxation Convention between Developed and Developing Countries (New York, 2011), p. 175.
64 It should be noted that the United Nations version of the threshold PE rule of Article 5 is also different in a number of respects, including a provision deeming a PE in the case of certain in-country services being rendered, a more limited approach to exemptions from PE status, a broader dependent agent rule and a narrower independent agent exemption.
65 Transfer Pricing and Multinational Enterprises, Three Taxation Issues, Paris: OECD, 1984, p. 3. The Three Taxation Issues volume contains three reports, being Transfer Pricing, Corresponding Adjustments and the Mutual Agreement Procedure; The Taxation of Multinational Banking Enterprises; and The Allocation of Central Management and Service Costs.
66 It is also recognized that similar issues arise in relation to the attribution of profit to permanent establishments, though that topic is not pursued. See Transfer Pricing, Corresponding Adjustments and the Mutual Agreement Procedure, included in Three Taxation Issues (n. 65), p. 14.
67 See Transfer Pricing, Corresponding Adjustments and the Mutual Agreement Procedure (n. 66), p. 18.
68 See Transfer Pricing, Corresponding Adjustments and the Mutual Agreement Procedure (n. 66), p. 18.
69 Three Taxation Issues (n. 65), pp. 19, 24, 26, 40.
70 The Taxation of Multinational Banking Enterprises, included in Three Taxation Issues (n. 65), p. 46.
71 The Allocation of Central Management and Service Costs, included in Three Taxation Issues (n. 65), p. 77.
72 The 1979 Report discussed above identifies three situations in which a profit mark-up is always appropriate. See Transfer Pricing and Multinational Enterprises (n. 36), pp. 80–1.
73 See, for example, Three Taxation Issues (n. 65), pp. 78, 79, 81.
74 Three Taxation Issues (n. 65), pp. 74, 83.
75 Thin Capitalisation, contained in the volume Issues in International Taxation No. 2, Paris: OECD, 1987. The report was compiled at a time when the leveraged buyout market was developing rapidly and amid concerns that generally levels of debt would skyrocket.
76 See, for example, Transfer Pricing and Multinational Enterprises (n. 36), which addressed the transfer pricing of loans and touched briefly on the topic of thin capitalization.
77 The 1987 Report was the first detailed discussion of the matter, though the earlier 1979 Report did also contain some tentative comments on the topic of thin capitalization. See Transfer Pricing and Multinational Enterprises (n. 36), pp. 86–9. On possible approaches to determining loan vs. equity characteristics, the 1979 report suggested the use of an approach under which the ‘special conditions of each individual case’ would be assessed (though it stopped short of referring unambiguously to an approach based on the ALP) but recognized that specialist staff and sophisticated analysis would be required. There is also a brief reference to recharacterization issues in the discussion of corresponding adjustments in the 1984 OECD Report and the discussion on banks also refers to the importance of the loan vs. equity discussion from 1979. See Three Taxation Issues (n. 65), pp. 12–13 and 52.
78 Thin Capitalisation (n. 75), pp. 9–11. The discussion, at pp. 10, 14, also includes a brief consideration of hybrid financing.
79 Thin Capitalisation (n. 75), p. 11.
80 See, for example, Thin Capitalisation (n. 75), pp. 11, 12, 13, 33.
81 Thin Capitalisation (n. 75), p. 13.
82 Thin Capitalisation (n. 75), p. 15. The Report comments that in an effort presumably to overcome these difficulties some countries have adopted a ‘fixed ratio’ approach under which interest on an amount of loan in excess of a certain proportion of equity is disallowed, though the Report is dismissive of this approach as inevitably arbitrary.
83 A number of other questions were raised relating to other articles of the Model Treaty. See Thin Capitalisation (n. 75), pp. 18–21 and 22–3. The Report also discussed the relative merits of a transfer pricing approach to excessive interest and debt as compared to a ‘fixed ratio’ approach, a matter which is discussed in chapter 7 at para. 7.97.
84 Thin Capitalisation (n. 75), pp. 21–2.
85 The report actually addressed the question whether Article 9 restricts or limits any domestic law adjustment made so that it cannot result in an adjustment exceeding the arm’s-length amount or whether Article 9 is ‘illustrative’ or ‘exemplary’ in the sense of providing a framework for the adjustment of profits, with the result that it would not prevent domestic law making an adjustment exceeding that which corresponds to an arm’s-length amount. On this point, countries seem to have had different perspectives but the Report states that there was general agreement that, in principle, thin capitalization rules ought not normally to increase taxable profits to any amount greater than the arm’s-length profit. Thin Capitalisation (n. 75), p. 22.
86 See Thin Capitalisation (n. 75), pp. 30–1. Consideration is also given to the possibility of a fixed ratio approach but the Report rejects this on the basis it ‘is bound to be arbitrary to some degree’ and it is also observed that ‘the majority of countries consider that the results would undoubtedly be inconsistent with the arm’s length principle’ (Thin Capitalisation (n. 75), pp. 31–2).
92 IRS Could Better Protect U.S. Tax Interests in Determining the Income of Multinational Corporations (GA0/GGD-81-81, 30 Sept. 1981, pp. 52–3). One commentator describes this period as involving ‘a gradual realisation by all parties concerned, but especially Congress and the IRS, that the ALS … did not work in a large number of cases, and in other cases its misguided application produced inappropriate results’ (Avi-Yonah (n. 9), p. 10).
93 A Study of Intercompany Pricing Under Section 482 of the Code, IRS Notice 88-123, 1988–2 C.B. 458 (hereafter, White Paper) which followed the suggestion of a conference committee working on TRA 1986 that the IRS should carry out a comprehensive study of the transfer pricing rules to determine if they could be improved. HR Conf. Rep. No. 841, 99th Cong., 2nd Session. NII-637 (1986), p. 638.
94 White Paper (n. 93), p. 472.
95 See, for example, White Paper (n. 93), p. 86. It may be argued that the one-sided testing approach started with the resale price and cost-plus methods of the 1968 Regulations.
96 The argument addressed is that the arm’s-length approach does not account for corporate synergies, or otherwise described, a return to the form of organization and this omission creates a ‘continuum price problem’ in which the sum of market-based returns for the individual group members is less than the actual return of the combined group. See White Paper (n. 93), ch. 10.
97 White Paper (n. 93), p. 83.
101 The 1993 Regulations considered five main methods for tangible property, being the CUP, resale price, cost plus, the comparable profit method, and profit split methods. For intangibles, the requirement was to use the comparable uncontrolled transaction method (CUT) (which combines the MTM and CATM of the 1992 Regulations) but if this could not be applied due to the absence of relevant data, then use of the comparable profit method, derived from the CPI of 1992, was required, though other methods such as a profit split method could be used if the reasons to support it were documented by the taxpayer.
103 The Regulations refer to the measure of ‘true taxable income’, being ‘the taxable income that would have resulted [to an entity] had it dealt with the other member or members of the group at arm’s length’. See (1994) 1.482–1(b)(1) and 1.482–1(i)(9).
104 The final report is Intercompany Transfer Pricing Regulations under U.S. Section 482 Temporary and Proposed Regulations, Paris: Committee on Fiscal Affairs, OECD (93)131, 1993. This preceded the 1994 Final Regulations.
105 See, for example, Transfer Pricing and Multinational Enterprises (n. 36), paras. 11, 15, 37, and 198. The ‘ex ante’ standard is stipulated in the 2010 OECD Transfer Pricing Guidelines. See Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, Paris: OECD, 2010, paras. 2.127–2.130 and 3.74 (hereafter, 2010 OECD Guidelines) and has remained the approach adopted in the overwhelming majority of countries, though Germany adopted rules similar to the US commensurate with income standard in section 1(3) of the Foreign Tax Act in 2007.
106 At this time, the use of profit-based methods outside the US was relatively limited. See IBFD survey discussed in Hubert Hamaekers, ‘An Introduction to Transfer Pricing: The American versus the European Approach’, INT-1563, paper prepared at the CEPAL seminar, 23–26 Jan. 1995, p. 19. There were also concerns that the US was repudiating its treaty obligations. See Clark (n. 23), pp. 1197–8.
107 In 1990, section 6662(a), (e), and (h)(2)(A) enacted accuracy-related penalties of up to 40 per cent for valuation misstatements and in 1993 these provisions were amended to operate according to whether contemporaneous documentation had been prepared.
108 UN Secretariat, ‘Transfer Pricing—History, State of the Art, Perspectives’, paper prepared at Tenth Meeting of the Ad Hoc Group of Experts on International Co-operation in Tax Matters, 26 June 2001, p. 11. See also N. Boidman, ‘The American Super Royalty Rule: A Canadian Perspective’, included in the 1988 conference report: Report of Proceedings of the Fortieth Conference, Canadian Tax Foundation, Toronto. However, the IRS has not in practice made extensive use of its power under the commensurate with income standard to make periodic adjustments. See Joseph L. Andrus, International Fiscal Association, Transfer Pricing and Intangibles, Vol. 92a, Cahiers de Droit Fiscal International (2007), US Branch report, p. 647.
109 1995 OECD Guidelines (n. 57).
110 The Guidelines also draw on the earlier OECD work on the US Proposed and Temporary Regulations. 1995 OECD Guidelines (n. 57), para. 14. For reasons which are not clear, the 1995 OECD Guidelines barely touch on the topic of thin capitalization.
112 The Guidelines have no mandatory force but the use of them is recommended by the Council of the OECD—see Recommendation of the Council on the Determination of Transfer Pricing between Associated Enterprises, C (95) 126/Final, as amended to take account of subsequent versions of the Guidelines.
113 1995 OECD Guidelines (n. 57), Preface, paras. 1 and 2.
114 1995 OECD Guidelines (n. 57), Preface, paras. 4 and 6.
115 1995 OECD Guidelines (n. 57), Preface, para. 7.
116 The rationale echoes comments first made more than fifty years earlier in the US case Advance Cloak Co., 1933 BTA Memo. at 108. ‘The purpose of [section 45] is to place transactions between related trades and businesses owned or controlled by the same interests upon the same basis as if such businesses were dealing at arm’s length with each other.’ A similar explanation was included in the 1935 Treasury Regulations on section 45, later to be section 482. See Article 45-1(b), Regulation 86 (1935).
117 1995 OECD Guidelines (n. 57), paras. 1.7, 1.8, and 1.12.
118 1995 OECD Guidelines (n. 57), paras. 1.45–1.48.
119 Transfer Pricing and Multinational Enterprises (n. 36), pp. 33–44.
120 1995 OECD Guidelines (n. 57), para. 1.42.
121 1995 OECD Guidelines (n. 57), paras. 1.17–1.35.
122 1995 OECD Guidelines (n. 57), paras. 3.2, 3.5, and 3.49–3.57. The increasingly common use by MNEs of highly interrelated global value chains does not seem to be reflected in the OECD discussion.
123 Transfer Pricing and Multinational Enterprises (n. 36), pp. 42–3. The Guidelines remain deeply opposed to formulary apportionment. See 1995 OECD Guidelines (n. 57), paras. 3.58–3.74. The main difference between profits-based methods that the OECD discusses and formulary apportionment is the absence in profits-based methods of a predetermined allocation formula.
124 1995 OECD Guidelines (n. 57), para. 3.5.
125 1995 OECD Guidelines (n. 57), para. 3.53. The concerns would presumably apply equally to the US CPM.
126 1995 OECD Guidelines (n. 57), paras. 3.55–3.56.
127 See, for example, the discussion in UN Secretariat (n. 108), pp. 25 and 35–6. Countries such as Germany remained deeply opposed to profit-based methods. See Opinion of the Ministry of Finance of 13 July 1995 (IStR 1995, 384).
128 The OECD position would be modified in BEPS but as recently as 2010 the use of hindsight was prohibited by the OECD. See 2010 OECD Guidelines (n. 105), paras. 9.56–9.57.
129 1995 OECD Guidelines (n. 57), paras. 1.68, 1.70, and 3.49.
130 1995 OECD Guidelines (n. 57), para. 1.5. See also Hubert Hamaekers, ‘Arm’s Length—How Long?’, International Tax Planning Journal (Mar./Apr. 2001), 34.
132 1995 OECD Guidelines (n. 57), para. 1.37. The first instance is a development of the position in the 1979 OECD Report. The second instance is introduced for the first time. The practical operation of these recognition rules is considered in Chapter 5.
133 The boundary, if any, between what is encapsulated in this first example and the (routine) application of transfer pricing provisions by reference to the real agreement between the parties is not explained and remains unclear. In legal terms, it is generally the case that the real agreement of the parties is determined by the intentions of the parties, taking account of all the relevant circumstances. This approach is adopted widely, in civil and common law countries.
134 The significant concerns with related party contracts first made in the 1979 OECD Report are also repeated in the 1995 OECD Guidelines, i.e. that associated enterprises are able to enter into contracts that would not be entered into by unrelated parties because the normal conflict of interest which would exist between independent parties is often absent. This includes the concern that ‘contracts within a MNE could be quite easily altered, suspended, extended, or terminated … and such alterations may even be made retroactively’ (1995 OECD Guidelines (n. 57), paras. 1.38–1.39). The discussion draws heavily on similar comments made in the earlier 1979 Report. See Transfer Pricing and Multinational Enterprises (n. 36), p. 20.
135 Given the remaining strong opposition on the part of some states to profits-based methods and the fact that the 1995 OECD Guidelines did not mandate any priority of methods, this opened the door to different views by states and therefore the possibility of double taxation.
136 The report is included in Three Taxation Issues (n. 65).
137 1995 OECD Guidelines (n. 57), paras. 4.78ff.
139 1995 OECD Guidelines (n. 57), paras. 5.16–5.27.
140 This is a revised version of the discussion in the 1984 OECD report, The Allocation of Central Management and Service Costs (n. 71), and generally recommends a cost-plus approach where a CUP method cannot be applied.
141 The chapters on intangible property (Chapter VI) and intra-group services (Chapter VII) were added in 1996 and Chapter VIII on cost contribution arrangements was added in 1997. There were no further additional chapters until 2010, when Chapter IX on business restructuring was also added.
142 See UN Secretariat (n. 108), pp. 37–40 and, generally, the compendium of country information on transfer pricing in International Transfer Pricing 2012, PwC (2012).
143 Significant additional information on the extent of transfer pricing regulations is available in Theresa Lohse, Nadine Riedel, and Christoph Spengel, The Increasing Importance of Transfer Pricing Regulations—A Worldwide Overview, Oxford University Centre for Business Taxation, Working Paper 12/27.
144 The tax systems of a number of countries have for many years contained transfer pricing requirements which have been in place as a result of provisions in the tax code or judicial decisions—e.g. Sweden (1928), Argentina (1932), France (1933), Australia and Italy (1936), and the Philippines (1939). The US and Canada were the first states to include a comprehensive arm’s-length provision in their domestic law in 1924. The country information draws on the compendium of country information on transfer pricing in International Transfer Pricing 2012, PwC (2012).
145 Lohse, Riedel, and Spengel (n. 143), pp. 7, 12. An interesting by-product of the expansion was that by the beginning of the BEPS Project, over fifty countries were regularly attending the transfer pricing discussions in the OECD’s Working Party No. 6.
146 The origin of the EU Arbitration Convention dates back to the European Commission’s 1976 proposal for a directive to eliminate double taxation in the case of transfers of profits between associated enterprises in different Member States (Official Journal C 301 of 21 Dec. 1976) reflecting earlier concerns on dispute resolution.
148 Earlier OECD work on permanent establishments included the 1984 OECD Report on the taxation of bank branches, The Taxation of Multinational Banking Enterprises (n. 70) and a more general report in 1994, Model Tax Convention: Attribution of Income to Permanent Establishments, Issues in International Taxation No. 5, Paris: OECD.
149 The original intention had been that the Transfer Pricing Guidelines would apply to both associated enterprises and permanent establishments, but the work on permanent establishments was postponed. OECD, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, Discussion Draft of Part I (8 July 1994), p. 16, para. 16. Even as late as 1997, the intention was for the work on permanent establishments to form a chapter of the Transfer Pricing Guidelines. See Summary Record of the Fifty First Meeting of Working Party No. 6, document DAFFE/CFA/WP6/M(97)1 p. 7, available at http://www.taxtreatieshistory.org [accessed 12 Apr. 2017].
150 One of the two broad approaches was the ‘relevant business activity’ approach which defines the ‘profits of an enterprise’ as referring only to the profits of the business activities in which the PE has some participation (hence, ‘relevant business activity’). Under this approach, the profits to be attributed to a PE would be based on the appropriate share of profits from the relevant business activity of the enterprise as a whole. Such profits would also include profits and also losses from such ‘relevant business activity’ of other parts of the enterprise. This meant that losses in the head office or in another PE of the enterprise would reduce the pool of profits to be apportioned. The other broad approach was the ‘functionally separate entity’ approach, which does not limit the profit attributed to the PE by reference to the profits of the enterprise as a whole or by reference to the profits from any particular business activity in which the PE has participated. The profits to be attributed to the PE are the profits 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. The latter approach is equivalent to the result intended by Mitchell B. Carroll in his work on the allocation of profits, and as reflected in the 1933 model treaty on the allocation of profits, as discussed in Chapter 1. For an extended discussion of the historic position, see Discussion Draft on the Attribution of Profits to Permanent Establishments, Paris: OECD, Feb. 2001, at pp. 8–12.
151 The degree to which these problems arose recurrently in practice remains unclear, casting some doubt on whether the long-running project, which dominated the work of the Committee on Fiscal Affairs for over a decade, was entirely warranted, particularly given the range of issues that were being experienced with transfer pricing while this work progressed. It has also been argued that, for various reasons, the higher level of latitude historically accorded to the application of Article 7 of the OECD Model was beneficial and therefore preferable to the new AOA. See Richard Vann, Reflections on Business Profits and the Arm’s Length Principle, Sydney Law School, Legal Studies Research Paper, No. 10/127, Nov. 2010.
152 Report on the Attribution of Profits to Permanent Establishments, Paris: OECD, 17 July 2008 and 2010 Report on the Attribution of Profits to Permanent Establishments, Paris: OECD, 22 July 2010 (hereafter, 2010 PE Report).
153 2010 PE Report (n. 152), p. 11.
154 In contrast to what had gone before, this new approach to capital introduced much greater clarity on the required capital attribution process to PEs under the OECD Model. The guidance also clarified that capital may be deemed to be held by a PE, irrespective of the actual capital held by that PE.
155 On the approach to step one, see 2010 PE Report (n. 152), pp. 14–20, 24–49.
156 On the approach to step two, see 2010 PE Report (n. 152), pp. 20, 49–57.
157 2010 PE Report (n. 152), pp. 19, 33.
159 The provision is similar to the ‘corresponding adjustment’ provisions first introduced in the London and Mexico Drafts and discussed in Chapter 1.
160 The OECD approach has been criticized by Richard Vann of the University of Sydney on the basis that, while it is appropriate to pursue a common approach under Article 9 and Article 7, it should be in the opposite direction to that adopted by the OECD (under which Article 9 is pre-eminent) with the result that it should be based on Article 7, recognizing the economic unity of the MNE group. See Vann (n. 151).
163 2010 PE Report (n. 152), pp. 22, 58–63, 165–8, and 196–8. The PE Report generally recommends the filing of two tax returns, one relating to the company concerned and one relating to the PE, which is created by the activities of the company.
166 For example, one of the issues addressed was the ‘all or nothing’ approach to the taxation of profits from activity by a local group entity participating in a global trading arrangement. See UN Secretariat (n. 108), p. 37.
168 1998 Global Trading Report (n. 167), p. 44.
169 1998 Global Trading Report (n. 167), pp. 7, 34, 41, and 44.
170 1998 Global Trading Report (n. 167), p. 27.
172 1998 Global Trading Report (n. 167), p. 46. With the development of the AOA and the use of the KERTs concept, the functions of risk assumption and risk management had together become the most important functions in the KERTs analysis for a global investment banking business.
173 2010 PE Report (n. 152), p. 135.
174 2010 PE Report (n. 152), pp. 139–41.
175 United Nations Practical Manual on Transfer Pricing for Developing Countries, New York, 2013, ST/ESA/347 (hereafter, UN Practical Manual). The 2013 version of the Manual is organized into ten chapters dealing with a variety of matters including the business and legal environment, comparability, transfer pricing methods, documentation, and audit. The 2017 revision can be found at: http://www.un.org/esa/ffd/wp-content/uploads/2017/04/Manual-TP-2017.pdf [accessed 12 Apr. 2017].
176 The wording cited appears at UN Practical Manual, para. 18.104.22.168. The general concerns referred to are illustrated, for example, in the letter of 9 Oct. 2012 from the US Council for International Business (USCIB) to Michael Lennard, the Chief of the International Tax Cooperation Section at the UN on the subject of the UN Transfer Pricing Manual. The letter raises a number of concerns and is available at http://www.un.org/esa/ffd/tax/eighthsession/USCIB_letter_on_UNTPM_October_9.pdf [accessed 12 Apr. 2017].
177 For example, the description of the approach in India puts significant emphasis on the concept of location savings, as regards both the scope of the concept and the value of such savings. See UN Practical Manual (n. 175), para. 10.4.7. The discussion on Chinese practice suggests that in some circumstances a ‘group’ rather than separate entity perspective may be required. See para. 10.3.5.2.
179 For an interesting summary of the challenges and issues relating to the Chinese approach to operating the ALP, see Jingyi Wang, ‘The Chinese Approach to Transfer Pricing: Problems Faced and Paths to Improvement’, British Tax Review 1 (2016), 89.
180 2010 OECD Guidelines (n. 105), paras. 9.1–9.2.
181 This has been seen as widening the scope of possible tax authority challenges to such arrangements, though the ground had already been covered in the US where the IRS already had avenues of attack under the detailed Sec. 482 Regulations. See Deloris R. Wright and Harry A. Keates, ‘Comments on the OECD Discussion Draft on Transfer Pricing Aspects of Business Restructurings’, International Transfer Pricing Journal (Mar./Apr. 2009), 115.
182 2010 OECD Transfer Pricing Guidelines (n. 105), paras. 1.23–1.27. The expanded discussion of risk appears in the 2010 OECD Guidelines (n. 105), paras. 9.10–9.47 and on the status of ‘control’ see paras. 9.20–9.21.
183 2010 OECD Guidelines (n. 105), paras. 9.59, 9.189, and 9.194. The discussion is a material expansion of the previous guidance.
184 2010 OECD Guidelines (n. 105), paras. 9.105, 9.141, and 9.164.
185 This is made clear by the 2010 OECD Guidelines (n. 105), para. 9.174.
186 2010 OECD Guidelines (n. 105), paras. 9.94 and 9.134–9.138. The comments on the aggregated approach are clearly an expansion of the approach initially contemplated for certain product lines or ‘groupings’. See Transfer Pricing and Multinational Enterprises (n. 36), para. 41 and 1995 OECD Guidelines (n. 57), paras. 1.42–1.44.
187 2010 OECD Guidelines (n. 105), para. 9.161.
188 2010 OECD Guidelines (n. 105), paras. 9.165–9.166, 9.171, and 9.180.
189 2010 OECD Guidelines (n. 105), paras. 9.183–9.187.
190 2010 OECD Guidelines (n. 105), paras. 9.181–9.182 and see also paras. 9.57 and 9.83.
191 2010 OECD Guidelines (n. 105), para. 9.179.
192 2010 OECD Guidelines (n. 105), para. 9.174. The difficulty of assessing commercial rationality in the case of alternative options available is brought out in the examples discussed at 9.69–9.73.
193 See, for example, 2010 OECD Guidelines (n. 105), paras. 9.19, 9.52, 9.138, 9.172, and 9.173.
194 The OECD position on this matter has remained broadly unchanged from the 1979 Report. See Transfer Pricing and Multinational Enterprises (n. 36), para. 38.
195 2010 OECD Guidelines (n. 105), para. 9.6. See also paras. 9.84, 9.134, and 9.173.
196 2010 OECD Guidelines (n. 105), paras. 2.1–2.11.