Canadian transfer pricing rules keep pace with international developments
J. Scott Wilkie
The transfer pricing rules in the Canadian Income Tax Act (the Act), and Revenue Canada’s practices in applying them, are being modified to more closely reflect the “arm’s length principle” as articulated by the Organisation for Economic Co-operation and Development (OECD), and to conform to and adopt by reference the revised transfer pricing guidelines(1) announced by the OECD in stages beginning in July 1995 (the OECD Guidelines). Canadian transfer pricing rules and practices have consistently assimilated and been applied according to the arm’s length principle and associated analytical and methodological guidelines(2) as set out by the OECD. Broadly, the reported income of Canadian taxpayers engaged in transactions with other members of controlled groups of which they are members in circumstances that are integrated functionally is measured with reference to transactional terms and conditions, including pricing, that would have been present in the absence of organizational and other unique relationships among the parties to the transactions. The changes, which do not substantially alter Canadian transfer pricing rules and practices, continue Canada’s adherence to this principle, but in a more contemporary fashion. These changes mirror recent developments internationally but perhaps most notably in the United States.
Importantly, the changes to the Act(3) include a specific penalty for deficiencies in reported income arising from insupportable and undocumented transfer pricing. Also included is a statutory requirement that transactional transfer prices be described in a materially “complete and accurate” manner in contemporaneous documentation, prepared when pricing determinations are made, that analyzes these determinations with reference to the corporate organizational, functional, and economic context in which transactions occur. The changes also expand the foreign reporting rules with respect to transactions between Canadians and non-residents, primarily to ensure that transactions involving partnerships and their members fall comprehensively within the ambit of these rules.
The legislative changes are included in a Notice and Ways and Means Motion to amend the Act (the Notice) tabled by the Minister of Finance in the House of Commons on December 8, 1997, and now Bill C-28, which was given first and second reading in the House on December 10, 1997, and February 4, 1998, respectively. The Notice introduces section 247 into the Act as the contemporary statutory embodiment of Canada’s transfer pricing rules, replacing its precursor in subsections 69(2) and 69(3) of the Act. In addition, the foreign reporting rules in section 233.1 of the Act are amended. The transfer pricing changes in the Notice replace an earlier version released on September 11, 1997, accompanied by a draft of proposed Revenue Canada Information Circular 87-2R (the “Draft Circular,” a revision of Revenue Canada’s longstanding statement of its administrative practice in this area). A variety of constructive modifications were made to the September 11 draft to reflect comments offered by taxpayers and their professional advisers. The Draft Circular remains subject to public comment and revision, but is expected to be released in its final form in 1998.
The new rules apply generally to taxation years or fiscal periods beginning after 1997,(4) although the penalty and its supporting rules, including the new contemporaneous documentation requirements, do not apply until taxation years or fiscal periods beginning after 1998 and in any event do not apply to transactions completed before September 11, 1997, when the rules first were announced in draft form.(5)
An Overview of the New Rules
Section 247 has three main elements: a statutory expression of the arm’s length principle as the basis on which adjustments to income and the cost of property may be made'(6) a specific transfer pricing penalty;(7) and documentation requirements that, if not satisfied at a threshold level, will engender the imposition of the penalty(8) in respect of adjustments to income and the tax cost of property.
A. The Arm’s Length Principle and Adjustments
Subsection 247(2) sets out the statutory basis for adjusting income with reference to an arm’s length analogue — effectively a contemporary incorporation in the Act of the OECD’s formulation of the arm’s length principle. It applies on a transaction (or series of transactions) basis and calls for adjustments to the “quantum or nature” of amounts where either transactional terms and conditions are different from those for transactions between arm’s length parties,(9) or the transaction (or series) itself would not have been undertaken by arm’s length parties and is primarily tax-motivated (i.e., “can reasonably be considered not to have been entered into primarily for bona fide purposes other than to obtain a tax benefit”).(10) In the first case, an adjustment will be based on the terms or conditions of a transaction that arm’s length parties would have adopted. In the second, the revenue authorities are authorized to redetermine amounts based on the transaction that arm’s length parties would have implemented on arm’s length terms — in other words to make the transfer pricing determination with reference to a substituted transaction rather than the actual transaction.(11)
Subsection 247(2) is a broad, descriptive formulation of the arm’s length principle. Despite the adoption for other purposes in section 247 of definitions of “arm’s length transfer price”(12) and “transfer price,”(13) which embellish the arm’s length principle by referring to the kinds of transactions that may be affected, subsection 247(2) is not so limited and in fact does not mention transactional pricing (though it does have a transactional orientation). Neither does it directly apply any reasonableness standard to the adequacy of transfer pricing determinations in affected circumstances. Although such a standard is probably implicit in the nature of transfer pricing analysis and the analytical methods commonly adopted for evaluating pricing and the implications of how they should be applied, the elimination of a specific statutory reference to such a test, as found in existing subsections 69(2) and (3) of the Act, makes this expectation less clear than would be desirable. The restatement of the transfer pricing rules without this qualification raises a question whether the restated rules allow less for the inherent imprecision of transfer pricing analysis than did its predecessors.
The Draft Circular contains, as before, the methodological basis for determining the adequacy of and adjusting transfer pricing. Revenue Canada’s general expectation that pricing determinations reflect the application of a transactional method (“comparable uncontrolled price,” “cost plus,” and “resale minus”), rather than a profit split or comparison, continues and, if anything, is more clearly stated, although to much the same effect as in the existing administrative practice.(14) Relying substantially on the OECD Guidelines(15) the Draft Circular reiterates the importance attached by Revenue Canada to evaluating transfer prices by reference to comparable transactions of the tested taxpayer with arm’s length parties (“internal comparables”) or of third parties at arm’s length with each other (“exact comparables”).(16) In short, Revenue Canada continues to express a strong preference for an analysis grounded in a “comparable uncontrolled price,” but in any event supported by either of the other two traditional methodologies, “cost-plus” and “resale-minus.”
Nevertheless, there is cautious recognition(17) of the transactional profit methods–residual and contribution profit-split, and the “transactional net margin method” (TNMM) developed as part of the OECD Guidelines (and, in the minds of some, very similar in concept to the U.S. “comparable profit method” (CPM))–that are acknowledged in a similarly guarded fashion in the OECD Guidelines.(18) Revenue Canada expresses a preference for a residual profit-split approach, rather than one based on relative contributions of the parties to an enterprise that includes the affected transactions, and clearly considers a profit-split approach to be superior to any kind of marginal or comparative profit analysis recognized in the guise of the TNMM, which is expressly considered to be a method of last resort. Indeed, the Draft Circular echoes the OECD’s view that these profit-based methods are intrinsically transactional. This assertion seemingly gives a transactional connotation to the profit-oriented approaches and therefore draws them closer to the typical methodologies by requiring that transfer prices among non-arm’s length parties be defensible by reference to comparable transactions and circumstances of independent counterparties. The application of profit-split methods, in principle, is confined by the stated expectation that they will only be useful where there is a considerable degree of corporate integration in the affected group of which the Canadian taxpayer is a member and a high degree of transactional comparability in relation to the data and circumstances referenced in a profit-based analysis.(19)
Subsection 247(2) replaces subsections 69(2) and (3) of the Act, which permitted amounts received to be increased and payments made to be decreased, in respect of various transfers of property and the performance of services involving arrangements between Canadian residents and non-residents. These provisions allowed for adjustments to conform to what “would have been reasonable in the circumstances if the non-resident person and the taxpayer had been dealing at arm’s length…for the purpose of computing the taxpayer’s income….” Although Revenue Canada has consistently characterized this formulation of the arm’s length principle as a mere reiteration of the OECD version, and accordingly insisted on the application of the typical methodological analyses in relation to transactional pricing (and, additionally, presumed that pricing respectful of this principle would reflect “fair market value”), it is not clear that this is what those provisions required. Indeed, the old formulation is in some respects closer to the statement of the arm’s length principle in Article IX of the OECD Model Convention (the associated enterprises article)(20) on which the OECD Guidelines depend for their significance than the statement of the principle in the OECD Guidelines as they commonly are understood.
Subsections 69(2) and (3) applied in respect of the computation of “income,” a net or aggregate concept to require that amounts paid and received be “reasonable in the circumstances.” This manner of expressing the arm’s length principle acknowledged the dependence of transfer pricing analysis on the facts and circumstances of specific taxpayers, notwithstanding the use of generally recognized methodological techniques or guidelines associated with transfer pricing analysis. Hence, it effectively provided for “income” rather than transaction-by-transaction pricing to be evaluated against the arm’s length standard. Indeed, in some respects this could be interpreted as an embryonic profit provision, though it was not generally acknowledged by the revenue authorities to have this significance. This interpretation reflects a close correspondence with Article IX, which contemplates a reallocation of “profits” to accord with what would have been appropriate for dealings between independent enterprises,(21) and with comments in this vein by the OECD in the OECD Guidelines. In Article IX, there is no specific mention of transactional pricing.
Although this point may be principally of academic interest in respect of the Canadian legislation, it is fundamental to an appreciation of the overall objectives of transfer pricing analysis and consequently how more transactional expressions of transfer pricing rules, informed by administrative guidelines, should be applied. The focus, despite the necessary attention to methodological analyses of transaction flows as elements of this determination, is on arriving at reported tax income of a taxpayer that makes sense in light of its contribution to a functionally integrated or combined enterprise. In part, this is why the legislative expression of transfer pricing rules is of much less significance than the manner in which reported income is evaluated with transfer pricing considerations in mind.
B. Transfer Pricing Penalty
Subsection 247(3) creates a specific penalty applied to transfer pricing adjustments if a taxpayer’s reporting practices are deficient or materially faulty. The basic rule is augmented by special rules dealing with transactions that involve partnerships and their members as well as an anti-avoidance rule to prevent distortions of the penalty calculation by manipulating revenue to maximize the threshold for the penalty’s application and a denial of the obligation of the revenue authorities to make offsetting adjustments in a taxpayer’s favor.
This penalty has the same sense and significance as its U.S. counterpart, though there are some notable differences. The Canadian provision applies a substantial — 10 percent — penalty rate to the net transfer pricing adjustment (the sum of positive income adjustments and negative cost reductions for capital and depreciable property owned directly or through a partnership) rather than simply to the additional tax that results from the adjustment. Subject to a “safe harbor” of sorts expressed with reference to a taxpayer’s gross revenue, which confines the application of the penalty to situations where the net positive transfer adjustment exceeds the lesser of 10 percent of the taxpayer’s gross revenue otherwise determined and $5 million, the penalty applies to adjustments and may be imposed even if the taxpayer is not in a net tax paying position. Furthermore, this “safe harbor” does not limit the penalty base if the threshold for the penalty’s application is exceeded; it merely sets out a “bright line” test that effectively switches the penalty “on” for all affected adjustments. The penalty should not apply to taxpayers who have made “reasonable efforts to determine arm’s length transfer prices or arm’s length allocations in respect of the transaction[s]” according to analytical and documentation standards anticipated in subsection 247(4).
Subsection 247(4) of the Act sets out a framework for specific analytical and documentation standards that taxpayers are expected to meet in reporting income with transfer pricing considerations in mind. In the main, this is a supporting rule for the penalty, although, in fact, it has broader significance. Because of its penalty orientation, however, the provision establishes positive documentary and recordkeeping requirements–“contemporaneous documentation ” standards –only by implication. This means that a failure to make “reasonable efforts” in determining transfer pricing–the linchpin to imposing a transfer pricing penalty — will be assumed unless a taxpayer analyzes and documents its transactions according to various categories of information “complete[ly] and accurate[ly] in all material respects.” On the other hand, the legislation provides no positive standard for evaluating when compliance will necessarily be considered to be reasonable measured against this broad expectation.
Section 233.1, which sets out the “normal course” foreign reporting obligation of Canadian taxpayers for their transactions with non-arm’s length non-resident counterparties, is modified principally to extend its application in respect of partnerships and their members, individuals, and trusts. Information returns will be required for transactions between reporting persons and non-arm’s length non-residents and partnerships of which such non-residents are members. Additionally, partnership reporting will be required with respect to transactions with nonarm’s length non-residents as well as non-residents with whom members of the partnership do not deal at arm’s length. In all cases, section 233.1 reporting is required only outside a “safe harbor”–specifically, if the cumulative fair market value of property transferred and services provided in relevant reportable transactions exceeds $1 million for the reporting person or partnership.(22)
D. Other Aspects
1. Definitions. For all three main elements of the new provision, subsection 247(1) contains supporting definitions of relevant terms, notably “arm’s length transfer,” “transfer price,” and various calculation concepts relevant principally to the penalty — namely, “transfer pricing capital adjustment,” “transfer pricing capital setoff adjustment,” “transfer pricing income adjustment,” and “transfer pricing income setoff adjustment.” Section 247 also applies specifically to “qualifying cost contribution arrangements,” a new legislative expression of arrangements already recognized in the relevant practice of Revenue Canada. It is interesting that this last development virtually coincided with the publication of Chapter VIII of the OECD’s transfer pricing guidelines which contain a new recognition of cost-sharing arrangements.
2. Adjustments Favoring Taxpayers. Potentially, one of the most significant supporting rules is contained in subsection 247(10), which provides that revenue authorities are not required to make any favorable transfer pricing adjustments in respect of taxpayer’s transactions. An adjustment that does not increase (1) a “transfer pricing capital adjustment” (by decreasing the tax cost of capital or depreciable property) or (2) a “transfer pricing income adjustment” (by increasing an amount of revenue or decreasing a cost or expense) is not required. In a sense this is not a surprising limitation, for it is entirely consistent with the discretionary nature of the “corresponding adjustment” or “correlative relief” aspects of Article IX of the OECD Model Convention and, generally, of Canada’s tax treaties. Notably, this is the only specific circumstance identified in the draft Information Circular in which this discretion is anticipated.(23)
Given the transactional orientation of the new Canadian transfer pricing rules, subsection 247(10) poses a potential limitation on the application of the penalty on a net adjustment basis. Furthermore, the statutory language adopted in subsection 247(10)–adjustments other than in favor of the revenue authorities “shall not be made” unless in the circumstances considered by them to be appropriate–coupled with certain objective characteristics of the adjustment in subsection 247(2)–if terms, conditions, or transactions depart from an arm’s length norm, the amounts “shall be adjusted” in accordance with that provision — may, at first blush, seem to suggest that transfer pricing adjustments unfavorable to a taxpayer must be made and are virtually automatic and, indeed, are mechanical.
The use of mandatory language in setting out what are essentially the limits of the exercise of administrative discretion seems somewhat odd, and perhaps may obscure the real limits that likely will apply in practice. Although there may be legitimate concern here about the phrasing of the legislation and the uncertainty it creates, it is nevertheless reasonable to expect that the new rules will not be applied so harshly in practice.
Some helpful and perhaps instructive inferences about the practical scope and significance of this provision may be drawn from the arm’s length principle itself and the nature of transfer pricing analysis generally. Even allowing for the possibility that this mandatory connotation of adjustments against taxpayers is an accurate implication based on the wording of subsection 247(10), the administrative latitude required to avoid mechanical transfer pricing adjustments is found in the analytical judgment required, and generally recognized to be present, in applying the arm’s length principle. This is not a complete answer, but it does explain why adjustments under subsection 247(2) are not as mechanical and perfunctory as they may seem. While the new rules and the restated administrative practices do not speak discursively to considerations such as range analyses and continuums of defensible transfer pricing results, the inherent imprecision of transfer pricing analysis generally makes it unlikely to expect irrational and ultimately unfair overall results. In this regard, the extensive discussion in the OECD Guidelines of factors relevant to making a pricing determination — which emphasizes the significance of taxpayers’ unique characteristics in modifying how a pricing analysis might otherwise take place — and the guarded recognition of transactional profit analyses suggest that arbitrary and unfair overall results should not be expected or, indeed, tolerable under the new rules if properly and thoughtfully applied. Furthermore, it would be surprising if Canadian revenue authorities did not acknowledge the unfairness of ignoring adjustments in a taxpayer’s favor except in unusual or special circumstances. On the basis of Revenue Canada’s historical approach to transfer pricing, the various manifestations of its guidelines, and the objectives and reasons for transfer pricing analysis, one should expect only rarely that the discretion in subsection 247(10) would be exercised to the detriment of a taxpayer.
As persuasive and sensible as this evaluation of subsection 247(10) is, the absence of any instructions or limitations in the legislation on how the discretion afforded to Revenue Canada may be exercised is troublesome. The elimination of a specific statutory “reasonableness” standard in subsection 247(2) magnifies this concern. It would be helpful if the final statement of Revenue Canada’s revised administrative practice in Information Circular IC 87-2R addressed the statutory interpretation and application issues discussed here and provided guidance to mitigate fundamentally unnecessary and unreasonable concerns that could develop about the latitude provided to Revenue Canada in subsection 247(10) in the context of section 247 generally.
3. Partnerships. Adjustments contemplated by subsection 247(2) apply in respect of transactions undertaken by partnerships and their members. There are three relevant cases for partnership dealing: (1) transactions between a partnership and a non-resident person with whom the partnership does not deal at arm’s length; (2) transactions between the partnership and a non-resident person with whom a member of the partnership does not deal at arm’s length; and (3) transactions between the partnership and another partnership of which the non-arm’s length non-resident (i.e., the person not at arm’s length with the tested partnership or a member of it) is a member. Tiers of partnerships are ignored for purposes of applying section 247 to partnerships.(24) Supporting rules for the penalty in subsection 247(3) apportion the tax cost of partnership property and the amount of partnership revenue among partners based on their relative entitlements to distributions of profit or allocations of loss. The expanded scope of foreign information reporting in section 233.1 mirrors the terms on which section 247 applies in respect of partnerships and their members.
A number of interesting issues can be anticipated in this area. First, it is far from clear how to determine whether two partnerships, persons and partnerships, and members of partnerships and the partnerships of which they are members are dealing at arm’s length. Partnerships are not themselves taxable under the Act; they are treated as taxpayers essentially as a computational device to determine income and loss allocable, and only taxable, to their members. Second, the scope of these rules, as in the case of the relatively new “tax shelter” rules in the Act, anticipates an awareness of and an ability to identify and evaluate relationships between participants in transactions with partnerships, and persons who do not deal at arm’s length with members of partnerships, even through tiers of partnerships. Both from taxpayer compliance and tax administration perspectives, it is difficult to envisage how affected taxpayers could or will be able to know about or discover relationships that may bring them within the ambit of these rules. While it may be prudent to seek protection by way of commercial representations and warranties, and perhaps covenants about ongoing behavior, these will not negate adverse tax results. At most, they will establish a basis for a civil remedy in respect of the tax liability attributable to the undisclosed relationship.
4. “Qualifying Cost Contribution Arrangement.” Almost contemporaneously with the announcement of new guidelines and expanded commentary by the OECD,(25) the Act is to be amended to apply section 247 to cost-sharing arrangements. A “qualifying cost contribution arrangement” involves the contribution, on a reasonable basis, by participants to the cost of “producing, developing, or acquiring any property, or acquiring or performing any services, in proportion to the benefits…reasonably expected…” to be realized by them. The Act has not previously recognized such arrangements though former Information Circular 87-2 implicitly addressed cost-contribution arrangements in connection with research and development programs. The Draft Circular explains how section 247 will apply to these arrangements.(26) The normative standard is what arm’s length parties would have done in comparable circumstances.(27) Parties are expected to benefit mutually from the pooling of their resources and relatively to their contributions. The Draft Circular sets out various ways — so-called “allocation keys”–in which the relative benefit may be measured, primarily in relation to income or cost savings expected to be achieved.(28) Mirroring notions adopted by the OECD, the Draft Circular anticipates the possible need for balancing payments in the event that parties’ contributions must be adjusted to properly reflect their relative benefits from participating in the arrangement; for similar reasons, buy-in payments are also contemplated.(29)
As noted in the Draft Circular, payments between the parties for the use of property are not required since each participant acquires a “beneficial interest in the property.” The OECD Guidelines also speak in terms of such an interest, and expand upon the circumstances in which beneficial interests in property and its legal ownership may be separate in circumstances that nevertheless satisfy the requirement of an acceptable cost-sharing arrangement.(30)
The attention paid to cost-sharing as an element of transfer pricing is interesting for various reasons. Cost-sharing is somewhat outside the scope of transfer pricing rules in a strict sense, since such arrangements are premised on the absence of any transfers between the participants, although embedded in the concept of sharing are notional economic transfers between, or joint dealing by, participants. Furthermore, describing a property interest in the outcome of a pooled enterprise by reference to a “beneficial interest” in property is curious. How a “beneficial interest” is determined and reflected commercially while satisfying the tax requirements usefully could be a subject of further consideration and development in completing the Draft Circular.
A. Application of the Arm’s Length Standard: Areas of Concern
Subsection 247(2) applies where either the terms or conditions of a transaction (or series) “differ” from an arm’s length analogue, or the transaction (or series) would not have been implemented by arm’s length parties and lacks a primary non-tax purpose. The “quantum or nature” of affected amounts may be adjusted, in the first case with reference to the terms or conditions that arm’s length parties would have adopted, and in the second by substituting, for this analysis, the appropriate arm’s length transaction.
The adoption in the Act and the administrative practices of Revenue Canada of the OECD’s contemporary conception of the arm’s length principle in itself is not problematic. Indeed, in principle it changes nothing. The adjustment rule in subsection 247(2), however, is of potentially wider scope than the current Canadian practices for three main reasons. First, it is not confined to particular or normative transactional circumstances. Second, it appears to accommodate, even require, transactional recharacterization even though transfer pricing analysis is generally grounded in the actual transactions undertaken by the parties. Third, unlike its predecessor provisions, subsection 247(2) does not incorporate any apparent limitation by way of a circumstantial reasonableness standard. How serious are these concerns?
1. Covered Transactions. Section 247 defines a transaction” to include “an arrangement or event.”(31) The scope of this notion is not confined by subsection 247(2). Accordingly, any intra-group transfers as well as other dealings within the group are conceivably affected by these rules.(32) Except with respect, possibly, to financial transactions between members of a multinational group, however, the scope of this rule is probably not more extensive than is currently the ease.
The possible extension of these rules to intra-group financial transactions, and in particular to loans by “foreign affiliates” of Canadian corporate taxpayers to their Canadian shareholders and to other “foreign affiliates” (both relatively common situations), has raised some concerns. An expansion of the scope of the transfer pricing rules in this regard would be a notable development. Canadian transfer pricing rules have historically not been applied to financial transactions, in part because the circumstances contemplated in subsections 69(2) and (3)(33) of the Act did not easily encompass transactions of this nature. Does the generality of the new statement of the arm’s length principle in–subsection 247(2) portend a change in this area?
For example, will subsection 247(2) apply to redetermine the nature of a financial transaction between parties to be equity rather than debt where a loan would not have been arranged in analogous arm’s length circumstances? Will interest be considered to have been paid at rates or on terms other than as specified by the parties? A reason for concern is the specific exclusion from the ambit of these rules of one particular formulation of an intra-group financial transaction. Specifically, subsection 247(7) provides that subsection 247(2) does not apply to loans referred to in subsection 17(3) of the Act (i.e., loans by Canadian resident corporations to “subsidiary controlled corporations” for use in their businesses). What is the need for such a specific exclusion if financial transactions are otherwise not generally covered by the statute? Furthermore, the Draft Circular acknowledges that the transfer pricing rules would accommodate recasting debt as equity based on an arm’s length comparison. A specific qualification on the likelihood of such a recharacterization where the thin-capitalization rules in subsections 18(4) to (8) of the Act apply also fortifies the possibility that financial transactions are covered.(34)
The answer to these concerns has two potential aspects. First, the transfer pricing practices applied in Canada have always anticipated the possibility that the actual transactions undertaken by taxpayers could be ignored in favor of a more accurate formulation of their substance. Canada adopted the OECD’s 1979 guidelines and has endorsed the revised OECD Guidelines. Both anticipate the possibility that the form of transactions could be recast. Both also, however, clearly state that usually the actual transactions undertaken by the parties are to be respected as the basis for a transfer pricing analysis. This expectation is repeated in the Draft Circular, which defers to parties’ actual transactions except where the form and substance differ or a transaction is not “commercially rational” with reference to the behavior of arm’s length parties. Although the present statement by the OECD of the limited circumstances in which recharacterization of a transaction or a payment may be justified is more descriptive than its 1979 predecessor, it is not materially different in its essence. Consequently, in principle, the new rules probably do not present any new or greater likelihood of recharacterization.
An important brake on this concern is also supplied by the Canadian legal context. Although there is much discussion now, in terms of the scope of the “general antiavoidance rule” in section 245, about the permissible limits of transactional recharacterization, the Canadian legal tradition is to respect a taxpayer’s transactions that have been properly implemented and are capable of being effective according to their terms to accomplish legitimate commercial and legal objectives. Furthermore, in the financial area, the Act respects the basic formal distinction between debt and equity, and applies debt-like rules to equity only by subjecting certain distributions to the same treatment as interest without recharacterizing them. Moreover, the Act is replete with rules to impose consideration for financial accommodations on a deemed basis to limit benefits in situations of special relationships, and to police constructive distributions effected by way of lending transactions. Arguably, the context of the Act suggests that it should be an unusual case in which transactions, and in particular financial transactions, are recast in a transfer pricing analysis.
It is understood that government authorities are aware of these concerns in the financial area and have no present intention to change their assessing practices to extend the application of the transfer pricing rules to inter–“foreign affiliate” financial dealings or to financial accommodations in favor of their Canadian corporate parents by “foreign affiliates.” The same outcome is generally expected where Canadian corporate parents or their “foreign affiliates” invest in preference shares of “foreign affiliates” rather than lend them money provided that offensive planning of a sort to which section 245 of the Act, the “general anti-avoidance rule” (GAAR), would apply is not present; in that event, while section 247 would be an assessment possibility, the normal recourse of revenue authorities would be to section 245 of the Act.
2. Tax Benefit and Transactional Recharacterization. The second main area of concern in subsection 247(2) is its adoption of a “tax benefit” test to determine whether the “quantum or nature” of amounts should be adjusted by substituting an arm’s length transaction that achieves the same effect for that undertaken by the parties. This feature, which is a modification of the provision originally proposed in September 1997, was introduced to allay a concern that transactions, as a matter of course, could or would be recharacterized in the absence of tax-based manipulation that affects the amount of reported income by Canadian taxpayers. In changing the proposed rule, the revenue and finance authorities have recognized that many intra-group transactions in fact would not occur but for the existence of the group, but that this, in itself, normally is an insufficient basis for insisting on transfer pricing adjustments grounded other than in the actual transactions undertaken by the parties.
The notion of “tax benefit” adopted here is exactly the same as that used in the general anti-avoidance rule in section 245 of the Act, without any specific statutory qualification on its application to change the tax outcome of transactions based on the existence of an abuse of the Act or a misuse of its provisions.(35) Here, and in section 245, a “tax benefit” is “a reduction, avoidance, or deferral of tax or other amount payable under this Act or an increase in a refund of tax or other amount under this Act.” In section 245, however, a transaction undertaken primarily to secure a “tax benefit” will not result in adjustments if the result is not contextually offensive — it does not entail an abuse of the Act as a whole or a misuse of any of its provisions.
Is there a material difference in the ultimate significance of “tax bereft”? Is the presence of “tax benefit” without an explicit contextual or other limit in section 247 on its potential scope a “strict liability offense” in a transfer pricing setting? The answer is “probably not.” Leaving aside important and unresolved questions about whether the “abuse” and “misuse” qualifications in section 245 apply as conditions to the original application of the general anti-avoidance rule or, alternatively, excuse its application, there is only one material outcome in a transfer pricing context of securing a tax benefit — namely, a reduction in reported income below what the arm’s length standard indicates is appropriate, evaluated with reference to a variety of well-accepted tests or methods of analysis. If a transaction would not have been undertaken but for the relationship between the parties and does not have a primary commercial motivation or otherwise is somehow unnatural commercially or at variance with how the parties to it in fact are conducting themselves, it is difficult to see how or why the resulting income and tax reductions would be sustained or considered acceptable with reference to the context supplied by the Act. Transfer pricing rules target the most fundamental notion in the Act — the accurate computation of income relative to how it was earned. There is no context in the Act that justifies distortion caused by transactions that arm’s length parties would not implement and that are targeted at distorting income (and resulting tax).
In effect, the “abuse” or “misuse” qualification on the effect of “tax benefit” in subsection 247(2) is supplied by the internal characteristics and inherent function of section 247.
The formal validity of transactions will not be disturbed where they are of the sort otherwise respected by the Act (even if arm’s length parties would not have adopted them in their arrangements) and are not undertaken to manipulate reported income according the arm’s length standard. This is consistent with an aspect of transfer pricing rules, generally, as a specific form of anti-avoidance proscription. That transactional recharacterization is excused by the absence of a primary tax benefit, however, does not mean that the other provisions of section 247 do not apply. If a transaction (or series) would not have been undertaken (making the “tax benefit” issue pertinent), it is difficult to see how the primary basis of an adjustment — transactional terms or conditions at variance with arm’s length dealing — would not be present. In making that evaluation, the transaction as cast by the parties is respected and the returns tested in the typical manner.
This view of the significance of “tax benefit” in section 247 is an added basis for not recharacterizing financial transactions according to the similarity of their economic characteristics or outcomes to other legal formulations. For example, a routine capitalization of a corporation by debt rather than equity does not, by itself suggest “tax benefit” even though the tax consequences of the two means of capitalization are different A specific reservation on recharacterization with reference to the Act is not required for such a fundamental distinction. As previously noted, the Act does not equate the treatment of debt and equity by recharacterization but rather by altering the extent to which equity returns bear tax. Accordingly, the taxpayer’s capital arrangements ought not to be disturbed unless it can be shown that the purpose was to manipulate the recognition of income by a Canadian taxpayer to an extent antithetical to the overall objective of the Act’s transfer pricing rules. At their core, transfer pricing restrictions are concerned with ensuring the correct amount of reported income in light of a taxpayer’s economic effort. They incorporate or reflect their own analytical expectations and are the relevant context of the Act for determinations such as this. This is one area, however, where refinement of the administrative commentary in the Draft Circular may be welcome.
Despite the compelling force of the analyses that would confine the breadth of the recharacterization aspect of subsection 247(2), the existence of this authority likely will continue to engender concern. The specific contemplation of transactional recharacterization does seem to extend the assessment tools of the revenue authorities. Moreover, the adoption in subsections 247(2) of aspects of the statutory expression of the general anti-avoidance rule — itself not well understood — without its “misuse” and “abuse” qualifications on which many rely as the main reason for not being concerned about it, has fueled an unreasonable concern about unchecked transactional recharacterization. In the transfer pricing area, recognizing the fundamentally transactional premise of the pertinent rules, some degree of understanding about the factual basis on which those rules will be applied is necessary. To the extent that the recharacterization aspect of subsection 247(2) may be incompletely circumscribed by its terms and pricing considerations generally, some interpretative guidance, perhaps with examples, in the Draft Circular would be helpful. It is important that a seeming difference in the use, in different places in the Act, of essentially similar tax avoidance concepts not provide the basis for an administrative or judicial construction of subsection 247(2) that does not adequately temper the application of the recharacterization rule.
3. Functional Analysis and Economic Context. Even if there are limits on the extent to which subsection 247(2) permits notional transactions to be substituted for those undertaken by taxpayers, the capacity of the basic adjustment rule to sustain a functional analysis of taxpayers’ transactions needs to be considered. Increasingly, an important consideration in many transfer pricing analyses is the extent to which seemingly simple transactions reflect a variety of embedded transfers each of which must be evaluated according to applicable transfer pricing standards. For example, in the consideration paid for a product, a complex of underlying economic transfers may be being compensated. Do these embedded economic transfers have significance for tax purposes as actual transactions? Is the payer, for example, actually paying product price and a royalty? On what basis is this determination to be made? Both the Act, in subsection 247(4), and the Draft Circular reflect Revenue Canada’s view that transactions should be separated and that, in any event, transactions and taxpayers’ circumstances should be evaluated on a functional basis with reference to their economic context. One aspect of this analysis is to detect underlying economic characteristics of transactions that are so significant that they rise to the level of legal transactions for tax purposes.
This is not transactional recharacterization, even though taxpayers may not perceive much of a distinction. In evaluating documentation requirements and engaging in the requisite underlying analysis, however, the identification of separate (embedded) transactions and the separation of transactions from each other are important considerations. Revenue Canada’s guidelines have always maintained the importance of applying transfer pricing analysis on a transactional basis and engaging in a functional evaluation of a taxpayer’s circumstances in this regard.(36) The present re-orientation of Canada’s transfer pricing rules and practices emphasizes the importance of functional and economic analysis.
In the Canadian context, there are two additional reasons to consider this aspect of a transfer pricing review carefully. First, Canada imposes non-resident withholding tax on a variety of payments to non-residents, including interest, and various fees and other periodic payments, subject of course to relief under applicable tax treaties. The disaggregation of economically complex transfers may expose embedded elements to non-resident withholding tax. On the other hand, recent developments in the customs area reflect an increasing inclination of revenue authorities and courts to consider various payments for services and intangible property to be included in product cost and thus to establish value for duty. The methodological techniques employed in transfer pricing and in evaluating transactions from a customs point of view are similar, even though customs determinations of value continue to not be determinative or even particularly significant for transfer pricing purposes.(37) It is interesting to observe, however, the possibility of contrary analytical tendencies of the two kinds of analysis by the same revenue authority dealing with essentially the same circumstances.
4. Strict Application or Reasonableness Standard? The adjustments contemplated by subsection 247(2) occur if transactional terms and conditions depart from those expected of arm’s length parties. As mentioned, subsection 247(10) seems to require positive adjustments to income to be made. In neither provision, however, is there an explicit recognition of the inherent imprecision of transfer pricing determinations, for example, as found in the “reasonable in the circumstances” formulation of subsections 69(2) and (3). An evident concern is whether the result of this legislative approach will be to make subordinate the analytical judgment required in a transfer pricing review by making the adjustments too mechanical.
Probably, the transfer pricing rules are no more mechanical in their operation under the new formulation than the old. Consequently, the same tolerances for imprecision, ranges of possibly acceptable results, and dependence on facts and circumstances remain.(38) A clue in this regard is found in the definition of “arm’s length transfer price.” Although the term is not used in subsection 247(2), the provision does refer to “a” transfer price that would have been appropriate had the parties to a transaction been dealing at arm’s length. Furthermore, the notion of arm’s length as it has been developed for transfer pricing analysis and as it is reflected, for example, in the OECD Guidelines and the Draft Circular, shows considerable sensitivity to implementing transfer pricing rules and practices in a way that sensibly reflects taxpayers’ circumstances. Even though an adjustment may be called for because transactional terms and conditions are not sustainable under the arm’s length standard, there remains considerable latitude, within that notion generally and in applying the relevant methodological analysis, to import an overriding qualification that the results be reasonable in the circumstances.
B. The Penalty
The penalty described in subsection 247(3) applies equally to transactions in the nature of transfers of property and the provision of services, and to “qualifying cost contribution arrangements.” The main characteristics of the penalty have been described, but there are several aspects relating to the calculation of the base that merit additional attention.
The penalty base is essentially the sum, on a transaction-by-transaction basis, of “transfer pricing capital adjustments” and “transfer pricing income adjustments,”(39) less the following:
* The amount of adjustments pertaining to transactions (including “qualifying cost contribution arrangements,” in which a diligence standard for documentation is directly included) in respect of which a taxpayer “made reasonable efforts to determine arm’s length transfer prices or arm’s length allocations”;(40) and
* Adjustments in favor of the taxpayer under subsection 247(2) (“transfer pricing capital setoff adjustments” and “transfer pricing income setoff adjustments”).(41)
The penalty applies if the base — the net adjustment amount — exceeds the lesser of (i) 10 percent of the unadjusted gross revenue of the taxpayer for the relevant year and (ii) $5 million.(42) In short, the base is the cumulative result of all transfer pricing adjustments against and in favor of the taxpayer if this result is greater than $5 million (or a lesser amount if the taxpayer’s unadjusted gross revenue is less than $50 million) for the year.
Significantly, nowhere in section 247 are positive standards set out to inform what are “reasonable efforts” in determining transfer pricing. This provision does use the phrase “arm’s length transfer price,” which together with the considerations discussed above in relation to the adjustment rule itself allow for some reasonable judgment. Nevertheless, the requirements, notably in relation to documentation and analysis, are open-ended. It is clear, under subsections 247(3) and 247(4), that a taxpayer will be deemed not to have made a “reasonable effort” if the documentation and analysis contemplated there are not available in the required time. But what is “reasonable” in a positive sense is not specified.
This is understood to be deliberate. There is a strong reservation on the part of the revenue authorities to being tied to or confined by prescriptive analytical standards when, in principle, what the analysis and reporting are designed to accomplish is an identification of the taxpayer’s relevant facts and circumstances which, as Revenue Canada has always acknowledged, the taxpayer is bestplaced to know.(43) The discussion of required documentation in the Draft Circular clearly anticipates that the degree and composition of the required transfer pricing analysis will depend on each taxpayer’s circumstances. Presumably, the complexity of the affected transactions and the relationship of those transactions to other arrangements between the transacting parties will be relevant. This will require considerable strategic planning and judgment by taxpayers. It also is a useful indication, implicitly, that the adjustments contemplated by subsection 247(2) are not as mechanical as they otherwise may seem. In principle, the new reporting obligation is no different from the more general statutory and administrative expectations of them under the old rules. As with the careful consideration and preparation of documentation generally, the latitude afforded taxpayers provides a planning opportunity. Taxpayers who understand that pricing must be explained from a rational business point of view(44) are more likely to avoid severe pricing adjustments arising from misunderstandings of their circumstances by revenue authorities than are those who regard these rules essentially as intruding on their commercial privacy.
A second aspect of the base computation may at first seem curious. Adjustments are excused from penalty if the underlying pricing determination reflects taxpayer diligence. In other words, only amounts that do not reflect taxpayer diligence are included in the penalty base. In contrast, the offset for positive adjustments in the taxpayer’s favor is available only where, in respect of transactions pertinent to them, the taxpayer has made reasonable efforts to determine transfer prices. Is it reasonable to expect that a taxpayer who fails the diligence required with respect to transactions that generate positive adjustments nevertheless will have been diligent for other transactions? Does this make sense? The answer is yes, since the offset is not necessary if the taxpayer has been diligent generally in determining prices. It would be odd if the taxpayer were relieved of penalty if its failure to make reasonable efforts to determine prices is comprehensive. The penalty is something to be imposed only where there are clear and severe reporting deficiencies. Effectively, there is an implicit assumption that there will be some attempt at compliance with the documentation requirements. Furthermore, section 247 both presumes and is meant to encourage a constructive response by taxpayers to explaining their pricing. In result, the only reasonable case in which an offset should be available is where reasonable pricing efforts are evident.
A persistently difficult and unresolved question in Canadian transfer pricing is whether some sort of annual or aggregative analysis to transfer pricing can apply, or whether transactional analysis necessarily prevails. A related issue is how narrowly transactions should be described — whether transactions with similar characteristics or related products with similar characteristics or markets should be grouped.
Section 247 essentially propounds a transactional rule. Curiously, however, there is an explicit acknowledgment of the possible relevance of a more global, profit-oriented approach in the reference in subsection 247(3) to “arm’s length allocations.”(45) “Arm’s length allocation” is itself a curious notion. It “means, in respect of a transaction, an allocation of profit or loss that would have occurred between the participants in the transaction if they had been dealing at arm’s length with each other.” “Profit” and “loss” are annual net determinations under the Act. Therefore, at least in evaluating whether a taxpayer has been diligent in pricing determinations, global profit-oriented analyses are specifically recognized legislatively, notwithstanding the guarded acceptance of such approaches in the revised administrative practice released contemporaneously with the new rules. Does this provide an avenue for some sort of aggregation or grouping analysis? Enthusiam for such a result must be tempered by another curiosity. Although generally reflective of net income concepts, the definition is expressed with respect to a transaction. What does profit or loss in relation to a transaction mean? By what mechanism would such a determination be made? Does this simply attempt to accommodate the profit-split methods and the TNMM given qualified recognition administratively? They are, in Revenue Canada’s mind, transactional concepts that, with the benefit of guidance by the OECD in this area, may require some aggregation of transactions with similar characteristics. This is another area where some refinement of Revenue Canada’s administrative statements in the Draft Circular may shed more light.
Another concern is the absence of any obligation on Revenue Canada to make positive adjustments from a taxpayer’s point of view. While it would seem unreasonable to assume that these will not be made in appropriate circumstances, this possibility could affect the utility of the setoffs incorporated in the penalty calculation. A more satisfying interpretation is that the revision of the original September 11,1997, proposal introduced the setoff aspects of the penalty calculation to ensure consistency between the overall adjustment to income arising from the cumulative effect of applying subsection 247(2) to all of a taxpayer’s transactions, and the penalty base (which, in the original version, lacked the same cumulative effect). Further, a reluctance on the part of the revenue authorities to adjust income in a taxpayer’s favor, except possibly where there are other considerations relating to the desirability of making or the ability to make corresponding adjustments as contemplated by Article IX of Canada’s tax treaties, presumes a degree of unfairness in the application of these rules by the revenue authorities that is not justified.
C. Documentation and Analysis
Subsection 247(4) enacts itemized recordkeeping, documentation, and analytical standards for transfer pricing. Neither it nor the penalty rule in subsection 247(3) establishes positive standards in this regard. Compliance with the documentary expectations set out in subsection 247(4) will result in a taxpayer’s not being presumed to have failed to make reasonable efforts to determine transfer prices according the arm’s length standard. Even so, the information anticipated to be relevant to a transfer pricing analysis as set out in this provision is implicitly instructive of what will be required in the positive sense.
The information required under subsection 247(4) is subject to two fundamental constraints, one temporal and the other qualitative in a analytical sense. First, there is a formative obligation imposed to “make or obtain” the required information by the taxpayer’s “documentation due date” for the relevant taxation year or fiscal period; for a person, this is the “filing due date” for tax returns and for a partnership, the date on which a partnership information return in required to be filed under section 229 of the Income Tax Regulations (or would be required to be filed if section 229 applied). This is the “contemporaneous” aspect of the rule. The assembled information, however, is not filed with returns but is to be retained by taxpayers. Upon a request by the revenue authorities, the information must be provided to the revenue authorities within three months. Moreover, material changes to information “made or obtained” as set out in subsection 247(4) must also be documented according to the same standards and limits as apply to the original information.
The overriding qualitative limitation is a requirement that the information be “complete and accurate in all material respects.” This may seem to be a difficult standard to achieve in the absence of any positive or prescriptive indication of what information is required and what “complete and accurate” and “material” mean. Some general guidance is provided in the draft Information Circular, which directly reflects the OECD Guidelines:
The documentation required in order for the taxpayer
to be considered to have made reasonable efforts to
determine an arm’s length transfer price or allocation
in respect of a transaction is that which the taxpayer
would have prepared or obtained, in respect of the
transaction, pursuant to principles of prudent
business management. A prudent businessperson
would attempt to weigh the significance of the
transactions in terms of its business with the
additional administrative costs required to prepare or
obtain such documentation. Therefore, the
obligation to find comparable transactions for
purposes of applying the arm’s length principle is
not an absolute one, but may take into account the
cost and likelihood of finding such comparable
relative to the significance of the transactions to the
taxpayer. On the other hand, the obligation to
comply with the arm’s length principle and the
documentation requirements should be taken into
account by a taxpayer when establishing internal
procedures and policies to document its
This statement of Revenue Canada’s documentation requirements is revealing in several respects. First, it reflects an expectation that a taxpayer will evaluate its transfer pricing at least contemporaneously with other business planning in advance of a transaction. Second, there is a suggestion that strategic or practical decisions about the availability, utility, and necessity of information will be judged by the revenue authorities with sensitivity to the main characteristics of affected transactions and to a variety of considerations that ultimately are concerned with reasonable, practical limits on a taxpayer’s resources. This, too, implies an inherent “circumstantial reasonableness” criterion underlying the adjustment rule in subsection 247(2).
The kinds of descriptive records or documents that taxpayers are expected to assemble are not very surprising and, for the most part, are probably not contentious. They are:
1. A description of the property or services to which a transaction relates.
2. A description of the terms and conditions of the transaction and the relationship of those terms and conditions, if any, to the terms and conditions of each other transaction entered into between the participants in the subject transaction.
3. The identity of the participants in the transaction and their relationship to each other at the time the transaction was entered into.
4. A description of the functions performed, the property used or contributed and the risks assumed, in respect of the transaction, by the participants in the transaction.
5. A description of the data and methods considered and the analysis performed to determine the transfer prices or the allocations of profits of losses or contributions to costs, as the case may be, in respect of the transaction.
6. A description of the assumptions, strategies, and policies, if any, that influenced the determination of the transfer prices or the allocations of profits or losses or contributions to costs, as the case may be, in respect of the transition.
In broad terms this list establishes the framework for a full functional and, in some respects, economic (which is not the same as a functional) analysis of a taxpayer’s circumstances on a transaction-by-transaction basis. The first three categories are not surprising, though some care may be required in relating terms and conditions of transactions to those of other transactions and not merely evaluating the interrelationships among transactions more generally. The fourth category is the essence of the new reporting standard, and reflects the kind of analysis that Revenue Canada already expects in an APA-defined setting. Essentially this is a description of how a taxpayer earned its income, with reference to particular transactions and the relationship of its involvement in an integrated or combined enterprise to the involvement’s of others in its corporate group.
The most contentious areas may be the last two. It should be expected that data underlying a comparable or other transfer pricing analysis should be available to explain the determination. The reference to methods, in the fifth category, is somewhat more problematic. Does a requirement to explain a transfer pricing determination with reference to the “methods considered” require taxpayers to consider more than one method and to discard all but the most appropriate — in effect a “best method” rule similar to that in the United States? Although there is no obligation to apply more than one method to explain pricing, and selection of the most apt method originally may represent substantial compliance, there are suggestions in the Draft Circular that at least in some cases methodological comparisons may be required. Certainly, where a taxpayer considers more than one method, it is expected to record and document this part of its analysis. Moreover, the Draft Circular requires the application of a recommended method and recognizes that the analysis required by one method may be significantly different from that required for others.(47)
Finally, the sixth category refers to “assumptions, strategies, and policies” that influenced a pricing determination. Reasonable questions arise about whether these essentially are general descriptions of objective market and related factors that affect how pricing determinations ought to be evaluated, particularly where departures from a strict application of the arm’s length standard may be justified for a period of time, or whether they reach further into the corporate mind to detect subjective planning motivation. In some respects, the three terms are words of art, and general comments in the OECD Guidelines may help to confine these references principally to objective commercial, market, and business factors affecting a taxpayer’s pricing. “Strategies” are explained in the OECD Guidelines to include a variety of factors essentially concerned with business dynamics.(48) For example, this term encompasses “new product development, degree of diversification, risk aversion, assessment of political changes input of existing and planned labor laws… [and] could include market penetration schemes.” Considerations such as these are relevant in evaluating whether, from time to time, a strict adherence to a transactional arm’s length result will be required. “Policies” may refer, as in the OECD Guidelines, to government policies.(49) And “assumptions” presumably relate to a description of the economic environment in which a taxpayer operates, as well as other matters that could affect the compilation of relevant data over multiple years for comparison purposes, the utilization or other effects of losses with reference to pricing and the significance of deliberate setoffs, to result in overall arm’s length income that in transactional terms departs from the arm’s length principle.(50)
There is no doubt that generally, in the early stages of becoming accustomed to documenting transfer pricing under the new rules, the required analysis may be burdensome. Moreover, taxpayers and their advisers will need to be more astute than ever to ensure that information is kept confidential as provided in section 241 of the Act. In this regard, taxpayer information may lose its protected status if another taxpayer successfully shows that the information was used by Revenue Canada to raise an assessment against that taxpayer. Then, once legal proceedings have commenced, subsection 241(3) allows the information to be released by Revenue Canada. Taxpayers may consider a variety of steps to mitigate this possibility, for example, by ensuring that Revenue Canada is aware of the sensitive nature of information provided to it voluntarily. Although not fail-safe, information delivered voluntarily to Revenue Canada under cover of express advice of its sensitive (e.g., proprietary) nature may benefit from judicial refinements of the scope of subsection 241(3) that would limit its accessibility to third parties.
In the main, the answers to the questions about satisfying these new reporting standards are found in the objectives of transfer pricing analysis and informed by the historical expectations of Revenue Canada, notably in connection with APAs. Difficult judgments may be required, but essentially what is expected is a reasonable, good faith explanation of pricing practices in light of the taxpayer’s characteristics and economic context. After the analysis has been done once, it should be less onerous to repeat.
Context Is Important –Transfer Pricing Generally
Considerable concern has been expressed about certain aspects of the new rules, including the authority accorded Revenue Canada to redetermine the nature of transactions out of step with typical arm’s length transactions. Also attracting attention is the expectation that taxpayers be prepared to explain and justify — even base — their transfer pricing with reference to documented functional analysis and a record of how pricing was established with reference to relevant data, the application of transfer pricing methodology, and corporate strategies and policies.
There is a danger that the new rules will be subjected to didactic, even scientific scrutiny that is interesting, but that does not reflect the ultimately practical and factual orientation of rules and indeed detracts from the attention that usefully could be paid to positive aspects or implications of the rules. In fact, focusing on legislative deficiencies, rather than the objectives underlying the legislation from both substantive income determination and reporting perspectives, probably obscures a constructive evaluation of their real significance and fuels an unreasonable degree of concern about their scope and the manner in which they will be applied by the revenue authorities. It would not be surprising if much of the anxiety about the perceived scope of these rules, the anticipated effort required to comply with the “contemporaneous documentation” requirement, and the specter of extraordinary penalties for deficient compliance has been fueled by memories of a similar debate that took place in the United States when its version of rules of this nature was introduced. It is, indeed, rather easy and almost sporting to parse these rules for interpretative or drafting uncertainties or seeming deficiencies by ignoring their internationally recognized context, which tends to accord less significance to the prescriptive rules themselves and, generally, more to the objectives that they serve and the transactional pricing and income measurement and allocation tests for which legislative rules are more a framework than a manual and that remain essentially unchanged despite the legislative reformulation. Transfer pricing is not scientific, as assisted as it may be by the application of various kinds of functional and economic analysis.
At its core, transfer pricing requires nothing more than the demonstration by a taxpayer that its reported tax and financial income reflects a meaningful and demonstrable relationship to its economic income where the relevant productive enterprise involves other members of a commonly controlled group. In principle, this is simply a particularization, but not an extension, of the expectation in a tax statute that a taxpayer accurately calculate and report its income, and be prepared to offer an explanation in this regard informatively documented in its financial records.(51) In short, the new transfer pricing rules and practices, as have their precursors, encompass the not-surprising expectation that a taxpayer be prepared to explain in a deliberate, organized, and coherent way how its income was earned and why its reported income is consistent with its characteristics and contribution — in relation to its activities and responsibilities in a tax jurisdiction — to a functionally integrated multinational enterprise, relative to other members of its multinational group.
Consistently, in its assessing guidelines,(52) the former Information Circular 87-2, and its more recent guidelines and practices with respect to APA,(53) Revenue Canada has required that Canadian taxpayers analyze, explain, and document their pricing practices in ways that are more or less the same as are anticipated in more specific and contemporary terms by the restated legislative requirements. The practical focus of Canada’s transfer pricing rules has always been on the accuracy of reported income with reference to a taxpayer’s organizational and functional characteristics in the economic context in which it transacts business, with general vestigial reference to the so-called arm’s length principle.
Quite simply, the revenue authorities have a legitimate interest in understanding taxpayers’ businesses in their industrial and economic contexts, and taxpayers have an equivalent and in part strategic interest in ensuring that reporting and related compliance efforts mitigate the possibility of difficult audit issues arising and not being resolved satisfactorily. In this regard, Revenue Canada has acknowledged that taxpayers generally are in the best position to explain how they earn income, and indeed has been prepared to defer to taxpayers’ explanations of their transfer pricing practices where they make sense, reflect underlying planning and analysis, and evidently are not merely designed to manipulate the arbitrary placement of multinational income. Taxpayers who have understood this and with their advisers applied relevant rules and practices as an opportunity to “tell their story” first have tended to fare better in this area than those who regard them as an intrusion on their commercial privacy. Indeed, as far as information gathering is concerned, the considerable ability of revenue authorities, generally, to share information through mechanisms incorporated in tax treaties and by other less formal means reduces the likelihood that ultimately salient information concerning a taxpayer’s circumstances will not be taken into account in evaluating its transfer pricing. A failure by a taxpayer to assist in–in fact to take the opportunity effectively to control–obtaining and organizing the presentation of this material in order to explain its pricing practices will increase the risk that revenue authorities will form incorrect but nevertheless persistent misimpressions about the taxpayer’s circumstances. This is not to suggest that somehow taxpayers ought to defer to revenue authorities’ views on how their businesses ought to be conducted. To the contrary, all–even the revenue authorities–acknowledge that acceptable transfer prices generally fall somewhere on a continuum of acceptable results; there is no “right” transfer price. The thrust of transfer pricing rules is to establish the parameters of acceptable pricing and consequently the reasonableness of reported income in the face of this imprecision.
The new rules retain the same substantive significance as those that they update. They do reflect, however, the common contemporary expectation among countries’ revenue authorities that taxpayers should apply transfer pricing analytical procedures and comply with tax reporting obligations much more diligently than may have been the case, and will be penalized if they fail to respond. If anything, this is the main thrust of the new rules, whatever their interesting but ultimately not very important legislative quirks and curiosities might be perceived to be. On this basis, the most significant effect of the new rules is to convey and reinforce the strategic importance of deliberate and considered transfer pricing planning and careful preparation to explain pricing practices, with the benefit of pertinent documentation, as elements of the development of prices. The Canadian rules may thus differ somewhat from their U.S. counterparts where the emphasis seems to be on justifying pricing after the fact. Not surprisingly, the procedures or methods of transfer pricing analysis, and the guidelines for their application remain of pre-eminent significance whatever the prevailing legislative statement of the rules may be.
Historically, the transfer pricing rules in the Act have been less important than statements by Revenue Canada about how it applies and expects taxpayers to apply analytical and reporting standards and guidelines associated with transfer pricing to ensure an accurate correspondence between taxpayers’ economic and tax income. Despite the present legislative restatement of the arm’s length principle, the essence of Canada’s transfer pricing regime will remain in the Information Circular, which adopts the OECD’s guidelines by reference generally and restates or paraphrases Revenue Canada’s view of them in particular transactional settings. It is expected that considerable attention will be devoted to refining the Information Circular in order that it is and remains informative of how the legislative rules will in practice be applied.
To What End? Some Concluding Comments
The revised transfer pricing legislation and Draft Circular do not change the substantive rules and practices applicable to transfer pricing, apart from according long overdue recognition to the reality of profit-based analyses. They do, however, reflect the increasing intensity with which revenue authorities generally are examining reported income with reference to its underlying functional and economic determinants. The new rules establish the basis for a reporting framework that is contemporary and coherent in terms of the OECD’s transfer pricing guidelines and the evolving transfer pricing standards of many of Canada’s trading partners. Despite the apparent breadth of the legislative formulation of the arm’s length principle and the expanded administrative practice set out in the Draft Circular, the focus of the new rules is on income reporting and analysis of reported income. Taxpayers will be expected to establish a reasonable correspondence between their tax, or financial, and economic income with reference to their entire commercial, organizational, and economic circumstances pertinent to affected transactions.
Complying with the documentation and analytical expectations underlying the new rules will require taxpayers to adopt a strategic approach to tax reporting. Prudent taxpayers will devote considerable attention to identifying the “transactions” that could be affected–that is to say, how broadly or narrowly they are described or evaluating their relative simplicity or complexity from an economic point of view and considering the proper methodological analysis to determine and support pricing practices. Furthermore, aspects of the functional and economic complexity of a transactional context (including the allocation of risk for such things as currency, product liability, financing, various market factors, research and development), as well as the significance of intangible property in property transfers that may be embedded in property transfers or the provision of services. will require attention. Given the substantial orientation of the rules and the analytical methodologies to explaining pricing in terms of comparable circumstances, even in applying profit-based “methods,” careful consideration will be required in assessing whether and to what extent relevant “comparables” exist.
Initially, taxpayers will need to develop a plan for ongoing compliance with the reporting requirements contemplated by the new rules. In view of uncertainties inherent in the nature of the analysis and the potential scope of section 247, taxpayers may wish to consider the APA procedure as an opportunity to reach agreement contractually with the revenue authorities on analytical and documentation standards and requirements. Indeed, the potential attractiveness of the APA program in this regard is implicitly acknowledged by an observation in the Draft Circular that transfer pricing adjustments under subsection 247(2), and consequently attendant penalties, are unlikely in the face of an APA.(54)
(1) Organisation for Economic Co-Operation and Development Transfer Pricing Guidelines for Multinational Enterprises and Tax Administration (1995, 1996, 1997).
(2) Revenue Canada Information Circular 87-2, paragraph 9; OECD, Transfer Pricing and Multinational Enterprises (1979); OECD, Transfer Pricing and Multinational Enterprises Three Taxation Issues (1984).
(3) References to the new rules are directly to the legislative changes, as distinguished from the administrative practice of Revenue Canada addressed in draft Information Circular 87-2R. References throughout to section 247 are to that section of the Income Tax Act as it will be amended.
(4) Notice, section 238(2).
(5) Notice, section 238(3).
(6) Section 247(2)
(7) Section 247(3).
(8) Section 247(4)
(9) Section 247(2)(a).
(10) Section 247(2)(b).
(11) This aspect of the new rules is controversial, largely because of its incorporation of concepts from the “general anti-avoidance rule” in section 245 of the Act. Although the implications of this apparently broad recharacterization authority are discussed in the text that follows, it is worth noting that even the OECD’s 1979 Guidelines, in paragraphs 23, 183 to 191 and 198 to 201, contemplated that taxpayers’ transactions might be recast as part of a transfer pricing determination; the present OECD Guidelines, in paragraphs 1.36 to 1.41, are to similar but more descriptive effect, and are mirrored in various provisions of the Draft Circular, notably in paragraphs 18 to 20. The recharacterization possibility is perhaps more prominent in contemporary transfer pricing rules, but it should not be considered any more likely than before that actual transactions will be ignored in deference to notional or substituted transactions of equivalent economic effect. It may, however, be important to distinguish between recharacterization, a dramatic and relatively unlikely occurrence, and a functional or economic disaggregation of composite or complex transactions to determine their inherent legal and commercial characteristics. This, too, has always been a possibility under historical analytical and reporting principles, but in practice may be more likely in view of the emphasis now placed on a functional analysis of a taxpayer’s transactions in their economic context.
(12) Section 247(1).
(13) Section 247(1).
(14) Draft Circular, paragraphs 21-27.
(15) Draft Circular, paragraph 6. “The Department supports and proposes to follow the OECD Guidelines. The contents of this Circular reflect the guidance provided by the OECD Guidelines. Readers should consult the OECD Guidelines for a more detailed discussion of the principles contained in Parts 2 to 6 of this Circular.” There is not, at this stage of the Draft Circular’s development, a perfect correspondence between the OECD Guidelines and the summaries or paraphrases of them in the Draft Circular, even thought the Draft Circular clearly is meant to reflect these Guidelines. It is expected that uncertainties will be resolved before the Draft Circular is released in its final edition.
(16) Draft Circular, paragraph 28.
(17) Draft Circular, paragraphs 37-46.
(18) OECD Guidelines, Chapter III, in particular paragraphs 3.2, 3.5, 3.15-3.20, 3.25, 3.26-3.29, 3.49 and 3.54.
(19) Revenue Canada discusses its views on profit techniques in paragraphs 37 to 46 of the Draft Circular. Generally, Revenue Canada’s reluctance to entertain profit, and in particular profit-comparison, methods owes itself to doubt about the comparability of data and circumstances needed for such an analysis and the difficulty in reliably taking into account non-transactional differences among taxpayers. This disinclination to accord much formal significance to profit-based analyses was foreshadowed by Revenue Canada’s criticism (which in principle persists) of the U.S. CPM. Despite reluctance to endorse the U.S. version of the TNMM and to employ profit methods generally, Revenue Canada historically has been prepared to entertain profit splits in appropriate circumstances and has acknowledged the potential utility of profit methods–even comparisons–at least as secondary tests of the results reached using the traditional methodologies. Furthermore, there have been some recent indications of a relaxation of views about the utility of the CPM, though taxpayers would be well advised to base and express the outcome of a transfer pricing analysis on the application of a traditional transactional method. Revenue Canada’s criticism of the CPM and the TNMM may have practical implications for how multinationals with U.S. connections prepare documentation for use in U.S. and Canadian tax reporting. It is common in the United States to limit exposure to penalties under the Internal Revenue Code by obtaining transfer pricing studies that frequently apply the CPM. Revenue Canada’s treatment of profit-based methods as subordinate to the traditional transactional methods and its persistent concern about the adequacy of the CPM heightens a concern that compliance with U.S. documentation standards alone may not be sufficient from a Canadian perspective.
(20) OECD, Model Tax Convention On Income and Capital (1995, as updated).
(21) Article IX, which is consistent both with the “associated enterprises” articles of Canada’s tax treaties generally and with the updated notion of the arm’s length principle incorporated in the restated Canadian transfer pricing rules provides, as summarized in paragraph 1.6 of the OECD Guidelines, that .”..[when] conditions are made or imposed between . . . two [associated] 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.” Interestingly, when the transactional arrangements of taxpayers were more closely concerned with transfers of simple goods, it would have been reasonable to expect a relatively close correspondence between transactional revenue and profit. With the increasing significance of intangible and financial property, notably in embedded intra-group transfers, this expectation is more difficult to sustain. The orientation of all these rules, however, is “profits” and not “price” and the concern of transfer pricing is ultimately the adequacy of reported income in relation to economic activity taking place in a jurisdiction regardless of what financial accounting principles applied to accounting income of legal entities might otherwise suggest. The OECD Guidelines acknowledge “profits” of members of an associated group as the reference point for the application of the comparability analysis for which the analytical methodologies are guidelines and tools. Even though the balance of the OECD Guidelines is expressed with respect to pricing determinations referable to a transaction, this “profit” focus provides an important indication of what the analytical objectives are and the tax policy and reporting outcomes should be.
(22) The Act contains a number of other provisions providing the revenue authorities with the ability to obtain information. The most notable among them is section 231.6, which allows the revenue authorities to require the delivery to them by Canadian resident taxpayers and non-residents carrying on business in Canada, upon written notice, of any “foreign” based information or document”–“any information or document that is available or located outside Canada . . . [that] may be relevant to the administration or enforcement of the Act.” Failure to comply with a reasonable requirement determined according to procedures set out in this provision will, on request of the authorities, result in that information not being available to assist the affected taxpayer in a proceeding under the Act.
(23) Draft Circular 87-2R, paragraph 96.
(24) Section 247(6). A member of an upper-tier partnership that itself is a partner of another lower-tier partnership is treated as a partner of the other lower-tier partnership, with a profit or loss share equal to the partner’s direct or indirect entitlement in that second partnership.
(25) OECD Guidelines, Chapter VIII (August 1997).
(26) Draft Circular, Part 4.
(27) Draft Circular, paragraphs 49 and 50.
(28) Draft Circular, paragraphs 52 and 53. The OECD Guidelines discuss methods for evaluating the correspondence between contributions and benefits in similar terms. See paragraphs 8.19-8.22.
(29) Draft Circular, paragraphs 54 and 55.
(30) OECD Guidelines, paragraphs 8.6, 8.10, and 8.19, in particular.
(31) Section 247(1)
(32) The provision refers to transactions between a “taxpayer” and a non-resident. Where withholding tax applies under Part XIII of the Act to payments received by a non-resident, the non-resident is the taxpayer. In certain cases, payments by non-residents to other nonresidents are subject to non-resident withholding tax (in particular, subsections 212(13)-(13.2) of the Act, dealing mainly with deductible payments made in respect of debts secured by Canadian real estate or in the course of certain Canadian business activities). In result, section 247 may in some cases apply to arrangements exclusively between non-residents.
(33) Subsections 69(2) and (3) of the Act referred to particular kinds of transactions and the consideration generated by them. Section 247 is general with respect to transactions covered. In addition even though not adopted in subsection 247(2), the definition of “transfer price” in subsection 247(1) refers to an “other payment . . . for property or as consideration for services.” Difficult legal questions relate to whether a financial accommodation is a service and whether in this context compensation for a loan is a payment for property (i.e., whether property, which under the Act includes money, unless circumstances require otherwise, includes money in a transfer pricing setting that generally has confined its application, or at least most commonly has been applied only, to non-financial tangible and intangible property).
(34) Draft Circular, paragraphs 18-20.
(35) Compare Section 245(4) of the Act, noting in particular the opportunity that the GAAR may provide to determine tax consequences based on a recharacterization of the “nature of any payment or amount.”
(36) For example, Information Circular 87-2, paragraphs 10-12, as well as guidelines that apply in connection with advance pricing agreements (APAs).
(37) Draft Circular, paragraph 91. See also paragraphs 1.65-1.67 in the OECD Guidelines.
(38) See OECD Guidelines, Chapter 1, in particular paragraphs 1.30, 1.31-1.35, 1.42-1.44, 1.45-1.48 (arm’s length range) and 1.60-1.64 (intentional set-offs).
(39) Section 247(3)(a)(i).
(40) Section 247(3)(a)(ii).
(41) Section 247(3)(a)(iii).
(42) Section 247(3)(b)
(43) This is evident, for example, in the comments contained in the Draft Circular relating to documentation, partly with deference to the OECD Guidelines. Paragraph 88 of the Draft Circular provides: “Paragraph 5.16 of the OECD Guidelines, which states that because of the variety of business scenarios encountered in practice it is not possible to produce an exhaustive list of the documentation required to support a particular pricing method. Instead only general guidance can be given to assist taxpayers in identifying documentation that provides evidence that has pricing that satisfies the arm’s length principle. Readers should consult the OECD Guidelines for a more detailed discussion of this issue.”
(44) Draft Circular, paragraph 80.
(45) Section 247(1).
(46) Draft Circular, paragraph 80. See also paragraphs 81-85.
(47) Draft Circular, paragraphs 26, 79, 80, 81, and 85.
(48) OECD Guidelines, paragraph 1.27(5).
(49) OECD Guidelines, paragraphs 155-159. In addition, taxpayer policies of various kinds may be relevant. See, e.g., OECD Guidelines, paragraphs 1.52-1.54 (losses) and 1.60-1.64 (intentional set offs).
(50) OECD Guidelines, paragraphs 1.29, 1.30, and 1.42-1.44.
(51) Section 230 of the Act generally requires that a taxpayer maintain records sufficient to reflect how its reported income was earned. Coupled with the analytical expectations that long have been a part of Canadian tax reporting practices, in former Information Circular 87-2, Canadian tax reporting arguably already incorporates the nature of tax reporting that is anticipated more specifically by the new rules and commonly described as “contemporaneous documentation.”
(52) Revenue Canada, TOM No. 14, “Audit Techniques,” section 14(53)0 Intercompany Pricing Guidelines.
(53) Information Circular 94-4, and related “Detailed Procedures and Guidelines.”
(54) Draft Circular, Part 12.
J. SCOTT WILKIE is a tax partner in the Toronto office of Ogilvy Renault. He is also an Adjunct Professor of International Taxation at Osgoode Hall Law School. He received his B.A. degree from McMaster University and his LL.B. degree from the University of Toronto. Mr. Wilkie is Chair of the National Tax Section of the Canadian Bar Association, Co-Chair of the Joint Committee on Taxation of the Canadian Bar Association and the Canadian Institute of Chartered Accountants, a member of the Canadian Branch Council of the International Fiscal Association, and a governor of the Canadian Tax Foundation. He is co-author of a book on transfer pricing entitled Transfer Pricing Strategies in a Global Economy.
COPYRIGHT 1998 Tax Executives Institute, Inc.
COPYRIGHT 2004 Gale Group