Income tax: using arm's length transfer pricing methodologies in international dealings between associated enterprises

This Ruling sets out the principles to be applied in selecting and applying a transfer pricing methodology. It discusses the methodologies acceptable to the Australian Taxation Office ('ATO'), our view on the definitional issues that arise in relation to the various methodologies, and the circumstances in which the various methodologies are considered appropriate. It links the concepts of the arm's length principle, comparability and methodological approaches with a process of analysis that can be used to establish an arm's length outcome. These concepts are important because under Division 13 of Part III of the Income Tax Assessment Act ('the ITAA') the Commissioner of Taxation can adjust the consideration used in the supply or acquisition of property in international agreements where inter-alia the consideration is not an arm's length consideration. Provisions to make adjustments to achieve an arm's length outcome are also contained in each of Australia's double taxation agreements ('DTAs').

This Ruling has a broad scope in order to keep the issues in context and to cover a wide variety of circumstances. This recognises the fact that dealings between associated enterprises involve many different types of property and services including tangible goods, the licensing of intangibles and financial and management services. It needs to be recognised at the outset that the application of the principles set out in this Ruling requires judgment. Transfer pricing issues cannot be easily resolved by the rigid and mechanical application of standardised rules or predetermined formulae.

This Ruling is limited to dealings between separate legal entities.

Definitions of terms used in this Ruling that require explanation are contained in the Definitions section.

This Ruling applies to years commencing both before and after its date of issue. However, this Ruling does not apply to the extent that it conflicts with the terms of a settlement of a dispute agreed to before the date of issue of this Ruling (see paragraphs 21 and 22 of Taxation Ruling TR 92/20).

With the exception of definitions marked '*' the definitions used here are from the 1995 Report of the OECD Committee on Fiscal Affairs, ' Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations ' ('the 1995 OECD Report'). Arm's length principle The international standard that OECD Members have agreed should be used for determining transfer prices for tax purposes. It is set forth in Article 9 of the OECD Model Tax Convention as follows: where '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'. Arm's length range A range of figures that are acceptable for establishing whether the conditions of a controlled transaction are arm's length and that are derived either from applying the same transfer pricing method to multiple comparable data or from applying different transfer pricing methods. Associated enterprises, entities or parties Two enterprises are associated enterprises with respect to each other if one of the enterprises meets the conditions of Article 9, subparagraphs 1 (a) or 1(b) of the OECD Model Tax Convention with respect to the other enterprise. * The expression also includes enterprises which do not meet the conditions of Article 9 but whose dealings can be adjusted under Division 13 of the ITAA. These enterprises may reside in non treaty countries. The consideration used in dealings between uncontrolled enterprises who do not deal at arm's length with one another may also be adjusted in some circumstances and the term 'associated enterprises' is intended to extend to these dealings. TR 94/14 discusses this situation at paragraph 50. Basket* A group or range of items such as transactions, functions, relationships, etc. Channel profits* The profits derived by all independent and associated enterprises from the flow of transactions linking production, distribution and sale to the final customer of the relevant product or services. Combined profits* That part of the channel profits derived by associated enterprises. Comparable uncontrolled price (CUP) method A transfer pricing method that compares the price for property or services transferred in a controlled transaction to the price charged for property or services transferred in a comparable uncontrolled transaction in comparable circumstances. Contribution analysis An analysis used in the profit split method under which the combined profits from controlled transactions are divided between the associated enterprises based upon the relative value of the functions performed (taking into account assets used and risks assumed) by each of the associated enterprises participating in those transactions, supplemented as much as possible by external market data that indicate how independent enterprises would have divided profits in similar circumstances. Controlled enterprises, entities or parties* See ' Associated enterprises, entities or parties ', above. Controlled transactions Transactions between two enterprises that are associated enterprises with respect to each other. Cost plus mark-up A mark-up that is measured by reference to margins computed after the direct and indirect costs incurred by a supplier of property or services in a transaction. Cost plus (CP) method A transfer pricing method using the costs incurred by the supplier of property (or services) in a controlled transaction. An appropriate cost plus mark-up is added to this cost, to make an appropriate profit in light of the functions performed (taking into account assets used and risks assumed) and the market conditions. What is arrived at after adding the cost-plus mark-up to the above costs may be regarded as an arm's length price of the original controlled transaction. Direct costs Costs that are incurred specifically for producing a product or rendering a service, such as the cost of raw materials. Enterprise * An entity organised for commercial purposes. Functional analysis* An analysis of the functions performed assets used and risks assumed by each associated enterprise in controlled transactions as a basis for examining the comparability of dealings by independent enterprises or for developing a view as to the economic significance of the taxpayer's activities. This analysis which can be performed at a series of different levels from a qualitative initial assessment to a more detailed analysis involving economic weightings, is generally referred to as a functional analysis. Global formulary apportionment method A method to allocate the global profits of a multinational enterprise ('MNE') group on a consolidated basis among the associated enterprises in different countries on the basis of a predetermined formula. Gross profits The gross profits from a business are the amount computed by deducting from the gross receipts of the transaction the allocable purchases or production costs of sales with due adjustment for increases or decreases in inventory or stock-in-trade, but without taking account of other expenses. Hurdle rate* The minimum rate of return expected of a long term investment opportunity for it to be adopted. Independent enterprises, entities or parties Two enterprises are independent enterprises with respect to each other if they are not associated enterprises with respect to one another. Indirect costs Costs of producing a product or service which although closely related to the production process may be common to several products or services (e.g., the costs of a repair department that services equipment used to produce different products). Multinational enterprise group (MNE group) A group of associated companies with business establishments in two or more countries. Multinational enterprise (MNE)* An enterprise that is part of an MNE group. Mutual agreement procedure* A procedure provided for in all of Australia's double taxation agreements ('DTAs') through which the ATO, at the behest of a taxpayer or on its own account, consults with other tax administrations to resolve disputes regarding the application of Australia's DTAs. The procedure can be used to eliminate double taxation that could arise from a transfer pricing adjustment. Profit comparison method* A transfer pricing methodology based on comparisons at the net profit level, on a single transaction level or in relation to some aggregation of dealings between associated enterprises, between the taxpayer and independent parties dealing wholly independently in relation to a comparable transaction or dealings. See also ' Transactional net margin method ' below. Profit split method* A transfer pricing method that identifies the combined profit to be split for the associated enterprises from a controlled transaction (or controlled transactions that it is appropriate to aggregate under the principles set out in this Ruling and then splits those profits between the associated enterprises according to 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 between independent parties. Related enterprises, entities or parties* See ' Associated enterprises, entities or parties ', above. Resale price margin A margin representing the amount of which a reseller would seek to cover its selling and other operating expenses and, in the light of the functions performed (taking into account assets used and risks assumed), make an appropriate profit. Resale price (RP) method A transfer pricing method based on the price at which a product that has been purchased from an associated enterprise is resold to an independent enterprise. The resale price is reduced by the resale price margin. What is left after subtracting the resale price margin can be regarded, after adjustment for other costs associated with the purchase of the product (e.g., customs duties), as an arm's length price of the original transfer of property between the associated enterprises. Residual analysis An analysis used in the profit split method which divides the combined profit from the controlled transactions under examination in two stages. In the first stage, each participant is allocated sufficient profit to provide it with a basic return appropriate for the type of transactions in which it is engaged. Ordinarily, this basic return would be determined by reference to the market returns achieved for similar types of transactions by independent entities. Thus, the basis return would generally not account for the return that would be generated by any unique and valuable assets possessed by the participants. In the second stage, any residual profit (or loss) remaining after the first stage division would be allocated among the parties based on an analysis of the facts and circumstances that might indicate how this residual would have been divided between independent enterprises. Traditional transaction methods The comparable uncontrolled price method, the resale price method, and the cost plus method. Transactional net margin method* OECD terminology for a transfer pricing method that examines the net profit margin relative to an appropriate base (e.g., costs, sales, assets) that a taxpayer realises from a controlled transaction (or transactions that it is appropriate to aggregate in accordance with the principles in this Ruling). See also ' Profit comparison method ', above. Transactional profit method* OECD terminology for a transfer pricing method that examines the profits that arise from particular controlled transactions of one or more of the associated enterprises participating in those transactions. The term is limited to the profit split method and the transactional net margin method. Uncontrolled enterprises, entities or parties See ' Independent enterprises, entities or parties ', above. Uncontrolled transactions Transactions between enterprises that are independent enterprises with respect to each other. Unrelated enterprises, entities or parties See ' Independent enterprises, entities or parties ', above. Whole of entity basis* A basis of analysis whereby the business operations of an entity are examined in their entirety rather than segmenting them into transactions or product, service or business lines.

Australia has endorsed the arm's length principle as the general benchmark for transfer pricing. This endorsement is reflected in our DTAs and Division 13 (paragraphs 236 - 241) .

Division 13 and Australia's DTAs do not prescribe any particular methodology or preference for the order in which methodologies might be applied to arrive at an arm's length outcome (paragraph 242) .

Within the context of definition of 'arm's length consideration' in Division 13 and the DTAs notion of 'independent parties dealing wholly independently with one another', Parliament intended to give the Commissioner the greatest possible scope to use methodologies appropriate to the circumstances (paragraph 243) .

This is a conscious decision by legislature which intended the arm's length principle to be the cornerstone of the law yet still allowing the Commissioner flexibility to administer the law especially in relation to the selection and application of methodologies, though clearly limited by the statutory purpose (paragraph 244) .

We agree that the statutory objective provides a guideline and limitation on the methodologies that are available under Division 13 and the DTAs - and in relation to how methodologies should be applied. The statutory objective of Division 13, is to enable the Commissioner to determine an amount of consideration in respect of an international agreement that would have been set if the dealing had occurred on an arm's length basis in cases where - judged against what independent parties dealing at arm's length might reasonably be expected to have received or paid in the taxpayer's circumstances - a taxpayer has received inadequate or no consideration for property (including services) supplied, or pays too much for purchases - and to use the amount so determined in the calculation of the taxpayer's taxable income. It has a further objective of allowing the Commissioner to estimate the arm's length consideration where for any reason (including an insufficiency of information available to the Commissioner) it is not possible or practicable for the Commissioner to ascertain the arm's length consideration (paragraph 245) .

The Associated Enterprise Articles in Australia's DTAs have the objective of allowing the Commissioner to adjust understatements of profits in cases where an enterprise resident in Australia and an enterprise resident in a country with which Australia has a DTA are under common management, control or ownership (whether direct or indirect) and have not dealt with each other on an arm's length basis (paragraph 246) .

Where complex issues arise that require specialist approaches in areas where there is no guidance as to what was intended by way of a method to achieve a stated objective, it is appropriate in determining the relevant principles and criteria, to have regard to the industry practice and any standards relevant to the issue. It would be appropriate in the context of statutory transfer pricing rules to consider the internationally agreed approaches as to what would generally be accepted as a reasonable way to determine the arm's length outcome, having regard to the intent of the transfer pricing provisions in Division 13 and the DTAs and the actual wording of the Australian provisions (paragraph 247) .

When applying Division 13 and the DTAs we will pay close attention to the OECD guidelines on transfer pricing methodologies and the operation of the Associated Enterprises Article of the OECD Model, being the considered view of many tax experts familiar with transfer pricing. However, they are not an interpretation of Division 13 which must be construed according to its terms and purpose. In relation to the application of the DTAs, it needs to be recognised that OECD Reports do not have as high a status in international law as an aid to interpretation as a document evidencing the intention of the Contracting States or the Commentary to the OECD Model Tax Convention. Nevertheless, the ' Report on Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrators ' that was approved by the OECD Council on 13 July 1995 ('the 1995 OECD Report') is seen as an important, influential document that reflects unanimous agreement amongst the member countries - an agreement that was achieved after an extensive process of consultation with industry and tax practitioners in member countries - and should be followed where relevant and in the absence of any intention to the contrary in Division 13 or the DTAs (paragraphs 248 and 249) .

It is accepted that the CUP method will provide the best reflection of an arm's length outcome where there is sufficient reliable data for its application. A flexible approach should be adopted to allow adjusted CUP analyses to be used in appropriate cases (paragraph 250) .

It is also accepted that the Resale Price (RP) and Cost Plus (CP) methods can be used in appropriate cases, subject to the general principles in paragraph 87 of TR 94/14 (paragraph 251) .

As stated in paragraph 100 of TR 94/14, the ATO will accept the use of a mixture, RP and CP methods or a profit split or profit comparison method in certain circumstances (paragraph 252) .

The transfer pricing rules in Division 13 and the profit reallocation rule in all of Australia's DTAs allow the Commissioner to use Division 13 to approximate an arm's length consideration in cases where the information available is inadequate to determine the income to be attributed to an enterprise - provided that when Division 13 is applied in conjunction with a DTA it is applied, so far as it is practicable to do so, consistently with the principles of the relevant Associated Enterprises Article of the DTA (paragraphs 253 and 254) .

It could be argued from the context of its operation and its wording that the Commissioner would be using a non-arm's-length methodology when applying subsection 136AD(4) of the ITAA. Differences in Australia's DTAs from the OECD Model and the enactment of subsection 136AD(4) reflect Parliament's intention to introduce provisions, which, while being as consistent as possible with the arm's length principle, allow the Commissioner to go beyond the strict confines of paragraph 136AA(3) (and the arm's length principle as found in the treaties) to deem an arm's length consideration (paragraph 255) .

However, having regard to the clear policy expressed in subsections 136AD(1) to (4) to use the arm's length principle and the fact that subsection 136AD(4) enables the deeming of the amount of the arm's length consideration which is then used in the application of subsection 136AD(1), (2) or (3) as appropriate, subsection 136AD(4) must be applied in a way that achieves the closest practicable estimate of an arm's length result. (See also paras 82 and 83 and 338 to 340 of TR 94/14) (paragraph 256) .

This does not mean that the arm's length principle is being ignored. Rather, the Australia law is designed to reflect the Australian experience in this area and to give the Commissioner sufficient flexibility to arrive at an answer in the wide range of cases that are likely to be encountered in practice. (See paragraphs 240 and 241 and 253 to 255) (paragraph 257) .

The aim in cases where subsection 136AD(4) applies will be to achieve as highly focused a comparison as is possible in the circumstances, consistent with the intention of Parliament as expressed in Section 136AD as a whole, the Explanatory Memorandum ('EM') introducing this provision into Division 13 and with the Associated Enterprises Articles in all of Australia's DTAs (paragraph 258) .

A similar limitation in terms of statutory purpose applies when the application of subsection 136AD(4) is authorised by a DTA because all of Australia's DTAs require the subsection to be applied, in so far as it is practicable to do so, consistently with the arm's length principle embodied in the Associated Enterprises Article (paragraph 259) .

It follows from the nature and purpose of subsection 136AD(4) that other methodologies may be used under that subsection than would be available under subsection 136AD(1) - (3) and Article 9 (the Associated Enterprises Article) of the OECD Model Tax Convention. These methods will depend on the facts and circumstances of each case but could include income and expense allocation on the basis of a formula, a return an assets method, a mixture of methods, or some form of profit comparison other than the profit split and profit comparison methods described later in this draft Ruling (paragraph 260) .

The selection and applicability of methodologies in the context of Division 13 was discussed in paragraphs 86 to 100 of TR 94/14 and regard should be had to the general principles expressed therein. Those principles are relevant to DTAs and should also be applied in that context (paragraph 261) .

The most appropriate method in a given case will depend on the facts and circumstances of the case and the extent and reliability of data on which to base a comparability analysis, the intention always being to select the method that produces the highest degree of comparability. In cases where there are no comparables or there is an insufficient information to determine the arm's length outcome, the method to be used should be a method that produces a reasonable estimate of an arm's length outcome on the basis of what is known in the case (paragraph 262) .

Having regard to the statutory objective of Australia's transfer pricing rules, the ATO takes the view that any transfer pricing methodology used to calculate an arm's length consideration in international dealings between associated enterprises must be applied in a way that will provide an arm's length outcome by closely reflecting commercial and economic reality and the economic contribution made by the enterprises in each jurisdiction (paragraph 263) .

The ATO does not see this Ruling putting into question internationally accepted methodologies that are used to determine an arm's length consideration. On the contrary the ATO endorses the CUP, RP, CP, profit split and profit comparison methods as the relevant approaches or criteria, the most appropriate of these depending on the nature of the case and the extent of reliable data to enable its proper application Nor is the ATO expressing any preference for particular methodologies since their suitability and reliability will depend on the facts and circumstances of each case. However, an understanding of the commercial and economic reality underlying any particular transaction or dealing will be reached by beginning with a search for and a close examination of comparable transactions or dealings between unrelated enterprises in an application of the traditional arm's length methodologies. If such comparables can be found and the resulting prices or terms would be acceptable to each of the parties concerned then a basis for an arm's length determination may have been reached. If not, then profit methods should be considered. In particular, the ATO agrees with the OECD view that profit methods (sometimes referred to as transactional profit methods) are methods of last resort where there is insufficient data on uncontrolled transactions (possibly because of unco-operative behaviour on the part of the taxpayer relative to these guidelines), or where such data is considered unreliable, or due to the nature of the business situation (paragraph 264) .

The arm's length principle requires a conclusion (and in the case of Division 13, a determination) of what might reasonably be expected if the parties were dealing at arm's length with one another. It does this by comparing what the taxpayer has done with notions of supply, demand and negotiation in an open market and uses the behaviour of independent parties dealing at arm's length as a benchmark. The notion of comparability is therefore central to the arm's length principle (paragraphs 265 and 266) . Implicit in the concept of 'the arm's length principle' and of the expressions 'arm's length consideration' in Division 13 and 'independent parties dealing wholly independently with one another' in Australia's DTAs is the notion that independent parties who are dealing at arm's length would each compare the options realistically available to them and seek to maximise the overall value of their respective entities from the economic resources available to or obtainable by them (paragraph 66 of TR 94/14 and paragraph 267 of this Ruling).

The issue of choice is important, because the question being asked by the arm's length principle is, 'What would have happened if the ownership link had been severed and the enterprise was motivated by its own economic interest?' This approach will involve a consideration of what a reasonable, independent business person might reasonably be expected to agree to in the same or similar circumstances (paragraph 268) .

It would not be expected that a seller would accept less or a buyer pay more than the open market price (bearing in mind that this could be a range of prices). In this sense, the search for comparable uncontrolled prices is a search for an open market based alternative. The cost plus method and the resale price method options can also be seen in this context as defining market based margins for functions performed (including assets used and risks assumed) and it could be argued that a reasonable prudent decision maker would look to open market in assessing the available alternatives. In this way the cost plus and resale price methods are also special cases that derive their validity from the fact that where they can be reliably applied they define reasonable courses of action by an uncontrolled enterprise which can be used as an arm's length benchmark for the taxpayer's dealings. The profit based methods are last resort options that identify reasonable alternatives when the more direct methods based on transactional comparability are not suitable or practicable. While the ATO has a preference for more direct methods based on transactional comparability, we recognise the overriding need to ensure that assessments reflect commercial and economic reality as well as the economic contribution made by each of the enterprises involved (paragraph 269) .

Where an open market exists from which one or more comparables can be inferred, the comparable will determine the options open to the enterprise in relation to how its dealings should be structured for tax purposes to accord with the arm's length principle. While the dealings between enterprises in the same multinational group demonstrate a strong emphasis on strategic long-term relationships based on the exchange of goods and services, finance, technology and know how to suit integrated production or marketing processes, organisational structure and strategies, it should not be automatically assumed that members of multinational groups are not dealing at arm's length with each other (paragraph 270) .

The decision to embark on a relationship is one that is profit driven both in the long and short term, with expectations that will be shaped by the experience of similar enterprises under similar circumstances. Whether or not the choice made is acceptable is best tested by looking to the options that can reasonably be held to exist. The starting point is therefore to consider the profit expectations at the time when a relationship is initiated, and what a reasonable business person would see as the critical assumptions at that time and how the relationships could reasonably be expected to develop should the conditions supporting the critical assumptions materially change (paragraphs 271 to 273) .

There will be situations where direct comparisons are impossible. There are other cases where the transactions are of a kind that would occur only between related parties. The arm's length principle still applies, but the focus is on functions, assets, and risks and the processes that parties dealing at arm's length would have adopted to allocate profits based on benchmark rates of return and economic weightings in comparable circumstances, using the concept of a joint venture between independent parties as a guide (paragraph 274) .

Parliament intended that the Commissioner still be empowered to apply the law in cases involving unique or highly differentiated dealings to ensure that Australia receives its fair share of tax. To that end, subsection 136AD(4) empowers the Commissioner to approximate an arm's length outcome where there is a lack of information that makes the determination of the arm's length consideration not possible or practicable. This residual power is preserved in all of Australia's DTAs (paragraph 275) .

Ideally, the arm's length principle should be applied to each international dealing in order to arrive at an arm's length consideration. However, in some cases this will not be feasible because of the absence of reliable data on which to assess comparability (paragraph 277) .

In practice, there are two fundamental approaches that may be taken to achieving an arm's length outcome. These are the use of analysis based on comparable uncontrolled dealings and other methods where there are no comparable uncontrolled dealings. Analysis based on reliable comparable uncontrolled dealings should be used to determine an arm's length consideration where reliable information is available to do so. There are some circumstances where it will be necessary to basket or aggregate transactions in order to achieve this end. Where there are no comparable uncontrolled dealings or there is insufficient reliable data to identify them, it will be necessary to use other methods to estimate or approximate an arm's length outcome (paragraph 278) .

The preferred arm's length methodologies are based on the concept of comparing the prices / margins achieved by associated enterprises in their dealings to those achieved by independent enterprises for the same or similar dealings. As there are many matters that may influence price / margins there is a need to closely examine the dealings being compared. In order for such comparisons to be useful, the economically relevant characteristics of the situations being compared must be sufficiently comparable. To be comparable means that none of the differences (if any) between the situations being compared could materially affect the condition being examined in the methodology (e.g., price or margin), or that reasonably accurate adjustments can be made to eliminate the effect of any such differences. In determining the degree of comparability, including what adjustments are necessary to establish it, an understanding of how unrelated companies evaluate potential transactions is required. Independent enterprises, when evaluating the terms of a potential transaction, will compare the transaction to the other options realistically available to them, and they will only enter into the transaction if they see no alternative that is clearly more attractive. Independent enterprises would generally take into account any economically relevant differences between the options realistically available to them (such as differences in the level of risk or other comparability factors discussed below) when valuing those options (paragraph 279) .

The objective in relation to comparability is always to seek the highest practicable degree of comparability, recognising though that there will be unique situations (which could be a result of business complexity) and cases involving valuable intangibles where it is not practicable to apply methods based on a high degree of direct comparability (see subparagraph 87(e) of TR 94/14 and paragraph 280 of this Ruling).

The standard of comparability that is practicable will be determined by the extent of reliable data on which to make comparisons with uncontrolled situations and dealings in the particular case. Comparisons with controlled dealings by other taxpayers cannot be regarded as arm's length comparisons (paragraph 281) .

Adjustments need to be made for any material differences so that the dealings can be as similar as is possible. If suitable adjustments cannot be made then the dealings cannot be considered to be comparable. Commonly, the use of methods other than the traditional transaction methods will produce results that without careful analysis and adjustment would not be sufficiently reliable to demonstrate the levels of comparability necessary to form an arm's length range. This is a particular issue where operating expenses are taken into account and the comparison is done at or approaching the net profit level. In these situations and in others where the level of comparability is affected, notwithstanding that adjustments to achieve true comparability cannot be made, the data should not be completely discarded as, in the absence of any true comparables, it may, in conjunction with other methods, provide insight as to what constitutes a reasonable approximation of an arm's length outcome (paragraph 282) .

Some dealings between taxpayers and unassociated enterprises may not be able to be accepted as reliable comparables because they may not be made in the ordinary course of business. An example would be a relatively insignificant sale made at the same price as charged to associated enterprises in order to create an internal comparable to justify the pricing to associated enterprises, but which, by open market standards required by the arm's length principle, was concessional to the unassociated enterprise (paragraph 283) .

It needs to be remembered that the ATO, when applying any method, may have more information available than a taxpayer has or can have reasonable access to through its own efforts. This data should be used where it enables a more reliable determination of the arm's length outcome, by producing a higher degree of comparability, though appropriate steps should be taken, subject to the need to protect the confidentiality of other taxpayers, to allow the taxpayer an adequate opportunity to defend its own position and generally safeguard the taxpayers' rights to have ATO decisions reviewed by the Administrative Appeals Tribunal ('AAT') or a court. In this regard the ability of the AAT and courts to hear evidence on a confidential basis in a closed hearing may be relevant. Not to use the more reliable information would undermine the statutory objective of the arm's length principle as expressed in Australia's legislation (paragraph 284) .

The assessment of comparability can be affected by: • the characteristics of the goods or services; • the terms and conditions of relevant agreements; • the relative risk assumed by the taxpayer, associated enterprises and any independent party being considered as a possible comparable; • economic and market conditions; and business strategies. All of these aspects need to be carefully considered in any comparability analysis in order to increase the reliability of the analysis. In this regard all the matters set out in paragraphs 285 to 304 below should be addressed.

To properly address comparability it is essential to analyse the functions performed, assets used and risks assumed by the taxpayer to identify the economically significant activities and to compare the results with a similar analysis of uncontrolled dealings or of uncontrolled enterprises that are being considered as possible comparables. The level of functional analysis that is needed will depend on the facts of the case. The matters discussed in paragraphs 286 to 292 and 509 to 591 should be taken into account in undertaking a functional analysis.

A valid conclusion as to what constitutes an arm's length outcome for a dealing usually requires examination of several years of dealings for both the controlled and uncontrolled parties. In this way differences due to business or product cycles can be more effectively taken into account and comparability more reliably determined (paragraphs 299 to 300) .

The number of years that need to be examined will depend on the facts and circumstances of the case, but as a starting point the ATO will consider the current year and the preceding four years (paragraph 301) .

In some cases it may be more appropriate to group products, services or activities in order to properly reflect the relative contributions to profit from the associated enterprise dealings. Grouping may be appropriate in the following situations: (a) Transactions/components of transactions Dealings between associated enterprises in a particular product may involve separate transactions for the product, the intangibles associated with the product, technical advice, management services and any other related matters. Where the independent dealings being considered as possible comparables cannot be disaggregated, it would generally be appropriate to group all the relevant transactions between associated enterprises so comparability to the uncontrolled party package deal transaction can be properly determined. (b) Integrated operations If it was decided to route the transaction through an associated enterprise it may be more appropriate to consider the dealing in its entirety rather than consider the component transactions on a separate basis. The combined/channel profit, the functions of each of the associates, the value added by each of the channel entities and the amount of profit appropriated to each entity would need to be considered when applying the arm's length principle to set or review prices or conditions. There could well be practical difficulties in determining the true value added by any intermediate company if it is considered in isolation. If it cannot be demonstrated in a particular case that the intermediate company either bears a real risk or performs an economic function in the chain that has produced the value of the goods or services, then any profit element that is claimed to be attributable to the activities of the intermediate company should be attributed elsewhere in the MNE group. (c) Product lines The business activities of a member of an MNE group are the importing and wholesaling of toasters, electric kettles, blenders and the provision of services in the form of advice on satellite communication. Although the MNE management may have a number of separate product lines it may be appropriate in analysing comparability to group the household electronic products together if the functions of wholesaling these products are similar. This could avoid possible distortions that may arise where indirect costs are allocated to individual products. (d) Prior dealings On occasions it will be important to look back over a sequence of transactions in order to put a current transaction in an appropriate context. This is important where there has been a substantial prior investment in the development of intangibles, or a prior sale of a relevant asset. There is a need to establish an appropriate setting or starting point for an identification of the economic alternatives that an uncontrolled decision maker would normally wish to consider and to identify comparables where these exist (paragraphs 305 and 306) .

Where dealings have been grouped there would be a need to allocate relevant operating, financial or other expenses across different product lines, divisions etc to reflect that grouping. Where it is not possible to allocate on a direct basis a soundly-based method of indirect allocation should be used that accords with accepted accounting principles and fits the particular circumstances (paragraph 307) .

In some cases, comparability can only be established by a further extension of the grouping approach. It may be necessary to aggregate the product or business lines so as to consider the matter in its proper business and economic context. This situation may arise where, for any reason, there is insufficient data available on comparable dealings to undertake a comparability analysis on any other basis (paragraphs 308 and 309) .

Comparability extends beyond product similarity. The special features of any relationships need to be taken into account. In this regard the matters discussed in paragraphs 310 and 311 need to be taken into account.

Where an aggregated analysis is undertaken and the taxpayer is dealing with related parties from different taxing jurisdictions it will still be necessary to allocate the income and expenses of the cross-border operations of the related parties to the different taxing jurisdictions on the basis of economic value added by each party (paragraph 312) .

There is a need to ensure that any differences in accounting treatment between entities being compared are adjusted so that an accurate comparison of costs and margins can be made. The basic rule is that while accepting that accounting standards will vary between countries, true comparability must be based on a consistent approach insofar as components of income and costs taken into account in comparing the performance of the taxpayer with that of the independent enterprise being considered as a possible comparable (paragraph 313) .

If data is not available to determine the basis of accounting of any enterprise being considered as a comparable then any comparable analysis should be at the net margin level or at a level that would include all relevant costs (paragraph 314) .

In the practical application of transfer pricing methodologies, an arm's length result may not always be a single point. The application of a single method or several different methods may produce a range of reliable results (paragraph 315) .

There are a number of considerations to be taken into account when constructing an arm's length range. Comparable uncontrolled dealings need to be identified and selected on the basis of criteria required to undertake the method being applied (paragraph 316) .

If material differences exist between the dealings by associated enterprises and the cases being considered as possible comparables, adjustments need to be made to reflect the differences in order to improve the comparability of the uncontrolled dealings. If reasonably accurate adjustments cannot be made to eliminate material differences then the case being considered as a possible comparable is not truly comparable (paragraph 317) .

The arm's length range will be constructed using only comparable uncontrolled dealings that have, or have been adjusted to, a similar level of comparability with the controlled dealings (paragraphs 318 and 319) .

In the absence of comparable uncontrolled dealings, it may be possible to infer from other industry information available whether dealing between the associated enterprises achieve an arm's length outcome. However, it should be noted that data which does not achieve the required level of comparablity cannot be used in constructing an arm's length range and, while it may be useful in terms of broad indications, cannot be given the same status in determining an arm's length outcome. Nevertheless, it does provide relevant information which, when combined with other information, may assist in determining an arm's length outcome (paragraph 320) .

Where a single methodology is used, it has to be capable of being applied with similar accuracy and reliability to each element of data constituting the range, having regard to all the factors relevant to comparability (paragraph 322) .

Where there is substantial divergence between data in the range it is doubtful that all the data in the range are truly arm's length outcomes. In such cases the reliability of the data in respect of each possible comparable, any adjustments made for material differences in comparability and the methodology itself should be reviewed (paragraph 323) .

There would be more confidence in ranges that are established by the use of different methodologies if those ranges, when overlayed, reflect common results (paragraph 324) .

A high level of comparability is required in order to apply a traditional transaction methodology (CUP, RP and CP methods). When using these methods, an outcome that falls within a properly constructed arm's length range should be regarded as being arm's length. The qualification to this statement is that the data used to construct the range must be truly comparable. However, if the dealing falls outside the arm's length range, it is a matter of judgment as to which point in the range the adjustment should be made. The ATO concurs with the view of the OECD that the adjustment should reflect the point in the range that best accounts for the facts and circumstances of the controlled transaction (paragraph 325) .

When applying a method other than a traditional transaction methodology (such as a profit comparison) it is not possible to give the same assurances in every case. The approximations used in applying these other methods which rely on broader measures of comparability can give extensive ranges, some of which may not be sufficiently accurate to permit the general statement that any point in the range may be regarded as arm's length (paragraph 326) .

Two categories of intangibles are recognised - production intangibles and marketing intangibles. Production intangibles typically include patents, trade secrets or unpatented know how. Marketing intangibles include trade marks, trade names and distribution networks. Hybrids of these categories can occur where intangibles obtain value through the activities of research and development/production and of marketing and sales. Intangibles may also include copyright protected software or specialised methods for providing service to customers, or of doing business (paragraphs 327 to 328) .

Intellectual property such as know how may be either a production or marketing intangible (paragraph 329) .

Australia's transfer pricing tax laws do not differentiate between the supply or acquisition of intangible property and any other transfer of property under an international agreement (paragraphs 330 and 331) .

It is necessary to clearly establish the existence and nature of the intangibles before attempting to attribute to them any value or taking them into account in applying an arm's length methodology (paragraph 332) .

Intangibles with different strengths will need to be rewarded differently. A patented production process may be useful, but it may be fairly simple to design around the patented aspects in order to achieve a similar outcome. This type of intangible should not receive the same level of relative reward as a breakthrough patent which may, say, significantly reduce production costs and improve the product so that there is greatly improved customer demand (paragraph 333 and 334) .

A functional analysis should identify each party's contribution to any manufacturing intangible or marketing intangible. A shared ownership of the intangibles derived from the economic contribution of the parties could result. This could influence the selection of a methodology (paragraphs 335 to 337) .

There are a number of internationally accepted methodologies - which are also accepted by the ATO - that test compliance with the arm's length principle in different ways (paragraph 338) .

These arm's length methodologies are divided into two groups: (a) the traditional transaction methods ('traditional methods') being the comparable uncontrolled price (CUP) method; the resale price (RP) method; and the cost plus (CP) method; and (b) the profit methods ('profit methods') which include the profit split methods and profit comparison methods (paragraph 339) .

It needs to be recognised that since international business is complex and dynamic, this process of refinement will be an ongoing one and the application of methodologies must remain flexible and receptive to those developments which may lead to refinement of the present arm's length methodologies or the development of new ones for particular categories of case. It is also recognised that enterprises may have either developed their own methodology or adapted the ones discussed in this Ruling to best suit their circumstances. Such approaches should not be automatically discounted as they may be appropriate in the enterprises particular circumstances (paragraphs 340 and 341) .

It is not possible to provide specific rules that will cover every case and no one method is suitable in every situation (paragraph 342) .

Since precise calculations cannot be made and the application of any methodology involves elements of judgment there is a need to avoid making adjustments to account for minor or marginal differences in comparability (paragraph 343) .

Since the issue of Taxation Ruling TR 94/14, the OECD has reviewed its description and discussion of the methodologies and has confirmed that certain profit methods are consistent with the arm's length principle when applied in cases where there is insufficient data on uncontrolled transactions or where such data is considered unreliable, or due to the nature of the business situation and in a way that has proper regard to comparability. The ATO agrees with this view. It is essential to always be mindful of what is reasonable and to adopt practical, flexible approaches in the application of the methodologies so that they achieve the highest degrees of comparability that is feasible and to avoid disputes about theoretical aspects that may not produce practical solutions in a given case (paragraph 344) .

As was stated in paragraphs 86 and 87 of TR 94/14, the ATO will select the method that is the most appropriate or best suited to the facts and circumstances of the particular case. The ATO is under no obligation to accept the particular methodology used by a taxpayer unless, on an objective analysis, it produces the most accurate calculation of the arm's length outcome. The most appropriate method will be the one that produces the highest practicable degree of comparability. Its selection will be driven by the extent to which reliable data on comparables is available in the particular case. Taxpayers and the ATO should have regard to all the principles in paragraph 87 of TR 94/14 when selecting a methodology (paragraph 345) .

Any transfer pricing methodology used to calculate an arm's length consideration in international dealings between associated enterprises must be applied in a way that will provide an arm's length outcome by closely reflecting commercial and economic reality and the economic contribution made by the enterprises in each jurisdiction (paragraph 346) .

Australia follows the OECD view that the CUP the RP and CP methods should be preferred over profit methods as a means of establishing whether a dealing is arm's length. It is also accepted that where it is possible to identify comparable uncontrolled transactions, the CUP method will provide a higher degree of comparability than the RP and CP methods (or any other method) (paragraphs 347 and 348) .

The choice of the most appropriate method should be based on a practical weighting of the evidence having regard to the nature of the activities being examined, the quality and reliability of the data and the nature and extent of any assumptions and the degree of comparability that exists between the controlled and uncontrolled dealings or between enterprises undertaking the dealings where the difference would effect conditions in arm's length dealings being examined (paragraphs 349 to 351) .

Where an analysis of comparability has been undertaken using one of the traditional transaction methods and there is some uncertainty as to the reliability of the outcome, perhaps due to comparability factors and the quality of the data used, it would be appropriate to check the outcome by using some other basis. One way this may be done is by comparing the result of the combined/channel profits achieved by applying the selected method with the result achieved by a method having regard to the matters like expected rates of return, risk levels, profitability, hurdle rates or other statistical analyses that independent parties would use to evaluate potential transactions (paragraph 352) .

The data necessary to apply the traditional transaction methods must be highly comparable. Data that is either not comparable or is only broadly comparable would not meet the comparability standards required of these methods. Nevertheless, this data may have a use in other methods but the result should not be preferred over those obtained by the application of a transaction net margin.

The CUP method compares the price for property or services transferred in a controlled transaction to the price charged for property or services transferred in a comparable uncontrolled transaction in comparable circumstances. This methodology could be used to arrive at an arm's length outcome for a wide range of dealings, not just prices for the transfer of tangible goods. For example, it may be appropriate to check whether a royalty rate for the use of intangible property, interest rate for funds supplied or acquired, or a management fee for services acquired or provided, complies with the arm's length principle (paragraphs 353 to 356) .

However, there may be cases where the dealings between associated enterprises involve a variety of transactions (tangible and intangible property, management services, interest, etc.) and it is not possible to obtain CUPs for all the transactions. In those cases the CUP method may be still suitable for some classes of dealings where it is supported by other methods that will reliably evaluate those transactions not determined by the CUP methodology (paragraph 359) .

The most important comparability factors are similarity of product, contract terms and economic/market conditions. While the application of the CUP methodology involves close product similarity, its application also requires a consideration of all other factors relevant to comparability. In this regard see paragraph 90 of Taxation Ruling TR 94/14 and paragraphs 360 to 364 of this Ruling.

Australia and other OECD member countries recognise that the CUP method provides the most direct comparison, and encourages its use even where adjustments to the data are required to be made, provided that reliable adjustments can be made for material differences. It needs to borne in mind that a minor difference in the property transferred in the controlled and uncontrolled transactions can materially affect the price. Nevertheless, the method should be seriously considered and not routinely dismissed because it may be difficult to make adjustments. However, if the differences have a material effect on price and adjustments cannot be made with any confidence, alternative methods will need to be considered. It should also be remembered that, since adjustments to controlled and uncontrolled dealings to develop a comparable will inevitable involve elements of judgment, the extent, number and reliability of such adjustments will affect the relative reliability of the CUP method analysis (paragraph 365) .

With the CUP method, comparisons need to be made for each dealing. Once an arm's length consideration has been determined there will be a need to monitor the correlation of the dealings of the taxpayer and those of the comparable over time to insure that the CUP initially selected remains valid and to reflect any price movement in the relevant market (paragraph 366) .

The CUP method is a particularly reliable method where an independent enterprise sells the same product as is sold between two associated enterprises. Where the circumstances surrounding controlled and uncontrolled sales are identical, except for the fact that the controlled sales price is a delivered price and the uncontrolled sales are f.o.b. The differences in terms of transportation and insurance generally have a definite and reasonably ascertainable effect on price. Therefore, to determine the uncontrolled sales price, adjustment should be made to the price for the difference in delivery terms (paragraph 367) .

Where, for whatever reason, it is not possible to use a CUP methodology the application of the other traditional transactional methods (other traditional methods) should be considered. These other traditional methods are: a) Resale Price (RP) Method; and b) Cost Plus (CP) Method (paragraph 368) .

As with any method, the availability and reliability of these other traditional methods depend on the availability of sufficient relevant data and may need to be applied on a basket or aggregated basis as discussed in paragraphs 306 to 312. The best comparisons will always be where the taxpayer has comparable dealings in comparable circumstances with both associated parties and uncontrolled parties (paragraphs 369 and 370) .

The ATO agrees with the OECD view that in making comparisons for the purposes of the RP and CP methods, fewer adjustments are normally needed to account for product differences than under the CUP method, because minor product differences are less likely to have as material an effect on profit margins as they do on price. However, closer comparability of products will produce a better result and significant differences in products or services is likely to be reflected in the functions performed (paragraphs 371 and 372) .

Fewer adjustments may be necessary to account for product differences under the other traditional methods than the CUP method, and it may be appropriate to give more weight to other factors, some of which may have a more significant effect on the margin than they do on price. Such differences may include: (a) the various stages of the business and product cycles; (b) the management strategies. For example, where some business assets are rented as opposed to being purchased; (c) the nature and extent of the functions performed assets employed and risks assumed; and (d) the cost structures of the enterprises being compared. In this regard see paragraphs 94 and 360 of Taxation Ruling TR 94/14 and 2.16 to 2.21 of the 1995 OECD Report (paragraph 373) .

Where there are differences in the functions or enterprises being compared that materially affect the margin earned in the controlled and uncontrolled transactions, adjustments should be made to account for such differences. The extent and reliability of any adjustments may affect the relative reliability of the analysis under the other transactional methods (paragraph 374) .

In some cases when applying other traditional methods, it may be difficult to obtain sufficient data or establish suitable comparable dealings to have any confidence in the outcome of the analysis. In such cases it may be appropriate to supplement the RP and CP methods by considering the results obtained from applying other methods, such as a profit method (paragraph 375) .

The resale price (RP) method begins with the consideration at which property (that has been purchased from an associated enterprise) is resold to an independent enterprise. The resale consideration is then reduced by an appropriate gross profit margin representing the amount out of which the reseller would seek to cover its selling and other operating expenses and, in the light of the functions performed, assets used and risks assumed, make an appropriate profit. Where the gross profit margin so calculated is deducted from the resale price the balance - subject to adjustment for other purchase costs like customs duties - is regarded as the arm's length price for the purchases from the associated enterprise (paragraph 376) .

When applying the RP or CP method, financing expenses are often excluded from general, administrative and selling expenses on the basis that the funding of the business is not a material consideration in comparing products, outputs or functions, and that the financials can in fact produce distortions. There will be other cases where the financing expenses are part of the transfer pricing examination. One aspect of this might be a consideration of whether all the arrangements between the associated parties leave the taxpayer with sufficient working capital compared to what arm's length parties would expect in those circumstances (paragraphs 377 and 378) .

This method is most useful where the reseller markets the product without adding substantial value by physically altering the goods and the resale is realised within a short time of the reseller's purchase of the goods (paragraph 379) .

It may be more difficult to use the resale price method to arrive at an arm's length price where, before resale, the goods are further processed or incorporated into a more complicated product so that their identity is lost or transformed (for example, where components are joined together in finished or semi-finished goods). Another example where the resale price margin requires particular care is where the reseller contributes substantially to the creation or maintenance of intangible property associated with the product (for example, trademarks or tradenames) which are owned by an associated enterprise. In such cases, the contribution of the goods originally transferred to the value of the final product cannot be easily evaluated. The amount of the resale price margin will be influenced by the level [and, we would add, nature] of activities performed by the reseller. The level of activity performed by the reseller, whether minimal or substantial, would need to be well supported by relevant evidence. This would include justification for marketing expenditures that might be considered unreasonably high; for example, when part or most of the promotional expenditure was clearly incurred as a service performed in favour of the legal owner of the trademark. In such a case the cost plus method may well supplement the resale price method (paragraphs 380 and 381) .

It is also more reliable where the reseller on-sells within a short time. The general equation would be best expressed as: Selling Price - Appropriate Gross Margin = Arm's Length Purchase Price (paragraphs 382 and 383) .

Of course an adjustment would have to be made for other costs associated with the purchase of the product to arrive at an arm's length purchase price. Examples of these would be insurance, transport costs and any other expenses involved in getting the products into store. Care would need to be taken where the expenses involve payments to associated enterprises. In such cases the transfer pricing review should cover both the acquisition of property and the related expenses (paragraph 384) .

It should be borne in mind that the starting point with RP method is to make a comparison of the gross profit with comparable independent enterprises. This will indicate any discrepancy arising in the cost of goods sold. Where there is a discrepancy below the gross profit line it is likely that the RP method is not the most appropriate method (paragraph 385) .

The appropriate gross margin is the amount of profit that, based on an arm's length comparison, is considered necessary to compensate the enterprise for its costs and make an appropriate profit that accords with the functions undertaken, assets employed and risks assumed (paragraph 386) .

The appropriate gross margin is usually measured at the gross profit level. However, in some circumstances it may be more accurate to undertake the comparison at some other (intermediate) profit level. The profit level at which to compare will be determined by the availability of sufficient reliable data, bearing in mind that there may also be a need to adjust for accounting differences between the associated enterprise and the uncontrolled enterprise being considered as a possible comparable (paragraph 387) .

Whenever the RP method is applied it would be appropriate to check whether the resale price margin so determined is realistic having regard to the operating expenses of the taxpayer (paragraph 388) .

Where company policies are determined or influenced by an associated enterprise then these type of costs may need to be taken into account when undertaking the comparability analysis in the course of applying the RP or CP method. For example, if it is corporate policy, for the benefit of any associated enterprises' intangible assets, to seek accommodation, staffing levels, etc., that are materially different to those levels required by comparable unrelated parties or to reduce profit margins for market penetration or market share, keep excessive stock levels or incur excessive marketing costs that materially differ from those policies of a comparable enterprise, then those costs may need to be taken into account as they would relate to management decisions imposed on the enterprise that may not have been accepted by an independent party operating wholly independently (paragraphs 389 and 390) .

The appropriate gross profit margin would be expected to vary according to the amount of value added by the reseller. It would be expected the appropriate gross profit margin would increase with the increased assets, functions and risks. In the cases where Division 13 is being applied it may be necessary to apply the arm's length gross profit margin on the basis of each international agreement. However, paragraph 23(b) of the Acts Interpretation Act 1901 could have the effect that this is not necessary in cases where the relevant dealings are of a similar type that could be treated as a basket of goods or services (see paragraphs 135 and 432 to 438 of TR 94/14). In relevant cases, Division 13 determinations would be prepared by the ATO relying on both bases so as to allow this issue to be considered in the event of a dispute between the taxpayer and the ATO proceeding to the Administrative Appeals Tribunal or a court. These comments on Division 13 also apply in relation to the CP method as explained in paragraph 433 (paragraphs 391 to 396) .

A methodology which adopts a margin which is calculated as a certain percentage of the resale price (for the purpose of determining the appropriate transfer price), where the percentage chosen is not benchmarked against comparable independent dealings is not a resale price methodology (paragraph 397) .

In extreme cases such a methodology might be able to be used for the purposes of subsection 136AD(4). However, every effort should be made to use other methods like profit methods or a mix of methods before resort is had to such an approach (paragraph 398) .

Where such a methodology has to be used in cases where no other approach is reasonably open, the fixed percentage should be calculated to produce a result that fairly reflects the functions performed, assets employed and risks undertaken, the intention always being to reasonably approximate an appropriate return for the economic value added in a way that is consistent as practicable with the arm's length principle (paragraph 399) .

The cost plus method begins with the costs incurred by the supplier of property (including services) in a controlled transaction for property transferred or services provided to a related purchaser. An appropriate cost plus mark up is then added to this cost, to make an appropriate profit in light of the functions performed and the market conditions. What is arrived at after adding the cost plus mark up to the above costs may be regarded as an arm's length price of the original controlled transaction. This method probably is most useful where semi-finished goods are sold between related parties, related parties have concluded joint facility agreements or long-term buy-and-supply arrangements or where the controlled transaction is the provision of services (paragraphs 402 to 406) . Examples of the application of the CP method appear at paragraphs 433 and 434.

The costs, in general, that need to be established for the CP method will be the direct and indirect cost of production of the relevant goods or services. It is important to remember that the costs are limited to those of the supplier of the goods or services. Any transfer pricing examination should have regard to the fact that the manipulation of the allocation of expenses could inappropriately increase or decrease the production costs of the taxpayer to which the cost plus margin is applied. These costs would include: direct costs of producing the goods or services such as the cost of raw materials or the salaries of skilled service staff; indirect costs of production, which although closely related to the production process may be common to several products or services (for example, the costs of a repair department that services equipment used to produce different products or deliver different services); and an appropriate allocation of other general, and administrative expenses, that can be related to the production activities the subject of the application of the CP method (paragraphs 407 and 408) .

This aggregation of direct and indirect cost is also known as absorption costing. These calculations should generally be done on the basis of historical cost. Although there are some exceptions to this general principle (for example, the use of marginal costing to dispose of marginal production), these exceptions are subject to strict controls. Consistent with the OECD approach and having regard to generally accepted accounting principles, as a general rule, the use of absorption costing will be required by the ATO where the cost plus method is used. The very limited exceptions are where replacement cost and marginal cost result in a more accurate measure of the appropriate profit margin. In such cases the justification for the view that replacement or marginal cost provide a higher integrity should be adequately and contemporaneously documented (paragraph 409) .

IT 2350 provides some guidance in relation to the use of absorption costing and in this respect would be relevant for the purposes of the application of the CP methodology (apart from the reference in paragraph 5 of IT 2350 which indicates that at that time direct costing was recognised for accounting purposes - which is not the case now) (paragraph 410) .

As historical costs such as materials, labour, depreciation, etc., may vary over a period it may be appropriate to average these costs when determining the appropriate level of costs when applying the CP method in relation to a limited period. Averaging may also be appropriate when determining costs across product groups or when applying the CP method in cases where grouping of dealings is needed to properly assess comparability (paragraph 411) .

The objective in allocating indirect costs is to determine the degree to which the indirect costs - which are a reflection of the taxpayer's functions, assets and risks - have contributed to the production of the goods or services being examined in the transfer pricing review. The ATO will evaluate the taxpayers' allocation of indirect costs against sound cost accountancy principles. In this regard, where a taxpayer allocates indirect costs using criteria that are key to the nature of the taxpayers' profit making activity and which can fairly apportion the particular costs on the basis of the extent of the activity subject to the transfer pricing examination relative to the other purposes for which the costs were incurred, the ATO will accept the allocation. The basis of allocation must, however, make sense in the context of the particular case and cannot produce significant distortions (paragraphs 412 and 413) .

In cases where a taxpayer uses a formula to allocate indirect expenses, ATO staff examining the allocation would need to establish whether the formula has been consistently followed over a number of years and whether there is any evidence of manipulation that produces an inappropriate loading of expenses in Australia. Where different types of indirect costs are being allocated it may be appropriate to use different allocation criteria (paragraphs 414 and 415) .

Care would need to be taken where the Australian taxpayer is part of an integrated production process involving a number of countries and the taxpayer has acquired partly finished goods from a foreign associated enterprise. In such cases the CP method would be more difficult to apply and its reliability would need to be carefully checked (paragraph 416) .

There may be occasions where questions would arise as to whether the costs incurred in relation to such acquisitions from associated enterprises were at arm's length. Before such costs were incorporated into the value of costs for the purposes of the CP methodology, an analysis would need to be undertaken to ascertain what an arm's length consideration for the associated enterprise acquisitions would be (paragraph 417) .

The above the gross profit line costs will need to be apportioned between the controlled dealings and the other business activities of the taxpayer on an appropriate basis (paragraph 418) .

Many taxpayers also have dealings with associated enterprises other than just in relation to the controlled transaction under review - which may effect the above the gross profit line of the controlled dealings. The nature of such dealings whether they are on an arm's length basis would be relevant to determining comparability between the controlled dealings and the uncontrolled dealings. An analysis of these other dealings between associated enterprises would therefore be necessary (paragraph 419) .

The appropriate mark-up is the amount of the mark-up on the relevant cost determined on the basis of an arm's length comparison (paragraph 420) .

The cost plus mark-up of the taxpayer in the dealings between associated enterprises should ideally be established by reference to the cost plus mark-up that the taxpayer earns in comparable uncontrolled dealings. Where the taxpayer has no comparable uncontrolled dealings, the cost plus mark-up may be able to be determined on the basis of comparable dealings by independent enterprises that are operating wholly independently. Functional analysis will assist this process (paragraph 421) .

The appropriate mark-up should be measured at the gross profit level. However, in some circumstances it may be more accurate to consider some intermediate profit level in order to make comparisons on a consistent basis (e.g., to adjust for accounting differences between the taxpayer and the company being considered as a comparable) (paragraphs 422 to 425) .

A methodology which applies a fixed percentage mark-up to a relevant cost base where that fixed percentage is not benchmarked against comparable independent dealings is not a cost plus methodology (paragraph 426) .

In extreme cases such a methodology might be able to be used for the purposes of subsection 136AD(4). However, every effort should be made to use other methods - like profit methods or a mix of methods - before resort is had to such an approach (paragraph 427) .

Where such a methodology has to be used, in cases where no other approach is reasonably open, the fixed percentage should be calculated to produce a result that fairly reflects the functions performed, assets employed and risks undertaken, the intention always being to reasonably approximate an appropriate return for the economic value added in a way that is as consistent as practicable with the arm's length principle (paragraph 428) .

Where valuable intangibles are involved in the transaction it may be appropriate to supplement the cost plus methods by considering the results obtained from applying other methods such as the profit split method in order to ensure that the profit contribution of the intangibles are properly reflected (paragraphs 429 and 430) .

Differences in the level and types of operating expenses and non-operating expenses (including financing) could indicate functional differences or additional functions between the taxpayer and the parties or dealings being compared that require adjustments to achieve a valid comparison. Subject to the earlier discussion in paragraph 120, no adjustment to the gross margin may be appropriate if the differences in expenses reflect different degrees of efficiency (paragraphs 431 and 432) .

Sometimes it is not possible or practical to use traditional transactional methods (traditional methods). Such situations may arise where: (a) there is insufficient reliable data to analyse comparability so as to determine an arm's length outcome other than through a profit split or a profit comparison at the net profit level; (b) the traditional methodologies are unable to establish an arm's length consideration because, for example, the product or service in question is unique or contains out-of-the-ordinary intangibles; or (c) while theoretically sound, the traditional methods are not practicable because of the complexity of the business situation or the extent and diversity of the taxpayer's cross-border dealings with associated enterprises. These cases present some significant difficulty in following traditional methods such that due administration of the law is seriously jeopardised beyond what a reasonable person would accept, and it is not merely a matter of convenience that supports a different approach. It would be expected that consideration would be given to the use of traditional methods on a basket approach and profit methods would be applied only where traditional methods are unreliable because they are not conceptually applicable or practicable in the particular case, or because there is insufficient reliable data to apply them. In these situations it may be more appropriate to consider the use of profit methods (paragraph 435) .

In the Australian economy certain industries are dominated by MNEs to the point where the parties operate in niches and the material differences between the mainstream market and the niches do not make adjustments to achieve comparability feasible (paragraphs 436 to 437) .

There is also the problem that in many cases there is a variety of transactions (transfers of tangible and intangible goods and services) back and forth between the associated enterprises - some of which may involve overlaps and there may be no comparables for the combination of transactions. In these cases profit methods may be a more reliable way to set or review the transfer pricing used in the dealings between the associated enterprises or to check findings made using traditional methods where there is doubt about the reliability of the data used (paragraph 438) .

Where, for whatever reason, and after taking account of the guidelines in this Ruling on selection of methods, it is not possible or practical to use the traditional methods of CUP method, RP or CP methods, the application of profit methods may be considered. The type of profit method used will depend on the facts and circumstances. These profit methods include: (a) the profit split methods; and (b) the profit comparison method (referred to by the OECD as the 'transactional net margin method' in the 1995 OECD Report) (paragraph 439) .

One of the main differences between the profit split and the profit comparison method is that the former is applied to all the relevant associated enterprises whereas the latter is applied to only one of the associated enterprises. A one-sided analysis potentially can attribute to one member of an MNE group a level of profit that implicitly leaves other members of the group with implausibly low or high profit levels. However, this is also a risk with the RP and CP methods, which are also one-sided analyses. Care needs to be taken to ensure in so far as practicable that the result produced by the one-sided application of any method makes commercial sense in the circumstances of the case (paragraph 440) .

There is a need in some cases to use profit methods so that the arm's length principle can be implemented as closely as practicable. Not to seek some means of testing taxpayers' dealings with associated enterprises and to devise some solution for transfer pricing problems would undermine the arm's length principle (paragraphs 441 to 443) .

While it is possible to apply a profit method in respect of a single transaction, these methods are generally applied in respect of a group or a 'basket' of transactions or on an aggregated basis. The important principles in this regard are: (i) to the extent transactions can be disaggregated without going to unreasonable lengths, then they should be analysed at the lower level; (ii) where transactions need to be analysed on a combined basis care should be taken to ensure that the profits, the subject of the transfer pricing examination, are limited to the profits that arise from controlled transactions. It would generally be inappropriate to apply a profit method on a 'whole of entity' basis unless all of the taxpayer's activities involved associates and, if a profit comparison is being used, the different types of controlled dealings can be approximately compared on a consistent basis with a similar basket of uncontrolled dealings by an independent enterprise operating wholly independently. (See paragraphs 135 and 432 - 438 of TR 94/14). It needs to be remembered, though, that the Commissioner is entitled to estimate the arm's length consideration in respect of international agreements if there is insufficient information to determine the arm's length amount (paragraph 444) .

The profit split method first identifies the combined profit or loss from the dealings between the associated enterprises. It then splits those profits or losses 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 combined profit may be the total profit from the transactions or a residual profit intended to represent the profit that cannot readily be assigned to one of the parties, such as the profit arising from high value, sometimes unique, intangibles. The contribution to each enterprise is based upon a functional analysis and valued to the extent possible by any available external market data (paragraph 445) .

In applying the profit split methods the object is to determine if the split of the combined profit from the dealings between associated enterprises is arm's length in light of the functions performed, assets used and risks assumed by the respective parties from the point of starting the related party manufacture to the sale to an arm's length party (paragraphs 446 to 448)

To determine if the split is arm's length an assessment has to be made 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. There is no one method to undertake this assessment as each case has to be decided on the data available and the facts and circumstances of that case (paragraph 449) .

The following factors need to be taken into account in undertaking a profit split: (a) The relevant dealings: (i) there is a need to determine if the profit split is to be undertaken on a particular product line, a basketing of products or where appropriate on the basis of the strict guidelines above, a whole of entity basis all of which will include a consideration of any intangible assets as well as financial assets (whether shown on the balance sheet or not). In all cases an allocation of general administration and similar costs to the relevant dealings will be required; (ii) where the taxpayer has dealings with more than one associated enterprise, care will be needed to identify the profits applicable to each party. Using the above example, if the reseller was supplied goods from two related manufacturers from different tax jurisdictions any profit split would need to identify the value added in each jurisdiction and the appropriate share of profit for each. (b) Consolidation of accounts: so that the combined profit can be determined, the accounts of the parties need to be put on a common basis as to accounting practice and currency and then consolidated. Once the split has been determined the accounts can then be rewritten on a separate entity basis, taking account of the relevant requirements in the taxpayer's home jurisdiction (paragraph 450).

A possible difficulty in attempting to undertake a profit split is obtaining the required information from foreign enterprises or tax administrations so that the combined profit can be determined. Nevertheless, reasonable attempts should be made to obtain the relevant information because there is more certainty that an arm's length share of profits has been reached when the economic contribution by all parties profit share has been examined than would be the case if a one-sided profit comparison was used (paragraph 451) .

There is a need to establish whether to apply the profit split to the projected or actual profits. The determining basis should be made as follows: (a) Projected Profits: where a taxpayer uses a profit split to establish (as opposed to 'review') transfer pricing for controlled transaction, this would necessarily be done on the basis of the projected profits because the actual profits would not be know at the time. This would produce a fractional allocation which would then be applied as the actual profit derived. However, if there are variances between projected and actual profits arm's length parties would make appropriate adjustments when reviewing their profit split projections for future years. In some cases, for example where the joint activity involves an intangible and the value is unclear, arm's length joint ventures might include a review clause in their agreement that would operate to review the profit split in the event of a major change in actual profit experience relative to their projections. (b) Actual Profits: where prices have been set using a basis other than a profit split (as will almost be the case) any evaluation would be undertaken on the actual profits achieved by the application of the other basis using the same information that was available at the time of the price setting thus avoiding the use of hindsight (paragraph 452) .

There are a number of approaches for estimating the division of profits, based either on projected or actual profits, as may be appropriate. The contribution analysis and the residual analysis are by far the most frequently used and are not mutually exclusive (paragraph 453) .

Under a contribution analysis, the combined profits, which are the total profits from the controlled dealings under examination, are divided between the associated enterprises based upon the relative value of the functions performed by each of the associated enterprises participating in the controlled dealings, supplemented as much as possible by external market data that indicate how independent enterprises would have divided profits in similar circumstances. In cases where the relative value of the contributions can be measured directly, it may not be necessary to estimate the actual market value of each participant's contributions (paragraph 454) .

It may be difficult to determine the relative value of the contribution that each of the related participants makes to the controlled transactions, and the approach will often depend on the facts and circumstances of each case. The determination might be made by comparing the nature and degree of each party's contribution of differing types (for example, provision of services, development expenses incurred, capital invested) and assigning a percentage based upon the relative comparison and external market data (paragraphs 455 and 456) .

A residual analysis divides the combined profit from the controlled transactions under examination in two stages: (a) first, each participant is allocated sufficient profit to provide it with a basic return appropriate for the type of transactions in which it is engaged. Ordinarily this basic return would be determined by reference to the market returns achieved for similar types of transactions by independent enterprises. Thus, the basic return would generally not account for the return that would be generated by any unique and valuable assets possessed by the participants; (b) secondly, any residual profit or loss remaining after the first stage division would be allocated among the parties based on an analysis of the facts and circumstances that might indicate how this residual would have been divided between independent enterprises. Indicators of the parties' contributions of intangible property and relative bargaining positions could be particularly useful in this context (paragraph 457).

Regard should be had to paragraphs 458 to 462 in applying the residual profit split method.

Another approach is to split the combined profit so that each of the associated enterprises participating in the controlled transactions earns the same rate of return on the capital it employs in that transaction. This method assumes that each participant's capital investment in the transaction is subject to a similar level of risk, so that one might expect the participants to earn similar rates of return if they were operating in the open market. It also assumes that return on financial capital is a relevant measure for each enterprise (paragraph 461) .

Another possibility is to determine the profit split based on the division of profits that actually results from comparable transactions among independent enterprises. In most cases where the CUP, RP and CP methods would not be used, it will be difficult to find independent enterprises engaged in transactions that are sufficiently comparable to use this approach as the primary method (paragraph 462) .

Formulary approaches are often appropriate for splitting profits from global trading. They can also be used reasonably reliably for allocating some types of head office expenses to associates, provided the principles in paragraphs 412 to 419 below are followed. The ATO will not prevent the use of a formulary approach in establishing arm's length outcomes, particularly where it is not possible or practicable to allocate an arm's length profit or to ascertain an arm's length price in accordance with other methods endorsed by the OECD. Reserving the option to use a formulary approach in appropriate cases on the basis of their facts and circumstances is consistent with the operation of subsection 136AD(4) (paragraph 463) .

Where an MNE is engaged in a global trading activity in financial products through various markets around the world, sometimes, on a twenty-four hour basis, there are some unique problems associated with the allocation of income and expenses amongst the contributing members - especially where the same pool of trading assets is used (paragraph 464) .

In some cases, the members have a fully integrated computerised trading network containing the various files (collectively known as a 'book') which is used to transact the deals. The authority to trade in this stock of financial instruments may move from one centre to the next as markets open and close, the authority continually being passed around the world to maximise the trading on the book. A deal may be opened in New York, continued in Sydney and Singapore and closed in London. During this time, the members of the multinational group who are dealing on the relevant book are taking orders from their clients at any time during the day, conducting trades for clients (from anywhere in the world) when their market is open, subject to the financial limits imposed on the dealers individually in relation the exposure they can take by setting an overall limit on risks and collectively by the multinational group - and perhaps regulators. Other functions would typically include overall strategic management, back office support like accounting, legal documentation and computer support, and sales and marketing (paragraph 465) .

In these circumstances, finding suitable comparables for the trading are not of real concern. The relevant financial markets would most likely provide the necessary arm's length considerations for the transactions entered into by each participant in the global trading of the multinational group. The real problem lies in how to allocate the income and expenses on an arm's length basis. For example, and as explained above, a deal may be opened in New York, continued in Sydney and Singapore and closed in London. By closing the deal in London, the group member in London will crystallise all the profit or loss. Yet, the profit or loss crystallised in London might effectively be attributable to all multinational group participants in the global trading. Generally speaking though, the group member in London should not recognise all the profit or loss in London because in the great generality of cases dealers will all have a certain level of skills in predicting markets, taking positions and hedging risks, there will be a level of prudential control and the activity by its nature contains an element of speculation. But more importantly, it is the fact that the profits arise largely from the integration of trading, hedging and marketing that force the conclusion that it is not the place where the profit or loss crystallises that is the determining factor in allocation. This integration is also a reason why dealers' salaries/bonuses are not a sufficient basis in themselves for allocating profit (paragraph 466) .

Primarily, the allocation of profit or loss is to be determined in accordance with the economic benefit each group member has contributed to the overall result from participating in the global trading. To determine the level of benefit, it would be necessary to complete a functional analysis establishing who does what and where, and to then allocate the income and expenses on some formula based on a suitable weighting of the assets, functions and risks of each contributing member (participant). The large volume of global trading may also prohibit anything other than a profit allocation on the basis of aggregating all trading transactions in financial instruments (paragraph 467) .

Where possible, the weighting should be based on some form of external market data. The outcome sought should be directed to reflecting what independent enterprises would have done if they were confronted with the similar allocation problem in comparable circumstances. Differences in functions assets and risks in different cases should be reflected in the allocation of profit (paragraph 468) .

A common form of global trading - perhaps the most common - is where the trading in a specific product is centralised in a particular jurisdiction. The decision on where to locate that book would be driven by commercial considerations like customer location, availability of trading skills and depth of market. There may be limited authority to trade outside the jurisdiction, but the main activity in other jurisdiction would be sales and marketing. It is also possible for back-office support to be located outside the 'book's' jurisdiction. This particular form of global trading does not necessarily require a profit split methodology - the argument for which is stronger where there is a very high level of integration - and it may be possible to use traditional methods (paragraph 469) .

Profit comparisons at the net margin level are an extension of the RP and CP methods. The profit comparison method examines the net profit margin, rather than the gross profit margin (or at some margin in between), relative to an appropriate base (e.g., costs, sales, assets), that an enterprise realises from a controlled transaction or from transactions that it is appropriate to combine (paragraph 470) .

As such, when applying the profit comparison method, care is needed to have regard to the previously discussed requirements for the application of the RP and CP methods and the principles regarding comparability (paragraph 471) .

This method requires the taxpayer (or the ATO) to compare the net margins obtained in its controlled dealings against either: (a) the net margins of the taxpayer's uncontrolled transactions in comparable circumstances; or (b) the net margins earned in comparable, uncontrolled transactions by an independent enterprise (paragraphs 472 and 473).

It is important to ensure the profit comparison is confined to the net profit from cross-border dealings with associated enterprises. Where the profit comparison method is applied on the basis of an aggregation of transactions, where such aggregation is appropriate, the reliability of such an approach relative to the application of the CUP, RP and CP methods on a similar basis would need to be rechecked and the most reliable method used (paragraph 474) .

If detailed comparable data is unavailable or there are accounting differences, that cannot be reliably adjusted to allow comparisons at the gross profit level, or the transfer pricing issues arise in respect of items below the gross profit line, comparisons at levels other than net profit may produce an incorrect outcome (paragraph 475) .

The net margins also may be more tolerant to some functional differences between the controlled and uncontrolled transactions than gross profit margins. Differences in the functions performed between enterprises are often reflected in variations in operating expenses. Consequently, enterprises may have a wide range of gross profit margins but still earn broadly similar levels of net profits. It is therefore important when applying the profit comparison method to carefully consider the functional analysis of the taxpayers and the entities being considered as possible comparables and to make adjustments whenever practicable to increase the degree of comparability, having regard to the previous discussion of the factors that could impact on comparability (paragraph 476) .

While the RP and CP methods are each based on a single ratio, there are several ratios that could be of assistance in applying the profit comparison method. The relative usefulness of the various ratios will depend on the facts of each case and the extent of reliable data available. For example, the rate of return on capital employed will be of greater importance if the taxpayer is operating in a capital- intensive industry. It would often be appropriate to have regard to more than one ratio in checking the reliability of the taxpayer's approach to determining transfer pricing. Return on assets could present particular problems when used in isolation (paragraph 477) .

An EBIT analysis can be helpful in the context of profit comparisons - bearing in mind the need to focus only on the relevant cross-border dealings between associated enterprises and to apply the EBIT approach consistently to the taxpayer and the other enterprises being considered as possible comparables (paragraph 478) .

Another possibility is the ratio of operating profit to sales, but the safeguards in relation to comparability that are needed when operating expenses are being considered should be carefully followed (paragraph 479) .

The ratio of gross profit to operating expenses (often referred to as a Berry ratio) can also be helpful in applying the profit comparison method, though care would be needed to ensure that comparisons between the taxpayer and other enterprises are limited to other enterprises that have a high degree of functional similarity (paragraph 480) .

It would also be essential to be aware of the possible distorting effects of methods of business financing, business strategies and the relative efficiency of managers when doing EBIT and Berry and operating profit to sales ratios (paragraph 481) .

It needs to be borne in mind that the application of these various ratios in appropriate cases will indicate potential transfer pricing risks at a primary level. However, further detailed analysis will be needed to identify particular transfer pricing problems. Any comparability analysis will need to have regard to the potential difficulties presented by a consideration of operating expenses (see paragraphs 476, 479 and 481 to 483).

Possible distortions through economic, market, business or product cycles would also need to be considered. The use of data covering a number of years will increase the reliability of profit comparison (see paragraphs 299 to 301).

As mentioned in the discussion on the RP and CP methods, a comparison at net profit level can sometimes be needed, but care is required when dealing with operating and financial expenses that relate to such things as marketing strategies, the management efficiencies, accounting policies (e.g., different depreciation methods) and business financing alternatives (paragraphs 485 to 489) .

An uncontrolled enterprise may not accept to pay a higher price resulting from the inefficiency of the other party. On the other hand, if the other party is more efficient than can be expected under normal circumstances, the other party should benefit from that advantage. However, where company policies are determined or influenced by an associated enterprise then the costs impacted by these policies would need to be carefully considered when undertaking the comparability analysis for the reasons mentioned in the discussion on the RP and CP methods (paragraph 486) .

Various factors will have a potential impact on the reliability of the profit comparison method and the comparability analysis should carefully address them. These factors would include how well the value of assets employed in the calculations is measured. In each case it is essential to determine the extent to which the value of the intangible property is not captured in the books of the enterprise. This as an important issue in analysing functions, assets and risks. The books may show assets at historical cost or as capitalised expenditure - which can be markedly different from their real value. Goodwill derived from product or service quality or from research and development, or from skills in distribution or the creation of marketing intangibles will often create value that is not shown on the balance sheet. These can be masked by bad management or poor financials. Some insights might be gained from market capitalisation and earnings history (e.g., earings per share). It is important that all these aspects be considered when doing the functional analysis as a good understanding can be developed of the taxpayer's strengths and weaknesses in terms of the profit drivers and features likely to reduce profitability. The factors affecting whether specific costs should be passed through, marked up, or excluded entirely from the calculation will be better understood if this approach is followed and when those factors are then taken into account the profit comparison should produce closer comparables than profit comparisons that ignore these issues (paragraph 487) .

The application of the profit comparison method requires a careful analysis of the taxpayer's operating expenses. These expenses reflect the taxpayer's functions, assets and risks and give insight into the possible distorting effects of methods of business financing and management approaches in relation to the net profit. These concerns are largely avoided with the RP and CP methods because they focus on the gross margin. However, where for the reasons outlined these methods cannot be used, the more closely operating expenses, market and business factors can be analysed and compared the more reliable the profit comparison method. Of course, care would also be needed to ensure that expenses above the gross profit line, market and business factors are also properly considered (paragraphs 488 and 489) .

Where there is a special relationship betweent the associated enterprises that produces unique dealings there may not be sufficient data or comparable dealings to apply CUP, RP, CP or the profit methods. In such cases their use should be reconsidered on the basis of possibly: (a) sufficiently broadening the comparability criteria to allow a comparison of the relevant dealings; and (b) sufficiently relaxing the normal conditions imposed in applying traditional and profit methods to allow the broadened comparability to be applied (paragraph 490).

This situation may come about where all the comparable enterprises in an industry are associated enterprises or where an associated enterprises have an industry monopoly. Where this is the case the appropriate arm's length comparison may be with enterprises in another industry segment or group of segments. However, great care is needed to ensure that the industry segments or groups of segments being compared are sufficiently similar, especially in relation to levels of profitability as well as functions performed (paragraph 491) .

Where the comparability criteria need to be broadened there may be a need to consider the dealings on a basketing, aggregated basis or, in extreme cases, on a whole of entity basis. However, the relative reliability of such an approach will need to be considered against the reliability of the applying the CUP, RP or CP method on an aggregated basis - if it is possible to use the traditional methods that way (paragraph 492) .

If the extended application of the traditional and profit methods cannot provide an answer it may be necessary to consider: (a) a mixture of the above methods; or (b) some other method or mixture of methods that is likely to lead to a result that is consistent as practicable with the arm's length principle (see paragraph 367 of TR 94/14). Where taxpayers find themselves in this category they should give serious consideration to an advanced pricing arrangement ('APA') (paragraph 493) .

Where none of the generally accepted methodologies can be used consideration would have to be given to a comparison of an enterprise's overall performance with that of other similar enterprises in the same of similar circumstances. By a somewhat similar process, the reasonableness of transfer prices may perhaps be assessed by comparing the yield or return on capital invested in the relevant associated enterprises with the yield or return on the capital invested in enterprises carrying on similar activities and requiring the same kind of capital investment. A further approach might be to look at the yield on the capital involved. The OECD's qualifications on the reliability of these approaches is noted. They are also repeated in the 1995 revision in the discussion of profit methods. Nevertheless, we recognise that they have on occasions been used by enterprises and they may have a role in assisting the establishing of an approximation of an arm's length consideration in some cases. In this regard, it needs to be remembered that all of Australia's DTAs would on the wording of various Associated Enterprises Articles allow methods that may not be permissible under the wording of the Associated Enterprises Article (Article 9) of the OECD Model Tax Convention (see paragraphs 19 to 28 above). It also needs to be remembered that subsection 136AD(4) of the ITAA, which is preserved by all of Australia's DTAs, also applies in appropriate cases where Australia does not have a DTA - and in DTA cases outside the scope of the relevant Associated Enterprises Article. However, as stated in paragraphs 11 to 13 above the choice and application of non-traditional or profit methods are limited to the statutory purpose of achieving an arm's length outcome (paragraph 494 and 495) .

The US CPM is a regulated process for developing an answer claimed as being arm's length. Because it relies upon data being used in a specified manner, it may not necessarily be flexible enough to deliver the best approximation of an arm's length result in the particular circumstances of a case (paragraphs 496 to 499)

If data from other countries is used, the application of the US. CPM may not properly reflect the circumstances of the market in which the enterprise is operating. Data that is not comparable can lead to inappropriate results (paragraph 500) .

Where the taxpayer in complying with the requirement of IRS Code 1.482-5(b)(3) develops an arm's length range in line with the IRS Code 1.482-1(e)(2) B and C, the excision of the upper and lower quartiles from the sample used to calculate the comparable profit tends to produce average outcomes. For some enterprises that are leaders in their industry segment, this method may produce figures which are not an accurate reflection of an arm's length outcome. Similar distortions could result in cases of less than average performance (paragraph 501) .

CPM calculations can undervalue certain types of functions and overvalue others where there is a high degree of sensitivity in the results to the data elements that have been selected (paragraph 502) .

CPM, or a substantially similar approach, may provide some help in very extreme cases where more reliable data is not able to be examined or does not exist (paragraphs 503 and 504) .

A global formulary apportionment method would allocate the global profits of an MNE group on a consolidated basis among the associated enterprises in different countries on the basis of a predetermined and mechanistic formula. There would be three essential components to applying a global formulary apportionment method: determining the unit to be taxed, i.e., which of the subsidiaries and branches of a MNE group should comprise the global taxable entity; accurately determining the global profits; and establishing the formula to be used to allocate the global profits of the unit. The formula would most likely be based on some combination of costs, assets, payroll, and sales (paragraph 505 and 506) .

The OECD member countries, including Australia, do not consider global formulary apportionment to be an acceptable alternative to the arm's length principle for a number of reasons. A principal reason is that global formulary apportionment can depart from the territorial connection that underpins the concept of source and may also raise issues about the timing of derivation. Equally important is the concern that predetermined formulas that are mechanistically applied do not have regard to the facts and circumstances and merits of the particular case - the result being that in many cases it results in either overtaxation or undertaxation. They also depend on a very high degree of international co-operation and coordination. The capacity for multinational groups to manipulate the formula and the inability of most formula to capture the particular circumstances of individual enterprises, their risks, geographical differences and differences in company efficiencies are serious drawbacks with this method. Also, currency exchange rate movements and inconsistent accounting standards between countries could lead to perverse profit allocations. Dispute over the acceptability and use of particular formulas which have different bases may mean that the expected benefits of no double taxation and lower compliance costs may not be realised (paragraph 507) .

In some cases, a formula developed by both tax authorities in cooperation with a specific enterprise after careful analysis of the particular facts and circumstances, such as might be used in an Advance Pricing Arrangement, would be appropriate to determine a fair allocation of revenue to the countries involved. However, these formulas are not instances of global formulary apportionment (paragraph 508) .

We strongly recommend that taxpayers and ATO staff adopt the following four step process for setting or reviewing transfer pricing for cross-boarder dealings between associated enterprises: 1. Develop an understanding of the cross-border dealings in the context of the taxpayer's business. 2. Select the methodology or methodologies. 3. Apply the methodology or methodologies. 4. Determine the arm's length outcome. Taxpayers should also implement a process that will support the chosen method(s), with a review mechanism to ensure an appropriate adjustment if material changes occur. They should also document the process. Some examples of the issues that arise are set out in paragraphs 509 - 591.

While other approaches may achieve reliable results, where taxpayers properly implement the four step process outlined they will be regarded as having taken reasonable care in relation to their transfer pricing for tax purposes.

The following points are made in relation to the process: (a) the four step process and the data collection and analysis outlined in this part are neither mandatory nor prescriptive approaches. The processes adopted for the review need to be tailored to the facts of the case. (b) for many enterprises which have relatively simple and/or low value international dealings with associated enterprises the extent of data collection and analysis may be minimal. For example, an enterprise may have an overseas subsidiary which conducts extensive business operations and deals in a wide range of goods and services. However, dealings between associated enterprises may be limited to the provision, by the parent, of a long term loan secured by the assets of the subsidiary. In these circumstances, the detailed issues and analysis contemplated in the following part will be largely irrelevant. Similarly, if a taxpayer has extensive dealings with associated enterprises but also has extensive dealings of the same kind and in similar circumstances with uncontrolled enterprises operating independently, a more limited analysis is sufficient. (c) it may be possible in some cases to adopt either a methodology or a specific price that has been developed and applied by a MNE on a global basis after some confirmatory analysis or consideration of its suitability and reliability in relation to the Australian enterprise. However, the data used to support the methodology will need to be carefully considered in terms of its relevance and reliability for Australian market conditions. (d) it needs to be remembered that associated enterprises will in many situations need to show that their association has not inappropriately impacted on the nature or terms of their dealings. Given the absence of the economic tension that exists between independent enterprises dealing wholly independently with each other, associated enterprises will often have to do analyses and keep records to show the arm's length nature of their dealings in circumstances where independent enterprises operating wholly independently could merely rely on their normal business records. This additional requirement cannot be removed without sacrificing the integrity of the arm's length principle and the underlying policy of the transfer pricing rules in Division 13 and Australia's DTAs (paragraph 513).

The taxpayer or ATO staff will need to understand the nature and extent of the dealings between the taxpayer and associated enterprises in the context of the taxpayer's business. It is important to be able to explain how the international related-party dealings of the enterprise are undertaken, the purpose or object of the dealings, what the taxpayer obtains from its participation in them and their significance to the taxpayer's overall business activities and those of the multinational group (paragraphs 514 to 516) .

It is useful to also identify relevant arm's length dealings of the taxpayer because it might be possible to use them as comparable uncontrolled dealings. Its dealings with associated and uncontrolled enterprises may be sufficiently similar in nature, frequency and size as to readily demonstrate that the dealings with associated enterprises are producing an arm's length outcome. Examples can be found at paragraphs 518 to 520. Many enterprises undertake a range of business activities or have a range of business lines. The key characteristics of these activities or business lines will need to be identified to enable the most appropriate method to be adopted in each case (paragraphs 517 and 521) .

In some cases the actual dealings can be different from the contracted terms. This can happen where the dealings extend over a long period and the parties modify their responsibilities but do not reflect these changes in the formal agreement. It would be important to determine the reasons why the original agreement was varied and whether the changes favour one or both parties and are in accordance with the reasons for the changes (paragraph 522) .

When examining the dealings it is also relevant to establish what systems, methods and procedures the taxpayer adopts for establishing the terms and conditions of dealings and whether they are applied uniformly in every case. Knowledge of how the dealings are conducted and the internal controls surrounding them can assist in gauging the likelihood that past dealings have been conducted in accordance with the arm's length principle (paragraphs 523 to 525) .

The formal organisational and capital structure of the enterprises that are parties to the dealings may also need to be reviewed. The corporate structure of the group would usually be established for both the formal lines of ownership, control, reporting and authority as well as the pattern of dealings between associated enterprises and how management performance is rewarded. Within particular enterprises in the group it may be necessary to review the organisational structure and decision making systems and processes. Examining these aspects can give an insight into the nature and purpose of the dealings between the taxpayer and other group companies and may indicate non-arm's length features of the relationships (paragraph 526) .

It may be useful to obtain information from a range of key managerial and supervisory staff to assist in obtaining an accurate perspective of the functions, assets, risks and operational aspects of the business (paragraph 527) .

It is also important to understand the nature of the industry and the markets within which the enterprise is conducting its business; the nature of competition experienced in its business dealings; and any broader economic and other factors affecting the taxpayer's business (paragraph 528) .

There is an expectation that the outcome from international dealings will reflect the conditions affecting the industry and the position of the enterprises within that industry. Changes in an industry should not be ignored. In general, in an arm's length situation it would be unlikely for one party to the dealings to be able to maintain its profitability regardless of changes in the industry, or the position of the enterprises within that industry, or the effect of government regulation on that industry (paragraph 529) .

Parent companies have entered into dealings with their subsidiaries that protect the parent's profit margin but have caused the subsidiaries to bear the full financial effects of changed industry or market conditions. Where this happens the wholesaler is forced to bear the cost of developing and maintaining the reseller network, but the end result is that the wholesaling subsidiary is relegated to accepting a residual profit margin or a loss. In these circumstances there are serious risks that the arm's length principle has not been complied with since the primary role of the subsidiary has become one of selling its parent's products and promoting its parent's brand name rather than making a profit in its own right. An independent enterprise operating wholly independently would seek to maximise the economic return from its functions, assets and risks. Overall, the wholesaling, marketing and distribution function is important to the MNE group and is often the source of its competitive advantage. This function should be rewarded on an arm's length basis (paragraphs 530 and 531) .

The effect of general economic factors, such as economic cycles, may mean that in order to evaluate or establish compliance with the arm's length principle it will be necessary to examine data for a number of years. It may be useful to collect data, where it is available, that reflects an entire business cycle. The length of the business cycle will of course be affected by conditions pertaining in the industry, such as the pace of technological change, and broader economic conditions (paragraph 532) .

The analysis of the functions, assets and risks of the associated enterprises engaged in the international dealings can take account of specific factors affecting the industry. However, care needs to be taken to identify and compensate for those decisions which artificially transfer risks between the related parties (paragraphs 533 and 534) .

Where comparability is difficult to assess or can only be approximated, it may be important to consider wider issues surrounding the dealings. This may include examining the circumstances surrounding the decision to enter into the dealings or, in some cases, how the property was dealt with in subsequent dealings (paragraph 535) .

An evaluation of the strategies of the taxpayer will also generally be necessary. The marketing and pricing strategies, the existence of relevant policies such as the provision of cross subsidies to parts of the business as well as any broader corporate objectives may need to be examined in order to understand the business context in which the enterprise operates. Information on the business strategies can assist in establishing the selection of methodologies and may be very important when addressing questions associated with comparability. An example can be found at paragraph 537 (paragraph 536) .

Non-arm's-length dealings may arise from a specific policy decision or a series of decisions of the board or senior management. For example, a subsidiary may have undertaken market development activities at its own expense and risk, and enhanced the value of a brand name owned by an associate which had no value prior to the market development activities of the enterprise. Senior management of the subsidiary may then agree to the payment of a royalty or management fee to a related foreign enterprise (paragraph 538) .

The payment of the royalty or management fee may significantly erode the profitability of the subsidiary. In evaluating whether the consideration conforms with the arm's length principle, it would be relevant to examine the decision making process of senior management or the board in arriving at the decision to agree to pay these fees and also incur market development expenses. It may be necessary to consider evidence as to whether the parties considered options realistically available to the enterprise. For example, in some situations it would be reasonable to conclude that an arm's length party would want its market development expenditure taken into account in the valuation of any royalty or by way of a reduced price for trading stock purchased from the owner of the brand name, and the consideration should be set accordingly (paragraph 539) .

Information on financial performance may be particularly important at a later stage if the methodology requires comparisons of the enterprise's performance over the relevant years or compared with other enterprises. The key ratios and statistics may vary depending upon the nature of the business being conducted. Usually, an application of methods (other than CUP) will require a comparison of the level of enterprise profit arising from dealings between associated enterprises with that achieved in its arm's length dealings or with the level of profit achieved by an uncontrolled enterprise (paragraph 540) .

This comparison is usually made in the form of some type of suitable accounting or statistical ratio analysis which will provide a basis to make the comparison. Such ratio analysis may include: (a) ratio of gross profit to operating expenses; (b) ratio of operating profit to sales; and (c) ratio of gross income / revenue to operating expenses (paragraph 541) .

What ratios, either from those mentioned above or others, is most appropriate needs to be established on the facts available. The use of ratios is discussed in relation to the profit comparison method at paragraphs 475 to 484 (paragraph 542) .

Trends would include general factors affecting the performance of an enterprise on a macro level, such as economic conditions as well as any significant features of the particular market or market segment within which an enterprise operates. Relevant trends at the enterprise level may include trends in gearing, dividend rate, non-performing assets and stock levels, as well as in other key financial ratios (paragraph 543) .

When considering trends as part of the ATO's overall approach, it would seem that those elements or factors which have a quantifiable impact on an enterprise's profit performance over time, or could reasonably have had an impact on pricing policy at the relevant time, should be taken into account. Projected trends and potential profit outcomes may be crucial in situations, such as APA's, and in those circumstances where taxpayers set the consideration on their dealings by reference to a profit split (paragraphs 544 to 547) .

In order to now select the most appropriate methodology or methodologies to use, the taxpayer may need to arrange the information that has been collected on its cross-border dealings with associated enterprises into an analysis of the: (a) the functions undertaken by each of the associated enterprises(including their nature and frequency); (b) the risks each of the parties assumed; and (c) the assets (both tangible and intangible) used by each of the parties and the nature and extent of that use (paragraph 548) .

This is sometimes referred to as a functional analysis. Some form of functional analysis will be necessary regardless of the methodology that has been selected (paragraphs 549 and 550) .

At its broadest level, such an analysis would result in the identification of categories such as manufacturing, wholesaling, financial services, mining, etc. However, such a broad description will not generally be sufficient (paragraph 551) .

It is essential to accurately establish the nature of the dealings of the enterprise (paragraphs 552 and 553) . For each of the main business activities of the enterprise, a detailed listing of the functions, assets and risks should be compiled. Such lists could be compiled for individual transactions, product or service lines, or for the enterprise as a whole. The analysis of function, assets and risks would be useful in: (a) determining the availability of comparables in relation to prices or functions; (b) assessing the degree of comparability with the functions, assets and risks in respect of the taxpayer's uncontrolled transactions or with those undertaken by other enterprises being considered as possible comparables; (c) assessing the relative weighting of the functions, assets and risks of each of the associated enterprises that are a party to the cross-border dealings in cases where an apportionment methodology, such as a profit split, is needed (paragraph 554) .

The compilation of such lists of functions, assets and risks, however detailed, does not in itself indicate which of the functions are the most significant, or economically the most important to the value added created by the business activities of the enterprise. A critical part of the analysis is to ascertain which are the most economically important functions, assets and risks and how these might be reflected in terms of an arm's length price, margin or profit on the dealings (paragraphs 555 and 556) .

It is generally not necessary to value each of the functions, assets and risks. The purpose of the examination is to understand the qualitative nature of the functions, assets and risks so that a comparison can be made with other enterprises that have similar functions, assets and risks. The allocation of actual income to assets may be far too difficult a task, and is likely to lead to undue complexities in the analysis. The value of some assets are not easily measured, notably intangible assets. In fact, if taken to extreme levels it could lead to an examination that becomes absurd. Many factors will simply be assessed as part of the business risks and comparisons made at that level. It needs to be remembered that the various methodologies work on the basis that the analysis is capable of being able to produce a quantifiable result. This can mean that it is necessary for some factors that cannot be quantified to be taken into account in some indirect way. For many cases, particularly where the international related-party dealings are relatively straightforward, it is desirable to avoid overly complex analyses (paragraphs 557 and 558) .

The reliability of the available material should be carefully evaluated to ensure it is capable of being used in practical manner (paragraphs 559 to 561) .

The information initially collected may include data that enables the calculation of gross profit ratios or ratio of gross income to operating expenses, or it may include market share data, etc. This might be sufficient to identify whether the taxpayer's case raises transfer pricing issues, and perhaps whether issues arise in relation to items above or below the gross profit line. However, in order to evaluate the proper return for the economically significant functions, assets and risks, other data may need to be collected to identify problem areas more clearly (paragraphs 562 and 563) .

The collection of further data should be done on a selective basis to identify the important value adding activities of the enterprise and to get a good sense of their relevant importance to the taxpayer's income earning activities (paragraph 564) .

In some cases the taxpayer or the ATO may select more than one appropriate methodology in order to either encompass the full range of its international dealings with associated enterprises or to obtain greater certainty that the primary method selected in fact produces reliable arm's length results (paragraphs 565 to 567) .

In some circumstances it may be possible to apply a particular method to only part of the relevant dealings of a taxpayer. In this situation, care will be needed to ensure the methodology is being legitimately applied. Sometimes an enterprise can have a mix of methods successfully applied to its dealings. However, 'cherrypicking', the selective application of a methodology to a limited range of dealings, can produce commercially absurd outcomes. In some situation it may be necessary to ultimately select a different method (or mix of methods)in preference because it has a wider application and produces a closer approximation of an arm's length result for all of the international dealings with associated enterprises. An example of the issues that can arise and their implications is set out in paragraph 568.

In Step 3 the taxpayer or the ATO will need to apply the chosen methodology using the information that has now been identified. This should be done in a way that tests the appropriateness of the selected methodology or methodologies and confirms its suitability. If this cannot be confirmed the taxpayer or the ATO will need to consider other methodologies, either alone or as a support for the method(s) initially selected. An example of this process is included in paragraph 571 (paragraphs 569, 570 and 572) .

The preliminary functional analysis that was prepared to select a methodology can now be extended. If a comparability methodology involving external benchmarking with independent enterprises is being used, the functional analysis assists in determining the comparability of the dealings or the enterprise with uncontrolled dealings undertaken by the independent parties. It is not necessary to value the functions, assets and risks of each of the enterprises since the main intention is to establish the degree of comparability. However, it is essential to ensure that where there are differences in the significance of the functions, assets and risks to each of the businesses that these differences are taken into account. An example of this issue is set out at paragraph 574 (paragraph 573) .

If an apportionment methodology, such as a profit split, is being used, it is also not essential to value each of the functions, assets and risks. The intention of an apportionment methodology is to establish the relative importance of the functions, assets and risks of the parties to the international dealings so that an arm's length apportionment of the consideration in the dealings (such as a profit split) can be undertaken (paragraph 575) .

Some of the functions, assets and risks may be shared between associated enterprises engaged in the international dealings. For example, research and development may be undertaken by both parties in a cost contribution basis. It will be necessary to recognise that the return to the research and development activity is not the sole property of one of the parties, and will need to be split in accordance with the relative contributions of each of the parties (paragraph 576) .

The functional analysis can be performed with varying levels of detail and can serve a variety of purposes. The analysis may be applied on a product or divisional basis for individual transactions, or it could be applied up to a corporate group basis. The scope of the analysis will be determined by the nature, value and complexity of the matters covered by international dealings and the nature of taxpayer's business activities, including the strategies which the enterprise pursues and the features of its products or services. It is difficult to set out guidelines here that are suitable for every possible type of enterprise. However, as has been noted, the material that is produced can be of assistance to an enterprise in developing documentation to support its views (paragraph 577) .

It is often important in evaluating the relative significance of the functions to analyse the staffing of the enterprise. Experienced and highly trained staff may be an important intangible assett of the company, and these staff may be undertaking essential functions which are generating considerable value added (paragraphs 578 and 579) .

Some of the data may need to be refined or adjusted to improve comparability. This may be particularly important in those cases where the enterprise is engaged in strategies (special conditions) which affect the arm's length consideration. It is possible that in some circumstances the practical implementation of the methodology may prove to be highly problematical. In this situation the selection of the particular methodology may need to be reconsidered and its reliability considered against the reliability of other methods that could be applied on the available data (paragraphs 580 and 581) .

Clearly the extent of analysis at this stage will vary depending on a range of factors, such as data availability, the extent and level of the international dealings and their importance to the enterprise's business. The methodology selected must be capable of practical application and must produce a result that is a reasonable approximation of what would result if the dealings were undertaken on an arm's length basis (paragraph 582) .

At this point, the data may suggest that there is a range of arm's length outcomes that might be applied to the dealings. If this is the case, the taxpayer or its advisers will need to evaluate the data to create an appropriate range that it believes reflects the range of outcomes that might reasonably be expected to occur if the dealings were undertaken on an arm's length basis (paragraph 583) .

In this step, the taxpayer will need to apply the data to the actual dealings to demonstrate how the outcomes conforms to the arm's length principle. This would be the way the ATO would check the taxpayer's dealings. How well this application can be done of course is highly dependent on the availability of relevant data, its reliability and the available documentation, either internal or external. If data availability is a problem, arguments will need to be developed to support the results achieved from the analysis, and a more indirect measurement methodology possibly may be needed (paragraphs 584 and 585) .

The process to date can deliver to the taxpayer an objective, documented and considered review of the available material and possible choices for arriving at an arm's length outcome. The determination of the arm's length consideration can now be made. This may be reasonably straightforward where the methodology has produced a high level of comparability - though judgment is still required - or it may require a greater degree of judgment where there is a range of results or a choice of answers is needed (paragraph 586) .

The ongoing reliability of a taxpayer's transfer pricing method(s) will depend on whether the necessary systems and procedures are put in place to collect the relevant data and ensure the proper analysis for tax purposes (paragraph 587) .

Where a methodology is being used on a continuing basis, the choice of methodology that has been made and the data that is used to establish the consideration need to remain valid (paragraph 588) .

The process of selection and application of a methodology is a dynamic exercise. How long a process or a method arising from the process remains appropriate before it ceases to give an accurate outcome will depend upon the nature and extent of the changes that occur to the factors that were initially taken into account (paragraph 589) .

To ensure that the processes, methods and data used continue to be appropriate, the taxpayer should incorporate reviews of these matters into its internal procedures (paragraph 590) .

Even where a methodology is first selected and applied, the results should be checked to ensure that they are commercially realistic having regard to the taxpayer's facts and circumstances. Where there is some doubt about the reliability of the outcome the earlier steps in the four step process should be revisited and the results checked against other approaches/indicators that may be available. It will be necessary to consider the relative reliability of another methodology using the known information. If this is not possible or practicable, the taxpayer should consider an APA with the ATO (paragraph 591) .

Australia has endorsed the arm's length principle as the general benchmark for transfer pricing. This endorsement is reflected in our DTAs and Division 13. This principle is adopted by all the member countries of the Organisation For Economic Co-Operation and Development ('OECD') and many non-members who have followed the OECD model in negotiating DTAs.

The arm's length principle is contained in each of Australia's DTAs through the Associated Enterprises Articles through the notion of 'independent parties dealing wholly independently with one another'. This is used as a benchmark against which to evaluate the 'commercial [and] financial relations' between associated enterprises (this categorisation in the DTAs being confined to cases where, broadly there is common management, control or capital ownership whether direct or indirect). These agreements are, in turn, incorporated into Australia's domestic law as schedules to the International Tax Agreements Act 1953.

Division 13 of the ITAA incorporates the arm's length principle in paragraphs 136AA(3)(c) and (d) and subsection 136AD(1) to (4) through the concept of the 'arms length consideration'. In the Explanatory Memorandum accompanying the introduction of the amended Division into the ITAA, it is stated: 'the Commissioner will be required to redetermine the taxpayer's assessable income or allowable deductions basically by using the internationally accepted "arm's length" principle ... The arm's length principle is also at the base of provisions of each of Australia's comprehensive double taxation agreements that enable the determination of profits attributable to business activities in one or other of the countries concerned' (page 4 of the EM).

Paragraph 136AA(3)(c) and (d) define arm's length consideration as: '...[arm's length consideration] ... the consideration that might reasonably be expected to have been received or receivable in respect of the supply of property (which includes services) - or given or agreed to be given in respect of an acquisition of property, if the property had been supplied or acquired as the case may be under an agreement between independent parties dealing at arm's length with each other in respect of the supply or acquisition' ( page 69 of the EM).

In reflecting on the potential limitations represented by the definition of 'arm's length consideration' found in paragraphs 136AA(3)(c) and (d), Parliament indicated that a situation in which a deeming under section 136AD(4) may be appropriate would be where: '... for example, the industry is so controlled and structured that there are no comparable arm's length dealings in relation to property of the same kind, or there are no comparable dealings in the same quantities as that supplied or acquired under the agreement.'

Another area where Parliament saw an application for subsection 136AD(4) was in cases where comparable dealings may exist but information about them is withheld or, for what ever reason, information about them is unobtainable by the Commissioner. Australia's DTAs and Division 13 also recognise the modified applications of the arm's length principle in cases involving highly differentiated or unique goods and services and intangibles. This extends its application in cases where markets have a very high proportion of controlled international dealings and a corresponding absence of independent benchmarks. Both these situations occur in the Australian economic scene and the flexible approach endorsed in the scheme of the legislation reflects Parliament's intention that the arm's length principle be given a practical application.

Division 13 and Australia's DTAs do not prescribe any particular methodology or preference for the order in which methodologies might be applied to arrive at an arm's length outcome. In Case No N69 (13 TBRD 270; 11 CTBR (NS) Case 53) the Board of Review said of the predecessor to Division 13, 'The goal to be aimed at in determining what "amount of the total receipts" shall constitute the taxable income is the amount which one would expect to arise as the taxable income under the general scheme of the Act if the business were independent of foreign control or share ownership. This conclusion states the objective to be achieved from the application of s.136, but the section puts no limitation on the methods to be used or on the considerations to be taken into account in determining the appropriate "amount of total receipts" so long as those methods or considerations do not give results beyond the statutory objective' (13 TBRD at 279; 11 CTBR (NS) at 274).

Within the context of definition of 'arm's length consideration' in Division 13 and the DTAs notion of 'independent parties dealing wholly independently with one another', it appears that Parliament intended to give the Commissioner the greatest possible scope to use methodologies appropriate to the circumstances. This is clear in the context of Division 13 in the statements made at page 63 of the EM where it is said: 'There are a number of methods ... Which of these or other methods (emphasis added) might appropriately be adopted, and the way in which it is applied, will depend on all the circumstances.'

This is a conscious decision by legislature which intended the arm's length principle to be the cornerstone of the law yet still allowing the Commissioner flexibility to administer the law especially in relation to the selection and application of methodologies, though clearly limited by the statutory purpose.

We agree that the statutory objective provides a guideline and limitation on the methodologies that are available under Division 13 and the DTAs - and in relation to how methodologies should be applied. The purpose of Division 13 is set out in paragraphs 10 and 13 and 154 to 157 of Taxation Ruling TR 94/14. Its statutory objective, in the sense used by the Board of Review in Case N69 , is to enable the Commissioner to determine an amount of consideration in respect of an international agreement that would have been set if the dealing had occurred on an arm's length basis in cases where - judged against what independent parties dealing at arm's length might reasonably be expected to have received or paid in the taxpayer's circumstances - a taxpayer has received inadequate or no consideration for property (including services) supplied, or pays too much for purchases - and to use the amount so determined in the calculation of the taxpayer's taxable income. It has a further objective of allowing the Commissioner to estimate the arm's length consideration where for any reason (including an insufficiency of information available to the Commissioner) it is not possible or practicable for the Commissioner to ascertain the arm's length consideration.

The Associated Enterprise Articles in Australia's DTAs have the objective of allowing the Commissioner to adjust understatements of profits in cases where an enterprise resident in Australia and an enterprise resident in a country with which Australia has a DTA are under common management, control or ownership (whether direct or indirect) and have not dealt with each other on an arm's length basis. The mechanism provided in the DTAs is the discretion to increase a taxpayer's profit to correct a misallocation to the other contracting state. Most of the DTAs have the additional objective of permitting the Commissioner to grant correlative relief in cases where another country makes an adjustment consistent with the Article.

Where complex issues arise that require specialist approaches in areas where there is no guidance as to what was intended by way of a method to achieve a stated objective, it is appropriate in determining relevant principles and criteria, as courts and tribunals have done, to have regard to the industry practice and any standards relevant to the issue. For example, tax payable on the value of property would, in the absence of statutory rules, raise questions of how the property should be valued and courts would have regard to relevant valuation and practice. In situations where it is necessary to allocate costs amongst various activities courts have regard to accepted accounting practice and any relevant standards. (See, for example, the statement of Kitto J in B P Refinery (Kwinana) Ltd v. FC of T [1961] ALR 52; 12 ATD 204 referred to in paragraph 412 below). It would be appropriate in the context of statutory transfer pricing rules to consider the internationally agreed approaches as to what would generally be accepted as a reasonable way to determine the arm's length outcome, having regard to the intent of the transfer pricing provisions in Division 13 and the DTAs and the actual wording of the Australian provisions. (See for example the reference to international tax literature by the Board of Review in Case N69 ).

When applying Division 13 and the DTAs we will pay close attention to the OECD guidelines on transfer pricing methodologies and the operation of the Associated Enterprises Article of the OECD Model, being the considered view of many tax experts familiar with transfer pricing. However, they are not an interpretation of Division 13 which must be construed according to its terms and purpose. In relation to the application of the DTAs, it needs to be recognised that OECD Reports do not have as high a status in international law as an aid to interpretation as a document evidencing the intention of the Contracting States or the Commentary to the OECD Model Tax Convention. Nevertheless, the ' Report on Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrators ' that was approved by the OECD Council on 13 July 1995 ('the 1995 OECD Report') is seen as an important, influential document that reflects unanimous agreement amongst the member countries - an agreement that was achieved after an extensive process of consultation with industry and tax practitioners in member countries - and should be followed where relevant and in the absence of any intention to the contrary in Division 13 or the DTAs.

Paragraphs 88 - 100 of Taxation Ruling TR 94/14 set out the methodologies that would be acceptable to the ATO for determining arm's length outcomes. Those paragraphs need to be read in the light of the following discussion on subsection 136AD(4).

It is accepted that the CUP method will provide the best reflection of an arm's length outcome where there is sufficient reliable data for its application. This method should not be routinely discarded. A flexible approach should be adopted to allow adjusted CUP analyses to be used in appropriate cases. It may be that in a given case an adjusted CUP analysis produces a more reliable result than any other method. This method is discussed in further detail below.

It is also accepted that the Resale Price (RP) and Cost Plus (CP) methods can be used in appropriate cases, subject to the general principles in paragraph 87 of TR 94/14. The RP and CP methods are discussed in detail below.

As stated in paragraph 100 of TR 94/14 the ATO will accept the use of a mixture, RP and CP methods or a profit split or profit comparison method in certain circumstances. The profit methods are also discussed below.

It is important to note that the transfer pricing rules in Division 13 and the profit reallocation rule in all of Australia's DTAs allow the Commissioner to use Division 13 to approximate an arm's length consideration in cases where the information available is inadequate to determine the income to be attributed to an enterprise - provided that when Division 13 is applied in conjunction with a DTA it is applied, so far as it is practicable to do so, consistently with the principles of the relevant Associated Enterprises Article of the DTA. Australia's position is expressed in a reservation to the OECD Model Tax Convention which is recorded in paragraph 18 of the OECD Commentary in the following terms: 'In negotiating conventions with other Member countries, Australia and New Zealand reserve the right to propose a provision to the effect that, if the information available to the competent authority of a Contracting State is inadequate to determine the profits to be attributed to an enterprise, the competent authority may apply to that enterprise for that purpose the provisions of the taxation law of that State, subject to the qualification that such law will be applied, as far as the information available to the competent authority permits, in accordance with the principles of this Article.'

Despite minor wording changes the effect of Australia's reservation is seen in the Associated Enterprises Articles of all our treaties. As discussed above, the result of this approach by Australia is to secure paramount application of the domestic law in cases of insufficient information, subject to the proviso referred to in paragraph 253 above. The relevant domestic law provision in such cases is subsection 136AD(4).

It could be argued from the context of its operation and its wording that the Commissioner would be using a non-arm's-length methodology when applying subsection 136AD(4). Differences in Australia's DTAs from the OECD Model and the enactment of subsection 136AD(4) reflect Parliament's intention to introduce provisions, which, while being as consistent as possible with the arm's length principle, allow the Commissioner to go beyond the strict confines of paragraph 136AA(3) (and the arm's length principle as found in the treaties) to deem an arm's length consideration.

However, having regard to the clear policy expressed in subsections 136AD(1) to (4) to use the arm's length principle and the fact that subsection 136AD(4) enables the deeming of the amount of the arm's length consideration which is then used in the application of subsection 136AD(1), (2) or (3) as appropriate, subsection 136AD(4) must be applied in a way that achieves the closest practicable estimate of an arm's length result. (See also paragraphs 82, 83 and 338 to 340 of TR 94/14).

This does not mean that the arm's length principle is being ignored. Rather, the Australia law is designed to reflect the Australian experience in this area and to give the Commissioner sufficient flexibility to arrive at an answer in the wide range of cases that are likely to be encountered in practice (see paragraphs 240, 241 and 253 to 255 above).

The aim in cases where subsection 136AD(4) applies will be to achieve as highly focused a comparison as is possible in the circumstances, consistent with the intention of Parliament as expressed in Section 136AD as a whole, the EM introducing this provision into Division 13 and with the Associated Enterprises Articles in all of Australia's DTAs. The operation of subsection 136AD(4) is discussed in detail in TR 94/14.

A similar limitation in terms of statutory purpose applies when the application of subsection 136AD(4) is authorised by a DTA because all of Australia's DTAs require the subsection to be applied, in so far as it is practicable to do so, consistently with the arm's length principle embodied in the Associated Enterprises Article.

It follows from the nature and purpose of subsection 136AD(4) that other methodologies may be used under that subsection than would be available under subsection 136AD(1) - (3) and Article 9 (the Associated Enterprises Article) of the OECD Model Tax Convention. These methods will depend on the facts and circumstances of each case but could include income and expense allocation on the basis of a formula, a return an assets method, a mixture of methods, or some form of profit comparison other than the profit split and profit comparison methods described later in this draft Ruling.

The selection and applicability of methodologies in the context of Division 13 was discussed in paragraphs 86 to 100 of TR 94/14 and regard should be had to the general principles expressed therein. Those principles are relevant to DTAs and should also be applied in that context.

It bears repeating that the most appropriate method in a given case will depend on the facts and circumstances of the case and the extent and reliability of data on which to base a comparability analysis, the intention always being to select the method that produces the highest degree of comparability. In cases where there are no comparables or there is insufficient information to determine the arm's length outcome, the method to be used should be a method that produces a reasonable estimate of an arm's length outcome on the basis of what is known in the case.

Having regard to the statutory objective of Australia's transfer pricing rules, the ATO takes the view that any transfer pricing methodology used to calculate an arm's length consideration in international dealings between associated enterprises must be applied in a way that will provide an arm's length outcome by closely reflecting commercial and economic reality and the economic contribution made by the enterprises in each jurisdiction.

The ATO does not see this Ruling putting into question internationally accepted methodologies that are used to determine an arm's length consideration. On the contrary, the ATO endorses CUP, RP, CP, profit split and profit comparison methods as the relevant approaches or criteria, the most appropriate of these depending on the nature of the case and the extent of reliable data to enable its proper application. Nor is the ATO expressing any preference for particular methodologies since their suitability and reliability will depend on the facts and circumstances of each case. However, an understanding of the commercial and economic reality underlying any particular transaction or dealing will be reached by beginning with a search for and a close examination of comparable transactions or dealings between unrelated enterprises in an application of the traditional arm's length methodologies. If such comparables can be found and the resulting prices or terms would be acceptable to each of the parties concerned then a basis for an arm's length determination may have been reached. If not, then profit methods should be considered. In particular, the ATO agrees with the OECD view that profit methods (sometimes referred to as transactional profit methods) are methods of last resort. In discussing this the OECD said in their 1995 report ' Transfer Pricing Guidelines For Multinational Enterprises And Tax Administrations ' (the 1995 OECD Report): '3.50. There are, however, cases where traditional methods cannot be reliably applied alone or exceptionally cannot be applied at all. These would be considered cases of last resort. Such cases arise only where there is insufficient data on uncontrolled transactions (possibly because of unco-operative behaviour on the part of the taxpayer relative to these guidelines), or where such data is considered unreliable, or due to the nature of the business situation.' In this regard see paragraphs 349 and 366 of Taxation Ruling TR 94/14.

We have discussed the arm's length principle in previous Rulings. (See paragraphs 54 and 64 to 66 of TR 94/14 and the explanations thereto and paragraph 20 of TR 92/11).

Conceptually, the arm's length principle requires a conclusion (and in the case of Division 13, a determination) of what might reasonably be expected if the parties were dealing at arm's length with one another. It does this by comparing what the taxpayer has done with notions of supply, demand and negotiation in an open market and uses the behaviour of independent parties dealing at arm's length as a benchmark. The notion of comparability is therefore central to the arm's length principle. The nature of this comparison with arm's length activity means that absolute precision and certainty is very difficult to achieve. There is an expectation, however, that taxpayers and ATO staff alike will make best endeavours in determining what might reasonably be expected if parties dealt at arm's length with one another, and the methodologies that have been developed are intended to systematically test a taxpayer's activity against that benchmark parties.

Implicit in the concept of 'the arm's length principle' and of the expressions 'arm's length consideration' in Division 13 and 'independent parties dealing wholly independently with one another' in Australia's DTAs is the notion that independent parties who are dealing at arm's length would each compare the options realistically available to them and seek to maximise the overall value of their respective entities from the economic resources available to or obtainable by them (paragraph 66 of TR 94/14). A statement to similar effect appears in paragraph 1.16 of the 1995 OECD Report where it is stated: 'All methods that apply the arm's length principle can be tied to the concept that independent enterprises consider the options available to them and in comparing one option to another they consider any differences between the options that would significantly affect their value. For instance, before purchasing a product at a given price, independent enterprises normally would be expected to consider whether they could buy the same product at a lower price from another party' (paragraph 1.16 of the 1995 OECD Report).

The issue of choice is important, because in most situations the question being asked by the arm's length principle is, 'What would have happened if the ownership link had been severed and the enterprise was motivated by its own economic interest?' This approach will involve a consideration of what a reasonable, independent business person might reasonably be expected to agree to in the same or similar circumstances.

If an open market exists that sets prices, or more generally contractual terms, profits and risk levels then this will provide the benchmark for a proposed transaction and it would not be expected that a seller would accept less or a buyer pay more than the open market price (bearing in mind that this could be a range of prices). In this sense, the search for comparable uncontrolled prices is a search for an open market based alternative. The cost plus method and the resale price method options can also be seen in this context as defining market based margins for functions performed (including assets used and risks assumed) and it could be argued that a reasonable prudent decision maker would look to open market in assessing the available alternatives. In this way the cost plus and resale price methods are also special cases that derive their validity from the fact that where they can be reliably applied they define reasonable courses of action by an uncontrolled enterprise which can be used as an arm's length benchmark for the taxpayer's dealings. The profit based methods are last resort options that identify reasonable alternatives when the more direct methods based on transactional comparability are not suitable or practicable. While the ATO has a preference for more direct methods based on transactional comparability, we recognise the overriding need to ensure that assessments reflect commercial and economic reality as well as the economic contribution made by each of the enterprises involved.

It is clear that where an open market exists from which one or more comparables can be inferred, the comparable will determine the options open to the enterprise in relation to how its dealings should be structured for tax purposes to accord with the arm's length principle. However, an open market is less likely to exist where relationships dominate. In some markets discrete transactions between a variety of buyers and sellers are no longer widely applicable. Rather the dealings are more likely to be long term in nature, and confined to a few players. In this situation there is a greater emphasis on strategic relationships rather than transactions, where buyers and sellers are bound to each other through numerous exchanges and linkages, including the exchange of goods and services, finance, technology and know-how. These exchanges are inevitably adapted to the specific needs of the buyer or seller, to suit integrated production or marketing processes, organisational structure and strategies. This is common in the dealings of multinational enterprises (MNEs) that are members of the same group, though it should not be automatically assumed that the members of multinational groups are not dealing at arm's length with each other.

The decision to embark on a relationship is one that is profit driven both in the long and short term, with expectations that will be shaped by the experience of similar enterprises under similar circumstances. Whether or not the choice made is acceptable is best tested by looking to the options that can reasonably be held to exist. The starting point is therefore to consider the profit expectations at the time when a relationship is initiated, and what a reasonable business person would see as the critical assumptions at that time and how the relationships could reasonably be expected to develop should the conditions supporting the critical assumptions materially change.

The arm's length principle has wide application and it should be considered in all international commercial and financial dealings between associated enterprises. Historically, tax enforcement in this area has centred on the determination of the arm's length outcome for transfers of tangible goods. Increasingly, the arm's length principle is being applied in examinations of the supply or acquisition of a wide range of services, and in the calculation of market rewards for the right to exploit intellectual or intangible property. Many cases involve a number of these elements.

We have also experienced situations where taxpayers have tried to distinguish their own facts and circumstances to the extent that they claim that it is not possible to determine an arm's length outcome. For example, some taxpayers have claimed that because their products or services are unique or highly differentiated, or because of the uniqueness of the organisational structure, management or synergies of their enterprise, the arm's length principle and analysis based on comparables is not applicable.

This is a misconception of the arm's length principle which has to be applied flexibly. It does not require the taxpayer's situation to be identical with independent dealings used for comparisons. It requires sufficient similarity between the taxpayer's circumstances and those being used as comparisons such that a reasonable business person in place of the taxpayer would regard those other cases, making appropriate allowances for material differences, as reliable benchmarks for the taxpayer's activities. Accordingly, there should be comparables for most cases. The question of whether the taxpayer can get access to those comparables is discussed later. However, there will be situations where direct comparisons are impossible. There are other cases where the transactions are of a kind that would occur only between related parties. The arm's length principle still applies, but the focus is on functions, assets, and risks and the processes that parties dealing at arm's length would have adopted to allocate profits based on benchmark rates of return and economic weightings in comparable circumstances, using the concept of a joint venture between independent parties as a guide.

Parliament intended that the Commissioner still be empowered to apply the law in cases involving unique or highly differentiated dealings to ensure that Australia receives its fair share of tax. To that end, subsection 136AD(4) empowers the Commissioner to approximate an arm's length outcome where there is a lack of information that makes the determination of the arm's length consideration not possible or practicable. This residual power is preserved in all of Australia's DTAs.

The subject matter of the comparison and the level at which this occurs together with the types of methodologies which may be used where there is an insufficiency of information, are issues which are discussed later in this Ruling.

Ideally, the arm's length principle should be applied to each international dealing in order to arrive at an arm's length consideration. However, in some cases this will not be feasible because of the absence of reliable data on which to assess comparability.

In practice, there are two fundamental approaches that may be taken to achieving an arm's length outcome. These are the use of analysis based on comparable uncontrolled dealings and other methods where there are no comparable uncontrolled dealings. The ATO considers that analysis based on reliable comparable uncontrolled dealings should be used to determine an arm's length consideration where reliable information is available to do so. As we discuss later in the Ruling, there are some circumstances where it will be necessary to basket or aggregate transactions in order to achieve this end. However, where there are no comparable uncontrolled dealings or there is insufficient reliable data to identify them it will be necessary to use other methods to estimate or approximate an arm's length outcome.

The preferred arm's length methodologies are based on the concept of comparing the prices / margins achieved by associated enterprises in their dealings to those achieved by independent enterprises for the same or similar dealings. As there are many matters that may influence price / margins there is a need to closely examine the dealings being compared. The OECD provides the following commentary concerning this matter: '1.15. Application of the arm's length principle is generally based on a comparison of the conditions in a controlled transaction with the conditions in transactions between independent enterprises. In order for such comparisons to be useful, the economically relevant characteristics of the situations being compared must be sufficiently comparable. To be comparable means that none of the differences (if any) between the situations being compared could materially affect the condition being examined in the methodology (e.g., price or margin), or that reasonably accurate adjustments can be made to eliminate the effect of any such differences. In determining the degree of comparability, including what adjustments are necessary to establish it, an understanding of how unrelated companies evaluate potential transactions is required. Independent enterprises, when evaluating the terms of a potential transaction, will compare the transaction to the other options realistically available to them, and they will only enter into the transaction if they see no alternative that is clearly more attractive. For example, one enterprise is unlikely to accept a price offered for its product by an independent enterprise if it knows that other potential customers are willing to pay more under similar conditions. This point is relevant to the question of comparability, since independent enterprises would generally take into account any economically relevant differences between the options realistically available to them (such as differences in the level of risk or other comparability factors discussed below) when valuing those options. Therefore, when making the comparisons entailed by application of the arm's length principle, tax administrations [and we would add taxpayers] should also take these differences into account when establishing whether there is comparability between the situations being compared and what adjustments may be necessary to achieve comparability' (paragraph 1.15 of the 1995 OECD Report).

The objective in relation to comparability is always to seek the highest practicable degree of comparability, recognising though that there will be unique situations (which could be a result of business complexity) and cases involving valuable intangibles where it is not practicable to apply methods based on a high degree of direct comparability (see subparagraph 87(e) of TR 94/14).

The standard of comparability that is practicable will be determined by the extent of reliable data on which to make comparisons with uncontrolled situations and dealings in the particular case. Comparisons with controlled dealings by other taxpayers cannot be regarded as arm's length comparisons.

With the many influences and competing priorities imposed on enterprises it is not always possible to find dealings that are exactly comparable to the dealings being examined. The question then becomes what matters need to be considered to determine if dealings are sufficiently comparable to be used in determining arm's length prices and margins. Adjustments need to be made for any material differences so that the dealings can be as similar as is possible. If suitable adjustments cannot be made then the dealings cannot be considered to be comparable. Commonly, the use of methods other than the traditional transaction methods will produce results that without careful analysis and adjustment would not be sufficiently reliable to demonstrate the levels of comparability necessary to form an arm's length range. This is a particular issue where operating expenses are taken into account and the comparison is done at or approaching the net profit level. In these situations and in others where the level of comparability is affected, notwithstanding that adjustments to achieve true comparability cannot be made, the data should not be completely discarded as, in the absence of any true comparables, it may, in conjunction with other methods, provide insight as to what constitutes a reasonable approximation of an arm's length outcome.

Some dealings between taxpayers and unassociated enterprises may not be able to be accepted as reliable comparables because they may not be made in the ordinary course of business. An example would be a relatively insignificant sale made at the same price as charged to associated enterprises in order to create an internal comparable to justify the pricing to associated enterprises, but which, by open market standards required by the arm's length principle, was concessional to the unassociated enterprise.

It needs to be remembered that the ATO, when applying any method, may have more information available than a taxpayer has or can have reasonable access to through its own efforts. This data should be used where it enables a more reliable determination of the arm's length outcome, by producing a higher degree of comparability, though appropriate steps should be taken, subject to the need to protect the confidentiality of other taxpayers, to allow the taxpayer an adequate opportunity to defend its own position and generally safeguard the taxpayers' rights to have ATO decisions reviewed by the Administrative Appeals Tribunal ('AAT') or a court. In this regard, the ability of the AAT and courts to hear evidence on a confidential basis in a closed hearing may be relevant. Not to use the more reliable information would undermine the statutory objective of the arm's length principle as expressed in Australia's legislation (discussed in paragraphs 238 - 247 above).

When comparing the margins, mark-ups and particularly price paid for goods or services (property) the characteristics of the property will often account, in some way, for the differences in their value in the open market. Characteristics that may need to be considered include the following: (a) tangible property: the physical features, quality, reliability, availability and the volume of supply; (b) services: the nature and extent of the services whether provided separately or bundled with tangible property; and (c) intangible property: the form and substance of the transaction in terms of the nature and extent of the rights transferred or contributed.

When comparing price and particularly margins or mark-ups between controlled and uncontrolled dealings, the extent of the combined value added by the each of the associated enterprises relative to the value added by independent parties in their uncontrolled dealings will often account for the differences in the value of the goods or services in the open market. The contributions to the value added can be determined by an analysis of the functions performed, assets used and risks assumed. This analysis which can be performed at a series of different levels from a qualitative initial assessment to a more detailed analysis involving economic weightings, is generally referred to as a functional analysis.

To properly address comparability it is essential to analyse the functions performed, assets used and risks assumed by the taxpayer to identify the economically significant activities and to compare the results with a similar analysis of uncontrolled dealings or of uncontrolled enterprises that are being considered as possible comparables. The level of functional analysis that is needed will depend on the facts of the case.

A functional analysis will identify the economically significant activities that are undertaken (functions performed, assets used and risks assumed) which will assist in the selection of a methodology and in evaluating comparables.

Prior to undertaking a functional analysis, it is necessary to obtain a broad understanding of the enterprise and the business it conducts. This would extend to understanding the overall relationship it has with other members of the group and the relative contributions of each of the parties to the economic value that has been created.

A functional analysis can be undertaken on a single taxpayer with both controlled and uncontrolled dealings as a basis for assisting a comparison between an associated enterprise with controlled dealings and an uncontrolled enterprise. However, it may need to be undertaken differently depending on the focus of the transfer pricing examination. Once a methodology has been selected on the basis of a functional analysis, continuing importance of the functional analysis is: (a) to further refine the examination of comparability on property and services, market terms and conditions and the presence of any special circumstances where transactional methodologies are employed; and (b) to better understand the economically significant activities identified in the first stage (functions performed, assets used and risks assumed) which are then used to refine the analysis of independent comparable transactions, business activities or enterprises in comparable markets (i.e., external profit comparisons between companies).

If identical uncontrolled dealings could be identified, the consideration, whether price or some other outcome, in such dealings is the most reliable point of comparison. In practice, identifying an identical situation can be difficult and most of the methodologies are based on a comparison with the most similar or comparable set of circumstances.

Simple cases, for example where the only dealing with an associated enterprise is a loan, not raising an issue of quasi equity, will not require a functional analysis. Where a taxpayer has significant dealings with uncontrolled parties and these are comparable to its dealings with associated enterprises and conducted on the same basis, a limited form of functional analysis is sufficient (see paragraphs 509 to 513 below).

When independent enterprises negotiate contracts or agreements the ultimate price / margin agreed is influenced by the terms and conditions of the proposed agreement. Terms and conditions that may influence the agreed price / margin include: credit and payment terms, volume, duration, product and service liabilities of the parties, warranties and exchange risk, and these matters will need to be taken into account when making any comparison.

When considering an agreement between associated enterprises there will be a need to have regard not only to the terms of the agreement but also to the actual conduct of the parties. The ATO concurs with the view of the OECD that where the economic substance of a dealing differs from its form, regard must be given to the conduct of the parties in determining what are the actual terms of the contract. In this regard see paragraphs 1.36 to 1.41 of the 1995 OECD Report.

The functional analysis should identify the risks associated with the combined dealings of the related parties and those of the comparables. As with any dealings the final price / margins agreed will be influenced by the sharing of the associated risks amongst the parties. These risks may include: (a) research and development risk; (b) product liability risk / professional indemnity risk; (c) financial risk: forex, interest rates, etc.; (d) market risk: cost fluctuations, the level of supply and demand, pricing and inventory levels, alternatives / substitutes, nature and degree of competition; (e) credit and collection risks; and (f) general business risks relating to business activities and the ownership of property, plant and equipment.

Therefore adjustments will need to be made for any material differences in the risks assumed by the associated enterprises.

As the nature of the MNE is that of a single entity, risks can be allocated amongst members without any real concern as it is the MNE as a whole who ultimately bears the risk. Therefore, there is a need to determine who actually bears the genuine risk associated with the dealings rather that who is purported to bear the risk. This may be determined by such things as whether the 'risk taker' has the control over the activities that directly influence the income or loss associated with the risk; or has the financial capacity to fund any loss that might arise from the risk.

Prices and margins may vary across different markets for the same or similar property. Therefore, comparability will be affected by any economic or market differences between associated enterprises dealings as compared to those of any comparable uncontrolled enterprise dealings and will need to be taken into account when making any comparison. The matters that may need to be taken into account include: similarity in geographic location; the level of the market (e.g., retail or wholesale); date and time of the dealing; market size, maturity, competition, level of supply and demand, risk of substitute technology; local consumer purchasing power, cost structures and regulation; and any other alternatives realistically available to the buyer and seller.

In attempting to determine an arm's length outcome for international dealings between associated enterprises, the results of any one year may be distorted by differences in economic, market or enterprise conditions affecting the controlled or uncontrolled dealings. Participants in an industry may not be uniformly affected by business and product cycles and therefore differences between dealings may reflect differences in circumstances, not the effects of non-arm's length dealings.

A valid conclusion as to what constitutes an arm's length outcome for a dealing usually requires examination of several years of dealings for both the controlled and uncontrolled parties. In this way differences due to business or product cycles can be more effectively taken into account and comparability more reliably determined (see paragraphs 1.49 - 1.51 of the 1995 OECD Report).

The number of years that need to be examined will depend on the facts and circumstances of the case, but as a starting point the ATO will consider the current year and the preceding four years.

Business strategies of an MNE group are as a rule formulated by one member of the group (usually the parent), often after consultation with and input from group members, and then put in place by all members. These strategies may include such things as: product innovation, degree of diversification, market level and location, worldwide pricing and general management policies.

These strategies may also include a market share strategy where it may be decided to enter new markets or increase market share by increasing market development costs, establishing a more effective distribution network or lowering the resale price to undercut the price of competing goods and services. The question then to be decided is which member of the group should bear the cost of such a strategy. This will need to be determined on the facts and circumstances of each case but for comparability purposes should only be taken into account if it can be found that uncontrolled enterprises engaged in similar strategies under similar circumstances.

Usually uncontrolled enterprises would come to some prior agreement before entering into these strategies that would detail the sharing of costs and resulting profits. This would normally involve the preparation of a budget or plan setting out the expected costs and profits, each parties share and the duration of such a strategy.

Ideally the traditional transaction methods should be applied on a transaction by transaction basis. However, it is also recognised that where it is impractical to assess individual transactions it may be more appropriate to consider a combination of transactions.

Consideration may also need to be given to the fact that dealings between associated enterprises are sometimes structured differently from uncontrolled dealings because members of an MNE group have an ongoing relationship that includes overarching strategies, a common strategic goal, objectives and policy commitments and they are unlikely to enforce contracts through the courts. In some cases, therefore, it may be more appropriate to group products, services or activities in order to properly reflect the relative contributions to profit from the associated enterprise dealings. Grouping may be appropriate in the following situations: (a) Transactions/components of transactions Dealings between associated enterprises in a particular product may involve separate transactions for the product, the intangibles associated with the product, technical advice, management services and any other related matters. In dealings with independent parties these various aspects may be rolled into a package deal with all the associated costs being included in the transfer price of the product. Where the independent dealings being considered as possible comparables cannot be disaggregated, it would generally be appropriate to group all the relevant transactions between associated enterprises so comparability to the uncontrolled party package deal transaction can be properly determined. Care would be needed to identify the value of any component of the package that is subject to different domestic tax treatment, e.g., items subject to interest or royalty withholding tax. (b) Integrated operations If it was decided to route the transaction through an associated enterprise it may be more appropriate to consider the dealing in its entirety rather than consider the component transactions on a separate basis. The combined/channel profit, the functions of each of the associates, the value added by each of the channel entities and the amount of profit appropriated to each entity would need to be considered when applying the arm's length principle to set or review prices or conditions. There could well be practical difficulties in determining the true value added by any intermediate company if it is considered in isolation. For example, a company may be licensing intangibles and supplying vital components to an associate as part of a highly integrated global manufacturing process (see paragraph 1.42 of the 1995 OECD Report). If it cannot be demonstrated in a particular case that the intermediate company either bears a real risk or performs an economic function in the chain that has produced the value of the goods or services, then any profit element that is claimed to be attributable to the activities of the intermediate company should be attributed elsewhere in the MNE group, because independent enterprises would not normally have allowed such a company to share in the profits from the dealing. In this regard the OECD states at paragraph 3.31 of its 1995 Report: 'A one-sided analysis may not take into account the overall profitability of the MNE group from the controlled transactions for the purposes of comparability. A one-sided analysis potentially can attribute to one member of an MNE group a level of profit that implicitly leaves other members of the group with implausibly low or high profit levels. While the impact on profit of the other parties to a transaction is not always a conclusive factor in determining the pricing of a transaction, it may act as a counter-check on the conclusions reached.' (c) Product lines The business activities of MNEs may be based on a single product or service, on a number of related products or services or on a variety of products or services. When applying the CP, RP and profit methods the emphasis is more on the comparability of the functions and product differences do not impact on the analysis the way they do when considering prices. For example, the business activities of a member of an MNE group are the importing and wholesaling of toasters, electric kettles, blenders and the provision of services in the form of advice on satellite communication. Although the MNE management may have a number of separate product lines it may be appropriate in analysing comparability to group the household electronic products together if the functions of wholesaling these products are similar. This could avoid possible distortions that may arise where indirect costs are allocated to individual products. (d) Prior dealings On occasions it will be important to look back over a sequence of transactions in order to put a current transaction in an appropriate context. This is important where there has been a substantial prior investment in the development of intangibles, or a prior sale of a relevant asset. There is a need to establish an appropriate setting or starting point for an identification of the economic alternatives that an uncontrolled decision maker would normally wish to consider and to identify comparables where these exist. Note that the identification of comparables flows from a careful and insightful specification of the 'transaction', i.e., from a determination of the point where the choice exists in an economic sense (see paragraph 1.15 of the 1995 OECD Report).

Where dealings have been grouped there would be a need to allocate relevant operating, financial or other expenses across different product lines, divisions, etc., to reflect that grouping. Where it is not possible to allocate on a direct basis a soundly-based method of indirect allocation should be used that accords with accepted accounting principles and fits the particular circumstances.

In some cases, comparability can only be established by a further extension of the grouping approach. It may be necessary to aggregate the product or business lines so as to consider the matter in its proper business and economic context. This situation may arise where, for any reason, there is insufficient data available on comparable dealings to undertake a comparability analysis on any other basis.

Lack of reliable data on comparable dealings may be due to the complexity of the dealings or the relationships between the parties. Where the resulting dealings are unique, the only option available for making transactional comparisons may involve some divisional comparisons or aggregation of a range of dealings.

Developments in relation to certain sectors seem to indicate that in arm's length dealings, relationships may be more influential in international trade and that transactions cannot be examined in isolation from those relationships. This is reflected in the previous discussion on comparability, particularly the fact that comparability extends beyond product similarity. The special features of any relationships need to be taken into account. The complexity of those relationships will often be dictated by the complexity of the deals that are being struck. While transactions suggest the possibility of markets or prices that may be of some help in benchmarking, a proper analysis of an arm's length relationship for the purposes of finding comparables has to have regard to the exercise of skill and power in the bargaining context. While price may serve as the single most important factor and as a logical focus for any negotiation in a transaction mode, bargaining to establish the relationship involves much more than price (which may indeed play a minor role). It will usually be approached by all parties on the basis of carefully thought through positions based on expected rates of return, profitability, hurdle rates (to establish bargaining thresholds), alternative agendas and the fact that the parties in dealing with each other will have regard to their continued participation in the relevant market. Business strategies therefore need to be addressed in the comparability analysis. Ultimately, profit in one form or another is the criterion for acceptability. A transaction by transaction approach tends to assume a degree of market stability, openness and information, at least sufficient to allow for comparables to emerge. The reality of many commercial situations today is best characterised as turbulent, and in this setting management is forced to think in terms of strategic relationships that exploit each party's distinctive competencies.

The trends in these sectors can be summarised as follows: From Transactions To Relationships Simple structure Complex Comparability Distinctiveness Markets Power, skills in bargaining Stability Turbulence Past orientation Future orientation One to one Industry networks. To the extent that these trends affect the market in which a taxpayer operates, the comparability analysis needs to take them into account.

Where an aggregated analysis is undertaken and the taxpayer is dealing with related parties from different taxing jurisdictions it will still be necessary to allocate the income and expenses of the cross-border operations of the related parties to the different taxing jurisdictions on the basis of economic value added by each party.

For various reasons enterprises record their transactions in their books of account in a manner that suits their reporting needs and the statutory requirements of the country in which they operate. As the majority of the methodologies rely upon reference to costs and comparable margins whether it be gross, net or some intermediate level - and the profit methods also have regard to them - there is a need to ensure that any differences in accounting treatment between entities being compared are adjusted so that an accurate comparison of costs and margins can be made. For example, some enterprises may include royalties paid or insurance and freight for purchased goods above the gross profit line while others may include them below the gross profit line. The basic rule is that while accepting that accounting standards will vary between countries, true comparability must be based on a consistent approach insofar as components of income and costs taken into account in comparing the performance of the taxpayer with that of the independent enterprise being considered as a possible comparable.

If data is not available to determine the basis of accounting of any enterprise being considered as a comparable then any comparable analysis should be at the net margin level or at a level that would include all relevant costs. For example a level may be selected whereby all relevant costs are included (direct and indirect costs) but certain operating and financial expenses related to management efficiencies and business financing alternatives are excluded because it is doubtful that these latter items have been accounted for consistently.

In the practical application of transfer pricing methodologies, an arm's length result may not always be a single point. Transfer pricing is not an exact science and the application of a single method or several different methods may produce a range of reliable results because: (a) in using a single method, application of the arm's length principle only produces an approximation of conditions that may be established between independent enterprises and for this reason the comparables examined may achieve varying outcomes in similar dealings; or (b) when using more than one method, differences in the nature of the methods and data relevant to applying each method may produce different outcomes. Each range has the potential to define an arm's length outcome.

There are a number of considerations to be taken into account when constructing an arm's length range. First, comparable uncontrolled dealings need to be identified and selected on the basis of criteria required to undertake the method being applied.

Secondly, if material differences exist between the dealings by associated enterprises and the cases being considered as possible comparables, adjustments need to be made to reflect the differences in order to improve the comparability of the uncontrolled dealings. A functional analysis should identify material differences and may offer a basis for determining any necessary adjustments. If reasonably accurate adjustments cannot be made to eliminate material differences then the case being considered as a possible comparable is not truly comparable.

Finally, the arm's length range will be constructed using only comparable uncontrolled dealings that have, or have been adjusted to, a similar level of comparability with the controlled dealings.

In some circumstances only a limited number of comparable uncontrolled dealings may exist and it may not be possible to construct an arm's length range either because of insufficient data or because reasonable adjustments cannot be made to account for differences between the dealings.

In the absence of comparable uncontrolled dealings, it may be possible to infer from other industry information available whether dealings between the associated enterprises achieve an arm's length outcome. However, it should be noted that data which does not achieve the required level of comparability cannot be used in constructing an arm's length range and, while it may be useful in terms of broad indications, cannot be given the same status in determining an arm's length outcome. Nevertheless, it does provide relevant information which, when combined with other information, may assist in determining an arm's length outcome.

As indicated in paragraph 315(a) above, a range may in some cases be produced by the use of more than one method. The principles set out in paragraphs 316 to 320 in relation to comparability should also be followed when using multiple methods.

A properly constructed arm's length range should only include data that is either comparable or which has been reliable adjusted for material differences to become comparable so that every element of data used in the range is a reliable and comparable arm's length outcome. Where a single methodology is used, it has to be capable of being applied with similar accuracy and reliability to each element of data constituting the range, having regard to all the factors relevant to comparability.

Where there is substantial divergence between data in the range it is doubtful that all the data in the range are truly arm's length outcomes. In such cases the reliability of the data in respect of each possible comparable, any adjustments made for material differences in comparability and the methodology itself should be reviewed. It may be that material differences in functions (including assets and risks) in the dealing between the associated enterprises and the comparables in the range have not been correctly identified or accurately reflected.

In order to test the reliability of outcomes, it may be helpful to apply a second methodology. There would be more confidence in ranges that are established by the use of different methodologies if those ranges, when overlayed, reflect common results.

A high level of comparability is required in order to apply a traditional transaction methodology (CUP, RP and CP methods). When using these methods, an outcome that falls within a properly constructed arm's length range should be regarded as being arm's length. The qualification to this statement is that the data used to construct the range must be truly comparable. However, if the dealing falls outside the arm's length range, it is a matter of judgment as to which point in the range the adjustment should be made. The ATO concurs with the view of the OECD that the adjustment should reflect the point in the range that best accounts for the facts and circumstances of the controlled transaction. In this regard see paragraph 1.48 of the 1995 OECD Report.

When applying a method other than a traditional transaction methodology (such as a profit comparison) it is not possible to give the same assurances in every case. The approximations used in applying these other methods which rely on broader measures of comparability can give extensive ranges, some of which may not be sufficiently accurate to permit the general statement that any point in the range may be regarded as arm's length. (Some revenue authorities have regulations which deal with this situation by statistical adjustment of the sample to exclude the outlying data.) Also, many applications of these broader methods will not produce data that is sufficiently reliable or comparable to constitute an arm's length range. However, such a range of results may be helpful in an indicative sense even though they don't constitute an arm's length range. In these situations the arm's length outcome which arises from the use of a method other than a traditional transaction methodology should reflect the point that best accounts for the facts and circumstances of the dealings between the associated enterprises. If such a point cannot be established it would be appropriate to consider seriously using another method to assist in approximating an arm's length outcome.

Intangible and intellectual property can present particular problems when examining comparability, usually because of the specialised nature of the property. There is a trend in the world economy for greater emphasis to be placed on the development and exploitation of intellectual property and intangible assets like patents, copyright and know how. For example, MNEs place increasing importance on brands and trademarks, including the development of global brands. Similarly, non financial assets can be of great strategic importance to an enterprise. Production processes and techniques and forms of organisation such as specialised distribution channels have assumed critical importance in some enterprises.

The term 'intangible property' 'includes rights to use industrial assets such as patents, trademarks, trade names, designs or models, literary and artistic property rights and intellectual property such as know how and trade secrets' (chapter IV of the 1995 OECD Report). Generally two categories of intangibles are recognised - production intangibles and marketing intangibles. Production intangibles typically include patents, trade secrets or unpatented know how. Marketing intangibles include trade marks, trade names and distribution networks. Hybrids of these categories can occur where intangibles obtain value through the activities of research and development/production and of marketing and sales. Intangibles may also include copyright protected software or specialised methods for providing service to customers, or of doing business.

The difference between production and marketing intangibles can be seen in a comparison of patents and trademarks. Patents are basically concerned with the production of goods while trademarks are used in promoting the sale of goods or services. Intellectual property such as know how may be either a production or marketing intangible.

Australia's transfer pricing tax laws do not differentiate between the supply or acquisition of intangible property and any other transfer of property under an international agreement. Both sets of transfers are governed by the arm's length principle. The reason that emphasis is sometimes placed on intangibles is due to the special difficulties they can present in the practical calculation of an arm's length consideration. Where there is insufficient comparable data for direct comparisons, this can lead to greater reliance being placed upon profit based or other indirect arm's length methods. This situation can arise due to: (a) the unique character of the intangible. There is commonly a scarcity of information on comparable property or dealings and difficult valuation questions arise; (b) the need for highly valuable intangibles to generally stay within the control of the MNE group to maximise its profitability; (c) the fact that certain intangibles can only be protected by keeping their attributes secret within the MNE group; or (d) the intangible being developed solely by the efforts of an enterprise and for its own purposes, these purposes being applicable only to it, for example some marketing intangibles.

However, the general principles and guidelines in relation to comparability and the methodological approaches applicable for transfers of other types of property are relevant.

It is necessary to clearly establish the existence and nature of the intangibles before attempting to attribute to them any value or taking them into account in applying an arm's length methodology. Sometimes research and development expenditures do not produce a valuable production intangible. Marketing activities may not result in the creation of a marketing intangible. Care should be taken to ensure that the relevant activity has in fact produced a business asset. Even where this is the case it can be difficult to evaluate the benefit to the business in a particular year from the use of the asset.

Similarly, intangibles with different strengths will need to be rewarded differently. For example, a patented production process may be useful, but it may be fairly simple to design around the patented aspects in order to achieve a similar outcome. This type of intangible should not receive the same level of relative reward as a breakthrough patent which may, say, significantly reduce production costs and improve the product so that there is greatly improved customer demand.

When examining whether an intangible has international application it may be appropriate to examine the reasons for an enterprise's success and to recognise that this can vary between countries. An enterprise does not necessarily succeed in the same way and for the same reason in all markets. For example, a consumer product may have a high level of acceptance in its home market due to its technical innovation. However, in an overseas market it may be able to demonstrate that local marketing intangibles were determinative in it achieving its market position.

A functional analysis can assist in identifying the intangibles and the way in which they are used. However, judgment is still needed to determine an appropriate reward for the owner but a better decision is likely to be made once the nature of the intangibles and their role in the profit making process are properly understood.

Many names, marks or patents have very little inherent value. What they require are successful business systems which can commercialise and add value to the property. The value can arise from the efforts and investment that are made in the assets, relationships and systems that the enterprise uses to deliver, position and differentiate its goods and services from competitors.

For example, a brand may become synonymous with quality, not because of the name itself, but because of the style and quality of the product, the marketing and promotion that is undertaken and the service associated with its delivery. A functional analysis can help to identify who contributed what to the overall creation of the intangibles. Extending this example, an enterprise may be the legal owner of a trade mark and name which it legally protects. It may attribute a high value to these marks for which it seeks a direct reward. Subsidiary enterprises in different countries may separately produce, market and support goods bearing this name and mark. A functional analysis should identify each party's contribution to any manufacturing intangible or marketing intangible. A shared ownership of the intangibles derived from the economic contribution of the parties could result. This could influence the selection of a methodology.

There are a number of internationally accepted methodologies - which are also accepted by the ATO -that test compliance with the arm's length principle in different ways.

These arm's length methodologies are divided into two groups: (a) the traditional transaction methods ('traditional methods') being the comparable uncontrolled price (CUP) method, the resale price (RP) method, and the cost plus (CP) method; and (b) the profit methods ('profit methods') which include the profit split methods and profit comparison methods (this second category referred to by the OECD as the 'transactional net margin methods').

The OECD, international tax administrations and MNEs have had to review transfer pricing methodologies and their application in order to take account of the increasing complexity and integration of cross-border dealings and to promote internationally consistent approaches to the application of the arm's length principle. There is an increasing volume of literature, and new ideas are emerging about how to better evaluate compliance. It needs to be recognised that since international business is complex and dynamic, this process of refinement will be an ongoing one and the application of methodologies must remain flexible and receptive to those developments which may lead to refinement of the present arm's length methodologies or the development of new ones for particular categories of case. It is also recognised that enterprises may have either developed their own methodology or adapted the ones discussed in this Ruling to best suit their circumstances. Such approaches should not be automatically discounted as they may be appropriate in the enterprises particular circumstances.

All of the internationally accepted methodologies are seeking to produce a reasonable approximation of an arm's length outcome. What varies between the methodologies is the type of comparison that is made and the manner and extent to which the methodologies rely upon information external to the enterprise.

The methods discussed in this Ruling provide a basis of determining if dealings between associated enterprises are consistent with the arm's length principle. However, it is not possible to provide specific rules that will cover every case and no one method is suitable in every situation. Sometimes, where the above methods are not suitable there may be a need to adopt some other form of comparison that will approximate an arm's length outcome. Accordingly, when undertaking the functional analysis any useful information should not be discarded simply because some rigid standard of comparability is not met.

Since precise calculations cannot be made and the application of any methodology involves elements of judgment there is a need to avoid making adjustments to account for minor or marginal differences in comparability.

The traditional methods are briefly outlined in Taxation Ruling TR 94/14 at paragraphs 353 - 365. Since issuing TR 94/14 the OECD has reviewed its description and discussion of the methodologies and has confirmed that certain profit methods are consistent with the arm's length principle when applied in cases where there is insufficient data on uncontrolled transactions or where such data is considered unreliable, or due to the nature of the business situation and in a way that has proper regard to comparability. The ATO agrees with this view. Each of the traditional methods together with profit methods are explained below. For the reasons explained when discussing the arm's length principle, it is essential to always be mindful of what is reasonable and to adopt practical, flexible approaches in the application of the methodologies so that they achieve the highest degrees of comparability that is feasible and to avoid disputes about theoretical aspects that may not produce practical solutions in a given case. This may mean considering alternatives. Some further approaches are also discussed in relation to the concept of aggregating some of the dealings of the associated enterprises in order to find an appropriate answer.

As was stated in paragraphs 86-87 of TR 94/14, the ATO will select the method that is the most appropriate or best suited to the facts and circumstances of the particular case. The ATO is under no obligation to accept the particular methodology used by a taxpayer unless, on an objective analysis, it produces the most accurate calculation of the arm's length outcome. The most appropriate method will be the one that produces the highest practicable degree of comparability. Its selection will be driven by the extent to which reliable data on comparables is available in the particular case. Taxpayers and the ATO should have regard to all the principles in paragraph 87 of TR 94/14 when selecting a methodology.

As was stated in paragraph 263 above, the ATO takes the view that any transfer pricing methodology used to calculate an arm's length consideration in international dealings between associated enterprises must be applied in a way that will provide an arm's length outcome by closely reflecting commercial and economic reality and the economic contribution made by the enterprises in each jurisdiction.

Australia follows the OECD view that the CUP, RP and CP methods should be preferred over profit methods as a means of establishing whether a dealing is arm's length. In this regard see paragraph 3.49 of the 1995 OECD Report. It is also accepted that where it is possible to identify comparable uncontrolled transactions, the CUP method will provide a higher degree of comparability than the RP and CP methods (or any other method).

The reliability of any method is affected by the completeness, accuracy and reliability of data and the degree of comparability that the data provides. The reliability of any method may also be affected by the extent and nature of any adjustments made to account for differences between a taxpayer and a enterprise being considered as a possible comparable or between controlled and uncontrolled dealings. For example: (a) the available comparable arm's length data may only cover a few products in the taxpayer's range, and this data may not be representative of the entity's overall business; (b) the available comparable arm's length data or the taxpayer's data may be recorded in such a manner that comparison on a product / service line, divisional line or otherwise is not possible; or (c) the comparable data may be available only for limited periods.

The choice of the most appropriate method should be based on a practical weighting of the evidence having regard to the nature of the activities being examined, the quality and reliability of the data and the nature and extent of any assumptions and the degree of comparability that exists between the controlled and uncontrolled dealings or between enterprises undertaking the dealings where the difference would effect conditions in arm's length dealings being examined.

All OECD member countries recognise that because of the complexities of real life business situations there may not be enough reliable data available to rely upon the use of a traditional transaction method. In this regard see paragraph 2.49 of the 1995 OECD Report. The lack of suitable data and comparables would hinder the application of all methods. However, CUPs rely on data of comparable products or services being transacted in similar circumstances; RP and CP methods rely more on data to demonstrate similarity of functions to enable comparison at the gross, net or at some other interval of profit; whereas profit methods consider similarity of functions and make the comparison at the net profit level. It may therefore be possible to apply one these methods in circumstances where the data is not complete or reliable enough to apply a method that conceptually would provide a more direct or reliable reflex of comparability.

Where traditional methodologies are inapplicable or not practicable, for example where there is insufficient data on uncontrolled transactions, where such data is considered unreliable, or due to the nature of the business situation other methods may need to be considered. Profit methods, such as a profit split method or the profit comparison method (which involves the comparisons of net margins for comparable activities of independent entities), may be appropriate.

Where an analysis of comparability has been undertaken using one of the traditional transaction methods and there is some uncertainty as to the reliability of the outcome, perhaps due to comparability factors and the quality of the data used, it would be appropriate to check the outcome by using some other basis. One way this may be done is by comparing the result of the combined/channel profits achieved by applying the selected method with the result achieved by a method having regard to the matters like expected rates of return, risk levels, profitability, hurdle rates or other statistical analyses that independent parties would use to evaluate potential transactions. In this regard see paragraph 1.15 of the 1995 OECD Report.

The ATO and other OECD member countries recognise three methods as traditional transaction methods. These methods are: the comparable uncontrolled price method or CUP method, the resale price or RP method and the cost plus method or CP method.

The CUP method compares the consideration paid in relation to the same or similar property in comparable circumstances. This analysis extends beyond product similarity. The RP method compares the profit margins in relation to similar functions where little physical value is added to the product, e.g., wholesaling. The CP method compares mark-ups on costs in relation to similar functions where there is physical value added, for example, an assembly plant.

If there is a difference between the taxpayer's and comparables' consideration or margin it may indicate that the dealings between the associated enterprises are not arm's length and that the comparables' consideration or margin may need to be substituted for the taxpayer's consideration or margin.

The data necessary to apply the traditional transaction methods must be highly comparable. Data that is either not comparable or is only broadly comparable would not meet the comparability standards required of these methods. Nevertheless, this data may have a use in other methods but the result should not be preferred over those obtained by the application of a profit split or profit comparison unless, on a relative basis, it is more reliable.

The CUP method compares: 'the price for property or services transferred in a controlled transaction to the price charged for property or services transferred in a comparable uncontrolled transaction in comparable circumstances' (paragraph 2.6 of 1995 OECD Report). Controlled and uncontrolled transactions are comparable if none of the differences between the transactions could materially affect the factor being examined in the methodology (e.g., price or margin), or if reasonable accurate adjustments can be made to eliminate the material effects of any such difference (see paragraph 2.7 of the 1995 OECD Report).

This methodology could be used to arrive at an arm's length outcome for a wide range of dealings, not just prices for the transfer of tangible goods. For example, it may be appropriate to check whether a royalty rate for the use of intangible property, interest rate for funds supplied or acquired, or a management fee for services acquired or provided, complies with the arm's length principle.

However, there may be cases where the dealings between associated enterprises involve a variety of transactions (tangible and intangible property, management services, interest, etc.) and it is not possible to obtain CUPs for all the transactions. In those cases the CUP method may be still suitable for some classes of dealings where it is supported by other methods that will reliably evaluate those transactions not determined by the CUP methodology.

The earlier discussion of the principles to be applied in determining comparability apply equally to the CUP method. The most important comparability factors are: similarity of product, contract terms and economic/market conditions. While the application of the CUP methodology involves close product similarity, its application also requires a consideration of all other factors relevant to comparability. In this regard see and paragraph 90 of Taxation Ruling 94/14 and paragraph 2.9 of 1995 OECD Report.

For example, the prices of internationally traded mineral commodities often differ because of differences in the geographic markets, the terms of the contractual arrangements (such as volumes, discounts, interest free periods, currency rate exposure), the particular time period of the contracts, or differences in the physical/chemical features of the commodity and the relative bargaining power and strategies of buyers and sellers. Business strategies like price competition and marketing intangibles like brand names can also impact on prices.

Data to determine the CUP comparability factors may be obtained from a functional analysis. This can produce four types of comparison of varying comparability. They are: (a) the same property or services sold or acquired in the same circumstances (contract terms, volume, economic/market conditions). For example, the entity may be involved in arm's length dealings which can be directly compared to its dealings with associated enterprises for the same products or services. Such a comparison will generally be highly reliable; (b) similar property in the same circumstances; (c) the same property in similar circumstances; (d) similar property in similar circumstances. This situation often results from a comparison of transactions undertaken by unrelated parties with third parties.

In earlier discussion on the concept of comparability, some general issues were identified that may also overlay this model creating additional areas for comparison. This could include the business strategies of the enterprise.

In cases where the functional analysis provides comparable data other than the same property and the same conditions the reliability of any comparison will depend on identifying the differences and the ability to make reliable adjustments for those differences. All comparisons should be considered and can be used to reinforce each other. It may be appropriate to examine more than one comparable uncontrolled transaction in order to arrive at an arm's length outcome.

Australia and other OECD member countries recognise that the CUP method provides the most direct comparison, and encourages its use even where adjustments to the data are required to be made, provided that reliable adjustments can be made for material differences. In some cases consideration may be given to applying a more flexible approach to enable the CUP method to be used and be supplemented as necessary by other appropriate methods. It needs to borne in mind that a minor difference in the property transferred in the controlled and uncontrolled transactions can materially affect the price (paragraph 2.8 of the 1995 OECD Report). Nevertheless, the method should be seriously considered and not routinely dismissed because it may be difficult to make adjustments (paragraph 2.9 of the 1995 OECD Report). However, if the differences have a material effect on price and adjustments cannot be made with any confidence, alternative methods will need to be considered. It should also be remembered that, since adjustments to controlled and uncontrolled dealings to develop a comparable will inevitable involve elements of judgment, the extent, number and reliability of such adjustments will affect the relative reliability of the CUP method analysis.

With the CUP method, comparisons need to be made for each dealing. Once an arm's length consideration has been determined, there will be a need to monitor the correlation of the dealings of the taxpayer and those of the comparable over time to ensure that the CUP initially selected remains valid and to reflect any price movement in the relevant market. Any variation in the dealings or the consideration given or received will need to be taken into account and adjustments will need to be made for any differences that have a material effect on terms and conditions.

The 1995 OECD Report provides the following examples to illustrate the application of the CUP method, including situations where adjustments may need to be made to uncontrolled transactions to make them comparable uncontrolled transactions: '2.11 The CUP method is a particularly reliable method where an independent enterprise sells the same product as is sold between two associated enterprises. For example, an independent enterprise sells unbranded Colombian coffee beans of a similar type, quality, and quantity as those sold between two associated enterprises, assuming that the controlled and uncontrolled transactions occur at about the same time, the same stage in the production/distribution chain, and under similar conditions. If the only available uncontrolled transaction involves unbranded Brazilian coffee beans, it would be appropriate to enquire whether the difference in the coffee beans has a material effect on the price. For example, it could be asked whether the source of coffee beans commands a premium or requires a discount generally in the open market. Such information may be obtained from commodity markets or may be deduced from dealer prices. If the difference does have a material effect on the price, some adjustments would be appropriate. If a reasonably accurate adjustment cannot be made, the reliability of the CUP method would be reduced, and it might be necessary to combine the CUP method with other less direct methods, or to use such methods instead. 2.12 One illustrative case where adjustments may be required is where the circumstances surrounding controlled and uncontrolled sales are identical, except for the fact that the controlled sales price is a delivered price and the uncontrolled sales are made f.o.b. factory. The differences in terms of transportation and insurance generally have a definite and reasonably ascertainable effect on price. Therefore, to determine the uncontrolled sales price, adjustment should be made to the price for the difference in delivery terms. 2.13 As another example, assume a taxpayer sells 1,000 tons of a product for $80 per ton to an associated enterprise in its MNE group, and at the same time sells 500 tons of the same product for $100 per ton to an independent enterprise. This case requires an evaluation of whether the different volumes should result in an adjustment of the transfer price. The relevant market should be researched by analysing transactions in similar products to determine typical volume discounts.'

Where, for whatever reason, it is not possible to use a CUP methodology the application of the other traditional transactional methods (other traditional methods) should be considered. These other traditional methods are: (a) Resale price (RP) method; and (b) Cost plus (CP) method.

As with any method, the availability and reliability of these other traditional methods depend on the availability of sufficient relevant data and may need to be applied on a basket or aggregated basis as discussed in paragraphs 306 to 312 above. The best comparisons will always be where the taxpayer has comparable dealings in comparable circumstances with both associated parties and uncontrolled parties.

The fundamental difference between the CUP method and other traditional methods is that the former compares the consideration for a comparable product or service in comparable circumstances whereas the RP and CP methods seek to establish the margin that the enterprise should achieve to reward it for functions undertaken, assets utilised and risks assumed.

The ATO agrees with the OECD view that in making comparisons for the purposes of the RP and CP methods, fewer adjustments are normally needed to account for product differences than under the CUP method, because minor product differences are less likely to have as material an effect on profit margins as they do on price. In this regard, see paragraph 2.16 of the 1995 OECD Report. However, closer comparability of products will produce a better result and significant differences in products or services is likely to be reflected in the functions performed. In this regard, see paragraphs 94 to 96 and 360 to 362 of Taxation Ruling TR 94/14, and paragraphs 2.16 to 2.18 and 2.34 of the 1995 OECD Report.

Where uncontrolled and controlled dealings are comparable in all characteristics other than the product or service itself, these other traditional methods might produce a more reliable measure of arm's length conditions than the CUP method. However, significant differences in product are likely to result in differences in functions. Moreover proper consideration should be given to whether adjustments could be made to account for differences in the products transferred and if so, whether the CUP method would produce a more reliable result.

Fewer adjustments may be necessary to account for product differences under the other traditional methods than the CUP method, and it may be appropriate to give more weight to other factors, some of which may have a more significant effect on the margin than they do on price. Such differences may include: (a) the various stages of the business and product cycles; (b) the management strategies. For example, where some business assets are rented as opposed to being purchased; (c) the nature and extent of the functions performed assets employed and risks assumed; and (d) the cost structures of the enterprises being compared. In this regard, see paragraphs 94 and 360 of Taxation Ruling TR 94/14 and 2.16 to 2.21 of the 1995 OECD Report.

Where there are differences in the functions or enterprises being compared that materially affect the margin earned in the controlled and uncontrolled transactions, adjustments should be made to account for such differences. The extent and reliability of any adjustments may affect the relative reliability of the analysis under the other transactional methods.

In some cases when applying other traditional methods it may be difficult to obtain sufficient data or establish suitable comparable dealings to have any confidence in the outcome of the analysis. In such cases it may be appropriate to supplement the RP and CP methods by considering the results obtained from applying other methods, such as a profit method.

The resale price (RP) method begins with the consideration at which property (that has been purchased from an associated enterprise) is resold to an independent enterprise. The resale consideration is then reduced by an appropriate gross profit margin representing the amount out of which the reseller would seek to cover its selling and other operating expenses and, in the light of the functions performed, assets used and risks assumed, make an appropriate profit. When the gross profit margin so calculated is deducted from the resale price the balance - subject to adjustment for other purchase costs like customs duties - is regarded as the arm's length price for the purchases from the associated enterprise. The following diagram broadly illustrates the RP method.

When applying the RP method, financing expenses are often excluded from general, administrative and selling expenses on the basis that the funding of the business is not a material consideration in comparing products, outputs or functions, and that the financials can in fact produce distortions. The RP method in such cases is akin to using an earnings before interest and taxation (EBIT) analysis to measure performance, bearing in mind, though, that the RP method focuses on the gross margin.

There will be other cases where the financing expenses are part of the transfer pricing examination. One aspect of this might be a consideration of whether all the arrangements between the associated parties leave the taxpayer with sufficient working capital compared to what arm's length parties would expect in those circumstances. Flexible approaches will therefore need to be adopted depending on the circumstances of the particular case.

This method is most useful where the reseller markets the product without adding substantial value by physically altering the goods and the resale is realised within a short time of the reseller's purchase of the goods.

The OECD described the RP method in the following terms in their 1995 Report: 'The resale price method begins with the price at which a product that has been purchased from an associated enterprise is resold to an independent enterprise. This price (the resale price) is then reduced by an appropriate gross margin (the "resale price margin") representing the amount out of which the reseller would seek to cover its selling and other operating expenses and, in the light of the functions performed (taking into account assets used and risks assumed), make an appropriate profit. What is left after subtracting the gross margin can be regarded, after adjustment for other costs associated with the purchase of the product (e.g., customs duties), as an arm's length price for the original transfer of property between the associated enterprises. This method is probably most useful where it is applied to marketing operations' (paragraph 2.14). 'An appropriate resale price margin is easiest to determine where the reseller does not add substantially to the value of the product. In contrast, it may be more difficult to use the resale price method to arrive at an arm's length price where, before resale, the goods are further processed or incorporated into a more complicated product so that their identity is lost or transformed (for example, where components are joined together in finished or semi-finished goods). Another example where the resale price margin requires particular care is where the reseller contributes substantially to the creation or maintenance of intangible property associated with the product (for example, trademarks or tradenames) which are owned by an associated enterprise. In such cases, the contribution of the goods originally transferred to the value of the final product cannot be easily evaluated' (paragraph 2.22). 'It should be expected that the amount of the resale price margin will be influenced by the level [and, we would add, nature] of activities performed by the reseller. This level of activities can range widely from the case where the reseller performs only minimal services as a forwarding agent to the case where the reseller takes on the full risk of ownership together with the full responsibility for and the risks involved in advertising, marketing, distributing and guaranteeing the goods, financing stocks and other connected services. If the reseller in the controlled transaction does not carry on a substantial commercial activity but only transfers the goods to a third party, the resale price margin could, in light of the functions performed, be a small one. The resale price margin could be higher where it can be demonstrated that the reseller has some special expertise in the marketing of such goods, in effect bears special risks, or contributes substantially to the creation or maintenance of intangible property associated with the product. However, the level of activity performed by the reseller, whether minimal or substantial, would need to be well supported by relevant evidence. This would include justification for marketing expenditures that might be considered unreasonably high; for example, when part or most of the promotional expenditure was clearly incurred as a service performed in favour of the legal owner of the trademark. In such a case the cost plus method may well supplement the resale price method' (paragraph 2.24). 'Where the reseller is clearly carrying on a substantial commercial activity in addition to the resale activity itself, then a reasonably substantial resale price margin might be expected. If the reseller in its activities employs reasonably valuable and possibly unique assets (e.g., intangible property of the reseller, such as its marketing organisation), it may be inappropriate to evaluate the arm's length conditions in the controlled transaction using an unadjusted resale price margin derived from uncontrolled transactions in which the uncontrolled reseller does not employ similar assets. If the reseller possesses valuable marketing intangibles, the resale price margin in the uncontrolled transaction may underestimate the profit to which the reseller in the controlled transaction is entitled, unless the comparable uncontrolled transaction involves the same reseller or a reseller with similarly valuable marketing intangibles' (paragraph 2.25).

Taxation Ruling TR 94/14 also made the following observations about the RP method at paragraphs 95 and 96: 'The resale price method is best suited to cases where there is a high degree of similarity of process between what the taxpayer does and the activities of independent parties engaged in comparable uncontrolled dealings. The resale method is generally a more reliable measure where there is little useable evidence of comparable uncontrolled sales, where the property or services sold are not used in a manufacturing process of the reseller, or the reseller does not add substantially to the value of the product, e.g., where the reseller, being merely a distributor, sells the product or service to an independent third party. Where the non-arm's length reseller adds substantial value to the property (e.g., where the products are further processed through manufacture or are incorporated into as components of amore complicated product so that the original products is lost or transformed or the taxpayer establishes, builds up or maintains a valuable trademark in the relevant market largely through its own expense and endeavour), a portion of the resale price is attributable to this effort. This addition would need to be assessed and accounted for, making it more difficult to establish an arm's length consideration and consequently, more difficult to apply this method.'

It is also more reliable where the reseller on-sells within a short time. The more time that elapses the more likely it is that other factors like changes in the market, in rates of exchange, in costs, etc., will need to be taken into account in any comparison (paragraph 2.23 of 1995 OECD Report).

As the dealing is the purchasing of property from an associated enterprise, the object of the exercise is to determine whether the purchase price is an arm's length price. As such the general equation would be best expressed as: Selling Price - Appropriate Gross Margin = Arm's Length Purchase Price

Of course an adjustment would have to be made for other costs associated with the purchase of the product to arrive at an arm's length purchase price. Examples of these would be insurance, transport costs and any other expenses involved in getting the products into store. Care would need to be taken where the expenses involve payments to associated enterprises. In such cases the transfer pricing review should cover both the acquisition of property and the related expenses.

When first approaching a transfer pricing problem using the RP method, there are two unknowns, i.e., the purchase price and the appropriate gross margin. These two components are inter-related and increases or decreases in one will have a corresponding reverse effect on the other. In applying this method, enquiries are directed at establishing the appropriate gross margin so that the arm's length purchase price can be deduced. A good starting point in this process might be to align the taxpayer's accounts with those of the possible comparables so that the comparisons can proceed on a consistent basis. It should be borne in mind that the starting point with RP method is to make a comparison of the gross profit with comparable independent enterprises. This will indicate any discrepancy arising in the cost of goods sold. Where there is a discrepancy below the gross profit line it is likely that the RP method is not the most appropriate method.

The appropriate gross margin is the amount of profit that, based on an arm's length comparison, is considered necessary to compensate the enterprise for its costs and make an appropriate profit that accords with the functions undertaken, assets employed and risks assumed.

The appropriate gross margin is usually measured at the gross profit level. However, in some circumstances it may be more accurate to undertake the comparison at some other (intermediate) profit level. The profit level at which to compare will be determined by the availability of sufficient reliable data, bearing in mind that there may also be a need to adjust for accounting differences between the associated enterprise and the uncontrolled enterprise being considered as a possible comparable.

Whenever the RP method is applied it would be appropriate to check whether the resale price margin so determined is realistic having regard to the operating expenses of the taxpayer.

An unrelated party may not accept to pay a higher price resulting from the inefficiency of the other party. On the other hand, if the other party is more efficient than can be expected under normal circumstances, this other party should benefit from that advantage. However, where company policies are determined or influenced by an associated enterprise then these type of costs may need to be taken into account when undertaking the comparability analysis. For example, if it is corporate policy, for the benefit of any associated enterprises' intangible assets, to seek accommodation, staffing levels, etc., that are materially different to those levels required by comparable unrelated parties or to reduce profit margins for market penetration or market share, keep excessive stock levels or incur excessive marketing costs that materially differ from those policies of a comparable enterprise, then those costs may need to be taken into account as they would relate to management decisions imposed on the enterprise that may not have been accepted by an independent party operating wholly independently.

The RP method requires the reseller to compare the functions and the resulting gross margin obtained in its controlled dealings against either: (a) the resale price margin that the same reseller earns on the same items purchased or sold in comparable uncontrolled dealings, e.g., the profit margin obtained by the taxpayer from a comparable purchase of goods (involving similar functions and risks) from an unrelated party which are resold to another unrelated party; or (b) the resale price margin earned by an independent enterprise in comparable uncontrolled dealings. A functional analysis should be used to assist this process.

The appropriate gross profit margin would be expected to vary according to the amount of value added by the reseller. There will be many different situations where the combination of assets, functions and risks will add the value to the product. This can be illustrated simply as follows: (a) where the reseller performs minimal services as a forwarding agent or broker. Here the comparable profit margin might be derived from an examination of commission or brokerage fees; (b) where the reseller takes property in the goods, assumes the business risks, warehouses and distributes them to customers. Here the profit margin applicable to a principal would be relevant; or (c) where the reseller not only carries out the functions and risks in (b) above but also undertakes marketing, education and other activities, assumes warranty and other risks and employs intangible assets such as a developed distribution network. The additional functions undertaken, risks assumed and intangibles used should result in higher returns.

As a general rule it would be expected the appropriate gross profit margin would increase with the increased assets, functions and risks. In some situations where the activity undertaken materially exceeds that of other cases being considered as possible comparables it may be more appropriate to consider the matter on a cost plus basis. For example, where a significant amount of marketing expenditure is incurred by the taxpayer for the promotion of a trade mark that is owned by an associated enterprise and risks its own resources in these activities, the taxpayer would be entitled to a commensurately higher expected return than an agent (see paragraph 96 of TR 94/14 and paragraph 1.25 of the 1995 OECD Report).

Under the terms of the agreement the reseller may have exclusive rights to resell the goods. In this case the appropriate gross margin may be influenced by such matters as: (a) the extent of the geographical market and the existence and relative competitiveness of possible substitute goods ( i.e., do the goods sell themselves or is there a need to win market share); (b) the level of activity undertaken by the reseller. For example, the reseller may commit large resources to market the property or may realise a monopolistic turnover without much effort; (c) the risk associated with having only one source of supply and being tied to the other enterprises product development cycle, etc.

As explained in paragraphs 95 and 96 of TR 94/14, it may be difficult to use the RP method to arrive at an arm's length consideration where, before resale, the goods are further processed or incorporated into a more complicated product so that their identity is lost or transformed. The OECD also cites a similar example, '... where components are joined together in finished or semi-finished goods. Another example where the resale price margin requires particular care is where the reseller contributes substantially to the creation or maintenance of intangible property associated with the product (for example, trademarks or tradenames) which are owned by an associated enterprise. In such cases, the contribution of the goods originally transferred to the value of the final product cannot be easily evaluated.' In this regard, see paragraph 2.22 of the 1995 OECD Report. In cases of this nature it may be more appropriate to supplement the RP method by considering the results obtained from applying other methods or consider using other methods such as the CP method where there has been physical changes in the property and profit split method where intangibles have been employed.

As a result of an application of the RP method it may be necessary to increase the taxpayer's assessable income . An example of this would be where the taxpayer has a gross profit margin of 35% calculated on the basis of: (Relevant gross profit / Relevant sales) * (100/1) The uncontrolled party has a profit margin of 70%. In the absence of a satisfactory explanation, the uncontrolled party's profit margin (70%) would then be applied to the taxpayer's sales figure in respect of the impugned dealings. This will result in an arm's length gross profit for the taxpayer in respect of its associated enterprise dealings. By deducting this figure from the taxpayer's sales figure for the associated enterprise dealings the arm's length purchase price for those dealings is deduced.

In the cases where Division 13 is being applied it may be necessary to apply the arm's length gross profit margin on the basis of each international agreement. However, paragraph 23(b) of the Acts Interpretation Act 1901 could have the effect that this is not necessary in cases where the relevant dealings are of a similar type that could be treated as a basket of goods or services (see paragraphs 135 and 432 to 438 of TR 94/14). In relevant cases Division 13 determinations would be prepared by the ATO relying on both bases so as to allow this issue to be considered in the event of a dispute between the taxpayer and the ATO proceeding to the Administrative Appeals Tribunal or a court.

A methodology which adopts a margin which is calculated as a certain percentage of the resale price (for the purpose of determining the appropriate transfer price), where the percentage chosen is not benchmarked against comparable independent dealings is not a resale price methodology.

In extreme cases such a methodology might be able to be used for the purposes of subsection 136AD(4) of the ITAA. Paragraphs 328 to 340 of TR 94/14 discuss the application of subsection 136AD(4). However, every effort should be made to use other methods like profit methods or a mix of methods before resort is had to such an approach.

Where such a methodology has to be used in cases where no other approach is reasonably open, the fixed percentage should be calculated to produce a result that fairly reflects the functions performed, assets employed and risks undertaken, the intention always being to reasonably approximate an appropriate return for the economic value added in a way that is consistent as practicable with the arm's length principle.

The 1995 OECD Report provides the following examples to illustrate the application of the RP method, including situations where adjustments may need to be made to uncontrolled transactions to make them comparable uncontrolled transactions: '2.29 Assume that there are two distributors selling the same product in the same market under the same brand name. Distributor A offers a warranty; Distributor B offers none. Distributor A is not including the warranty as part of a pricing strategy and so sells its product at a higher price resulting in a higher gross profit margin (if the cots of servicing the warranty are not taken into account) than that of Distributor B, which sells at a lower price. The two margins are not comparable until an adjustment is made to account for that difference. 2.30 Assume that warranty is offered with respect to all products so that the downstream price is uniform. Distributor C performs the warranty function but is, in fact, compensated by the supplier through a lower price. Distributor D does not perform the warranty function which is performed by the supplier (products are sent back to the factory). However, Distributor D's supplier charges D a higher price that is charged to Distributor C. If Distributor C accounts for the cost of performing the warranty function as a cost of goods sold, then the adjustment in the gross profit margins for the differences is automatic. However, if the warranty expenses are accounted for as operating expenses, there is a distortion in the margins which must be corrected. The reasoning in this case would be that, if D performed the warranty itself, its supplier would reduce the transfer price, and therefore, D's gross profit margin would be greater. 2.31 A company sells a product through independent distributors in five countries in which it has no subsidiaries. The distributors simply market the product and do not perform any additional work. In one country, the company has set up a subsidiary. Because this particular market is of strategic importance, the company requires its subsidiary to sell only its product and to perform technical applications for the customers. Even if all other facts and circumstances are similar, if the margins are derived from independent enterprises that do not have exclusive sales arrangements or perform technical applications like those undertaken by the subsidiary, it is necessary to consider whether any adjustments must be made to achieve comparability.'

To illustrate the basic concepts of the RP method a chart, with supporting notes, is set out below. RESALE PRICE METHOD In Relation To Associated Enterprise's Dealings Arm's Length Amount Explanation SALES - To Uncontrolled Parties 1000 Arm's Length Sales PURCHASES - From Associated Enterpises 600 Amount determined by RP Method GROSS PROFIT 400 Appropriate gross margin, based on comparable analysis, to cover selling, general and administration expenses and in light of the functions, assetts and risks make an appropriate profit. Selling, General and Admin, (SG and A) Expenses 300 (See note below) Net operating Profit 100 Appropriate Profit Formula: Selling Price less Appropriate Gross Margin = Arm's Length Purchase Price 1000 - 400 = 600 Key: To determine the appropriate gross margin so as to deduce the arm's length purchase amount. Note: If, by way of associated enterprise dealings, amounts are included in selling, general and administration expenses the arm's length value of these amounts will need to be determined prior to or in conjunction with the determination of the gross profit. If these dealings are extensive or complex the resale price method may not be the most appropriate method.

Paragraphs 97, 98 and 363 to 365 of TR 94/14 contain a description of the cost plus method and certain factors that can affect its suitability and reliability. Paragraphs 2.32 to 2.48 of the 1995 OECD Report contain a similar discussion and provide some examples of its operation.

In paragraph 2.32 of the 1995 OECD Report the cost plus methodology is described in the following terms: 'The cost plus method begins with the costs incurred by the supplier of property (including services) in a controlled transaction for property transferred or services provided to a related purchaser. An appropriate cost plus mark up is then added to this cost, to make an appropriate profit in light of the functions performed and the market conditions. What is arrived at after adding the cost plus mark up to the above costs may be regarded as an arm's length price of the original controlled transaction. This method probably is most useful where semi-finished goods are sold between related parties, related parties have concluded joint facility agreements or long-term buy-and-supply arrangements or where the controlled transaction is the provision of services.' A similar description is contained in paragraph 363 of TR 94/14.

The following diagram broadly illustrates the CP methodology.

The CP method is based on the following equation: Costs of Producing the Goods and Services + Appropriate Mark-up = Arm's Length Selling Price

As the relevant transaction is the taxpayer supplying goods or services to an associated enterprise the object of the exercise is to determine whether the sale price is an arm's length price. As there are two unknowns in the equation (i.e., the appropriate mark-up and the selling price) the appropriate mark-up must be determined to enable the arm's length selling price to be deduced.

As the mark-up is based on costs there is a need to establish which costs will be taken into account when applying the mark-up.

The costs, in general, that need to be established for the CP method will be the direct and indirect cost of production of the relevant goods or services. The functional analysis completed on the taxpayer will assist this determination. It is important to remember that the costs are limited to those of the supplier of the goods or services. Any transfer pricing examination should have regard to the fact that the manipulation of the allocation of expenses could inappropriately increase or decrease the production costs of the taxpayer to which the cost plus margin is applied. These costs would include: direct costs of producing the goods or services such as the cost of raw materials or the salaries of skilled service staff; indirect costs of production, which although closely related to the production process may be common to several products or services (for example, the costs of a repair department that services equipment used to produce different products or deliver different services); and an appropriate allocation of other general, and administrative expenses, that can be related to the production activities the subject of the application of the CP method.

This aggregation of direct and indirect cost is also known as absorption costing. These calculations should generally be done on the basis of historical cost (see paragraph 2.42 of the 1995 OECD Report). Although there are some exceptions to this general principle (for example the use of marginal costing to dispose of marginal production discussed in paragraphs 2.42 and 2.44 of the 1995 OECD Report), these exceptions are subject to strict controls. Consistent with the OECD approach and having regard to generally accepted accounting principles, as a general rule, the use of absorption costing will be required by the ATO where the cost plus method is used. The very limited exceptions are where replacement cost and marginal cost result in a more accurate measure of the appropriate profit margin. In such cases the justification for the view that replacement or marginal cost provide a higher integrity should be adequately and contemporaneously documented.

Some taxpayers already use absorption costing for the purposes of valuing trading stock on a 'cost' basis (Philip Morris Ltd v. FC of T 79 ATC 4352). Taxation Ruling IT 2350 provides some guidance in relation to the use of absorption costing and in this respect would be relevant for the purposes of the application of the CP methodology (apart from the reference in paragraph 5 of that Ruling which indicates that at that time direct costing was recognised for accounting purposes - which is not the case now).

As historical costs such as materials, labour, depreciation, etc., may vary over a period it may be appropriate to average these costs when determining the appropriate level of costs when applying the CP method in relation to a limited period. Averaging may also be appropriate when determining costs across product groups or when applying the CP method in cases where grouping of dealings is needed to properly assess comparability.

In relation to the most appropriate bases for the allocation of indirect costs, Kitto J in B P Refinery (Kwinana) Ltd v. FC of T [1961] ALR 52 at 56;12 ATD 204 at 208, albeit in the context of section 56, said: 'The only admissible criticism (of the taxpayer's choice of allocation of indirect costs), in my opinion, is that the method of apportionment employed is not justified in the particular case by sound accountancy principles.' He went on to say (at 208): 'But it cannot be in accordance with s.56 that one should chase the rainbow of absolute accuracy beyond the point at which the practical accountant would stop. In my opinion it is at that point that the requirement of the word "cost" is satisfied, so that the question in each case is, what is the figure reached by the proper application of the recognised principles of costing to the particular circumstances.' In our view, these statements are equally applicable to the allocation of indirect costs in applying the cost plus method for the purposes of Division 13 and the DTAs. The objective in allocating indirect costs is to determine the degree to which the indirect costs - which are a reflection of the taxpayer's functions, assets and risks - have contributed to the production of the goods or services being examined in the transfer pricing review. The ATO will evaluate the taxpayers' allocation of indirect costs against sound cost accountancy principles.

In this regard, where a taxpayer allocates indirect costs using criteria that are key to the nature of the taxpayers' profit making activity and which can fairly apportion the particular costs on the basis of the extent of the activity subject to the transfer pricing examination relative to the other purposes for which the costs were incurred, the ATO will accept the allocation. The basis of allocation must, however, make sense in the context of the particular case and cannot produce significant distortions.

In cases where a taxpayer uses a formula to allocate indirect expenses, ATO staff examining the allocation would need to establish whether the formula has been consistently followed over a number of years and whether there is any evidence of manipulation that produces an inappropriate loading of expenses in Australia. Where different types of indirect costs are being allocated it may be appropriate to use different allocation criteria.

The functional analysis should assist in identifying the profit drivers in a particular case and the development of allocation criteria based on the key functions, assets and risks that have generated the relevant profits.

Determination of the costs involved in the controlled transaction (to which the arm's length gross margin will be applied) should in many cases present few difficulties for taxpayers as the relevant costs will generally be able to be equated to the calculation of cost of goods sold used for the trading stock provisions of the ITAA or deductible and non-deductible costs for service providers. Often the relevant expenditure will have been incurred in Australia in dealings with independent enterprises (for example, the cost of direct labour and direct materials). However, care would need to be taken where the Australian taxpayer is part of an integrated production process involving a number of countries and the taxpayer has acquired partly finished goods from a foreign associated enterprise. In such cases the CP method would be more difficult to apply and its reliability would need to be carefully checked.

However, there may be occasions where questions would arise as to whether the costs incurred in relation to such acquisitions were at arm's length. For example, where raw materials used in manufactured or expert services used to provide other services are purchased from associated enterprises. In these cases and before such costs were incorporated into the value of costs for the purposes of the CP methodology, an analysis would need to be undertaken to ascertain what an arm's length consideration for the associated enterprise acquisitions would be.

The above the gross profit line costs will need to be apportioned between the controlled dealings and the other business activities of the taxpayer on an appropriate basis.

Many taxpayers also have dealings with associated enterprises other than just in relation to the controlled transaction under review - which may affect the above the gross profit line or below the gross profit line costs of the controlled dealings. The nature of such dealings and whether they are on an arm's length basis would be relevant to determining comparability between the controlled dealings and the uncontrolled dealings. An analysis of these other dealings between associated enterprises would therefore be necessary.

The appropriate mark-up is the amount of the mark-up on the relevant cost determined on the basis of an arm's length comparison.

The cost plus mark-up of the taxpayer in the dealings between associated enterprises should ideally be established by reference to the cost plus mark-up that the taxpayer earns in comparable uncontrolled dealings. Where the taxpayer has no comparable uncontrolled dealings, the cost plus mark-up may be able to be determined on the basis of comparable dealings by independent enterprises that are operating wholly independently. Functional analysis will assist this process.

The appropriate mark-up should be measured at the gross profit level. However, in some circumstances it may be more accurate to consider some intermediate profit level in order to make comparisons on a consistent basis (e.g., to adjust for accounting differences between the taxpayer and the company being considered as a comparable).

It is important to ensure that a comparable mark-up is applied to a comparable cost basis. Distortions caused by different approaches to business financing between the taxpayer and a company being considered as a comparable would need to be removed (see paragraph 2.37 of the 1995 OECD Report). When applying the CP method, financing expenses are often excluded from costs on the basis that the funding of the business is not a material consideration in comparing products, outputs or functions, and that they can in fact produce distortions. Such an approach is akin to using an earnings before interest and taxation (EBIT) analysis to measure performance.

There will be other cases where the financing expenses are part of the transfer pricing examination. One aspect of this might be a consideration of whether all the arrangements between the associated parties leave the taxpayer with sufficient working capital compared to what arm's length parties would expect in those circumstances. Flexible approaches will therefore need to be adopted depending on the circumstances of the particular case.

An uncontrolled enterprise may not accept to pay a higher price resulting from the inefficiency of the other party. On the other hand, if the other party is more efficient than can be expected under normal circumstances this other party should benefit from that advantage (see paragraph 2.45 of the 1995 OECD Report and paragraph 486 below). However, where company policies are determined or influenced by an associated enterprise then these types of costs may need to be taken into account when undertaking the comparable analysis. For example, if it is corporate policy, for the benefit of any associated enterprises' intangible assets, to seek accommodation, staffing levels, etc., that are materially different to those levels required by comparable unrelated parties, or reduce profit margins for market penetration or increased market share, keep excessive stock levels or incurs excessive marketing costs that materially differ from those policies of a comparable enterprise then those costs may need to be taken into account as they would relate to management decisions imposed on the enterprise that may not have been accepted by an arm's length party.

A methodology which applies a fixed percentage mark-up to a relevant cost base where that fixed percentage is not benchmarked against comparable independent dealings is not a cost plus methodology.

In extreme cases such a methodology might be able to be used for the purposes of subsection 136AD(4). Paragraphs 328 to 340 of TR 94/14 discuss the application of subsection 136AD(4). However, every effort should be made to use other methods - like profit methods or a mix of methods - before resort is had to such an approach.

Where such a methodology has to be used, in cases where no other approach is reasonably open, the fixed percentage should be calculated to produce a result that fairly reflects the functions performed, assets employed and risks undertaken, the intention always being to reasonably approximate an appropriate return for the economic value added in a way that is as consistent as practicable with the arm's length principle.

We agree with paragraph 2.36 of the 1995 OECD Report where it is stated: 'The cost plus method presents some difficulties in proper application, particularly in the determination of costs. Although it is true that an enterprise must cover its costs over a period of time to remain in business, those costs may not be the determinant of the appropriate profit in a specific case for any one year. While in many cases companies are driven by competition to scale down prices by reference to the cost of creating the relevant goods or providing the relevant service, there are other circumstances where there is no discernible link between the level of costs incurred and a market price (e.g., where a valuable discovery has been made and the owner has incurred only small research costs in making it).' Another example would be where a taxpayer has to keep a smelter in production rather than incur significant expenditure from a shutdown, even though base metal prices have fallen significantly on world markets.

Where valuable intangibles are involved in the transaction it may be appropriate to supplement the cost plus methods by considering the results obtained from applying other methods such as the profit split method in order to ensure that the profit contribution of the intangibles are properly reflected.

We agree with the statement in paragraph 2.38 of the 1995 OECD Report that differences in the level and types of operating expenses and non-operating expenses (including financing) could indicate functional differences or additional functions between the taxpayer and the parties or dealings being compared that require adjustments to achieve a valid comparison. We also agree, subject to the earlier discussion in paragraph 419, that no adjustment to the gross margin may be appropriate if the differences in expenses reflect different degrees of efficiency.

Where, as a result of applying the CP method, an increase is required in the mark-up used by the taxpayer in a particular income year, for the purposes of adjusting the taxpayer's sales figure it will be necessary to convert the percentage mark-up into a dollar figure on the basis of the taxpayer's costs (adjusted as necessary for the purposes of comparability). In making a determination in cases where Division 13 is being applied, the matters discussed in paragraph 393 above should be borne in mind and a similar approach adopted.

The 1995 OECD Report provides the following examples to illustrate the application of the CP method, including situations where adjustments may need to be made to uncontrolled transactions to make them comparable uncontrolled transactions: '2.46 A is a domestic manufacturer of timing mechanisms for mass-market clocks. A sells this product to its foreign subsidiary B. A earns a 5 percent gross profit mark-up with respect to its manufacturing operation. X, Y and Z are unrelated, domestic manufacturers of timing mechanisms for mass- market watches. X, Y and Z sell to unrelated foreign purchasers. X, Y and Z earn gross profit mark-ups with respect to their manufacturing operations that range from 3 to 5 percent. A accounts for supervisory, general and administrative costs as operating expenses, and thus these costs are not reflected in cost of goods sold. The gross profit mark-ups of X, Y and Z, however, reflect supervisory, general and administrative costs as part of costs of goods sold. Therefore, the gross profit mark-ups of X, Y and Z must be adjusted to provide accounting consistency. 2.47 Company C in country D is a 100% subsidiary of company E, located in country F. In comparison with country F, wages are very low in country D. At the expense and risk of company E, television sets are assembled by company C. All the necessary components, know how, etc., are provided by company E. The purchase of the assembled product is guaranteed by company E in case the television sets fail to meet a certain quality standard. After the quality check the television-sets are brought - at the expense and risk of company E - to distribution centres company E has in several countries. The function of Company C can be described as a purely cost manufacturing function. The risks company C could bear are eventual differences in the agreed quality and quantity. The basis for applying the cost plus method will be formed by all the costs connected to the assembling activities. 2.48 Company A of an MNE group agrees with company B of the same MNE group to carry out contract research for company B. All risks of a failure of the research are borne by company B. This company also owns all the intangibles developed through the research and therefore has also the profit chances resulting from the research. This is a typical set-up for applying a cost plus method. All costs for the research which the related parties have agreed upon, have to be compensated. The additional cost plus may reflect how innovative and complex the research carried out is.'

To illustrate the basic concepts of the CP method a chart, with supporting notes, is set out below. COST PLUS METHOD In Relation To Associated Enterprise's Dealings Arm's Length Amount Explanation SALES - To Associated Enterprises 1000 Amount determined by CP Method Cost of Production (Goods & Services) - Direct Costs - Indirect Costs 600 Arm's Length Costs (See Note below) Gross Profit 400 Appropriate mark-up, based on comparable analysis, to cover general, selling and administration expenses and in light of the functions assets and risks make an appropriate profit Selling, General and Admin (SG and A) Expenses 300 (See note below) Net operating Profit 100 Appropriate Profit. Formula: Costs of Goods/Services plus Appropriate Gross Mark-up = Selling Price 600 + 400 = 1000 Key: To determine the appropriate gross mark-up so as to deduce the arm's length selling amount. Note: If, by way of associated enterprise dealings, amounts are included in production costs or SG & A expenses, the arm's length value of those amounts will need to be determined prior to or in conjunction with the determination of the gross profit. If these dealings are extensive or complex the cost plus method may not be the most appropriate method.

Sometimes it is not possible or practical to use traditional transactional methods (traditional methods). Such situations may arise where: (a) there are insufficient reliable data to analyse comparability so as to determine an arm's length outcome other than through a profit split or a profit comparison at the net profit level; (b) the traditional methodologies are unable to establish an arm's length consideration because, for example, the product or service in question is unique or contains out-of-the-ordinary intangibles; or (c) while theoretically sound, the traditional methods are not practicable because of the complexity of the business situation or the extent and diversity of the taxpayer's cross-border dealings with associated enterprises. These cases present some significant difficulty in following traditional methods such that due administration of the law is seriously jeopardised beyond what a reasonable person would accept, and it is not merely a matter of convenience that supports a different approach. It would be expected that consideration would be given to the use of traditional methods on a basket approach and profit methods would be applied only where traditional methods are unreliable because they are not conceptually applicable or practicable in the particular case, or because there is insufficient reliable data to apply them. In these situations it may be more appropriate to consider the use of profit methods.

Global industries, such as the finance, communications, transport, pharmaceutical, chemical, motor vehicle and computer industries are based on highly sophisticated technology, involve valuable production, distribution or marketing intangibles and are generally vertically and horizontally integrated. The global networks in such industries are complex, have their own unique structures, and their products are also often unique and contain valuable intangibles (which may have been supplied by a number of the associated enterprises). Where this is the case the use of traditional methods is limited as reliable comparable data is not readily available.

Also, in the Australian economy certain industries are dominated by MNEs to the point where the parties operate in niches and the material differences between the mainstream market and the niches do not make adjustments to achieve comparability feasible.

There is also the problem that in many cases there is a variety of transactions (transfers of tangible and intangible goods and services) back and forth between the associated enterprises - some of which may involve overlaps and there may be no comparables for the combination of transactions. In these cases profit methods may be a more reliable way to set or review the transfer pricing used in the dealings between the associated enterprises or to check findings made using traditional methods where there is doubt about the reliability of the data used.

Where, for whatever reason, and after taking account of the guidelines in this Ruling on selection of methods, it is not possible or practicable to use the traditional methods of CUP method, RP or CP methods, the application of profit methods may be considered. In this regard, see the discussion starting at paragraph 3.5 of the 1995 OECD Report. The type of profit method used will depend on the facts and circumstances. These profit methods include: (a) the profit split methods; and (b) the profit comparison method (referred to by the OECD as the 'transactional net margin method' in the 1995 OECD Report).

One of the main differences between the profit split and the profit comparison method is that the former is applied to all the relevant associated enterprises whereas the latter is applied to only one of the associated enterprises. A one-sided analysis potentially can attribute to one member of an MNE group a level of profit that implicitly leaves other members of the group with implausibly low or high profit levels. However, this is also a risk with the RP and CP methods, which are also one-sided analyses. Care needs to be taken to ensure in so far as practicable that the result produced by the one-sided application of any method makes commercial sense in the circumstances of the case.

The OECD has provided considerable guidance on the use of these methods in paragraphs 3.1 to 3.57 of the 1995 OECD Report.

We agree with the OECD preference for traditional methods over profit methods. Concerns that profit methods need to be applied with caution and the difficulties that are encountered in their application are explained in detail (see paragraphs 3.8, 3.9, 3.17, 3.22, 3.24, 3.29, 3.32, 3.38, 3.39 and 3.49 of the 1995 OECD Report). We agree that profit methods should be applied in 'cases of last resort', to be used where '... there is insufficient data on uncontrolled transactions ... or where such data is considered unreliable, or due to the nature of the business situation' (paragraph 3.50 of the 1995 OECD Report and compare paragraph 435 above).

However, as with other approaches, there can be a need in some situations for an aggregation of dealings in order to obtain a reliable answer. In discussing profit methods and other less direct approximations of arm's length outcomes it is our view that there is a need in some cases to use profit methods so that the arm's length principle can be implemented as closely as practicable. Not to seek some means of testing taxpayers' dealings with associated enterprises and to devise some solution for transfer pricing problems would undermine the arm's length principle.

While it is possible to apply a profit method in respect of a single transaction, these methods are generally applied in respect of a group or a 'basket' of transactions or on an aggregated basis. Paragraphs 305 to 307 above, paragraphs 135 and 432 to 438 of TR 94/14 and paragraph 1.42 of the 1995 OECD Report set out the various ways that transactions may be combined for purposes of analysis. The important principles in this regard are: (i) to the extent transactions can be disaggregated without going to unreasonable lengths, then they should be analysed at the lower level; (ii) where transactions need to be analysed on a combined basis care should be taken to ensure that the profits, the subject of the transfer pricing examination, are limited to the profits that arise from controlled transactions. It would generally be inappropriate to apply a profit method on a 'whole of entity' basis unless all of the taxpayer's activities involved associates and, if a profit comparison is being used, the different types of controlled dealings can be approximately compared on a consistent basis with a similar basket of uncontrolled dealings by an independent enterprise operating wholly independently (see paragraphs 135 and 432 to 438 of TR 94/14). It needs to be remembered, though, that the Commissioner is entitled to estimate the arm's length consideration in respect of international agreements if there is insufficient information to determine the arm's length amount.

The profit split method first identifies the combined profit or loss from the dealings between the associated enterprises. It then splits those profits or losses 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 combined profit may be the total profit from the transactions or a residual profit intended to represent the profit that cannot readily be assigned to one of the parties, such as the profit arising from high value, sometimes unique, intangibles. The contribution to each enterprise is based upon a functional analysis and valued to the extent possible by any available external market data. In this regard, see paragraph 43 above.

The diagram on the following page broadly illustrates the profit split method.

The profit split method can be best demonstrated by the following simple example: A and B are related parties. A manufactures goods and sells them to B who re-sells (wholesales) them to unrelated parties. (A) Manufacturer (B) Reseller Sales to Reseller 100 Sales to Customers 150 Less: Less Direct materials Labour and oncost 60 Purchases from the manufacturer (A) Indirect costs Indirect costs 100 Gross Profit 40 Gross Profit 50 Selling and other costs 20 Administration and other costs 30 Administration and other costs 10 30 Net Profit 10 Net Profit 20

For the purpose of the example traditional methods will not be used. Some reasons why this might happen would be different accounting treatments that cannot be sufficiently identified in terms of particular costs to enable adjustments for comparisons on a consistent basis, or the transfer pricing issue relates to a number of 'below the gross profit line' items which may have overlaps. In applying the profit split methods the object is to determine if the split of the combined profit from the dealings between associated enterprises is arm's length in light of the functions performed, assets used and risks assumed by the respective parties from the point of starting the related party manufacture to the sale to an arm's length party. The split as originally disclosed is 20/10 in favour to the reseller. However, if the goods require a relatively small amount of marketing because of a high value intangible embedded in the product which causes buyers to seek them out and demand exceeds supply, the manufacturer's 1/3 share may not be sufficient reward for its value added. Conversely, if the product is 'yesterday's technology' and an arm's length party would have usually discontinued stocking the item then the reseller's 2/3 share may not be sufficient. If the stock is unsaleable but the taxpayer has been required by its parent to buy the stock, the purchase price should be reduced to nil. Where the stock can be sold at a much reduced price, but with considerable effort, the purchase price should be reduced to a level that would allow a reasonable return for the marketing and distribution effort and holding costs.

To determine if the split is arm's length an assessment has to be made 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. There is no one method to undertake this assessment as each case has to be decided on the data available and the facts and circumstances of that case.

The following factors need to be taken into account in undertaking a profit split: (a) The relevant dealings (i) There is a need to determine if the profit split is to be undertaken on a particular product line, a basketing of products or where appropriate on the basis of the strict guidelines above, a whole of entity basis all of which will include a consideration of any intangible assets as well as financial assets (whether shown on the balance sheet or not). In all cases an allocation of general administration and similar costs to the relevant dealings will be required. (ii) Where the taxpayer has dealings with more than one associated enterprise, care will be needed to identify the profits applicable to each party. Using the above example, if the reseller was supplied goods from two related manufacturers from different tax jurisdictions any profit split would need to identify the value added in each jurisdiction and the appropriate share of profit for each. (b) Consolidation of accounts So that the combined profit can be determined, the accounts of the parties need to be put on a common basis as to accounting practice and currency and then consolidated. Once the split has been determined the accounts can then be rewritten on a separate entity basis, taking account of the relevant requirements in the taxpayer's home jurisdiction.

A possible difficulty in attempting to undertake a profit split is obtaining the required information from foreign enterprises or tax administrations so that the combined profit can be determined. Nevertheless, reasonable attempts should be made to obtain the relevant information because there is more certainty that an arm's length share of profits has been reached when the economic contribution by all parties profit share has been examined than would be the case if a one-sided profit comparison was used.

There is a need to establish whether to apply the profit split to the projected or actual profits. The determining basis should be made as follows: (a) Projected profits Where a taxpayer uses a profit split to establish (as opposed to 'review') transfer pricing for controlled transaction, this would necessarily be done on the basis of the projected profits because the actual profits would not be know at the time. This would produce a fractional allocation which would then be applied as the actual profit derived. However, if there are variances between projected and actual profits arm's length parties would make appropriate adjustments when reviewing their profit split projections for future years. In some cases, for example where the joint activity involves an intangible and the value is unclear, arm's length joint ventures might include a review clause in their agreement that would operate to review the profit split in the event of a major change in actual profit experience relative to their projections (see paragraphs 3.11 - 3.12 and 3.25 of the 1995 OECD Report). (b) Actual profits Where prices have been set using a basis other than a profit split (as will almost always be the case) any evaluation would be undertaken on the actual profits achieved by the application of the other basis using the same information that was available at the time of the price setting thus avoiding the use of hindsight (see paragraph 3.14 of the 1995 OECD Report).

There are a number of approaches for estimating the division of profits, based either on projected or actual profits, as may be appropriate. The contribution analysis and the residual analysis are by far the most frequently used and are not mutually exclusive.

Under a contribution analysis, the combined profits, which are the total profits from the controlled dealings under examination, are divided between the associated enterprises based upon the relative value of the functions performed by each of the associated enterprises participating in the controlled dealings, supplemented as much as possible by external market data that indicate how independent enterprises would have divided profits in similar circumstances. In cases where the relative value of the contributions can be measured directly, it may not be necessary to estimate the actual market value of each participant's contributions (paragraph 3.16 of the 1995 OECD Report).

It may be difficult to determine the relative value of the contribution that each of the related participants makes to the controlled transactions, and the approach will often depend on the facts and circumstances of each case. The determination might be made by comparing the nature and degree of each party's contribution of differing types (for example, provision of services, development expenses incurred, capital invested) and assigning a percentage based upon the relative comparison and external market data (paragraph 3.18 of the 1995 OECD Report).

In the above example, the $30 combined profit on a contribution analysis would be split between A and B based on the their contribution to the total assets employed (including non financial assets), functions undertaken and the risked assumed to achieve that combined profit. The value of their contribution would be determined, where possible, by external market data, i.e., from comparable uncontrolled enterprises performing comparable functions and operating wholly independently.

A residual analysis divides the combined profit from the controlled transactions under examination in two stages: (a) First, each participant is allocated sufficient profit to provide it with a basic return appropriate for the type of transactions in which it is engaged. Ordinarily, this basic return would be determined by reference to the market returns achieved for similar types of transactions by independent enterprises. Thus, the basic return would generally not account for the return that would be generated by any unique and valuable assets possessed by the participants. (b) Secondly, any residual profit or loss remaining after the first stage division would be allocated among the parties based on an analysis of the facts and circumstances that might indicate how this residual would have been divided between independent enterprises. Indicators of the parties' contributions of intangible property and relative bargaining positions could be particularly useful in this context.

The following example illustrates how a residual profit split is calculated: Foreign Co manufactures goods which it sells to its associated enterprise, Aust Co, which re-sells the goods to unrelated parties. The total combined profit from the operations is $1000, Aust Co is rewarded $150 for the marketing, distribution and other functions undertaken (based upon an analysis of typical returns for that type of business activity) while the Foreign Co is rewarded $250 (based upon an analysis of returns for similar manufacturing functions). The remaining profit of $600 is then allocated on the basis of the contribution of each of the enterprises to the value of the intangible, say 10% (being $60) to Aust Co and, say, 90% being $540 to Foreign Co. Profits Aust Co Foreign Co Total Profits A, F & R* 150 250 400 Intangibles 60 (10%) 540 (90%) 600 Total 210 790 1000 *basic tangible assets, functions and risks. While there is usually agreement on the concept of the methodology when allocating profits, enterprises often fail to use the same logic where an overall loss has been incurred. In the following example where the total loss from operations is $500, Aust Co is still rewarded $150 for the marketing, distribution and other functions undertaken while Foreign Co is still rewarded $250 for the manufacturing function undertaken. The residual loss of $900 is then allocated on the basis of the contribution of each of the enterprises to the value of the intangible, say, 10% being $90 to Aust Co and, say, 90% being $810 to the Foreign Co. Losses Aust Co Foreign Co Total Profits A, F & R* 150 250 400 Intangibles 90 (10%) 810 (90%) 800 Total 60 560 500

Referring to the earlier example in paragraph 447, it appears at first glance that the amount of combined profit to be split is $30 (the $10 of manufacturer's profit plus the $20 of reseller's profit). This assumes there are no other transfer pricing issues other than the appropriate profit split. However, the basic return on an arm's length basis could result in either a different split of the $30, or one party making a loss and the other making more than $30. If the combined basic return is less than $30 this may indicate that there are intangible assets which are adding value to the goods. However, other factors like corporate strategies, management efficiency and financials would need to be reviewed before firming up on this view. If the combined basic return is more than $30 it would indicate that either the intangible assets (or perhaps liabilities) are the source of the loss or that other factors like financials, management strategies and efficiencies need to be reconsidered.

There is no one basis of determining the basic return or a division of the residual profits. The most appropriate basis should, as closely as is possible, split the profits on a basis that would have been realised had the parties been independent enterprises operating at arm's length. The model to have in mind here is a joint venture between independent parties operating wholly independently in the way they seek to protect and enhance their own economic interests. The following may serve as a guide to undertake the residual profit split: (a) One approach to a residual analysis would seek to replicate the outcome of bargaining between independent enterprises in an open market. In this context, the basic return provided to each participant would correspond to the lowest price an independent seller reasonably would accept in the circumstances and the highest price that the buyer would be reasonably willing to pay (see paragraph 66 to 68 of TR 94/14). Any discrepancy between these two figures could result in the residual profit over which independent enterprises would bargain. The residual analysis therefore could divide this pool of profit based on an analysis of any factors relevant to the associated enterprises that would indicate how independent enterprises might have split the difference between the seller's minimum price and the buyer's maximum price (paragraph 3.21 of the 1995 OECD Report); (b) In some cases an analysis could be performed, perhaps as part of a residual profit split or as a method of splitting profits in its own right, by taking into account the discounted cash flow to the parties to the controlled transactions over the anticipated life of the business. This may be an effective method where a start-up is involved, cash flow projections were carried out as part of assessing the viability of the project, and capital investment and sales could be estimated with a reasonable degree of certainty. However, the reliability of such an approach will depend on the use of an appropriate discount rate, which should be based on market benchmarks (see paragraph 3.22 of the 1995 OECD Report); (c) The development expenditures incurred to create the factors which gave rise to the residual may also provide an indication of the relative contributions of the enterprises and may in some cases be an appropriate basis for allocating the residual. When adopting this basis for a profit split regard should be had to those matters discussed in paragraph 450 above.

Another approach is to split the combined profit so that each of the associated enterprises participating in the controlled transactions earns the same rate of return on the capital it employs in that transaction. This method assumes that each participant's capital investment in the transaction is subject to a similar level of risk, so that one might expect the participants to earn similar rates of return if they were operating in the open market. It also assumes that return on financial capital is a relevant measure for each enterprise (paragraph 3.24 of the 1995 OECD Report).

Another possibility is to determine the profit split based on the division of profits that actually results from comparable transactions among independent enterprises. In most cases where the CUP, RP and CP methods would not be used, it will be difficult to find independent enterprises engaged in transactions that are sufficiently comparable to use this approach as the primary method (see paragraph 3.25 of the 1995 OECD Report).

Formulary approaches are often appropriate for splitting profits from global trading. They can also be used reasonably reliably for allocating some types of head office expenses to associates, provided the principles in paragraphs 412 to 419 above are followed. The ATO will not prevent the use of a formulary approach in establishing arm's length outcomes, particularly where it is not possible or practicable to allocate an arm's length profit or to ascertain an arm's length price in accordance with other methods endorsed by the OECD. Reserving the option to use a formulary approach in appropriate cases on the basis of their facts and circumstances is consistent with the operation of subsection 136AD(4) of the ITAA.

Where an MNE is engaged in a global trading activity in financial products through various markets around the world, sometimes on a twenty-four hour basis, there are some unique problems associated with the allocation of income and expenses amongst the contributing members - especially where the same pool of trading assets is used.

In some cases, the members have a fully integrated computerised trading network containing the various files (collectively known as a 'book') which is used to transact the deals. The authority to trade in this stock of financial instruments may move from one centre to the next as markets open and close, the authority continually being passed around the world to maximise the trading on the book. A deal may be opened in New York, continued in Sydney and Singapore and closed in London. During this time, the members of the multinational group who are dealing on the relevant book are taking orders from their clients at any time during the day, conducting trades for clients (from anywhere in the world) when their market is open, subject to the financial limits imposed on the dealers individually in relation the exposure they can take by setting an overall limit on risks and collectively by the multinational group - and perhaps regulators. Other functions would typically include overall strategic management, back office support like accounting, legal documentation and computer support, and sales and marketing.

In these circumstances, finding suitable comparables for the trading are not of real concern. The relevant financial markets would most likely provide the necessary arm's length considerations for the transactions entered into by each participant in the global trading of the multinational group. The real problem lies in how to allocate the income and expenses on an arm's length basis. For example, and as explained above, a deal may be opened in New York, continued in Sydney and Singapore and closed in London. By closing the deal in London, the group member in London will crystallise all the profit or loss. Yet, the profit or loss crystallised in London might effectively be attributable to all multinational group participants in the global trading. Generally speaking though the group member in London should not recognise all the profit or loss in London because in the great generality of cases dealers will all have a certain level of skills in predicting markets, taking positions and hedging risks, there will be a level of prudential control and the activity by its nature contains an element of speculation. But more importantly, it is the fact that the profits arise largely from the integration of trading, hedging and marketing that force the conclusion that it is not the place where the profit or loss crystallises that is the determining factor in allocation. This integration is also a reason why dealers' salaries/bonuses are not a sufficient basis in themselves for allocating profit.

Primarily, the allocation of profit or loss is to be determined in accordance with the economic benefit each group member has contributed to the overall result from participating in the global trading. To determine the level of benefit, it would be necessary to complete a functional analysis establishing who does what and where, and then to allocate the income and expenses on some formula based on a suitable weighting of the assets, functions and risks of each contributing member (participant). The large volume of global trading may also prohibit anything other than a profit allocation on the basis of aggregating all trading transactions in financial instruments.

Where possible, the weighting should be based on some form of external market data. The outcome sought should be directed to reflecting what independent enterprises would have done if they were confronted with the similar allocation problem in comparable circumstances. Differences in functions, assets and risks in different cases should be reflected in the allocation of profit.

A common form of global trading - perhaps the most common - is where the trading in a specific product is centralised in a particular jurisdiction. The decision on where to locate that book would be driven by commercial considerations like customer location, availability of trading skills and depth of market. There may be limited authority to trade outside the jurisdiction, but the main activity in other jurisdictions would be sales and marketing. It is also possible for back-office support to be located outside the 'book's' jurisdiction. This particular form of global trading does not necessarily require a profit split methodology - the argument for which is stronger where there is a very high level of integration - and it may be possible to use traditional methods.

Profit comparisons at the net margin level are an extension of the RP and CP methods. The profit comparison method examines the net profit margin, rather than the gross profit margin (or at some margin in between), relative to an appropriate base (e.g., costs, sales, assets), that an enterprise realises from a controlled transaction or from transactions that it is appropriate to combine in accordance with the principles referred to in paragraphs 309 to 314 above, paragraphs 135 and 432 to 438 of TR 94/14 and paragraph 1.42 of the 1995 OECD Report.

As such, when applying the profit comparison method, care is needed to have regard to the previously discussed requirements for the application of the RP and CP methods and the principles regarding comparability.

The following diagrams broadly illustrate the profit comparison method.

Data to apply the profit comparison method and to determine the appropriate net margin will be obtained from the functional analysis of the taxpayer and the comparable enterprises. This method requires the taxpayer (or the ATO) to compare the net margins obtained in its controlled dealings against either: (a) the net margins of the taxpayer's uncontrolled transactions in comparable circumstances; or (b) the net margins earned in comparable, uncontrolled transactions by an independent enterprise.

It is important to ensure the profit comparison is confined to the net profit from cross-border dealings with associated enterprises. The OECD has termed the profit comparison method the 'transactional ' net margin method to emphasise the need to confine the analysis to a particular transaction or transactions, viz, those between associated enterprises. However, this terminology by the OECD does not preclude the application of the profit comparison method on the basis of an aggregation of transactions where appropriate, though the reliability of such an approach relative to the application of the CUP, RP and CP methods on a similar basis, would need to be rechecked and the most reliable method used (see paragraph 3.50 of the 1995 OECD Report).

As mentioned above, when applying the RP and CP methods the appropriate profit margin is usually measured at the gross profit level. However, in some circumstances it may be more accurate to undertake the comparability analysis at the net profit level. One of the possible difficulties in transfer pricing analysis is the lack of adequate reliable data. If detailed comparable data is unavailable or there are accounting differences, that cannot be reliably adjusted to allow comparisons at the gross profit level, or the transfer pricing issues arise in respect of items below the gross profit line, comparisons at levels other than net profit may produce an incorrect outcome. For example, if selling, general and administrative costs that are treated as part of costs of goods sold for an independent enterprises cannot be identified so as to adjust the gross margin in a reliable application of cost plus, it may be necessary to examine net margins in the absence of more reliable comparisons.

The net margins also may be more tolerant to some functional differences between the controlled and uncontrolled transactions than gross profit margins. Differences in the functions performed between enterprises are often reflected in variations in operating expenses. Consequently, enterprises may have a wide range of gross profit margins but still earn broadly similar levels of net profits. It is therefore important when applying the profit comparison method to carefully consider the functional analysis of the taxpayers and the entities being considered as possible comparables and to make adjustments whenever practicable to increase the degree of comparability, having regard to the previous discussion of the factors that could impact on comparability.

While the RP and CP methods are each based on a single ratio, there are several ratios that could be of assistance in applying the profit comparison method. The relative usefulness of the various ratios will depend on the facts of each case and the extent of reliable data available. For example, the rate of return on capital employed will be of greater importance if the taxpayer is operating in a capital- intensive industry. It would often be appropriate to have regard to more than one ratio in checking the reliability of the taxpayer's approach to determining transfer pricing. Return on assets could present particular problems when used in isolation (see paragraph 487 below).

An EBIT analysis can be helpful in the context of profit comparisons - bearing in mind the need to focus only on the relevant cross-border dealings between associated enterprises and to apply the EBIT approach consistently to the taxpayer and the other enterprises being considered as possible comparables.

Another possibility is the ratio of operating profit to sales, but the safeguards in relation to comparability that are needed when operating expenses are being considered should be carefully followed.

The ratio of gross profit to operating expenses (often referred to as a Berry ratio) can also be helpful in applying the profit comparison method, though care would be needed to ensure that comparisons between the taxpayer and other enterprises are limited to other enterprises that have a high degree of functional similarity.

It would also be essential to be aware of the possible distorting effects of methods of business financing, business strategies and the relative efficiency of managers when doing EBIT and Berry and operating profit to sales ratios.

It needs to be borne in mind that the application of these various ratios in appropriate cases will indicate potential transfer pricing risks at a primary level. However, further detailed analysis will be needed to identify particular transfer pricing problems. Any comparability analysis will need to have regard to the potential difficulties presented by a consideration of operating expenses (see paragraphs 476, 479 and 480 above).

The four step process set out below will assist an understanding of the taxpayer's business and help identify the key profit drivers and the aspects of the case which might present transfer pricing problems. This information will in turn assist the selection and application of a methodology. In particular, it will assist in determining the applicability and reliability of the various ratios in the circumstances of the taxpayer's case.

Possible distortions through economic, market, business or product cycles would also need to be considered. The use of data covering a number of years will increase the reliability of profit comparison (see paragraphs 46 - 47 above).

As mentioned in the discussion on the RP and CP methods, a comparison at net profit level can sometimes be needed, but care is required when dealing with operating and financial expenses that relate to such things as marketing strategies, the management efficiencies, accounting policies (e.g., different depreciation methods) and business financing alternatives.

An uncontrolled enterprise may not accept to pay a higher price resulting from the inefficiency of the other party. On the other hand, if the other party is more efficient than can be expected under normal circumstances, the other party should benefit from that advantage. However, where company policies are determined or influenced by an associated enterprise then the costs impacted by these policies would need to be carefully considered when undertaking the comparability analysis for the reasons mentioned in the discussion on the RP and CP methods.

Various factors will have a potential impact on the reliability of the profit comparison method and the comparability analysis should carefully address them. For example, these factors would include how well the value of assets employed in the calculations is measured. In each case it is essential to determine the extent to which the value of the intangible property is not captured in the books of the enterprise. This as an important issue in analysing functions, assets and risks. The books may show assets at historical cost or as capitalised expenditure - which can be markedly different from their real value. Goodwill derived from product or service quality or from research and development, or from skills in distribution or the creation of marketing intangibles will often create value that is not shown on the balance sheet. These can be masked by bad management or poor financials. Some insights might be gained from market capitalisation and earnings history (e.g., earings per share). It is important that all these aspects be considered when doing the functional analysis as a good understanding can be developed of the taxpayer's strengths and weaknesses in terms of the profit drivers and features likely to reduce profitability. The factors affecting whether specific costs should be passed through, marked up, or excluded entirely from the calculation will be better understood if this approach is followed and when those factors are then taken into account the profit comparison should produce closer comparables than profit comparisons that ignore these issues.

The application of the profit comparison method requires a careful analysis of the taxpayer's operating expenses. These expenses reflect the taxpayer's functions, assets and risks and give insight into the possible distorting effects of methods of business financing and management approaches in relation to the net profit. These concerns are largely avoided with the RP and CP methods because they focus on the gross margin. However, where for the reasons outlined the RP and CP methods cannot be used, the more closely operating expenses, market and business factors can be analysed and compared the more reliable the profit comparison method. Of course care would also be needed to ensure that expenses above the gross profit line, market and business factors are also properly considered (see also paragraphs 3.35 - 3.36 of the 1995 OECD Report).

To illustrate the basic concepts of the profit comparison methods two charts (one applied on the net resale price basis and the other on the net cost plus basis), with supporting notes, are set out below: PROFIT COMPARISON (PC) METHOD APPLIED ON A NET RESALE PRICE BASIS In Relation to Associated Enterprise Dealings - Arm's Length Amount Explanation SALES - To Uncontrolled Party 1000 Arm's length Sales PURCHASES - From Associated Enterprises 600 Amount determined by PC Method GROSS PROFIT 400 Selling, General and Admin., (SG and A) Expenses 300 Amount determined by PC Method if expenses from associated enterprise dealings are included in SG and A Net operating Profit 100 Appropriate net margin in light of the functions assets and risks based on the analysis of comparables. Formula: (1) If no expenses from Associated Enterprise dealing included in SG and A Arm's length Sales less (Appropriate net margin plus SG and A) = Purchase Price 1000 - (100 + 300) = 600 (2) If expenses from Associated Enterprise dealings included in SG and A Sales less Appropriate net margin = Arm's length (Purchase price and SG and A expenses) 1000 - 100 = 600 + 300 Key: To determine the appropriate net margin so as to deduce the arm's length amounts for all associated enterprise dealings. Note: This method would be used where there is insufficient reliable data or the complexities of real life business put practical difficulties in the way of the application of traditional transaction methods. PROFIT COMPARISON (PC) METHOD APPLIED ON A NET COST PLUS BASIS In Relation to Associated Enterprise Dealings - Arm's Length Amount Explanation SALES - To Associated Enterprise 1000 Amount determined by PC Method Cost of Production (Goods & Service) - Direct Costs - Indirect Costs 600 Amount determined by PC Method if costs from associated enterprise dealings are included in costs of production GROSS PROFIT 400 Selling, General and Admin., (SG and A) Expenses 300 Amount determined by PC Method if expenses from associated enterprise dealings are included in SG and A Net operating Profit 100 Appropriate net margin in light of the functions assets and risks based on the analysis of comparables. Formula: (1) If by way of associated enterprise dealings, amounts are included in sales, cost of production and SG & A) Appropriate net margin = Selling price less (Costs of Production + SG & A) 100 = 1000 - (600 + 300) (2) If associated enterprise dealings are confined to Sales Cost of production plus SG & A plus Appropriate net margin = sales 600 + 300 + 100 = 1000 Key: To determine the appropriate net margin so as to deduce the arm's length amount for all associated enterprise dealing. Note: This method would be used where there is insufficient reliable data or the complexities of real-life business puts practical difficulties in the way of the application of traditional transaction methods.

Where there is a special relationship between the associated enterprises that produces unique dealings, there may not be sufficient data or comparable dealings to apply CUP, RP, CP or the profit methods. In such cases their use should be reconsidered on the basis of possibly: (a) sufficiently broadening the comparability criteria to allow a comparison of the relevant dealings; and (b) sufficiently relaxing the normal conditions imposed in applying traditional and profit methods to allow the broadened comparability to be applied.

This situation may come about where all the comparable enterprises in an industry are associated enterprises or where an associated enterprises have an industry monopoly. Where this is the case the appropriate arm's length comparison may be with enterprises in another industry segment or group of segments. However, great care is needed to ensure that the industry segments or groups of segments being compared are sufficiently similar, especially in relation to levels of profitability as well as functions performed (see paragraph 3.34 of the 1995 OECD Report).

Where the comparability criteria need to be broadened there may be a need to consider the dealings on a basketing, aggregated basis or, in extreme cases on a whole of entity basis. However, the relative reliability of such an approach will need to be considered against the reliability of the applying the CUP, RP or CP method on an aggregated basis - if it is possible to use the traditional methods that way.

If the extended application of the traditional and profit methods cannot provide an answer it may be necessary to consider: (a) a mixture of the above methods; or (b) some other method or mixture of methods; that is likely to lead to a result that is consistent as practicable with the arm's length principle (see paragraph 367 of TR 94/14). Where taxpayers find themselves in this category they should give serious consideration to an advanced pricing arrangement ('APA').

The 1979 OECD Report (' Transfer Pricing and Multinational Enterprises, Report of the OECD Committee on Fiscal Affairs ', 1979) discusses some other methodological approaches. Some enterprises have adopted these approaches and in some circumstances they may remain appropriate even though they have not been discussed in the same terms in the 1995 revision of the Report. In considering other approaches the 1979 OECD Report noted: '70. Cases will often occur where there is no satisfactory evidence of comparable uncontrolled prices and where it may also be difficult to apply either the resale price method or the cost plus method. In such instances, it may be necessary to adopt other reasonable approaches for arriving at an arm's length price. It may also be necessary to have recourse to additional calculations and computations to cross check a price derived according to one or other of the methods previously described. Some other methods used in practice are discussed in the following paragraphs, though the list is by no means exhaustive. They all need to be used with care. Comparable profits 71. Tax authorities may find some help in a comparison of an enterprise's overall performance with that of other similar enterprises in the same of similar circumstances. Levels of profit in an industry may for example conform to a pattern and an exception to the pattern might indicate that profits were being shifted by artificial transfer prices. But comparisons of this sort would need to be made with care. It does not necessarily follow that exceptional profits are artificial. Moreover there are many ways of looking at the profit situation of an enterprise - analysing for example what relation the profit bears to total receipts from sales, or to the operating expenses incurred, and a method which is appropriate for one company may not be appropriate for another. It may be helpful nevertheless to make comparisons of this sort in relation to the gross profits to sales of particular products or groups of products but even so the results of the comparison could normally be regarded only as pointers to further investigation. 73. By a somewhat similar process the reasonableness of transfer prices may perhaps be assessed by comparing the yield or return on capital invested in the relevant associated enterprises with the yield or return on the capital invested in enterprises carrying on similar activities and requiring the same kind of capital investment. This too presents difficulties. One problem is that the financial structuring of the relevant entities would need to be essentially similar if this comparison was to be useful and if the structure was not essentially similar (because for example of the different impact of obligations to pay interest) then difficult adjustments would have to be made to calculations in order to ensure comparability. Net yield expectations 74. A further approach might be to look at the yield on the capital involved, for example, to assume in the case of transactions between a parent and a subsidiary company that the subsidiary's arm's length profit would be related in some way to the return which would be expected from investing the relevant capital in other ways. But this too is an arbitrary assumption: the profit made by a comparable independent concern would not necessarily bear any particular relationship to the return on investment in other ways. Although therefore such an approach may have some value in indicating a reasonable range of possible margins the concept of a "normal return" on capital is too imprecise to be likely to be useful in isolation.'

The OECD's qualifications on the reliability of these approaches is noted. They are also repeated in the 1995 revision in the discussion of profit methods. Nevertheless, we recognise that they have on occasions been used by enterprises and they may have a role in assisting the establishing of an approximation of an arm's length consideration in some cases. In this regard, it needs to be remembered that all of Australia's DTAs would, on the wording of various Associated Enterprises Articles, allow methods that may not be permissible under the wording of the Associated Enterprises Article (Article 9) of the OECD Model Tax Convention (see paragraphs 18 to 28 above). It also needs to be remembered that subsection 136AD(4) which is preserved by all of Australia's DTAs, also applies in appropriate cases where Australia does not have a DTA - and in DTA cases outside the scope of the relevant Associated Enterprises Article. However, as stated in paragraphs 9 to 11 above, the choice and application of non-traditional or profit methods are limited to the statutory purpose of achieving an arm's length outcome.

CPM is a form of profit comparison which was included in the final regulations issued by the US Treasury Department in July 1994 to cover section 482 of the Internal Revenue Code. It reflects an assessment of the total returns on the business activities achieved by broadly similar but uncontrolled taxpayers.

There has been international debate about CPM including argument that it does not satisfy the arm's length principle. The method is not directly discussed in the OECD's 1995 Report. However, the earlier proposed US 482 Regulations were considered and commented upon in 1993 in a Task Force Report of the OECD which contributed to the increased flexibility now found in the current US Regulations.

The US CPM is a regulated process for developing an answer claimed as being arm's length. Because it relies upon data being used in a specified manner, it may not necessarily be flexible enough to deliver the best approximation of an arm's length result in the particular circumstances of a case. The reservations already discussed about the use of profit methods should also be noted when considering the use of the US CPM.

However, CPM can be distinguished from non arm's length methods which use administrative formulae, because it involves an examination of the financial performance, functional characteristics and industry grouping of the enterprise to benchmarks set by other enterprises on the basis of comparability.

What constitutes a broadly similar enterprise is an important consideration when using the method. Also, if data from other countries is used, the application of the US CPM may not properly reflect the circumstances of the market in which the enterprise is operating. Data that is not comparable can lead to inappropriate results.

Where the taxpayer in complying with the requirement of IRS Code 1.482-5(b)(3) develops an arm's length range in line with the IRS Code 1.482-1(e)(2) B and C, the excision of the upper and lower quartiles from the sample used to calculate the comparable profit tends to produce average outcomes. For some enterprises that are leaders in their industry segment, this method may produce figures which are not an accurate reflection of an arm's length outcome. For example, an analysis of the functions, assets and risks of a successful enterprise may suggest that its financial results should be above the industry average, even though its reported results are below the industry average. By only using the inter quartile results under CPM, the method may not enable comparisons with similarly successful enterprises that are consistent with the market success, functions, assets and risks of the enterprise being examined. Similar distortions could result in cases of less than average performance.

CPM calculations can undervalue certain types of functions and overvalue others where there is a high degree of sensitivity in the results to the data elements that have been selected. For example, an enterprise may undertake complex manufacturing functions which embody significant intangibles. A comparable rate of return that is based exclusively on the results of contract manufacturers may significantly undervalue the return that would be expected from a complex manufacturing operation. Similarly, an enterprise may undertake significant value added activities through its marketing and distribution functions. However, if the method results in a comparison only with high volume/low margin distributors, the value added of the enterprise through its marketing and distribution function may not be properly rewarded. Nevertheless, difficulties associated with questions of comparability are not confined to CPM.

The need for the development of an approach such as CPM highlights the obstacle that insufficient reliable data can present for decision makers in this area. In this context, CPM, or a substantially similar approach, may provide some help in very extreme cases where more reliable data is not able to be examined or does not exist.

In earlier discussion about the arm's length principle it was noted that an enterprise may need to establish an approximation of what is arm's length even in cases where close comparables do not exist or the dealings are not of a type that arm's length enterprises would conduct. Because CPM, or approaches based on similar profit comparison techniques, can use external data which is available, there is a role for this type of analysis. It is also recognised that enterprises which fall within the jurisdiction of the US Regulations need to comply both with the US laws that have been enacted and the Australia/USA Double Tax Convention and competent authority processes should be used in appropriate cases to ensure the closest practicable approximation of an arm's length result and the lowest risk of double taxation.

Global formulary apportionment is a method that has sometimes been suggested as an alternative to the arm's length principle as a means of determining the proper allocation of profits across competing national tax jurisdictions.

Paragraphs 3.59 - 3.60 of the 1995 OECD Report make the following comments: 'A global formulary apportionment method would allocate the global profits of an MNE group on a consolidated basis among the associated enterprises in different countries on the basis of a predetermined and mechanistic formula. There would be three essential components to applying a global formulary apportionment method: determining the unit to be taxed, i.e., which of the subsidiaries and branches of a MNE group should comprise the global taxable entity; accurately determining the global profits; and establishing the formula to be used to allocate the global profits of the unit. The formula would most likely be based on some combination of costs, assets, payroll, and sales. 'Global formulary apportionment methods should not be confused with the transactional profit methods [which equate to profit splits and the profit comparison method in the case of Australia] ... The former methods would use a formula that is predetermined for all taxpayers to allocate profits whereas transactional profit methods compare, on a case by case basis, the profits of one or more associated enterprises with the profit experience that comparable independent enterprises would have sought to achieve in comparable circumstances. Global formulary apportionment methods also should not be confused with the selected application of a formula developed by both tax administrations in co-operation with a taxpayer or MNE group after careful analysis of the particular facts and circumstances, such as might be used in a mutual agreement procedure, advance transfer pricing agreement (sic), or other bilateral or multilateral determinations. Such a formula is derived from the particular facts and circumstances of the taxpayer and thus avoids the globally pre-determined and mechanistic nature of global formulary apportionment methods.'

The OECD member countries, including Australia, do not consider global formulary apportionment to be an acceptable alternative to the arm's length principle for a number of reasons. A principal reason is that global formulary apportionment can depart from the territorial connection that underpins the concept of source and may also raise issues about the timing of derivation. Equally important is the concern that predetermined formulae that are mechanistically applied do not have regard to the facts and circumstances and merits of the particular case - the result being that in many cases they result in either overtaxation or undertaxation. They also depend on a very high degree of international co-operation and co-ordination. The capacity for multinational groups to manipulate the formula and the inability of most formulae to capture the particular circumstances of individual enterprises, their risks, geographical differences and differences in company efficiencies are serious drawbacks with this method. Also, currency exchange rate movements and inconsistent accounting standards between countries could lead to perverse profit allocations. Dispute over the acceptability and use of particular formulae which have different bases may mean that the expected benefits of no double taxation and lower compliance costs may not be realised.

In some cases, a formula developed by both tax authorities in co-operation with a specific enterprise after careful analysis of the particular facts and circumstances, such as might be used in an Advance Pricing Arrangement, would be appropriate to determine a fair allocation of revenue to the countries involved. However, these formulas are not instances of global formulary apportionment.

The arm's length principle, the question of comparability and the appropriate methodological approach can be incorporated into a structured approach for setting and reviewing the outcomes of international dealings between associated enterprises.

This part links the discussion on the application of the arm's length principle, the question of comparability and the selection of the most appropriate method with a four step process by which a taxpayer, its advisers or the ATO can address conformity with the arm's length principle. When used by taxpayers, this process can assist the development of the reasoning and documentation needed to support their evaluation.

A taxpayer may wish to adopt this approach in several situations. First, by adopting a particular methodology, the taxpayer may be able to set the terms and condition for their dealings with associated enterprises at the time they were contemplating or entering the arrangements. Secondly, where other approaches for arriving at a consideration were used for management purposes, these may need to be reviewed at the time tax returns are being prepared and adjusted, if necessary, to the arm's length consideration for tax purposes. Finally, the taxpayer may wish to satisfy themselves, or be asked by the ATO to demonstrate, that the commercial practices or other approaches used in the international dealings between associated enterprises achieve an outcome consistent with the arm's length principle.

Documentation will be especially important for tax purposes where the enterprise is experiencing a regular turnover of key staff who can explain the dealings, or where the enterprise adopts strategies that have a measurable effect on the arm's length outcome (e.g., particular pricing strategies) or where those strategies have not previously been documented.

In suggesting an approach, four points need to be made: (a) The four step process and the data collection and analysis outlined in this part are neither mandatory nor prescriptive approaches. The processes adopted for the review need to be tailored to the facts of the case. (b) The approach outlined below assumes that the nature of the international dealings is fairly extensive and necessitates a thorough analysis. For many enterprises which have relatively simple and/or low value international dealings with associated enterprises the extent of data collection and analysis may be minimal. For example, an enterprise may have an overseas subsidiary which conducts extensive business operations and deals in a wide range of goods and services. However, dealings between associated enterprises may be limited to the provision, by the parent, of a long term loan secured by the assets of the subsidiary. In these circumstances, the detailed issues and analysis contemplated in the following part will be largely irrelevant. Similarly, if a taxpayer has extensive dealings with associated enterprises but also has extensive dealings of the same kind and in similar circumstances with uncontrolled enterprises operating independently, a more limited analysis is sufficient. (c) It may be possible in some cases to adopt either a methodology or a specific price that has been developed and applied by a MNE on a global basis after some confirmatory analysis or consideration of its suitability and reliability in relation to the Australian enterprise. However, the data used to support the methodology will need to be carefully considered in terms of its relevance and reliability for Australian market conditions. (d) From the perspective of an enterprise that is seeking to set, review or justify the terms and conditions of its international related-party dealings, the analysis outlined in this part complements the documentation created by enterprises in the normal course of their business dealings. It needs to be remembered that associated enterprises will in many situations need to show that their association has not inappropriately impacted on the nature or terms of their dealings. Given the absence of the economic tension that exists between independent enterprises dealing wholly independently with each other, associated enterprises will often have to do analyses and keep records to show the arm's length nature of their dealings in circumstances where independent enterprises operating wholly independently could merely rely on their normal business records. This additional requirement cannot be removed without sacrificing the integrity of the arm's length principle and the underlying policy of the transfer pricing rules in Division 13 and Australia's DTAs.

The taxpayer or ATO staff will need to understand the nature and extent of the dealings between the taxpayer and associated enterprises in the context of the taxpayer's business. It is important to be able to explain how the international related-party dealings of the enterprise are undertaken, the purpose or object of the dealings, what the taxpayer obtains from its participation in them and their significance to the taxpayer's overall business activities and those of the multinational group.

The insight developed in this first stage will assist in determining the extent of any functional analysis that might be needed and in doing the analysis of comparability that is central to any application of the arm's length principle.

The size, scope, type and value of the taxpayer's international dealings with associated enterprises will need to be documented and the records available in Australia. Some of this information may be in marketing or cost accounting areas of the business. Much of this information will be available from the financial records of the enterprise, or be recorded in the business records of the enterprise. The enterprise is already obliged to collate some information of this type in order to complete the Overseas Transaction Information section of Schedule 25A which forms part of the income tax return. In many cases more detailed information will need to be sought to more thoroughly identify aspects of the property involved, the nature of the dealings and which part of the taxpayer's organisation conducts those dealings.

It is useful to also identify relevant arm's length dealings of the taxpayer because it might be possible to use them as comparable uncontrolled dealings. Its dealings with associated and uncontrolled enterprises may be sufficiently similar in nature, frequency and size as to demonstrate readily that the dealings with associated enterprises are producing an arm's length outcome. Establishing the arm's length consideration in this situation should be relatively straightforward where the enterprise does not distinguish between these types of dealings in its internal processes, including in the allocation of costs.

The following two examples demonstrate how different approaches would be necessary in determining the arm's length outcome, depending on the degrees of complexity of the relevant businesses and the availability (or absence) of data on comparability.

Company 1 is an Australian company which has two markets of similar size and characteristics in the USA. It sells its finished goods to a subsidiary in San Francisco and significant quantities of the same goods to an unrelated distributor in Los Angeles on the same terms and conditions. The arm's length distributor performs essentially the same functions as the subsidiary. There are no other features that might affect comparability. Establishing the arm's length consideration in this situation should be relatively simple because there are arm's length sales of an identical product under the same terms and conditions in comparable circumstances. It should be noted that a limited form of functional analysis would be needed to ensure that the internal comparable was truly comparable in all material respects.

Company 2 has an exclusive agreement to import and distribute finished goods obtained from associated enterprises. The company also manufactures finished goods which incorporate components supplied by an associated enterprise, and it exports some of its own manufactured components and finished goods exclusively to other associated enterprises analyses and data collection will generally be necessary for each of those business lines to establish the arm's length outcomes. Depending upon the availability of reliable data on comparable dealings, one result may be the application of different methodologies to each of the business lines to determine the arm's length consideration.

Many enterprises undertake a range of business activities or have a range of business lines. The key characteristics of these activities or business lines will need to be identified to enable the most appropriate method to be adopted in each case. Some of this information should be recorded in the management records of the enterprise, or would be known to the senior operational staff. The manner of collection and assembly of this information will vary. However, it could be expected that the use of most approaches in cases involving significant cross-border dealings with associated enterprises will require this information to be collated in a form that enables the preparation of a functional analysis.

The extent of the dealings between associated enterprises will need to be identified. Any documentation relevant to these dealings should be identified and the terms of any undocumented agreements should be established from the business records and the conduct of the parties. In some cases the actual dealings can be different from the contracted terms. This can happen where the dealings extend over a long period and the parties modify their responsibilities but do not reflect these changes in the formal agreement. It would be important to determine the reasons why the original agreement was varied and whether the changes favour one or both parties and are in accordance with the reasons for the changes. A position that accurately reflects the facts and circumstances of the dealings will need to be established.

When examining the dealings it is also relevant to establish what systems, methods and procedures the taxpayer adopts for establishing the terms and conditions of dealings and whether they are applied uniformly in every case. Knowledge of how the dealings are conducted and the internal controls surrounding them can assist in gauging the likelihood that past dealings have been conducted in accordance with the arm's length principle.

Enquiries into these processes should also reveal the extent and nature of the plans, performance reports, statistics, etc., produced by the taxpayer. It may also be necessary to examine a range of other material such as strategy documents or marketing plans, forecasts, costings, bids, capital expenditure requests and budgets. This helps develop an understanding of the business and the context in which the dealings are conducted. It also establishes the scope and depth of material upon which later analysis can draw.

Some enterprises may find that documents (such as annual return filings, prospectus, etc.) lodged with corporate regulatory authorities are a useful source of information. For example, the level of disclosure required in the USA by corporate regulators and taxation authorities may mean that, in some cases, useful descriptions and summaries of the information discussed in this part may be readily available in an authoritative and useful form.

The formal organisational and capital structure of the enterprises that are parties to the dealings may also need to be reviewed. The corporate structure of the group would usually be established for both the formal lines of ownership, control, reporting and authority as well as the pattern of dealings between associated enterprises and how management performance is rewarded. Within particular enterprises in the group it may be necessary to review the organisational structure and decision making systems and processes. Frequently this information is formally recorded by Human Resource managers. Examining these aspects can give an insight into the nature and purpose of the dealings between the taxpayer and other group companies and may indicate non-arm's length features of the relationships.

The staff of the enterprises obviously will have the most detailed knowledge about the business and its operations. It may be useful to obtain information from a range of key managerial and supervisory staff to assist in obtaining an accurate perspective of the functions, assets, risks and operational aspects of the business. These staff will generally be able to provide or prepare information that will help identify those factors that are affecting organisational performance (for better or worse) and whether these influence the dealings between the associated enterprises in a way that is inconsistent with the arm's length principle.

It is also important to understand the nature of the industry and the markets within which the enterprise is conducting its business; the nature of competition experienced in its business dealings; and any broader economic and other factors affecting the taxpayer's business. This knowledge can assist in the preparation of a functional analysis and in determining comparability. For example, the pricing structures and profitability of enterprises in the Australian pharmaceutical industry are affected by the Pharmaceutical Benefits Scheme. The effect of the government regulation on the profitability of enterprises in the industry would need to be taken into account in any analysis of comparability.

There is an expectation that the outcome from international dealings will reflect the conditions affecting the industry and the position of the enterprises within that industry. Changes in an industry should not be ignored. In general, in an arm's length situation it would be unlikely for one party to the dealings to be able to maintain its profitability regardless of changes in the industry, or the position of the enterprises within that industry, or the effect of government regulation on that industry. The effects of government policies are also discussed at paragraphs 1.55 - 1.56 of the 1995 OECD Report and at paragraphs 70 and 319 of TR 94/14.

There have been circumstances where parent companies have entered into dealings with their subsidiaries that protect the parent's profit margin but have caused the subsidiaries to bear the full financial effects of changed industry or market conditions. For example, there are enterprises engaged in the importation of products from related parties for wholesaling and distribution through a third party reseller network. It is important for these wholesalers to maintain the financial viability of their reseller networks. Where this happens, the wholesaler is forced to bear the cost of developing and maintaining the reseller network, but the end result is that the wholesaling subsidiary is relegated to accepting a residual profit margin or a loss. In similar circumstances, an arm's length party might seek to renegotiate the terms of their arrangements. In these circumstances there are serious risks that the arm's length principle has not been complied with since the primary role of the subsidiary has become one of selling its parent's products and promoting its parent's brand name rather than making a profit in its own right. An independent enterprise operating wholly independently would seek to maximise the economic return from its functions, assets and risks. Overall, the wholesaling, marketing and distribution function is important to the MNE group and is often the source of its competitive advantage. This function should be rewarded on an arm's length basis.

The impact of general economic factors on individual enterprise performance may also need to be explained. For instance, the financial performance of enterprises in some industries is affected by the general level of economic activity to a greater extent than others. Some industries are less affected by general economic conditions and are strongly influenced by the nature of competition or technological change. The conclusions drawn from this information can influence the extent to which historical data on the group is relevant and may affect the selection of data.

The effect of general economic factors, such as economic cycles, may mean that in order to evaluate or establish compliance with the arm's length principle it will be necessary to examine data for a number of years. It may be useful to collect data, where it is available, that reflects an entire business cycle. The length of the business cycle will of course be affected by conditions pertaining in the industry, such as the pace of technological change, and broader economic conditions.

The analysis of the functions, assets and risks of the associated enterprises engaged in the international dealings can take account of specific factors affecting the industry. However, care needs to be taken to identify and compensate for those decisions which artificially transfer risks between the related parties.

The following example shows how an obsolescence risk can be transferred from the parent to the subsidiary. The market for the product is characterised by significant changes in product technology which results in the build up of excess stock of products embodying outdated technology in the inventories of the parent enterprise. The product has become less competitive in the retail market. Associated distributors of these products are directed by the parent to take part of this excess inventory at the usual inter-company price and on the same payment terms. Because of the need to offer discounts in order to sell the stock, the dealings erode the subsidiary's profitability. In an arm's length situation, the distributor may choose not to purchase the product, or may attempt to negotiate changes in the terms of the purchases including volumes, price, rebates, etc., in order to protect its own profitability.

Where comparability is difficult to assess or can only be approximated, it may be important to consider wider issues surrounding the dealings. This may include examining the circumstances surrounding the decision to enter into the dealings or, in some cases, how the property was dealt with in subsequent dealings. For example, the right to exploit property protected by copyright may be assigned to a related foreign enterprise. If there were subsequent reassignments to other associated enterprises in third countries, (perhaps in a treaty shopping arrangement) prior to ultimate licensing to a third party, it may be relevant to examine these subsequent dealings in the course of establishing an arm's length consideration that has regard to the value of the intangible. This may include an examination of the consideration that each received, the functions they performed, the risks they undertook and the assets they employed.

An evaluation of the strategies of the taxpayer will also generally be necessary. The marketing and pricing strategies, the existence of relevant policies such as the provision of cross subsidies to parts of the business as well as any broader corporate objectives may need to be examined in order to understand the business context in which the enterprise operates. Information on the business strategies can assist in establishing the selection of methodologies and may be very important when addressing questions associated with comparability.

For example, the information on business strategies can demonstrate whether the enterprise is engaged in market penetration activities. It can also assist in establishing the relative importance of marketing and distribution strategies compared with research and development activities, new product technology or specialised services.

Non-arm's length dealings may arise from a specific policy decision or a series of decisions of the board or senior management. For example, a subsidiary may have undertaken market development activities at its own expense and risk, and enhanced the value of a brand name owned by an associate which had no value prior to the market development activities of the enterprise. Senior management of the subsidiary may then agree to the payment of a royalty or management fee to a related foreign enterprise.

The payment of the royalty or management fee may significantly erode the profitability of the subsidiary. In evaluating whether the consideration conforms with the arm's length principle, it would be relevant to examine the decision making process of senior management or the board in arriving at the decision to agree to pay these fees and also incur market development expenses. It may be necessary to consider evidence as to whether the parties considered options realistically available to the enterprise. For example, in some situations it would be reasonable to conclude that an arm's length party would want its market development expenditure taken into account in the valuation of any royalty or by way of a reduced price for trading stock purchased from the owner of the brand name, and the consideration should be set accordingly.

Information on financial performance may be particularly important at a later stage if the methodology requires comparisons of the enterprise's performance over the relevant years or compared with other enterprises. The key ratios and statistics may vary depending upon the nature of the business being conducted. Usually, an application of methods (other than CUP) will require a comparison of the level of enterprise profit arising from dealings between associated enterprises with that achieved in its arm's length dealings or with the level of profit achieved by an uncontrolled enterprise.

This comparison is usually made in the form of some type of suitable accounting or statistical ratio analysis which will provide a basis to make the comparison. Such ratio analysis may include: (a) ratio of gross profit to operating expenses; (b) ratio of operating profit to sales; and (c) ratio of gross income / revenue to operating expenses.

What ratios, either from those mentioned above or others, are most appropriate needs to be established on the facts available. The use of ratios is discussed in relation to the profit comparison method at paragraphs 477 to 482 above.

Trends would include general factors affecting the performance of an enterprise on a macro level, such as economic conditions as well as any significant features of the particular market or market segment within which an enterprise operates. Relevant trends at the enterprise level may include trends in gearing, dividend rate, non-performing assets and stock levels, as well as in other key financial ratios.

When considering trends as part of the ATO's overall approach, it would seem that those elements or factors which have a quantifiable impact on an enterprise's profit performance over time, or could reasonably have had an impact on pricing policy at the relevant time, should be taken into account. Projected trends and potential profit outcomes may be crucial in situations such as APA's and in those circumstances where taxpayers set the consideration on their dealings by reference to a profit split.

A number of commentators have found difficulty in identifying the type of information that may be relevant to understanding an enterprise, its structure and business. A useful guide to the issues that might need to be taken into account is included in the Appendix to Statement of Auditing Practice AUP 34, ' Knowledge of the Client's Business ', which was prepared by the Auditing Standards Board of the Australian Accounting Research Foundation.

An extract from the Statement of Auditing Practice AUP 34 is reproduced as Appendix 1 to this Ruling (see paragraph 593 below).

Statement of Auditing Practice AUP 34 was not created for the purposes of resolving practical problems surrounding arm's length methods. Its purpose is to illustrate the knowledge that an auditor needs to conduct an effective audit of financial reports. However, it also provides a useful guide to any person seeking to understand the business of an enterprise and the factors that determine its competitive advantage.

In order to now select the most appropriate methodology or methodologies to use, the taxpayer may need to arrange the information that has been collected on its cross-border dealings with associated enterprises into an analysis of: (a) the functions undertaken by each of the associated enterprises (including their nature and frequency); (b) the risks each of the parties assumed; and (c) the assets (both tangible and intangible) used by each of the parties and the nature and extent of that use.

This is sometimes referred to as a functional analysis. Some form of functional analysis will be necessary regardless of the methodology that has been selected.

The functional analysis can assist in identifying issues associated with comparability in the selection of a methodology. An example of how this preliminary analysis assists is shown on the following page.

At its broadest level, such an analysis would result in the identification of categories such as manufacturing, wholesaling, financial services, mining, etc. However, such a broad description will not generally be sufficient. For some enterprises that are engaged in a number of business activities the range of functional areas that need to be analysed may be quite numerous.

It is essential to establish accurately the nature of the dealings of the enterprise. Some of the categories used in the industrial classification systems may not accurately reflect the actual activities undertaken by each enterprise. For example, an enterprise may be classified in the wholesaling industry even though its main business activities are more in the nature of a trading company or trade facilitator.

There are often significant differences between the nature of dealings of enterprises within the same industry. For example, some enterprises in the wholesaling industry may take title to goods and assume full commercial risks. Other wholesalers may take title in back-to-back transactions or may not take title to goods. These differences will need to be taken into account when comparable data based on these industrial classifications is used.

For each of the main business activities of the enterprise, a detailed listing of the functions, assets and risks should be compiled. Such lists could be compiled for individual transactions, product or service lines, or for the enterprise as a whole. The analysis of function, assets and risks would be useful in: (a) determining the availability of comparables in relation to prices or functions; (b) assessing the degree of comparability with the functions, assets and risks in respect of the taxpayer's uncontrolled transactions or with those undertaken by other enterprises being considered as possible comparables; (c) assessing the relative weighting of the functions, assets and risks of each of the associated enterprises that are a party to the cross-border dealings in cases where an apportionment methodology, such as a profit split, is needed.

The compilation of such lists of functions, assets and risks, however detailed, does not in itself indicate which of the functions are the most significant, or economically the most important to the value added created by the business activities of the enterprise. A critical part of the analysis is to ascertain which are the most economically important functions, assets and risks and how these might be reflected in terms of an arm's length price, margin or profit on the dealings.

A more colloquial approach would be to consider some direct questions of the following type: (a) what do you do? (b) what assistance do you receive and what transfers have been made to and from the enterprise? (c) how does this enterprise add value? (d) why is this enterprise successful? (e) are there any unique factors in your success? (f) what examples are there of cases where the strategies/success factors did/did not work?

It is generally not necessary to value each of the functions, assets and risks. The purpose of the examination is to understand the qualitative nature of the functions, assets and risks so that a comparison can be made with other enterprises that have similar functions, assets and risks. The allocation of actual income to assets may be far too difficult a task, and is likely to lead to undue complexities in the analysis. The value of some assets are not easily measured, notably intangible assets. In fact, if taken to extreme levels it could lead to an examination that becomes absurd. Many factors will simply be assessed as part of the business risks and comparisons made at that level. It needs to be remembered that the various methodologies work on the basis that the analysis is capable of being able to produce a quantifiable result. This can mean that it is necessary for some factors that cannot be quantified to be taken into account in some indirect way. For many cases, particularly where the international related-party dealings are relatively straightforward, it is desirable to avoid overly complex analyses.

The functional analysis will be developed and extended in Step 3 depending upon the methodology that is selected in Step 2. At that stage it is possible that the type of analysis that has been undertaken on some or all of the dealings between associated enterprises may be undertaken using the available information on uncontrolled dealings. This will allow comparisons to be drawn between the two different types of dealings.

Step 2 mostly requires the compilation and assessment of data to be used and in the selection of a methodology for establishing acceptable standards of comparability or in determining the appropriate allocation of profits or income between the associated enterprises.

At this stage it is also important to ascertain the extent of reliable uncontrolled data that is available to help determine an arm's length outcome. The nature of the available data, especially the amount of detail on prices and functions and its reliability, will be very important factors in the selection and future application of a methodology. The types of data that might be sought will vary from case to case. For example, it could mean investigating the availability of open market prices and terms for comparable transactions in particular types of commodities. However, in other cases it may mean enquiring into the availability of information about the gross or net profit margins or business risks of enterprises that may be comparable.

In Step 2 the reliability of the available material should be carefully evaluated to ensure it is capable of being used in practical manner. In Step 3 it will actually be used to extend and refine the functional analysis by making comparisons between the relevant dealings of the associated enterprises and those of the uncontrolled enterprises.

The information initially collected may include data that enable the calculation of gross profit ratios or ratio of gross income to operating expenses, or it may include market share data, etc. This might be sufficient to identify whether the taxpayer's case raises transfer pricing issues, and perhaps whether issues arise in relation to items above or below the gross profit line. However, in order to evaluate the proper return for the economically significant functions, assets and risks, other data may need to be collected to identify problem areas more clearly.

The data to be collected may be important in evaluating the role of intangible assets used in the business. For example, the distribution/reseller network utilised by an importer may be essential in creating and realising the value from a particular product. If the particular methodology being considered necessitates an evaluation of the proper return on such intangible assets, then the nature of the data to be collected will be different from that needed to do a comparison of financial ratios.

The collection of further data should be done on a selective basis to identify the important value adding activities of the enterprise and to get a good sense of their relevant importance to the taxpayer's income earning activities. This will facilitate comparisons with arm's length parties or to evaluate the appropriateness of profit splits. It is also important to assess the reliability of the data and to ensure that it is sufficient to allow the practical application of the methodology selected.

Based on the type of information and documentation available, the taxpayer and/or the ATO need to decide which of the arm's length methodologies is the most appropriate for its business. Further, it will be important to be able to show how the application of the selected methodology or methodologies is able to quantify an arm's length outcome for the taxpayer's cross-border dealings with associated enterprises.

The information obtained from Step 1 can assist with the: (a) determination of comparability when transactional methodologies are appropriate; and/or (b) determination of comparability between enterprises when methodologies using profit comparison are appropriate; and/or (c) allocation of the consideration between the enterprises when a profit split or other apportionment methodology is appropriate.

In some cases the taxpayer or the ATO may select more than one appropriate methodology in order to either encompass the full range of its international dealings with associated enterprises or to obtain greater certainty that the primary method selected in fact produces reliable arm's length results.

In some circumstances it may be possible to apply a particular method to only part of the relevant dealings of a taxpayer. In this situation care will be needed to ensure the methodology is being legitimately applied. Sometimes an enterprise can have a mix of methods successfully applied to its dealings. However, 'cherrypicking', the selective application of a methodology to a limited range of dealings, can produce commercially absurd outcomes. In some situation it may be necessary ultimately to select a different method (or mix of methods) in preference because it has a wider application and produces a closer approximation of an arm's length result for all of the international dealings with associated enterprises. For example, it may be possible to determine an arm's length consideration for some of the taxpayer's dealings with associated enterprises by using an analysis of uncontrolled dealings of third parties. However, an application of this methodology to only some of those products or services may produce a commercially unrealistic outcome due to the existence of other dealings with associated enterprises for which an arm's length consideration cannot easily be determined. This may be due to the uniqueness of the products or services or the presence of intangibles. Where this occurs it could raise a serious question as to whether the initial analysis in fact produced an accurate comparable. To gain assurance that the arm's length principle is consistently applied across all the relevant international dealings between associated enterprises it may be more appropriate in these circumstances to apply a different methodology (or mix of methods) to the relevant part or the whole entity rather than adjusting the price or profit margin on only some of the enterprise's products or services.

By the completion of Step 2, the taxpayer will have documented and examined the international dealings between the associated enterprises, collected relevant internal and external data, and undertaken a preliminary functional analysis of the enterprise or enterprises engaged in the international dealings, or of the comparables. Based on this information the taxpayer would have selected a methodology or methodologies most appropriate to the facts and circumstances of the dealings.

In Step 3 the taxpayer or the ATO will need to apply the chosen methodology using the information that has now been identified. This should be done in a way that tests the appropriateness of the selected methodology or methodologies and confirms its suitability. If this cannot be confirmed the taxpayer or the ATO will need to consider other methodologies, either alone or as a support for the method(s) initially selected.

For example, if the taxpayer have selected a methodology that involves the use of data derived using comparable uncontrolled transactions, then an analysis might need to be prepared to demonstrate how these transactions are comparable to the transactions the taxpayer has with associated enterprises. This might involve a discussion of the comparability in terms of the functions, assets, risks of the entities undertaking the comparable uncontrolled transactions, the comparability of the contractual arrangements, and the similarities of the market and economic conditions. If the taxpayer has selected a profit split methodology, then the functional analysis should be used to demonstrate the relative economic and other contributions of the parties to the dealings.

The chart on the following page shows how the functional analysis may be used differently depending upon the methodology that is used.

The preliminary functional analysis that was prepared to select a methodology can now be extended. If a comparability methodology involving external benchmarking with independent enterprises is being used, the functional analysis assists in determining the comparability of the dealings or the enterprise with uncontrolled dealings undertaken by the independent parties. It is not necessary to value the functions, assets and risks of each of the enterprises since the main intention is to establish the degree of comparability. However, it is essential to ensure that where there are differences in the significance of the functions, assets and risks to each of the businesses that these differences are taken into account.

For example, two enterprises may appear to be comparable when a listing of the functions, assets and risks is compiled. Both enterprises may undertake product innovation activities and marketing and distribution activities. For one of the enterprises product innovation activities may be more critical for maintaining the competitive position of the business, while for the other company marketing and distribution activities may be more important. While the broad functions of the enterprises may appear to be similar, in fact the economic importance of some of the functions may differ substantially. This needs to be taken into account when an analysis of comparability is undertaken.

If an apportionment methodology, such as a profit split, is being used, it is also not essential to value each of the functions, assets and risks. The intention of an apportionment methodology is to establish the relative importance of the functions, assets and risks of the parties to the international dealings so that an arm's length apportionment of the consideration in the dealings (such as a profit split) can be undertaken.

Some of the functions, assets and risks may be shared between associated enterprises engaged in the international dealings. For example, research and development may be undertaken by both parties in a cost contribution basis. It will be necessary to recognise that the return to the research and development activity is not the sole property of one of the parties, and will need to be split in accordance with the relative contributions of each of the parties.

The functional analysis can be performed with varying levels of detail and can serve a variety of purposes. The analysis may be applied on a product or divisional basis for individual transactions, or it could be applied up to a corporate group basis. The scope of the analysis will be determined by the nature, value and complexity of the matters covered by international dealings and the nature of taxpayer's business activities, including the strategies which the enterprise pursues and the features of its products or services. It is difficult to set out guidelines here that are suitable for every possible type of enterprise. However, as has been noted, the material that is produced can be of assistance to an enterprise in developing documentation to support its views.

It is often important in evaluating the relative significance of the functions to analyse the staffing of the enterprise. Experienced and highly trained staff may be an important intangible asset of the company, and these staff may be undertaking essential functions which are generating considerable value added. Data on staff levels, experience, educational qualifications, remuneration, performance evaluation and duties may be collected, and information obtained from key operational staff will often be required. The type of staff and their duties and skills may also be a reliable guide to the nature and type of the activities that the taxpayer undertakes.

Considerable data may have been collected by the taxpayer or its advisers from both internal and external sources. This data may extend over a number of years. The data would need to be organised in such a way as to demonstrate, for example, how economic and business cycles have impacted on the business in the past. However, some of this data may involve projections of profit splits or outcomes from dealings. This data would need to be gathered and organised in such a way as to demonstrate how the projections are consistent with what an arm's length party might obtain.

Some of the data may need to be refined or adjusted to improve comparability. This may be particularly important in those cases where the enterprise is engaged in strategies (special conditions) which affect the arm's length consideration. Some of the data may be incomplete, and some of the data may be irrelevant to determining an arm's length outcome. It is possible that further more detailed data may need to be collected at this stage to supplement the functional analysis. In practice, there may be gaps in the available data which necessitate further enquiries of the type already conducted. It is possible that in some circumstances the practical implementation of the methodology may prove to be highly problematical. In this situation the selection of the particular methodology may need to be reconsidered and its reliability considered against the reliability of other methods that could be applied on the available data.

The taxpayer may also find it useful to attempt to check the results from one methodological approach against the results of another methodology if sufficient reliable data is available. For example, depending upon the reliability achieved, the outcome from a cost plus methodology may need to be checked against the results from a profit split or a method which uses profit comparisons.

Clearly the extent of analysis at this stage will vary depending on a range of factors, such as data availability, the extent and level of the international dealings and their importance to the enterprise's business. The methodology selected must be capable of practical application and must produce a result that is a reasonable approximation of what would result if the dealings were undertaken on an arm's length basis.

At this point, the data may suggest that there is a range of possible arm's length outcomes that might be benchmarks for the dealings. If this is the case, the taxpayer or its advisers will need to evaluate and refine the data to create an appropriate range that it believes reflects the range of outcomes that might reasonably be expected to occur if the dealings were undertaken on an arm's length basis.

The taxpayer will be required to demonstrate how the methodology used actually produces an arm's length outcome for the dealings between the associates. That is, the enterprise will be required to show how the data has been used in the application of the chosen methodology to determine the arm's length result. When this step is completed, there should be sufficient documentation and reasoning to explain how the outcome is consistent with what arm's length parties would have achieved.

In this step, the taxpayer will need to apply the data to the actual dealings to demonstrate how the outcomes conforms to the arm's length principle. This would be the way the ATO would check the taxpayer's dealings. How well this application can be done of course is highly dependent on the availability of relevant data, its reliability and the available documentation, either internal or external. If data availability is a problem, arguments will need to be developed to support the results achieved from the analysis, and a more indirect measurement methodology possibly may be needed.

The process to date can deliver to the taxpayer an objective, documented and considered review of the available material and possible choices for arriving at an arm's length outcome. The determination of the arm's length consideration can now be made. This may be reasonably straightforward where the methodology has produced a high level of comparability - though judgment is still required -or it may require a greater degree of judgment where there is a range of results or a choice of answers is needed. The earlier discussion on the nature of the arm's length principle highlighted some of the practical problems in its application which may now be evident. Nevertheless, the evidencing of the imprecision involved in the application of the arm's length principle should assist in reducing the risk of a taxpayer's adopting a non-arm's length consideration in its international dealings.

The ongoing reliability of a taxpayer's transfer pricing method(s) will depend on whether the necessary systems and procedures are put in place to collect the relevant data and ensure the proper analysis for tax purposes.

Where a methodology is being used on a continuing basis, the choice of methodology that has been made and the data that are used to establish the consideration need to remain valid. The methodology and the data can become outdated and unreliable if there are material changes in the business or if the data available to apply the methodology change. For example, data on comparable dealings may cease to be available or better sources of data may be found.

The process of selection and application of a methodology is a dynamic exercise. How long a process or a method arising from the process remains appropriate before it ceases to give an accurate outcome will depend upon the nature and extent of the changes that occur to the factors that were initially taken into account.

To ensure that the processes, methods and data used continue to be appropriate, the taxpayer should incorporate reviews of these matters into its internal procedures. Factors which may influence such a review are discussed in more detail in TR 95/D23.

Even where a methodology is first selected and applied, the results should be checked to ensure that they are commercially realistic having regard to the taxpayer's facts and circumstances. Where there is some doubt about the reliability of the outcome the earlier steps in the four step process should be revisited and the results checked against other approaches/indicators that may be available. It will be necessary to consider the relative reliability of another methodology using the known information. If this is not possible or practicable, the taxpayer should consider an APA with the ATO.