Income tax: international transfer pricing - cost contribution arrangements

This Ruling explains how the arm's length principle applies to international dealings in relation to cost contribution arrangements ('CCAs') for purposes of section 136AD of Division 13 of Part III of the Income Tax Assessment Act 1936 ('ITAA 1936') and the Associated Enterprises Article in Australia's double tax agreements.

A CCA is a contractual arrangement between business enterprises to share the costs and risks of developing, producing or obtaining assets, services or rights, and to define the interests of each participant in those assets, services or rights.

This description of a CCA is that used in Chapter VIII of the OECD's Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations ('the 1995 OECD Report'). Chapter VIII provides a broad framework of guidelines for the application of the arm's length principle to CCAs. This Ruling accepts and builds upon the views in Chapter VIII in addressing how we consider they apply in the context of the relevant provisions of the Australian income tax law.

This Ruling focuses on the use of CCAs by multinational enterprises ('MNEs'), which is most commonly in respect of research and development ('R&D') activities, mining exploration and development ventures and group management services.

Subject to the specific guidance in this Ruling, the general principles for using and documenting arm's length transfer pricing methodologies, as set out in Taxation Rulings TR 97/20 and TR 98/11, apply to CCAs. [F1]

This Ruling deals with arrangements between separate legal entities, not dealings between parts of a single entity. However, it is relevant to the application of Australia's permanent establishment attribution rules [F2] where a CCA is considered an appropriate separate enterprise analogy for applying the arm's length principle in attributing income and expenses to a permanent establishment. [F3]

This Ruling deals only with transfer pricing issues related to the application of the arm's length principle to CCAs. It does not address domestic tax issues related to the application of provisions of the ITAA other than those dealing with transfer pricing.

This Ruling does not specifically discuss the issue of how share options provided to a CCA participant's employees who perform CCA activity might impact the value of the participant's contribution to the arrangement. The OECD has recently commenced considering transfer pricing issues related to employee share option plans. We are currently considering our position on these issues, which arise in a significantly broader context than just CCAs. [F4]

The Ruling and Explanation part of this Ruling is presented in four parts: • A - Concept of a CCA • B - Applying the arm's length principle • C - Consequences if a CCA is not arm's length • D - Documenting CCAs.

It is proposed that when the final Ruling is issued it will apply both before and after its date of issue. However the final Ruling will not apply to taxpayers to the extent that it conflicts with the terms of a settlement of a dispute agreed to before the date of issue of the Ruling. [F5]

The concept of a CCA is broad enough to cover any arrangement under which the parties agree to share the costs and risks of developing, producing or obtaining assets, rights or services in return for a share of the expected benefits from what is developed, produced or obtained.

The concept of a CCA addressed in Chapter VIII of the 1995 OECD Report contemplates an arrangement that has several key characteristics: (a) it is a contractual arrangement rather than necessarily a distinct juridical entity or permanent establishment of all the participants; [F6] (b) each participant in the arrangement, in return for agreeing to make a specified contribution towards the activity performed under the arrangement ('the CCA activity'), acquires a specified interest in the results of that activity; (c) a participant independently exploits its interest in the results of the CCA activity; [F7] and (d) a participant's rights to exploit its interest are free of obligation to pay royalties or other consideration additional to its contribution.

A CCA is thus best described as a form of joint venture arrangement. A CCA with the above characteristics is a 'development-only' joint venture rather than an 'income sharing' joint venture. The arrangement is limited to sharing the costs and risks of jointly developing, producing or obtaining assets, rights or services; it does not extend to joint exploitation of the results of this activity and sharing of any resulting profits. Each participant separately exploits its specified interest and is entitled to all of the profits from this exploitation.

Most commonly, a CCA does not extend to joint exploitation of the results of the CCA activity. [F8] However, the concept of a CCA is flexible enough to include an arrangement under which there is both joint development activity and joint exploitation of the results of that activity. In practice, participants to a CCA commonly obtain and exploit results during continuation of development activity. Depending upon the terms of the CCA agreement, exploitation may be a part of the joint activity performed under the arrangement, so that the resulting income and profits are shared between the participants.

Two major types of CCA are most commonly encountered in practice. Each is fundamentally different as regards its commercial rationale and characteristics, particularly in respect of the relationship between cost, risk and benefit. These differences have significant implications for the application of the arm's length principle.

The discussion in this Ruling relates to the first type of CCA, unless otherwise stated. Our transfer pricing guidelines on intra-group services, Taxation Ruling TR 1999/1, apply to pure service arrangements. While that Ruling does not specifically address CCAs, [F9] it states that if a service arrangement does not result in any property being produced, developed or acquired, the principles in that Ruling apply for dealing with intra-group services, whether the arrangement is described as a CCA or not. CCAs for pure service arrangements are discussed at paragraphs 105-113 and 158.

Not all CCAs are simply one or other of the above types. A particular CCA may be a variation or hybrid of one or both of these types. For instance, a CCA may relate to both development and ongoing technical support of a new product or process, so that the CCA activity includes both R&D and technical services. A CCA may relate to multiple activities. For instance, a single CCA might cover more than one aspect of a MNE's business, such as R&D, marketing, centralised product or raw materials purchasing, management, administrative and technical services. [F10]

In general terms, determining whether the conditions of a CCA are consistent with the arm's length principle requires a consideration of whether the arrangement accords with what independent parties dealing at arm's length might be expected to have entered into in comparable circumstances.

In addressing this, we will have regard to the following matters, to the extent that each is relevant in a particular case.

The actions that we may take where we consider that the conditions of a CCA are not consistent with the arm's length principle are discussed at paragraphs 199-213.

Our expectations in relation to documenting the application of the arm's length principle to CCAs are set out at paragraphs 214-217.

TR 97/20 states several key notions that underlie our approach to applying the arm's length principle: (a) an arm's length outcome is one that makes business sense in the circumstances of the particular taxpayer; [F11] (b) an independent party dealing at arm's length would seek to protect its own economic interest; [F12] (c) an independent party dealing at arm's length would compare the options realistically available and seek to maximise the overall value derived from its economic resources; [F13] and (d) one option might be not to enter into a transaction because it does not make commercial sense for the particular taxpayer. [F14]

Thus, a taxpayer's participation in a CCA should make business sense in its particular circumstances. The terms of a CCA should be consistent with what would have been agreed by the taxpayer as a party acting in its own economic interests, and reflect outcomes that make business sense in its particular circumstances.

Ordinarily, this requirement will be satisfied provided the terms of a CCA relating to the sharing of costs and expected benefits, and to any necessary buy-in, buy-out and balancing payments, satisfy the arm's length principle in accordance with this Ruling. However, particular circumstances may give rise to a threshold issue as to whether it makes business sense for a taxpayer to enter into a CCA, notwithstanding that the CCA otherwise reflects arm's length terms for what is supplied and acquired under the arrangement. In some circumstances, even if contributions were valued and shared under a CCA in accordance with the views in this Ruling, an independent party might not enter into the arrangement because it is not in its economic interest or does not make business sense for it to do so given other available options.

The arm's length principle calls for a consideration of the commercial imperative for a taxpayer to enter into a CCA, given its particular circumstances. In deciding whether to enter into a CCA, a taxpayer may have a choice between taking all of the risk for all of the potential profit, or sharing the risk and sharing the potential profit. The commercial need for a taxpayer to seek others to jointly participate in a venture through a CCA may be readily apparent in some circumstances. For instance, where the taxpayer alone would not be commercially or financially able to either undertake the venture or exploit the expected results. This might be the case where the cost or risk of failure of the venture is high, or where the taxpayer lacks the necessary assets, skills or capital resources. A CCA may be the most appropriate and advantageous commercial strategy for the taxpayer to obtain such inputs to the venture.

A taxpayer's decision as to whether to enter into a CCA may involve a choice between investment options. For instance, where a taxpayer is deciding whether to enter into a CCA to develop intangible property that will be used in the taxpayer's business, there may be a choice between investing in the CCA and acquiring an interest in the property, or simply licensing the right to use the property. An independent party would be expected to decide between these investment options based upon what would best promote its economic interest.

The arm's length principle requires that a taxpayer's decision as to whether to enter into a CCA be made having regard to its own economic interest, and that the decision makes commercial sense in the context of the taxpayer's business circumstances. In the absence of circumstances that explain the commerciality of a CCA as a business strategy in the taxpayer's circumstances, an independent party, having performed a cost benefit analysis of the options available, might be expected not to enter into a CCA.

AusCo, [F15] a member of a MNE group, owns existing technology for a highly profitable product. AusCo manufactures and sells the product itself, and has licensing arrangements with other group members to use the technology to manufacture the product for sale in their local markets. AusCo performed the R&D that created the original technology, and is about to commence intensive R&D to enhance the technology for the next generation of the product. The risk of this R&D being unsuccessful is considered relatively low. AusCo has the necessary resources, expertise and financial capacity to perform the R&D and exploit any new technology produced. The MNE board decides that the new R&D will be performed under a CCA, whose participants will be AusCo and a newly established non-resident group company, ForCo. AusCo's contributions will be in the form of existing technology and ongoing R&D services, while ForCo's contributions will be cash. In return, ForCo will have the right to license the new technology to group members other than AusCo to manufacture the product for sale in their local markets.

In these circumstances, the commercial need for AusCo to enter into the CCA is not readily apparent. An issue arises as to why it would make business sense for AusCo, acting in its own economic interests, to enter into the CCA rather than choosing to develop and exploit the technology itself. This is not likely to be satisfactorily explained simply by demonstrating that the participants' contributions are appropriately valued and shared relative to the sharing of their expected benefits from the arrangement. Even if ForCo has a high level of expected benefits that is appropriately reflected in a high share of costs, an issue arguably remains as to why an independent party in AusCo's position would agree to any sharing of the expected benefits with ForCo. If it were concluded that an independent party in AusCo's position might be expected not to enter into the CCA, we would apply the arm's length principle to determine AusCo's taxable profits on this basis. [F16]

AusCo, a member of a MNE group, currently licenses technology that it uses to manufacture and sell a certain product range in Australia. The foreign group member that owns the technology, ForCo, is about to commence intensive R&D to enhance the technology for the next generation of the product. The MNE board decides that the new R&D will be performed under a CCA, whose participants will be AusCo and ForCo. ForCo's contributions will be in the form of existing technology and ongoing R&D services, while AusCo's contributions will be cash. The risk of the R&D failing to produce commercially exploitable results is considered relatively high. The profitability of the product range in the Australian market is low due to heavy and increasing competition. The importance of the product range to AusCo's business is expected to steadily diminish. Based upon the R&D budget, the estimated value of the new technology, and AusCo's projected sales figures for the new product, it can be determined that AusCo could expect that financially it would be significantly better off if it either licensed the use of the new technology or had no involvement with the new product, rather than participate in the CCA.

In these circumstances, an issue arises as to why it would make business sense for AusCo, acting in its own economic interests, to enter into the CCA compared to other options realistically available to it. This is not likely to be satisfactorily explained by demonstrating that AusCo's share of costs under the CCA is appropriate relative to its share of the expected benefits. Even if AusCo's low level of expected benefits is appropriately reflected in a low share of costs, an issue remains as to why an independent party in AusCo's position would agree to any sharing of the costs and risks of the CCA, given that it has other more financially advantageous options. If it were concluded that an independent party in AusCo's position might be expected not to enter into the CCA, we would apply the arm's length principle to determine AusCo's taxable profits on this basis. [F17]

The importance of the commerciality of a CCA is also reflected in the requirement that its form (i.e. agreed terms) accord with its economic substance, as evidenced by the conduct of the parties and what parties dealing at arm's length would be expected to have agreed in similar circumstances. [F18] Where the terms purportedly agreed by the participants do not accord with the commercial reality of the arrangement, those terms may be disregarded. [F19]

AusCo and ForCo, two members of a MNE group, enter into a CCA to develop new technology. AusCo is to contribute existing technology and cash. ForCo is to contribute by performing R&D services. Each is given an interest in any results of the CCA activity. ForCo's interest is to be exploited by licensing to other group members. AusCo makes all major decisions regarding performance of the CCA activity, including its scope, what is and is not to be performed, whether particular research is to be pursued or abandoned, and the program's resourcing and budgeting. ForCo simply follows AusCo's instructions.

Having regard to the conduct of the parties and what independent parties would be expected to have agreed in similar circumstances, an issue arises as to whether the economic substance is of a contract R&D arrangement rather than a CCA. [F20] An independent party would ordinarily not be expected to agree to share in risk if it does not have a say in the extent to which that risk is assumed or control over managing the risk. Given the level of AusCo's control over the R&D activity, ForCo as an independent party might not be expected to have assumed any of the entrepreneurial risk of that activity (ie. the risk of its success or failure), as it has no control over that risk and is not in a position to manage that risk. If it were concluded that the commercial reality is of a contract R&D arrangement and not a CCA, AusCo would be treated as sole owner of the results of the R&D activity, with ForCo treated as performing the activity at the risk of and for the benefit of AusCo.

AusCo, ForCo1 and ForCo2 are members of a MNE group who enter into a CCA to jointly develop technology. The agreement provides that AusCo will contribute existing technology, ForCo1 will provide R&D services, and ForCo2 will make cash contributions. ForCo2 makes no actual payments when its contributions are due. Instead, its intercompany accounts with AusCo and ForCo1 are debited for the amounts due. No payments are made by ForCo2 during the course of the CCA to reduce the balances of these accounts. The expectation of the parties is that, if the CCA activity is successful, ForCo2 will be able to finance repayment of the intercompany loans and accrued interest out of profits from exploiting its interest in the results of the CCA activity.

In these circumstances it might be argued that the agreed terms are inconsistent with the commercial reality or economic substance of the arrangement. On this view, the contributions ForCo2 has agreed to make are in substance made indirectly by the other participants. Therefore, it might be argued that AusCo and ForCo1, as parties dealing at arm's length, would not be expected to agree to ForCo2's participation in the arrangement, as there is no commercial need for that participation in the circumstances. However, this argument should not lead to ForCo2's participation being disregarded, provided it has the financial capacity to be able to assume its share of the risk of loss of the venture. [F21] ForCo2's participation in the CCA does not lack commercial reality or economic substance simply because its contributions are funded by other participants. While AusCo and ForCo1 assume risk as lenders to ForCo2, this risk is separate from their risk as CCA participants, and it is this latter risk that can legitimately be shared with ForCo2 provided it has the financial capacity to assume that risk.

An important notion underlying the arm's length principle is that independent parties dealing at arm's length would be expected to undertake a process of real bargaining and agreement of terms prior to entering into a transaction or arrangement. Accordingly, the terms of a CCA should be agreed at the outset of the arrangement. There should be evidence that the parties' intention to share the costs and risks and expected benefits of the CCA activity existed prior to commencement of that activity.

As the benefits expected from a CCA are to be derived at some time after entry into the arrangement and sharing in its costs and risks, independent parties would not be expected to enter into such an arrangement without a written agreement. It will be difficult for a taxpayer to demonstrate the commerciality of the purported terms of a CCA, and we will be more likely to disregard those terms, where they are not evidenced by a written agreement executed prior to commencement of the CCA activity.

The views in TR 98/11, [F22] as to why a taxpayer would be well advised to contemporaneously document its efforts to comply with the arm's length principle, apply to CCAs. The documenting of CCAs is further discussed at paragraphs 214-217.

Consistent with the above, whether the terms of a CCA accord with what independent parties dealing at arm's length would be expected to agree should be judged by reference to circumstances known or reasonably foreseeable at the time of entry into the arrangement. [F23]

The performance of the CCA activity or the derivation of expected benefits from the arrangement commonly involves a considerable period of time after entry into the arrangement. This makes it difficult, at the time of initially agreeing the terms of the CCA, to anticipate later events and project future benefits. Given this, independent parties might be expected in such cases to agree that those terms provide for adjustment to the sharing of costs and expected benefits in certain specified events. [F24]

The arm's length principle requires that, to be a participant in a CCA, an entity must have a reasonable expectation that it will benefit by exploiting its interest in the results of the CCA activity. [F25] This gives rise to the following principles that are discussed below: (a) a participant must have an interest in the results of the CCA activity; and (b) a participant should have a reasonable expectation of benefit from exploiting its interest in the results of the CCA activity.

To be a participant in a CCA, an entity's expected benefit must come from exploiting an interest in the results of the CCA activity. An entity whose only expected benefit is from performing some part of the CCA activity is not regarded as a participant in the arrangement. [F26] Thus, where an entity operates as an R&D centre performing research activities, it is not a participant in a CCA if it has no interest in the results of those activities. There is a fundamental distinction between a CCA for R&D and a contract R&D arrangement. A participant in a CCA that performs R&D activity shares the risk of failure of the activity and has an interest in any results of the activity. In contract R&D, the party performing the R&D activity does so as a service; it does not bear any of the risk of failure of the activity, and does not have an interest in any results of the activity.

An essential characteristic of a CCA is that a participant, in consideration for contributions made, acquires an interest in the results of the CCA activity. The legal ownership of those results may, for various reasons, be vested in only one participant, some participants or all participants. However, the economic ownership vests in all participants, to the extent of their specified interests. For instance, R&D performed under a CCA may produce intellectual property that can be patented. While one participant may be registered as the holder of the patent, all of the CCA participants are economic owners of the property.

A participant has the right to exploit its interest in the results of the CCA activity by using those results without further consideration. [F27] For instance, where a CCA develops intangibles, the participants have the right to use their interests in those intangibles without payment of a royalty.

The interest that each participant has in the results of the CCA activity should be such as to have expected benefits that are capable of reliable measurement. It is implicit in the requirement that the sharing of contributions is consistent with the sharing of expected benefits that each participant's interest is sufficiently specified that its expected benefits from exploiting that interest can be reliably estimated when entering into the arrangement.

As previously discussed, [F28] each participant in a CCA that is a development-only venture has a specified economic interest that it separately exploits and is entitled to all of any resulting profits. It follows that the participants' specified interests must be able to co-exist, so that each participant's interest is able to be separately exploited by that participant to the exclusion of the other participants. Typically, participants to a CCA will operate in different geographic, product or other markets, and each will have a right to exploit the results of the CCA activity in the market in which it operates. However, the interests of the participants may be competing, in the sense that one may take market share from another.

Where participants to a CCA have dealings in the course of exploiting the results of the CCA activity, it is important to ensure that charges for those dealings take account of the participants' economic interests in the results of the CCA activity.

AusCo and ForCo are members of a MNE group who are participants in a CCA that has developed a new product, Product A. ForCo and AusCo both manufacture and distribute the group's products in their local markets. Under the CCA, AusCo has the right to manufacture and sell Product A in Australia and ForCo has a similar right in its home country. ForCo sometimes sells Product A to AusCo, for instance when customer demand temporarily exceeds AusCo's production capacity.

The transfer price that ForCo charges AusCo should not give a return to ForCo for use of Product A intangibles that is covered by the economic interest that AusCo has in those intangibles as a participant in the CCA. AusCo has effectively paid for such usage of the intangibles through its CCA contributions, and should not pay again through the price of the product purchased from ForCo.

It would be inconsistent with the arm's length principle for an entity to participate in a CCA by sharing in the costs and risks of the arrangement without a reasonable expectation of deriving benefit from the arrangement.

The concept of 'benefit' in a CCA context is similar to that used in relation to charging for intra-group services: [F29] In general terms, a benefit is something of economic or commercial value that an independent entity might reasonably expect to pay for, or to obtain consideration for supplying. For example, a benefit is an economic or commercial advantage that would assist the recipient's profitability or net worth by enhancing, assisting or improving its income production, profit making or the quality of its products. Alternatively, a benefit could result in a reduction of the recipient's expenses or otherwise facilitate its operations.

An independent party would have two types of benefits in mind when considering whether to become a participant in a CCA. These can be described as the benefits to which the CCA activity as a whole is directed ('activity benefits') and the benefits that come from undertaking the activity with other parties ('process benefits'). The CCA activity benefits include the services, assets or rights that the participants are seeking to develop, produce or acquire through the CCA activity. The CCA process benefits could include the sharing of risks, access to more or better resources, the acceleration of projects, economies of scale, or improved efficiency and productivity, perhaps from the combination of different individual strengths and spheres of expertise.

The concept of expected benefit should be viewed from a commercial perspective, and thus should not be narrowly confined simply to a measurable increase in future profits. For instance, the enhanced skills and expertise gained by a participant's staff working on R&D in respect of a particular product may be more valuable for that participant than future sales of the product. Participation in a CCA may give the right to ongoing access to know-how and technology that the participant could not itself produce or replicate except at a prohibitive cost. This in itself may be a valuable benefit, even if the participant has no immediate objective of using any particular know-how or technology on a particular project.

An entity that has no reasonable expectation of being able to exploit its interest in the results of the CCA activity has no expectation of benefit from the arrangement. Thus, for an entity to be a participant in a CCA it must reasonably expect to be commercially and financially capable of exploiting its interest in the results of the CCA activity, in the event of the success of the venture. The capability need not exist at the time of entering into the CCA, however the extent to which it does is relevant to considering the reasonableness of the expectation that it will exist by the anticipated time for exploitation of any results. It is not necessary that a participant have the capability to exploit its interest by directly or physically using the results of the CCA activity itself. [F30]

Each CCA participant should be financially able to assume its share of the risk of loss, were the venture ultimately unsuccessful. Given that a commercial rationale for a CCA is to share the risk of loss amongst the participants, [F31] there is no such rationale where a purported participant lacks the financial capacity to assume its share of such risk. As a matter of economic substance, there is no real transfer of risk to such a party. It is inconsistent with the arm's length principle for a CCA to have as a participant an entity that is only able to participate on the basis of the expected success of the venture being achieved.

All CCA participants should share in the costs and risks of all CCA activity from which they might expect to derive a benefit if the activity were successful. Independent parties entering into a CCA would not agree to a participant having the right to 'cherry pick' so as to share only in the costs of successful activity.

Whether it makes business sense and is in an entity's economic interests to participate in a CCA may be queried where the expected fruits of the CCA have either no connection or insufficient connection with the entity's existing business at the time of entry into the CCA. For instance, where a CCA is for development of technology related to Products A, B and C, it may be appropriate for an entity whose business relates only to Product C not to participate in the arrangement if the CCA activity in large part relates to Products A and B. If the entity does participate, its share of costs must reflect its relatively limited share of the overall expected benefits of the arrangement.

It is common for a MNE to have a permanently on-going R&D program related to its business overall or that of specific business segments. The program is intended to continuously work on developing new or enhanced technologies or products. Within the program, a series of discreet and more specifically focused projects or parts of the program typically exist. Where the program is conducted as a CCA, it is common for group members who participate to share the costs, risks and expected benefits of the overall program.

Participation on this basis may be consistent with the arm's length principle, even though not all individual projects may be expected to benefit all of the participants. The arm's length principle does not require that a participant expect to benefit from each and every CCA activity. Ordinarily, it does not require that projects within the overall program be separately evaluated, so that the appropriateness or commerciality of the participants and their contributions are assessed on the basis of their expected benefits from each and every project. It is acceptable to evaluate such things by looking at the program overall, provided any variation in relative levels of benefit for participants from individual projects is reflected in the outcome of the method used to estimate their relative expected benefits from the overall program.

A CCA covering a broad range of activities may be commercially impractical, and so less likely to be entered into by independent parties dealing at arm's length, given the likely difficulty of consistently and reliably measuring the various and differing expected benefits and appropriately relating these to contributions. It might be possible to address this by using more than one allocation key to estimate the relative expected benefits. [F32] Each activity might have its own pool of costs in which only those participants that expect to benefit from that activity share. Or, in some circumstances, multiple CCAs may be more commercially realistic than a single CCA for multiple activities.

AusCo is a member of a MNE that operates an oil products business. The MNE has a R&D program, conducted as a CCA, relating to all of the group's types of businesses, including fuels, LPG, lubricants, bitumen, aviation and marine products. Under the CCA the costs and risks of R&D related to each type of business are separately pooled and shared amongst such of the participants as operate that business and thus expect to benefit from the results of that R&D. Costs and risks are shared using an allocation key to measure the expected benefits of the participants in respect of the particular type of business to which the R&D relates. In return for contributing on this basis as a participant, AusCo obtains a right to use the results of the R&D in those businesses in which it is involved.

In these circumstances, the general framework prescribed under the CCA for the sharing of costs, risks and expected benefits of the R&D program might be expected to accord with the arm's length principle. The terms of the CCA specifically seek to ensure that a participant only shares in the costs and risks of the overall program to an extent that has regard to its expected benefits from the program.

The arm's length principle does not mandate that all participants to a CCA must expect to benefit by exploiting their interests in the results of the CCA activity in the same way. For instance, all participants to a CCA to develop manufacturing technology need not intend to exploit their interests by using the technology to manufacture. There is no requirement implicit in the arm's length principle that a CCA participant must use the results of the CCA activity in a particular way in its business. It is sufficient if there is economic use through which the participant receives the economic benefit of the results of the CCA activity. Thus, an entity can participate in a CCA even though it cannot benefit directly from exploitation of the results of the CCA activity. A participant can benefit through transferring or licensing the use of its interest in the results of the CCA activity to others.

For instance, an entity that is purely a product distributor might ordinarily not have sufficient expectation of benefit to be a participant in a CCA for development of manufacturing technology, given that it cannot physically use the technology itself. However, in some circumstances there may be a commercial explanation for such participation. For example, the distributor may expect to use the technology by contracting with a manufacturer to make the relevant product for the distributor.

It is not unusual for MNEs to have a company or companies in the group whose main or only purpose is to hold intangibles. These companies do not directly use the intangibles but license their use to others. They exist for various non-tax related reasons, including centralising management and control, limiting legal liability, and avoiding excessive regulatory requirements. They may also be used to obtain tax advantages. For instance, through reducing the tax borne on intangibles profits by locating the company holding intangibles in a jurisdiction with relatively low tax rates or other favourable taxation characteristics.

Where a resident taxpayer has an interest in a company holding intangibles, Australia's Controlled Foreign Companies rules in Part X of the ITAA 1936 may apply to attribute income of that company to the taxpayer. Where a taxpayer participates in a CCA with the purpose of obtaining a tax benefit, the general anti-avoidance provisions in Part IVA of the ITAA 1936 may apply.

The commercial rationale for using companies to hold intangibles may explain the decision to have intangibles owned by a different entity to the user of the intangibles, and hence to have the intended owner (i.e. the company holding intangibles) rather than the intended users (e.g. manufacturing entities) participate in a CCA for development of the intangibles. However, while this may explain the commerciality of a company's participation in a CCA from a group perspective, the arm's length principle requires a consideration of commerciality from the taxpayer's perspective. Thus, as the example at paragraphs 28-29 above illustrates, it is necessary to consider whether the taxpayer, as an independent party acting in its own commercial interests, would agree to share costs, risks and expected benefits with another party or parties in such circumstances.

In appropriate circumstances, one group member may participate in a CCA on behalf of other group members. The participating member's interest in the results of the CCA activity must then be transferred to or made available for use by the other members on an arm's length basis.

In appropriate circumstances, a CCA whose participants are to exploit their interests in differing ways may be consistent with what independent parties might be expected to agree. Important qualifications would include: (a) each party's participation must be in its own economic interests and make business sense; [F33] (b) the parties' differing exploitation rights must be able to legally and commercially co-exist, such that each is able to separately exploit its interest to the exclusion of the other participants; [F34] and (c) the relative values of the parties' differing expected benefits can be reliably measured for determining their relative contributions. [F35]

The arm's length principle does not necessarily mandate that all group members who might be expected to benefit from a CCA activity be participants in the arrangement. For instance, if a CCA is established to develop manufacturing technology, all group members who are manufacturers and expect to use the technology need not be participants in the arrangement. As would be the case if there were no CCA, any group members who have a need to use the technology can enter into arm's length arrangements for such use, e.g. by licensing from those who own the technology.

However, in considering the commerciality of a CCA, we will look at all group members who might be expected to benefit from the arrangement. The possibility of significant leakages of benefits to such members who are non-participants, without appropriate arm's length compensation, will impact our view on the commerciality of the arrangement. An independent party would not enter into a CCA where non-participants are so able to share in the fruits of the arrangement.

In addition, the extent to which non-participants are expected to use the results of the CCA activity, and what arrangements will be needed to enable such use, are likely to impact the sharing of the expected benefits between the participants, and the reliability and commerciality of the methods used to estimate such benefits. Thus, if such use will be through licensing, the CCA must take account of the extent to which each participant is expected to benefit as licensor. This will commonly require projections of royalty income, and hence royalty rates and sales volumes of the non-participants. [F36]

Where a CCA is a pure service arrangement under which the expected benefits derive solely from performance of the CCA activity, [F37] all group members who are expected to benefit should participate in the arrangement. Participation may be either direct or by one group member participating on behalf of another. [F38] In such cases, contributions made through participation in the CCA are the only way of compensating the services that constitute the CCA activity.

It is characteristic of a CCA that there is no guarantee that expected benefits will actually be derived. However, the timeframe within which benefits could reasonably be expected to be derived must make business sense for the taxpayer. An independent party would not be expected to participate in a CCA unless it were satisfied that it could obtain an acceptable rate of return on its CCA investment within a timeframe that had regard to its financial and business circumstances.

If a participant has derived no significant actual benefit from the CCA activity over a considerable period, it is necessary to look at the reasons for this outcome, and also to consider what response, if any, might be expected under the terms of a CCA between independent parties. Long lead times are common in R&D projects, and therefore a lack of exploitable results over a considerable period may not be unusual. [F39] This feature affects all participants to a CCA, and may only require a response if it is unexpected, for which the terms of a CCA between independent parties might be expected to provide for adjustment of benefit projections. [F40] This may also be the appropriate response where only a particular participant derives no benefit over a longer than expected period. Alternatively, all of the relevant circumstances may indicate that the participant did not have a sufficient or reasonable expectation of benefit when entering into the CCA to be a participant.

An entity may have a reasonable expectation of benefit when it enters into the CCA, but may later lose that expectation through a change in circumstances. A CCA agreed between independent parties would be expected to provide for a party's withdrawal from participation in the arrangement in such an event.

As a general rule, for a CCA to satisfy the arm's length principle each participant's proportionate share of the overall contributions to the CCA should be consistent with the participant's proportionate share of the overall expected benefits. [F41]

This is simply a presumption as to what independent parties would require. [F42] It may therefore be possible in particular circumstances to demonstrate that real bargaining between independent parties would produce an outcome that does not fully accord with this presumption. In other words, it might be possible to demonstrate, by reference to commercial factors, that a CCA is arm's length notwithstanding that the sharing of costs is not fully consistent with, or solely based upon, the sharing of expected benefits. For instance, a taxpayer may obtain collateral benefits from a CCA that are not part of the expected benefits shared under the arrangement, but which compensate for a share of costs that otherwise exceeds its share of the expected benefits. For this reason, it is important to recognise that there can be a range of different results in terms of the sharing of costs and/or expected benefits, each of which may nevertheless be consistent with the arm's length principle. [F43] The essential thing is that a participant's expected benefits from the arrangement are consistent with what an independent enterprise would have agreed to receive, given the contributions it agrees to make, in comparable circumstances.

Given the general need for consistency in the sharing of costs and expected benefits, it is necessary to estimate the relative or comparative values of each participant's contributions and expected benefits.

The value assigned to each participant's contributions should accord with that which independent parties would have assigned in comparable circumstances. [F44] Thus, what is the most appropriate basis for valuing contributions must be determined case-by-case; no single basis, for example market value or historical cost, is necessarily appropriate in all cases. [F45]

The valuation of contributions in this context is part of determining whether a CCA is on arm's length terms, and in particular whether the sharing of cost and risks is consistent with the sharing of expected benefits. It is not for determining whether income is generated by making contributions to a CCA. [F46]

A CCA is an arrangement where resources and skills are pooled and the consideration received is, in part or whole, the reasonable expectation of mutual benefit. [F47] It is the existence of mutual expected benefit, and its inclusion in the consideration received for contributions made, that warrants a different approach to determining an arm's length value for those contributions than ordinarily applies outside a CCA context.

In measuring each participant's contribution, it should be borne in mind that parties to a CCA expect to make their return from being able to exploit over time the results of the CCA activity rather than from an immediate mark-up on their contributions. The concept of a CCA is an arrangement into which independent enterprises would agree to enter without the intention of earning a profit directly from the conduct of the activity under the arrangement, but rather from exploiting the results of the activity if successful. In applying the arm's length principle to a CCA between associated enterprises, the appropriate comparison is with similar types of joint venture arrangements in which independent enterprises do not intend or seek to make a profit from charging each other in respect of their inputs to the arrangement. It is not a requirement of the arm's length principle that a participant receives an arm's length consideration for property or services contributed to a CCA in the form of immediate receipt, at the time of contribution, of its market value or price.

In this regard, a CCA avoids the difficulties involved in requiring the separate determination of arm's length prices for the two-way flows of benefits provided and received by the participants. For instance, take the case of a CCA where one participant performs research activities and another provides funds. Instead of the first participant being rewarded at a market price of cost plus a margin for the research services it has performed for the benefit of the second, and that participant being rewarded by a margin on the funds it has supplied to the benefit of the first, the costs of both might simply be shared and rewarded not through any margins but through commensurate sharing in the expected benefits from use of the results of the CCA activity.

To ensure an arm's length relationship between the sharing of contributions and expected benefits, contributions should be measured in a way that reliably determines their relative value. Accurately determining the relative arm's length values of all participants' contributions may be a difficult exercise in practice, particularly where those contributions are of different types, ie. tangible or intangible property, services or cash. Recognising this, the aim of the exercise should be to measure the relative values of the contributions using the most reasonable, practical and reliable basis for estimating values using available data.

Contributions to a CCA may be in cash or in kind (ie. tangible property, intangible property or services). Where a CCA involves a contribution in kind by a participant, independent parties would ordinarily be expected to agree that the contributions of all participants are valued on a consistent basis so as to reflect their real relative values. For instance, if one participant's contribution is measured using market value, independent parties would not ordinarily be expected to agree that the contribution of another participant be measured using historical cost, where this materially differs from market value.

In some circumstances, the most reliable basis for determining the relative values of contributions to a CCA may be market value. This may be the case where valuing participants' contributions on some other basis, e.g. historical cost, would fail to adequately reflect their real relative values. Bearing in mind what is said at paragraph 84 above, the use of market values or prices to measure a participant's contributions to a CCA does not imply that other participants should immediately remunerate the participant through payment to it of such amounts. For instance, valuing a contribution of services on a cost plus a mark-up basis does not imply immediate remuneration of the services at a profit to the contributor. It simply measures the cost of a participant's contribution for purposes of relating this to its expected benefits from the CCA. Provided the participant's share of the expected benefits is consistent with its share of the costs, appropriately measured, those benefits are sufficient compensation for those costs.

The use of market value measures the cost of a contribution by its true economic cost to the contributor. This 'opportunity cost' is the loss to the contributor from not using its contribution in an alternative way to the CCA. By entering into a CCA, a participant is relinquishing profits that it could have derived from alternative uses of the property or personnel contributed, for expected benefits to be derived from successful completion of the CCA. Given that an independent party contributing to a CCA would expect benefits in return that are commensurate with what it had given up in order to make its contribution, the cost of the contribution might be measured by estimating the price the contributor could have received had it instead sold the contribution on the open market.

It is not merely contributions in kind that have an opportunity cost. A cash contribution may be viewed as having a similar cost. Arguably, the opportunity cost of a cash contribution is its face value marked-up by a margin or profit to reflect the return that the contributor could otherwise have earned had it invested the money in an alternative way. This may be viewed as analogous to, for instance, treating the opportunity cost of a service contribution as the costs of performing the service marked-up by a profit to reflect the price that the contributor could otherwise have earned had it supplied the service on the open market.

Given that all types of contribution may be considered to have an opportunity cost in this way, differences between opportunity cost or market value and historical cost might be expected to exist for all types of contribution. Thus, to the extent that the concept of a contributor forgoing a current return on its contribution for expected benefits from the CCA activity commercially justifies use of historical cost rather than opportunity cost to value contributions, this applies to all types of contribution.

A participant's cost contributions represent its investment in the CCA venture. In determining the extent to which a CCA participant shares in the risk of the venture, and hence should share in its expected benefits, it is important to take account of the participant's total investment risk in the CCA activity. A participant's share of the expected benefits should reflect an arm's length return on its investment in the CCA. If one party is putting itself at risk to a greater extent than another, then as an independent party dealing at arm's length it would be expected to demand a commensurately greater return through a greater share of the expected benefits.

Different levels of risk are inherent in different forms of CCA contribution. The risk of loss attaching to a contribution of services differs from that attaching to a contribution of cash. The risk of loss associated with a CCA may not simply be the risk that the CCA activity is unsuccessful, so that the participants do not recover the cost of their contributions and receive a return on their investment. Other risks are assumed from performance of the CCA activity. The extent to which the participants to the arrangement share these risks must be taken into account in valuing the contribution of a participant that performs the CCA activity. [F48]

A participant in a CCA that is not a pure service arrangement [F49] may contribute by performing some or all of the CCA activity. In accordance with paragraph 84 above, the arm's length principle does not require inclusion of a profit mark-up in measuring such a contribution, provided the associated costs and risks are jointly shared amongst all participants consistent with their sharing of expected benefits from the activity.

This proviso means that the extent to which the risks assumed in performing the CCA activity are shared by the participants to the arrangement must be taken into account in valuing the contribution of a participant that performs the activity. For instance, an accident during R&D activities conducted under a CCA into development of hazardous chemicals may result in liability for fines and civil damages for environmental pollution. Depending upon the terms of the CCA agreement, the party to the CCA that is performing the activities may be solely liable in such an event, or all participants may have some degree of joint liability. However this risk is shared between the participants, such sharing should accord with the sharing of the expected benefits of the venture. Where all participants jointly share this risk in accordance with their sharing of expected benefits, the service provider as an independent party might agree to use historical cost in measuring its contribution. Where the participant performing the services solely assumes such risk, the cost of its contributions should be measured on a basis that reflects such risk. This might be historical cost if it includes the cost of relevant insurance, or might be a market price for the service on the basis that this would be expected to reflect the value of the risks assumed in providing the service.

Thus, historical cost may be used in appropriate circumstances to measure a participant's contribution of services to a CCA. However, in some circumstances it may be necessary to use market value. This may be the case where independent parties would be expected to agree that this basis more reliably determines the relative values of the service contribution and other contributions to a CCA. [F50]

Where historical cost is the most appropriate basis for measuring a contribution of services, the general accounting rules adopted by the participant making the contribution should be used in measuring the costs. For instance, where more than one participant performs some of the CCA activity, each would use the generally accepted accounting rules applicable in its jurisdiction to determine its costs.

AusCo, its parent ForCo and several affiliates resident in other countries are members of a MNE group operating in the automotive industry. Each group member produces vehicles for sale in its local market. ForCo operates an R&D facility responsible for all aspects of design and technology for all vehicle models produced by the group. This activity takes place under a CCA, which provides that the costs and risks of operating the facility and performing the R&D activity are shared amongst the participants based upon their expected benefits from use of the R&D results in the production of vehicles.

In these circumstances, given that the costs and risks of performing the R&D activity are jointly shared amongst all participants consistent with their sharing of expected benefits from the activity, it accords with the arm's length principle that ForCo's contribution of services is measured using historical cost (i.e. without a profit mark-up).

AusCo, its parent ForCo and several affiliates resident in other countries are members of a MNE group operating in the automotive industry. Each group member produces vehicles for sale in its local market. Each operates its own R&D centre responsible for all aspects of design and technology for its locally produced vehicle models. Each member shares the knowledge and results of its R&D activity with other members. This information is stored in a global database to which all members have unlimited access. ForCo's R&D centre is by far the largest in the group, and produces the majority of the shared results. Any centre may develop technology that can be used if desired by another member in developing its product. This activity takes place under a CCA, which provides that the costs and risks of operating the R&D centres are pooled and shared amongst the participants based upon their expected benefits from use of the R&D results in the production of vehicles.

In these circumstances, the participants all jointly perform the CCA activity. It accords with the arm's length principle for the total costs and risks of the activity to be pooled and then shared, irrespective of which participant performed the activity, in a manner that corresponds to the sharing of the expected benefits for the individual participants arising from the joint activity.

AusCo and ForCo are members of a MNE group and participants in a CCA for R&D activity. Each expects to benefit by using results of the R&D in the manufacture and sale of products in its local market. Each operates an R&D facility that will perform parts of the CCA activity. AusCo's facility is in Australia, and ForCo's facility in another country. There are significant cost of living differences between the two countries, so that major costs of the R&D, such as wages and rents, are significantly higher in the other country.

As in the previous two examples, it might accord with the arm's length principle for the CCA to provide for the costs and risks of the R&D activity to be pooled and shared amongst the participants based upon their expected benefits from use of the CCA results. If this is the case, costs are pooled and shared, irrespective of who performed particular activity and who incurred the costs of that activity. All participants share the cost and risk of all of the activity. This includes the geographical market risk that impacts upon the cost of performing the activity in a particular location. Therefore, if due to location cost differences ForCo incurs higher cost than AusCo in performing an otherwise comparable activity, this is not directly relevant or taken into account in measuring their relative contributions. However, it may impact the sharing of costs if such market factors also impact the participants' expected benefits (i.e. revenues or cost savings). [F51] Thus, as ForCo and AusCo will each be using the R&D results to manufacture and sell in their local markets, the market differences may mean that ForCo will expect to derive relatively higher revenues or cost savings per unit from use of the results than AusCo. This greater expected benefit should result in ForCo being allocated a commensurately greater share of the pooled costs on a per unit basis than AusCo.

In such circumstances, we would look at why activity is being performed in a relatively high cost location in assessing the commerciality of the arrangement. A possible commercial rationale may be that the facilities, resources or expertise needed to perform the activity are available in that location and not in Australia. Absent some such explanation, a question arises as to whether an independent party in AusCo's position would enter into an arrangement that unnecessarily increases the costs of the activity.

As previously discussed, [F52] a CCA may relate to activity that, unlike R&D, is not expected to benefit the participants by producing exploitable assets or rights. This may give rise to an issue as to whether the views expressed at paragraphs 83-84 and 94 above should apply to such a CCA. The answer will depend upon the particular circumstances, and whether there is both a sharing of the costs and risks of the CCA activity and an expected benefit for a participant other than from performing that activity.

In some cases the expected benefits of a participant in a CCA that is a pure service arrangement may come solely from performance of the CCA activity. If this is so, then arguably the service provider should not be treated as a participant in a CCA. [F53] For instance, a participant may perform typical 'head office activity' of a MNE, such as management, accounting, HR and IT support, for the benefit of itself and other group members as participants. In such a case, as an independent enterprise the service provider would expect to obtain its return on a current basis and at a market price, because the CCA provides no other means to obtain an arm's length reward for its contribution. Unlike a CCA under which contributions are in some part rewarded through benefits expected from exploiting what is developed from the CCA activity, the service provider in such a CCA has no way of getting a return other than charging a market price for performing the service. Similarly, the other participants as independent parties receiving the service would expect to incur costs on a current basis since they obtain their benefits on this basis.

This means that the result for the participants under such a CCA will be the same as where the services are performed outside a CCA context. In other words, the result will accord with the application of the arm's length principle as per TR 1999/1 and Chapter VII of the 1995 OECD Report. Those guidelines indicate that an arm's length charge for services will normally include a mark-up on the costs of performing the services. [F54] However, there may be circumstances where an arm's length charge will not exceed the costs incurred by the service provider. [F55]

The benefit of this type of arrangement for a MNE group is that it lowers the total cost of obtaining the services, relative to purchasing them in the market, because the market return for the functions, assets and risks is retained within the group. However, it will not lower the cost to individual members who do not undertake any of the activity, apart from cost savings that may arise from the pooling arrangement.

There may be other cases where a participant in a CCA that is a pure service arrangement expects benefits other than from performing the CCA activity. In such a case the service provider might expect to obtain an arm's length return on its contribution other than from an immediate mark-up on its costs of performing the activity. The views expressed at paragraphs 83-84 and 94 above may appropriately apply to such a CCA.

AusCo, ForCo1 and ForCo2 are members of a MNE group. Each is a distributor, in its own geographic market, of a range of consumer goods purchased from third party suppliers. Each has its own purchasing department. It is decided to enter into a CCA, so that all purchasing will in future be handled by just one department. The remaining departments will be closed, resulting in significant cost savings. In addition, all of the companies can expect to benefit from significantly higher volume discounts allowed by suppliers. The CCA provides for ForCo1 to perform the purchasing department activity and AusCo and ForCo2 to make cash contributions. All participants are to share the costs and risks of the activity, including associated operating risk and inventory risk. The costs and risks are to be shared, based upon the sharing of expected benefits, as follows: AusCo 20%, ForCo1 40%, and ForCo2 40%. In the relevant income year, ForCo1 incurs relevant costs of $900,000.

In these circumstances, given that the costs and risks of performing the purchasing activity are jointly shared amongst all participants consistent with their sharing of expected benefits from the activity, it may accord with the arm's length principle that ForCo1's contribution of services is measured using historical cost (ie. without a profit mark-up). As an independent party, ForCo1 might agree to a share of the expected benefits from obtaining higher volume discounts on stock purchases as a sufficient return on its contribution. On this basis, in order to achieve a sharing of contributions proportionate to the sharing of expected benefits AusCo and ForCo2 would make payments to ForCo1 of $180,000 and $360,000 respectively.

The facts are as in the previous example, except that instead of ForCo1 operating the purchasing department, it will be operated by ForCo3, a group service centre. The contributions of AusCo, ForCo1 and ForCo2 will all be in cash. ForCo3's only benefit is from performing the purchasing services, and it is therefore not a participant in the CCA. In the relevant income year, ForCo3 incurs costs of $900,000 in performing the purchasing department activity. An arm's length charge for the services is determined to be $1 million.

In these circumstances, AusCo, ForCo1 and ForCo2 will contribute to the $1 million due to ForCo3, by making payments in proportionate to their sharing of expected benefits of $200,000, $400,000 and $400,000 respectively.

To determine the value or amount that independent parties would have assigned to contributed property, tangible or intangible, it is first necessary to establish the basis upon which the property has been contributed to the CCA. This is determined by the intention of the parties to the CCA, and by what economic or legal rights and interests in the contributed property, if any, have been transferred by the contributor of the property to the other CCA participants.

For instance, it may be the intention of the parties that a building or piece of machinery contributed by one of the participants is to be made available for use in the CCA activity but remain the property of the party contributing it. This may be the most common situation in practice. Where historical cost is the most appropriate basis for measuring the contribution, its value would be the sum of the allowance for depreciation in the income year plus any other costs incurred in that year relating to the use of the asset, such as repair and maintenance expenses. If market value were the most appropriate basis for measuring the contribution, a market rental might represent the value of providing the asset.

Alternatively, in some cases it may be intended that all participants share joint economic ownership of the asset contributed. Where historical cost is the most appropriate basis for measuring the contribution, its value might be either the cost incurred to acquire it for contribution to the CCA or its written down value, as appropriate. If market value were the most appropriate basis for measuring the contribution, then the market value or price of the asset at the time of its contribution would be used.

Where historical cost is the most appropriate basis for measuring a contribution of property, the general accounting rules adopted by the participant making the contribution should be used in measuring the cost.

A participant's contribution to a CCA may be of intangible property developed or acquired outside the CCA ('pre-existing intangibles'). This may commonly occur upon establishment of a CCA for enhancement of, or development of the next generation of, these intangibles.

A participant's contribution of pre-existing intangibles may take different forms, depending upon the facts and circumstances, and in particular the terms of the arrangement and conduct of the parties. For instance, the contribution may be either the granting of a contractual right, a disposal of a part interest in the intangibles, or a licence to use the intangibles. These alternatives, and their implications for how the contribution is appropriately valued, are discussed at paragraphs 121-126.

While bearing in mind what is said at paragraphs 81 and 86 above, independent parties might ordinarily be expected to agree to use market value to measure a contribution of pre-existing intangibles. The market value of an intangible commonly significantly exceeds or falls short of its development cost. In either case, independent parties might not ordinarily be expected to agree that development cost reliably estimates the relative value of the contribution. Pre-existing intangibles may be acquired or licensed for contribution to a CCA, rather than developed by the contributor. In such circumstances, the value of the contribution might be based upon the acquisition or licensing costs, provided these are arm's length amounts.

The contributor of pre-existing intangibles may simply grant a non-exclusive right to the other participants to have the intangibles made available for use in the CCA activity and to share in any income from exploiting the results of that activity. For example, the contributor may wish to share the costs and risks of further development of the intangibles. In this case, which may be the most common in practice, the contributor has not provided the other participants with any right to use those intangibles outside the context of the CCA activity (i.e. to independently exploit the intangibles). In such cases, and subject to the contractual arrangement between the participants, the contributor retains the existing rights it has in respect of the intangibles (e.g. to earn income from the sale of products incorporating the intangible, to dispose of some or all of its legal and economic interests in the intangible, to earn royalty income from licensing others to use the intangible, etc). In such cases, the participants do not have joint economic ownership of the pre-existing intangibles. Their interests are limited to an economic interest in what is developed from the CCA activity.

In valuing such a contribution, it would not be appropriate to determine market value on the basis of the discounted present value of expected future profits from exploitation of the pre-existing intangibles. This is because the arrangement under which those intangibles are contributed to the CCA activity does not give the other participants a right to use the intangibles outside the CCA, which is retained by the contributor. Thus the opportunity cost of making the contribution in these circumstances is not the price that the contributor could have obtained for the intangibles if it had sold them in the open market instead of contributing them into the CCA. Subject to the terms of the CCA agreement, the contributor remains free to deal with the entirety of the rights and interests in the intangibles as it sees fit.

In some cases, which appear to be less common in practice, the contribution of pre-existing intangibles to a CCA activity may also involve an effective disposal of the intangibles to the other participants, so that they acquire a legally recognised economic interest in the intangibles. This interest may carry an entitlement to exercise ownership rights in relation to the intangibles, including the right to use them independently of the CCA activity, and to derive profits or income from this. The contribution in such cases, depending on the facts and circumstances, may involve the transfer of a partial interest in those intangibles from the contributor to one or more of the other participants or the granting of an exclusive licence to them.

In such cases, and as an additional amount to that discussed at paragraph 122 above, it might be appropriate to determine the market value of this component of the total contribution on a basis which would take account of the discounted present value of expected future profits from use of the intangibles outside the CCA activity. The opportunity cost of making the additional contribution in these circumstances is the price that the contributor could have obtained for the intangibles if it had disposed of a similar interest in the open market.

A contribution of pre-existing intangibles might also involve one or more of the other participants obtaining a right to use the intangibles for its own benefit outside the CCA activity. For example, under a licensing arrangement for which an arm's length royalty would be payable.

In such cases, as for those discussed at paragraph 124, it might be appropriate to determine the market value of this component of the total contribution on a basis which would take account of the discounted present value of expected future profits from use of the intangibles outside the CCA activity. The opportunity cost of making the additional contribution in these circumstances is the price that the contributor could have obtained for the intangibles if it had licensed them in the open market.

A participant may be entitled to some form of government assistance measure (e.g. a subsidy, grant, cash incentive or tax benefit) in connection with its involvement in the arrangement. For example, a taxpayer participating in a CCA for R&D may be entitled to the concessions under section 73B of the ITAA 1936. This may impact the valuation of that participant's contribution. If that contribution is valued after taking account of the benefit of the subsidy, ie. the cost of the contribution is measured net of the subsidy, then all participants effectively share in that benefit. If not, then the benefit is wholly retained by the participant to whom the subsidy is granted.

How the subsidy should be taken into account in valuing a participant's contributions depends upon what independent parties would have agreed to in comparable circumstances. [F56] If a participant receives a subsidy as a result of undertaking the CCA activity, it might ordinarily be expected that independent parties would share that benefit. This reflects that the activity is performed for the joint benefit of all participants. On the other hand, if a participant receives a subsidy for making a contribution to the CCA activity, an independent party might not share that benefit.

Where the results of the CCA activity are used to generate income, an issue may arise as to the correct treatment of that income. For instance, a CCA for R&D may create know-how that is made available to non-participants in consideration for receipt of royalties. The CCA activity may produce such exploitable results that are used by the participants to derive income during continuation of development activity under the CCA. In such circumstances, an issue arises as to whether the royalty income can appropriately be deducted from the costs to be shared, so that it is taken into account in determining the value of cost contributions and the amounts of any balancing payments to be made.

This treatment is inappropriate in a CCA that is a development-only venture under which each participant has a separate interest in the results of the CCA activity and is solely entitled to any income from exploiting that interest. It may be appropriate where the CCA extends to joint exploitation of the results of the CCA activity, so that each participant is beneficially entitled to a share of the resulting income. In other words, where the income is shared on the same basis as the costs are shared.

A taxpayer, AusCo, is a participant in a CCA for R&D. The other participants are ForCo, a non-operating holding company that is the ultimate parent company of the group, and several subsidiaries located in various countries. The subsidiaries all manufacture and sell a similar range of products in their respective countries. The subsidiaries, including AusCo, all perform R&D activity that produces know-how relating to the products. Under the CCA, AusCo and the other subsidiaries each have an exclusive royalty-free right to exploit the know-how, either directly or by licensing, in their respective countries. ForCo has a similar right in respect of anywhere else in the world. ForCo receives royalties from licensing the know-how to group affiliates who are non-participants in the CCA. The costs of the R&D activity are pooled and shared between the participants, with AusCo's share determined by the following formula: AusCo net sales / Participants' net sales * (R & D costs - royalties received)

In these circumstances, while the pooling and sharing of the costs of the R&D activity may be appropriate, [F57] the formula used is unlikely to produce an arm's length outcome due to inappropriate treatment of the participants' rights to license the results of the activity. First, the denominator in the sales allocation key fraction is understated. It should be worldwide net sales, ie. include also sales of non-participants who use the know-how under license from the participants, not merely the total sales of the participants. This properly reflects the totality of the rights to the know-how that are shared between the participants, and of their expected benefits from exploitation of those rights. Secondly, the amount to which the sales allocation key fraction is applied should simply be the total R&D costs, without netting off against royalty income. That income belongs solely to the participant whose interest is exploited to derive it, i.e. ForCo. It is not appropriately shared among all participants by reducing the pool of costs to be shared. In addition to these issues, there is a threshold issue as to the appropriateness of using sales as an allocation key in these circumstances, given that the participants may expect to use the results of the R&D activity in different ways. [F58]

Balancing payments may be needed to allow for differences in contributions made by participants. Such payments may also be required where an up-front payment is made based upon an estimate of a participant's share of the costs for a period, which is later balanced when actual costs are known.

A balancing payment increases the value of the contributions of the payer and correspondingly reduces the value of the contributions of the payee. [F59]

A CCA may relate to advertising and promotional activities. For instance, members of a MNE group may jointly contribute to the cost of developing and implementing a global marketing campaign to promote brands and logos of products that each member sells in its own local geographic market.

An aim of such marketing activity is to develop, maintain or increase the value of relevant marketing intangibles, e.g. trade names, brand names and trademarks. Such marketing intangibles typically cost little to create, but have little value until marketing activity develops the value. Thus, most of the cost and risk attaching to marketing intangibles relates to the marketing activity needed to develop, maintain or increase their value. In this respect they differ from trade intangibles, e.g. patents, where most of the cost and risk relates to creation of the intangible, ie. through R&D activity. Marketing activity, like R&D, inherently involves risk, as there is no guarantee that the marketing intangible will have value as a result of the activity.

A participant that shares in the costs and risks of marketing activity under a CCA is entitled to a proportionate share of the results of that activity, in the form of an economic interest in the resulting value of the marketing intangible. Its CCA contributions are the consideration it gives for this interest, and it is not required to pay any further consideration, e.g. a royalty, for that interest or its use.

Where a participant performs marketing activity as its contribution to a CCA, the views at paragraphs 94-97 apply to measuring such a contribution of services.

As previously discussed, [F60] how a contribution of a pre-existing marketing intangible is appropriately valued will depend upon the form in which the contribution is made.

A marketing intangible may have different values in different geographic or product markets. It may be well known and have great value in one market, but relatively unknown and of little value in another. In a CCA for marketing activity, any such market differences should be accounted for in measuring expected benefits, [F61] and therefore in the sharing of the costs and risks of the activity. In this way, the contributions of a participant in a CCA for marketing of a brand name, trademark, etc. who has rights to exploit that intangible in a particular market are appropriately based upon its expected benefits from exploiting the resulting value of the intangible in that market.

Marketing activity may have a relatively short term effect on demand for a product. This is relevant to valuing the expected benefit from the activity. Estimates of expected benefit from a CCA for marketing activity may not require projections over a timeframe as long as that for a CCA for R&D. If this is so, current data is likely to be a more reliable indicator of expected benefits for a CCA for marketing activity than for a CCA for R&D. [F62]

As previously discussed in relation to intangibles created through R&D, [F63] there should be no 'double charging' for the right to use a marketing intangible. For instance, if a distributor has the right, as a CCA participant, to exploit marketing intangibles in its local market, the transfer price it pays for products should not include the value of that right.

As R&D activity generally does not yield present benefits, a participant's proportionate share of contributions to a CCA for such activity must be determined on the basis of projections as to that participant's expected future benefits from the activity. 'Expected benefits' may be defined for these purposes as projected revenues or cost savings to be generated from use of the participant's interest in the CCA results. [F64]

The accurate projection of expected benefits is commonly a difficult exercise in practice. One reason is that there is often a long period between the commencement of the CCA activity and the development of an exploitable result. This requires significant projections over time. Also, where a CCA governs only the development phase of a project, benefits are not derived under the CCA, but from the exploitation phase performed by each participant individually and independently of the CCA and its other participants. The extent to which each participant can successfully and profitably exploit its interest in the results of the CCA activity will ordinarily be impacted by many economic, commercial and market factors peculiar to that participant. In such circumstances, the projecting of relative benefits is likely to be significantly more difficult than under a CCA where the participants are also jointly exploiting the results of the CCA activity and entitled to fixed shares of whatever income or profits might be derived.

Recognising these difficulties, it is important to bear in mind that: (a) the aim is to measure the relative expected benefits of the participants using the most reasonable, practical, and reliable basis for estimating the projected revenues or cost savings from exploiting their interests in the results of the CCA activity, having regard to the circumstances of each participant; (b) in many cases, it may not be possible or appropriate to insist that a single projected result is the only one that satisfies the arm's length principle. Rather, there may be a range of projected results, all of which are equally reasonable and reliable as estimates given available data at the time the projections are made; and (c) the terms of a CCA might appropriately provide for adjustment to the sharing of contributions to account for changes in circumstances that result in changes in the sharing of expected benefits. [F65]

Expected benefits may be estimated directly as the projected revenues or cost savings to be generated from use of the participant's interest in the CCA results. However, in many cases such direct estimation may be impractical. It may then be necessary to resort to an indirect basis for measuring relative expected benefits, and which may therefore appropriately be used as an allocation key for costs. The principles that apply to the selection and use of allocation keys in this context are similar to those in applying an indirect method of charging for intra-group services. [F66] Possible indirect bases that might be used include sales values, production or sales volumes, gross or net profit, numbers of employees, asset values, and capital invested. What is the most appropriate allocation key will depend upon the facts and circumstances, and in particular the nature of the CCA results, how each participant expects to benefit from using those results, and the availability and reliability of the data needed to apply the key.

It may be necessary to make adjustments to an allocation key to account for differences in the benefits expected by the participants. [F67] The most appropriate allocation key will be that which can reliably be used either without adjustment or with the least or most accurate adjustment to account for such differences. In some cases, it may be appropriate to use more than one key to reliably estimate the differing expected benefits of the participants.

Projected production or sales volumes or sales values are commonly used to measure expected benefits in CCAs for R&D to develop new products or production processes. The use of such an allocation key, without adjustment, assumes that a comparable relative benefit (i.e. contribution to revenue or cost savings) is derived by each participant from use of the results of the CCA activity for each unit produced or sold or dollar of sales revenue generated. Therefore, an essential condition to reliably using such a key on an unadjusted basis is that this assumption holds true in the particular circumstances.

The assumption may not hold true if there are differences in the circumstances of the participants that could materially affect their expected revenues or cost savings from use of the CCA results, and those differences are not taken into account in projecting production or sales volumes or sales values. Each participant's expected revenues or cost savings from use of the CCA results may be impacted by a variety of business, economic and market factors peculiar to that participant. For instance, the participants' interests in the CCA results may relate to markets that differ significantly as to factors such as production costs, government regulation or other risks related to exploitation of the CCA results. Such factors should be accounted for in using a sales-based allocation key to make projections as to sales volumes or values, to the extent that they impact the volume or value of sales made. Any significant differences in the impact of such factors on the participants' relative revenues or cost savings from use of the CCA results that are accounted for in this way should not adversely affect the reliability of a sales-based allocation key. On the other hand, any such differences that are not so accounted for may affect the reliability of such a key.

The above assumption may also not hold true where the participants are not all expecting to use the results of the CCA activity in the same way. For instance, use of an unadjusted sales-based allocation key in a CCA to develop production intangibles may not reliably measure the relative expected benefits of the participants where one expects to use those intangibles to manufacture and another expects to benefit by licensing the use of the intangibles. A manufacturer that owns production intangibles generally keeps all income from their use, whereas a licensor owner of such intangibles generally shares some of that income with the licensee. Accordingly, such a key would be expected to favour the manufacturer participant, by understating its share of expected benefits, and hence share of costs, relative to the share of expected benefits and costs of the licensor participant. In such circumstances it might be possible to make suitable adjustments to the allocation key to enable its reliable use, or it might be more appropriate to use either a different key or more than one key.

AusCo, ForCo1 and ForCo2 are members of a MNE group and participants in a CCA to jointly develop manufacturing technology. AusCo expects to exploit the technology by licensing its use to affiliate manufacturers in the Asia Pacific region. ForCo1 will exploit the technology by using it to manufacture and sell products to retail customers in its local market. ForCo2 will exploit the technology by using it to manufacture products that it will sell to affiliates for distribution to wholesalers in their local markets. The CCA agreement provides for the participants' contributions to be based upon relative sales values. Thus, AusCo's contributions are based upon projected sales revenues of the relevant products by the manufacturers to whom AusCo licenses the technology. The contributions of ForCo1 and ForCo2 are based upon their projected sales revenues for the relevant products.

In these circumstances, it is unlikely that use of this measure will result in a sharing of contributions in proportion to the sharing of expected benefits. The relationship between the use of the CCA technology and the profit from that use, measured by reference to sales value, is not the same for all of the participants. A participant using the CCA technology to manufacture products is likely to derive a different level of profits from that use, as a percentage of sales revenue, to a participant that uses the technology by licensing it to derive royalty income. Thus, unadjusted sales values may not reliably measure the relative expected benefits of AusCo and the other participants. As ForCo1 and ForCo2 are selling at different levels of the market (i.e. ForCo1 to retailers and ForCo2 to wholesale distributors) the unadjusted values of their sales may not reliably measure their relative expected benefits.

AusCo, ForCo1 and ForCo2 are members of a MNE group and participants in a CCA to jointly develop manufacturing technology. Each expects to exploit the technology by using it to manufacture and sell products to retail customers in its local market. The CCA agreement provides for the participants' contributions to be based upon relative sales values. The MNE group operates three distinct business segments: consumer products, medical products and pharmaceutical products. The CCA covers R&D relevant to all of these segments. Operating margins vary significantly between business segments, with pharmaceutical products in the circumstances significantly more profitable than the other products. Pharmaceutical products are a significantly lower percentage of AusCo's total sales than for ForCo1 and ForCo2.

In these circumstances, it is unlikely that use of this measure will result in a sharing of contributions in proportion to the sharing of expected benefits. Given the differences in product profitability and sales composition, relative sales values may not reliably measure the relative expected benefits of the participants.

It might be argued that it is not commercially realistic to expect CCA participants to share information as to their projected benefits from the arrangement, as independent parties would not divulge such confidential information. In other words, the application of the arm's length principle to a CCA should not assume a degree of information and knowledge sharing that would not exist between independent parties to such an arrangement.

However, an independent party would not be expected to agree to a basis for sharing costs unless it had access to sufficient information to be able to satisfy itself that this basis produces an outcome that is in its economic and commercial interests. For instance, where the participants agree to use of a sales-based allocation key, it is implicit that they have agreed to divulge sufficient sales data that the outcome of this key for the sharing of costs can be evaluated and known by all participants. It is not uncommon for independent joint venturers to share information on the basis of a strict confidentiality agreement. The arm's length principle requires that the participants to a CCA have the right of access to a certain level of information essential to the operation of the terms of the arrangement. [F68]

It may, in appropriate circumstances, be possible to use data as to the participants' current sales volumes, values, profit margins or other indirect measures to estimate their relative expected benefits from the CCA. Given that it is often difficult to make accurate projections, the use of current data may be a more reliable alternative in some cases. To use such data, it would be necessary to demonstrate its reliability as an indicator of future results. The use of such data for this purpose assumes that the current relativity between the results of the participants will be maintained, and will not materially alter either through impact of the use of the results of the CCA activity or other factors (e.g. changes in economic, business or market circumstances).

The guidelines for intra-group services provide that in appropriate circumstances an indirect charge method (e.g. using allocation keys) may be used to estimate an arm's length charge for expected benefits provided. [F69] As previously discussed, these guidelines apply to a CCA that is a pure service arrangement, [F70] so that similar allocation bases are relevant to determining the sharing of cost contributions in such an arrangement.

Events after the making of benefit projections or agreement as to benefit estimation methods (e.g. allocation keys) may give rise to questions about the reliability or commerciality of the projections or methods. This may be due to changes in relevant economic, business or other circumstances. Or actual benefits realised may differ from projected benefits. In such situations, issues arise as to whether a benefit projection or estimation method, and the resulting sharing of cost contributions, should be adjusted.

As previously stated, [F71] whether the terms of a CCA accord with what independent parties dealing at arm's length would be expected to agree should be judged by reference to circumstances known or reasonably foreseeable at the time of entry into the arrangement. [F72] Therefore, where benefit projections are consistent with this standard and later events impacting their reliability or commerciality were unanticipated or unforeseeable at the time the projections were made, there will be no retrospective adjustment in respect of past use of the projections to share contributions. This would be a use of hindsight that is inconsistent with the arm's length principle.

On the other hand, adjustment of the original projections, with effect as from the time they were made, would be appropriate where: (a) there is evidence that they were not made in a genuine attempt to estimate expected benefits; or (b) the later events were reasonably foreseeable at the time the projections were made.

In such unusual cases, where the projections are considered to lack commerciality, it is appropriate that they be disregarded and projections substituted that accord with those that independent parties might be expected to have made in similar circumstances.

In some cases, later events may call for a prospective adjustment or modification of benefit projections or estimation methods, and the resulting sharing of cost contributions. This will be the case if independent parties would be expected to have agreed to take account of such eventualities by making such adjustments.

In this regard, a CCA should provide that the sharing of contributions be subject to review and prospective adjustment to account for material changes in economic or expected circumstances, [F73] if this is what independent parties would be expected to agree when negotiating the arrangement. This is a commercially realistic recognition and response to the difficulty in many cases of reliably projecting future benefits. For instance, where a project is to be performed over a relatively long timeframe, independent parties might be expected to agree milestones at which the future conduct of the project, including the sharing of contributions, is reviewed.

A party's entry into or withdrawal from an active CCA may alter the contractual basis of the arrangement in a way that the terms of the CCA call for prospective adjustment of benefit projections or estimation methods used for determining the future sharing of costs for any or all of the participants.

The guidance at paragraphs 159-164 above also applies where actual benefits realised differ from projected benefits. Given the difficulties previously discussed in projecting benefits, such differences will not be uncommon. The arm's length principle does not require that projections of benefits used to share costs be shown to accord with benefits actually realised. Even if there is a material difference between projected and actual benefits, this does not necessarily warrant disregarding the projections or the arrangement as non-arm's length. It is necessary to examine the reasons for this difference.

At the time of entry into a CCA whose benefits are expected to be realised in the future, the parties could not know what actual benefits might eventually be realised. Accordingly, it is ordinarily an inappropriate use of hindsight to disregard benefit projections and retrospectively apply actual benefits to adjust the sharing of contributions.

However, evidence of actual benefit, and the extent to which it is consistent with projections made by the parties, is relevant, but not conclusive, in determining the reasonableness of those projections, and whether they are likely to have been made by independent parties. [F74]

The experience as to actual benefit may also be relevant where the terms of a CCA between independent parties might be expected to provide for adjustment to the sharing of contributions to account for changes in circumstances that result in changes in expected benefits. [F75]

It is only in extreme cases where the difference between the projected and actual benefits warrants a conclusion that the projections lack commerciality, that it is appropriate that the projections be disregarded. This will be the case where: (a) the difference between projected and actual benefits is of such a degree as to evidence that the projections were not made in a genuine attempt to estimate expected benefits; or (b) the difference between projected and actual benefits is due to circumstances that should have been known or reasonably foreseeable at the time the projections were made.

One type of case where expected and actual benefits differ is where the activity performed under a CCA produces unexpected results. For instance, a CCA intended to develop a drug for treatment of a certain medical condition may result in discovery of other drugs, either instead of or additional to the intended drug. Or work on developing a product may have unexpected spin-off applications that can be commercially exploited in other areas. These unexpected outcomes may be of greater benefit (i.e. more profitable) than the expected outcomes.

AusCo, a member of a MNE group, is a manufacturer of widgets. AusCo enters into a CCA to develop technology for a new generation of widgets. Under the CCA, AusCo has the right to use the CCA technology to manufacture for sale in the Australian market. The other participants, who are also widget manufacturers, have similar rights in their local markets. Cost contributions are shared based upon projected benefits from use of the results of the R&D in the production of widgets. The R&D performed under the CCA unexpectedly develops technology that can also be used to manufacture gadgets, a product totally different to widgets.

In these circumstances, the actual benefits derived exceed the expected benefits. The additional benefits are in respect of a new and different product to that upon which the sharing of expected benefits and costs was originally based. Independent parties might be expected to have originally agreed to a CCA that provides for review and prospective adjustment of the sharing of contributions to account for such a material change affecting their expected benefits. Otherwise, independent parties might be expected to renegotiate the CCA. While there may be no need to revise the original projections related to widgets, new projections related to gadgets may also be required for the sharing of future contributions. If each participant has the same relative interest in the CCA technology in respect of its use for gadgets as it does for widgets, then it will have obtained that interest by sharing costs only on the basis of projected benefits in respect of widgets. The participants' relative projections of benefits for gadgets may be very different to those for widgets. However, independent parties might not agree to account for this by providing for retrospective adjustment of contributions towards past CCA activity.

Where an entity joins or leaves an active CCA, an arm's length result is required for any transfer of a valuable interest in the results of past CCA activity. [F76]

A 'buy-in' payment by a new participant joining a CCA compensates the existing participants for the new participant obtaining an interest in the results of past CCA activity. [F77] Likewise, a 'buy-out' payment compensates a departing participant for a transfer of its interest in the results of past CCA activity to the benefit of the remaining participants.

The interest transferred as a result of an entity's entry into or withdrawal from an active CCA typically comprises an economic interest in presently existing intangible property developed under the CCA, in the work-in-progress being undertaken within the CCA at the time of entry or withdrawal, and rights to the knowledge resulting from past CCA activity.

R&D activity is inherently a cumulative process in which future activity builds upon past activity. The results obtained in future activity are in part the fruit of experience and knowledge obtained from past activity. Given that past CCA activity has a value for future CCA activity in this way, independent parties would be expected to agree to the making of an adjustment payment for transfer of an interest in the results of past CCA activity upon a party's entry or withdrawal during the continuance of a CCA.

The making of a buy-in payment by a new participant reflects that it obtains the benefits of know-how and other valuable intangible property resulting from past CCA activity that it will not be paying for through its future cost contributions. A new entrant to an active CCA need not obtain an interest in the results of past CCA activity. If those results remain exclusively available to the existing participants and the entrant benefits only from the results of CCA activity performed after its entrance, then there should be no buy-in payment.

The results of past CCA activity may have value even if this activity is regarded as 'failed R&D, such as work into developing a new product or technology that is abandoned as unsuccessful before any exploitable results are achieved. Know-how from failed R&D may have value for future R&D. Knowing what does not work is often an important step in the process of finding out what may or does work. For every successfully exploitable R&D result there may be many failed development activities. For instance, for every pharmaceutical drug successfully developed there may typically be thousands of unsuccessful experiments on various compounds.

The commerciality of the overall R&D effort is based upon the expected benefits (i.e. revenues or cost savings) from any successful results being sufficient to recover the costs of all related activity, including failed activity. A new participant joining a CCA after it has undertaken failed R&D shares in any expected benefits from such activity for future CCA activity. Accordingly, the making of a buy-in payment by the new participant in respect of such benefits would ordinarily be expected under the terms of a CCA between independent parties. Conversely, a participant leaving a CCA after it has undertaken failed R&D may give up to the remaining participants any expected benefits from such activity for future CCA activity. A buy-out payment to the departing participant in respect of such benefits should be made, provided independent parties would be expected to have agreed to a CCA that provides for such a payment in similar circumstances.

The results of past CCA activity may have no value, in which case the terms of a CCA between independent parties would not be expected to require a buy-in or buy-out payment. [F78] However, for the reasons discussed above it would not be expected that this would often be the case. In particular, it cannot be said that failed R&D necessarily has no value. Where past CCA activity is failed R&D, it will have no value only if there is no knowledge or other benefit obtained that is expected to have value for any future R&D activity.

Where a participant leaves a CCA, it may forfeit any interest in the results of CCA activity performed after its withdrawal. However, it may agree with the remaining participants to retain some or all of its interest in the results of past CCA activity. [F79] The departing participant may be able to exploit that interest, without needing any interest in the results of CCA activity performed after its withdrawal. Where it retains its interest it may later exploit that interest without payment to the remaining participants. For instance, it may exploit its rights under the CCA to use information, know-how or other intangible property resulting from past CCA activity without payment of a royalty to the remaining participants. In this case there is no buy-out payment to the departing participant in respect of such rights.

Alternatively, the departing participant may transfer or dispose of some or all of its interest in the results of past CCA activity to one or more of the remaining participants. In this case, to the extent that the interest is transferred to the benefit of a remaining participant, so that the value of its interest in the results of the past CCA activity is increased, the terms of the CCA should require a buy-out payment from that participant. Where a departing participant receives a buy-out payment for transfer of an interest in the results of the CCA activity, any later use of that interest by it should be compensated by payment of an arm's length consideration (e.g. royalty).

As a matter of commercial reality, the use by a departing participant of its knowledge resulting from past CCA activity is something that the remaining participants may have little ability to deny or verify. This may particularly be so with regard to legally unprotected know-how. Given this, independent parties might ordinarily be expected to agree that the departing participant has retained its interest in and rights to use such knowledge, so that no buy-out payment in respect of it is warranted.

In some cases the interest transferred by a departing participant to the remaining participants may have no value. Where the withdrawal does not benefit a remaining participant by increasing the value of its interest in the results of the past CCA activity, no buy-out payment from that participant is required. [F80]

For instance, a participant may withdraw because it no longer has an expectation of benefit. [F81] Depending upon the circumstances, any transfer of that participant's interest to a remaining participant either may or may not increase the value of its interest. [F82]

A participant's withdrawal may result in a reduction in the value of what was being developed under the CCA and of the continuing CCA activity. The importance of the departing participant to the CCA may mean that the new interests of the remaining participants in the results of the CCA activity are of lesser value than their former interests. For instance, the absence of contributions to future CCA activity that the departing participant would otherwise have made (e.g. highly skilled technical staff) may adversely impact the completion, and hence the value, of work in progress at the time of withdrawal. If the withdrawal of a participant disadvantages a remaining participant by reducing the value of its interest, the terms of a CCA between independent parties might be expected to call for a payment from the departing participant to the remaining participant.

However, a remaining participant is not necessarily disadvantaged in a relevant sense by another participant's withdrawal simply because it results in an increase in future contributions that may be required. There will be no such disadvantage if there is an increase in a remaining participant's share of the costs of undertaking the CCA activity that is properly compensated by a corresponding increase in its share of the expected benefits from that activity.

AusCo, ForCo1 and ForCo2 are members of a MNE group and participants in a CCA for R&D related to production process intangibles. AusCo has rights to exploit the results of the CCA activity through manufacture and sale of the products in the Australian market, ForCo1 has similar rights in its local market, and ForCo2 has rights to license the results for use in all other markets. A decision is made to close AusCo's manufacturing plant due to operating cost inefficiencies. Product for sale into the Australian market is to be manufactured in future by ForCo1. As AusCo will no longer be able to exploit the expected benefits attaching to its interest, it withdraws from the CCA.

In these circumstances, subject to the terms of the CCA agreement either ForCo1 or ForCo2 would be expected to agree to a transfer of AusCo's interest in the results of the CCA activity. Either ForCo1 might acquire AusCo's interest, or ForCo2 might acquire it for licensing to ForCo1. Whichever of the remaining participants acquires the interest, the value of that participant's interest in the results of the CCA activity is increased through AusCo's withdrawal. Accordingly, the terms of the CCA should require a buy-out payment from that participant to AusCo.

AusCo, ForCo1 and ForCo2 are members of a MNE group and participants in a CCA for product-related R&D. AusCo has rights to exploit the results of the CCA activity in the Australian market, and ForCo1 and ForCo2 have similar rights in their respective local markets. A stage is reached in the CCA activity when it becomes apparent that the benefits originally expected will not eventuate, but that other benefits are now expected. AusCo will not be able to exploit these new expected benefits because they relate only to products that are not marketable in Australia. AusCo therefore withdraws from the CCA.

In these circumstances, assuming the Australian rights to the results of the CCA activity have no value, ForCo1 and ForCo2 would not be expected to agree to a transfer of AusCo's interest. The values of the interests of ForCo1 and ForCo2 in the results of the CCA activity are not increased through AusCo's withdrawal. Accordingly, the terms of the CCA should not require a buy-out payment to AusCo.

If a buy-in or buy-out payment is required, the amount should reflect the arm's length value of the interest transferred. [F83] The basis used to determine an arm's length value will depend upon the particular circumstances and what independent parties might be expected to have agreed in similar circumstances.

To the extent that a buy-in or buy-out payment is for the transfer of an interest in intangible property, independent parties might commonly be expected to agree to base the payment upon the market value of the results of past CCA activity. The market value of an intangible rarely equates with its development cost. Rather, it is broadly based upon perceptions of the intangible's profit potential. Thus, the market value of the results of past CCA activity is ordinarily not simply determined by the costs of that activity.

It is recognised that it may not always be possible or practicable to accurately and reliably estimate the market value of the results of past CCA activity for these purposes. For this reason, or for other commercial reasons, independent parties might be expected in some circumstances to agree to value a buy-in or buy-out payment on some other basis.

For instance, independent parties might in some circumstances agree to use the cost contributions made in respect of past CCA activity as the basis for determining a buy-in or buy-out payment. Thus, a buy-in payment might be based upon reimbursing cost contributions made by the existing participants, so that it represents the total value of the contributions that the new entrant would have made if it had been a participant since commencement of the CCA activity. In other words, it may be commercially realistic that a buy-in payment is intended to put a new entrant in the same position as if it had been a participant since commencement of the CCA activity. Conversely, this may be the case in basing a buy-out payment on reimbursing cost contributions made by the departing participant, so as to put it in the same position as if it had never been a participant in the CCA. This would effectively value the results of the past CCA activity as approximating their development cost.

A taxpayer as an independent party looking to enter an active CCA may be able to use the benefit of hindsight to achieve a more advantageous position than if it had been a participant since commencement of the CCA activity. It therefore might ordinarily not be expected to agree to a buy-in payment determined on a cost basis if it is able to determine that this exceeds the market value of the results of the CCA activity at the time of its entry into the arrangement. However, the imperative for a taxpayer to enter or depart a CCA must be considered in the context of its business and financial circumstances to determine whether it makes commercial sense for the taxpayer acting in its economic interests to agree to the amount of a buy-in or buy-out payment.

It is consistent with the arm's length principle that upon termination of a CCA, each participant obtains an interest in the results of the CCA activity commensurate with its share of cost contributions made. [F84] If a participant surrenders or otherwise transfers that interest to another participant, similar principles to those discussed above in respect of buy-in and buy-out payments will apply in determining what, if any, consideration should be received for that transfer in accordance with the arm's length principle.

As previously stated, [F85] this Ruling deals only with whether the conditions of a CCA accord with the arm's length principle. It does not discuss the legal consequences of a CCA, or the taxation consequences related to the application of provisions such as those dealing with partnerships, deductibility of R&D expenditure, capital gains tax and royalty withholding tax. The language used in this Ruling has been carefully chosen with this in mind. For instance, we have intentionally referred to a CCA participant having an 'economic interest' in 'the results of the CCA activity', rather than a 'beneficial interest' in 'property'.

It is not possible to be highly prescriptive as to the conditions that are required for a CCA to comply with the arm's length principle. This is ultimately a matter dependent upon the facts in a particular case, and hypothesising what independent parties might be expected to do in similar circumstances. In most cases there are unlikely to be data as to comparable arrangements between independent parties. Given this, a CCA may be particularly appropriate for an Advance Pricing Arrangement.

Where we consider that the conditions of a CCA are not consistent with the arm's length principle, and this has resulted in detriment to the Australian revenue, we will seek to take the action that is most appropriate to produce an arm's length outcome for the taxpayer. The action taken in a particular case must depend upon the facts and circumstances. It will most often involve reducing the taxpayer's net contributions through making or imputing receipt of a balancing payment. [F86] In some cases it may be necessary to disregard part or all of the terms of the CCA. [F87]

Discussed below is the action that we may seek to take where a CCA is inconsistent with the arm's length principle in respect of each of the specific matters listed at paragraph 19.

Where we consider that the terms of a CCA are inconsistent with what would have been agreed by the taxpayer as an independent party acting in its own economic interests, or do not reflect outcomes that make business sense for the taxpayer, this can usually be rectified by adjusting contributions. [F88]

In some cases such action will not be sufficient to produce an arm's length outcome for the taxpayer. Where it does not make business sense for the taxpayer, acting in its own economic interests, to enter into a CCA compared to other available options, we may disregard the arrangement. The specific action we may take to produce an arm's length outcome for the taxpayer in such circumstances may include the following. The purported contributions made by the taxpayer may be disregarded to disallow deductions for costs incurred in respect of those contributions. Where other parties have used tangible or intangible property of the taxpayer, including any interest in that property that such parties have obtained as participants in the CCA, then receipt of an arm's length consideration for that use may be imputed. Where the taxpayer has performed activities for the benefit of other CCA participants, then receipt of an arm's length consideration for such services may be imputed.

Where the agreed terms of a CCA are inconsistent with the economic substance and commercial reality of the arrangement, we will base any adjustments needed to increase the assessable income or reduce the deductible expenditure of a taxpayer participant upon the terms that parties dealing at arm's length would be expected to have agreed in similar circumstances.

We may disregard terms of a CCA where there is no evidence that those terms were agreed prior to commencement of the CCA activity. We may also disregard the agreed terms where they are inconsistent with what independent parties would be expected to have agreed when entering into the arrangement (i.e. prior to commencement of the CCA activity).

The action we may take where the taxpayer as an independent party would not be expected to have participated in a CCA is as discussed at paragraph 203.

Where a party other than the taxpayer would not have been a participant had the CCA been arm's length, the action we may take to produce an arm's length outcome for the taxpayer may include the following. The taxpayer's interest in the results of the CCA activity and share of expected benefits and costs might be adjusted to reflect what independent parties might be expected to have agreed to if the relevant party had not participated. If that party has used the results of the CCA activity, then an arm's length consideration for that use (e.g. a royalty) may be imputed to the taxpayer, based upon its adjusted interest in those results. In some cases it may be either necessary or more appropriate to entirely disregard the CCA. An obvious instance would be where the taxpayer and the other party were the only purported participants.

Where the non-arm's length feature of a CCA is a disproportionate sharing of cost contributions and expected benefits, this situation can ordinarily be satisfactorily rectified by reducing the taxpayer's net contributions through making or imputing receipt of a balancing payment. [F89]

Where benefit projections or estimate methods lack reliability or commerciality, so as to be inconsistent with what independent parties might be expected to have agreed, they may be adjusted or disregarded for purposes of reducing a taxpayer's contributions.

In extreme cases where the sharing of contributions is significantly disproportionate to the sharing of expected benefits, simply reducing a taxpayer's contributions may not sufficiently redress the non-arm's length features of the arrangement. It may be necessary to increase the taxpayer's interest in the results of the CCA activity and impute receipt of an arm's length consideration for use of that interest by the other participants. [F90] This may be the case if the commercial reality is that it is how the interests in the results of the CCA activity are shared, rather than how the costs are shared, that does not accord with what independent parties might be expected to have agreed.

Where a CCA does not accord with the views at paragraphs 173-196, we may take action to impute or adjust a buy-in or buy-out payment where a taxpayer has either: (1) not received a buy-in or buy-out payment in circumstances where a payment is appropriate, or has received a payment that is less than an arm's length amount, resulting in an understatement of its assessable income; or (2) made a buy-in or buy-out payment in circumstances where no payment is appropriate, or has made a payment that exceeds an arm's length amount, resulting in an overstatement of its deductible expenses.

Similar action will be taken, as appropriate, in making any adjustment required to ensure an arm's length consideration for a transfer of an interest in the results of the CCA activity upon termination of the arrangement.

As previously discussed, [F91] independent parties would not be expected to enter into a CCA without a written agreement.

The general guidelines in Taxation Ruling TR 98/11 on the need to document the application of the arm's length principle to arrangements or dealings apply to CCAs. Thus, in general terms, the nature and extent of the documentation needed in respect of a CCA should be judged by reference to what a reasonable business person would consider appropriate, having regard to the relative importance and complexity of the arrangement. [F92]

The Pacific Association of Tax Administrators (PATA), whose members include Australia, Canada, Japan and the United States, has recently released the 'PATA Transfer Pricing Documentation Package'. The Package is intended as a uniform prescription of the documentation needed to evidence a taxpayer's efforts to comply with the arm's length principle, for purposes of satisfying each PATA member's documentation requirements and avoiding the imposition of transfer pricing penalties. In respect of CCAs, the Package lists the following, which is consistent with paragraphs 8.42 and 8.43 of the 1995 OECD Report: • 'A copy of the CCA agreement that is contemporaneous with its formation (and any revision) and any other agreements relating to the application of the CCA between the CCA participants • A list of the arrangement's participants, and any other associated enterprises that will benefit from the CCA • The extent of the use of CCA property by associated enterprises which are not CCA participants, including the amounts of consideration paid or payable by these non-participants for use of the CCA property • A description of the scope of the activities to be undertaken, including any intangible or class of intangibles in existence or intended to be developed • A description of each participant's interest in the results of the CCA activities • The duration of the arrangement • Procedures for and consequences of a participant entering or withdrawing from the agreement (ie. buy-in and buy-out payments) and for the modification or termination of the agreement • The total amount of contributions incurred pursuant to the arrangement • The contributions borne by each participant and the form and value of each participant's initial contributions (including research) with a description of how the value of initial and ongoing contributions is determined and how accounting principles are applied • A description of the method used to determine each participant's share of the contributions including projections used to estimate benefits, any rationale and assumptions underlying the projections, and an explanation of why that method was selected • The consistent accounting method used to determine the contributions and benefits (including the method used to translate foreign currencies), and to the extent that the method materially differs from accounting principles accepted in the relevant PATA member's country, an explanation of the material differences • Identification of each participant's expected benefits to be derived from the CCA, the extent of the benefits expected, and the formula and projections used for allocating or sharing the expected benefits, and the rationale and assumptions underlying the expected benefits • Where material differences arise between projected benefits and actual benefits realised, the assumptions made to project future benefits need to be amended for future years, and the revised assumptions documented • Procedures governing balancing payments, e.g. where payments are required to reflect differences between projected benefits and actual benefits realised'.

To reiterate the point made at paragraph 214, the extent to which each category of documentation listed above is needed for a CCA will depend upon the particular circumstances, and what is sufficient to demonstrate a reasonable effort to comply with the arm's length principle.

We invite you to comment on this draft Taxation Ruling. We specifically invite comments to assist in developing our views on the treatment of employee share option plans in a CCA context. We are allowing 6 weeks for comments before we finalise the Ruling. If you want your comments to be considered, please provide them to us within this period. Comments by date: 10 September 2003 Contact officer details have been removed following publication of the final ruling.

Below is a detailed contents list for this draft Taxation Ruling: Paragraph What this Ruling is about 1 Date of effect 10 Ruling and explanation 11 A. Concept of a CCA 11 Types of CCAs 15 B. Applying the arm's length principle 18 Arrangement should make business sense 22 Terms of a CCA should accord with its economic substance 32 Terms of CCA should be agreed up-front 37 Participants should have reasonable expectation of benefit 42 A participant must have an interest in the results of the CCA activity 43 A participant should have a reasonable expectation of benefit from exploiting its interest in the results of the CCA activity 51 A participant should have a reasonable expectation of being able to exploit its interest in the results of the CCA activity 55 A participant need not expect to benefit from all parts of the CCA activity 59 A participant may expect to exploit its interest in the results of the CCA activity either directly or indirectly 64 Expected benefits of non-participants must be considered 71 A participant's expectation of benefit should be within a commercially realistic timeframe 75 Sharing of contributions should be consistent with sharing of expected benefits 78 Cost contributions should be measured on an arm's length basis 81 Valuing contributions of services 94 Valuing contributions to a pure service arrangement 105 Valuing contributions of tangible property 114 Valuing contributions of intangible property 118 Granting of a non-exclusive right to use the intangibles in the CCA activity 121 Effective disposal of part interest in the intangibles in addition to use in CCA activity 123 Licensing of intangibles in addition to use in CCA activity 125 Treatment of government subsidies and incentives in valuing contributions 127 Treatment of receipts from non-participants in valuing contributions 129 Balancing payments may be required 133 CCAs for marketing activities 135 Expected benefits should be measured using reasonable estimates of revenues or cost savings from use of the results of the CCA activity 143 Use of benefit projections to share costs requires participants to share relevant data 155 Use of current data may be an alternative to projections in appropriate cases 157 Pure service arrangements 158 Benefit projections may need to be adjusted 159 Expected benefits may not accord with actual benefits 166 Entry, withdrawal and termination should be on arm's length terms 174 Buy-in and buy-out payments may be required upon entry and withdrawal 174 Payment should reflect arms's length value of interest transferred 193 Termination of a CCA 198 C. Consequences if a CCA is not arm's length 199 Arrangement does not make business sense 203 Terms do not accord with economic substance 205 Terms not agreed up-front 206 Participant has no reasonable expectation of benefit 207 Sharing of contributions inconsistent with sharing of expected benefits 209 Entry, withdrawal or termination not on arm's length terms 212 D. Documenting CCAs 214 Your comments 218 Detailed contents list 219