Coordinated Issue Paper Addresses Cost-Sharing Arrangement Buy-In Adjustments

I. Executive Summary

This coordinated issue paper (CIP) provides guidance to IRS personnel concerning methods that may be applied to evaluate the arm's length charge for pre-existing intangible property that is made available, for purposes of research, to a qualified cost sharing arrangement (a CSA). In this context, the payment in question is generally referred to as a "buy-in payment."¹ Treas. Reg. § 1.482-7(g)(2). This CIP will refer to the intangible asset made available to the CSA for research as either the "buy-in intangible" or the "platform intangible." See further discussion at section III.A.

A CSA is an arrangement by which the participants agree to share the cost of developing one or more intangibles (cost shared intangibles) that will be separately exploited by each of the participants. See Treas. Reg. § 1.482-7. By participating in a CSA, each participant obtains a separate interest in such cost shared intangible. Treas. Reg. § 1.482-7(b)(4)(iv). Consequently, each participant may separately exploit the cost shared intangible in a manner consistent with that interest, without owing additional compensation to the other participants.

The CIP addresses two major types of buy-ins. The first type, referred to as an "initial buy-in," involves a U.S. consolidated group making a self-developed buy-in intangible available to a CSA at the time the participants, the U.S. group and its controlled foreign corporation or CFC, originally enter into the arrangement. In the typical scenario, the U.S. group contemporaneously also licenses "make-sell" rights to the CFC in the current generation of the platform intangible. The second type, referred to as a "subsequent acquisition buy-in," involves a buy-in intangible that is acquired in an asset or stock acquisition by the U.S. group from an uncontrolled taxpayer and made available to an ongoing CSA.²

The best method rule in Treas. Reg. § 1.482-1(c) requires application of the transfer pricing method that, under the facts and circumstances, provides the most reliable measure of an arm's length result. The best method rule thus requires that such method be used to measure the buy-in, albeit an unspecified method. In determining whether an allocation is appropriate under section 482, IRS personnel evaluate the result achieved, rather than the specific "method or procedure" that the taxpayer used to reach that result. Treas. Reg. § 1.482-1(f)(2)(v)(A). This CIP describes certain factual settings in which approaches put forward by taxpayers may not provide a reliable measure of an arm's length result.

A. INITIAL BUY-IN: INCOME METHOD IS GENERALLY THE BEST METHOD

This CIP analyzes the factual scenarios in which an unspecified method known as the income or foregone profits method will generally constitute the most reliable method for measuring an initial buy-in (in the aggregate with the tandem license of make-sell rights).³ This method determines the value of the buy-in intangible (along with any licensed make-sell rights) as the present discounted value of the stream of projected operating profits of the CFC (or affiliated CFCs), after reduction by routine returns and projected cost sharing payments for the CFC's share of the R&D projected under the CSA. This method reliably takes into account the risks that the CFC assumes when it commits to fund a portion of research activities by means of cost sharing payments.

Where the CFC participant (or affiliated CFCs) has foreign operations, and owns pre-existing marketing, manufacturing, or other operating intangibles, the analysis must also take into account any contributions by those foreign operating intangibles to the total operating profits from exploitation of the cost shared intangibles. Consequently, this CIP describes several refinements to the income method that may be utilized to back out contributions by such preexisting operating intangibles to the CFC's (or affiliated CFCs') operating profit.

This CIP also includes exhibits that illustrate the application of the income method in certain typically encountered situations. The exhibits, being illustrative, are not intended to be used as templates or to replace the judgment necessary to the appropriate application of any transfer pricing method, including the income method. The details of proper application of the method, for example the discount rate, the projections, and terminal value assumptions, are all subject to variation as appropriate in light of the specific factual circumstances of a case, in the exercise of good judgment by the analyst.

B. SUBSEQUENT ACQUISITION BUY-IN: ACQUISITION PRICE IS GENERALLY THE BEST METHOD

This CIP analyzes the factual scenarios in which an unspecified method known as the acquisition price method will generally constitute the most reliable method for measuring a subsequent acquisition buy-in. This method determines the value of the buy-in intangible by reference to the acquisition price of a contemporaneous acquisition of that intangible in an asset or stock acquisition from an uncontrolled party. As with the income method, the exhibits, being illustrative, are not intended to be used as templates or to replace the judgment necessary to the appropriate application of any transfer pricing method, including the acquisition price method.

II. Facts -- Typical Buy-In Scenarios

A. TYPICAL INITIAL BUY-IN SCENARIO

The typical initial buy-in fact pattern involves an established U.S. group with significant self-developed intangibles. The U.S. group has a R&D capability consisting of fixed assets, an experienced workforce, and a record of successfully developing products or technologies. The U.S. group and its CFC enter into a CSA to further develop the U.S. group's platform intangibles. Contemporaneously with entering into the CSA, the U.S. group often licenses to the CFC make-sell rights in the current generation of a product incorporating the platform technologies. In combination, the CSA and the license convey to the CFC a significant portion of the intangible rights of the U.S. group. Under the CSA, the CFC bears a proportionate share of the costs of the ongoing intangible development, based on its reasonably anticipated benefits from the resulting cost shared intangibles.

Accordingly, the CFC shares with the U.S. group the risk of developing intangible property on a going-forward basis. In other respects, the CFC participant in a CSA generally performs limited functions and has limited capabilities. That is, subject to the exceptions discussed below, the CFC typically does not own valuable intangible assets that are made available to the CSA, or that are otherwise exploited in conjunction with the cost shared intangibles. Similarly, the CFC typically does not perform R&D in its own right. As a consequence, the

U.S. group generally performs all R&D under the CSA. Often, the CFC is a special purpose entity formed for entering into the CSA, acquiring rights in resulting cost shared intangibles, and on-licensing or otherwise enabling those rights to be used by other foreign affiliates (these may be CFCs, disregarded entities, or commissionaires). In such cases, the CFC's assets at the outset of cost sharing consist of only the cash required to pay the CSA buy-in and to fund its proportional share of the intangible development costs under the CSA.

In many cases, the initial buy-in under a CSA occurs at a time when the CFC (or its affiliated CFCs) and the foreign operations are in startup or early stage operations. IRS personnel may also encounter situations in which the CFC (or affiliated CFCs) has fully developed operations and owns pre-existing marketing, manufacturing, or other intangibles related to such operations. This CIP will consider the base case (described above) as well as certain more complex scenarios that may call for refinements or adjustments in an otherwise applicable method to provide the most reliable measure of an arm's length result.

B. TYPICAL SUBSEQUENT ACQUISITION BUY-IN SCENARIO

The typical subsequent acquisition buy-in fact pattern involves the acquisition of the buy-in intangible in an asset or stock acquisition from a third party. This scenario is especially common in the high technology sector, where one business often acquires another in order to obtain intangible "building blocks" that the target possesses and the acquirer believes will contribute to its own technology.

III. Taxpayer Methods and Positions

This Section III addresses certain arguments commonly advanced by taxpayers in support of the amounts reported for the cost sharing buy-in and their shortcomings in the factual scenarios described in this CIP.

A. IDENTIFICATION AND CATEGORIZATION OF INTANGIBLES

Taxpayers may argue that the only buy-in intangible(s) made available to a CSA are those the taxpayer expressly designated. Furthermore, taxpayers may take the position that the determination of what intangibles were, and were not, so made available may be made in light of hindsight, e.g., concerning which intangibles panned out, and which did not, in the R&D pursuant to the CSA.

Under the regulations, a taxpayer's actual transactions, contractual terms, and allocations of risk will be respected to the extent consistent with economic substance.⁴ However, to the extent the express terms are inconsistent with the economic substance, the IRS may impute terms that are consistent with the economic substance. For purposes of such imputation, greatest weight will be given to the actual conduct of the parties and their respective legal rights.

Thus, due regard will be given to taxpayer's express designation of the buy-in intangibles. However, such express designation must be consistent with the economic substance of the arrangement. If it is not, the IRS may supplement or otherwise identify the buy-in intangibles consistent with the economic substance.

The following are guidelines for evaluating the economic substance of the taxpayer's designation of the buy-in intangibles, or, where necessary, for imputing the identification of such intangibles. The fundamental economic facts of the typical CSA buy-in transaction are that the controlled participants are committing to long-term R&D based on the buy-in intangible with a view to gaining separate ownership of intangibles based on the buy-in intangible. Accordingly, to be consistent with this economic substance, identification of the buy-in intangibles to be valued for purposes of determining the buy-in must be accomplished upfront -- at the time the CSA is entered into in the case of an initial buy-in and at the time of the subsequent acquisition in the case of a subsequent acquisition buy-in.⁵ The identification must be made, however, on the basis of reasonable expectations at such time. The test is whether at the time of the buy-in it is reasonably anticipated that an intangible will benefit the R&D under the CSA. This identification cannot be revisited based on hindsight in light of subsequent events not reasonably anticipated at such time. Thus, for example, one cannot contend that an intangible, initially reasonably anticipated to benefit the R&D under the CSA, is not subject to a buy-in (or is subject only to a lesser-valued buy-in), because in light of events not reasonably anticipated it turns out not to have advanced the R&D actually performed.

This CIP refers to intangibles other than the buy-in intangibles, or the cost shared intangibles resulting from the R&D pursuant to the CSA, as "operating intangibles." Additional intangibles that are reasonably expected to contribute to the value of the operations of the U.S. group are referred to as "U.S. operating intangibles," and those that are reasonably expected to contribute the value of the operations of the CFC(s) are referred to as "foreign operating intangibles." If the categorization of intangibles as operating intangibles, as distinct from buy-in intangibles is relevant it must be made upfront at the time of the buy-in, in light of what is reasonably anticipated at such time.

The categorization in a particular case of intangibles as "buy-in intangibles," or "U.S (or foreign) operating intangibles," is based on the facts and circumstances. Operating intangibles may, depending on the facts and circumstances, consist of marketing intangibles (including goodwill) or manufacturing intangibles (including current make-sell rights and going concern value) or both. Types of intangibles that are operating intangibles for one CSA, may be buy-in intangibles for another. For example, in a CSA to develop technology, any independent marketing intangibles (i.e., other than to the extent merely by-products of the technology itself )⁶ would be operating intangibles. As such, they would not be subject to a buy-in determined under Treas. Reg. § 1.482-7.⁷ By contrast, in a CSA to develop marketing intangibles benefiting both the U.S. group and CFC (s), any preexisting marketing intangibles serving as the platform for development would be buy-in intangibles subject to a buy-in determined under Treas. Reg. § 1.482-7.

As will be explained, in some cases it will be appropriate to evaluate the buy-in under Treas. Reg. § 1.482-7 in the aggregate, together with the arm's length compensation under Treas. Reg. § 1.482-4 for the transfer or license to the CFC(s) of current generation make-sell rights, or other foreign operating intangibles owned by the U.S. group. In such cases, the categorization of the aggregated intangibles (e.g., make-sell rights vs. buy-in intangible) does not affect the analysis. For purposes of clarity, however, IRS personnel are advised to explain and document their conclusions with respect to the ownership of operating intangibles and the scope of the CSA (i.e., including whether operating intangibles have been aggregated for purposes of evaluating the buy-in).

B. SCOPE OF BUY-IN INTANGIBLE RIGHTS -- THE "SALE" VS. "LICENSE" ISSUE; USEFUL LIFE ASSUMPTIONS

The value of a transfer of intangible rights varies in direct proportion to the nature and overall scope of the rights that are transferred. In some cases, taxpayers may seek to limit the apparent scope of the transfer of intangibles pursuant to the CSA, in an effort to reduce the buy-in payment.

In this regard, taxpayers may point to language in the regulation indicating that "[t]he buy-in payment . . . is the arm's length charge for the use of the intangible under the rules of §§ 1.482-1 and 1.482-4 through 1.482-6." Treas. Reg. § 1.482-7(g)(2) (emphasis added). They infer, for example, that the term "use" implies that the buy-in must be analyzed as a "license," not a "sale," and argue the Second Circuit's decision in Nestle Holdings, Inc. v. Commissioner, 152 F.3d 83, 87-88 (2d Cir. 1998), implies that the rights received in a "license" necessarily are less valuable than those received from a "sale." They then deduce that buy-in payments should be structured similarly to, and measured by comparison to, licenses for "use," often referred to as "make-sell" licenses. Thus, they assert that buy-in payments should consist of rapidly declining royalties over the period in which the current make-sell rights obsolesce. The value of all residual rights other than those make-sell/"use" rights are then said to fall outside the scope of the buy-in.

The foregoing argument is unpersuasive. First, the single instance of the word "use" is taken out of context. Viewed in context, the term does not suggest that the rights subject to the buy-in are limited to make-sell/"use" rights. The preceding sentence of the regulation clearly indicates the referent of "use" to be the utilization of "pre-existing intangible property . . . for purposes of research in the intangible development area." The preceding paragraph of the regulations provides that the participant making pre-existing intangible property available to the CSA "will be treated as having transferred interests in such property to the other controlled participants." Treas. Reg. § 1.482-7(g)(1). The CSA participants become coowners of "any intangible property that is developed as a result of the research and development undertaken under the cost sharing arrangement (intangible development area)." Treas. Reg. §§ 1.482-7(a)(1) (end of 1st sentence -- "interests in the intangibles assigned to them under the arrangement"); 1.482-7(b)(4)(iv) ("each participant's interest in any covered intangibles . . . any intangible property that is developed as a result of the research and development undertaken under the cost sharing arrangement (intangible development area)"). Accordingly, the rights for which the buy-in payments are due consist of the rights to use the pre-existing intangible property for purposes of research and development so as to acquire co-ownership of resulting intangible property.⁸ These rights are obviously and materially distinguishable from make-sell rights in the platform intangible. Moreover, in the typical scenario described above, the U.S. group already licenses the make-sell rights to the CFC, as well as making the platform intangible available for purposes of the R&D under the CSA.

Second, properly understood, the Nestle decision is entirely consistent with the above textual construction of the regulations. The issue in Nestle was the value of a trademark that had been sold to Nestle. The IRS expert had valued the trademark on the basis of the stream of royalties which could be generated by licensing the mark. The Court found this "relief-from-royalty" model to be inadequate in that it failed "to capture the value of all of the rights of ownership, such as the power to determine when and where a mark may be used, or moving a mark into or out of product lines." By analogous reasoning, determining the value of the buy-in on the basis of make-sell royalties⁹ fails to capture the value of the rights the CFC receives to co-own intangibles developed on the basis of the platform intangible.

Third, valuation of an intangible transfer depends primarily on the nature and scope of the rights actually transferred, in light of economic substance, and application of the most reliable method under the best method rule. For purposes of this analysis, the label that is attached to the transfers is insignificant. Thus, the critical question is the arm's length price for the buy-in intangibles made available to the CFC, which is entirely independent of whether the transaction is characterized as a "license" or a "sale." Depending on the scope of the "interest in" -- i.e., rights in -- an intangible that is transferred, a given transfer of interest may be either a "sale" or a "license." Technically a "sale" requires the transfer of an interest consisting in all substantial rights in an intangible, while a "license" is a transfer of an interest consisting in less than all substantial rights in an intangible.¹⁰ The character of a transfer is unaffected by the form of payment (e.g., a "sale" character is possible even though a contingent form of payment is adopted and a "license" character is possible even if a lump sum form of payment is adopted).¹¹ The transfer pricing valuation turns on the nature and scope of the rights transferred, not on the technical characterization of the transaction as a "sale" or a "license." The rights/interest in the buy-in intangible that is the subject of the buy-in, as explained, are the rights to conduct R&D based on the platform and to acquire co-ownership of any intangibles resulting from such R&D. While standing alone these rights may not represent all substantial rights in the buy-in intangible and so their transfer may technically constitute a "license" rather than a "sale," that does not have any inherent implication for the valuation. Indeed, combined with a transfer of the make-sell rights in the buy-in intangible that is part of the typical buy-in scenario, the transfer of the combined rights is economically tantamount to a "sale."¹²

Fourth, in ascertaining the actual rights conveyed in the buy-in transaction, the specific terms of the buy-in provisions of the CSA are critical, but such terms are given effect only to the extent they are consistent with the economic substance of the underlying cost sharing transaction.¹³ Treas. Reg. § 1.482-1(d)(3)(ii)(B) and (iii)(B). The fundamental economic facts of the typical CSA buy-in transaction are that the controlled participants commit to long-term R&D based on the buy-in intangible, with a view to gaining separate ownership of intangibles based on the buy-in intangible. Any contractual limitations on the CFC participant's rights must be evaluated in that light. For example, the taxpayer may claim that it is making the buy-in intangible available only on an at-will or other short-term basis. Alternatively, the taxpayer may say that the license was terminated or not renewed, because the buy-in intangible proved not to be valuable or "was never used" -- a contention made in connection with subsequent acquisition buy-ins. These positions belie the mutual commitment of the related parties to engage in a long-term R&D effort in which both intend to make their risky contributions available for the duration, without regard to success or failure of particular elements of the project. Moreover, it is exceedingly difficult, if not impossible, to trace, after-the-fact, the degree to which outcomes are attributable to one or another contribution. Such limitations, and others like them, fail the economic substance test and should be disregarded.¹⁴

Fifth, the useful life of make-sell rights in the pre-existing intangible is irrelevant to evaluation of the buy-in intangibles. The object of a CSA is typically to develop a next generation intangible based on the platform intangible. Therefore, the current generation make-sell rights may and often do become less valuable as the fruits of the research come on line. That, however, does not imply that the benefits from making the platform intangible available "for purposes of research in the intangible development area" are limited to the period preceding the expected decline in use of the current generation make-sell rights. On the contrary, the benefits of the R&D rights conveyed in the platform intangible typically are expected to span the lives of the intangibles developed as the result of the R&D.

This is not to suggest, however, that buy-in payments necessarily must continue indefinitely. The form and duration of buy-in payments are independent of the scope and valuation of the rights transferred. Thus, for example, although the benefits of using the buy-in intangible for purposes of R&D may be expected to continue over the entire period of exploitation of intangibles resulting from the R&D, the buy-in payment for those expected benefits may be discounted and paid in a lump sum, or in contingent amounts over a finite period as the result of royalty rates determined by standard econometric techniques.¹⁵

Another variant of the "limited rights" argument is that the CSA specifies that the CFC obtains "nonexclusive" rights in the cost shared intangibles. Such nonexclusive agreements, by their terms, might allow the U.S. group to sell into the CFC's territory and vice versa, or might allow the U.S. group to grant to uncontrolled parties the right to sell into the CFC's territory. Taxpayers may argue that such restrictions significantly reduce the value of the rights transferred and the amount of the buy-in. Again, it is incumbent upon the IRS to evaluate these and other nominal restrictions in light of the economic reality of the parties' activities under the cost sharing arrangement. Rather than indicate an intent to limit the CFC's rights, the controlled parties' actual conduct may instead show that a worldwide nonexclusive arrangement, when coupled with a similar worldwide nonexclusive make-sell license, conveyed to the CFC the economic equivalent of joint ownership in the buy-in intangible, together with the U.S. group.¹⁶ In this context, the economic reality of a nominal reservation between the related parties of a right to sublicense to third party competitors in their respective markets may also be suspect, particularly if that right is never in fact exercised.

C. TRANSACTIONS CLAIMED AS "CUTS" LACK SIMILAR PROFIT POTENTIAL AND SIMILAR RISKS/ECONOMIC CONDITIONS

Taxpayers sometimes assert that their buy-in valuation is justified based on transactions that they identify as "comparable uncontrolled transactions" or "CUTs." They may argue that a make-sell license of the buy-in intangible to a third party, a co-development arrangement between third parties, or an uncontrolled license coupled with a (limited) right to produce derivative works, represent CUTs. The position taken is that these "CUTs" support quickly ramped-down buy-in royalties, or even no buy-in royalty at all, e.g., based on the observation that no buy-in is due in these uncontrolled transactions.

The regulations specify two comparability conditions that must be met if an uncontrolled transaction is to qualify as a CUT. First, the uncontrolled transaction must involve the same intangible property as the controlled transaction, or involve comparable intangible property having a similar profit potential. Second, the controlled and uncontrolled transactions must involve similar contractual terms and economic conditions. Treas. Reg. § 1.482-4(c)(2)(iii)(A) & (B).

IRS personnel should carefully scrutinize any transactions put forward as CUTs for the CSA buy-in transaction, to ascertain whether they satisfy, or fall short of the comparability criteria identified above. An uncontrolled make-sell license of the fully-developed current generation of a platform intangible is ordinarily distinguishable from the controlled transfer to a CSA participant that gives rise to the obligation to pay a buy-in. A "make-sell" license conveys the rights to exploit an existing intangible. Such a license sometimes provides the transferee a limited right to modify the existing intangible (such as to adapt it for use in a particular geographic market), which may in some cases be classified as "derivative works" under intellectual property law, but importantly it conveys no rights to perform significant further development of the intangible. In contrast, a CSA provides to each cost sharing participant full access to the buy-in intangible for purposes of research, and a CSA is specifically intended to give rise to new intangibles, in which each of the participants has separate ownership rights.

Uncontrolled co-development arrangements also generally fail to qualify as CUTs for a cost sharing arrangement and therefore cannot provide a reliable measure for the buy-in. Such co-development arrangements typically involve a materially different division of costs, risks, and benefits among the parties as compared to a CSA, and also involve significantly different intangible property under development with dissimilar profit potentials. For example, codevelopment arrangements may contemplate joint exploitation of intangibles developed under the arrangement, and may tie the division of actual results to the magnitude of each party's contributions (such as by way of preferential returns). These arrangements are not comparable to a CSA in which the participants divide contributions in accordance with the reasonably anticipated benefits from separate exploitation of the resulting intangibles. Similarly, in contrast to the typical CSA scenario, all participants in co-development arrangements typically contribute valuable intangible property to the venture. Thus, payments for transfers of intangible property in such co-development arrangements reflect offsetting of obligations that will not be present in the typical CSA. Moreover, co-development arrangements often are targeted to specifically circumscribed research, while the typical CSA often involves the entire gamut of the participants' core technologies or, in some cases, all technology related to an entire segment or product line.

D. RPSM EQUATES FUNDAMENTALLY DIFFERENT PAST AND FUTURE RISKS

Another method commonly relied upon by taxpayers to support buy-in valuations is an application of the residual profit split method (RPSM). Under this application, each year the parties first obtain a portion of the CFC's operating income based on market returns for their respective routine contributions, and then split the residual profit (i.e., after the first-step compensation of the participants' routine contributions) in proportion to: (a) the CFC participant's capitalized and amortized cost sharing payments as compared to (b) the sum of the U.S. group's capitalized and amortized past development costs of the buy-in intangible, plus the group's capitalized and amortized share of ongoing R&D costs. The buy-in royalty owed by the CFC to the U.S. group is the sum of the market return to the routine contributions (if any) the group makes to the CFC's operations, plus the group's share of the CFC's residual profit. Taxpayers claim that this application is consistent with the specified method and the example under the regulations. See generally Treas. Reg. § 1.482-6(c)(3) and Example.

Notably, under this application, the buy-in royalty that the CFC owes to the U.S. group declines significantly over time as the U.S. group's past development costs are fully amortized. The ramp-down effect may be more or less pronounced, depending upon the specific assumptions regarding the useful life of past development costs. Here, as discussed elsewhere, the distortion of the results is often exacerbated by the assumption that the useful life of the past development costs is limited to the life of the make-sell rights in the current generation of the platform intangible -- an intangible that is claimed to be displaced by the intangibles developed under the CSA.

Putting aside the doubtful reliability of such assumptions regarding useful life, however, this application of the RPSM is undermined by the fundamental incomparability of the risks associated with past development costs as compared to ongoing and future development costs. This application in effect places the risks that are assumed by the CFC in making cost sharing payments on a par with the risks assumed by the U.S. group in prior years when it developed the underlying platform intangible. Yet, there is unlikely to be a parity of risks between the funding of past R&D and the funding of current R&D.

At the time a CSA is entered into, the past development costs are already sunk, in contrast to the ongoing R&D costs. The return to past development costs may have largely been resolved and, given that development is ongoing, "worst-case scenarios" apparently avoided. In contrast, the returns to future development costs may be uncertain. There is no reason generally to believe that the (unknown) returns in the future will be the same as the (known) returns in past. Indeed, given that previous development was probably both risky and successful, if anything one might expect returns to future development costs to be lower. Stated another way, comparing past costs to ongoing/future costs is comparing apples and oranges.

The RPSM provisions in the regulations, including the example, address the determination of an appropriate make-sell royalty for a license of a fully-developed intangible between controlled parties, both of whom contribute pre-existing intangibles to the CFC's operations. Even in such a case, the regulations prefer a split of the CFC's residual profit on the basis of the relative value of the intangible property contributed by each controlled taxpayer as measured by external market benchmarks that reflect the fair market value of such intangible property. Alternatively, the regulations permit the use of capitalized and amortized costs of development of the parties' respective intangibles. Importantly, however, in such a case past development costs are compared to past development costs, i.e., apples to apples.

Thus, the specified RPSM provisions are ill-suited to measure the buy-in regarding a CSA for the development of intangibles when only one of the parties contributes a pre-existing intangible. Although both parties bear risks, the CFC bears only prospective risk, whereas the U.S. group brings both its past risks as well as its commitment to bear a share of the prospective development risks. An application of the RPSM that tries to compare these disparate risks is of doubtful reliability and should yield, in any event, under the best method rule, to a transfer pricing method, even if unspecified, that is more reliable under the circumstances.¹⁷

E. APPROPRIATE TREATMENT OF GOODWILL AND GOING CONCERN VALUE; MARKETING AND RESEARCH TEAM INTANGIBLES

Taxpayers often assert that substantial residual intangible value associated with the right to exploit foreign markets may be made available to the CFC without giving rise to any buy-in obligation to the U.S. group. Taxpayers may view this claim as consistent with the exclusion of "foreign goodwill or going concern value" in the section 367(d) regulations. See Temp. Treas. Reg. § 1.367(d)-1T(b). Such claims must be evaluated by reference to the facts and circumstances of the particular case. The value of such foreign goodwill or going concern value, to the extent it exists, does not include substantial residual intangible value associated with the right to exploit foreign markets that, instead, belongs to other identifiable intangibles.

1. "Foreign Goodwill or Going Concern Value" is Narrowly Defined

The regulations under section 367(d) do not support the notion that a substantial residual value associated with the right to exploit foreign markets must be excluded from the CSA buy-in. The regulations under that provision narrowly define foreign goodwill or going concern value as "the residual value of a business operation conducted outside of the United States after all other tangible and intangible assets have been identified and valued." Treas. Reg. § 1.367(a)-1T(d)(5)(iii). This definition has two salient features.

First, to be within the scope of this exception, there must be a business operation conducted outside the United States. Thus, no foreign goodwill or going concern value attaches to assets other than those associated with a business operation conducted outside the United States.

Second, goodwill or going concern value attaches to the entire business and not to any particular individual asset or group of assets that forms a part of the business. As a result, assets that can be separately identified, such as marketing intangibles, research workforce in place, or other operating intangibles (including for this purpose synergy effects among a group of such assets), cannot be classified as foreign goodwill or going concern value. Thus, separately identifiable operating intangibles -- that may have some resemblance to what may colloquially be called goodwill or going concern value -- do not, in any event, fall within the technical compass of "foreign goodwill or going concern" within the meaning of the regulations. Rather, the taxpayer must contemporaneously identify and carve out intangible assets (or groups of such assets) available to the CFC. Only the residual value (if any) that remains after this process is potentially classified as foreign goodwill or going concern value.

The scope of the U.S. group's pre-existing intangibles that are the subject of the buy-in depends on which intangible, or combination of intangibles, is made available to the CSA for purposes of research in the intangible development area. See section I.A.

Technological intangibles associated with one or more of the U.S. group's business lines are typically buy-in intangibles covered by the buy-in. Marketing intangibles may also be developed under some CSAs, either as the sole covered intangibles or in addition to technological covered intangibles. In such cases, pre-existing marketing intangibles may constitute buy-in intangibles covered by the buy-in.

Even in those cases in which the U.S. group's marketing intangibles do not constitute buy-in intangibles, the U.S. group may nevertheless separately license marketing intangibles it owns to the CFC, to be used by the CFC to exploit the cost shared intangibles resulting from the CSA (or to exploit make-sell rights in the current generation intangible). In that case, arm's length consideration is due for any such license and, as discussed further below, an aggregate analysis may provide the most reliable measure of the arm's length charge for the combination of: (1) the buy-in intangible; (2) the make-sell rights; and (3) the foreign marketing intangibles. See Treas. Reg. § 1.482-1(f)(2)(i) (Aggregation of transactions).

To the extent operating intangibles -- e.g., marketing intangibles, including goodwill, and manufacturing intangibles, including going concern value -- relate solely to the U.S. group's operations, they are subject neither to the buy-in nor other arm's length charge, because they are not reasonably expected to benefit the CFC's operations. On the other hand, operating intangibles such as so-called global marketing intangibles (trademarks, brands, etc.) that belong to the U.S. group but jointly benefit the U.S. group and the CFC(s), are distinguishable, and their use by the CFC (or affiliated CFCs) would require arm's length compensation.

2. Accounting Value of "Goodwill" Not Reliable Basis to Determine the Buy-In

Arguments that a portion of the U.S. group's intangibles constitute non-compensable goodwill items generally rely on a goodwill value shown in the financial statements or in appraisals conducted in connection with an acquisition. Such valuations, however, seldom attempt to break out the separate value of component intangibles entering into such "goodwill." While in such cases separate identification of the different intangibles may not be material in terms of the accounting disclosures, this plainly is material to the income tax issues.

The need to look past accounting treatment of goodwill is particularly acute in the case of subsequent acquisition buy-ins. For example, consider an acquisition whose purpose is to acquire self-developed intangibles of a target that is in startup phase or early stage operations. The self-developed intangibles are valuable in the view of the acquirer, but the R&D costs were written off, and consequently all the value may be put in the "goodwill" column. For income tax purposes, however, a more refined allocation is required to separately identify intangibles. See generally Treas. Reg. §§ 1.338-6, 1.338-7, and 1.1060-1. (See also the discussion above of the narrow definition of foreign goodwill in section III.E.1.)

A similar issue may arise where a target possesses several intangibles but the acquirer is only interested in one or several of those intangibles and intends to abandon or otherwise not utilize the remaining intangibles. The accounting treatment may spread the acquisition value over all the intangibles, but for income tax purposes, again, a more refined allocation is required. Id.

3. Research Team Intangible Contribution is Part of Buy-In Intangible

As noted, to the extent operating intangibles -- e.g., manufacturing intangibles, including going concern value -- relate solely to the U.S. group's operations, they are subject neither to the buy-in nor other arm's length charge, because they are not reasonably expected to benefit the CFC's operations.

Taxpayers may attribute a significant portion of the value derived by a CFC from a CSA to the U.S. group's research "workforce in place," which they assert is not separate from U.S. going concern value¹⁸ and so is not compensable under section 482. They also argue: (1) that workforce in place is not an item listed in section 936(h)(3)(B) or Treas. Reg. § 1.482-4(b) having substantial value independent of the services of any individual, and so is not an intangible for purposes of section 482; and (2) that workforce in place is not an intangible interest for purposes of Treas. Reg. § 1.482-7(a)(2), because it is not "commercially transferable." These arguments are addressed below.

Workforce in place must be analyzed separately from going concern value. While the Court of Appeals in Ithaca Industries affirmed the Tax Court decision, it expressly declined to adopt the Tax Court's view that assembled work force is not an asset separate and distinct from going concern value in light of Newark Morning Ledger Co. v. United States, 507 U.S. 546 (1993). The Circuit Court stated that "it is no longer appropriate to classify an intangible asset based on its resemblance to the classic conception of goodwill or going concern value." 17 F.3d at 688.¹⁹ Indeed, section 197(d)(1)(C)(i) now expressly recognizes workforce in place as a separately identifiable intangible distinct from goodwill and going concern value.

Regarding value independent of an individual's services, it may be that a given individual's employment contract or know how may lack substantial value independent of the services of such individual, as the gross value is effectively offset by the liability to pay the individual's compensation.²⁰ The U.S. group's assembled research team, however, may be expected to have a substantial value independent of the services of any individual member of the team attributable to the team's collective contracts and know how, as no one or several individuals may be able to bargain compensation sufficient to eliminate a premium. Thus, the value of a research team workforce in place is either derived from the "contract" or "know how" items expressly listed in section 936(h)(3)(B) and Treas. Reg. § 1.482-4(b), or represents a "similar item" to such items.

Finally, the research team is an intangible interest for purposes of Treas. Reg. § 1.482-7(a) (2).²¹ The reference to "a commercially transferable interest" is only given as an example (". . . includes any commercially transferable interest . . ."). For the general definition of an intangible, a cross reference is made to Treas. Reg. § 1.482-4(b) just discussed above. Notably, a commercial transferability requirement, which appeared in the 1993 temporary regulations § 1.482-4T(b), was eliminated from the 1994 final regulations as "superfluous." To the extent the CFC (or affiliated CFCs) benefits from the contribution of the research team intangible, therefore, it must pay compensation to the U.S. group via a buy-in.²²

4. The Synergy Value of the Research Team Related to the Buy-In Intangible Must Be Included in the Buy-In

The value of research in process and the value of the research team conducting the research are linked. Promising research tends to heighten the expected value of the research team conducting the research. Committing a talented team to a research project tends to heighten the expected value of the research.

Taxpayers who argue that the research workforce is not an intangible for purposes of the buy-in attempt to assign the value of the foregoing synergy to the excluded research workforce, and on that basis seek to minimize the buy-in.

Whether or not the research workforce is an intangible asset that must be compensated in its own right, the above synergy value of the research team related to the research in process is properly part of the buy-in. Workforce in place is valued independently from particular research activities.²³ Thus, for example, case law and the valuation literature provides for valuing workforce in place by reference to the cost of assembling a replacement workforce.²⁴

Thus, even if the routine value of the research team conducting the CSA R&D is excluded from the buy-in, the synergy value related to that R&D belongs to the buy-in intangible in process and must be compensated in the buy-in.

F. COMMENSURATE WITH INCOME -- NOT OPEN TRANSACTION WITH BENEFIT OF HINDSIGHT²⁵

The second sentence of section 482, added by the Tax Reform Act of 1986, provides: "In the case of any transfer (or license) of intangible property (within the meaning of section 936(h) (3)(B)), the income with respect to such transfer or license shall be commensurate with the income attributable to the intangible."

Many controlled parties with CSAs interpret this provision as authorizing payment of the buy-in with the benefit of hindsight. For example, they may initially set the buy-in at a lump sum of, say, $100 million. If the actual results of the cost sharing activity are favorable, the taxpayer lets stand the lump sum buy-in. However, if the actual results -- or what are represented as the actual results - - are unfavorable, these taxpayers claim the entitlement to make an affirmative adjustment to the prior buy-in so that it can be made "commensurate with the income" -- here the losses -- "attributable to the intangible."

If accepted, this approach would sanction the type of abuse the commensurate with income (CWI) principle was enacted to police. In 1986, Congress indicated a significant degree of skepticism about related-party transfers of high-profit potential intangibles for relatively insignificant lump sum or royalty consideration that effectively place all the intangible development downside risk in one controlled taxpayer and all the upside profit potential in another. See H.R. Rep. 99-426, at 424-25 (1985). See also Notice 88-123 (the White Paper), 1988-2 C.B. 458, 472-74, 477-80.

Congress envisioned the CWI principle as ensuring that controlled taxpayers would price an intangible transfer consistent with the expectation that the income with respect to that transfer be commensurate with the relative risks the parties undertake in the development and exploitation of the intangible. As regards the buy-in payment under a cost sharing arrangement, Congress emphasized that the significance of CWI lies in assuring the pricing is consistent with the expectation of appropriate compensation for the risks the contributor of the buy-in intangible undertook in its development prior to entering the arrangement:

Thus, the CWI principle mandates that the long-term commitments and risks associated with a CSA be reflected in upfront pricing that appropriately accounts for the expected benefits of those risks over time. The proper accounting for risk may take different forms and, as elsewhere, the taxpayer's choice of form will be respected, provided that it is clearly specified and conforms with economic substance.²⁶ So, for example, a taxpayer may adopt a lump sum form of buy-in payment, provided that the lump sum amount reflects a conscientious upfront valuation of the risks and expected benefits. Having made that choice, the taxpayer must stick with it and accept the actual outcome. A taxpayer may not rely on the CWI principle to attack its own pricing of an intangible transfer or its choice of payment form. By no means does the CWI principle allow the taxpayer to back out of a deal, without consideration, when the deal plays out differently than expected.

Alternatively, a taxpayer may adopt a contingent form of buy-in payment, provided that such choice is clearly specified, and provided further that the expected present value of future payments (appropriately discounted to reflect risk) is consistent with the taxpayer's conscientious upfront valuation. If appropriately structured and valued, a contingent payment may reflect the profit potential and risks, while also permitting actual buy-in payments to vary with later outcomes. In such arrangements, less income will properly result if the outcomes are less successful than reasonably anticipated, and more income will result if the outcomes are more successful.

Accordingly, the taxpayer has flexibility in the first instance to structure its buy-in payment in a manner consistent with a conscientious upfront valuation. However, the taxpayer's choice must be adequately documented. Importantly, taxpayers cannot make so-called affirmative CWI adjustments to undo their upfront deal in light of hindsight. See also section VI (Form of Buy-in Payment).

By contrast, if the taxpayer's pricing fails to reflect a conscientious upfront valuation effort, the IRS has the ability to make periodic adjustments to bring the pricing in line with the CWI principle.²⁷ Although the IRS necessarily examines the taxpayer's transaction after-the-fact, it should exercise its periodic adjustment authority consistent with what would have been a conscientious upfront valuation -- had the taxpayer in fact made one. Thus, the IRS should not make a periodic adjustment to a buy-in on the basis of outcomes that could not be reasonably anticipated at the time the CSA was entered into.²⁸ The regulations reflect the intent that the IRS exercise restraint in making periodic adjustments based only on the upfront reasonable expectations and not based on subsequent events which could not be reasonably anticipated.²⁹

On the other hand, in making appropriate periodic adjustments, the IRS may adopt a form of payment different than that the taxpayer adopted for its buy-in payment -- such as a contingent royalty rather than a lump sum form. The legislative history plainly intends that the IRS may make periodic adjustments as a contingent royalty when the taxpayer provides for only an inadequate lump sum. See H.R. Rep. 99-426 at 424-25 (1985). The regulations similarly authorize the IRS to make periodic adjustments in a different form than that adopted by the taxpayer, e.g., an equivalent royalty in lieu of a lump sum. See Treas. Reg. § 1.482-4 (f)(5) (lump sum payments adjusted on the basis of equivalent royalty).

Taxpayers may cite Treas. Reg. § 1.482-1(a)(3) in support of the claimed entitlement to make affirmative CWI adjustments. That regulation does authorize taxpayer use of section 482 to report results on a timely return based on prices different from those actually charged, but only "[i]f necessary to reflect an arm's length result." So, for example, suppose that a CSA unconditionally provides for a lump sum buy-in payment in Year 1, but the Year 3 outcome proves to be less profitable than expected. The taxpayer might assert that a taxpayer-initiated section 482 adjustment may be made pursuant to the cited regulation on a timely Year 3 income tax return. This argument is misleading and mistaken. As noted, the regulation imposes as an express condition to a taxpayer-initiated adjustment that such adjustment must be "necessary to reflect an arm's length result." As explained, a hindsight adjustment that permits the taxpayer to walk away, without consideration, from the deal it struck for itself cannot qualify as "necessary to reflect an arm's length result." Such an adjustment would in fact depart from an arm's length result.

Treas. Reg. § 1.482-1(g)(4) (Setoffs) also fails to provide support for taxpayer affirmative CWI adjustments. That regulation addresses a situation in which the IRS upon examination has made an original allocation under section 482 regarding a given non-arm's length transaction (the primary transaction). The regulation provides substantive and procedural rules under which a taxpayer's claim for a "setoff" will be allowed by the IRS against its original section 482 allocation on account of a section 482 allocation raised by the taxpayer regarding another non-arm's length transaction going in the opposite direction (the setoff transaction). The substantive requirements are that the setoff transaction must be "between the same controlled taxpayers" and "in the same taxable year" as the primary transaction. Furthermore, the taxpayer must establish "that the transaction that is the basis of the setoff was not at arm's length" and must establish "the amount of the appropriate arm's length charge" for such setoff transaction. The procedural requirements relate to documentation and timely notification to the IRS of the basis for any claimed setoff. Nothing in this provision suggests that the taxpayer may alter the terms or prices of the primary transaction itself (i.e., the subject of the IRS original section 482 allocation) based on hindsight in contravention of the arm's length standard.

G. ATTRIBUTING A SUBSTANTIAL POSITIVE NET PRESENT VALUE TO THE CFC'S PROJECTED R&D EXPENDITURES

Certain valuation methods employed by taxpayers erroneously attribute a substantial positive net present value (NPV) to the CFC's projected R&D expenditures. These methods generally reflect the notion that, because the buy in intangible is anticipated to obsolesce, over time the cost shared intangibles resulting from the R&D should account for an increasing share of the residual profits of future operations.³⁰ However, as explained below, such methods violate established principles of transfer pricing and corporate finance.

In the typical early stage scenario, the CFC generally has little functionality and contributes nothing more than its commitment to fund its pro rata share of the R&D. The applicable discount rate must reflect the risky return expected for the R&D project, both in absolute terms and in relation to competing investments available elsewhere. At arm's length, the U.S. group would generally refuse to share the R&D project with the CFC on the basis of a significant premium above such risky return, and would instead seek to retain that potential future premium return for its own benefit, unless the CFC provided compensation sufficient to make the NPV of the CFC's anticipated expenditures and returns approximately equal to zero.³¹ Indeed, given the depth of the capital markets and competition, at arm's length if the CFC sought a significant premium above the applicable discount attributable to the risk of the R&D project, it would be expected that competing capital providers would undercut the premium sought by the CFC, with a tendency to approach a zero NPV.

Moreover, the anticipated NPVs, to both USP and the CFC, of the R&D project in the CFC's territory must sum up to a constant, which is the anticipated positive NPV of the project as a whole in the CFC's territory. Therefore, the notion that USP would accept a buy-in with an expected net present value significantly less than the expected NPV of the project as a whole in the CFC's territory, would violate an established principle of corporate finance. USP would in that case be investing an asset (the buy-in intangible as it pertains to the CFC's territory) for an anticipated return with a smaller expected NPV than the value of the asset invested.³² USP's investment thus would have a negative anticipated NPV. In general, however, a corporation will undertake a project only if its NPV is greater than zero, i.e., projected returns from the project are in excess of its cost, when both are appropriately discounted for risk. See Franklin Allen, Stewart C. Myers, Richard A. Brealey, Principles of Corporate Finance, 17-31, 222-231 (McGraw-Hill, 8th Ed., 2006).

In the mature foreign operations scenario, where the CFC or its foreign affiliates possess significant operating intangibles, one would expect that the CFC would share in the residual profits of future operations, based on those intangibles as well as the buy-in intangible. This CIP provides for sharing in residual profits in such case.³³

IV. Initial Buy-In: Best Method Analysis -- Income Method is Generally the Best Method

A. AN UNSPECIFIED METHOD MAY BE THE BEST METHOD

The arm's length result of a controlled transaction, and any IRS adjustment in the event a taxpayer failed to report an arm's length result, must be determined under the method that, taking into account the specific facts and circumstances, provides the most reliable measure of an arm's length result. Treas. Reg. § 1.482-1(c). There is no strict priority of methods, and no method -- whether specified or unspecified -- will invariably be considered more reliable than others. An arm's length result may be determined under any method -- specified or unspecified -- without establishing the inapplicability of another method, but if another method subsequently is shown to produce a more reliable measure of an arm's length result, such other method -- whether specified or unspecified -- must be used.

Accordingly, the best method rule contemplates the possibility that an unspecified method may provide the most reliable measure of an arm's length result. This CIP concludes that unspecified methods provide the most reliable measure of an arm's length result, both for the initial buy-in and the subsequent acquisition buy-in. In the case of the typical initial buy-in scenario, this CIP concludes that the income method is generally the best method. This CIP also describes the circumstances under which a market capitalization method either may be the best method standing alone, or may be reliable as an additional corroborating method.³⁴

B. NONE OF THE SPECIFIED METHODS PROVIDE THE BEST METHOD FOR DETERMINING THE INITIAL BUY-IN

Where the CFC participant is in startup or early-stage operations, the CFC, as the simpler of the controlled taxpayers, will generally be the controlled party whose income is evaluated under one or more transfer pricing methods. The primary analytical challenge in these cases is that the U.S. group and the CFC are jointly bearing the risks of developing the cost shared intangible under the CSA, while also performing other functions that must be evaluated under the arm's length standard.

Under these circumstances, none of the specified methods in Treas. Reg. §§ 1.482-3 though 1.482-6 provides a reliable means of taking into account the shared risk of the intangible development undertaken by the U.S. group pursuant to the CSA, much less the risks that the U.S. group bore when it performed prior-year R&D to develop the platform intangible. Further, the specified methods in Treas. Reg. § 1.482-4, which apply to transfers of intangible property, are not suited for the joint development and subsequent separate exploitation of in-process intangibles -- the specific scenario at issue under a CSA. Given that the contemporaneous transfer pricing documentation prepared by taxpayers frequently applies one of these specified methods, this paper considers the analytical shortcomings of each of these methods, as they are most commonly applied by taxpayers.

1. Transactions Claimed as CUTs Are Not Comparable³⁵

Taxpayers sometimes assert that "make-sell" licenses to third parties of the current generation application of a platform intangible, afford a basis to use the "comparable uncontrolled transaction" or CUT method to determine the initial buy-in. Moreover, a "useful life" analysis based on product life is frequently also employed that tends to limit the period during which buy-in compensation is paid by the CFC. Sometimes taxpayers argue that codevelopment arrangements between third parties represent CUTs. These and similar applications of the CUT method raise serious reliability concerns.

Make-sell licenses involving the exploitation of fully-developed intangibles are ordinarily distinguishable from a controlled transfer for purposes of developing a successor generation of the platform intangible that gives rise to the obligation to pay an initial buy-in. A make-sell license conveys the rights to exploit an existing intangible. Although such a license sometimes provides the transferee a limited right to modify the existing intangible (such as to adapt it for use in a particular geographic market), it conveys no rights to perform significant further development of the intangible. In contrast, a CSA contemplates major research and development rights involving the platform intangible. A CSA is specifically intended to give rise to new intangibles in which each of the participants will own separate rights of exploitation.

Uncontrolled co-development arrangements also fail to qualify as CUTs for a CSA and, therefore, cannot be said to provide a reliable measure for the buy-in. Such co-development arrangements typically involve materially different division of costs, risks, and benefits among the parties as compared to a CSA, and involve significantly different intangible property under development with dissimilar profit potentials. For example, co-development arrangements may contemplate joint exploitation of intangibles developed under the arrangement, and may tie the division of actual results to the magnitude of each party's contributions (such as by way of preferential returns). These arrangements are not comparable to a CSA in which the participants divide contributions in accordance with the reasonably anticipated benefits from separate exploitation of the resulting intangibles. Similarly, in contrast to the typical CSA scenario, all participants in co-development arrangements typically contribute valuable intangible property to the venture. Thus, payments for transfers of intangible property in such co-development arrangements reflect offsetting of obligations that will not be present in the typical CSA. Moreover, co-development arrangements often are targeted to specifically circumscribed research, while the typical CSA often involves the entire gamut of the participants' core technologies or, in some cases, all technology related to an entire segment or product line.

2. CPM Data Publicly Available Do Not Reflect Similar Risks

The comparable profits method or CPM, which generally evaluates the CFC as the tested party, likewise does not provide an appropriate means for evaluating an initial buy-in. It is not possible, especially from the public data usually used in CPM analyses, to identify uncontrolled entities that perform business activities involving risks similar to those borne by the CFC when it funds its proportionate share of R&D costs under a CSA.³⁶ Neither is the U.S. group a good candidate to be the tested party under a CPM analysis, because the U.S. group performs complex functions and has unique intangibles (including but not limited to an assembled R&D workforce).

3. RPSM that in Effect Equates Incomparable Past and Future Risks³⁷

Frequently, taxpayers use the residual profit split method (RPSM) as a means of evaluating the CSA buy-in. Under many such applications, each year the parties first obtain a portion of the CFC's operating income based on market returns for their respective routine contributions, and then split the residual profit (i.e., after the first-step compensation of the participants' routine contributions) in proportion to: (a) the CFC participant's capitalized and amortized cost sharing payments as compared to (b) the sum of the U.S. group's capitalized and amortized past development costs of the buy-in intangible, plus the group's capitalized and amortized share of ongoing R&D costs. In some cases, taxpayers apply the RPSM in a way that in effect places the risks assumed by the CFC in making cost sharing payments on a par with the risks assumed by the U.S. group in prior years, when it developed the underlying platform intangible. Such an application of the RPSM is unreliable because, among other things, it incorrectly assumes that a parity of risks exists between the funding of past R&D and the funding of current R&D.³⁸

C. THE INCOME METHOD GENERALLY PROVIDES THE BEST METHOD FOR DETERMINING THE INITIAL BUY-IN

This CIP concludes that in the typical initial buy-in scenario an unspecified method known as the income or foregone profits method will generally constitute the most reliable method for measuring an initial buy-in (in the aggregate with the tandem license of make-sell rights). This method determines the value of the buy-in intangible (along with the licensed make-sell rights) as the present discounted value of the stream of projected operating profits of the CFC (or affiliated CFCs), after reduction for routine returns and projected cost sharing payments for the CFC's share of the R&D projected under the CSA. This method has a number of significant advantages over the specified methods described above. Importantly, it reliably takes into account the risks that the CFC assumes when it commits to fund a portion of research activities by providing an expected return to cost sharing payments. Furthermore, since the income method looks only to the projected operating profits of the CFC, it automatically excludes the effect of any operating intangibles that benefit only the U.S. Group (including U.S. goodwill and going concern value).³⁹ In addition, the method may simultaneously determine, in the aggregate with the buy-in, the arm's length charge for other pre-existing operating intangibles owned by the U.S. group that are used by the CFC(s) in the foreign operations. In the typical scenario, the U.S. group transfers make-sell rights, and sometimes marketing intangibles, it owns to the CFC along with making available the buy-in intangible at the inception of a CSA. This method determines the arm's length charge for all such intangibles, on an aggregate basis. See Treas. Reg. § 1.482-1(f)(2) (i).

Various elements of the income method are specifically examined further below. Exhibits A.1 through A.6 illustrate the application of the income method.⁴⁰ The exhibits, being illustrative, are not intended to be used as templates or to replace the judgment necessary to the appropriate application of any transfer pricing method, including the income method. The details of proper application of the method, for example the discount rate, the projections, and terminal value assumptions, are all subject to variation as appropriate in light of the specific factual circumstances of a case, in the exercise of good judgment by the analyst.

1. Appropriate Discount Rate for Present Valuing Income and Expense Flows

Under the income method, the projected income and expense flows in connection with the CFC's operations in exploiting the intangibles resulting from R&D under the CSA (as well as in exploiting the licensed make-sell rights in the existing generation of the platform intangible), and the CFC's projected stream of cost sharing payments, must be discounted to present value. The discount rate (or range of discount rates, see below) employed needs to reflect the risks particular to the CFC's business and the intangible development area under the CSA.

Where the scope of the CSA is relatively broad and encompasses most or all lines of business and intangibles of the U.S. group, the weighted average cost of capital (WACC) of the U.S. parent will provide a reasonable starting-point for evaluation of the discount rate. In contrast, where the CSA deals with a narrowly focused segment of the group's business and intangibles, the discount rate appropriate to the CFC's operations and the intangible development area likely will not reflect the same mix of risks as is reflected in the WACC; in such case, accordingly, the WACC will be less reliable as an estimate of the appropriate discount rate (or range).

2. Projections: Begin with Taxpayer's; Extrapolation from Actual Results

It is often the case that the taxpayer, contemporaneously with entering into a CSA, prepared projections of intangible development costs, other operating costs, and revenues anticipated from exploiting intangibles resulting from the R&D (as well as from exploiting licensed make-sell rights in the existing generation intangible). The taxpayer is generally in the best position to make projections of likely results from cost sharing. IRS personnel should therefore use taxpayer generated projections to the extent possible, and should be reluctant to substitute their judgment in place of a reasonable application of judgment by the taxpayer. The use of taxpayer generated projections in applying the income method to calculate an initial buy-in is illustrated in Exhibit A.1.⁴¹

In some cases, IRS personnel may need to develop projections or may need to supplement or adjust projections prepared by the taxpayer (e.g., taxpayer projections that covered only a limited number of years). Actual experience under the CSA may provide evidence concerning projections that could have been made for those past years, and may indicate appropriate projections for taxable years that are still in the future as of the time of the IRS examination.⁴² If a CSA has been in place for several years at the time of an examination, several years of actual results may be available. In such cases, the income method could be applied by substituting actual experience for projections of revenues and expenses in the past years of the CSA that were not generated by the taxpayer at the CSA's inception. For periods for which actual results are not yet available, extrapolations could be performed. For example, reasonable growth factors could be applied to historical revenues and net margin data. Exhibit A.2 illustrates such an extrapolation under the income method.

3. Arm's Length Range of Initial Buy-In Results

While the regulations recognize the possibility of a single arm's length result, the norm is that application of a method may produce a number of results from which an arm's length range of reliable results may be derived. The IRS will not adjust a taxpayer's income if it falls within the arm's length range established by the IRS based on application of the best method. Treas. Reg. § 1.482-1(e)(1).

An arm's length range of buy-in results may be derived from application of the income method by varying the factors going into the determination within arm's length tolerances. Factors susceptible to such treatment include the routine returns to functions, the appropriate discount rate applied to the income and expense flows, and the projected terminal values. Exhibit A.5 illustrates the derivation of an arm's length range of initial buy-in amounts though variation in the discount rates and assumptions as to terminal values.⁴³

D. INITIAL BUY-IN FOR CFC HAVING MATURE FOREIGN OPERATIONS

In some cases, the CFC (or affiliated CFCs) has fully-developed foreign operations at the time it enters into the CSA. There may be pre-existing operating intangibles -- marketing, manufacturing, or other intangibles -- that are associated with those operations and that will contribute to the profit from exploiting the intangibles resulting from the R&D (or from exploiting the make-sell rights in the current generation intangible). In such cases, it is important to determine who -- the CFC or the U.S. group -- owns such pre-existing foreign operating intangibles.⁴⁴

If it is determined that the U.S. group owns such foreign operating intangibles, then the analysis under the income method is similar to that already described above with respect to transfers of make-sell rights at the inception of the CSA. The analysis will evaluate in the aggregate the compensation owing to the U.S. group for the combination of the initial buy-in, any license of make-sell rights in the current generation of the buy-in intangible, and the license of the foreign operating intangibles.

If, on the other hand, the CFC is determined to own the pre-existing foreign operating intangibles, the analysis under the income method must be modified to take into account the contributions that those foreign operating intangibles make to the total operating profits of the CFC. Application of the income method without such an adjustment yields an amount that is attributable in part to the buy-in intangible and in part to the CFC's foreign operating intangibles. Accordingly, the income method should be supplemented by an additional step that distinguishes the contribution of the pre-existing intangibles to the residual profit. This step will typically use an approach similar to that used to divide nonroutine profits under step 2 of the residual profit split method. See Treas. Reg. § 1.482-6(c)(3)(i)(B) (Allocate residual profit). The balance constitutes the payment for the buy-in intangible.

Exhibit A.6 illustrates application of the income method, adjusted to account for foreign operating intangibles owned by the CFC participant (or affiliated CFCs).

E. INDEPENDENT OR CORROBORATING METHOD FOR DETERMINING AN INITIAL BUY-IN: MARKET CAPITALIZATION

The market capitalization method is another unspecified method that may be used to evaluate an initial buy-in. Depending on the circumstances discussed below, a market capitalization method either may be the best method standing alone, or may be reliable as an additional corroborating method, for this purpose. Typically, the U.S. parent of the U.S. group is a publicly traded corporation. The U.S. group typically has established U.S. operations and possesses significant operating intangibles in addition to those made available to the CSA.

Under the market capitalization method, the initial buy-in is determined as the result of the following three steps. The first step is to determine the total value of all intangibles of the U.S. group as the difference between the market capitalization of the U.S. parent's stock, as grossed up by corporate liabilities, minus the value of the tangible property of the U.S. group.⁴⁵

Unlike the income method, which automatically excludes the effect of operating intangibles that benefit only the U.S. taxpayer, the market capitalization method requires an adjustment for such intangible assets. Therefore, application of the market capitalization method involves a second step, which reduces the step one intangible value by the value of the U.S. group's operating intangibles other than the rights in the buy-in intangible.⁴⁶ Where the U.S. group owns significant operating intangibles other than the buy-in intangible, the reliability of the application of this step warrants careful consideration.⁴⁷

The third step is to set the initial buy-in amount as the CFC's pro rata share of the step two buy-in intangible amount. The pro rata share for this purpose is the CFC's reasonably anticipated benefits share (RAB share).

An alternative approach is available that effectuates the second and third steps simultaneously. In this approach, the denominator of the RAB share, for purposes of determining both cost sharing and buy-in payments, includes the reasonably anticipated benefits of the U.S. operating intangibles -- say, U.S. marketing intangibles and U.S. goodwill. The buy-in payment is then set as the product of the foregoing adjusted RAB share and the step one intangible value -- i.e., without a separate second step segregation of the U.S. operating intangible value. The aggregate expected present value of the cost sharing and buy-in payments borne by the CFC can be shown to be the same under this approach as in the three-step method described previously.⁴⁸ This, of course, does not mean that the cost share and buy-in payments, considered individually, will be the same under the two approaches. Depending on the circumstances, it may, nevertheless, be as reliable or more reliable to perform an aggregate valuation of the cost sharing and buy-in payments under the alternative approach.

The above simultaneous solution may be administratively easier than independent second and third steps, since independent steps would require the segregation of the U.S. operating intangible value from the buy-in intangible value (and the value of any tandem make-sell rights). Whether one approach or the other provides a more reliable measure of an arm's length result will depend on the facts and circumstances of the specific case.

V. Subsequent Acquisition Buy-In: Best Method Analysis -- Acquisition Price Method is Generally the Best Method

A subsequent acquisition buy-in involves an intangible that is acquired by the U.S. group from an uncontrolled party and made available to the CSA contemporaneously with the acquisition.⁴⁹ The subsequent acquisition buy-in is the payment which the CFC owes the U.S. group attributable to such intangible.⁵⁰

A. AN UNSPECIFIED METHOD MAY BE THE BEST METHOD

The arm's length result of a controlled transaction, and any IRS allocation in the event that a taxpayer failed to report an arm's length result, must be determined under the method that, under the facts and circumstances, provides the most reliable measure of an arm's length result. Treas. Reg. § 1.482-1(c). There is no strict priority of methods, and no method -- whether specified or unspecified -- will invariably be considered to be more reliable than others. An arm's length result may be determined under any method -- specified or unspecified -- without establishing the inapplicability of another method, but if another method subsequently is shown to produce a more reliable measure of an arm's length result, such other method -- whether specified or unspecified -- must be used.

Accordingly, the best method rule contemplates the possibility that an unspecified method may provide the most reliable measure of an arm's length result.

B. THE ACQUISITION PRICE METHOD GENERALLY PROVIDES THE BEST METHOD FOR DETERMINING A SUBSEQUENT ACQUISITION BUY-IN

In the case of the typical subsequent acquisition buy-in scenario, this CIP analyzes the factual scenarios in which an unspecified method known as the acquisition price method will generally constitute the best method for measuring the buy-in payment. This method determines the value of the buy-in intangible by reference to the acquisition price of a contemporaneous acquisition of that intangible in an asset or stock acquisition from an uncontrolled party.⁵¹

1. Determination of Subsequent Acquisition Buy-In Under the Acquisition Price Method

Under the acquisition price method, the subsequent acquisition buy-in is determined as the result of the following three steps. The first step is to determine the total value of all the acquired intangibles. In an asset acquisition, that is the acquisition price of the target assets minus the value of the target's tangible property; and in a stock acquisition, that is the difference between the acquisition price of the target's stock, as grossed up by corporate liabilities, minus the value of the target's tangible property.⁵²

The acquisition price method, like the market capitalization method but unlike the income method, does not automatically exclude the effect of operating intangibles that benefit only the U.S. group. For this reason, a second step is required to reduce the step one intangible value by the value of acquired operating intangibles (if any) other than the rights in the buy-in intangible, to the extent such other intangibles are neither transferred nor made available to the CFC (such as U.S. marketing intangibles, including any goodwill of the target's U.S. operations, or U.S. manufacturing intangibles, including any going concern value of the target's U.S. operations).⁵³ Where the target owns significant operating intangibles other than the buy-in intangible, the reliability of the application of this step warrants careful consideration.

The third step is to set the subsequent acquisition as the CFC's pro rata share of the step two buy-in intangible amount. The pro rata share for this purpose is the CFC's reasonably anticipated benefits share.

Exhibit B.1 illustrates the application of the acquisition price method to calculate a lump sum subsequent acquisition buy-in payment. As with the income method, the exhibits, being illustrative, are not intended to be used as templates or to replace the judgment necessary to the appropriate application of any transfer pricing method, including the acquisition price method.

2. Arm's Length Range of Subsequent Acquisition Buy-In Results

Exhibit B.4 illustrates the derivation of an arm's length range of subsequent acquisition buy-in amounts through variation in the values assigned to assets other than the buy-in intangible and discount rates.⁵⁴

3. Adjustments to Reflect Timing Differences on Subsequent Acquisition Buy-In Payments

The U.S. group in some cases may experience a distortive mismatch, such as where an immediate inclusion of income results from lump-sum payment of a subsequent acquisition buy-in, while amortization of the basis in the subsequent acquisition target intangible extends over a term of years. IRS personnel should be aware that timing mismatches of this type may work a significant hardship on taxpayers. IRS personnel may consider permitting the taxpayer appropriate relief in these cases, such as by offsetting an appropriate portion of the acquisition basis in the subsequent acquisition target intangible against the inclusion on account of the buy-in, as a means of clearly reflecting income. See Treas. Reg. § 1.482-1(a) (2). Exhibit B.5 illustrates one possible means of performing such an adjustment.

VI. Form of Buy-In Payment

A. INITIAL BUY-IN: LUMP SUM VS. CONTINGENT ROYALTIES

A controlled taxpayer has substantial freedom to structure the CSA buy-in as a lump sum, a series of installment payments, or royalty payments contingent on the use of the intangible. Treas. Reg. § 1.482-7(g)(7). The characterization of the buy-in transaction (e.g., as a sale, license etc.) does not restrict the taxpayer's ability to choose a form of payment. However, where only nominal or no consideration has been paid, the regulations dictate the arm's length consideration shall be in the form of a royalty, unless a different form is demonstrably more appropriate. Treas. Reg. § 1.482-4(f)(1).

The transfer pricing methods described in this CIP yield distinct payment forms for the CSA buy-in. The income method, for example, yields a lump sum present value of the initial buy-in. Exhibit A.2 illustrates use of the income method to derive a lump-sum buy-in. Where the payment form yielded by the selected transfer pricing method differs from the form chosen by the controlled taxpayer, a conversion will be necessary.⁵⁵ The lump sum present value of an initial buy-in generated under the income method may be reliably converted into a royalty by dividing the lump sum amount by the expected present value of an appropriate royalty base, e.g., sales or revenues from the CFC's exploitation of intangibles resulting from the R&D (and, in some cases, from exploiting make-sell rights in the current generation intangible). Exhibit A.3 and Exhibit A.4 illustrate the conversion of a lump-sum initial buy-in into a contingent royalty payable either over ten years or in perpetuity.

Of course, because the buy-in is determined by reference to the anticipated present value of the royalty base, the present value of the actual royalty payments may not equal the present value of the lump sum buy-in. The present value of the actual royalties will be either more or less than the present value of the lump sum buy-in, reflecting the fact that exploitation of the cost shared intangible is more or less successful than originally anticipated.⁵⁶ Thus, so long as the present value used for purposes of either the lump sum or the contingent royalty is reached via the most reliable method based on the reasonable expectations upfront at the time of the buy-in determination, it is not pertinent, if a royalty form is adopted, whether or not the contingent royalties over time actually add up to such present value. Indeed, actual results may prove to be worse (a "dry hole") or better (a "gusher") than projected, in which case the contingent royalties may turn out to be less (even zero) or more (even by several orders of magnitude) than such present value.

A variety of forms of payment are possible that will be respected, if consistent with economic substance. For example, taxpayers could conceivably adopt a fixed outcome payment form structured as preliminary royalties with subsequent true-up/down to a liquidated amount agreed in advance before the actual results were known. Conversely, taxpayers could conceivably adopt a contingent outcome payment form structured as a preliminary lump sum with true-up/down in light of subsequently arising actual results (i.e., a prepaid royalty). However, taxpayers must clearly adopt these deals - - the Commissioner is not obligated to write in such contractual terms for the taxpayer's benefit (i.e., as drafter, the taxpayer may and should be held to its deal). By the same token, if the taxpayer conscientiously adopted and lived by a truly contingent form of payment, and if subsequent events do not turn out as reasonably anticipated, the Commissioner may not come in and propose a "true-up" adjustment on the grounds that such an adjustment is necessary to reach the originally-projected present value.⁵⁷

IRS personnel may encounter contractual provisions that call for the CSA participants to reevaluate, on a periodic basis, the CSA buy-in. Such a provision might indicate, for example, that the buy-in royalty rate, although agreed to in advance based on initial projections, is potentially subject to adjustment on a prospective basis by reference to sales, profits, or other data. Such a contractual term normally would make the buy-in payments vary more with the success of the project, and thus would increase the risk of the participant receiving those payments. Therefore, in computing the present value to that participant of the anticipated royalty payments, a higher discount rate should be used than would apply in absence of such a contractual term.⁵⁸ This higher discount rate means that higher nominal anticipated buy-in payments will be required to achieve the same present value (the value of the buy-in intangible at the time of the buy-in, which does not change). Thus, the presence of such a contractual term means that the initial royalty rate established (i.e., the rate that is now subject to adjustment) would be higher than it would be if the term were absent.

As any other contractual term in a controlled transaction, a provision of this type is subject to economic substance considerations. See Treas. Reg. § 1.482-1(d)(3)(ii)(B)(1). In the context of a CSA under Treas. Reg. § 1.482-7, a periodic reevaluation clause would need to be in writing, agreed to in advance, and would need to contain terms that could be applied in an objective manner. Importantly, taxpayer's conduct must be consistent with the reevaluation, such that it applies the provision to increase the royalty as well as to decrease it, depending on actual experience. Moreover, such a provision may not fundamentally contravene the intended allocation of risk under cost sharing, in particular the requirement that each CSA participant must bear the risk of and be exposed to adverse outcomes with respect to its interest in the cost shared intangibles. Finally, in evaluating the economic substance of such a term, the IRS may take into account that the payment form specified for the buy-in also calls for an uneconomic, rapid ramp-down of the royalty rate to zero, which would indicate that the royalty may not be subject to meaningful adjustment as contemplated by the reevaluation provision.

B. SUBSEQUENT ACQUISITION BUY-IN: LUMP SUM VS. CONTINGENT ROYALTIES

The taxpayer may adopt a contingent royalty payment form for a subsequent acquisition buy-in. As with the income method, conversion to a royalty requires projections of the base (e.g., operating sales or profits) on which royalties will be paid. Unlike the income method, the conversion process under the acquisition price method requires projections that are not otherwise necessary in applying the method. Exhibits B.2 and B.3 illustrate the conversion of a lump-sum subsequent acquisition buy-in into a contingent royalty payable either over 10 years or in perpetuity.

C. BASE FOR APPLICATION OF CONTINGENT ROYALTY

As just described, the lump sum present value may be converted into a contingent royalty payment form by dividing the lump sum by an appropriate royalty base amount, such as the present value of the CFC's projected sales or revenues from exploitation of the intangibles that result from the R&D (and from exploitation of the make-sell rights in the current generation intangible, under an aggregate analysis). This calculation yields a royalty rate on the chosen base that is thereafter applied to the actual base amounts from year to year, over the course of the contingent royalty period. If the selected royalty base is sales of a product or group of products, the royalty rate is applied to such sales each year, regardless whether they materialize to be greater or less than projected. Any variation of actual sales (or royalties) from the amounts that were projected constitutes the playing out of the risks inherent in the taxpayer's adoption of a contingent form of payment.

The royalty base selected by the taxpayer, to which the net present values will be applied, should be one that the taxpayer uses in its ordinary course of business or, alternatively, one as to which data can be readily generated, either in the course of the taxpayer's preparation of contemporaneous transfer pricing documentation or in response to standard information requests in an IRS examination. Thus, the royalty base should generally correspond to total sales or profits from a particular product, product line, or a rational grouping of products.

Consider the following example.

Company X is a designer and manufacturer of looms for industrial and home use. Company Y is a small company that holds a patent on a revolutionary new control mechanism that has not been successfully applied to looms, but has been adapted to other textile-production machines. Company X anticipates that it may be able to adapt various features of this patented technology, perhaps to its entire product line of looms. It also anticipates that, if this development effort is successful, Company X will experience increased sales volume or total revenues. Company X, in its ordinary course of business, tracks sales and revenues to industrial and home loom product divisions, but it does not maintain detailed data at the level of product lines, much less specific products. Company X acquires the patented technology from Company Y and contributes it an existing CSA with its CFC, and elects a royalty form of payment for the buy-in for the technology. Company X may specify that the CFC participant will pay a buy-in royalty based on any reasonable basis, which would include: (1) sales of all looms, (2) sales of industrial looms, or (3) sales of home looms.

The obligation on the part of taxpayer to perform a conscientious, upfront valuation of the buy-in intangible generally cannot be reconciled with the taxpayer's adoption of payment term that calls for determination, after-the-fact, of the portion of sales or profits that are attributable to the specific buy-in intangible. As a practical matter, such "tracing" to actual transactions would be extraordinarily difficult and resource-intensive, from the perspective of the IRS examination function. This approach would also give rise to a potential for abuse, as taxpayers could readily claim that specific items of intangible property that were contributed to the CSA in fact failed to result in successful derivative products. Finally, under a valid upfront valuation, potential variations in the royalty base (i.e., on account of different potential outcomes) should properly have been taken into account when the taxpayer made the projections and determined the applicable royalty rate, in the first instance.⁵⁹

For example, assume that the CFC forecasts that it will develop its own marketing or manufacturing operating intangibles that it will utilize in conjunction with intangibles resulting from the cost shared R&D (as well as licensed make-sell rights in the current generation intangible). In such cases, the taxpayer's projections should include projected expenditures relating to development of such operating intangibles. Those expenses would reduce the projected operating profits and, hence, result in a lower net present value of the buy-in. Dividing that amount by the present value of sales or revenues (likewise reflecting the development of the operating intangibles) will yield a royalty rate. The resulting royalty rate, by definition, takes into account the potential successful development of such operating intangibles.

VII. Transfer Pricing Documentation Penalties

Where an adjustment proposed by the IRS exceeds the applicable penalty threshold, the transfer pricing documentation penalty is nonetheless inapplicable if the taxpayer produces contemporaneous documentation indicating that it had a basis to reasonably conclude that its selection and application of a transfer pricing method, whether specified or unspecified, produced the most reliable measure of an arm's length result.

This CIP has considered cases where, under the facts and circumstances, the taxpayer's application of specified methods would not produce as reliable a measure of an arm's length result for initial or subsequent acquisition buy-ins as would applications of unspecified methods such as the income or acquisition price methods. Even if this is in fact the case and a buy-in adjustment in excess of the penalty threshold is proposed, the taxpayer may still be able to show under the facts and circumstances that its application of a specified method provided a basis for a reasonable conclusion that its method was the best method. In some cases, however, it may not be possible for the taxpayer to demonstrate the basis for such a conclusion, under the facts and circumstances.

IRS personnel should be familiar with procedures used to submit proposed applications of transfer pricing documentation penalties to the Penalty Oversight Committee. See Announcement 96-16 and subsequent administrative guidance. It is anticipated that dispositions of cases involving CSA buy-ins will be handled in accordance with the procedures generally applicable to transfer pricing cases, including, as appropriate, referral to the Penalty Oversight Committee.

VIII. List of Exhibits

A. INITIAL BUY-IN (INCOME METHOD)

² Solely for ease of presentation, the CIP is phrased in terms of U.S.-based groups, but the analysis is symmetrically applicable to foreign-based groups. In such cases, one should read "U.S. parent" or "U.S. group" as "foreign parent" or "foreign-based group" and the "CFC" or "foreign participant" as the "U.S. subsidiary" or "U.S. participant." Similarly, the CIP refers to a single CFC, although multiple CFCs may be CSA participants. Also, the CFC participant may not be the same as the affiliated CFC or CFCs that actually exploit the cost shared intangibles. For simplicity, however, the CIP generally refers to a single CFC participant, which is assumed to exploit the cost shared intangibles directly.

³ A license of make-sell rights is often executed at the same time as a transfer of buy-in intangibles under a CSA. This practice is particularly common in the high technology sector (e.g., computer software and hardware), but less common (or absent) in other sectors. Make-sell rights, if transferred, do not constitute buy-in intangibles, because they are not "made available for purposes of research" and thus do not give rise to an obligation to make a buy-in payment. Instead, the arm's length payment for such make-sell rights is determined under Treas. Reg. § 1.482-4. Nonetheless, it is often possible to value the buy-in intangibles and the make-sell rights on an aggregate basis, and this may provide the most reliable measure of an arm's length result. See Treas. Reg. § 1.482-1(f)(2)(i) (Aggregation of transactions).

⁴ Treas. Reg. § 1.482-1(d)(3)(ii)(B), (d)(3)(iii)(B), and (f)(2)(ii).

⁵ Thus, weight is more likely to be given to the taxpayer's express designation to the extent such designation is evidenced contemporaneously with entering into the CSA in the case of an initial buy-in, or contemporaneously with the subsequent acquisition in the case of a subsequent acquisition buy-in.

⁶ Technology and marketing intangibles may be more or less interdependent, depending on the particular facts. To the extent a marketing intangible and technology intangible are economically linked, they would not be separately distinguished for purposes of the analysis.

⁷ They possibly could be the subject of an arm's length charge for a transfer or license determined under Treas. Reg. § 1.482-4, e.g., in the case of so-called worldwide marketing intangibles owned by the U.S. group, where those also benefit the operations of the CFC(s).

⁸ Any suggestion that the CFC's undertaking to share the funding of future R&D compensates for the rights to use the pre-existing intangibles as a research platform, must be rejected. Sharing of ongoing R&D entitles the CFC to share in the future incremental value expected from the R&D. For example, the present value of the reasonably projected R&D is subtracted in the income method. It is not correct to say that these R&D cost sharing payments also compensate for the pre-existing platform value of the intangible. The separateness of the two values is evidenced in two distinct obligations: one to make cost sharing payments and the second to make buy-in payments. The compensation for the platform rights clearly falls into the latter category. Treas. Reg. § 1.482-7(g)(2) (buy-in payment for "pre-existing intangible property [made available] . . . for purposes of research in the intangible development area").

⁹ Note that the buy-in typically relates to technology intangibles, although, less commonly, marketing intangibles may also be within the scope of the cost sharing arrangement. Nestle, of course, dealt with a marketing intangible -- a trademark.

¹⁰See Treas. Reg. § 1.861-18(f)(1).

¹¹See I.R.C. § 881(a)(4); Commissioner v. Wodehouse, 337 U.S. 369, 1949-2 C.B. 62 (1949).

¹² Technically the license of make-sell rights, as assumed in the typical buy-in scenario, is governed by Treas. Reg. § 1.482-4, while the transfer of the interest in the buy-in intangible is governed by Treas. Reg. § 1.482-7(g). As explained, however, a more reliable result may often be achieved by evaluating both these transactions in the aggregate. See Treas. Reg. § 1.482-1(f)(2)(i) (Aggregation of transactions). The transfer of the buy-in intangible, combined with a license to the CFC of make-sell rights, effectively provides the CFC with full rights to the platform intangible in the CFC's territory or field of use. Thus, together, the transfer of buy-in intangibles under the CSA along with the tandem grant of a make-sell license, may be viewed as economically indistinguishable from the §sale§ of an interest in the buy-in intangible.

¹³ Buy-in provisions of the CSA may be contained in separate agreements contemporaneous with the CSA, for example, a technology license agreement or "TLA" which also license make-sell rights. For simplicity, within this CIP when referring to buy-in provisions, the term "CSA" includes any associated agreements containing such provisions such as TLAs.

¹⁴ "If the contractual terms are inconsistent with the economic substance of the underlying transaction, the [IRS] may disregard such terms and impute terms that are consistent with the economic substance of the transaction." Treas. Reg. § 1.482-1(d)(3)(ii)(B)(1) (last sentence). See also section VI.C, below.

¹⁵ See Exhibits A.4 and B.3. As discussed further in section VI, so long as the present value used for purposes of either the lump sum or the contingent royalty is reached via the most reliable method based on the reasonable expectations upfront at the time of the buy-in determination, it is not pertinent, if a royalty form is adopted, whether or not the contingent royalties over time actually add up to such present value. Indeed, actual results may prove to be worse (a "dry hole") or better (a "gusher") than projected, in which case the contingent royalties may turn out to be less (even zero) or more (even by several orders of magnitude) than such present value.

¹⁶ In some cases, over time the U.S. group may even surrender a significant share of its prior business to the CFC.

¹⁷ Where both the U.S. group and the CFC(s) contribute pre-existing intangibles -- e.g., the U.S. group contributes a pre-existing platform intangible and the CFC(s) contribute preexisting operating intangibles that will enhance the exploitation of cost shared intangibles in the foreign operations -- this CIP describes an unspecified RPSM that may provide a reliable measure of the buy-in. See section IV.D.

¹⁸See Ithaca Industries, Inc. v. Commissioner, 97 T.C. 253, 272 (1991), aff'd on other grounds, 17 F.3d 684 (4th Cir. 1994).

¹⁹ Nevertheless, the Fourth Circuit upheld the Tax Court on the grounds that the taxpayer's evidence was insufficient to demonstrate a limited useful life for the workforce intangible at issue.

²⁰ Even then, there may be an independent value to a long term contract with an individual who possesses special know how.

²¹ Note that the value of the research team intangible, as explained above, is net of the expected liabilities for compensating the team members. The latter compensation is part of the R&D costs that the CFC shares via cost sharing payments. There is thus no double counting as between the values compensated by the buy-in and the cost sharing payments.

²² It is recognized that certain administrative statements may be understood as inconsistent with this CIP's conclusion, but, to the extent they are so interpreted, they are erroneous. See, e.g., Report on the Application and Administration of Section 482 (April 1992) ("it is not necessary to compensate a participant for the going concern value of its research operations under the buy-in provisions"). Even these administrative statements may be reconciled with the position set forth below in section III.E.4 that, in any event, the going concern value of the taxpayer's research operations excludes the synergy value of the research team related to the buy-in intangible.

²³ Presumably the 1992 IRS Report, supra note 22, intended to refer only to such generic value when it spoke of the going concern value of research operations.

²⁴ See the case law cited in the text accompanying note 18, supra. "The cost approach is generally favored for valuing assembled workforces." Gordon V. Smith and Russell L. Parr, Valuation of Intellectual Property and Intangible Assets, 321 (John Wiley & Sons, 3rd Ed., 2000).

²⁵See generally "Taxpayer Use of Section 482 and the Commensurate with Income Standard," AM-2007-007, Mar. 15, 2007.

²⁶See generally Treas. Reg. § 1.482-1(d)(3)(ii)(B) & (iii)(B).

²⁷ I.R.C. § 482 (". . . the Secretary may . . . allocate gross income, deductions, credits, or allowances . . ."); Treas. Reg. § 1.482-4(f)(2)(i) (". . . the district director . . .").

²⁸ For example, assume that the cost sharing participants undertake research on Compound A, which is projected to cure minor dermatological ailments. The resulting compound unexpectedly provides an effective treatment for most types of skin cancer. The taxpayer must be given an opportunity to demonstrate that the outcome of the Compound A project could not have been reasonably anticipated when the CSA was executed.

²⁹ Treas. Reg. § 1.482-4(f)(2)(ii)(D). See also Treas. Reg. § 1.482-4(f)(2)(i) ("Adjustments made pursuant to this paragraph (f)(2) shall be consistent with the arm's length standard and the provisions of § 1.482-1.").

³⁰ Some analysts attempt to identify and separately consider: (1) returns to existing product intangibles or sunk R&D costs; (2) in-process R&D made available to CSA participants; and (3) R&D under cost sharing that leads to the creation of future intangibles. Others apply a cut-off or a ramp-down for purposes of identifying the stage at which the CFC can capture all or a portion of the extraordinary returns to the cost-shared R&D.

³¹See D. Frisch, Frisch Offers Method for Computing Royalty Under White Paper's BALRM, Tax Notes Today, 89 TNT 107-30 (May 19, 1989). Dr. Frisch states, in pertinent part:

³³ In both the early stage and mature foreign operation scenarios, this CIP provides for a range of results derived, inter alia, from a possible range of arm's length discount rates available for projects of comparable risk. The arm's length range in effect recognizes that at arm's length there may be a range of returns that capital providers might expect in a competitive environment.

³⁴ On August 29, 2005, proposed regulations on cost sharing arrangements were published in the Federal Register which, inter alia, provided guidance under which both the income method and the market capitalization method, as well as a modified three-step residual profit split method, would become specified methods for purposes of valuing the buy-in (there referred to as the amount charged in a preliminary or contemporaneous transaction (PCT)). See Prop. Treas. Reg. § 1.482-7(g)(4), (6), and (7). This CIP does not rely on the proposed regulations method specification. Rather, as indicated, the CIP position follows from the application of the best method analysis under the existing regulations. These methods have had widespread use both for tax and non-tax purposes and, accordingly, are eligible to be found to be the best method under the standards of Treas. Reg. § 1.482-1(c).

³⁵ See also section III.C.

³⁶ See discussion of the income method, below. That method does allow for reliable adjustments to take into account the R&D cost contribution risk.

³⁷ See also section III.D.

³⁸ In addition, some taxpayers incorrectly assume that R&D expenses may be amortized (often by reference to the useful life of a particular generation of finished product) over a very short period. This assumption can dramatically reduce the amount of the CSA buy-in.

³⁹ Thus, the income method avoids the adjustment necessary under the market capitalization method to carve out pre-existing operating intangible value (including U.S. goodwill and going concern value). The circumstances and reliability of this carveout adjustment under the market capitalization method is discussed further below. See section IV.E.

⁴⁰ Although the income method is technically an unspecified method, the method is nevertheless well-grounded in the regulations. Regarding the key comparability requirement for a CUT, namely the requirement of that comparable intangible property must have a similar profit potential, the regulations provide in pertinent part:

⁴¹ Application of the income method requires projections for the CFC's operating income, and projections that focus solely on the CFC are typically more reliable. Projections for the CFC may be readily generated when, as in Exhibit A.2, they are based on actual experience with the CSA. Nevertheless, in some cases, separate projections for the CFC are not available and must instead be derived from data for the controlled group. As a matter of presentation, Exhibits A.1 through A.6 start with projections of the controlled group's income and generate projections for the CFC, based on the assumption that revenues and costs are pro rata to RAB share. This simplifying assumption may not hold true in particular cases, for example, if the operating margin or sales growth rates for the CFC differ from those applicable to the controlled group as a whole.

⁴² When performing any extrapolation, IRS personnel should guard against impermissible application of hindsight, in contravention of the CWI principle. See discussion above in section III.F. In applying the income method, the touchstone is the initial, upfront valuation, which should take account of all information known at the time of the valuation.

⁴³ Exhibit A.5, as the other exhibits to this CIP, is illustrative only. No general inference is intended concerning the factors that might be treated as varying across some range in a reliable application of the income method, the appropriate range of such factors, or the range of results that might reasonably be expected.

⁴⁴ This might require examination of pre-CSA transfer pricing, as well as business restructurings, transfers or licenses of intangibles, or changes in the respective risk-profiles of the controlled parties that took place contemporaneously with entry into the CSA. This may also require an analysis of the identification and categorization of the relevant intangibles. See section III.A. For example, a foreign installed customer base may be attributable to the U.S. group's technology or reputation, rather than the CFC(s)' marketing intangibles. See note 6 and accompanying text.

⁴⁵ The discussion in the text assumes a wholly-owned U.S. group. Further refinements may be necessary for more complex structures.

⁴⁶ If make-sell rights in the current generation intangible are transferred to the CFC at the inception of the CSA, the value of the make-sell rights and the buy-in intangible may often be reliably determined, on an aggregate basis. In that case, in the second step, the step one intangible value should not be reduced by the value of the make-sell rights.

⁴⁷ Where the CFC(s) also own foreign operating intangibles, the second step adjustment must also carve out their value in determining the amount of the buy-in owed to the U.S. group.

⁴⁸ This equivalence can be expressed mathematically as follows. Assume --

• m0 must equal U. S. group's anticipated benefit from marketing, since upfront value equals anticipated benefit

• t0+t1 must equal ustb+cfctb, since the worldwide technology investments must, upfront, equal the worldwide technology return

                             t1 *    cfctb

                          (t0+m0) *     cfctb

                       (t0+m0+t1) *     cfctb

⁴⁹ Whether such an asset is made available to the CSA "for purposes of research" is a factual question. As a practical matter, an acquired intangible is often made available to the CSA contemporaneously with its acquisition from a third party, although the contractual provisions adopted by the related parties may be unclear concerning the intended treatment of the asset. As elsewhere, such provisions need to be evaluated in light of economic realities and, in the absence of economic substance, the IRS may impute provisions in line with the substance. Treas. Reg. § 1.482-1(d)(3)(ii)(B).

⁵⁰ Where the CFC makes the subsequent acquisition, the subsequent acquisition buy-in would be owed to the CFC by the U.S. group.

⁵¹ The proposed regulations on cost sharing arrangements, published on August 29, 2005, provided guidance under which the acquisition price method would become a specified method for valuing the buy-in (amount charged in a PCT). See Prop. Treas. Reg. § 1.482-7 (g)(5). This CIP does not rely on the proposed regulations method specification. Rather, as indicated, the CIP position follows from the application of the best method analysis under the existing regulations. This method has widespread use both for tax and non-tax purposes and, accordingly, is eligible to be found to be the best method under the standards of Treas. Reg. § 1.482-1(c).

⁵² Determining the total acquisition price for the target assets or target stock may require further analysis. Where the U.S. parent's stock is publicly traded, there may be a readily ascertainable market value of the stock.

⁵³ However the target intangibles may also include intangibles that are made available to the CFC (including, possibly, make-sell rights in current generation products). It may be more reliable to evaluate such other transferred intangibles, such as make-sell rights, and the buyin intangible on an aggregate basis. In that case, the value of such other transferred intangibles should not be backed out in this second step.

⁵⁴ Exhibit B.4, as the other exhibits to this CIP, is illustrative only. No general inference is intended concerning the factors that might be treated as varying across some range in a reliable application of the acquisition price method, the appropriate range of such factors, or the range of results that might reasonably be expected.

⁵⁵ The regulations recognize that a fixed lump sum or installment payment may be converted to a contingent royalty using projections together with an appropriate discount rate. Treas. Reg. §§ 1.482-4(f)(5) (lump sum payments).

⁵⁶ Upfront, at the time of the buy-in, the probabilities must be consistent with the present value reached for the buy-in. After-the-fact, events may or may not actually depart from what was reasonably projected upfront. Thus, there is no general proposition that the actual after-the-fact amounts of a buy-in in a contingent form will or will not line up with the upfront present value.

⁵⁷ Again, however, the present value projected for the consideration under these payment form variants must be consistent with the present value reasonably expected for the buy-in. As just one example, consider a form of payment under which the CFC agrees to pay a lump sum to the U.S. group at the time of the buy-in, but with the right to demand a refund subsequently, if results actually realized are less favorable than reasonably projected at the time of the buy-in. In such a case, the CFC must purchase the refund option for arm's length consideration. So, the upfront lump sum payment under this form will be greater than for a purely fixed lump sum buy-in, because the arm's length refund option price must be added to the arm's length fixed buy-in price.

⁵⁸ Similarly, substituting a royalty-on-sales payment form for a fixed-installment payment form, or substituting a profit-split payment form for a royalty-on-sales payment form, increases the risk that the participant will not receive the amount of payments that was projected, and requires use of a higher discount rate to compute a present value of those for the anticipated payments. The contractual term discussed above is simply one of several permutations in payment form that increase risk and thus increase the discount rate that should be used in the NPV calculation.

⁵⁹ On the one hand, the taxpayer might reliably take account of various outcomes, for example by using a probability analysis in making the initial projections of sales revenues. As noted in the example concerning operating intangibles, such assumptions would be reflected in the projected sales, profits, etc. On the other hand, attempting to take into account possibly-divergent outcomes of R&D directly by means of contingent payment terms (e.g., royalty payments are dependent on the outcome of cost-shared R&D), is inconsistent with the concept that the taxpayer must perform a conscientious, up-front valuation of the buy-in.