Coke Concentrate: A Recipe for Understanding the IRS's Biggest Win in 40 Years
Clark Armitage, Heather Schafroth, Elizabeth Stevens, and David Rosenbloom authored the January 28, 2021 article "Coke Concentrate: A Recipe for Understanding the IRS’s Biggest Win in 40 Years" for IBFD International Transfer Pricing Journal Volume 28, No.2. Below is the full article, and please visit this link to view the article as it appears on IBFD's website.
The United States Tax Court’s November 18, 2020 opinion in The Coca-Cola Company & Subsidiaries v. Commissioner marks an inflection point in U.S. transfer pricing jurisprudence. Writing for the court, Judge Albert Lauber ruled for the U.S. Internal Revenue Service (“IRS”) on most issues, including those for which the most dollars were at stake. The taxpayer must include in income more than U.S. $7 billion in transfer pricing adjustments over three taxable years (2007-09), with more at stake for later years.
The decision is a watershed not only because it is the biggest IRS win in decades, but because it debunks the myth that the Tax Court is uncomfortable accepting IRS adjustments based on applications of the comparable profits method (the “CPM”) and other profits-based methods. This view arose in recent years when, in case after case, the Tax Court rejected IRS applications of the CPM (a method analogous to the OECD’s Transactional Net Margin Method) and other profits based methods in favor of the comparable uncontrolled transactions (“CUT”) method. Our view is that the Coca-Cola court’s acceptance of the CPM is not a departure from precedent, because it turns on findings of fact which, unlike in other recent decisions, allowed for the application of the CPM. As should be the case in transfer pricing, the facts mattered.
Perhaps more fundamentally, the court’s decision seems to embrace substance concepts akin to the OECD’s “DEMPE” doctrine. While the court did not expressly adopt DEMPE (or even mention it), its reasoning reflects an unwillingness to allow a mere cash box, with no real decision making ability or authority, to earn non-routine profits from its cash funding. As we explain below, the court may in fact have gone too far in this effort by suggesting that it would disallow even a risk-adjusted return to the cash box.
Both of these apparent shifts in approach turn on a core conceptual underpinning of the case – the central importance of legal ownership in allocating profits from intangible property. Judge Lauber refused to treat foreign affiliated “supply points” as beneficial owners of intangible property or as owners of discrete derivative assets, when the supply points lacked legal ownership of the intangibles and had only short-term contracts with the U.S. parent company (which did have legal ownership). As the court emphasized, doing so would have been inconsistent with the form of the taxpayer’s own transactions, and in the circumstances of this case, could not be legally supported.
The case involves numerous other important determinations:
- That a transfer pricing method (“TPM”) provided in a closing agreement for prior years (1986-1995) generally will not bind the IRS for later years (2007-09), regardless of whether the taxpayer consistently applied the TPM during the intervening period;
- That a court may not be willing to evaluate the combined profits of all foreign related entities in reviewing adjustments relating only to some of those foreign entities (taking a different tack than a 2012 Canadian Supreme Court decision);
- That the IRS cannot treat foreign income tax paid by a foreign branch of a U.S. company as non-compulsory (and therefore non-creditable) prior to adjudication of the transfer pricing issues that might reduce the income of the branch for U.S. purposes, especially in circumstances where IRS refuses access to the mutual agreement procedure of the applicable tax treaty; and
- That a court will give effect to IRS adjustments to foreign exchange gain or loss of a foreign branch upon an allocation of income from that branch to the U.S. head office which allocation has no U.S. tax consequence.
The Coca-Cola decision is a lengthy roadmap (244 pages) for future courts to follow in supporting IRS adjustments based on the CPM, given appropriate facts. In this article, we explain the critical factual and legal determinations that allowed the court to reach its decision on the CPM and discuss other important aspects of the case.
Coca-Cola’s international operations during the 2007-09 audit period involved transactions among four categories of parties – three controlled (so that their inter-entity transactions are subject to section 482) and one uncontrolled (which in the end provided comparables):
- The Coca-Cola Company, the U.S. parent of the multinational enterprise, based in Atlanta, Georgia (including its U.S. affiliates, “Coke US”).
- Seven foreign “supply points,” which were controlled by Coke US, manufactured concentrate for sale to independent bottlers, and were located in Brazil, Chile, Costa Rica, Egypt, Ireland, Mexico, and Swaziland. All were foreign subsidiaries of Coke US except the Mexican supply point, which was a foreign branch. These seven reflected the consolidation over many years of numerous supply points that had been located in nearly every country where Coca-Cola products were sold.
- Foreign independent “bottlers,” which almost always were unrelated to Coke US, purchased concentrate from the supply points, used it to manufacture finished Coca-Cola products, and marketed, sold and distributed those products outside the United States.
- About 60 foreign “ServCos,” which were controlled by Coke US, and supported local marketing efforts of the bottlers in one or more local markets.
The IRS and Coke US (and the court) seemed generally to agree on the functions performed by each of these parties:
- Coke US performed some of the most important functions. It:
- “set detailed guidelines for brand identity, visual identity of products, quality assurance, business goals, and marketing strategies[;]”
- determined budgets and set strategies for annual budget plans for the ServCos and bottlers;
- “took principal responsibility for R&D and quality assurance;” and
- “was chiefly responsible for supply chain management regarding concentrate.” 
- The supply points made concentrate for Coke beverages and sold the concentrate to the bottlers, using recipes, formulas, processes and quality control standards developed and provided to them by Coke US.
- “The bottlers used this concentrate to produce finished beverages that they marketed (directly or through distributors) to millions of retail establishments throughout the world (excluding the United States and Canada).”
- “The ServCos were responsible for local advertising and in-country consumer marketing, which they carried out with assistance from third-party media companies and creative design firms. The ServCos were also responsible for liaison with local bottlers, a function petitioner called ‘franchise leadership.’ A few ServCos had research and development (R&D) centers, which served multiple national markets.”
Coke US characterized the supply points and the ServCos together as the foreign “Field.” This characterization was important to Coke US’ framing of the transfer pricing analysis, which, as explained below, was rejected by the court.
Assets, Risks, and Contractual Rights:
The parties likewise agreed on some of the risks and resulting assets of the related parties—but not all. The key difference relates to the supply points’ risk profile and intangible asset ownership.
The parties did not dispute that Coke US legally owned almost all valuable intangibles for the group: “[Coke US] was the registered legal owner of virtually all trademarks and other intangible assets used to manufacture and produce [Coke US]-branded beverages.” Any residual profit from the business that was not properly retained by the ServCos and the supply points would thus flow to Coke US.
The parties likewise did not dispute that the ServCos neither legally owned valuable intangibles nor had long-term contractual rights. All agreed that the ServCos, contractually and substantively, were routine functionaries that should be compensated on a cost-plus basis.
Finally, the parties did not dispute the material facts regarding the bottlers. The bottlers contracted directly with Coke US, and their contractual relationships reflected an overall effort to align financial interests, with each expected to earn about 50% of overall system profit. “[Coke US’] contracts with its bottlers explicitly granted them long-term and generally exclusive rights to produce and sell [Coke US’] products within their respective territories.”
The IRS and Coke US disagreed, however, on the allocation of risk between Coke US and the supply points and the effect of that risk allocation on the supply points’ ownership of valuable intangible property. Coke US asserted that the supply points, as part of the foreign “Field,” were risk takers with respect to certain marketing payments made to and by the ServCos and, therefore, owned “immensely valuable intangible assets that do not appear on their balance sheets or in any written contract. These assets, which petitioner calls ‘marketing intangibles’ or ‘IP associated with trademarks,’ allegedly were created when the supply points financed consumer advertising in foreign markets.” The IRS vehemently disagreed.
As we explain below, the court sided with the IRS, and its determination that the supply points did not own valuable intangible assets was key to the overall outcome of the case.
IRS Audit History and Compensation of the Supply Points:
To conclude an audit of Coke US’ 1986-1995 tax years, the IRS and Coke US executed a closing agreement that applied a “10-50-50 method” to share profits between the supply points and Coke US. Under this method, the supply points were entitled to retain profits equal to 10% of their sales plus 50% of their remaining profit, with the other 50% share of remaining profit due to Coke US as a royalty. Coke US generally continued to apply this method after 1995, including for the 2007-09 audit period.
Under the closing agreement, the supply points were permitted to pay the royalty due to Coke US in the form of other repatriation payments, including through dividends:
During 2007-2009 more than $1.8 billion of the income petitioner received from its foreign supply points pursuant to the 10-50-50 method took the form of dividends rather than royalties. Petitioner claimed “deemed paid” foreign tax credits (FTCs) under section 902 with respect to these dividends, as the closing agreement had permitted for 1987-1995.
The Coca-Cola decision already is seen as the case that revived the CPM (and other profits-based TPMs). In time, the decision also may come to be known as the first U.S. court application of the OECD’s DEMPE doctrine. But the court’s application of the CPM and its informal adoption of DEMPE were possible only because of the emphasis the court placed on legal ownership of intangible property.
Legal Ownership is the Lodestar
Coke US’ case turned on its contention that the supply points owned highly valuable marketing intangibles that they acquired by funding a portion of the non-U.S. consumer marketing campaigns conceived by Coke US and customized and implemented by the ServCos. The Tax Court found “no support for [this] . . . argument in law, fact, economic theory, or common sense.”
The regulations under section 482 provide that:
The legal owner of intangible property pursuant to the intellectual property law of the relevant jurisdiction, or the holder of rights constituting an intangible property pursuant to contractual terms (such as the terms of a license) or other legal provision, will be considered the sole owner of the respective intangible property for purposes of this section unless such ownership is inconsistent with the economic substance of the underlying transactions.
Thus, the inquiry begins and ends with legal ownership unless the economic substance of the parties’ dealings (contractual ownership) conflicts with it.
For the Tax Court, the outcome of this inquiry was clear-cut: Coke US itself “was the registered legal owner of virtually all trademarks and other intangible assets used to manufacture and produce” its branded beverages, while the supply points were not “the legal owners of any distinct marketing intangibles” under any jurisdiction’s law. The supply points had no greater claim under their contracts with Coke US, which granted them “only a limited right to use [Coke US’] intangibles in connection with their production and sales activities.”
Moreover, the court observed, even if the agreements on their face had purported to grant the supply points a greater interest in Coke US’ intangible property, “such rights would be illusory” because Coke US could (and often did) revoke the agreements. Further, many of Coke US’ agreements with the ServCos—which, unlike the supply points, actually participated in and/or led consumer marketing efforts through their own employees and with the assistance of third-party marketing professionals—provided that any marketing intangibles emerging from those efforts were the sole property of Coke US.
In short, Coke US had successfully pursued a “consistent strategy for protection of its ‘crown jewels’—centralizing ownership of all intangible assets under the U.S. parent to ensure protection under U.S. law.” Because Coke US had not shown that the supply points held legal rights in any intangibles under any jurisdiction’s law or pointed to any contract recognizing or granting such rights, the supply points owned no intangible property cognizable for transfer pricing purposes.
But what about intangibles that cannot be legally owned in either sense contemplated in the regulation? When the entities that became the supply points were incorporated after World War II, they acquired “tangible operating assets, associated goodwill, and similar items” but “no meaningful intangible property in the form of trademarks, tradenames, copyrights, franchises, licenses, or bottler agreements.” Citing private letter rulings obtained in connection with some of these transactions, Coke US insisted that the supply points had “inherited goodwill and similar assets from their predecessors.” The Tax Court swatted away this evidence because Coke US had pointed to “no authority for the proposition that ordinary corporate goodwill should be treated as an ‘intangible asset’ for purposes of analysis under sec. 1.482-4(b), Income Tax Regs” and suggested instead that such goodwill derives from the contractual relationships of the parties.
The Taxpayer Is Stuck With Its Form
Having concluded that the supply points lacked legal or contractual ownership of valuable intangibles, the court turned next to Coke US’ contention that the court should ignore legal and contractual ownership of intangibles in favor of the real economic substance of the transactions. Coke US asserted that the supply points owned intangible assets by virtue of their payments to the ServCos for local marketing expenditures: “According to petitioner, these intangible assets arose from the ServCos’ expenditures for consumer marketing, which [the supply points] invoiced to [Coke US].”
The court rejected Coke US’ economic substance claims: “First, only the Commissioner, and not the taxpayer, may set aside contractual terms as inconsistent with economic substance. Second, even if petitioner could set aside the terms of its own contracts, it has failed to establish that the economic substance differs from the contractual form.”
The court supported the first of these conclusions with references to the IRS transfer pricing regulations and to long-standing federal court jurisprudence. The regulations permit only the IRS to set aside contractual terms “and impute terms that are consistent with the economic substance” or, in the absence of a written contract, “impute a contractual agreement between the controlled taxpayers consistent with the economic substance of the transaction.” Similarly, the court-developed Danielson rule, which narrowly circumscribes when a taxpayer may set aside contracts, did not permit Coke US to do so here:
The supply points’ agreements with [Coke US] endow them with no ownership rights in marketing intangibles, and the ServCos’ contracts with [Coke US] dictate that any rights to marketing intangibles “are the property of [Coke US].” Under Danielson petitioner cannot disregard these contract terms unless it can show that they would be judicially unenforceable. Petitioner has not attempted to make such a showing.
The Tax Court then concluded that, even if Coke US could disregard its form, its substance was consistent with the form. While this determination was unnecessary to the decision (and essentially dictum), as discussed in the next section, the court’s findings on economic substance allowed it to readily accept IRS application of the CPM to the supply points and to reject Coke US’ alternative TPMs.
The CPM Wins Acceptance
Many commentators have identified the court’s acceptance of an IRS application of the CPM as the most important development of the case. In recent years, the Tax Court has consistently rejected IRS adjustments based on the CPM and other purely profits-based measures, including in its Amazon (2017), Medtronic (2016), and Veritas (2009) decisions. This consistent trend gave rise to the view that Tax Court judges will accept almost any application of the CUT method rather than rely on a profits-based measure to determine pricing. The question then, is whether the Coca-Cola court’s acceptance of the CPM heralds a change in how the Tax Court will look at the CPM and other profits-based TPMs. We believe the court’s acceptance of the CPM stems from its findings on economic substance and that proper application of the CPM remains a case-by-case determination.
Several findings of fact—and one important legal finding—were key to the court’s acceptance of the CPM. First, Judge Lauber found that the supply points performed only routine manufacturing functions and no higher-value functions, and that Coke US in effect agreed:
[The supply points] engaged almost exclusively in manufacturing, and petitioner’s experts agreed that this was a routine activity that could be benchmarked to the activities of contract manufacturers. Two of petitioner’s experts, Drs. Cragg and Unni, applied an 8.5% markup on costs to determine an appropriate arm’s-length return for the supply points’ concentrate manufacturing function.
Second, not only did Coke US own all valuable intangibles in form, it owned them in substance because of the relative bargaining powers of Coke US and the supply points: “[Coke US], not the supply points, had nearly all the bargaining power in the market for concentrate production. The supply points . . . were easily replaceable--and regularly were replaced--at relatively low or zero cost to petitioner. A party thus circumstanced would have little leverage in arm’s-length negotiations.”
Judge Lauber also rejected Coke US’ contention that the supply points’ unreimbursed payments to the ServCos for local marketing expenses entitled them to a share of non-routine profits: “Since petitioner permitted the supply points to enjoy astronomical levels of profitability, their agreement to the quid pro quo of absorbing marketing costs is perfectly explicable in economic terms: Because the forgone royalty expense would vastly exceed the marketing costs, any rational economic actor would have accepted this deal.”
Finally, as a legal matter, the court addressed Coke US’ assertion that the CUT method is always preferable to the CPM. Coke US cited regulatory history indicating that the CPM was to be a method of last resort: “The preamble notes that, ‘[g]iven adequate data, methods that determine an arm’s length price (e.g., the CUP method) * * * generally achieve a higher degree of comparability than the CPM.’ . . . ‘In this regard,’ Treasury said, ‘the CPM generally would be considered a method of last resort.’”
But the Court concluded that Treasury intended for the CPM to be a method of last resort only when adequate data are available to apply the CUT method. Such data were not available here:
Petitioner has identified no pricing data for transactions with unrelated parties that “involve the transfer of the same intangible”--viz., the trademarks, brand names, patents, logos, secret formulas, and proprietary manufacturing processes used to produce Coca-Cola, Fanta, Sprite, and the Company’s other branded beverage products. Thus, the circumstances that caused Treasury to refer to the CPM as a “method of last resort” do not exist here.
Given these factual (and legal) findings, the court’s acceptance of the IRS application of the CPM was all but inevitable: “First, a CPM analysis was appropriate given the nature of the assets owned and the activities performed by the controlled taxpayers.” The court then also found that the IRS had selected appropriate comparable companies – the bottlers – and had applied the CPM (with a return on assets profit level indicator) using reliable data and making reliable assumptions and adjustments.
Having found the IRS’ proposed “bottler CPM” to be reliable, such that its adjustments were not arbitrary and capricious, the Court need not have considered Coke US’ three proposed alternative transfer pricing methodologies. Nevertheless, the Tax Court evaluated and rejected all three, for two principal reasons. First, as discussed below, all three alternatives tested the foreign “Field” and not the supply points alone. Second, all three alternatives suggested that the supply points had contractual rights—intangibles, in effect—that (as discussed above) they simply did not own.
The Coca-Cola court is the first in some years to apply a CPM to price transfers of intangibles. Given the Tax Court’s historical hesitancy to accept IRS adjustments based on the CPM and other profits-based method, is the Coca-Cola opinion out of sync with those earlier decisions (and a change in direction for the Tax Court), or can it be distinguished? We favor the latter view. Notably, Judge Lauber also penned the Tax Court’s Amazon decision, in which he evaluated whether to defer to the IRS’s pure discounted cash flow method (i.e., a profits-based method) or to adopt the taxpayer’s proposed CUT methods for pricing transfers of website technology, marketing intangibles and customer information. Judge Lauber explained that the profits-based discounted cash flow method was inappropriate because:
[The IRS] computes the buy-in payment not by valuing the specific intangible assets transferred . . . but by determining an enterprise value for Amazon’s entire European business. [The IRS] deems the pre-existing intangibles to have a value equal to AEHT’s enterprise value, less its initial tangible assets. By employing an enterprise valuation, [the IRS] necessarily sweeps into [its] calculation assets that were not transferred . . . and assets that were not compensable “intangibles” to begin with.
In contrast, Judge Lauber accepted the CUT method because both parties agreed to could be reliably applied:
If an uncontrolled transaction involves transfer of the same intangible under the same or substantially similar circumstances, the CUT method will generally yield the most reliable measure of the arm’s-length result. . . . Respondent’s and petitioner’s experts agree that the CUT method may reliably be used to value separately the website technology, the marketing intangibles, and the customer information. . . . We conclude that the CUT method provides the best method for determining the fair market value of all three species of intangible property, but we do not wholly agree with the results reached by either party in implementing this approach.”
The facts of Coca-Cola and Amazon – in particular, the availability of reliable CUTs – are distinguishable and, in transfer pricing, facts matter.
Cash Boxes, DEMPE Functions, and the Limits of a Sweetheart Deal
As discussed above, the contention that the supply points “acquired ‘marketing intangibles’ worth tens of billions of dollars” by paying the marketing expenses of the ServCos was central to Coke US’ challenge to the IRS’s transfer pricing analysis. If the supply points had ownership interests in these intangible assets, Coke US asserted, then the IRS’ bottler CPM inappropriately ignored the supply points’ contribution of these marketing intangibles to the business. The Tax Court rejected this assertion based on substance principles that are closely akin to those relevant to application of the DEMPE doctrine in the OECD Guidelines: the supply points were not entitled to non-routine profits from the exploitation of these intangibles because they were mere cash boxes that lacked effective control over the expenditures incurred to develop the intangibles and faced little to no market or financial risk in making the expenditures.
The OECD Guidelines provide that, when analyzing controlled transactions involving intangibles, it is “necessary to determine, by means of a functional analysis, which member(s) perform and exercise control over development, enhancement, maintenance, protection, and exploitation [“DEMPE”] functions, which member(s) provide funding and other assets, and which member(s) assume the various risks associated with the intangible.” The OECD Guidelines treat ownership interests in intangibles as “not confer[ring] any right ultimately to retain returns derived…from exploiting the intangible.” Instead, they emphasize that “[a]ssessing the capacity of a particular entity to exert control and the actual performance of such control functions will be an important part of the analysis.” Specifically, “a party that provides funding, but does not control the risks or perform other functions associated with the funded activity or asset” generally receives lower returns than parties that exercise more control.
The Tax Court did not explicitly apply the OECD Guidelines to analyze the economic substance of the transactions, but its control-focused analysis parallels the OECD Guidelines’ emphasis on control. Indeed, control—who exercises it, over which decisions, about which risks—may be the key to a deeper understanding of the Tax Court’s economic substance analysis in Coca-Cola. The Tax Court noted that “[Coke US] dictated the bottlers to which [the supply points’] concentrate was sold,” “directed” and “instructed” each supply point where to sell and ship beverage concentrate, “controlled how much revenue each supply point received . . . and how much expense each supply point was charged,” and “initiated inter-company charges that placed on the books of each supply point, as [Coke US] deemed appropriate” a portion of the ServCos’ marketing expenses. Coke US similarly “controlled the ServCos’ annual budgets, provided major inputs to their marketing efforts, and supplied final approval for all business plans.”
In contrast, the supply points exercised no control over the ServCos’ marketing expenses they bore. They “played no role in arranging consumer marketing,” “did not themselves engage in significant marketing activities,” “had no legal obligation to defray the cost of the ServCos’ consumer marketing,” and in fact “had nothing to do with consumer marketing.” The Tax Court found “no evidence that the supply points received invoices” for any ServCo marketing expenses, that the supply points “reviewed the propriety of the amounts they were charged,” or that any supply point (with the possible exception of the Irish supply point) had ever “explicitly agreed to bear financial responsibility for these charges.” Instead, the supply points were “passive recipients of charges that [Coke US] put on their books,” with “no voice in selecting or evaluating the services for which they were…made financially responsible” by Coke US, and “simply had marketing costs assigned to them.”
Under the OECD Guidelines, a related party that provides funding will not be treated as assuming financial risk unless that party has “the capability to make the relevant decisions” about that risk, such as deciding whether and how much to invest. Although exercising control over whether to take a financial risk may entitle a party to a risk-adjusted return on the funding it provides, the risk involved in providing funding is “lower when the party to which the funding is provided has a high creditworthiness.” Noting that “the supply points had no operational responsibility for consumer marketing,” the Tax Court concluded that “they thus bore no risk in the sense of ‘mission failure.’“ The supply points also bore no real marketing risk because Coke US controlled “the flow of revenue and marketing expenses to the supply points,” and “the revenue invariably exceeded the marketing expenses by a very wide margin.” The Tax Court also found that the supply points bore very little financial risk: “the risk of bottler default was very low.”
The Tax Court’s reasoning implies that the supply points’ payment of marketing expenses, without the assumption of any real marketing or financial risk, should not entitle the supply points to any profits on those expenditures. Thus, to the extent that the supply points paid marketing expenses incurred by the ServCos, the Tax Court appeared to view the supply points as “cash boxes.” Cash boxes are “capital-rich entities without any other relevant economic activities.” Under the OECD transfer pricing approach for intangibles, cash boxes are “entitled to no more than a risk-free return.”
Coke US responded that, in an arm’s-length transaction, the supply points’ payments of the ServCos’ marketing expenses would be “inexplicable” unless the supply points expected to enjoy some form of ownership interest in Coke US’ intangibles. But the Tax Court responded that the supply points accepted responsibility for the payments because they were, in effect, receiving a sweetheart deal: “[a]fter paying the consumer marketing expenses that petitioner allocated to them,” several of the supply points enjoyed average ROAs between 215% and 143%, which was higher than any of the nearly one thousand comparable companies identified by the IRS. Indeed, even after the IRS’s transfer pricing adjustments, most supply points remained “in the top quartile of food and beverage companies worldwide in terms of profitability.” Given the supply points’ “astronomical levels of profitability, their agreement to the quid pro quo of absorbing marketing costs [was] . . . perfectly explicable in economic terms: Because the forgone royalty expense would vastly exceed the marketing costs, any rational economic actor would have accepted this deal.”
The court doubled down on this conclusion by explaining that the supply points received a sweetheart deal even after the IRS adjustments:
Under [the IRS’s] bottler CPM, the four Latin American supply points would have ROAs of 34.3%--higher than those enjoyed by 863 of the 996 food and beverage companies in [the IRS] comparison group. The Irish and Swazi supply points would have ROAs of 20.9%--higher than those enjoyed by 794 of the 996 companies in [the IRS] comparison group.
The court then went even further by suggesting that the supply points were in fact entitled to no more than an 8.5% contract manufacturing return on total costs: “As petitioner’s experts conceded, the supply points’ manufacturing activity was a routine activity, benchmarkable to the activities of contract manufacturers and meriting no more than a cost-plus return.”
This hypothetical determination – one not needed to accept the IRS adjustments – arguably went too far as it provides no reward to the supply points for their very substantial payments of marketing expenses. It does seem inexplicable that the supply points would agree to incur these expenses without some return. The question then becomes “how much” is enough?
Other Determinations by the Court
Prior Deals Do Not Bind the IRS
As a threshold argument, Coke US contended that the IRS’ adjustments for 2007-2009 were inherently arbitrary because they deviated from the terms of a 1996 closing agreement in which the IRS agreed to Coke US’ use of the 10-50-50 method. Coke US was not the first taxpayer to raise this kind of argument in a transfer pricing case—Medtronic, for example, presented a similar argument to the Tax Court, with some success—but it may prove to be the last. The Tax Court firmly rejected the notion that the prior deal reflected anything more than a one-off dispute settlement.
As the Tax Court observed, closing agreements are governed by contract law and, absent ambiguity, must be construed in accordance with their express terms. Thus, Coke US’ extrapolation from the closing agreement of “factual underpinnings” to which the IRS had allegedly agreed (e.g., that the supply points played an important role in generating consumer demand and were accordingly entitled to a share of nonroutine profit) was unpersuasive. “[N]othing within the four corners of the closing agreement” suggested any such meeting of the minds with regard to the supply points’ functional profiles or profit entitlements. Rather, the parties had simply agreed to a formula “in settling the dispute before them at that moment.”
Likewise fatal to Coke US’ argument was the closing agreement’s silence regarding “the transfer pricing methodology that was to apply for years after 1995.” By its terms, the agreement applied only to Coke US’ 1987-1995 tax years. The agreement did provide that, if Coke US continued to apply the 10-50-50 method for years after 1995, it would be considered to have met relevant exceptions for accuracy-related penalties. As the Tax Court noted, however, this provision demonstrated that “the parties knew how to make the closing agreement conclusive for future years when they wished to do so”—and with respect to the appropriateness of the 10-50-50 method, they made no such choice. Indeed, the provision could be viewed as foreshadowing future IRS transfer pricing adjustments if the 10-50-50 method were continued. In sum, the court refused to allow Coke US “to estop the Government on the basis of a promise the Government did not make.”
The message for taxpayers is sobering. A taxpayer executing a closing agreement at the end of a transfer pricing examination faces a difficult decision. Typically, the agreement will have imposed adjustments to a taxpayer’s as-filed positions, so the taxpayer should reasonably not revert to its prior practice, even if that practice was based on in-depth functional, comparability, and economic analyses prepared by qualified experts. Yet, neither can a taxpayer reasonably rely on the closing agreement TPM because the IRS reserves the right to change its theory and approach—“gotcha!”—for later years.
Moreover, the Court’s analysis elides a nuance of Coke US’ argument. The IRS may not have been estopped or otherwise legally bound by the terms of the closing agreement for years after 1995, but its entry into the agreement, followed by acquiescence to Coke US’ application of the 10-50-50 method for its 1996-2005 tax years, all of which were audited, suggests a certain caprice in the IRS’ rejection of that method in the 2007-2009 audit cycle. To be sure, a reader of the opinion cannot know what may have been communicated by the IRS to Coke US in the intervening years. However, the Tax Court’s opinion divulges no material change in the facts or the law that could account for the IRS’ change in position, or that could have put Coke US on notice that the 10-50-50 method would no longer pass muster. The IRS may not have reneged on a prior deal, but from the taxpayer’s perspective (at least), the IRS’ attack on the 10-50-50 method after 20 years of express or implicit assent was quintessentially “arbitrary and capricious.”
Aggregation Applies to Transactions, Not Taxpayers
The regulations under section 482 generally call for a transaction-by-transaction analysis of dealings among related parties. Where, however, two or more “transactions, taken as a whole, are so interrelated that an aggregate analysis of the transactions provides the most reliable measure of an arm’s length result,” such transactions should be analyzed in the aggregate. Coke US, however, took the aggregation principle a step further, in effect asserting that taxpayers should be aggregated for purposes of analyzing the company’s royalty arrangements.
In its pleadings, Coke US urged the Tax Court to focus broadly on the activities of its foreign business units, management reporting constructs primarily led by employees of the ServCos, in lieu of analyzing the supply points as a discrete set of “controlled taxpayers,” separate from the ServCos. Similarly, all three expert witnesses who presented alternatives to the IRS’ bottler CPM on Coke US’ behalf constructed and applied their transfer pricing analyses to transactions between Coke US and an aggregate of its foreign affiliates (“the Field”). Coke US’ arguments, and those of its experts, conflated the functions, assets, and risks of the supply points with those of the ServCos. The supply points had acquired immensely valuable marketing intangibles, Coke US asserted, because marketing expenses incurred by the ServCos through activities of their employees were cross-charged to the supply points. In general, Coke US’ experts validated the supply points’ supra-normal profits by treating the supply points as performing all activities performed by all foreign affiliates (and, in at least one case, some activities performed by Coke US, itself).
The Tax Court dispensed with this argument quickly. As the Court explained, the proper unit of analysis under the 482 regulations is a “controlled taxpayer”—here, each of the supply points. Applying the rules to an aggregate of all Coke US foreign affiliates, or to “boxes on an organizational chart,” would “ignore the separate legal and taxable entities involved.” This welcome reaffirmation of the separate entity principle for U.S. transfer pricing purposes contrasts somewhat with global tax policy trends. To the extent it implies that distinct treatment of legally separate foreign affiliates flows directly and definitively from the regulations under section 482, however, the opinion oversimplifies the U.S. regulatory landscape.
For the years at issue, the regulations permitted an aggregate analysis of closely interrelated transactions. A temporary regulation in effect for certain later years requires that transactions be analyzed in the aggregate in certain instances, with the scope of aggregation delimited only by the mandate that the selected mode of analysis be that which provides the “most reliable measure of an arm’s length result.” Regulations under section 482 also emphasize the centrality of comparability to the transfer pricing analysis. For transfer pricing methods based on the results of transactions between uncontrolled parties (like the IRS’ bottler CPM and Coke US’ comparable uncontrolled transaction method), reliability “depends on the degree of comparability between the controlled transaction or taxpayers and the uncontrolled comparables,” taking into account, inter alia, their respective economic circumstances and contractual arrangements. To achieve appropriate comparability in the analysis of a particular transaction between particular controlled taxpayers, it may be necessary to consider the effects of other transactions involving other controlled taxpayers.
The Coca-Cola System as described in the Tax Court’s opinion seems a paradigmatic example of interrelated transactions most reliably evaluated as a whole. And, indeed, the court examined the relevant functions, assets, risks, contractual relations, and economic circumstances of the entire multinational group, including the ServCos, in conducting its functional analysis. Only when it came to the mechanics of pricing did the court narrow its focus to the supply points. The ultimate question before the court concerned the arm’s length royalty due to Coke US for the supply points’ use of its intangible property. The IRS’ bottler CPM (and the Tax Court’s opinion) approached this question in the reverse, evaluating the profit retained by the supply points after payment of the royalty. This approach was consistent with accepted modes of transfer pricing analysis where one controlled party is much less complex than the other and makes only routine (i.e., benchmarkable) contributions. Because the court had already dismissed Coke US’ assertion that the supply points owned valuable intangible assets, the supply points—which operated as routine contract manufacturers—were the appropriate unit of analysis.
The Coke US-Field dichotomy makes more sense in the context of the three alternative TPMs proposed by Coke US’ experts, all of which presupposed that the Field had acquired non-routine intangibles through advertising, marketing, and other expenditures and marketing activities. Had this premise been found correct, the question before the court would not have been turned on its head. The selected TPM would need to evaluate whether Coke US—the principal entrepreneur for the controlled group and the U.S. taxpayer whose income had been adjusted—had reported an appropriate amount of system profit as compensation for the whole of its functions performed, assets employed, and risks assumed. Distinctions between the supply points and ServCos would not have been as relevant to that inquiry. It may then have been appropriate to consider Coke US’ foreign affiliates in the aggregate, for example, in applying a RPSM. Ultimately, Coke US lost on the issue of whether the supply points or “the Field” was the appropriate unit of analysis because the Tax Court didn’t buy its marketing intangibles theory.
Foreign Tax Credit Questions
The Coca-Cola decision is important not only for its discussion of transfer pricing issues but for several ancillary issues as well. The most important of these are certain foreign tax credit questions common to transfer pricing disputes and the correlative consequences under section 987 of a reallocation of income pursuant to section 482.
Both issues stemmed from a reallocation from the Mexican branch of a U.S. subsidiary to Coke US. Such reallocations within a consolidated return are relatively rare, but they do occur.
Approximately three years prior to the major opinion rendered on November 18, 2020, the Tax Court released a separate opinion addressing the foreign tax credit questions. Those questions arose from an argument by the Internal Revenue Service that, in light of its proposed challenge to the amount of royalties paid to the parent by the Mexican branch, there had been an excess — and therefore noncompulsory and non-creditable — payment of Mexican tax by the branch. Coke US responded that the branch’s method of computing royalties had been in place for many years prior to the IRS challenge, that it had received advice from a reputable Mexican tax attorney to the effect that its method was acceptable for Mexican tax purposes, and that it had been told any greater amount of royalties would not be allowed as a deduction by local tax authorities. Coke US had attempted to raise the issue in a Mutual Agreement Procedure under the tax treaty between the United States and Mexico, but IRS had refused to participate in the proceedings because the pending transfer pricing case had been designated for litigation.
Upon Coke US’ motion for summary judgment, the Tax Court ruled in its favor. The Court observed that regulations relating to “non-compulsory payments” of foreign taxes allow a taxpayer to rely on advice obtained in good faith from a competent tax professional. There was no reason at this juncture to suppose that IRS’s proposed reallocation of income would be sustained or, if it was, that Mexico would agree with it. Taxpayers are not required to pursue refund claims that would be futile. Furthermore, Coke US could hardly be faulted for failing to seek relief under the treaty when IRS had refused to participate in discussions with Mexican tax authorities. The law allows taxpayers to claim a foreign tax credit when foreign tax is paid, even if the tax is disputed or may be disputed in the foreign country.
The Service next argued that Coke US should not be allowed a credit until after the transfer pricing issues had been decided. It maintained that if the decision went in the Government’s favor, Coke US would have no incentive to pursue competent authority proceedings at that time. The Court found the suggestion of a deferred foreign tax credit not to be in accord with “the procedure that Congress envisioned when it enacted the Code.” If the proposed reallocation of the branch’s income was eventually upheld, the Government could, if necessary, seek competent authority proceedings on its own. And if Mexico should agree in whole or part with a reallocation, the foreign tax could be refunded, and the foreign tax credit reduced, in accordance with the procedures envisioned by section 905(c).
In short, there was no reason in 2017, prior to the transfer pricing adjudication, to conclude that the payment of Mexican tax was anything other than compulsory, and no reason to believe the Government would be whipsawed if it later prevailed on the transfer pricing reallocation and Mexico eventually refunded some or all of the tax. The decisions on each of these points are lucid, sensible, and clearly correct. They do not make new law, but they do provide a clear path through a maze of obfuscatory arguments that IRS deemed appropriate to advance.
The ancillary issue pertaining to foreign exchange gain or loss of the Mexican branch was discussed, but less clearly described, in the November 18, 2020 opinion. Since the functional currency of the branch was the Mexican peso, its income was computed in pesos and translated into dollars at the “appropriate exchange rate” under section 987. Remittances of cash or property from the branch, for example as royalties paid to the parent, would give rise to foreign exchange gain or loss measured by the difference between the historic dollar value of peso earnings and the value of remittances out of those earnings.
Upon the Court’s decision upholding a reallocation of additional royalty income from the branch to the parent, governing regulations called for “appropriate correlative allocations.” One correlative allocation was a reduction of the income of the Mexican branch. This, in turn, resulted in greater exchange losses in 2007 and 2008 but exchange gains in 2009 in an amount exceeding the additional losses for the earlier years.
The calculations are, as the Court noted, “mechanical (albeit complex)” and “follow more or less automatically once correlative allocations have been made to the income of the Mexican branch.” Coke US argued there could be no section 987 adjustments because the primary transfer pricing reallocation (greater income for the parent) had no tax consequences, since income of both the branch and the parent fell in a U.S. consolidated return both before and after the reallocation. The Court rejected this argument as posing no jurisdictional bar and as “irrelevant” to the Court’s determination. Coke US’ attempt to analogize the section 987 adjustment to the foreign tax credit issue was also rejected, since the “compulsory” nature of foreign tax paid bore no relevance to either a correlative reduction of the branch’s income or the foreign exchange consequences of that reduction. Nor was the adjustment premature. The IRS could wait to make correlative adjustments until the Court’s decision on a primary transfer pricing adjustment became final, but nothing in the regulations precluded it from making those adjustments in the notice of deficiency, as it had done in this instance.
The calculations underlying the section 987 adjustments do not appear in the Court’s opinion. It may be supposed that they flowed from one or more of (a) lower peso earnings of the branch as a correlative result of the transfer pricing reallocation from the branch; (b) foreign exchange consequences of an additional remittance from the branch in the form of the reallocation of income to the parent; and (c) use of a different dollar basis for that remittance than the one employed for the pre-allocation remittance.
The decision is, again, clearly correct on the law. The section 987 issue is only relevant in the event of a primary transfer pricing adjustment reallocating income from a U.S. company with a branch employing a foreign functional currency to another U.S. company. As noted, however, such reallocations are not unknown.
In transfer pricing, facts matter. The Tax Court’s conclusions in the Coca-Cola case were fact-specific and fact-intensive. The CPM applied because it was the most reliable method in the circumstances. A DEMPE-like analysis supporting the CPM made sense because of the limited risk faced by the supply points and the court’s conclusions regarding legal ownership of intangibles. Agree or disagree with the decision, it now sets the standards for future courts who might wish to adopt the CPM to back into prices for valuable intangibles.
© Copyright 2021 IBFD: No part of this information may be reproduced or distributed without permission of IBFD.
Disclaimer: IBFD will not be liable for any damages arising from the use of this information.
 The Coca-Cola Company & Subs. v. Commissioner, 155 T.C. No. 10 (Nov. 18, 2020) (“COCA-COLA”).
 The “IRS . . . made adjustments that increased petitioner’s aggregate taxable income by more than $9 billion,” and the court agreed that the IRS’s reallocations should be offset by $1.8 billion in dividends that subsidiaries had paid. COCA-COLA at 6; see also COCA-COLA at 219.
 Reuven S. Avi-Yonah and Gianluca Mazzoni, Coca-Cola: A Decisive IRS Transfer Pricing Victory, at Last, 100 Tax Notes Int'l 1419 (Dec. 14, 2020).
 DEMPE stands for “development, enhancement, modification, protection and exploitation” of intangibles. OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations 2017, Chapter VI.B.2 (“OECD Guidelines”), ¶ 6.48. While U.S. Treasury officials have long indicated that DEMPE is implicit in U.S. transfer pricing rules, many commentators have questioned the assertion. See Ryan Finley, BEPS Reports Retain Concept of Control, Treasury Official Says, 80 Tax Notes Int’l 231 (Oct. 19, 2015) (“The OECD's final base erosion and profit-shifting reports do not fundamentally depart from the prior OECD guidelines or the U.S. section 482 regulations, according to Brian Jenn, attorney-adviser, Treasury Office of International Tax Counsel. . . . He said the concept of control adopted in the final report on actions 8 through 10 was already included in chapter 9 of the OECD guidelines on business restructurings and that the concept of economic substance in the section 482 regulations implicitly applies a similar standard.”); Mark Martin, Mark Horowitz, and Thomas Bettge, INSIGHT: Transfer Pricing Substance in Flux—DEMPE, BEPS 2.0, and Covid-19, Bloomberg Law Daily Tax Report (July 31, 2020) (“[T]he IRS (including exam teams, the Transfer Pricing Practice, and the Advance Pricing and Mutual Agreement program) has begun to make arguments based on these OECD substance rules, or at least consider these issues. . . . At least for the moment, U.S. transfer pricing law contains no analogue to the DEMPE rules, and continues to respect contractual assumption of risks except in extreme cases.”)
 Given the number of issues raised in the case and in the court’s opinion, our treatment necessarily omits some points, such as the “dividend offsets” issue, that may interest particular readers. We also do not address contentions relating to research efforts and the Brazilian supply point, which Judge Lauber rejected based on application of the developer-assister rules. In addition, Judge Lauber did not yet address blocked income issues relevant to the Brazilian supply point because the validity of the blocked income regulation is currently before the Tax Court in 3M Co. & Subs. v. Commissioner, T.C. Dkt. No. 5816-13 (filed Mar. 11, 2013).
 See Canada v. GlaxoSmithKline Inc., 2012 SCC 52,  3 S.C.R. 3 (Can.).
 Section 482 is the main transfer pricing provision of the U.S. Tax Code, known as the Internal Revenue Code of 1986, as amended (the “Code”).
 COCA-COLA at 14 (“[Coke US] initially established affiliates in virtually every country to manufacture and supply concentrate.”).
 COCA-COLA at 18.
 COCA-COLA at 20.
 COCA-COLA at 20.
 COCA-COLA at 21.
 COCA-COLA at 7.
 COCA-COLA at 15.
 COCA-COLA at 156.
 COCA-COLA at 50–51.
 A functional analysis of the bottlers is important to the decision because the IRS used the bottlers as CPM comparables for the supply points.
 COCA-COLA at 41, 62.
 COCA-COLA at 57.
 COCA-COLA at 11.
 COCA-COLA at 8–9.
 COCA-COLA at 9.
 See Ryan Finley, U.S. Tax Court’s Coca-Cola Ruling: Early Sign of a New Approach?, 100 Tax Notes Int’l 1279. 1280 (Dec. 7, 2020) (“Unlike its predecessors, the Coca-Cola decision roundly rejects the taxpayer’s favored comparable uncontrolled transaction method analysis in favor of the IRS’s comparable profits method analysis.”); Aysha Bagchi, Isabel Gottlieb and Jeffery Leon, Coca-Cola Ruling Sends Warning on Facebook, Medtronic Tax Fights, Bloomberg Law Daily Tax Report (Nov. 20, 2020) (“While courts have traditionally favored transactional methods, the [Coca-Cola] ruling . . . ‘at least breathes some life back into CPM as a viable method.’” (quoting Professor Blaine Saito)); Ryan Finley, A Look Ahead: Transfer Pricing to Remain in Focus in 2021, 101 Tax Notes Int’l 16 (Jan. 4, 2021) (noting that the Coca-Cola case was “a rare almost-complete win for the IRS in a transfer pricing case,” “provided a strong vindication of the IRS’s interpretation of the regulations’ best method and comparability standards,” and “upheld the IRS’s selection and application of the comparable profits method.”)
 See COCA-COLA at 150–51. Taxpayers that have typically taken the opposite position, fighting transfer pricing adjustments to their non-U.S. marketing and distribution affiliates predicated on their supposed ownership of valuable local marketing intangibles, will find irony in Coke US’ contention.
 COCA-COLA at 152.
 Treas. Reg. § 1.482-4(f)(3)(i)(A). As the Tax Court observed, this regulation was initially adopted in temporary form and made applicable for tax years beginning after December 31, 2006, so the temporary regulation applied for the years at issue. The final regulation was later finalized without change and made applicable for tax years beginning after July 31, 2009. Similarly, under the OECD Guidelines, “[t]he legal owner will be considered to be the owner of the intangible for transfer pricing purposes,” but “[i]f no legal owner of the intangible is identified under applicable law or governing contracts, then the member of the [multinational enterprise] group that, based on the facts and circumstances, controls decisions concerning the exploitation of the intangible and has the practical capacity to restrict others from using the intangible will be considered the legal owner of the intangible for transfer pricing purposes.” OECD Guidelines, ¶ 6.40.
 COCA-COLA at 156.
 COCA-COLA at 157.
 See COCA-COLA at 157–58.
 See COCA-COLA at 158–59.
 COCA-COLA at 159.
 COCA-COLA at 13.
 COCA-COLA at 159 n.47.
 Id. Goodwill, of course, is now expressly included in the definition of intangible property in section 367(d)(4), as are going concern value, workforce in place, and any “other item the value or potential value of which is not attributable to tangible property or the services of any individual.” Yet, goodwill is not legally protectable, often not expressly documented or protected by contract, and not reflected on a taxpayer’s books. Given the Tax Court’s reading of the regulatory standard for legal ownership—no pertinent law or contract assigning ownership, no intangible—it is unclear how a taxpayer would justify remunerating a foreign affiliate for value attributable to intangibles such as goodwill. This problem is compounded for more esoteric intangibles, transfers of which the IRS nonetheless deems compensable. See, e.g., Amazon.com v. Commissioner, 148 T.C. 108, 157–59 (2017) (explaining that IRS’ enterprise valuation of intangibles transferred in cost-sharing buy-in (improperly) captured value of “‘growth options’ and corporate ‘resources’ and ‘opportunities’” that “cannot be bought and sold independently [and] . . . are an inseparable component of an enterprise's residual business value”), aff’d 934 F.3d 976 (9th Cir. 2019).
 COCA-COLA at 159, n.47.
 COCA-COLA at 167.
 COCA-COLA at 160.
 COCA-COLA at 160–61.
 COCA-COLA at 164 (based on the 11th Circuit decision in Peterson v. Commissioner, 827 F.3d 968, 987–88 (11th Cir. 2016) (applying Commissioner v. Danielson, 378 F.2d 771 (3d Cir. 1967) (en banc) and Spector v. Commissioner, 641 F.2d 376 (5th Cir. 1981) (adopting the Danielson rule.
 See supra note 26.
 Amazon.com v. Commissioner, 148 T.C. 108 (2017), aff’d 934 F.3d 976 (9th Cir. 2019); Medtronic Inc. v. Commissioner, T.C. Memo. 2016-112, vacated, 900 F.3d 610 (8th Cir. 2018); Veritas Software Corp. v. Commissioner, 133 T.C. 297 (2009), nonacq. AOD 2010-05.
 See Ryan Finley, U.S. Tax Court’s Coca-Cola Ruling: Early Sign of a New Approach?, 100 Tax Notes Int’l 1279, 1282 (Dec. 7, 2020) (“The court’s recognition of this distinction [regarding CUT method unreliability] — along with its corresponding rejection of the hierarchy-of-methods approach that the regulations no longer follow — contrasts sharply with its prior decisions, arguably suggesting the possibility that its near-universal preference for transactional methods may be eroding.”); see also supra note 26.
 COCA-COLA at 109.
 COCA-COLA at 173–74. For example, Coke US’ consolidation of concentrate manufacturing capacity in its Irish supply point might appear to provide the Irish supply point with negotiating leverage. However, as the court noted, Coke US constructed a new concentrate plant in Singapore and “caused the Irish supply point to ship to Singapore 30 containers” of equipment, illustrating that “[e]very supply point faced the perpetual risk that its production would be shifted, in whole or in part, to a competing supply point.” Id. at 29, 174.
 COCA-COLA at 171. We discuss below whether this determination amounts to acceptance of the OECD’s DEMPE doctrine and, if so, whether the court properly applied that doctrine here.
 COCA-COLA at 106–07 (citing T.D. 8552, 1994-2 C.B. 93, 109).
 COCA-COLA at 107–08.
 COCA-COLA at 115.
 For example, a proposed CUT method, based on master franchising agreements, failed because: “At no point in his analysis did [Coke US’ expert] compare the respective contractual terms under which the supply points and his master franchisees operated.” COCA-COLA at 196. Similarly, the premise for Coke US’ proposed residual profit split method (“RPSM”) “is that the supply points owned intangible assets. . . . But the supply points were neither ‘[t]he legal owner[s] * * * of [intangible property]’ nor ‘holder[s] of rights constituting * * * [intangible property] pursuant to contractual terms.’” COCA-COLA at 202.
 148 T.C. 108 (2017), aff’d 934 F.3d 976 (9th Cir. 2019).
 Amazon, 148 T.C. at 156–57.
 Amazon, 148 T.C. 163–64.
 COCA-COLA at 152.
 OECD Guidelines ¶ 6.48.
 OECD Guidelines ¶ 6.42.
 OECD Guidelines ¶ 6.53.
 OECD Guidelines ¶ 6.59 (“See the guidance in…paragraphs 1.85 and 1.103, which illustrate a situation where the party providing funding does not control the financial risk associated with the funding.”); see also OECD Guidelines ¶ 1.103 (If a company “does not have the capability to make decisions to take on or decline the financing opportunity, or the capability to make decisions on whether and how to respond to the risks associated with the financing opportunity,” then the company “would not be entitled to any more than a risk-free return.” (footnote omitted))
 COCA-COLA at 128.
 COCA-COLA at 186.
 COCA-COLA at 129.
 COCA-COLA at 152.
 COCA-COLA at 52.
 COCA-COLA at 152, 131, 167, 170.
 COCA-COLA at 55–56.
 COCA-COLA at 152.
 COCA-COLA at 131. In provisions particularly relevant to the Coca-Cola case, the Guidelines discuss the importance of control and risk when a related party performs marketing functions for a party that owns a trademark. A related party “being reimbursed for its promotional expenditures and being directed and controlled in its activities by the owner of the trademarks and other marketing intangibles” generally would “not be entitled to additional remuneration” under the Guidelines, because the related party “does not assume the risks associated with the further development of the trademark and other marketing intangibles.” OECD Guidelines ¶ 6.77
 OECD Guidelines ¶ 6.63.
 OECD Guidelines ¶ 6.60, 6.61.
 COCA-COLA at 69; see also COCA-COLA at 126, 128.
 OECD/G20 Base Erosion and Profit Shifting Project, Aligning Transfer Pricing Outcomes with Value Creation, Actions 8-10 - 2015 Final Reports, at 11.
 COCA-COLA at 171.
 COCA-COLA at 172.
 COCA-COLA at 171.
 COCA-COLA at 172. As explained above, Coke US identified an 8.5% return on total costs as an arm’s length contract manufacturing return. The Tax Court accepted that analysis: “[The supply points’] return is almost seven times higher than the 8.5% return that represents the arm’s-length value of the manufacturing activities they performed.” COCA-COLA at 111.
 See Medtronic, Inc. v. Commissioner, T.C. Memo. 2016-112 (where taxpayer argued IRS’ adjustments were arbitrary in light of the terms of a closing agreement for earlier years, after rejecting the IRS’ proposed TPM and taxpayer’s proposed TPM, constructing a TPM that produced results eerily similar to those under the closing agreement).
 COCA-COLA at 98.
 COCA-COLA at 95.
 COCA-COLA at 94.
 COCA-COLA at 96.
 COCA-COLA at 97.
 COCA-COLA at 98.
 A taxpayer desiring prospective certainty could, of course, apply for an advance pricing agreement (APA). See generally Rev. Proc. 2015-41, 2015-35 I.R.B. 263.
 COCA-COLA at 227.
 One factual change disclosed in the Tax Court’s opinion—Coke US’ shift of substantial production to the Irish supply point, which reported a 1.4% income tax rate during the years at issue, from the Mexican supply point, the results of which were included in Coke US’ U.S. return, see COCA-COLA at 28—would not, alone, justify a change in the TPM.
 See generally Treas. Reg. § 1.482-1(a), (b).
 See Treas. Reg. § 1.482-1T(f)(2)(i)(B). For the years at issue, a final regulation provided that “[t]he combined effect of two or more separate transactions . . . may be considered, if such transactions, taken as a whole, are so interrelated that consideration of multiple transactions it the most reliable means of determining the arm’s length consideration for the controlled transactions.” Treas. Reg. § 1.482-1(f)(2)(i)(A) (prior to September 2015). The temporary regulation expired in 2015 without being finalized. Post-TCJA, section 482 itself expressly permits aggregation in the valuation of transfers of intangible property.
 COCA-COLA at 100.
 COCA-COLA at 152.
 See COCA-COLA at 192, 198, 207.
 See COCA-COLA at 98–100.
 See COCA-COLA at 100–101.
 See generally OECD (2020), Tax Challenges Arising from Digitalisation – Report on Pillar One Blueprint: Inclusive Framework on BEPS, OECD/G20 Base Erosion and Profit Shifting Project, OECD Publishing, Paris https://doi.org/10.1787/b3ba0634-en. U.S. tax policy is not wholly free of infringements on the separate entity principle. See I.R.C. § 951A (providing for calculation of a U.S. taxpayer’s global low-taxed intangible income (GILTI) by netting the tested income and tested loss of all of taxpayer’s controlled foreign corporations (CFCs), by aggregating the qualified business asset investment of all tested income CFCs, and by aggregating the specified interest expense of all CFCs).
 See COCA-COLA at 98–100.
 Treas. Reg. § 1.482-1T(f)(2)(i)(B); see also Treas. Reg. § 1.482-1(c)(1) (“The arm’s length result of a controlled transaction 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)(2)(I), (d)(3)(ii), (iv).
 See, e.g., former Treas. Reg. § 1.482-1(f)(2)(i)(B) Examples 2–3 (where multiple members of a controlled group make interrelated contributions to a computer manufacturing and sales business, “it may be appropriate to consider the combined effects” of separate transactions between group members in evaluating particular transfer prices, and similarly interrelated transactions among another group of controlled taxpayers may provide a more reliable measure of the arm’s length result than wholly independent transactions among uncontrolled taxpayers); Treas. Reg. § 1.482-1T(f)(2)(i)(E) Examples 2-3 (same); Canada v. GlaxoSmithKline Inc., 2012 SCC 52 (holding that, under then-applicable Canadian transfer pricing law and the OECD’s Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (OECD Guidelines), lower court could—and should—have considered the existence, terms, and conditions of a license agreement between the Canadian taxpayer and its U.K. parent in evaluating the transfer prices it paid to a Swiss affiliate for supplies used in manufacturing products under the license).
 See COCA-COLA at 18–77.
 In general, where one party performs only routine functions, bears routine risks, and holds no non-routine intangibles, that party is selected as the “tested party,” and the transfer pricing analysis seeks to determine whether the tested party’s income from controlled transactions or activities is arm’s length. See, e.g., Treas. Reg. § 1.482-5(b)(2)(i).
 For example, in Canada v. GlaxoSmithKline Inc., where the Canadian taxpayer was in effect the tested party, the Supreme Court of Canada held that obtaining “a realistic picture of the profits of Glaxo Canada” required acknowledging that “Glaxo Canada was paying for at least some of the rights and benefits under the License Agreement as part of the purchase prices for ranitidine from Adechsa.” Canada v. GlaxoSmithKline Inc., 2012 SCC 52,  3 S.C.R. 3 ¶¶ -.
 Coca-Cola Co. & Subs. v. Commissioner, 149 T.C. 446 (2017).
 See Treas. Reg. § 1.901-2(e)(5)(i).
 IBM Corp. v. United States, 38 Fed. Cl. 661 (1997); Rev. Rul. 70‑290, 1970‑1 C.B. 160.
 Coca-Cola, 149 T.C. at 462.
 Treas. Reg. § 1.987‑5(d).
 Treas. Reg. § 1.482‑1(g)(2)(i).
 COCA-COLA at 217.