With the long-awaited entry of a final Tax Court decision in Coca-Cola Co. v. Commissioner, 155 T.C. 145 (2020), the company’s appeal to the Eleventh Circuit can finally begin, and the stakes could hardly be higher.
It’s been a while since the Tax Court released its November 2020 opinion deciding the transfer pricing methodological issues raised in Coca-Cola. (Prior coverage: Tax Notes Int’l, Nov. 23, 2020, p. 1097; prior analysis: Tax Notes Int’l, Dec. 7, 2020, p. 1279.) But after an overhaul of the company’s legal strategy and representation, an eccentric motion for reconsideration, a supplementary opinion on the blocked income regulations (T.C. Memo. 2023-135), and months’ worth of Rule 155 calculations, Judge Albert Lauber entered a final decision in the case on August 2. (Prior coverage: Tax Notes Int’l, Aug. 12, 2024, p. 1103.) So begins the 90-day window for the long-awaited appeal to the Eleventh Circuit, which Coca-Cola has made abundantly clear it intends to promptly pursue. (The Coca-Cola Co., “Coca‑Cola to Move Forward With Appeal” (Aug. 2, 2024).)
The monetary stakes in Coca-Cola are staggering. Lauber’s final decision, which concerns only the 2007 through 2009 tax years, resulted in a total tax liability of about $2.7 billion and a further $3.3 billion in accrued interest. According to Coca-Cola’s most recent Form 10-Q filing with the SEC, the company’s total tax and interest exposure would rise to about $17 billion through the first half of 2024 if the method endorsed by the Tax Court were applied to all subsequent tax years. The broader effects of the case’s outcome on the integrity of the section 482 regulations, and tax administration in general, are orders of magnitude greater.
However, despite what others may consider unfavorable circumstances, Coca-Cola evidently remains convinced that it will eventually be vindicated. The company has projected complete confidence through its public relations announcements, continued use of the method rejected by the Tax Court, and strikingly small provision for income taxes under Accounting Standards Codification 740. Coca-Cola remains so confident in its position that, according to its most recent Form 10-Q, the company recorded only a $456 million ASC-740 reserve against its $17 billion tax and interest exposure.
To put it in simplified sports-betting terms, a $456 million reserve translates into a -3723 “moneyline” and a 97.4 percent implied probability of victory. In other words, the company believes it has roughly Tyson-over-Spinks odds of success. It may well be that “fortune favors the bold,” but it’s equally true that fortune disfavors the reckless. Clinging to a transfer pricing approach firmly rejected by the Tax Court in 2020 as of June 2024, and incurring a further $1 billion in exposure every year the practice continues, seemingly strays far past the boundary separating the two.
Bold, reckless, or otherwise, Coca-Cola’s defiant response to the Tax Court’s 2020 decision was apparent almost immediately. The company announced in January 2021 that J. Michael Luttig, former judge for the Fourth Circuit, would serve as a counselor and special adviser in the ongoing tax dispute and eventual appeal. The announcement praised Luttig, perhaps best known as a conservative constitutional law expert and star witness during House hearings on the January 6, 2021, Capitol Hill riot, for his experience in corporate law and complex litigation. No reference to Luttig’s transfer pricing expertise found its way into the announcement, and Coca-Cola’s post-opinion Tax Court filings suggest why.
Joining Luttig in spearheading Coca-Cola’s post-2020 tax litigation efforts is Harvard Law School’s Laurence Tribe, another legal expert whose impressive CV is heavy on constitutional law and light on anything relevant to transfer pricing. Coca-Cola’s reconstituted legal team still included some proven section 482 litigators, but the company seemingly left its ultimate strategic direction and public fanfare for lawyers with minimal transfer pricing litigation experience.
Every indication is that Coca-Cola will largely rely on its new legal strategy in the forthcoming appeal. It offered a detailed preview of this strategy in a June 2021 motion for reconsideration, which the company requested leave to file months after the normal 30-day window to do so had expired. (Prior coverage: Tax Notes Int’l, June 14, 2021, p. 1554; Tax Notes Int’l, Dec. 6, 2021, p. 1170.) The motion’s focus was on constitutional issues allegedly overlooked by the Tax Court in 2020, but it also tried to address some of the section 482 regulatory interpretation questions at stake. The result was about what one would expect from a team of esteemed legal experts in entirely different areas of law who just spent a weekend reading the transfer pricing regulations for the first time. (Prior analysis: Tax Notes Int’l, July 26, 2021, p. 422; Tax Notes Int’l, July 26, 2021, p. 425.)
In many ways, the Coca-Cola appeal is a perfect storm of timely and important transfer pricing issues. It poses pressing questions regarding the viability of a method that at least imposes a limit on the profit that can be offshored through controlled cross-border transactions. The case also raises critical questions about the extent of taxpayers’ reliance interests, if any, in informal IRS representations and the validity of the blocked income regulations under reg. section 1.482-1(h)(2). This article deals only with the methodological issues at stake in Coca-Cola; other questions will be the subject of a subsequent article.
Overblown Intangibles
According to Coca-Cola’s 2021 reconsideration motion, the Tax Court legally erred by failing to recognize that the company’s foreign supply points held intangible property through their intercompany licenses with Coca-Cola. These licenses, which were the mispriced transactions at issue in the case, gave the supply points rights to produce beverage concentrate using Coca-Cola’s proprietary product formulas and manufacturing process know-how. They also conveyed rights to sell the concentrate to independent bottling companies, which sold finished products to retailers and distributors, under Coca-Cola’s trademarks and brand names. The failure to recognize these licensee rights as IP led the Tax Court to erroneously accept the IRS-favored comparable profits method, the motion argued, which by design assigns a routine return to the controlled entity chosen as the “tested party.”
The underlying assumption was that any intangible ownership is a per se disqualifier for a potential tested party candidate, meaning the CPM couldn’t have been the best method under the circumstances. As Coca-Cola’s motion repeatedly claims, the “CPM applies only if no intangible assets contribute to a company’s revenue and profits.” Because the supply points held something considered an “intangible” under reg. section 1.482-4(b), Coca-Cola’s motion insisted, the CPM couldn’t have been reliably applied using the supply points as tested parties.
This argument is based on a caricature of reg. section 1.482-5(b)(2)(i), which states that the tested party under the CPM typically “will not own valuable intangible property or unique assets that distinguish it from potential uncontrolled comparables” (emphasis added). (Prior analysis: Tax Notes Int’l, July 8, 2024, p. 153; Tax Notes Int’l, Nov. 21, 2022, p. 934.) Reg. section 1.482-5 would be completely inert if intangible ownership were an absolute disqualifier because virtually all enterprises hold some kind of IP. And the independent bottlers used in the IRS’s CPM analysis were no exception: The Tax Court found that they owned or controlled valuable distribution networks, sales forces, and customer lists. The real question under the regulations is whether the tested party’s intangibles are “nonroutine,” which means they have no equivalents among potentially comparable independent enterprises.
The flaw in Coca-Cola’s reasoning was particularly glaring given the specific intangible asset at issue. It would come as an unwelcome surprise to transfer pricing practitioners, most of whom often apply the CPM using licensees as the tested parties, to learn that they’ve been using disqualified tested parties for years. Licenses and the rights they convey undeniably qualify as IP, as recognized in both section 367(d)(4)(D) and reg. section 1.482-4(b)(4). But of all the forms of IP that could disqualify a potential tested party, licensee rights are among the least likely.
If tested parties really can’t have any intangibles at all, including contractual rights acquired under a license, then it would make little sense for reg. section 1.482-4(a)(2) to include the CPM as a specified method for IP transfers. Licenses of IP account for the vast majority of controlled IP transfers that can be reliably priced using the CPM. This makes economic sense: Routine uncontrolled enterprises with limited self-developed intangibles often need to license nonroutine IP rights that they don’t own themselves.
Consistent with this premise, the regulations’ examples clearly indicate that a controlled licensee can be a routine entity suitable for use as the tested party. In Example 9 under reg. section 1.482-8(b), a U.S. parent licenses rights to use a “unique and highly valuable” manufacturing process to a foreign subsidiary for use in the production of compact discs. Despite the subsidiary’s rights as licensee of the parent’s valuable IP, the example concludes that the CPM is the best method and the subsidiary should be the tested party.
Other instances include Example 1 under reg. section 1.482-4(f)(2)(iii), which deals with periodic adjustments. In the example, a U.S. pharmaceutical company licenses rights to manufacture and regionally market a new, potential blockbuster antimigraine drug to its European subsidiary (Eurodrug). After observing that no uncontrolled licenses of comparable IP exist, the example identifies the CPM as the best method to determine an arm’s-length royalty rate with Eurodrug as the appropriate tested party. Owning unique and valuable IP can disqualify a controlled enterprise from consideration as the tested party, but as these examples make clear, licensing the right to use another party’s IP while performing a routine activity plainly does not.
Coca-Cola’s other intangible ownership argument, which featured more prominently in the Tax Court proceedings than in the 2021 motion, was that the supply points’ marketing expenses made them co-investors in brand development within their geographic territories. These alleged local brand development investments were allegedly vital to maintaining and expanding the local value of Coca-Cola-branded products. And they manifested themselves in so-called marketing intangibles: nonroutine intangible assets that, once created by the supply points’ brand development investments, generated substantial revenue and profit despite their lack of legal protection or accounting recognition.
The problem with this argument was its conflict with Coca-Cola’s intercompany contractual relationships, which expressly confirmed that Coca-Cola Co. or another U.S. entity owned all intangibles associated with the group’s trademarks and associated brand intangibles. U.S. group entities maintained a tight grip on the group’s IP under the group’s intercompany contracts, which reserved sole ownership and control over trademarks and marketing concepts for Coca-Cola Co. or Coca-Cola Export Corp. The marketing intangibles argument also forced the company to make the awkward claim that intangibles worth billions of dollars were invisible from a legal or accounting perspective.
The company has thus advanced its argument largely based on economic substance, undeterred by the unilateral nature of the concept under the section 482 regulations. Under reg. sections 1.482-1(d)(3)(ii)(B), 1.482-1(d)(3)(iii)(B), and 1.482-4(f)(3)(A), only the district director (that is, the IRS) may set aside the controlled parties’ legal ownership and contractual arrangements that lack economic substance. The section 482 regulations grant no reciprocal right to controlled taxpayers, which typically have the unfettered power to allocate assets and contractual rights across group entities (almost always with tax planning considerations in mind) however they see fit. Once they do so, there is little policy justification for letting them ignore the legal rights and obligations they unilaterally put in place. And as Lauber notes, the Eleventh Circuit adopted the related “Danielson rule” in Peterson v. Commissioner, 827 F.3d 968 (11th Cir. 2016).
Coca-Cola’s 2021 reconsideration motion also made a convoluted attempt to reformulate the marketing intangibles argument by relabeling the intangibles “goodwill,” which was squarely in unresolvable conflict with the holding in Amazon.com v. Commissioner, 934 F.3d 976 (9th Cir. 2019), aff’g 148 T.C. No. 8 (2017). Unsurprisingly, the judge who held in Amazon that goodwill isn’t a so-called compensable intangible under reg. section 1.482-4(b) revealed little sympathy for this argument in his order denying the motion. Definitional questions aside, the Tax Court found that no compelling evidence presented at trial corroborated the existence of the valuable marketing intangibles Coca-Cola claimed.
The 2020 opinion’s rejection of these intangible-ownership arguments also preemptively rejected the 2021 reconsideration motion’s criticism of the return on assets (ROA), the profit-level indicator used by the IRS to apply the CPM. As is standard in CPM cases, the motion failed to acknowledge that the returns generated by IP are captured in the ROA’s numerator. But if, as the Tax Court found in Coca-Cola, the tested parties held no significant IP, then excluding IP value from the ROA denominator cannot distort the results either.
Manufactured Risk
In an attempt to reach a similar result by different means, the 2021 reconsideration motion also claimed that the marketing expenses were risky investments that entitled them to greater returns than the CPM can provide. Coca-Cola argued that if the supply points’ marketing expenses weren’t incurred to develop the value of their own licensee rights or other nonroutine intangibles, then they must have been investments to further develop the trademarks owned by Coca-Cola. The supply points were thus “assisters” within the developer-assister rules in the 1968 regulations, and an assister would never make such investments without a reasonable prospect of commensurate returns. Coca-Cola alleged that the Tax Court’s fixation on legal ownership, and its corresponding failure to recognize that risky investments entitle controlled parties to nonroutine returns regardless, was another legal error underlying its acceptance of the CPM.
The suggestion, seemingly lurking beneath the surface in Section II.A.2 of Coca-Cola’s 2021 motion, that contributing to group revenue and profitability in any way entitles a controlled party to greater returns than the CPM can offer has no regulatory basis. Under reg. section 1.482-4(f)(4)(i), the compensation required for contributions to another controlled party’s IP depends on the contributions associated with the comparables. The examples that follow in reg. section 1.482-4(f)(4)(ii) affirm the principle that the sales-generating activities typically performed by licensees and distributors, if properly accounted for in the royalty rate or purchase price, do not require separate compensation. Even if they do, an imputed contract for additional routine marketing services may be sufficient to resolve the issue.
It follows that if the CPM is the best method, as the regulations make clear that it very well may be, then the relevant baseline is the level of marketing expenditure incurred by the independent licensees used as comparables. Example 6 in reg. section 1.482-4(f)(4)(ii), which involves a trademark licensee responsible for marketing the group’s products in the United States, cites the CPM as a potential method for pricing the intercompany royalty. The Eurodrug example in reg. section 1.482-4(f)(2)(iii) specifies that the European subsidiary is a reliable tested party, notwithstanding its duty to market the product in its territory and maximize sales.
For any controlled licensee, performing marketing activities to develop the local market and increase sales is par for the course. Virtually all these entities strive to increase local sales through advertising, marketing, and promotion, and they are generally subject to a contractual obligation or at least an informal expectation to do so. Assuming proper application of the arm’s-length standard, the value of these activities is priced into the royalty rate.
It’s true that incurring unreimbursed marketing investments beyond what otherwise comparable independent licensees would incur, and assuming the potentially nonroutine risk that these investments won’t pay off, could justify a share of nonroutine returns. But this can only be the case if the controlled party’s investments exceed the baseline set by the CPM comparables. This excess over the market baseline is what the regulations mean by “incremental,” a critical qualifier either overlooked or conveniently glossed over in Coca-Cola’s 2021 motion.
Example 4 under reg. section 1.482-1(d)(3)(ii)(C) illustrates this principle, noting that the hypothetical comparable licenses “obligate the licensors and licensees to undertake without separate compensation specified types and levels of marketing activities.” Separate compensation is only warranted to the extent that the licensor’s marketing activities are incremental. If the best method is the CPM, then the activities performed by the comparables establish the baseline for “incremental.” And the tested party’s activities must exceed this baseline to justify “separate compensation,” potentially in the form of nonroutine returns. Whether the Coca-Cola supply points’ marketing expenses were incremental is a factual question the Tax Court answered in the negative, in part based on the similar levels of marketing expenses incurred by Coca-Cola entities and the independent bottlers under an informal 50-50 “true up” policy.
The second relevant point concerns the controlled licensee’s right to a “premium return,” which is an approximate synonym for “residual return” or “nonroutine return” and the antonym of “routine return.” The CPM limits the controlled party’s share of profit to a routine return, and taxpayers have deftly misrepresented these concepts to give the misleading impression that a routine return is no return at all. (Prior analysis: Tax Notes Int’l, Mar. 20, 2023, p. 1593.) But “routine return,” which means only that the return can be benchmarked by market data, says nothing whatsoever about magnitude. And as the Tax Court found in Coca-Cola, a 21 to 34 percent ROA — whether routine, nonroutine, or otherwise — was ample inducement for the supply points to make significant brand development expenses.
The Tax Court found that the seven relevant supply points’ pre-CPM-adjustment ROAs were (with one exception) astronomically high compared with a baseline set of 996 food and beverage businesses. Although the company’s Egyptian supply point was an outlier, the other six supply points’ ROAs were far higher than any other company in the set, which included heavyweights like the Pepsi-Cola Co., Nestlé, and the Coca-Cola group as a whole.
Even after their income was drastically reduced by the IRS’s CPM-based adjustments, the six other supply points were highly profitable. Coca-Cola’s Irish and Swazi supply points each had a CPM-based ROA of 20.9 percent, which put them just outside the 80th percentile of the 996-company set. And Coca-Cola’s even more profitable Latin American supply points had post-CPM ROAs of 34.3 percent, which placed them between the 86th and 87th percentiles of the baseline set.
And even this comparison understates the supply points’ profitability. The 996-company sample consisted of corporate groups, not just their manufacturing subsidiaries, so each ROA observation in the set blended the higher returns associated with IP with the lower returns attributable to routine functions. For example, the ROAs for Pepsi and Nestlé included the returns for their unique and highly valuable IP and the returns on routine manufacturing and distribution. As the Tax Court found, the notion that the supply points were entitled to higher ROAs than Pepsi, Nestlé, and the overall Coca-Cola group without owning any of the group’s core IP strains credulity. The supply points performed routine manufacturing functions and duly recorded their headquarters-assigned share of global expenses, but in almost every case they neither developed nor owned Coca-Cola’s trademarks, brand names, product formulas, or manufacturing process know-how.
Merely carrying out contract manufacturing functions and recording an allocated share of marketing expenses on their books was, according to Coca-Cola, enough to rightfully award the six profitable supply points with ROAs ranging from 94 to 215 percent. Even the lower bound of this range (Mexico’s 94 percent ROA) was above the 99th percentile for the 996-company reference set, and it far exceeded the Coca-Cola group’s 97th-percentile ROA of 55 percent. That the IRS would have the audacity to push these gravity-defying ROAs down to between 21 and 34 percent was, according to Coca-Cola’s 2021 motion, an unmistakable step along the road to serfdom. The Tax Court, understandably, wasn’t particularly moved by this argument.
An even greater potential problem for Coca-Cola is the dubious attribution of the marketing activities and risks that, according to the company, created the supply points’ rights to nonroutine returns in the first place. The Tax Court found, and Coca-Cola conceded at trial, that the marketing investments the company credited with creating the supply points’ CPM-disqualifying intangibles and risks were essentially accounting constructs. The activities were performed by separate marketing service companies whose contracts were with Coca-Cola or a U.S. group subsidiary, and U.S. group entities directly compensated them and oversaw their efforts. At the unilateral discretion of U.S. group executives, these costs were then charged out and recorded on the supply points’ books as marketing expenses.
The intercompany contracts with the marketing service companies provided for compensation of cost plus 5 to 7 percent, which the parties stipulated was arm’s length. That stipulation led to the odd, if not necessarily outright contradictory, suggestion that activities worth 105 to 107 percent of the underlying service costs became astronomically more valuable upon their accounting allocation by U.S. headquarters personnel to the supply points. The only way to reconcile this apparent disparity would be to show that the marketing service companies’ cost-plus compensation arrangement shielded them from the risk that the marketing investments would fail, and that this risk was passed on to the supply points and impossible to reliably address through adjustments. (Prior analysis: Tax Notes Int’l, Aug. 12, 2024, p. 979.)
The Tax Court found that Coca-Cola Co. arbitrarily put the marketing expenses on the supply points’ books in such a way that each supply point’s gross profit would consistently and dramatically exceed its nearly contemporaneous expense allocations. On average, the seven supply points at issue in Coca-Cola had about $10.6 billion in annual revenue, $2.1 billion in cost of goods sold, and $4.6 billion in allocated marketing expenses and other operating expenses. The U.S.-parent-dictated marketing expense allocations thus left the supply points with average operating margins of 37 percent and net cost-plus markups of 57 percent.
An effective guarantee of year-after-year returns of this magnitude doesn’t reflect significant “risk” as the term is commonly understood. And it doesn’t reflect the meaning of risk under the regulations either: As specified in reg. section 1.482-1(d)(3)(iii)(B), an allocation of risk between controlled taxpayers after the outcome of the risk is known or reasonably knowable lacks economic substance. Allocating marketing expenses after the supply points’ gross profitability would have been known or reasonably knowable, and in amounts that consistently left the supply points with considerable operating profit, is an apparent ex post risk allocation. This prompted Lauber’s memorable quip that “risk is not something that can be assigned after the fact.”
Coca-Cola’s return-on-investment argument ran into the further obstacle that it effectively credited one set of group entities with contributions made by another group. According to Coca-Cola’s 2021 motion, it was a legal error for the Tax Court to consider which entities actually performed the marketing functions “because the only question to be answered is which of the parties as between Coca-Cola and its supply points bore the risks.” Lauber rejected this argument based on the well-established, and usually taxpayer-favorable, premise that the separate legal existence of corporate taxpayers must be respected in all but exceptional circumstances:
We consider the fact that [the supply points] reported on their books most of the marketing and related costs that the [marketing service companies] incurred and invoiced. . . . But we will not conflate the [marketing service companies] with the supply points, attribute the activities of the [marketing service companies’] employees to the supply points, or otherwise combine them for purposes of our transfer pricing analysis.
In effect, Lauber refused to let Coca-Cola sham its own subsidiaries.
Focusing on Distortions
As Lauber noted in his 2020 opinion, the futility of Coca-Cola’s arguments didn’t necessarily imply that the IRS’s allocation was reasonable. However, he concluded that “the tested activity is highly susceptible to a CPM and that the tested parties (supply points) and uncontrolled comparables (independent Coca-Cola bottlers) engaged in similar business activities under similar circumstances.” He further concluded that the data and assumptions underlying the IRS’s application of the CPM were reliable.
Lauber’s acceptance of the CPM in principle was based on a series of key factual findings and legal interpretations. But his most important conclusion, which dealt with a mixed question of law and fact posed by the specific application of the CPM at issue in Coca-Cola, may have been his approach to the CPM’s comparability threshold.
As the section 482 regulations and Lauber’s 2020 Coca-Cola opinion both recognize, the quality of the comparables is fundamental to any method’s reliability. According to reg. section 1.482-5(c)(2)(ii), comparability under the CPM principally hinges on similarity in assets employed, risks borne, and functions performed (to the extent they cast further light on assets and risks). A largely duplicative list of general comparability criteria appears in reg. section 1.482-1(d)(3), which provides a list that also includes contractual terms, economic conditions, and the features of any transferred property or services. In a careful review spanning 14 pages, Lauber concluded that the independent bottlers and supply points were comparable according to these criteria because they:
operated in the same industry, faced similar economic risks, had similar (but more favorable) contractual and economic relationships with petitioner, employed in the same manner many of the same intangible assets (petitioner’s brand names, trademarks, and logos), and ultimately shared the same income stream.
Of course, as Lauber acknowledged, the supply points and independent bottlers were not the same. The supply points were responsible for producing beverage concentrate using the formulas and processes they licensed from U.S. group entities. After the supply points sold the concentrate to the independent bottlers, the bottlers mixed the concentrate with water, bottled and packaged the final product, and sold it to unrelated distributors and retailers. This level-of-market distinction made exact equivalence in asset composition, functional profile, and risk assumption a practical impossibility.
But exact equivalence, contrary to some interested taxpayers’ and litigators’ claims, isn’t the CPM’s comparability threshold. (Prior analysis: Tax Notes Int’l, Oct. 2, 2023, p. 59.) Lauber found that the supply points and independent bottlers, which both heavily relied on nondiversifiable assets and depended on the same general income stream, were “highly comparable” in risks assumed. He further found that the two sets of companies were otherwise similar in assets employed, functions performed, economic conditions, and contractual terms. The notion that marketing expenses set the supply points apart from the independent bottlers was belied, according to Lauber, by the parties’ policy of splitting marketing costs 50-50.
Critically, the Tax Court also found that the qualitative differences between the supply points and independent bottlers, including the difference in market level, caused no ROA distortions that would undermine the CPM’s reliability to Coca-Cola’s disadvantage. The independent bottlers’ de facto territorial exclusivity, noncaptive customer base, and ownership of “genuine intangible assets in the form of retail distribution networks, sales forces, and customer lists” justified higher returns than the less risky and intangible-intensive supply points, the Tax Court found. And as the opinion explained, it would follow that the IRS’s CPM likely inflated the supply points’ returns:
Comparability under the regulations is principally judged on the basis of functions performed, contractual terms, risks assumed, economic conditions, and assets employed. We have found the bottlers highly comparable to the supply points in all five respects. Although concededly there are differences between the two sets of companies, we find on balance that these differences tend to make the bottlers deserving of a higher ROA than the supply points. To that extent Dr. Newlon’s CPM will tend to overcompensate rather than to undercompensate the supply points, and it is therefore conservative.
Translating this finding into the conclusion that the IRS’s CPM analysis selected reliable comparables requires a specific understanding of the regulations’ concept of reliability — that the IRS can reliably apply the CPM, even when there are material differences in recognized comparability criteria, as long as the resulting distortions (if any) on balance favor the taxpayer. Although the factor does not appear in the CPM regulations, reg. section 1.482-1(d)(3) expressly recognizes market level as a potentially relevant economic condition. But Lauber found “on balance that these differences tend to make the bottlers deserving of a higher ROA than the supply points,” leading him to conclude that this difference didn’t taint the CPM’s reliability.
The opposing interpretation, which could be described as the “distinction theory” of CPM reliability, is that comparability is a qualitative concept: A difference is a difference, and there’s no need to establish its significance by proving that it distorts operating profitability. The standard of judicial review is a mere formality. The difference is the distortion, and its existence automatically renders the IRS’s application of the CPM arbitrary, capricious, and unreasonable.
The concept can be illustrated with a simple example. Frisbees and styrofoam minifootballs clearly differ in shape, aerodynamic properties, and throwing motion. They’re also produced using different materials in different facilities, likely subject to subtly distinct regulatory requirements. However, the two items are used in almost identical ways, the functions and assets used to manufacture them significantly overlap, and the risks associated with producing and selling them are highly similar.
But it doesn’t matter. You simply can’t meaningfully compare the profitability of a manufacturer of proverbial Class II recreational projectiles to the profitability of a manufacturer of proverbial Class III recreational projectiles. They’re just not the same. The companies’ product designs, manufacturing processes, and quality control protocols are different, and that should be the end of the discussion. This was the gist of the unsolicited microeconomics lecture delivered by Glenn Hubbard during the Medtronic II trial, which has thus far won the day in a case pending a second appeal. (Prior analysis: Tax Notes Int’l, Jan. 1, 2024, p. 51.)
Harping on differences analogous to those between manufacturers of Frisbees and manufacturers of styrofoam footballs is surely a highly lucrative enterprise for tax advisers and litigators whose tax minimization tool of choice is the comparable uncontrolled transaction method. But the Coca-Cola opinion’s “distortion theory” of CPM reliability is more consistent with the regulatory scheme and the principles it represents. The section 482 regulations’ overarching objective is to obtain a reliable measure of an arm’s-length result, not to obsess over qualitative differences with no significant quantitative effect on the results.
Reg. sections 1.482-1(d)(2) and 1.482-1(e)(2) and a host of related regulatory provisions imply that differences are relevant to a method’s reliability only if they materially affect the quantitative measure of an arm’s-length result. And as noted in reg. section 1.482-5(b)(4), the reliability of a profit-level indicator depends on the tested party’s activities, the available data, and the extent to which it reliably measures “the income that the tested party would have earned had it dealt with controlled taxpayers at arm’s length.”
The regulations’ examples concerning the CPM, which make clear that the method can accommodate comparability shortcomings that other methods cannot, confirm this antidistortion principle. In Example 6 under reg. section 1.482-8, the CPM is the best method despite significant differences between the tested party and the comparables in manufacturing process complexity. In Example 9 of the same section, generally similar companies that perform routine manufacturing and marketing activities provide a sufficient basis to apply the CPM. And the interquartile range is the arm’s-length range in every relevant CPM example throughout reg. sections 1.482-1 and 1.482-5, which in accordance with reg. section 1.482-1(e)(2)(iii), implies that there must be undetected or unresolved comparability shortcomings in the comparables set.
None of the qualitative differences identified or implied in these examples prevent comparability under the CPM. The underlying principle, it would seem, is that the differences that matter are differences that lead to a mispriced controlled transaction. Qualitative differences that don’t interfere with the determination of an arm’s-length result are of no particular significance.
Prepare for a Bad Review
Most of the rest of Lauber’s methodological review scrutinized the comparability adjustments and computational details of the IRS’s CPM analysis and provided a scathing overview of Coca-Cola’s alternative proposed methods. These aspects of the opinion generally concern granular questions of fact and almost universally accepted interpretations of law; thus, they’re unlikely to garner significant attention on appeal.
As for the more important and contentious aspects of the 2020 opinion, Coca-Cola has an obvious incentive to cast them as questions of law or mixed questions of law and fact. A concerted attempt to do exactly that is apparent in the company’s 2021 motion, which repeatedly refers to the Tax Court’s adverse determinations as legal errors. If successful, these characterizations would allow Coca-Cola to obtain de novo appellate review — as opposed to review for clear error — of decisive aspects of the Tax Court opinion. Unfortunately for Coca-Cola, most of the adverse conclusions in the Tax Court opinion fall under one of two categories: legal contentions that are so obviously flawed that de novo review conveys no practical value, and well-documented factual findings reviewable for clear error.
Regarding interpretations of law, one important question addressed by the 2020 opinion is the applicable standard of review. (Prior analysis: Tax Notes Int’l, May 16, 2022, p. 856.) The premise that the IRS’s section 482 allocations are subject to deferential judicial review was clear from the statutory text since its enactment under section 45 of the Revenue Act of 1928. It was also universally accepted by courts decades before Chevron U.S.A. Inc. v. Natural Resources Defense Council Inc., 467 U.S. 837 (1984), required general judicial deference to all agency interpretations of ambiguous statutes. As Lauber noted in Coca-Cola, the holding in Asiatic Petroleum Co. v. Commissioner, 31 B.T.A. 1152, 1157 (1935), aff’d, 79 F.2d 234 (2d Cir. 1935), was that the taxpayer has the burden of proving that the commissioner’s exercise of statutory discretion “was purely arbitrary.”
After observing that the Asiatic Petroleum “standard of review continues to apply today,” Lauber concluded (without mentioning Chevron) that an IRS allocation concerning transfers of unique and valuable IP must stand unless the method used to compute it was unreasonable. An unreasonable method, according to Lauber, is one that either implicates “significant legal error” or makes faulty assumptions, uses flawed data, contains internal inconsistencies, or is otherwise unreasonable in its practical application.
The 2020 Coca-Cola opinion tried to clarify a standard of review that, despite its repeated incantation in section 482 cases over the years, has generally remained nebulous and underexamined. Although Lauber’s analysis was supported by judicial precedent, his attempt to fill a precedential void was necessarily somewhat novel. But the absence of anything atextual, illogical, or eccentric in Lauber’s formulation makes it unlikely to be a successful target on appeal. The text of section 482 clearly delegates broad discretionary authority, and these delegations must be given effect even after Loper Bright Enterprises v. Raimondo, 144 S. Ct. 2244. Assuming the Eleventh Circuit complies with this obligation, a deferential standard of review like the one articulated in Coca-Cola would be difficult to avoid.
Another important legal interpretation reviewable de novo in Coca-Cola concerns the hierarchy of transfer pricing methods or, more accurately, the lack thereof. Lauber rightly and firmly dispatched the contention that the regulations generally favor transactional methods like the CUT method and disfavor the CPM. But legally baseless as it may be, the notion that the CPM sits at the bottom of some phantom hierarchy of transfer pricing methods retains a devoted (if financially interested) following, and its validity is one of the pending questions before the Eighth Circuit in Medtronic II. (Prior analysis: Tax Notes Int’l, June 17, 2024, p. 1733.)
The hierarchy-of-methods argument probably doesn’t meet the threshold for a frivolous argument (at least not yet) so it would be surprising if it didn’t appear in Coca-Cola’s briefs on appeal. But it would be more surprising still if two out of the three Eleventh Circuit judges who hear the case fell for it. And as for the seemingly uncontested factual contention that no transactional comparables sufficient to apply the CUT method were available, proving clear error on appeal seems highly unlikely.
Regarding the more specific regulatory interpretations at issue, Coca-Cola’s flimsy arguments for rejecting the supply points as tested parties under the CPM are likely all doomed. That the regulations’ standard is whether the tested party’s intangibles are similar to those held by the potential CPM comparables, not whether the tested party has any intangibles at all, is clear as day from the text of reg. section 1.482-5(b)(2)(i). It’s especially fanciful to claim that licensee rights could be per se disqualifiers, especially considering the CPM’s inclusion as a specified method in reg. section 1.482-4(a)(2) and the multiple regulatory examples in which licensees are selected as tested parties. The 2021 motion’s claim that it was “legally erroneous” to apply the CPM, and thus to deny the supply points’ rightful returns simply because of their contractual rights as licensees, would be unlikely to succeed under any standard of appellate review.
The Tax Court’s rejection of Coca-Cola’s more amorphous marketing intangibles argument, which creates conflict with the terms of its own intercompany contracts, raises a further legal obstacle regarding taxpayers’ ability to disregard their own contractual arrangements based on economic substance. Reg. section 1.482-1(f)(2)(ii)(A) commits the IRS to respect controlled transactions as structured unless they lack economic substance, and reg. section 1.482-1(d)(3)(ii)(B) does the same specifically for intercompany contractual terms. But no regulatory, statutory, or judicial authority has acknowledged or implied that taxpayers have a reciprocal right, and it’s hard to imagine how any Eleventh Circuit judge could be induced into hallucinating one into existence. The circuit’s endorsement of the Danielson rule, which prevents taxpayers from disregarding their own contractual terms unless they are judicially unenforceable, only reinforces this assessment.
Of course it could have been that, as a factual matter, the supply points’ licensee rights, goodwill, or so-called marketing intangibles were valuable nonroutine intangible assets that should have excluded the entities from contention as tested parties. However, unfortunately for Coca-Cola, the Tax Court rejected some formulation of all these factual claims in its 2020 opinion. Persuading the Eleventh Circuit to overrule the Tax Court on these inherently factual questions would require proof of clear error, evidence of which is scarce if not entirely absent from Lauber’s meticulously detailed 2020 opinion.
Coca-Cola’s regulatory misinterpretation that incurring any unreimbursed costs to increase territorial sales automatically entitles a controlled party to more than a CPM-based return is seemingly hopeless as well. Section II.A.2 of the 2021 motion repeatedly states that compensation is required for contributions to the value of another controlled party’s IP, which is obviously true. But the regulations and examples clearly falsify the premise that this compensation must, as a matter of law, include a share of nonroutine returns. The regulations are replete with examples of tested parties that bear responsibility for marketing and local sales-generating activities, either as licensees or resellers of tangible goods.
The regulations are clear that only incremental contributions can entitle controlled parties to nonroutine returns. In the context of the CPM, that means the tested party’s contributions must exceed the baseline established by the comparables. The only legal argument for rewarding the supply points with nonroutine returns is that unreimbursed marketing investments disqualify a potential tested party based on the assumption of risk, and that clashes with the CPM’s comparability principles and the economic principles that underlie them. Only nonroutine risk that distinguishes the tested party from the potential comparables undermines the CPM’s reliability under the regulations, and Lauber made a factual finding that no such risk distinction existed in Coca-Cola.
The general legal premise that Coca-Cola cannot conflate its supply points and marketing service companies, and thus effectively sham its own subsidiaries, is likely also on solid ground even if reviewed de novo. But one of the few mixed factual and legal questions on which there may be a sliver of doubt, albeit a very narrow one, is that a risk allocation lacks economic substance under the specific circumstances present in Coca-Cola.
The recipients of the allocated costs in Coca-Cola didn’t perform, control, coordinate, or directly pay for the underlying marketing “investments,” and the parent company unilaterally dictated the amount of each allocation without direction from any governing policy or principle. The timing of the allocations meant that the subsidiaries’ other financial results, including revenue flows that were also dictated by the parent, were already known or reasonably knowable. Without fail, at least for the six profitable supply points at issue, the result was that the supply point’s gross profit far exceeded the sum of its cost allocation and other operating expenses. This effectively ensured that the risk purportedly borne by the supply points could only play out in their favor.
This reasoning is inextricably intertwined with factual determinations. But its legal basis, according to many critics, is an ultra vires endorsement of the OECD transfer pricing guidelines’ risk control threshold and its guidance on development, enhancement, maintenance, protection, and exploitation functions. (Prior coverage: Tax Notes Int’l, May 24, 2021, p. 1133.) But the more natural interpretation, and the one seemingly implied by Lauber’s remark about allocating risk after the fact, is that the practices described in the opinion resulted in an ex post risk allocation.
However, the most consequential aspect of Lauber’s analysis may be his rejection of the Medtronic-like distinction (or Frisbee-versus-styrofoam football manufacturer) theory of CPM reliability in favor of a distortion-based theory. The distortion theory is certainly the more logical choice, and it’s more consistent with the regulations’ stated goal of identifying an arm’s-length result. It also better reflects the regulations’ concept of comparability, according to which the differences of concern are those that have a material effect on price.
However, the section 482 regulations don’t directly address the question, and the sheer volume of regulatory text all but guarantees that motivated readers will find excerpts seemingly more aligned with the distinction theory. Considering its critical importance to the future viability of the CPM and its less immediately obvious answer, this may be the most significant point of review on appeal. If the Eleventh Circuit and Eighth Circuit both endorse the misguided distinction theory of CPM comparability, then we may be left with a method suitable only for proverbial widget manufacturers.
But these potential unknowns aren’t nearly enough to justify confidence in Coca-Cola’s position, much less its apparent certainty of eventual victory. The company’s appeal will still largely be directed by transfer pricing neophytes, and the series of legal and factual errors that Coca-Cola will have to identify in Lauber’s meticulously detailed opinion to discredit the CPM’s selection is daunting. Even the most vulnerable aspects of the Tax Court’s opinion are far more compelling than Coca-Cola’s counterarguments, and the IRS and Tax Court determinations on these issues are almost entirely insulated from the Loper Bright fallout.
Coca-Cola will not only have to establish that at least one of the Tax Court’s conclusions was wrong as a matter of law or clearly erroneous as a matter of fact, it will also have to show that the error is sufficient to void the Tax Court’s acceptance of the CPM using the supply points as tested parties. Otherwise, the most the company can hope for is a remand directing the Tax Court to adjust the particulars of the IRS’s CPM analysis. That means Coca-Cola’s hopes for a successful appeal may ultimately hinge on its “reliance interests” argument, according to which the IRS had a duty to forewarn the company of the change in its enforcement position starting with the 2007 tax year.
But Coca-Cola’s reliance interests argument, which is dubious even for 2007 through 2009, would be nonsensical as a justification for applying the same method more than a decade after notice was unambiguously tendered in 2011. Coca-Cola will thus have to prevail on its anti-CPM argument to avoid the bulk of the $17 billion potential exposure reported on its most recent Form 10-Q, and the company's prospects for doing so are far bleaker than public posturing suggests.