Coke Criticizes IRS’s Comparable Profits Method

THE COCA-COLA COMPANY AND SUBSIDIARIES,
Petitioner,
v.
COMMISSIONER OF INTERNAL REVENUE,
Respondent

UNITED STATES TAX COURT

Judge Albert G. Lauber

PETITIONER'S REPLY BRIEF

As explained in petitioner's Motion for Leave to File Reply Brief, petitioner is filing this reply brief for the limited purpose of addressing arguments that respondent improperly raised for the first time in Respondent's Answering Brief.

I. RESPONDENT'S BELATED ARGUMENTS REGARDING HIS INCONSISTENT TREATMENT OF PETITIONER'S CANADIAN AFFILIATE AND THE SIX FOREIGN LICENSEES ARE FALSE AND MISLEADING

Petitioner explained before trial and in Petitioner's Opening Brief why respondent's inconsistent treatment of petitioner's Canadian affiliate, Coca-Cola, Ltd. (“CCL”), and the Six Foreign Licensees demonstrates that respondent's position in this case — that legal ownership of trademarks entitles licensors to all returns related to the licensed intangibles — is arbitrary and capricious. Petitioner's Pretrial Memorandum at 20-21; Petitioner's Opening Brief at 471-474. After unsuccessfully attempting to exclude evidence of his treatment of CCL, respondent simply ignored the issue altogether in Respondent's Opening Brief.

Before turning to his improperly withheld arguments, respondent rehashes previously rejected evidentiary arguments. Respondent first asserts that evidence relating to his “settled” Canadian adjustments is inadmissible, observing in a footnote that the Court already ruled against him twice on the issue (without prejudice before trial and ultimately near the end of trial).¹ Respondent's Answering Brief at 698 n.82. Respondent ends his argument on a similar note, asserting that “[e]vidence of respondent's settlement position is inadmissible and irrelevant.” Id. at 704. Though it is unclear what respondent means by “settlement position,” respondent's focus on the competent authority resolution is a red herring.² The basis and details of the competent authority resolution are indeed irrelevant. It is, however, relevant that respondent's notice of deficiency contains transfer-pricing adjustments that rest on conflicting and irreconcilable positions. The Court should not be distracted by respondent's repeated recourse to evidentiary issues the Court has already decided.

Respondent's account of CCL's acquisition of trademark rights is false and misleading. Respondent asserts that “CCL paid one dollar to TCCC for the Company's Canadian trademark rights.” Id. at 700. But respondent neglects to mention that the trademark transfer to which he refers occurred nearly 100 years ago, in 1923 (the same year in which the Company incorporated CCL). See Ex. 2921-P (Form 886-A Attached to Canada NOP A) at CCADMIN0121223 (“U.S. Parent sold the 'Coke' name to Canadian Subsidiary in 1923 for $1.00.”); Ex. 155-J (TCCC Annual Report for 1923) at EXHJ_00000524 (reflecting that TCCC incorporated its Canadian interests in 1923). Moreover, respondent's assertion that the 1923 transfer was solely in exchange for $1 is false. TCCC received the capital stock of CCL when it incorporated the Canadian subsidiary and transferred trademark rights to that subsidiary. See Ex. 155-J (TCCC Annual Report for 1923) at EXHJ_00000524 (reflecting that TCCC received $1,000,000 of capital stock in CCL when it incorporated its Canadian interests in 1923); see also Coca-Cola Co. v. United States, 47 F. Supp. 109, 116 (Ct. Cl. 1942) (reflecting that “[i]n 1923 [TCCC] organized the Coca-Cola Company of Canada, Ltd., hereinafter referred to as the 'Canadian Company,' and since that time has owned the entire stock of the Canadian Company”).

Respondent's assertions that “[p]etitioner did not contend, let alone show, that an uncontrolled party would have permitted CCL to use its trademarks for one dollar” and that respondent was “[f]aced with TCCC's sale of bare legal title for a nominal amount,” Respondent's Answering Brief at 701, are likewise false.³ The transfer of CCL's stock meant that TCCC received an amount that was necessarily arm's length. See Eli Lilly & Co. v. Commissioner, 856 F.2d 855, 866 (7th Cir. 1988) (explaining that “the stock of a corporation is by definition equal in value to that corporation's assets” and would therefore equal the arm's-length value of assets transferred in exchange for stock).

While presenting his warped narrative of events that occurred in the distant past, respondent attempts to impugn Ms. Lopes's testimony that she was aware of no reasons under trademark law for registering trademarks in TCCC's name rather than in the name of a subsidiary.⁴ Respondent's Answering Brief at 700 n.85. Respondent asserts that this testimony is not supported by credible evidence. Respondent bases this assertion on a vague statement in an appendix to petitioner's U.S. Mutual Agreement Procedure request reflecting that the “registration of the trademarks in the name of CCL [in Canada] was undertaken on the advice of intellectual property counsel to ensure and facilitate the legal protection of trademark rights in the Canadian market.” Id. (quoting Ex. 2923-J (Appendix 6 to U.S. MAP Request) at CCADMIN0121322).

Respondent's attack is off-base for two reasons. First, the fact that there might have been Canadian law reasons for registering the Canadian trademarks in the name of a local Canadian affiliate in no way means that there were trademark law reasons for registering foreign trademarks in TCCC's name. Second, Ms. Lopes was speaking based on her knowledge of the Company's practices during the years in issue — not to events from three or four generations ago. In addition, respondent's baseless attack on Ms. Lopes's testimony regarding the Company's trademark-registration practices is grounded in part in improper objections to petitioner's proposed findings of fact.⁵ See Respondent's Answering Brief at 700 n.85 (citing OPFF ¶¶ 514-516).

Respondent tries to misdirect the Court with the assertion that “[a]fter the years at issue, TCCC owned Canadian trademarks related to Trademark Coke.” Respondent's Answering Brief at 702 n.88. Respondent references Exhibit 1034-J, which is an Anaqua printout.⁶ That printout is irrelevant because it relates to years after those in issue. And respondent's assertion is false: CCL (not TCCC) was the registered owner of the trademarks to which respondent refers.⁷

Respondent then pivots to an assertion that he “took into account” other TCCC contributions to the Canadian business. Respondent's Answering Brief at 702. The “contributions” respondent mentions include formulas, which the record shows cannot explain a firm's ability to earn premium prices.⁸ See Petitioner's Answering Brief at 101-102 (explaining that formulas do not explain the Company's ability to earn premium prices). The record likewise debunks respondent's other claimed “contributions” from TCCC, including “final decisions on core global brand frameworks and . . . control over global marketing” and ownership of “the Canadian Bottler network.” Id. at 702-703; see Petitioner's Answering Brief at 101-102 (explaining that respondent's “distribution network” is a flawed concept) and id. at 78-79 and 79 n.68 (explaining that basic trademark-and marketing-related guidelines are standard in arm's-length license transactions and do not affect royalty rates and that corporate headquarters marketing activity did not drive value creation in any foreign market); Petitioner's PFF ¶¶ 77-83 (explaining that business units were responsible for leading the Company's relationships with bottlers) and Petitioner's PFF ¶¶ 182-184 (explaining that business units developed local marketing strategies). Respondent's assertion is also flatly contradicted by the fact that Dr. Newlon excluded “brands” owned by Atlantic Industries in computing the Ireland Licensee's financial results, which is reflected in respondent's own notice of deficiency.⁹

Respondent then moves to assertions regarding Caribbean Refrescos, Inc. (“CRT"), a wholly owned U.S. subsidiary of TCCC that (like Export) filed federal consolidated returns with TCCC. Seventeenth Stip. of Facts ¶ 2840. Though respondent never explains why his intra-consolidated-return (i.e., between CRI and TCCC) adjustments matter in this context, his assertion highlights another glaring inconsistency in his position. Elsewhere in Respondent's Answering Brief, respondent falsely asserts that TCCC operated in the foreign markets before the licenses were executed with the Six Foreign Licensees. See Respondent's Answering Brief at 758 (falsely stating that TCCC operated a Brazilian branch before 1963; branches of Export Sales and Export operated in Brazil); OPFF ¶ 453 (falsely stating that Company operations expanded internationally through branches of TCCC). TCCC was always a licensor that relied on licensees (either controlled foreign corporations or branches of other affiliates) to develop the foreign markets at issue. Petitioner's PFF ¶ 453; Petitioner's Objection to Respondent's Proposed Finding of Fact (“RF”) ¶ 233. When it suits respondent, he emphasizes the separateness of TCCC and its U.S. affiliates. And when it does not, he simply conflates the two.¹⁰

In the end, the arbitrary and capricious nature of respondent's conflicting notice positions is apparent from the notices of proposed adjustment that explain his Canada adjustments in the notice of deficiency. Those notices of proposed adjustment reflect respondent's disregard of legal ownership and emphasis on the expenses and risks associated with the continued development, enhancement, and maintenance of the trademarks that drove the Company's business in Canada. Petitioner's PFF ¶¶ 668, 673, and 674.¹¹ Respondent's Answering Brief simply ignores respondent's actual position on Canada and fabricates a new one devoid of credibility.¹²

II. RESPONDENT INSERTS IMPROPERLY WITHHELD ARGUMENTS INTO HIS OBJECTIONS TO PETITIONER'S PROPOSED FINDINGS OF FACT

A. Respondent belatedly offers convoluted calculations that rely on false assumptions in a failed attempt to save Dr. Newlon's return-on-assets profit-level indicator

Dr. Newlon admitted that if the ratio of intangible assets to tangible operating assets is lower for the bottlers than for the supply points, then his bottler-based application of the comparable profits method (“CPM”) understates the operating income the supply points would have earned at arm's length. Tr. (Newlon) at 8404:11-20; Ex. 8295-R (Newlon Rebuttal Report) at 3. Prof. Willig calculated the Foreign Licensees' and bottlers' tangible assets in his opening report and determined that the bottlers had twelve times the tangible assets of the Foreign Licensees. Ex. 7251-P (Willig Opening Report), Ex. 5. Dr. Newlon did not dispute Prof. Willig's tangible-asset calculation. Tr. (Newlon) at 8537:19-8538:5. Dr. Newlon also admitted that his bottler-based CPM understates the supply points' income unless the bottlers had intangible assets worth twelve times those of the supply points.¹³ Tr. (Newlon) at 8542:8-8543:6.

Respondent waited until Respondent's Answering Brief to attack Prof. Willig's estimation of the Foreign Licensees' and bottlers' tangible-asset ratios. Respondent now contends for the first time that Prof. Willig should have included the tangible assets of service companies in the supply points' tangible-asset bases because service companies “were also involved in generating revenues for the markets at issue.”¹⁴ OPFF ¶ 944. Respondent then concocts a brand new estimation of service company tangible assets to try and show that including those assets in the supply points' tangible asset bases reduces the twelve-to-one difference between bottler and supply point tangible assets to a five-to-one difference. Id.

That estimation — based primarily on respondent's misleading interpretation of data from petitioner's expert Dr. Luehrman — is irreparably flawed. Respondent makes the unsupported assertion that the asset profiles of the Chile and Brazil Licensees are reflective of the asset profiles that would have resulted if other supply points were combined with service companies. Id. Respondent then infers that because (1) the Chile and Brazil Licensees had “similar revenue to asset turnover ratios” as bottlers¹⁵ and (2) there is evidence in the record that bottlers earned about five times the revenues of supply points, bottlers must have had five times the tangible assets of the combined supply points and service companies. Id. Respondent is incorrect.

As Dr. Luehrman explained at trial, the differences between the revenue-to-asset-turnover ratios of the Chile and Brazil Licensees and the other Foreign Licensees were driven by securities and other investments — i.e., nonoperating assets that neither Prof. Willig nor Dr. Newlon used in their analyses — that the Chile and Brazil Licensees possessed and the other Foreign Licensees did not. Tr. (Luehrman) at 6189:4-6190:8.¹⁶ If those nonoperating assets were removed, then the Chile and Brazil Licensees' revenue-to-asset-turnover ratios would have resembled those of the other Foreign Licensees (which were significantly higher than bottlers'). Tr. (Luehrman) at 6190:9-22. This demonstrates that the securities and other investments held by the Chile and Brazil Licensees — not service company assets reflected on the Chile and Brazil Licensees' balance sheets — accounted for the differences between their revenue-to-asset-turnover ratios and those of the other Foreign Licensees.¹⁷ 

In the end, respondent's logic fails. Respondent is stuck with what his experts did not challenge — the twelve-to-one difference between the bottlers' and Foreign Licensees' tangible operating assets. There is no evidence that the bottlers had twelve times or more intangible assets than the Six Foreign Licensees. Petitioner's ¶ 946. This is fatal to Dr. Newlon's bottler-based CPM.

B. Respondent misportrays expert testimony and documentary evidence to create the false impression that arm's-length licensors charge astronomical royalty rates that soak up all profits in excess of “benchmarked” low margins

Faced with unrefuted evidence that arm's-length licensing parties seek to split profits equitably and do not use margins from one transaction to benchmark a licensee's profits in another, in OPFF ¶ 822 respondent belatedly cobbles together disparate evidence in an attempt to support the admittedly theoretical approaches advocated by his economists.¹⁸ Respondent begins by falsely asserting that Dr. Barbera confirmed that Dr. Becker's method was a valid method for determining arm's-length royalties. Id. Dr. Barbera did no such thing. He was a rebuttal witness who attacked Dr. Becker's approach (in a section of his rebuttal report titled “Dr. Becker's Analysis is Unreliable”) before he sought (in a separate section of his rebuttal report) to rebut Dr. Becker's conclusion that the Foreign Licensees' profit margins were too high by showing that licensees in the concentrate business could earn high margins even after paying substantial royalties. Ex. 7376-P (Barbera Rebuttal Report) at ¶¶ 47-54.

Later in OPFF ¶ 822, respondent strangely pivots from attempting to defend the use of operating margins to determine royalty rates to arguing for astronomical royalty rates in the abstract.¹⁹ Respondent does so by pointing to a single spreadsheet relating to impairment testing regarding a trademark for a product called “Joya” and drawing the conclusion that, “in testing its intangible assets for impairment, the Company determined implied royalty rates that exceeded 50% of sales.” OPFF ¶ 822. Although respondent examined many fact witnesses, including the Company's former CFO (whose deposition transcript was admitted into evidence), respondent failed to question a single witness about Joya or the referenced spreadsheet. If respondent had done so, then he would have learned that the purpose of impairment testing is not to identify an arm's-length royalty rate.

This is apparent from the spreadsheet to which respondent points. The 79.2% royalty rate used in the impairment testing exceeds the operating income from Joya products over the period considered (i.e., it would have left a hypothetical licensee in a loss position over a ten-year period). See Ex. 8787-R (Joya Trademark Spreadsheet) at Tab “Enterprise Value-Trademarks” (row 34) and Tab “Royalty Rate” (row 19) (indicating that the “royalty” calculated as part of the impairment test exceeded operating income during the tested period). Moreover, respondent fails to mention that a note next to the royalty rate indicates that “[a]s a reasonableness test, the royalty rate should be in the general range of approximately 25% of the normalized gross profit margin and 50% of the normalized operating income margin.”²⁰ Id. at Tab “Royalty Rate” (rows 67-69). The 79.2% royalty rate obviously flunks that reasonableness test. In the end, OPFF ¶ 822 is nothing but an attempt to mislead the Court via a belated argument by subterfuge.²¹.

III. RESPONDENT'S BELATED ARGUMENTS RELATING TO THE OWNERSHIP OF BRAZILIAN INTANGIBLE PROPERTY ARE WITHOUT MERIT

A. Respondent's attempt to apply some of the secondary factors for determining developer status under the 1968 transfer-pricing regulations is meritless

Under the 1968 transfer-pricing regulations, the developer of intangible property is the owner of that intangible property regardless of whether another affiliate is the legal owner. The primary determinant of developer status is the relative development cost borne by each group member. Treas. Reg. § 1.482-2(d)(1)(ii)(c) (1968). Secondary factors are the location of development activity, the capabilities of various members to carry out development independently, and the degree of control over the project exercised by the various group members. Id.

In Petitioner's Pretrial Memorandum, petitioner explained why all of the factors for determining developer status support a determination that the Brazil Licensee was the developer of the Company's Coca-Cola, Fanta, and Sprite trademarks in Brazil that were in existence and licensed to it before November 17, 1985 (“the grandfathered trademarks”). Petitioner's Pretrial Memorandum at 106-110. In Respondent's Opening Brief, respondent noted but did not apply the secondary factors for determining developer status. Respondent's Opening Brief at 491; Petitioner's Answering Brief at 148 n.117. For the first time in Respondent's Answering Brief, respondent purports to apply the secondary development factors to argue that those factors do not support treating the Brazil Licensee as the developer of the grandfathered trademarks. Respondent's Answering Brief at 760-761.

Although there are three secondary development factors, respondent belatedly addresses only two.²² To support his claim that the development of Brazilian trademarks occurred in the United States (and not Brazil), respondent notes that “TCCC contributed to Brazilian sales by providing the framework of marketing guidelines and advice for its global core products.” Id. But providing frameworks and advice is a far cry from the decision-making and local marketing activity that actually occurred in Brazil (and developed the Company's Brazilian trademarks). The Brazil Licensee was responsible for developing local consumer marketing and had to adapt the Company's broader marketing ideas to local conditions. Petitioner's PFF ¶¶ 162-189, 192, 202, 206, and 213.

In addressing whether the Brazil Licensee was capable of carrying out the development of the grandfathered trademarks independently, respondent asks the Court to ignore the proven capabilities of the Brazil Licensee because the clearest example in the record reflecting those capabilities — which involved Kuat — is from the 1990s (rather than the 1960s). Respondent's Answering Brief at 761. The evidence reflects that local knowledge and capabilities were always required for the Company to succeed in foreign markets. Petitioner's PFF ¶¶ 11-21. Moreover, it is telling that respondent never characterizes TCCC as having had the capability to develop the Brazilian trademarks independently.

B. Respondent's untimely argument regarding Kuat and other local Brazilian products ignores the relevant inquiry and is counterfactual

Respondent has long been on notice of petitioner's argument that it is inconsistent with economic substance to treat TCCC as the owner of intangible property associated with Kuat and other local Brazilian products. In their analyses, experts (including Dr. Newlon) excluded the revenues from Kuat and other local Brazilian products. But Respondent's Opening Brief does not address the issue on its substantive merits, noting only that respondent was not conceding that TCCC was entitled to no royalties on sales of those products. Respondent's Opening Brief at 482 n.187. Respondent belatedly debuts a new argument in Respondent's Answering Brief at 763-764.²³

According to respondent, TCCC owned the intangible property associated with Kuat and other local Brazilian products because TCCC owned intangible property associated with branded products. Id. Respondent's argument simply assumes away the relevant inquiry. The regulations look through legal ownership where it is inconsistent with economic substance. Treas. Reg. § 1.482-4(f)(3)(i)(A).²⁴ Because respondent never examined the economic substance of the license transaction with respect to Kuat and the other local Brazilian products, he cannot credibly contend that TCCC's legal ownership of intangible property associated with Kuat and other local Brazilian products is consistent with economic substance.

It would be inconsistent with economic substance to require the Brazil Licensee to pay TCCC royalties on sales of concentrate relating to Kuat and other local Brazilian products. Petitioner's Opening Brief at 534-535. The evidence reflects that corporate headquarters played no role in the development of local Brazilian products. Petitioner's PFF ¶ 256. Unable to dispute the relevant factwitness testimony, respondent bizarrely (and improperly) declares it irrelevant and pivots to an observation that corporate-headquarters approval was required for all new products. OPFF ¶ 256. Aside from the fact that sound corporate controls (i.e., approval processes) present in every competently managed multinational enterprise are of no moment in a transfer-pricing analysis, Petitioner's Answering Brief at 75-78, unrefuted fact-witness testimony reflects that approval for new formulas developed by regional R&D centers (like the Brazil Licensee's facility in Rio de Janeiro) took just a few days and was granted in a rubber-stamp process, Petitioner's PFF ¶ 254.²⁵

IV. CONCLUSION

Respondent's improperly withheld arguments rest on significant falsehoods and misstatements and lack merit. Respondent's Answering Brief exhibits the same kind of distortion as his other post-trial brief and further confirms that respondent's factual assertions and use of legal authorities require close scrutiny.

Respectfully submitted,

John B. Magee
T.C. Bar No. MJ0060
Morgan, Lewis & Bockius LLP
1111 Pennsylvania Avenue, N.W.
Washington, DC 20004
(202) 373-6229
john.magee@morganlewis.com

Kevin L. Kenworthy
T.C. Bar No. KK0147
Miller & Chevalier Chartered
900 16th Street, N.W.
Washington, DC 20006
(202) 626-5848
kkenworthy@milchev.com

Sanford W. Stark
T.C. Bar No. SS0902

Saul Mezei
T.C. Bar No. MS0625 

Michael D. Kummer
T.C. Bar No. KM0588 

C. Terrell Ussing
T.C. Bar No. UC0015 

Hans D. Gerling-Ritters
T.C. Bar No. GH0415 

John F. Craig III
T.C. Bar No. CJ1811

Steven R. Dixon
T.C. Bar No. DS0259

Lamia R. Matta
T.C. Bar No. ML0538

Jarrett Y. Jacinto
T.C. Bar No. JJ0473

Lisandra Ortiz
T.C. Bar No. OL0109

Nicholas R. Metcalf
T.C. Bar No. MN0152

Dated: March 15, 2019