DOJ Urges Eighth Circuit to Reverse Tax Court in Medtronic Case

MEDTRONIC, INC. & CONSOLIDATED SUBSIDIARIES
Appellee/Cross-Appellant
v.
COMMISSIONER OF INTERNAL REVENUE
Appellant/Cross-Appellee

IN THE UNITED STATES COURT OF APPEALS
FOR THE EIGHTH CIRCUIT

ON APPEAL FROM THE DECISION
OF THE UNITED STATES TAX COURT

BRIEF FOR THE APPELLANT/CROSS-APPELLEE

DAVID A. HUBBERT
Deputy Assistant Attorney General

FRANCESCA UGOLINI (202) 514-3361
JENNIFER M. RUBIN (202) 307-0524
JUDITH A. HAGLEY (202) 514-8126
Attorneys
Tax Division
Department of Justice
Post Office Box 502
Washington, D.C. 20044

SUMMARY OF THE CASE AND REQUEST FOR ORAL ARGUMENT

This federal income-tax case concerns the proper method to determine an arm's-length price for an intercompany transfer of high-profit intangibles by a U.S. corporation (Medtronic) to its controlled foreign subsidiary under Section 482 of the Internal Revenue Code and the related regulations. After a trial, the Tax Court determined a price utilizing Medtronic's transfer-pricing method. The Commissioner appealed, and this Court remanded for findings that were “necessary” to its determination whether the Tax Court had adopted the “best transfer pricing method.” Medtronic, Inc. v. Commissioner, 900 F.3d 610, 615 (8th Cir. 2018). On remand, the Tax Court (i) made some, but not all, of the required findings, (ii) rejected the Commissioner's transfer-pricing method for reasons contrary to the regulations, and (iii) adopted a method that conflicts with the regulations and allocates profits to the foreign subsidiary in an amount that far exceeds its relative economic contribution to Medtronic's operations.

Counsel for the Commissioner respectfully inform the Court that they believe that oral argument may be beneficial here, and suggest that 20 minutes per side would be adequate to address the issues.

---

TABLE OF CONTENTS

Summary of the case and request for oral argument

Table of contents

Table of authorities

Glossary

Jurisdictional statement

Statement of the issue

Statement of the case

Summary of argument

Argument

Conclusion

Certificate of compliance

Certificate of service

TABLE OF AUTHORITIES

Cases:

Coca-Cola Co. v. Commissioner, 155 T.C. 145 (2020)

G.D. Searle & Co. v. Commissioner, 88 T.C. 252 (1987)

Medtronic, Inc. v. Commissioner, 900 F.3d 610 (8th Cir. 2018)

Smoky Hills Wind Project II, LLC v. City of Indep., Mo., 889 F.3d 461 (8th Cir. 2018)

Sundstrand Corp. v. Commissioner, 96 T.C. 226 (1991)

Statutes:

Internal Revenue Code (26 U.S.C.):

Tax Reform Act of 1986, P.L. 99-514

Regulations, Rules, and Rulings:

Treasury Regulations (26 C.F.R.):

Fed. R. App. P. 13(a)(1)

58 Fed. Reg. 5,263 (1993)

59 Fed. Reg. 34,971 (1994)

Other Authorities:

Bittker & Lokken, Fed'l Tax'n of Income, Estates and Gifts (2023)

Ryan Finley, What's at Stake in the Medtronic II Appeal, 112 Tax Notes Int'l 59 (Oct. 2, 2023)

Ryan Finley, Medtronic, Part 2: A High Risk of Bad Precedent, 175 Tax Notes 703 (May 2, 2022)

H.R. Rep. No. 99-426 (1985)

H.R. Rep. No. 99-841 (1986)

Joint Committee on Tax'n, Gen'l Explanation of the Tax Reform Act of 1986, JCS-10-87 (1987)

GLOSSARY

Add.	Commissioner's Separate Addendum
App.	Commissioner's Separate Appendix
CPM	Comparable profits method
CUT	Comparable uncontrolled transaction
FDA	Food and Drug Administration
I.R.C.	Internal Revenue Code (26 U.S.C.)
IRS	Internal Revenue Service
R&D	Research and development
Reg.	Treasury Regulation (26 C.F.R.)

---

JURISDICTIONAL STATEMENT

Appellee/cross-appellant Medtronic, Inc. (Medtronic), a U.S. corporation and common parent of a multinational group of consolidated U.S. corporations and foreign affiliates, filed consolidated federal income-tax returns for its 2005-2006 tax years. (App. 748, Add. 7.) In 2011, it petitioned the Tax Court for a redetermination of deficiencies determined by the Commissioner of Internal Revenue for those years. (App. 2.) The Tax Court had jurisdiction under Sections 6213(a) and 7442 of the Internal Revenue Code (26 U.S.C.).

Following an appeal to, and a remand from, this Court, the Tax Court entered its decision on June 14, 2023. (App. 1415-1416, Add. 222-223.) That decision disposed of all the parties' claims. This Court has jurisdiction over the Commissioner's appeal under I.R.C. §7482(a)(1). The Commissioner's notice of appeal was timely filed on September 8, 2023 (App. 1421), within the 90 days allowed by I.R.C. §7483. See Fed. R. App. P. 13(a)(1).

STATEMENT OF THE ISSUE

Medtronic developed highly profitable intangibles, which it licensed to an affiliated manufacturing operation in Puerto Rico (Medtronic-P.R.). Although Medtronic controlled it, Medtronic-P.R.'s income was not subject to U.S. tax. Section 482 and its implementing regulations authorize the Commissioner to revise the intercompany (transfer) pricing so that it clearly reflects income subject to U.S. taxation. The question presented is:

Whether the Tax Court's rejection of the Commissioner's transfer-pricing method, and its adoption of a method that relies on a transaction involving what the court found to be wholly different intangibles, violates Treasury's Section 482 regulations.

The most apposite authorities are:

STATEMENT OF THE CASE

A. Procedural overview

Medtronic priced intercompany transactions with Medtronic-P.R. in a manner that did not clearly reflect its income subject to U.S. taxation. Section 482 of the Internal Revenue Code is designed to prevent such behavior and allows the Commissioner to reallocate income among related parties by determining the arm's-length price for intercompany transactions. In 2010, the Commissioner issued Medtronic a notice of deficiency for tax years 2005-2006 that increased its U.S. taxable income by $500 million for 2005 and $750 million for 2006, to reflect reallocation of income for certain intercompany transactions between Medtronic's U.S. companies and Medtronic-P.R. (App. 357.) Medtronic petitioned the Tax Court challenging the deficiencies. (App. 2.)

After a trial, the Tax Court concluded that neither party had determined an arm's-length price for the transaction at issue (an intercompany license of technology intangibles), made adjustments to Medtronic's transfer-pricing method, and entered a decision largely affirming Medtronic's tax-reporting position. The Commissioner appealed. This Court vacated the decision and remanded for additional fact findings. Medtronic, 900 F.3d at 615. On remand, the Tax Court concluded that the transfer-pricing method that it previously adopted was not the best method, rejected the Commissioner's method, and adopted a new “unspecified” method that was (i) developed by Medtronic on remand in its post-trial briefs and (ii) largely based on a transaction that the court found unreliable for pricing the intangibles. The Commissioner has appealed, and Medtronic has cross-appealed.

B. Background: Transfer pricing

U.S. corporations operating through related enterprises, including affiliated foreign corporations, have long attempted to manipulate their internal allocations and transactions in order to avoid U.S. income. For example, a U.S. corporation which sells a product generated through the joint efforts of U.S. and Puerto Rican subsidiaries could artificially lower its U.S. taxable income by allocating most of the costs to the U.S. companies and most of the income to the Puerto Rican subsidiaries. To combat such abuse, and to ensure that transactions between related parties reflect economic reality, Congress — for almost 100 years² — has given the IRS “broad authority” to evaluate pricing of transactions between commonly controlled parties (App. 817, Add. 76), and to allocate certain tax items (including gross income) “if [it] determines that such . . . allocation is necessary in order to prevent evasion of taxes or clearly to reflect the income of any such” entities. I.R.C. §482. Under regulations implementing Section 482, taxable income of commonly controlled entities will be determined as if they had conducted their affairs in the manner of unrelated entities “dealing at arm's length.” Reg. §1.482-1(b)(1).

This case concerns the transfer of the right to use intangibles between related parties. For many years, the arm's-length price for such transfers was determined primarily by reference to transactions between unrelated parties involving “the same or similar intangible property under the same or similar circumstances.” 26 C.F.R. §1.482-2(d)(2)(ii) (1985). By the mid-1980s, however, Congress became concerned that this transaction-based approach was insufficient to properly allocate income when related parties transferred “high-profit” intangibles because there were no comparable transactions for such transfers. Joint Committee on Tax'n, Gen'l Explanation of the Tax Reform Act of 1986, JCS-10-87, at 1014-1016 (1987).

In particular, and as relevant here, Congress was concerned about taxpayers “transferring relatively high profit intangibles to Puerto Rico” without requiring the Puerto Rican affiliate to pay the U.S. affiliate a royalty that was “commensurate with the income attributable to the intangible.” H.R. Rep. No. 99-426, at 425 (1985). As Congress explained, there “is a strong incentive for taxpayers to transfer intangibles to related [corporations] in a low tax jurisdiction, particularly when the intangible has a high value relative to manufacturing or assembly costs.” Id. at 423.

To remedy this problem, Congress added the following sentence to Section 482 in 1986:

In the case of any transfer (or license) of intangible property (within the meaning of section 936(h)(3)(B)), the income with respect to such transfer or license shall be commensurate with the income attributable to the intangible.

Tax Reform Act of 1986, P.L. 99-514, §1231(e)(1) (codified at I.R.C. §482). The stated “objective” of the 1986 amendment — known as the commensurate-with-income requirement — was to ensure “that the division of income between related parties reasonably reflect the relative economic activity undertaken by each.” H.R. Rep. No. 99-841, at II-637 (1986) (Conf. Rep.). Congress also directed Treasury to evaluate its transfer-pricing regulations and consider alternative methods for pricing intercompany intangible transfers that did not depend on identifying comparable transfers between unrelated parties. Id. at II-638.

In response, Treasury overhauled its transfer-pricing regulations. Intercompany Transfer Pricing Regulations Under Section 482, 59 Fed. Reg. 34,971 (1994). The changes were directed at remedying the prior inappropriate pricing of high-profit intangibles by reference to transactions that were not truly comparable. To remedy the problem, the 1994 regulations (i) heightened the comparability standards for reliance on purportedly comparable transactions between unrelated parties involving intangibles, and (ii) provided alternative, profit-based methods that, in many instances, could more reliably provide an arm's-length price by valuing the intangibles indirectly. See Coca-Cola, 155 T.C. at 218 (explaining how, for “high-value intangibles,” the “most reliable transfer pricing method is often one that avoids any direct valuation of those intangibles”).

The 1994 regulations provide several alternative methods for determining an arm's-length price for intercompany transactions, and require that the “best method” — that is, the method that provides the most reliable arm's-length result for the transaction at issue — be utilized. Reg. §1.482-1(c). For intercompany transfers or licenses of intangible property (such as patents or trade secrets), the alternative methods include (as relevant to this appeal) the comparable-uncontrolled-transaction method (CUT), Reg. §1.482-4(c), the comparable-profits method (CPM), Reg. §1.482-5, and the unspecified method, Reg. §1.482-4(d). The latter (as the name suggests) includes any method not specified in the regulations that otherwise satisfies the regulatory requirements.

The CUT method directly values the intangibles and determines the arm's-length consideration for the related-party transfer (the “controlled transaction”) based on the price utilized in a transaction between unrelated parties (the “uncontrolled transaction”), and is limited to situations where the intangibles transferred, and the circumstances of the transfer, are comparable. Reg. §1.482-4(c)(2)(iii)(B)(1)&(2). Differences in circumstances between the controlled and uncontrolled transactions may be accommodated by making appropriate adjustments to the price in the uncontrolled transaction, Reg. §1.482-4(c)(2)(iii)(A), so long as the number and magnitude of the adjustments are small enough for the method to be reliable, Reg. §1.482-1(c)(2)(i). But differences in intangible property — that is, property that is not “comparable” (as that term is used in Reg. §1.482-4(c)(2)(iii)(A)) — cannot be accommodated by adjustments, because a CUT “requires” “comparable intangible property.” Reg. §1.482-4(c)(2)(iii)(A). In particular, intangibles that do not have “similar profit potential” are not “comparable intangible property.”³ Reg. §1.482-4(c)(2)(iii)(B)(1)(ii).

Unlike the CUT method, the CPM does not depend on finding a comparable uncontrolled transaction, which may not exist for high-profit intangibles. Instead, the CPM determines the arm's-length consideration for a controlled transaction indirectly, “based on objective measures of profitability (profit level indicators) derived from uncontrolled taxpayers that engage in similar business activities under similar circumstances.” Reg. §1.482-5(a). This method “relies on the general principle that similarly situated taxpayers will tend to earn similar returns.” 59 Fed. Reg. at 34,985. The regulatory standards for evaluating a “comparable” company under the CPM are less “strict” than those used to evaluate a “comparable” transaction for transaction-based pricing methods, and “diversity in terms of the functions and products is permitted.” Id. at 34,974. The CPM generally is applied to the participant in the controlled transaction that has the least complex operations and lacks unique intangibles (the tested party). Reg. §1.482-5(b)(2).

The regulations demonstrate (through an example) how the CPM may be used to indirectly price controlled transfers of intangibles. Reg. §1.482-5(e), Ex. 4. Assume U.S. Company licenses intangibles to Foreign Subsidiary, which uses them to manufacture goods that it sells to other subsidiaries of U.S. Company. Under the CPM, royalties paid to U.S. Company by Foreign Subsidiary may be tested by comparing the operating profit reported by Foreign Subsidiary to that of companies engaged in similar manufacturing operations, based, for example, on their rates-of-return on operating assets. If Foreign Subsidiary's rate-of-return differs from that of the comparable companies, the royalties must be adjusted to an amount that produces a comparable return for Foreign Subsidiary.

C. Medtronic's operations

Medtronic is a leading medical device company headquartered in the United States with operations and sales worldwide. (App. 748, Add. 7.) As relevant here, Medtronic produced and marketed implantable cardiac pulse generators, neurological stimulators (Devices), and related medical-therapy-delivery devices (Leads) (collectively “Devices/Leads”) to U.S. customers. See App. 623-634 (describing Devices/Leads).

Medtronic's Devices/Leads business engages several different functions within the Medtronic organization: research and development (R&D), clinical trials and regulatory approvals, manufacturing, marketing, and sales and distribution. (App. 754-762, Add. 13-21; App. 509-540.) The clinical and regulatory functions are required because the Devices/Leads are class III devices regulated by the Food and Drug Administration (FDA). The FDA classifies all medical devices into class I, II, or III, and applies the highest level of control to class III devices, requiring those devices to go through a premarket-approval process prior to distribution in the United States. (App. 748-750, Add. 7-9.) Medtronic's U.S. operations bore the high costs of obtaining those regulatory approvals. (App. 757, 1345-1346, Add. 16, 150-151.)

During 2005-2006, all the functions required to produce and market the Devices/Leads were performed by Medtronic's U.S. operations (Medtronic and its consolidated U.S. subsidiary, Med USA), except for the portion of the manufacturing performed by Medtronic-P.R. (discussed below). (App. 754-762, 789-796, Add. 13-21, 48-55; App. 509-540.) Although Medtronic-P.R. assisted its U.S. affiliates with some of the non-manufacturing functions, the responsibility for those functions rested with Medtronic's U.S. operations. (App. 754-758, 771, 785-791, 851, Add. 13-17, 30, 44-50, 110.) During the years at issue, Medtronic's U.S. operations incurred approximately 90% of the external costs (i.e., costs excluding intercompany costs) for the Devices/Leads, and Medtronic-P.R. incurred the rest. (App. 1025.)

Product quality is essential to Medtronic's ability to promote and sustain its brand. (App. 751, Add. 10; App. 503.) Accordingly, Medtronic emphasizes quality throughout its operations, and has created quality manuals and policies for each of its business units. (App. 755-756, Add. 14-15; App. 503-504, 534-536.) Although concerns about quality affect all functional areas, it has the greatest impact on Medtronic's R&D, component manufacturing, and finished-product manufacturing. (App. 510.)

D. Manufacturing of Devices/Leads

During 2005-2006, Medtronic's manufacturing function was split between its U.S. and Puerto Rican operations. (App. 758-765, 794-796, Add. 17-24, 53-55; App. 510.) Medtronic's U.S. manufacturing divisions made the critical, complex components for the Devices/Leads. (App. 758-761, Add. 17-20; App. 525-527.) Medtronic-P.R. then incorporated those components into the finished Devices/Leads that it manufactured.⁴ (App. 528-533.)

The U.S. component-manufacturing operations and the Puerto Rican finished-product-manufacturing operations were both responsible for, and strove to implement, Medtronic's quality standards. (App. 755-761, Add. 14-20; App. 534-536.) Almost 90% of the personnel responsible for quality were employed by Medtronic's U.S. operations. (App. 1002.)

E. Intercompany agreements

During 2005-2006, the following intercompany agreements created by Medtronic were in effect with regard to its production of the Devices/Leads (App. 815-817, Add. 74-76):

• components supply agreement: Medtronic's U.S. manufacturing divisions provided Medtronic-P.R. the critical components needed to manufacture the Devices/Leads. (App. 778-779, Add. 37-38.)

• distribution agreement: Med USA distributed and sold the Devices/Leads that Medtronic-P.R. manufactured. (App. 779, Add. 38.)

• trademark license: Medtronic granted Medtronic-P.R. the right to use Medtronic's trademarks in exchange for a royalty. (App. 780, Add. 39.)

• technology licenses: Medtronic granted Medtronic-P.R. the exclusive right to use Medtronic's clinical data, designs, know-how (both existing technical manufacturing information and future improvements), copyrights, patents, regulatory approvals, and trade secrets to make and sell Devices/Leads in exchange for Medtronic-P.R.'s agreement to pay a royalty based on its net sales to Med USA. (App. 776-778, Add. 35-37; App. 177-194.)

This appeal concerns the technology licenses (Intercompany-Licenses).

During 2005-2006, Medtronic was the “market leader” regarding the Devices/Leads, and its operating profit margins for those products were extraordinary. (App. 877, 1370-1371, Add. 136, 175-176.) Medtronic's exceptional profitability was due in large part to its intangibles. (Id.; App. 541.)

The profit generated by the Devices/Leads was allocated among Medtronic, Med USA, and Medtronic-P.R. through Medtronic's pricing of its intercompany agreements. (App. 367, 371.) Pursuant to those agreements, Medtronic allocated over 65% of the profit to Medtronic-P.R. for its final-product manufacturing and the remainder to Medtronic's U.S. operations for all of the other functions. (Id.) In particular, Medtronic allocated 4% of the profit to its U.S. component-manufacturing operations (App. 599), even though Medtronic's U.S. and P.R. manufacturing operations engaged in similar activities and incurred similar costs (App. 370).

During the years at issue, Medtronic self-insured against product-liability risk. (App. 1350, Add. 155.) That risk exposed Medtronic to both “direct” and “indirect” costs. (App. 1358, Add. 163.) The Intercompany-Licenses required Medtronic-P.R. to pay the direct costs due to product recalls and defects. (App. 1353, Add. 158; App. 181, 190.)

F. Audit of Medtronic's transfer pricing

After auditing Medtronic's consolidated U.S. tax returns for 2005 and 2006, the IRS determined that Medtronic's pricing of its intercompany agreements did not clearly reflect income and that the best method for pricing those agreements was the CPM. (App. 811-812, Add. 70-71.) Applying the CPM, the IRS further determined that the royalty rates that Medtronic-P.R. paid for the intangibles under the Intercompany-Licenses — and thus Medtronic's U.S. income — were too low, resulting in tax deficiencies for 2005-2006. (Id.) Medtronic contested those determinations. (App. 813-814, Add. 72-73.)

G. Original Tax Court proceedings

During the Tax Court proceedings, the parties disputed whether the CPM or the CUT method was the best method for determining an arm's-length royalty rate for the Intercompany-Licenses.

1. The Commissioner's transfer-pricing method (CPM)

The Commissioner argued that the CPM is the best method for determining an arm's-length price for the Intercompany-Licenses. (App. 829-830, Add. 88-89.) Under the CPM, an arm's-length result is determined by computing the profits that would have been earned by the “tested party” if its profitability were the same as that of uncontrolled comparable companies performing similar activities under similar circumstances. (App. 558-559.) To apply that method, the Commissioner's expert (Michael Heimert) selected Medtronic-P.R. as the tested party because Medtronic owned virtually all of the intangibles and performed more functions. (Id.; App. 1175.)

The first step under the CPM is to select companies that are comparable to the tested party (i.e., Medtronic-P.R.), focusing on the assets utilized and the risks incurred rather than the products made or the functions performed. Reg. §1.482-5(c)(2)(ii)&(iii). Heimert selected companies that devoted most of their resources to manufacturing implantable or invasive medical products, were responsible for product quality, and faced similar risks that Medtronic-P.R. faced, including product-liability risk. (App. 592-593, 635-643, 653-666.) Like Medtronic-P.R., all were subject to FDA regulation and, except one, engaged in finished-product manufacturing. (Id.; App. 943.)

The second step under the CPM is to select a profit-level indicator and apply it to the financial information of the comparable companies to determine their profitability. Reg. §1.482-5(b)(4)(i)-(iii). Because manufacturers like Medtronic-P.R. depend on their operating assets to generate profits, Heimert selected the return-on-operating-assets, Reg. §1.482-5(b)(4)(i), as the profit-level indicator. (App. 463-464.)

Applying the return-on-operating-assets ratio to the financial information for the selected companies and for Medtronic-P.R, Heimert computed a range of returns earned by the comparable companies and compared it to the returns earned by Medtronic-P.R. (App. 463-468, 595-598.) He determined that, in 2005, the comparable companies had a median return of 28.1% on their assets compared to Medtronic-P.R.'s 210.7% return on its assets, and that, in 2006, the comparables had a median return of 26.0% on their assets compared to Medtronic-P.R.'s 300.2% return on its assets. (App. 596-597.) Relying on the median return, Heimert calculated an arm's-length return for Medtronic-P.R. and the resulting royalty rates for the Intercompany-Licenses for 2005 and 2006. (App. 600.)

2. Medtronic's transfer-pricing method (CUT)

Medtronic argued that the CUT method was the best method for determining an arm's-length price for the Intercompany-Licenses. (App. 813-814, Add. 72-73.) To apply that method, its expert (Louis Berneman) applied its “best comparable,” a cross-license between Medtronic and one of its competitors, Siemens Pacesetter (the Pacesetter Agreement). (App. 862, Add. 121.) Medtronic and Pacesetter entered into that cross-license in 1992 in order to settle several pending lawsuits. (App. 1356-1357, Add. 161-162; App. 195-266.) The cross-license involved only certain Medtronic patents and specifically excluded all of its other intangibles. (App. 1371, Add. 176; App. 200-201.) Pacesetter agreed to pay Medtronic $75 million plus a 7% royalty on future sales of devices covered by Medtronic's cardiac patents. (App. 1356-1357, Add. 161-162.) In computing a royalty rate for the Intercompany-Licenses, Berneman adjusted the Pacesetter 7% rate to 17%. (App. 865, Add. 124.)

3. Tax Court opinion

The Tax Court determined that neither party's transfer-pricing analysis was reasonable. (App. 871-872, Add. 130-131.) The court first rejected the Commissioner's CPM analysis because (in the court's view) it relied on medical-device manufacturers that performed more functions than Medtronic-P.R. and made different products. (App. 850-858, Add. 109-117.)

The Tax Court next rejected Medtronic's CUT analysis, finding that it failed to account for the significant differences between the Pacesetter Agreement and the Intercompany-Licenses. (App. 867-870, Add. 126-129.) In particular, the court faulted Berneman for “unacceptably lack[ing] an examination of the profit potential of his comparable transactions,” as required by the regulations. (App. 870, Add. 129.)

Having rejected both parties' transfer-pricing analyses, the Tax Court nevertheless concluded that the CUT method, with adjustments, was the best method and that the “Pacesetter Agreement is an appropriate CUT” for the Intercompany-Licenses. (App. 874, Add. 133.) The court increased Berneman's adjusted royalty rate of 17% to 30% to account for certain differences. (App. 878, Add. 137.) The Tax Court issued a decision consistent with its opinion. (App. 886-887, Add. 220-221.)

H. Prior appeal

On appeal, this Court vacated and remanded the decision, holding that the “factual findings are insufficient” to allow the Court to evaluate whether the Tax Court correctly accepted the Pacesetter Agreement as a CUT and whether it “applied the best transfer pricing method.” Medtronic, 900 F.3d at 614-615. In particular, the Tax Court “did not evaluate how the different treatment of intangibles affected the comparability of the Pacesetter agreement” and the Intercompany-Licenses, given that the former was “limited to patents and excluded all other intangibles” and the latter included the full panoply of intangibles required for manufacturing and sale of the Devices/Leads. Id. The Court also directed the Tax Court to make a “specific finding as to what amount of risk and product liability expense was properly attributable” to Medtronic-P.R. so as to allow the Court to evaluate the Tax Court's (i) rejection of the CPM, and (ii) allocation of almost 50% of the relevant profits to Medtronic-P.R. when (according to the Commissioner) it bore only 11% of the costs. Id. at 615. The Court remanded the case and directed the Tax Court to make the required findings.

Judge Shepherd wrote a concurring opinion, stating that he had “serious doubts that the Medtronic-Pacesetter agreement can serve as an appropriate CUT.” Medtronic, 900 F.3d at 618. He emphasized (among other things) that the intangibles included in the Intercompany-Licenses but not the Pacesetter Agreement were “varied — and valuable.” Id. at 617.

I. Tax Court proceedings on remand

On remand, Medtronic continued to assert that the Pacesetter Agreement is a CUT and that the CUT method is the best method to determine an arm's-length royalty rate for the Intercompany-Licenses. (App. 1386, Add. 191.) Medtronic proposed different adjustments and resulting royalty rates than in the original proceeding. (Id.)

The Commissioner argued that the Pacesetter Agreement was ineligible under the regulations to be a CUT because it involved different intangibles with different profit potential. (App. 1121-1122, 1326-1327.) He continued to advance the CPM as the best method, and proposed a modified version of that method. (App. 1381-1382, Add. 186-187.) To address the Tax Court's concern that some of the companies did not make class III implantable medical devices and to account for the product-liability risk assumed by Medtronic-P.R., the Commissioner proposed that the CPM could be modified as follows (App. 1328-1339):

• Limit Comparables: The comparable companies utilized in the modified CPM were limited to the five that made class III implantable medical devices as a significant portion of their manufacturing operations (Implantables).⁵ (App. 1381, Add. 186; App. 943, 955-956 & n.144.) Like Medtronic-P.R., the Implantables (i) bore quality, product-liability, and market risk, (ii) were subject to FDA regulations, and (iii) had routine manufacturing intangibles. (App. 938-948, 1130-1131, 1179-1181, 1240-1247.)

• Adjust for product-liability risk: The Commissioner proposed allocating an additional $25 million of profit per year to Medtronic-P.R. to account for product-liability risk borne by Medtronic-P.R. not otherwise captured by his CPM analysis. (App. 1380, Add. 185; App. 1265-1266, 1273.) That adjustment was designed to reflect the fact that Medtronic-P.R. self-insured against potential product-liability claims for the class III Devices/Leads, whereas the Implantables (which also bore product-liability risk) had insurance that covered some of their claims.⁶ (Id.)

• Use the upper quartile return-on-assets. To increase the profit allocable to Medtronic-P.R., and to account for any remaining material differences, the Commissioner adjusted his CPM analysis by utilizing the upper quartile return-on-assets rather than the median. (App. 1272-1274, 1339.)

Applying these adjustments, the Commissioner (i) calculated the Implantables' return on their assets (using the book value of those assets), (ii) produced an interquartile range of results from those calculations, (iii) selected the upper quartile return-on-assets, and (iv) applied that percentage to Medtronic-P.R.'s operating assets, after incorporating a $25 million per-year adjustment for product-liability risk. (App. 1273-1274.) The result allocated to Medtronic-P.R. 14% of the total Devices/Leads profits for 2005 and 12% for 2006, amounts that exceeded Medtronic-P.R.'s share of third-party costs for those years (i.e., 9% and 12%). (App. 1024-1025, 1274.)

During post-trial briefing, Medtronic proposed, as an alternative, an “Unspecified Method” — a method that incorporated aspects of the parties' CUT and CPM methods and then allocated residual profits between Medtronic and Medtronic-P.R. (using two alternative profit splits — a 35/65 allocation favoring Medtronic-P.R. and a 50/50 split). (App. 1276-1298.) Pursuant to Reg. §1.482-4(d), pricing methods that are not specified in the regulations are permissible so long as they satisfy the applicable regulatory requirements. (Id.) As proposed by Medtronic, the Unspecified Method had the following steps:

• CUT-Method Step: Treat the Pacesetter Agreement as if it were a CUT and use its royalty rate (with certain adjustments) to price Medtronic's R&D function;

• CPM-Method Step: Modify the Commissioner's CPM by increasing the amount of Medtronic-P.R.'s operating assets and then applying the Implantables' median return-on-assets to those assets to price Medtronic-P.R.'s finished-manufacturing function; and

• Profit-Split Step: Split the remaining profit from the Devices/Leads between Medtronic and Medtronic-P.R. on either a 35/65 basis (favoring Medtronic-P.R.) or 50/50 basis, using the Pacesetter Agreement as supporting evidence.

(App. 1392-1399, Add. 197-204.) Under this proposal, Medtronic-P.R. would receive 49% or 43% of the profit attributable to the Devices/Leads (depending on the profit split used). (App. 1296-1297.)

The Commissioner objected to Medtronic's proposed Unspecified Method on numerous grounds, including that it did not qualify as the “best method” under the regulations. (App. 1300-1325.) Specifically, the Method was disqualified because (i) steps one and three were based on an uncontrolled transaction (Pacesetter Agreement) that failed the regulatory requirements applicable to pricing based on uncontrolled transactions (including Reg. §1.482-4(c)), and (ii) step two erroneously (and grossly) inflated Medtronic-P.R.'s assets and lacked any evidentiary support.⁷ (App. 1256-1259, 1300-1320.)

After hearing from ten experts on remand, the Tax Court rejected the outcome reached in Medtronic I. (App. 1372, Add. 177.) As the court explained, “the Pacesetter agreement is not a CUT” for the Intercompany-Licenses (App. 1369, Add. 174), and therefore “the CUT is not the best method” (App. 1388, Add. 193). The court found that the two agreements were substantially different in three critical ways. (App. 1369-1371, Add. 174-176.) First, the intangible property was materially different, because Pacesetter obtained only patents, whereas Medtronic-P.R. obtained the “full array of intangible property” necessary to manufacture Devices/Leads for sale, including patents, know-how, regulatory approvals, secret processes, and copyrights for the software. (App. 1351-1352, 1371, Add. 156-157, 176.) Second, the Pacesetter Agreement lacked “the same profit potential as defined in Treasury Regulation §1.482-4(c)(2)(iii)(B)(1)(ii).” (App. 1386-1387, Add. 191-192.) In so ruling, the court relied on expert testimony that the Intercompany-Licenses' profit potential was 34.6% greater than that of the Pacesetter Agreement. (App. 1391, Add. 196 (citing App. 1055).) Third, the agreements involved substantially different functions, because Medtronic-P.R. performed only finished-product manufacturing whereas Pacesetter also performed component manufacturing, R&D, and distribution. (App. 1370, Add. 175.)

The Tax Court next determined that the Commissioner's modified CPM could not be the “best method” because it relied on companies that made products different than those made by Medtronic-P.R. (App. 1386, 1389, Add. 191, 194.) The court further concluded that the Implantables had different asset bases, functions, and risks than those of a class III medical-device manufacturer. (App. 1383, Add. 188.) The court also rejected the Commissioner's proposed CPM adjustment for the product-liability risk and expense assumed by Medtronic-P.R. (App. 1380-1381, Add. 185-186.) But it made no finding regarding the amount of that risk and expense, other than finding that the amount proposed by the Commissioner ($25 million per year) was insufficient. (Id.)

Instead of adjusting the Commissioner's CPM, as envisioned by the regulations, the Tax Court instead adopted Medtronic's Unspecified Method, with adjustments. (App. 1405, Add. 210.) It adjusted the third step by allocating 80% of the residual profit to Medtronic rather than the 35% or 50% proposed by Medtronic. (Id.) The court found that, without that adjustment, the Unspecified Method would have violated Section 482's “commensurate with income” requirement because it would have allocated over 40% of the profit to Medtronic-P.R. even though it bore less than 15% of the costs. (App. 1403, 1407-1408, Add. 208, 212-213.) The royalty rates determined by the court under the adjusted Unspecified Method resulted in 31.28% of the profit from the Devices/Leads being allocated to Medtronic-P.R. (App. 1407, Add. 212.) The court briefly concluded that Medtronic lacked realistic alternatives to ceding this profit to Medtronic-P.R. based on its determination that replicating Medtronic-P.R. would require “substantial time and costs.” (App. 1399, Add. 204.)

The Tax Court adopted Medtronic's Unspecified Method even though it found that the first two steps were unreliable and flawed in ways that allocated too much income to Medtronic-P.R. In this regard, the court found that step one (i) treated the Pacesetter Agreement as a CUT even though it could “not result in a reliable CUT” given the different intangibles involved (App. 1387, Add. 192) and (ii) included a royalty-rate adjustment for Medtronic's non-patent intangibles that was not “high enough” (App. 1401, Add. 206). The court found that step two inflated” Medtronic-P.R.'s assets in a manner that the “evidence does not support,” resulting in an asset computation that was “too high.” (App. 1402, 1405, Add. 207, 210.) But in the court's view, adjusting step three's residual-profit split offset the acknowledged “imperfections” and unreliability of the first two steps. (App. 1405-1406, Add. 210-211.) The court cited no evidence supporting its assertion that this adjustment in fact offset the problems it had identified.

SUMMARY OF ARGUMENT

This dispute involves the classic case of a U.S. multinational taxpayer (Medtronic) shifting income from its highly profitable U.S. operations to an offshore subsidiary operating in a tax haven (Medtronic-P.R.), by charging an artificially low royalty rate for extraordinarily profitable intangibles. The Commissioner reallocated that income using a CPM analysis, a transfer-pricing method designed to price highly profitable intangibles in a manner that accurately reflects the relative economic contributions of related parties. The Tax Court rejected the Commissioner's CPM method, and adopted a method that was developed by Medtronic's attorneys in a post-trial brief. In doing so, the court erred as a matter of law.

1. The Tax Court's rejection of the Commissioner's transfer-pricing method conflicts with the Section 482 regulations. In particular, the court failed to apply regulatory rules that were designed to facilitate the Commissioner's profit-based approach for situations where (as here) there is no comparable uncontrolled transaction. Moreover, the court failed to make the necessary findings to support its legal conclusions, including findings specifically requested by this Court in the prior appeal, or consider adjustments that the regulations envision.

2. The Tax Court's adoption of the Unspecified Method also conflicts with the regulations. That method cannot — as a matter of law — be the best method because it (i) relies on a transaction that the regulations prohibit using to price the Intercompany-Licenses, (ii) is facially unreliable according to the court's own findings, and (iii) does not satisfy the mandatory realistic-alternative principle.

The Tax Court's decision should be reversed and the case remanded so that the court can correct its legal errors, make the required findings, and, to the extent necessary under those findings, further adjust the Commissioner's modified CPM, the only method considered by the court that is eligible under the regulations to price Medtronic's uniquely valuable intangibles.

ARGUMENT

The Tax Court's rejection of the CPM as the best method for pricing Medtronic's highly valuable intangibles and adoption of the Pacesetter-based Unspecified Method violate Treasury's Section 482 regulations

Standard of review

This Court reviews the Tax Court's “decision in the same manner as a district court's civil bench ruling.” Medtronic, 900 F.3d at 613. Findings of fact are reviewed for clear error, and legal conclusions and “mixed questions of law and fact that require consideration of legal concepts” are reviewed de novo. Smoky Hills Wind Project II, LLC v. City of Indep., Mo., 889 F.3d 461, 470 (8th Cir. 2018).

A. Introduction

As noted above (at pp.6-7), Congress amended Section 482 to require that consideration for intangible property transferred between related parties be “commensurate with the income attributable to the intangible.” I.R.C. §482. Through that amendment, Congress sought to remedy the “recurrent problem” of parties pricing certain “high value” intangibles transferred to “low tax jurisdiction[s]” based on uncontrolled  transactions that were not in fact “comparable” to controlled transactions. H.R. Rep. No. 99-426, at 423-426. Congress directed “that the division of income between related parties reasonably reflect the relative economic activity undertaken by each” and that Treasury develop alternative methods for pricing high value intangibles that did not require finding comparable transfers of those intangibles. H.R. Rep. No. 99-841, at II-637-638.

Treasury implemented that mandate in 1994 by promulgating regulations that (i) tightened the standards for utilizing uncontrolled transactions to determine arm's-length pricing for intangibles (as with the CUT method) and (ii) provided alternative profit-based methodologies (such as the CPM) that could be used to ensure that income allocations related to intangibles are commensurate with the parties' economic contributions. See 59 Fed. Reg. 34,971 (promulgating Reg. §§1.482-1, 1.482-4, and 1.482-5).

The Pacesetter Agreement is exactly the type of unreliable “comparable” transaction that prompted Congress and Treasury to overhaul Section 482 and its implementing regulations. On remand, the Tax Court correctly found that the Pacesetter Agreement could not be used to reliably price Medtronic's highly valuable intangibles because (among other reasons) it involved wholly different intangibles with vastly different profit potential. (App. 1369-1371, 1386-1387, Add. 174-176, 191-192.) But the court nevertheless used that unreliable transaction as the foundation for its Unspecified Method, an unsound artifice cobbled together by Medtronic in a post-trial brief (without any evidentiary support) that “most commentators have criticized.” Bittker & Lokken, Fed'l Tax'n of Income, Estates and Gifts ¶79.8.3 at n.55 (2023); e.g., Ryan Finley, What's at Stake in the Medtronic II Appeal, 112 Tax Notes Int'l 59, 59-60 (Oct. 2, 2023) (observing that on remand, “the Tax Court again rejected the comparable profits method on dubious grounds” and adopted a method that was a “mutation of the CUT method”). The court's rejection of the Commissioner's CPM, and its reliance on the Pacesetter Agreement as a comparable transaction under the Unspecified Method, conflicts with Treasury's 1994 regulations and thwarts Congressional intent.

The Tax Court was required to select the “best method” to price Medtronic's unique and highly valuable intangibles. Reg. §1.482-1(c). It failed to do so here. Its rationale for rejecting the CPM as the best method conflicts with the regulations and lacks supporting findings, including findings requested by this Court in the prior appeal. See, below, §B. And the court's adoption of Medtronic's Unspecified Method as the best method is wrong as a matter of law, because that Method (i) is predicated on an agreement that is categorically ineligible for pricing the intangibles here, (ii) is facially flawed even under the court's own findings, and (iii) violates the regulatory realistic-alternative principle. See, below, §C.

B. The Tax Court rejected the CPM based on a demanding threshold of comparability that conflicts with the regulations and undermines their purpose

As explained above, the CPM evaluates whether the price charged in a controlled transaction (here, the royalty charged to Medtronic-P.R. for use of Medtronic's intangibles) is arm's length according to objective measures of profitability derived from uncontrolled companies that engage in similar business activities under similar circumstances. Reg. §1.482-5. In selecting comparable companies, the “focus [is] on their assets utilized and their risks incurred” rather than on product or functional similarities. Medtronic, 900 F.3d at 612 n.1 (citing Reg. §1.482-5(c)(2)(ii)). If there are differences between the controlled party and the uncontrolled companies, and those differences “would materially affect the profits” determined under the relevant profit-level indicator, then “adjustments should be made” to account for the differences. Reg. §1.482-5(c)(2)(iv).

The Commissioner applied the CPM as provided for in Reg. §1.482-5, and determined an arm's-length profit for Medtronic-P.R. consistent with the returns earned by comparable companies. During the remand proceedings, the Commissioner proposed several modifications to his CPM analysis to address concerns raised by the Tax Court. See, above, pp.23-25. The court's rationales on remand for rejecting outright the Commissioner's modified CPM, rather than making adjustments, conflict with Reg. §1.482-5.

What the regulations make clear — and what the Tax Court failed to recognize — is that the CPM, by design, is a flexible method that best accommodates differences. See 59 Fed. Reg. at 34,974 (explaining that the CPM's “standard of comparability is not as strict as other methods. Some diversity in terms of the functions and products is permitted.”). Unlike a CUT — which cannot be used to price intangibles if it does not involve comparable property, Reg. §1.482-4(c)(2)(iii)(A)&(B)(1) — the CPM permits comparable companies to be used to price intangibles even if there are “significant differences” that cannot be adjusted for, Reg. §1.482-8, Ex. 6. By imposing comparability standards on the CPM that far exceed those in the regulations, the court not only rewrote binding Treasury regulations, but it also weakened a tool critical to tax enforcement that was designed to address Congress's dissatisfaction with methods resting on purportedly comparable transactions like the Pacesetter Agreement.

1. The Tax Court's determination that product differences precluded the CPM from being the “best method” is legally erroneous

The Tax Court concluded that the “adjusted CPM” is not the “best method because of the lack of class III comparables” (App. 1389, Add. 194), focusing on the fact that the Implantables and Medtronic-P.R. made different “medical devices” (App. 1386, Add. 191). Although the court acknowledged that the Implantables made class III medical devices (App. 1384, Add. 189), it found that they did not “solely” make such devices (App. 1386, Add. 191) and did not make the exact same class III products that Medtronic-P.R. made, i.e., “cardio or neuro devices” (App. 1383, Add. 188). The court's assumption that product differences could render a CPM ineligible as a “best method” is wrong as a matter of law.

To begin with, the general comparability factors applicable to all transfer-pricing methods do not require the comparables to make the same products, only that the comparables be “sufficiently similar” so as to provide “a reliable measure of an arm's length result.” Reg. §§1.482-1(d)(2), 1.482-5(c)(2)(i). Moreover, the regulations emphasize that the general comparability factors set out in Reg. §1.482-1 should be tailored to the specific transfer-pricing method at issue, and that when applying the CPM, certain comparability factors are less important. Reg. §§1.482-1(c)(2)(ii)(C), 1.482-1(d)(1), 1.482-5(c)(2)(ii)&(iii). The comparability factor emphasized by the Tax Court — related to the products made by the Implantables — is less important under the CPM than under other transfer-pricing methods. Reg. §1.482-5(c)(2)(iii). As the regulations provide, “differences in product characteristics will ordinarily have a greater effect” on transaction-based methods “than on a comparable profits method analysis.” Reg. §1.482-1(c)(2)(ii)(C). The comparability analysis under the CPM focuses instead on “resources employed and risks assumed,” Reg. §1.482-5(c)(2)(ii), as this Court previously recognized, Medtronic, 900 F.3d at 612 n.1.

The regulations explain why the reliability of a CPM is less sensitive to product comparability than are other transfer-pricing methods, such as the CUT method where it is “particularly relevant.” Reg. §1.482-4(c)(2)(iii)(B). “[P]roduct similarity” is less important because the CPM determines an arm's-length price based on operating profits, and “operating profit usually is less sensitive than gross profit to product differences.” Reg. §1.482-5(c)(2)(iii); see Reg. §1.482-8, Ex. 6. The Tax Court's single-minded focus on product differences conflicts with these regulations.

The Tax Court failed to acknowledge — let alone apply — these CPM-specific provisions, relying instead on the opinion of Medtronic's financial-economics expert (Glenn Hubbard) that “from an economic perspective certain product characteristics should be considered in assessing the comparability of companies.” (App. 1378, Add. 183.) But his perspective was grounded on the legally incorrect premise that the CPM required comparables “exactly like” the tested party and in the “exact same business.” (App. 1138, 1155-1156.) Hubbard's personal opinion cannot displace the mandatory regulations that should have been followed here.⁸

Moreover, the Tax Court was required to select the “best method” — not a perfect method (or even a good method). Reg. §1.482-1(c). The regulations make clear that even a flawed CPM analysis — involving materially different products — can be the best method. Reg. §1.482-8. Example 6 of Reg. §1.482-8 illustrates this point and highlights the intentional flexibility of the CPM. In that example, the CPM was the best method even though there were “significant differences” between the products manufactured by the tested party and those of the comparables, as well as “significant” functional differences between the parties. Id. As this regulation illustrates, even material product differences do not render the CPM per se ineligible from being the “best method,” as the Tax Court incorrectly concluded.⁹ (App. 1389, Add. 194.)

2. The Tax Court's remaining critiques of the CPM cannot withstand scrutiny

Although the Tax Court determined that the CPM was not the “best method” solely on the basis of product differences, it also cited other “problems” with the CPM: in the court's view, the Implantables had “fundamentally different asset bases and involve different functions and risks” than Medtronic-P.R. (App. 1383, Add. 188.) The court, however, failed to make the findings necessary for this Court to evaluate that conclusion and failed to consider — as required by the regulations — adjustments for any relevant differences.¹⁰

a. The Tax Court failed to make findings regarding the amount of product-liability risk and expense attributable to Medtronic-P.R., in contravention of this Court's directive

In the prior appeal, this Court held that the Tax Court failed to make findings that were “essential” to its review of the lower court's determination that the CPM was not the best method because the Commissioner's comparables “did not incur the same amount of risk incurred by” Medtronic-P.R. Medtronic, 900 F.3d at 615. As the Court explained, the Tax Court rejected the CPM “without making a specific finding as to what amount of risk and product liability expense was properly attributable to” Medtronic-P.R. Id. The Court directed the Tax Court to make those specific findings. The court, however, failed to do so.

On remand, the Tax Court again concluded that the Commissioner's CPM is problematic because the Implantables had different “risks” than Medtronic-P.R. (App. 1383, Add. 188.) But it did so — again — without any specific findings regarding (i) the “amount” of Medtronic-P.R.'s risk and product-liability expense, (ii) whether that amount differed materially from that borne by the Implantables, and (iii) whether any material risk could reliably be adjusted for, as the regulations require. Reg. §1.482-5(c)(2)(iv).

i. Amount of Risk. The Tax Court found that Medtronic-P.R. bore “all the product liability risk” because the Intercompany-Licenses state that Medtronic-P.R. is “liable for all costs and damages arising from recalls and product defects.” (App. 1379, Add. 184.) But it did not make findings regarding the amount of that risk and expense during the years at issue. The court recognized that the parties had quantified the amount of product-liability risk borne by Medtronic-P.R., with the Commissioner quantifying it as $25 million per year (assuming that Medtronic-P.R. bore all the product-liability costs for products that it manufactured) and Medtronic quantifying it as $220-235 million per year.¹¹ (App. 1380-1381, Add. 185-186.) The court however did not resolve that dispute, finding only that the Commissioner's proposed adjustment was too low. (Id.)

ii. Risk as differentiating factor. The Tax Court should also make findings that establish whether “risk” is a factor that materially differentiates Medtronic-P.R. from the Implantables. The court's conclusory assertion that Medtronic-P.R. bore materially different risk than the Implantables bore (App. 1383, Add. 188) was not based on any specific findings.

The Tax Court did not — and could not — find that the Implantables were less concerned about product quality or liability than Medtronic-P.R. It is undisputed that the Implantables were class III medical-device manufacturers that were — like Medtronic-P.R. — subject to FDA regulation, registration, and inspection, and that all were concerned about product-liability risk, as their public filings make clear. (App. 267-351, 1152-1153, 1163-1165.) Medtronic's medical-device-industry expert (McCoy) admitted that the “major elements [of quality management]” at the Implantables and their concern about quality would be “quite similar” to that at Medtronic-P.R. because any defect in product quality would expose them to product liability. (App. 1127.) Although the court observed that Hubbard — who was not an expert in the medical-device industry — speculated that “quality is more important for a manufacturer solely of class III devices than for the companies Heimert selected as comparables” (App. 1385, Add. 190), that speculation conflicted with McCoy's testimony¹² and lacked any supporting data. By relying on the returns earned by companies that were concerned with product quality, operated under the same FDA quality standards, and bore all product-liability risk and expense (App. 938-948, 1191-1196, 1241-1242), the Commissioner's CPM analysis properly accounted for risk and the importance of quality in the production process.

Indeed, Hubbard conceded that he could not demonstrate that there was any difference in product quality between the Implantables and Medtronic-P.R., let alone a material difference. (App. 1144-1145.) He further conceded that he did not quantify Medtronic-P.R.'s risk, compare its risk to that of the Implantables, or evaluate the relative cost of a product-liability event for class III products as compared to non-class III products. (App. 1136-1137, 1146-1151; see also App. 1246-1247.)

The fact that Medtronic self-insures against product-liability claims (App. 1350, Add. 155) does not set Medtronic-P.R. apart from the Implantables, which were also “self-insured for a significant portion” of their product-liability exposure. (App. 947; e.g., App. 334.) Any medical-device manufacturer that purchases product-liability insurance self-insures for losses up to the initial retention amount and for catastrophic losses above the policy limits. (App. 1182-1185, 1197-1199.)

iii. Adjustments. Even if self-insurance, or the manufacturing of only class III devices, did present a material difference in risk, the CPM could be adjusted to account for those differences, as Hubbard acknowledged. (App. 1154.) Different levels of risk necessitate an adjustment but do not break comparability under the CPM, as Heimert explained. (App. 1199-1201.) And, as noted above, the parties provided the court dueling computations that quantified the risk of self-insuring class III products made by Medtronic-P.R. (App. 1380, Add. 185.) On remand, the Tax Court should make specific findings regarding the amount of risk assumed by Medtronic-P.R., whether that amount set Medtronic-P.R. apart from the other Implantables, and the amount of any adjustment needed to account for material risk differences.

b. The Tax Court failed to make findings regarding any asset difference between Medtronic-P.R. and the Implantables

The Tax Court also concluded that Medtronic-P.R. and the Implantables had “fundamentally different asset bases.” (App. 1383, Add. 188.) The court did not, however, make any findings regarding what those differences were, whether they negatively or positively impacted the profit allocated to Medtronic-P.R., or whether reliable adjustments could be made to account for any material differences that negatively impacted Medtronic-P.R's profit allocation.

With regard to the operating assets utilized by Medtronic-P.R. and the Implantables, Heimert demonstrated that adjusting for differences in asset composition had no “material” impact on the royalty rate. (App. 1021-1022, 1187-1190.) Hubbard, for his part, complained that Medtronic-P.R. had more account receivables (relative to fixed assets, such as plants and equipment) than the Implantables, but he acknowledged that companies expect smaller returns from account receivables than from fixed assets because they are less risky. (App. 889-892, 897, 1139-1141.) Therefore, this difference would tend to overcompensate Medtronic-P.R. by giving it the same return earned by a company holding riskier assets. See Coca-Cola, 155 T.C. at 226, 229 (applying a CPM and observing that comparables with fewer accounts receivable relative to capital equipment assumed “greater risks” than a controlled party that had relatively more accounts receivable and that applying their return-on-assets to the controlled party's less-risky assets under a CPM “will tend to overcompensate rather than to undercompensate the [controlled party]” and was “therefore conservative”).

As noted above, the Tax Court made no findings to support its conclusion that Medtronic-P.R. and the Implantables had “fundamentally different asset bases.” (App. 1383, Add. 188.) If, after making specific findings on remand, the court were to determine that there are material differences that negatively impact Medtronic-P.R.'s profits, then it should make adjustments to the CPM. See Reg. §1.482-5(c)(2)(iv). Hubbard conceded that adjustments can be made to the CPM to account for asset differences, albeit without proposing any such adjustments. (App. 891-892, 1141-1142.) And the court itself utilized an (albeit grossly incorrect) asset adjustment in the CPM-based step of its Unspecified Method. (App. 1402, 1405, Add. 207, 210.) See, below, §C.2.

c. Functional differences are permitted under the CPM regulations

The Tax Court also concluded that the Implantables performed “different functions” than those performed by Medtronic-P.R. (App. 1383, Add. 188.) Like Medtronic-P.R., the Implantables all performed finished-product manufacturing to a significant degree (App. 936, 943), but, unlike Medtronic-P.R., they also performed additional functions, such as R&D (App. 1384, Add. 189). Those additional functions do not, however, render the CPM unreliable, and, if anything, tend to overcompensate Medtronic-P.R. for its singular function by allocating it the same return on assets that other companies earn for doing more work involving additional valuable functions.¹³ (App. 1019, 1157-1158, 1171-1172, 1185-1187.)

As with product differences, the regulations provide that functional differences are “less” important under the CPM than under other transfer-pricing methods. Reg. §1.482-5(c)(2)(ii); see Reg. §1.482-8, Ex. 6 (illustrating the lesser importance of functional comparability for the CPM). The CPM is more accommodating of functional differences than other methods because “differences in functions performed often are reflected in operating expenses,” so that companies “performing different functions may have very different gross profit margins but earn similar levels of operating profit.” Reg. §1.482-5(c)(2)(ii). This stands in contrast to the CUT method, where functional comparability is “particularly relevant.” Reg. §1.482-4(c)(2)(iii)(B)(2).

The Tax Court ignored Reg. §1.482-5(c)(2)(ii)&(iii) when evaluating the Implantables under the CPM. That those companies performed additional functions, as compared to Medtronic-P.R., does not preclude use of the CPM. Indeed, if it did, the CPM would be useless for transfer pricing. See Finley, 112 Tax Notes Int'l at 64. The controlled party typically performs fewer functions than a comparable because the controlled party is — by definition — part of a larger integrated entity, a point emphasized by Medtronic's transfer-pricing expert (Jonathan Putnam) to explain why he did not adjust for functional differences in applying the CUT method. (App. 1134-1135.) What is important under the CPM, and what the Tax Court failed to appreciate, is that the Implantables all engaged in the finished-manufacturing function to a significant degree. (App. 943.)

The Tax Court also concluded that Medtronic-P.R.'s manufacturing function was “nonroutine” and therefore could not be “benchmarked” to that of the Implantables because Medtronic-P.R. “ensur[ed] product quality and assum[ed] the risk for product liability.” (App. 1378, Add. 183.) But the court did not — and could not — find that the Implantables did not do the same. To the contrary, ensuring product quality and assuming product-liability risk is a routine function performed by the Implantables and others in the medical-device industry. (App. 938-939, 1238-1239, 1248-125.) See, above, pp.45-47; App. 1144-1145, 1238-1249. Accordingly, Medtronic-P.R.'s manufacturing function — including its role in ensuring product quality and assuming product-liability risk — can be “benchmarked” to that of the Implantables.

3. Measured against the objective evidence, the Commissioner's CPM profit allocation is reasonable

Finally, the Tax Court rejected the CPM because (in its view) allocating 12-14% of the profit from the Devices/Leads to Medtronic-P.R. was “unreasonable.” (App. 1390, Add. 195.) The court cited no support for that conclusory determination, stating only that it was “unreasonable for the same reasons” reached by the court in Medtronic I. (Id.) But in Medtronic I, the court provided no reasons for rejecting the CPM's profit allocation, stating only “[i]t is difficult to place an exact value on what [Medtronic-P.R.] contributed to the manufacturing of devices and leads, but it is certainly more than” the amount allocated by the Commissioner. (App. 858, Add. 117.) In so ruling, however, the court failed to measure the profit split resulting from the CPM against the objective evidence or address the numerous formulary metrics that support the Commissioner's allocation of 12-14% of the profit to Medtronic-P.R.

For example, the arm's-length profit allocated to Medtronic's U.S. component-manufacturing operations under the Tax Court's decision was only 4%. (App. 582-583, 599.) Although the court found that component manufacturing and finished-product manufacturing were not identical (App. 853, Add. 112), the court could not — and did not — find that they were so different as to justify allocating almost 8 times as much profit to Medtronic-P.R. for its finished-product manufacturing as was allocated to Medtronic for its component manufacturing, particularly given that Medtronic spent more resources on its component manufacturing than Medtronic-P.R. spent on its finished-product manufacturing (App. 370).

The relative costs borne by Medtronic-P.R. and Medtronic's U.S. operations provides another objective measurement that supports the Commissioner's profit allocation. During the years at issue (2005-2006), Medtronic-P.R. bore approximately 9% (2005) and 12% (2006) of the external costs related to Medtronic's Devices/Leads business. (App. 1025.) Although relative annual profits and costs may not be “strictly proportional” in every case (as Heimert acknowledged (App. 1023)), one premise underlying transfer pricing is that there is an important relationship between the two (App. 1226-1227). The 1986 amendment to Section 482 (which is what the regulations at issue implement) was designed to ensure “that the division of income between related parties reasonably reflect the relative economic activity undertaken by each.” H.R. Rep. No. 99-841, at II-637. Indeed, the Tax Court criticized Medtronic's transfer-pricing method because it allocated more than 40% of the profit to Medtronic-P.R. even though it bore less than 15% of the costs.¹⁴ (App. 1408, Add. 213.) But that same critique applies to the court's method, which allocated more than 30% of the profit to Medtronic-P.R., an amount more than double its share of the costs.

In rejecting the Commissioner's profit allocation as unreasonable, the Tax Court emphasized that Medtronic-P.R. manufactured class III medical devices. (App. 1390, Add. 195.) But the fact that the Devices/Leads were classified by the FDA as class III products weighs in favor of allocating more, not less, in total profit to Medtronic. As the court found, class III products are required to go through a lengthy and expensive preapproval process that can take 5-10 years to complete before the device can be sold in the United States. (App. 749-750, 1345-1346, Add. 8-9, 150-151.) And, as the court further found, Medtronic “bore the significant costs” related to that process. (App. 757, Add. 16.) Accordingly, because Medtronic — not Medtronic-P.R. — heavily invested in the “rigorous, costly, and time consuming” preapproval process required of class III devices (App. 1346, Add. 151), and licensed those valuable regulatory approvals to Medtronic-P.R., the royalty rate should compensate Medtronic for its extensive economic contribution that allowed these class III devices to be manufactured for sale in the first place.¹⁵

C. The Tax Court erred as a matter of law in adopting the Pacesetter-based Unspecified Method as the best method

Section 1.482-4(d) of the regulations permits the use of “unspecified methods” (i.e., methods not specified in the regulations) but imposes certain mandatory requirements that are designed to operate as guardrails against pricing that violates the arm's-length standard. An unspecified method “must” be applied in accordance with the provisions of Reg. §1.482-1. Reg. §1.482-4(d). In particular, an unspecified method “will not be applied unless it provides the most reliable measure of an arm's length result under the principles of the best method rule,” including the reliability of the data and assumptions used (best-method rule). Id. In addition, an unspecified method cannot be applied without first (i) determining the “prices or profits” that a controlled party could have realized in a “realistic alternative” to the controlled transaction and (ii) comparing those to the prices or profits realized under the unspecified method. Id. If the pricing determined under the unspecified method is not “preferable” — that is, if the controlled party could have obtained more profits under the realistic alternative — then the pricing realized under the unspecified method is not arm's length (realistic-alternative principle). Id. The Tax Court failed to properly apply either rule (see, below, §C.2-3), and inexplicably relied on an unreliable transaction (the Pacesetter Agreement) that was ineligible to support any method (see, below, §C.1)

1. The Unspecified Method cannot be the best method because it is predicated on a transaction that is categorically ineligible for pricing the Intercompany-Licenses

The Unspecified Method is fatally flawed because the foundation for the method — the Pacesetter Agreement — was ineligible to price the Intercompany-Licenses. Step one of the Unspecified Method is a “modified version of [Medtronic's] CUT method” and relies on the Pacesetter Agreement's “modified CUT royalty” to price Medtronic's intangibles, as the Tax Court found. (App. 1392-1393, Add. 197-198.) But to be eligible to price intangibles as a CUT (modified or otherwise), the Pacesetter Agreement “must” involve “comparable intangible property” with comparable “profit potential.” Reg. §1.482-4(c)(2)(iii)(A), (c)(2)(iii)(B)(1)(ii) (emphasis added).

The Tax Court's findings make clear that the Pacesetter Agreement cannot satisfy Reg. §1.482-4(c)(2)(iii)'s mandatory requirements. In this regard, the court found that the Intercompany-Licenses involved wholly different intangibles with materially different profit potential than the Pacesetter Agreement:

• The Intercompany-Licenses — unlike the Pacesetter Agreement — involved secret processes, technical information, technical expertise, know-how, regulatory approvals, and over 1,400 additional patents. (App. 1352, 1371, Add. 157, 176.)

• The adjusted Pacesetter royalty rate was not even “high enough” to compensate Medtronic for the “know-how” that was licensed to Medtronic-P.R., let alone all of the other additional intangibles (such as regulatory approvals) licensed to Medtronic-P.R. but not to Pacesetter. (App. 1401, Add. 206.)

• The “magnitude” of the difference in “profit potential” between the Pacesetter Agreement and the Intercompany-Licenses caused the former to fail Reg. §1.482-4(c)(2)(iii)(B)(1)(ii)'s mandatory comparability requirement (App. 1386-1387, Add. 191-192), as well as Reg. §1.482-1(d)'s comparability factor (App. 1370-1371, Add. 175-176).

Those findings regarding the Unspecified Method's foundation preclude it — as a matter of law — from being the “best method.” In order for an uncontrolled transaction to be considered “comparable” so as to be eligible for pricing intangibles in a controlled transaction, it “must” involve “comparable intangible property,” Reg. §1.482-4(c)(2)(iii)(A) (emphasis added) that “must . . . [h]ave similar profit potential,” Reg. §1.482-4(c)(2)(iii)(B)(1)(ii) (emphasis added). Full stop. Although the Tax Court attempted to adjust the Pacesetter Agreement's royalty rate through the subsequent steps in the Unspecified Method, the regulations do not permit adjustments to account for differences between intangible property that is not “comparable,” including because of different profit potential.¹⁶ Id.

The Tax Court also found that the Pacesetter Agreement could not satisfy Reg. §1.482-1(d)'s threshold comparability factors because of differences in functions, economic conditions, and property. (App. 1368-1371, Add. 173-176.) Those differences similarly preclude the Pacesetter Agreement from being used to price intangibles under the Unspecified Method. In this regard, the unspecified-method regulation, Reg. §1.482-4(d), explicitly provides that “[a]ny method used under this paragraph (d) must be applied in accordance with the provisions of §1.482-1.” (emphasis added). Although the court opined that the Pacesetter Agreement could nevertheless be a “starting point” for the Unspecified Method (App. 1388, Add. 193), that conclusion conflicts with Reg. §1.482-4(d)'s instructions. The disqualifying characteristics of the Pacesetter Agreement do not disappear simply by changing the name of the method.

The damage done by the Tax Court's failure to heed the regulatory limitations extends well beyond flawed transfer pricing in this case. By allowing a transaction (the Pacesetter Agreement) that flunks the rules applicable to uncontrolled transactions to serve as the foundation for an unspecified method — essentially rebranding an ineligible CUT — the court invites other parties to “sidestep” Treasury's detailed and mandatory requirements for such transactions. See Finley, Medtronic, Part 2: A High Risk of Bad Precedent, 175 Tax Notes 703, 707 (May 2, 2022) (observing that the Unspecified Method adopted by the Tax Court renders meaningless the “stringent comparability standards in reg. section 1.482-4(c)(2)(iii)” by allowing taxpayers to make them “optional” by “simply tacking on the 'unspecified method' label”).

2. The Unspecified Method is facially flawed and violates the best-method rule

Even if the regulations permitted the Pacesetter Agreement to be used for pricing Medtronic's intangibles, the Unspecified Method is nevertheless precluded from being the best method as a matter of law because each step of that Method is facially flawed. The best-method rule provides that an arm's-length result must be determined under the method that, given the facts and circumstances, provides the “most reliable measure” of an arm's-length result. Reg. §1.482-1(c)(1). Accordingly, even if (as the Tax Court found) there were problems with the CPM, the court should nevertheless have adopted that method, rather than the Unspecified Method, unless the latter was found to be more reliable than the CPM with adjustments. The court did not — and could not — make that finding. First, the court never considered the CPM with all the possible adjustments, as explained above, and thus could not fairly compare it to the Unspecified Method. Second, and based on the court's findings, each step of the Unspecified Method is unreliable, precluding that method from being the “best method” under the regulations.

The Tax Court found that the first step of Medtronic's Unspecified Method was unreliable and undercompensated Medtronic for its intangibles. In this regard, the court found that the royalty rate applied in step one was not “high enough” to compensate Medtronic for even a portion of its intangibles. (App. 1401, Add. 206.) More fundamentally, the court found that the foundation for step one — the Pacesetter Agreement — was not “reliable” to price Medtronic's intangibles, even with adjustments. (App. 1387, Add. 192.)

The Tax Court also found that the second step of Medtronic's Unspecified Method was unreliable and overcompensated Medtronic-P.R. for its manufacturing function. That step utilized the CPM but included an asset-intensity adjustment that was — as the court found — “too high” and wholly “unsupported.” (App. 1405-1406, Add. 210-211.) As the court explained, the adjustment “inflated” Medtronic-P.R.'s revenues to include “sales attributable to contributions by Medtronic US.” (App. 1402, Add. 207.) The court concluded that it could not correct this unreliable step.¹⁷(App. 1405, Add. 210.)

Finally, the Tax Court found that the proposed third step of Medtronic's Unspecified Method was unreliable because it allocated an unreasonable amount of profit to Medtronic-P.R, even using the proposed residual-profit split that gave more profit to Medtronic (i.e., the 50-50 split). (App. 1407-1408, Add. 212-213.) Specifically, if the court applied the 50-50 residual-profit split, the overall Unspecified Method would allocate 43% of the total profits to Medtronic-P.R. even though it bore less than 15% of the costs — which did not reflect the parties' relative economic contributions as required by Section 482. (Id.) But the court failed to realize that its profit allocation under the adjusted Unspecified Method (discussed below) was similarly unreasonable, allocating Medtronic-P.R. 31.28% of the profits when its economic contributions for the years at issue were less than half that amount. (Id.)

The Tax Court attempted to offset the unreliability and arbitrariness of the Unspecified Method's first two steps — errors that shifted income to Medtronic-P.R. and away from the U.S. taxing jurisdiction — by allocating less residual profit to Medtronic-P.R. in step three than Medtronic had proposed. According to the court, “making a greater allocation of the remaining profits in step three to Medtronic US than to [Medtronic-P.R.]” would eliminate “concerns” about steps one and two. (App. 1405-1406, Add. 210-211.) But the court did not explain how it could reliably correct in step three errors that it could not even quantify (let alone reliably correct) in the other steps. To be sure, as the court observed, “allocating more of the remaining profits in step three to Medtronic US” results in a “higher royalty rate.” (App. 1405, Add. 210.) Indeed, the entire exercise was premised on the concept that allocating profits to Medtronic would set the royalty rate. But the court made no findings establishing that the ultimate rate was high enough.

In short, without quantifying the magnitude of the errors in step one and step two, the Tax Court made it impossible to evaluate whether its adjustment in step three adequately addressed the “concerns” with those flawed steps or whether its purported “best method” analysis (App. 1409-1410, Add. 214-215) was sound. That evaluation would require findings regarding (i) the value of the various intangibles not covered by the Pacesetter Agreement (even though those differences in intangibles should bar consideration of the Pacesetter Agreement altogether, see (§C.1)), and (ii) the amount by which step two overinflated Medtronic-P.R.'s assets. Neither finding was made. And, as explained above, step three itself is unreliable because it results in Medtronic-P.R. receiving a profit share that more than doubles its economic contribution.

3. The Tax Court failed to make the findings required by the realistic-alternative principle

Before applying the Unspecified Method, the Tax Court was required to evaluate Medtronic's “realistic alternatives” to allocating more than 31% of the profit to Medtronic-P.R. for its manufacturing function. Reg. §1.482-4(d)(1). This mandatory rule is based on the principle that “uncontrolled taxpayers evaluate the terms of a transaction by considering the realistic alternatives to that transaction, and only enter into a particular transaction if none of the alternatives is preferable to it.” Id. This rule required the court to make specific factual findings regarding the “prices or profits” that Medtronic could have realized if it had chosen an alternative, such as replacing Medtronic-P.R. with another of its manufacturing facilities. Id.; see Reg. §1.482-1(d)(3)(iv)(H).

The Tax Court failed to make any findings regarding “prices or profits” related to Medtronic's alternatives. It acknowledged that Medtronic-P.R.'s manufacturing function “could have been replicated.” (App. 848, Add. 107.) But it rejected that alternative out of hand, stating that replacing Medtronic-P.R. would involve “substantial time and costs.” (App. 1400, Add. 205.) It did not, however, quantify those costs and compare them to the steep cost of retaining Medtronic-P.R. as its manufacturer, despite the fact that the Commissioner submitted evidence on remand that “quantified the expense and time required.” (App. 1303; see also App. 1253-1255, 1266-1267, 1300-1306.) It is that economic comparison that gives meaning to the regulatory realistic-alternative principle. Skipping that analytical step altogether was legal error. See Reg. §1.482-4(d)(2) (illustrating required analysis). What the regulations require, and what the Tax Court failed to provide, is a quantitative corroboration that the controlled transaction at issue is economically preferable to the posited alternative, ceding millions of dollars to Medtronic-P.R.

CONCLUSION

The decision of the Tax Court should be reversed and the case remanded for the court to determine an arm's-length royalty rate for the Intercompany-Licenses utilizing the CPM.

Respectfully submitted,

DAVID A. HUBBERT
Deputy Assistant Attorney General

Judith A. Hagley

FRANCESCA UGOLINI (202) 514-3361
JENNIFER M. RUBIN (202) 307-0524
JUDITH A. HAGLEY (202) 514-8126
Attorneys
Tax Divison
Department of Justice
Post Office Box 502
Washington, D.C. 20044

DECEMBER 14, 2023