The recent decision from the U.S. Tax Court results in an outcome that is anything but tidy. What happened? Read our commentary in this edition of Insights:

MEDTRONIC PART DEUX: THE BEST METHOD IS YET TO COME?

The purpose of the most recent decision in the Medtronic saga[1] extends and refines the prior analysis of one of five connected controlled transactions within Medtronic’s controlled group of multinational medical device producers and suppliers. The transactions may be described as follows:

Medtronic licensed patented and unpatented intangible property related to the design and production of sophisticated medical devices to a controlled Puerto Rican subsidiary (“MPROC”), which served as the manufacturer. The intangible property related to (i) implantable pacemakers, cardioverter defibrillators, cardiac resynchronization devices, neurostimulation devices, and (ii) connective leads.
Medtronic licensed its trademark intangible assets to MPROC.
Medtronic sold manufactured product components and sub-assemblies to MPROC.
MPROC sold finished medical devices to a U.S. group company for resale worldwide.
Medtronic licensed the same intangible assets related to products (i) and (ii) described above to a controlled Swiss manufacturer that began device and leads production operations after MPROC. The Swiss affiliate paid royalties at the same rates as MPROC to Medtronic.[2]

The first two transactions – license of manufacturing intangible property and license of trademarks – were the main subject of a period of examination controversy that concluded with the I.R.S. adjusting the royalty income of the U.S. Medtronic licensor for tax years 2005 and 2006. The adjustments included additional income necessary for the royalty to be arm’s length as determined under the comparable profits method (“C.P.M.”) analyses performed by the I.R.S.

Medtronic’s 2005 and 2006 position originated in an M.O.U. settlement with the I.R.S. involving the same transactions and issues, but in respect of Medtronic’s 2002 tax year. That settlement was based on an agreed division of profit between the U.S. and MPROC but not on a specific transfer pricing method. The settlement outcome nonetheless informed the transfer pricing position of Medtronic for the tax years 2005 and 2006. Medtronic applied the comparable uncontrolled transaction (“C.U.T.”) method to determine its 2005 and 2006 non-trademark income using a licensing agreement between Medtronic and Pacesetter, a Siemens group company active in the cardiac rhythm disease management business, concluded for the purpose of settling a medical device patent litigation matter between the two competing companies in the early 1990’s. Medtronic continued to rely on its C.U.T. method application to argue that the M.O.U. outcome was arm’s length and petitioned the Tax Court to vacate the proposed I.R.S. income adjustments for 2005 and 2006.

The value of the second transaction was resolved easily by the Tax Court using expert evidence of trademark royalty rates. The decision noted there was an error in the calculation of the intercompany inventory sales revenue that would be corrected during the proceedings.[3] Though the focus of the controversy became the non-trademark licensing transaction, all five transactions excepting the Swiss licensing transaction were referenced by Medtronic and the I.R.S. in making arguments about the division of profit between Medtronic and MPROC.

2016 Tax Court Decision

In 2016, the Tax Court decided for the most part for Medtronic, though several adjustments were made to the non-trademark royalty rates (one rate for devices, a different rate for leads) to improve the comparability of the terms and circumstances of the independent Pacesetter agreement primarily relied on by Medtronic with the terms and circumstances of the licensing agreement between Medtronic and MPROC. Much of the 2016 Tax Court decision dealt with the facts of the transactions and the Medtronic burden of demonstrating that the I.R.S. position was arbitrary, capricious, or unreasonable. The I.R.S. was also found to have abused its authority in making the allocation.

On the question of whether the allocations used by Medtronic satisfied the arm’s length method, the Tax Court reasoned that three adjustments were required to meet the comparability standard. These adjustments related to (i) the supply of non-patented design and production know-how, (ii) profit potential, and (iii) the licensed property (cardiac product intangibles as distinct from neurology product intangibles). After adjusting for differences, the Tax Court concluded that the adjusted C.U.T. was the best method. The outcome of the first decision raised the devices royalty rate by 15% and the leads rate by 7%. These rates, the court noted, were not significantly different from those agreed between Medtronic and the I.R.S. in the 2002 M.O.U.

The I.R.S. argued exclusively for its comparable profits method (“C.P.M.”) approach and against the Medtronic C.U.T. method during the examination of tax years 2005 and 2006, during Medtronic’s appeal to I.R.S. Appeals Office, and during the Tax Court proceedings that culminated in the 2016 decision. The I.R.S. provided the court with no expert testimony to assist it in estimating an adjustment to the Medtronic royalty rate to remedy several identified comparability shortcomings identified in the Medtronic C.U.T. analysis.[4]

The I.R.S. appealed the 2016 Tax Court decision, arguing that (i) the decision did not adequately dispose of the question of whether Medtronic’s C.U.T. was the best method, and (ii) the Tax Court did not make sufficient findings of fact to conclude that Medtronic’s C.U.T. met the comparability standards of Treas. Reg. §1.482-4(c)(2). The I.R.S. was successful on appeal to the 8th Circuit Court of Appeals, and the case was remanded to the Tax Court to make the required findings of fact and determine the best method.

The tasks before the Court

As an initial matter, the Tax Court listed the tasks assigned by the 8th Circuit:

Determine whether the Pacesetter agreement is a C.U.T.
Determine whether the Tax Court made appropriate adjustments to the Pacesetter agreement, if it were a C.U.T.
Determine whether the circumstances between Pacesetter and Medtronic were comparable to the licensing agreement between Medtronic and MPROC and whether the Pacesetter agreement was an agreement created in the ordinary course of business
Determine the degree of comparability of the Pacesetter agreement’s contractual terms and those of the MPROC licensing agreement
Determine how the different intangibles in the two agreements affected the comparability of the Pacesetter agreement and the MPROC licensing agreement
Compare and contrast the results under the C.U.T. method using the Pacesetter agreement with or without adjustments with those under the C.P.M., and determine which of the two methods is the best method.

As an introduction to these tasks and its analysis, the Tax Court provided some general commentary to place its assigned tasks in the context of the transfer pricing regulations and the case law.

The Tax Court cited Sundstrand[5] to establish its role in determining whether the I.R.S. position underlying a notice of deficiency was arbitrary, capricious, or unreasonable. When these conditions are met, the next step is for the taxpayer to show that the allocations of income or expense among the related parties that satisfy the arm’s-length standard.[6] The decision explains the role of the court when the position of the I.R.S. is viewed to be unreasonable but the taxpayer does not meet its burden of demonstrating the proper method to be used:

If neither party has proposed a method that constitutes “the best method,” the Court must determine from the record the proper allocation of income. Sundstrand Corp., 96 T.C. at 354. After hearing expert witnesses during further trial and reviewing the parties’ positions, we conclude that there are some benefits to the CUT, and the Pacesetter agreement is an appropriate comparable as a starting point. We are concerned that there is only one comparable, that adjustments need to be made, and that if too many adjustments are made, the Pacesetter agreement might cease to be useful even as a starting point.[7]

Consistent with the decision in Medtronic I, the Tax Court found that the Pacesetter agreement by itself, without any adjustment for comparability, was not a C.U.T. The Tax Court rethought the appropriateness of the adjustments it made in the first trial, and on remand, made different adjustments. Ultimately choosing the application of an alternate, unspecified method but with Pacesetter as its cornerstone despite its stated misgivings. The Tax Court made findings of fact and reached conclusions on the questions posed in tasks 3-5. This set the court up to answer question 6 concerning the best method.

Best method, ideal method, or the method that gets the right answer?

Given the recent decision reads as a retelling of fact mixed with functional, financial, and legal analysis, a grounding in the basic principles of the transfer pricing regulations is a logical place to explain the reasoning of the court. The regulations implement the goal of Code §482, “to prevent evasion of taxes or clearly to reflect the income of any of such organizations, trades, or businesses” by determining the true taxable income of a controlled taxpayer.[8] When determining the true taxable income of a controlled taxpayer, the standard to be applied in every case is the arm’s length standard.[9] Evaluation of whether a controlled transaction produces an arm’s length result is made pursuant to a method selected under the best method rule described in Treas. Reg. §1.482-1(c).[10]

The best method rule is a familiar standard used to evaluate all transfer pricing positions in practice. It is fundamental to all transfer pricing analysis, and has been mimicked as the “most appropriate method” in the most recent multilateral guidance from the O.E.C.D. The regulations define the best method to be the transfer pricing method that provides the most reliable measure of an arm’s length result under the facts and circumstances of the taxpayer.[11] The best method rule is applied by evaluating five criteria:

The degree of comparability between the controlled transaction or taxpayers and the uncontrolled comparables, after making adjustments for differences.
The completeness and accuracy of data.
The reliability of assumptions.
The sensitivity of results to deficiencies in data and assumptions.
Confirmation of results by application of another method.[12]

In Medtronic I and Medtronic III, the Tax Court decided that the C.P.M. applied by the I.R.S. did not meet any of the criteria of the best method rule, and accordingly, was not the best method. Medtronic’s C.U.T. method was the only other candidate specified method, making the Court’s analysis an assessment of reliability of the taxpayer’s adjusted C.U.T. method under the best method rule as distinct from a determination of which method, adjusted C.U.T. or C.P.M., was relatively more reliable. The outcome of this analysis would have been the Court’s reply to the sixth task from the Eighth Circuit. The recent decision applied the best method criteria to Medtronic’s adjusted C.U.T. as summarized below:

The Degree of Comparability Between the Controlled Transaction or Taxpayers and the Uncontrolled Comparables, After Making Adjustments FOR Differences

The Court found that the MPROC and Pacesetter agreements were not comparable. Medtronic did not perform research and development functions in connection with the intangible property licensed to Pacesetter, but performed this function in connection with intangible property licensed to MPROC.

A significant effort was spent by the court to address a perceived difference in profit potential for the parties in the MPROC and Pacesetter licensing transactions. The intangible property regulations are clear that profit potential should be measured from reasonable expectations generated at the point of negotiation, rather than a look-back analysis a profit outcome from a point in the future.[13] Consequently, profit potential is defined in terms of the net present value of the stream of reasonably anticipated profits. The I.R.S. and the Court seemingly referenced only actual profits or sales as opposed to forecasted or expected profitability when assessing comparability of the profit potential under the MPROC and Pacesetter nontrademark licensing arrangements. In other words, the prescience of the taxpayer was assumed.

The court also noted that the intangible property licensed to MPROC consisted of 1,800 patents and know-how, while only 342 Medtronic patents were licensed to Pacesetter.

The terms of the MPROC and Pacesetter agreements were found to be sufficiently comparable, and the fact that the Pacesetter agreement arose out of litigation settlement was not found to constitute a sufficiently different economic condition for the purpose of determining comparability. No changes in technology or in industry conditions were found between the MPROC and Pacesetter agreements.

Medtronic made adjustments to its C.U.T. to quantify the difference in royalty rate that resulted from profit potential, the supply of know-how, different license portfolios (and numbers of patents), differences in the licensing terms among cardio and neuro products, and the Pacesetter cross-license. While little detail of the adjustments is discussed in Medtronic III, no material difference goes unadjusted under the proposed adjusted C.U.T. method.

Completeness and Accuracy of Data

No mention of shortcomings was made as to data quantity, completeness, or quality used in applying the adjusted C.U.T. in Medtronic III.

Reliability of Assumptions

None of Medtronic’s assumptions underlying its C.U.T. adjustment calculations were directly faulted. All I.R.S. criticisms were appropriately weighed and dismissed by the Tax Court.

Sensitivity of Results to Deficiencies in Data and Assumptions

While there were ranges of estimates used to adjust the C.U.T. method outcome to account for noted differences, the usual criticism of breadth of ranges appears somewhat unfounded in the Tax Court’s decision, given the ranges suggested from other method applications and from C.U.T. analysis examined in Medtronic I that were similarly broad.

Confirmation of Results by Another Method

This criterion is not a mandatory item but was noted as a means of demonstrating the reliable nature of one method by using a secondary method. Medtronic applied an unspecified method in addition to its adjusted C.U.T. Its wholesale royalty rate result using a 35/65 residual profit split as a final step in a multi-step calculation was 2.6% higher (35.7%) than its high value of the adjusted C.U.T. range (33.1%).

If two or more methods produce inconsistent results (as the C.P.M. and the adjusted C.U.T. did), the best method rule will be applied to select the method that provides the most reliable measure of an arm’s length result.[14] The similarity of the unspecified method and adjusted C.U.T. outcomes points to the adjusted C.U.T. being highly reliable.

Despite the reporting of the outcome of the best method rule when applied to Medtronic’s adjusted C.U.T., the Tax Court concluded that Medtronic’s adjusted C.U.T. was not the best method. It proceeded to amend Medtronic’s unspecified method in a way that can only be described as bizarre as the foundation for its opinion.

The Tax Court analysis seems to search for the ideal method or an absolute standard, contrary to the best method rule set out in the regulations. Several mentions of the division of profit between MPROC and Medtronic resulting from different method applications suggest the court prioritized an arbitrary, but somehow intuitively appealing, profit split to determine the method applications that produced reasonable outcomes and other method applications that fell short. The court stated its “goal was to find the right answer,”[15] but this goal is stated neither by the regulations nor the 8th Circuit in its instructions on remand. More importantly, the court’s reasoning for identifying the best method and comparability subfactor determinations adopted a standard that evaluates results as “too high,” “too low,” or “too many.” In other words, the court adopted a standard under which the ends justified the means, which differs from the norm under which the means justify the ends. The court knew where it wanted to end up, and simply looked for a method that was consistent with its destination.

An old saying of trial lawyers is that bad facts make bad law and Medtronic III is rife with bad facts.

A wide margin existed between the positions of the I.R.S. and the taxpayer.
The dispute between the parties was endless. The dispute involved the tax years 2005 and 2006 and Medtronic III was decided in 2022.
The amounts in question were high.
The precedential value of the final determination could be enormous.
The presence of an unusual fact subsequent to the initial decision in Medtronic I – the taxpayer sought a refund as a result of the overturn of the M.O.U. outcome.

The recently decided controversy in Eaton[16] over the I.R.S. cancelling an APA (not immediately relevant, but these were disputes of the same nature in many ways) may have encouraged the court to fashion a compromise decision that intends to disappoint both sides while limiting the scope for appeal.

Unique unspecified method use

The use of an unspecified method is without I.R.S. field guidance and will likely lead to future controversy in examinations if the court’s approach is adopted by field examiners. Relying on an unspecified method to support a transfer pricing position prior to examination is a risky course of action for a taxpayer. Calculating a tax provision that employs an unspecified method involving an affiliate based in a treaty partner jurisdiction through the Competent Authority process would involve a series of novel and uncertain steps.

Will unspecified methods be used to impose reasoned settlements in exams, cases before the Tax Court, and Competent Authority proceedings? It is possible we will see greater use of unspecified methods as parties grow weary from the volume of factual documentation required even before addressing the substantive questions of method selection, especially as tax authorities continue to take “method-agnostic” examination positions that propose income adjustments solely using a functional analysis, rely on unsupported normative arguments, or argue for largely unsupported profit-split positions.

As was true of Medtronic I, all cases are fact-specific, and not all fact patterns or method selection questions will lend themselves to the use of an unspecified method.

The “No-recipe recipe”

In retrospect, the question that remains after reading the decision in Medtronic III is whether the court adopted an unspecified method to resolve a transfer pricing dispute, or did it simply adjust a C.U.T. in a subjective way to obtain a result that seemed to be fair? In the view of the author, the method used by the court was in substance an adjusted C.U.T., a specified method. The profit split method uses more than one method, strictly speaking, to arrive at an income allocation and remains squarely in the category of specified methods.

Nonetheless, the problem with Medtronic III, is that two transfer pricing methodologies that were each deemed to be unreliable were packaged together using a no-recipe recipe as an unspecified method replete with a number of largely unsupported adjustment factors to reach a conclusion. The second trial did not produce practical guidance that was any better than the very limited guidance that emerged from the first.

If not appealed, the only virtue of the decision in Medtronic III is that it will have the force of law strictly between the taxpayer involved and the I.R.S. In a way its importance may be similar to a written determination of the I.R.S. that is covered by Code §6110(k)(3).[17] “Unless the [I.R.S.] otherwise establishes by regulations, a written determination may not be used or cited as precedent.” Nonetheless, advisers often point out that a written determination tends to indicate the position of the national office of the I.R.S. at the time issued and can be cited as authority for purposes of eliminating a penalty for the understatement of tax.

More than anything, companies want transfer pricing certainty when calculating tax provisions and disclosing uncertain positions or transactions with B.E.P.S. hallmarks. The decision in Medtronic III demonstrates that this certainty may not be obtainable even at a high price. Arbitration emerges again as a strong dispute resolution candidate mechanism given the practical alternatives

[1] Medtronic, Inc. & Consol. Subs. v. Commr., T.C. Memo. 2022-84 (“Medtronic III”), on remand from 900 F.3d 610 (8th Cir. 2018) (“Medtronic II”), vacating and remanding T.C. Memo. 2016-112(“Medtronic I”).

[2] It was agreed by the parties that the Swiss royalty rate(s) would be adjusted to equal the MPROC royalty rate(s) determined by the court. No detailed facts of this transaction or of any transfer pricing analysis were presented.

[3] This sales revenue was critical to the final determination of the arm’s length amount of the royalty because the royalty was a function of the sales revenue and an acceptable royalty rate.

[4] The I.R.S. did not present an adjusted C.P.M. in the second Tax Court trial, though the Tax Court commented that such an analysis would have been considered if presented.

[5] Sundstrand Corp. v. Commr., 96 T.C. 226, 353 (1991) (citing G.D. Searle & Co. v. Commr., 88 T.C. 252, 358 (1987), and citing Eli Lilly & Co. v. Commr., 84 T.C. 996, 1131 (1985), aff’d. on this issue, rev’d in part and rem’d, 856 F.2d 855 (7th Cir. 1988)).

[6] Eli Lilly & Co. v. Commr., 856 F.2d 855, 860 (and the cases cited therein).

[7] Medtronic III at p. 49.

[8] Treas. Reg. §1.482-1(a)(1)

[9] Treas. Reg. §1.482-1(b)(1)

[10] Id.

[11] Treas. Reg. §1.482-1(c)(1)

[12] Treas. Reg. §1.482-1(c)

[13] Treas. Reg. §1.482-4(c)(2)(iii)(B)(ii)

[14] Treas. Reg. §1.482-1(c)(2)(iii)

[15] T.C. Memo. 2022-84, p. 71

[16] Eaton Corp. & Subsidiaries v. Commr., T.C. Memo. 2017-147, supplemented by 153 T.C. 119 (2019) and an earlier decision, 140 T.C. 410 (2013), aff’d and rev’d in part 6th Cir. (Docket No. 21-1569/2674, 8/25/2022) reported unofficially at 30 A.F.T.R. 2d 2022-5746. Eaton involved the unsuccessful attempt by the I.R.S. to cancel an advance pricing agreement with the taxpayer based on certain errors that were self-corrected by the taxpayer and in which substantial transfer pricing penalties of 40% of the asserted increase in tax were unsuccessfully asserted in the post decision computation of tax by the parties.

[17] Examples of a written determination include a private letter ruling, a Technical Advice Memorandum, and Chief Counsel Advice.

Medtronic III: The best method is yet to come?