In Medtronic v. Commissioner, 900 F.3d 610 (8th Cir. 2018), vacating and remanding TC Memo 2016-112, the Eight Circuit considered whether a license arrangement between Medtronic Inc (Medtronic) and Medtronic Puerto Rico Operations Co (MPROC) was arm’s-length under section 482 of the US Internal Revenue Code.
MPROC used the patents and know-how that were the subject of the license agreement to manufacture sophisticated medical devices and leads that it sold to Medtronic USA Inc, which distributed the devices and leads in the US. The IRS had previously raised concerns about the royalty arrangements between Medtronic and MPROC, but Medtronic settled that dispute by entering into a memorandum of understanding (MOU) with the IRS that adjusted the terms of the license agreement.
The IRS subsequently repudiated the MOU, asserting deficiencies of more than $1.3 billion for the two years at issue in the case. Rather than transactionally pricing the royalty arrangement, as was the case in the MOU, the IRS relied on a comparable profits method (CPM) analysis, which benchmarked the overall profit that it believed MPROC should earn for its manufacturing activities. This approach effectively characterised MPROC as a contract manufacturer of medical devices and leads.
In contrast, Medtronic believed that the license agreement could be reliably priced using the comparable uncontrolled transaction (CUT) method set forth in Treas. Reg. § 1.482-4. Specifically, Medtronic asserted that a cross-license agreement that it had with a competitor, Pacesetter, could be used to price the license arrangement with MPROC.
The Tax Court rejected the IRS’s use of the CPM, which the Tax Court found did not take into account the substantial risk relating to quality and product liability regarding the highly technical and sensitive devices and leads that MPROC manufactured. However, the Tax Court also disagreed with aspects of the taxpayer’s CUT analysis, noting that such analysis failed to make adjustments to account for significant differences between the license of devices and leads to MPROC and the Pacesetter agreement, which arose from a litigation settlement. Thus, the court had to fashion an outcome that it believed was arm’s-length. Favouring the transactional method applied by the taxpayer, the Tax Court made certain adjustments to the Pacesetter agreement to arrive at a determined arm’s-length royalty rate.
The Eighth Circuit vacated and remanded the Tax Court’s decision, finding that the lack of certain factual findings prevented it from evaluating whether the CUT was the best method, and whether proper adjustments were made. That decision led to a second Tax Court trial addressing these issues. That trial recently ended, and we are now awaiting a second decision from the Tax Court.
The second trial focused on the selection of the most reliable TP method, as well as the determination of any necessary adjustments to the results of applying the CUT method. In the second trial, Medtronic and its expert witnesses continued to assert that the CUT method was the most reliable method to price the license agreement between Medtronic and MPROC, while the IRS and its experts continued to assert that the CPM was the best method. Specifically, the IRS and its experts asserted that the differences between the Pacesetter agreement and the royalty transaction between Medtronic and MPROC were too significant to be resolved through adjustments and that the application of the CPM to determine the profits earned by MPROC most accurately reflected an arm’s-length result.
At the conclusion of the second trial, Tax Court Judge Kathleen Kerrigan suggested that the use of the Pacesetter agreement as a CUT with adjustments may be the best method. However, she also indicated that it may be possible to use that agreement to price the related party license agreement as an unspecified method. This suggests that Judge Kerrigan remains convinced that the CPM is an unreliable method in this case, but the adjustments needed to make the Pacesetter agreement reliable may be so significant that it would no longer be appropriate to consider such transactional pricing as a valid CUT.
Treas. Reg. §1.482-4(d) recognises that an unspecified method may be the best method to price a related party licensing agreement. Moreover, that regulation and an example in the regulation indicate that such method will be more reliable if it is based on an uncontrolled arrangement, rather than internal data. Thus, Judge Kerrigan may elect to use the Pacesetter agreement as the basis for applying an unspecified method to determine the arm’s-length royalty rate between MPROC and Medtronic, rather than characterising the pricing as an adjusted CUT.
This approach would be consistent with the general principle in Treas. Reg. § 1.482-1(d)(2) that inconsistencies between controlled transactions and comparables are not grounds for discarding the comparables outright. In the Tax Court’s view, an unspecified method grounded in the Pacesetter agreement may still be more reliable than the CPM proposed by the IRS, and using an unspecified method could allay some of the Eighth Circuit’s specific concerns regarding the CUT regulations.
It will be interesting to see how the Tax Court resolves this dispute, including potentially its approach to applying adjustments to the Pacesetter agreement as a CUT, applying an unspecified method relying on the Pacesetter agreement, or rejecting reliance on such agreement fully and applying the CPM.
Mark Martin
Principal, KPMG
Thomas Bettge
Manager, KPMG
