On Thursday, August 18, 2022, Chief Judge Kathleen Kerrigan of the U.S. Tax Court published her 75-page decision on the Medtronic and I.R.S. controversy (“Medtronic III”).[1] This decision resulted from a remand by the Eighth Circuit Court of Appeals (“Medtronic II”) of her original 144-page 2016 decision (“Medtronic I”).[2]
The IRS and Medtronic had agreed to an MOU to apply to the tax years in question of 44 percent for the sales of devices and 26 percent for the sales of leads. The MOU also required Medtronic to undertake an annual profit split analysis with a system profit target for the Puerto Rico operation of 38 percent for devices and 45 percent for leads, Medtronic making additional adjustments if the target was missed by 3 percent on either side. However, during the audit of 2005-06, the IRS did not like the result generated by the MOU and instead applied a CPM. The IRS’ CPM resulted in a reduction of the Puerto Rico operation operating profits to 8.1 percent and 5.6 percent for 2005 and 2006 respectively. This operating profit constraint resulted in a two-year additional tax deficiency of $1,358,481,810.
In Medtronic I, Judge Kerrigan initially determined, using a CUT analysis, a 44 percent royalty rate for Medtronic’s Swiss supply agreement as well as the wholesale royalty rate for its other medical devices – the same result as the MOU. But Judge Kerrigan’s CUT analysis led to an incremental reduction of the MOU’s 26 percent for leads to 22 percent. The much lower royalty rate for the leads is partly explained in that each manufactured lead requires over 100 manufacturing and quality control steps undertaken by the workforce of the manufacturing facility. The CUT analysis was based upon a patent settlement and resulting license agreements of Medtronic and Siemens Pacesetter. The Tax Court, applying its CUT, reduced the $1.3 billion deficiency to merely $14,251,848.
In Medtronic II, the Eighth Circuit Appellate Court stated that the Tax Court did not provide a comparability analysis for the circumstances of the patent settlement and resulting license agreements of Medtronic and Siemens Pacesetter to the licensing agreement of Medtronic and its Puerto Rico subsidiary. The Eighth Circuit stated that the Tax Court should determine whether the Medtronic and Siemens Pacesetter license agreement to resolve litigation was created in the ordinary course of business.[3] The Eighth Circuit cited Treasury Regulation § 1.482-1(d)(4)(iii)(A)(1) that transactions not made in the ordinary course of business will not generally be considered reliable for arm’s length purposes. For a deep dive into the facts and analysis of Medtronic I and II, see William Byrnes, Practical Guide to U.S. Transfer Pricing (4th Ed.).
This brings us up to speed for Medtronic III. The license rate outcome of the Tax Court’s 75- page analysis of Medtronic III:
“A wholesale royalty rate of 48.8% for both devices significantly bridges the gap between the parties.”[4]
The Tax Court asked about possible alternative methods to apply to the facts, including averaging the CUT and the CPM, which neither Medtronic nor the IRS supported. Pursuant to the Tax Court’s request to the parties during and post-trial, Medtronic proposed an unspecified method that combines elements of the CUT with adjustments and the CPM focused on a return on assets, followed by a residual profit split pursuant to the Siemens Pacesetter agreement. The Tax Court adopted Medtronic’s three-step approach albeit not Medtronic’s proposed result. Most divergent, for Step 3 Medtronic proposed a 7 percent retail royalty rate starting point whereas the Tax Court based on the evidence of the record decided an 80 / 20 split of the residual in favor of Medtronic’s U.S. operations bridged the gap.
The Tax Court relied on a three-step unspecified method to apportion the $3,333,823,544 device and lead system profit between Medtronic’s US tax base and that of Puerto Rico.[5] For Step 1 the Tax Court generated a modified CUT profit result plus trademark license fees for Medtronic US ($674,352,148). For Step 2 the Tax Court generated a modified CPM result, net of component and distribution costs, on behalf of Puerto Rico ($1,344,326,942). Finally, in Step 3 the Court split the residual of the system profits 80 percent in favor of Medtronic US ($1,052,115,563). The sum of the profit apportionment of $2,290,970,127 (representing 68.72 percent of total profits) to the U.S. base and $1,042,853,417 to Puerto Rico’s (representing 31.28 percent) generated a wholesale royalty rate of 48.8 percent.
The Tax Court CUT analysis addressed the Siemens Pacesetter settlement and licensing agreement, providing much more detail.[6] To dispel uncertainty about whether the agreement represented one that would have formed in the ordinary course of business, the Tax Court cited the Medtronic experts’ that patent litigation and settlement licenses are common for the industry. The Tax Court stated that royalty negotiations are often based upon the outcome that the parties would expect in litigation.[7] Yet, the Tax Court determined that the degree of comparability of the Pacesetter licensing agreement’s contractual terms to that of Medtronic Puerto Rico’s licensing agreement was insufficient to establish a CUT but enough to be used as a starting point for determining the proper royalty rate.[8] Note that in Medtronic I, the Tax Court had already adjusted the CUT to account for variations in profit potential.[9]
St. Jude, a competitor of Medtronic, acquired Pacesetter from Siemens, accepting without change the terms of the original license agreement and royalty rates. St Jude further confirmed the agreement in other transactions with third parties. However, the Tax Court also noted that Pacesetter was licensed 342 of Medtronic US’s patents whereas Medtronic Puerto Rico by comparison received licenses for 1,800 of Medtronic US’s patents. The Tax Court compared that as of May 2004 approximately 9 percent of the patents licensed by Medtronic US to Puerto Rico overlapped with its Pacesetter 1992 agreement, reduced to approximately a 6.2 percent overlap by April 2006.
The IRS compared the Medtronic relationship to that of the Coca-Cola decision wherein the Tax Court found that a CPM was appropriate.[10] For Medtronic III, the IRS expert modified the CPM generating a 14 and 12 percent profit margin for Puerto Rico for 2005-06 instead of 8.1 and 5.6 percent.[11] The Medtronic III Tax Court solidly struck the Coca-Cola comparison down because Medtronic Puerto Rico contributed more than a routine quality control function.[12]
“In Coca-Cola Co. the manufacturing process entailed forms of extraction, filtration, mixing, blending, aging, and precision filing. The affiliates performed routine quality control pursuant to detailed specifications from the U.S. parent. With one exception, the affiliates had no employees of their own specifically dedicated to quality assurance. The taxpayer’s experts agreed that the affiliates’ manufacturing activity was a routine activity that could be benchmarked to the activities of contract manufacturers, meriting compensation no greater than cost plus 8.5%.”
The Tax Court also noted the expert testimony that the gross margins of Medtronic US, Boston Scientific, Guidant, and St. Jude did not show any dramatic change from 1992 to 2006 and that Medtronic US’s 2004 gross profit margin was 75.2 percent.
Is Medtronic III a big win for the IRS as is being commonly reported a day after the decision? In the alchemy of transfer pricing intangibles valuation disputes, the quantum needle moved upward in the IRS’ direction. The IRS pulled a larger share of Medtronic’s available tax base into current taxation for the 2005 and 2006 tax years. So very simply put, without the background controversy as context, the IRS appears to have won, albeit the inevitable dollar figure of tax collected will probably be in the range of a hundred million to two hundred million resulting from the Tax Court nearly doubling the leads royalty rate of the MOU at 26 percent to 48.8 percent. The incremental 4.8 percent adjustment from the MOU for device sales will add much less to the additional U.S. tax to be collected. Medtronic III is certainly nowhere near a Coca-Cola quantum.
Yet, the issue that irked the IRS with Medtronic I was the Tax Court’s rejection of the IRS stance to apply a CPM to Medtronic’s Puerto Rico operations. Because the Tax Court’s analysis of the evidence determined the Puerto Rico operation contributed significantly to the production of the medical equipment and leads via the workforce’s quality control, the Tax Court held the IRS’ CPM application inappropriate for allocating transactional value among the parties. The IRS’ stance is that it holds the ultimate discretion to choose and apply a transfer pricing method to a taxpayer’s transactions and operations. The Tax Court’s rejection of the IRS’ attempted application of the CPM is a more significant IRS loss than the quantum gained. The IRS may appeal, and we will be able to read a Medtronic IV and perhaps, even a V.
However, has Medtronic III set the IRS back regarding the CPM? No. The Tax Court rejected the CPM because the evidence clearly established that Medtronic’s Puerto Rico operation contributed more than routine quality control. Note that a “lead” is a highly complex “wiring” system that transmits therapies from Medtronic’s device to the heart via electrical signals and information about the heart’s activity from the heart back to the device. The Tax Court noted that it takes several weeks to manufacture a single lead consisting of over 100 steps in the process. The manufacturing process for even the subassembly of a single portion of a lead, such as the outer assembly of the lead, comprised approximately 20 steps. In addition to interim quality reviews, there were as many as 50 quality tests throughout the lead manufacturing process.[13]
The Tax Court’s three-step other method continues the long line of alchemic cases bridging the gap between the IRS and the taxpayer by splitting the difference. The Chief Judge’s outcome likely accomplishes the Eighth Circuit’s objective unless the Eighth Circuit intends to interject its own weighing of the evidence and command the CPM to be applied. I think that the IRS should take its quantum win and not appeal this decision. It may convince the Eighth Circuit on Appeal to mandate the CPM. But equally, the Eighth Circuit can let the decision stand cementing in its Appellate Circuit that the CPM should not be applied when the related foreign party has significant quality control work force contribution, not a desirable outcome for the IRS.
Below I provide more information about the MOU and Pacesetter agreement. I am submitting a thorough and substantive analysis for the 2023 edition of Practical Guide to U.S. Transfer Pricing (2,800 pages available via Lexis or in print).
Medtronic’s License Rates Before the Current Dispute. Medtronic’s Puerto Rico operation agreed to pay, effective in 2001, based on a CUT analysis by EY, a royalty of 29 percent to Medtronic US on its U.S. net intercompany sales of devices, 15 percent to Medtronic US on its net intercompany sales of leads, and 8 percent for use of the Medtronic trademark.[14] But for an audit of the 2002 tax year, the IRS contended that a CUT was not the best method to value the license arrangement.
Pursuant to an MOU between the IRS and Medtronic to resolve the 2002 audit, Medtronic amended its agreements to include significantly higher royalty rates of 44 percent for the sales of devices and 26 percent for the sales of leads.[15] Moreover, the MOU required that the results of the amended license agreement be tested in future years using a profit split methodology.[16] Medtronic and the IRS agreed that the acceptable Puerto Rico operation system profit target would be 38 percent for devices and 45 percent for leads with an allowance for a 3 percent deviation of this target. i.e. 35 to 41 percent for devices and 42 to 48 percent for leads.
Medtronic and the IRS agreed the MOU would apply to 2002 and all future years if there are no significant changes in any underlying facts.[17] For the tax years of 2005-06, the MOU increased Medtronic US’ total royalty income from the device and leads licenses by an additional $581,747,668 (total royalty payments of $1,773,389,542).[18] However, the IRS challenged the 2005 and 2006 results, determining initially that $539 million of additional royalties should be paid (i.e. a total of $2,312,389,542 for 2005-06), leading to a proposed tax adjustment of $84 million.[19] Medtronic protested based on the MOU. The IRS then set aside the MOU and instead applied a CPM. The IRS CPM resulted in the Puerto Rico operation being limited to 8.1 percent and 5.6 percent of the operating profits for 2005 and 2006, and thus, the remaining 91.9 percent and 94.9 percent respectively accruing to Medtronic US’ operations. The CPM resulted in a two-year additional tax deficiency of $1,358,481,810.[20]
In Medtronic I the Tax Court found that a CUT could be determined partly because Medtronic has settled third-party patent litigation with Siemens Pacesetter, Inc. regarding licensing of comparable intangibles. The Tax Court’s analysis also considered the factors of the Medtronic industry’s background and competitive market[21], Medtronic’s need to self-insure for tort remedies resulting from the quality of its medical devices[22], the input of the Medtronic employees and manufacturing workforce at the stages of the value chain[23], among other factors.
Siemens Pacesetter / Medtronic License. The Siemens Pacesetter and Medtronic patent litigation concerned Medtronic’s patents for many of its cardiac rhythm stimulation devices, including patents underlying its pacemakers’ rate-responsive technology that monitors and adapts to changes in cardiac rhythm.[24] As part of the patent litigation, Medtronic US had won a successful court ruling that its patents were valid and being infringed on by Pacesetter. Medtronic US and Siemens Pacesetter negotiated a patent license and a settlement agreement that resolved patent, antitrust, and unfair competition litigation (“the Pacesetter license”).[25] Pursuant to the Pacesetter license, Pacesetter agreed to pay Medtronic a 1.8 percent “portfolio access fee” with $75 million paid upfront. Thereafter, Pacesetter agreed to pay Medtronic US a 7 percent royalty on CRDM devices and leads sales in the United States and Japan, and a 3.5 percent royalty on all other international sales. The parties agreed a maximum rate clause whereby each party could compel a license to any of the other’s CRDM patents developed during the agreement’s term for an aggregate rate of no more than 15 percent. This meant that Siemens, Pacesetter’s parent company, was entitled to license all of Medtronic US’s CRDM patents for an aggregate rate not higher than 15 percent which included the 7 percent royalty that Pacesetter was already paying for current patents. The outcome of the Pacesetter License was the most lucrative deal Medtronic US had achieved and remains one of the highest royalty rates in the pacemaker and defibrillator industry to date.[26]
[1] Medtronic, Inc. and Consolidated Subsidiaries v. Commissioner of Internal Revenue, T.C. Memo, 2022 Tax Ct. Memo LEXIS 83 (T.C. Aug. 18, 2022).
[2] Medtronic, Inc v. Comm’r., 900 F3d 610, 2018 US App LEXIS 22835 (Aug 16, 2018) (“Medtronic II”), remanding Medtronic, Inc. v. Comm’r of Internal Revenue T.C. Memo 2016-112 (June 9, 2016) (“Medtronic I”).
[3] Medtronic II at 9.
[4] Medtronic III at 72.
[5] Medtronic III at 67.
[6] Medtronic III at 17-19, 34-35.
[7] Medtronic at III at 34, citing to Jonathan S. Masur, The Use and Misuse of Patent Licenses, 110 Nw. U. L. Rev. 115, 120 (2015).
[8] Medtronic III at 35.
[9] Medtronic III at 38, citing Medtronic I at 129.
[10] Coca-Cola Co. v. Comm’r, 155 T.C. 145 at 217–18, 2020 U.S. Tax Ct. LEXIS 27 (Nov. 18, 2020).
[11] Medtronic III at 43-4. The expert reduced the number of comparable companies from 14 to 5 to generate higher profit margins: Bard, Orthofix, Stryker, Wright Medical Group, and Zimmer. But Medtronic Puerto Rico only manufactured class III (highest value) medical devices whereas these five companies made class I and II devices as well.
[12] Medtronic III at 38.
[13] Medtronic I at 43.
[14] Medtronic I at 36, 65-66.
[15] Medtronic I at 66.
[16] Medtronic I at 66-67.
[17] Medtronic I at 67.
[18] Medtronic I at 69.
[19] Medtronic II at 4.
[20] The IRS amended answer sought $548,180,115 for 2005 and $810,301,695 for 2006.
[21] Medtronic I at 9-11.
[22] Medtronic I at 12.
[23] Medtronic I at 27.
[24] Medtronic I at 58.
[25] Medtronic III at 17-18 for details.
[26] Medtronic III at 18.
________________________
To make sure you do not miss out on regular updates from the Kluwer International Tax Blog, please subscribe here.
Responding to Paul Sutton on LinkedIn (too many characters for LinkedIn to allow the post) https://www.linkedin.com/feed/update/urn%3Ali%3Aactivity%3A6967468054402191361/
Howdy Paul, I can elaborate on your point. What I am getting at is that, in the IRS’ application of the best method, at least for the taxpayers that we know about because their TP controversies have reached public court dockets, and via the annual APA Report, CPM is its choice of the best method. For a TP audit to have percolated into a docketed controversy, the Service has probably set aside the taxpayer’s TP method choice (as not the best method) because mere adjustments would be settled out. And if the taxpayer’s method is set aside as not the best method, the IRS is afforded deference by the Tax Court in its selection of the TPM. But as we read in Medtronic I and III, that deference is rebuttable.
It would be an interesting academic study to have the results from blind access to a number (perhaps five) years of LB&I TP audits to determine when taxpayers’ methods are accepted as the best method, and when rejected, the method proposed by the IRS (my guess is the substantial majority will be CPM. Like the annual APA Report, but across all TP audits. What we know is that 85 percent of APAs regarding the sale of tangible property and the use of intangible property use the CPM, according to the 2022 APA Report. There are some big transfer pricing controversies we all know of, but not nearly enough on an annual basis to draw data inferences. Maybe the lack of controversies at the tax court level results in the IRS accepting the taxpayer method in the vast majority of transfer pricing audits? Maybe it’s the lack of IRS resources to litigate each case – perhaps solved with the new funding to hire (my graduating transfer pricing Aggies!).
I am not knocking the CPM/TNMM. I think PLIs are highly useful for transfer pricing and for financial KPIs when examining a business’ performance relative to the market. Practically speaking, I think that undertaking a one-sided CPM is easier than undertaking a two-sided residual profit split that seeks to take into account and benchmark the value, of all the inputs for each party. I understand that tax compliance resources are not infinite for either the IRS or taxpayers. One-sided publicly available financial data-driven methods, without the need for undertaking a FA/GVC are less resource intensive, albeit taxpayers to comply with the regulations need to undertake an FA. I get all this and OK if this is driving the reasoning. But state these reasons if these are them, instead of trying to paint the CPM results of complex intangible relationships between two intragroup or multiple intragroup members at arm’s length.
The CPM is just the method. Which PLI or PLIs will be used? We know from the 2022 APA Report that for the 85 percent of covered transactions involving tangible and intangible property that used the CPM/TNMM, the operating margin (OM) was used 65 percent of those times. The APA Report also notes as it has for many years, the Berry Ratio. Then, we still need to determine the comparability of companies that we can obtain data about to establish the benchmark range. The selection of the pool, as we read in Medtronic III (dropping the large pool to just 5 companies), changes the outcome of the benchmark range. And will comparability adjustments need to be made (the answer was yes in Medtronic)?
If the best method regulations require a validating method for the taxpayer’s method, then a CPM may offer such validation or point out problems with the taxpayer-selected method. But a two-sided method (eg residual profit split), albeit more resource intensive, I think is more likely to approximate an arm’s length result. I’ve thought about whether it could be interesting to undertake a CPM for both intragroup parties and establish a range based on some combination of the results. The Tax Court Judge asked the IRS and taxpayer for alternative methods to apply to the facts, including averaging the CUT and the CPM. What about averaging a two-sided CPM? I doubt the IRS would like such an approach because the Service wants to treat foreign manufacturing operations as OEM/routine (low value) functions (and they might be, but not in the Medtronic PR situation).
I think that most intangibles, even HTVIs, can have CUTs with benchmarkable adjustments that represent a market-based return. As we know from the Treasury White Paper, it is not of this opinion. How substantially similar must a comparability factor be to qualify to potentially establish a CUT for the tested party’s transaction (with or without adjustments)? If comparability is set hyper restrictively, then the result is that the IRS can throw out nearly all CUTs. Back to the CPM as the best method. Perhaps we need (a PhD student) to be generating more academic literature on comparability with data-backed benchmarking adjustments for categories of intangibles so that the IRS, Taxpayers, and Tax Courts have something to refer to?
“I think that undertaking a one-sided CPM is easier than undertaking a two-sided residual profit split that seeks to take into account and benchmark the value, of all the inputs for each party.”
Section 1.482-6 tells us when need to consider the residual profits that should accrue to the licensee if it owns a portion of the valuable intangible assets. I have often noted that even if the licensee did not own any valuable intangible assets, it would still deserve a portion of residual profits from its bearing of additional commercial risk when using a valuable intangible owned by another entity. Medtronic could have been addressed using this well recognized financial economics idea but alas the advocates for both sides were not interested in a reasonable middle ground position, I have another paper on this issue that specifically addresses the convoluted approach in Medtronic III>
Here is the Cliff Notes version of what I sent to Tax Notes for possible publication. The unspecified method was a weird profits based approach in 3 steps. Step 1 suggest the profits from the licensed IP was less than 11% of sales based on flawed CUTs which ignored profit potential differences. Step 2 assigned routine returns including an absurdly high routine return for assembly functions done in Puerto Rico. This left a lot of “remaining profits” which were not captured in step 1. Step 3 gave a lot of these remaining profits to Puerto Rico even though it does not appear to own product or marketing IP. The supposed justification might have come from the profits to processed IP or compensation for commercial risk taking but the taxpayer’s “analysis” measured neither claim. So it was rather illogical and entirely resulted orientated.