Medtronic used the comparable uncontrolled transaction (CUT) method to determine the arm’s length royalty rates received from its manufacturing subsidiary in Puerto Rico for the use of intellectual property (IP) under an inter-group licence agreement.
The tax authorities found that Medtronic had left too much profit in Puerto Rico. Using a ‘modified comparable profit method’, the IRS concluded that, at arm’s length, 90 per cent of Medtronic’s ‘devices and leads’ profit should have been allocated to the US parent company, with only 10 per cent allocated to operations in Puerto Rico.
Medtronic took the case to the Tax Court.
The Tax Court conducted its own analysis and determined that the Pacesetter agreement was suitable for application of the CUT method. This decision from the Tax Court was then appealed by the tax authorities to the Court of Appeals.
In 2018, the Court of Appeal found that the Tax Court’s factual findings were insufficient. The Court of Appeals stated that that the Tax Court determined that the Pacesetter agreement was an appropriate CUT because it involved similar intangible property and licensing circumstances. However, the Court determined that the Tax Court’s factual findings were insufficient to enable it to evaluate that determination and consequently, the case was remanded to the Tax Court for further consideration.
On remand, the Tax Court in 2022 rejected both parties’ primary proposals and instead applied Medtronic’s alternative three-step method with adjustments. This produced a wholesale royalty rate of 48.8 per cent and a profit split of 68.7 per cent to Medtronic US/Med USA and 31.3 per cent to Medtronic Puerto Rico.
The tax authorities appealed this decision, arguing that the Tax Court had incorrectly rejected the comparable profits method and had incorrectly adopted an unspecified method that was not permitted under the regulations. Medtronic cross-appealed, maintaining that the Pacesetter agreement was in fact a valid comparable, or that the unspecified method should be used with modifications.
Judgment:
The Court of Appeal – again – set aside the Tax Court’s order and remanded the case for a fresh determination. On remand, the Tax Court was instructed to reconsider the application of the comparable profits method and to make full factual findings on the disputed comparability issues. It was then to determine which of the specified methods yielded the most reliable arm’s length result.
According to the Court of Appeal, the Tax Court had not clearly erred in finding the Pacesetter agreement non-comparable, as it covered only patents whereas the Medtronic licences included a full array of intangibles, such as know-how, regulatory approvals, secret processes, and technical expertise, which carried greater profit potential. However, it held that the Tax Court was wrong to reject the comparable profits method on the grounds cited.
The regulations permit the use of the comparable profits method even in the presence of significant functional and product differences, provided that profitability measures are reliable and adjustments are considered. However, the Tax Court had not made adequate findings on issues such as functional differences, product liability risks, differences in the asset base, and whether Medtronic could realistically have replaced Puerto Rico’s manufacturing capacity.
As the comparable uncontrolled transaction method was not the most appropriate and the comparable profits method had been incorrectly discarded, the Tax Court had erred in adopting the three-step unspecified method.
Click here for other translation