Introduction

In some recent transfer pricing audits the Internal Revenue Service (IRS) has 'flipped the tested party' after examining transactions between related US and foreign companies. Typically, this practice results in the attribution of a larger portion of profit from the relevant business activity to the United States. On its face, the expression 'flipping (or switching) the tested party' suggests that the IRS and the taxpayer have essentially similar views of the transactions (and the entities involved), but have approached the pricing from different angles: the taxpayer priced the transaction by comparing the profits of one party (usually the US entity) to those of comparable unrelated parties, whereas the IRS priced the transaction by comparing the profits of the other party (usually the foreign entity) to those of comparable unrelated parties. In some cases, however, it appears that in purporting to flip the tested party ,the IRS may in fact have misconstrued the substance and/or form of the actual transaction and disregarded the taxpayer's use of an appropriate and more reliable transfer pricing method. As these cases reveal, taxpayers can effectively respond to – and quash – IRS attempts to flip the tested party when those attempts reflect methodologically unsound approaches. This update suggests potential challenges.

Transfer pricing methods and tested parties

The Internal Revenue Code and the Treasury Regulations oblige related parties to price an intercompany transaction such that its results are consistent with those that would have been realised had unrelated parties engaged in the same transaction – that is, the price charged between related parties must be an arm's-length price. To arrive at the arm's-length price, a taxpayer must apply a transfer pricing method; but not just any transfer pricing method. Under the best method rule, the taxpayer must select the transfer pricing method that provides the most reliable measure of an arm's-length result.

The regulations allow a taxpayer to use whichever method is best in light of the taxpayer's facts and circumstances, and specifically define several transfer pricing methods that should be considered. For example, the comparable uncontrolled price (CUP) method for tangible goods, the comparable uncontrolled services price (CUSP) method for services and the comparable uncontrolled transaction (CUT) method for intangibles all determine the arm's-length price by first identifying comparable transactions between unrelated parties and then computing the interquartile range of prices observed in those "uncontrolled" transactions. The taxpayer's intercompany transaction price typically must fall within this arm's-length range.

The CUP, CUSP and CUT methods rely on direct observations of market prices. In contrast, the comparable profits (CP) method arrives at the arm's-length price indirectly. To apply the CP method, a taxpayer must choose one of the two parties to its intercompany transaction to serve as the tested party. The CP method first determines the tested party's arm's-length profitability on the basis of the profitability of comparable uncontrolled taxpayers and then uses that profitability measure to demonstrate that the prices charged are arm's length.

Whether a direct method (eg, the CUT method) or an indirect method (eg, the CP method) is the best method for pricing a given transaction primarily depends on two factors:

  • the quality of the data and assumptions used in the analysis; and
  • the degree of comparability between the controlled transaction (or under the CP method, the tested party) and any uncontrolled comparables.

The regulations detail the factors relevant to evaluating comparability. For the CUP, CUSP and CUT methods, the most important factor is the similarity of the products or services that are the subject of the transactions. For the CP method, it is most important that the tested party and uncontrolled comparables perform similar functions, assume similar risks and invest similar resources. To help to ensure a high degree of similarity on these points, the regulations provide that the tested party should generally be the least complex of the two related parties and the one that does not own unique or valuable intangibles.

Troubling trend

Ongoing and recently concluded transfer pricing disputes illustrate a troubling trend. In these cases the IRS has adopted different views of the transactions' substance from those of the taxpayers and may consequently have applied the CP method to an inappropriate tested party.

Abbott Laboratories

In Abbott Laboratories v Commissioner(1) a Tax Court petition filed by Abbott Laboratories suggests that, in Abbott's view, the IRS mischaracterised the role of Bermudan subsidiary Abbott Ireland in transactions between Abbott Ireland and Abbott Cardiovascular Systems (ACS), a US subsidiary of Abbott. Abbott Ireland licensed intangibles from ACS and used these intangibles in manufacturing drug-eluting stents and other vascular intervention devices, at least some of which it then sold to ACS for distribution. According to the petition, ACS and Abbott Ireland determined the royalty rate for the intangibles licence using the CUT method. They set the transfer price for the finished devices sold to ACS using the CP method and, because in the sale of goods transaction ACS was a mere distributor while Abbott Ireland manufactured highly complex, heavily regulated products, they chose ACS as the tested party.

The petition alleges that after examining the group's consolidated US income tax return, the IRS adjusted the transfer prices for both transactions to attribute more income to ACS. First, the IRS applied the resale price method in lieu of the parties' chosen CP method to the finished devices sales and concluded that ACS should earn a margin 11 times higher than what it reported, thus decreasing the transfer price and Abbott Ireland's income. Second, in contrast to the parties' use of a CUT method for the intangibles licence, the IRS applied the CP method. It chose Abbott Ireland (the licensee) as the tested party and concluded that Abbott Ireland was too profitable and so should be paying a higher royalty rate to ACS.

In the petition, Abbott explains in detail the functions performed and risks borne by Abbott Ireland in manufacturing medical devices. The petition emphasises, for example, Abbott Ireland's assumption of substantial product liability and regulatory risks and its engineers' role in process research and development. Reading between the lines, it appears that the IRS adjustments rest on characterisation of Abbott Ireland as akin to a risk-stripped contract manufacturer rather than an independent, risk-bearing manufacturer. The petition does not reveal what flaws, if any, the IRS identified with the taxpayer's methodology or selected comparables. Thus, the petition illustrates both challenges posed by the IRS practice of flipping the tested party:

  • possible mischaracterisation of the transaction's substance (in particular, the tested party's functions, assets and risks); and
  • apparent disregard of a potentially more reliable transfer pricing method.

Despite its advanced age – the petition was filed December 22 2011 – the case has not yet been set for trial.

Medtronic Inc

Another long-pending Tax Court petition in Medtronic Inc v Commissioner(2) filed by Medtronic, Inc on March 23 2011 –tells a similar story. Medtronic licensed intangibles to non-US subsidiary Medtronic Puerto Rico Operation Co (Med PR). Med PR used the intangibles to manufacture pulse generators and medical therapy delivery devices, which Med PR then sold to a US distribution affiliate. Med PR also used some of the intangibles licensed from Medtronic to manufacture spinal screws and sold the screws to yet another US subsidiary.

According to the petition, Medtronic determined the royalty rate that it received from Med PR using the CUT method, supporting that conclusion with the profit split method, and used the CP method for the other intercompany transactions. The petition does not disclose which party Medtronic selected as the tested party in the transactions for which it used the CP method, but strongly suggests that it did not choose Med PR. In the petition, Medtronic repeatedly asserts that Med PR was "an entrepreneurial, risk-bearing, and functionally autonomous licensed manufacturer" and recites at length the substantial risks that Med PR assumed and the extreme complexity of the manufacturing it performed. Med PR, in other words, would generally not be the least complex party to the intercompany transactions.

The IRS apparently took a different view. According to the petition, it disregarded Medtronic's CUT method and instead applied the CP method to the Medtronic-Med PR licence, with Med PR as the tested party. It did the same with respect to Med PR's sales of spinal screws to a US distribution affiliate. For both transactions, the IRS analysis produced a transfer price that increased the Medtronic group's US income. The IRS position, Medtronic contends, is "based on how [it] believes Medtronic should have, could have, or might have structured its business operations... not... on how Medtronic, in fact, structured its business operations". As in the Abbott petition, Medtronic's petition alleges that the IRS rejected the CUT method and applied the CP method using inappropriate comparables based on a mischaracterisation of the tested party's role. The petition suggests that the IRS viewed Med PR as a contract manufacturer, not as an "entrepreneurial, risk-bearing and functionally autonomous manufacturer" that would, at arm's length, negotiate a lower royalty rate and insist on retaining a larger share of the final price for the goods it manufactured. Thus, at least from the taxpayer's point of view, the case exemplifies both potential methodological errors raised by the practice of flipping the tested party – the IRS:

  • disregarded the substance of the taxpayer's transactions; and
  • inexplicably rejected use of a potentially more reliable transfer pricing method.

Judge Kathleen Kerrigan tried Medtronic's case in February and March 2015 and set a briefing schedule that concludes in October 2015. Depending on the proof presented at the non-public trial, her decision, expected in 2016, may address the potential issues raised by flipping the tested party.

Guidant LLC

On March 11 2011, in Guidant LLC, formerly Guidant Corporation v Commissioner,(3) Guidant LLC filed a Tax Court petition that raises issues similar to those in the cases discussed above. It licensed intangibles to two foreign subsidiaries: Guidant Puerto Rico BV (Guidant PR), which operated a manufacturing branch in Puerto Rico; and Guidant Luxembourg SARL (Guidant Ireland), which operated a manufacturing branch in Ireland. Both subsidiaries used the intangibles in manufacturing complex medical devices which they sold to other related parties.

According to the petition, Guidant established the royalty rate charged to Guidant PR and Guidant Ireland using the CUT method, an unspecified method or a profit-split method, and used the CP method to price the two subsidiaries' sales of finished products. The IRS used the CP method to reprice all of the intercompany transactions and, the petition contends, failed to select the least complex party as the tested party. The petition strongly suggests that the IRS chose Guidant PR or Guidant Ireland as the tested party for each transaction. It characterises the two subsidiaries as "risk-bearing entrepreneurial manufacturer[s] and developers[s]" that bear significant regulatory and product liability risks and that participate with Guidant in product research and development. It can be presumed that in applying the CP method to the finished goods sales, Guidant treated the purchasers, not Guidant PR and Guidant Ireland, as the tested parties. Hence, in addition to the issues of disregard for substance and the best method raised by the petitions described above, Guidant's petition presents a pure example of the IRS flipping the tested party.

No trial date has yet been set.

Tools for taxpayers

The pending Tax Court petitions summarised above suggest that the IRS may, in flipping the tested party:

  • mischaracterise the parties' roles in related party transactions;
  • reflexively apply the CP method even when a reliable CUT method is available; and
  • misapply the CP method by selecting the more complex party as the tested party.

Taxpayers have tools to respond to IRS auditors' attempts to inappropriately flip the tested party.

Rely on the regulations

The best method rule prescribes adoption of the transfer pricing method for which available uncontrolled comparables exhibit the greatest degree of comparability to the controlled transaction (or taxpayer), taking into account the quality of the data and assumptions.

The regulations require evaluation of comparability in light of "all factors that could affect prices or profits in arm's length dealings", but emphasise the importance of certain factors to certain transfer pricing methods. For the CP method, the most important factors are the tested party's functions performed, risks assumed and resources invested. A thorough analysis of these factors with respect to the tested party is thus a necessary predicate to the selection of uncontrolled comparables and the CP method will yield a reliable measure of the arm's length result only if the selected comparables are genuinely similar to the tested party.

The regulations state that "in most cases the tested party will be the least complex of the controlled taxpayers and will not own any valuable intangible property or unique assets that distinguish it from potential uncontrolled comparables". In other words, as the tested party becomes more complex, the pool of potential uncontrolled comparables shrinks and the analysis becomes less reliable.

Prove up your functional analysis

Taxpayers should be prepared to prove the accuracy of their functional analyses. Rather than simply reciting what a party does and what risks it bears, a taxpayer should be able to present witness testimony and documentary evidence consistent with its representations. If the IRS mistakenly assumes that a limited-risk distributor enjoys a high degree of independence, the taxpayer should be prepared to show – not merely state – that the distributor in fact exercises little autonomy, performs few functions and bears few risks. Note the number of pages devoted to descriptions of functions and risks in the Abbott, Guidant and Medtronic petitions.

Defend your chosen transfer pricing method

Taxpayers must always be prepared to support their chosen comparables with factual analysis; but where high-quality, reliable uncontrolled comparables data for the CUP, CUSP or CUT method is available, those methods will necessarily be superior to other methods because they rest on direct observations of market prices. The regulations do not incorporate this self-evident proposition, but perhaps because it seems self-evident, it bears remembering – and reminding.

Consider profitability down the value chain

A taxpayer always makes the initial choice of a transfer pricing method and, if applicable, tested party. However, as the cases described above illustrate, the IRS may challenge those choices and foreign tax administrations may do the same. Taxpayers should thus examine and be prepared to explain how the profitability of each participant in the chain of related-party transactions aligns with its functions, risks and assets.

For further information on this topic please contact Peter A Barnes, J Clark Armitage or H David Rosenbloom at Caplin & Drysdale by telephone (+1 202 862 5000) or email (pbarnes@capdale.com, carmitage@capdale.com or drosenbloom@capdale.com).