New Developments in Transfer Pricing Rules for Intangibles

A 2016 survey conducted by EY revealed that 72 percent of corporate respondents identified transfer pricing as the most important international tax issue they face.¹ The survey also projected that transfer pricing would be the most controversial tax policy topic in the next two years. The transfer price charged to a related party is generally either based on the price that would be charged to an unrelated party or determined by reference to sales of similar products or services by other entities to unrelated parties.

As the arm’s-length standard implies, the efficient setting of appropriate transfer prices depends on the existence of competitive markets. However, it is not always possible to find comparable market transactions to set an acceptable transfer price, especially for intangible properties for which there is no established market. Therefore, it can be difficult to determine which transfer pricing method provides the most reliable measure of an arm’s-length price for intangible properties. Also, it is not always easy to determine which cost should be included in determining cost-sharing arrangements.

Two recent court cases, Medtronic and Altera, focused on these issues and the opinions in the cases will cause substantial changes in global transfer pricing and are likely to affect the commercial decisions made by multinational companies.² This article informs CPAs about which transfer pricing method should apply to intercompany transactions of intangibles and which costs should be included in determining cost-sharing arrangements. The remainder of the article discusses (1) court decisions affecting transfer pricing, (2) cost-sharing arrangements, and (3) the OECD action plan affecting base erosion and profit shifting.

Court Decisions Affecting Transfer Pricing

In allocating appropriate transfer pricing costs, courts usually rely on the most basic principle underlying any transfer pricing analysis: the arm’s-length standard. The arm’s-length standard, by definition, requires an analysis of what unrelated parties would do in similar circumstances under the section 482 regulations.³ However, because similar transactions with unrelated parties under similar circumstances often do not exist, the regulation allows the use of uncontrolled transactions that are comparable with the controlled transactions rather than identical to the controlled transactions. That is, a determination of an arm’s-length price frequently involves reference to similar transactions in comparable markets. The efficient setting of appropriate transfer prices thus is subject to the existence of comparable markets. The following scenario illustrates the application of the arm’s-length standard under section 482.

Example 1: Company A, a U.S. corporation, sold property to one of its shareholders for $500,000 in 2016. In an audit of the entity’s 2016 tax return, the auditor would try to ascertain what price would have been applied between the company and an unrelated party in determining whether the $500,000 is an arm’s-length price. If the property had been sold to an unrelated party, the price would have been $1 million. In this situation, the taxing authority would require Company A to adjust the price of the property to $1 million under the arm’s-length standard of section 482.

However, it is not always possible to find similar market transactions to determine an acceptable transfer price, especially for intangible properties for which there are no established markets.⁴ Thus, it can be difficult to determine which transfer pricing method provides the most reliable measure of an arm’s-length price. This problem can be clearly seen in Medtronic, in which the Tax Court rejected the IRS’s use of the comparable profits method to determine the appropriate royalty rate between Medtronic US and its Puerto Rican subsidiary, Medtronic Puerto Rico Operations Co. (MPROC).

Medtronic US was the parent company of a consolidated medical technology firm, with operations and sales worldwide. As a parent company, Medtronic US was responsible for routine functions, including accounting, tax, finance, and legal, and was greatly involved in ensuring product quality. MPROC, an affiliate of Medtronic US, was responsible for the quality of the medical device pulse generators (devices) and physical therapy delivery devices (leads) it manufactured. To further its business, Medtronic US and MPROC entered into four separate intercompany agreements, with two key agreements at issue during trial: (1) Medtronic US’s grant of a trademark license to MPROC and (2) Medtronic US’s grant of technology and know-how licenses for devices and leads to MPROC. In general, if a U.S. entity licenses intangible property to a related foreign subsidiary, it is subject to transfer pricing rules that require the transaction to be arm’s length.

The best method rule described in reg. section 1.482-1(c) states that taxpayers should choose the method that provides the most reliable measure of an arm’s-length consequence, taking into account all the facts and circumstances available. For these two agreements Medtronic US selected the comparable uncontrolled transaction method as the best method to benchmark each transaction.⁵ The CUT method determines an arm’s-length royalty for a controlled transfer of an intangible by referring to uncontrolled transfers of “comparable intangible property” under “comparable circumstances.”⁶ Therefore, the CUT method requires both comparability of the transactions and an arm’s-length range for the results. Example 2 illustrates the application of the CUT method.

Example 2: Company A, a U.S. corporation, licenses its know-how to its Irish subsidiary Company B as well as to Company C, an unrelated corporation. The circumstances and license agreements affecting companies A and B are nearly identical. The license to Company C offers a royalty of 5 percent of sales. If Company B pays a 3 percent royalty to Company A, the license rate to Company B would not be considered within an arm’s-length measurement. Thus, the royalty rate applied to Company C determines whether the license to Company B is within an arm’s-length measurement.

On the other hand, if Company C incurs significantly more risks than Company B while performing significantly different functions, with different license terms, the CUT method does not apply. This means that a similar level of risk for developing intangibles should be borne by Company B and Company C to apply the CUT method. Thus, when significant differences in intangibles arise, the CUT method will not apply.

The IRS examined Medtronic US’s 2002 tax return and reviewed several intercompany manufacturing and sales transactions related to the devices and leads. The IRS’s examining agent accepted Medtronic US’s use of the CUT method, which established arm’s-length royalty rates of 8 percent for trademarks, 29 percent for intercompany devices transactions, and 15 percent for intercompany leads transactions to be paid by MPROC on its sales. However, the IRS’s examining agent also proposed some adjustments⁷ on the transactions to increase their “profit potential.” A profit potential of intangibles is an essential factor under the CUT method, and the profit potential of intangibles for controlled transactions is compared with the profit potential for uncontrolled transactions. The examiner noted that MPROC’s profits were extremely overstated. Finally, the IRS and Medtronic US entered into a memorandum of understanding, adjusting royalty rates to 44 percent for intercompany devices transactions and 26 percent for intercompany leads transactions. The MOU also included an agreement that system profit attributable to MPROC should be between 35 percent and 41 percent for devices and between 42 percent and 48 percent for leads. These agreements in the MOU were applicable to 2002 and subsequent tax years.

Using the CUT method in accordance with the MOU, Medtronic’s 2005 and 2006 tax returns first calculated arm’s-length royalty rates of 29 percent for devices and 15 percent for leads. Medtronic US then reported additional royalty income totaling more than $0.58 billion on its Schedule M-3 for the 2005 and 2006 tax returns based on the increased royalty rates of 44 percent for devices and 26 percent for leads on intercompany sales under the profit-split method. However, the IRS issued Medtronic US notices of deficiency for both 2005 and 2006. Using the CPM, the IRS made adjustments to Medtronic US’s royalty income totaling approximately $1.3 billion. While the CUT method applies to a transfer of intangibles, the CPM can be used for a transfer of both intangible and tangible property. The CPM applies “profit-level indicators,” objective measures of profitability, in determining the arm’s-length amount. Example 3, abstracted from Transfer Pricing Methods: An Applications Guide by Robert Feinschreiber,⁸ illustrates the application of the CPM.

Example 3: Company A, a U.S. corporation, licenses its patented technologies to its Irish subsidiary, Company B. Using these technologies, Company B manufactures a high-tech device and sells it at an arm’s-length price to Company A. Company B pays a 7 percent royalty to Company A.

In an audit of the entity’s 2016 tax return, Company B is selected as a tested party because it engages in relatively routine manufacturing activities, whereas Company A has unique and valuable activities and engages in relatively complex activities. Because Company B is a manufacturer, the ratio of operating profits-to-operating assets is selected as the most appropriate profit-level indicator. The interquartile range of profits derived from uncontrolled comparables (foreign manufacturers) for 2016 is $1 million to $2.5 million (the median is $1.75 million), while Company B’s operating profit for 2016 is $12.75 million (which lies outside the arm’s-length range). Thus, an adjustment is required.

The median of the interquartile range of uncontrolled comparable operating profits of $1.75 million is determined to be the arm’s-length profit for Company B. Thus, the royalties Company B paid Company A in 2016 are increased by $11 million (the difference between Company B’s reported operating profit of $12.75 million and the arm’s-length profit of $1.75 million). This adjustment increases Company A’s U.S. taxable income by $11 million, and decreases Company B’s taxable income by the corresponding amount.

At trial, Medtronic US argued that the MOU royalty rates on the intercompany sales of devices and leads from MPROC exceeded what arm’s-length royalty payments would have been based on its own CUT analysis. The IRS, in accordance with the MOU, calculated that in 2005, the proper royalty rates were 37.6 percent for devices and 25.5 percent for leads after the profit-split calculation; in 2006, the proper royalty rates were 44 percent for devices and 26 percent for leads after the profit-split calculation. However, Medtronic US asserted that the appropriate arm’s-length royalty rates were 29 percent for devices and 15 percent for leads. Also, Medtronic US alleged that the IRS’s allocations using the CPM were “arbitrary, capricious, or unreasonable” because the IRS’s aggregate CPM value chain analysis failed to treat separately the intercompany transactions between Medtronic US, MPROC, and Medtronic USA Inc. (a U.S. affiliate of Medtronic US). Generally, a taxpayer has the burden of proving that the IRS’s adjustment is “arbitrary, capricious, or unreasonable” to challenge its pricing adjustment. Medtronic US also contended that the IRS did not give appropriate weight to the importance of quality control in MPROC manufacturing, meaning that the IRS’s analysis did not assign MPROC an appropriate return equal to the risk it bore as the manufacturer of the medical devices, which resulted in unreasonable adjustments.

In response to Medtronic US’s claims, the IRS contended that the CPM, rather than CUT, was the best method to determine the arm’s-length royalty rates on the intercompany transactions of devices and leads and that its section 482 adjustments did not result in an abuse of discretion. The IRS argued that quality control was not the primary factor in considering the success of medical devices, indicating that MPROC did not perform economically significant functions for leads and devices. Therefore, the IRS asserted that it used an analysis that assigned MPROC an appropriate return.

To challenge the IRS’s imposition of adjusted royalty income based on the CPM method, Medtronic US needed to meet the burden of proving that the IRS abused its discretion by making “arbitrary, capricious, or unreasonable” allocations. First, in response to Medtronic US’s claim that the IRS abused its discretion by using the CPM when allocating the appropriate royalty amount of system operating profit in the several intercompany transactions, the Tax Court noted that the IRS’s economic analysis that served as the basis for its position used a value chain analysis. Under the IRS’s value chain approach, Medtronic US’s operations were segmented into functional activities. However, because this analysis is not prescribed under the section 482 regulations, the Tax Court was tasked with conducting a facts and circumstances analysis to determine whether the use of the CPM was “arbitrary, capricious, or unreasonable.”

Also, in response to the IRS’s claim that quality was not the primary factor when considering the success of medical devices, the Tax Court concluded that the IRS failed to give enough weight to MPROC’s quality control function. In its report, the IRS contended that MPROC was merely responsible for assembling finished medical devices, but the Tax Court disagreed and found factors demonstrating that MPROC was “an integral part of [Medtronic US that] not only made the finished product; it made sure that the finished product was safe and could be implanted in the human body.”⁹ Accordingly, the Tax Court concluded that the IRS’s report failed to give enough weight to MPROC’s quality control functions.

Further, the Tax Court noted that many comparable companies that the IRS’s report used in its CPM analysis were not indeed comparable under section 482 regulations. Some of the 14 companies that the IRS’s report used as comparable entities in its CPM analysis were not manufacturers of similar devices or were manufacturers on a much smaller scale. Many of the comparable companies were also engaged in sales and research and development functions — functions in which MPROC was not engaged. The Tax Court further noted that regardless of an insufficient degree of comparability, the IRS’s report used these comparable companies to benchmark Medtronic US’s return for the sale of component parts to MPROC and MPROC’s arm’s-length manufacturing return, calling into question the reliability of the IRS’s analysis.

Finally, the Tax Court noted that the IRS’s use of the return on assets (ROA) as the profit-level indicator was incorrect because the IRS’s ROA estimates did not take into account the value of MPROC as an operation; the IRS’s ROA estimates included only MPROC’s buildings, equipment, and inventory. Since MPROC had “valuable intangible assets that were obtained through the devices and leads licenses [that] are not recorded on [the company’s] balance sheet,” the Tax Court concluded that the IRS’s use of the ROA as the profit-level indicator was “misleading.”¹⁰ The Tax Court also noted that section 482 regulations do not mandate aggregation for transactions, despite perhaps being appropriate in some cases. The Tax Court found that aggregation was unnecessary in this case because it did not result in an appropriate arm’s-length allocation of income to MPROC. Thus, the Tax Court concluded that Medtronic US met its burden of proving that the IRS’s allocations were “arbitrary, capricious or unreasonable.”

To challenge the IRS’s royalty income adjustments, Medtronic US also needed to meet the burden of proving that its application of the CUT method results in allocations aligned with the arm’s-length standard. The Tax Court noted that Medtronic US’s royalty rates of 29 percent of intercompany sales for devices and 15 percent for leads using the CUT method did not produce an arm’s-length result for several reasons. One reason the Tax Court pointed out was that Medtronic US’s expert failed to include a profit potential analysis, a component that is required under the section 482 regulations in calculating the arm’s-length royalty rates. Thus, the Tax Court found that Medtronic US did not meet its burden of proving that its CUT method produced an arm’s-length result.

Although the IRS’s allocations using the CPM represented an abuse of discretion, Medtronic US’s CUT method also failed to produce an arm’s-length result. Thus, the Tax Court constructed an arm’s-length analysis itself, determining what an allocation of profit between Medtronic US and MPROC on the devices and leads licensees should have been under the arm’s-length standard. The Tax Court adopted Medtronic US’s CUT method as a starting point and made several adjustments for know-how and profit potential factors that increased a royalty rate for device licenses from 29 percent to 44 percent. Also, the Tax Court found that a reasonable royalty rate for the leads business should be 22 percent because the leads business was not very profitable when compared with the devices business. The Tax Court, using its own analysis, thus found appropriate royalty rates for the licenses. Example 4 highlights the issues litigated in Medtronic Inc.

Example 4: Company A, a U.S. corporation, granted technology and know-how licenses for its products to Irish subsidiary Company B. Using the CUT method, Company A calculated an arm’s-length royalty rate of 15 percent for one of its products in consideration of Company B’s risk. However, during audit, the taxing authority calculated, by using the CPM, that the proper royalty rate was 26 percent.

If this issue were disputed, based on the Medtronic decision, a court would likely determine that Company A’s use of the CUT method is considered the most appropriate method. The court would likely reject the taxing authority’s use of CPM and show a preference for the CUT method because this method would be considered more carefully incorporating each individual transaction.

Similarly, the Tax Court in Bausch & Lomb¹¹ rejected the IRS’s aggregate value chain analysis and showed a preference for CUT. In this case, the Tax Court was faced with the challenge of determining the arm’s-length price for soft contact lenses purchased by B&L US, a U.S. parent company, from its Irish subsidiary, B&L Ireland. B&L US and B&L Ireland entered into a nonexclusive licensing agreement, including B&L US’s grant of both a trademark license and a technology license for soft contact lenses to B&L Ireland, in return for a 5 percent royalty on sales. The Tax Court found that the IRS expert’s suggested royalty rate between 27 and 33 percent for the use of B&L US’s intangibles by B&L Ireland was “unreasonable,”¹² but it also found that the royalty of 5 percent was too low. The Tax Court was unable to identify a “sufficiently similar transaction involving an unrelated party” and thus decided to proceed with its own arm’s-length analysis.¹³ Finally, the Tax Court held that a royalty of 20 percent of B&L Ireland’s sales price for soft contact lenses established arm’s-length consideration for the use of B&L US’s intangibles. In both Medtronic and Bausch & Lomb, the judges rejected the IRS’s aggregate value chain analysis and instead applied the taxpayer’s CUT that more carefully incorporated each individual transaction.

Cost-Sharing Arrangements

One method used by controlled parties to avoid both uncertainties and the administrative burdens related to transfer pricing for intangibles is to enter into a cost-sharing arrangement. For example, a U.S. parent company and its foreign subsidiary may agree to equally share the costs of development of intangibles. Under this agreement, the parent company might obtain the rights to manufacture and market the new product in the United States, while the foreign subsidiary company may obtain the rights to manufacture and market the new product abroad. A bona fide cost-sharing arrangement must allocate research and development costs in proportion to the profits earned by each controlled party from the intangibles. However, it is not always easy to determine which costs should be included in calculating costs for a cost-sharing arrangement. This can be clearly seen in Altera,¹⁴ in which the Tax Court rejected the IRS’s inclusion of stock-based compensation (SBC) costs in the cost pool for a qualified cost-sharing arrangement (QCSA).

The Tax Court in Altera¹⁵ unanimously decided that the IRS erred in allocating Altera US’s SBC costs to its QCSA cost pool. Altera US and its Cayman Islands-based subsidiary, Altera International, entered into QCSAs that mandated sharing the risks and costs of R&D activities. In the tax years from 2004 through 2007, Altera US granted stock options and other SBC to employees who were responsible for performing the R&D activities covered under its QCSA. In allocating the costs to be shared according to the QCSA, Altera US did not include its SBC costs in the shared cost pool under the QCSA, but included the cash compensation of its employees involved in R&D. Under this agreement, Altera International made substantial cost-sharing payments between $129 million and $192 million from 2004 to 2007, but the IRS issued notices of deficiency that allocated additional income from Altera International to Altera US by increasing Altera International’s cost-sharing payments to include the stock-based compensation under reg. section 1.482-7(d)(2). Under reg. section 1.482-7(d)(2), parties in QCSAs must share SBC costs to achieve an arm’s-length result.

Altera US petitioned the Tax Court, asserting that the requirement in reg. section 1.482-7(d)(2), which requires parties to share SBC costs as a basis for achieving arm’s-length results, was “arbitrary and capricious”; therefore, the regulation was argued to be “invalid.”¹⁶ The Tax Court noted that the definition of the arm’s-length standard requires an analysis of what unrelated parties do under similar circumstances. The Tax Court found that unrelated parties would not share the SBC costs because the value of SBC is too speculative and completely outside the parties’ control. The Tax Court also found that reg. section 1.482-7(d)(2) was finalized without adequate responses to significant public comments from interested parties.¹⁷ Therefore, the Tax Court held that reg. section 1.482-7(d)(2) mandating controlled parties that enter QCSAs to include SBC costs in the shared costs is “arbitrary and capricious and therefore invalid.” The Altera opinion indicates that taxpayers may not need to include SBC costs in the cost pool for QCSAs. This case also suggests that taxpayers may be able to challenge regulations if they believe the regulations do not reflect reasoned decision-making supported by empirical evidence.

OECD Action Plan Affecting BEPS

Another matter for taxpayers to watch is the effect of the OECD’s action plan within the base erosion and profit-shifting initiative. BEPS refers to tax planning strategies that use gaps and mismatches in tax rules to shift profits to low-tax jurisdictions where there is little economic activity. To reduce instances of abuse arising from the inappropriate use of transfer pricing rules and methods, on July 4, 2016, the OECD published a discussion draft as part of its BEPS project.¹⁸ The OECD’s BEPS project has resulted in the worldwide implementation of more stringent requirements for justifying transfer pricing methods.

The new discussion draft points out, among other items, the important role of “a value chain analysis” in determining the most appropriate transfer pricing method, which is not the approach taken by the courts in Medtronic or Bausch & Lomb. Specifically, the OECD’s action item 8 attempts to align transfer pricing outcomes of intangibles with value creation. Even though in Medtronic the IRS argued that value chain should have been considered in pricing arrangements between the related parties,¹⁹ the United States has not implemented action item 8 on the basis that substantial changes to the U.S. transfer pricing regulations need not be made to conform to the BEPS standards, and that the current regulations include a description of a value chain approach.²⁰ It will be interesting to see if the United States makes a move toward adopting the action item and how courts will react. If the United States adopts this action item, companies’ ability to develop and document a value chain analysis will be important when they plan and defend their transfer pricing methods in the future. Until then, taxpayers may rely on the court decisions reached in Medtronic and Bausch & Lomb for changes that could be caused by the OECD’s BEPS project.

Conclusion

The Medtronic and Altera opinions are likely to significantly affect multinational companies’ transfer pricing policies. Because they may be called on to help their clients revise their transfer pricing and cost-sharing strategies for intangibles, CPAs should understand the implications of these opinions so they can offer clients ideas for revising strategies.