A multinational enterprise’s internal transfer pricing policies determine how it allocates its worldwide taxable income to its subsidiaries or branches in the multiple taxing jurisdictions in which it operates. These transfer pricing rules – always a significant compliance issue for multinational enterprises – have seen significant developments in the OECD and the United States in the past year. The 2009 developments in transfer pricing coincide with the global recession and the resulting efforts by countries’ treasuries to shore up their shrunken receipts. As a result, multinational enterprises should review their internal transfer pricing policies to ensure compliance with the latest rules and proposed rules.
A multinational enterprise’s allocation of its profits to operations in different jurisdictions is significant because the majority of the world’s trade takes place between related companies, often called “intra-group transactions” and not between unrelated producers and consumers. For example, an automobile manufacturer in the United Kingdom that sells cars to customers in Ireland may first buy parts from its parent company in Japan who in turn purchases the parts from its manufacturing subsidiary in China. Alternatively, a U.S. computer software development company may purchase software engineering services from its Irish subsidiary, incorporate the services into its software, and sell a license to the software to its subsidiaries in Brazil and Mexico that will market the software to customers throughout Latin America and the Caribbean.
Taxing authorities are aware that multinational enterprises will seek to avoid income recognition in one jurisdiction in favor of recognizing it in another jurisdiction that has either lower income tax rates or that provides the enterprise with tax benefits that reduce its tax burden (such deductions from income or credits against income tax). Unlike unrelated parties who are economically inclined to strike a fair bargain when transacting for goods, rights or services, multinational enterprises engaging in intra-group transactions have no such incentive. To shift income to favorable taxing jurisdictions, multinational enterprises may prefer to undercharge for goods, rights or services produced by a subsidiary or branch in a high tax jurisdiction and sold to a subsidiary or branch in a low tax jurisdiction, and vice versa, overpaying for goods, rights or services if the purchaser is in a higher tax jurisdiction relative to the provider.
Recognizing the disinterest of multinational enterprises to strike a fair bargain in intra-group transactions, taxing authorities hold multinational enterprises’ intra-group transactions to an “arm’s length principle.” Applying the arm’s length principle, a taxing authority may tax a local subsidiary or branch of a multinational enterprise according to the profits the subsidiary or branch would have earned had it conducted its intra-group transactions according to the same economic terms as unrelated parties who have an economic incentive to strike a fair bargain. Accordingly, the taxing authority will determine whether goods, rights or services provided in an intra-group transaction were provided for a price that reflects the fair market value of the goods, rights or services.
Taxing authorities and taxpayers employ “transfer pricing methods” to appraise whether subsidiaries or branches have conducted their intra-group transactions in accordance with the arm’s length principle (i.e., have they charged a fair market value price for goods, rights or services). There are “traditional methods” that compare the economic results of an intra-group transaction to the economic results of a comparable transaction between unrelated parties to identify terms in the intra-group transaction that violate the arm’s length principle. There are also “transactional profit methods” that examine the profit from a particular intra-group transaction and determine an arm’s length allocation of the profit between the two related subsidiaries or branches. Under the “best method rule,” taxpayers are to apply the method that best evaluates whether a transaction was conducted according to the arm’s length principle. The OECD guidelines have another caveat: the transactional profit methods are only to be used in lieu of traditional methods in exceptional circumstances.
Taxing authorities are continually updating their transfer pricing rules, and 2009 was no exception. In response, enterprises must continually update their internal transfer pricing policies to conform to the transfer pricing rules in the jurisdictions in which they operate.
In 2009, the OECD approved for publication a new compilation of its Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations. The 2009 compilation of the OECD guidelines incorporates supplements and updates to the 1995 OECD compilation. Although just released, the OECD intends to review and revise the 2009 guidelines on an ongoing basis, and has already done so.
During September 2009, after approving the 2009 compilation of its transfer pricing guidelines, the OECD released for public comment its Proposed Revision of Chapters I-III of the Transfer Pricing Guidelines. The September 2009 proposed revisions note that comparable transactions are not always reliable or available for analytical use by taxpayers and taxing authorities seeking to apply one of the traditional transfer pricing methods. Comparable transactions may be unreliable or unavailable because one of the parties contributes unique intangibles to the transaction.
In circumstance where comparable transactions are unreliable or unavailable, a traditional method may not satisfy the best method rule. Accordingly, the OECD proposes changing the hierarchy among the two categories of transfer pricing methods. Under the proposal, transactional profit methods would lose their status as methods of last resort only to be used in exceptional circumstances where a traditional method cannot be applied. However, the traditional method would still be preferable where a traditional method and a transactional profit method can be applied in an equally reliable manner.
The transactional profit methods are comprised of (i) the “Transactional Profit Split Method” (the “TPS Method”) and (ii) the “Transactional Net Margin Method” (the “TNM Method”). The TPS Method seeks to eliminate the effects on profit caused by special conditions imposed on a transaction between related parties, conditions that would not exist if the parties each had an economic incentive to maximize profits in the transaction. The TPS Method examines the transaction, and it determines the profits that the transaction participants would have expected to realize from engaging in the transaction had they been unrelated parties. To perform the transfer pricing analysis under the TPS Method, first determine the total profit that is to be split between the parties, which may be the sum of the separate profits realized by each participant in the transaction. Second, divide the total profit based upon the participants’ functional contributions to the transaction. A party’s contributions to a transaction include its risks assumed and assets employed.
The TNM Method examines the net profit margin relative to the appropriate base. The OECD guidelines note costs, sales and assets as appropriate bases, but this is not an exclusive list. Drawing a parallel between the traditional methods of cost plus and the resale price, under the TNM Method, the appropriate net margin is ideally established by reference to the same taxpayer’s comparable transaction with an unrelated party. If the taxpayer only engages in the transaction with related parties, then the TNM Method is less likely to be the best method. However, the OECD guidelines do provide for an examination of comparable transactions of other unrelated parties, but this would be a less reliable application of the TNM Method.
The OECD is not the only organization shaking up transfer pricing in 2009. The U.S. Treasury with the U.S. Internal Revenue Services and an appellate court in the United States contributed to transfer pricing developments in 2009. The U.S. Treasury and Internal Revenue Service released Treasury Decision 9456 (4 August 2009). The Treasury Decision releases final regulations on “controlled services transactions.” A controlled service transaction is the provision of services between related parties as opposed to the provision of goods and rights (i.e., automobile parts or a software license).
The 2009 transfer pricing rules provide six transfer pricing method:
- Services Cost Method.
- Comparable Uncontrolled Services Price Method
- Gross Services Margin Method
- Cost of Services Plus Method
- Comparable Profits Method
- Profit Split Method
The second, third and fourth methods are comparable to the traditional methods under the OECD Guidelines. The fifth and sixth methods are transactional profit methods, comparable to the TNM Method and the TPS method, respectively, under the OCED Guidelines.
The most unique of the six new methods is the Services Cost Method; it is not mirrored by traditional methods or transactional profit methods. The Comparable Profits Method evaluates whether the amount charged in a controlled transaction satisfies the arm’s length principle by reference to the “total services cost” with no markup. Total services cost means all costs of rendering the service paid in cash or in kind, including stock based compensation paid to employees rendering the controlled service, and any other resource used, made available, or expended in providing the service.
No markup is required under the services cost method; the arm’s length price is the service’s cost with no markup. The service provider merely needs to recoup its total services costs from the service recipient. The Services Cost Method is most useful for a taxpayer in a high tax jurisdiction that provides controlled services to a branch or related taxpayer in a low tax jurisdiction. Under the Services Cost Method, the service provider would not need to report a profit in its high tax jurisdiction from the services it provides, and the service recipient would retain more profit in its low tax jurisdiction.
The Services Cost Method is subject to limitations that should prevent taxpayers from applying the method to their enterprises’ key entrepreneurial activities. The Services Cost Method only applies if
- The taxpayer concludes that the services do not contribute significantly to the competitive advantages, core capabilities, or fundamental risks of business success or failure.
- The taxpayer must keep adequate books and records related to the service.
- The taxpayer’s books and records must contain a statement stating that the taxpayer intends to apply the service cost method and
- The services must be “covered services.”
Covered services are “specified covered services” and “low margin covered services.” A specified covered service is a service identified by the IRS in an IRS Revenue Procedure. The U.S. Internal Revenue Service promulgated guidance identifying 101 specified covered services. These are mostly general administrative type services, including payroll, invoicing, accounting and auditing, budgeting, treasury activities, information technology services, etc. A “low margin covered service” is a controlled service for which the “median comparable markup” on total services costs is 7 percent or less. The median comparable markup is determined by examining uncontrolled comparable uncontrolled service transactions. The U.S. Treasury Regulations provide that the following transactions are not low margin covered services: (i) manufacturing, (ii) production, (iii) extraction, exploration or processing of natural resources, (iv) construction, (v) reselling, distributing, acting as a sales or purchasing agent, or acting under a commission or other similar arrangement, (vi) research, development or experimentation, (vii) engineering or scientific, (viii) financial transactions including guarantees, and (ix) insurance or reinsurance.
Multinationals often use “Cost Sharing Agreements” (“CSA”) in lieu of licensing intangible property between related companies or branches. The related parties that enter a CSA each contribute to the costs of developing the intangible property and obtain the right to use it. In early 2009, the U.S. Treasury and Internal Revenue Service released temporary regulations in Treasury Decision 9441 (6 January 2009). The temporary regulations note that intangible development costs the parties are to share include stock-based compensation paid to employees participating in the intangible development activity.
There had been debate among taxpayers whether intangible development costs include stock-based compensation. The issue was addressed in a significant 2009 court ruling, Xilinx, Inc. v. Commissioner, U.S. Court of Appeals for the Ninth Circuit. A California company, Xilinx, had allocated some of its research and development costs to its Irish subsidiary, but not stock based compensation of employees participating in the intangible development activity. Xilinx argued that including stock-based compensation as an intangible development cost violated the arm’s length principle because unrelated enterprises negotiating at arm’s length would not agree to share stock-based compensation costs. Xilinx obtained a favorable ruling in the U.S. Tax Court. The Ninth Circuit reversed the U.S. Tax Court and ruled for the U.S. Internal Revenue Service. The Ninth Circuit ruled that the relevant U.S. Treasury Regulations required the parties to share “all costs,” a different standard than the arm’s length principle.
The Ninth Circuit’s ruling focuses only on the Treasury Regulations. It does not focus on the long history of the arm’s length principle in preventing multinationals from engaging in profit shifting to tax favorable jurisdiction. The Ninth Circuit’s ruling is significant because the arm’s length principle has always been the norm for multinational enterprises. Deviating from the arm’s length principle would set the United States apart from the rest of the OECD countries.
Xilinx has petitioned the Ninth Circuit for a rehearing. As of early January 2010, the Court has not ruled on whether it will grant a rehearing. If the Court upholds its Xilinx ruling, it is not certain whether other United States appellate courts would apply the “all costs” standard to cost sharing agreements. However, it is certain that “all costs” and “arm’s length” are irreconcilable. Assuming the worst case scenario, for United States tax purposes, a multinational enterprise would allocate stock-based compensation costs globally to satisfy the “all costs” standard, but other jurisdictions would apply the arm’s length principle to reject the allocation of stock-based compensation costs because independent enterprises would not share those costs.
Fortunately for multinational enterprises operating in the United States, the Ninth Circuit’s ruling in Xilinx is restrained by the Court’s narrow focus on cost sharing agreements. The ruling does not affect transfer pricing of intra-group transactions outside of cost sharing agreements. The Ninth Circuit may reverse itself on rehearing, or other U.S. appellate courts may not follow the ruling. The U.S. Internal Revenue Service has not tried to deviate from the arm’s length principle for intra-group transactions performed outside of cost sharing agreements. The transfer pricing rules in the U.S. Treasury Regulations, including the new Controlled Service Transaction regulations, still hold intra-group transactions to the arm’s length principle. The United States cannot deviate far from the arm’s length principle if it wants to remain an attractive country for multinational enterprises to operate.