On December 10, 2009, the United States Tax Court rejected the position taken in an Internal Revenue Service (IRS) deficiency notice in the transfer pricing area in a decision that has been viewed as favourable to taxpayers. In Veritas Software Corp. et. al. v. Commissioner,1 the Court concluded that the taxpayer had not understated the "buy-in" value of intangible property transferred to its Irish affiliate in conjunction with establishing a cost-sharing arrangement (CSA). While all transfer pricing cases are inherently fact-specific, and Veritas relates to Treasury regulations which have since been superseded,2 this IRS defeat coupled with a similar defeat in Xilinx v. Commissioner3 suggests that IRS theories which place a high value on outbound transfers of intangible property are being met with a skeptical audience in the U.S. Tax Court. Looking ahead, Veritas may inform the treatment of CSAs under the new regulations and influence whether certain pending legislative changes regarding the transfer of intangibles will be prospective or retroactive in nature.
Veritas, a Silicon Valley-based company specializing in electronic data management and storage, entered into a CSA with a newly-formed Irish affiliate pursuant to which the Irish affiliate acquired the right to enhance existing base or "platform" technology and develop European markets for its software products. Generally, a CSA involves the licensing of U.S.-owned intangible property to a foreign cost-sharing participant, often efficiently located in a low-cost jurisdiction, which then finances the cost of enhancing the platform technology in exchange for the right to exploit the enhanced technology in predetermined markets. In order for the CSA to qualify under the existing U.S. Treasury regulations, the foreign participant must make a "buy-in" payment to the U.S. owner commensurate with the value of the platform technology. Thereafter, costs are shared in proportion to the anticipated exploitation benefits from marketing the resulting product.
Veritas entered into such an arrangement and agreed to accept a buy-in payment from its Irish affiliate calculated using a "make-sell" license valuation in 1999, shortly before the significant market decline attributable to the collapse of the hi-tech industry. Generally, a "make-sell" license represents the right to manufacture and sell products utilizing an existing intangible without any rights to further development. Pursuant to the terms of the CSA, the Irish affiliate made a $6.3 million payment in 1999 and a subsequent $166 million lump sum buy-in payment which the parties adjusted downward to $118 million in 2002. Upon audit, the IRS challenged Veritas' valuation methodology and asserted that the base technology was actually worth $2.5 billion. The IRS subsequently adjusted downward its calculation of the buyout payment, based on a revised methodology and an alternate expert report, to $1.675 billion. Based on the position outlined in a published coordinated issue paper.4 (CIP) on cost sharing buy-in adjustments, the IRS argued that the make-sell license methodology was inappropriate for ascertaining a comparable uncontrolled transaction (CUT) and that the assets needed to be valued on a aggregate basis.5 The IRS's economist calculated an aggregate value of Veritas using a discounted cash flow analysis for the enterprise and concluded that such value was attributable to the base technology because, like many software companies, Veritas had very few tangible assets.
The Court rejected the IRS's internal financial analysis methodology as "arbitrary, capricious and unreasonable" and took issue with several of the IRS economist's underlying assumptions. The Judge concluded that the expert's discount rate used in the cash flow analysis was too low, his assumption that the pre-existing intangibles had a perpetual useful life failed to acknowledge the reality that such technology is a rapidly deteriorating asset and finally, the aggregate valuation methodology included items that had insignificant value or that weren't transferred to the Irish affiliate. Specifically, the Court found that, due to the rapidly-changing nature of the industry, Veritas platform technology had a useful life of approximately four years. Further, there was insufficient evidence to show that access to Veritas' research and development and marketing teams was transferred to the Irish affiliate. Instead, the Court favoured the taxpayer's CUT methodology which was based on an analysis of historical transactions in which original equipment manufacturers (OEMs), such as Sun, Hewlett Packard and Dell, licensed the use of existing Veritas software to be bundled with the OEMs' hardware. Such comparables represented the arm's-length licensing of intangibles that were substantially similar to those transferred pursuant to the CSA.
The significance of the Tax Court's decision in Veritas remains to be seen. The decision contains much to suggest that the IRS arguments and expert witness were not well received. The IRS still has the right to appeal the decision to the Ninth Circuit Court of Appeals. However, the Court's holding in Veritas rests primarily on findings of fact which may call into question the likelihood of an appeal. In particular, the availability of historical unrelated third-party comparables, coupled with the substantial growth of the Irish affiliate's real business operations and its successful development of foreign markets, may provide significant factual hurdles to an IRS appeal. In any case, it may be some time before any appeal is heard. Part of the Veritas decision deals with the issue of the inclusion of stock option expenses as a shared expense for purposes of a CSA, and is contingent upon the ultimate outcome of the appeal in Xilinx. As such, the final Tax Court decision cannot be entered until any Xilinx appeals have been resolved.
Many practitioners have regarded the IRS's approach to valuing outbound transfers of intangible property pursuant to a CSA or Section 367(d) transfers as heavy-handed. The Tax Court's decision in Veritas coupled with the Ninth Circuit's affirmation of the Tax Court's decision in Xilinx suggests that IRS approaches to transfer pricing that diverge from arm's-length principles may be vulnerable to judicial review. The Tax Court's preference for valuation methodologies based on third-party transactional comparables, even if adjusted to account for some factual differences, may provide taxpayers with a more practical approach to valuing outbound transfers of intangibles rather than applying the conceptual framework espoused by the IRS. Moreover, statements in the Obama administration's 2011 budget bill which describe aggregate pricing of transferred intangibles as a "clarification" of existing law appear to be inconsistent with Veritas, suggesting that such a change might only be prospectively applied. On the other hand, to the extent that Veritas represents an admonition of the IRS's steadfast adherence to the valuation methodologies outlined in the CIP, its broader precedential value may be limited. It is also worth considering whether the increased international focus on transactional profits methods6 would influence the outcome of a similar set of facts argued on different grounds. Canadian tax advisors long frustrated by the opacity of U.S. transfer pricing rules may find comfort in the common sense underlying these recent decisions.
1. 133 T.C. No. 14 (Dec. 10, 2009) (Veritas).
2. Veritas involves the application of Treas. Reg. § 1.482-7 which has since been supplanted by new Temp. Treas. Reg. §1.482-7T effective from January 5, 2009.
3. 125 T.C. 37 (2005) (Xilinx).
4. "Section 482 CSA Buy-In Adjustments," LMSB-04-0907-62 (Sept. 27, 2007).
5. Aggregate basis valuation is important because that is the methodology used in the current regulations as well as a method proposed for valuing all outbound transfers of intangible property under the 2011 Obama administration budget proposals. Contrast the IRS approach with the Canada Revenue Agency's preference for applying the arm's-length principle on a "transaction-by-transaction" or "unbundled" basis as expressed in Transfer Pricing Memorandum TPM-06 (May 16, 2005) and Information Circular IC 87-2R (Sept. 27, 1999).
6. See e.g. Proposed Revision of Chapters I-III of OECD Transfer Pricing Guidelines, (OECD Committee on Fiscal Affairs, Sept. 9, 2009).
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.