On December 10, the U.S. Tax Court released an important decision in Veritas Software Corp. v. Commissioner that addresses a number of current “hot button” transfer pricing issues. The decision has broad implications for taxpayers and the IRS alike, and serves as a valuable lesson for taxpayers in terms of preparing and defending transfer pricing documentation.
The Veritas decision, authored by Judge Foley, undermines a number of theories the IRS frequently uses to assert large adjustments to buy-in payments under cost sharing arrangements. Beyond cost sharing structures, the decision also highlights a number of critical diligence issues taxpayers should consider when preparing transfer pricing reports and defending the same during audit, Appeals, or litigation.
The amount at issue in the case—an adjustment of approximately $1.5 billion—makes this a highprofile decision and demonstrates the staggering magnitude of adjustments that can result purely from a buy-in transaction and from transfer pricing cases in general.
The decision has broad importance because the theories employed by the IRS are not unique to the facts of this case. Most of these theories were outlined in the IRS’s Coordinated Issue Paper on Cost Sharing Buy-In Adjustments (Sept. 27, 2007). We have seen the same theories, with minor variations, asserted time and again. In addition, the court goes beyond buy-in theories and addresses the Obama administration’s recent proposal to include workforce in place, goodwill, and going concern value in the definition of “intangible property” for section 4821 purposes as an effort to expand the application of sections 367(d) and 482.
The decision should be of great service to taxpayers. It provides strong support to the arguments that taxpayers have been making in Examination and Appeals in defense of cost sharing buy-in valuations, the valuation of intellectual property (IP) transfers in general, the use of comparables—specifically the Comparable Uncontrolled Transaction (CUT) method—and lastly, the IRS’s use of hypothetical transactions to increase adjustments associated with IP transfers.
Veritas entered into a cost sharing agreement (CSA) and technology license agreement (TLA) with a group of its overseas subsidiaries, collectively referred to as “Veritas Ireland” in the opinion. The buy-in for the transfer of the preexisting intangibles associated with the CSA consisted of royalties and a lumpsum prepayment of royalties. The TLA provided that the parties “shall adjust the royalty rate prospectively or retrospectively as necessary so that the rate will remain an arm’s length rate.” In 1999, Veritas Ireland made an initial buy-in payment of $6.3 million and agreed to prepay the remaining consideration relating to the preexisting intangibles in 2000. In 2000, Veritas Ireland made a $166 million lump-sum buy-in payment to Veritas, which was retroactively adjusted in 2002 to $118 million based on updated projections.2
The IRS originally asserted a $2.5 billion adjustment, but at trial reduced it to $1.675 billion. The IRS made little to no attempt at defending its original notice of deficiency and in fact relied on a new economist at trial.
The Court’s Decision
As a threshold matter, the U.S. Tax Court noted that the IRS seemed to rely on various pricing methodologies found in both the new temporary cost sharing regulations—in some cases the IRS’s economist actually referred to the transferred intangibles as “platform contribution” intangibles—and the Obama administration’s recent proposal to include workforce in place, goodwill, and going concern value in the definition of “intangible property” for purposes of section 482. The court noted that this was inappropriate, stating that taxpayers “are merely required to be compliant, not prescient.” We have often seen the IRS assert theories—such as a market capitalization or acquisition price method—that were not included under the prior regulations.
But the court went even further in addressing the administration’s proposals. In a footnote, the court acknowledges that the administration views these changes to the definition of “intangible property” as “clarifications,” yet notes that the administration expects to raise an additional $3 billion over 10 years from these and other clarifications. This echoes the sentiments of a number of taxpayers and commentators, who view the administration proposals as a wholesale change to a longstanding definition.
To value the buy-in, the IRS used an “akin to a sale” argument, which appears to be a variation of the “foregone profits” model that has been incorporated into the new temporary cost-sharing regulations. The IRS’s economist used a discounted cashflow analysis that appeared to value the transferred intangibles based on similar acquisitions that Veritas had made in previous years. The court does not explain this method in great detail, but it appears to have elements of both an income-based discounted cashflow method and the acquisition price method (for intangibles that do not appear to have actually been acquired by Veritas Ireland). In a footnote, the court notes that the “akin to a sale” theory “may” violate Treas. Reg. section 1.482-1(f)(2)(ii)(A), which requires that the IRS evaluate a transaction as actually structured unless the structure lacks economic substance. The court states that the structure in this case was a license of preexisting intangibles, not a sale of a business. The difference between a license and a sale is an important point that has been a specific area of frustration for taxpayers.
The IRS economist valued the transferred intangibles in the aggregate, in part because the intangibles “collectively possess synergies that imbue the whole with greater value than each asset standing alone.” At trial, when the economist was asked if his valuation captured the value of synergies, he replied, “I really don’t have an opinion. It may have. It may have not.” The court found the IRS’s “akin to a sale” theory “specious” and that an aggregate valuation was not the most reliable means of valuation.
Although not specifically stated by the court, we believe that it is likely that the IRS’s attempt to value the transferred intangibles in the aggregate and to capture synergies was a veiled attempt to exclude any value attributable to goodwill (that was not transferred to Veritas Ireland) in the acquisitions analyzed by the IRS. We have seen the IRS frequently assert that when a technology company is acquired, virtually the entire purchase price is attributable to the intangible property. Goodwill, the IRS asserts, is merely an accounting concept. Taxpayers should be aware that this is frequently a matter of contention under audit. It is helpful that the court found the IRS’s aggregate valuation improper.
The court also found that the certain “nonproduct items” were either not transferred or had insignificant value. These included “distribution channels,” “customer lists and customer base,” “access to research and development team,” and “access to marketing team.” Although the IRS asserted that these items were transferred to Veritas Ireland, the IRS’s economist stated that he did not know whether they had in fact been transferred or had value.
In a particularly interesting footnote, the court noted that even if there were evidence that “access to research and development team” and “access to marketing team” (both of which were referred to as “workforce in place” by the IRS at trial) had been transferred, these items would not be taken into account for purposes of the buy-in payment because they did not have “substantial value independent of the services of any individual” under section 936(h)(3)(B) and Treas. Reg. section 1.482-4(b). This has been a controversial area and directly impacts whether “workforce in place” is treated as an intangible for section 482 purposes. Note that in the Joint Committee on Taxation’s “General Explanations of the [Obama] Administration’s FY 2010 Revenue Proposals,” the Joint Committee asserts that workforce-inplace is an intangible because its value is independent of the services of any particular individual. That is, workforce-in-place derives its value from the collective services of individuals and not from any one individual. The court’s language directly contradicts this reasoning and is extremely useful to taxpayers.
As part of its theory, the IRS asserted that the transferred intangibles had a perpetual life. The court disagreed, finding that the useful life of Veritas’s software products was in general four years. The court specifically noted that Veritas was “in a perpetual mode of innovation,” but nonetheless found that its products still had “finite lifecycles.” Again this represents a very useful rebuttal to a frequent position taken by the IRS—that the preexisting intangibles have long useful lives because they are incorporated into the next generation of the product (i.e., a “platform intangible”). The Tax Court’s conclusion concerning the IRS’s attempt to assign a perpetual value on the transferred intangibles is interesting in the context of the new cost-sharing regulations at Treas. Reg. section 1.482-7T—the facts of Veritas demonstrate that many classes of technology are quickly obsolete.
The court also found that the IRS used the wrong beta (a well-established measure of stock volatility) and the wrong equity risk premium (30-day Treasury yields are appropriate as the “risk free rate,” rather than 20-year Treasury yields). These errors led the IRS to use an incorrect discount rate.
Finally, the court found that the CUT method, which was used by the taxpayer, was in fact the best method. Veritas had licensed the same intangibles to various unrelated parties in a variety of contexts and used these licenses as the basis for a CUT for the buy-in transaction. The court agreed, but applied various adjustments. Again, this demonstrates that a frequent IRS audit position—that the taxpayer’s licenses to unrelated parties are not made under comparable circumstances and cannot be the basis for applying the CUT method—is incorrect. Even if the unrelated party licenses are not exact CUTs, they are similar enough such that adjustments can be made, and the result is the best method. IRS audit teams frequently take the position in practice that absent an exact CUT, the CUT methodology is too unreliable compared to other available methodologies and thus should be rejected under the Best Method Rule. Veritas illustrates that an inexact CUT, subject to appropriate adjustments, can provide the best method.
One issue not explicitly addressed by the decision is the commensurate with income standard. It has been the IRS’s position that taxpayers are not permitted to make commensurate with income standard adjustments; taxpayers are in effect stuck with their buy-in payments if the projections turned out to have been overly optimistic. While the decision does not discuss the taxpayer use of the commensurate with income standard directly, two facts are of interest. First, Veritas filed an amended return in 2002 lowering the amount of the buy-in (which was a “prepayment” of royalties) from $166 million to $118 million. As stated above, the TLA had a clause that enabled the parties to make retroactive adjustments to the buy-in amount. Second, the IRS position itself used actual income data rather than projected income data. This position in effect incorporates the commensurate with income standard.
Although the Tax Court could have provided more discussion in several areas and occasionally conflates separate issues, the decision nonetheless is extremely useful to taxpayers facing similar assertions under audit. We do note, however, that an appeal would fall to the Ninth Circuit, which recently reversed a taxpayer-favorable decision in another important cost-sharing case, Xilinx v. Commissioner.