In Facebook Case, the Cost-Sharing Regulations Pass Their First Test

Forbes

09-06-2025

Although Tax Court Judge Cary Douglas Pugh rejected the IRS's position on almost every key methodological detail in Facebook Inc. v. Commissioner, 164 T.C. No. 9 (2025), her opinion vindicated the general legitimacy of the method and the regulatory scheme that introduced it.
Like most other major section 482 cases, especially those involving contributions to cost-sharing agreements (CSAs), Facebook was largely a clash between the parties' economists. Both sides leaned heavily on valuation reports by expert witnesses to support their vastly differing valuations of Facebook's platform contributions to a 2010 CSA with its Irish subsidiary, which included core technologies and a preexisting user base. In this battle of economists, the IRS clearly lost.
Pugh's opinion repeatedly expressed frustration with both parties' valuation experts for acting more like advocates than experts. However, she found the IRS's expert witness to be particularly unreliable. The two most critical inputs when applying the income method under the 2009 temporary cost-sharing regulations (T.D. 9441) are financial projections and discount rates, and Pugh rejected the IRS lead witness's testimony on both. Pugh was even more critical of the IRS's reliance on an unusual and aggressive variation of the income method.
The cumulative monetary effect of these setbacks for the IRS will be dramatic. The final platform contribution transaction (PCT) value won't be official until Rule 155 computations are complete, but Pugh estimated that applying the income method with reliable inputs would yield a roughly $7.8 billion PCT value — far closer to Facebook's $6.3 billion valuation than to the $19.9 billion value derived from the IRS's method. Based on this estimate, less than 11 percent of the IRS's total PCT value adjustment would stand. The tax revenue gain will still exceed the IRS's litigation costs, but in purely monetary terms, Facebook hardly looks like a major IRS victory.
However, a win for Facebook on the facts isn't necessarily a loss for the IRS. The $12.2 billion difference between the IRS's PCT valuation and the value tentatively cited by Pugh was entirely attributable to inputs and other methodological details. These were ultimately questions of fact, and their relevance is limited to the facts of this case. On questions of statutory and regulatory interpretation, which have ramifications that extend far beyond this case, Facebook was a resounding IRS win.
Billions of dollars are at stake in Facebook, and the case pairs one of the world's best-known and most polarizing companies with one of U.S. tax law's most widely exploited and frequently criticized profit-shifting tools. But Facebook is, above all, the first judicial test of the cost-sharing regulatory regime created by the 2009 temporary regulations, which was the product of a painstaking effort to fix the loopholes (real or perceived) responsible for the IRS's losses in Veritas Software Corp. v. Commissioner, 133 T.C. 297 (2009), nonacq., AOD 2010-05, and Amazon.com Inc. v. Commissioner, 148 T.C. No. 8 (2017), aff'd, 934 F.3d 976 (9th Cir. 2019).
In Veritas and again in Amazon, which was later affirmed by the Ninth Circuit, the Tax Court held that the buy-in requirement for 'preexisting intangible' contributions in the 1995 regulations (T.D. 8632) excluded residual business assets like goodwill and intangibles developed over the life of the CSA. This was a consequence of restrictions that, according to the Tax Court and Ninth Circuit, followed from reg. section 1.482-4(b)'s definition of intangible and the modifier 'preexisting.' Because the buy-in payments derived from the IRS's discounted cash flow (DCF) valuations included the value of residual business assets and later developed intangibles, courts held that the 1995 regulations prohibited their use.
To stop the uncompensated transfers of valuable assets allowed by this interpretation, the 2009 temporary regulations and the substantially identical 2011 final regulations (T.D. 9568) replaced the term 'preexisting intangible' with 'platform contribution.' This decoupled CSA participants' PCT payment obligations from any definitional limitations inherited from reg. section 1.482-4(b). It also clarified that the PCT value must include the value of intangibles developed under the CSA to the extent that their development benefited from access to the platform contribution.
Arguably more important than these terminology changes was the introduction of specified PCT valuation methods that mechanically prevent artificial exclusions. The 1995 cost-sharing regulations cross-referenced reg. section 1.482-4 for pricing methods, and DCF valuations are permitted by reg. section 1.482-4 only as 'unspecified methods.' Courts thus consistently turned to the comparable uncontrolled transaction method, which allows for the kinds of exclusions that DCF valuations do not. It also indulged the historical judicial preference for transactional methods.
The specified PCT valuation method at issue in Facebook is the income method, and it is a close cousin of the DCF valuations rejected in Veritas and Amazon. Similar to a DCF valuation, it derives the PCT value by discounting projected income to present value, and it provides no plausible basis for carving out excluded items. The IRS's ability to defend its selection of the income method in Facebook was thus the first test of the new regulatory scheme's ability to prevent another Veritas or Amazon.
To some, it seemed hard to conceive of any plausible basis for reading the old exclusions and methodological preferences into the new law. It would be counterintuitive, to say the least, if the interpretations of the 1995 regulations that compelled Treasury and the IRS to draft a more elaborate version of reg. section 1.482-7 from scratch somehow remained viable under the new regulatory scheme. Construing the 2009 cost-sharing regulations to exclude residual business asset value and prioritize transactional methods would be like reading Prohibition into the 21st Amendment. It would turn the regulatory scheme on its head.
But this view was never unanimous. Undaunted tax advisers proposed ways to resurrect the old loopholes, and skeptics criticized the whole effort for legitimizing an irredeemably flawed profit-shifting technique. Until the Tax Court released its Facebook opinion, the only real signal of how courts would interpret the new regulations was in a footnote to the Ninth Circuit's Amazon opinion:
"If this case were governed by the 2009 regulations or by the 2017 statutory amendment, there is no doubt the Commissioner's position would be correct."
What's correct may not have been in doubt. What would actually happen when a taxpayer tested this prediction certainly was. Facebook's briefs attest to what must have been a determined and all-encompassing search for a winning legal theory, and some of Pugh's comments at trial arguably suggested skepticism toward the income method.
The legal questions presented in Facebook won't be definitively resolved until at least one or, more likely, multiple appeals courts confirm the answers. But in Facebook, the income method and the regulatory scheme that introduced it passed their first major test. What was clearly correct to the Ninth Circuit panel that decided Amazon was clear to Pugh as well. Any inferences to the contrary drawn from Pugh's questions at trial and dissent in 3M Co. v. Commissioner, 160 T.C. No. 3 (2023), were apparently misguided.
Pugh found that the IRS derived its PCT value from a severely flawed application of the income method, and these findings had a drastic effect on the PCT value. But she unequivocally affirmed the general validity of the income method, and by extension the regulatory scheme, despite Facebook's determined effort to discredit it. As the opinion concludes:
'Applying the statute and regulations, we conclude that using the income method to determine the requisite PCT Payment value and resulting payments for 2010 produces an arm's-length result if the correct inputs are used. . . . The regulations themselves are not invalid merely because they impose a limit on the expected return on [intangible development costs] at a discount rate reflecting market-correlated risks.'
The basic premise underlying the income method is that, for CSAs in which only one party makes any nonroutine platform contribution, the PCT value should equal the difference between the net present value (NPV) of entering the CSA and the NPV of entering the best realistic alternative transaction. The PCT value is thus the difference between the NPV of the PCT payer's reasonably anticipated operating income under the cost-sharing alternative and the NPV of its operating income under the hypothetical best realistic alternative. In general, although not in Facebook, the PCT payer's best realistic alternative is to license the right to exploit the cost-shared intangibles from an independent developer. The NPV of the best realistic alternative is thus the present value of the PCT payer's expected returns as a hypothetical licensee, as determined using either the comparable profits method or the CUT method.
In effect, the income method forces the PCT payer to hand the expected NPV excess associated with CSA participation back to the participant responsible for the nonroutine platform contribution. This leaves the PCT payer with an expected return on the cash it invests in the CSA equal to the cost-sharing alternative discount rate, which represents the return that a market investor could expect to earn on an investment with the same risk profile as the CSA activity. If the PCT payer makes nonroutine contributions specific to its own territory, the income method requires that the payer's best realistic alternative be adjusted to reflect a return on its contributions.
This approach follows from the general 'investor model,' which provides that all PCT valuation methods should offer CSA participants a return on their aggregate net investment commensurate with the CSA activity's risk profile. A corollary of this principle is that all cash contributions included in 'aggregate net investment' should have a uniform expected rate of return, regardless of whether they take the form of a PCT payment or a cost contribution. As reg. section 1.482-7(g)(2)(ii) explains:
'The relative reliability of an application of a method also depends on the degree of consistency of the analysis with the assumption that, as of the date of the PCT, each controlled participant's aggregate net investment in the CSA Activity (including platform contributions, operating contributions . . . and cost contributions) is reasonably anticipated to earn a rate of return (which might be reflected in a discount rate used in applying a method) appropriate to the riskiness of the controlled participant's CSA Activity over the entire period of such CSA Activity.'
The method's logic is sound. The income method is appropriate only when the PCT payer makes no nonroutine platform contributions, so the payer's principal contribution to the CSA will be the cash it invests through the PCT payment and cost contributions. The arm's-length return for a cash contribution is the expected return available to market investors for bearing the risk associated with the CSA activity, and this is what the CSA discount rate represents. Unless the PCT payer makes nonroutine contributions of its own, any excess in its expected returns over the CSA discount rate must be attributable to the PCT payee's platform contribution.
One could reasonably suggest that the discount rate for developing sophisticated and potentially extraordinarily valuable technologies, which Pugh found to be 17.7 percent in Facebook, overcompensates PCT payers that don't functionally contribute to cost-shared intangible development. But it prevents the far more egregious results made possible under the 1995 regulations by the residual business assets exclusion, the use of decay curves and finite useful lives, and the general judicial aversion to income-based valuation methods. The income method generates an aggregate PCT value that reflects the full NPV difference between alternatives, and it provides no basis for carving out value attributable to excluded assets. If the income method didn't establish a meaningful limit on profit shifting, a taxpayer like Facebook wouldn't make such a determined effort to invalidate it.
Facebook upheld the general validity of a method that aggregates the value of all platform contributions, and in doing so, it implicitly rejected the notion that a residual business asset exclusion survives hidden somewhere in the 2009 and 2011 regulations. Pugh expressly rejected that notion as well, and in short order:
'Petitioner also spends a couple of pages in its opening brief on an argument that Facebook Ireland should not be required to compensate Facebook US for residual business assets. It is true that the definition of intangible property in the second sentence of section 482 in 2010 was limited to the intangible property listed under section 936(h)(3)(B). But the first sentence of section 482 has no such limits; the statute does not constrain the contributions to a CSA that might be compensable through a PCT Payment.'
Facebook's more intricate methodological objections, including its 'zero NPV' critique, fared no better. During briefing, Facebook's zero NPV argument seized on the arithmetic relationship between the PCT value and the NPVs of the two alternatives. Because the PCT value is equal to the NPV of the cost-sharing alternative minus the NPV of the licensing alternative, the post-PCT NPV difference between the two alternatives is, by definition, zero. In other words, the income method requires that the PCT payer transfer all of the incremental value associated with entering the CSA back to the PCT payee.
According to Facebook, the arm's-length standard entitles cost-sharing participants to retain some of the NPV excess associated with CSA participation. However, as Pugh rightly observed in her opinion, claiming that a PCT payer's cost contributions have an expected rate of return in excess of the discount rate would discredit every PCT method based on the investor model. This claim, the opinion explains, implies that PCT payers should receive a preferential rate of return in excess of what a market investor would receive on the same investment:
'Petitioner's objection that a generic investor would seek a return that is greater than its cost of capital proves too much. It necessarily assumes that this investment should be more attractive than another similar investment. The arm's-length standard does not require a preferred return (a positive NPV); it requires a return comparable to returns on other similar investments. Moreover, petitioner does not explain why in a controlled transaction, such as this, a positive NPV for Facebook Ireland would not result in a negative NPV for Facebook US.'
Another way in which the income method allegedly shortchanges PCT payers is by denying them a return for the entrepreneurial risks and functions associated with exploiting the cost-shared intangibles in their territory. Echoing reg. section 1.482-7(g)(4)(vi)(E), Pugh explained that any such contributions can be accounted for by properly valuing the licensing alternative:
'To the extent petitioner's objection is that Facebook Ireland receives no return for its entrepreneurial contributions, that is addressed by proper comparables for the licensing alternative. . . . It is incorrect therefore to conclude that the income method denies an economic profit for any entrepreneurial efforts of the PCT Payor. Petitioner's objections are addressed through selection of the proper inputs into the income method.'
Consistently applying this reasoning also led Pugh to reject the way in which the IRS applied the income method in Facebook. The regulations generally assume that the PCT payer's best realistic alternative transaction will be to license the cost-shared intangibles from the developer. This assumption shifts all development risk to the developer, but it leaves the risks associated with exploiting the cost-shared intangibles with the PCT payer. As the final cost-sharing regulations provide (in reg. section 1.482-7(g)(4)(i)):
'In general, the best realistic alternative of the PCT Payor to entering into the CSA would be to license intangibles to be developed by an uncontrolled licensor that undertakes the commitment to bear the entire risk of intangible development that would otherwise have been shared under the CSA. [Emphasis added.]
But the IRS valuation expert instead used a 'services alternative' as the best realistic alternative, which treated Facebook Ireland as though it were a low-risk marketing services provider. Under the services alternative, Facebook Ireland received a cost-plus markup of 8 percent, which was nominally based on a set of marketing services companies that bore none of the exploitation risk typically associated with the licensing alternative. Although the 8 percent markup was within the interquartile range (6.8 to 14.2 percent) for the comparables set, it was well below the median value (13.9 percent).
Whether it's necessary or appropriate to reward PCT payers with an expected return consistent with the returns of real risk-bearing licensees is open for debate. But the reason that Facebook's theoretical criticism of the income method failed is also the reason that, at least under the regulations, the IRS's services alternative approach was inappropriate. Calculating the PCT value by reference to alternatives with drastically different risk profiles also raises major practical problems, including those associated with a wide discount rate differential that cannot be attributed to a specific risk. Unlike a services alternative, the licensing alternative can differ from the cost-sharing alternative in narrow and predefined ways that relate only to development risk.
It's unclear why the IRS opted to use a novel and more aggressive variation of the income method when the method's overall validity was at stake. But it was logically consistent for Pugh to uphold the income method in general while rejecting the method's application in Facebook, and the trade-off for the IRS was a favorable one.
By confirming the income method's general validity, Facebook tentatively vindicates the foundations of the current cost-sharing regulations. Pugh rejected Facebook's attempts to create a new residual business asset exclusion and invalidate the investor model, both of which were critical for the regulatory scheme to function. But Pugh's endorsement of the income method's arm's-length bona fides in Facebook followed from her interpretation of the arm's-length standard in general, which could have implications that extend far beyond cost sharing.
For Facebook, the income method's zero-NPV effect is invalidating because it creates a conflict between reg. section 1.482-7(g)(4) and the arm's-length standard. This assumes that Treasury and the IRS had an obligation to conform reg. section 1.482-7(g)(4) to the arm's-length standard. It also assumes that the arm's-length standard is a transactional and comparables-based concept, regardless of what the regulations say on the matter. As noted in Facebook, this interpretation implies that any transactional evidence at all takes priority over the methodological reliability standards specified by regulation:
'Where there are no uncontrolled comparables, petitioner maintains, the arm's-length standard requires a 'method that is expected to most closely approximate the way in which unrelated parties price transactions.' Petitioner submits that this approximation can be accomplished through sources such as peer-reviewed academic literature and broad industry standards.'
The two assumptions underlying Facebook's argument are related, and the distinction between the two is often blurred. But they are distinct. Whether Treasury and the IRS have a statutory obligation to adhere to something that falls within the ambit of the arm's-length standard is one question, and whether they have to interpret the arm's-length standard in a narrow and archaic way is another.
On the first question, Pugh emphasized that the applicable statutory standard established by the first sentence of section 482 is a clear reflection of income. Her opinion observes that 'neither sentence of section 482 expressly adopts the arm's-length standard,' which 'originated in the regulations promulgated under the Revenue Act of 1934.' Only the sentence added by the Tax Reform Act of 1986 directly addresses controlled intangible transfers, Pugh said, and it does not support Facebook's contention:
'The only statutory touchstone relating to intangibles in section 482 is the 'commensurate with the income' requirement. That addition seems to move the statute away from, not toward, an 'arm's length' standard, at least as petitioner defines it; it requires compensation commensurate with the income earned in the transaction. [Emphasis added.]
The Facebook opinion doesn't directly say whether Treasury and the IRS could issue regulations that openly repudiate the arm's-length standard. But if the statute doesn't bind Treasury and the IRS to the arm's-length standard, then any obligation to apply it would be a self-imposed regulatory restraint on their broader statutory authority. It would follow that Treasury and the IRS have the right to specify the terms of that self-imposed restraint.
However, in Facebook and other best method cases, the authority to openly abandon the arm's-length standard is less important than the discretion to interpret it. On the second question, Pugh was unequivocal. Drawing heavily on the Ninth Circuit majority's reasoning in Altera Corp. v. Commissioner, 926 F.3d 1061 (9th Cir. 2019), rev'g 145 T.C. 91 (2015), Pugh rejected the antiquated interpretation of the arm's-length standard favored by Facebook and other taxpayers:
'In Altera, the Ninth Circuit expressly held that in the light of concerns over third-party comparables, a focus on internal allocations that follow economic activity is an appropriate method to reach an arm's-length result.'
Pugh's unqualified reliance on Altera in Facebook is significant in multiple respects. Although Altera is binding circuit precedent in Facebook, Pugh's opinion reflects a broader acceptance of the Ninth Circuit's reasoning. It also confirms that, at least in the Tax Court's view, the Ninth Circuit's holding was unaffected by Loper Bright Enterprises Inc. v. Raimondo, 603 U.S. 369 (2024). Therefore, all taxpayer validity challenges targeting the cost-sharing regulations' treatment of stock-based compensation, including in Abbott Laboratories v. Commissioner, No. 20227-23, and McKesson Corp. v. United States, No. 3:25-cv-01102, should fail.
Perhaps even more significant, Pugh's reliance on Altera in a best method case thwarts a ubiquitous and foundational element of taxpayers' arguments in methodological disputes. As the Facebook opinion explains:
'Petitioner attempts to convert the arm's-length standard, as defined in Treas. Reg. section 1.482-1, into an independent rule. But nothing in the text of section 482 bars Treasury from prescribing what arm's length means when no comparable transactions can be identified. Section 482 does not contain the words 'arm's length'; rather, its focus is on clear reflection of income and preventing tax evasion in controlled transactions.'
In other words, neither section 482 nor reg. section 1.482-1's general articulation of the arm's-length standard provides any basis for invalidating the method-specific provisions that apply them. This is critical because manufacturing such conflicts has become the basis for taxpayer attacks on all income-based methods, including the CPM in Medtronic Inc. v. Commissioner, T.C. Memo. 2022-84. If legitimized by courts, those conflicts would twist the section 482 regulations into an ineffectual knot.
The significance of upholding one of the centerpieces of the 2009 cost-sharing regulations, and by extension the regulatory scheme itself, in Facebook can't be understated. But the Tax Court's broader acceptance of Altera, and the corresponding rejection of an inappropriately narrow interpretation of the arm's-length standard, is arguably even more important. For the IRS, these victories on the law far outweigh its loss on the facts in Facebook.