It’s an old Tax Court maxim that Constitutional arguments don’t cut any Glasshouse ice, but Amgen Inc. & Subsidiaries, 16017-21, filed 7/3/24, claim IRS violated Due Process when they disavowed seven (count ’em, seven) closing agreements covering nine (count ’em, nine) tax years.

IRS gave Amgen annual audits for each of those years, with the adjustments culminating in the abovementioned closing agreements. But shortly thereafter, IRS audited another six (count ’em, six) subsequent years, adjusting all and adding chops to the last three, using a transfer pricing method deviating from that in the settled years’ agreements. Amgen petitioned all, but seeks summary J only as to chops in the last three. IRS cross moves as to all six, but Judge Travis A. (“Tag”) Greaves sorts it out.

There’s caselaw that government cannot change the rules justifiably relied upon without giving notice, but that applies to guidance. Here, there’s a specific agreement, covering only the years then at issue. Moreover, the agreements do not discuss the Section 482 transfer pricing methodology.

And Supreme Court learning says IRS isn’t bound by prior years. Another old Tax Court maxim is each year stands on its own.

“Petitioner had no legitimate reliance interest for future years derived from the closing agreements. The closing agreements unambiguously do not cover future tax years. The agreements are silent as to what transfer pricing methodology was to apply for years after [last year]. In fact, the closing agreements related to [last three] tax years made it clear that the IRS could make future transfer pricing adjustments regardless of any alleged prior approval. These closing agreements specifically stated ‘This agreement does not prevent further allocations under section 482 with respect to taxable events involving Amgen and [sub] that are attributable to taxable periods of Amgen for which allocations are not determined by this agreement.’ This clause put petitioner on notice that the IRS might make transfer pricing adjustments in future tax years. Additionally, none of the closing agreements used the phrases ‘best method’ or ‘arm’s length’ to describe the reallocation. Instead, the adjustments are simply those to which the parties agreed in settling the disputes before them at that moment. The closing agreements unambiguously do not cover tax years past [last year], and therefore, petitioner does not have a legitimate reliance interest created by the closing agreements.” Order, at pp. 6-7.

And Amgen could always have entered into advance pricing agreements.

“If petitioner sought to apply its transfer pricing methodology to future years, it could have attempted to negotiate a closing agreement that made the method applicable for future years. Petitioner likewise could have applied for an advanced pricing agreement that would have set forth a ‘binding agreement’ between petitioner and the IRS as to ‘the best transfer pricing method (‘TPM’) within the meaning of § 482 of the Code and the regulations.’ Rev. Proc. 2006-9, §§ 2.04, 10.01, 2006-2 I.R.B. 278. Had petitioner sought one of these options, it would have had a genuine reliance interest grounded in a binding contract. However, the closing agreements fall significantly short of creating a legitimate reliance interest.” Order, at p. 7.

Summary J to IRS.

Of course, the leading case on IRS mind-changing, Dickman v. Commissioner, 465 U.S. 330 (1984), just happened to be decided by the same Court in the same year that decided Chevron.

Taishoff says, post-Loper Bright and post-Boechler, P. C,  exactly what is any Rev. Proc. worth? Is Mayo Foundation still good law? What is any IRS Reg. worth? The Supremes are bringing “discipline” to tax law, all right all right. Yeah, roger that.