On August 25, 2022, the Sixth Circuit Court of Appeals answered a fundamental question about advance pricing agreements (APAs): Is an APA a contract between the IRS and a taxpayer? If it is a contract, is it subject to contract law? Circuit Judge John Nalbandian answered this question for the unanimous three-judge Appellate panel, opening the decision with the statement:
Taxes may well be “what we pay for civilized society,” … but that doesn’t mean the tax collector is above the law. This case arises from the IRS’s efforts to circumvent basic contract law.
My two cents opinion is that the Sixth Circuit’s holding is sound and important to preserve the integrity of the APA program as a tax risk management for both the IRS and taxpayers. If a taxpayer calculation error allows the IRS to cancel an APA and interject a new transfer pricing method (TPM), then taxpayers, especially foreign taxpayers, will no longer have a tax risk management incentive to agree on an APA with the IRS. As I tell my transfer pricing students each year, “Math is hard, tax math is harder, and transfer pricing math is guesswork”.
Background of the case is necessary to appreciate Eaton’s overlapping accounting and financial systems from which it may pull data, and the complexity of the adjustments to that data to arrive at transfers prices that resemble those of comparable uncontrolled parties as well as to validate that the transfer prices lead the tested party to fall within a profit range using the IRS’ suggested Berry Ratio transfer pricing method. I think a deep dive reading about the multiple data systems of Eaton and the manual constructions necessary (contained in the 2017 Tax Court decision) is necessary to fully understand the challenges a multi billion dollar MNEs transfer pricing or tax director faces when attempting to generate arm’s length prices for intricate products produced a manufacturing operation.
Moreover, I think that this case has a similarity to the Medtronic case. I think that the IRS takes an approach of ‘heads I win, tails you lose.’ If an IRS-taxpayer agreement’s TPM and methodology for calculation results in an outcome the IRS finds favorable, then the IRS will respect the agreement. But if the IRS does not like the outcome, it is allowed to retroactively nullify the agreement and impose another TPM. The taxpayer is not afforded the same deference by the IRS when the taxpayer does not like the result.
I do not understand the IRS’ decision-making regarding it canceling the APAs from the perspective of the IRS-Eaton relationship. I must be missing something here? Was it a case of a future audit team not liking the result generated by an agreement with a past audit team? Said another way, ‘second guessing’. Was it that the IRS believed that Eaton, a significant NYSE listed corporation, committed fraud by lying to the IRS? I just do not see the benefit to a tax director to ‘cheat’? Was this case chosen for litigation by the IRS to prove a point (one that goes over my head)? If a reader knows the IRS thinking about why canceling the APAs, of all of its options, made the most sense, please relieve my curious mind.
The Background of the Case
Background of the case is necessary to understand how the APAs came about, Eaton’s system errors in calculating the transfer price pursuant to the agreed methodology, and the IRS’ canceling of the APAs.
Eaton is a power management company doing business in more than 175 countries, U.S.-founded in 1911. In 2012 Eaton redomiciled to Ireland its New York Stock Exchange listed parent via an acquisition of Dublin-headquartered Cooper Industries. Eaton developed, manufactured, and sold circuit breaker and electrical control products (“breaker products”) through various manufacturing plants in Puerto Rico and the Dominican Republic (the Puerto Rico operations). A single breaker product can have as many as 100 component parts, and its quality is heavily regulated to be sold in the U.S. In 2005 and 2006 Eaton’s Puerto Rico operations manufactured most of the component parts that went into the breaker products.
In 2000, Eaton and its Puerto Rico operations entered into a license agreement whereby Eaton granted a nonexclusive license to use, including the right to sublicense, a broad class of intangible property used to manufacture and assemble certain breaker products for a royalty of 4 percent of net sales of the licensed breaker products. The royalty was increased to 6.45 percent in exchange for a license to manufacture additional breaker products.
Eaton used a financial reporting and management system called Hyperion for various purposes, including financial and legal consolidation of its several accounting ledgers. Eaton’s financial reporting to the Securities and Exchange Commission (SEC) included financial results segmented by business area and geographic region. Eaton used Encore, a system that received data from its ledgers, and Corptax, a system that consolidated its ledger data for U.S. tax reporting purposes. Accounting ledger data was maintained in an Oracle database. Eaton prepared its tax returns using a Path8 Hyperion report that mapped individual ledgers into legal entities. For internal purposes only such as managerial unit cost and profit accounting, Eaton set the price for breaker products transferred internally to be 1.3 times the cost of manufacturing the products.
The IRS examined Eaton’s tax returns for the taxable years 1994 and 1997 and proposed adjustments related to transfer pricing for certain products manufactured by Eaton’s Puerto Rico operations. Eaton and the IRS settled that matter in 2002 using either a “cost-plus” or “gross profit markup” transfer pricing methodology (TPM). Using the settlement, in 2002 Eaton sought an APA based on its contention that the comparable unrelated parties (CUP) method would serve as the best TPM for taxable years going forward.
After 18 months of meetings, providing data, and negotiating, Eaton Corporation and the IRS agreed, initially in 2004 and then a renewal, a pair of APAs to establish the transfer pricing method for Eaton’s tax calculations from 2001 through 2010 (the 2001–2005 tax years and 2006–2010 tax years respectively). Eaton’s initial APA addressed the prices at which Eaton’s Puerto Rican operations sold finished products to Eaton’s U.S. affiliates, the license of manufacturing technology by Eaton’s U.S. operations to the Puerto Rican operations, and a cost-sharing payment made by the Puerto Rican operations to Eaton’s U.S. operations. The second APA only addressed Eaton’s breaker product sales from its Puerto Rican operations to its U.S. affiliates.
For the APAs Eaton and the IRS agreed on a transfer price of 1.8 times the cost that was derived via a complicated CUP analysis due to the number of component parts involved and the availability of granular data that could be generated by the accounting systems. Eaton and the IRS agreed to apply as a validation TPM the Berry Ratio, gross profit divided by operating expenses, equal to a range of 1.20 to 1.27 (i.e. operating profit equals 20 – 27 percent of operating expenses). If the yearend Berry ratio analysis did not fall within the range of range of 1.20 to 1.27, then Eaton would make an adjustment to the purchase price to bring the yearend within the range. Eaton’s mirror ledgers were distinct from the constructed income statement required by the APA TPM. Unlike the mirror ledgers, which exist as part of EEI’s accounting system, the constructed income statement did not actually exist outside of the APA TPM. Eaton and the IRS agreed in the APAs that the APAs’ legal effect and administration were governed by Revenue Procedure 96-53 and Revenue Procedure 2004-40 respectively.
In 2010, Eaton reviewed its APA tax records and caught inadvertent calculation errors. The data and computational errors fell into two categories: (1) an error in the APA multiplier that caused the transfer price computed under the APA TPM to be reflected incorrectly in the books and records, and therefore reflected incorrectly on its tax returns, and (2) errors that affected the computation of the transfer price under the APA TPM. The errors were caught because of new employees taking over the transfer pricing tasks and challenges with recording data disparately in the accounting and tax systems. For example, Eaton’s VISTA system did not record actual manufacturing costs. VISTA’s standard cost needed to be adjusted for variances to determine actual manufacturing costs. The adjustment was manually determined using the Puerto Rico operations ledgers to derive how each plant’s actual manufacturing costs varied from its expected standard costs.
Eaton disclosed the errors to the IRS, corrected the mistakes, and in 2010 filed amended returns with the additional tax that was owed for past years. However, in 2011, the IRS decided that Eaton’s errors were serious and canceled both APAs, the first for its last year of 2005, and the second one from its start date. The IRS then issued a notice of deficiency for 2005 and 2006 of $75 million of additional income taxes and $51 million of penalties based upon it applying an alternative transfer pricing methodology that led to a reallocation of income of $102,014,000 for 2005 and $266,640,000 for 2006 (a total of $368 million) to Eaton from its Puerto Rican subsidiaries.
Four Tax Court Cases
Eaton filed suit to the Tax Court. Eaton argued that an APA is a contract and thus governed by contract law. If governed by contract law, Eaton argued that the IRS had a burden to show that its unilateral cancellations are appropriate under contract law. The IRS countered that an APA is governed by a revenue procedure that gives the APA legal effect. If APAs are procedural instruments, then the IRS argued that the administration and cancellation are IRS administrative determinations for which Eaton must show the IRS abused its discretion in administering.
On June 26, 2013, the Tax Court addressed its jurisdiction to review the IRS’s cancellations of an APA and the burden of proof that must be met by a taxpayer challenging such APA cancellation. The Tax Court held it had jurisdiction to review an APAs cancellation but that a taxpayer must show the IRS abused its discretion. Abuse of discretion is a question of fact to show the IRS action was arbitrary, capricious, or without sound basis in fact.
On April 6, 2015, the Tax Court ruled that Eaton’s internal emails, memos, and data compilations, generated before the APA in relation to the APA are protected by the attorney work product doctrine stating:
Considering all of the circumstances, we conclude that the petitioner subjectively believed that litigation with the IRS regarding the appropriate TPM for its Puerto Rico operations was a real possibility at the time it requested and negotiated the APA.
Moreover, the Tax Court found that the communications were protected by the attorney-client privilege because “… petitioner’s tax attorneys and tax practitioners were actively engaged in advising petitioner how best to present and defend its position that the “CUP” method was the best TPM in respect of its operations in Puerto Rico. The record shows that the communications in question were intended to be confidential.” However, the Tax Court ruled that Eaton waives its privilege protecting the communications from discovery because Eaton argued a reasonable cause/good faith defense for its errors that puts into contention the subjective intent and state of mind of those who acted for Eaton.
On July 26, 2017, the Tax Court rendered its decision as to whether the IRS had abused its discretion in its determination to cancel the two APAs. The Tax Court found that the IRS abused its discretion and thus could not cancel the APAs because canceling an APA to change the underlying methodology is inappropriate. The Tax Court stated that Eaton’s errors were not pertinent to the APAs being agreed but rather were made after the fact in applying the APA methodology. Eaton’s errors were not an attempt by it to change the agreed methodology. Eaton thought it was using the correct data, and when it discovered it was not, disclosed the errors and why they occurred, then corrected the calculations with the correct data. The Tax Court concluded its analysis:
An APA is a binding agreement and it should be canceled only according to the terms of the revenue procedures. It should not be canceled because of a desire to change the underlying methodology of a method that would result in a significantly different profit split.
Without regard of the Tax Court holding that the APAs continued through 2010, the IRS pressed forward with its imposition of 40 percent penalties for its issued deficiencies for 2005 and 2006. On October 28, 2019, the Tax Court found for Eaton because the Tax Court had ordered that the APAs remained in effect and the deficiencies pursuant to the APAs were not transfer pricing adjustments but rather calculation error adjustments.
The Sixth Circuit Appellate Decision
Addressing de novo the Tax Court’s determination of burden of proof, the Sixth Circuit stated:
The threshold question is this: Who has the burden? Both APAs incorporated the Revenue Procedures, which say that the IRS “may cancel” the agreement for various enumerated grounds, including, among other things, “mistake as to a material fact” or the “failure to state a material fact.” … Must the IRS prove these grounds consistent with contract-law principles, or is it Eaton’s burden to show that the cancellation was “plainly arbitrary”?
And the Sixth Circuit forcefully enunciated its answer:
The IRS has the burden. In arguing to the contrary, the IRS hides behind administrative deference to avoid the consequences of its bargain. In its view, after the parties spent years negotiating a bargain of this complexity, the government can simply rip up the contract unless the taxpayer can show that doing so is “plainly arbitrary.” … Cancelling a contract is just like any other agency determination—forget about what contract law demands, says the IRS. We disagree. Neither the available cases nor the IRS’s own regulations and procedures support the IRS’s argument.
Pursuant to the Revenue Procedure, an APA is a binding agreement between the taxpayer and the Service. The Sixth Circuit found that the case law clearly establishes that when the United States enters into contract relations, its rights and duties therein are governed generally by the law applicable to contracts between private individuals. Because contract law applies, the IRS must prove the exception it alleges that allows it to back out of its contractual promises. For example, the Court cited Tax Court decisions that the party, either taxpayer or the IRS, seeking to set aside a closing agreement bears the burden of proving fraud, malfeasance, or misrepresentation.
After the Court applied canons of contract interpretation, it found that the grounds for the IRS’ ability to cancel do not extend beyond the APA’s cancellation subsection. The cancellation subsection does not include errors in “the supporting data and computations” used in applying the TPM. The subsection allows cancellation for the failure of a critical assumption, the taxpayer’s misrepresentation, mistake as to a material fact, failure to state a material fact, failure to file a timely annual report, or lack of good faith compliance with the terms and conditions of the APA.
The Revenue Procedure states that material facts are those that, if known by the Service, would have resulted in a significantly different APA (or no APA at all). The IRS will consider facts as material if, for example, knowledge of the facts could reasonably have resulted in an APA with significantly different terms and conditions. The Sixth Circuit held that after-the-fact miscalculations cannot thus be “material” pursuant to the IRS’ definition of materiality.
Concluding about the IRS’ appeal of the Tax Court decision regarding penalties, the Sixth Circuit stated that:
If the penalty doesn’t cover the same adjustments, it’s a different penalty. After all, the same statute that authorizes these penalties makes clear that each penalty pegs to specific “portion[s] of the [taxpayer’s] underpayment. 26 U.S.C. § 6662(h)(1).
The IRS asks us to forget that it advanced one set of calculations before trial and an entirely different set after trial. But the IRS can’t do that—its changeup clearly implicates “the presentation of different evidence.” Under the IRS’s proposed rule, the government is free to mix and match an infinite number of theories—before trial, after trial, and anywhere in between—so long as everyone agrees that Eaton made some sort of error in its transfer pricing for 2005 and 2006. That is a bridge too far.
… the IRS explicitly rejected Eaton’s APA-based self-corrections at first. It did this by rejecting Eaton’s amended tax returns, which were calculated using the TPM. Instead, the IRS continued to insist on its own, non-APA method. That is, of course, until after trial. And now, the IRS must take the bitter with the sweet. It cannot reject Eaton’s adjustment method before trial, but then claim after trial that it was arguing the same thing all along.
Consequently, the Sixth Circuit held that the IRS may not impose penalties on Eaton’s adjustment due to its calculation errors.
I will continue to think about this case and its ramifications, and complete my thoughts in my 2023 edition of Practical Guide to U.S. Transfer Pricing (Matthew Bender).
 Eaton Corp. v. Comm’r, 2022 U.S. App. LEXIS 23853 (6th Cir.).
 Eaton Corp. & Subsidiaries v. Comm’r, T.C. Memo 2017-147, 13, 2017 Tax Ct. Memo LEXIS 147, *9.
 Eaton Corp. & Subsidiaries v. Comm’r, 2017 Tax Ct. Memo LEXIS 147.
 Eaton Corp. & Subsidiaries v. Comm’r, 2017 Tax Ct. Memo LEXIS 147,
 Eaton Corp. v. Comm’r, 2015 U.S. Tax Ct. LEXIS 66, *4.
 Eaton Corp. v. Comm’r, 140 T.C. 410, 2013 U.S. Tax Ct. LEXIS 19.
 Eaton Corp. & Subsidiaries v. Comm’r, 2017 Tax Ct. Memo LEXIS 147, *64
 Eaton Corp. & Subsidiaries v. Comm’r, T.C. Memo 2017-147, 91, 2017 Tax Ct. Memo LEXIS 147, *94.
 Eaton Corp. v. Comm’r, 2022 U.S. App. LEXIS 23853 (6th Cir.) *2.
 Eaton Corp. v. Comm’r, 140 T.C. 410, 411, 2013 U.S. Tax Ct. LEXIS 19, *3.
 Eaton Corp. v. Comm’r, 140 T.C. 410, 2013 U.S. Tax Ct. LEXIS 19.
 Eaton Corp. v. Comm’r, 2015 U.S. Tax Ct. LEXIS 66, *1, *8.
 Eaton Corp. v. Comm’r, 2015 U.S. Tax Ct. LEXIS 66, *10.
 Eaton Corp. v. Comm’r, 2015 U.S. Tax Ct. LEXIS 66, *18-19.
 Eaton Corp. & Subsidiaries v. Comm’r, T.C. Memo 2017-147, 193, 2017 Tax Ct. Memo LEXIS 147, *204.
 Eaton Corp. & Subsidiaries v. Comm’r, 153 T.C. 119, 2019 U.S. Tax Ct. LEXIS 25, 153 T.C. No. 6.
 Eaton Corp. v. Comm’r, 2022 U.S. App. LEXIS 23853, *11-12, 2022 FED App. 0202P (6th Cir.).
 Rev. Proc. 2004-40, § 9.01.
 Eaton Corp. v. Comm’r, 2022 U.S. App. LEXIS 23853, *24, 2022 FED App. 0202P (6th Cir.).
 Eaton Corp. v. Comm’r, 2022 U.S. App. LEXIS 23853, *36-37, 2022 FED App. 0202P (6th Cir.).
To make sure you do not miss out on regular updates from the Kluwer International Tax Blog, please subscribe here.
Kluwer International Tax Law
The 2022 Future Ready Lawyer survey showed that 78% of lawyers think that the emphasis for 2023 needs to be on improved efficiency and productivity. Kluwer International Tax Law is an intuitive research platform for Tax Professionals leveraging Wolters Kluwer’s top international content and practical tools to provide answers. You can easily access the tool from every preferred location. Are you, as a Tax professional, ready for the future?
Learn how Kluwer International Tax Law can support you.
Thanks for the detailed discussion of the financials from the previous court decision. The big issue is whether the intercompany royalty rate was arm’s length. That Berry ratio issue comes down to the fact that distribution operating expenses had to be 13% of sales so a Berry ratio near 1.23 translates into a 16 percent gross margin and hence a 3 percent of sales. No controversy there. Now maybe I need to model out the consolidated financials but it is hard for me to believe that a reasonably applied profits based approach would lead to a higher royalty rate than the one reached in the most recent APA. So in a way – this is sort of like Medtronics but something leaves me puzzled how the IRS expected a significantly higher royalty rate than they got from the APA.