Menu
Tax Notes logo

A Question of Identity – Interest Netting, Part 2

Posted on Jan. 9, 2019

Today, guest blogger Bob Probasco brings us the second part of his post on interest netting. At the end of this part he refers to an article in the most recent edition of The Tax Lawyer. If you are a member of the ABA Tax Section you can link through to the article after signing in as a member. If you are not a member but have access to Heinonline, Westlaw or Lexis you can also get to articles of The Tax Lawyer. As the editor of that law review, I invite you to look at the articles there which explore issues in much more depth than we are able to do on the blog. I hope you would agree with me that after 70 years it continues to be a premier tax law review. If you have a law review article on tax you want to publish, consider sending it to The Tax Lawyer. Keith

We’re continuing to explore the “same taxpayer” issue for interest netting under Section 6621(d), for which the Federal Circuit issued an important decision in November. In Part 1 of this two-part series, I discussed other approaches to netting, the background of Section 6621(d), early IRS guidance, and the first of four major cases that have addressed this question. In Part 2, I’ll wrap up with the remaining cases plus some thoughts about the future.

The evolution of “same taxpayer”

Section 6621(d) allows netting only of equivalent overpayments and underpayments “by the same taxpayer.” Part 1 covered what I consider one of the more interesting attempts at IRS guidance on this issue, Field Service Advice 2002-12028,. It concluded that to qualify for this benefit, one must “both be liable for the underpayment of tax, and entitled to the overpayment of tax.” This kept the restrictive adjective “same” while creating exceptions for mergers and consolidated returns by attributing overpayments or underpayments of one corporation to another. As taxpayers starting filing cases for expansive interpretations of “same taxpayer,” the DOJ retreated from the IRS guidance to very narrow interpretations that would, in practice, make interest netting virtually impossible for the large corporations that most need it.

Balances that arose prior to consolidation: Energy East Corp. v. United States, 645 F.3d 1358 (Fed. Cir. 2011), aff’g 92 Fed. Cl. 29 (2010). This case involved acquisitions in which the subsidiaries survived as members of an affiliated group and tried to net overpayments and underpayments for pre-acquisition tax years. The taxpayers lost. That was probably the right result, although I still have some reservations. The case was decided by a “temporal requirement” that later cases borrowed. Read Part I for the gory details.

And now on to the other three cases.

Attribution of an overpayment on a consolidated return to its members: Magma Power Co. v. United States, 101 Fed. Cl. 562 (2011). Magma Power had an underpayment for its 1993 tax return; it was acquired by CalEnergy on February 24, 1995; it was included thereafter on consolidated tax returns with CalEnergy and other subsidiaries; and the consolidated group overpaid its taxes for 1995-1998. The question was whether Magma’s 1993 underpayment could be netted against the consolidated group’s overpayments. If the situation were reversed – a 1993 overpayment by Magma and consolidated group underpayments for 1995-1998 – it would be an easier decision because all group members are severally liable for group underpayments. But this fact pattern is the one that FSA 200212028 answered as “theoretically possible.”

Magma provided an affidavit to the effect that a substantial portion of the overpayments by the consolidated group were attributable to Magma. How substantial? More than 100% of the reduction of consolidated taxable income resulting in the 1995 and 1997 overpayments; 92% of the reduction of consolidated taxable income resulting in the 1996 overpayment; and 79% of the reduction of consolidated taxable income resulting in the 1998 overpayment. The government disputed the plaintiff’s methodology and amounts but conceded that some portion of the overpayments were attributable to Magma.

The Court of Federal Claims’ opinion is one of the best and most comprehensive analyses of this issue I’ve read – kudos to the court and the litigants. I think it’s also potentially the most important for future developments in interpretation of Section 6621(d). The court interpreted “same taxpayer” as “same TIN.” In that respect it accepted the government’s argument. But the court also concluded that overpayments could be attributed to individual members of a consolidated group, not just underpayments (for which all members are severally liable per regulation); the government appeared to concede neither. The court also rejected the “complete identity” or “exact DNA identity” argument of the government. Key to the decision was the court’s observation that the group is not itself a taxpayer, merely a method of combining all the members for computing tax liability; the members of the group are the taxpayers. Further, the tax liability of the consolidated group must be allocated to individual members as part of tracking E&P for each member pursuant to Section 1552. Either individual members pay their allocated share of the tax liability or the amount they don’t pay would be treated as a contribution from the other member who did pay, Treas. Reg. section 1.1502-33(d)(1). Because the tax liability is allocated and payments by the group are allocated, that should be enough to allocate overpayments to individual members as well, shouldn’t it?

After a careful consideration of legislative history, previous IRS guidance, and the remedial nature of the legislation, the court held that pre-merger Magma Power and post-merger Magma Power “should be properly considered the same taxpayer to the extent the consolidated group’s overpayment can be traced to the company” (emphasis added). Because there had not yet been an agreement by the parties or determination by the court of how much of the overpayments could be attributed, the court ordered the parties to propose further proceedings to resolve the case.

The parties entered a stipulation as to the amount owed about eleven months later – interest computations can be difficult and securing the required government approvals for settlement can drag out – and the court entered judgment. The government filed a notice of appeal on November 20, 2012, and then filed a motion in the Federal Circuit to dismiss the appeal on December 12, 2012. We have no assurance of how the Federal Circuit would rule on this issue but apparently, DOJ was not confident or at least wanted to avoid the risk of an adverse precedent.

This analysis seems to follow the “attribution to a single entity” framing of FSA 200212028 rather than a “two entities are treated as the same taxpayer” framing.

Statutory mergers, various scenarios: Wells Fargo & Co. v. United States, 827 F.3d 1026 (Fed. Cir. 2016), aff’g in part and rev’g in part 119 Fed. Cl. 27 (2014). Stephen Olsen posted here a few years ago when the Court of Federal Claims and Federal Circuit opinions came out.

This case included many different factual circumstances, resulting from a series of seven mergers and 64 separate refund claims. The government and Wells Fargo identified three “test claims”; the principles would govern all of the claims. Wells Fargo (represented by the same firm that represented Magma Power) argued that merged corporations are always treated as the “same taxpayer,” regardless of the timing of the payments. The government argued that taxpayers are only the “same taxpayer” if they have the same TIN at the time of the payments.

Scenario One: Wachovia had an overpayment for 1993; First Union had an underpayment for 1999. The two merged in a statutory merger in 2001, and First Union survived. The government argued that the netting was not available because the two corporations had different TINs and were unaffiliated at the time of both the overpayment and the underpayment.

Scenario Two: First Union had an overpayment for 1993, underwent four statutory mergers between 1993 and 1999 (in each of which it was the surviving corporation), and First Union had an underpayment for 1999. The government conceded the availability of netting in this situation; “the underpaying and overpaying company retained the same TIN because it was the surviving corporation in the mergers.”

Scenario Three: CoreStates had an overpayment for 1992 and merged with First Union in 1998 with First Union surviving. Then First Union had an underpayment for 1999. The government argued that netting was not available because the two corporations had different TINs.

The Federal Circuit mentioned the “same taxpayer = same TIN” rule from Magma Power, without explicitly adopting the rule. But it mischaracterized the Court of Federal Claim’s application of the rule as that “the consolidated group or corporations met the ‘same taxpayer’ requirement because they shared a single TIN.” As discussed above, that is not what the CFC did in Magma Power. The court focused on the TIN of the subsidiary, Magma Power, rather than the consolidated group. It allowed netting, but only if the consolidated return’s overpayment could be traced or attributed to the company.

Based on merger law, the Federal Circuit concluded that two merging corporations are the “same” regardless of which survives. The Court of Federal Claims declined to apply the temporal requirement from Energy East because joining an affiliated group (when both corporations maintain their separate identity) differs from a statutory merger (in which only once corporation survives). But the Federal Circuit disagreed, applied the temporal requirement, and allowed netting in Scenario Three but not in Scenario One.

In Energy East, the Federal Circuit seemed to say that the taxpayer must be the same before both the underpayment and the overpayment. Of course, that was the situation being decided; the court wasn’t dealing with a situation with an overpayment made prior to the acquisition. But in Wells Fargo, the court allowed netting in Scenario Three, when the overpayment was made prior to the merger. Apparently, the requirement is that the taxpayer must be the same when the overpayments or underpayments are made. That seems plausible; there are no overlapping balances to be netted until the second balance comes into existence.

I don’t find the temporal requirement imposed in Energy East and Wells Fargo completely persuasive. The court interpreted an overpayment or underpayment as being associated with a particular date rather than a period. That’s understandable, given the language in Sections 6601(a) and 6611(b). But I think the antiquated language of the Code has effectively been superseded by Avon Products Co. v. United States, 588 F.2d 342 (2d Cir. 1978), and the realities of tax administration. In Avon Products, the Second Circuit concluded that individual transactions must be netted into a single balance before computing interest. The IRS has acquiesced in not only the result of that case but also the reasoning. Over a period of time, the single balance required by Avon Products may change from underpayment to overpayment back to underpayment, and the traditional determinations of the “date” of an overpayment and underpayment no longer fit well. They are better considered in terms of a period rather than a particular date when they arise. Admittedly, no court has yet reached the same conclusion.

Further, the “last antecedent rule” is simply an interpretative standard, not an ironclad rule any more than other canons of statutory construction. Given the remedial nature of netting, it would have been possible to interpret the provision more broadly. Even if both the underpayments and overpayments began before the merger or joining into an affiliated group, netting might be permissible if both are still outstanding afterward. At that point, once the two corporations become the “same taxpayer,” the harm that Section 6621(d) is intended to remedy exists. Netting might be allowed from that point. This argument is likely stronger for mergers than for affiliated groups that file consolidated returns.

But that’s not what the court ruled.

Not part of the parent’s consolidated return: In Ford Motor Company v. United States, 908 F.3d 805 (Fed. Cir. 2018), aff’g 132 Fed. Cl. 104 (2017), the court concluded that Ford Motor Company (“Ford”) and a wholly-owned subsidiary were not the “same taxpayer” for purposes of interest netting. Against the backdrop of the earlier cases, Ford seems an even harder case for the taxpayer to win. Ford formed Ford Export Services B.V. (“Export”), its wholly-owned subsidiary, in 1984 as a foreign sales corporation. Because FSCs must be foreign rather than domestic corporations and a consolidated group cannot include foreign corporations, Ford and Export filed separate U.S. income tax returns between 1990 and 1998. Ford had an overpayment for the 1992 tax year and Export underpaid its taxes for 1990-1993 and 1995-1998. Ford claimed that Ford and Export were the same taxpayer because Ford “exercised near complete control over Export’s operations” and “Export never performed any activity that Ford did not direct.”

Instead of relying on a dictionary, the Federal Circuit concluded that the meaning of “same taxpayer” depended on “background legal principles” at the time Congress enacted Section 6621(d). One of those background legal principles was that a parent corporation and its subsidiaries are separate taxable entities. (Cases cited in the opinion concluded that even if the parent and subsidiaries join to file a consolidated return, the group is not as a single entity and individual members retain their identity.) The court identified “the unique legal effects of a merger” as an exception to that general rule, citing Wells Fargo. Ford argued that the FSC statute was another relevant background legal principle that would constitute an exception to the general rule, but the court disagreed. There was no statutory provision for FSCs comparable to the continuation of the identity of the acquired corporation in the successor corporation after a merger.

The plaintiff’s arguments here were based on control and direction because of the factual situation: Export was not a member of the affiliated group that filed a consolidated return. That suggests that the decision may be relatively narrow in scope and may also help explain why the court reached this decision. This level of “control and direction” will most often occur with a subsidiary that meets the 80% voting and value test for an includible corporation in an affiliated group (even if not includible for other reasons such as being a foreign corporation). Consider three categories such subsidiaries might fall into: (a) included in the affiliated group; (b) not included in the affiliated group but a foreign corporation with no US income; and (c) not included in the affiliated group but a corporation that files a US tax return. If the Federal Circuit eventually blesses netting involving attribution from consolidated returns, similar to FSA 2001-12028 and Magma Power, the effect of the decision in Ford may be limited to category (c). Further, allowing netting based on control and direction, rather than attribution from a consolidated return, would impose more of an administrative burden on the IRS because the factual determination is more difficult. It’s not well suited for a general background legal principle to apply as an interpretation of “same taxpayer.” On the other hand, netting with subsidiaries who meet the 80% test but are not includible for other reasons, might be a viable test.

What does the future hold?

The boundaries of “same taxpayer” are still not entirely clear. Because most netting claims include a request for additional overpayment interest, these cases will usually be brought in the Court of Federal Claims rather than district court. Review by the Supreme Court and legislative action both seem unlikely in the extreme. So the Federal Circuit’s decisions carry a lot of weight. But there are some possibilities the Federal Circuit has not yet ruled on.

In a statutory merger, the Federal Circuit will allow netting of a pre-merger balance against a post-merger balance but not if both balances are from pre-merger years. If a taxpayer or the government wants to change that, it will be difficult to do just through litigation.

The Federal Circuit has never directly ruled on netting in the context of consolidated returns. Energy East and Wells Fargo involved mergers; a Magma Power appeal was dismissed on the government’s request; and Ford involved a subsidiary that was not consolidated. The opinion in Ford seems to suggest that these situations are not an exception to the general rule that a parent and its subsidiaries are separate entities and therefore are not the same taxpayer. I don’t think that necessarily precludes netting, though.

First, I suspect that as a practical matter the IRS has been allowing netting administratively, and thus there have been no lawsuits, in circumstances where one member has an overpayment based on a separate tax return and the consolidated return has an underpayment. We haven’t seen such a case in the Federal Circuit and I would expect to by now if the IRS were disallowing the claims. If that’s correct, I’m not sure that the opinion in Ford will be enough for the IRS to change its administrative practice.

Second, treating a parent and its subsidiaries as separate entities does not preclude netting if you allow attribution of underpayment and overpayments to individual members of the affiliated group. Then, as in Magma Power, you are dealing with a single taxpayer; the attribution results in one taxpayer having equivalent overpayments and underpayments outstanding at the same time. The Federal Circuit has not addressed the attribution theory and, based on its mischaracterization of Magma Power in Wells Fargo, may not have considered it yet.

Netting in the context of consolidated returns, under an attribution theory, will still create administrative issues. Even in simple netting claims, the taxpayer cannot use balances if it has previously used them for netting. For example, if Corporation A nets a 2001 underpayment against a 2004 overpayment, Corporation A cannot later net the same 2001 underpayment against a 2006 overpayment. This would apply for netting under an attribution theory in the context of consolidated returns. Further, the netting claim would have to provide documentation to support the attribution. But such administrative issues should not be an impediment. Revenue Procedure 2000-26 simply shifts the burden to the taxpayer requesting netting. A more significant problem with netting in the context of consolidated returns might arise with respect to disputes between members of the group about the attribution – particularly after a member has left the group but still wants to net past balances as part of the group against post-departure balances.

Are there situations other than mergers and consolidated returns that might supply “background legal principles” to justify netting? Contractual assignments of tax liabilities and the right to refunds of overpayments might be a possibility, although courts are often reluctant to be bound by those when deciding tax issues. There may be others.

Even if the Federal Circuit is unlikely to approve netting in the consolidated return context or other situations, it’s worth still contesting issues on which the Federal Circuit has not yet ruled. Interest disputes for the largest corporations can involve significant amounts, which alone justifies taking a shot. Ford, as an example, involved a $20 million claim while Wells Fargo involved a $350 million claim (although that included items other than netting).

Postscript

As I was finishing this post up, I received the latest volume of The Tax Lawyer, which includes an article “More of the ‘Same’: Section 6621(d) in the Federal Circuit” by David Berke, an associate at Skadden, Arps. He and I are not in complete agreement, but for those with an interest in a different perspective on this topic, it’s worth perusing.

DOCUMENT ATTRIBUTES
Copy RID