Santander – the taxpayer wins a STARS case (for now)Posted: December 21, 2015
On November 13th, the District Court in Massachusetts issued its opinion in Santander Holdings USA, Inc. v. Commissioner. The case involved a “Structured Trust Advantaged Repackaged Securities” (“STARS”) transaction by Sovereign Bancorp, Inc. (predecessor of Santander). The government earlier won cases involving this transaction against Bank of N.Y. Mellon and American International Group, in the Second Circuit, and Salem Financial (a subsidiary of BB&T) in the Federal Circuit. There is also a pending case by Wells Fargo in the District Court of Minnesota, which is heading to trial and looks favorable for the government.
The Massachusetts District Court, however, found for the taxpayer in Santander. It also pushed back a bit against the economic substance doctrine that the government increasingly relies on in tax cases. We will be watching the case to see what happens on appeal and for possible impact in future cases.
The facts of the typical STARS transaction are complex. A brief summary is available in an article I wrote (with Lee Meyercord) on the Tax Court’s decision in the Bank of N.Y. Mellon. But the essentials are fairly straightforward. As summarized in the recent Wells Fargo decision:
The plan had four key elements: (1) the American bank would voluntarily subject some of its income-producing assets to U.K. taxation; (2) the American bank would offset those U.K. taxes by claiming foreign-tax credits on its U.S. returns; (3) Barclays would enjoy significant U.K. tax benefits as a result of the American bank’s actions; and (4) Barclays would compensate the American bank for engaging in the STARS transaction.
The compensation paid by Barclays to the American bank is referred to in the cases as the “Barclays payment” or the “Bx payment.” It plays a key role in the analysis of the transaction under the economic substance doctrine.
In a previous ruling, the Santander court held that the Bx payment should be counted as revenue to Sovereign when determining whether the transaction has a reasonable possibility of earning a profit. In the most recent opinion, the court reaffirmed that conclusion and also ruled that the foreign taxes paid to the U.K. on income earned by the trust should not be treated as an economic cost for the profit analysis. This last point follows the earlier cases of Compaq Computer Corp. v. Commissioner and IES Industries, Inc. v. United States, in the Fifth Circuit and the Eighth Circuit respectively. Decisions in other STARS cases had come to a different conclusion. The Santander court concluded that the transaction satisfied the economic substance test and should be respected. Thus, the bank was entitled to foreign tax credits to offset the foreign taxes it had paid to the U.K.
It’s still too early to determine what effect the case will have. This was a district court decision and the government will almost certainly appeal. And, of course, the Supreme Court may eventually weigh in. Cert petitions were filed in the Salem Financial, American International Group and Bank of N.Y. Mellon cases. (The government opposed cert in the Salem and AIG cases; its response in the BNY case is due January 4th.) If the Santander ruling is upheld on appeal, or if the taxpayer unexpectedly wins in the Wells Fargo case, that will increase the likelihood that the Supreme Court will decide to hear the issue. For now, this is just another step in the journey, in this case in the taxpayer’s favor.
What I found most interesting about the Santander opinion, though, was the court’s general observations about the economic substance doctrine. This judicial doctrine, dating back to Gregory v. Helvering, is an important tool for the government in combating tax shelters. But the Santander court issued an implicit call for carefully constraining that powerful tool; using it, but remaining aware of its problems and potential for abuse.
The Salem Financial, Bank of New York Mellon, and Wells Fargo cases illustrate, I think, that the judicial anti-abuse doctrines – whether substance over form or economic substance – can themselves be susceptible to abuse. . . . [Whether the transaction should be respected] can turn in large part on whether a court subjectively thinks the transaction being examined is “the sort that Congress intended to be the beneficiary of the foreign tax credit provision.”
[emphasis added] And later:
Throughout the government’s arguments in this case there has been an undertone of indignation, suggesting that the issues in the case are as much a matter of moral judgment as legal. . . What seems to bother the government is not so much that Sovereign does not qualify for foreign tax credits as that it does not deserve them. It is almost as if the government thinks that, under a sort of aiding and abetting theory, Sovereign should be punished by taking away its credit for helping Barclays manipulate its benefits under the U.K. tax laws.
The judicial anti-abuse doctrines are important, but their employment should be analytical and not visceral. Among other things, too-ready resort to the government’s “trump card” . . . may lead to the ad hoc development of novel principles of judgment solely on the basis of their utility for the particular case at hand. One serious risk is that the ultimate standard of decision becomes a kind of smell test, with the judge’s nose ending up the crucial determinant of the outcome. The more that is the case, the less predictability there is in the law, and predictability is a high value in tax law.
The court essentially concluded that if there were any abuse involved in STARS, it was Barclays’ manipulation of U.K. tax law to allow it to get an inappropriate benefit. Barclays was able to treat the U.K. taxes on the trust income as though it had paid the taxes, although the trust/Sovereign had in fact actually paid the taxes. But the U.K. tax authorities were aware of the transaction and did not challenge Barclays’ treatment of it.
A couple of weeks earlier, I had read Confidence Games: Lawyers, Accountants, and the Tax Shelter Industry. The authors criticize how lawyers and accountants analyze economic substance:
Economic substance became a formal requirement like any other and was therefore treated technically. As a consequence, satisfying the requirement became one of technique . . . rather than ascertaining the economic reality . . . .
Whether BLIPS had economic substance became a narrow technical question. . . . Treated as a technical question, the economic substance problem became decoupled from the broader question of whether the strategy would perpetrate tax fraud by allowing clients to claim losses for transactions that did not reflect economic reality.
The description of result-oriented technical manipulation of the rules aptly describes what some tax practitioners did during the late 1990’s and early 2000’s. But it struck me at the time because it describes not only how tax practitioners have approached economic substance analysis but also how the IRS and courts approach it. And that can contribute to the potential for abuse that the Santander court was concerned about. The judge is not the first to make those observations, and he won’t be the last, but his analysis does point out some concerns.
Economic substance analysis is familiar to tax practitioners and Tax Court judges. But there are several things about the analysis of economic substance in the other STARS cases that likely would seem a bit strange to those who are not tax specialists. In fact, some of these aspects may seem a bit strange even to tax specialists.
When analyzing potential probability, the courts bifurcated the STARS transaction into the “Loan” component and the “Trust” component. (Barclays contributed certain assets to the trust in return for units. The American bank agreed to repurchase all of Barclays’ units at the end of a certain period of time, for a pre-arranged price. These portions of the overall transaction are characterized for U.S. tax purposes as a loan rather than Barclays actually owning an equity interest.) This bifurcation approach to economic substance analysis originally arose because tax shelters, to avoid the doctrine, would “combine” two functionally unrelated activities, one that created profits and one that generated tax benefits, into a single transaction. The rationale for bifurcation is not as strong, though, when dealing with an integrated transaction. Here, the parts depended on each other and the parties would not have entered into either part with the other. But bifurcation of integrated transactions seems to be broadly accepted by the courts now.
When analyzing potential profitability of the “Trust” component, some courts have ignored the income earned by the assets placed within the trust. The rationale was that the assets would have earned the same income if they had not been placed in the trust. But putting assets in the trust did not, by itself, result in the American bank paying foreign taxes and claiming foreign tax credits. It was the income earned by the trust that was taxed by the U.K. There would have been no foreign taxes paid, and therefore no claim for foreign tax credits, without that income earned by the trust.
Some courts have concluded that the “Bx payment” should not be considered when analyzing the potential profitability of the transaction. This was justified because, in the words of the Wells Fargo decision, “[t]he focus of the objective inquiry is whether the transaction ‘offers a reasonable opportunity for economic profit, that is, profit exclusive of tax benefits.” But “tax benefits” in that formulation of the test originally meant the U.S. tax benefits claimed by the taxpayer and disputed by the government – here, the foreign tax credits. The STARS cases expanded the scope of this formulation by treating payments by a third party (sharing the third party’s U.K. tax benefits) as “tax benefits” rather than subsidies or compensation that are part of the American bank’s economic profit.
And, of course, the BNY decision concluded that the foreign taxes paid by the American bank should be treated as an economic cost for purposes of analyzing potential profitability, disagreeing with the earlier IES and Compaq cases.
Add all of this together, and the analysis includes as part of potential profit only: (a) transaction costs; and (b) the foreign taxes paid by the American bank. Potential profit does not include: (c) the subsidy from Barclays for participating in the transaction; and (d) the American bank’s income on which those foreign taxes were paid.
Virtually any overseas business activity will be unprofitable for purposes of the economic substance doctrine under those circumstances.
Arguably, these STARS cases have demonstrated “the ad hoc development of novel principles of judgment solely on the basis of their utility for the particular case at hand” mentioned by the Santander court. And some of the cases have a whiff of results-oriented manipulation of the technical standard that the Santander court referred to as a “smell test.”
The problem, I suspect, is that the economic substance doctrine was originally (Gregory v. Helvering) focused on the question of “whether what was done, apart from the tax motive, was the thing which the statute intended.” The Court in that case concluded that “the transaction upon its face lies outside the plain intent of the statute. To hold otherwise would be to exalt artifice above reality and to deprive the statutory provision in question of all serious purpose.”
The economic substance test that has developed over the years, though, doesn’t really match all that well with the question of what “the statute intended.” If a Son-of-BOSS tax shelter was structured in such a way that there was a reasonable possibility of profit, it would still not be what “the statute intended.” The real abuse is that the taxpayers’ claimed tax losses bore no relation to economic reality, not simply that the transaction as designed would not earn a profit. But, I suspect, the IRS and courts find it easier to just apply the mechanical test – and manipulate it with “novel principles of judgment” to reach the desired result – rather than grapple with statutory intent. Particularly when there often is no indication that Congress ever considered the potential abuse that could occur.
That leaves both sides – taxpayer and IRS/court – trying to manipulate the technical standard so that the transaction fits or does not fit within it. That creates the potential for results-oriented abuse by both sides and allow both sides to avoid the difficulty of arguing about statutory intent.
The government’s real concern about the STARS transaction is not that the taxpayer seeks to claim foreign tax credits for foreign taxes it did not actually pay. The real concern is that the taxpayer has deprived the U.S. fisc of tax dollars by shifting business activity overseas, paying taxes on the income to a foreign government and claiming a credit against its U.S. taxes on that income. If we focused on the statutory intent of the foreign tax credit instead of the technical test for economic substance, the IRS and courts would have to identify under what circumstances, exactly, shifting business activity overseas is not what Congress intended to qualify for the foreign tax credit.
Is it any time that a taxpayer shifts existing business activity to a foreign country? That seems too broad. Is it only when certain types of business activity are shifted offshore? If so, which kinds? Is it when the taxpayer does not have a “good reason” to shift the business activity offshore, and how do we determine what is a “good reason”? (And what would Congress have considered a good reason?) How should third-party subsidies that increase the overall profit factor in? What types of subsidies can be considered? How is the “Bx payment,” for example, different from an implicit subsidy in the form of lower wages?
Promulgating regulations can answer these questions in a way that carefully targets what is really abusive and provides an opportunity for public comment. In fact, the IRS did issue temporary regulations in 2007 and final regulations in 2011 that limit taxpayers’ ability to engage in transactions like STARS. But an amorphous doctrine like economic substance bypasses the procedural safeguards of the rule-making process and applies retroactively, not giving taxpayers advance warning of what is not acceptable. It also is not well suited to identifying specifically what is abusive, and why, and setting clear boundaries. It may only be used properly, against real cases of abuse. But it can be used more broadly, if the IRS chooses. In effect, the economic substance doctrine gives the IRS very broad prosecutorial discretion. Just as with very broadly drawn criminal statutes, even as we recognize the need for such an approach, we should be worried about potential abuse.