On February 12, the CEO of Barclays announced that the bank is shuttering its U.K. tax-structuring unit. The unit had previously been extremely lucrative and high-profile in structuring and executing tax-advantaged transactions in the United Kingdom.1
It is unclear if the move was the result of a change in the appetite for such transactions in the United Kingdom market or represented a change of view by Barclays about the transactions.
Barclays also announced that, going forward, all transactions must meet each of five criteria:
1. support genuine commercial activity
2. comply with generally accepted custom and practice, domestic law and the United Kingdom Code of Practice on Taxation of Banks
3. be of a type that tax authorities would expect
4. take place with only customers and clients sophisticated enough to assess their risks
5. be consistent with, and be seen as consistent with, the bank’s purpose and value.2
The third criterion is particularly interesting: “be of a type that tax authorities would expect”. Few, if any, tax advisors in the United States or the United Kingdom would articulate such a standard as either a legal or ethical requirement. Nonetheless, there are indications that similar principles are finding their way into U.S. tax law.
For instance, if tax planning is based on interpretation of a statute that is ambiguous on its face, the Supreme Court has held that legislative history is to be considered.3 In the context of the tax law, determining whether a transaction is supported by legislative history often implicitly means considering whether the transaction is of a type that Congress intended to address with the law in question; such a consideration is quite close to whether a transaction is “a type that tax authorities would expect.”
Another example of this standard creeping into our tax law is the Federal Circuit’s decision in the ConEdison “lease-in/lease-out” case. In that case, the Federal Circuit disallowed substantial up-front rental deductions to a New York utility that had contracted to be the lessee of a power plant in the Netherlands.4
The court did not find that the transaction violated any statute, regulation or specific existing common law doctrine. Instead, the court created an entirely new doctrine: a leasing transaction fails if there is a fixed- price purchase option that is “reasonably likely” to be exercised. It cited no precedent supporting such a standard for analyzing a purchase option in a lease. Further, it offered no explanation as to how the standard jibed with (i) the seminal Supreme Court case upholding a lease with multiple fixed-price purchase options5 or (ii) a revenue ruling holding that a fixed-price purchase option will not undermine true lease treatment so long as the purchase price is not nominal in relation to the expected value of the property, as determined at the time of entering into the agreement, or relatively small compared to the total payments due under the lease.6
The best rationale I have for understanding the Federal Circuit’s holding is that the court was influenced by the principle cited by Barclays: a transaction must be of the type that the tax authorities would expect. The court did not say so, but it appears to me that the outcome was driven by the Federal Circuit’s reaction to the transaction as being outside the bounds of the types of transactions to which the tax accounting rules for rent were intended to apply.
1 David D. Stewart, U.K. Banking Giant Barclays Shutters Tax Planning Unit, 2013 World Tax Daily 30-6 (Feb. 13, 2013).
3 See Wisc. Public Intervenor v. Mortrier, 501 U.S. 597 609, n. 4 (1991).
4 Consolidated Edison Co. of N.Y. v. United States, 703 F.3d 1367 (Fed. Cir. 2013).
5 Frank Lyon v. U.S., 435 U.S. 561 (1978).
6 Rev. Rul. 55-540, 1955-2 C.B. 39.