Tax Court Holds STARS Transaction Lacked Economic Substance

Jill A. Frisch, Curt M. Rubin, Anne O'Brien Hintermeister, Matthew J. Avon, Justin L. Campolieta, and Michael A. Sienkiewicz, for respondent.

KROUPA, Judge: Respondent determined deficiencies in petitioner's Federal income tax of $100 million² and $115 million for 2001 and 2002 (years at issue), respectively. There are three issues for decision. The first issue is whether petitioner is entitled to foreign tax credits under section 901³ claimed in connection with a Structured Trust Advantaged Repackaged Securities transaction (STARS transaction or STARS). We hold that petitioner is not because the STARS transaction lacked economic substance. The second issue is whether petitioner is entitled to deduct certain expenses incurred in furtherance of the STARS transaction. We hold petitioner is not for the same reason. The final issue is whether income attributed to a trust with a U.K. trustee used to effect the STARS transaction is U.S. source income rather than foreign source income. We hold that the income is U.S. source income.⁴

The affiliated group through BNY entered into the STARS transaction in 2001 with Barclays Bank, PLC (Barclays), a global financial services company headquartered in London, United Kingdom. The STARS transaction generated approximately $199 million in foreign tax credits for the combined years at issue.

STARS was represented as a "below market loan" in KPMG's initial presentation. KPMG indicated that STARS required a U.K. counterparty and a certain trust structure holding income-producing assets. KPMG explained that the below-market cost would be achieved by the U.K. counter party "sharing" U.K. tax benefits from STARS through an offset to the cost of the loan. Finally, KPMG indicated that the U.K. tax benefits would be generated by subjecting income-producing assets held by a trust to U.K. tax and thus generating foreign tax credits that BNY could use to offset its U.S. tax liability.

BNY notified KPMG in August 2001 that it was prepared to move forward with a STARS transaction with Barclays as the U.K. counterparty. BNY proposed that it would contribute assets that would generate $93 million of annual U.K. tax costs and expected Barclays to reduce the loan's annual cost by half that amount. Shortly thereafter, BNY agreed to supplement STARS by engaging in a "stripping transaction." The effect would be to accelerate and increase the tax benefits STARS produced (i.e., foreign tax credits). And just before STARS closed, BNY indicated to Barclays that it had decided to increase the targeted benefit.

A. The STARS Structure

BNY used existing subsidiaries and created special-purpose entities to create a structure (STARS structure) to carry out the STARS transaction. BNY accomplished this by engaging in the following steps.

The DelCo class 1 shares held all the voting rights in DelCo. The DelCo class 2 shares had the right to receive approximately 99% of DelCo's distributions. The holders of DelCo class 1 shares had the exclusive right to appoint DelCo's managers. DelCo's income was distributable in the absolute discretion of DelCo's managers.

InvestCo transferred the remaining STARS assets (approximately $1.2 billion) and the DelCo class 2 shares to the trust in exchange for the class A units and the class B unit, which had stated values of $6.3 billion and $1.494 billion, respectively.

The initial trustee was BNY, acting through its London branch (U.S. trustee). The Bank of New York (DE), a wholly-owned subsidiary of BNY Parent, served as the trust manager. Only the holder of all the class A units could nominate a replacement trustee.

In sum, the above steps moved approximately $7.86 billion in net assets into DelCo and the trust. The following chart summarizes steps 1 through 5.

 

 

BNY and Barclays entered into the following agreements and transactions, the net effect of which was to create a $1.5 billion loan to BNY from Barclays.

Investco and Barclays entered into another forward sale agreement obligating InvestCo to purchase the class D unit (class D unit forward sale agreement ) from Barclays within 90 days of the purchase by InvestCo of the class C unit, for $25 million plus any additional amount for any accrued but unpaid distributions on the class D unit. The sale price under the class D unit forward sale agreement was the same as the original subscription price of the class D unit.

The following diagram broadly reflects the terms of the various agreements making up the loan.

 

 

The trust immediately redeemed InvestCo's class B unit with the $1.494 billion the trust received from Barclays' purchase of the class C unit and the class D unit. InvestCo then placed $1.5 billion on deposit with a BNY branch office in the Cayman Islands (Cayman branch). After an initial 11-day term, the money was held on deposit at the Cayman branch in 1-month terms for the duration of STARS. The Cayman branch booked this deposit as a liability to Barclays.

D. Replacement of the U.S. Trustee

BNY and Barclays replaced the U.S. trustee with The Bank of New York Trust and Depositary Company Limited (U.K. trustee), which was treated as a U.K. resident for U.K. tax purposes. The U.K. trustee was a wholly-owned subsidiary of BNY parent.

E. The Stripping Transaction

The parties entered into a series of agreements slightly over a month after STARS closed to accelerate the U.K. taxes due on trust income by converting periodic cashflows into an up-front taxable lump-sum payment (stripping transaction). These agreements contemplated the following steps.

First, BNY would contribute $402 million to DelCo through REIT Holdings, InvestCo and the trust. Second, the U.K. trustee would transfer the trust collateral securities to BNY as "custodian" in exchange for principal-only receipts and interest-only receipts. Third, DelCo would use the contributed funds to purchase the interest-only receipts from the trust for $402 million. Fourth, the collateral arrangements would be amended so that Barclays obtained a security interest in the principal-only receipts and the interest-only receipts.

To effect the stripping transaction, the trust exchanged the trust collateral securities for the interest-only receipts and principal-only receipts, which represented beneficial ownership in the interest payments and principal payments, respectively. DelCo then purchased the interest-only receipts for $402 million from the trust. The funds used to purchase the interest-only receipts were not transferred in accordance with steps contemplated in the transaction documents. Instead, BNY transferred $402 million directly to the trust's bank account.

Barclays was granted a security interest in the principal-only and interest-only receipts that were transferred to a trust security account and DelCo security account, respectively.

For U.K. tax purposes, the trustee treated the $402 million from the sale of the interest-only receipts as taxable income at the time of the sale. The U.K. trustee was required to pay U.K. taxes on the taxable income. Under U.S. tax rules, however, the trust did not report a gain or loss.

The post-tax income from the stripping transaction was distributed to the class C unit holder in the next monthly period. BNY received its benefit, a portion of the spread, over a period of 14 months.

Net the stripping transaction added $402 million in income to the trust, over and above the monthly target amount, that was taxable in the United Kingdom and generated additional trust taxes and foreign tax credits of $88 million in 2001.

BNY ignored the stripping transaction in managing and disposing of the stripped securities. When the trust manager sold a stripped security, the trust manager would reconstitute the security and withdraw it from the collateral pool.

F. Management and Control of Trust Assets and DelCo Assets

The trust manager held absolute discretion in managing the trust assets. The trust manager delegated its authority to BNY through a servicing agreement for which BNY received a monthly fee. BNY also agreed to manage DelCo's assets for a monthly fee and was authorized to take any necessary action.

G. Allocation of STARS Risk

BNY and Barclays also took steps to apportion risk associated with STARS. These steps are as follows.

Barclays engaged in transactions substantially similar to STARS with other U.S. banks. Barclays disclosed one of those transactions to the U.K. tax authority in June 2001 while it was negotiating STARS with BNY. Barclays disclosed the STARS transaction in April 2002 to the U.K. tax authority. The U.K. tax authority advised Barclays that it agreed with Barclays' tax reporting of the STARS transaction in June 2002. The STARS transaction increased tax revenue for the United Kingdom.

B. U.K. Tax Treatment of the Trust

The trust was treated as an unauthorized unit trust under U.K. law that qualified as a collective investment scheme under the applicable U.K. regulatory laws. When the U.K. trustee replaced the U.S. trustee, the trust became subject to U.K. tax as a collective investment scheme for purposes of U.K. law. As a result, the income arising from the trust assets was treated as income of the U.K. trustee, which was subject to a 22% U.K. income tax under section 469 of the U.K. Income and Corporations Taxes Act 1988. That U.K. income tax paid was a liability of the U.K. trustee and not of any of the trust unit holders. The U.K. trustee owed the U.K. income tax whether the trust made actual distributions to the trust unit holders.

C. U.K. Tax Treatment of Barclays

Under U.K. law, Barclays, as a trust unit holder, was deemed to receive annual payments from the trust. Barclays owed U.K. corporation tax at a 30% rate on those deemed annual payments even if Barclays did not receive any trust distributions. The deemed annual payments were equal to the income available for distribution from the trust to Barclays as holder of the class C unit and class D unit, grossed up for 22% U.K. income tax. Barclays was entitled to a U.K. tax credit of 22% on the deemed annual payment. Barclays could also claim a U.K. deduction for contributing the class C unit distributions and for the spread amount paid to InvestCo through the zero coupon swap.

A. DelCo Distributions

DelCo held most of the STARS assets at closing. DelCo made monthly distributions to InvestCo (class 1 shareholder) and the trust (class 2 shareholder) with InvestCo receiving 1% and the trust receiving the remaining 99%. The monthly distributions to the trust were sufficient for the trust to meet the target distributions to Barclays. When DelCo's income did not meet projected DelCo distributions, DelCo satisfied the difference from its cash on hand. BNY also arranged for the contribution of more income-producing assets to DelCo.

B. Trust Distributions

The trust generated income from the trust assets and DelCo class 2 distributions. The trust set aside 22% of the trust income in reserves for U.K. taxes, which were periodically sent to the U.K. tax authority. The remaining income was distributed monthly to trust unit holders.

The trust made monthly class C unit distributions to the Barclays blocked account. Those distributions were approximately equal to the corresponding target distribution amounts. Barclays immediately contributed these distributions to the trust to satisfy its obligation to pay further subscription amounts. The trust also made the required monthly class D unit distributions to Barclays. Barclays retained all of these distributions, totaling $7 million over the term of STARS.

Finally, the trust made monthly contributions to DelCo of amounts at least equaling but often substantially exceeding the corresponding contributed income amount from the Barclays blocked account starting after the first nine months of STARS.

C. Zero Coupon Swap and Credit Default Swap Payments

InvestCo or Barclays made monthly payments as required under the zero coupon swap. LIBOR was 2.09% when STARS closed and stayed below 3% until almost mid-2005. During that period, the spread due from Barclays under the zero coupon swap was greater than the LIBOR plus 30 basis points amount due from InvestCo. Barclays made net payments to InvestCo under the zero coupon swap of $12 million for 2001 and $51 million for 2002. Over the life of STARS, Barclays made net payments to BNY of $82.6 million under the floating leg of the zero coupon swap. Additionally, Barclays made all required payments to BNY under the credit default swap.

BNY treated the payments made to Barclays on the class D unit distributions as a component of interest on the loan. With respect to the floating leg of the zero coupon swap, BNY netted the spread component and the LIBOR plus 30 basis points component of the zero coupon swap. This treatment effectively resulted in BNY claiming an interest deduction for the LIBOR plus 30 basis points interest amount (zero coupon swap interest) for the years at issue as the spread component exceeded the zero coupon swap interest component for each year. BNY reduced unrelated interest expense by the net payments Barclays made to InvestCo under the zero coupon swap for the years at issue.

BNY claimed $835,100 and $6,753,720 as deductible expenses, fees and transaction costs for 2001 and 2002, respectively.

We agree with respondent. The STARS transaction was structured to meet the relevant requirements in the Code and the regulations for claiming the disputed foreign tax credits. The STARS transaction in essence, however, was an elaborate series of pre-arranged steps designed as a subterfuge for generating, monetizing and transferring the value of foreign tax credits among the STARS participants. We now turn to the merits of the STARS transaction under the economic substance doctrine.

Taxpayers may structure business transactions in a manner that results in the least amount of tax. See Boulware v. United States, 552 U.S. 421, 430 n.7 (2008) (citing Gregory v. Helvering, 293 U.S. 465, 469 (1935)); Gerdau Macsteel, Inc. v. Commissioner, 139 T.C. ___, ___ (slip op. at 163-164) (Aug. 30, 2012). Courts have also long recognized, however, that even if a transaction complies literally with the Code, it does not necessarily follow that Congress intended to cover the transaction and allow a tax benefit. Knetsch v. United States, 364 U.S. 361, 365 (1960); Helvering v. Gregory, 69 F.2d 809, 810 (2d Cir. 1934), aff'd, 293 U.S. 465 (1935). In Frank Lyon Co. v. United States, 435 U.S. 561, 583-584 (1978), the Supreme Court explained the circumstances in which a transaction should be respected for tax purposes:

There is a split among the Courts of Appeals, however, as to the proper application of the economic substance doctrine, and alternative approaches have emerged. Some Courts of Appeals require that a valid transaction have economic substance or a non-tax business purpose. See, e.g., Horn v. Commissioner, 968 F.2d 1229, 1236-1238 (D.C. Cir. 1992), rev'g Fox v. Commissioner, T.C. Memo. 1988-570; Rice's Toyota World, Inc v. Commissioner, 752 F.2d 89, 91 (4th Cir. 1985), aff'g in part, rev'g in part 81 T.C. 184 (1983). Other Courts of Appeals require a valid transaction have both economic substance and a non-tax business purpose. See Dow Chem. Co. v. United States, 435 F.3d 594, 599 (6th Cir. 2006); Winn-Dixie Stores, Inc. v. Commissioner, 254 F.3d 1313, 1316 (11th Cir. 2001), aff'g 113 T.C. 254 (1999); United Parcel Serv. of Am., Inc. v. Commissioner, 254 F.3d 1014, 1018 (11th Cir. 2001), rev'g T.C. Memo. 1999-268. Still other Courts of Appeals adhere to the view that a lack of economic substance is sufficient to invalidate a transaction regardless of the taxpayer's subjective motivation. See, e.g., Coltec Indus., Inc. v. United States, 454 F.3d 1340, 1355 (Fed. Cir. 2006). And still other Courts of Appeals treat the objective and subjective prongs merely as factors to consider in determining whether a transaction has any practical economic effects beyond tax benefits. See, e.g., ACM P'ship v. Commissioner, 157 F.3d 231, 248 (3d Cir. 1998), aff'g in part, rev'g in part T.C Memo. 1997-115.

An appeal in this case would lie to the Court of Appeals for the Second Circuit absent stipulation to the contrary, and, accordingly, we follow the law of that circuit to the extent it is directly on point. See Golsen v. Commissioner, 54 T.C. 742 (1970), aff'd, F.2d 985 (10th Cir. 1971). The Court of Appeals for the Second Circuit has endorsed applying a flexible analysis in assessing economic substance. Gilman v. Commissioner, 933 F.2d 143 (2d Cir. 1991), aff'g T.C. Memo. 1989-684; Long Term Capital Holdings v. United States, 330 F. Supp. 2d 122 (D. Conn. 2004), aff'd, 150 Fed. Appx. 40 (2d Cir. 2005). The analysis evaluates both the subjective business purpose of the taxpayer for engaging in the transaction and the objective economic substance of the transaction. Gilman v. Commissioner, 933 F.2d at 148; Long Term Capital Holdings, 330 F. Supp. 2d at 171. These distinct aspects of the economic substance inquiry do not, however, constitute discrete prongs of a rigid two-step analysis. Long Term Capital Holdings, 330 F. Supp. 2d at 171 n.68; see also Gilman v. Commissioner, 933 F.2d at 148. They are instead simply more precise factors to consider in the overall inquiry of whether the transaction had any practical economic effects other than the creation of tax losses. Altria Grp. Inc. v. United States, 694 F. Supp. 2d 259, 282 (S.D.N.Y. 2010), aff'd, 658 F.3d 276 (2d Cir. 2011); Long Term Capital Holdings, 330 F. Supp. 2d at 171 n. 68. A finding of a lack of either economic substance or a non-tax business purpose can be but is not necessarily sufficient for a court to conclude that a transaction is invalid for Federal tax purposes. Altria Group, Inc., 694 F. Supp. 2d at 282; Long Term Capital Holdings, 330 F. Supp. 2d at 171 n.68. The ultimate determination of whether a transaction lacks economic substance is a question of fact. See Nicole Rose Corp. v. Commissioner, 320 F.3d 282, 284 (2d Cir. 2003), aff'g 117 T.C. 328 (2001). We now turn to the scope of the economic substance doctrine.

B. Scope of the Economic Substance Inquiry

The first step in the economic substance inquiry is to identify the transaction to be analyzed. See, e.g., Sala v. United States, 613 F.3d 1249, 1252 (10th Cir. 2010). Petitioner argues that we should analyze the components of the STARS transaction as an integrated arrangement for purposes of testing economic substance. In contrast, respondent argues that we should bifurcate the STARS transaction and focus on the STARS structure, not the loan, for purposes of testing economic substance. We agree with respondent.

The relevant transaction to be tested is the one that produces the disputed tax benefit, even if it is part of a larger set of transactions or steps.⁸See Nicole Rose Corp. v. Commissioner, 320 F.3d at 284; Kipnis v. Commissioner, T.C. Memo. 2012-306; Country Pine Fin., LLC v. Commissioner, T.C. Memo. 2009-251; Long Term Capital Holdings, 330 F. Supp. 2d at 183; see also Sala, 613 F.3d at 1252; Klamath Strategic Inv. Fund v. United States, 568 F.3d 537, 545 (5th Cir. 2009); Coltec Indus., Inc., 454 F.3d at 1352-1355; Black & Decker Corp. v. United States, 436 F.3d 431, 436 (4th Cir. 2006); ACM P'ship v. Commissioner, 157 F.3d at 260 n.57. Stated another way, the requirements of the economic substance doctrine are not avoided simply by coupling a routine transaction with a transaction lacking economic substance. See, e.g., Long Term Capital Holdings, 330 F. Supp. 2d at 183; see also ACM P'ship v. Commissioner, 157 F.3d at 260 n.57. A contrary application would undermine the flexibility and efficacy of the economic substance doctrine.

Accordingly, we focus our economic substance inquiry on the transaction that gave rise to the disputed foreign tax credits. The disputed foreign tax credits were generated by circulating income through the STARS structure. In contrast, the loan was not necessary for the STARS structure to produce the disputed foreign tax credits. It is the use of the STARS structure then that is relevant and that we test for economic substance.

C. Economic Substance of the STARS Structure

The record reflects that BNY did not have a reasonable expectation that it would make a non-tax economic profit from using the STARS structure. First, the STARS structure did not increase the profitability of the STARS assets in anyway. To the contrary, it reduced their profitability by adding substantial transaction costs, e.g., professional service fees and foreign taxes incurred as result of using the STARS structure.⁹

Additionally, the activities or transactions that the STARS structure was used to engage in did not provide a reasonable opportunity for economic profit. The STARS structure's main activity was to circulate income between itself and Barclays. Every month, as pre-arranged, DelCo would transfer pre-determined amounts of income to the trust. Substantially all of the trust income was distributed to the Barclays blocked account, which in turn was immediately recontributed to the trust and then passed back to DelCo where it was available for BNY's use. These circular cashflows or offsetting payments had no non-tax economic effect.

Courts have consistently recognized that the presence of circular cashflows strongly indicates that a transaction lacks economic substance. See Altria Group, Inc., 658 F.3d at 289 (citing AWG Leasing Trust v. United States, 592 F. Supp. 2d 953, 983 (N.D. Ohio 2008)) (circular payments from and back to foreign bank "strongly indicate" that SILO transaction "has little substantive business purpose other than generating tax benefits"); Merryman v. Commissioner, 873 F.2d 879, 882 (5th Cir. 1989) (tax structuring disregarded where "money flowed back and forth but the economic positions of the parties were not altered"), aff'g T.C. Memo. 1988-72; Prof'l Servs. v. Commissioner, 79 T.C. 888, 928 (1982) (disregarding prearranged circular cashflows through a trust); see also Knetsch, 364 U.S. at 366 (offsetting payments on annuity bond and notes resulted in sham). This follows from the common sense proposition that a taxpayer is not entitled to benefits from circular transfers the net result of which is effectively nothing.

The STARS structure was also used in connection with the stripping transaction. The stripping transaction too resulted in a circular cashflow and did not provide a reasonable opportunity for economic profit. In particular, the trust sold its right to interest income from the trust collateral securities to DelCo for a lump-sum payment taxable in the United Kingdom, which DelCo made with funds provided by BNY. This reallocated the income and principal payments associated with the trust collateral securities within the STARS structure. It did not alter the amount and timing of the cashflows generated by the underlying assets. And because the sale of the interest rights was funded by BNY and between entities within the STARS structure, the stripping transaction had no potential to generate a non-tax economic profit on the aggregate.

Petitioner argues that we should consider the income generated by the STARS assets in evaluating whether the STARS structure had a reasonable opportunity for economic profit. We disagree. Economic benefits that would result independent of a transaction do not constitute a non-tax benefit for purposes of testing its economic substance. See Gerdau Macsteel, Inc. v. Commissioner, 139 T.C. at ___ (slip op. at 174-175). Stated otherwise, benefits that are unrelated to the transaction cannot be what motivates a taxpayer to engage in the transaction and therefore are of no aid in determining whether the taxpayer would have engaged in the transaction absent the tax effects. Id.

Here, BNY's control and management over the STARS assets did not materially change as a result of their transfer to the STARS structure.¹⁰ Additionally, the STARS structure had no effect on the income stream generated by the STARS assets. Accordingly, the STARS assets would have generated the same income regardless of being transferred to the trust. Thus, income from the STARS assets was not an incremental benefit of STARS.

BNY's obligation with respect to the loan was more than adequately secured by other arrangements independent of the trust. Barclays held a security interest in a pool of high quality assets valued at $2.25 billion, creating a collateralization level of 150%. Respondent's expert Steven Schwarcz concluded that the collateralization level (e.g., securitization) in a structured finance transaction is usually around 10% and that the loan was substantially over collateralized. In addition to the collateral arrangements, Barclays effectively had full recourse to BNY itself for repayment through the credit default swap.

Petitioner's expert W. Clifford Atherton suggested that the special-purpose entities making up the STARS structure served a project financing (a type of structured financing transaction) function. We disagree. Respondent's expert Mr. Schwarcz emphasized that special-purpose entities are typically used in connection with a structured financing transaction to efficiently reallocate risk and reduce information asymmetry.¹² Mr. Schwarcz also highlighted that structured financing transactions generally involve special-purpose entities incurring debt and using the proceeds to finance the acquisition of income-producing assets. And the lenders look to the cash produced by those assets for repayment, bearing the risk that the cash will be insufficient to repay the debt.

These common indicia of a structured financing transaction are not present in STARS. The loan proceeds were not used to purchase the STARS assets, and Barclays did not look to any assets purchased with the financing proceeds for repayment. And unlike a typical structured financing transaction, the special-purpose entities in STARS did not function to efficiently reallocate risk.

In this regard, Mr. Schwarcz observed that STARS simply involved a full-recourse secured financing. Mr. Schwarcz correctly concluded that, given the characteristics of the loan, Barclays could have made the same $1.5 billion loan to BNY, secured by the same assets constituting the collateral for the loan, using only a loan agreement and a security agreement. Such an arrangement would have been much simpler, avoided the use of the special-purpose entities and had substantially lower transaction costs than STARS.

Efficiency aside, Mr. Schwarcz concluded that the special-purpose entities used in the STARS transaction did not "realistically" function to transfer risk between the parties. Mr. Schwarcz opined that the over-collateralization level of the loan and the other security arrangements minimized Barclays' risk with respect to the loan. Accordingly, there was no significant risk for the special-purpose entities to transfer.

Finally, Mr. Schwarcz concluded that the STARS structure did not reduce information asymmetry between the parties. In contrast, he opined that STARS was excessively complex given the economics of the loan and arguably increased information asymmetry. We agree with Mr. Schwarcz that the STARS structure did not perform a structured financing function.

Petitioner finally argues more generally that Barclays made the loan contingent on the STARS structure and therefore the two transactions were "commercially linked." Again, we are not persuaded. Making a routine business transaction contingent on an economically meaningless transaction, like the STARS structure, is insufficient to establish that the nexus between the two is reasonable. See, e.g., Long Term Capital Holdings, 330 F. Supp. 2d at 183.

In sum, the record does not support that the STARS structure performed any significant banking, commercial or business function with respect to the loan. Consequently, we find that the STARS structure did not bear a reasonable relationship to the loan. This lack of reasonableness indicates BNY's true motivation -- tax avoidance.

We are mindful in evaluating the substance of the spread that labels and characterizations do not determine the tax consequences where they are inconsistent with economic realities. Frank Lyon Co., 435 U.S. at 583-584 (labels must be economically meaningful); TIFD III-E, Inc. v. United States, 459 F.3d 220 (2d Cir. 2006); Saba P'ship v. Commissioner, T.C. Memo. 2003-31 (payments characterized as consulting fees held to be a guaranteed return to a purported partner).

The stated interest rate on the loan was LIBOR plus 30 basis points less the spread. The spread was a fixed amount equal to one-half the present value of the U.K. taxes the trust was expected to pay on the target class C unit income each month. We acknowledge the spread was part of the formula for calculating the interest expense on the loan. Its substance did not match, however, its form.

Respondent's expert Anthony Saunders opined on the commerciality of the loan's pricing. Mr. Saunders noted that the pricing of a loan generally depends on the time value of money and the risks presented to the lender through the particular loan transaction. Mr. Saunders also noted that here the loan's cost was such that Barclays could not reasonably expect that the return (i.e., interest) it received from BNY would exceed Barclays' cost of funds. He further noted that, independent of Barclays' cost of funds, the interest rate was "negative" for most of the tenure of the loan. That is, the "lender" (Barclays) was paying the "borrower" (BNY) to borrow its funds. Mr. Saunders concluded that the loan's pricing did not reflect the risk inherent in the STARS transaction and more generally that the loan fundamentally deviated from attributes of a standard banking transaction. He further concluded that there were no unique economic conditions that might explain the non-economic pricing.

Respondent's expert Mr. Schwarcz also opined on the commerciality of the loan. Like Mr. Saunders, Mr. Schwarcz noted that the loan had a "negative interest rate." Mr. Schwarcz opined that, in an arm's-length commercial lending transaction, a loan would not bear a negative interest rate, absent unique circumstances external to the loan, e.g., to avoid a loss or to effect government policy, such as stimulus. He noted that it makes no economic sense for a lender to pay a borrower interest on a loan absent such a circumstance. He concluded there were no special circumstances that warranted the loan bearing a "negative interest rate" and the loan was not commercially reasonable.

Respondent's expert Michael Cragg analyzed the pricing of the loan, including the economics of the spread. He concluded that circulating income through the STARS structure generated the economic benefit labeled the "spread" by combining certain U.S. and U.K. tax effects. Mr. Cragg's analysis showed that it would not have been economically beneficial for Barclays to pay BNY the spread absent the U.K. tax benefits from STARS. Similarly, Mr. Cragg's analysis showed that the STARS arrangement would not have been beneficial to BNY absent the foreign tax credits arising from the payment of U.K. tax on the trust income. Mr. Cragg ultimately concluded that the spread was economically derived and contingent on the parties receiving certain U.S. and U.K. tax treatment with respect to the STARS structure and as a result was not a pre-tax cashflow.

Petitioner denies that the spread was a tax effect because it was not expressly contingent on either Barclays or BNY receiving any particular U.S. or U.K. tax treatment or benefit. In this regard, petitioner asserts that BNY could under certain circumstances keep the spread Barclays paid even if Barclays did not realize its expected U.K. benefits.¹³ And petitioner asserts that Barclays' obligation to pay the spread did not vary depending on whether BNY's U.S. tax treatment was respected.

Petitioner's argument is unpersuasive. The manner in which the parties agreed to allocate tax risk does not preclude the spread from being a tax effect. The spread's value was derivative of expected U.S. and U.K. tax effects. And it would not have been paid going forward if either of those effects had been foreclosed. Indeed, STARS would no longer be economically beneficial to either BNY or Barclays and each could terminate STARS on short notice.¹⁴

In sum, we agree with respondent's experts. The spread artificially reduced the loan's cost and lacked economic reality. In substance the spread was contingent on the parties' anticipated tax treatment and was unrelated to the time value of money or the attendant risks associated with the loan. We conclude, on the record as a whole, that the spread was in substance not a component of loan interest.

The spread rather was a tax effect. It was embedded in the loan to serve as a device for monetizing and transferring the value of anticipated foreign tax credits generated from routing income through the STARS structure. That the generated tax savings were used to offset the cost of the loan does not provide a valid non-tax purpose. Indeed, courts have consistently recognized that intending to use tax savings from a transaction lacking economic substance to underwrite or enhance the commercial terms of a legitimate business transaction does not constitute a valid non-tax purpose. See Winn-Dixie Stores, Inc. v. Commissioner, 113 T.C. at 287; see also Am. Elec. Power Co., Inc. v. United States, 326 F.3d 737, 744 (6th Cir. 2003) ("Money generated by means of abusive tax deductions can always be applied to beneficial causes, but the eventual use of the money thus generated is not part of the economic sham analysis.").

D. Economic Substance of the Integrated STARS Arrangement

The STARS transaction still lacks economic substance even if the STARS structure and the loan are evaluated as an integrated transaction. Petitioner contends that the integrated STARS transaction has objective economic substance because it offered a reasonable opportunity for pre-tax profit. Petitioner asked its expert Mr. Atherton to calculate the pre-tax profitability of the STARS transaction. Mr. Atherton concluded that BNY reasonably could have expected a profit of more than $1.6 billion before taking into account U.K. or U.S. income taxes over the life of the STARS transaction.

We find that Mr. Atherton's analysis of STARS' pre-tax profitability contains several critical flaws and is therefore not helpful to the Court. One such flaw with Mr. Atherton's pre-tax profitability calculation is that he includes income from the STARS assets as revenues arising from the STARS transaction. As we previously held, the pre-existing cashflows from the trust assets are not incremental to the STARS transaction and therefore irrelevant to the objective economic substance analysis.

Mr. Atherton's pre-tax profitability analysis is also flawed because he includes returns on asset-backed securities he assumes BNY contemplated acquiring with the loan proceeds. Only cashflows arising from the transaction whose economic substance is at issue are relevant to the pre-tax profitability analysis. See Nicole Rose Corp. v. Commissioner, 320 F.3d at 284 (rejecting the taxpayer's argument that profits from an unrelated asset sale should be attributed to a lease transaction generating the tax benefits at issue); ACM P'ship v. Commissioner, 157 F.3d at 260 (disregarding profits from funds acquired in a transaction and invested outside of the structure being evaluated for economic substance); see also Kipnis v. Commissioner, T.C. Memo. 2012-306 (economic substance should be reviewed without reference to expected profit from an intended real estate investment that the taxpayer expected to make with proceeds from the "CARDS" transaction); Long Term Capital Holdings, 330 F. Supp. 2d at 183 (requirements of economic substance are not avoided by coupling a routine profitable economic transaction with no inherent tax benefits to a unique transaction that otherwise lacks profit potential).

Here, the integrated STARS transaction's net pre-tax effect was to create a $1.5 billion loan at LIBOR plus 20 basis points. It did not generate any revenue, only an obligation to repay the loan principal and interest. Any income from investing the loan proceeds was not a cashflow arising from the integrated STARS transaction. Rather, it resulted from a separate and distinct transaction. Thus, income from investing the loan proceeds is not relevant to the economic substance analysis of the integrated STARS transaction and should have been excluded from the pre-tax profitability analysis. We note that even if the projected yield on the loan proceeds Mr. Atherton assumed was relevant that yield is insufficient to offset the foreign tax costs¹⁶ of the transaction.

The last critical flaw in his analysis is his including the spread in calculating the cost of the loan, the effect of which is to reduce the cost. As we previously found, the spread is a tax effect of the STARS structure and its value is effectively funded by the foreign tax credits. Mr. Atherton therefore should not have reduced the cost of the loan by the spread in his pre-tax profitability analysis.

Mr. Atherton's analysis substantially inflates pre-tax income by including the non-incremental income from the STARS assets, the projected yield from the loan proceeds and the spread as pre-tax income. When these items are omitted, all that remains is the loan at LIBOR plus 20 basis points. As we previously, discussed, Mr. Cragg's analysis shows that the loan was overpriced and therefore not profitable on a pre-tax basis. Mr. Cragg concluded more generally that it would have been economically irrational for BNY to enter into the integrated STARS transaction without the foreign tax credits BNY derived from it. Accordingly, we find that the integrated STARS transaction lacks economic substance.

We now address the subjective economic substance of the integrated STARS transaction. Here, petitioner argues that it was motivated to enter the STARS transaction to obtain "low cost" financing. As we previously held, we reject that business purpose because it lacks merit. Aside from that claimed business purpose, petitioner contends it was motivated to enter into STARS by a realistic expectation of pre-tax profit. Specifically, petitioner claims that BNY intended to use the loan proceeds to grow its "investment portfolio" and earn a profit by investing in asset-backed securities. As we previously held, any income from the investment of the loan proceeds is not income from the integrated STARS transaction and therefore is irrelevant to the objective economic substance analysis. Similarly, any profit petitioner expected to earn from investing the loan proceeds is not relevant to the subjective economic substance analysis.¹⁷

E. Congressional Intent

We now consider whether the disputed tax benefits are what Congress intended in establishing the foreign tax credit. Petitioner contends that the economic substance doctrine does not warrant disallowing the disputed tax benefits because Congress intended the foreign tax credit for transactions like STARS. We disagree.

The United States taxes income of its citizens, residents and domestic entities on a worldwide basis. A U.S. corporation must include foreign source income in its U.S. taxable income even though that income may also be subject to foreign tax. Congress enacted the foreign tax credit to alleviate double taxation arising from foreign business operations. See United States v. Goodyear Tire & Rubber Co., 493 U.S. 132, 139 (1989); Am. Chicle Co. v. United States, 316 U.S. 450, 451 (1942); Burnet v. Chicago Portrait Co., 285 U.S. 1, 7 (1932). Congress intended the foreign tax credit to neutralize the effect of U.S. tax on the business decision of where to conduct business activities most productively. 56 Cong. Rec. App. 677-678 (1918) (statement of Rep. Kitchin). The enactment of the foreign tax credit was also informed by fairness. See National Foreign Trade Council, Inc., International Tax Policy for the 21st Century, (Dec. 15, 2001).

The STARS transaction was a complicated scheme centered around arbitraging domestic and foreign tax law inconsistencies. The U.K. taxes at issue did not arise from any substantive foreign activity. Indeed, they were produced through pre-arranged circular flows from assets held, controlled and managed within the United States. We conclude that Congress did not intend to provide foreign tax credits for transactions such as STARS.

Expenses incurred in furtherance of a transaction that is disregarded for a lack of economic substance are not deductible. See Winn-Dixie Stores, Inc. v. Commissioner, 113 T.C. at 294 (observing that "a transaction that lacks economic substance is not recognized for Federal tax purposes" and that "denial of recognition means that such a transaction cannot be the basis for a deductible expense"); see also Gerdau Macsteel, Inc. v. Commissioner, 139 T.C. ___ (slip op. at 188). The claimed transactional expenses, the zero coupon swap interest expense and the U.K. taxes were all incurred in furtherance of the STARS transaction, which we previously held lacks economic substance. Consequently, they are not deductible.

Petitioner contends that the resourcing provision applies and that respondent's foreign source income adjustment was improper. We disagree. U.S. tax laws and treaties do not recognize sham transactions or transactions that have no economic substance as valid for tax purposes. Del Commercial Props., Inc. v. Commissioner, T.C. Memo. 1999-411 (citing Gregory v. Helvering, 293 U.S. 465, 470 (1935), and Johansson v. United States, 336 F.2d 809, 813 (5th Cir. 1964)), aff'd, 251 F.3d 210 (D.C. Cir. 2001). Because we previously held that the STARS transaction is disregarded for U.S. tax purposes, BNY is treated for U.S. tax purposes as owning the STARS assets and the income is treated as being derived by BNY within the United States. Consequently, the U.S.-U.K. tax treaty, including the resourcing provision, does not apply. We therefore sustain respondent's adjustment of petitioner's foreign source income.

We have considered all remaining arguments the parties made and, to the extent not addressed, we find them to be irrelevant, moot or meritless.

To reflect the foregoing,

Decision will be entered for respondent.

² All monetary amounts have been rounded to the nearest million unless otherwise indicated.

³ All section references are to the Internal Revenue Code (Code) for the years at issue, unless otherwise indicated.

⁴ There is also a question of whether respondent properly adjusted interest expenses allocated to the foreign source income. We need not address this issue because of our holding that the trust income reported as foreign source income is U.S. source income.

⁵ "LIBOR" is an acronym for "London Interbank Offering Rate." See generally Bank One Corp. v. Commissioner, 120 T.C. 174, 189 (2003), aff'd in part, vacated in part and remanded sub nom. J.P. Morgan Chase & Co. v. Commissioner, 458 F.3d 564 (7th Cir. 2006).

⁶ For simplicity, we net the zero coupon swap floating leg, the credit default swap payment and the class D unit distributions in referring to the interest rate (LIBOR plus 20 basis points) on the loan.

⁷ Respondent also argues that the foreign tax credits BNY claimed are disallowed under substance over form doctrines (including the step transaction doctrine) and under the statutory anti-abuse rule in sec. 269(a). We need not decide these arguments because of our other holdings.

⁸ Congress noted when codifying the economic substance doctrine in sec. 7701 in 2010 that under present law courts could "bifurcate a transaction in which independent activities with non-tax objectives are combined with an unrelated item having only tax avoidance objectives to disallow those tax motivated benefits." Staff of Jt. Comm. on Taxation, Technical Explanation of the Revenue Provisions of the "Reconcilliaton Act of 2010" as amended, in combination with the "Patient Protection and Affordable Care Act" 153 & n.352 (J. Comm. Print 2010).

⁹ We have previously held that foreign taxes are economic costs for purposes of the economic substance doctrine. See Compaq Computer Corp. v. Commissioner, 113 T.C. 214 (1999), rev'd, 277 F.3d 778, 785 (5th Cir. 2001). We are mindful that the Courts of Appeals for the Fifth and Eighth Circuits have subsequently held that foreign taxes should not be taken into account in evaluating pre-tax effects for purposes of the economic substance analysis. See IES Indus., Inc. v. United States, 253 F.3d 350 (8th Cir. 2001); Compaq Computer Corp. v. Commissioner, 277 F.3d 778, 785 (5th Cir. 2001), rev'g 113 T.C. 214 (1999). Nevertheless, the Supreme Court and the Court of Appeals for the Second Circuit have yet to consider the issue, and we are not bound by Fifth and Eighth Circuit precedent here.

We maintain the position we took in Compaq Computer with respect to foreign taxes in the economic substance context. Economically, foreign taxes are the same as any other transaction cost. And we cannot find any conclusive reason for treating them differently here, especially because substantially all of the foreign taxes giving rise to the foreign tax credits stemmed from economically meaningless activity, i.e., the pre-arranged circular cashflows engaged in by the trust.

Additionally, excluding the economic effect of foreign taxes from the pre-tax analysis would fundamentally undermine the point of the economic substance inquiry. That point is to remove the challenged tax benefit and evaluate whether the relevant transaction makes economic sense. See In re CM Holdings, Inc., 301 F.3d 96, 105 (3d Cir. 2002).

¹⁰ DelCo held most of the income-generating STARS assets with the trust holding the remaining STARS assets. BNY directly or indirectly held all the voting rights of DelCo, the initial and successor trustee of the trust and the trust manager, and thus effectively controlled those entities. In addition, BNY executed servicing agreements that gave BNY control over the management of the STARS assets the trust and DelCo held.

¹¹ Petitioner's experts opined on several other potential business purposes at trial. The record does not support, however, that BNY contemplated those suggested business purposes at the time it participated in STARS. We therefore reject these after-the-fact rationalizations. See, e.g., Winn-Dixie Stores, Inc. v. Commissioner, 113 T.C. 254, 285-286 (1999), aff'd, 254 F.3d 1313 (11th Cir. 2001).

¹² Mr. Schwarcz defined "information asymmetry" as a scenario in which one party has more information than the other party. According to Mr. Schwarcz, structured financing transactions reduce information asymmetry by allowing parties taking on risk to more efficiently assess that risk, typically by creating well-defined, easily-valued and bankruptcy-protected sources of repayment.

¹³ If the U.K. tax authority determined that the trust was not a collective investment scheme before the due date of the U.K. trust tax, BNY could keep the spread paid to date even though Barclays would not realize the anticipated U.K. tax benefits. After the first due date of U.K. tax on trust income, however, BNY was obligated to pay Barclays half of any STARS trust tax that would be refunded if the U.K. tax authority did not respect Barclays' tax position with respect to the STARS structure. That amount would be roughly equal to spread payments BNY had received.

We note that the risk of having to pay the spread before the first due date without realizing the anticipated U.K. benefits was likely minimal. According to petitioner's U.K. regulatory expert, Michael Brindle, Q.C., the U.K. tax authority would likely view the STARS structure as a collective investment scheme. In addition, the STARS structure was promptly submitted to the U.K. tax authority for approval and a similar transaction was already under review. And on net the STARS transaction added revenue to the U.K. relative to its position without the STARS transaction, increasing the likelihood that the U.K. tax authority would view the STARS structure favorably.

¹⁴ Petitioner concedes that Barclays agreed to pay the spread based upon expected U.K. tax benefits and that the spread was calculated as a percentage of the present value of those benefits. Those U.K. tax benefits depended on the vitality of the trust structure whose economic rationality for BNY depended on BNY receiving a U.S. foreign tax credit for U.K. tax paid on trust income as the spread equaled only half of the U.K. tax. Accordingly, Barclays' U.K. tax benefit could not be achieved without BNY achieving its U.S. tax benefit.

¹⁵ We note that, regardless of how the spread is characterized, the benefit of the spread was more than offset by the additional transaction costs that BNY incurred to obtain the spread.

¹⁶See supra note 9.

¹⁷ We note that petitioner failed to substantiate the claimed business purpose otherwise. None of the STARS transactional documents or any other persuasive contemporaneous evidence show that BNY considered investing the loan proceeds in asset-backed securities. Nor did BNY consider any projected returns from such an investment in evaluating whether to enter into STARS. And the record does not reflect that loan proceeds were in fact used to purchase such securities.