MERCK & CO., INC. v. UNITED STATES OF AMERICA, No. 2:2005cv02575 - Document 196 (D.N.J. 2009)

Court Description: OPINION filed. Signed by Judge Katharine S. Hayden on 8/28/09. (rg, )

Download PDF

UNITED STATES DISTRICT COURT DISTRICT OF NEW JERSEY SCHERING-PLOUGH CORPORATION, Civ. Action No. 05-2575 (KSH) Plaintiff, v. OPINION UNITED STATES OF AMERICA, Defendant. Katharine S. Hayden, U.S.D.J. I. INTRODUCTION In this taxpayer refund action, the Court examines a domestic corporation s assignment of future income streams derived from interest rate swaps with a third party to its offshore subsidiaries, in exchange for lump-sum payments from the subsidiaries. In so doing, the Court must decide whether the structured transactions were in essence a loan from, or a sale to, the subsidiaries. Should it find the former, the Court must uphold as valid the tax levied by the government. Should it conclude the latter, plaintiff Schering-Plough Corporation ( ScheringPlough ) is entitled to a $473 million refund, plus interest. Before discussing the facts of the case more closely, the Court briefly describes the critical loan/sale distinction underlying the dispute, the operation of the relevant taxation policy, and the transactions at issue. Generally, the earnings of a domestic corporation s foreign subsidiaries are not taxed until the money is distributed to the parent corporation via a dividend. 1 See I.R.C. § 451(a).1 Under the international taxation scheme in effect at all relevant times here, however, an intercompany loan from an offshore subsidiary to its domestic parent is immediately taxable. This is a departure from the traditional rule that a loan is not income. See James v. United States, 366 U.S. 213, 219 (1961) (accepted definition of gross income excludes loans ). Instead, Congress has enacted its policy judgment, discussed more fully below, that when a foreign subsidiary invests in the corporate debt of its domestic parent, the use of the loaned funds is no different than a dividend to the parent shareholder, and should be taxed accordingly. This regime, known as Subpart F of the tax code, is intended to prevent United States corporations from sheltering their subsidiaries income in so-called tax-haven countries, while simultaneously putting the money to domestic use. Conversely, the sale of future income rights under an interest rate swap transaction triggers a different principle of taxation. The federal tax code mandates that when a taxpayer s method of accounting does not clearly reflect the income the taxpayer actually generates, the method of computing taxable income as prescribed by the Commissioner of the Internal Revenue Service ( IRS )2 should nonetheless be reflective of such income. See I.R.C. § 446(b); United States v. Hughes Props., Inc., 476 U.S. 593, 603 (1986). The Commissioner determined in Notice 89-21, a revenue notice published on February 7, 1989, that money received from the sale of rights to future income streams under an interest rate swap transaction did not clearly reflect income if it was reported in the year received. Instead, the notice stated 1 2 Citations to Title 26 of the United States Code (the Internal Revenue Code) will appear in this opinion as I.R.C. The Court s references to the Commissioner and the IRS herein are interchangeable. 2 that income is more accurately reflected when it is reported as having been received over the lifetime of the swap contract.3 Turning to the transactions at issue: in 1991 and 1992, Schering-Plough, an international pharmaceutical conglomerate, wishing to repatriate its subsidiaries foreign earnings back to the United States, entered into two 20-year interest rate swap transactions with Algemene Bank Nederland, N.V. ( ABN ), a Dutch investment bank. Under the swaps, the two counterparties agreed to exchange periodic interest payments based on a hypothetical amount (the notional principal ) and two different interest rate indices. The swap agreements obligated ScheringPlough and ABN to make periodic payments to each other reflecting the movement of the particular interest rate assigned to their respective sides of the transaction. Under the swaps, Schering-Plough had the right to assign or otherwise transfer its right to receive interest payments from ABN (the receive legs ). It did, in fact, assign the majority of the receive legs to two of its foreign subsidiaries. In return, the subsidiaries made lump-sum payments to Schering-Plough totaling approximately $690 million. Schering-Plough did not report the lump sums as present income. Instead, it deferred reporting income until later years, relying on Notice 89-21. Specifically, because Notice 89-21 required ratable taxation of payments received in exchange for the assignment of future income streams from notional principal contracts, Schering-Plough reported income for the lump sums by amortizing them over the period in which the future income streams had been assigned.4 3 The applicable text of Notice 89-21, entitled Deferred Recognition of Income from Lump-Sum Payments in Connection with Notional Principal Contracts, appears in Section IV.C below. 4 Throughout this litigation, the transactions have been given various names. The transactions were originally proposed and are now referred to by the government as Strippable Increasing Principal Swaps, or STRIPS transactions. Schering-Plough refers to the transactions as swap-and-assignment transactions. For simplicity, the 3 Notice 89-21 also states that [n]o inference should be drawn from this notice as to the proper treatment of transactions that are not properly characterized as notional principal contracts, for instance, to the extent that such transactions are in substance properly characterized as loans. So if the Schering-Plough transactions are deemed loans (as the government eventually deemed them to be), the amortization provision does not apply and the entire lumpsum payments are immediately taxable. In 2004, characterizing the transactions as loans, the Commissioner assessed a tax deficiency upon Schering-Plough because it had not reported the lump-sum payments as present income in 1991 and 1992, the years in which they had been received. Schering-Plough paid the $473 million tax bill and thereafter filed this action seeking a refund. The Court s decision requires an examination of the transactions for their economic reality that is, regardless of how a given transaction was characterized by the taxpayer, is it in reality a loan or is it in reality a sale? Put another way, the Court scrutinizes for substance over form. The Court then tests the economic substance of the transaction: Does it have sufficient economic substance despite the existence of tax avoidance objectives, or is it a sham transaction ? Finally, the Court must assess whether the transactions, anchored as they were in Notice 89-21, duly comported with the relevant taxation scheme implemented by Congress. For Schering-Plough to prevail, the Court must find the following: (1) that the transactions were the economic equivalent of sales of future income streams (that is, they were not loans dressed up as sales); (2) that Schering-Plough entered into them with objectives beyond tax avoidance and that its net economic position was appreciably altered as a result (that is, that Court will modify Schering-Plough s terminology and hereafter refer to the transactions as swap-and-assign transactions. 4 the transactions were not shams); and (3) that the tax shelter that Schering-Plough alleges Notice 89-21 provides is consistent with Congress s legislative intent. Should Schering-Plough falter on any of these grounds, the Court must render judgment for the government. Because the Court conducted a bench trial during which the parties took full opportunity to present their respective positions, there is plenty of evidence, factual and opinion, to examine in making the ultimate decision. Throughout this opinion the Court will be citing the testimony of: (a) fact witnesses the people who made the decisions and signed the documents to determine what they did and why they did it; and (b) the opinions of experts called by the parties both their reports and their testimony for guidance about the best reasoned, least strained interpretations of the facts. Pursuant to Rule 52(a) of the Federal Rules of Civil Procedure, the Court has organized this opinion around a full discussion of the facts surrounding the transactions, based on the evidence adduced at trial. Following its factual findings, the Court examines the transactions economic reality, economic substance, and, finally, their assimilation with applicable tax laws. II. JURISDICTION The Court has jurisdiction concurrent with the United States Court of Federal Claims to hear all actions against the United States seeking the recovery of any internal-revenue tax alleged to have been erroneously or illegally assessed or collected . . . . 28 U.S.C. § 1346(a); I.R.C. § 7422. Venue is proper in this District pursuant to 28 U.S.C. § 1402(a)(2) because Schering-Plough s principal place of business is located in New Jersey. 5 III. STANDARD OF REVIEW A tax assessed by the IRS is presumed to be correct; a taxpayer challenging the assessment shoulders the burden of proving that the assessment is legally erroneous.5 Welch v. Helvering, 290 U.S. 111, 115 (1933); Francisco v. United States, 267 F.3d 303, 319 (3d Cir. 2001); Sullivan v. United States, 618 F.2d 1001, 1009 (3d Cir. 1980); Psaty v. United States, 442 F.2d 1154, 1159 (3d Cir. 1971). The Court s review of the IRS determination is de novo. Lewis v. Reynolds, 284 U.S. 281, 283 (1932). IV. FACTS Schering-Plough seeks a tax refund of $472,870,042.69 in paid federal income taxes for taxable years 1989, 1991, and 1992. The taxes were assessed and collected by the IRS as a result of the two interest rate swap-and-assign transactions that Schering-Plough entered into with its foreign subsidiaries and ABN in 1991 and 1992. Before turning to the specifics of the transactions, the Court describes the applicable taxation scheme, Schering-Plough s corporate structure, and the financial condition in which it found itself in the late 1980s. A. Subpart F Taxation The independent taxing identity of corporations and the worldwide taxation of income before President John F. Kennedy assumed office led to creative tax planning structures that enabled corporations to shelter revenue in subsidiaries headquartered outside the United States. Office of Tax Policy, Dep t of the Treasury, The Deferral of Income Earned Through U.S. 5 For audits conducted after July 22, 1998, the burden of proof regarding factual issues may be shifted to the government upon presentation of credible evidence by the taxpayer. See 26 U.S.C. § 7491; Neonatology Assocs., P.A. v. Comm r, 299 F.3d 221, 228 n.6 (3d Cir. 2002). Schering-Plough has not argued that the burden-shifting regime applies here, and in any event the statute has real significance only in the rare event of an evidentiary tie. Blodgett v. Comm r, 394 F.3d 1030, 1039 (8th Cir. 2005). Because that is not the case here, the standard burden of proof continues to rest with the taxpayer. 6 Controlled Foreign Corporations: A Policy Study, at x, 1-5 (2000), available at www.ustreas.gov/offices/tax-policy/library/subpartf.pdf (last visited Aug. 28, 2009). In response, Congress passed the Revenue Act of 1962, Pub. L. No. 87-834, 76 Stat. 1006, § 12(a) (1962), which was designed to prevent United States corporations from stockpiling their foreign subsidiaries earnings and profits ( E&P ) offshore in an effort to avoid domestic income tax. The objective was to end artificial arrangements between related corporations that exploit the multiplicity of foreign tax systems and international agreements in order to reduce sharply or eliminate completely their tax liabilities both at home and abroad. Id. at 13 (quoting Message from the President of the United States Relative to our Federal Tax System, H.R. Doc. No. 140, 87th Cong., 1st Sess. 6 (1961)). The legislation Subpart F of the Internal Revenue Code mandates taxation of foreign E&P upon repatriation to the United States. Mechanically, Subpart F assesses a tax on any United States shareholder (as defined in I.R.C. §§ 951(b) & 958(b)) of a controlled foreign corporation ( CFC ) (as defined in I.R.C. §§ 957 & 958) when the United States shareholder invests previously untaxed foreign E&P in United States property. An obligation by a United States shareholder acquired by a CFC is deemed to be such an investment in United States property under I.R.C. § 956(c)(1)(C). When a CFC makes a loan to its domestic parent, the amount of the loan is presently taxable under Subpart F. (Joint Trial Stipulation ( JTS ) 12, ¶¶ 100-01); see also Ludwig v. Comm r, 68 T.C. 979, 983-84 (1977) ( [I]f a controlled foreign corporation makes a loan to its shareholder, a United States person, the obligation to repay the loan is United States property and the shareholder thereby realizes income under section 951. ). B. Schering-Plough s Corporate Structure During the tax years in question, Schering-Plough was a New Jersey corporation, which, through its wholly-owned subsidiary Schering Corporation (a New Jersey corporation), owned 7 all of the voting stock of Schering-Plough International, Inc. ( International ) (a Delaware corporation). International owned a majority of the voting stock of Schering-Plough Ltd. ( Limited ), which, in turn, owned a majority share in Scherico, Ltd. ( Scherico ). Finally, Scherico was the majority owner of Essex Chemie, A.G. ( Essex Chemie ). Limited, Scherico, and Essex Chemie are Swiss corporations (collectively, the Swiss subsidiaries ). (JTS 11, ¶¶ 73-85.) Due to a favorable Irish corporate income tax on manufacturers at the time, Limited, like many other pharmaceutical companies, conducted significant manufacturing operations in Ireland. (JTS 11-12, ¶¶ 80, 81, 88; Nichols Dep. 90:10-23; Ex. 2000.) Given the above, Schering-Plough (through its domestic subsidiaries) was a United States shareholder within the meaning of I.R.C. §§ 951(b) and 958(b). (JTS 12, ¶¶ 95-96.) Moreover, there is no dispute that the Swiss subsidiaries were CFCs within the meaning of I.R.C. §§ 957 and 958. (JTS 12, ¶ 94.) The taxation format implemented by Subpart F therefore applied in all relevant respects during the tax years at issue. The sole question is whether the particular transactions at issue were subject to immediate taxation as investments in United States property (i.e., loans) under the Subpart F regime. C. Background At trial, Schering-Plough presented testimony from some of its senior management working at the company during the relevant time periods, including Schering-Plough s Chief Executive Officer Robert Luciano, and finance executive Dan Nichols, who headed the company s tax department. These senior managers testified about Schering-Plough s hierarchy and general objectives during the late 1980s and early 1990s. Also testifying was Jay Ludwig, who served in Schering-Plough s treasury department and worked on the team that designed the swap-and-assign transactions. 8 During the late 1980s, Schering-Plough s Swiss subsidiaries (particularly Limited) generated and held substantial amounts of untaxed E&P from their operations in Ireland. (JTS 11, ¶¶ 88-89; 1/16/08 Nichols Test. 74:25-75:9.) By the end of 1990, Limited had accumulated $391.5 million of this Irish cash, of which only $41 million was previously taxed income ( PTI ). Under the tax code, investments in United States property are not taxed to the extent that the funds used have already been subjected to Subpart F taxation. See I.R.C. § 959(a), (c)). By 1991, the earnings figure had grown to $498 million, of which only $16 million was PTI; and by 1992 it had ballooned to $829.7 million, of which only $29.6 million was PTI. (Exs. 33, 2068.) Schering-Plough housed its Irish cash generated by Scherico and Limited in a cash investment pool located within Essex Chemie on the theory that it was administratively and economically more efficient to manage the E&P under one entity. (JTS 13, ¶ 103; 1/17/08 Ludwig Test. 70:25-71:11; Ex. 75.) Ludwig testified that when Schering-Plough engineered inter-subsidiary loans and advances, it used intercompany payables and receivables. But he never saw accompanying formalities such as loan documentation, promissory notes, or other documents containing any covenants, warranties, or default events. Nor did he ever see a transferee post security or collateral for an advance. (1/17/08 Ludwig Test. 69:25-73:2.) Schering-Plough wanted to use the mostly untaxed, offshore Irish cash in the United States for multiple purposes, including R&D programs, normal operating expenses, and particularly a $1 billion stock repurchase program. Stock repurchase programs, which reduce the number of shares outstanding, were common among Schering-Plough s competitors in the late 1980s and early 1990s for several reasons, not least of which was that they created value for shareholders by increasing earnings per share, which usually results in a higher stock price. 9 Stock repurchase programs also help protect companies from hostile takeovers. (1/15/08 Luciano Test. 56:16-57:12; 1/15/08 Wyszomierski Test 115:14-22; 1/16/08 Nichols Test. 24:325:11.) Schering-Plough s stock repurchase program, which it had announced in September 1990, required leveraged financing (i.e., loans). (1/15/08 Wyszomierski Test. 128:4-18; Ex. 211 at 3, 11, 14; 1/16/08 Nichols Test. 21:16-24:9.) Luciano and Nichols testified that Schering-Plough was eager to not incur debt where cash was otherwise accessible to fund domestic operations, including its stock repurchase programs. (1/15/08 Luciano Test. 65:1-67:2; 1/16/08 Nichols Test. 23:25-24; 27:24-28:7.) Schering-Plough continually attempted to maintain a debt-to-capital ratio comparable to those of competitor pharmaceutical companies. (1/15/08 Wyszomierski Test. 128:23-129:20; 1/17/08 Ludwig Test. 10:18-11:8; 1/29/08 Moore Test. 12:3-22.) Pharmaceutical companies typically had low debt-to-capital ratios, and, in order to keep its favorable credit rating, Schering-Plough needed to keep its debt ratio below 50 percent. (1/15/08 Luciano Test. 65:1-67:2.) Despite its efforts, by the end of 1990, Schering-Plough s debt-to-capital ratio was higher than any of its main competitors. (1/17/08 Ludwig Test. 10:18-11:8.) Because Schering-Plough found increasing debt and a ballooning 6 balance sheet undesirable, it consulted with ABN in the late-1980s about how it could decrease the cash and debt on its balance sheet. (1/15/08 Wyszomierski Test. 127:19-129:5; 1/18/08 Ludwig Test. 31:23-33:8; 1/17/08 Ludwig 76:11-77:25.) According to Ludwig, Schering-Plough engaged 6 Schering-Plough describes ballooning of the balance sheet as the reporting of increasing amounts of cash and debt in a consolidated balance sheet. (Pl. Proposed Findings of Fact ¶ 25.) Jack Wyszomierski, Treasurer of Schering-Plough at the time the transactions were created (and later, Schering-Plough s Chief Financial Officer), testified that [p]articularly in this case . . . ballooning [refers to] cash resources on the asset side of the balance while simultaneously increasing debt on the other side of the balance sheet, so both assets and liabilities go up. (1/15/08 Wyszomierski Test. 127:19-128:2.) 10 ABN in the late 1980s under a consulting agreement to provide advisory services on how Schering-Plough could reduce the ballooning of its balance sheet. (1/17/08 Ludwig Test. 76:1177:25.) Schering-Plough also consulted with Merrill Lynch, its long-time investment banker and principal financial advisor. (1/18/08 Ludwig Test. 8:3-9; 31:23-33:8; 1/16/08 Nichols Test. 19:13-20:19; 1/15/08 Wyszomierski Test. 115:23-116:17.) Luciano testified that a principal responsibility Schering-Plough placed on Merrill Lynch was to design tax-beneficial investment vehicles. (1/15/08 Luciano Test. 69:24-71:8.) In the past, Merrill Lynch had developed a tax vehicle for Schering-Plough and others known as a contingent installment note sale. Under this product, Schering-Plough invested in a partnership known as the Kralendijk Partnership in which ABN had participated as a counterparty. The Kralendijk Partnership generated capital losses that offset certain capital gains Schering-Plough had received through a sale of one of its divisions. (1/16/08 Nichols Test. 100:1-12, 128:16-25, 129:1-14; 1/17/08 Ludwig Test. 78:19-79:9).7 It is against this backdrop that Schering-Plough again sought counsel from Merrill Lynch on how to repatriate its offshore cash without incurring the bite of Subpart F. In addressing Schering-Plough s need for cash and ballooning balance sheet, Luciano enlisted Schering-Plough executive Nichols to help find a tax-efficient domestic funding solution. (1/16/08 Nichols Test. 19:4-12.) Nichols considered, but ultimately rejected, the options of direct loans and so-called back-to-back loans. (1/16/08 Nichols Test. 26:2-28:7; 2/25/08 Foster Test. 158:20-159:16; JTS 13, ¶¶ 108-110.) In a back-to-back loan, a foreign subsidiary with excess cash makes a deposit at a foreign branch of a bank, from which the 7 The Third Circuit later invalidated this very type of product marketed by Merrill Lynch for another client in ACM Partnership v. Commissioner, 157 F.3d 231 (3d Cir. 1998). The IRS likewise disallowed Schering-Plough s claimed capital losses from the Kralendijk Partnership. 11 domestic parent company would subsequently borrow money. (JTS 13, ¶ 109.) Despite the fact that back-to-back loan arrangements would have had no federal income tax consequences, Schering-Plough did not pursue this mode of domestic funding because it considered the financial accounting treatment inferior to other alternatives. (JTS 13-14, ¶ 110.) According to Ludwig, the company rejected intercompany lending from subsidiary to parent, as well as a dividend issuance from the Swiss subsidiaries to their parent shareholder because these proposals would have been presently taxable to the extent the offshore cash had not been previously taxed. (1/17/08 Ludwig Test. 128:11-24.) See I.R.C. §§ 316, 956, 959, Merrill Lynch proposed to Nichols and other Schering-Plough executives within Schering-Plough s treasury, accounting, legal, and tax departments another alternative, this time an interest rate swap-and-assign transaction that would enable Schering-Plough to obtain cash to finance its domestic operations, would not balloon Schering-Plough s balance sheet, and would be tax efficient. (1/16/08 Nichols Test. 28:8-32:13; 1/17/08 Ludwig Test. 15:13-25; Exs. 2, 22.) The transactions, Merrill Lynch explained, would be governed by IRS Notice 89-21. Notice 89-21, entitled Deferred Recognition of Income from Lump-Sum Payments in Connection with Notional Principal Contracts, was published in the Internal Revenue Bulletin on February 7, 1989. Notice 89-21 provide[d] guidance with respect to the federal income tax treatment of lump-sum payments received in connection with interest rate and currency swap contracts. (Ex. 224, p. 651.) Notice 89-21 functioned as the prevailing guidance for tax reporting of receipt of a lump-sum payment in connection with an interest rate swap where swap counterparties were obligated to make swap payments in future taxable years. (Ex. 224 at 651; Treas. Regs. § 1.446-3(g)(4), and (j).) The notice states in pertinent part: Under . . . the Internal Revenue Code, if a taxpayer s method of accounting does not clearly reflect income, the computation of 12 taxable income shall be made under such method as, in the opinion of the Secretary, does clearly reflect income. . . . In the case of a payment received during one taxable year with respect to a notional principal contract where such payment relates to the obligation to make a payment or payments in other taxable years under the contract, a method of accounting that properly recognizes such payment over the life of the contract clearly reflects income. Moreover, including the entire amount of such payment in income when it is received or deferring the entire amount of such payment to the termination of the contract does not clearly reflect income and is an impermissible method of accounting. . . . In the case of lump-sum payments made or received with respect to notional principal contracts entered into, or assignments made, prior to the effective date of the regulations (including contracts entered into prior to the publication of this notice), a method of accounting used by a taxpayer is a method that clearly reflects income only if the payments are taken into account over the life of the contract using a reasonable method of amortization. . . . No inference should be drawn from this notice as to the proper treatment of transactions that are not properly characterized as notional principal contracts, for instance, to the extent that such transactions are in substance properly characterized as loans. This notice serves as an administrative pronouncement as that term is described in section 1.6661-3(b)(3) of the Income Tax Regulations and may be relied upon to the same extent as a revenue ruling or revenue procedure. IRS Notice 89-21, 1989-1 C.B. 651, 1989 IRB LEXIS 91 (emphasis added). Relying on Notice 89-21, Skadden, Arps, Slate, Meagher & Flom ( Skadden Arps ), Schering-Plough s outside counsel, and Deloitte & Touche ( Deloitte ), Schering-Plough s independent auditor, reviewed the swap structure, after which Deloitte opined that there is no need to accrue U.S. income taxes on the proceeds of the sale of the swap receive legs to Scherico and Limited. (1/16/08 Nichols Test. 29:25-30:13; 32:14-33:17; 34:6-10; 38:23-39:11; 39:25-40:6; 102:12-104:6; Ex. 21.) Schering-Plough and Merrill Lynch executed a standardform engagement letter memorializing Schering-Plough s business objectives and outlining 13 Merrill Lynch s supporting role. (Ex. 1424; 1/17/08 Ludwig Test. 17:6-18; 2/1/08 Pepe Test. 12:5-15:7.) A report prepared for Schering-Plough s Finance and Audit Committee confirmed that the swap-and-assign transactions effectively repatriated $728 million while deferring U.S. tax. (Exs. 216, 34.) Deloitte concurred, stating that the transactions were used as a means of repatriating money from Europe without having it taxed as a dividend. (Ex. 15.) Luciano answered affirmatively when asked at trial whether Mr. Das [of Merrill Lynch] represented to [him] that the sole purpose of this swap and sale transaction was to give ScheringPlough access to its foreign cash, without paying taxes, correct? (1/15/08 Luciano Test. 76:417.) Luciano also responded yes without qualification when asked if the STRIPS transactions were opportunities allowing the repatriation of additional funds from non-U.S. forces to accelerate future remittances by several years. (1/15/08 Luciano Test. 78:19-23.) Luciano also agreed that the only problem with accessing cash earned by the foreign subsidiaries was paying the tax upon repatriation. (1/15/08 Luciano Test. 75:10-13.) Jack Wyszomierski, who was a vice president and treasurer at Schering-Plough at the time the transactions at issue were executed, testified that at the time, Schering-Plough had been suffering from a ballooning balance sheet due to the cash-flush Swiss subsidiaries and Schering-Plough s domestic borrowing. (1/15/08 Wyszomierski Test. 124:4-18.) D. The 1991 and 1992 Swap-and-Assign Transactions 1. 1991 Swap-and-Assign Transaction Consistent with its plan detailed above, in 1991, Schering-Plough entered into an interest rate swap with ABN on January 2, 1991. (JTS 4, ¶ 9; Compl. ¶ 16.) The counterparties agreed to make interest payments to each other based on a notional amount of principal, and to make payments under a different interest rate for a set term of years. (Compl. ¶ 5.) The parties only exchanged the interest payments, not the notional principal. (Id.) Schering-Plough s 1991 14 transaction with ABN ( 1991 swap ) called for each party to make payments every six months from January 2, 1991 until December 15, 2010, based on a principal amount of $650 million. (JTS 5, ¶ 11; Compl. ¶ 16.) The agreement called for Schering-Plough to pay ABN interest based upon the London Interbank Offered Rate ( LIBOR ), while ABN would pay ScheringPlough interest based on the federal funds rate.8 (JTS 5, ¶ 13; Compl. ¶ 16.) The standardized swap terms permitted ABN and Schering-Plough to offset ( net ) the two payments, such that the party owing the higher amount paid only the difference. (JTS 5, ¶ 15.) Because the swaps were entered into at market, the present value of Schering-Plough s receive leg at inception approximated the present value of its pay leg. (1/24/08 Taylor Test. 21:13-20; 1/30/08 Parsons Test. 53:21-54:3.) Under the 1991 swap, Schering-Plough was permitted to assign its right to receive payments from ABN to a designated assignee, which included the Swiss subsidiaries. (JTS 5, 18.) Significantly, upon any assignment, the parties could no longer net payments; rather, each periodic payment would be due in full to the party owning the right to the particular income stream. (JTS 5, ¶ 19.) Thus, Schering-Plough s obligations to ABN remained unchanged, irrespective of any re-routing of incoming payments to Schering-Plough s offshore subsidiaries. In other words, upon assignment of its receive leg rights, Schering-Plough remained duty-bound to make the entire periodic pay leg distributions to ABN, notwithstanding ABN s parallel obligation to make the payments to Schering-Plough s third-party designee. Another provision in the 1991 swap permitted ABN to terminate the swap upon default by Schering-Plough. One 8 LIBOR is the market interest rate that major banks offer to pay each other on deposits in Europe of U.S. dollars for a given maturity of the deposit. The federal funds rate is the interest rate, set by the Federal Reserve, that U.S. banks charge each other for overnight deposits. (Compl. ¶ 16.) 15 of the events triggering default the so-called 60-day credit trigger permitted ABN to terminate upon the following occurrence: If [Schering-Plough s] credit rating by Standard & Poor s Corporation shall fall below AA- and its credit rating by Moody s Investors Service shall fall below Aa3 and within 60 days of the later of such credit ratings so to fall [Schering-Plough] shall not have reinstated one of such credit ratings . . . . (JTS 6, ¶ 20.) Upon entering into the $650 million notional principal interest rate swap with ScheringPlough, ABN entered into a mirror swap with Merrill Lynch with the same notional principal of $650 million. (1/24/08 Taylor Test. 35:24-37:11.) Under the mirror swap, ABN obligated itself to make the same LIBOR-based payments that it was to receive under its swap with Schering-Plough. In exchange, it received from Merrill Lynch the same federal funds rate payments that it was obligated to make to Schering-Plough, plus a premium of ten basis points for serving as intermediary to the swap-and-assign transaction.9 (1/24/08 Taylor Test. 37:2-5.) The purpose of this mirror swap was to eliminate the market (i.e., interest rate) risk ABN faced under its swap with Schering-Plough. At the same time, however, ABN faced (at least nominally) credit risk exposure were Schering-Plough to default on its swap obligations. (1/24/08 Taylor Test. 37:18-24.) On February 6, 1991, Schering-Plough assigned its right to receive income streams on $60 million of the notional principal to Banco di Roma, a third-party bank. The purpose of this assignment was to establish an arms-length pricing arrangement for the assignments to the Swiss subsidiaries that followed. (1/17/08 Ludwig Test. 81:21-83:5; 1/18/08 Ludwig Test. 33:21- 9 One basis point is equal to 1/100 of one percent. Thus, ABN could expect to receive annual compensation of 1/10 of one percent for participating in the transaction. 16 34:19; Exs. 662-667.) Banco di Roma paid Schering-Plough $26.4 million for the assignment on February 8, 1991; it funded the lump-sum payment with a $27 million zero coupon time deposit provided by ABN. (Exs. 667, 2038.) In contrast with the assignments to the Swiss subsidiaries, ABN and Banco di Roma agreed that ABN would have an option to call (i.e., terminate) its payment obligation, and Banco di Roma would have the corresponding ability to sell the same income streams back to ABN. As part of the agreement, ABN agreed to compensate Banco di Roma 15 basis points (.015%) per year until the option was exercised. The option was, in fact, exercised in full by March 1993. (Ex. 2038; 1/23/08 Den Baas Test. 62:7-63:2.) Unlike the Swiss subsidiaries, Banco di Roma did not receive the plenary option to sell the assigned receive legs back to Schering-Plough. (1/17/08 Ludwig Test. 88:14-20.) See infra. After entering into the benchmark assignment with Banco di Roma, Schering-Plough assigned its right to receive payments under the 1991 swap to Scherico for years 6-20, which would commence on December 15, 1995, and consist of payments for $460 million of the notional principal relating to the 1991 swap ( 1991 Scherico Assignment ). (JTS 6, ¶ 21.) Using the benchmark price established by the Banco di Roma assignment, Scherico paid Schering-Plough $202.4 million for the assignment of the receive leg under the 1991 swap. (JTS 6, ¶¶ 22-23.) ABN assented to the assignment by way of a Letter of Consent to Schering- Plough. (JTS 6, ¶ 28.) ABN also confirmed that it would make the assigned payments to Scherico independently of, and without reference to, the performance by . . . [Schering-Plough] of . . . [Schering-Plough s] obligations in respect of the [1991 swap]. (JTS 6, ¶ 27.) At the same time, Scherico and Schering-Plough entered into a Put Option Agreement that would permit Scherico to assign back to Schering-Plough the receive leg of the 1991 swap that Scherico 17 had purchased, and would thereby compel Schering-Plough to pay the fair market value of the remaining receive leg income streams on the date the put option was exercised. (JTS 6-7, ¶¶ 2930.) Also on February 6, 1991, Schering-Plough assigned Limited the right to receive payments from the 1991 swap as to $100 million of the notional principal amount, for which Limited paid Schering-Plough $44 million (again using the Banco di Roma benchmark price) on February 8, 1991 ( 1991 Limited Assignment ). (JTS 7, ¶¶ 32-34.) As with Scherico s assignment, ABN also consented to the assignment to Limited, and a put option was granted to Limited to re-assign the receive pay leg back to Schering-Plough. (JTS 7, ¶¶ 32-34.) The Court will refer to the 1991 Scherico Assignment and the 1991 Limited Assignment collectively as the 1991 assignments. Schering-Plough paid Merrill Lynch a fee of $2.2 million for its services in connection with the 1991 swap and subsequent assignments to Scherico and Limited. (JTS 8, ¶ 44.) It accounted for those fees by capitalizing them over the intended 20-year length of the swaps, analogizing to Statement of Financial Accounting Standards ( SFAS ) No. 91: Accounting for Nonrefundable Fees and Costs Associated with Originating or Acquiring Loans and Initial Direct Costs of Leases. (JTS 118; Exs. 52, 501, 2058.) Schering-Plough reported the assignments as sales for federal income tax purposes and applied an amortization method set forth in Notice 89-21. (JTS 8, ¶ 45; 1/16/08 Nichols Test. 38:23-39:7; Ex. 31.) Schering-Plough did not report as income in 1991 the consideration received from the offshore subsidiaries for the assignment of payment streams. (JTS 8, ¶ 46.) Instead, in each year starting in 1996, Schering-Plough reported, for federal income tax purposes, 18 a ratable portion of the consideration received by reducing its deductions for payments made under the swap contracts. (JTS 8, ¶ 46; Compl. ¶¶ 17, 18.) 2. 1992 Swap-and-Assign Transaction On October 1, 1992, Schering-Plough and ABN entered into a second interest rate swap ( 1992 swap ).10 (JTS 8, ¶ 47.) Under the 1992 swap, Schering-Plough was required to make payments to ABN beginning on October 1, 1992 through October 1, 2012 based upon a total notional principal amount of $950 million and a 12-month LIBOR index for the period running for the term of the swap. (JTS 8, ¶ 48.) In return, ABN would make payments from October 1, 1992 to October 1, 2012 on the same notional principle using an interest rate that would reset every 2 years based upon a 30-day commercial paper rate, plus 0.05%.11 (JTS 8, ¶ 49.) Like the 1991 swap, the 1992 swap permitted Schering-Plough and ABN to net the periodic payments due. (JTS 8, ¶ 51.) The 1992 swap also contained a 60-day credit trigger identical to the default trigger contained in the 1991 swap agreement. (JTS 9, ¶ 58.) As in the 1991 swap, ABN entered into an offsetting mirror swap with Merrill Lynch in which ABN converted its right to receive LIBOR payments into a commercial paper-based income stream, plus a premium of ten basis points. This eliminated ABN s interest rate risk and simultaneously compensated ABN for serving as the transaction intermediary. (1/24/08 Taylor Test. 35:24-39:10; Ex. 206.) On October 30, 1992, Schering-Plough assigned its right to receive income streams on $25 million of the notional principal to Rabobank Nederland ( Rabobank ), a third-party bank, 10 Schering-Plough entered into the 1992 swap one day after having been advised by outside tax counsel that the tax window provided by Notice 89-21 purportedly relieving it of Subpart F taxation would close by year end 1992. (Ex. 25; Nichols Dep. 189:24-194:7.) 11 Commercial paper is a short-term, unsecured promissory note issued by a corporation. See Federal Reserve Release, Commercial Paper, available at http://www.federalreserve.gov/releases/CP/about.htm (last visited Aug. 28, 2009). 19 for approximately $12 million (the sum was paid to Schering-Plough on November 2, 1992). (Ex. 702.) The purpose of this assignment was again to establish an arms-length pricing arrangement for the assignments to the Swiss subsidiaries. (1/17/08 Ludwig Test. 55:15-21, 92:11-21; 1/18/08 Ludwig Test. 34:20-35:7; 1/15/08 Wyszomierski Test. 134:1-4; Ex. 692; Exs. 700-702; Ex. 711.) Like the arrangement between ABN and Banco di Roma, ABN paid Rabobank $28,000 for its participation in the transaction, and repurchased the receive legs from the third-party bank one week after Schering-Plough had assigned them to Rabobank. (1/17/08 Ludwig Test. 94:21-23; 1/23/08 Den Baas Test. 66:6-67:4; Exs. 18, 245, 246.) Also on October 30, 1992, Schering-Plough assigned to Scherico the right to receive payments from ABN from years 6-20 on a notional principal amount of $925 million, an assignment for which Scherico paid a sum of $444 million to Schering-Plough on November 2, 1992 ( 1992 Scherico Assignment ). (JTS 9, ¶¶ 59-61.) ABN consented to the assignment, and acknowledged that it would make the payments to Scherico independently of, and without reference to, the performance by [Schering-Plough] in making its payments. (JTS 10, ¶ 66.) Unlike the 1991 swap, in the 1992 swap, Scherico and its parent Schering-Plough did not enter into a put option agreement. (JTS 10, ¶ 67.) Schering-Plough paid Merrill Lynch a $2.0 million fee for its work on the 1992 swap and subsequent assignment. (JTS 10, ¶¶ 63, 70.) ScheringPlough accounted for the advisement fees by amortizing them over the contemplated 20-year swap, again referencing SFAS No. 91. Schering-Plough adopted the same tax approach for the 1992 transaction as it did with the 1991 transaction, and reported the assignments to the Swiss subsidiaries as sales for federal income tax purposes, and applied an accounting method prescribed by Notice 89-21. (JTS 10, ¶ 71; 1/16/08 Nichols Test. 38:23-39:7; Ex. 31.) Schering-Plough did not report the consideration 20 received from Scherico for the assignment as income in 1992, but instead began reporting as income a ratable portion of the consideration received by reducing its deductions for payments made under the swap contracts. (JTS 10, ¶ 72.) All told, under the 1991 and 1992 swap-and-assign transactions, Schering-Plough received $690.4 million in repatriated lump-sum payments from the Swiss subsidiaries in exchange for the future income streams from ABN. (JTS 6, ¶ 22; 7, ¶ 33; 9, ¶ 61.) E. Structuring and Funding the Transactions 1. The Transactions Were Structured to Maximize the Amount Repatriated to the United States Luciano s trial testimony established that before determining the notional principal amounts for the swap transactions, Merrill Lynch needed to know how much cash ScheringPlough wanted to repatriate. (1/15/08 Luciano Test. 80:22-81:8.) After determining the amount to repatriate, Merrill Lynch then back-solved for the notional principal amount. (1/15/08 Luciano Test. 80:22-81:8.) The more money Schering-Plough wanted to repatriate, the higher the notional principal amount would go. Macauley Taylor, a member of Merrill Lynch s swaps group, explained at trial that by entering into a longer-term swap, Schering-Plough was able to more easily manage a larger notional principal amount, and at the same time assign a receive leg to the Swiss subsidiaries with a larger present value, thereby enabling it to receive a larger lumpsum payment. (1/24/08 Taylor Test. 5:7-8, 25:24-26:3.) Thus, by increasing the length of the swap-and-assign transactions and the amount of the notional principal, Schering-Plough was able to bring larger amounts of the offshore E&P to the United States under the swap-and-assign mechanism. 21 2. Schering-Plough Routed Limited s Irish Cash Through Scherico to Fund the Lump-Sum Payments Schering-Plough s corporate hierarchy is reflected by the following diagram: Schering-Plough Corporation Schering Corp. International Limited Scherico Essex Chemie As indicated, Limited had accumulated substantial amounts of previously untaxed foreign E&P that was held in the Essex Chemie cash investment pool. Scherico, by contrast, was in a more favorable tax position: It held relatively little E&P, and most of what it did hold was PTI. Scherico did not report any accumulated earnings and profits, and had $48.5 million in PTI at the end of 1991. At the end of 1992, Scherico had $32.8 million total E&P and $115 million in PTI (Exs. 2000, 2064.) Thus, the majority of the cash that Scherico held was PTI that could be repatriated back to the United States without incurring Subpart F taxation. See I.R.C. § 959. To make the lump-sum payments to Schering-Plough in exchange for future receive leg rights, Limited advanced to Scherico funds that it owned in the cash investment pool. As Schering-Plough explained it: Essentially, all Irish cash of Schering-Plough, Ltd. is advanced to Scherico. There are no notes or other evidences of indebtedness. It is actually capital contributions which have not been formally declared, as such, in order to avoid significant transactional taxes (stamp taxes) in Switzerland. The funds were 22 used to enable Scherico to make investments, including the swap transactions.12 (Ex. 20; Nichols Dep. 146:11-153:20). The $646.4 million that Scherico paid to ScheringPlough in exchange for the receive leg assignments was funded by Limited s accumulated E&P. Taking into account the $44 million that Limited paid for the assignment that it received, Limited s Irish cash funded the entire lump sum payments that were repatriated to the United States as a result of the 1991 and 1992 swap-and-assign transactions. (JTS, Proc. Agreement, ¶ B.2.) F. Aftermath of the Swap-and-Assign Transactions Parker Douglas, a member of ABN s workout group for managing distressed credit, testified for Schering-Plough that, while he could not recall the exact timing, in the early 2000s, Schering-Plough s credit ratings dipped, which led it to begin negotiations with ABN about amending the 60-day credit trigger provisions in the swap agreements. (1/22/08 Douglas Test. 81:22-82:13.) Kevin Moore, a Schering-Plough vice-president and treasurer responsible for managing the company s debt positions, testified that the parties agreed to amend the provisions from the original AA-/Aa3 threshold to A/A, and changed the trigger from an event of default to an event permitting either party to terminate the running agreement. (1/29/08 Moore Test. 47:20-48:20; Exs. 118, 274.) Schering-Plough agreed to compensate ABN $36 million over the remainder of the swaps duration for this accommodation. (1/22/08 Douglas Test. 105:20106:22; 1/29/08 Moore Test. 71:10-13; Ex. 274.) As Moore recounted, the original swap agreement required Schering-Plough to pay ABN a gross amount of $748 million if it breached 12 As noted above, Schering-Plough used intercompany payables and receivables sans formalities to account for the funds advanced from Limited to Scherico. 23 the credit default provision.13 (1/29/08 Moore Test. 64:17-65:8; Ex. 119.) Moore agreed that Schering-Plough would fund such a payment using a taxable dividend from its Swiss subsidiaries, which would result in a $185 million tax liability, and that avoiding such a taxable event was a consideration for not having [the] Swiss subsidiaries declare a dividend to Schering-Plough. (1/29/08 Moore Test. 68:13-16.) Despite the renegotiated credit trigger, Schering-Plough s credit ratings did fall below the amended threshold in 2004, and the company decided to terminate the swap agreement. (1/22/08 Douglas Test. 109:3-13; 1/29/08 Moore Test. 73:19-74:7; Ex. 280.) Moore testified that Schering-Plough first reacquired the receive legs from the Swiss subsidiaries for $395 million so that it could avoid making the $748 million gross payment to ABN. (1/29/08 Moore Test. 50:2451:8, 77:3-78:18, 80:10-18; Exs. 43, 123, 126, 127, 128.) The swap agreements were ultimately terminated on May 6, 2004. (1/29/08 Moore Test. 79:3-13.) G. IRS Audits and Assessments Following audits for the years of 1991 and 1992, the IRS issued a Notice of Deficiency, dated April 8, 2004, which stated that the 1991 and 1992 transactions were either constructive dividends or loans under § 956 of the Internal Revenue Code. According to the Notice of Deficiency: For the taxable years 1991 and 1992, it has been determined that the substance of the transactions between [Schering-Plough] and [ABN], which resulted in the transfer, in part through conduits, by [Limited] to [Schering-Plough], of $246,400,000.00 in 1991 and $444,000,000.00 in 1992, was not consistent with the form of these transactions, and that these transactions lacked economic 13 Under such a scenario, Schering-Plough would receive a gross payment of $731 million in return. While the net result was not a major concern to the company s balance sheet, the cash needed in the United States to make the $748 million gross payment and the resulting tax liability for financing such a payment was the driving force in renegotiating the credit trigger. (Ex. 119; 1/29/08 Moore Test. 64:17-65:8.) 24 substance. Therefore, the characterization of these transactions as sales of portions of the notional principal contracts should not be respected. Further, these transactions should properly be characterized, consistently with their substance, or under the step transaction doctrine, as either loans, for purposes of section 956(c)(1)(C), or constructive dividends. As a result, Schering-Plough International, Inc. must include in gross income for the taxable years 1991 and 1992, its pro rata share of the increase in earnings invested in United States property by [Limited], (deemed distribution) or in the alternative [ScheringPlough] must include in gross income as a constructive dividend, the following amounts . . . . IRS Notice of Deficiency (Apr. 8, 2004.) The IRS thereafter increased Schering-Plough s 1991 taxable income by $242,331,758.00 and its 1992 taxable income by $462,587,882.00. (JTS 17, ¶ 130.) For the 1991 taxable year, the IRS determined that International should have included within its gross income the pro rata share of the increase in earnings invested in United States property by Limited ($242,331,758.00), or that Schering-Plough alternatively should have included that amount in gross income as a constructive dividend. (JTS 15, ¶ 125.) The IRS also examined other smaller-scale adjustments to Schering-Plough s 1991 tax return, and adjusted Schering-Plough s taxable income upward by disallowing the amortization of $83,000.00 in fees paid for the 1991 swap. (JTS 15, ¶ 125.) With respect to the 1992 taxable year, the IRS determined Schering-Plough s taxable income was to be increased by $462,587,882.00 because International was required to include that amount in gross income as its pro rata share of investment in United States property by Limited. (JTS 16, ¶ 126.) Like the 1991 swap, it determined that the deduction of $110,000.00 for fees was improper because the fees did not qualify as an ordinary and necessary business expense. (JTS 16, ¶ 126.) 25 As a result of the audit, the IRS assessed a tax deficiency as follows: For 1989, $8,624,448.00, plus interest of $11,855,683.21, totaling $20,480,131.2114; for 1991, it assessed a deficiency of $70,146,376.00, plus interest of $109,529,102.86, totaling $179,675,478.86; and for 1992, it assessed $114,843,573.00, plus interest of $157,870,859.62, totaling $272,714,432.62. (JTS 18, ¶ 131.) Thus, the grand total the IRS required Schering-Plough to pay in back taxes and interest relating to the 1991 and 1992 swap-and-assign transactions was $472,870,042.69.15 (JTS 18, ¶ 131.) Schering-Plough paid the assessment under protest on September 13, 2004, and on December 23, 2004 filed a timely claim for refund of the back taxes, plus interest. (JTS 18, ¶¶ 132-33.) On February 16, 2005, the IRS denied Schering-Plough s claims for a refund. (JTS 18, ¶ 134.) H. Commencement of Refund Action and Prior Motion Practice On May 16, 2005, Schering-Plough filed suit in this Court under 28 U.S.C. § 1346(a), seeking a refund of the 1989, 1991 and 1992 tax assessments, plus interest and costs.16 (JTS 18, ¶ 135.) The Court has previously found that Schering-Plough s complaint raises essentially two grounds for refund: The government mischaracterized the transactions as loans or 14 There was a 1989 assessment because the correlative fee adjustments for 1992 eliminated $8,624,448 of ScheringPlough s general business credit carryback from 1992 to 1989. (JTS 18, ¶ 131.) 15 The parties have agreed that, to the extent necessary, issues relating to computation of the correct amounts of Schering-Plough s federal income tax owed will be addressed after the Court renders judgment. (JTS Proc. Agreement, ¶ B.1.) 16 A taxpayer wishing to contest a tax assessment in court may pay the deficiency and sue for a refund in United States District Court or the United States Court of Federal Claims. 28 U.S.C. § 1346(a); I.R.C.. § 7422. Alternatively, the taxpayer may contest a notice of deficiency in the United States Tax Court without first paying the tax. I.R.C. § 6213. Schering-Plough considered all three options, but ultimately stated to the SEC (upon an inquiry regarding this litigation) that its chances of prevailing in the Tax Court [were] low, that its chances of prevailing in the Court of Federal Claims were less than 50/50, and in this Court as in our favor. (Ex. 38.) 26 constructive dividends; and the government treated Schering-Plough differently from other similarly situated taxpayers. The parties refer to the latter ground as the disparate treatment claim, and it was the basis of a motion for summary judgment brought by the government. The Court filed a written opinion [D.E. # 124], finding in favor of the government. In identifying the issue at the core of the disparate treatment claim, the Court wrote: An unrelated taxpayer ( First Taxpayer ), who is a direct competitor to Schering-Plough, entered into a transaction that was identical in all material respects to the swap and assignment transactions entered into by Schering-Plough. (Gov t. Statement of Material Facts ¶¶ A, B.) IRS auditors sought legal advice about First Taxpayer s swap and assignment transaction from the IRS s Office of the Chief Counsel, which was given in the form of a Field Service Advice Memorandum ( FSA ). (Schering-Plough Statement of Material Facts ¶¶ 4-5.) The Commissioner of the IRS determined that IRS Notice 89-21 applied, and therefore First Taxpayer[ ]s transaction was not taxable in the year in which it was executed. (Gov t. Statement of Material Facts ¶ C.) See also 1997 FSA LEXIS 206 (Aug. 29 1997). IRS auditors for ScheringPlough had also requested legal advice about swap and assignment transactions from the Office of the Chief Counsel. (ScheringPlough Statement of Material Facts ¶ 7.) As the parties explain it: Notice 89-21 was in effect until the IRS adopted new regulations in 1993. Those new regulations reversed the basic conclusion of Notice 89-21, and provided that such a lump-sum payment may be recharacterized as a loan that is reported as income in the year received under 26 U.S.C. section 956. See Treas. Reg. § 1.446-3(g)(4). (Ltr. from Schering-Plough and Gov t to Judge Shwartz, April 13, 2007 (docket entry # 94).) While Schering-Plough, as part of this lawsuit, disputes this new characterization, the significant fact related to the disparate treatment claim is that the IRS applied the regulations retroactively. Schering-Plough Corp. v. United States, No. 05-2575, 2007 U.S. Dist. LEXIS 88387, at *5-7 (D.N.J. Dec. 3, 2007). 27 The Court examined the cases upon which Schering-Plough relied, and ultimately determined that: (1) International Business Machines Corp. v. United States, 343 F.2d 914 (Ct. Cl. 1965), cert denied, 382 U.S. 1028 (1966), did not have the precedential value ScheringPlough attributed to it; and (2) that Schering-Plough s reliance on a field service advice memorandum did not preclude the government from changing its interpretation of how Notice 89-21 applied to the swap transactions and from applying the interpretation retroactively. Rather, the Court found that Dickman v. Commissioner, 465 U.S. 330 (1984) controlled, and on the basis of that case, and the earlier decision upon which Dickman relied, Dixon v. United States, 381 U.S. 68 (1965), agreed that the Commissioner may change an earlier interpretation of the law, even if such a change is made retroactive in effect. See Schering-Plough, 2007 U.S. Dist. LEXIS 88397, at *12, 17 (quoting Dickman, 465 U.S. at 343). V. DISCUSSION To summarize: Thus far, the Court has recounted the position in which Schering-Plough found itself during the late 1980 s and early 1990 s its foreign subsidiaries were flush with cash and it desired to bring the money back to the United States in a tax-efficient manner to fund domestic operations. The Court has also noted the various transactions reflected in the record that Schering-Plough considered in consultation with ABN and Merrill Lynch, its financial advisors. The Court has also: (1) described the details of the transactions Schering-Plough ultimately chose notional principle swaps with ABN, followed quickly by assignments of future swap income to its foreign subsidiaries in exchange for up-front payments; and (2) reviewed the government s rejection of Schering-Plough s tax treatment and its assessment of a significant tax liability. Finally, the Court has given the procedural history of the refund litigation before it, which led to the bench trial on the claim that Schering-Plough was 28 improperly taxed because the government mischaracterized the transactions at issue as loans (or constructive dividends) instead of sales to the Swiss subsidiaries. The Court now takes a closer look at the economics of the transactions, using expert testimony and reports and other evidence illustrating the nature of the transactions. After this economic analysis, the Court will apply the legal frameworks earlier identified: What is the economic reality of the transactions? Do they have economic substance? How do they fit within the scope of the applicable tax laws? A. Economic Analysis 1. Cash Flows According to Schering-Plough, the assignments to the Swiss subsidiaries were present sales of intangible property the future income streams from the stripped swap receive legs. The government, on the other hand, contends that the true economic nature of the swap-andassign transactions is that of a loan: The lump sums paid by the Swiss subsidiaries represent the traditional principal loaned, and the future income streams from the pay legs Schering-Plough paid to ABN (which were later re-routed back to the Swiss subsidiaries) represent repayment of the lump-sum principal, plus interest. The 1991 and 1992 swaps called for Schering-Plough to make payments every six months to ABN, based on the LIBOR. In return, ABN was obligated to make reciprocal payments to Schering-Plough; for the 1991 swap, the payments were based on the federal funds rate, and for the 1992 swap, they were based on a two-year commercial paper index. After the assignments to the Swiss subsidiaries, Schering-Plough remained obligated to make the pay leg distributions to ABN. ABN, however, after assenting to the assignments, became obligated to re-route its distributions to the Swiss subsidiaries for an aggregate $1.485 billion notional 29 principal ($560 million for the 1991 assignments and $925 million for the 1992 assignments) during years 6-20 of the swap agreements. The government analogizes this to an amortizing home mortgage loan. Dr. Ira Kawaller, a Ph.D economist, testified for the government that, based on the cash flows from the swap-and-assign transactions, many significant economic indicia of loans were present, including: (1) fixed maturity dates (1/31/08 Kawaller Test. 23:18-24:7); (2) set principal amounts (1/31/08 Kawaller Test. 24:8-25:7); (3) interest payments by Schering-Plough, routed through ABN to the Swiss subsidiaries (1/31/08 Kawaller Test. 25:17-26:9); and (4) fixed repayment schedules. (1/31/08 Kawaller Test. 25:8-16.) Dr. Kawaller made the following synthetic loan analysis at trial: By entering into a basis swap and replacing future cash receipts from the swap with a cash amount equal to the present value of the future cash receipts, Schering-Plough became the debtor in a synthesized loan. (Ex. 538 at 2; 1/30/08 Kawaller Test. 157:18-158:3.) The Swiss subsidiaries were the synthetic lenders to ScheringPlough because it was they who provided the up-front lumpsum payment to the parent. (Ex. 538 at 10; 1/30/08 Kawaller Test. 158:4-14.) The presence of ABN as a counterparty to the swaps and an intermediary to the synthetic loan did not disturb the fact that Schering-Plough received $690.4 million from the Swiss subsidiaries. If A loans B money and then B directs C to repay A, the transaction is nonetheless a loan. (Ex. 538 at 8; 1/31/08 Kawaller Test. 16:11-17:14, 22:8-23:11.) From an economic perspective, whether Schering-Plough repaid its Swiss subsidiaries directly, or chose to forgo the future income streams it was initially entitled to receive from ABN, does not affect the swap-and-assign transactions status as loans. In either case, the subsidiaries made an upfront advance to Schering-Plough, and thereafter received that money, plus interest. (1/31/08 Kawaller Test. 8:3-16.) In other words, Schering-Plough s lack of any legal obligations to its 30 Swiss subsidiaries did not affect the economic nature of the transactions. Dr. Kawaller also articulated a parallel between the swap-and-assign transactions and other swap varieties that are appropriately classified as loans. Dr. Kawaller explained the differences between three different types of swaps: a spot swap, in which swap counterparties become obliged to make periodic interest payments immediately; a deferred swap, in which one counterparty delays the date on which it must make a payment; and a prepaid swap, in which one counterparty accelerates its payment while the other makes payments pursuant to the contract. (1/30/08 Kawaller Test. 170:3-173:5; 1/31/08 Kawaller Test. 4:10-6:3.) Because prepaid swaps are considered loans from an economic perspective, Dr. Kawaller stated that there is really only one reason to do a pre-paid swap. . . . To serve as a financing mechanism. That s the only reason somebody would . . . accelerate the payments of one stream . . . to move cash up front. (1/30/08 Kawaller Test. 172:2-17.) Dr. Kawaller concluded that based on the cash flows resulting from assigning the income streams for swap years 6-20, Schering-Plough effected a financing with an embedded and deferred/prepaid swap. (1/30/08 Kawaller Test. 179:16-180:23; 1/31/08 Kawaller Test. 4:1010:4, 16:1-24.) In other words, because Schering-Plough executed the assignments for years 620 at the outset of the transactions, it received a pre-paid advance from the Swiss subsidiaries, and deferred ultimate repayment to the subsidiaries until the time in which the receive legs had been assigned. (1/31/08 Kawaller Test. 16:1-24.) Thus, by delaying the assignment until years 6-20 of the swap contract, Schering-Plough was able to embed a deferred financing mechanism for the latter fifteen years, while retaining a spot swap on the income streams for the initial five years. According to Dr. Kawaller, the only reason to bundle the spot swap for years 1-5 with the deferred swap for years 6-20 was to create window dressing to disguise the transactions true 31 nature. (1/31/08 Kawaller Test. 7:19-22.) Dr. Kawaller emphasized that structuring the transaction in this manner had no net effect on the cash flows inherent in the transaction the Swiss subsidiaries still advanced their corporate parent a lump sum, and the Swiss subsidiaries were still repaid the nominal amount, plus interest. (1/31/08 Kawaller Test. 9:21-10:4.) Consistent with Kawaller s testimony, Dr. John Parsons, another expert called by the government, discussed the importance of cash flows when examining the economic reality of a transaction: Q. When you re looking at whether a transaction has economic impact, what do you look at? A. I am primarily [focusing] on the cash flows, the amount and risk[] necessary, and pattern, anything in the bottom line, would yield a measure of value or an impact on the firm. Q. Are the cash flows an important consideration? A. Yes. That s the center of my focus. Q. Based on your review of the 1991 and 1992 swap and assignment transactions, what is their economic impact? A. There the basically they transfer the cash from the subsidiary to the parent without tax and they ultimately, the cash is promised back to the subsidiaries, so it s like a loan. Q. Is the transaction[] as a loan consistent with the cash flows of the transaction? A. Yes. I mean from a cash flow perspective, I don t know about a legal perspective or otherwise. Financially it s the same thing. That s how I would evaluate it, it s the same thing for all of the tools that I have at my disposal. Q. From a financial economist perspective, what is ABN s role in the 1991 and 1992 swap and assignment transactions? 32 A. They basically function as an intermediary to help the loan happen. (1/30/08 Parsons Test. 94:10-95:7.) Dr. Parsons further opined that ABN s status as an intermediary between Schering-Plough and its offshore subsidiaries did not materially affect the transactions status as loans. (1/30/08 Parsons Test. 94:25-95:20.) Another government expert, Joel Finard, an experienced capital markets executive, testified that from a capital markets standpoint, the swap-and-assign transactions were substantively loans because the purpose of the transactions, from the outset, was to appropriate the foreign cash up front to Schering-Plough, and that the transactions were nothing more than pre-paid swaps. (2/19/08 Finard Test. 86:2-21.) Finard agreed with Dr. Kawaller s explanation that pre-paid swaps are treated as loans from a capital markets perspective. (2/19/08 Finard Test. 86:3-22, 91:23-92:5, 96:3-97:2; Ex. 2133 at 40.) Moving to Schering-Plough s evidence about the economics of the transactions, Robert Clarke, who formerly served as Comptroller of the Currency, testified that from a banking perspective, the swap-and-assign transactions were not loans. (1/23/08 Clarke Test. 100:1100:16.) According to Clarke, swap transactions have none of the characteristics of loans because there is no indebtedness and there are no borrowers or lenders. (1/23/08 Clarke Test. 103:19-22.) Rather, the transactions involve two counterparties exchanging cash flows. (1/23/08 Clarke Test. 101:12-17.) Clarke opined that Schering-Plough s swap transactions were just plain and simple swap[s] and not loans. (1/23/08 Clarke Test. 100:11.) Further, Clarke found that Schering-Plough s sale of the receive-leg did not transform the swap into a loan (1/23/08 Clarke Test. 102:4-6), inasmuch as the Swiss affiliates paid an up-front and fixed-price for an asset in the receive-leg and there was no repayment obligation to those Swiss subsidiaries after the assignment. (1/23/08 Clarke Test. 103:12-18.) 33 Clarke pointed out that before the assignment, if Schering-Plough defaulted, ABN could have discontinued its reciprocal payments under the swap. (1/23/08 Clarke Test. 104:17-19.) After Schering-Plough assigned the receive-leg, however, if it reneged on its payments to ABN, ABN would have had the undiminished obligation to pay the Swiss subsidiaries. (1/23/08 Clarke Test. 104:20-24.) Thus, according to Clarke, the pay leg and receive leg payments were delinked post-assignment (1/23/08 Clarke Test. 105:5-7), such that Schering-Plough could not enforce ABN s payment to the Swiss subsidiaries. (1/23/08 Clarke Test. 105:19-25.) In support of his position that the receive leg was an asset for the Swiss subsidiaries, Clarke explained that owning an asset gives the owner the full benefit and detriment of its change in value, whereas a promise to pay remains at a constant value. (1/23/08 Clarke Test. 107:4-9.) According to Clarke, ABN s future payments were assets and not promises, because if the value of the payments increased due to a change in the interest rates, the Swiss subsidiaries would benefit by receiving additional cash; and if the payments went down in value the Swiss subsidiaries were harmed by losing expected money. (1/23/08 Clarke Test. 106:4-9.) Clarke considered tenuous the government s position that the swap-and-assign transactions were bargained-for loans, because the loan might not be paid off if the subsidiaries receivable interest rate slipped. (1/23/08 Clarke Test. 107:22-108:2.) Clarke also noted that the transactions lacked other loan characteristics, such as a fixed maturity date, lending of principal, or promissory note all hallmarks of loans. (1/23/08 Clarke Test. 114:11-20.) He explained that loan structures generally use collateral, which was missing in these transactions because the subsidiaries owned the receive legs. (1/23/08 Clarke Test. 106:17-21.) On these bases, Clarke concluded that each swap assign transaction was a sale, not a 34 loan. (1/23/08 Clarke Test. 115:7-23.)17 Reviewing the parties expert testimony on the economics of the transactions, the government s position, examined closely, fits. In a normal mortgage loan, the lender makes an up-front advance of principal, after which the mortgagor makes periodic principal and interest payments until the principal is reduced (amortized) to zero and the loan is paid off. The present value of the future payments made by the mortgagor to the lender (which, again, reflect principal and interest) equals precisely the amount initially loaned. Put another way, the total nominal amounts paid back by the mortgagor exceed the amount loaned; the difference represents interest. And so it was with the swap-and-assign transactions. The fairly priced swaps obligated Schering-Plough and ABN to exchange payments in which their present values, at the outset, were equivalent. When Schering-Plough assigned the receive legs to its Swiss subsidiaries, it received an up-front lump sum, then made continuing progress payments (albeit to ABN) reflecting the movement in the LIBOR index that it was required to pay. The Swiss subsidiaries then received interest payments (based on either federal funds or commercial paper indices) from ABN. Critically, the present values of these receivables were virtually identical to the amount advanced to Schering-Plough by the Swiss subsidiaries. Moreover, on any given date, the present values of the subsidiaries assigned receive leg rights equaled the principal amount that Schering-Plough had received from the subsidiaries, minus the amounts that it had already paid. 17 Other expert witnesses testified at trial and submitted reports on Schering-Plough s behalf. These experts, Andrew Onslow, Dr. Christopher Culp, and David Ross, testified regarding a myriad of issues, including, inter alia: (1) the manner in which the swap-and-assign transactions would be governed under English law; (2) the role ABN played as intermediary; (3) the effects (if any) the transactions had on ABN; (4) the absence of any direct payment obligations Schering-Plough owed to the Swiss subsidiaries; and (5) Schering-Plough s motivations for structuring the transactions in the manner that it did. The Court discusses more fully below this expert evidence as it becomes relevant to the Court s application of the facts to the various legal frameworks. See infra. 35 As Dr. Kawaller explained, Schering-Plough s current principal owed was determinable at any given moment by computing the present value of the subsidiaries receive leg rights. (1/31/08 Kawaller Test. 24:23-25:7.) The total amount to be repaid to ABN exceeded the nominal value of the lump-sum payments to Schering-Plough. The difference between the total amount ultimately recouped by the Swiss subsidiaries and the cash repatriated to Schering-Plough represents the economic equivalent of interest. This point is corroborated by the report of Schering-Plough s own expert witness, Dr. Christopher Culp. In his report, Dr. Culp concludes that the Swiss subsidiaries were entitled to receive anywhere between $33 million to $54 million in excess of the amounts advanced to Schering-Plough as a result of the receive leg assignments. (Ex. 615 at 1.) Acknowledging this point at trial, Dr. Culp testified that the Swiss subsidiaries got a good deal as a result of the return from the lump sums they advanced to Schering-Plough. (1/29/08 Culp Test. 106:24-107:18.) The Court also notes that, as described above, in terminating the swap-and-assign transactions in 2004, Schering-Plough paid the Swiss subsidiaries $395 million to reacquire the receive legs. (1/29/08 Moore Test. 50:24-51:8, 77:1-78:18, 80:8-16; Exs. 43, 128.) Consistent with the amortizing mortgage-loan model, this reacquisition payment reflects the everyday equivalent of an accelerated principal-and-interest reduction payment that extinguished the loan. Whether analogized with a deferred swap, a prepaid swap, a synthetic loan, or any other economic device, the transactions created an inflow of cash from the Swiss subsidiaries to Schering-Plough, followed by a later outflow of cash from Schering-Plough to ABN, thereafter followed by an inflow of cash back to the Swiss subsidiaries. This fits the description of a threeparty loan. 36 2. Documentary Evidence In addition to the experts opinions about the proper characterization of the transactions, the record evidence supports the government s argument that Schering-Plough considered the transactions to be loans during the period they were being designed and implemented: An internal ABN credit proposal, dated December 21, 1990, explaining the proposed transactions with respect to ABN s participation in the swaps with Schering-Plough. The document states in pertinent part: The reason for this structure is the fact that the parent through this mechanism receives a 20-year amortizing loan from it s [sic] subsidiary without incurring any negative tax implications in the U.S. (Ex. 238 at SB/SP0000575.) Internal notes by Daniel Filiberto, Schering-Plough s director of financial reporting and compliance. Filiberto participated in accounting for the 1991 and 1992 swap-and-assign transactions. Filiberto s notes state that We are really accounting for the net deferred income as a loan, but tax could not have us record it as a loan. (Ex. 168 at SP09471.) An amortization schedule Filiberto prepared for the 1992 swapand-assign transaction. The schedule contains such terms as beginning balance, interest, principal reduction, and ending balance. (Ex. 195 at P014781.) Schering-Plough s Treasury Department s first report to the company s Finance and Audit Committee of the Board of Directors. The report stated that after executing the swap-andassign transactions, the Swiss subsidiaries owned financial assets which earned interest. The report does not label the transactions as sales of future income streams. (Ex. 212 at P000114.) A Schering-Plough investment policy implemented shortly before the 1991 swap-and-assign transaction was executed. The policy required approval by the Finance & Audit Committee of the Board of Directors for any investment having a maturity of 360 days or more. The policy did not, however, apply to intercompany loans or dividends. Despite the new policy, the Finance & Audit Committee approved neither the 1991 nor the 1992 swap-and37 assign transactions. (Ex. 75.) Jay Ludwig, Schering-Plough s Assistant Treasurer, admitted at trial that there were only two possible conclusions: [E]ither Schering-Plough violated its own brand-new policy or Schering-Plough didn t think these were really investments. (1/18/07 Ludwig Test. 29:15-18.) Schering-Plough urges the Court not to give much or any weight to these documents. First, it claims that because the internal credit proposal generated by ABN was largely written in Dutch and the author was not identified, the Court should discount its probative value. (Pl. Proposed Findings of Fact ¶ 176(4).) But the document: (1) clearly contains ABN logos on multiple pages; (2) gives an accurate description of the transactions into which Schering-Plough subsequently entered; and (3) repeatedly refers to Schering-Plough as the client. Furthermore, it is rather unsurprising that portions of an internal ABN document proposing massive and complex transactions to the upper echelon of the investment bank are written in Dutch after all, ABN is a Dutch bank! Regardless, the Court need not employ an interpreter to understand the document s import; the provision demonstrating ABN s contemporaneous understanding that the swap-and-assign transactions would provide a 20-year amortizing loan . . . without incurring any negative tax implications in the U.S. is written in English and is thus readily comprehensible. (Ex. 238 at SB/SP0000575.) Schering-Plough next asserts that Filiberto s notes documenting the transactions as loans are unpersuasive because [he] denied that Schering-Plough was accounting for the assignment proceeds as loans. (Pl. Proposed Findings of Fact ¶ 176(2).) Without providing a citation of trial or deposition testimony, Schering-Plough essentially argues that Filiberto did not really believe what his own contemporaneous notes plainly state. Even if Filiberto did later disavow the import of his notes, however, this inconsistency would be reason to discredit rather than credit his trial testimony. 38 Schering-Plough also argues that Filiberto s amortization schedule for the payoff of the lump sums is unpersuasive because the schedule contains the terms Notional principal purchased. (Ex. 195 (emphasis added); Pl. Proposed Findings of Fact ¶ 176(3).) But other terms in the exhibit clearly militate against a sale: interest, amortize, beginning and ending balance, principal reduction, etc. Schering-Plough argues further that because the transaction documents themselves, Schering-Plough s financial statements for 1991 and 1992, and the testimony of Schering-Plough officials and advisors, refer to the transactions as sales, the transactions must be sales. But calling a club a spade does not make it one. Permitting a taxpayer to control the economic destiny of a transaction with labels would, as the Court discusses more completely below, exalt form over substance, thereby perverting the intention of the tax code. B. Legal Analysis 1. Introduction The Court begins its legal examination with two basic principles of American tax law, which are both expounded in Gregory v. Helvering, 293 U.S. 465 (1935). There, the Supreme Court stated that [t]he legal right of a taxpayer to decrease the amount of what otherwise would be his taxes, or altogether avoid them, by means which the law permits, cannot be doubted. But the question for determination is whether what was done, apart from the tax motive, was the thing which the statute intended. Id. at 469 (internal citations omitted). In other words, a taxpayer is free to arrange his financial affairs to minimize his tax liability; thus, the presence of tax avoidance motives will not nullify an otherwise bona fide transaction. Estate of Stranahan v. Comm r, 472 F.2d 867, 869 (6th Cir. 1973). However, while it is well settled that a corporate taxpayer may enter into transactions that minimize its taxes, at the same time, taxation is based upon substance, not form. Lerman v. Comm r, 939 F.2d 44, 54 (3d Cir. 1991). For that reason, 39 a court must scrutinize the transaction to ensure that its particular structure accurately reflects its economic essence. This substance-over-form analysis requires a reviewing court to decipher for itself the true nature of the transaction: Where the Commissioner attacks the formal agreement the Court involved is required to examine the substance and not merely the form of the transaction. This is so for the very good reason that the legitimate operation of the tax laws is not to be frustrated by forced adherence to the mere form in which the parties may choose to reflect their transaction. Comm r v. Danielson, 378 F.2d 771, 774 (3d Cir. 1967); see also Comm r v. Court Holding Co., 324 U.S. 331, 334 (1945) ( To permit the true nature of a transaction to be disguised by mere formalisms, which exist solely to alter tax liabilities, would seriously impair the effective administration of the tax policies of Congress. ); ACM P ship v. Comm r, 157 F.3d 231, 247 (3d Cir. 1998) ( [W]e must look beyond the form of the transaction to determine whether it has the economic substance that its form represents. ) (citing Kirchman v. Comm r, 862 F.2d 1486, 1490 (11th Cir. 1989) (internal quotations omitted)); Kirchman, 862 F.2d at 1490 ( If a transaction s form complies with the Code s requirements for deductibility, but the transaction lacks the factual or economic substance that form represents, then expenses or losses incurred in connection with the transaction are not deductible. ). In 1977, the Court of Claims (now the Court of Federal Claims) reached essentially the same question that this Court faces today: Whether the assignment of future income in exchange for a lump-sum payment is, in substance, a loan or a sale. See Mapco, Inc. v. United States, 556 F.2d 1107 (Ct. Cl. 1977). The Court of Claims described its analysis as follows: In determining whether the instant transaction was a bona fide sale, we have concentrated on the economic substance of the transaction rather than the mere form in which it was cast. To us, the facts that valuable consideration was present and that the assignment had legal effect between the parties are insufficient to show that a 40 substantive sale occurred. We readily admit that the distinction is narrow between selling a property right to future income and assigning anticipated income as collateral to secure financing. Nevertheless, we feel that the distinction seems logically and practically to turn upon an out-and-out economically realistic transfer of a substantial property interest. Where the line of demarcation should finally be placed we need not try to anticipate here. But we are certain that distinctions attempted on the basis of the various legal names given a transaction, rather than on its actual results between the parties, do not afford a sound basis for delimitation. Mapco, 556 F.2d at 1110.18 The substance-over-form standards enunciated above are also incorporated into Notice 89-21 itself. The penultimate clause in the notice states specifically that [n]o inference should be drawn from this notice as to the proper treatment of transactions that are not properly characterized as notional principal contracts, for instance, to the extent that such transactions are in substance properly characterized as loans. IRS Notice 89-21, 1989-1 C.B. 651, 1989 IRB LEXIS 91 (emphasis added). The legal analysis below addresses several variations of the broader tax concept that substance should prevail over form. Am. Potash & Chem. Corp. v. United States, 399 F.2d 194, 207 (Ct. Cl. 1968). First, the Court analyzes its factual findings that the swap-and-assign transactions substantively and economically embodied loans under the general substance-overform doctrine. In doing so, it discusses ABN s role as a pass-through intermediary between Schering-Plough and the Swiss subsidiaries in completing the transactions, and compares Mapco to the facts presented here. The Court also reviews the step transaction doctrine in relation to the structure of the transactions. Second, it applies the economic substance doctrine to determine whether the transactions had sufficient objective economic effects on the parties 18 Further discussion of Mapco appears infra. 41 business positions, and whether Schering-Plough had legitimate subjective business motivations, such that the transactions should be respected. Finally, the Court discusses whether the structured transactions comport with the overall intent of Subpart F. Each iteration of analysis yields the conclusion that the government correctly levied a repatriation tax upon ScheringPlough. 2. Substance-over-Form Doctrine For a given transaction to constitute [a] true loan[] there must have been, at the time the funds were transferred, an unconditional obligation on the part of the transferee to repay the money, and an unconditional intention on the part of the transferor to secure repayment. Geftman v. Comm r, 154 F.3d 61, 68 (3d Cir. 1998) (quoting Haag v. Comm r, 88 T.C. 604, 615-16 (1987), aff d, 855 F.2d 855 (8th Cir. 1988) (table)). Absent direct evidence of intent, the nature of the transaction may be inferred from its objective characteristics. Id. Such indicia include the presence or absence of debt instruments, collateral, interest provisions, repayment schedules or deadlines, book entries recording loan balances or interest payments, actual repayments, and any other attributes indicative of an enforceable obligation to repay the sums advanced. Id. (citing Fin Hay Realty Co. v. United States, 398 F.2d 694, 696 (3d Cir. 1968)). Because Schering-Plough and its Swiss subsidiaries are related entities not acting at arms length, the Court applies particular scrutiny because the control element suggests the opportunity to contrive a fictional [transaction]. Id. (quoting In re Uneco, Inc., 532 F.2d 1204, 1207 (8th Cir. 1976)); see also E. Coast Equip. Co. v. Comm r, 222 F.2d 676, 678 (3d Cir. 1955) ( [W]hen relationships are as close as these, intercorporation transactions are subject to rather close scrutiny if the result is to effect tax obligations. ); Alterman Foods, Inc. v. United States, 611 F.2d 866, 870 (Ct. Cl. 1979) (in a related party transaction, [t]he possible tax 42 evasion motive invites most careful scrutiny (quoting Electric & Neon, Inc. v. Comm r, 56 T.C. 1324, 1339 (1971), aff d, 496 F.2d 876 (5th Cir. 1974)). Because the parties to these transactions occup[ied] both sides of the bargaining table, the form of [the] transaction[s] does not necessarily correspond to the intrinsic economic nature of the transaction[s], for the parties may mold [them] at their will in order to create whatever appearance would be of benefit to them despite the economic reality of the transaction[s]. Geftman, 154 F.3d at 75 (internal citations and quotations omitted).19 Courts employ an objective test of economic reality, considering whether the evidence as to the contemporaneous intent at the time of the transfers and the objective attributes and economic realities of the transaction . . . gave rise to a bona fide debt. Id. at 68 (citing Fin Hay Realty, 398 F.2d at 697). Schering-Plough lists 17 factors purporting to establish why the transactions were really sales. See Pl. Pre-Trial. Br. at 14-21. These factors are: The transactions were documented, accounted for, and reported as sales, not loans. Id. at 14; Schering-Plough had no payment or repayment obligations to the Swiss subsidiaries. Id. at 15; The receive legs were transferred absolutely, and not as collateral. Id.; There was no reserve or holdback of any portion of the sale price from the Swiss subsidiaries to Schering-Plough. Id. at 16; The Swiss subsidiaries paid fixed prices that were not contingent upon their receipt of future amounts under the swap receive legs. Id.; ABN was notified of the assignments and was required to acknowledge the same. Id.; 19 While the Court in Geftman emphasized the ability of related parties to structure a transaction to be loan in form, sale in substance, the converse is no less true. Related parties are, as here, perfectly capable of structuring a loan transaction so as to create the appearance of a sale. See Geftman, 154 F.3d at 68; see also TIFD III-E, Inc. v. United States, 459 F.3d 220, 233 (2d Cir. 2006). 43 The Swiss subsidiaries, and not Schering-Plough, were entitled to receive leg income streams from ABN. Id. at 17; Only the Swiss subsidiaries had enforcement rights relative to the assigned receive leg streams. Id.; The Swiss subsidiaries only mode of receiving payment was through ABN. Id. at 18; Schering-Plough was not obligated to reimburse the Swiss subsidiaries if ABN failed to make receive leg payments. Id.; Schering-Plough was not obligated to generate or produce revenues that would be used to make payments to the Swiss subsidiaries under the swap receive legs. Id. at 19; Schering-Plough was not obligated to make Swiss subsidiaries whole if ABN s creditworthiness declined or if the floating rate indices applicable to the receive leg declined. Id.; The Swiss subsidiaries were not obligated to return any payments or excess payments to Schering-Plough once a certain amount had been received. Id.; The assignments imposed no restrictions or covenants on ScheringPlough s operations. Id. at 20; The assignments did not contain loan covenants binding ScheringPlough. Id.; The Swiss subsidiaries, and not Schering-Plough, had the right to sell or dispose of the assigned swap receive legs. Id.; There were no other objective factors of a loan, such as promissory notes, collateral, security agreement, or a loan agreement. Id. at 21. These factors fit into one or more of three general categories: (1) the obligations (or purported lack thereof) of Schering-Plough vis-Ã -vis the Swiss subsidiaries; (2) the rights of the Swiss subsidiaries (or purported lack thereof) vis-Ã -vis Schering-Plough and ABN; and (3) the stated terms of the transaction documents. But these three categories are easily manipulated by the related parties appearing on both sides of the swap-and-assign transactions. The evidence demonstrates that the transaction documents were drafted specifically to avoid reference to corresponding rights and obligations between Schering-Plough and the Swiss subsidiaries. 44 The 17-factor argument lodged by Schering-Plough over-emphasizes the form of the swap-and-assign transactions. Moreover, the cases Schering-Plough cites that ostensibly support these factors did not involve meticulously crafted, tax-motivated, related-party transactions between a parent corporation and its controlled subsidiaries. See Pl. Pre-Trial Br. at 14-21 (citing E. Coast Equip. Co., 222 F.2d at 677 (concerning transactions between taxpayer company and two separate companies); Burford-Toothaker Tractor Co. v. United States, 262 F.2d 891, 892 (5th Cir. 1959) (concerning whether funds obtained by Taxpayer from its bank under a transfer of customer s conditional installment sales contracts and promissory notes was a part of borrowed capital under § 439(b)(1), 26 U.S.C.A. Excess Profits Taxes, § 439 ); United Surgical Steel Co. v. Comm r, 54 T.C. 1215, 1218 (T.C. 1970) (concerning assignment to a bank of receivables from installment sales as collateral for a loan); Town & Country Food Co. v. Comm r, 51 T.C. 1049 (T.C. 1969) (concerning sale of goods on installment contract and method of reporting income)). Schering-Plough asks the Court to hold that the transactions were sales because it owed no direct repayment obligation and because the transaction documents denominated them as sales. This gives undue weight to interposing ABN between Schering-Plough and the Swiss subsidiaries. The related-party structure of the transactions establishes that Schering-Plough controlled the direction of the cash flows consistent with an indirect obligation to repay the Swiss subsidiaries for the lump sums advanced. Finally, any lack of formal documentary evidence for these transactions gives way to considerable de facto (but no less objective) indicia of Schering-Plough s indebtedness to the Swiss subsidiaries. 45 a. Contemporaneous Intent and Objective Indicia As to intent, the determinative fact is the intention as it existed at the time of the transaction. Saigh v. Comm r, 36 T.C. 395, 420 (T.C. 1966). First, the 1990 ABN credit proposal, dated before the execution of the swap-and-assign transactions, unambiguously proves that, at the very least, ABN knew and believed that its role in the transactions would be that of an accommodation party to a loan. See Ex. 238. The credit proposal shows that the transactions that ABN believed to be loans were undertaken by ABN s Financial Engineering Group at the behest of its client, Schering-Plough. Second, Daniel Filiberto s notes show that ScheringPlough itself understood the transactions to be loans, but treated them as sales because the tax [department] could not have us treat [the transactions] as . . . loan[s]. See Ex. 168 at SP09471. Third, Schering-Plough s brand-new investment policy, in which certain investments but not intercompany loans had to be approved by the Finance & Audit Committee, did not prompt scrutiny of the swap-and-assign transactions. See Ex. 75. This confirms that Schering-Plough and ABN did not truly believe the transactions were accomplishing anything more than intercompany indebtedness.20 Also present are objective indicia of a loan structure. In addition to the parties contemporaneous intent, the presence vel non of notes or other debt instruments, security or collateral, interest charges, repayment schedules or deadlines, book entries recording loan 20 In any event, it is immaterial if one might view the evidence of actual intent as ambiguous. Where there is an absence of evidence from which to find that a loan was intended or . . . in the absence of any showing of the true intention at the time of the transaction, Saigh, 36 T.C. at 419-20, a court is entitled to presume that the actual economic effect of the transactions reflects the parties true intent. See id. (citing Waldheim v. Comm r, 244 F.2d 1, 5 (7th Cir. 1957)). Thus, even if the evidence as to intent was ambiguous, having already found that the cash flows from the transactions mimic a loan structure, the Court may presume that the parties contemporaneously understood the transactions to be loans. 46 balances or payments, actual repayments, or any other factors indicative of an unconditional obligation to repay are sufficient indicators of a loan structure. Geftman, 154 F.3d at 70. As previously discussed, the difference between the gross amounts routed through ABN to the Swiss subsidiaries and the lump sums initially repatriated is the economic equivalent of interest repaid to the subsidiaries. Such a conclusion is supported by Schering-Plough s report to its Finance & Audit Committee that, as a result of the swap-and-assign transactions, the Swiss subsidiaries owned financial assets which earned interest. Ex. 212 at P000114 (emphasis added). Also consistent with this assessment, Filiberto created an amortization schedule for the 1992 swap-and-assign transaction which calculated total payments based on an assumed receivable interest rate of 4%, part of which comprised periodic interest payments using a constant effective interest rate of 2.9284%. Ex. 195 at P014781. Thus, the schedule shows that Schering-Plough could expect to make annual $37 million progress payments in which the interest component would successively decrease and the principal component would simultaneously increase until the principal balance amortized down to zero in 2012. Aside from the interest payments reflected in the amortization schedule, the transaction documents themselves also envision specified repayment schedules (six-month receive leg intervals for the 1991 assignments and annual receive leg intervals for the 1992 assignments). See Geftman, 154 F.3d at 70; ASA Investerings P ship v. Comm r, No. 96-27320, 1998 WL 513167, at *17 (T.C. Aug. 20, 1998). Given these repayment schedules, the transactions were also accompanied by definite principal balances, see Geftman, 154 F.3d at 68; TIFD III-E, Inc. v. United States, 459 F.3d 220, 236 (2d Cir. 2006), and fixed maturity dates (the dates on which the repatriated sums (i.e., principal) amortized down to zero), see Geftman, 154 F.3d at 68, 70-71; ASA Investerings, 1998 WL 513167 at *17. 47 The fact that Schering-Plough consistently, materially, and timely made repayments to the Swiss subsidiaries (through ABN) further demonstrates an objective lending arrangement. Geftman, 154 F.3d at 70-72. Unlike cases where the existence of partial repayment is insufficient for the finding of a loan, in this case the periodic repayments were proportionate to the total amount advanced, and Schering-Plough ultimately made payment in full. See id. at 7172 (citing In re Uneco, 532 F.2d at 1204); Georgiou v. Comm r, 70 T.C.M. (CCH) 1341, 1351 (T.C. 1995)). Nor is this a case where Schering-Plough made repayments long after the initial transfer (indicating that such repayments did not correspond to an established schedule). See id. (citing Gilbert v. Comm r, 74 T.C. 60, 65-66 (T.C. 1980)21). For instance, Schering-Plough consistently made repayments to ABN of $37 million every year on the 1992 swap-and-assign transaction until it eventually repurchased the receive leg from Scherico. Each payment independently represented approximately 8.33% of the total amount advanced. See Ex. 195 at P014781. Additionally, by repurchasing the receive leg from Scherico, Schering-Plough repaid the repatriated lump sum in full, plus interest (further distinguishing this case from Uneco, Gilbert, and Georgiou); as noted above, this is consistent with a typical accelerated repayment of an outstanding debt. Schering-Plough argues that because its payments to ABN did not necessarily equate with the streams flowing from ABN to the Swiss subsidiaries, such pay leg streams cannot be considered repayments to the subsidiaries. 21 But any differences between the respective In Gilbert, the Tax Court emphasized the fact that the transferee initiated repayment 2½ years after the transfer had been made, thus suggesting that no repayment schedule truly existed. Gilbert, 74 T.C. at 65-66. Here, Schering-Plough did not begin making repayments until year 6 of the assignments (when the Swiss subsidiaries began receiving income from the assignments). However, unlike Gilbert, the repayments were made pursuant to a regimented and pre-ordained schedule created at the time the lump sums were advanced by the subsidiaries. The Court does not read Geftman or Gilbert to defeat proper characterization of a lending arrangement merely because repayment was deferred to later years by prior agreement. 48 receivable amounts were due to the spread between the fixed/floating interest rate indices assigned to each side of the transactions. The fact that ABN might not make payments to the Swiss subsidiaries identical to those which it received from Schering-Plough does not change the character of the transaction. The Court agrees with the government that a floating rate loan . . . is nonetheless a loan. Def. Pre-Trial Br. at 41-42. It is undisputed that intercompany loans between Schering-Plough and its subsidiaries were not accompanied by formal documentation, but this fact is not determinative. Robert Clarke conceded that a loan could be written up and disguised as a sale. (1/23/08 Clarke Test. 176:12-178:12.) Importantly, he admitted that bona fide loans sometimes are not documented by written loan agreements and that not all loans have collateral. (1/23/08 Clarke Test. 174:24175:10.) He further acknowledged that he was not aware that Schering-Plough in fact did not use customary loan documentation for inter-company (e.g., subsidiary to parent) loans. (1/23/08 Clarke Test. 179:12-14; 182:4-7.) Based on the record before it, the Court rejects the argument that because the swap-and-assign transactions were not accompanied by formal loan documentation, they cannot be loans. If that were the case, then there has never been a loan between Schering-Plough and its foreign subsidiaries. No principle of taxation exists that permits related parties to drive the nature of a transaction simply by avoiding the execution of a formal document denominating a loan as such. Thus, this case is unlike others where courts have placed significance on the absence of formal documentation in a particular transaction. See, e.g., Baird v. Comm r, 43 T.C.M. (CCH) 1173, 1180-81 (T.C. 1982); Donisi v. Comm r, 26 T.C.M. (CCH) 327 (T.C. 1967), aff d, 405 F.2d 481 (6th Cir. 1968). Instead, the Court attaches little weight to the fact that ScheringPlough and the Swiss subsidiaries did not evidence the transactions with notes or other debt 49 instruments, particularly when they never executed such documents for similar endeavors. The presence of customary loan documentation is not a prerequisite to a bona fide loan, but only one consideration under a totality analysis; the absence of [such] . . . documentation merely left [the government] with one less string to strum for the factfinder. Crowley v. Comm r, 962 F.2d 1077, 1082 (1st Cir. 1992); see also Frank Lyon Co. v. United States, 435 U.S. 561 (1978) (citing Comm r v. Tower, 327 U.S. 280, 291 (1946) ( The Court has never regarded the simple expedient of drawing up papers as controlling for tax purposes when the objective economic realities are to the contrary. ); Helvering v. Lazarus & Co., 308 U.S. 252, 255 (1939) ( In the field of taxation, administrators of the laws, and the courts, are concerned with substance and realities, and formal written documents are not rigidly binding. ); Gilbert, 74 T.C. at 65 ( [I]t is clear that a valid debt may exist even where no formal debt instrument exists. ) (citing Joseph Lupowitz Sons, Inc. v. Comm r, 497 F.2d 862, 868 (3d Cir. 1974)). The unsecured character of the transactions is also not determinative. Had the funds been advanced to an unrelated, high-risk transferee in an uncertain market, as was the case in Hambuechen v. Commissioner, 43 T.C. 90, 103 (T.C. 1964), the Court might face a different question. But here, as the Court discusses below, any risk of a Schering-Plough default was negligible.22 Collateral was therefore not a necessary component of the transactions, and its absence does not carry the weight that Schering-Plough gives it. Nor does the lack of a direct obligation (on paper) to the Swiss subsidiaries defeat the transactions proper characterization as loans. Schering-Plough surely had a repayment obligation, albeit to ABN. And the Swiss subsidiaries surely had an unconditional intent to recoup the money they advanced to their 22 The Court uses default here merely in terms of solvency or ability to make payments, and not contractual default under the transactions as a result of a downgraded credit rating. 50 parent. Simply by breaking up the obligations using an intermediary, Schering-Plough cannot defeat the reality that its subsidiaries advanced it a lump sum and were later funneled money via a third party. See Mapco, 556 F.2d at 1111-12, 1115-17 (discussed infra). Supporting the Court s conclusion is the method by which Schering-Plough financed the transactions, routing the Irish Cash used to purchase the assignments from Limited to Scherico (and ultimately back to Schering-Plough). As the government points out, Treasury Regulation § 1.956-1T(b)(4) states: [A] controlled foreign corporation will be considered to hold indirectly . . . investments in U.S. property acquired by any other foreign corporation that is controlled by the controlled foreign corporation, if one of the principal purposes for creating, organizing, or funding (through capital contributions or debt) such other foreign corporation is to avoid the application of section 956 with respect to the controlled foreign corporation. Treas. Reg. § 1.956-1T(b)(4). For this reason, the parties have stipulated that should the Court find the transactions to be, in substance, loans, then the loans to Schering-Plough are from Limited and not Scherico. (JTS, Proc. Agreement, ¶ B.2) No Schering-Plough witness could explain why the untaxed Limited money was re-routed through Scherico to purchase the receive leg assignments. If Schering-Plough s true motive was to repatriate money from an entity whose cash had already been taxed in order to create the appearance that PTI was used to acquire the receive legs, then the re-routing could and did disguise what was going on. But if the transactions were substantively sales, purportedly relieving the offshore cash of Subpart F taxation, why would it matter whether the cash was PTI or not?23 Schering-Plough has not adequately answered this question. 23 The government also points out that in a further effort to disguise the source of the funds used to acquire the assignments, Schering-Plough neglected to account for the shifting of funds on Schedule M to Form 5471 (the form used to document certain transactions between CFCs and related parties) for Scherico and Limited. As a result, 51 Dr. David LaRue, an expert witness for the government, testified that form takes on importance when unrelated parties act in their own independent economic self-interest because those parties have countervailing economic interests and work to protect those interests in drafting governing agreements. (2/27/08 LaRue Test. 57:5-19.) But where a parent and a subsidiary enter into a transaction, Dr. LaRue testified that the form and labels adopted may not comport with the intrinsic economic nature of the transaction because the absence of countervailing self-interest allows the parties to create whatever appearances they choose, regardless of economic reality. (2/27/08 LaRue Test. 60:7-61:15.) This commonsense explanation exposes the economic reality of the transactions. Schering-Plough s expert witness Andrew Onslow emphasized the absence of direct obligations between Schering-Plough and the Swiss subsidiaries. He discussed the obligations of Schering-Plough, ABN, and the Swiss subsidiaries under the swap-and-assign contracts according to English Law. (1/22/08 Onslow Test. 22:18-21.) Onslow also testified more generally about whether and how English law would govern the swap contracts. (1/22/08 Onslow Test. 22:12-18.) Onslow opined that the obligations for the swap and receive legs were entirely independent of each other. (1/22/08 Onslow Test. 22:8-32:9.) He explained that after the assignments became effective, Schering-Plough did not have repayment obligations to the Swiss subsidiaries because the assignment contracts contained no language that would suggest such an obligation. (1/22/08 Onslow Test. 28:8-12.) He also stated that Schering-Plough was not contractually obligated to pay the Swiss subsidiaries if ABN failed to make the payments on the receive leg of the transaction. (1/22/08 Onslow Test. 28:22-29:2.) In addition, Onslow there was no way of knowing that the re-routing of cash had occurred until 1996, when Schering-Plough began reporting income from the swap-and-assign transactions. (Exs. 2064-2065, 2067-2068, 638; 2/27/08 LaRue Test. 162:18-164:23, 166:1-4; 2/28/08 LaRue Test. 5:20-7:8; 9:16-11:1, 12:5-23.) 52 testified that the assignments required ABN to make payments to the Swiss subsidiaries even if Schering-Plough did not make its payments. (1/22/08 Onslow Test. 31:2-10.) Finally, Onslow stated that the Swiss subsidiaries were not obligated to return any payments to Schering-Plough if the ABN payments exceeded the value of the supposed loan. (1/22/08 Onslow Test. 31:1118.) On cross-examination, however, it became clear that Onslow did not consider any underlying subjective purpose that the parties may have had that is not expressed in the contractual documents. (1/22/08 Onslow Test. 38:13-17.) Onslow s testimony also did not take into account the complete control of the Swiss subsidiaries exercised by Schering-Plough. (1/22/08 Onslow Test. 37:19-38:2.) Onslow s testimony is competent to explain English contract law, but the Court believes it is limited in utility because it failed to recognize the purpose for which the transactions were specifically structured. Moreover, the Swiss subsidiaries were unconditionally empowered to enforce ultimate repayment by the put option that they held on the 1991 assignment (even though it was ScheringPlough that dictated when the option would be exercised ).24 The put option gave the subsidiaries the unconditional power to recoup and the exercise of the put option imposed the unconditional obligation on Schering-Plough to repay the lump sums repatriated. ScheringPlough posits that it had legal obligations to its subsidiaries under the put option when it argues infra that the transactions had sufficient economic effect so as to be respected. See Pl. Post-Trial Br. at 33. But Schering-Plough cannot cherry-pick the existence of obligations to its liking. This 24 While there was no put option granted on the 1992 assignment, the fact that Schering-Plough did in fact repurchase the 1992 assignment from Scherico is also indicative of an implied obligation to repay. 53 material inconsistency regarding obligations reveals both the true economic nature of the transactions and the tax-driven motivations behind them. Schering-Plough emphasizes that there is no transaction document that explicitly obligates it to repay the Swiss subsidiaries. But the put option granted as part of the assignment meant that the Swiss subsidiaries did have the contractual right to enforce ultimate repayment, and Schering-Plough did have a corresponding obligation to repay the subsidiaries upon exercise of the put. The absence of an express repayment obligation can be seen as part of the sophistication of the overall arrangement; it is not determinative. In an analogous case, the Second Circuit reversed a district court determination that a partnership s allocation of income to a Swiss bank in return for an initial investment merely reflected the bank s equity interest in the partnership. TIFD III-E, 459 F.3d at 237-38. The Second Circuit vacated the district court s determination, relying, in part, on the fact that the partnership agreements empowered the bank to terminate the agreement and recover the investment at an agreed rate of return. Like TIFD IIIE, Schering-Plough s subsidiaries could force repayment under the put. The Court concludes that Schering-Plough s efforts to structure the transactions as sales fail to overcome the significant objective indicia of the company s indebtedness to the Swiss subsidiaries. b. Conduit Theory How does ABN fit into the picture? The government argues that ABN was a mere passthrough in routing payments from Schering-Plough to its subsidiaries. It points to the conduit theory of the substance over form doctrine, under which a court may disregard an entity if it is a mere conduit for the transaction at issue. Def. Post-Tr. Br. at 42 (citing Enbridge Energy Co., 54 Inc. v. United States, 553 F. Supp. 2d 716 (S.D. Tex. 2008)); see also Comm r v. Court Holding Co., 324 U.S. 331, 334 (1945). Explaining Schering-Plough s position, David Ross argued against ABN s status as a conduit, and concluded that the swap assign transactions were, as a matter of economics, asset sales and not dividends or loans. (1/24/08 Ross Test. 56:22-59:9.) Ross first examined the common dictionary definition of a conduit, defined as a means by which something is transmitted. (1/24/08 Ross Test. 64:23-65:19.) Ross testified that ABN was not a conduit under the standard dictionary definition of the word because under the terms of the swap there was no requirement that ABN transmit Schering-Plough s payments through to the Swiss subsidiaries. (1/24/08 Ross Test. 65:20-66:10.) Ross also agreed with Andrew Onslow s testimony regarding the independence of ABN s obligation to pay the subsidiaries and Schering-Plough s obligation to pay ABN. (1/24/08 Ross Test. 68:7-12.) He stated that if ABN could not pay the Swiss subsidiaries, it was under no obligation to transmit the money it received from Schering-Plough to the Swiss subsidiaries to fulfill its obligation, thereby defeating a conduit characterization. (1/24/08 Ross Test. 68:17-70:14.) Ross testified that ABN was not a conduit under the banking definition of a mortgage conduit. (1/24/08 Ross Test. 65:20-66:21.) He testified that mortgage conduits make distributions only if they receive payments first, and that they also separate and isolate the assets involved. (1/24/08 Ross Test. 76:5-77:18.) In this case, ABN commingled the payments from Schering-Plough, and commingled its obligations. (1/24/08 Ross Test. 77:19-78:12.) Ross suggested that, unlike the pass-through requirement in the banking definition of a conduit, ABN was not required to pass through Schering-Plough s payments if it could not meet its obligations to the subsidiaries. (1/24/08 Ross Test. 68:7-17.) Ross pointed to responses to interrogatories 55 and stipulations that confirmed these facts. (1/24/08 Ross Test. 67:6-70:14.) Further, Ross explained that ABN and Schering-Plough s obligations were different in kind because ABN and Schering-Plough frequently made payments of different values and, in the second swap, ABN made payments annually while Schering-Plough made them monthly. (1/24/08 Ross Test. 72:474:15.) Ross also testified that ABN s credit risk is irrelevant to the question of whether it is a conduit because the transaction s independent obligations alone make it impossible for ABN to be a pass-through, regardless of its credit risk. (1/24/08 Ross Test. 79:8-22.) Under a conduit analysis, courts have applied several related factors: (1) whether there was an agreement between the principals to do a transaction before the intermediary participated; (2) whether the intermediary was an independent actor; (3) whether the intermediary assumed any risk; (4) whether the intermediary was brought into the transaction at the behest of the taxpayer; and (5) whether there was a nontax-avoidance business purpose to the intermediary s participation. Enbridge, 553 F. Supp. 2d at 730. The Court first considers whether the Schering-Plough entities (parent and subsidiaries) had agreed to perform a transaction before involving ABN. The facts show that ScheringPlough s corporate treasury department was empowered to direct the subsidiaries in their management of funds and investments. (See 1/15/08 Luciano Test. 82:7-14; 1/17/08 Ludwig Test. 69:1-70:23.) Schering-Plough decided on Merrill Lynch s advice, without input from ABN to enter into the swap transactions with its Swiss subsidiaries. As to the first factor, the evidence indicates that ABN had no substantive involvement in Schering-Plough s initial decision to effect the swap-and-assign transactions. The Court finds the second Enbridge factor whether ABN was an independent actor to encompass broadly the entire conduit analysis. To the extent that the factor can be broken 56 down, however, and insofar that ABN is a corporate entity distinct from the Schering-Plough family, this factor leans slightly in favor of the taxpayer. But ABN s participation, albeit in its own self-interest, was limited to facilitating the swap-and-assign transactions for its client, Schering-Plough, an endeavor which earned it ten basis points annually. The third factor whether the intermediary assumed risk is a pivotal point, and one on which Schering-Plough rests its entire argument.25 Because Schering-Plough s argument focuses solely on this factor, the Court discusses it more fully below. In short, the Court finds that ABN minimized its risk using: (1) mirror swaps; (2) credit default provisions permitting it to escape the transactions; and (3) other contractual provisions designed to limit its regulatory capital requirements. Based upon those facts and the testimony of Johannes den Baas, as well as Finard, Kawaller and Parsons, the Court concludes that ABN s real risk was de minimis. The fourth factor related to the first goes to the manner in which ABN became involved in the transactions. The evidence demonstrates that ABN became involved with the swap transactions because Schering-Plough wanted to repatriate its foreign-earned cash in a taxefficient manner, not because it had an independent business motivation for itself. Den Baas, who served as ABN s negotiator in the transactions, testified that Merrill Lynch approached ABN to participate in the transactions as an intermediary between Schering-Plough and Merrill Lynch. (1/23/08 Den Baas Test. 31:11-18; 32:4-6.) Thus, it cannot seriously be disputed that Schering-Plough, as taxpayer, with the aid of Merrill Lynch, as counselor, brought ABN into the transactions for its own tax purposes. 25 The government argues that Schering-Plough erroneously concludes that ABN is not a conduit solely on the [risk] factor, and that it conveniently ignores the other [s] . . . . Def. Post-Trial Br. at 42. While the Court will not rule out the possibility that a conduit argument can be rejected solely on the showing of significant risk, it agrees with the government that application of the other four Enbridge factors here militates in favor of finding that ABN was a conduit. 57 The fifth factor is whether there was any non-tax avoidance purpose in including ABN. Schering-Plough could not have executed the purported tax-efficient transactions without a thirdparty intermediary to act as counterparty to the swap. Absent ABN, Schering-Plough s obligations argument its central premise in this case would not exist. And Schering-Plough could have just as easily, for significantly lower costs, effected a direct loan from the Swiss subsidiaries to the parent or it could have paid a dividend between those entities. ABN s presence in the transactions cannot be explained without referencing the tax implications Schering-Plough faced without the intermediary s participation. i. Risk As stated above, Schering-Plough argues that trial testimony revealed that ABN in fact bore material risk in the swap transactions, and that ABN therefore could not be a conduit. As to credit risk, the Court does not doubt that ABN carried some limited measure of credit risk in the transactions, but none that rose to the level of significance to be material. To show how it could have been a risky prospect, Schering-Plough points to the jump-to-ruin scenarios torn out of the front pages of newspapers in 2008, such as Bear Stearns s collapse or the Pacific Gas and Electric bankruptcy of 2001, but fails to account for the unique risks posed to those companies business models, the former fatally laden with risky bets on subprime mortgages and the latter unable to purchase energy and resell it for a profit. See Pl. Post-Trial Br. at 13-14. Schering-Plough s business model differs qualitatively from those entities. It is a diversified conglomerate in the pharmaceutical and consumer products sector. In many cases, people cannot live without Schering-Plough s products, and the prospects for a jump-to-ruin scenario were decidedly slim. More convincing is the government s statistical evidence, unrebutted by Schering-Plough, that ABN s risk of losing money due to a Schering-Plough 58 default registered at 0.0005%. (Ex. 2133 at 33-37; 2/20/08 Finard Test. 28:7-40:7.) Even were the Court to accept the theoretical possibility of a jump-to-ruin scenario, the historical ratings data published by Moody s from 1982 to 2005, and put forth by the Government, estimated the risk of loss within a brief two-month window highly unlikely. (2/20/08 Finard Test. 44:16-19.) Robert Clarke also acknowledged that the 60-day credit ratings trigger significantly reduced ABN s credit risk. (1/23/08 Clarke Test. 165:14-166:5.) Schering-Plough argues that ABN hedged substantial interest rate risk only to assume material credit risk by entering into mirror swaps with Merrill Lynch. Thus, Schering-Plough asserts that the limitation of ABN s interest rate risk came at the expense of taking on greater credit risk by exposing itself to Merrill Lynch s ability to pay. To be sure, Schering-Plough faced some credit exposure to Merrill Lynch as a result of the mirror swaps. But given the fact that it was also receiving substantial remuneration for its involvement, the Court rejects the proposition that any additional credit risk posed by Merrill Lynch rises to a material level. Schering-Plough also argues that ABN could not have been a conduit because it was exposed to litigation and reputation risks over the suitability of the swaps or for failure to disclose financial risks associated with the swaps to Schering-Plough. Pl. Post-Trial Br. at 15. Specifically, Schering-Plough argues that ABN believed that it might be sued by ScheringPlough, and that such a lawsuit would harm its reputation. However, Schering-Plough has presented no evidence of litigation risk or possible reputational damage over and above what it faced daily as an ongoing business. Consistent with Finard s testimony, whatever risk ABN faced of being sued (or any resulting reputational damage) is too minimal and attenuated from meaningful risk factors to register in the conduit analysis. (2/20/08 Finard Test. 51:3-21.) 59 Presenting the testimony of Robert Clarke, Schering-Plough maintains that the capital set-aside requirements fixed by the Basel Accord evidenced material risk, thereby proving ABN was not a conduit. Schering-Plough argues that the compensation ABN received from participating in the swap-and-assign transactions fell short of its internal 10% return-on-capital guideline due to the capital set-aside requirement. Clarke testified that the capital set-aside requirement caused ABN to experience significant economic consequences from the transactions. (1/23/08 Clarke Test. 126:11-22.) After the assignment, the Basel Accord required ABN to set aside capital equal to 8% of the notional principal. (1/23/08 Clarke Test. 122:11-13.) According to Clarke, this set-aside created opportunity cost for ABN, particularly because it could have used that capital for more lucrative purposes. (1/23/08 Clarke Test. 126:8-127:9.) Schering-Plough points to 1999, a year in which the Basel Accord required ABN to set aside $72 million for the swap transactions, equating to a $7.2 million internal profit projection for ABN. Pl. Post-Trial Br. at 17. Because ABN was only projected to make $32.4 million over the entire duration of the swaps, an amount significantly below its internal profit margin guidelines, Schering-Plough argues that ABN suffered as a result of the swaps. It claims that ABN s lawyers and financial experts struggled for a decade to try to solve these capital solvency problems. Id. at 18. Den Baas testified that the transaction documents permitted ABN to alleviate the effects of the regulatory capital requirement by assigning its rights and responsibilities under the transactions to a non-consolidated subsidiary. (1/23/08 Den Baas Test. 18:15-20, 54:8-21.) By making such an assignment to a special purpose vehicle, ABN s regulatory capital requirements would have been greatly limited, if not eliminated. (1/23/08 Den Baas Test. 56:5- 60 15.) Den Baas believed that ABN s failure to assign the obligations could only be attributed to oversight. (1/23/08 Den Baas Test. 56:24-58:15.) Clarke rejected Den Baas s testimony, suggesting that ABN could not have freed up the capital by assigning its obligation because the proposed assignment was within the ABN family, and such an assignment would not change ABN s credit exposure. (1/23/08 Clarke Test. 131:12132:7.) Based on his experience, Clarke believed that regulators would have found the transaction internal and required the capital freeze anyway. (1/23/08 Clarke Test. 133:6-18.) It was clear at trial, however, that Clarke s conclusion arose primarily from the fact that ABN did not take advantage of the special purpose vehicle. (1/23/08 Clarke Test. 131:18132:24, 167:3-168:20.) Despite confess[ing] that he was unsure exactly how the special purpose vehicle would function, he surmised that the ABN team must have come to the conclusion that it would not work because such an option was never invoked. (1/23/08 Clarke Test. 132:1-3, 20-24.) On cross-examination, his testimony regarding the efficacy of the special purpose vehicle was equivocal at best: Q. Was there any testimony that you are aware of that indicated that ABN was paying attention to this [capital] problem before the time period had expired? A. Before which time period? Q. The five year period . . . they had to exercise the special purpose vehicle provision? ... A. It depends on when that was set up. I can t it s hard for me to answer the question if I don t have the time frame. Q. But you re rendering an opinion here today that it was possible for ABN to take this step and that they chose not to[,] is that correct? 61 A. That s not my opinion at all. I have no idea whether it was possible for them to do it. (1/23/08 Clarke Test. 168:1-20.) Given the above, Clarke s post hoc, ergo propter hoc determination cannot be credited. More convincing was ABN s negotiator with respect to the provision Den Baas who opined that ABN s special purpose vehicle option would indeed have worked. (1/23/08 Den Baas Test. 55:12-56:4; 1/23/08 Clarke Test. 133:1-15; 167:3-11.) That ABN chose not to invoke it (or fell asleep at the switch ) is not a reason why the solution to ABN s capital commitment problems would not have worked. (1/23/08 Clarke Test. 167:19.) The evidence demonstrates that rather than being indicative of risk, the proportion of capital that the Basel Accord required ABN to set aside represented only an opportunity cost for ABN engaging in the transaction. (1/22/08 Douglas Test. 99:16-100:6; 1/23/08 Den Baas Test. 16:5-10.) The amount to be set aside was based on general categories of transactions and not the actual terms of Schering-Plough s swaps. In particular, as Den Baas explained, the regulatory capital requirements do not take into consideration specialized provisions, such as the credit trigger. (1/23/08 Den Baas Test. 16:5-10.) Further, ABN opted for certainty in using that capital not risk in that it expected 10 basis points from Schering-Plough for the duration of the swaps, a fee that totaled over $1.5 million on both. (1/24/08 Ross Test. 82:13-24; 1/24/08 Taylor Test. 38:3-39:5.) While perhaps not optimal under ABN s internal profit guidelines, this figure represents a safe and not insubstantial return. Schering-Plough argues that Frank Lyon Co. v. United States, 435 U.S. 561 (1978) conclusively proves that commitment of capital is adequate to show that an entity is not a conduit. But the argument overemphasizes the weight of the Supreme Court s capital- commitment rationale. In Frank Lyon, the Supreme Court reviewed whether a three-party transaction, in which the taxpayer entered into a sale and leaseback of a new bank building, 62 permitted it to deduct certain expenses and depreciation from its ownership of the building. Frank Lyon, 435 U.S. at 563-69. Specifically, the taxpayer, which was selected among a number of potential investors, obtained financing to purchase the building (then under construction) from the bank that was developing it, and then immediately leased it back to the bank. Id. at 564-65. When the taxpayer sought to deduct the interest on the loan used to finance the purchase, the monthly depreciation for the building, and certain related expenses, the IRS determined that the taxpayer was not the owner of the building for tax purposes, arguing that the taxpayer was merely a conduit for mortgage payments from the bank to the construction lender. Id. at 568-71. The Supreme Court ultimately disagreed, finding that the taxpayer s sole liability on the loan from the construction lender established that it did indeed legitimately own the building, and that the rental payments from the bank to the lender were not pass-through payments. Id. at 578-80. It is true, of course, that the Frank Lyon Court believed that [t]o the extent that [the taxpayer] . . . used its capital in this transaction, it [wa]s less able to obtain financing for other business needs. Id. at 577. But that was not the extent of the Court s analysis. Far from zeroing in on capital commitment, the Court relied on no less than 23 case-specific factors in finding that the taxpayer had far the better of the case. Id. at 582-83. Several of the factors relied upon by the Frank Lyon Court are noticeably present (or absent, as the case may be) here, except that they suggest in this case that the government has the better argument: (1) the bank s inability, as a matter of legal restraint and as effected by regulators, to carry its building plans into effect by conventional financing26; (2) the presence of several finance organizations 26 The Supreme Court emphasized on a number of occasions that the sale and leaseback arrangement was not the first option that the bank entertained. Instead, its original financing arrangement was blocked by state and federal regulators. Frank Lyon, 435 U.S. at 563-64, 575, 577-78. For that reason, the Court thought it significant that the taxpayer was not intending that the interests involved were allocated in a way other than that associated with a saleand-leaseback. Id. at 578. As this Court has found above, this rationale is decidedly not the case here. 63 seriously interested in participating in the transaction and in the resolution of the bank s predicament; (3) the submission of formal proposals by several of those financiers; (4) the competitive bargaining process that ensued; (5) the bona fide character of the negotiations; (6) the nonfamily and nonprivate nature of the entire transaction ; and (7) the absence of any differential in tax rates and of special tax circumstances for one of the parties. Id. Frank Lyon involved a sale and leaseback, a common transaction, unlike the swap-andassign transactions here. (See 2/20/08 Finard Test. 7:10-23) (stating that the assignment of a single stripped leg of a swap was very, very unusual ); see also In re Moreggia & Sons, Inc., 852 F.2d 1179, 1186 (9th Cir. 1988) (noting that a sale and leaseback is a common financing scheme ). Conventional financing was freely available for Schering-Plough to meet its objectives, but it did not undertake those taxable transactions. Furthermore, unlike the relatedparty transaction between Schering-Plough and its subsidiaries, the Supreme Court expressly stated that the transaction in Frank Lyon was not a familial one between counterparties with the same incentives, but one that involved formal and competitive bidding between several disinterested investors. Frank Lyon, 435 U.S. at 564, 575. The Supreme Court did not make a universal statement in Frank Lyon that every time capital is committed the intermediary ipso facto cannot be a conduit. Even crediting Schering-Plough s assertion that the special purpose vehicle could not have prevented an ABN capital commitment under the Basel Accord, the Court finds that, on balance, the related-party engineering of the transaction and ABN s otherwise minimal risk exposure befit a conduit characterization.27 27 To the extent the Supreme Court found in Frank Lyon the existence of a three-party structure indicative of a legitimate transaction, see id. at 575, 582, the sale-and-leaseback transaction arrangement there does not parallel the transactions this Court now scrutinizes. See Mapco, discussed infra (finding a loan structure in a three-party transaction). In any event, Schering-Plough s affiliation with the Swiss subsidiaries weighs against finding 64 At bottom, the facts demonstrate that ABN s risk was not so significant to alter its character as a pass-through for the swap-and-assign transactions. Furthermore, the other Enbridge factors indicate that ABN s accommodation role in the transactions existed solely to break up payment flow and obligations. This is insufficient to establish a bona fide participatory role. The Court concludes that ABN was a mere conduit for the transactions, further supporting its holding that the swap-and-assign transactions were, in substance, loans. c. Mapco v. United States The Court finds the three-party transaction in Mapco Inc. v. United States, 556 F.2d 1107 (Ct. Cl. 1977), briefly referenced above, instructive. In Mapco, the Court of Claims affirmed a trial court judgment determining that an assignment of future income in exchange for a lumpsum payment was, in economic substance, a loan. Mapco, 556 F.2d at 1112. Mapco, a corporate taxpayer suing the government for a refund, had entered into a sale of future income streams from its business in order to take advantage of significant loss carryovers that would expire if offsetting income was not realized in the relevant taxable years. Id. at 1114-15. To accelerate the future income it expected to earn, Mapco made an arrangement with another company, Rock Creek, in which Rock Creek would make a $4 million lump-sump payment to Mapco in return for 75% of Mapco s future revenues, until such income repaid the $4 million (thus accomplishing the goal of realizing immediate income). Id. at 1115. The transaction documents specifically assigned 75% of future Mapco revenues until such income amounted to $4 million, plus a stated amount of interest and costs. Id. Notably, the documents purporting to sell the future income streams stated that Buyer [Rock Creek] shall look solely to [the assigned] legitimacy in these putative three-party transactions, especially given Frank Lyon s heavy emphasis on a nonfamilial and arms-length arrangement. 65 [r]evenues . . . for liquidation and discharge of the [lump-sum payment] and seller [Mapco] shall not be personally liable for the payment thereof. Id. Under the agreement, Mapco was required to use its best efforts to actually earn the income which was assigned back to Rock Creek. Id. Rock Creek financed the advance payment through a $4 million loan with another entity, Chemical Bank. Id. Linking the three parties together, Mapco used the $4 million lump sum received from Rock Creek to purchase certificates of deposit ( CDs ) through Chemical Bank. Id. The CDs matured in two-week intervals which corresponded with Rock Creek s repayment obligations under the Chemical Bank-Rock Creek loan. Id. at 1115-16. Chemical Bank held the CDs on Mapco s account, and upon being notified by Mapco of the amount of revenue it had earned during the prior period, Chemical Bank would take 75% of such amount from the CD funds it held for Mapco and do the following: (1) credit Rock Creek with its profit margin (the difference between the interest Rock Creek earned on its loan to Mapco and the interest it paid to Chemical Bank); and (2) credit the balance back to itself as repayment of its loan to Rock Creek.28 Id. at 1116. Instead of reporting the future revenues in the years actually earned, Mapco reported the income derived from the sale of the future revenues in the year received, and offset such income against the loss carryovers it had accumulated. Id. Characterizing the transaction as a loan not eligible to be offset against the loss carryovers, the IRS assessed a deficiency against Mapco. Id. The trial court29 agreed with the government, characterizing the transaction as a 28 Rock Creek also pledged its own interest in Mapco s future revenues to secure the loan from Chemical Bank. Mapco, 556 F.2d at 1109. 29 Before the Court of Claims was dissolved in 1982, it was comprised of a trial division and an appellate division. The opinion of the trial court appears, as modified on appeal, in the appellate opinion of the Court of Claims. See Mapco, 556 F.2d at 1112. All quotations from the trial court s opinion are taken from that appellate opinion. 66 loan. Id. In doing so, the court discussed three cases with similar facts; it distinguished Estate of Stranahan v. Commissioner, 472 F.2d 867 (6th Cir. 1973), and analogized Martin v. Commissioner, 56 T.C. 1255 (T.C. 1971), aff d, 469 F.2d 1406 (5th Cir. 1972), and Hydrometals, Inc. v. Commissioner, 31 T.C.M. (CCH) 1260 (1972), aff d, 485 F.2d 1236 (5th Cir. 1973), cert. denied, 416 U.S. 938 (1974). Illuminating the principal difference between Stranahan on the one hand and Martin and Hydrometals on the other, the court explained: [I]n Stranahan the seller of the future income was not under any obligation whatever to produce such income for the benefit of the purchaser, who was compelled to look solely to a third person . . . for the future income that he had purchased, whereas in the Martin and Hydrometals cases the purported sellers of future income were themselves obligated to produce future income for the benefit of the purported buyers. Thus, in substance, the transactions in the Martin and Hydrometals cases were more like loans that were to be repaid, with interest, by borrowers out of their own productive efforts, than like true sales agreements. Here, the plaintiff was obligated to produce the future . . . revenues that Rock Creek was to receive under the agreement . . . . Although the plaintiff did not, in terms, guarantee the repayment of the [$4 million] principal sum to Rock Creek [plus interest and expenses], the plaintiff, in substance obligated itself to such a program. The plaintiff specifically obligated itself to continue to operate [its revenue-generating products] properly during the life of the agreement, and also obligated itself to use its best efforts not only to maintain [its] revenues at their then-current level . . . but to increase such revenues, if possible. Thus, it was certain, as a practical matter, that the plaintiff would itself repay the [$4 million] principal sum [plus interest and expenses] . . . . The agreement . . . was more in the nature of a loan-and-repayment transaction than it resembles a sale-and-purchase transaction. Moreover, it is at least worthy of some consideration that the transaction . . . was contrived solely for income tax purposes. Mapco, 556 F.2d at 1116-17 (emphasis added). The appellate court affirmed. Id. at 1117. As it interpreted the facts of the case, the Court of Claims held that the assignment was actually in the nature of a nonrecourse loan. Id. 67 at 1110. Emphasizing the direction of the cash flows, the court agreed with the government that Rock Creek made a $4 million loan to Mapco, and thereafter received a portion of Mapco s revenues until it was repaid. Id. The court found insignificant that Rock Creek had no rights against Mapco because Rock Creek had a subsidiary entitlement to receive assigned income streams through the CDs on account at Chemical Bank until it was repaid, and Mapco had a corresponding duty to produce the revenues that ultimately repaid Rock Creek. Id. at 1111. Additionally, because Mapco had a consistent record of earning such revenues, any chance of Mapco defaulting appeared remote. Id. Significantly, the court further found that by depositing the CDs which it had purchased from Chemical Bank with the funds loaned to it by Rock Creek, Mapco indirectly guaranteed repayment to Rock Creek (and ultimately to Chemical Bank). Id. The court stated: [W]e do not feel that the [CDs] were merely a handsel to Chemical Bank from Mapco. Quite the contrary, by depositing in Chemical Bank the proceeds of the MapcoRock Creek transaction, Mapco indirectly assured Chemical Bank that it would comply with the . . . agreement. Id. The court concluded, in other words, the consideration was not consideration for a sale but principal loaned to a debtor. Id. The Court finds the reasoning in Mapco sound and applicable here, where the Swiss subsidiaries, like Rock Creek, advanced a lump sum and subsequently were repaid through an assignment of future income streams. Schering-Plough was not directly obligated to repay the transferors, nor did the subsidiary transferors have specific enforcement rights against the parent. But like Mapco, Schering-Plough owed an unconditional obligation to produce the payments (which included interest) that were essentially routed through a conduit intermediary back to the Swiss subsidiaries. See id. at 1117. Also, as in Mapco, the risk of Schering-Plough becoming unable to pay was slight, thus supporting the Swiss subsidiaries ultimate expectation of full 68 repayment of the principal loaned, plus interest. Schering-Plough s continuing payments to ABN the middle-man in the transaction which was responsible for directing payment to the Swiss subsidiaries indirectly and recurrently ensured the subsidiaries that they would recoup the lump-sum advancements. The Court finds that this indirect assurance is identical in substance to Mapco s depositing the CDs which were used to repay the Rock Creek loan with Chemical Bank, the third-party in the transaction. See Mapco, 556 F.2d. at 1111. d. Step Transaction Doctrine The government also submits that Schering-Plough s transactions are taxable under the step transaction doctrine, which is applied to ensure that ostensibly separate transactions do not obscure an improper tax-avoidance scheme. Essentially, [a] given result at the end of a straight path [must] not [be] made a different result because reached by following a devious path. Minn. Tea Co. v. Helvering, 302 U.S. 609, 613 (1938); see also Long-Term Capital Holdings, LP v. United States, 150 F. App x 40, 43 (2d Cir. 2005) ( [U]nder the step transaction doctrine, a particular step in a transaction is disregarded for tax purposes if the taxpayer could have achieved its objective more directly, but instead included the step for no purpose than to avoid U.S. taxes. ). Under the step transaction doctrine, a court will treat[] a series of formally separate steps as a single transaction if such steps are in substance integrated, interdependent, and focused toward a particular result. Penrod v. Comm r, 88 T.C. 1415, 1428 (1987). There are three tests used by courts to apply the step transaction doctrine: the end result test, the interdependence test, and the binding commitment test.30 The step transaction doctrine will operate where the circumstances satisfy only one of the tests. Long 30 The binding commitment test is rarely applied and will not be addressed here. See, e.g., Falconwood Corp. v. United States, 422 F.3d 1339, 1349 n.5 (Fed. Cir. 2005). 69 Term Capital Holdings v. United States, 330 F. Supp. 2d 122, 191 (D. Conn. 2004), aff d sub nom. Long-Term Capital Holdings, LP v. United States, 150 F. App x 40 (2d Cir. 2005). Under the end result approach, courts can collapse or re-order steps after looking to the end result of a series of transactions, so that purportedly separate transactions will be amalgamated into a single transaction when it appears that they were really component parts of a single transaction intended from the outset to be taken for the purpose of reaching the ultimate result. Assoc. Wholesale Grocers v. United States, 927 F.2d 1517, 1523 (10th Cir. 1991) (internal citation omitted); see also Kanawha Gas & Util. Co. v. Comm r, 214 F.2d 685, 691 (5th Cir. 1954) (end result test is useful where there is a series of transactions designed and executed as parts of a unitary plan to achieve an intended result ). The second approach, the interdependence test, centers on the degree of mutual interdependence between the steps taken by a taxpayer. Courts must ask whether the steps [were] so interdependent that the legal relations created by one transaction would have been fruitless without the completion of the series [of steps]. ACF-Brill Motors Co. v. Comm r, 189 F.2d 704, 707 (3d Cir. 1951). According to the Fifth Circuit, the interdependence test is especially appropriate where it is unlikely that any one step would have been undertaken except in contemplation of the other integrating acts. Kuper v. Comm r, 533 F.2d 152, 156 (5th Cir. 1976). Both tests demand that the Court explore how the steps were interrelated and what ultimate purpose they served. Applying this framework, the evidence here establishes that the swaps and subsequent assignments of the 1991 and 1992 transactions were pre-arranged and indispensable parts of a broader initiative. Macauley Taylor testified regarding Merrill Lynch s planning and marketing of the swap-and-assign vehicles. (1/24/08 Taylor Test. 13:21-19:8.) 70 The participants in the transaction knew that the swaps would precede assignments to ScheringPlough s overseas subsidiaries. (1/16/08 Nichols Test. 37:8-37:16; 1/17/08 Ludwig Test. 81:411; 1/23/08 Den Baas Test. 41:18-22, 61:25-62:6.) Further, Schering-Plough s CEO Robert Luciano testified that the principal purpose for entering into the entire swap-and-assign transactions was to bring the cash back . . . in [a] tax effective manner so that we could consummate . . . transactions that included stock repurchase programs. (1/15/08 Luciano Test. 63:19-21.) The steps were really component parts of a single transaction intended from the outset to be taken for the purpose of reaching the ultimate result of repatriating funds from the Swiss subsidiaries. Assoc. Wholesale Grocers, 927 F.2d at 1523 (quoting King Enters., Inc. v. United States, 418 F.2d 511, 516 (Ct. Cl. 1969) (internal quotations omitted)). Accordingly, under the end result test, the steps of Schering-Plough s 1991 and 1992 swaps can be collapsed, as they all functioned to achieve the underlying goal of repatriating funds. As to the interdependence of the steps, the amount Schering-Plough sought to repatriate would be used to back-solve for the notional principal amount and the amount of the assignments. (See 1/15/08 Luciano Test. 80:22-81:8.) The interlocking figures for the swaps, which temporally preceded the assignments, had to be governed by the amount ultimately repatriated to Schering-Plough, which was the end result. The amount of the assignments also dovetailed neatly, and not accidentally. All of the numbers for the putatively separate steps in the swap-and-assign transactions were predetermined. This result-driven conduct illustrates that the goal of the interlocking transactions was to repatriate foreign-earned funds such that the interest rate swaps would have been pointless had Schering-Plough not entered into the assignments with its subsidiaries just one month later. 71 Schering-Plough argues that use of the step-transaction analysis depends on a fictitious step that the Swiss subsidiaries initially loaned Schering-Plough funds. Pl. Post-Trial Br. at 5556. True enough, the government characterizes the payment from the Swiss subsidiaries to the parent as a loan, but why would it not? The step transaction doctrine is specifically designed to define correctly each component step as part of a larger transaction. As the government points out, it is undeniable that the Swiss subsidiaries transferred cash to their domestic parent. That the government calls these lump-sum payments loans is not inventing steps; the government is only reordering and recharacterizing them. Schering-Plough also argues that by referencing the step where Schering-Plough was ultimately obligated to repay Limited (under the 1991 swap) based on a notional principal of $560 million rather than the $650 million swap it actually entered into, the government invents a step that never happened. This is unpersuasive. The government does not deny that Schering-Plough actually entered into a $650 million notional principal swap. Instead, it uses a portion of the entire swap notional principal $560 million as a repayment benchmark because that was the notional amount for which the receive legs were assigned to the Swiss subsidiaries.31 The government only cites the $560 million amount because it does not refer to the fact that Schering-Plough assigned receive-leg rights on $60 million notional principal to Banco di Roma and retained for itself receive leg rights on $30 million. When the $90 million that was not notionally assigned to the subsidiaries is added to the $560 million that the government references, the sum is the entire $650 million notional principal. So, the $560 million figure is the relevant reference point for establishing the amount that Schering-Plough was effectively obligated to repay the Swiss subsidiaries under the loans. 31 Receive leg rights for $460 million notional principal were actually assigned to Scherico in exchange for money funneled from Limited, and Limited used $100 million of its own funds to purchase the receive leg rights. As previously discussed, the parties have stipulated that the repatriated money came from Limited. 72 In sum and in short on this point, the Court rejects the argument that the government has conjured steps to shoehorn the swap-and-assign transactions into the step transaction doctrine. Schering-Plough asserts that the government identified no meaningless or unnecessary steps, and thus the step transaction doctrine is inapplicable. Pl. Post-Trial Br. at 56-57. But the step transaction doctrine includes re-ordering the steps of the transactions to illustrate the de facto loan structure that the Swiss subsidiaries used to pass funds to the parent corporation. The evidence establishes that Schering-Plough s management and its treasury department transferred monies from Swiss subsidiaries to the main corporate treasury in a circuitous fashion and later repaid the money through an intermediary, with the goal of tax avoidance. Both the end result test and the interdependence test reveal that the formal structure of the transactions can be disregarded and the steps taken to achieve the transactions objectives can be reordered without disturbing the transactions economic reality. Under these tests, the Court concludes that Schering-Plough obtained a loan from its Swiss subsidiaries. e. Summing Up: The Swap-and-Assign Transactions were, in Economic Substance, Loans The economic characteristics of the swap-and-assign transactions do not pass muster as sales under substance-over-form analysis. The evidence demonstrates that the transactions were, in true economic substance, loans, with the direction of the cash flows constituting the most persuasive indicator of the transactions economic essence. There is also substantial evidence demonstrating that Schering-Plough officials contemporaneously understood the transactions aside from their labels to be loans. Furthermore, despite the absence of customary loan documentation, the related-party transactions here were accompanied by substantial objective indicia militating in favor of characterizing them as loans. ABN, a paid participant in the transaction which faced no material (and minimal gross) risk, was a pass-through that routed 73 Schering-Plough s repayments to the Swiss subsidiaries. The Court concludes that ScheringPlough has failed to meet its heavy burden of proving that the government incorrectly characterized the transactions as loans. The Court holds that the income deferral provision in Notice 89-21 did not provide Schering-Plough a shelter from immediate Subpart F taxation. 3. Economic Substance Doctrine Having found that under the various branches of the substance-over-form doctrine, the 1991 and 1992 swap-and-assign transactions were loans not eligible for tax-deferred treatment under Notice 89-21, the Court need go no further to render judgment for the government. Nonetheless, the transactions also fail the related economic substance analysis.32 That is, the transactions were no more than they were tax vehicles lacking legitimate business reasons for entering into them, other than the repatriation of $690 million of previously untaxed foreign income. The economic substance examination is guided by, and a product of, the Third Circuit s decision in ACM Partnership v. Commissioner, 157 F.3d 231, 247 (3d Cir. 1998), and that case will be referenced throughout this section. Per ACM, the Court must look to both the objective economic substance of the transactions and the subjective business motivation behind them. ACM, 157 F.3d at 247 (quoting Casebeer v. Comm r, 909 F.2d 1360, 1363 (9th Cir. 1990)). These two inquiries do not constitute discrete prongs of a rigid two-step analysis, but rather 32 The parties agree (correctly) that the economic substance and substance-over-form doctrines are analytically distinct. The former, also known as the sham-transaction and the economic sham doctrine, applies where the economic or business purpose of a transaction is relatively insignificant in relation to the comparatively large tax benefits that accrue (that is, a transaction which actually occurred but which exploits a feature of the tax code without any attendant economic risk, Horn v. Comm r, 968 F.2d 1229, 1236 n.8 (D.C. Cir. 1992)); in that situation, where the transaction was an attempted tax shelter devoid of legitimate economic substance, the doctrine governs to deny those benefits. Neonatology Assocs., P.A. v. Comm r, 299 F.3d 221, 231 n.12 (3d Cir. 2002). The latter, already discussed, is applicable to instances where the substance of a particular transaction produces tax results inconsistent with the form embodied in the underlying documentation, permitting a court to re-characterize the transaction in accordance with its substance. Id. Both doctrines are relevant here. 74 represent related factors both of which inform the analysis of whether the transaction[s] had sufficient substance, apart from its tax consequences, to be respected for tax purposes. Id. In ACM, the court employed this rational continuum scrutiny and found that contingent installment note sales did not have economic substance to be respected under the tax code. Id. at 263. Scrutinizing the transactions here the same way, this Court concludes that they lack the asserted objective effects and subjective non-tax business motivations. a. Objective Economic Effects In its objective inquiry, the Court asks whether the transaction has any practical economic effects other than tax avoidance and will reject the tax consequences of transactions that were devoid of nontax substance [if] they did not appreciably affect the taxpayer s beneficial interest except to reduce his tax. ACM, 157 F.3d at 248 (internal citations omitted). [T]ransactions which do not vary control or change the flow of economic benefits are to be dismissed from consideration where they do not alter significantly the taxpayer s financial situation. ACM, 157 F.3d at 249. As ACM clarified, courts are empowered to examine superficially legitimate dispositions in their broader economic context, and to decline to recognize them for tax purposes where other aspects of the taxpayers transactions offset the consequences of the disposition, resulting in no net change in the taxpayers economic position. Id. A court can disregard even facially passable dispositions of property, for example, dispositions done at a loss, ha[ving] no net economic effect on the taxpayer s economic position, either because [the] taxpayer retained the opportunity to reacquire the property at the same price, or because the taxpayer offset the economic effect of the disposition by acquiring assets virtually identical to those relinquished. 75 Id. at 249. Overall, a bona fide transaction must have economic substance separate and distinct from economic benefit achieved solely by tax reduction. Id. Schering-Plough submits that the transactions held objective substance because they objectively affected Schering-Plough s net economic position in that Schering-Plough received over $728 million it did not have before the assignments. Pl. Post-Trial Br. at 33. But this argument is circular: Schering-Plough would have the Court believe that the very taxsheltered money it repatriated can itself provide the objective economic effect it seeks to prove. The repatriated funds in themselves do not constitute an appreciable economic effect any more than shifting money from one pocket to another, to borrow Dr. LaRue s analogy. (2/27/08 LaRue Test. 79:25-80:25.) Inasmuch as Schering-Plough faced interest rate risk by entering into the swaps, it entered into add-on mirror swaps with Merrill Lynch to hedge the risk entirely. (1/29/08 Moore Test. 36:16-23; 1/29/08 Moore Test. 60:2-4; 1/29/08 Moore Test. 45:16-22, 61:1-2.) The perfect hedge worked so well that it prompted Schering-Plough s tax counsel, Paul Oosterhuis, to recommend restoring interest rate risk back to the company. (1/29/08 Moore Test. 61:13-21). Any interest rate risk exposure to Schering-Plough was purposely inflicted and does not create objective economic effects. Schering-Plough correctly notes that the objective inquiry focuses heavily on the economic effects on third parties, and it cites the Supreme Court s decisions in Frank Lyon, 435 U.S. at 575, and United Parcel Service of America, Inc. v. Commissioner, 254 F.3d 1014, 101819 (11th Cir. 2001), for the principle that where genuinely unrelated parties are involved, and obligations are created for those parties, then economic substance is present. Pl. Post-Trial Br. at 32 . As stated above, Frank Lyon concerned a sale-and-leaseback transaction not shaped solely 76 by tax-avoidance features that have meaningless labels attached, and the parties there were genuine and did not include a prearranged, compensated counterparty like ABN. Frank Lyon, 435 U.S. at 584. In United Parcel Service, the Eleventh Circuit explicitly noted that [e]ven if the transaction has economic effects, it must be disregarded if it has no business purpose and its motive is tax avoidance. United Parcel Serv., 254 F.3d at 1018. The Court has already addressed the transactions minimal impact on ABN s net financial position beyond the reliable ten-basis-point remuneration it received for its participation. Regulatory capital requirements under the Basel Accord were avoidable (and in any case were inconsequential), and those requirements do not qualify as adverse economic consequences. As to credit risk, ABN was an accommodation party which carried little risk, and what it did carry was hedged with a 60-day termination provision upon a credit downgrade, leaving less than a 0.0005% chance of a Schering-Plough default. Schering-Plough argues that the presence of profit-seeking motivations is not a rigid requirement for economic substance, citing Sacks v. Commissioner, 69 F.3d 982, 991 (9th Cir. 1995) ( Absence of pre-tax profitability does not show whether the transaction had economic substance beyond the creation of tax benefits. ) (internal quotation omitted)). But ScheringPlough was willing to incur significant cost in undertaking the 1991 and 1992 transactions. Joel Finard s testimony for the government established that in addition to the $4.2 million it incurred in advisory fees to Merrill Lynch for the 1991 and 1992 transactions, Schering-Plough could have expected to spend well in excess of $30 million in trading costs of carrying out the transactions. (2/19/08 Finard Test. 87:12-88:16; 2/20/08 Finard Test. 100:13-22.) Adding advisory fees and trading costs, Finard put the price tag at nearly $35 million, which went uncontradicted by Schering-Plough. Thus, while it may be true that profit-seeking is not a 77 prerequisite for economic substance, in this case it is the willingness to spend significantly that is persuasive. b. Subjective Business Purpose Before addressing the non-tax business motivations proffered by Schering-Plough, a few words concerning the manner and scope of the Court s subjectivity inquiry are in order. First, the Court is confronted with apparently divergent statements contained in ACM. On one hand, the government cites ACM for the proposition that because it is the assignments themselves that generated the disputed tax consequences, the Court should ignore the stated business motivations behind the swap portion of the 1991 and 1992 transactions. See ACM, 157 F.3d at 256 n.48, 258 n.54, 260 & n.57 (subjective business motivations behind an independent transaction that [were] executed independently of, did not further, and in fact impeded ACM s pursuit of its non-tax . . . objectives should not be considered); see also Coltec Indus, Inc. v. United States, 454 F.3d 1340, 1356 (Fed. Cir. 2006) ( the transaction to be analyzed is the one that gave rise to the alleged tax benefit ); Black & Decker Corp v. United States, 436 F.3d 431, 441 (4th Cir. 2006) ( specific transaction whose tax consequences are in dispute is the proper focus of the analysis). On the other hand, the Court must view the transactions as a whole, and each step, from the commencement to the consummation is relevant. ACM, 157 F.3d at 247 (quoting Weller v. Comm r, 270 F.2d 294, 297 (3d Cir. 1959)). Here, it is undisputed that the swaps alone generated no tax benefit to Schering-Plough. Only upon the assignments did Schering-Plough obtain the repatriated cash that it used to purportedly take advantage of Notice 89-21. While the Court does agree with the government that the assignments and not the swaps are the driving force behind the transactions tax effects, it will address the asserted business purposes behind the swaps as well. The result is largely academic, as the Court finds in any 78 event that the asserted purposes with respect to both the swaps and assignments do not disturb its conclusion that these transactions were solely tax-motivated. Second, the government asserts that how the repatriated sums were used is an irrelevant inquiry when assessing the legitimacy of Schering-Plough s proffered business uses for the swap-and-assign transactions. Citing Internal Revenue Service v. CM Holdings, Inc. (In re CM Holdings, Inc.), 254 B.R. 578, 638-39 (D. Del. 2000), aff d, 301 F.3d 96 (3d Cir. 2002); American Electric Power Co. v. United States, 326 F.3d 737, 744 (6th Cir. 2003); and WinnDixie Stores, Inc. v. Commissioner, 113 T.C. 254, 287 (T.C. 1999), the government argues that the Court must focus on the purpose of the underlying transaction[s] at issue here, not what [Schering-Plough] intended to do with the proceeds. Def. Post-Trial Br. at 57 (quoting In re CM Holdings, 254 B.R. at 638-39). Schering-Plough attempts to distinguish these cases on the ground that they focus on the manner in which tax savings from sham deductions were used, as opposed to the manner in which cash obtained from a tax-motivated transaction (as in this case) was used. But the use of tax savings through deductions and the use of cash obtained are flip sides of the same coin. In the cases cited by the government, extra money (albeit in the form of reduced tax) was obtained using deductions; here, cash was acquired from the assignments. Differentiating the tax viability of transactions based on the particular mode of the tax benefit conferred puts too fine a point on the inquiry. 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. Am. Elec. Power Co. v. United States, 326 F.3d at 744 (citing In re CM Holdings, 254 B.R. at 638-39. This statement is no less true merely because money was generated by an transaction in the form of cash instead of deductions. 79 Schering-Plough cites only Goldstein v. Commissioner, 364 F.2d 734, 738-42 (2d Cir. 1966) and Rubin v. United States, 304 F.2d 766, 770 (7th Cir. 1962) to support its argument that the Court must consider the uses of the repatriated funds. In Goldstein and Rubin, the taxpayers used loaned funds to purchase Treasury securities, and thereafter claimed deductions for the interest payments on the loans. In both cases, however, the courts of appeals affirmed the lower court s disallowance of the taxpayers claimed deductions. It is apparent that the use of the funds obtained was not the turning point on which the courts holdings rested; quite the opposite, the transactions were not legitimated by otherwise valid uses of the monies obtained. In conducting the sham transaction analysis, the courts placed little emphasis on the fact that the taxpayers purchased assets with the loaned funds. Turning to the subjectivity inquiry, Schering-Plough gives three non-tax reasons for the swap-and-assign transactions. i. Financial Reporting, Cash Management, and Balance Sheet Motivations The government argues that because the results of the transactions from a financial reporting perspective remained the same irrespective of whether they were accounted for as sales, loans, or dividends, Schering-Plough could not have had the subjective business purpose of serving cash management and financial reporting objectives. Schering-Plough responds by arguing that it is not required to take the path that leads inevitably to the greatest tax. But this position misconstrues the argument. The point is that the manner in which the repatriated money is reported demonstrates that financial reporting could not have been a consideration for entering into the transactions, because the transactions whether loan, dividend, or sale would not appear on Schering-Plough s consolidated financial statements. John Neel Foster, another expert witness for the government, explained in his report the reason why. 80 The financial statement of a corporation that consolidates its own activities with those of its subsidiaries collapses intercompany transactions to prevent them from entering into transactions with favorable reporting effects, but no real economic effect on the consolidated entity. (Ex. 501 ¶ 21.) Thus, Foster concluded that financial reporting effects could not have been a motivating factor for the transactions because the entry in Schering-Plough s consolidated statement would be the same irrespective of the manner in which they were classified. (Ex. 501 ¶ 24.) If cash management, Schering-Plough s second asserted motivation, means the domestic parent s physical access to the cash in the United States, the Court finds that purpose deficient because it would be nothing more than moving the cash from one pocket to the other. (2/27/08 LaRue Test. 79:25-80:25.) The Court arrives at the same conclusion if cash management means that Schering-Plough was able to prevent ballooning either way, the asserted non-tax purpose is unpersuasive. To be sure, the repatriated cash was used to gross down the balance sheet by either paying off debt or using the money for its stock repurchase programs without incurring new debt (the testimony on this point conflicted at trial). But again, the use for which a disputed transaction is put is not relevant in determining whether the transaction itself has sufficient substance. The Court therefore rejects Schering-Plough s asserted balance sheet motivations for entering into the transactions. ii. 1991 Transaction: Hedging Potential Interest Rate Spread Schering-Plough argues that aside from its tax motivations for entering into the 1991 swap-and-assign transaction, it wanted to hedge against the possibility that the historically positive spread between the LIBOR it earned on its foreign cash positions and the commercial paper it paid on its domestic debt would decline. By entering into a swap with a LIBOR pay leg and a commercial paper receive leg, Schering-Plough presumably would then be hedged against 81 any future downturn in the LIBOR-commercial paper spread. And on a superficial level this makes sense. A closer inspection, however, reveals that the proffered hedging objective, in Dr. Parson s words, makes no sense. Ex. 539A ¶ 35. First, the swap was designed to hedge the LIBOR positions the subsidiaries held overseas. But upon execution of the swap, the cash was immediately transferred to ScheringPlough under the assignments. As Joel Finard stated in his expert report, after the assignments were effected, locking-in a positive interest rate spread between LIBOR and CP [commercial paper] might no longer have been applicable because there might no longer have been cash to hedge. Ex. 2133 at 20 n.29. Second, the manner in which the notional principal for the swap was determined is inconsistent with a true hedging objective. As discussed above, the terms of the swap were determined by back-solving for the maximum notional principal that Schering-Plough desired to repatriate. As Finard explains, this is contrary to standard market practice with respect to the implementation of hedges. Typically the notional amount of a swap being used to implement a hedge is determined by matching the amount of the exposure being hedged. Id. But here, there was no reference to the potential exposure Schering-Plough faced on deviations in the LIBORcommercial paper spread. Moreover, the 20-year length of the swaps is incompatible with the short-term nature of the Swiss subsidiaries LIBOR positions and Schering-Plough s domestic debt.33 Third, the evidence fails to demonstrate that the 1991 swap was implemented to accomplish that objective. Jay Ludwig admitted at trial that the last briefing book for the 1991 33 At the time of the 1991 swap, the median duration of its investment was 30 to 90 days, the median duration of its commercial paper debt was 30 to 180 days, and 88% of its investments were under 6-months in maturity. (1/17/08 Ludwig Test. 120:2-13,122:1-123.) 82 swap-and-assign transaction presented to Schering-Plough s Finance & Audit Committee not only failed to mention the potential spread decline, but in fact stated that the exposure to the spread had been reduced by decreasing its gross amount of domestic commercial paper debt. (1/17/08 Ludwig Test. 111:5-114:8; Ex. 211 at P005758.) Only after entering into the 1991 transaction did it refer ex post to a potential arbitrage problem between the LIBOR earnings and its commercial paper debt. (1/17/08 Ludwig Test. 110:16-111:4; 113:17-114:8.) Fourth, the asserted hedge, on a substantive level, did not and could not provide an effective hedge against the supposed risk of a LIBOR-commercial paper spread decline. Schering-Plough allegedly entered into the swap to sustain its then-favorable spread. As government expert witness Dr. Parsons explains in his report: If financial markets are predicting a future decline in the spread, Schering-Plough will not be able to lock in the current spread. Instead, they will be offered terms that lock in a fixed spread that is less than the current spread, reflecting the predicted decline. Instead of protecting themselves against the predicted decline in the spread, Schering-Plough will simply have fixed its terms to lock in whatever is the predicted decline. Without the swap, they will on average find themselves earning the lower spread, but depending upon how rates move, could get lucky and find that the spread improves, or could get unlucky and find that the spread declines even more than expected. With the swap they avoid the uncertainty, but still suffer the lower forecasted spread not just on average, but certainly. . . . The hedging motivation advanced above refers to exposure to the forecasted decline in the level of the spread. And the swap has not protected the company against this forecasted decline. It is not possible. Indeed, what the swap has done is to lock in the value consequences of the forecasted decline!34 34 Schering-Plough attempts to undercut Dr. Parsons s analysis, arguing that it was not attempting to lock in a profit, but simply attempting to protect against adverse changes in the spread. See Pl. Proposed Findings of Fact ¶ 324. But the above quote shows that Dr. Parsons candidly acknowledged that Schering-Plough locked in certainty. And it still does not blunt the fact that Schering-Plough locked in a certainly lower rate, nor does it affect Dr. Parsons s and the Court s ultimate opinion that this particular swap made little sense in light of the circumstances. 83 Ex. 539A ¶¶ 35, 43. In fact, Luciano admitted that not only did the swap not effectively provide the hedge desired, but it actually exposed Schering-Plough to fluctuations in domestic interest rates. (1/15/08 Luciano Test. 79:11-24.) This is supported by Ludwig s testimony that the swap did not function effectively during its first six months and after its second six months. (1/17/08 Ludwig Test. 123:16-124:4; Exs. 213 at P000070, 214, 222.) The government also presented the Court with additional evidence that Schering-Plough: (1) looked to terminate the 1991 swap as early as November 1991; and (2) incurred significant losses while attempting to mitigate the adverse effects of the swap. (1/17/08 Ludwig Test. 124:2-125:5; Exs. 214, 22.) Further adding to the inadequacy of the proffered hedge is the fact that Schering-Plough received a federal funds rate rather than a commercial paper rate on the receive leg. Before executing the 1991 swap, Merrill Lynch informed Schering-Plough that ABN was unwilling to enter into a swap under which it was obligated to pay a commercial paper rate. (1/17/08 Ludwig Test. 21:9-13, 27:18-28:12, 118:16-119:14; Ex. 13.) The parties agreed upon an average federal funds rate as a substitute, with Schering-Plough paying a spot LIBOR rate in return. This resulted in an imperfect hedging mechanism. Schering-Plough argues that the historical federal funds rate tracked the commercial paper rates, and thus the substitution of an imperfect proxy does not defeat its subjective purpose. When examined carefully, however, Schering-Plough s argument is disingenuous. The company did not have any information concerning the historical relationship between commercial paper and federal funds rates until five months after it entered into the 1991 swap, when Merrill Lynch provided this information to it. (1/17/08 Ludwig Test. 117:20-118:6, Ex. 89.) Thus, even if the indices had any correlative relationship, and even if the average federal funds rate was a sufficient substitute, the fact that it did not know the historical relationship between commercial paper and federal funds rates shows that Schering-Plough s 84 hedging motive came after the fact. While a taxpayer may have ulterior tax avoidance purpose[s] along with its non-tax intentions, ACM, 157 F.3d at 247 (quoting Gregory v. Helvering, 293 U.S. at 469), Schering-Plough may not invoke business purposes ex post, even if those purposes might have been legitimate ex ante. Finally, Schering-Plough performed no preliminary analysis of its own regarding the probable effectiveness of the hedge. Instead, relying on its trusted advisor Merrill Lynch, Schering-Plough argues that because it, and companies like it, did not have the wherewithal at the time to do their own financial analysis of hedges, the failure to do minimal review of the swap does not undermine subjective adequacy. But Ludwig testified that even a pharmaceutical company like Schering-Plough would have wanted to see an analysis showing how the indices being used tracked one another before entering into a hedging swap. (1/17/08 Ludwig Test. 116:25-117:4.) He further admitted that Merrill Lynch provided its client neither interest rate projections nor historical comparisons regarding LIBOR and the federal funds rates that were used for the 1991 swap. (1/1708 Ludwig Test. 22:7-24, 103:5-104:8.) Finally, Ludwig conceded that Merrill Lynch did not prepare a proprietary model of the 1991 swap-and-assign transaction before it was executed, and did not share such modeling with its client until 1993, well after the transaction s execution. (1/17/08 Ludwig Test. 104:9-15). This testimony supports Finard s conclusion that Schering-Plough failed to perform then market-standard analysis of historical rates and possible scenarios. Ex. 2133 at 21-22. iii. 1992 Transaction: Yield Enhancement Schering-Plough asserts a different motivation for the 1992 swap-and-assign transaction yield enhancement. Schering-Plough s witnesses testified at trial about hedge accounting. Under these reporting rules, the legs of the 1991 swap were not reported on 85 Schering-Plough s balance sheet; only the netted receivable or payable would be reported on the income statement. (1/17/08 Ludwig Test. 114:12-21.) Ludwig testified that Schering-Plough would not have entered into the transactions without this special type of accounting. (1/17/08 Ludwig Test. 114:23-115:1) Later, when hedge accounting was unavailable to Schering-Plough as a result of changed accounting rules, it did not cite hedging as a purpose for entering into the 1992 transaction. (1/17/08 Ludwig Test. 115:17-20.) Instead, after Schering-Plough s ability to use special hedge accounting rules for the 1991 swap disappeared, Schering-Plough shifted course, and now asserts that the 1992 swap accomplished a yield enhancement function. That is, by entering into a swap in which it received effectively a two-year federal funds rate and paid a one-year LIBOR rate, it wanted to enhance the yield the Swiss subsidiaries earned on their overseas LIBOR investments by moving farther out on the interest rate yield curve to a longerterm interest rate. Pl. Post-Trial Br. at 40. First, as with the asserted non-tax business purpose of the 1991 swap, Ludwig admitted that Schering-Plough undertook no pre-transaction analysis regarding the 1992 swap. (1/17/08 Ludwig Test. 130:23-131:23.) The failure to conduct any analysis lends significant support for the conclusion that the yield enhancement function of the 1992 swap is an insufficient post hoc justification. See ACM, 157 F.3d at 257 n.51. Second, as Dr. Parsons testified, Schering-Plough could have very easily accomplished its yield enhancement objective by liquidating its short term LIBOR positions and re-investing in longer-term assets, rather than entering into illiquid and costly 20-year swaps. (1/30/08 Parsons Test. 79:13-84:13, Ex. 539A ¶ 60.) Schering-Plough s only answer to Dr. Parson s analysis is that the government may not second-guess the business judgment regarding the manner in which its executives accomplished the yield enhancement objective. It cites Boyd Gaming Corp. v. Commissioner, 177 F.3d 1096, 86 1100 (9th Cir. 1999), and Network Systems Corp. v. United States, 814 F. Supp. 778, 781 (D. Minn. 1993). See Pl. Post-Trial Br. at 42. But neither Boyd nor Network Systems address whether a transaction should be respected under the economic substance doctrine. As the Court has already discussed, the longer the term of the swap, the more Schering-Plough could repatriate under the auspices of Notice 89-21. This is more consistent with a tax objective than a costly and awkward yield enhancement mechanism. c. The Swap-and Assign Transactions Fail ACM In addition to the above, other factors suggest that the transactions were purely taxmotivated and that their economic effects were limited or absent altogether. Highly persuasive is the manner in which the prices the Swiss subsidiaries paid to Schering-Plough were obtained: Using third-party banks that understood their role from the beginning, that were compensated for their participation, and that were quickly relieved of their newly purchased assets. Further, the general atmosphere under which the transactions were implemented, along with the manner in which the transactions played out in fact, suggest that cash management, hedging, and yield enhancement had no contemporaneous force behind Schering-Plough s decision to enter into the transactions. The government s theory in this case is not that Schering-Plough was obliged to chart its course so as to arrive at the greatest tax. Rather, the government asserts that the relative ease, efficiency, and economy with which Schering-Plough s alternative options presented themselves prove simply that the company could not have had the proffered business motivations when it entered into the transactions. The Court agrees. A reasonable taxpayer would not have undertaken such roundabout means to arrive at such straightforward ends. Ultimately, the Court finds that the transactions do not survive the economic substance analysis. Were the swap-and-assign vehicles genuine multiple-party transaction[s] with 87 economic substance which [were] compelled or encouraged by business or regulatory realities, . . . imbued with tax-independent considerations, and . . . not shaped solely by tax-avoidance features that have meaningless labels attached, Frank Lyons, 435 U.S. at 583-84, the Court could hold differently. But the economic effects of the transactions themselves on both Schering-Plough and ABN were, in the end, insignificant, and Schering-Plough has failed to establish a genuine purpose for the transactions other than tax avoidance. The transactions thus fail ACM. 4. The Big Picture: Subpart F vs. Notice 89-21 The Court is mindful not to miss the forest for the trees, as cautioned in Prabel v. Commissioner, 882 F.2d 820, 828 (3d Cir. 1989). Wholly apart from the Court s economic and legal analysis above is the powerful fact that Schering-Plough desired from the outset to bring $690 million of previously untaxed foreign income back to the United States without paying an up-front tax. Subpart F of the Internal Revenue Code was specifically designed to prevent this. Congress endeavored to deter United States taxpayers from using related foreign base companies located in tax haven countries to accumulate earnings that could have been accumulated just as easily in the United States. Koehring Co. v. United States, 583 F.2d 313, 317 (7th Cir. 1978). More specifically, it enacted the statute in 1962 to curb perceived abuses by taxpayers who, through controlled foreign corporations, would repatriate otherwise nontaxable foreign income for the use and benefit of domestic shareholders. Jacobs Eng g Group, Inc. v. United States, No. 96-2662, 1997 WL 314167, at *3 (C.D. Cal. Mar. 5, 1997) (citing Gulf Oil Corp. v. Comm r, 87 T.C. 548, 571 (T.C. 1986)). To credit plaintiff s position 88 on the evidence before the Court would mean disregarding this pervasive legislation the bread and butter of international tax practice. 35 Notice 89-21 does not supplant, qualify, or displace Subpart F. The statutory scheme and the administrative pronouncement are not on equal footing. The former reflects congressional will, and the latter merely represents the Commissioner s position with respect to a specific factual situation, and as such does not constitute authority for deciding a case in this Court. Stark v. Comm r, 86 T.C. 243, 250-51 (T.C. 1986); see also Cont l Ill. Corp. v. Comm r, 58 T.C.M. (CCH) 790, at *45-46 (T.C. 1989) (declining to give precedential weight to Notice 8921). To the extent the IRS previously determined that consideration received in exchange for the sale of income rights under a notional principal contract should be amortized over the life of the contract, it did not do so with the purpose of granting corporate taxpayers an out, so to speak, from Subpart F taxation. In that respect, the government is correct that Schering-Plough misframes the issue. This is not a case about the mechanics of timing receipt of income, an issue to which Notice 89-21 was entirely devoted. Rather, the core issue here is whether the notice permitted avoidance of the scheme implemented by the statute. Schering-Plough argues that as a result of the Court s holding today, Notice 89-21 was meaningless from its inception. The Court, ironically, cannot disagree with this sentiment insofar as Notice 89-21 could be construed to allow repatriation of foreign E&P free from immediate taxation. Subpart F was aimed at preventing repatriation of income to the United States in such a way that the income is not subject to U.S. taxation. Jacobs, 1997 WL 314167, at *1 (quoting Gulf Oil, 87 T.C. at 571). In a case such as this, where intricate semantic 35 Micah Levy, Impact of the American Jobs Creation Act of 2004 on the Internal Revenue Code Section 956(c)(2)(a) Exception from U.S. Property for Deposits with Persons Carrying on the Banking Business, 24 ANN. REV. BANKING & FIN. L. 295, 303 (2005). 89 distinctions can blur the dividing line between loan and sale, the Court is ultimately tasked with deciding whether the facts fall within the intended scope of the Internal Revenue Code provision at issue. Stewart v. Comm r, 714 F.2d 977, 988 (9th Cir. 1983). The Court concludes that Notice 89-21 was not intended to permit United States shareholders of controlled foreign corporations to repatriate offshore revenues without incurring an immediate repatriation tax. In a clash between the short-lived notice and the enveloping Subpart F regime, the Court is faced with an easy choice, especially where the consequences of elevating the notice over the legislative regime would be to permit the schemes of taxpayers to supersede legislation in the determination of the time and manner of taxation. Higgins v. Smith, 308 U.S. 473, 477-78 (1940); see also Levy, supra note 35, at 303 (discussing legislative history of Subpart F and citing H.R. Rep. No. 87-1447, at 405, 461 (1962); S. Rep. No. 87-1881, at 707, 784 (1962)). Notice 89-21 was never intended as a reprieve from Subpart F taxation. C. Ultimate Conclusions of Law The Court holds that the 1991 and 1992 swap-and-assign transactions into which Schering-Plough entered should not be respected as sales of future income streams. transactions were, in substance, loans. The Furthermore, the transactions had no appreciable economic effect on the parties, and Schering-Plough lacked sufficient subjective non-tax motivations for entering into them; it therefore cannot reap the benefit of the tax-driven vehicle. Finally, by repatriating $690 million in offshore earnings, Schering-Plough cannot avoid under the pretext of Notice 89-21 the obvious intent of Congress to capture a portion of such sums under Subpart F. The Supreme Court has said: As to the astuteness of taxpayers in ordering their affairs so as to minimize taxes . . . the very meaning of a line in the law is that you intentionally may go as close to it as you can if you do not pass it. 90 This is so because nobody owes any public duty to pay more than the law demands: taxes are enforced exactions, not voluntary contributions. Estate of Stranahan, 472 F.2d at 867 n.3 (quoting Superior Oil Co. v. Mississippi, 280 U.S. 390, 395-96 (1930); Atl. Coast Line v. Phillips, 332 U.S. 168, 172-73 (1947)). Subpart F does not leave room for a corporate taxpayer to opt out of making a voluntary contribution. Subpart F prescribes an exaction which the law demands upon repatriation of foreign-earned revenue. The repatriation transactions became immediately taxable when the Swiss subsidiaries advanced the lump-sum payments to their corporate parent in the United States. The Court therefore holds that Schering-Plough is not entitled to a refund. VI. CONCLUSION For the aforementioned reasons, Schering-Plough has failed to meet its burden demonstrating legal error in the tax assessed by the IRS. The Court will enter judgment in favor of the government and dismiss plaintiff s complaint with prejudice. An appropriate judgment and order will accompany this opinion. /s/ Katharine S. Hayden Hon. Katharine S. Hayden United States District Judge Date: August 28, 2009 91

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.