ACM Pshp. v. Commissioner

1997 T.C. Memo. 115, 73 T.C.M. 2189, 1997 Tax Ct. Memo LEXIS 118

• Court: United States Tax Court
• Decided: March 5, 1997
• Docket: Docket No. 10472-93.
• Status: Unpublished

Opinion

ACM PARTNERSHIP, SOUTHAMPTON-HAMILTON COMPANY, TAX MATTERS PARTNER, Petitioner v. COMMISSIONER OF INTERNAL REVENUE, Respondent

ACM Pshp. v. Commissioner

Docket No. 10472-93.

United States Tax Court

T.C. Memo 1997-115; 1997 Tax Ct. Memo LEXIS 118; 73 T.C.M. (CCH) 2189;

March 5, 1997, Filed

---

Decision will be entered under Rule 155.

In 1988, C reported a $ 105 million capital gain. In 1989, M, an investment banking firm, approached C with an elaborate scheme to shelter that gain from Federal income tax. Pursuant to M's advice, A, C, and M created an offshore partnership (P) in which their respective initial interests were 82.63, 17.07, and .29 percent. P served as the vehicle for a contingent installment sale transaction (CINS transaction) that would create approximately $ 100 million of capital losses for C, a domestic corporation, and corresponding capital gains for A, a foreign corporation that was not subject to U.S. tax. Pursuant to the scheme, P purchased securities and, approximately 3 weeks later, sold most of the securities for cash and LIBOR Notes. The value of the total consideration received, in the form of cash and LIBOR Notes, equaled the price that P had paid for the securities sold. The transactions and the returns connected thereto were the result of a carefully crafted and faithfully executed sequence of sophisticated and costly financial maneuvers that left little to chance or market opportunities. P used the contingent payment sale provisions of sec. 15a.453-1(c), Temporary Income Tax Regs., 46 Fed. Reg. 10711 (Feb. 4, 1981), to report the sale for Federal income tax purposes. In accordance therewith, P reported a large capital gain in the year of sale; most of this gain was allocated to A. In a later year, after P redeemed A's entire interest, P sold the notes and reported a corresponding capital loss, most of which was allocated to C. The loss was carried back to 1988 by C to offset its gain. Held: The Court will disregard the CINS transaction for Federal income tax purposes because it lacked economic substance.

Fred T. Goldberg Jr., Albert H. Turkus, Pamela F. Olson, William L. Goldman, Christopher Kliefoth, and Joni Lupovitz, for petitioner.

Jill A. Frisch, Patricia A. Donahue, Edward D. Fickess, Sheila Olaksen, Elizabeth P. Flores, Brian J. Condon, and James M. Guiry, for respondent.

LARO, Judge

---

LARO

CONTENTS

Findings of Fact

1. The Contingent Installment Sale Transaction

2. Development of Colgate's Liability Management Partnership

3. The Partners

4. The Partnership Agreement

5. Initial Stage of Colgate's Partnership Strategy

6. Tax and Financial Accounting for the Results

7. Final Stage of Colgate's Partnership Strategy

8. Merrill's Collateral Swap Transactions

9. ABN's Investment Management

Opinion

1. Mechanics of a Contingent Payment Sale

2. Economic Substance

a. Introduction

b. Profit

c. Hedging Within the Four Corners of the Partnership

d. Interim Use for Idle Cash

e. The Pattern of Ostensibly Market-Driven Decisions

MEMORANDUM FINDINGS OF FACT AND OPINION

LARO, Judge: ACM Partnership (ACM or the partnership), Southampton-Hamilton Co. (Southampton), Tax Matters Partner, petitioned the Court under section 6226 to readjust respondent's adjustments of partnership items flowing from the partnership. Respondent issued ACM a notice of final partnership administrative adjustment (FPAA) that reflects adjustments to ACM's partnership return of income for its taxable years ended November 30, 1989 (FYE 11/30/89), November 30, 1990 (FYE 11/30/90), November 30, 1991 (FYE 11/30/91), and December 31, 1991 (FYE 12/31/91). In relevant part, respondent eliminated the capital gain reported by ACM in FYE 11/30/89 as resulting from the transaction described herein, and she disallowed the corresponding capital loss reported in FYE 12/31/91.

Respondent asserted a number of alternative theories in the FPAA to support the adjustments. Primarily, respondent asserted, the purchase and sale of the debt instruments at issue herein were prearranged and predetermined, devoid of economic substance, and lacking in economic reality. Alternatively, respondent asserted, ACM's activities must be disregarded under the step transaction doctrine, ACM's activities were not engaged in for profit within the meaning of section 183, and the sale of the subject debt instruments did not satisfy the formal requirements for a contingent payment sale under section 15a.453-1(c) (1), Temporary Income Tax Regs., 46 Fed. Reg. 10711 (Feb. 4, 1981).

Following respondent's concession of a number of these alternative theories, the parties ask the Court to decide the following issues:

(1) Whether respondent's adjustments to items of income and loss reported by ACM on the subject transactions should be sustained on the ground that the transactions lacked economic substance. We hold they should.

(2) Whether, as alleged by respondent in her amendment to answer, the foreign partner should be treated as a lender for Federal income tax purposes. In view of our disposition of the first issue, we do not decide this issue. Consistent with the FPAA, as well as the manner in which ACM reported the foreign partner on its returns, we assume that the foreign partner is not a lender.

(3) Whether ACM's allocation of taxable gain on the sale had substantial economic effect or was otherwise in accordance with the partners' interests in the partnership. In view of our disposition of the first issue, we need not and do not decide the validity of this allocation.

Unless otherwise indicated, section references are to the Internal Revenue Code in effect for the years at issue, and Rule references are to the Tax Court Rules of Practice and Procedure. Throughout this Opinion, we use the terms "purchase", "sale", "contingent installment sale", and "contingent payment sale" solely for purposes of convenience and clarity. Our use of these terms is not meant to give legal significance to the underlying and surrounding transactions.

Some of the facts have been stipulated. The stipulations of fact and attached exhibits are incorporated herein by this reference. When the petition was filed, ACM's principal place of business was in Wilmington, Delaware.

ACM is one of 11 partnerships (section 453 partnerships) formed over a 1-year period from 1989 to 1990 by the Swap Group at Merrill Lynch & Co., Inc. ¹ Each section 453 partnership was intended to be a vehicle for sheltering capital gains of one of its partners. For purposes of this Opinion, the principal transactions in which ACM engaged are collectively referred to as the contingent installment sale transaction (CINS transaction).

The design of the CINS transaction appears to have originated in discussions in early 1989 between Macauley Taylor (Taylor), a managing director of Merrill's Swap Group, and James Fields (Fields), a member of his staff. From the spring of 1989 through the summer of 1990, the Swap Group and Merrill's investment bankers promoted the idea among Merrill's clients.

Colgate-Palmolive Co. (Colgate) was one of Merrill's clients that Taylor and his staff approached. Colgate's treasury department regularly consulted Henry Yordan (Yordan), a managing director in Merrill's Capital Markets Group, concerning developments in the debt markets. Yordan was aware that Colgate had reported a sizeable capital gain (approximately $ 105 million) for its 1988 taxable year on its sale of the Kendall Co. (Kendall), and that Colgate might be receptive to the CINS transaction. Through Yordan's introduction, a meeting was held on May 15, 1989, at which Taylor and his staff described the CINS transaction to Colgate's assistant treasurer, Hans Pohlschroeder (Pohlschroeder). Merrill's representatives stated that, apart from the few elements that were essential to secure the desired tax consequences, the partnership structure could be adapted to suit a variety of investment objectives.

Colgate's initial reaction to the proposal was skeptical. Pohlschroeder explained that Colgate did not have the required cash to invest in the partnership, and that the cost of borrowing to finance the investment was likely to exceed the return on a pretax basis. Pohlschroeder also was not persuaded that the partnership would serve a business purpose of Colgate. When Pohlschroeder related the proposal to Steve Belasco (Belasco), Colgate's vice president of taxation, the latter agreed: But for the tax benefits, the transaction did not accomplish anything useful for the company. Belasco also was concerned that the transaction did not have sufficient economic substance to withstand scrutiny, and that the transaction's legal, financial, and accounting complexities would require broad interdepartmental support within Colgate. Absent a connection to Colgate's business, Belasco believed, the necessary support would not be forthcoming.

Merrill's proposal was not the first that Colgate considered to minimize the tax impact of the Kendall sale. During the previous summer, while the sale was pending, a cross-functional team from Colgate's treasury, accounting, and tax departments had considered at least 11 tax-saving proposals, including investing in low-income housing or property eligible for rehabilitation credits and creating a charitable foundation. All of these proposals were rejected.

After the initial meeting between Colgate and Merrill, Fields, on behalf of Merrill, contacted a law firm for advice on the tax consequences of a CINS transaction. In relevant part, the firm summarized the contemplated transaction as follows: A (a foreign entity), B, and C form the ABC Partnership (ABC) on June 30, 1989, with respective cash contributions of $ 75, $ 24, and $ 1. Immediately thereafter, ABC invests $ 100 in short-term securities which it sells on December 30, 1989, to an unrelated party. The fair market value and face amount of the short-term securities at the time of the sale is still $ 100. In consideration for the sale, ABC receives $ 70 cash and an installment note that provides for six semiannual payments, commencing 6 months after the sale. Each payment equals the sum of a notional principal amount multiplied by the London Interbank Offering Rate (LIBOR) at the start of the semiannual period. ² ABC uses the $ 70 cash and the first payment on the installment note to liquidate A's interest in ABC and uses the subsequent interest payments to purchase long-term securities.

Relying on section 15a.453-1(c), Temporary Income Tax Regs., 46 Fed. Reg. 10711 (Feb. 4, 1981), the law firm advised Fields that ABC would be entitled to report the sale of the short-term securities on the installment method, and that ABC would recover an equal portion of its basis in the short-term securities in each year in which a payment on the note could be received. The law firm advised Fields that ABC would recover $ 25 of its basis in each of the 4 taxable years from 1989 through 1992, and that ABC would have to recognize gain in each of these years to the extent that the year's payments exceeded $ 25. To the extent that the year's payments were less than $ 25, the law firm advised, ABC would not be allowed to recognize a loss in that year, but ABC would have to carry over that "loss" to a later year in which it would otherwise recognize enough gain on the sale to absorb all or part of the "loss". The law firm advised that any unrecognized loss on the sale would be recognized in the final year of payment.

In a series of telephone calls in early July 1989, Pohlschroeder revisited Merrill's proposal with Yordan, Taylor, and Fields. Pohlschroeder communicated a number of concerns that Colgate had regarding the management of its debt. Pohlschroeder wondered whether there was a way to combine Colgate's financial objectives with Merrill's proposal. On July 18, 1989, Taylor called Pohlschroeder back with a suggestion for resolving the problem. The gist of the conversation can be reconstructed from Pohlschroeder's handwritten notes:

Mac

Invest. partnership

Based on bus. purpose

Economic profit

Is this partnership profitable?

Every single step to be substantiated

Invest in your own debt

Consolidation of effective control but not majority ownership

The notion that Colgate could use a partnership to acquire its own debt was the breakthrough that overcame Colgate's reservations, for it provided the opportunity to design an elaborate superstructure of liability management functions around Merrill's original tax shelter transaction. To understand the extent to which ACM was designed to serve these functions, we first review the concerns of Colgate's treasury department in this period.

A number of developments during 1988 and 1989 posed special challenges for the management of Colgate's debt. In this period, Colgate radically altered the maturity profile of its debt through two actions. First, it used the proceeds from the sale of Kendall in October 1988 to retire over half a billion dollars of commercial paper constituting all of its U.S. short-term debt. Second, it established an employee stock ownership plan (ESOP) in June 1989, financed by issuing $ 410 million in long-term debt.

The substitution of long-term debt for short-term debt caused Colgate's average debt maturity to exceed substantially the norm in its industry and increased its exposure to interest rate risk. ³ Colgate's treasury department expected the Federal Reserve to ease monetary policy, causing interest rates to fall in late 1989 or the first half of 1990. In a falling interest rate environment, Colgate would earn a lower return on its cash balances and short-term investments; yet, unlike its competitors with relatively greater amounts of short-term debt, it would be unable to cut its interest expense by refinancing. The establishment of the ESOP had the further consequence of prompting Moody's to downgrade Colgate's long-term debt from A1 to A2 on the ground that the addition of so much long-term debt reduced the company's financial flexibility. In the summer of 1989, Colgate's treasury department was exploring ways to rebalance the term structure of its debt and lower its exposure to falling interest rates. Pohlschroeder raised these issues in his discussions with Merrill's representatives in July 1989.

The discussions also concerned the credit spread at which Colgate's long-term debt was trading. The market's perception of the credit worthiness of a corporation is reflected in the extent to which the yield on the corporation's bonds exceeds the yield on U.S. Treasury instruments of comparable maturity. The "spread to Treasury" of Colgate's long-term debt had exceeded the average for high and medium grade industrials throughout 1988 and, after narrowing in the early part of 1989, had widened markedly during the summer. One reason for this change was the downgrade in Colgate's credit rating in June. Colgate's treasury believed that another factor was widespread speculation that Colgate could become the target of a hostile takeover or leveraged buyout. This led to the emergence of an "event risk" premium that caused Colgate debt to trade at a discount relative to the price that would otherwise obtain. In Colgate's opinion, the market was overestimating the risks of holding Colgate's debt. Thus, Colgate's debt was undervalued, and an opportunity existed to capture subsequent improvements in its perceived credit quality by repurchasing the debt. Yet, Colgate's flexibility to respond to this arbitrage opportunity was constrained by the prospect that a significant reduction in its balance sheet liabilities would enhance its appeal to a potential acquirer.

Through the collaboration of Merrill's Swap Group and Colgate's treasury department, from late July to early October 1989, the partnership gradually took shape. Merrill's first written exposition of the concept, entitled "Colgate Partnership Transaction Summary", dated July 28, 1989, states: "the primary mission of the Partnership is the acquisition and control of Colgate debt". "Colgate Sub.", "A Corp.", and "B Corp." would contribute $ 30 million, $ 169.3 million, and $ 0.7 million, respectively. Colgate Sub. would act as managing general partner with the authority to determine partnership investments. Over a period of several months, the partnership capital would be used to acquire long-term Colgate debt from investors. The partnership would then exchange some of the long-term debt for newly issued Colgate medium-term debt. Merrill noted that the accounting treatment of the partnership was unclear. Despite Colgate's minority interest, Merrill believed, the partnership might have to be consolidated on Colgate's financial statements if Colgate were deemed to control the partnership. Merrill thought that either result might be advantageous.

By the beginning of October 1989, the design had been revised in two important respects. First, it had been determined that the partnership would be most useful if its transactions were initially kept off of Colgate's balance sheet and its consolidation with Colgate for financial accounting purposes was deferred until such time as Colgate acquired a majority interest in the partnership from the foreign partner. This would enable Colgate to conceal its activities from the market as well as choose more advantageous market conditions for retiring and reissuing the debt. Second, Merrill had devised a mechanism by which Colgate and the foreign partner could share the credit risk with respect to partnership holdings of Colgate debt in different proportions from the so-called treasury (i.e., interest rate) risk. The efficiency of "allocating to each partner the risks that it could bear" would make it possible for Colgate to receive greater benefits from the partnership at less cost. Thus, it was expected that Colgate could negotiate for the right to appropriate all the benefit of the improvement in its credit quality that it expected to occur over time, while negotiating for an option to vary the partners' relative shares of the treasury risk inherent in the debt so as to capitalize on expected changes in interest rates.

A document entitled "Liability Management Partnership Executive Summary", dated October 11, 1989, purports to identify the main non-tax advantages of the contemplated partnership structure at about the time that it was approved by Colgate's senior management.

The proposed Liability Management Partnership (the "Partnership") has been developed specifically for Colgate-Palmolive ("Colgate") to enable it to most efficiently manage the term structure of its liabilities, using predominantly its partners' capital. Normally an issuer's acquisition of its own debt involves three events, the acquisition of the debt, the retirement of the old issue and the issuance of substitute financing. The Partnership provides the opportunity to separate the timing of these events * * * by (i) acquiring Colgate debt in the market today, while it remains available, and (ii) placing such debt in "friendly hands," to be retired, modified or exchanged at an advantageous time in the future.

* * * *

Despite the current opportunity to acquire its debt, Colgate does not wish to immediately retire all of such debt and issue substitute financing. This reluctance is based in part on Colgate's current rate outlook (i.e., anticipation of gradual return to a positively-sloped yield curve) and in part on Colgate's desire not to permanently restructure all of such debt immediately.

The Partnership provides Colgate with flexibility to exchange the Colgate debt held by the partnership for newly issued Colgate debt of different maturity. Such exchanges may be effected as often and rapidly as Colgate deems appropriate. If Colgate attempted to refinance existing debt within a short time frame by repurchasing it and issuing new debt, transactions costs would rise dramatically. * * *

The Partnership also allows Colgate to effectively retire its debt, while leaving the debt outstanding for accounting purposes, and to take a position on rates by adjusting the relative sharing of Treasury risk by the partners. As Colgate bears a relatively greater share of the Treasury risk (i.e., losses in value of the Colgate debt attributable to interest rate increases) with respect to its debt, it has economically retired an increasing percentage of such debt and effectively changed its position with respect to interest rates.

The partnership's fulfillment of the liability management purposes for which it was designed would depend on the identity of Colgate's partners. Merrill undertook to procure them. During the summer of 1989, Taylor approached Hans den Baas (den Baas), the head of the Financial Engineering Group at ABN Bank New York (ABN New York), ⁴ concerning the possibility of ABN's participation in a partnership with Colgate. Taylor explained that the partnership would be used to acquire Colgate long-term debt for liability management purposes. He also stated that a contingent payment sale was contemplated, and that ABN's participation would be limited to 2-3 years. Den Baas forwarded Taylor's inquiry to Peter de Beer (de Beer), head of the legal department of ABN Trust Co., Curacao N.V. (ABN Trust), who would be responsible for structuring the legal aspects of the participation and negotiating the agreements. ABN Trust was engaged in the business of forming and managing Netherlands Antilles' entities to facilitate financial transactions. De Beer agreed to meet Colgate representatives in Bermuda during the middle of October 1989. He learned of the liability management aspects of the proposed partnership only when actual negotiations with Colgate began.

Based on prior dealings with Merrill, both den Baas and de Beer were already familiar with the CINS transaction and the defined role of the participating foreign partner. Taylor had discussed the transaction with den Baas during its development phase early in 1989. Taylor had previously approached den Baas to solicit ABN's participation in a CINS transaction on behalf of at least one other client. In that case too, den Baas had referred him to de Beer, who had represented ABN in the ensuing negotiations.

For a number of reasons, ABN was well suited for the role of majority partner in Colgate's liability management partnership. An ABN affiliate created and managed the foreign partners for each of the 11 section 453 partnerships promoted by Merrill. ABN New York provided financial engineering and other services to the foreign partners in each of the partnerships. Whether or not an understanding that ABN would collaborate as a copromoter existed from the outset, ABN would have had an interest in assuring the satisfaction of Merrill's clients in order to ensure the continuity of a valuable relationship with Merrill. It is unclear whether Colgate was aware of Merrill's relationship with ABN, but Colgate already had an established relationship with ABN. Acting as Colgate's lead bank in the Netherlands, ABN had underwritten a large foreign bond issue and performed other services in connection with Colgate's foreign operations. For these reasons, ABN could be trusted to cooperate in keeping the partnership "friendly", by yielding effective control to Colgate, by protecting the confidentiality of Colgate's debt restructuring activities, and by agreeing to relinquish its partnership interest at such time as Colgate might wish to acquire it. ABN's experience and sophistication in regard to European capital markets would assist the partnership in acquiring Colgate's Eurodollar debentures. As a major international bank, ABN possessed the liquidity needed to finance the venture, and, as a major derivatives dealer, it could accommodate, at little or no cost, Colgate's desire for an option to adjust their relative shares of interest rate exposure.

The third partner was to be an affiliate of Merrill. This provided Colgate with further reassurance. An equity interest would reinforce Merrill's incentive to continue to provide support and to act in a manner consistent with Colgate's interest when arranging the contemplated partnership transactions. Merrill would receive an advisory fee and transaction-based fees for initiating the partnership's asset transfers.

The ultimate challenge for Merrill in designing the liability management partnership was to find a way to integrate each step of the CINS transaction convincingly so that the transaction, as a whole, would stand up for tax purposes. The Swap Group devoted considerable effort to this task. Although the basic insight was incorporated in the initial "Colgate Partnership Transaction Summary" of July 28, 1989, it was refined in subsequent revisions of this document. The version entitled "XYZ Corporation: Revised Partnership Transaction Summary", dated August 17, 1989, set forth an outline of 10 steps to be taken by the partnership summarized as follows:

Step 1: The partnership is formed with contributions from XYZ Sub., A Corp. and B Corp. of $ 30 million, $ 169.3 million and $ 0.7 million, respectively.

Step 2: The partnership invests $ 200 million cash in short-term, floating-rate private placement securities pending acquisition of long-term XYZ debt. The private placement notes will be issued by highly rated issuers and will provide the partnership a return greater than comparably rated commercial paper or bank deposits.

Step 3: The partnership sells the private placement notes for a combination of cash, which will be used to acquire XYZ long-term debt over a period of 6 months, and LIBOR-based notes. "The purpose of the LIBOR notes will be to partly hedge the interest rate sensitivity of the long-term XYZ debt acquired by the Partnership." Depending on the maturity of the XYZ debt acquired, Merrill anticipated a ratio of 70-percent cash ($ 140 million) to 30-percent LIBOR Notes ($ 60 million).

Step 4: Some long-term XYZ debt is exchanged for newly issued medium-term XYZ debt.

Step 5: If a substantial amount of long-term debt was exchanged, the partnership would likely reduce its holding of the LIBOR Notes in order to rebalance its hedge. "Such a reduction would be necessary because the Medium-Term Debt, received in exchange for long-term XYZ debt, is less interest rate sensitive than the long-term XYZ debt. LIBOR Notes may either be sold directly or distributed to one or more Partners in a non-liquidating distribution."

Steps 6 and 7: Partnership assets are disposed of in the event that the desired investments cannot be made.

Step 8: A Corp.'s partnership interest is "possibly" redeemed at any time after 1 year following formation.

Step 9: The partnership is consolidated with XYZ for financial accounting purposes. The document advises that

it would be most reasonable for the Partnership to sell the LIBOR Notes and any other LIBOR-based assets if A Corp. is redeemed. Since the principal asset of the Partnership, other than LIBOR Notes and LIBOR-based assets, is likely to be XYZ debt and XYZ would be a 98% partner, the hedge protection provided by the LIBOR Notes and LIBOR-based assets is no longer necessary.

Step 10: B Corp. is eventually retired after a period of years.

In support of its characterization of the LIBOR Notes as a risk management tool, Merrill performed a series of quantitative analyses of the effect of a given change in the level of interest rates on the value of Colgate debt and LIBOR Notes in the partnership portfolio. These analyses purport to demonstrate that the interest rate sensitivity of the interest-only LIBOR Notes greatly exceeds that of fixed rate debt instruments of equal maturity and is comparable to that of long-term fixed rate debt. Thus, a 100 to 200 basis point increase or decrease in interest rates would produce roughly equal and offsetting changes in the value of $ 1 of LIBOR Notes, $ 2.34 of 9 percent 5-year Colgate debt, and $ 0.88 of 9-5/8 percent 30-year Colgate debt.

Pohlschroeder was impressed with Merrill's analysis. In an October 3, 1989, memorandum written for the purpose of recommending the "ABN Liability Management Partnership" to his superior, Colgate treasurer Brian Heidtke (Heidtke), Pohlschroeder explained how the composition of the partnership's portfolio would be planned to serve the purpose of "risk management within the partnership". "One aspect of importance is the interest rate exposure on the asset of the partnership which consists of Colgate debt. To minimize the exposure to ABN and Colgate, it is planned to convert a portion of the short-term notes to contingent LIBOR Notes as a hedge of the partnership's fixed rate assets." Although the hedge ratio would be determined through negotiations with ABN, he was confident that the partnership could acquire $ 140 million of Colgate debt, and that $ 60 million of LIBOR Notes would provide an appropriate level of protection. The plan was to adjust "the LIBOR note hedge" as needed in order "to achieve the ideal Colgate liability structure." Pohlschroeder envisioned "two possible situations arising in the future" which would call for the disposition of some of the LIBOR Notes. One was the exchange of long-term debt for medium- or short-term debt. "Because a shorter term instrument is less volatile, a smaller notional amount of the LIBOR Note is required for hedging purposes." A second situation was a change in the treasury risk sharing ratios. "The partnership is overhedged when Colgate decides to take more of the treasury risk and ABN reduces its share of the treasury risk. Conversely, as ABN's participation goes up, it needs more of a hedge in [the] form of the LIBOR notes."

Merrill provided Colgate with estimates of the expected costs of the contemplated partnership transactions. The "Perpetual Partnership Cost Component Analysis" reproduced in modified form below was prepared based on market conditions prevailing on September 1, 1989, and evidently assumed that the partnership would remain in existence indefinitely after these transactions were completed.

Perpetual Partnership Cost Component Analysis
($ millions)
	After Tax	Pretax 1
Net present value before	$ 25.47	---
transaction costs & advisory fee
Cost Components:
Origination of Citicorp Notes	1.32	$ 2.00
Remarketing of LIBOR Notes	1.29	1.95
Preferred returns to partners	0.74	1.12
Premium on debt tender	0.48	0.73
Legal expenses	0.17	0.25
Advisory fee	1.32	1.75
Total	5.32	7.80
Net present value of partnership	20.15	---
investment

The "origination" cost refers to the transaction cost that the partnership would incur on the exchange of private placement notes for cash and LIBOR Notes. The remarketing cost represents the transaction cost that would be incurred on the sale of the LIBOR Notes. The preferred return was an estimate of the additional allocation of income that the majority partner was expected to require. The advisory fee was payable to Merrill for its services. Colgate's management understood that most, if not all, of these costs would be borne by Colgate because all the liability management and tax benefits of the partnership transactions would enure to Colgate. They believed that the costs, though high in absolute terms, were reasonable in relation to the benefits that Colgate expected to receive from the partnership.

Liability management benefits would have been difficult to quantify for purposes of this comparison. The tax benefits, however, were calculable and greatly exceeded the expected transaction costs. Although the Perpetual Partnership Cost Component Analysis does not explain the derivation of the $ 25.47 million net present value that appears on the top line, this figure must be attributable almost entirely to tax benefits. A succession of summaries, cash-flow projections, and flip-chart presentations that Colgate received from Merrill between August and mid-October 1989, demonstrated how the sale of $ 200 million private placement notes for $ 140 million cash and $ 60 million market value of LIBOR Notes would result in $ 107 million taxable gain for the partnership and a net taxable loss for Colgate of approximately $ 90 million. If the foreign partner's interest were acquired and the LIBOR Notes sold within the 2-year period remaining for carryback of capital losses to the year of the Kendall divestiture, the present value of the tax savings achieved by this transaction, discounted at prevailing interest rates of 8-1/2 to 9-1/2 percent, would be roughly $ 25 million.

In a series of internal meetings and meetings with Merrill representatives during September and early October 1989, the liability management partnership proposal was presented to successively higher levels within Colgate's management. The vice president of taxation was now comfortable with the economic substance of the partnership. The treasurer concluded that this was a "uniquely suitable transaction for us." They, in turn, presented the tax and treasury aspects of the proposal to the chief financial officer and to the president of the company, who approved it. The decision was made to enter into negotiations with ABN.

ABN chose a form for its participation that would appear on its consolidated balance sheet as a loan to a third party rather than an equity investment. A Netherlands Antilles corporation named Kannex Corp., N.V. (Kannex), would be formed to borrow approximately $ 170 million from a bank and contribute it to the partnership. Kannex's stock would be held by two Netherlands Antilles stichtingen named Coign and Glamis. Stichtingen are foundations under Dutch law, have no owners, and conduct no commercial activities. Their sole purpose in this transaction would be to hold Kannex's stock. Control over the foundations would be exercised by their respective boards, of which de Beer would serve as chairman and other ABN Trust employees as members. The foundations would appoint ABN Trust to act as sole managing director of the corporation.

Financial arrangements for Kannex's participation were initiated by den Baas at ABN New York. Based on information about the proposed partnership that den Baas had received from Taylor, ABN New York prepared a credit proposal on behalf of Kannex, dated October 3, 1989. Since the borrower's only asset would be an interest in a portfolio expected to consist largely of Colgate long-term debt, ABN New York assessed Colgate's creditworthiness. Under the terms of the proposed credit facility, the bank would loan Kannex $ 170 million for 1 year at an interest rate of LIBOR plus 30 basis points, corresponding to the rate that the bank would have charged Colgate or a similarly rated company for a line of credit. Colgate was listed as the "client" on the credit proposal. This was because ABN New York viewed the financing transaction as a means of fostering closer banking relations with Colgate. As the credit proposal explained:

Colgate has been an important prospect for ABN New York Branch because of its strong financial condition and extensive international operations. Establishing a relationship has proven difficult because of the company's loyalty to its line banks. ABN's past involvement has been limited to facilities for Colgate subsidiaries. * * * We believe that the proposed transaction would provide an excellent entry into the parent's banking relationship.

Although the interest rate on the loan would provide an acceptable return commensurate with the level of the credit risk involved, ABN New York expected that the total returns to the bank from the loan transaction would be appreciably higher. The bank would also earn sizeable profits off the bid-ask spread on swaps necessary to stabilize Kannex's return from the assets in the partnership portfolio so that it could repay the loan. Because of the size of the loan, approval was required at three levels within the bank: The credit committee at ABN New York, the North American Credit Committee (NACC) in Chicago, and the Risk Management Dept. (RMD) in Amsterdam. ⁵

After approval by ABN New York, NACC reviewed the proposal together with a memorandum describing the partnership. On October 11, 1989, sent an advice to RMD recommending approval subject to a number of conditions, of which three are noteworthy:

1) The timing of the purchases and sales of the various securities be adhered to as proposed such that the credit risk is no greater than as outlined in partnership memo.

2) Interest rate risk is fully hedged.

3) Colgate's obligation to purchase Kannex's interest in the partnership by 11/30/89 [sic] is unconditional (will those proceeds be assigned to ABN?)

RMD advised NACC and ABN New York of its decision: "We agree on the condition that Merrill again verbally states to the partners that they will buy the MTN's at par on November 29, 1989." The reference to "MTN's", or medium-term notes, evidently denotes the private placement notes in which the partnership was expected to invest the partners' contributions pending acquisition of Colgate debt. The earlier oral assurance to which RMD refers may have been one that Merrill made to the first section 453 partnership in which ABN collaborated, the Nieuw Willemstad Partnership. Failing to locate a buyer for the partnership's private placement notes within the time frame required by the partners, Merrill itself became the counterparty, buying the private placement notes and issuing LIBOR Notes. A second condition was that the loan to Kannex be syndicated in order to reduce the credit risk.

ABN records indicate that the credit proposal was "approved per RMD". There is no record of any modification to the NACC and RMD conditions. Under ABN procedures, if credit conditions had been changed, the changes should be reflected in NACC files. Although there are cases in which a branch fails to advise NACC of changes in credit conditions or changes are made without documentation, such cases are rare.

Kannex was incorporated in the Netherlands Antilles on October 25, 1989, and issued shares with a total par value of $ 6,000, held in equal proportions by Coign and Glamis. Kannex's financial statements reflect accounts receivable for loans to the foundations in the amount of $ 6,000, indicating that they borrowed from the corporation the funds they used to acquire its stock. By "Revolving Credit Agreement" dated November 2, 1989, ABN's Cayman Islands Branch (ABN Cayman Islands) agreed to make loans available to Kannex in the aggregate amount of $ 180 million from November 2, 1989, through August 1, 1990. The shares of Kannex stock held by Coign and Glamis were pledged to ABN as security for the loans. Kannex entered into a management agreement with ABN Trust and a financial services agreement with ABN New York, executed by den Baas, under which ABN New York agreed to provide advice on hedging strategies to reduce Kannex's interest rate exposure and to provide other services at Kannex's request. The agreement does not make provision for either the amount or calculation of ABN New York's compensation.

Southampton, a wholly owned subsidiary of Colgate, was incorporated under Delaware law on October 24, 1989, for the purpose of becoming a partner in Colgate's liability management partnership. Belasco served as Southampton's president and Pohlschroeder as its vice president and treasurer. During the taxable years at issue, Southampton filed a consolidated return with Colgate.

Merrill Lynch MLCS, Inc. (MLCS), was incorporated under Delaware law on October 27, 1989. MLCS is the wholly owned subsidiary of Merrill Lynch Capital Services (Merrill Capital), which operates as the swap dealer for the Merrill Lynch Group. Taylor was MLCS's president and Paul Pepe (Pepe), a member of his staff, its vice president.

Negotiations were conducted at two meetings held in Bermuda on October 18 through 19, and October 27, 1989. The meetings were attended by, inter alia, Heidtke, Pohlschroeder, and Belasco from Colgate; Taylor and Fields from Merrill; de Beer and den Baas from ABN. By agreement dated as of October 27, 1989 (the Partnership Agreement), ACM was formed as a general partnership under New York law with its principal place of business in Curacao, Netherlands Antilles. ⁶ The partners' initial capital contributions were determined to be as follows:

Partner	Capital Contribution	Percentage of Total
Kannex	$ 169,400,000	82.63
Southampton	35,000,000	17.07
MLCS	600,000	.29
	205,000,000	1 100.00

The conduct of the business and affairs of the partnership would be under the direction of a partnership committee (the Partnership Committee) composed of a representative of each of the three partners. In general, action by the Partnership Committee required the assent of partners having an aggregate capital account balance equal to at least 99 percent of the total partners' capital. The affirmative concurrence of both Kannex and Southampton was therefore necessary for most partnership decisions. As its representative, Southampton appointed Pohlschroeder. Kannex appointed de Beer, and MLCS appointed Taylor.

The Partnership Agreement provided that, in general, income, gain, expense, and loss, as reported by the partnership for Federal income tax purposes, would be allocated among the partners in proportion to their respective capital accounts. As subsequent events would demonstrate, this general sharing provision did not fully reflect the partners' original understanding of the manner in which they would share the economic costs of partnership transactions.

Upon the occurrence of specified "Revaluation Events", the partnership would revalue its assets on its books, and any unrealized income, gain, expense, or loss inherent in its assets would be allocated among the partners as if realized in a sale of the assets at their fair market value. These Revaluation Events included: (i) a change in a partner's proportionate interest in partnership capital; (ii) a sale or exchange by the partnership of any Colgate debt instrument; (iii) an adjustment to the Yield Component (as defined below) with respect to Colgate debt; (iv) a contribution or distribution of partnership assets; (v) liquidation of the partnership; (vi) the last business day of each fiscal year; and (vii) after November 30, 1989, the properly executed request of any partner.

To allocate gains and losses arising in connection with Colgate debt instruments in the partnership portfolio for each revaluation period, the Partnership Agreement distinguished between that portion of any change in value attributable to changes in the general level of interest rates (the Yield Component) and that portion of any change in value attributable to changes in the market's perception of risks specifically associated with Colgate's credit quality (the Quality Component). Together, the Yield Component and Quality Component would capture all of the fluctuation in market value of the Colgate debt held by the partnership.

The Yield Component was initially allocated among the partners based on their respective capital interests. ⁷ Southampton could elect, however, to change its and Kannex's relative shares of the Yield Component to any level it desired within a specified range, on 5 days notice. It could increase its own share to as much as 49.7 percent, thereby reducing Kannex's share to 51 percent, and it could reduce its own share to as little as 10 percent, causing Kannex to take 89.7 percent.

The allocation of the Quality Component depended on whether Colgate's credit had improved or deteriorated during the relevant revaluation period. Improvement or deterioration was measured by the change in the implied spread of the Colgate debt yield over an index of the yield on U.S. Treasury securities. If Colgate's credit improved, the spread would narrow; if Colgate's credit deteriorated, the spread would widen. The Quality Component was the change in the value of the Colgate debt attributable to this change in the spread. The Partnership Agreement provided for the following Quality Component allocations: (a) For the first 50 basis point decline in value, 84.7 percent of the decline was allocated to Southampton, 15 percent to Kannex, as were subsequent increases within this 50 basis point range; (b) all declines beyond 50 basis points were allocated 99.7 percent to Southampton, and all other increases were allocated 99.7 percent to Southampton. MLCS's share of all changes was 0.3 percent.

The substantial risk shifting potential of the Yield Component option, which was of substantial value to Colgate's liability management scheme, proved relatively unproblematic for ABN because of the bank's ability to hedge interest rate risks outside the partnership through routine techniques employed by financial intermediaries in the derivative markets. Indeed, in its design of this option mechanism, Merrill's Swap Group took for granted ABN's ability to make accommodations in this manner.

The Quality Component provision was a bone of contention for the same reason that the Yield Component provision was not. A credit derivative that could be used by the bank to hedge the share of spread risk allocated to it under this provision was not available in the market at that time. ABN was loath to accept any spread risk for Kannex. On the advice of its tax lawyers, Colgate insisted, and the parties finally agreed, on a sharing formula that limited Kannex's exposure to 7-1/2 basis points (15 percent of a 50 basis point range).

The parties agreed on one further special allocation under the Partnership Agreement. From the date of the initial capital contributions through February 28, 1992, the first $ 1,241,000 of partnership income and gain for each fiscal year otherwise allocable to Southampton would be allocated to Kannex. This preferred return was not cumulative and was prorated daily. For this purpose, gains otherwise allocable to Southampton did not include unrealized gains resulting from revaluations of partnership assets. ABN had insisted on a preferred return as compensation to Kannex for participating in the spread risk of the Colgate debt. ABN intended that the amount would also include a small service fee for the adjustments that the bank would have to make to accommodate Southampton's discretionary management of interest rate exposure under the Yield Component provision. As the price for these benefits and as a substitute for the covenants and other legal protections that a lender in the position of Kannex would require as a condition for investing a great deal of money in Colgate debt obligations, Colgate considered the $ 1.24 million preferred return to be reasonable.

Southampton was required to maintain at least 2 percent of partnership capital. In the event that a substantial widening of the credit spread on Colgate debt caused Southampton's capital account to fall below the 2-percent threshold, unless prevented by insolvency, Southampton would contribute enough additional capital to continue to finance at least a certain minimum amount of the preferred return.

Section 4.03 of the Partnership Agreement governed the maintenance of the partners' capital accounts. The capital accounts would be increased by the amount of the partners' contributions, adjusted for allocations of partnership income, gain, expenses, and loss, and reduced by the fair market value of distributed property. Upon the occurrence of Revaluation Events, the capital accounts would be adjusted to reflect the mark-to-market revaluation of partnership assets.

Each of the partners was entitled to have its interest redeemed at fair market value upon request. Kannex could request redemption at any time after February 28, 1992. The other two partners could request redemption 1 year later. The redemption provision apparently was not the subject of negotiation. It was the intention of the parties that Kannex would be redeemed within 2 years, before its formal right under the Partnership Agreement ripened. The planned duration of Kannex's participation was dictated by the period prescribed for carryback of the capital loss to Colgate's 1988 taxable year. Colgate's plan afforded ABN the convenience of limiting the extent of Kannex's risk exposure.

The first meeting of the Partnership Committee (First Partnership Meeting) was held in Bermuda on October 27, 1989. The first noteworthy item of business was to appoint Merrill as qualified appraiser of partnership assets and to authorize both Merrill and ABN to make necessary arrangements for the purchase of three specified issues of Colgate debt: (1) $ 100 million principal amount of 8.4 percent private placement notes due in 1998 (Met Note) held by the Metropolitan Life Insurance Co. (Met Life); (2) $ 35 million principal amount of 9.625-percent notes due in 2017 (Long Bonds); (3) $ 5 million principal amount of 9.5-percent Eurodollar notes due in 1996 (Euro Notes).

Next, the Partnership Committee resolved that "in order to maximize the investment return on its assets pending the acquisition of Colgate-Palmolive Bonds", the partnership authorized Merrill to arrange for the purchase, in the form of a private placement, of $ 205 million of 5-year floating rate notes with an investor put option exercisable after about 15 to 24 months. Finally, according to the minutes, Pohlschroeder reported that he had communicated an offer to Met Life to purchase the Met Notes at a price within a stated price range, and that Met Life undertook to consider the proposal and review it with tax and legal advisers and, if interested, would come to Bermuda on November 17 in order to complete negotiations. The Partnership Committee authorized ABN Trust to conduct "such further discussions from outside the U.S. as are necessary with Metropolitan prior to such meeting."

During the proceedings in Bermuda, Taylor and Fields, on two separate occasions, presented Pohlschroeder and Belasco with revised estimates of the present value of transaction costs that were likely to be incurred in connection with the anticipated partnership transactions. According to one estimate, the total amounted to $ 6.95 million before tax, including $ 1.31 million origination cost on the sale of the private placement securities and issuance of the LIBOR Notes and $ 1.0 million for remarketing of the LIBOR Notes. The other estimate was higher: A total of $ 7.91 million before tax, including origination and remarketing costs of $ 2.0 million and $ 1.1 million, respectively. Colgate and Merrill did not discuss the costs of alternative short-term investments for the partnership's cash balances pending acquisition of Colgate debt.

On November 2, 1989, the partners' cash contributions in the amount of $ 205 million were deposited in the partnership bank account at ABN New York paying interest at a rate of 8.75 percent annually. The funds were withdrawn, at no cost, on the following day. By Private Placement Note Purchase Agreement between ACM and Citicorp, dated November 3, 1989, ACM acquired from Citicorp at par $ 205 million principal amount of floating rate notes due October 19, 1994 (Citicorp Notes or the Notes). The Citicorp Notes paid interest at the commercial paper rate plus 15 basis points, paid and reset monthly. The initial coupon was set at 8.78 percent and the first reset date was November 15. The Notes were rated AA by Standard & Poors. The holder had the option of tendering the Citicorp Notes for repayment on October 16, 1991, at 100 percent of the principal amount. The Citicorp Notes were not registered under the Securities Act, 15 U.S.C. sec. 77a (1933) and were not traded on an established securities market.

At the time of purchase, it was contemplated that the Citicorp Notes would be sold at the end of the month. Indeed, arrangements to sell the notes were already well underway. In several meetings beginning in late October, Pepe and other Merrill representatives discussed a proposed structure for the sale with the Capital Markets Group of the Bank of Tokyo's (BOT) New York Agency. Parallel discussions were held with the New York Branch of Banque Francaise du Commerce Exterieure (BFCE). During the first week of November, Merrill disclosed the specific terms of its proposal to each bank. The banks would purchase $ 175 million of the Citicorp Notes, paying 80 percent of the price ($ 140 million) in cash and the remainder with an installment purchase note providing for a 5-year LIBOR cash flow having a present value of $ 35 million. In addition, the banks would enter into collateral swaps with Merrill Capital that provided the banks with risk protection and an attractive return. Merrill had already prepared the legal documentation for the transactions. By facsimile dated November 9, BOT Capital Markets Group sent an urgent request for credit approval to the head office in Tokyo, attaching "all details of the transaction". Merrill required that the agreements be executed within a few days and any delay was likely to result in loss of the deal. On November 10, Merrill informed the banks that, at the asset seller's request, the transaction would be divided between them: BOT would purchase $ 125 million of the Citicorp Notes and BFCE would purchase $ 50 million.

If the amount and timing of the partnership's cash needs were so clearly foreseen at the beginning of November, it was in large part because by this time preparations for the acquisition of Colgate debt were also well advanced. The Met Note, Long Bonds, and Euro Notes that the Partnership Committee directed Merrill and ABN to acquire had been targeted for acquisition months earlier. Merrill's first "Partnership Transaction Summary", prepared in July, had contemplated that the partnership would purchase these three issues, using approximately $ 140 million cash from the sale of the private placement notes. During the summer, Pohlschroeder had told Fields that he knew that Met Life would be willing to sell the Met Note and could probably be induced to sell it immediately. He had arrived at the conclusion as a result of recent unsuccessful attempts by the insurance company to renegotiate the loan agreement. Both the Long Bonds and Euro Notes were identified as good candidates because substantial amounts of these public issues were held by institutions. Based upon his own study of market activities and consultation with traders during the first 6 to 9 months of 1989, Pohlschroeder was able to estimate how much of the Long Bonds and Euro Notes were available. Colgate's treasury department had Yordan perform further research on availability and price. By the beginning of October, Pohlschroeder felt confident that the partnership would meet Colgate's debt purchase target of approximately $ 140 million.

The only genuine question in regard to the Met Note was price. In late September, Pohlschroeder contacted Met Life to indicate a possible interest in purchasing the Met Note. On October 23, a few days before he returned to Bermuda to conclude the Partnership Agreement, Pohlschroeder prepared himself for negotiations with Met Life by conferring with Yordan. His notes from that conversation conclude with a reference to the date November 17, which is circled. As the minutes of the First Partnership Meeting reflect, Pohlschroeder contacted Met Life again from Bermuda to invite a representative of the insurance company to negotiate a sale of the note in Bermuda on November 17. The statement in the minutes that Pohlschroeder had communicated an offer on specific terms appears to have no basis in fact, however. It is clear that Pohlschroeder refused to enter into any discussion of terms on that occasion. During the 3 weeks prior to the meeting scheduled for November 16 and 17, Pohlschroeder received a telephone message from Met Life stating the insurance company's asking price. He did not return the call. There were no negotiations prior to the scheduled meeting, either by Colgate within the United States, or by ABN Trust, the partnership's authorized representative for this purpose, outside the United States.

Yordan attended the meeting of the Partnership Committee in Bermuda on October 27 in order to advise the partnership concerning availability and prices of Colgate's Long Bonds and Euro Notes. At this time, Pohlschroeder prepared notes regarding standing orders that ACM intended to issue to Merrill for the purchase of the Long Bonds and Euronotes. The notes apparently reflect a decision as to the timing of these transactions: "Peter de Beer, Curacao will give instructions from C to M.L. after Citi's purchase".

The second partnership meeting was held in Bermuda on November 17, 1989. A representative from Met Life came to Bermuda at this time to negotiate the sale of the Met Note. The negotiation was not lengthy; price was the only issue, and the parties split the difference between their respective offers. By Note Purchase Agreement dated November 17, 1989, and effective December 4, 1989, ACM purchased $ 100 million principal amount of the Met Note for the aggregate purchase price of $ 99,291,000 plus accrued interest.

Pohlschroeder reported the successful conclusion of the agreement to the Partnership Committee. According to the minutes, he pointed out that the partnership would now require cash in order to perform its obligations under the Note Purchase Agreement with Met Life. In addition, this investment "would create a risk to the Partnership in the event that interest rates increased because the Met Bonds had a fixed rate of interest." Pohlschroeder recommended "that the Partnership hedge its risk by purchasing notional principal contracts with a floating rate of interest." By resolution of the Partnership Committee, Merrill was authorized to arrange the sale of $ 175 million principal amount of the Citicorp Notes to one or more of BOT, BFCE, and Mitsubishi Bank "for cash and other LIBOR-based consideration, upon substantially the terms of a draft Installment Purchase Agreement presented to the meeting".

One other significant item of business at the second partnership meeting was the adoption of the "Investment Policy Guidelines" (Investment Guidelines). Weeks before the formation of the Partnership, Pohlschroeder had reported to Heidtke that Colgate would ensure in the Partnership Agreement that the company's own cash management policies would be used as guidance to maintain "liquidity * * * required to facilitate the buyback of long-term debt". As it turned out, the partners were not yet ready to adopt such policies at the time the Partnership Agreement was executed. The primary objective of the belated Investment Guidelines was "to preserve principal". To this end, temporary cash balances were to be invested in a portfolio of short-term money market instruments selected so as to achieve both a high degree of liquidity and diversification. Upon the liquidation of most of its investment in unregistered 5-year notes of a single issuer, the partnership would be in a position to implement its Investment Guidelines.

On November 27, 1989, ACM sold $ 175 million principal amount of the Citicorp Notes to BOT ($ 125 million) and BFCE ($ 50 million). The aggregate consideration consisted of cash in the amount of $ 140 million and eight notes requiring quarterly payments of 3-month LIBOR for 20 quarters commencing March 1, 1990, on a notional principal amount of $ 97.76 million (LIBOR notes). ⁸ The LIBOR notes were not registered under the Securities Act of 1933 and were not readily tradable on an established securities market. At the time of the transaction, Standard & Poors rated the senior debt of BOT AA and that of BFCE AAA.

The aggregate amount of the consideration paid by the banks included the discount, or origination cost, that Merrill determined it would need to charge for its role in the arrangement and intermediation of the transaction. The discount was 5/8 percent of the par value of the Citicorp Notes, or $ 1,093,750. The banks issued the LIBOR Notes at a price equal to the aggregate consideration less the cash. The notional principal amount of the Notes was the amount that was required at current market swap rates to give the expected LIBOR cash flows a present value equal to this price.

The following table summarizes the various costs associated with the Citicorp Notes and LIBOR Notes:

Citicorp Notes aggregate par amount	$ 175,000,000
Transaction price	99.375%
Transaction value	173,906,250
Accrued interest (12 days @ 8.65 percent)	504,564
Total consideration	174,410,814

	BOT	BFCE	TOTAL
Citicorp Notes par value	$ 125,000,000	$ 50,000,000	$ 175,000,000
Accrued interest	360,403	144,161	504,564
Cash payment	(100,000,000)	(40,000,000)	(140,000,000)
Cost of LIBOR Notes	25,360,403	10,144,161	35,504,564
Origination cost	(781,250)	(312,500)	(1,093,750)
Issue price/present value	24,579,153	9,831,661	34,410,814
of partner Notes
Notional principal of	69,850,000	27,910,000	97,760,000
LIBOR Notes

On the same day that the partnership acquired the LIBOR Notes for the stated purpose of hedging the partners' exposure to interest rate risk associated with the Colgate debt, Southampton served notice of an adjustment to the Yield Component sharing ratio. Desiring greater exposure, Southampton increased its share of the Yield Component from 16.7 percent to 29.7 percent.

ACM invested the $ 140 million cash received in the sale in several commercial paper issues (time deposits and certificates of deposit (CD's)) maturing December 4, 1989, and bearing interest at 8.15 to 8.20 percent. Upon maturity, these funds became available at no transaction cost to finance the following purchases of Colgate debt between December 4 and 8:

$ 100 million principal amount of the Met Note for $ 99,291,000 plus accrued interest;

$ 1 million principal amount of Euro Notes for $ 1,025,500 plus accrued interest;

$ 4 million principal amount of Euro Notes for $ 4,102,000 plus accrued interest;

$ 31 million principal amount of Long Bonds for $ 31,493,396 plus accrued interest.

During November, the groundwork was being laid for the disposition of some of the LIBOR Notes that ACM would acquire in the sale. A memorandum that Merrill prepared for Colgate entitled "Analysis of Partnership Hedging Activity," dated November 13, 1989, purports to demonstrate quantitatively how either an increase in Southampton's share of the interest rate volatility of the Colgate debt from 30 percent to 50 percent or an exchange of the Long Bonds for a new issue of 5-year Colgate debt would warrant a reduction in the amount of the LIBOR Note hedge in the partnership portfolio by approximately $ 10 million. Merrill reasoned that, in either case, ABN's interest rate exposure would fall by about 30 percent, and a 30-percent reduction in the size of the partnership's hedge would leave the bank's net exposure unchanged. Sometime in November, Pepe approached Neil Schickner (Schickner), head of the Capital Markets Desk at the New York Branch of Sparekassen SDS (Sparekassen). ⁹ Pepe proposed a transaction involving the purchase of the BFCE LIBOR Notes by Sparekassen and collateral swaps that provided Sparekassen with risk protection and an attractive return. Schickner was already familiar with the transaction structure; at about the same time, Pepe offered him one or two similar deals in connection with other section 453 partnerships. On December 5, in order to conclude the deal, Schickner notified the bank's headquarters in Copenhagen that he was reserving a credit line in the amount of $ 10 million.

The third partnership meeting took place on December 12, 1989, in Curacao. On behalf of Southampton, Pohlschroeder served notice of an adjustment in the Yield Component, whereby Southampton elected to increase its share of interest rate exposure to 39.7 percent. Next, the Committee voted to accede to a Colgate proposal to exchange $ 4.7 million aggregate principal amount of the Long Bonds plus $ 4,165 cash payment for $ 5 million aggregate principal amount of new 3-1/2 year fixed rate debt.

Macauley Taylor next stated that the debt exchange contemplated by the foregoing resolutions would reduce the Partnership's exposure to the risk of interest rate fluctuations and recommended that the Partnership reduce its position in the variable rate instruments purchased to hedge against such exposure. He reported that a reduction of approximately 30 percent in the hedging provided by the Installment Purchase Agreements executed by the Partnership on November 27, 1989 would be economically advisable. He noted that this reduction would not adversely affect Kannex because of the adjustment of sharing of Yield Component effected by the notice dated December 12, 1989, from Southampton to the Partnership Committee.

It was decided that the BFCE Notes would be distributed to Southampton as a partial return of capital. ACM assigned the BFCE Notes to Southampton as of December 13. By Assignment Agreements dated December 22, 1989, Southampton agreed to assign the notes to Sparekassen for aggregate consideration of $ 9,406,180. The discrepancy between the issue price at which the Notes had been acquired ($ 9,831,661) and the price that Southampton received on their sale ($ 9,406,180) was largely attributable to a bid-ask spread of $ 390,000. The bid-ask spread reflected the margins above and below mid-market value that Merrill deemed necessary in order to originate and sell the Notes. Estimating cash flows under the Notes from ask-side swap rates and discounting at a spread below LIBOR in its valuation of the Notes at issuance, Merrill was able to create an attractively priced liability for BFCE. Estimating cash flows under the Notes from bid-side swap rates and discounting at a spread above LIBOR in its valuation of the Notes for purposes of the assignment transaction, Merrill was able to create an attractively priced asset for Sparekassen. The remaining portion of the discrepancy, $ 35,481, was due to a decline in market interest rates over the 3-week period since the issuance of the LIBOR Notes, which caused them to lose value.

For Federal income tax purposes, ACM treated the sale of the Citicorp Notes as a contingent payment sale, governed by section 15a.453-1(c)(3), Temporary Income Tax Regs., 46 Fed. Reg. 10714 (Feb. 4, 1981). As there was no stated maximum selling price and all payments on the LIBOR Notes would be received over a fixed period of 6 taxable years, ACM recovered its basis in the Citicorp Notes ratably over 6 years. On Form 1065, U.S. Partnership Return of Income, for FYE 11/30/89, the partnership reported capital gain of $ 110,749,239.¹⁰ The gain was allocated among the partners in proportion to their capital accounts as shown on the November 30, 1989, revaluation worksheet: $ 91,516,689 to Kannex, $ 18,908,407 to Southampton, and $ 324,144 to MLCS. The parties to this proceeding have agreed that the partnership's tax basis in the LIBOR Notes immediately after the sale was $ 146,253,803, an amount that exceeded the cost of the Notes by the gain recognized on the sale.

Kannex paid neither U.S. nor foreign tax on its 82.63 percent distributive share of the partnership capital gain. On its consolidated Federal income tax return for 1989, Colgate reported a net capital loss attributable to Southampton in the amount of $ 13,521,432, representing the difference between Southampton's distributive share of the partnership capital gain ($ 18,908,407) and the capital loss that Southampton recognized on the sale of the BFCE Notes to Sparekassen ($ 32,429,839). ¹¹

During the years at issue, Colgate retained Arthur Andersen & Co., as its accountants. In connection with the audit of Colgate's consolidated financial statement for 1989, the audit engagement team and Arthur Andersen's tax team discussed with Colgate's treasury, financial, and tax department personnel how to report the partnership and its activities for financial accounting purposes. Representatives of Merrill were also present. An outline was presented of the planned sequence of steps by which the partnership would borrow to redeem ABN's interest in October 1991 and recognize the remainder of the total $ 100 million capital loss. The auditors were concerned that recognition of the large tax loss without a corresponding book loss would leave Colgate with an outside basis considerably lower than the value of the partnership assets. ¹² The deferred tax liability associated with this built-in gain would have to be recognized for financial accounting purposes, unless the company could demonstrate an "exit tax strategy". With Merrill's assistance, Colgate explained how the low outside basis and deferred tax liability would be eliminated through a series of contemplated tax-free asset and stock transfers among Colgate affiliates some time after 1992. The auditors were of the opinion that until it became clear that they would be sustainable, for the most part the tax benefits of the transaction should not be recognized for financial accounting purposes. They understood from Colgate's account of the partnership, however, that sizable transaction costs would be incurred in connection with its activities. Colgate explained that only a minor amount of these costs would be shared with the other partners. Colgate would bear approximately $ 5 million, including all of Merrill's advisory fee of $ 1.7 million as well as approximately $ 2 million to originate and remarket the LIBOR Notes. The auditors agreed with Colgate that tax benefits from the partnership could be recognized to the extent of the net-of-tax amount of these transaction costs.

On the issue of consolidation, the auditors endorsed Colgate's position. Consolidation would not be required until ABN's retirement, chiefly because the Colgate debt was not effectively retired to the extent that ABN was sharing changes in its market value. In the meantime, since Colgate was using its position in the partnership essentially as a hedge of its liabilities, and would otherwise have used swaps or other conventional hedging operations to accomplish the same purposes, its investment in ACM should be treated in the same manner for financial accounting purposes as a swap. This would entail the recognition of mark-to-market changes in the value of its equity interest on its financial statements.

The Curacao office of Arthur Andersen served as accountants for ACM. In the course of their review of the results for FYE 11/30/89, the auditors noted two problems with the partnership's financial statements. The first problem was that the $ 1,093,750 discount on the sale of the Citicorp Notes was not reflected in the income statement. The second problem was that the partnership had included this discount in the book value of the LIBOR Notes, contrary to provisions of the Partnership Agreement that required partnership assets to be restated at fair market value on the last day of the fiscal year. Following consultations with the New York office of Arthur Andersen and with Colgate, in February 1990, the audit engagement manager briefed his colleague on the status of the problem:

Colgate does not want the cost to sell of US $ 1,093,750 * * * in the November 30, 1989 income statement of ACM. The reasons are mainly tax driven, as inclusion might set the IRS on top of the reasons why the partnership was constructed in the first place and thus the planned tax losses may be denied by the IRS. We, in cooperation with Steve Rossi of our New York office, were requested to think with Colgate in order to keep the cost to sell out of the balance sheet. [Emphasis added.]

One proposal under consideration was as follows:

Leave the LIBOR notes on the balance sheet as they are and reason that one third of the notes will be distributed to Colgate by 1990 and that the remainder of the notes is eventually for the account of Colgate too. This would require a side letter to the partnership agreement stating that the LIBOR notes are the one exception to the valuation rules which now state valuation at market and would state valuation at market and would then state valuation at market increased by the cost to sell the original Citicorp notes.

The partnership followed this approach. Pursuant to the "Summary of Financial Accounting Policies" (Accounting Policies), adopted 2 weeks later at the fourth partnership meeting, the LIBOR Notes would be:

carried on the books of the Partnership at cost, and adjusted * * * (I) for amortization of principal on a straight-line basis; and (ii) for movements in interest rates upon the following events: (a) distribution of any * * * [LIBOR] notes; (b) redemption of any Partner; and liquidation of the Partnership.

Thus, the LIBOR Notes were initially booked at a cost that included the $ 1,093,750 transaction costs incurred on their origination. The cost would be amortized over the life of the investment. This amortization would constitute a charge against income, offset by accrued payments on the Notes. If any of the LIBOR Notes were distributed or a partner was redeemed, the amortized balance would be adjusted for changes in value due to interest rate movements and increased by the previously amortized portion of the origination cost. This convention had the effect of ensuring that the origination cost would be borne solely by the partner(s) that held an interest in the Notes, directly or indirectly, at the time they matured or were sold.

The Accounting Policies do not specify the methodology to be used in revaluing the LIBOR Notes to reflect changes in interest rates. The methodology would differ depending on whether the book value was meant to reflect the minimum price at which the Notes could be purchased in the market (ask value), the maximum price at which they could be sold in the market (bid value), or the midpoint between the two (mid-market value). The understanding among the partners on this issue is revealed by the partnership's actual accounting practice. In pricing the LIBOR Notes at issuance, Merrill used an ask-side valuation methodology. The Notes were originally booked at a value based on this price; the bid value of the Notes at that time, as determined by Merrill, was about $ 1.3 million lower. Thereafter, book value was consistently adjusted to reflect the current ask price. This convention had the effect of ensuring that the bid-ask spread would be borne solely by the partner(s) that held an interest in the Notes, directly or indirectly, at the time they matured or were sold.

Finally, unlike the policies governing the revaluation of Colgate debt, there is no provision in any agreement for adjusting the book value of the LIBOR Notes to reflect changes in the credit quality of the issuers. As a result, any credit risk would be borne only upon the sale of the Notes to a third party.

As a corollary to the Accounting Policies described above, the partners agreed that in the event that any of the LIBOR Notes were distributed to a partner before maturity, they would be distributed at book value. As a result, the distributee partner's capital accounts and outside basis would be reduced. This reduction would result in the distributee in effect paying the full origination cost and bid-ask spread attributable to the distributed LIBOR Notes. In connection with the distribution of the BFCE Notes to Southampton, as of December 13, the partnership's assets were revalued. The book value of the BFCE Notes was adjusted to $ 10,133,540. For financial and tax accounting purposes, Southampton's capital account was reduced by this amount, resulting in a decrease in its ownership percentage from 16.89 percent to 12.60 percent.

7. Final Stage of Colgate's Partnership Strategy

ACM made additional purchases of Colgate debt from the marketplace as follows:

			Aggregate
Issue		Principal	Purchase
Acquired	Date	Amount	Price
Euro Notes	6/1/90	$ 5,000,000	$ 5,154,861
Long Bonds	9/6/90	4,000,000	3,864,622
Euro Notes	9/11/90	1,750,000	1,859,132
Long Bonds	9/12/90	6,000,000	5,852,290
Euro Notes	10/23/90	2,000,000	2,159,389

There were also exchanges between ACM and Colgate of the Met Note and approximately one-third of the Long Bonds. In January 1990, ACM exchanged the Met Note for a new Colgate Note with substantially identical terms. This new note was, in turn, exchanged on July 26, 1990, for the purpose of rescheduling certain payments.

ACM made two exchanges of the Long Bonds, which totaled $ 10 million. On December 13, 1989, ACM exchanged $ 4.7 million principal amount of Long Bonds for $ 5 million principal amount of Colgate 8.72-percent notes due June 13, 1993. On March 1, 1991, the partnership exchanged $ 4.85 million principal amount of Long Bonds for $ 5 million principal amount of Colgate Notes due in 1994. The exchanges of the Long Bonds had the effect of reducing Colgate's original average debt maturity of 13 years by only 2 months (or 1 percent).

At the end of August 1990, Colgate's treasury concluded that a significant change had occurred in the interest rate environment. Inflationary expectations and the prospect of war in the Persian Gulf were causing a rise in long-term interest rates and a steepening of the yield curve. Under these conditions, the value of Colgate debt held by the partnership would fall. Reversing its policy over the past 10 months of accepting substantially greater interest rate exposure than its pro rata share, Colgate caused Southampton to reduce its share of the Yield Component to 10 percent, effective September 6. Thereafter, Southampton adjusted the Yield Component Sharing ratio on two more occasions, maintaining its exposure between 10 and 20 percent.

Contrary to the expectations of Colgate's management, long-term interest rates declined. By the spring of 1991 Colgate's treasury department identified a constellation of factors favoring consolidation of the partnership and retirement of its Colgate debt holdings in the near future. Not only were general interest rates lower, but the credit spreads on Colgate debt had narrowed appreciably, reflecting stronger prices for the company's stock and diminished takeover risk. Moreover, efforts to locate Colgate debt available for purchase were no longer successful.

By Partnership Interest Purchase Agreements dated June 25, 1991, Colgate acquired a 38.31-percent interest in ACM from Kannex for $ 85,897,203, and Southampton acquired a 6.69-percent interest in ACM from Kannex for $ 15 million. As a result of these transactions, Kannex's ownership percentage declined to 43.04 percent. The shift in ownership was accompanied by a revaluation of partnership assets. Changes in asset values were allocated among the partners' respective capital accounts and the purchase price was determined based upon the balance of Kannex's account. In this process, the book value of the BOT LIBOR Notes was adjusted to reflect their current market value increased by $ 781,250, the full amount of the origination cost attributable to the notes, and 88 percent of the adjustment was allocated to Kannex's capital account. Although not specifically provided for by the partnership's Accounting Policies, a revaluation of the LIBOR Notes under these circumstances was evidently consistent with the agreement among the partners that Kannex would bear none of the origination cost.

By agreement dated November 27, 1991, ACM redeemed the remainder of Kannex's partnership interest for $ 100,775,915. The redemption was financed in part with cash and in part with the proceeds of a loan from Citibank secured by the partnership's holdings of Colgate debt. In accordance with the Accounting Policies, partnership assets were revalued and unrealized income, gains, and losses were allocated among the partners. For this purpose, a value of $ 13,974,304 was assigned to the BOT LIBOR Notes, reflecting their current market value increased by the $ 781,250 origination cost attributable to them. The liquidating distribution that Kannex received was equal to the resulting balance in its capital account.

At the twelfth partnership meeting, held on December 5, 1991, it was observed that

as Colgate and a subsidiary, Southampton, owned 99.4% of the Partnership, the principal Partners' net economic exposure to the risk of interest rate fluctuations in the value of the Colgate debt was effectively minimal, and the Partnership need not maintain its position in the instruments purchased to hedge against such exposure.

Moreover, the LIBOR Notes "were a highly volatile investment and * * * without the need to hedge interest rate risk, it was unwise for the Partnership to hold them." "[Short-term interest rates had declined steadily in recent months, thereby reducing the value of the instruments." It was resolved that the partnership would sell the LIBOR Notes. The final substantive comment of the meeting was delivered by Belasco, representing Colgate, who noted that "the Partnership had achieved substantially all of its objectives in connection with the acquisition of Colgate bonds and related debt management."

On December 17, 1991, shortly before the close of Colgate's 1991 taxable year, ACM sold the BOT LIBOR Notes to BFCE for $ 10,961,581. The notes had fallen considerably in value owing to the decline in market interest rates. Eight and one-half percent at the time the first payment on the notes had been determined, 3-month LIBOR was below 5.7 percent when the last payment was determined. The price at which the BOT LIBOR Notes were sold also reflected a remarketing cost corresponding to the bid-ask spread, equal to $ 440,000.

The economic loss incurred on the sale of the LIBOR Notes was more than compensated for by the tax loss. On its Form 1065 for FYE 12/31/91, ACM reported a capital loss in the amount of $ 84,997,111. Colgate claimed $ 84,537,479 as its own and Southampton's combined distributive shares of this loss on its consolidated corporation tax return for the 1991 taxable year. By amended return, Colgate carried this loss back to 1988. The total net tax loss that Colgate achieved through the CINS transaction exceeded $ 98 million.

As a result of the consolidation of ACM on Colgate's financial statements for 1991, Colgate's reported outstanding long-term indebtedness declined by $ 124.1 million, ¹³ approximately one-half of the overall decline in long-term debt during this year. As of December 31, 1991, the value of Southampton's and Colgate's capital accounts plus the proceeds that had been received from sale of BFCE LIBOR Notes exceeded the costs of their combined investment in the partnership by approximately $ 5.42 million, representing a pre-tax internal rate of return of 4.7 percent. More than 2 percentage points of this return was attributable to the appreciation of the partnership's Colgate debt caused by further declines in interest rates in the month following Kannex's redemption.

The origination and remarketing costs of nearly $ 2 million that Colgate incurred through its partnership strategy represented the costs of a highly complex structure of collateral swaps arranged and executed by Merrill for the purpose of accommodating the investment in and divestment of assets qualifying for contingent payment sale treatment. This section outlines the transactions that Merrill entered into with BOT, BFCE, and Sparekassen between the issuance of the LIBOR Notes in November 1989 and the partnership's sale of the BOT LIBOR Notes in December 1991.

To secure the participation of BOT and BFCE in the contingent payment sale desired by ACM, Merrill's Swap Group offered each of the banks a "structured transaction." ¹⁴ The structured transaction consisted of two swaps to be executed in conjunction with the contingent payment sale, a basis swap related to the asset that the banks would be purchasing and a hedge swap related to the liability that they would be issuing to finance the purchase. The banks' counterparty in these swaps was Merrill Capital. Both sets of swaps were entered into on November 27, 1989.

Under the basis swaps, BOT and BFCE were obligated to make monthly payments to Merrill Capital at the 1-month commercial paper rate plus 15 basis points on notional amounts of $ 125 million and $ 50 million, respectively. These payments were equivalent to the interest that the banks received on the Citicorp Notes. In exchange, Merrill Capital was required to make monthly payments to the banks at a rate of 1-month LIBOR plus 25 basis points on identical notional amounts. After 3 months the spread over LIBOR that Merrill Capital was required to pay increased to 40 basis points and in the case of BOT, to 50 basis points after another month, unless on any payment date Merrill Capital elected to terminate the basis swaps and purchase the Citicorp Notes from the banks at par.

The basis swaps served a risk management function for the banks. The net cash flows resulting from the combination of the Citicorp Notes with the basis swaps were tied to LIBOR, the index in terms of which BOT and BFCE, like international banks generally, conducted most of their business. The step-up provisions were negotiated at the request of the banks and were designed to give Merrill Capital a financial incentive to make arrangements for resale of the notes as quickly as possible. Merrill Capital would forgo the exercise of its call option only in the event of a substantial decline in Citicorp's credit that caused the value of the Citicorp Notes to fall by more than the cost of paying the premium.

Under the hedge swaps, Merrill Capital was obligated to make quarterly payments over 5 years equivalent to the LIBOR Note payments that the banks were required to make to ACM. In return, BOT agreed to pay the sum of $ 25 million in 20 equal quarterly installments plus interest on the unpaid balance at a rate of LIBOR minus 18.75 basis points. BFCE agreed to pay the sum of $ 9,831,661 in 20 equal quarterly installments plus interest on the unpaid balance at a rate of LIBOR minus 25 basis points. In addition, there were two upfront payments: Merrill Capital paid $ 35,000 to BOT, and BFCE paid $ 168,339 to Merrill Capital. Like the basis swaps, the hedge swaps served a risk management function for the banks. They were designed to replicate the portfolio effects of partly financing the purchase of the Citicorp Notes with a conventional amortizing loan, whose value would not be affected by changes in LIBOR, rather than with the highly volatile LIBOR Notes. ¹⁵

The structured transactions were designed to be remunerative for the dealer, Merrill Capital. Under the basis and hedge swaps, the present value of the banks' payment obligations exceeded the present value of Merrill Capital's obligations. In this way, the swaps were expected to result in the transfer from the banks to Merrill Capital of the 5/8 discount incurred by ACM on the contingent payment sale. To the extent that the basis swap continued beyond 3 months, Merrill Capital would return some or all of the discount to the banks through the stepped up LIBOR payments.

BOT and BFCE would not have participated in the hedge swaps if they did not also perceive an opportunity to profit. Internal bank documents confirm that those who negotiated the structured transactions with Merrill believed that they offered "very attractive", "extremely favorable" terms. According to calculations performed by petitioner's expert Tanya Beder (Beder), ¹⁶ the transactions effectively provided both banks with funding at a cost 39 basis points lower than that available in the direct interbank market. The 39 basis points in savings represents each bank's net present value gain from the structured transaction expressed in relation to the amount of the financing involved. Beder's valuation analysis is useful for identifying how the banks expected to gain overall while losing money on both the basis and hedge swaps.

Valuation of the Positions of
BOT and BFCE as of 11/27/89
($ millions = mm)
	BOT	BFCE
LIBOR Notes
Price rec'd from ACM	$ 24.58 mm	$ 9.83 mm
Mid-market value	(24.05) mm	(9.61) mm
Citicorp Notes
Price paid to ACM	(124.58) mm	(49.83) mm
PV of expected sale proceeds
rec'd by banks	125.39 mm	50.15 mm
Hedge Swap
Liability leg	(24.88) mm	(9.77) mm
Asset leg	24.08 mm	9.62 mm
Basis Swap
Asset leg	18.77 mm	7.43 mm
Liability leg	(18.22) mm	(7.29) mm
Merrill's cancellation option	(0.89) mm	(0.29) mm
Up-front payment	0.04 mm	(0.17) mm
Net Present Value	222,586	88,323
Implied Funding Spread
Under LIBOR	1 0.39%	1 0.39%

This analysis indicates that the source of the banks' expected gains was Merrill's pricing of the Citicorp Notes and LIBOR Notes for purposes of the contingent payment sale. These prices reflect sizeable bid-side and ask-side spreads. Transaction spreads generally tend to be wider for structured transactions than for direct market transactions because structured transactions are customized to meet the needs of the end users and often incorporate a premium to the dealer for innovations that competitors are unable to replicate. The spreads implied in Merrill's pricing of the Citicorp Notes and LIBOR Notes represented the costs of the financial engineering that the contingent payment sale required. Accordingly, the costs were charged to ACM. The banks acquired the Citicorp Notes at the bid price and issued the LIBOR Notes at the ask price. The spreads on these two instruments could have been expected, at the time of the contingent payment sale, to result in the transfer of a total of about $ 1.8 to $ 1.9 million in value from ACM to the banks. The banks could have expected to retain approximately $ 300,000 of this value. See diagram 1 infra p. 63. ¹⁷

It was the understanding of BFCE that Merrill would arrange for the resale of the Citicorp Notes after only 1 month, well in advance of the date that the step-up in Merrill's payments took effect. The written agreement contained no such provision, but Merrill found a buyer, and BFCE sold its $ 50 million principal amount of Citicorp Notes on December 22, 1989. At the same time, the basis swap between Merrill Capital and BFCE was canceled. In January 1990, the basis swap with BOT was terminated, and the remaining $ 125 million principal amount of Citicorp Notes was resold.

Merrill arranged another structured transaction to facilitate Southampton's sale of the BFCE LIBOR Notes to Sparekassen on December 22, 1989. Under the hedge swap between Merrill Capital and Sparekassen, Sparekassen was obligated to make quarterly payments equivalent to those it was entitled to receive from BFCE under the LIBOR Notes. In return, Merrill Capital was required to pay $ 9,406,180, an amount that corresponded to the purchase price of the notes, in 20 equal quarterly installments, together with interest on the unpaid balance at a rate of LIBOR plus 35 basis points. The spread over LIBOR increased to 85 basis points after March 1, 1990, if Merrill did not first exercise its right to call the notes at a price equal to the unpaid principal balance and terminate the swap. From Sparekassen's perspective, the structured transaction was similar to investing in an amortizing loan that paid a margin over LIBOR, rather than in volatile LIBOR Notes. From Merrill Capital's perspective, the transaction provided an asset whose volatility matched and offset the volatility of its liability under the hedge swap with BFCE or BOT. The step-up in Merrill Capital's payment obligations provided it a financial incentive to exercise its call right and cancel the swap. Petitioner's expert, Beder, concluded that as of the time of its acquisition of the BFCE Notes, Sparekassen could have expected a net present value benefit of $ 7,208, equivalent to a return on its investment of 41 basis points more than that available in the direct interbank market.

[SEE Diagram 1 Flow of Benefits in 11/17/89 Structured Transaction IN ORIGINAL]

[SEE Diagram 2 Flow of Benefits in 12/22/89 Structured Transaction IN ORIGINAL]

[SEE Diagram 3 Flow of Benefits in 12/17/91 Structured Transaction IN ORIGINAL]

Valuation of Sparekassen's
Position on 12/22/89
( $ millions = mm)
LIBOR Notes
Price paid to Southampton	(9.41) mm
Mid-market value	9.63 mm
Hedge Swap
Asset leg	9.58 mm
Liability leg	(9.63) mm
Merrill's cancellation option	(0.17) mm
Net Present Value	7,208
Implied Return Over LIBOR	1 0.41%

As in the structured transaction that Merrill designed for the other two banks, Sparekassen could expect to lose money on the swap; the source of its gain is the bid-side spread implied in Merrill's pricing of the LIBOR Notes. The transaction pricing resulted in the transfer from Southampton to the bank of more than $ 200,000 in value, most of which would ultimately enure to the benefit of Merrill Capital. See diagram 2 supra p. 63.

By agreements among BFCE, Sparekassen and Merrill Capital, the BFCE LIBOR Notes and the two hedge swaps related to them were terminated during 1990.

Merrill arranged another hedge swap for BFCE in conjunction with the bank's purchase of the BOT LIBOR Notes from ACM for $ 10,961,581 on December 17, 1991. The structure and function of this swap were for the most part identical with those of the hedge swap between Merrill Capital and Sparekassen. BFCE agreed to pay Merrill Capital amounts equal to the flows it was entitled to receive under the BOT Notes. Merrill Capital agreed to make 12 equal quarterly payments aggregating $ 10,961,581, together with interest on the unpaid balance at LIBOR plus 35 basis points. The interest rate was stepped up after the first year unless Merrill elected to terminate the swap and acquire the notes at a price equal to the unpaid principal balance remaining on the amortizing leg. For BFCE, the hedge swap effectively created a synthetic asset paying an attractive margin over LIBOR, and, for Merrill Capital, a hedge for its payment obligations under the outstanding swap with BOT.

According to Beder's calculations, the midmarket value of the BOT LIBOR Notes at the time of their sale to BFCE was $ 11.18 million. The bid-side spread of $ 220,000 implicit in the purchase price that BFCE paid ACM for the notes financed the gains shared by Merrill and the bank from the transaction. See diagram 3 supra p. 63. Ultimately, the cost of engineering this structured transaction, like the two before it, was borne almost entirely by Colgate.

In conformity with the requirements for approval of Kannex's loan, den Baas and his colleagues at ABN New York took steps to protect the bank from the risks of Kannex's participation in ACM and to ensure the bank an adequate return. ABN New York had the authority to implement a comprehensive financial management program for Kannex by virtue of ABN New York's financial services agreement. First, Kannex's exposure to the intrinsic interest rate risk of partnership assets would be "fully hedged". Den Baas never considered relying on the partnership's LIBOR Notes for this purpose. He made no attempt to evaluate their hedging effect within the partnership portfolio. It was clear to him that effect would not be adequate, and hedging instruments of greater precision and reliability were available. Accordingly, ABN New York arranged to neutralize the effect of the LIBOR Notes on Kannex's interest. The structure that it employed for this purpose consisted of back-to-back swap transactions with Kannex on the one hand and Merrill Capital on the other. ABN New York assumed the role of intermediary on the assumption that neither Merrill Capital nor any other third party would accept Kannex's credit risk.

By swap confirmations effective November 27, 1989, the issue date of the LIBOR Notes, ABN New York entered into a hedge swap agreement with Merrill Capital. Under the swap, ABN New York was required to make to Merrill Capital quarterly payments of 3-month LIBOR over 5 years equivalent to Kannex's 82.63 percent pro rata share of the payments owed to ACM under the LIBOR Notes. Merrill Capital was required to pay to ABN New York the sum of $ 28,433,655 in 20 equal quarterly installments together with interest on the unpaid balance at a rate of LIBOR minus 25 basis points. This amortizing principal amount was equal to 82.63 percent of $ 34,410,814, Kannex's pro rata share of the issue price of the LIBOR Notes. ABN New York entered into a matching hedge swap with Kannex under which Kannex's rights and obligations vis-a-vis ABN New York corresponded to those of ABN New York vis-a-vis Merrill Capital. When Kannex's indirect interest in the LIBOR Notes held by the partnership changed significantly as a result of the distribution of the BFCE Notes to Southampton on December 13, 1989, the partial purchase of Kannex's partnership interest on June 27, 1991, and the redemption of its remaining interest on November 27, 1991, both legs of the hedge swaps were adjusted proportionately. At these times, the portion of the swap that was to be terminated would be marked to market, and the counterparty that would otherwise have benefitted from the change in market interest rates would receive a compensatory termination payment. The back-to-back hedge swaps satisfied complementary needs. Kannex was able to stabilize its return on $ 28 million of its partnership investment. Likewise, Merrill Capital was able partly to offset the interest rate exposure that it incurred in connection with its hedge swaps with BOT and BFCE.

The back-to-back hedge swaps relating to the LIBOR Notes also served an additional function that can be understood only by reference to the terms of the structured transactions in which the LIBOR Notes were issued. According to the analysis of petitioner's expert, the transaction spreads implied in Merrill's pricing of the Citicorp Notes and LIBOR Notes for purposes of the contingent payment sale could be expected to result in the transfer of between $ 1.8 and $ 1.9 million of value from ACM to the foreign banks. The banks could have expected to retain only about $ 300,000 of this value, because their basis and hedge swaps with Merrill Capital were structured in such a way that the present value of the swap payments they were entitled to receive from Merrill Capital was less than the present value of the swap payments they were obligated to pay to Merrill Capital. Thus, the value of BFCE's right to quarterly payments of 3-month LIBOR Notes over 5 years on a notional principal amount of $ 27.91 million was $ 9.62 million, while the value of its obligation to pay $ 9,831,661 in equal quarterly installments over 5 years together with interest on the unpaid balance at LIBOR minus 25 basis points was $ 9.77 million. As a result of the discrepancy in the value of these two legs of the hedge swap, Merrill Capital could have expected to realize a net gain, and BFCE a net loss, of $ 150,000.

The hedge swap between Merrill Capital and ABN was structured in a manner similar to the hedge swap between BFCE and Merrill Capital. The ABN swap differed from the BFCE swap in only two respects. First, the payment obligations on both sides of the ABN swap were proportionately larger. In the BFCE swap, the notional principal amount of the fixed notional leg was set at an amount ($ 27.91 million) equal to 50/175, or 28.5 percent, of the combined total notional principal amount of the BOT and BFCE Notes ($ 97.76 million); in the ABN swap, it was set at an amount ($ 80,779,000), equal to Kannex's 82.63 percent share of the combined total notional principal amount of the BOT and BFCE Notes. Likewise, in the BFCE swap, the principal amount of the amortizing leg ($ 9,831,661) was equal to 50/175, or 28.5 percent of the combined total issue price of the BOT and BFCE Notes ($ 34,410,814); in the ABN swap, the principal amount of the corresponding leg was $ 28,433,655, an amount approximately equal to Kannex's 82.63 percent share of the combined total issue price of the BOT and BFCE Notes. If, as Beder concluded, the amortizing leg was worth more than then fixed notional leg in the BFCE swap, that asymmetry in value would necessarily have been magnified in the larger, but structurally identical, ABN swap. The second respect in which the swaps differed was that Merrill Capital occupied the position of the net creditor in the BFCE hedge swap but that of the net debtor in the ABN swap. The hedge swap between ABN and Kannex was in all respects identical to the hedge swap between Merrill Capital and ABN, except that ABN now assumed the position of net debtor.

The effect of the back-to-back hedge swaps would have been to transfer from Merrill Capital to ABN and from ABN to Kannex a portion of the value extracted from the partnership through the transaction spreads it was charged in the contingent payment sale. This transfer partly indemnified Kannex for its share of the partnership's economic loss.

By separate swap confirmations effective November 27, 1989, Merrill Capital agreed to pay ABN, and ABN agreed to pay Kannex, interest at the rate of LIBOR minus 25 basis points on a notional principal of $ 903,765, an amount that corresponded to Kannex's share of the 5/8 discount incurred by the partnership in the sale of the Citicorp Notes and origination of the LIBOR Notes. Following the distribution of the BFCE Notes to Southampton, the notional principal was reduced to $ 680,156. This revised amount represents the product of Kannex's then current percentage interest as reflected on a preliminary draft revaluation worksheet (87.06 percent) multiplied by the portion of the discount attributable to the BOT Notes retained by the partnership ($ 781,250). The documentation characterized these agreements as "swaps". This is a misnomer, however, because the payment obligations were unilateral. The parties' characterization reflects the fact that these "one-sided swaps" were negotiated in conjunction with the back-to-back hedge swaps and were intended to complement them. Like the hedge swaps, the one-sided swaps had the effect of compensating Kannex for a loss that it would otherwise have borne in connection with the contingent payment sale.

We have previously discussed how the partnership chose to account for the 5/8 discount incurred in the contingent payment sale for financial and tax accounting purposes. Rather than recognizing this transaction cost, the partnership included it in the carrying cost of the LIBOR Notes. Although this method of accounting was calculated to result eventually in the allocation of all of the transaction cost to Kannex's partners, as long as recognition of the cost was deferred, the capital accounts of Kannex's partners were overstated, and Kannex's share of partnership income was understated. According to the revaluation worksheets, the partners' capital account balances as of the end of FYE 11/30/89, were restated at fair market value as follows:

Kannex	MLCS	Southampton	Total
$ 170,617,686	$ 603,976	$ 35,145,281	$ 206,366,943
(82.68%)	(0.29%)	(17.03%)	(100%)

Had the $ 1,093,750 discount been recognized and allocated, say, entirely to Southampton at this time, Kannex's pro rata interest in partnership assets and share of partnership income would have been .4402742 percentage points higher and Southampton's .4402742 percentage points lower:

Kannex	MLCS	Southampton	Total
$ 170,617,686	$ 603,976	$ 34,051,531	$ 205,273,193
(83.12%)	(0.29%)	(16.59%)	(100%)

This .4402742 percentage point discrepancy corresponds to Kannex's allocable share of the discount:

$ 205,273,193 x .4402742% = $ 903,765 = $ 1,093,750 x 82.63%.

Under the one-sided swaps, ABN received from Merrill Capital and Kannex received from ABN a return on this .4402742 percentage point discrepancy in the capital accounts. When the transaction cost was subsequently recognized in part and charged to Southampton's capital account upon the distribution of the BFCE Notes, the understatement of Kannex's capital account was partly corrected and the notional principal amount on which the one-sided swap payment obligations were based was accordingly reduced. This compensatory arrangement appears to be critical to an understanding of why ABN agreed to an accounting policy that caused the partners' capital accounts to misrepresent the agreed allocation of costs to Kannex's detriment.

An unexecuted version of the one-sided swap between Merrill and ABN ran for a 5-year period coterminous with the hedge swap. In the executed agreements, the termination date was December 1, 1990. At the expiration of this term, the one-sided swap between ABN and Kannex was extended for a second year. There is no record of any similar extension of the corresponding one-sided swap between ABN and Merrill.

Through another series of swaps arranged by ABN New York, Kannex effectively eliminated its risk of loss and opportunity to gain from allocations of the Yield Component of the Colgate debt. The counterparty in these swaps was ABN Cayman Islands, but it was den Baas and others at ABN New York who executed the transactions on behalf of both counter parties. With respect to each issue of fixed-rate Colgate debt acquired by the partnership, Kannex entered into a fixed-for-floating interest rate swap on a notional principal amount corresponding to the dollar amount of Kannex's exposure to interest rate risk on the debt. Whenever Southampton elected to adjust the Yield Component sharing ratio or Kannex's partnership interest changed, the notional principal amounts of Kannex's swaps were adjusted to cover the amount of its exposure. The net effect for Kannex resembled an investment in a portfolio of LIBOR-based assets whose value would not vary in relation to the value of its LIBOR-based liability under the Revolving Credit Agreement.

The swaps with ABN Cayman Islands effectively offset Kannex's losses and gains from the intrinsic treasury risk of the Colgate debt held by the partnership. The swaps also offered Kannex the opportunity to profit from the spread risk of the Colgate debt. Kannex was required to pay interbank swap rates on its swaps. The fixed interbank swap rates were determined by adding a spread to the prevailing yields on comparable Treasury securities. For every piece of Colgate debt purchased, there was a referenced Treasury rate. To the extent that the yields on the partnership's Colgate debt exceeded these rates, Kannex kept the difference. ABN profited from the spreads that it earned in hedging its swap positions through coordinated trading of Treasury securities or futures, or through matching swaps with third parties.

In order for the hedging of Kannex's risks to be both effective and lucrative, the selection of Treasury securities used in the construction of hedge positions had to be consistent with the selection of Treasury securities used in the revaluation of the Colgate debt within the partnership. Aware of these hedging operations, Merrill accommodated them by consulting with ABN on the valuation of ACM's Colgate debt whenever changes in value were likely to affect Kannex's capital accounts. Thus, one Merrill internal memorandum described the procedures for an upcoming revaluation:

Since Kannex must actually trade Treasuries based upon the Base Treasury yields, Kannex would determine yields on Base Treasuries for each Note. These yields, along with previously determined spreads, are used by ML to set prices of each Note.

Under its Revolving Credit Agreement with Kannex, ABN reserved the right to sell participations, provided that it would remain solely responsible for performance of the obligations owed to Kannex under the Agreement. ¹⁸ Beginning in the fall of 1989, ABN offered a number of banks the opportunity to participate in its loans to Kannex as well as to other special purpose corporations that ABN Trust had organized for section 453 partnerships. The participations ABN proposed were short-term and renewable. ABN would guarantee an interest rate of LIBOR plus 35 basis points or 50 basis points. ABN would possess the exclusive right to enforce the loan.

ABN's relationship to Kannex was a source of some confusion. An internal memorandum of Banco di Roma outlining the syndication proposal described ABN as a "shareholder in Kannex together with another major U.S. Corporation". In the attempt to reassure prospective investors that their principal would be secure, den Baas went further than the terms of the formal Participation Agreement in defining ABN's position in the arrangements: "Since there is neither a scheduled interest payment on the notes held in the portfolio nor a principal repayment you would look even more to ABN to take you out at the maturity date of the loan". Within Banco di Roma, the participation was recommended for approval with the following explanatory gloss: "The repayment source of our advance is the committed facility provided by ABN through its Curacao or Grand Cayman Branch." The memorandum concludes: "Taking into consideration: The de facto guarantee of ABN, * * * we recommend your authorization to participate". An internal credit proposal of Banco Espirito Santo E Comercial De Lisboa (Banco Espirito Santo) reflects a similar understanding. Beside the heading "Guarantor", the following explanation appears: "Subsidiary of ABN will borrow against a firm takeout at maturity". Considering its reliance on the repeated participation of a small group of banks to sustain its involvement in numerous section 453 partnerships, it is not surprising that ABN would wish to imply, and that the investors would be prepared to infer, that they could look to ABN for repayment.

Generale Bank, Banco Espirito Santo, and Banco di Roma acquired participations in Kannex's loan in amounts between $ 25 million and $ 75 million. All participations were repaid by July 1991. The loan from ABN Cayman Islands was ultimately repaid out of the liquidating distribution that Kannex received at the end of November 1991. Owing to the preferred return that Kannex received from Southampton and appreciation of Colgate debt as a result of the decline in interest rates, there was a sizeable surplus remaining after repayment of the loan, as shown on Kannex's balance sheet for the period ended November 30, 1991. Kannex did not retain this surplus. Kannex also did not distribute this surplus to its nominal shareholders when Kannex was liquidated shortly thereafter.

Following the redemption, Kannex's swaps with ABN were terminated. The benefit that Kannex had enjoyed from a fall in interest rates for purposes of the valuation of its partnership interest was offset by the appreciation of the fixed-rate cash flows that it was obligated to pay relative to the floating rate cash flows it was entitled to receive under the Colgate debt swaps. Kannex owed ABN Cayman Islands $ 3,180,453. For reasons that the record does not disclose, the amount Kannex paid was higher by $ 1,655,000, and this excess was credited to den Baas' Financial Engineering Group. The back-to-back hedge swaps between Kannex and ABN New York and ABN New York and Merrill Capital were also terminated at the same time. Although the terms of the swaps were identical, for reasons not disclosed in the record, the termination payment that ABN New York made to Kannex was $ 500,000 less than the termination payment that was received from Merrill Capital. Kannex's balance sheet for the period ended January 27, 1992, shows remaining stockholder's equity of $ 17,278. Of this amount, $ 6,000 was attributable to the loans that Kannex had originally made to the foundations to finance their contributions and the rest may have been attributable to a capitalized loan from ABN. All the proceeds of Kannex's participation in ACM were, in one way or another, remitted to ABN. Liquidation procedures commenced in the following month.

ACM structured its sale of the Citicorp Notes to fall within the contingent payment sale provisions of section 15a.453-1(c), Temporary Income Tax Regs., 46 Fed. Reg. 10711 (Feb. 4, 1981). On November 3, 1989, ACM purchased $ 205 million of Citicorp Notes, and, 3 weeks later, it sold $ 175 million of the notes to BOT and BFCE for $ 140 million in cash and eight LIBOR Notes with a present value of $ 35 million. The LIBOR Notes did not provide for the payment of a stated principal amount. For FYE 11/30/89, ACM applied the ratable basis recovery rules of section 15a.453-1(c), Temporary Income Tax Regs., supra, recovering only $ 29,250,761 of its basis in the notes and recognizing $ 110,749,239 of capital gain. ACM allocated $ 91,516,689 of the gain to Kannex, an entity that was not subject to U.S. tax.

In FYE 12/31/91, after ACM redeemed Kannex's partnership interest, ACM sold the BOT LIBOR Notes to BFCE for $ 10,961,581, and, under section 15a.453-1(c), Temporary Income Tax Regs., supra, recognized a capital loss of $ 84,997,111. ACM allocated $ 84,537,479 of this loss to Colgate and Southampton.

We must decide whether ACM's planned sequence of investments and dispositions should be respected for tax purposes. We sometimes refer to ACM's planned sequence of investments and dispositions calculated to create the capital losses that were the objective of the CINS transaction as the "section 453 investment strategy".

Section 15a.453-1(c), Temporary Income Tax Regs., supra, provides installment sale treatment for "contingent payment sales". A "contingent payment sale" is "a sale or other disposition of property in which the aggregate selling price cannot be determined by the close of the taxable year in which such sale or other disposition occurs." Id. Where the sales agreement provides for no maximum aggregate selling price but fixes the period over which payments may be received, the temporary regulations generally require the seller to allocate an equal portion of its basis in the sale property to each of the taxable years in which payments may be received. Sec. 15a-453-1(c)(3), Temporary Income Tax Regs., 46 Fed. Reg. 10714 (Feb. 4, 1981). The seller computes its income for each year in respect of a contingent payment sale as the excess of the payments received in that year over the portion of the basis allocated to that year. Id.

The temporary regulations anticipate that application of the general rule for basis recovery will create distortions of income in some cases, and they provide certain remedies. The Commissioner may require an alternate method of basis recovery if the Commissioner finds that the general rule will "substantially and inappropriately accelerate recovery of basis." Sec. 15a.453-1(c)(7)(iii), Temporary Income Tax Regs., 46 Fed. Reg. 10716. Conversely, if application of the general rule "will substantially and inappropriately defer recovery of basis," the taxpayer may request an alternate method, but the Commissioner is not granted explicit authority by the temporary regulations to require the use of an alternate method in that situation. Sec. 15a.453-1(c) (7) (ii), Temporary Income Tax Regs., 46 Fed. Reg. 10716. The Commissioner may prescribe an alternate method if she determines that the taxpayer's method of accounting with respect to the sale does not "clearly reflect income". Sec. 446(b). In general, the Commissioner has broad discretion to determine whether an accounting method clearly reflects income. See Thor Power Tool Co. v. Commissioner, 439 U.S. 522, 532-533 (1979); Commissioner v. Hansen, 360 U.S. 446, 467 (1959); Ferrill v. Commissioner, 684 F.2d. 261, 264 (3d Cir. 1982), affg. T.C. Memo. 1979-501; Hudson v. Commissioner, T.C. Memo. 1996-106. A taxpayer's method of accounting does not clearly reflect income when it does not represent "economic reality". See Prabel v. Commissioner, 882 F.2d 820, 826-827 (3d Cir. 1989), affg. 91 T.C. 1101 (1988). In this case, the Commissioner has not exercised her discretion by raising the clear reflection of income issue in her pleadings or in her brief.

In his opening statement, petitioner's counsel aptly characterized the role of economic substance in this case: "Both parties agree that the question of substance is critical to the outcome. At the most fundamental level, this case is about very different views of commercial reality and very different views of the tax law's concept of substance."

ACM sold the $ 175 million aggregate principal amount of Citicorp Notes for $ 140 million in cash and eight LIBOR Notes, and, in connection therewith, reported a capital gain for FYE 11/30/89 and a corresponding capital loss for FYE 12/31/91. Respondent eliminated this gain and disallowed the loss. Respondent determined that the underlying transactions should not be given effect for Federal income tax purposes because it was tax-driven and devoid of economic substance. Respondent argues that the formation of the partnership and its activities during the relevant years were merely prearranged steps in a contrived, tax-motivated transaction that was carried out in accordance with Merrill's pursuit of approximately $ 100 million in taxable losses for Colgate. Respondent states that the liability management functions ascribed to ACM in documentation prepared by Merrill and Colgate were spurious. Respondent alleges that the structured transactions in which the LIBOR Notes were created and sold formed a "tax shelter market" that was controlled by Merrill, and that was operated in accordance with unwritten understandings. Respondent asserts that this "market" was supported by subsidizing the participating banks, as well as by circular payment flows and premature terminations that insulated the banks from a material risk with respect to the LIBOR Notes. Respondent alleges that structured transactions involving substantially the same patterns, timetables, and many of the same banks were involved in the issuance and sale of LIBOR Notes for each of the other section 453 partnerships.

Petitioner's account of the CINS transaction bears little resemblance to respondent's view. Petitioner argues that ACM was rationally designed to address genuine liability management needs. Petitioner alleges that all partnership transactions were negotiated at arm's length, priced at fair market value, conducted in accordance with standard commercial practices, and had practical effects wholly apart from their tax consequences. Petitioner asserts that the partnership and each of its partners had reasonable prospects for profit and risk of loss. Petitioner contends that, in arranging the structured transactions, Merrill acted in the customary role of a market maker, bringing counterparties together on terms that suited their respective needs. Petitioner argues that the swaps are irrelevant to the legal analysis because ACM was not a party to any of the swaps.

Following our review of the record, with due regard to our view and perception of the witnesses, we do not find any economic substance in the section 453 investment strategy. ¹⁹ We are convinced that tax avoidance was the reason for the partnership's purchase and sale of the Citicorp Notes. We do not suggest that a taxpayer refrain from using the tax laws to the taxpayer's advantage. In this case, however, the taxpayer desired to take advantage of a loss that was not economically inherent in the object of the sale, but which the taxpayer created artificially through the manipulation and abuse of the tax laws. A taxpayer is not entitled to recognize a phantom loss from a transaction that lacks economic substance.

In analyzing whether the CINS transaction had economic substance, we have been mindful that for some businesses there is little, if any, meaningful difference between an improvement in financial performance achieved by cutting operating expenses and one that results from reducing taxes. Both reductions improve the financial statement. The tax law, however, requires that the intended transactions have economic substance separate and distinct from economic benefit achieved solely by tax reduction. The doctrine of economic substance becomes applicable, and a judicial remedy is warranted, where a taxpayer seeks to claim tax benefits, unintended by Congress, by means of transactions that serve no economic purpose other than tax savings. Yosha v. Commissioner, 861 F.2d 494, 498-499 (7th Cir. 1988), affg. Glass v. Commissioner, 87 T.C. 1087 (1986); see also Estate of Thomas v. Commissioner, 84 T.C. 412, 432-433 (1985), and the cases cited therein.

Our conclusion is supported by well-settled judicial jurisprudence. In the seminal case of Gregory v. Helvering, 293 U.S. 465, 469 (1935), the Court recognized an individual's right to decrease her taxes in any way permitted by law. As held by the Court, however, this right is not absolute. The Court held that a reorganization that met the literal requirements of the Code would not be respected for Federal income tax purposes because "what was done, apart from the tax motive, was [not] the thing which the statute intended". The Court stressed that the transaction had "no business or corporate purpose", but was "a mere device which put on the form of a corporate reorganization as a disguise for concealing its real character". Id.

In the 60 years since the U.S. Supreme Court first expounded this doctrine of "business purpose", the doctrine's application has proved a perennial challenge to the courts to set boundaries between acceptable tax planning and abuse, while taking into account the importance of maintaining public confidence in the integrity of the tax system. In Knetsch v. United States, 364 U.S. 361 (1960), for example, the Court applied the Gregory v. Helvering case to disallow an interest deduction. In so doing, the Court stated that "there was nothing of substance to be realized * * * from this transaction beyond a tax deduction." Knetsch v. United States, supra at 366. Similarly, in Frank Lyon Co. v. United States, 435 U.S. 561 (1978), the Court stated that economic substance is a necessary requirement of any transaction. In Frank Lyon, the Court looked to "the objective economic realities of a transaction rather than to the particular form the parties employed", id. at 573, and stated that the Government should honor the allocation of rights and duties effectuated by the parties "where, as here, there is a genuine multiple-party transaction with economic substance which is compelled or encouraged by business or regulatory realities, is imbued with tax-independent considerations, and is not shaped solely by tax-avoidance features that have meaningless labels attached", id. at 583-584.

The Court of Appeals for the Second Circuit applied an economic substance analysis in Goldstein v. Commissioner, 364 F.2d 734 (2d Cir. 1966), affg. 44 T.C. 284 (1965). In that case, Mrs. Goldstein won the Irish Sweepstakes. In an attempt to shelter her winnings from tax, she borrowed from two banks and invested the loan proceeds in Treasury notes. The loans required her to pay interest at 4 percent, while some Treasury notes yielded one-half percent and others yielded 1-1/2 percent. Her financial advisers estimated that these transactions would produce a pretax loss of $ 18,500 but a substantial after-tax gain. This Court sustained the Commissioner's disallowance of the interest deductions. In affirming the decision of this Court, the Second Circuit stressed that this Court had found that Mrs. Goldstein's purpose in entering into the loan transactions "'was not to derive economic gain or to improve here [sic] beneficial interest; but was solely an attempt to obtain an interest deduction as an offset to her sweepstakes winnings.'" Id. at 738 (quoting Goldstein v. Commissioner, 44 T.C. at 295). The Second Circuit stated further that the loan arrangements did not "have purpose, substance, or utility apart from their anticipated tax consequences", and that the transactions had no "realistic expectation of economic profit". Id. at 740.

The Goldstein case marks an important step in the development of the economic substance doctrine. ²⁰ Unlike many purported tax shelters, the tax-motivated transactions in that case were not fictitious. Goldstein v. Commissioner, supra at 737-738. They were real and conducted at arm's length.²¹ Mrs. Goldstein's indebtedness was enforceable with full recourse and her investments were exposed to market risk. Yet, the strategy was not consistent with rational economic behavior in the absence of the expected tax benefits. Other courts have applied the teaching of Goldstein in varied settings. In Sheldon v. Commissioner, 94 T.C. 738 (1990), for example, this Court analyzed the financial transactions in issue there in a manner similar to that employed in Goldstein. The Court first determined that the transactions at issue were real, rather than fictitious. The Court then evaluated economic substance, stating that "the principle of * * * [Goldstein] would not, as petitioners suggest, permit deductions merely because a taxpayer had or experienced some de minimis gain." Id. at 767. The Court held that a transaction resulting in gain that was "infinitesimally nominal and vastly insignificant when considered in comparison with the claimed deductions" had no economic substance. ²²Id. at 768. The Court noted that "if the transactions had been fully offset, straddled, or hedged to obviate the possibility of any loss or gain, the form of the transaction could have been more readily attacked by respondent." Id. Accord Merryman v. Commissioner, 873 F.2d 879, 881 (5th Cir. 1989), affg. T.C. Memo. 1988-72; Levin v. Commissioner, 87 T.C. 698, 699, 728 (1986), affd. 832 F.2d 403 (7th Cir. 1987); Julien v. Commissioner, 82 T.C. 492, 509 (1984).

In Lerman v. Commissioner, 939 F.2d 44 (3d Cir. 1991), affg. Fox v. Commissioner, T.C. Memo. 1988-570, the Court of Appeals for the Third Circuit analyzed the economic substance doctrine. In Lerman, the taxpayers claimed to be commodities dealers and sought to deduct losses resulting from their option-straddle transactions. Id. at 45. The Third Circuit held that the transactions were "shams, devoid of economic substance, and thus any losses generated thereby cannot be the basis for deductions." Id. at 56. The court noted that "Per Gregory v. Helvering * * * it is settled federal tax law that for transactions to be recognized for tax purposes they must have economic substance." Id. at 52.

More recently, the Third Circuit reiterated that "the general rule on sham transactions in this circuit is well-established: 'If a transaction is devoid of economic substance * * * it simply is not recognized for federal taxation purposes, for better or for worse. This denial of recognition means that a sham transaction, devoid of economic substance, cannot be the basis for a deductible loss.'" United States v. Wexler, 31 F.3d 117, 122 (3d Cir. 1994) (quoting Lerman v. Commissioner, supra at 45). In Wexler, the taxpayer claimed deductions resulting from financial arrangements known as "repo to maturity" transactions. Id. at 118. The taxpayer argued that the economic substance doctrine did not apply to the deduction of interest payments pursuant to section 163 if the taxpayer's obligation to pay the interest is binding and enforceable. Id. at 122. The Third Circuit analyzed a series of related cases and noted that the key requirement that permeated each of those cases was that the financial transaction be "economically substantive". Id. at 127 (emphasis omitted). The Third Circuit stated that "transactions with no economic significance apart from tax benefits lack economic substance." Id. at 124.

The "principle laid down in the Gregory case is not limited to corporate reorganizations, but rather applies to the federal taxing statutes generally." Weller v. Commissioner, 270 F.2d 294, 297 (3d Cir. 1959), affg. Emmons v. Commissioner, 31 T.C. 26 (1958) and Weller v. Commissioner, 31 T.C. 33 (1958); see also Knetsch v. United States, 364 U.S. 361 (1960) (interest deduction); Higgins v. Smith, 308 U.S. 473 (1940) (loss deduction on sale to wholly owned corporation); Weyl-Zuckerman & Co. v. Commissioner, 232 F.2d 214 (9th Cir. 1956), affg. 23 T.C. 841 (1955) (mineral rights transferred to a wholly owned subsidiary); Braddock Land Co. v. Commissioner, 75 T.C. 324 (1980) (shareholders-employees' forgiveness of accrued salaries, bonuses, and interest owed by corporation in complete liquidation); David's Specialty Shops v. Johnson, 131 F. Supp. 458 (S.D.N.Y. 1955) (affiliated corporations). The tax statutes apply only "to transactions entered upon for commercial purposes and 'not to * * * transactions entered upon for no other motive but to escape taxation.'" Weller v. Commissioner, 270 F.2d supra at 297 (quoting Commissioner v. Transport Trading & Terminal Corp., 176 F.2d 570, 572 (2d Cir. 1949), revg. 9 T.C. 247 (1947)). Thus, transactions will only be recognized for tax purposes if there is some "tax-independent purpose" for the entire transaction. See Sheldon v. Commissioner, supra at 759. Only after we conclude that a transaction has economic substance will we consider the transaction's tax consequences under the Code. See Rice's Toyota World, Inc. v. Commissioner, 752 F.2d 89, 95 (4th Cir. 1985), revg. on a different issue 81 T.C. 184 (1983).

Whether a transaction has economic substance is a factual determination. United States v. Cumberland Pub. Serv. Co., 338 U.S. 451, 456 (1950). Key to this determination is that the transaction must be rationally related to a useful nontax purpose that is plausible in light of the taxpayer's conduct and useful in light of the taxpayer's economic situation and intentions. Both the utility of the stated purpose and the rationality of the means chosen to effectuate it must be evaluated in accordance with commercial practices in the relevant industry. Cherin v. Commissioner, 89 T.C. 986, 993-994 (1987). A rational relationship between purpose and means ordinarily will not be found unless there was a reasonable expectation that the nontax benefits would be at least commensurate with the transaction costs. See Yosha v. Commissioner, 861 F.2d 494, 498 (7th Cir. 1988), affg. Glass v. Commissioner, 87 T.C. 1087 (1986) (explaining the teaching of Goldstein); cf. Seykota v. Commissioner, T.C. Memo. 1991-234, amended T.C. Memo. 1991-541. "Deliberately to incur an expense greater than the expected gain -- to pay 4 percent for the chance to make 2 percent -- is the antithesis of profit-motivated behavior; such a transaction lacks economic substance." Yosha v. Commissioner, supra at 498.

Since the overall transaction must have economic substance for the Federal tax statutes to apply, we first consider whether the section 453 investment strategy had economic substance. Petitioner concedes that the section 453 investment strategy was tax motivated, but argues that tax-independent considerations informed and justified each step of the strategy. Petitioner explains ACM's investment in the Citicorp Notes as follows: "The ACM partners believed the Citicorp Notes offered a reasonable return on ACM's investment until such time as ACM might require cash for the purchase of Colgate debt". The Citicorp Notes were sold after 24 days to enable the partnership to invest in Colgate debt and LIBOR Notes. Petitioner argues that the investment in LIBOR Notes had two purposes. First, unlike an interest rate swap, which ACM could have used as an alternative hedging instrument, the LIBOR Notes provided the partners with an investment return. According to petitioner, "there was a realistic prospect that ACM would have made a profit on the LIBOR Notes." Petitioner contends that ACM disposed of the BFCE Notes and the BOT Notes when the hedging protection was no longer needed. Second, ACM invested in LIBOR Notes because it was "within the four corners of the partnership to operate as a hedge".

In light of each of these stated purposes, we examine the economic substance of the section 453 investment strategy.

The following colloquy at trial sheds some light on how Colgate's management arrived at the conclusion that the section 453 investment strategy promised a reasonable return and realistic prospect for profit. Pohlschroeder was the witness.

Q: In determining whether you should cast a vote or recommend that the partnership purchased (sic) the Citicorp Notes, did you take into account the transaction cost that would be incurred upon the sale of those notes?

A: It was known that there were transaction costs.

I really didn't know at that time what that exact amount was going to be, and basically, the initial part was just to get a reasonable return on the Citicorp Notes and make sure that the cash that we had received as a contribution was invested as quickly as possible.

Q: So, in determining whether you were going to earn a reasonable return, did you take into account the transaction costs that might be incurred upon the sale?

A: Not at that point. It was just basically an investment decision.

Q: So you did not compare those transaction costs that might have to be incurred upon the sale of the Citicorp note to the transaction cost on other instruments?

A: That is right, yes.

When the Partnership Committee formally authorized the purchase of the Citicorp Notes, Merrill informed Colgate that the section 453 investment strategy would result in transaction costs of between $ 2.3 and $ 3.1 million on a pretax present value basis, of which $ 1.3 to $ 2.0 million would be incurred in the contingent payment sale. The cash contributions that had to be "invested as quickly as possible" in Citicorp Notes yielding 8.78 percent in order for the partners to earn a reasonable return were already earning 8.75 percent in an ABN deposit account before the notes were acquired.

That Colgate's treasury department did not attach importance to the relative costs of the section 453 investment strategy is particularly significant because Colgate would bear both the transaction and remarketing costs. Pepe testified concerning the mutual understanding with respect to the five-eighths discount incurred in connection with the contingent payment sale:

The transaction was performed and put together, organized, on behalf of Colgate-Palmolive; therefore, the partners understood that the cost related to setting the transaction up should be borne by Colgate-Palmolive, whether that's through the partnership or through one of its partners.

The allocation of these costs to Colgate was accomplished by including them in the value at which the LIBOR Notes were carried on the partnership books and in the partners' capital accounts. When the BFCE Notes were distributed to Southampton, the other partners' allocable shares of these costs were charged to Southampton's capital account. When Southampton, Colgate and, to a nominal extent, MLCS acquired Kannex's partnership interest, they effectively purchased Kannex's share of the BOT Notes at a price that included Kannex's allocable share of these costs. Because the LIBOR Notes were acquired for Colgate's benefit, the partners provided that the remarketing costs would be borne almost entirely by Colgate as well. This was accomplished by selling the LIBOR Notes only after Colgate, Southampton and, to nominal extent, MLCS, had acquired all of Kannex's interest in them.

Kannex's interest in the BOT Notes could be acquired by Colgate alone or together with MLCS. If only Colgate purchased Kannex's interest, Colgate would bear all origination and remarketing costs allocable to that interest. If Colgate and MLCS purchased or redeemed Kannex's interest pro rata, each would bear a pro rata share of these costs. Acquisition of Kannex's interest by a combination of these methods would result in the sharing of these costs in some intermediate ratio. This was the approach that the parties actually adopted, but the evidence suggests that this decision may not yet have been made in November 1989. Nevertheless, it is unlikely that Colgate would have acquired any less than a pro rata share of Kannex's interest, since the opportunity cost of foregoing valuable tax benefits would have been too great. ²³ The nontax benefits of holding and selling Kannex's share of the notes would be shared in the same ratio as the costs associated with that share, but, in all events, these benefits would necessarily be less than the costs.

If the section 453 investment strategy was economically justifiable in part on the basis of expected pretax returns, and the partners understood that Colgate, as the beneficiary of the strategy, would bear virtually all transaction costs, then the strategy must have provided Colgate a realistic possibility of recovering these costs for the section 453 investment strategy to be deemed profitable. We examined the proposition that Colgate could reasonably have expected to recover the transaction costs of the strategy through cash flows from the LIBOR Notes, and we now set forth our analysis with respect thereto.

Colgate's return was a function of two variables. First, the credit quality of the issuers of the LIBOR Notes could have affected Colgate's returns. The possibility of benefitting from a credit improvement, however, was negligible. BFCE's rating was AAA and could not have improved. BOT was rated AA. If an improvement in BOT's credit could have increased the sale price of the notes, then one would expect that the difference between the banks' respective ratings would have affected the pricing of the notes at issuance. It had no effect.

The second and more important factor was interest rates. Based on its assumption that future interest rates would equal the levels predicted by the yield curve used to price the LIBOR Notes at their issuance, Merrill estimated that the issue price for the notes exceeded by approximately $ 1.3 million the bid price at which the notes could be sold to a third party. Hence, the partnership, and ultimately Colgate, would almost certainly lose money.

One must wonder what were the nontax benefits that the partnership hoped to achieve through its acquisition of the notes at that price level. Interest rates would have had to rise by at least 400-500 basis points, to a level of 13 percent or more, soon after the partnership acquired the LIBOR Notes and be expected to remain at that level throughout the 5-year life of the notes in order for Colgate to earn a sufficient return from the notes to cover the transaction costs of the section 453 investment strategy. Had the partners' economic arrangement contemplated a pro rata allocation of these costs, Colgate still could not have earned a profit on a net present value basis unless interest rates exceeded their expected levels, but a much smaller increase would have been sufficient to break even.

We reviewed historical data to assess the likelihood that 3-month LIBOR would have risen by the requisite amount for Colgate to break even. The record includes published records of market interest rates extending back to January 1984. There are 71 observations of 3-month LIBOR as of the first day of each month between January 1984 and November 1989. Not one of the 71 monthly quotations is 300 basis points or more above the quotations for the 1 to 6 previous months. Only three of the quotations represent a level 200 basis points or more above any quotations during the previous 6 months. There is no month for which 3-month LIBOR was above 12.13 percent. It reached or exceeded 11 percent in 6 months, all in mid-1984. In 30 months, it fell within the range of 8 to 9.99 percent, and it fluctuated between 10.31 and 8.56 percent during the first 11 months of 1989. The longest that 3-month LIBOR remained at or above 10 percent was 9 consecutive months in 1984. Thereafter, the longest period was 2 consecutive months in early 1989. In the late summer and early autumn of 1989, Colgate's treasury department confidently expected that interest rates would follow a downward trend for the foreseeable future.

Colgate could not have achieved a non-negative net present value under any reasonable forecast of future interest rates. A major war, an oil crisis, a resurgence of double digit inflation or other economic catastrophe might have been capable of inducing a sudden rise in interest rates by 400-500 basis points and might perhaps have sustained such levels for a period of months or years. But nothing in the record suggests that anyone involved in planning the section 453 investment strategy anticipated, or had any reason to anticipate, the extraordinary economic conditions which would have been necessary in order to make Colgate's investment in the LIBOR Notes profitable.

Appreciation of the LIBOR Notes was not the only source of potential profit from the section 453 investment strategy. Petitioner and its experts contend that some or all of the transaction costs of the strategy could have been recovered out of returns from the Citicorp Notes. They identify three sources of potential profit: (1) Gain on the sale of the Citicorp Notes attributable to an improvement in Citicorp's credit, (2) gain attributable to an increase in the commercial paper rate to which the coupon on the notes was linked, and (3) accumulation of interest income over the period the partnership held the notes.

With respect to petitioner's first claim that an improvement in Citicorp's credit could produce a profit, petitioner states that "ACM's exposure to Citicorp's credit was real, not theoretical": There was a significant risk that Citicorp's credit could deteriorate, but a significant possibility of improvement as well. The Citicorp Notes were rated AA by Standard & Poors and A1 by Moody's, which implies that there was some room for improvement in the issuer's credit quality. Data for the 5-year period ending in December 1991 confirm many instances in which the credit spread on publicly traded Citicorp debt declined by large amounts over short periods of time. To conclude from this that there was a reasonable possibility that ACM could have sold the Citicorp Notes at a price above par would not be warranted, considering the terms of the structured transaction in which they were sold.

Under the terms of the basis swap between Merrill and the purchasing banks, Merrill had a right to call the Citicorp Notes at a strike price equal to their par value. This option was exercisable on any payment date and the step-up in the amount of Merrill's payment obligations under the basis swap after 3 months effectively guaranteed that Merrill would exercise the option unless Citicorp's credit quality had substantially declined. Internal documents of BOT and BFCE indicate that both banks expected Merrill to purchase the notes from them within 1 to 3 months under this arrangement. Even if Citicorp's credit quality had improved over the period that ACM held the notes, it is unlikely that the banks would have been willing to pay any more than par for them, since all the increase in the value of the notes would only be appropriated by Merrill. It appears from the BOT and BFCE documents that the terms for Merrill's call option had already been worked out, along with most of the other details of the transaction structure, within 1 week of ACM's acquisition of the Citicorp Notes. Thus, Merrill designed the Citicorp Note transactions in a manner that effectively left no opportunity for ACM, or Colgate, to benefit from an improvement in Citicorp's credit. We reject petitioner's first contention.

Turning to petitioner's second claim that the Citicorp Notes, as floating rate notes (FRN's), could increase in value by way of an increase in the related commercial paper rate, we note that the value of a FRN is generally invariant to changes in market interest rates. Indeed, this is the source of its appeal to investors. Because the coupon payable on the Citicorp Notes was reset each month at the current commercial paper rate, the value of the notes should not have deviated significantly from par. This appears to have been the understanding of those who planned and approved the Citicorp Note investment. A memorandum of ACM's accountants recites that "as per explanation of Mr. Hans Pohlschroeder * * * the Citicorp Notes were floating rate notes * * * and can thus by definition not fluctuate in value because of changes in interest rates as the interest on the notes follows these changes". Under the partnership's Accounting Policies, the Citicorp Notes were treated as a cash equivalent for this reason.

According to petitioner, this understanding is subject to significant qualifications. Petitioner relies on the observations of one of its experts, Joseph Grundfest (Grundfest) of Stanford University. Grundfest notes that the decision to purchase a FRN locks in a return tied to a specified floating rate index. There are several indices, LIBOR, treasury bill, Federal funds rates, etc., and their relationship is not stable over time. Payments on FRN's can vary substantially depending on the choice of the underlying index. Grundfest goes on to cite actual examples of significant discrepancies between certain floating rate indices that occurred during and around the years at issue.

We cannot quarrel with these observations. How much significance we should attach to the potential for such market discrepancies as a basis for a reasonable expectation of profit is another matter. FRN's are commonly used by investors as a substitute for short term money market instruments such as certificates of deposit (CD's). Historical interest rate data introduced in evidence confirm that changes in the 1-month commercial paper rate and CD rate are not perfectly correlated. Over the 71 months from January 1984 to November 1989, the two rates fluctuated, but generally remained within 15 basis points of one another. In only 4 months did the difference between them equal or exceed 40 basis points. During the period that ACM planned to hold the Citicorp Notes, the coupon would be reset only once. The historical data provide no basis for concluding that there was any significant likelihood that an appreciable change in the historical relationship between the 1-month commercial paper rate and other money market indices would have arisen on this single occasion. Accordingly, we are not persuaded by petitioner's claim that it expected the Citicorp Notes to increase in value by way of an increase in the related commercial paper.

We now consider petitioner's third and final claim that it had a high probability of recovering its transaction costs through accumulation of interest income on the Citicorp Notes over the period that petitioner held the notes. Petitioner and its experts take the position that a substantial portion of the transaction costs of the section 453 investment strategy were likely to be recovered dollar-for-dollar through the accumulation of interest income from the Citicorp Notes: The longer ACM held the notes, the greater the amount of interest it received from Citicorp, and, all other things being equal, the greater the likelihood of earning a profit. ACM could reasonably have expected to receive, and did receive, about $ 1.2 million in interest on the Citicorp Notes over the 24 days that it held them. Colgate's share of this income (through Southampton) was about 17 percent (approximately $ 204,840), significantly less than the transaction costs incurred in the CINS transaction.

The initial coupon on the Citicorp Notes offered a three basis point advantage over the yield that the partners' contributions were currently earning in an ABN deposit account. Had the Citicorp Notes retained that yield advantage for the duration of the 24-day holding period, they would have provided ACM with $ 3,500 more income, of which Colgate's share would be about $ 600. Another alternative investment for the partnership cash was a portfolio of short-term money market instruments like those which it acquired with the $ 140 million cash proceeds of the contingent payment sale and which matured 1 week later on the settlement date for the purchase of the Colgate debt. These commercial paper issues provided yields ranging from 8.15 to 8.20 percent, 45-50 basis points less than the 8.65-percent coupon payable on the remaining $ 30 million Citicorp Notes for the second reset period. This yield differential was likely to have been attributable in part to a declining trend in short-term interest rates throughout the fall of 1989, which the coupon rate on the Citicorp Notes reflected only after a lag. Assuming, however, that at the time ACM acquired the Citicorp Notes they would have provided the same 50 basis point advantage over alternative commercial paper investments over the 24-day holding period, this advantage would have resulted in $ 58,000 more income for the partnership and less than $ 10,000 more income for Colgate. In short, any yield advantage that the Citicorp Notes may have offered over less costly alternatives would not significantly have improved Colgate's prospects for recovering the $ 2-3 million present value of transaction costs that it expected to incur in connection with the section 453 investment strategy. Accordingly, we reject petitioner's third contention.

We conclude that the partnership did not undertake the section 453 investment strategy with a reasonable expectation that it would be profitable, on a pretax basis, for Colgate. We also conclude that the strategy was not pursued with a realistic expectation of realizing an economic profit for ABN. Petitioner's expert, Beder, concedes that the expected rate of return in an environment with a 50-percent probability on a rising rate and a 50-percent probability on a falling rate would only equal 2.3 percent. Moreover, as the excerpt from Pepe's testimony quoted above confirmed, the agreed allocation of transaction costs reflected the fact that ABN did not expect to derive any significant profit from the strategy. To the extent that interest on the Citicorp Notes may have exceeded the interest that could be earned on money market instruments, Kannex would have shared in this premium pro rata, but given the short holding period, the accumulation would not have been significant. Through the back-to-back hedge swaps that ABN arranged with Kannex and Merrill with respect to the LIBOR Notes, ABN relinquished the opportunity to gain from Kannex's interest in the LIBOR Notes.

Petitioner's experts correctly point out that it has become common in the capital markets to enter into one transaction only for the purpose of using it as the basis for a profitable swap opportunity. The fact that the swap effectively forecloses the possibility of gain from the underlying transaction does not mean that the transaction serves no profit objective. On the contrary, the underlying transaction is an indispensable component of the arbitrage scheme. Arbitrage, however, is not a plausible explanation for ABN's behavior in this instance. Based upon testimony of Merrill witnesses, petitioner emphatically maintains that ABN did not approach Merrill with the proposal for the LIBOR Note hedge swap until shortly before the contingent payment sale. This was after the decision had been made, with Kannex's approval, to authorize the sale. If the partnership had authorized the section 453 investment strategy with the expectation that it would provide ABN with an arbitrage opportunity, presumably there would be evidence that ABN had planned, and attempted to arrange, its swap with Merrill beforehand.

More importantly, the terms of the back-to-back hedge swaps with respect to the LIBOR Notes are inconsistent with the arbitrage interpretation. In our Findings of Fact, we discussed at length the structural correspondence between these swaps and the hedge swaps between Merrill Capital, BFCE, and BOT, and we discussed the functional implications of that correspondence. Thus, although it appears that ABN could reasonably have expected to derive gain from these swaps, this gain represented value transferred, through the network of structured transactions growing out of the contingent payment sale, from the partnership to the banks, to Merrill Capital, and back to ABN and Kannex. The section 453 investment strategy was not designed to provide ABN with an opportunity for profitable LIBOR Note swaps. On the contrary, the swaps were calculated to compensate ABN in part for Kannex's share of the economic loss sustained by the partnership through the section 453 investment strategy.

Considering the high costs of the financial engineering it required and ABN's unwillingness to have Kannex share any of these costs or be exposed to any of the entrepreneurial risks it entailed, the section 453 investment strategy would not have been consistent with rational profit-motivated behavior in the absence of the expected tax benefits.

The theory of the LIBOR Note hedge was carefully developed in contemporaneous documents and argued in these proceedings. It forms the linchpin of petitioner's economic substance argument. It is, however, false. It is false even if we assume arguendo that there was as high a negative correlation between the interest rate sensitivity of the LIBOR Notes and that of the Colgate debt as petitioner asserts, a proposition that respondent and her experts vigorously contest. To recognize why the theory is false it is necessary to grasp this central insight: Neither ABN nor Colgate needed a hedge inside the partnership for the Colgate debt because both were effectively fully hedged outside the partnership - ABN through swaps and Colgate by virtue of being the issuer of the debt. Employing an additional hedging instrument within the partnership was not only redundant, but also flatly inconsistent with the manner in which both principals were otherwise managing their interests in the partnership.

In his opening argument at trial, petitioner's counsel began his analysis of the case as follows:

ACM, the partnership, is before the Court, and the tax treatment of its transactions is at the heart of the dispute. In many respects, however, the real party in interest is the Colgate-Palmolive Company and the impact of ACM's transactions from Colgate's vantage point is critical to understanding the substance of this case.

In justifying the partnership, petitioner argues that it was designed to perform functions integral and useful to Colgate's liability management strategy. On the other hand, petitioner argues that to evaluate whether the LIBOR Notes served a useful hedging function it is the effects "within the four corners of the partnership" that are relevant. The implication is that we should treat the position that Colgate held within the partnership through the instrumentality of Southampton as functionally unrelated to Colgate's liability management strategy: The utility of the LIBOR Notes is to be judged without regard to the primary purposes for which the partnership was created. It should be borne in mind that we are inquiring not whether a partnership should be treated as an entity or an aggregate for tax purposes or whether Southampton and ACM are entitled to be respected as separate legal entities, but whether there is any coherence to petitioner's economic explanation for the existence of the partnership and Southampton's role in it. The shift in focus that petitioner proposes is simply a sophistical sleight of hand. With a little analysis, the absurdity of the implications of this proposition can be appreciated. In any event, we emphasize that while we make this analysis we nevertheless decline to unbundle the transaction in order to isolate one element that might have economic substance. Rather, we view the transaction as bundled and judge it in its entirety.

Colgate's position within the partnership was functionally analogous to an interest rate swap. This is the way contemporaneous documents of its treasury department analyzed Colgate's overall interest rate exposure, the way Colgate's accountants recommended that the investment in ACM be treated for financial reporting purposes, and the way Pohlschroeder described Colgate's intentions in designing ACM. The swap analogy is apt and useful for purposes of our economic substance analysis. Suppose that Colgate issues fixed-rate debt and, in order to reduce its exposure to interest rate movements, enters into a "plain vanilla" interest rate swap in which it receives fixed and pays floating interest. As a result, Colgate is hedged. Now suppose that Colgate modifies the swap agreement such that whenever interest rates fall or rise the fixed rate that it is entitled to receive on the asset leg of the swap will be lowered or raised by some specified proportion of the notional principal amount. The reason offered for this modification is that Colgate wants to limit its exposure to interest rates "within the four corners of the swap", by ensuring that both its rights and obligations under the swap will move in tandem. There is a major fallacy in this proposition. The only effect of modifying the swap in this way is to defeat its very purpose as a hedge against Colgate's exposure to the underlying debt issuance.

From Colgate's perspective, the partnership's investment in the LIBOR Notes had the same effect as the modification of the swap in this hypothetical. To the extent that changes in their value were inversely correlated with changes in the value of the Colgate debt, the LIBOR Notes counteracted the hedging effect that Colgate was trying to achieve through its position in the partnership and thereby increased Colgate's exposure to interest rate risk.

Pohlschroeder's October 3, 1989, memorandum contains quantitative projections that show this clearly. Pohlschroeder analyzes the effects of a 200 basis point parallel shift in the Treasury yield curve on Colgate's financial position. The table presents his results.

Colgate's Financial Exposure to Partnership Portfolio
(Based on Pohlschroeder's Projections)
( $ millions)
	Base Level	200 Basis Pt.	Change
		Decline
Long Bonds	(42.00)	(51.45)	(9.45)
Met Note	(98 00	(104.66)	(6.66)
Total liabilities	(140.00)	(156.11)	(16.11)
Partnership interest (15%)	21.00	23.42	2.42
Net liabilities	(119.00)	(132.69)	(13.69)
LIBOR Notes	60.00	48.99	(11.01)
Partnership interest (15%)	9.00	7.35	(1.65)
Net position	(110.00)	(125.34)	(15.34)
		200 Basis Pt.	Change
		Rise
Long Bonds		(34.88)	7.12
Met Note		(92.18)	5.82
Total liabilities		(127.06)	12.94
Partnership interest (15%)		19.06	(1.94)
Net liabilities		(108.00)	11.00
LIBOR Notes		69.90	9.90
Partnership interest (15%)		10.48	1.48
Net position		(97.52)	12.48

The parentheses in the table reflect that the Long Bonds and Met Note are liabilities for Colgate. Changes in the value of these liabilities are offset in part by changes in the value of Colgate's 15-percent interest in the partnership portfolio comprising these bonds and LIBOR Notes. When interest rates fall, Colgate's bonds appreciate, resulting in a $ 16.11 million decrease in the market value of Colgate's net worth. This loss represents the opportunity cost to Colgate of being locked into a fixed rate liability that now exceeds the prevailing cost of capital in the market. By virtue of its proposed 15-percent ownership share in the partnership portfolio, Colgate realizes a gain that offsets this loss in part: The net effect on Colgate is a $ 13.69 million loss. But the partnership also holds LIBOR Notes, which decrease in value when interest rates fall. The effect of holding a 15-percent share of the LIBOR Notes through the partnership is to magnify the net effect of a fall in interest rates: If the partnership did not hold LIBOR Notes the market value of Colgate's net worth would decline by $ 13.69 million; the LIBOR Notes increase this loss to $ 15.34 million.

Now consider the effects of an increase in interest rates on Colgate's net worth. The Colgate bonds decrease in value by $ 12.94 million. The benefit to Colgate of having lower financing costs than the prevailing market rates is partially offset by Colgate's 15-percent share of the capital loss experienced by the partnership. But the net effect for Colgate is a gain. Colgate also benefits from the appreciation of the LIBOR Notes: If the partnership did not hold LIBOR Notes, the market value of Colgate's net worth would increase by $ 11 million; the LIBOR Notes increase Colgate's gain to $ 12.48 million. Thus, once again, the effect of the LIBOR Notes is to magnify Colgate's exposure to interest rates. From the perspective of Colgate's overall financial position, the LIBOR Notes do not function as a hedge at all.

There is a curious inconsistency in Pohlschroeder's memorandum between his discussion of how the partnership will serve Colgate's liability management objectives and his discussion of the function that the LIBOR Notes will perform. In the section entitled "Risk Management Within the Partnership", he calls attention to the importance of the partners' exposure to the interest rate volatility of the Colgate debt in the partnership portfolio, and states that the partnership will acquire LIBOR Notes "to minimize the exposure to ABN and Colgate". Based on the process of negotiation, a hedge ratio is going to be negotiated with ABN which may not be a perfect hedge." This might be taken to imply that a perfect hedge would be desirable, if possible.

But Colgate would not really have wanted a perfect hedge. Indeed, in Pohlschroeder's view, for the foreseeable future, Colgate did not want to reduce its interest rate exposure within the partnership at all. On the contrary, consistent with his forecast of falling interest rates over the next 3 to 9 months, in a different section of the memorandum Pohlschroeder states that Colgate will use the flexibility of the partnership structure to increase its exposure within the partnership substantially above its pro rata share:

One of the most important aspects of the partnership structure relates to the risk management of the interest rate risk as negotiated between Colgate and ABN. Colgate will attempt to negotiate a close to 50/50 sharing of the treasury risk.

To demonstrate how this arrangement would benefit the company, he examines how a 200 basis point decline in interest rates would affect the principals under different sharing ratios. He concludes:

The more treasury risk is assumed by Colgate, i.e. 85/15 to 51/49, the better off Colgate is. The value of Colgate's share in the partnership is roughly $ 30 MM using the 85/15 example and increases to $ 36 MM if we were to assume 49% of the treasury risk and interest rates dropped by 200 b.p.

At some point in the future, Colgate might wish to reduce its exposure: "As an example, if we started with a 50/50 sharing ratio and see interest rates bottom out, in the future we could switch at the bottom of the interest rate cycle to a 100%/0% ratio."

The difficulty of reconciling the LIBOR Note hedge with Colgate's liability management strategy becomes more apparent in the light of events that unfolded over the next 11 months. In his memorandum, Pohlschroeder assumed that the partnership would "establish a hedged capital structure with approximately $ 140 MM of Colgate debt and $ 60 MM of LIBOR Note hedge." The ratio of $ 140 million Colgate debt to $ 60 million LIBOR Notes originated in Merrill's first effort to integrate the CINS transaction into a liability management framework, the Partnership Transaction Summary dated July 28, 1989. Thereafter, all of Merrill's revisions of this document, its cash-flow projections and flip chart presentations to Colgate management through late October assumed that $ 200 million private placement notes would be sold for $ 140 million cash and $ 60 million LIBOR Notes. Around the time of the formation of ACM, however, it was decided that the partners could afford to do without a substantial amount of this internal hedge: $ 175 million private placement notes would be sold for $ 140 million cash and $ 35 million LIBOR Notes. No explanation was provided at trial, and none is to be found in the documentary evidence, of the reasons for the decision. But the effect was a reduction by 42 percent in the planned level of interest rate hedging protection and the retention of assets whose value would not vary with interest rates in a manner that undercut the effectiveness of Colgate's liability management strategy. ²⁴

At the time the LIBOR Notes were acquired, Colgate had no intention of using them to reduce Southampton's interest rate exposure. Its management of Southampton belies any such claim. Over the first 6 weeks after formation of ACM, Colgate increased Southampton's share of the Yield Component to 39.7 percent, more than double its original pro rata share and more than triple its pro rata share after the distribution of the BFCE Notes. In conformity with the original plan for a falling interest rate environment outlined by Pohlschroeder in his memorandum, it held Southampton's exposure at this level until September 1990.

One might suppose that if the LIBOR Notes were acquired for their utility to Colgate as a hedge within the partnership it was because, even if Colgate might have desired leveraged exposure to treasury risk at the outset, at some point in the future when a rise in interest rates appeared imminent it would wish to minimize its exposure. Yet, before the LIBOR Notes were acquired, Colgate and Merrill had planned for the immediate disposal of 30 percent of them. The timing of the acquisition and disposition of the LIBOR Notes bore no relationship to Colgate's interest rate expectations.

If Colgate had intended to use the LIBOR Notes for protection against rising interest rates, they would not have been a cost-effective instrument for this purpose. Colgate appears to have had no reason to believe otherwise. In an undated document entitled "Risk Allocation Analysis" that seems to have been prepared for Colgate in late October or November, before the LIBOR Notes were acquired, Merrill estimated that a 200 basis point increase in interest rates would cause $ 35 million market value of LIBOR Notes to appreciate to $ 40.31 million. This appreciation of just over 15 percent would offset less than half of the devaluation of the Colgate bonds. Southampton's original 17.07 percent pro rata share of the gain on the LIBOR Notes would offset approximately $ 906,000 ($ 5.31 million x .1707) of its share of the loss. After distribution of the BFCE Notes, a 200 basis point increase in interest rates would have generated a $ 3.79 million offsetting gain on the remaining LIBOR Notes, of which Southampton would have been entitled to only $ 478,000, in proportion to its 12.6 percent post-distribution partnership interest. When Colgate would have reviewed the results of Merrill's analysis and planned with Merrill the distribution and sale of the BFCE Notes, it would have understood that the discounted present value of the transaction costs that it would bear in connection with the acquisition and sale of the LIBOR Notes would be in the vicinity of $ 2-3 million. The potential hedging benefits would properly be discounted for uncertainty. Let us assume, for example, that there was a weighted average probability of 50 percent that interest rates would rise by an average of 200 basis points during the foreseeable future. A 50-percent probability is still clearly an overstatement, given the declining interest rate environment predicted in the implied forward rates that Beder estimated, in the market swap rates that Merrill used to price the LIBOR Notes, and in the Colgate treasury department's own forecasts. Nevertheless, even under this extreme assumption, the maximum hedging benefit that could be expected during the foreseeable future would have been less than 1/10 the expected cost (.5 x $ 478,000 / $ 2.5 million). ²⁵

In December 1991, after the redemption of Kannex's interest, the partnership concluded that it no longer needed the BOT Notes. The explanation recited in the minutes of the twelfth partnership meeting is that since Colgate and Southampton now owned over 99 percent of the partnership, "the principal Partners' net economic exposure to the risk of interest rate fluctuations in the value of the Colgate debt was effectively minimal," and with their usefulness exhausted, so volatile an investment could not be justified. Heidtke's explanation at trial was as follows: "At that point in time, the need for the - originally for the LIBOR notes as a hedge of the debt had basically gone away because now we owned all of the debt basically, so it was no longer outstanding, it was effectively retired".

It was reasonable for Colgate to be indifferent about exposure to the volatility of its own debt in the partnership portfolio at this time. Yet, it had always been the case that to the extent Colgate held its own debt through the partnership that debt was economically retired and there was no exposure to hedge. This was among the principal advantages of the liability management partnership identified in the Executive Summary dated October 9, 1989:

The Partnership also allows Colgate to effectively retire its debt, while leaving the debt outstanding for accounting purposes, * * * As Colgate bears a relatively greater share of the Treasury risk * * * with respect to its debt, it has economically retired an increasing percentage of such debt * * *

Petitioner's expert, Kenneth Singleton of Stanford University, makes the same point:

Exposure to Colgate debt through Southampton would have fully hedged an equal amount of liabilities on Colgate's balance sheet * * * From this particular perspective, Colgate's investment in Southampton had an impact similar to the consolidation of the bonds owned by ACM onto Colgate's balance sheet * * *

If the LIBOR Notes were not necessary as a hedge for Colgate in December 1991, they had never been necessary.

It is true that the hedging effect of Colgate's investment in its own debt did not appear on Colgate's consolidated financial statements until ACM was actually consolidated for financial reporting purposes. All the same, the Colgate bonds were stated on the balance sheet at their historic cost and were not revalued to reflect changes in the market cost of capital. Yet, Colgate's investment in the partnership would be marked to market, in accordance with the convention for reporting swaps or other hedging activities. This asymmetrical accounting treatment could have been expected to add an insignificant volatility to the consolidated financial performance of the Colgate group. The question arises whether this undesirable accounting byproduct of Colgate's liability management strategy would have provided reasonable grounds for hedging within the partnership.

We do not think so. If the standard financial accounting treatment of hedging activities was a cause for concern warranting countervailing positions designed to eliminate the effects from the financial statements of the business, businesses would routinely offset their own hedges and receive little or no net economic benefit from them. In July 1989, Colgate had entered into $ 300 million notional principal amount of interest rate swaps for liability management reasons similar to those that actuated its investment in ACM. That these swaps were also marked to market for financial reporting purposes evidently did not trouble Colgate, for it took no action to counteract their economic effects. Thus, a desire to stabilize the value of the ACM investment on its financial statements could not have provided a rational basis for the decision to hedge inside the partnership.

ABN never had any intention of using the LIBOR Notes as a hedge for Kannex's interest in the partnership. Instead, it hedged Kannex's exposure to the Colgate debt by means of swaps outside the partnership and, by separate swaps, eliminated the superfluous and deleterious effects of the volatile LIBOR Notes. However, that is not what the minutes of the third partnership meeting on December 12, 1989, suggest. According to the minutes, Taylor recommended that the partnership dispose of 20 percent of the LIBOR Notes because the planned exchange of some of the Long Bonds for new Colgate debt of shorter maturity would reduce the partners' interest rate exposure. "He further noted that such reduction would not adversely affect Kannex because of the Adjustment of sharing of Yield Component effected by the notice dated December 12, 1989, from Southampton-Hamilton Company", which lowered Kannex's share of the Yield Component. One would not gather from Taylor's explanation that, 2 weeks before the meeting, Kannex, ABN, and Merrill entered into the back-to-back hedge swaps that rendered the LIBOR Notes utterly ineffectual as a risk management instrument for Kannex, or that Kannex and ABN were also hedging Kannex's exposure to the Colgate debt so that Kannex would not be affected by Southampton's adjustments of the Yield Component.

The explanation for the decision to dispose of the BOT Notes provided in the minutes of the twelfth partnership meeting in December 1991 is likewise misleading. Pepe is reported to have said:

As Colgate and a subsidiary Southampton, owned 99.4% of the Partnership, the principal Partners' net economic exposure to the risk of interest rate fluctuations in the value of the Colgate debt was effectively minimal, and the Partnership need not maintain its position in the instruments purchased to hedge against such exposure.

Although there is no explicit assertion here that the partnership believed the LIBOR Notes to be necessary so long as Kannex was one of the principal partners, that is the implication.

Petitioner contends that Merrill and the Partnership Committee could honestly and reasonably have represented that the LIBOR Notes actually served as a hedge for Kannex's benefit. Petitioner denies that Merrill and Colgate knew of Kannex's swaps with ABN.

Even though Merrill entered into swaps with ABN relating to Kannex's share of the LIBOR notes owned by ACM, Merrill was not specifically informed of the Kannex/ABN swaps relating to the LIBOR notes. [Emphasis added.]

Although Merrill may have suspected that Kannex and ABN had entered into similar swaps, there is no evidence that Merrill knew, in fact, that such a transaction had taken place. Consequently, there is nothing about Taylor's representation at the third Partnership meeting that is inaccurate or misleading.

It is true that there is no evidence in the record that ABN specifically apprised Merrill of its swaps with Kannex. But this misses the point. Petitioner seems to think that Merrill's understanding as to the utility of the LIBOR Notes would be significantly affected by specific information or lack thereof that ABN was engaging in a swap with Kannex that mirrored the swap between ABN and Merrill. There was no doubt why ABN entered into its hedge swap with Merrill. The purpose and effect of that swap were to neutralize the impact of the LIBOR Notes on ABN's investment in the partnership. Nor was there any illusion that Kannex could pursue its own risk management strategy independent of the purposes of ABN. That knowledge alone would have been sufficient to enable Merrill to conclude that, in managing Kannex's participation, ABN had no use for a hedge within the partnership. There is also unequivocal evidence that Merrill was in fact aware of the activities ABN was conducting outside the partnership to hedge exposure to the Colgate debt on Kannex's behalf. Merrill consulted with ABN on revaluations of the Colgate debt, as an internal memorandum of the Merrill Swap Group explains, "since Kannex must actually trade Treasuries based upon the Base Treasury Yields".

The misleading explanations we find in the minutes were prepared long before the events they describe, during the planning of the section 453 investment strategy. Therefore, they raise the more fundamental issue of whether Merrill and Colgate could honestly and reasonably have planned to have ACM acquire the LIBOR Notes for ABN's use in managing the risks of Kannex's participation. The evidence is overwhelming that from the early stages in the planning of the liability management partnership at least Merrill, if not Colgate as well, expected that, as a matter of course, ABN would take steps to manage these risks independently.

To secure a partner for Colgate that would bear most of the interest rate risks of the liability management partnership's investments, Taylor approached the Financial Engineering Group of a major foreign commercial bank. As a rule, financial institutions like ABN do not expose themselves to interest rate risk; it is a common practice of such institutions to hedge their positions as promptly and fully as practicable. Taylor and his Swap Group knew this well enough to offer structured transactions that eliminated interest rate risks to BOT, BFCE, and Sparekassen, as well as to all the banks that issued or purchased LIBOR Notes in connection with each of the section 453 partnerships that Merrill promoted. Taylor's Swap Group would not need to offer similar services to ABN. That would be the responsibility of den Baas and his Financial Engineering Group, whose regular business was to devise sophisticated structures for hedging interest rate and currency risks. As Kannex's financial adviser, den Baas's Group performed the function for which they had been recruited.

As a witness, Taylor was asked about his expectations concerning the manner in which ABN would handle the risks of Kannex's participation:

Q: * * * Mr. Taylor, based on your experience with complicated transactions, why did you not expect or believe that ABN was hedging its risks with respect to ACM partnership?

A: I didn't-did I say I didn't-I never said they wouldn't hedge.

Q: Okay. Did you believe that ABN would hedge its risks?

A: Did I believe that they would hedge their risks? Yes, in some way, sure.

Q:And hedge their risk with respect to their investment in ACM partnership?

A: However they saw fit.

The expectation that ABN would manage the risks of Kannex's participation "however they saw fit" does not square with the notion that a hedge within the partnership was designed for the principals' mutual benefit. That expectation, however, was a cornerstone in Merrill's design for the liability management partnership. The concept of creating a mechanism to separate treasury risk from credit risk and "allocating to each partner the risks that it is best able to bear" presupposed that the foreign partner would make use of the risk management capabilities in which it possessed a comparative advantage; it would have made no sense if the foreign partner were expected to rely upon risk management conducted at the partnership level. There was no attempt to make the LIBOR Note hedge into the kind of flexible and precise hedging instrument that Kannex would have required in order to provide Southampton with the Yield Component option it desired at an affordable cost. Taylor knew that this was unnecessary because ABN would not be relying on the LIBOR Notes in any case. Yordan testified that before the formation of ACM he asked Taylor why ABN would be willing for Kannex to bear the burden of the flexible interest rate risk allocation mechanism that Merrill contemplated. "His answer was that they intended to - to enter into some hedge transactions to neutralize that risk". As Taylor expected, Southampton was able to negotiate for the Yield Component option at little or no cost. The partnership was successful because ABN exploited its comparative advantage in a manner consistent with rational economic behavior, and did not behave in the manner implied by the theory of the LIBOR Note hedge.

The LIBOR Notes served no useful risk management function for the partnership. Nor was there any genuine expectation on anyone's part that they would. The theory of the LIBOR Notes as a hedge "within the four corners of the partnership" was nothing other than an elaborate tax avoidance scheme that had no economic substance.

Petitioner explains the investment in the Citicorp Notes on November 3, 1989, in part by the need for an interim use for the partners' cash contributions during the indefinite period during which efforts were made to identify and acquire Colgate debt. This is supported by the account that Taylor gave of the sequence of events in his trial testimony:

The partnership was funded on November 2nd. From that date forward, Colgate or Southampton-Hamilton was - was negotiating for the repurchase of a prior [sic-private] placement note from Met.

Merrill Lynch was trying to identify, locate, and purchase Colgate long bonds, and ABN Bank was charged with identifying, locating, and purchasing Euro notes * * * so, * * * the cash needed to be invested and it was invested in these notes. [Emphasis added.]

The weight of the evidence indicates that the search for Colgate debt had begun long before the partnership was funded, and that by the beginning of November the timing of the partnership's purchase of the debt was largely within its control. Between December 4 and 8, 1989, ACM acquired the Met Note, Euro Notes, and Long Bonds in an aggregate principal amount of $ 135.9 million. Prior to ACM's formation, Merrill prepared a series of cash-flow projections with respect to the investment activities of a liability management partnership under various assumptions. In the six projections between August 8 and September 7, 1989, the amount of Colgate debt in the partnership portfolio is arbitrarily assumed to be $ 50 million. The actual target contemplated since July was $ 140 million. In the October 24, 1989, projection, the amount of Colgate debt is assumed to be $ 138.95 million. In the October 27, 1989, projection, it is assumed to be $ 134.96 million. The later numbers are not arbitrarily selected for purposes of illustration. They clearly purport to be estimates. Both the precision and the accuracy of the estimates suggest strongly that by the time of the formation, at least several days before the Citicorp Notes were acquired, not only had Colgate debt been identified, but Merrill already had a very clear expectation of the prices.

In our Findings of Fact, we described in detail the pattern of studied hesitation and postponement calculated to hold up progress in consummating the purchases of Colgate debt. A brief summary will suffice. Weeks before his negotiations with Met Life on November 17, 1989, Pohlschroeder was ready, but unwilling, to negotiate. His reluctance to enter into discussion of specific terms before the appointed date was attributable at least in part to concern that the partnership's activities be conducted entirely offshore. Yet, the Partnership Committee authorized ABN Trust to proceed with the negotiations in an offshore location for that very reason, and evidently it made no efforts to do so. Final arrangements for the purchase of Long Bonds and Euro Notes in the marketplace were similarly delayed. In late October 1989, Pohlschroeder drafted standing orders which, on their face, purport to instruct Merrill not to make any purchases until the partnership had acquired the Citicorp Notes. Within 1 week after acquisition of the Citicorp Notes, the amount of cash that would be needed for purchases of the Colgate debt and the time it would be needed were definite enough that Merrill could press BOT and BFCE to conclude arrangements concerning the sale of the Citicorp Notes. The investment in the Citicorp Notes was not made to accommodate the timing of the acquisition of Colgate debt; rather, it was the reverse: The acquisition of the Colgate debt was timed so as to accommodate the requirements of the section 453 investment strategy.

If the timing of ACM's acquisition of Colgate debt was largely within the principals' control, and they were confident that negotiations could be concluded and sales closed within a short time, what the partnership needed for its temporary cash balances was a portfolio of short-term highly liquid investments. That need was not served by the decision to acquire an undiversified portfolio consisting of Citicorp's unregistered 5-year notes. Nor can that need explain the decision to liquidate the portfolio by means of a complex structured transaction in which a substantial amount of the principal would be consumed by dealer fees.

ACM's conduct subsequent to the Citicorp Note purchase belies the claim that the use of the Citicorp Notes as a temporary store for partnership cash was economically sound. The Citicorp Note investment would not have met the criteria for management of temporary cash balances set forth in the partnership's Investment Guidelines, had they been in effect at the time. But the adoption of the Investment Guidelines 2 weeks later, like so many partnership decisions, appears to have been scheduled to accommodate the section 453 investment strategy. Once the Citicorp Notes had been sold, the partnership was at liberty to follow sound investment principles. The $ 140 million cash generated in the sale was invested in a diversified portfolio of commercial paper instruments maturing after 7 days. No liquidation costs were incurred to obtain the cash needed for settlement of the Colgate debt purchases. But financial assets that could be converted into cash without a sale and registered financial assets that could be traded on an exchange at relatively little transaction cost would not have satisfied Colgate's tax needs.

Petitioner argues that the choice of private placement notes allowed ACM to negotiate for a put option, "a valuable option it could not otherwise have obtained". The logic appears to be backwards. The partnership did not choose the Citicorp Notes because they offered a put. If ACM had invested in short-term money market instruments or otherwise in accordance with the criteria in its belated Investment Guidelines, it would not have needed a put option. The option was valuable because the partnership chose to invest all of its cash in 5-year notes of a single issuer that were not tradeable on an exchange.

Petitioner sums up the manner in which the partnership executed the section 453 investment strategy as follows:

Although the evidence clearly indicates that the transactions ACM entered into were contemplated from the outset, it is equally plain from the record evidence that none of the ACM transactions was "pre-wired" or certain to occur. Moreover, it is clear that none of the parties was ever under any obligation to undertake any of the transactions: Ultimately market events and conditions dictated whether the transactions went forward and the terms on which they went forward.

The pattern of market-driven decisions that petitioner describes cannot be found in the record. On the contrary, the record reveals only a series of inconsistencies between the steps actually taken and the decisions that tax-independent considerations would have implied. There is no evidence that the occurrence or timing of any of these steps was a function of anything other than tax planning.

The documents that were drafted to explain the liability management partnership proposal for purposes of Colgate's internal review exhibit a scrupulous regard for the need to justify the proposal by reference to the relative costs of alternative structures. Petitioner presented expert opinion to the effect that ACM was a cost-effective vehicle for accomplishing Colgate's liability management objectives. By contrast, Pohlschroeder's account of how the decision was made to invest in the Citicorp Notes reveals a striking indifference to cost considerations. Petitioner points out, in support of its position that the consequences of ACM's transactions were not predetermined, that the partners' exposure to Citicorp's credit was "real, not theoretical". If the purchase of the Citicorp Notes confirms that market forces could have affected the economic outcomes for the partners, it also illustrates how little market considerations actually affected partnership decisions. Investing all $ 205 million of the partners' capital in Citicorp Notes, most of which would be sold at market price rather than held until they could be put back to the issuer at par, did subject the partnership to risk. The Investment Guidelines reflect the judgment that such risk normally would not be justifiable. In order to explain the acquisition of the Citicorp Notes as an interim use for idle cash, preparations to acquire the Colgate debt were suspended. Over the short period that the partnership planned to hold the notes, any premium yield it may have earned from them could not have covered an appreciable amount of the costs that the partnership expected to incur upon their sale.

The acquisition of the LIBOR Notes, ostensibly to minimize the partners' exposure to a rise in interest rates, was planned and carried out at a time when Colgate expected interest rates would fall over the next several months, and accordingly desired leveraged exposure to interest rate risk within the partnership. Instead of initially setting Southampton's share of the Yield Component at the target level of approximately triple its pro rata share, Colgate caused Southampton to increase its share in two steps over a period of 6 weeks to provide part of the rationale for why the partnership no longer needed the hedging effect of the BFCE Notes, in accordance with scenarios developed several weeks beforehand. The acquisition of the LIBOR Notes was planned with the expectation that Kannex would not in fact have any net exposure to hedge, and the acquisition proceeded as planned even after ABN and Merrill had entered into an agreement that would offset their effect on ABN's interest altogether.

Each of the steps in the section 453 investment strategy was planned and arrangements commenced considerably in advance of execution. ²⁶ Before the negotiations to form ACM, Merrill had already begun negotiations to purchase the Citicorp Notes. Before their purchase, Merrill was negotiating for their disposition. By the time ACM acquired the LIBOR Notes, Merrill was arranging with Sparekassen the terms on which some of them would be sold. The contingent payment sale was scheduled to take place before the end of ACM's first taxable year in order to permit the partnership to spread its tax basis in the Citicorp Notes over 6 years instead of 5. The distribution and sale of the BFCE Notes was scheduled to occur before the end of Colgate's 1989 taxable year in order to offset Southampton's share of the contingent payment sale gain on Colgate's consolidated return. It was the understanding of the principals that Kannex would retire from the partnership by the fall of 1991 so that the LIBOR Notes could be sold in time for Colgate to carry back the taxable loss to its 1988 taxable year. No supervening market forces or other nontax considerations disrupted the scheduled execution of these steps. "'If we stood at the top of the world and looked down on this transaction', we would see events unfolding during the year[s] * * * about as they were contemplated when the plan was adopted." Braddock Land Co. v. Commissioner, 75 T.C. 324, 331-332 (1980) (quoting Mathews v. Commissioner, 520 F.2d 323, 325 (5th Cir. 1975)).

But for the $ 100 million of tax losses it generated for Colgate, the section 453 investment strategy would not have been consistent with rational economic behavior. The section 453 investment strategy lacked economic substance. It served no useful nontax purpose. Accordingly, the pertinent adjustments made by respondent to ACM's reported items of income and loss are sustained.

To reflect the foregoing,

---

Footnotes

• 1. During the period at issue, Merrill Lynch & Co., Inc., was a holding company that, through subsidiaries and affiliates, provided various financial services. We use the name "Merrill" to refer generally to the affiliated group or a member thereof.

• 2. LIBOR is the primary fixed income reference rate used in Euro markets.

• 3. All other things being equal, the longer the maturity of a debt instrument the more sensitive its value will be to fluctuations in market interest rates. Hence, long-term debt tends to carry greater risk than short-term debt of the same issuer.

• 4. During the period at issue, ABN New York was a subsidiary of Algemene Bank Nederland, N.V., one of the Netherlands' largest financial institutions. ABN Trust Co., Curacao, N.V., was another subsidiary. For purposes of this Opinion, the name "ABN" refers to Algemene Bank Nederland, N.V., or any one of its subsidiaries, affiliates or branches.

• 1. In its review of these costs, as part of a separate document, Colgate translated aftertax amounts into pretax amounts using a 34-percent marginal rate. The original aftertax estimateof Merrill's advisory fee ($ 1.32 million) would imply a pretax amount of $ 2 million. The discrepancy between this and the $ 1.75 million figure reflected in this separate document was not explained.

• 5. A bid-ask spread is the spread between the price at which an instrument is bought and sold. The bid price is the price at which dealers buy the instrument, and the ask price is the price at which dealers sell the instrument.

• 6. The original name of the partnership was CAM Partnership. At the first meeting of the Partnership Committee, for reasons not disclosed in the record, the initials of Colgate and ABN (A) were reversed, and the name became ACM.

• 1. Includes rounding error of .01

• 7. Kannex's share was set slightly higher (83 percent) and Southampton's slightly lower (16.7 percent) than their respective capital interests.

• 8. The term "notional principal amount" means that the principal amount is not actually exchanged; rather, parties agree to exchange payments based on the notional amount.

• 9. During 1989, Sparekassen was the largest savings bank in Denmark. Later, in the same year, it merged with the two other banks to form Unibank. We refer to the bank at all times as Sparekassen.

• 10. ACM computed the gain as follows:

  Payments received in FYE 11/30/89		$ 140,000,000
  Basis recovered in FYE 11/30/89
  Citicorp Note basis plus
  accrued interest	175,504,564
  Portion allocable to
  FYE 11/30/89 (1/6)		(29,250,761)
  Capital gain		110,749,239

• 11. Colgate computed the loss as follows:

  Cash proceeds	$ 9,406,180
  Imputed interest on contingent
  payments	(48,693)
  Amount realized	9,357,487
  Citicorp Note basis plus
  accrued interest	50,144,161
  Basis allocable to LIBOR
  Notes (5/6)	41,786,801
  Section 1274 interest accrued
  by ACM	525
  Adjusted basis allocable to
  LIBOR Notes	41,787,326
  Capital loss	32,429,839

• 12. "Outside basis" refers to a partner's basis in its partnership interest.

• 13. This figure represents the aggregate face amount of Colgate long-term debt held by the partnership ($ 136.6 million) minus the decline that would have occurred in any case during 1991 owing to a scheduled principal payment ($ 12.5 million).

• 14. In financial terminology, a "structured transaction" is one that combines two or more financial instruments or derivatives. Most structured transactions, like those in this case, include at least one derivative.

• 15. The banks did not actually pay Merrill Capital the full amount of the interest coupons they received from Citicorp, nor did Merrill Capital pay them the full amounts payable to ACM under the LIBOR notes. On each payment date amounts owed by each counterparty to a swap were offset, and only the net payments were made. The netting of payments is standard practice in the swap market and was provided for in all of the swap agreements discussed hereafter.

• 16. Beder is affiliated with the New York consulting firm of Capital Market Risk Advisors, and serves on the faculty of the Yale School of Management.

• 1. The approximate calculations are: $ 222,586 savings divided by $ 25 million in principal, spread over 2.3 year duration of principal payments; $ 88,323 savings divided by $ 9,831,661 in principal, spread over 2.3 year duration of principal payments.

• 17. As will be seen hereafter, Merrill Capital did not retain all of the remaining $ 1.5 to $ 1.6 million of value extracted from the partnership. Some of this value was transferred back to ABNs and Kannex through a separate set of swaps relating to the LIBOR notes.

• 1. The approximate calculation is: $ 7,208 gain divided by $ 9,406,180 invested, spread over 0.189 year duration of payments. The calculation assumes that Merrill Capital will cancel the swap on March 1, 1990, when the opportunity to do so first arises: At the inception of the swap, the prospect of a decline in BFCE's credit sufficient to warrant retention of the option at the large cost that this would impose was highly unlikely.

• 18. Details of the syndication of the loan to Kannex and details of Kannex's ultimate liquidation, which are related hereafter, shed light on the character of the relationship between Kannex and ABN.

• 19. We need not, and do not, delve into the appropriateness of reporting the transaction on the installment method. We are compelled to note, however, that the installment method reports income, sec. 453(a), and the partnership sold the Citicorp Notes for consideration equal to the notes' purchase price.

• 20. In United States v. Wexler, 31 F.3d 117, 123 (3d Cir. 1994), the Court of Appeals for the Third Circuit described Goldstein as "the seminal sham transaction case".

• 21. We believe the CINS transaction also was real and not fictitious. In Rice's Toyota World, Inc. v. Commissioner, 752 F.2d 89 (4th Cir. 1985), affg. in part and revg. in part 81 T.C. 184 (1983), the Court of Appeals for the Fourth Circuit concluded that a transaction was a sham because it lacked business purpose and economic substance. In Lerman v. Commissioner, 939 F.2d 44, 53-54 (3d Cir. 1991), affg. Fox v. Commissioner, T.C. Memo. 1988-570, the Court of Appeals for the Third Circuit adopted the Second Circuit's definition of a sham transaction as "a transaction that 'is fictitious or * * * has no business purpose or economic effect other than the creation of tax deductions.'" (quoting DeMartino v. Commissioner, 862 F.2d 400, 406 (2d Cir. 1988), affg. 88 T.C. 583 (1987)). The CINS transaction was not a sham in the sense that it was fictitious but it was a sham in the sense that the sec. 453 investment strategy lacked economic substance.

• 22. The Court of Appeals for the Third Circuit commented that "Sheldon actually expanded the sham transaction doctrine because it barred interest deductions from arrangements motivated by tax benefits even if the transactions could have generated a profit." United States v. Wexler, 31 F.3d 117, 124 n.9 (3d Cir. 1994).

• 23. The BOT Notes had a tax basis of $ 104.467 million. Even if we assume that interest rates rose by 500 basis points, causing an increase in the cost to acquire Kannex's interest in the notes from $ 20.955 million ($ 25.361 million x .8263) to $ 29.283 million ($ 35.439 million x .8263) and a decrease in the taxable loss recognizable on the sale of Kannex's interest in the notes from $ 66.842 million (($ 104.467 million - $ 23.574 million) x .8263) to $ 58.622 million (($ 104.467 million - $ 33.521 million) x .8263), each $ 1 that Colgate paid to acquire Kannex's interest would still produce more than $ 2 of taxable losses.

• 24. The change did not materially affect the size of the anticipated tax loss.

• 25. The hedging benefit is maximal if it is realized immediately. The longer it takes for interest rates to rise, the lower the present value of this benefit.

• 26. Respondent's expert, Irving H. Plotkin, concluded that:

  In judging the economic rationality of the Partnership, it must be remembered that the complex financial transactions and the profits realized by the parties did not occur as a reaction to or consequence of random economic factors. Likewise, the very low pretax rate of return suffered by Colgate was not the result of poorly chosen investments or of any unexpected adverse market conditions. Rather the transactions and the returns were the result of a carefully crafted and faithfully executed sequence of sophisticated and costly financial maneuvers that left little to chance or market opportunities. The score for the Partnership's actions was very detailed and the libretto even included the writing of the minutes of the Partnership meetings weeks before those meetings occurred. The actual Partnership transactions conformed to each of the seven steps choreographed in Merrill Lynch's September 1989 presentation to Colgate.