Full opinion text
MEMORANDUM OF DECISION UNDERHILL, District Judge. TIFD III-E Inc. (“TIFD III-E”) has sued the United States of America to recover approximately $62 million that TIFD III-E deposited with the Internal Revenue Service (“I.R.S.”) in satisfaction of an alleged tax liability. That tax liability arose from the I.R.S.’s determination that TIFD III-E had incorrectly calculated and reported the amount of income TIFD III-E earned as a partner in Castle Harbour-I Limited-Liability Company (“Castle Har-bour”). The case was tried to the court over eight days. The court’s findings of fact and conclusions of law are set forth below. I. Procedural History Castle Harbour was a Nevada limited liability company. In every year from 1993 to 1998, it filed a partnership tax return, also known as a Form 1065. (J. Exs.22, 40, 50, 56, 59, 70) TIFD III-E, a Delaware corporation, was one of the owners of Castle Harbour. Because Castle Harbour was treated as a partnership for tax purposes, TIED III-E paid United States tax on the income attributed to it on Castle Harbour’s Form 1065. In 2001, the I.R.S. issued two notices of Final Partnership Administrative Adjustments (“FPAAs”) concerning Castle Har-bour. (Pl.’s Exs. 377, 378) The FPAAs attributed approximately $310 million of additional income to TIFD-III E, resulting in an additional tax liability of $62,212,010. Pursuant to section 6226(a) of the Internal Revenue Code (“I.R.C.”), TIFD III-E filed a complaint against the United States challenging the FPAAs. Before filing, TIFD III-E deposited the disputed sum of $62,212,010 with the I.R.S. as required by I.R.C. section 6226(e). Because it is the tax matters partner of Castle Harbour, TIFD III-E is authorized to bring this suit. I.R.C. § 6226(a) II. Findings of Fact On the basis of the testimony, exhibits, and designated depositions, I make the following findings of fact. A. Background TIFD III-E is a wholly owned subsidiary of the General Electric Capital Corporation (“GECC”), a subsidiary of the General Electric Company (“GE”). Among other things, GECC is in the business of commercial aircraft leasing. (July 21, O’Reilly, 48; July 21, Lewis, 134) Typically, airlines do not own aircraft, principally because airlines do not ordinarily produce sufficient income to take advantage of the tax depreciation deductions generated by commercial aircraft. (July 21, Brickman, 204-05) Instead, a company with greater taxable income, such as GECC, will buy the planes and lease them to airlines, thereby giving the airlines the use of the aircraft and the lessor company the tax deductions. (July 21, Brickman, 204^05) In the early 1990s, the airline industry experienced a number of setbacks, including several bankruptcies. (July 21, O’Reilly, 49; July 21, Lewis, 137; July 22, Dull, 299; July 22, Parke, 385-86; July 23, Nay-den, 451) These events caused GECC concern for its own prospects in the aircraft leasing business. (July 21, O’Reilly, 66-72; July 21, Lewis, 141-42; July 22, Parke, 386) In 1992, at least partially in response to this concern, GECC executives began looking for ways to reduce GECC’s risk in the aircraft leasing business. (July 21, Lewis, 155-56) To do this, GECC initiated what it referred to as a “sell-down” effort — an attempt, among other things, to raise immediate cash against GECC’s aircraft assets. (July 21, O’Reilly, 77 — 78; July 21, Lewis, 158-59; July 26, Nayden, 455-56) In other words, rather then simply awaiting return in the form of the — now less certain' — rental income, GECC wanted to lower its risk by raising immediate cash against that future stream of income. Selling the aircraft or borrowing money against them, two straightforward ways of raising capital, were not options. Sale was not a realistic option because, in general, the secondary market for aircraft was weak. (July 21, O’Reilly, 70) This was particularly true with respect to GECC’s older aircraft — known as “Stage II” aircraft — which did not meet certain regulatory standards, including those for noise. (July 21, O’Reilly, 50; July 22, Dull, 308) Non-recourse debt was not an option for two reasons. First, in order to maintain its AAA credit rating, GECC had an agreement with credit rating agencies that prohibited GECC from borrowing more than eight times its common equity. (July 21, O’Reilly, 95-96; July 22, Parke, 379) In 1993, GECC’s debt-to-common-equity ratio was 7.96 to 1, giving it little room to borrow. (July 22, Parke, 381) Second, a number of GECC’s medium-term and long-term debt instruments contained a “negative pledge” — a covenant prohibiting GECC from using its assets to secure debt other than purchase money debt. (July 21, O’Reilly, 96; July 22, Parke, 384) With the two most obvious options for raising money off the table, GECC sought outside advice concerning other possibilities. In May 1992, GECC submitted Requests for Proposal (“RFPs”) to seven investment banks. (Pl.’s Exs. 141, 142, 143, 146; J. Exs. 8, 10) The RFPs sought proposals that would, in essence, allow GECC to solicit outside investment in at least part of its aircraft leasing business. All of the investment banks submitted proposals; none of them met all of GECC’s objectives. (July 21, O’Reilly, 88-89) Nevertheless, in March 1993, after some back and forth, the investment bank Babcock & Brown submitted a revised proposal that GECC found acceptable. (July 21, O’Reilly, 90) Babcock & Brown eventually received a $9 million fee for its work. (July 21, Brick-man, 225) Babcock & Brown’s final proposal called for the creation of a separate entity to which GECC would contribute a number of aircraft. (D.’s Ex. 22) Investors would then be solicited to purchase ownership shares in the new entity, (J. Ex. 17) The result would be that GECC would trade some of the risks and returns of those aircraft to the outside investors in exchange for a cash contribution to the newly created entity. The proposal also called for the investors to be foreign tax-neutral entities, an arrangement that would offer lucrative tax savings to GECC. After GECC senior management approved the proposal, it was implemented in two stages. First, on July 26, 1993, GECC formed a Nevada limited liability company known as GE Capital Summer Streei>-I Limited Liability Company (“Summer Street”), which was owned by three GECC subsidiaries: TIFD III-E, TIFD III-M, and General Electric Capital AG. (Pl.’s Ex. 118; July 22, Dull, 315) Through these subsidiaries, GECC contributed to Summer Street: (a) 63 “Stage II” aircraft worth $530 million, but subject to $258 million of non-recourse debt (a net value of $272 million); (b) $22 million of rents receivable on the aircraft; (c) $296 million in cash; and (d) all the stock of GECC subsidiary TIFD VI, which had a value of $0. (July 22, Dull, 314-15, 331) Second, on October 6, 1993, the GECC subsidiaries sold $50 million of their interest in Summer Street, which included all of GE Capital AG’s interest, to two Dutch banks, ING Bank N.V. and Rabo Merchant Bank N.V. (collectively, the “Dutch Banks”). (July 22, Dull, 322-23) The Dutch Banks also contributed an additional $67.5 million, bringing their total investment to $117.5 million. (July 22, Dull, 322-23) Summer Street then changed its name to Castle Harbour-I Limited Liability Company (“Castle Harbour”) (Pl.’s Exs. 118, 119), and TIFD VI changed its name to Castle Harbour Leasing, Inc. (“CHLI”) (July 22, Dull, 331) B. Structure of Castle Harbour When the final stage of the Castle Har-bour transaction was completed on October 6, 1993, the parties entered into an Amended and Restated Operating Agreement (“the Operating Agreement”). (J. Ex. 1) The Operating Agreement dictated the way gains and losses were allocated between TIFD III-E, TIFD III-M (collectively the “GECC entities”), and the Dutch Banks. Consequently, understanding the terms of the Operating Agreement is essential to understanding the tax consequences at issue in this litigation. 1. Overview Castle Harbour was a self-liquidating partnership. Through two entities, GECC contributed to Castle Harbour a net $246 million in cash and, more importantly, approximately $294 million worth of leased aircraft. The Dutch Banks contributed approximately $117.5 million in cash. Each partner received an allocation of the net income of the partnership. The Dutch Banks were referred to as Class A partners, the GECC entities as Class B partners. (J. Ex. 1 at Bates 101430-31) Over eight years, the Dutch Banks’ ownership interest was to be almost entirely bought out with the income of the partnership. This self-liquidation mechanism can best be understood by considering an actual year in the partnership. Castle Harbour’s first full year of operation was 1994. At the beginning of 1994, each partner’s ownership interest — recorded in a “capital account” entry on Castle Harbour’s financial statements (J. Ex. 1 at Bates 101403-31) — was approximately the same as at the start of the partnership. GECC’s capital account was approximately $540 million. (J. Ex. 24 at Bates 8812) The Dutch Banks’ combined capital accounts were approximately $112 million. (J. Ex. 24 at Bates 8812) In 1994, Castle Harbour received approximately $100 million in gross income— mostly rent from aircraft leases. In accordance with the Operating Agreement, the net income for 1994, approximately $9.8 million, was distributed $9.6 million to the Dutch Banks and $0.2 million to GECC. (J. Ex. 37 at Bates 16344) (The mechanics of net income calculation and allocation are discussed below.) Accordingly, the Dutch Banks’ combined capital accounts increased to approximately $122 million, and the GECC entities’ accounts increased negligibly, remaining at around $540 million. The Dutch Banks’ capital accounts, however, did not actually increase because part of the gross rent of $100 million was used to “buy-out” approximately $40 million of the banks’ combined ownership interest. (J. Ex. 37 at Bates 16344) Thus, at the end of 1994, the Dutch Banks’ capital accounts, originally at $112 million, had increased by $10 million (allocation) but decreased by $40 million (buy-out), resulting in final combined capital accounts of approximately $82 million. This general pattern was to be repeated for eight years. Each year the Dutch Banks were to have their capital accounts debited or credited, depending on whether the partnership had received a gain or suffered a loss, and each year the Dutch Banks were to have a significant portion of their ownership interest bought out by the partnership. The amount of the annual buy-out payment was set forth in the Operating Agreement at Exhibit E, giving rise to the name “Exhibit E payments.” (J. Ex. 1 at Bates 101662) The Exhibit E payments were scheduled to pay cash annually in amounts that would provide the Dutch Banks with an internal rate of return of 9.03587% over eight years. At the end of eight years, if the Dutch Banks’ capital accounts had actually earned a rate of return 9.03587%, the Dutch Banks’ capital accounts, i.e., ownership interests, would be decreased to near zero — in other words, Exhibit E payments would have cancelled out the Dutch Banks’ capital account increases and returned the Dutch Banks’ initial investment. Similarly, if the Dutch Banks’ capital accounts were credited with partnership income at a rate less than 9.03587%, the capital accounts would be negative after eight years; if the capital accounts were credited at a rate greater than 9.03587%, the capital accounts would be positive. Positive capital accounts would result in payments to the banks when the partnership wound up; negative accounts would mean the banks owed money to the partnership. (Op. Agmt. § 12.2-12.3, J. Ex. 1 at Bates 101492-93). If the banks’ interests were not liquidated after eight years, the banks would still have their capital accounts credited or debited by allocations of income or loss in successive years. This arrangement provided the Dutch Banks, in return for their initial cash investment, with (a) an ownership interest in Castle Harbour that was increased or diminished by allocations of income or loss and (b) a stream of payments over eight years that would repay the Dutch Banks’ initial investment at an internal rate of return of 9.03587%. Any discrepancies between these two items — the Dutch Banks’ ownership shares and the total payments made to the banks over the eight years— would be reconciled upon liquidation of the banks’ partnership interests. The arrangement allowed GECC to reduce part of its interest in fixed assets to cash, i.e., it “monetized” or “securitized” part of the assets. Before the formation of Castle Harbour, GECC — through its wholly owned subsidiaries — owned 100% of its $272 million of aircraft (net), $22 million in associated rent receivables, and $240 million in cash. After the formation of Castle Harbour, GECC owned approximately 82% of a partnership that itself owned the aircraft, receivables, GECC’s cash, and an additional $117.5 million in cash. In other words, pre-Castle Harbour, GECC owned 100% of $272 million in aircraft, 100% of $22 million in receivables, 100% of $240 million cash, and 0% of $117.5 million cash. Posi^Castle Harbour, GECC owned 82% of each asset, i.e., $223 million of aircraft, $18 million of receivables, and $290 million in cash. Over the first eight years, however, while the Dutch Banks’ ownership interest was being bought out using Castle Harbour’s income, GECC’s interest was increasing inversely. Consequently, at the end of the eight years (had the partnership lasted that long), GECC would have regained almost complete ownership of the aircraft. In sum, through Castle Harbour, GECC received $117.5 million in cash from the Dutch Banks, in return for which the banks received a self-liquidating (i.e., limited-term) ownership interest in aircraft and an allocation of the rental income from those aircraft. Put another way, GECC sold the Dutch Banks what amounted to an eight-year investment in a commercial leasing partnership. 2. Allocations The general structure of Castle Harbour just described is fairly simple and perhaps not very different from that of a great many partnerships. The complexity of the transaction, and the source of this litigation, comes principally from the way in which the partnership’s income was allocated between the GECC entities and the Dutch Banks. Crucial to its allocation scheme, the Operating Agreement defined two categories of income (or loss): Operating Income and Disposition Gains/Losses. a. Operating Income Operating Income was comprised of income less expenses. Income was rent and interest on investments. Expenses consisted of normal administrative expenses, interest owed on aircraft debt, depreciation of the aircraft, and guaranteed payments to GECC entities, also known as Class B Guaranteed Payments. (J. Ex. 1 at Bates 101425-26) Depreciation was calculated on a straight-line basis, starting with the fair-market value of the aircraft at the time of Castle Harbour’s formation and assuming a useful life for each aircraft of the greater of 7 years or 125% of the time remaining on the then outstanding lease of the aircraft. (J. Ex. 20 at Bates 8785) The Class B Guaranteed Payments were made annually to the GECC entities in the amount of either $500,000 or $2 million and did not reduce the capital account of the receiving GECC entity. (Op. Agmt. § 4.1, J. Ex. 1 at Bates 101447) Notably, expenses did not include payment of principal on the airplane debt or Exhibit E payments to the Dutch Banks. Once Operating Income had been calculated, it was allocated to the capital accounts as follows. If Operating Income was positive, i.e., an Operating Gain, it was allocated 98% to the Dutch Banks and 2% to the GECC entities. (Op. Agmt. § 3.1, J. Ex. 1 at Bates 101435) If Operating Income was negative, i.e., an Operating Loss, then it was (a) first allocated in an amount sufficient to offset the cumulative Disposition Gains allocated to any of the partners in previous years, (b) the remainder was then allocated 98% to the Dutch Banks until they had been allocated, cumulatively, $3,854,493 of Operating Losses, and (c) the remainder was allocated 99% to the GECC entities and 1% to the Dutch Banks. (Op. Agmt. § 3.2, J. Ex. 1 at Bates 101435-36) Operating Gain allocation was straightforward. For example, in 1997 Castle Harbour’s net Operating Income was $2,304,000, which was allocated $2,258,000 to the Dutch Banks and $46,000 to the GECC entities — a simple 98/2 split. (J. Ex. 61 at Bates 18002) Castle Harbour never experienced an Operating Loss, but a hypothetical situation will illustrate how one would have been allocated. Assume that in 1994 Castle Harbour had an Operating Gain of $10 million and a Disposition Gain (described below) of $10 million. The Operating Gain would have been allocated $9.8 million to the Dutch Banks and $200,000 to the GECC entities. The Disposition Gain, assuming it was Castle Harbour’s first ever, would have been allocated (for reasons explained later) approximately $3 million to the Dutch Banks and $7 million to GECC. If in 1995 Castle Harbour had a — very unlikely — Operating Loss of $15 million, it would have been allocated as follows. First, $3 million would be allocated to the Dutch Banks and $7 million to the GECC entities to offset their prior Disposition Gains. Second, the remaining $5 million would be allocated 90% to the Dutch Banks and 10% to the GECC entities until the Dutch Banks had been allocated $3,854,493 of cumulative Operating Income losses. Assuming the Dutch Banks in previous years had never been allocated an Operating Loss, this second step would mean that $4,282,770 would be allocated 90/10, resulting in $3,854,493 going to the Dutch Banks and $428,277 going to the GECC entities. Third, the remaining approximately $717,000 would be allocated 99% to the GECC entities and 1% to the Dutch Banks. b. Disposition Gain/Loss A Disposition Gain or Loss was the result of the difference between the sale price of an asset, usually an aircraft, and its book value. (Op.Agmt. § 3.3(h),(j), J. Ex. 1 at Bates 101438-39) Alternatively, if an aircraft was distributed back to one of the GECC entities, the fair market value was deducted from the receiver’s capital account and the difference between the aircraft’s fair market value at the time of distribution and its book value was treated as a Disposition Gain or Loss. (Op.Agmt. § 10.8(a)(i)(B), J. Ex. 1 at Bates 101488) Similarly, if the GECC entities were to buy out entirely the Dutch Banks’ interest in the partnership, the difference between ■ the fair market value of all held assets and their book value was to be treated as a Disposition Gain or Loss. (Op.Agmt. § 10.8(b)(i)(B), J. Ex. 1 at Bates 101488) Disposition Gains and Losses were allocated much like Operating Losses: (a) first, Disposition Gains were allocated to offset prior Disposition Losses and prior Operating Losses; Disposition Losses offset prior Disposition Gains, (b) the remainder was then allocated 90% to the Dutch Banks until they had been allocated, $2,854,493 of either Disposition Gains or Losses, (c) the remainder was allocated 99% to the GECC entities and 1% to the Dutch Banks. (Op.Agmt. § 3.3(h),(j), J. Ex. 1 at Bates 101438-41) For example, in 1995 Castle Harbour disposed of a number of aircraft, some to TIFD III-E and some to third parties. The aircraft distributed to TIFD III-E had a fair market value of approximately $27 million, and consequently TIFD III— E’s capital account was reduced by that amount. (J. Ex. 49 at Bates 8969) The aircraft sold to third parties were sold for approximately $21 million. (J. Ex. 49 at Bates 8969) The book value of all these aircraft was approximately $74 million, causing Castle Harbour a Disposition Loss of approximately $26 million. (J. Ex. 49 at Bates 8969) That loss was distributed as follows. First, because in prior years cumulative Disposition Gains of approximately $3 million had been allocated to the Dutch Banks, and cumulative Disposition Gains of approximately $1.5 million had been allocated to the GECC entities (J. Ex. 37 at Bates 16347), those amounts of loss were allocated to each respectively. Second, the Dutch Banks were allocated 90% of the remaining losses and GECC, 10%, until the Dutch Banks had been allocated $2,854,493 in losses, and GECC had been allocated about $0.3 million. Third, the remainder was allocated 99% to GECC (approximately $18 million) and 1% to the Dutch Banks (approximately $0.1 million). In total, therefore, the Dutch Banks were allocated approximately $6 million in Disposition Losses, and the GECC entities, approximately $20 million. (J. Ex. 49 at Bates 8970) c. Operating Income vs. Actual Income Operating Income, as defined by the Operating Agreement, might at first glance look like a simple measure of the net cash received by Castle Harbour in its normal operations, i.e., gross non-disposition income less expenses. That was not the case. Operating Income in fact defines a non-obvious category of income, primarily because it includes as expenses items not clearly considered expenses, e.g., Class B Guaranteed Payments, and excludes items that appear to be expenses, e.g., debt payments and Exhibit E payments. It is worth pointing out some of the effects of this definition. As noted above, the Class B Payments guaranteed to the GECC entities were treated as expenses and did not reduce the receiver’s capital account. Consequently if one were to consider such payments as allocations — which would make sense given that they represented income going directly to GECC entities — the income allocated to GECC would then be significantly more than 2%. For example, in 1997, as discussed above, Operating Income was approximately $2.3 million, and was allocated 98% to the Dutch Banks and 2% to GECC. If one considers that GECC also received a Class B distribution of $2,000,000 prior to any allocation and not deducted from its capital account, it might make sense to consider its allocation to be $2,046,000, making the actual split closer to 50/50. The treatment of aircraft depreciation as an expense, and its consequent deduction from Operating Income, also had some interesting effects. The depreciation schedule for the aircraft was fairly aggressive, usually coming out to between 60 to 70 percent of the rental income for a given year. The effect of this depreciation was that a large portion of the cash that came into Castle Harbour was not reflected in Operating Income. Of course, aggressive depreciations meant Castle Harbour would be more likely to realize a gain when the assets were sold (or the Dutch Banks were bought out), but this gain would be a Disposition Gain and therefore allocated more favorably to GECC. 3. Other Provisions The principal features of the Castle Har-bour transaction have been described: it was a self-liquidating partnership with a complex scheme for allocating gains and losses. There are three other features of the partnership relevant to this case. a. CHLI Most of the cash invested in Castle Har-bour was not held by Castle Harbour. Instead it was transferred to Castle Har-bour Leasing Inc. (“CHLI”), a domestic corporation and wholly owned subsidiary of Castle Harbour. (July 22, Dull, 322-23, 358; July 22, Parke, 394) This arrangement allowed income from any asset (cash or aircraft) to be recognized only as a Disposition Gain rather than as Operating Income, simply by moving that asset to CHLI. For example, interest earned on the reinvestment of the approximately $360 million cash initially invested by the partners would have counted as Operating Income had the cash been held in Castle Harbour. Because the cash was moved to CHLI, interest accumulated there and was allocated to the partners as a Disposition Gain when the Dutch Banks were ultimately bought out. Similarly, CHLI purchased several aircraft during Castle Har-bour’s existence. (July 22, Dull, 325) Rental income generated by those aircraft did not count towards Operating Income, as it would have if the aircraft were owned by Castle Harbour, and that income was not allocated to the partners until the buyout of the Dutch Banks, when it was allocated as a Disposition Gain. (July 22, Dull, 370) b. Investment Accounts Under the Operating Agreement, Castle Harbour was required to maintain “Investment Accounts” for the Dutch Banks. (J. Ex. 1 at Bates 101405-06) No cash was actually paid into these accounts; they merely kept track of a hypothetical balance. Id. The opening balance of these accounts was the initial investment made by the Dutch Banks. Id. That balance was to be recalculated at the time the Dutch Banks exited the partnership as if every year the balance had been increased by a defined Applicable Rate but also reduced by the Exhibit E payments. Id. The Applicable Rate was either 9.03587% or 8.53587%, depending on the reason for the Dutch Banks’ exit. Id. If at the time of the Dutch Banks’ exit from Castle Harbour the Investment Account balance exceeded the algebraic sum of the Dutch Banks’ allocation of (a) Operating Gains, (b) Operating Losses that had been allocated at the 98% rate, which could not exceed approximately $4 million, (c) Disposition Gains, and (d) Disposition Losses that had been allocated at the 90% rate, which could not exceed approximately $3 million, that amount would be paid to the Dutch Banks, instead of the amount in their capital accounts. That payment, if made, was labeled a Class A Guaranteed Payment. Id. c. Core Financial Assets As discussed above, Castle Harbour put most of its cash in its subsidiary, CHLI. CHLI was not free to dispose of that cash as it chose. Under the Operating Agreement, CHLI was required to keep high-grade commercial paper or cash, referred to as “Core Financial Assets,” in an amount equal to 110% of the current value of the Investment Accounts. (J. Ex. 1 at Bates 101408; Op. Agmt. § 5.8(b), J. Ex. 1 at Bates 101471-72) As it turns out, CHLI ended up investing most of its $360 million in GECC commercial paper. (July 22, Dull, 324) This benefited GECC because, by having a GECC subsidiary — Castle Harbour — purchasing GECC commercial paper, GECC was buying back or “retiring” its debt, thereby decreasing its debt-to-equity ratio and freeing GECC to borrow more money. (July 22, Parke, 393) d.Management Rights Castle Harbour was managed primarily by the GECC entities who elected the partnership’s managers. The Dutch Banks’ role was minimal. They did not vote for managers and none of their employees worked for Castle Harbour. They did participate in annual member meetings (July 22, Dull, 362), but, for the most part, the only control they had was negative (July 22, Dull, 361). The actual day-to-day operations of Castle Harbour, such as financing and accounting activities, were outsourced to other GE entities' — -first, GE Capital Advisory Services Ltd., then, General Electric Capital Aviation Services, Ltd. (“GECAS Ltd.”). (July 22, Dull, 412-13; July 26, Tewell, 646-47) 4. Risks and Returns Having explained Castle Harbour’s structure, it is worth summarizing how this complex structure allocated the risks and returns of the Castle Harbour business. The Dutch Banks received the lion’s share of Operating Income, though this number was greatly reduced from gross rental income because depreciation was treated as an expense. By contrast, the Dutch Banks were not likely to receive much upside from the disposition of assets. As a practical matter, their return on asset dispositions was capped at about $3 million. Although they also received 1% of Disposition Gains above $3 million, that amount was insignificant compared to their overall investment. For example, even if Castle Harbour had sold all its assets at approximately $303 million over book value, the Dutch Banks would only have recognized an additional $3 million of return. By comparison, Castle Harbour actually disposed of the assets at a cumulative gain of $137 million, which netted the Dutch Banks only $1.3 million above their initial $3 million allocation. Similarly, under the Operating Agreement, the Dutch Banks were exposed to little more than a $3 million risk of Disposition Losses and a $4 million risk of Operating Losses, i.e., a total risk of little more than $7 million. Again, that risk was capped because, for amounts above that range, the Dutch Banks were exposed to only 1% of the risk for each type of loss (Disposition or Operating). For example, if the aircraft had simply been given away (i.e., a Disposition Loss of roughly $530 million) in 1994, the Dutch Banks would only have been exposed to approximately $5.27 million dollars of loss above their initial $3 million allocation of Disposition Losses. Accordingly, under the Operating Agreement, the Dutch Banks were entitled to any Operating Income upside, but probably little more than approximately $3 million of any Disposition upside. On the other hand, they were not likely to be allocated much more than $7 million in losses, $4 million from Operating Losses and $3 million from Disposition Losses. The Dutch Banks were actually protected against the possibility of even that $7 million in losses by their Investment Accounts. The Operating Agreement provided that, if the Dutch Banks’ Investment Accounts exceeded the sum of their Operating Gains, Disposition Gains, Operating Losses up to $4 million, and Disposition Losses up to $3 million, the banks would be paid the difference. Thus, even if the Dutch Banks were allocated up to $4 million of Operating Losses and $3 million of Disposition Losses, they would still receive the full amount of their Investment Accounts. That is, the only negative effects of these losses would be to limit the Dutch Banks’ payout to the return rate of their Investment Account (either 9.03587% or 8.53587%). That is not to say that the Dutch Banks were guaranteed to receive at least their Investment Account amounts. Those amounts could be reduced by the 1% allocation of Disposition Losses or Operating Losses, because such loss amounts were not included in the determination whether a Class A Guaranteed Payment was required. To illustrate the risks to the Dutch Banks, it is useful to start with what actually happened. When the Dutch Banks were actually bought out, their capital accounts were worth $31.1 million (J. Exs.71, 72) and their Investment Accounts were worth approximately $29 million (using the applicable interest rate of 8.53587%) (July 22, Dull, 366). Because their allocated gains, $31.1 million, exceeded their Investment Accounts, no Class A Guaranteed Payment was made; in other words, the Investment Accounts were irrelevant. (July 22, Dull, 365) The situation would have been different if Castle Harbour had done worse. If, for example, Castle Harbour had a cumulative Operating Loss of $1 million and no Disposition Gain or loss, the Dutch Banks would have received a lower return. They would not, however, have received a negative return, that is, they would not have been required to pay more money into Castle Harbour. Castle Harbour would have suffered a loss, and that loss would have been allocated 98% to the Dutch Banks, bringing their capital account to negative $980,000. The result would have been that their Investment Accounts ($29 million) would have exceeded the sum of their Disposition Gains and Losses ($0) and their Operating Losses below $3 million ($980,-000). Consequently, the Dutch Banks would have been bought out for only $29 million. Thus, though they would have received a lower return than if Castle Har-bour had done better, their return would still have been positive. The Dutch Banks would only have received less than the amount in their Investment Accounts if Castle Harbour had done badly enough to cause losses to be distributed to the Dutch Banks at the 1% allocation rate, i.e., greater than approximately $4 million in Operating Losses or $3 million in Disposition Losses. Even then the effect in most scenarios would be minimal because the allocation would be only 1%. For example, the Dutch Banks would only have received a zero return, i.e., an allocation of $28.8 million in losses, if Castle Harbour experienced approximately $2.8 billion in Disposition or Operating Losses, an unlikely scenario given that the combined assets of Castle Har-bour never exceeded a value of $700 million. In short, the Castle Harbour Operating Agreement shifted to the Dutch Banks principally the upside of the aircraft rentals and the risk that the upside would not exceed their Investment Accounts. The Dutch Banks also received a small portion of any upside from disposition of assets. Theoretically, the Dutch Banks also bore some risk from Disposition and Operating Losses, but, for the reasons just given, the risk was minimal. 5. Tax Consequences The tax consequences of the Castle Har-bour partnership allocations were significant. They are also relatively simple to understand. As described above, 98% of Operating Income was allocated to the Dutch Banks. Operating Income was reduced by expenses, including asset depreciation, which in most years equaled close to 70% of gross rental income. For tax purposes, the same allocation was made; Operating Income was allocated 98% to the Dutch Banks. Again, Operating Income was reduced by expenses, including depreciation, but all the aircraft in Castle Harbour had already been fully depreciated for tax purposes. Consequently, although nominally depreciation was an expense for tax purposes, it did not actually reduce taxable income. Accordingly, the taxable income allocated to the Dutch Banks was greater than their book allocation by the amount of book depreciation for that year. The Dutch Banks, however, did not pay United States income taxes. Thus, by allocating 98% of the income from fully tax-depreciated aircraft to the Dutch Banks, GECC avoided an enormous tax burden, while shifting very little book income. Put another way, by allocating income less depreciation to tax-neutral parties, GECC was able to “re-depreeiate” the assets for tax purposes. The tax-neutrals absorbed the tax consequences of all the income allocated to them, but actually received only the income in excess of book depreciation. Thus, the full amount of book depreciation was available, pre-tax, to Castle Harbour to use. C. Operation of Castle Harbour Having explained the structure and mechanics of Castle Harbour, I will now summarize the actual results of its five years of operation. Castle Harbour operated from its formation on October 6, 1993 until December 31, 1998 when GECC liquidated the Dutch Banks’ interest. Its principal place of business was in Bermuda. For most of its period of operation, Castle Harbour’s lease portfolio was managed by GECAS under the terms of an administrative services agreement. (J. Exs. 26, 31, 32; July 22, Dull, 412-13; July 26, Tewell, 646-47) GE-CAS also helped Castle Harbour place aircraft when their existing leases expired. (J. Ex. 32; July 26, Hyde, 534-35) From 1993 to 1998, Castle Harbour’s cumulative Operating Income was approximately $28.6 million. (J. Exs.20, 24, 37, 49, 53, 61, 68) Approximately $28 million of that income was allocated to the Dutch Banks. Id. During its period of operation, Castle Harbour disposed of a number of aircraft at a cumulative loss of about $24 million. Id. When the Dutch Banks were bought out, the value of all aircraft and of CHLI exceeded their respective book values by approximately $161 million. (J. Ex. 71 at Bates 24679) Consequently, Castle Harbour had a cumulative Disposition Gain of approximately $137 million. Approximately $4 million of that was allocated to the Dutch Banks. Exhibit E payments were made through 1997 in an amount totaling approximately $118.5 million. At the time of the buyout, the Dutch Banks had a positive balance in their capital accounts of approximately $31 million. They were bought out at that price. In total, the Dutch received nearly $150 million ($118.5 million Exhibit E payments, plus $31 million buyout) over five years for their $117.5 million investment. Stated more usefully, they received an internal rate of return of approximately 9.1%. (July 27, Myers, 817) The GECC entities were allocated approximately $600,000 of the $28.6 million of Operating Income, and approximately $133 million of the $137 million Disposition Gains. (J. Exs.20, 24, 37, 49, 53, 61, 68) Over the period of Castle Harbour’s operation, GECC received $6 million in Class B payments, approximately $20 million in distributions from its capital accounts, and distributions of aircraft worth about $41 million. Id. In 1998, after GECC bought out the Dutch Banks for approximately $31 million, it became the sole owner of the assets of Castle Harbour’, worth approximately $692 million. In total, GECC received nearly $728 million ($6 million Class B payments, plus $20 million distributions, plus $41 million aircraft, plus $692 assets, minus $31 million buyout) over five years for its initial investment of around $591 million. These payments gave GECC a pre-tax internal rate of return of approximately 5.5%. (July 27, Myers, 762) For the reasons given above, the Dutch Banks were allocated much more taxable income than book income. Specifically, the Dutch Banks were allocated approximately $310 million in taxable income. Had this income been allocated to GECC, GECC would have been required to pay approximately $62 million in taxes on that income. III. Conclusions of Law The government argues three alternative theories in defense of its reallocation of Castle Harbour’s income. First, the government argues that Castle Harbour was formed with no non-tax purpose, making its formation a “sham” transaction to be disregarded when calculating taxes. Second, the government argues that, even if the arrangement had a business purpose, the Dutch Banks were, for tax purposes, only lenders to Castle Harbour, not partners, and could not, therefore, be allocated any partnership income. Third, the government argues that, even if Castle Harbour should be treated as a partnership for tax purposes, the way it allocated income violated the “overall tax effect” rule of section 704(b) of the Internal Revenue Code. A. Standard of Review The parties do not dispute that I am to review TIFD III-E’s tax liability de novo, ignoring the factual findings and legal analysis of the Commissioner of Internal Revenue (“the Commissioner”). R.E. Dietz Corp. v. United States, 939 F.2d 1, 4 (2d Cir.1991). The ultimate determination of the Commissioner, however, is presumptively correct. Therefore the taxpayer, TIFD III-E, bears the burden of persuading me that the determination is incorrect. Id. B. Was Castle Harbour’s Formation a “Sham” Transaction? Regardless of its literal compliance with the tax code, a transaction will be deemed a “sham” and disregarded when calculating taxes if it has no business purpose or economic effect other than the creation of tax benefits. Jacobson v. Commissioner, 915 F.2d 832, 837 (2d Cir.1990). There is no dispute that the Castle Har-bour transaction created significant tax savings for GECC. The critical question, however, is whether the transaction had sufficient economic substance to justify recognizing it for tax purposes. Newman v. Commissioner, 902 F.2d 159, 163 (2d Cir.1990). To determine whether a transaction has economic substance or is, instead, a “sham,” a court must examine both the subjective business purpose of the taxpayer for engaging in the transaction and the objective economic effect of the transaction. Gilman v. Commissioner, 933 F.2d 143, 148 (2d Cir.1991). TIFD III-E takes this to mean that if I find either a subjective business purpose or objective economic effect, the transaction is not a sham. The government adopts another reading of the law, arguing that I should apply a more flexible standard that considers both factors but makes neither dispositive. The decisions in this circuit are not perfectly explicit on the subject. Recently, for example, Judge Arterton adopted the more flexible standard, but acknowledged some potentially contrary, or at least ambiguous, language in Gilman. Long Term Capital Holdings v. United States, 330 F.Supp.2d 122, 171-72 (D.Conn.2004). That ambiguity, however, does not affect the decision of this case. As I will explain, under either reading I would conclude that the Castle Harbour transaction was not a “sham.” The transaction had both a non-tax economic effect and a non-tax business motivation, satisfying both tests and requiring that it be given effect under any reading of the law. 1. Economic Effect In light of my findings of fact, I have little trouble concluding that the Castle Harbour transaction had a real, non-tax economic effect. In return for a significant portion of Castle Harbour’s Operating Income, the Dutch Banks contributed approximately $117 million dollars, which was used by Castle Harbour’s subsidiary CHLI either to purchase aircraft or to retire GECC debt. The economic reality of such a transaction is hard to dispute. The Dutch Banks gave up $117 million and received part of Castle Harbour’s Operating Income in return. Castle Harbour received the Dutch Banks’ $117 million, assigned part of its Operating Income in return, and put the $117 million to use in an ongoing, substantial business. The government does not dispute that, if this is what occurred, the transaction would have economic effect. Instead, the government argues that various provisions of the Operating Agreement stripped these apparent realities of any substance. Specifically, the government argues that: (1) the Dutch Banks did not really give up their $117 million, because the combination of Investment Accounts and Exhibit E payments guaranteed return of that money; and (2) Castle Harbour did not really gain the use of the $117 million because the “Core Financial Assets” provision of the Operating Agreement essentially “froze” that money. As explained in my findings, the Investment Accounts and Exhibit E payments served different purposes. The Exhibit E payments provided the Dutch Banks with a guarantee that they would receive fixed payments over eight years, resulting in the ultimate buyout of most of their ownership interest. The net result of the Investment Accounts, on the other hand, was to provide the Dutch Banks with some amount of security for their investment, namely, that they would almost certainly receive no less than an 8.5% return. In other words, Exhibit E payments guaranteed the manner of payment, whereas the Investment Accounts guaranteed the amount. It is hard to see how an assurance about the manner in which returns would be paid could undercut the economic reality of an initial investment. In year one, Castle Harbour received, and used, $117 million from the Dutch Banks. Over subsequent years, that amount was paid back in installments out of the gross rental proceeds of the aircraft leasing business. The latter fact does not change the former — Castle Harbour received an economically real, upfront payment of $117 million. In truth, I am not sure that the government takes issue with this point. Rather, the government appears to believe that Exhibit E payments somehow guaranteed the Dutch Banks an amount of return. They did not. The Exhibit E payments reduced the banks’ capital accounts, while Operating Income increased them. If, when the Dutch Banks were bought out, the Exhibit E payments had “overpaid” the banks’ accounts they would owe money (and conversely, if the banks were underpaid they would be owed money). In other words, it was quite possible for the Dutch Banks’ allocation to fall short of the sum of Exhibit E payments and, in that case, the banks would be required to make a payment to Castle Harbour. The Investment Accounts, by contrast, did provide the Dutch Banks with some guarantee of return. As explained earlier, the Dutch Banks were almost entirely certain of at least an 8.5% internal rate of return on their investment. Latching on to this fact, the government asserts that, because of this guarantee, there was no risk to the Dutch Banks and, therefore, no economic reality to their investment. I do not agree. First, a lack of risk is not enough to make a transaction economically meaningless. Even with an 8.5% guaranteed return, the Dutch Banks still participated in the — economically real — upside of the leasing business. The better the leasing business did, the more money the Dutch Banks made. In fact, they made a return of approximately 9.1% on their investment (i.e., an amount greater than 8.5%), and, had things gone better, they would have made even more. Participating in upside potential, even with some guarantee against loss, is economically substantial. Second, the government’s premise, that a guarantee of a positive return indicates no risk, is simplistic. Whether an investment is “risky” to the investor depends on a number of factors, including the investor’s cost of capital and opportunity costs. Had the Dutch Banks, for example, borrowed the $117 million at 8.6% or foregone an opportunity to lend that money at 8.6%, the chance that they would only earn an 8.5% internal rate of return would have unquestionably represented a real risk. In short, entirely open-ended risk is not the only economically real risk. The government also argues that Castle Harbour did not really raise any money, because the Operating Agreement required CHLI to maintain 110% of the Dutch Banks’ Investment Accounts in “Core Financial Assets,” i.e., high-grade commercial paper or cash. Thus, says the government, the $117 million was effectively “frozen,” not available for use. This argument is missing a step. It is true that the Operating Agreement contained the provision in question, but that provision only restricted the money’s use; it did not forbid it. At the very least, CHLI was permitted to invest in GECC commercial paper, which it did. That investment was not meaningless, because it allowed GECC to retire that amount of commercial paper, thereby improving its debt-to-equity ratio and creating borrowing capacity under GECC’s agreement with the credit rating agencies. 2. Btisiness Purpose TIFD III-E contends that GECC’s non-tax purpose in entering into the Castle Harbour transaction was to raise capital and, more importantly, to demonstrate to investors, rating agencies, and GECC senior management, that it could raise capital on its fleet of aging Stage II aircraft. The most direct evidence in support of this contention was the testimony of five GECC executives, who all swore that “demonstrating liquidity” and “monetizing” Stage II aircraft were important motivations. In evaluating the economic substance of a transaction, courts are cautioned to give more weight to objective facts than self-serving testimony. Lee v. Commissioner, 155 F.3d 584, 586 (2d Cir.1998). Were the executives’ testimony the only evidence before me, I am not sure how persuaded I would be of GECC’s motives. As things stand, however, against the backdrop of the objective economic reality of the Castle Harbour transaction — i.e., that GECC did raise $117 million and increase its liquidity by retiring debt — I find the testimony of GECC’s executives persuasive. Consequently, I find that GECC was subjectively motivated to enter into the Castle Harbour transaction, at least in part, by a desire to raise capital and a desire to demonstrate its ability to do so. C. Were the Dutch Banks Partners in Castle Harbour? The government takes the alternative position that, even if the Castle Harbour transaction as a whole had economic substance, for tax purposes the Dutch Banks were not partners of the GECC entities but rather were their creditors. Though neither party has put it exactly this way, there are two circumstances under which the Dutch Banks would not be considered partners: (1) if there was no economic reality to the label “partner;” and (2) if, regardless of the economics of the situation, the tax code simply classifies them as something else. In either case, the practical effect of declaring that the Dutch Banks were not partners would be to reassign all of Castle Harbour’s income to GECC. In the first circumstance, the analysis is essentially the same as the “economic substance” analysis just undertaken, but, rather than examining the substance of the entire transaction, I would only address the narrow question whether there was any economic reality to the choice of the partnership form. Indeed, there is a narrow subset of case law, derived in part from the general economic substance doctrines and in part from the Supreme Court’s more specific partnership analysis in Commissioner v. Culbertson, 337 U.S. 733, 742, 69 S.Ct. 1210, 93 L.Ed. 1659 (1949), that address just this point. In the second circumstance, the question is not whether the form chosen has economic reality, but whether tax law requires a different choice of form, i.e., a different classification. I have not seen a case in which a court has undertaken this type of “classification” analysis of a partnership, and I am skeptical that under the current tax code there is much chance that an economically substantial partnership would ever be classified as something else. Nevertheless, the government urges me to examine the Dutch Banks’ partnership interest using the same standard used in determining whether a commercial instrument is classified as “debt” or “equity” for the purpose of taxing distributions. For reasons I will explain below, I do not think the debt/equity test is relevant to classifying a partnership — the Tax Code’s definition of a partnership is extremely broad and easily met in this case. Moreover, even applying the test urged by the government, there is no question in my mind that the Dutch Banks, for tax purposes, held equity in Castle Harbour. 1. Economic Substance of the Dutch Banks’ Partnership Interest The decision to form a partnership may be economically insubstantial, even though the partnership undertakes a legitimate business. In other words, a transaction may have economic effect, yet there may be no non-tax reason to join together with a third party to engage in that transaction. This situation has received particular attention in a recent line of D.C. Circuit decisions starting with ASA Investerings Partnership v. Commissioner, 201 F.3d 505 (D.C.Cir.2000). In ASA Investerings, the D.C. Circuit examined a complex partnership and concluded that, even if the taxpayer had a legitimate non-tax business purpose for engaging in the transaction in question, the formation of a partnership with various foreign entities to accomplish that goal served no non-tax purpose. Specifically the court asked the question (derived from the Supreme Court’s analysis in Culbertson ), “whether, all facts considered, the parties intended to join together as partners to conduct business activity for a purpose other than tax avoidance.” ASA Investerings, 201 F.3d at 513. The D.C. Circuit answered the question in the negative, noting that “[tjhere is no reason to believe that AlliedSignal[, the taxpayer,] could not have realized Matthews’s interest rate play[, one of the transaction’s benefits,] without the partnership at far, far lower transaction costs.” Id. at 516. The court also concluded that the foreign investor’s interest did not look much like a partner’s interest because the investor was guaranteed an exact return — “A partner whose risks are all insured at the expense of another partner hardly fits within the traditional notion of partnership.” Id. at 515. In a subsequent case, involving a nearly identical transaction, the D.C. Circuit was even more explicit that a partnership would not be recognized if its formation served no business purpose. “The only logical explanation then, for the partnership’s formation was the exploitation of Temp. Treas. Reg. § 15A.453-l(c)(3)(l) .... The absence of a non-tax business purpose is fatal to the recognition of the entity for tax purposes.” Boca Investerings Partnership v. United States, 314 F.3d 625, 632 (D.C.Cir.2003). Neither ASA Investerings nor Boca In-vesterings enunciate a new standard of review. Both cases are only applications of the general economic substance or “sham transaction” doctrine. The ASA Investerings court said as much, noting, “[courts] treat ‘sham entity’ cases the same way the law treats ‘sham transaction’ cases.” ASA Investerings, 201 F.3d at 512. In fact, the D.C. Circuit directly rejected the argument that different standards might apply: “Although the Tax Court said it would not consider whether the transactions at issue lacked ‘economic substance,’ its decision rejecting the bona fides of the partnership was the equivalent of a finding that it was, for tax purposes, a ‘sham.’ ” Id. (emphasis in original). I have already explained why the Castle Harbour transaction did not lack economic substance. Nevertheless, ASA Invester-ings raises the possibility that, though the transaction as a whole has economic substance, the formation of a partnership to accomplish the transaction might, absent tax considerations, be economically meaningless. On the facts of this case, however, such a holding is not possible. If I had concluded, for example, that Castle Harbour was not a sham because it was engaged in the business activity of leasing aircraft, or if I concluded that it was not a sham because it earned GECC a pre-tax profit, then I might very well need to consider separately the question whether the use of a partnership with foreign banks to conduct such activity was a sham, serving no non-tax purpose. In a transaction where a part of an ongoing business is spun off into a separate partnership, the fact that the underlying business has economic substance does not necessarily preclude a finding that the creation of the spin-off was a sham transaction. In other words, when two parties form a new entity, the fact that the newly created entity has a business purpose does not always mean that the act of creating the entity was not a sham. Here, however, I did not conclude that the Castle Harbour transaction had economic substance because Castle Harbour engaged in a legitimate business. Rather, I concluded that the transaction that created Castle Harbour was not a sham. In other words, I concluded there was valid business purpose and economic reality in the arrangement by which the GECC entities and the Dutch Banks came together to form Castle Harbour, i.e., there was economic substance in not only the actions, but also the formation, of the partnership. In this respect, the Castle Harbour transaction differs significantly from the transaction at issue in ASA Investerings. In that case the court was principally concerned that (1) the outside “investors” appeared to have absolutely no stake in the partnership and (2) there seemed to be no reason to form a separate entity to engage in the underlying transactions, other than to avoid taxes. Neither situation is present here. The Dutch Banks had a very real stake in the transaction because their return was tied directly to the performance of the aircraft leasing business. If the business did better, their return was greater, if the business did worse, their return was less. By contrast, the foreign banks in ASA Investerings were guaranteed an exact amount of return regardless of the business’s performance. It is true that the Dutch Banks’ risk was mitigated by the existence of the Investment Accounts, which guaranteed them a minimum return, but in ASA Investerings the court was careful to distinguish the case where an investor’s exact return is guaranteed, making the investor entirely indifferent to the partnership’s activities, from the case where downside is limited but not upside, giving the investor an obvious interest in the performance of the partnership. The former situation indicates a sham, the latter — the situation present here — is not. See ASA Investerings, 201 F.3d at 514; see also Hunt v. Commissioner, 59 T.C.M. (CCH) 635, 1990 WL 66551 (1990) (holding valid partnership existed even though one partner guaranteed a minimum return of 18%). With respect to the reason for forming a separate entity, not only was there a legitimate non-tax reason to create a separate entity — so that the Dutch Banks and GECC could share an investment in a specific business — but it is actually hard to imagine an alternative to creating a separate entity. As already discussed, a non-recourse loan on the aircraft was not possible because of the “negative pledge” and eight-to-one debt/equity covenant. Accordingly, given that GECC wanted to raise money against its aircraft, and given that it could not borrow against them, it is difficult to see what else it could have done other than create a separate entity and seek investments in that entity. Of course, even if there had been another way of achieving this financing, it would not change my analysis; the creation of a partnership was one — even if not the only— legitimate way of achieving the non-tax purpose of raising capital against some of GECC’s Stage II aircraft. That is all the economic substance test requires. 2. Characterization of the Dutch Banks’ Interest in Castle Harbour The government argues that, even if the Dutch Banks had an economically substantial interest in Castle Harbour, their interest should, as a matter of tax law, be characterized as a creditor’s interest, not a partner’s. In support of this argument, the government cites exclusively to the line of cases stemming from the Supreme Court’s holding in John Kelley Co. v. Commissioner, 326 U.S. 521, 66 S.Ct. 299, 90 L.Ed. 278 (1946). Those cases discuss whether various commercial instruments are considered debt or equity for tax purposes — usually in order to know whether to treat return on those instruments as deductible interest or non-deductible dividends. Debt/equity analysis differs fundamentally from “sham” transaction analysis. John Kelley, 326 U.S. at 523, 66 S.Ct. 299 (“There is not present in either situation the wholly useless temporary compliance with statutory literalness which this Court condemned as futile, as a matter of law, in Gregory v. Helvering, 293 U.S. 465, 55 S.Ct. 266, 79 L.Ed. 596 (1935).”). In fact, the two analyses are exclusive. “Classification” analysis asks the question what formal classification is appropriate for tax purposes, whereas “sham transaction” analysis asks whether an otherwise appropriate formal classification should be disregarded. Accordingly, rather than asking me to apply the reasoning of ASA Investerings, as it did in its previous argument, the government here asks me to question what ASA Investerings presupposed, namely, that the entity in question was, as a, formal matter, correctly classified as a partnership. See ASA Investerings, 201 F.3d at 511 (noting that despite compliance with the tax code’s formal definition of a partnership, the question was “whether the formal partnership had substance”). Given that the question is how to classify the Dutch Banks’ interest in Castle Har-bour, I am more than a little puzzled by the cases cited. If the question is whether Castle Harbour was formally a partnership, it makes sense to look not to the case law distinguishing between debt and equity for the purpose of determining appropriate deductions but to the tax code’s definition of a “partnership.” Section 761 of the Internal Revenue Code states that “the term ‘partnership’ includes a syndicate, group, pool, joint venture, or other unincorporated organization through or by means of which any business, financial operation, or venture is carried on.” 26 U.S.C. § 761. The section further provides, “[f]or purposes of this subtitle, the term ‘partner’ means a member