SECOND CIRCUIT REVERSES DECISION IN CASTLE HARBOUR CASE.

In a corporate income tax dispute involving the use of the partnership form for the purpose of allocating income among partners, the Internal Revenue Service adjusted the taxpayer's return, disallowed the allocation, and assessed $62 million in additional taxes based on a finding that the foreign banks to which the income was allocated were not bona fide equity partners. Upon the taxpayer's suit challenging the adjustment, the United States District Court for the District of Connecticut (Stefan R. Underhill, J.) ruled that the banks were bona fide equity partners, and accordingly disallowed the Commissioner's adjustments. We reverse the judgment of the district court and remand for further proceedings.

Reversed and remanded.

This appeal tests the power of the Internal Revenue Service to examine and recharacterize an interest which accords with its ostensible classification only in illusory or insignificant respects. The taxpayer-plaintiff TIFD III-E, Inc. (the "taxpayer" or "TIFD III- E"), a subsidiary of General Electric Capital Corporation ("GECC"), brought suit in the United States District Court for the District of Connecticut (Stefan R. Underhill, J.) to challenge adjustments made by the Internal Revenue Service ("IRS") to the tax returns for 1993 to 1998 of a partnership named Castle Harbour Limited Liability Company ("Castle Harbour" or the "partnership"), for which the taxpayer was the tax-matters partner. The IRS's adjustments added $62 million to the taxpayer's tax bill. After an eight-day bench trial, the court ruled in favor of the taxpayer, TIFD III-E Inc. v. United States, 342 F. Supp. 2d 94 (D. Conn. 2004), and the government brought this appeal. We reverse the judgment of the district court.

The litigation turns primarily on the propriety of the partnership's ostensible allocation of its income as between the taxpayer and two Dutch banks, ING Bank N.V. and Rabo Merchant Bank N.V. (the "Dutch banks" or the "banks"), which invested in the partnership. The Dutch banks do not pay taxes to the United States. The partnership's "Operating Agreement" allocates to them 98% of the "Operating Income," which comprises the great majority of the partnership's income. The Operating Income, calculated for tax purposes, however, vastly exceeded the amounts the banks would actually receive. The net Operating Income, of which 98% would go to the banks, was drastically reduced by huge depreciation deductions which the IRS would not recognize, as the assets in question had already been fully depreciated. The effects of the ostensible allocation of the majority of the partnership's income to the non- taxpaying Dutch banks were to shelter most of the partnership's income from taxation and to redirect that income tax-free to the taxpayer. What the Dutch banks were in fact to receive from the partnership was dictated by provisions of the partnership agreement calling for the reimbursement of their initial investment at an annual rate of return of 9.03587% (or, in some circumstances, 8.53587%), subject to the possibility of small adjustments and to the possibility of a slight increase in the event of unexpectedly great partnership earnings. The banks' reimbursement at the agreed rate of return was formidably secured by a variety of contractual undertakings by the taxpayer and its parent GECC.

Castle Harbour filed a federal partnership return every year from 1993 to 1998. In 2001, the IRS issued two notices of Final Partnership Administrative Adjustments ("FPAAs"), the first covering 1993 through 1996 and the second covering 1997 through 1998. Each provided for reallocation of Castle Harbour's income. The IRS rejected the partnership's treatment of the two banks as bona fide equity partners for tax purposes and accordingly rejected the partnership's allocations. The effect of the reallocation was to assign a far greater percentage of Castle Harbour's income to the taxpayer, rather than the banks. The FPAAs attributed approximately $310 million in additional income to the taxpayer, imposing on the taxpayer an additional tax liability of $62,212,010.

The taxpayer deposited this sum with the IRS and, pursuant to 26 U.S.C. § 6226, brought suit in 2001 against the United States challenging the validity of the FPAAs. The IRS claimed two primary justifications for its adjustment -- first, that under the authority of cases such as ASA Investerings v. Commissioner, 201 F.3d 505 (D.C. Cir. 2000), and Boca Investerings Partnership v. United States, 314 F.3d 625, 632 (D.C. Cir. 2003), the arrangement was a "sham"; and, second, that the interest of the Dutch banks was not, for tax purposes, a bona fide equity partnership participation because the banks had no meaningful stake in the success or failure of the partnership. The IRS argued that their investment was in the nature of a secured loan, and that whatever aspects of their interest resembled an equity partnership participation were either illusory or insignificant.

The court conducted a bench trial in the District of Connecticut. By memorandum and order dated November 1, 2004, the district court ruled that the FPAAs were invalid and ordered the IRS to refund the taxpayer's deposit. TIFD III-E, 342 F. Supp. 2d at 121-22. The court entered judgment on November 3, 2004.

The provisions of the immensely complex partnership agreements are analyzed in the district court's thorough, comprehensive, and detailed opinion. See id. The court essentially acknowledged that the creation of the partnership was largely tax-motivated. The court nonetheless found that the partnership had other bona fide purposes, and some genuine economic effect. It therefore rejected the government's contention that the partnership's construct was a sham, which should be disregarded. The court concluded that the Dutch banks were, for tax purposes, partners of Castle Harbour.

While the government raises several arguments on appeal, we focus primarily on its contention that the Dutch banks should not be treated as equity partners in the Castle Harbour partnership because they had no meaningful stake in the success or failure of the partnership.¹ In its analysis of this question, the district court made several errors of law which undermined the soundness of its conclusions. The facts of the allocation of partnership resources, as set forth in the partnership documents (and as found by the district court), compel the conclusion that the IRS correctly determined that the Dutch banks were not bona fide equity participants in the partnership. We accordingly reverse the judgment.

A. Overview

GECC has long been in the business of owning commercial aircraft which it leased to airlines. The economic benefits of aircraft ownership are primarily two -- the revenues produced by employment of the aircraft in the business of air transportation, and the depreciation deductions the aircraft generate, which can substantially reduce the owner's tax liability. However, once an owner has taken depreciation covering its full investment in the aircraft, no further depreciation deductions may be claimed. GECC had long produced revenue by leasing the aircraft to airlines and realized beneficial tax relief through depreciation. Eventually, however, GECC found itself in the position of owning a fleet of aircraft that had been fully depreciated and could thus no longer serve as the basis for depreciation deductions. By reason of its inability to take further depreciation, its ownership of these aircraft had become less remunerative. GECC solicited proposals from investment banks for financing of these aircraft.

The Castle Harbour partnership, which is the subject of this litigation, resulted from a proposal presented to GECC by Babcock & Brown, to which GECC paid $9 million for its assistance in the creation and execution of the plan. Following Babcock & Brown's proposal in 1993, GECC caused the formation of an eight-year partnership, later named Castle Harbour, and solicited foreign financial companies, which are not subject to tax in the United States, as investors in the partnership. GECC caused its subsidiaries to transfer the ownership of a fleet of fully depreciated aircraft under lease to airlines to the Castle Harbour partnership, in which its subsidiary TIFD III-E, the taxpayer in this action, became the tax-matters partner. The assets transferred by GECC entities to the partnership were the fleet of aircraft, with a market value of $272 million, $22 million in receivables from aircraft-rental agreements, and $296 million in cash, making a total investment of $590 million. Shortly thereafter, the two Dutch banks contributed $117.5 million in cash to the partnership.² The district court found that the taxpayer contributed 82% of the partnership's capital and the Dutch Banks contributed 18%. See TIFD III-E, 342 F. Supp. 2d at 100. The Dutch banks' participation in the partnership was to be passive. They were to exercise no role in the management, which was assigned entirely to the taxpayer.³ The documents of the partnership characterize the Dutch banks as equity partners. This characterization is the main focus of this dispute.

B. Castle Harbour's Division of Assets, Revenues, and Losses

An extraordinarily complex maze of partnership provisions dictates how the revenues, losses, and assets of the partnership would be allocated over the eight-year duration of the partnership.

Using complex definitions, the partnership agreement allocated 98% of what the parties and the district court referred to as the "Operating Income" of the partnership to the Dutch banks. See TIFD-III E, 342 F. Supp. 2d at 102-03. Operating Income was a flexible classification. It included most of the partnership's taxable income, while allowing the taxpayer when it so desired to reclassify an income stream, taking it out of Operating Income and designating it instead as a "Disposition Gain." Disposition Gains were allocated (after a threshold amount) primarily to the taxpayer. For tax purposes, the allocation of 98% of the partnership's Operating Income to the non-taxpayer Dutch banks meant that only a tiny portion of the income of the partnership would be subject to tax.⁴

The partnership's Operating Income was reduced by expenses, the largest of which was the aggressive depreciation of its aircraft. Because the aircraft had already been fully depreciated for tax purposes, however, this depreciation did not serve to reduce the partnership's taxable income. As a result, the 98% allocation of Operating Income to the Dutch banks created an enormous discrepancy between the banks' share of the partnership's taxable income and their share of its book value. When it came to the actual division of the assets, revenues, and losses, the partnership did not credit the Dutch banks' capital accounts with the same 98% of the taxable Operating Income described above, but rather with 98% of a much smaller figure, drastically reduced by depreciation charged against the already fully depreciated aircraft. The partnership agreement was designed essentially to guarantee the Dutch banks the reimbursement (according to a previously agreed eight-year schedule) of their initial investment of $117.5 million at an annual rate of return of 9.03587% (or, in some circumstances, 8.53587%), referred to in the agreement as the "Applicable Rate." The rate of return was subject to the possibility of a small increase in the event of unforeseen, extraordinary partnership profits. The scheduled reimbursement of the Dutch banks, at the Applicable Rate of annual return, was in no way dependent on partnership performance. These payments to the Dutch banks were assured in numerous ways -- most importantly, by the guaranty of the taxpayer's parent, GECC, and accordingly were virtually certain to be made, regardless of whether the partnership earned profits or suffered losses.

The taxpayer and the banks were each permitted to terminate the arrangement prior to its scheduled termination date by paying a small penalty. If, for example, the Dutch banks terminated their participation prematurely, the agreement provided that the Applicable Rate would drop from 9.03587% to 8.53587%. By the same token, if the taxpayer elected to terminate the arrangement early, exercising its continuing right to buy out the Dutch banks' interest, the banks would receive a premium of approximately $150,000. Finally, through the mechanism of Disposition Gains, as described below in greater detail, the Dutch banks would receive a slightly higher return if the profits of the partnership were unexpectedly great.

In short, the Dutch banks, which contributed about 18% of the partnership's capital and contributed nothing to its management, were allocated, for tax purposes, 98% of most of its taxable income. The actual distributions to be made to the banks, however, were arranged so that they would receive, according to a previously agreed schedule, the reimbursement of their investment, plus an annual return at an agreed rate near 9%, plus a small share in any unexpectedly large profits. See TIFD III-E, 342 F. Supp. 2d at 105, 107-08.

The above facts were either found by the district court or are not in dispute (or both). What is in dispute is whether for tax purposes the Internal Revenue Service acted within its power in refusing to recognize the Dutch banks' interest in the partnership as a bona fide equity participation. We conclude for reasons set forth at greater length below that the IRS properly refused to accept the partnership's characterization because the banks did not meaningfully share in the business risks of the partnership venture and their interest was overwhelmingly in the nature of secured debt.

C. Factors Affecting the Proper Characterization of Dutch Banks' Interests

The partnership interests of the Dutch banks were designed to have a superficial appearance of equity participation, but in the end (in all but a negligible part) to function in the manner of a repayment of a secured loan. This was achieved either by reason of the requirements of the partnership agreement that trumped the provisions that appeared to create an equity interest, or by reason of the powers given to the taxpayer and to the Dutch banks to protect their respective interests. The banks, as a consequence of these arrangements, did not meaningfully share the risks of the partnership business. We focus on aspects of the partnership documents that had this effect.

Exhibit E of the Operating Agreement specified the amounts the Dutch banks would receive in annual cash distributions. The Exhibit E payments were calculated to reimburse the banks' $117.5 million investment plus an annual rate of return at the agreed Applicable Rate. Whether Castle Harbour had a profit or loss did not affect the Exhibit E payments. Although the making of Exhibit E payments was not ostensibly required by the Operating Agreement, as a practical matter, it was required because in the event of a failure to pay an Exhibit E payment, the Dutch banks could unilaterally force the partnership's dissolution and would receive in such dissolution an amount which, when combined with the previous payments, would provide, subject to the possibility of small adjustments, reimbursement of the initial investment at the Applicable Rate of return.

Each Castle Harbour partner had a "capital account" and an "Investment Account." The capital account represented each partner's ostensible share of the partnership capital. Annually, the Dutch banks' capital accounts were to be credited or debited with the amount of their allocable shares of Castle Harbour's income or loss, and debited to reflect distributions of cash or property. The "Investment Account" for each Dutch bank did not hold money, but instead kept track of the minimum balance that the Dutch banks would receive upon dissolution. The opening balance was the Dutch banks' investment, and, at the time the Dutch banks exited the partnership, the "balance was to be recalculated . . . as if every year the balance had been increased by a defined Applicable Rate but also reduced by the Exhibit E payments." TIFD IIIE, 342 F. Supp. 2d at 104. If, at the dissolution of Castle Harbour, the Dutch banks' Investment Accounts exceeded the amount in their capital accounts, the Operating Agreement required that the Dutch banks receive a "Class A Guaranteed Payment" virtually equal to the difference between those two figures. Id. The effect of the Investment Accounts, as the district court found, was therefore to ensure that, were Castle Harbour to experience losses, the Dutch banks would nonetheless recover the reimbursement of their initial investment at the Applicable Rate of annual return.⁵Id. Thus, reimbursement at a minimum of the Applicable Rate of return was assured independent of the operating results of the partnership.

Furthermore, the Dutch banks' receipt of reimbursement, at the annual Applicable Rate of return, was elaborately protected. First, the taxpayer was required by the partnership agreement to keep "Core Financial Assets," consisting of high-grade commercial paper or cash, in an amount equal to 110% of the current value of the banks' Investment Accounts. The partnership, in addition, was obliged for the banks' protection to maintain $300 million worth of casualty-loss insurance. Finally, and most importantly, GECC -- a large and very stable corporation -- gave the banks its personal guaranty, which effectively secured the partnership's obligations to the banks. As the district court correctly found, these protections collectively ensured there was no realistic chance that the Dutch banks would receive less than the reimbursement of their initial investment at the Applicable Rate of annual return. TIFD III-E, 342 F. Supp. 2d at 105-06. The Dutch banks thus incurred no meaningful downside risk.

On the other hand, the banks' investment ostensibly had unlimited upside potential. If, upon dissolution of Castle Harbour, the banks' capital accounts exceeded their Investment Accounts, the partnership's Operating Agreement provided that the banks would receive the entire balance of their capital accounts. Because the agreement placed no limits on the amount of Operating Income that could be credited (at a rate of 98%) to the banks' capital accounts, it might appear that the banks would be entitled to a significant share of any unexpectedly high income earned by the partnership. As a practical matter, however, the ability of the banks to receive a share of unexpectedly large partnership returns was severely limited.

The taxpayer could limit the Dutch banks' participation in partnership profits through its power to transfer assets into a subsidiary of the partnership named Castle Harbour Leasing, Inc. ("CHLI"). The partnership agreement provided that the income or loss produced by any asset transferred into CHLI would no longer be classified as Operating Income but instead as "Disposition Gain" or "Disposition Loss."⁶ After an initial allocation of 90% of Disposition Gains to the Dutch banks up to a maximum of $2,854,493, Disposition Gains were then allocated 99% in favor of the taxpayer.⁷ Thus by reclassification of an income-producing asset, the taxpayer could reduce the banks' participation in the income produced by the asset from 98% to 1%. The taxpayer had every incentive to make such a reclassification in the case of sufficiently large unexpected profits, as it would increase the taxpayer's share of the income produced by the asset to 99%. Finally, the taxpayer had the ability, on payment of a negligible premium, to terminate the Dutch banks' interest in the partnership and thereby prevent allocation of earnings to the banks.⁸

In summary, the Dutch banks' interest in the eight-year partnership was essentially as follows: (1) They were promised the reimbursement, on a previously agreed schedule, of their initial $117.5 million investment at an agreed annual rate of return; (2) their repayment at the agreed rate of return was secured by the personal guaranty of GECC; (3) they were fully protected against risk of loss, except as to a tiny amount in highly unlikely circumstances; and (4) their ability to earn in excess of the agreed annual rate by participation in unexpected gains was, as a practical matter, capped at $2,854,493 (less than 2.5% of their investment), plus 1%.

D. The District Court's Opinion

The district court ruled in favor of the taxpayer, rejecting the government's argument that the Castle Harbour partnership was a sham. The court stated that in applying the sham test, the two considerations are whether any non-tax business purpose motivated the taxpayer to engage in the disputed transaction and whether that transaction had any objective economic effect. The district court found that GECC was motivated to enter into the Castle Harbour transaction, at least in part, (i) by a desire to raise capital otherwise than by borrowing, as GECC was prohibited from borrowing additional amounts by covenants with its lenders; and (ii) to demonstrate to investors, rating agencies, and GECC senior management that it could raise equity capital on its fleet of fully depreciated aircraft. TIFD III-E, 342 F. Supp. 2d at 111. Because the achievement of these ends depended on the Dutch banks being equity participants, rather than lenders, the court found a business purpose to the banks' equity partnership designation. Id. at 112-14. The court also found a genuine economic effect in that the taxpayer raised $117.5 million and used it to retire debt, while the banks participated in the partnership's profits. Id. at 111. See also id. at 109-10. The court further found the opportunity to participate in "upside potential, even with some guarantee against loss, [to be] economically substantial." Id. at 110.

The court then inquired into whether there was a basis to consider the banks not to be equity partners for tax purposes. As the court saw the question, the banks might properly be deemed not to be equity partners in two circumstances: (1) if there was no "economic substance" to the label "partner," and (2) if the tax code classified their interest as other than a partner's interest. Id. at 111. As for lack of economic substance, the court reasoned that it had already answered this question in the taxpayer's favor in its resolution of the sham analysis. As for the classifications commanded by the tax code, the court found that the banks' interest conformed to I.R.C. Section 761's definition of a partnership as including "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." TIFD III-E, 342 F. Supp. 2d at 115 (quoting I.R.C. § 761). The court thus resolved the question of the propriety of the partnership designation on the basis of its sham analysis, coupled with the formal definitions of the Code.

While questioning its relevance, the court went on also to consider the government's contention that the Dutch banks' interest was more akin to debt than to equity. The court observed that "the only possible relevance of the debt/equity analysis is as an aid to performing 'sham transaction' analysis." TIFD III-E, 342 F. Supp. 2d at 115. The court rejected the government's contention, based on its conclusion that (1) the banks were not owed a "sum certain"; (2) they enjoyed "unlimited upside" potential to participate in the partnership's profits; (3) the payment of the banks was subordinate to payment of general creditors; (4) the banks' interests were treated by the GECC companies as an equity interest (notwithstanding that the Dutch banks had at times described their interests as debt); and (5) insofar as GECC's lenders allowed GECC to obtain this financing notwithstanding its exhaustion of the debt limitations of its negative covenants, those lenders accepted the characterization of the banks' interest as other than debt. See TIFD III-E, 342 F. Supp. 2d at 115-17.⁹

We find that the district court erred as a matter of law in several respects. In rejecting the government's contention that the Dutch banks were not bona fide equity partners for tax purposes, the court relied essentially upon the sham-transactions test to the exclusion of the test of totality-of-the-circumstances set forth by the Supreme Court in Commissioner v. Culbertson, 337 U.S. 733, 742 (1949). In examining the related question whether the banks' interest was more in the nature of debt or equity,¹⁰ the court erred in several respects -- primarily by accepting at face value the appearances and labels created by the partnership, rather than assessing the underlying economic realities. The court's conclusion was impaired by these errors of law.

We further find that the facts of the partnership agreement, as properly found by the district court, generally conceded by the taxpayer, and unquestionably established by the partnership documents, compel the conclusion that the Internal Revenue Service was on sound ground in rejecting the partnership's purported characterization of the Dutch banks' interest as bona fide equity participation.¹¹

In sum, the Dutch banks' interest was overwhelmingly in the nature of a secured lender's interest, which would neither be harmed by poor performance of the partnership nor significantly enhanced by extraordinary profits. The banks had no meaningful stake in the success or failure of Castle Harbour. While their interest was not totally devoid of indicia of an equity participation in a partnership, those indicia were either illusory or insignificant in the overall context of the banks' investment. The IRS appropriately rejected the equity characterization.

The IRS, however, is entitled in rejecting a taxpayer's characterization of an interest to rely on a test less favorable to the taxpayer, even when the interest has economic substance. This alternative test determines the nature of the interest based on a realistic appraisal of the totality of the circumstances. We do not mean to imply that it was error to consider the sham test, as the IRS purported to rely in part on that test. The error was in failing to test the banks' interest also under Culbertson after finding that the taxpayer's characterization survived the sham test.

In Culbertson, the Supreme Court ruled that a partnership exists when, "considering all the facts -- the agreement, the conduct of the parties in execution of its provisions, their statements, the testimony of disinterested persons, the relationship of the parties, their respective abilities and capital contributions, the actual control of income and the purposes for which it is used, and any other facts throwing light on their true intent -- the parties in good faith and acting with a business purpose intended to join together in the present conduct of the enterprise." Culbertson, 337 U.S. at 742. This test turns on the fair, objective characterization of the interest in question upon consideration of all the circumstances. See, e.g., Estate of Kahn v. Commissioner, 499 F.2d 1186, 1189 (2d Cir. 1974) (identifying factors a court might consider); Luna v. Commissioner, 42 T.C. 1067, 1077-78 (1964) (same). The IRS's challenge to the taxpayer's characterization is not foreclosed merely because the taxpayer can point to the existence of some business purpose or objective reality in addition to its tax-avoidance objective.¹³Cf. Gilbert v. Commissioner, 248 F.2d 349, 406 (2d Cir. 1957). The district court recognized the existence of the Culbertson test, see TIFD III-E, 342 F. Supp. 2d at 111, but did not conduct a Culbertson analysis of whether the banks' interest was a bona fide equity partnership participation. This was error.

Neither the Internal Revenue Code nor the Regulations provide for definitions of debt and equity. We have noted that Congress appears to have intended that "the significant factor" in differentiating between the two be whether "the funds were advanced with reasonable expectations of repayment regardless of the success of the venture or were placed at the risk of the business." Gilbert, 248 F.2d at 406; see also Nassau Lens Co. v. Commissioner, 308 F.2d 39, 47 (2d Cir. 1962) (Marshall, J.).

We recognize that the cases in which this standard has developed generally differ from the present case in two respects. First, they usually involve the characterization of an investment in a corporation, rather than in a partnership. See, e.g., John Kelley Co. v. Commissioner, 326 U.S. 521, 524-25 (1946) (analyzing whether certain payments to note-holders constituted "dividends" or "interest"); Matthiessen v. Commissioner, 194 F.2d 659, 661 (2d Cir. 1952) (analyzing whether advances to a corporation by stockholders were contributions to capital or loans). Second, conventionally in those cases, the taxpayer has characterized the investment as debt, with the objective of securing the tax benefit of the deductibility of interest payments, while the IRS has contended that because the investor will not receive either the recovery of the investment or the purported interest payments unless the venture succeeds, the investment is properly seen as a capital contribution. See, e.g., John Kelley Co., 326 U.S. at 524-25; Matthiessen, 194 F.2d at 660; see also Gilbert 248 F.2d at 402 ("Generally we find an effort by the taxpayer to induce the Commissioner and the courts to make a finding that [the relevant] transactions are loans.").

Given the facts and circumstances of the present case, we see these as distinctions without substantial difference. In all such cases, a taxpayer has cast a transaction representing an investment as equity or as debt with a view to obtaining tax benefits resulting from that characterization, and the government has challenged the characterization. We see no reason why the standard for distinguishing between debt and equity should not be focused in all such cases on whether "the funds were advanced with reasonable expectations of repayment regardless of the success of the venture or were placed at the risk of the business." Gilbert, 248 F.2d at 406; see also Hambuechen v. Commissioner, 43 T.C. 90, 99 (1964).

Our decision in O'Hare v. Commissioner, 641 F.2d 83 (2d Cir. 1981), is illustrative. In O'Hare, we found that the Tax Court had correctly determined that the taxpayer, who provided financial backing in securing a loan, had not become "sufficiently involved with the profitability of the venture to warrant treatment as a joint venturer," notwithstanding the fact that the taxpayer was not indemnified against certain liabilities or losses. Id. at 85. We affirmed the Tax Court's finding that upon a thorough analysis and appraisal of the circumstances, the taxpayer's investment was more in the nature of debt financing than a joint venture. Noting that any risks alleged by the taxpayer were "more imagined than real, and not materially different from those borne by other lenders," id., we found that, in sum, the "total effect" of the transaction in question "was that of a financing scheme rather than a joint venture," id. at 87.

It is true that the partnership agreement placed no formal upper limit on the banks' receipt of 98% of the Operating Income of the partnership. On the other hand, as a practical matter, the banks enjoyed only a narrowly circumscribed ability to participate in profits in excess of their Applicable Rate of return. The limitations of the banks' ability to share meaningfully in the profits of the partnership resulted, as noted above, from three provisions of the agreements. First, the taxpayer, which held the full power to manage the partnership, had the right, by transferring productive assets to the CHLI subsidiary of the partnership, to reclassify the income produced by those assets from the category of Operating Income (in which the banks would take 98%) to Disposition Gains (in which the banks' share was 1%, over and above approximately $2.85 million).¹⁴ Second, the taxpayer could reduce drastically the net Operating Income, in which the banks took a 98% share, by redepreciating the already fully depreciated aircraft, which had the effect of transferring the revenue covered by the depreciation from the banks to the taxpayer. Finally, the taxpayer could at any time, and at negligible cost, terminate the partnership.

As a practical matter, therefore, the Dutch banks' opportunity to participate in unexpected and extraordinary profits (beyond the reimbursement of their investment at the Applicable Rate of return) was capped at $2.85 million, plus 1%. Full realization of the $2.85 million potential, on an investment of $117.5 million, would have increased the banks' total return by less by 2.5% -- a relatively insignificant incremental return over the projected eight- year life of the partnership. In short, the district court's conclusion that the banks enjoyed a meaningful and unlimited share of the partnership's upside potential was erroneously based on the fictions established by the partnership documents, and did not take account of the practical reality of the taxpayer's powers to restrict the banks' share of profits. See, e.g., ASA Investerings, 201 F.3d at 514-15 (dismissing possibility that the interest of a purported equity participant would be meaningfully affected by another participant's failure to act in its own economic self-interest); O'Hare, 641 F.2d at 86 (same).

It is not entirely clear what the court meant by this analysis. The court could have meant either that (i) any potential share in profits results in an obligation to pay something other than a sum certain, and therefore tends to indicate an equity interest, rather than debt; or (ii) that a substantial share in potential profits, such as represented by the banks' 98% share of the Operating Income, indicates an equity interest rather than debt, even though downside risk may be eliminated. We think the district court probably meant the latter, both because it emphasized the 98% share and spoke of an investor with unlimited upside potential, and because the court repeatedly asserted that trivial interests should play no role in its analysis. Either meaning would be mistaken, although for different reasons.

If the court intended the first meaning (which we think unlikely) -- that any deviation from a sum certain, no matter how trivial, argues against a finding of debt -- this would misunderstand the basic purpose of the analysis. The purpose of the analysis is to bypass appearances and focus instead on practical realities. See, e.g., John Kelley Co., 326 U.S. at 530; Gilbert, 248 F.2d at 402; Gregg Co. of Delaware v. Commissioner, 239 F.2d 498, 502 (2d Cir. 1956); Dixie Dairies Corp. v. Commissioner, 74 T.C. 476, 493-98 (1980); Hambuechen, 43 T.C. at 98; see also ASA Investerings, 201 F.3d at 514-15 (dismissing "de minimis" risk as irrelevant). While an obligation to pay a sum certain indicates debt, it does not follow that any insignificant deviation from a sum certain indicates equity. The purpose of the test is to determine as a practical matter whether the interest created is more akin to equity or debt. Thus, the closer the amount owed comes to being a sum certain, the more it would tend to indicate debt. Trivial or insignificant deviations from a sum certain would do little to argue against a finding of debt. See, e.g., Gilbert, 248 F.2d at 402 (noting that "some variation" from the "classic debt" formula does not necessitate treatment as equity); Gregg Co., 239 F.2d at 499 (indicating that the same analysis applies even when an advance has "some of the characteristics of a conventional debt issue and some of the characteristics of a conventional equity issue").

If the court meant the latter -- that the banks' unlimited upside potential, represented by their 98% share in the partnership's Operating Income, indicated a meaningful stake in the profits of the venture and thus argued against a finding of debt -- this was error for the reasons discussed above: It failed to take account of the power of the taxpayer to direct the profits to itself by various devices, rather than to the banks.

The banks had essentially bargained for and received a secure guaranty of the reimbursement of their investment at the agreed Applicable Rate of return. Their apparent 98% share of partnership income was largely defeasible by the taxpayer, and was more in the nature of window dressing designed to give ostensible support to the characterization of equity participation, which was essential to the dominant tax objective, than a meaningful stake in the profits of the venture. The possibility of a small share in extraordinary profits was not a significant feature of their investment. While the amount owed to the banks was not exactly a "sum certain," it was not significantly different; it was more akin to the characteristic repayment of debt than to a real equity stake in the venture.

It is true that no partnership document overtly provided that repayment of the banks would take priority over payment of general creditors of the partnership. If the inquiry were limited to disposition of assets of Castle Harbour, we see no reason to doubt that general creditors would take prior to the banks. In fact, however, the repayment of the banks' interest was even more securely protected than by priority over the general creditors. The banks did not need to rely on assets of Castle Harbour or the taxpayer for reimbursement. They had the guaranty of the taxpayer's far more solvent parent GECC. Upon consideration of all the facts and circumstances, it is clear that, far from being subordinate to the general creditors, the Dutch banks were secured in such manner that they would be repaid in full with interest from a source to which the general creditors had no access. The apparent subordination found by the district court was a fiction overridden by GECC's guaranty.

The issue raised by the factor, however, is not whether the Dutch banks had a right to force a buyout of their share, giving account to the profits they had gained and losses they had suffered during their participation. It is rather whether they had the right to force the payment of what was effectively their principal and interest. They did. The partnership agreements gave the banks the power to terminate and receive the reimbursement of their $117.5 million investment at the agreed annual rate of return. As the district court elsewhere recognized, the banks were protected against any meaningful diminution of such repayment. And to the extent they were entitled to a share of profits over and above the return of their principal plus interest, this was, as explained above, a negligible interest (which, furthermore, did not detract from the fact that they were empowered to force the payment of their invested principal and interest). The position of the Dutch banks was thus very different from an ordinary equity partner's ability to force liquidation of a partnership. It was error for the district court to treat this factor as one without significance in the debt/equity analysis. See, e.g., Estate of Mixon v. United States, 464 F.2d 394, 405 (5th Cir. 1972) ("If there is a definite obligation to repay the advance, the transaction [will] take on some indicia of a loan."); Rosenberg v. Commissioner, 79 T.C.M. (CCH) 1769 (2000) ("A taxpayer's right to enforce repayment of an advance suggests that the advance is a loan.").

The explanation is inapposite. In the first place, the Dutch banks were not like the "average stockholder" as they had invested nearly 20% of the partnership's assets. Furthermore, if the average stockholder of a publicly traded corporation has no practical influence on management, that is not because he possesses no management rights, but rather because his voting right is small in relation to the total electorate. Every share of stock in the typical publicly traded corporation carries with it one vote for the directors of the corporation. The banks, in contrast, although they had contributed a sizeable share of the partnership's capital, exercised no such vote. The failure to exercise management rights is certainly not conclusive. A holder of a bona fide minority equity interest in a partnership or corporation may well have no practical ability to influence management and may even have no vote as a formal matter. Nonetheless, it seems clear that the denial of participation in management to an investor is a factor which tends, even if only slightly, to favor a conclusion that the interest resembles debt. See, e.g., Estate of Mixon, 464 F.2d at 406 (noting that lack of "increased voting power or participation in the Bank's affairs by virtue of the advance . . . serve[d] as cumulative evidence that the advances here involved were loans, rather than risk capital").

The taxpayer's $60 million tax objective depended on successfully characterizing the interest of the Dutch banks as an equity partnership participation. There could be no conceivable doubt that the taxpayer had a vital interest in the acceptance by the IRS and the courts of the banks' participation as equity, and had taken pains in the design of the partnership to promote that characterization. The main issue in the litigation was whether that characterization was entitled to be respected, or was so overwhelmed by debt-like features, and by the banks' lack of meaningful stake in the partnership venture, as to compel a conclusion different from that advocated by the taxpayer. In these circumstances, the fact that the taxpayer, in accordance with its strong self-interest, consistently described the banks' interest as equity would seem to be of only slight probative value. As the tax court has noted, a taxpayer's "self-serving statements that the advances were intended to be loans bear little weight in the face of the other facts of record." Hambuechen, 43 T.C. at 104 (internal quotation marks omitted). In contrast, the Dutch banks treated their interest as debt both for financial accounting and Dutch taxation. The fact that the Dutch banks repeatedly contradicted the taxpayer's characterization would seem entitled in logic to at least as much weight as the taxpayer's highly self-interested characterization. Yet the district court appears to have attached no significance to the banks' description of their interest as debt, and to have accorded unquestioning credit to the taxpayer's self-interested characterization. In our view, the district court's analysis of this factor cannot be sustained. At best, factor (g) is equivocal.¹⁶

First, as we explained in Gilbert, one factor indicative of debt is "a fixed percentage in interest payable [to the creditor] regardless of the debtor's income or lack thereof." 248 F.2d at 402. See also Nassau Lens Co., 308 F.2d at 47; Jewel Tea Co. v. United States, 90 F.2d 451, 453 (2d Cir. 1937) (L. Hand, J.). As discussed above, the mechanisms of the partnership agreements ensured that the Dutch banks would receive precisely this: an annual return at the Applicable Rate, regardless whether Castle Harbour was experiencing profits or losses.

Another factor identified by courts as indicative of debt is the "reasonableness of expectation of payment." Hambuechen, 43 T.C. at 99; see also Nassau Lens Co., 308 F.2d at 47. As explained above, features of the Castle Harbour agreements combined to provide the Dutch banks with not only a reasonable expectation, but an ironclad assurance that they would receive repayment of their principal at the Applicable Rate of return, regardless of the success of the Castle Harbour venture. These features included (a) the Exhibit E payment schedules; (b) the Investment Accounts; (c) the Class A Guaranteed Payments; (d) the requirement for the benefit of the Dutch banks that CHLI maintain Core Financial Assets of 110% of the obligation owed to the Dutch banks; (e) the banks' ability to liquidate the partnership in certain circumstances and receive reimbursement at the Applicable Rate of return; (f) the $300 million worth of casualty loss insurance, which was obtained by Castle Harbour for the benefit of the Dutch banks; and, most importantly, (g) GECC's personal guaranty of the obligations owed by the partnership to the Dutch banks.

Finally, courts have noted that, in determining whether an interest is more akin to debt or equity, it is relevant to consider the "use to which the [invested] funds were put" and "whether payment [to the purported creditor] can only be paid out of future profits." Hambuechen, 43 T.C. at 99. For the Dutch banks' benefit, the Operating Agreement required that CHLI maintain 110% of the Dutch banks' Investment Accounts in Core Financial Assets, thereby precluding the partnership from using the banks' investment in the partnership's aircraft-leasing business. TIFD III-E, 342 F. Supp. 2d at 110. With regard to whether payment "can be made only out of profits," as noted above, the repayment of the banks was assured even in the event the partnership suffered catastrophic loss.

Each of these factors, which the district court failed to consider, would have tended to show that the banks invested "with reasonable expectations of repayment regardless of the success of the venture," and were not meaningfully "at the risk of the business." Gilbert, 248 F.2d at 406.

B. Whether the IRS Properly Determined that the Interest of the Banks Was Not a Bona Fide Equity Participation.

The ultimate question is whether the IRS properly determined in its FPAAs that the interest of the banks was not a bona fide equity participation. The errors of law made by the district court in considering this question invalidate its conclusion. We are accordingly compelled to vacate the judgment. The question remains whether we should remand for new findings or undertake ourselves to answer in the first instance the question of bona fide equity participation. We conclude that consideration of this question under Culbertson's mandate to appraise the totality of the circumstances compels the conclusion that, for tax purposes, the banks were not bona fide equity partners in Castle Harbour. Accordingly, there is no reason to remand for new findings.

The transaction consisted, as a practical matter, of an advance by the Dutch banks of $117.5 million. The partnership undertook to repay the advance at an agreed rate of return, pursuant to a previously agreed payment schedule. In the event of a missed payment, the banks had the power to liquidate the partnership and to receive the return of their principal at the agreed rate of return. These payments were to be made regardless of the fortunes of the partnership. Furthermore, because the payments were guaranteed by GECC, the banks were secured in the receipt of the payments, regardless of whether the partnership profited or lost money. Because of this guaranty, the banks had effective priority over the general creditors of the partnership. In all these respects, the banks were, for all intents and purposes, secured creditors. It is true that the banks' interest took on some aspect of equity participation in that they ran a small risk of a shortfall in event of catastrophic loss -- a risk so small that the district court disregarded it¹⁷ -- and that they were given in addition a participation in extraordinary unforseen profits of the partnership. As for participation in unexpected profits, for the reasons explained above, the taxpayer was as a practical matter empowered to limit the banks' participation to an amount that was insignificant in the context of the banks' $117.5 million eight-year investment.¹⁸ This was not a significant participation in the profits of the partnership. See, e.g., Gilbert, 248 F.2d at 402; O'Hare, 641 F.2d at 85; see also ASA Investerings, 201 F.3d at 514-15 (acknowledging that the taxpayer failed to argue that the relevant payments received were "related to the success of the partnership's investments," dismissing the relevance of "de minimis" investment risks, and noting that "[a] partner whose risks are all insured at the expense of another partner hardly fits within the traditional notion of partnership").

We recognize that if the Dutch banks had a sufficiently sizable share in the profit potential of the partnership, they might appropriately be deemed equity participants for tax purposes, notwithstanding the guaranteed repayment of their initial investment at an agreed rate of return. See Hunt v. Comm'r, 59 T.C.M. (CCH) 635 (1990) (noting that a bona fide equity interest can exist even where a minimum return is guaranteed). On different facts a difficult question would arise whether an investor's right to a share of profits was sufficient to make its interest a bona fide equity participation for tax purposes notwithstanding the secured guaranty of the return of its principal plus interest. This is not such a case. Here the banks were accorded the appearance of a meaningful interest in the potential profits of the partnership, which was effectively nullified by the taxpayer's ability to limit their participation in such profits to an insignificant amount.

In following the approach of Culbertson, we are compelled to look not so much at the labels used by the partnership but at true facts and circumstances. The Dutch banks' interest was in the nature of a secured loan, with an insignificant equity kicker. Only in a negligible fashion was their well-secured interest intertwined with the fortunes of the business. The facts and circumstances presented, considered in their totality, compel the conclusion that the Dutch banks' interest was, for tax purposes, not a bona fide equity participation.

² A part was used to buy out the interest of a GECC subsidiary, and a part was contributed directly.

³ As the district court explained,

⁴ The so-called Ceiling Rule, derived from I.R.C. § 704(c), which has since been repealed, was the statutory provision upon which the partnership relied to justify its allocations.

⁵ The district court found a small risk that the Dutch banks would receive less than the amount in their Investment Accounts, insofar as the Dutch banks' return could be reduced by a third-tier allocation of losses (i.e., by the 1% of losses that would eventually be allocated to the Dutch banks if those losses were large enough). TIFD IIIE, 342 F. Supp. 2d at 105-06. However, the district court found the likelihood of loss to be so remote as to be irrelevant. Id.; see also id. at 116- 17.

⁶ Disposition Gains and Losses included the results of the disposition of any asset as well as the operating revenues and expenses produced by any asset held by CHLI.

⁷ The Disposition Gains and Losses were allocated in the following manner. First, Disposition Gains were allocated to offset (i) prior Disposition Losses and (ii) prior Operating Losses. Disposition Losses were first allocated to offset (i) prior Disposition Gains, but not (ii) prior Operating Gains. Second, Disposition Gains and Losses were allocated 90% to the Dutch banks and 10% to the taxpayer until the Dutch banks had been allocated $2,854,493. Finally, the Operating Agreement allocated 99% of the gain or loss to the taxpayer and 1% to the Dutch banks. After approximately $541 million had been allocated, the allocation regime switched. However, as the district court found, "these allocations never came into play, were highly unlikely to ever come into play, and, accordingly, are not relevant." TIFD III-E, 342 F. Supp. 2d at 101 n.16.

⁸ The taxpayer ended up terminating the Dutch banks' participation and paying this premium on December 31, 1998, three years earlier than anticipated. This buyout coincided with a change in U.S. tax law that "made it possible that Castle Harbour would no longer be treated as a partnership" -- a change that threatened to trigger a tax indemnification whereby the taxpayer would assume liability for any U.S. taxes assessed against the banks. TIFD III- E, 342 F. Supp. 2d at 107 n.30.

⁹ The district court rejected additional government arguments not raised in this appeal and others that we have no need to address.

¹⁰ It is true that in conducting its debt/equity analysis, the court looked at facts and circumstances. This, however, did not cure the failure to consider the question under the Culbertson test, because the court observed that "the only possible relevance of the debt/equity analysis is as an aid to performing 'sham transaction' analysis," TIFD III-E, 342 F. Supp. 2d at 115, and because, in conducting the debt/equity analysis, the court made the errors of law described supra.

¹¹ Our disposition makes it unnecessary for us to consider whether the district court correctly determined that the characterization of the banks' interest as equity was not a sham.

¹² Although the district court explained that its decision to rely upon the sham-transaction doctrine stemmed from the broad definition of partnership found in I.R.C. § 761, this provision of the Code does not supercede or otherwise alter the analysis necessary under Culbertson. See Culbertson, 337 U.S. at 751 n.1 (Frankfurter, J., concurring) (quoting I.R.C. § 3797, the direct precursor to § 761).

¹³ It is true that the Court of Appeals in ASA Investerings noted that the Tax Court's decision "rejecting the bona fides of the partnership was the equivalent of a finding that it was, for tax purposes, a 'sham.'" ASA Investerings, 201 F.3d at 512 (emphasis omitted). We do not believe that this language should be interpreted as suggesting that the two underlying inquiries are identical. While a classification that fails the sham test may be certain also to fail the Culbertson analysis, a classification that passes the sham test would not necessarily survive Culbertson.

¹⁴ The district court, recognizing that the management of the partnership, in relevant part, remained with the taxpayer, TIFD III-E., 342 F. Supp. 2d at 104, found that the partnership 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 [the subsidiary]," id. at 103. As an illustration, the district court pointed to aircraft held in CHLI, rather than in Castle Harbour, and noted that the rental income from those aircraft were credited to the banks only as Disposition Gain, and not as Operating Income. See id. at 103-04 ("CHLI purchased several aircraft during Castle Harbour's existence. 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." (citations omitted)).

On appeal, the taxpayer attacks this finding of the district court, contending it is clear error, unsupported by the record. It argues that the general manager of Castle Harbour "acted independently and not at the direction of GECC," that the Operating Agreement "contemplated" that aircraft contributed to Castle Harbour would be held directly by Castle Harbour, and that the general manager did not regard such a transfer as one of his options, in part because it would have been "contrary to the legitimate business expectations of [the Dutch banks]," and in part because such a transfer would have been inconsistent with a so-called "representation" GECC had made to its auditor. In support of the latter claim, the taxpayer's brief points to a document in which GECC's tax counsel, soon after the formation of the partnership, provided the auditor with "a list of assumptions you may rely on in developing your position regarding the tax structure," which included that "assets transferred to LLC sub by LLC will consist of only cash and cash equivalents whose adjusted tax basis is equal to fair market value."

We consider the taxpayer's argument that the district court's finding was clear error to be frivolous. In most respects, the taxpayer's contentions are not supported by its citations to the record. As for the argument that such a transfer would not have been made because it would have been inconsistent with legitimate expectations of the banks, it is illogical. There is no reason to think that the "legitimate expectations" of the banks were not shaped by their agreement, which granted the powers in question to the taxpayer. Indeed, as the district court expressly found, the taxpayer did make use of the CHLI device, limiting the banks' participation in certain aircraft rentals to their share of Disposition Gain, rather than their share of Operating Income. And as for the so-called "representation" to the auditor, in the first place it was no more than a working assumption. Moreover, even assuming that GECC did not anticipate at the outset making such transfers, so far as the record reflects, if extraordinary unexpected profit potential in an asset made its transfer to CHLI desirable, the earlier statement of assumption to the auditor would not have prevented it.

¹⁵ The factors are:

¹⁶ Factor (h) asks "whether the instruments are intended to be treated as debt or equity for non-tax purposes." I.R.S. Notice 94-47. It is true that GECC intended that its lenders treat the Dutch banks' interest as equity for non-tax purposes, including GECC's contractual debt limitations. We acknowledge that factor (h) favors the taxpayer. It appears to be the only factor listed in Notice 94-47 that favors the taxpayer.

¹⁷See supra note 5.

¹⁸ An additional return of $2.85 million over eight years would have increased the banks' rate of return by approximately one-third of one percent.

¹⁹ We have not considered the taxpayer's argument that the partnership was a family partnership under the provisions of I.R.C. § 704(e). We leave this question for consideration in the first instance by the district court.