Bristol-Meyers' Disappearing Gain

III. Bristol Transaction

A. Offshoring Built-In Gain

Like other pharmaceutical companies, Bristol licenses rights to use and manufacture its most lucrative drugs to related foreign entities located in low-tax jurisdictions, allowing shifting of profits from the U.S. to foreign jurisdictions such as Ireland. Prior to 2012, Bristol (and its consolidated subsidiaries) held the worldwide intellectual property rights (the “licensed intangibles”) to three successful drugs. Using a series of licensing agreements, Bristol granted exclusive rights to sell and manufacture the licensed intangibles (in exchange for a stream of royalty payments) to an entity (“Holdings”) domiciled in Ireland and treated as a corporation for U.S. tax purposes.

In 2012, Bristol reorganized its European operations, ostensibly for the purpose of “better aligning the geographical and operational focus” of its business. Pursuant to the reorganization, Bristol’s wholly-owned foreign affiliate (which owned 100% of the equity interests in Holdings) contributed its entire interest in Holdings to ForeignCo, an Irish corporation which indirectly was wholly owned by Bristol. Holdings then elected to be treated as a disregarded entity for U.S. tax purposes, with the result that ForeignCo was treated as directly owning Holdings’ assets. Bristol then transferred its non-U.S. rights (i.e., the rest of the worldwide rights) in the licensed intangibles to Holdings, which thereby sprang into being as a partnership for U.S. tax purposes with Bristol and ForeignCo as its two partners. The following simplified diagram illustrates the post-reorganization structure for holding Bristol’s intangibles:

Consultants valued Bristol’s contributed property (the “contributed intangibles”) at approximately $4.6 billion and ForeignCo’s contributed property at approximately $27.6 billion; the partners’ interests in the partnership (with total assets of $32.2 billion) were 14.27% and 85.73%, respectively. Bristol’s transferred assets had a zero tax basis (and a built-in gain of $4.6 billion). By contrast, ForeignCo’s transferred assets were either (1) nondepreciable and nonamortizable or (2) depreciable or amortizable but with a tax basis equal to fair market value. Over one-half of ForeignCo’s contribution consisted of a zero-basis nonamortizable intangible (worth $15.3 billion) labelled by the parties as “goodwill,” presumably self-created. Since the zero-basis goodwill was nonamortizable in the partnership’s hands, it did not give rise to any disparity in tax and book amortization.

The partners shared the partnership’s expected revenue of $32.2 billion based on their percentage interests. To account for the disparity between the basis and value of the contributed assets, the partnership elected to use the traditional method under section 704(c). The choice of the traditional method, coupled with the wasting nature of the contributed intangibles and the large spread between basis and value, set the stage for a pre-arranged shift of built-in gain as the contributed intangibles were exhausted in the partnership’s business. The shortfall in the amortizable basis of the contributed intangibles ($4.6 billion) meant that, over the relevant amortization period, $3.9 billion of Bristol’s built-in gain (85.73% × $4.6 billion) would inexorably shift to ForeignCo. The shift was virtually ironclad, provided the partnership continued to own and use the property—a decision entirely within Bristol’s control. In effect, Bristol took advantage of ForeignCo’s zero-tax rate to reduce its own U.S. tax liability. The transaction would never have occurred between unrelated parties because the noncontributing partner would have demanded compensation from Bristol for absorbing the shifted taxable income in excess of economic income.

Bristol’s contributed intangibles were not expected to produce any tax depreciation or amortization, while ForeignCo’s contributed property would predictably produce significant amounts of tax depreciation. In accordance with section 704(b), the partnership amortized Bristol’s contributed intangibles over their remaining useful life for book purposes. During the first short year of operation, the partnership allocated its book (economic) income ($333,667) to the partners in accordance with their percentage interests. Taxable income was allocated in the same percentages as book income, but the partnership’s taxable income substantially exceeded its book income. In the case of Bristol’s contributed intangibles, book amortization exceeded tax amortization by $116 million; ForeignCo was allocated book amortization of $99 million (and zero tax amortization), creating a disparity of $99 million between ForeignCo’s book and tax amortization. The remaining $17 million of book amortization (and zero tax amortization) was allocated to Bristol. Thus, under the traditional method, Bristol shifted $99 million of built-in gain ($116 million ×.8573) to ForeignCo during the partnership’s first short year. By contrast, ForeignCo’s depreciable contributed property had sufficient tax basis to ensure that Bristol would be allocated equal amounts of book and tax depreciation; since ForeignCo would bear the entire burden of any shortfall in tax depreciation, no built-in gain was shifted from ForeignCo’s depreciable property to Bristol.

To illustrate the built-in gain shift over the partnership’s life, assume that Bristol contributes amortizable intangibles with a basis of zero and a value of $4.6 billion and ForeignCo contributes nondepreciable, nonamortizable property with a basis and fair market value of $27.6 billion. The partnership uses the traditional method under section 704(c) and maintains tax and book capital accounts in accordance with the section 704(b) Regulations. Prior to liquidation, the partnership has operating income of $32.2 billion, equal to the discounted present value of the partners’ respective contributions. The partnership allocates the operating income based on the partners’ percentage interests in the partnership and distributes all of its operating income currently. Bristol’s zero-basis contributed intangibles are assumed to have no value at the end of the amortization period. Book amortization of the contributed intangibles (based on their fair market value) reduces the partners’ book capital accounts in proportion to their percentage interests, reducing the amount each partner receives upon liquidation. Immediately prior to liquidation, the partners would have the following tax and book capital accounts:

ForeignCo’s book capital account would be reduced by its share of the partnership’s book amortization ($3.9 billion) attributable to Bristol’s contributed intangibles, and Bristol’s book capital account would be reduced by the remaining book amortization ($.7 billion). Upon liquidation, Bristol would recognize a gain of $3.9 billion, equal to the excess of the amount distributed over its outside basis (zero). ForeignCo would recognize a mirror loss of $3.9 billion equal to the excess of its outside basis ($27.6 billion) over the amount distributed ($23.7 billion). Thus, Bristol shifted $3.9 billion of built-in gain to ForeignCo equal to ForeignCo’s share of book amortization ($4.6 billion ×.8573) in excess of zero tax amortization, while the remaining book amortization of $.7 billion ($4.6 ×.1427) reduced the amount received by Bristol upon liquidation.

The shift of built-in gain from Bristol to ForeignCo was reasonably expected to occur over the remaining useful life of the contributed intangibles. According to representations in the outside adviser’s tax opinion, the partnership had no intention of transferring or assigning the contributed intangibles to a third party, nor did Bristol have any intention of selling its partnership interest. Although Bristol’s built-in gain was merely deferred, Bristol controlled whether the partnership would eventually be liquidated. Under certain circumstances, the partnership agreement provided for curative allocations to offset the effect of the ceiling rule limitation, but these curative allocations were illusory.

Under the partnership agreement, a curative allocation could be triggered only upon sale of Bristol’s contributed intangibles. Since a sale would terminate income-shifting, it would be economically unattractive from Bristol’s perspective. Given the wasting nature of Bristol’s contributed intangibles, any gain upon sale would likely be minimal. Moreover, a curative allocation of gain upon sale of a particular intangible could only be used to reverse the ceiling rule limitation with respect to that specific property. For purposes of section 704(c), the curative allocation provision created the illusion that future gain on sale might reverse the shifted built-in gain of $3.9 billion. Bristol’s financial reporting, however, required minimizing any realistic possibility of such a future reversal that would undercut Bristol’s position concerning indefinite deferral of the shifted built-in gain.

B. Specific Anti-Abuse Rule

The government has asserted that the Bristol transaction violated the specific anti-abuse rule of section 704(c). Under the anti-abuse rule, an allocation method (or combination of methods) is not reasonable if the contribution of property (or the revaluation event) and the corresponding allocation of tax items with respect to the section 704(c) property “are made with a view to shifting the tax consequences of built-in gain or loss among the partners in a manner that substantially reduces the present value of the partners’ aggregate tax liability.” As illustrated by two examples under the section 704(c) Regulations, the anti-abuse rule clearly encompasses attempts to exploit ceiling rule limitations using the traditional method or the traditional method with curative allocations.

In the traditional method example, C contributes equipment with a fair market value of $10,000 and a basis of $1,000, and D contributes cash of $10,000 in exchange for equal partnership interests. The contributions are made “with a view to taking advantage” of the fact that the equipment has a short remaining recovery period (one year) but a “significantly longer” economic life. D has a substantial net operating loss carryforward which will otherwise expire unused. The partnership sells the equipment during the second year for $10,000, when its tax and book basis have both been reduced to zero, and the partnership allocates tax and book gain equally to the partners ($5,000 each). The interaction between the property’s short book-recovery period and the traditional method allows C to shift built-in gain of $4,000 to D (who is allocated tax gain of $5,000 on sale but receives tax depreciation of only $1,000); D’s share of tax gain is offset by D’s net operating losses. Thus, the arrangement shifts a significant amount of taxable income to low-bracket D (and away from high bracket C). Under these circumstances, use of the traditional method is unreasonable, since the contributions and allocations were made with a view to reducing the partners’ aggregate tax liability. Nevertheless, the Regulations state that use of the traditional method would not have been abusive if the partnership agreement had provided that any tax gain from the sale of the property would be allocated to the contributing partner “to offset the effect of the ceiling rule limitation.”

In the curative allocation example, the facts are similar except that the partner contributing the depreciable equipment (J) has expiring net operating losses, while the cash partner (K) is in a high bracket. The partnership does not sell the equipment; instead, it invests the cash in inventory which is sold for an $8,000 book and tax profit at the end of the first year. To remedy the $4,000 shortfall in K’s share of tax depreciation attributable to J’s equipment ($5,000 book depreciation less $1,000 tax depreciation), the partnership makes a curative allocation of $4,000 of income from the sale of the inventory to J (and away from K); the entire $8,000 of gain on sale of the inventory is allocated to J, but such gain is offset by J’s expiring net operating losses. The use of the curative allocation method is unreasonable because the contributions and allocations were made with a view to shifting a significant amount of taxable income away from the high bracket partner (K) to the loss partner (J) over a period of time significantly shorter than the equipment’s economic life. As clarified in the final section 704(c) Regulations, the curative allocations would nevertheless be reasonable if stretched over a longer period of time, such as the property’s remaining economic life (rather than its short one-year recovery period).

In each example, a section 704(c) method that would otherwise be valid is found to be unreasonable because of the mismatch between the contributed property’s short one-year recovery period and longer useful life. Given the partners’ differing tax profiles, the choice of section 704(c) methods gives rise to unreasonable tax results. The Regulations come close to indicating that curative allocations spread over the true economic life of the property will always be considered reasonable when used to remedy a shortfall in depreciation or other cost recovery deductions. Nevertheless, the Treasury intentionally declined to provide any anti-abuse safe harbors, apart from indicating that the traditional method is reasonable if used for all contributed property when no ceiling rule limitations arise. While the lack of safe harbors has prompted taxpayer concerns about the potentially expansive scope of the anti-abuse rule, one commentator has concluded that “ceiling rule shifts will be tolerated, unless taxpayers manipulate the timing of the ceiling rule shifts” or attempt to cure ceiling rule limitations with curative allocations occurring “over a period substantially shorter than the contributed property’s economic life.” Under this view, the traditional method (or traditional method with curative allocations) will be considered “reasonable in all but the most unusual circumstances.”

Interpreting the anti-abuse rule narrowly, Bristol may claim that use of the traditional method with curative allocations was not unreasonable. Since there was no mismatch between the remaining amortization period for the contributed intangibles and their remaining economic life, the transaction can be distinguished from the examples illustrating the anti-abuse rule. Indeed, Bristol may point to the curative allocation provision in the partnership agreement as coming within a putative safe harbor, since curative allocations (if any) would not occur over a shorter period than each contributed intangible’s remaining economic life. Finally, Bristol may assert that the anti-abuse rule is not violated because no built-in gain can be shifted to ForeignCo in the absence of a sale (or other disposition) of the partnership’s assets, which has not occurred (and may never occur). From Bristol’s perspective, all that was needed to exploit the ceiling rule was to defer a sale indefinitely, while shifting operating income to a tax-indifferent partner as the partnership used the contributed intangibles in its business. While the tax opinions from the outside adviser and law firm offered a potential penalty shield if the larger transaction were challenged, Bristol was apparently unable to obtain a tax opinion on the specific anti-abuse rule under section 704(c).

In the case of wasting assets used in a partnership’s business, the shift in taxable income is economically realized as the assets generate income, whether or not the assets are eventually sold. A contribution of property to a partnership is generally accorded nonrecognition treatment under section 721, with gain deferred until a subsequent triggering event. Such a tax-free contribution is nevertheless an economic exchange of the contributing partner’s property for cash or other property contributed by the other partners. Under the traditional method, the exchange of properties locked in the future shift of built-in gain from Bristol to ForeignCo even though the tax consequences of the shift were deferred until income was realized from business operations. The partners did not realistically contemplate a reversal of the shifted built-in gain. The partnership’s curative allocation provision was negated by Bristol’s representations and by Bristol’s financial reporting, which treated the shifted gain as permanently deferred. Even if reasonable curative allocations may insulate a ceiling rule shift, the partnership’s curative allocation provision lacked any substance.

Under the traditional method, a known mismatch between tax and economic depreciation may accelerate a shift in taxable income, as the two anti-abuse examples illustrate. The anti-abuse rule should not be interpreted narrowly, however, as simply a backstop to uneconomic cost recovery rules that are particularly susceptible of manipulation. In crafting anti-abuse rules, the regulation drafters cannot be expected to foresee all possible opportunities for abuse. Instead, such rules are drafted in light of particular types of transactions known to the drafters when the regulations are promulgated. As planning evolves, it may be appropriate to apply specific anti-abuse rules to novel transactions that were not necessarily contemplated by the drafters. The absence of safe harbors enhances the efficacy of anti-abuse rules in preventing aggressive planning from frustrating the intended effects of a highly-complex statutory scheme. By contributing low-basis, high-value wasting assets that would be exhausted in the partnership’s business, Bristol shifted taxable built-in gain in the contributed intangibles to a related tax-indifferent party. Moreover, the limited curative allocation provision virtually ensured that, even in the unlikely event of a sale, the shifted built-in gain would not be reversed. Even if the anti-abuse rule is limited to rare circumstances, the Bristol transaction provides an apt illustration of the need for such a rule.

The anti-abuse rule employs present value concepts to determine whether a reduction in the partners’ aggregate tax liability is substantial. While quantifying the present value of Bristol’s tax benefit from shifting built-in gain to ForeignCo depends on various assumptions, the tax savings would appear to be substantial under any plausible scenario. Each dollar of shifted income potentially saved Bristol taxes equal to the difference between the 35% marginal U.S. tax rate (under prior law) and ForeignCo’s zero marginal rate. The shift in built-in gain might never be corrected because Bristol could indefinitely defer liquidation of the partnership (or sale of its partnership interest). Bristol’s financial statements further evidenced its intent that the shifted built-in gain would never be subject to U.S. taxes. Regardless of whether “substantial” is defined in relative or absolute terms, the transaction clearly produced substantial tax savings in present value terms.

Under the anti-abuse rule, the contributions and allocations must be undertaken “with a view” to the proscribed purpose of substantially reducing the partners’ aggregate tax liabilities. Although the section 704(c) Regulations do not provide specific guidance, the “view” requirement establishes a significantly lower threshold than the “principal purpose” standard under the general partnership anti-abuse rule of section 701. In the analogous context of the collapsible corporation provisions of former section 341, the Regulations provided that the requisite view existed if the proscribed activity “was contemplated, unconditionally, conditionally, or as a recognized possibility.” When sophisticated parties structure a transaction to achieve the prohibited result, the requisite view may exist even if the taxpayer has other valid business purposes. Since the parties were all related and the amortization schedule of the intangibles was known in advance, Bristol was in a position to appreciate that the traditional method would foreseeably shift built-in gain to a zero-rate taxpayer. The substantial tax savings should also have been evident to Bristol’s outside advisers who opined on other aspects of the restructuring and partnership formation. Since the transaction was structured in such a manner as to facilitate the income shift which actually occurred, Bristol should be considered to have acted with the proscribed view, even if the restructuring may also have served other non-tax business purposes.

C. Government’s Proposed Remedy

If the Bristol transaction is determined to have violated the anti-abuse rule, the government has a choice of remedies. Under the anti-abuse rule, the government may exercise its authority to require curative allocations to remedy the ceiling rule distortion but cannot force the partnership to use the remedial method. Over the life of the partnership, ForeignCo was allocated taxable income, including operating income from Bristol’s contributed intangibles, in excess of the items limited by the ceiling rule. The government has proposed to reallocate taxable income to Bristol equal to the shortfall in ForeignCo’s tax amortization attributable to the contributed built-in gain property. The proposed reallocation of $3.9 billion of taxable income away from ForeignCo would suffice to fully remedy the shift in built-in gain:

Since the government’s proposed reallocation completely eliminates any book/tax disparities, neither partner would recognize gain or loss upon liquidation.

In effect, the proposed reallocation would force Bristol to recognize additional ordinary income of $3.9 billion from partnership operations, rather than deferring such income until liquidation and converting it into capital gain. Bristol would be treated as having sold a built-in gain asset (with a basis of zero and a fair market value of $3.9 billion) to ForeignCo; the reallocation of taxable income away from ForeignCo would remedy the shortfall in ForeignCo’s tax amortization deductions. In addition, Bristol would continue to recognize its percentage share of the partnership’s operating income, leaving Bristol with total ordinary income of $8.5 billion ($4.7 billion plus $3.9 billion reallocation). As required under the section 704(c) Regulations, the proposed reallocation would generally be of the same type, and have substantially the same tax effect, as the ceiling rule limited items. Under the general anti-abuse rule of section 701, the government could also seek to disregard Bristol’s (or ForeignCo’s) contributed assets.

The section 704(c) Regulations do not prescribe a particular remedy when the specific anti-abuse rule is violated. The government clearly has authority, however, to require a reasonable method to remedy the abuse, including the traditional method with curative allocations. The anti-abuse examples under the section 704(c) Regulations do not specifically sanction mandatory curative allocations of taxable operating income. As amended in 1984, section 704(c)(1)(A) requires allocation of “income, gain, loss, and deduction” to take into account the difference between the fair market value and basis of contributed property. According to the legislative history, the expanded scope of section 704(c) was intended primarily to address situations in which a cash-method partner contributes accounts receivable, accounts payable or other accrued but unpaid items to a partnership. In explaining this change, the Conference Report stated that “[b]y making reference to items of income and deduction, the conferees do not intend for the Treasury to require such variations to be eliminated by allocations of operating income and loss attributable to the contributed property (other than depreciation, depletion, and similar items).” Simultaneously, Congress amended section 704(c) to deal specifically with accrued but unpaid items of a cash-method contributor, thereby potentially obviating the need to expand the scope of section 704(c)(1)(A).

The 1984 legislative history should not be read as prohibiting allocations of operating income to cure ceiling rule abuses. Not only did Congress fail to fully apprehend the role of the ceiling rule, it also failed to consider the unique problem of wasting assets. In the case of assets whose value declines as taxable income is generated, abusive use of the ceiling rule reflects not just the contributed property’s lack of sufficient basis but also the ability to shift pre-contribution gain through use of the contributed property in the partnership’s business. Curative allocations of gain on sale will often fail to reverse a prior ceiling rule shift because the economic value of wasting assets declines as such property is depreciated or amortized. Since reasonable curative allocations of operating income are already permissible by agreement, mandating such curative allocations to address ceiling rule abuses represents only a minor incremental step. The existing “lack of symmetry between the treatment of gain from sale and the treatment of income from a wasting asset” may reflect Congress’ failure to fully appreciate that contributions of wasting assets often result in precisely the type of income-shifting that section 704(c) is intended to prohibit. Rectifying the ceiling rule abuse should not be left to the fortuitous circumstance of whether the contributed property generates sufficient gain from sale to remedy the abuse.

Nevertheless, some commentators maintain that the government may only cure ceiling rule abuses through curative allocations of “tax basis derivative” items—namely, items determined with respect to the tax basis of contributed property to the extent such items have no corresponding book (economic) corollary. By contrast, items such as partnership operating income give rise to corresponding book and tax items, allocated in accordance with the partners’ economic interests under the section 704(b) Regulations. Accordingly, mandatory curative allocations of operating income arguably lack economic effect under the section 704(b) Regulations. The argument that the government may require only the use of tax basis derivative items—not partnership operating income—to cure ceiling rule problems is consistent with the underlying analysis of former section 704(c)(2), the predecessor of section 704(c)(1)(A). The scope of current section 704(c)(1)(A) is broader, however, than former section 704(c)(2); moreover, the book-tax analysis mandated under the section 704(b) Regulations has radically altered the mechanics of eliminating section 704(c) disparities.

Former section 704(c)(2) referred only to “depreciation, depletion, or gain or loss with respect to property contributed to the partnership.” Under the elective rule of section 704(c)(2), partners could specially allocate gain on sale or depreciation deductions, subject to the ceiling rule, to prevent shifting of tax consequences related to contributed property. To compensate for a shortfall in the basis of contributed property, former section 704(c)(2) effectively allocated a disproportionate share of the partnership’s inside basis to the noncontributing partners. Thus, the former elective rule permitted partners to achieve equitable results by reallocating the tax basis of contributed property. Since partnership operating income bears no relation to the difference between the basis and value of contributed property, it was understandably outside the purview of former section 704(c)(2).

When Congress amended section 704(c) in 1984, the Final Regulations under section 704(b) had not yet been issued. The final section 704(b) Regulations require contributed property to be credited to a partner’s capital account based on its fair market value (rather than its tax basis) at the time of contribution. Because unrealized appreciation in contributed property must be reflected in the partners’ book capital accounts, subsequent allocations of taxable income corresponding to the booked-up gain cannot have economic effect. By requiring tax items to follow the corresponding book items, the final section 704(b) Regulations inflict any book-tax disparities on the contributing partner. Both sections 704(b) and 704(c) are intended to operate in tandem. In 1984, Congress could not have foreseen the extent to which capital account analysis under the final section 704(b) Regulations would “revolutionize” Subchapter K by permitting deferred gain in specific partnership assets to be linked to particular partners.

Under current law, analyzing section 704(c) in terms of the partners’ shares of common basis often produces a similar result as eliminating book-tax disparities under the section 704(b) Regulations. Focusing on tax basis (and basis derivative items) may serve, however, to obscure shifting of built-in gain when wasting assets produce taxable income in the partnership’s business. The analysis under the final section 704(b) Regulations focuses on reducing or eliminating book-tax disparities attributable to contributed property, rather than on reallocating inside basis. Thus, there is no reason why mandatory curative allocations should consist only of tax basis derivative items, even if such a limitation was sensible under former section 704(c)(2). Nor would allocating operating income violate the Treasury’s self-imposed stricture against requiring use of the remedial method or creating notional items in exercising its authority under Reg. section 1.704-3(a)(10).

IV. Solving the Problem

A. Stubborn Persistence of Ceiling Rule

In 1954, Congress adopted an entity approach as the general rule for allocating items attributable to contributed property. Former section 704(c)(1) allowed partners to allocate items relating to contributed property in any manner they chose. The partners’ ability to freely shift taxable income among themselves reflected the 1954 consensus that such income-shifting was “not a matter involving revenue considerations.” Simultaneously, Congress offered partners an optional aggregate method, under former section 704(c)(2), for avoiding ceiling rule distortions, but modified the aggregate approach by incorporating the “entity theory vestige of the ceiling rule.” The optional aggregate approach was generally consistent with a 1954 proposal by the American Law Institute (ALI)—referred to as the “deferred sale” or “credited value” approach—except that the latter had no ceiling rule counterpart. While it did not appear in the 1954 statutory language, the ceiling rule was embodied in Regulations issued in 1956.

In 1984, Congress changed course and required mandatory adjustments under section 704(c) “to prevent an artificial shifting of tax consequences between the partners with respect to pre-contribution gain or loss.” Although Congress granted the Treasury broad authority to implement section 704(c), the 1984 legislative history suggests that Congress intended to preserve a role for the ceiling rule. Apart from avoiding complexity and valuation difficulties, the ceiling rule has been justified based on the entity theory of computing a single partnership-level gain or loss. According to one commentator, “repeal of the Ceiling Rule (or even relaxation of such a rule through an anti-abuse mechanism) is not simply elimination of a statutory rule of convenience—it is a substantial erosion of Sections 703, 721, 722 and 723.” Under this view, the Treasury’s decision to retain the ceiling rule in the final section 704(c) Regulations was dictated by the need to “prevent section 704(c) from becoming more of a recognition provision than an allocation provision,” thereby curtailing partners’ ability to defer pre-contribution gain.

The notion that repealing the ceiling rule would somehow conflict with the basic nonrecognition policy of section 721 is untenable. Section 721 grants nonrecognition treatment upon an initial contribution of property but does not, by its terms, address subsequent recognition of gain or loss with respect to contributed property. The ALI’s 1954 partial deferred sale model contemplated that noncontributing partners would receive essentially a cost basis in the purchased portion of contributed property and corresponding tax deductions as the property’s basis was recovered; the contributing partner’s deferred gain attributable to the sold portion would be recognized in the form of reduced depreciation deductions, even though the partnership had not disposed of the property. In 1984, the ALI proposed an alternative version of the deferred sale model (the “full” deferred sale approach) that treated the partnership as obtaining a cost basis in the entire contributed property, rather than just the portion deemed sold to the noncontributing partners. Both the partial and deferred sale approaches achieve quite similar results. Under either approach, the contributing partner’s built-in gain is fully taxed; the main difference is that the partial approach is premised on the aggregate theory while the full approach reflects an entity theory.

In 1999, an ALI Reporters’ Study considered but did not recommend a mandatory full deferred sale approach as a possible solution for section 704(c) allocations. As the ALI Reporters’ Study recognized, the deferred sale approach had the advantage of eliminating the ceiling rule problem and potentially allowing repeal of ancillary provisions that reinforce section 704(c)(1). The full deferred sale method was viewed as too favorable to taxpayers, however, since the contributing partner received increased ordinary depreciation deductions at the cost of capital gain recognition. Also, the full deferred sale approach would give rise to significant issues on a subsequent transfer of the contributed property in a nonrecognition transaction. Because the partnership would be accorded a cost basis in deferred sale property before the contributor recognized the corresponding built-in gain, a subsequent nonrecognition disposition of the property could give rise to excessively generous or harsh treatment.

Similar considerations prompted the Treasury’s decision to adopt the remedial method as a substitute for the deferred sale approach under the proposed section 704(c) Regulations. As the Treasury indicated, use of the remedial method “achieves results substantially similar to the results under the deferred sale model. . . without the complexity of that method.” Given the shortcomings of the full deferred sale method, the remedial method likely represents the only feasible alternative if Congress decides to eliminate the ceiling rule. Because the section 704(b) Regulations now require contributed property to be reflected at its fair market value for purposes of maintaining book capital accounts, valuation uncertainties may no longer be a compelling reason for retaining the ceiling rule. Nevertheless, partners may still have an incentive to misvalue specific assets even when the total value of each partner’s contribution can be accurately determined. But even apart from valuation difficulties, mandatory remedial allocations are likely to be perceived as too complex for nearly all partnerships. The 1999 ALI proposal would have required mandatory remedial allocations only for sophisticated partnerships as the cost of gaining access to the flexible allocation provisions of Subchapter K. Similarly, the Treasury limited mandatory remedial allocations under the section 721(c) Regulations primarily to large corporate partners with a high overall level of related ownership.

B. Targeting Abuse

The specific anti-abuse rule under the section 704(c) Regulations may be viewed as a rough compromise between eliminating the ceiling rule and preventing taxpayers from achieving results that are clearly “too good to be true.” When the Treasury finalized the section 704(c) Regulations, it was well aware that the ceiling rule presented opportunities for shifting built-in gain. Prior to 1993, tax-avoidance schemes that exploited ceiling rule distortions had already received widespread publicity within the tax community. Although the litigated cases would not reach the courts until nearly a decade later, some of these shelter transactions were hastily undertaken to avoid the effective date of the anti-abuse rule. Once the anti-abuse rule was finalized, it was expected that “most arrangements that rely on the ability to use the ceiling rule, no matter how bad the apparent abuse, will disappear.” While “theoretically troubling” because of its uncertain reach and lack of a baseline against which to measure abuse, the anti-abuse rule was expected to apply only infrequently.

Despite the ubiquity of ceiling rule limitations, abuse of the ceiling rule generally arises only when partners with different tax profiles have the ability to trade tax characteristics in a manner that reduces their aggregate tax liability solely at the government’s expense. Unless the parties are related, transactions structured to manipulate the ceiling rule clearly require the contributing partner to provide economic compensation to the other partner—in the form of enhanced allocations of partnership income, guaranteed payments, priority returns and distributions, or low valuation of the contributed property—in exchange for absorbing the shifted taxable income. Once detected, such arrangements are unlikely to survive scrutiny, since it should be relatively easy to conclude that the arrangement was undertaken with the prohibited view of substantially reducing the present value of the partners’ aggregate tax liabilities.

Indeed, the drafters of the anti-abuse rule focused primarily on ceiling rule problems related to cost recovery deductions for depreciable or amortizable property. If a contribution of property is reflected on the partnership’s books at its fair market value, as required under the section 704(b) Regulations, the requisite view to shift taxable income upon a subsequent sale is unlikely to be present. The ceiling rule implicitly assigns any post-contribution decrease in the value of built-in gain property to the contributor. The effect of the ceiling rule depends, however, on the future sale price of the contributed asset, which will generally be impossible to predict at the time of contribution. Hence, abusive transactions involving an anticipated decline of the contributed property’s value prior to sale are likely to be quite rare.

The aggregate approach of section 704(c) attempts to link the tax consequences associated with contributed property to particular partners. In hindsight, it is clear that the entity-based ceiling rule frequently prevents the provision from accomplishing its intended purpose. Described as the “heart” of the section 704(c) Regulations, the anti-abuse rule of section 1.704-3(a)(10) represents the necessary antidote to the entity-aggregate confusion created by Congress in 1984. Despite the apparently flexible general rule of section 704(c)—allowing taxpayers to choose any reasonable method of section 704(c) allocations—such flexibility is not unlimited. The anti-abuse rule allows the government to remedy overly aggressive use of the ceiling rule to achieve unintended results. The section 721(c) Regulations now address specific abuses identified by the the Service and the Treasury involving contributions to related foreign partnerships. Nevertheless, such transactions (as well as other tax-avoidance techniques) are also vulnerable under the section 704(c) anti-abuse rule.

C. Reassessing the Anti-Abuse Rule

Nearly 30 years after its promulgation, the section 704(c) anti-abuse rule is ripe for reassessment. In an article that appeared shortly after the final section 704(c) Regulations were issued, one commentator expressed frustration with the anti-abuse rule, describing the abuse problem as one “of the Treasury’s own making: Had the Treasury abandoned the ceiling rule, the remedial method could be required of all partnerships, and the section 704(c) Regulations could be substantially simpler.” While arguing for a narrow scope for the anti-abuse rule, the commentator also stated that the Treasury “apparently believes that it lacks authority to eliminate the ceiling rule.” As the 1984 legislative history demonstrates, the primary responsibility for retaining the ceiling rule lies with Congress, not the Treasury. If the remedial method is simply not feasible for most partnerships, speculation about the Treasury’s authority (or lack of authority) to abolish the ceiling rule misses the point.

The drafters’ decision not to mandate use of the remedial method in the final section 704(c) Regulations very likely reflected significant uncertainty about how well the novel method would work, particularly given the shortcomings of the proposed deferred sale method. While the ambiguous legislative history may also have played a role, the drafters’ cautious approach was entirely understandable in light of administrability concerns that persist even today. In the Treasury’s view, a specific anti-abuse rule had the advantage of targeting particular types of transactions that have significant potential for abuse of the ceiling rule, rather than burdening all partnerships with remedial allocations. Because section 704(c) is such an integral part of Subchapter K, proposals to reform section 704(c) risk “unwittingly becom[ing] a subterfuge for challenging the most basic rules of subchapter K,” including tax-free contributions and distributions. While the pervasive effect of the ceiling rule should not be ignored, broader changes in section 704(c) should be addressed in the context of fundamental reform of Subchapter K.

The anti-abuse rule was intended to stop transactions such as those engaged in by Merck, General Electric, and Dow Chemical prior to its effective date. Like these earlier transactions, Bristol sought to abuse the ceiling rule through the simple expedient of contributing high-value, low-basis wasting assets to a partnership and shifting built-in gain to a tax-indifferent related party. In such transactions, the inappropriate shifting of built-in gain can be eliminated by reallocating operating income to the contributing partner as built-in gain is realized through use of the property in the partnership’s business. The existing Regulations do not provide any specific mechanism for reallocating built-in gain realized in the form of partnership operating income. Nevertheless, with the benefit of hindsight the government may readily determine the annual ceiling rule shift of built-in gain to the noncontributing partner. A corresponding reallocation of operating income to the contributing partner is needed to compensate for the shortfall in the tax basis of the contributed property. Without mandatory curative allocations of operating income, there would often be no effective cure for such ceiling rule abuses.

Interpreting the specific anti-abuse rule narrowly to prohibit curative allocations of operating income is unwarranted. It would frustrate the purpose of the anti-abuse rule to prevent tax-motivated shifting of built-in gain under section 704(c). Reallocating operating income would affect relatively few partnerships—those that cannot show that they lack the requisite view to abuse the ceiling rule in a manner that substantially reduces their aggregate tax liability—but would cure the ceiling rule problem in nearly all abusive situations involving wasting assets. If a partnership lacked sufficient operating income to rectify the ceiling rule distortion, the cure might nevertheless be imperfect. But such a problem would likely arise only if, subsequent to formation of the partnership, the value of the contributed property unexpectedly declined. Unlike ceiling rule distortions attributable to post-contribution changes in the contributed property’s value, shifting of built-in gain attributable to wasting assets is entirely predictable in advance. Not surprisingly, the most egregious tax shelters structured to exploit the ceiling rule have involved contributions of low-basis, high-value depreciable or amortizable property used in a partnership’s business.

In effect, the government’s proposed reallocation of operating income would treat Bristol as if it had sold a portion of the contributed intangibles for a series of ordinary income installment payments realized as ForeignCo’s purchased share of the property is amortized. This treatment is appropriate since, economically, Bristol exchanged an interest in the contributed zero-basis intangibles for the property contributed by ForeignCo with a tax and book basis equal to fair market value. Requiring Bristol to recognize ordinary operating income equal to the shifted built-in gain matches the ordinary amortization deductions that ForeignCo would be allowed upon an actual purchase of the contributed intangibles. Except for potential timing differences, the overall result is generally the same as would occur under the remedial method which requires matching of ordinary remedial income and ordinary remedial deductions. Since the remedial method is the theoretically correct method for handling contributions of built-in gain property, it should also serve as a benchmark for determining whether use of the traditional method (or traditional method with curative allocations) is abusive.

Once the problem of low-basis, high-value wasting assets is properly understood, there is no reason why Congress should not require the contributor’s built-in gain to be recognized as such property is depreciated or amortized. No actual sale or disposition would be required on the theory that depreciation or amortization of a wasting asset approximates the income to be generated by the asset. Consistent with section 721 nonrecognition treatment and carryover of basis in the partnership’s hands, the contributed property should be bifurcated into two separate assets. The partnership would be treated as acquiring an asset with a tax basis equal to its book value; this asset would be depreciated or amortized as under existing law. The remaining built-in gain asset would have a book basis equal to its fair market value and a tax basis of zero; depreciation or amortization of the built-in gain asset would give rise to mandatory curative allocations of ordinary income items sufficient to reduce or eliminate ceiling rule distortions. Thus, the contributing partner would be forced to recognize built-in gain as the property is depreciated or amortized, with any remaining built-in gain recognized upon sale (or liquidation of the partnership). Unlike under the remedial method, such mandatory curative allocations would consist of actual partnership items of the appropriate character. Although such curative allocations are already elective under the traditional method, Congress should make them mandatory to prevent shifting of built-in gain from wasting assets used in the partnership’s business.

V. Conclusion

When partners clearly have the requisite view to substantially reduce their aggregate tax liability by exploiting the ceiling rule, the anti-abuse rule would be a nullity if no mechanism were available to remedy the abuse. Fortunately, reallocation of partnership operating income from Bristol’s wasting assets would fully reverse the shift of built-in gain to ForeignCo. The apparent failure of Bristol’s sophisticated tax advisers to opine on the anti-abuse rule, while opining on related issues, is glaring. Those tax advisers may well have relied on a technical argument that the government was helpless to cure the obvious abuse because operating income is not basis derivative, an argument that should be foreclosed by the broader scope of section 704(c)(1)(A) and the book-tax analysis required under the current section 704(b) Regulations. Bristol may seek to bolster this technical argument by misreading the 1984 legislative history as flatly prohibiting use of operating income to cure ceiling rule abuses. The legislative history does not purport, however, to restrict the government’s ability to remedy 704(c) abuses; moreover, Congress clearly failed to perceive the unique problem of wasting assets where reallocation of operating income will often be the only possible fix. If the Bristol transaction is litigated, its tax advisers will be forced to address the anti-abuse rule squarely.

Recently, controversial proposals to reform Subchapter K would have required all partnerships to adopt the remedial method. Ironically, opponents of mandatory remedial allocations suggested that such reform was overly complex and unnecessary because the government already possessed adequate authority to remedy abuses. If the government is powerless to address the Bristol transaction under the specific anti-abuse rule, Congress may find it necessary to mandate the remedial method, greatly increasing complexity for all partnerships. In the case of wasting assets, a more targeted approach would require curative allocations of operating income to remedy ceiling rule abuses, while obliging taxpayers to disclose and justify the reasonableness of their choice of section 704(c) methods. Such disclosure would help to identify and deter abusive transactions, and it would provide a salutary warning to tax advisers who might be tempted to render opinions on transactions involving an unreasonable choice of section 704(c) methods.