Section 704(b) and its Regulations allows partnerships a great deal of flexibility on how items of income and deduction are allocated to partners. This flexibility has been heavily criticized over the years. The article reviews section 704(b) and its Regulations, including the partner’s-interest-in-the-partnership test, the substantial-economic-effect safe harbor, and “target allocations.” (Target allocations are widely used notwithstanding the lack of clear legal underpinnings). The article discusses the shortcomings of the existing scholarship, argues for a flexible section 704(b) allocation regime, but acknowledges that reform is necessary. The article proposes a new definition of substantiality, limiting section 704(b) to “bottom-line” items, and adding a safe harbor for target allocations.
Keywords: Section 704(b), allocations, partner’s interest in the partnership, substantial economic effect, target allocations, Wynn, related partners.
There have been many articles written on the allocation rules of section 704(b) over the years. Many of these articles suffer from a disconnect with the real world. It is common, for example, for articles to recommend that allocations essentially be made the same way they are for S corporations, rigidly based on the partners’ capital contributions. Some have even recommended that we should (directly or indirectly) eliminate Subchapter K. As I will discuss, these views often are not only politically unrealistic but fail to meet the real-world needs of many businesses.
In September 2021, Senate Finance Committee Chair Ron Wyden (D. OR) released a draft proposing what would be the most substantial changes to Subchapter K since 1954 (“Wyden Proposals”). While taking a hard look at Subchapter K is commendable, I will discuss where its proposals to reform section 704(b) allocations fall short.
I agree, however, that reform is necessary. But any reform must align with how the real world operates. My proposal, discussed in detail below, would redefine substantiality, limit allocations to “bottom line” items, and add “target allocations” as a safe harbor. This approach should limit abusive use of allocations, while still permitting legitimate deals to take place, and would align the Regulations with current legitimate practices.
It is worth recalling that partnerships play a pivotal role in our economy. The most recent year for which tax filing data are available is 2020. In that year there were 4.3 million partnerships with over 28.2 million partners, almost $43.2 trillion in assets, and more than $760.2 billion in net income. In 2020, about 1% of partnerships held about 79% of partnership assets, and about 10% of partnerships held about 95% of partnership assets. Therefore, the bottom 90% of partnerships only held about 5% of partnership assets. It is tempting to focus on the largest partnerships after seeing these figures. Indeed, the Wyden Proposals seem to do just that. One should tread carefully, however; 5% of $43.2 trillion of partnership assets is still a big number in absolute terms and can have a substantial economic impact. Any changes to Subchapter K should take small and mid-sized partnerships into account. They should also be made carefully with appreciation for their economic impact given the economic size of the partnership universe. Ill-advised reform has the potential to wreak real economic havoc.
In the main, I will assume that the partners are fully taxable and acting at arm’s length. Allocations between related partners raise very different concerns and are not a focus of this article. Nonetheless, in light of the mostly sensible Wyden Proposal in this regard, I will discuss them briefly. Tax exempt partners and items the allocation of which cannot have substantial economic effect, such as nonrecourse deductions and tax credits, are mostly beyond the scope of this article, though some discussion is unavoidable. I will assume that the at-risk rules of section 465, the passive loss rules of section 469, and the excess business loss rules of section 461(l) do not limit the deduction of losses. More often than not, this is a bad assumption, but it creates excessive complications to have to fold these loss limitation rules into my discussion. Further, given how common it is to use partnerships, the loss limitation rules do not seem to provide much of a barrier in ways that are relevant to this article. Finally, by partnerships, tax partnerships are meant, which typically includes, under the default rule, LLCs with two or more members.
Part I provides a basic review and analysis of the relevant law, Part II discusses target allocations and why they should be included in any safe harbor, Part III addresses related party allocations, Part IV discusses misguided criticisms of section 704(b) allocations, Part V critiques the Wyden Proposals on section 704(b) allocations, Part VI provides my proposals for reforming section 704(b) allocations, and Part VII contains the conclusion.
A partnership is a flow-through entity. Income and deductions are passed through to the partners. A mechanism needs to exist, therefore, for determining each partner’s allocable share of partnership income and deductions. Section 704(b) and its Regulations generally allow partners a great deal of flexibility in this regard. The allocations do not necessarily need to be in proportion to the underlying ownership of the partnership interests or capital contributions. Someone who contributed 10% of the partnership’s capital could be allocated 90% of depreciation deductions, for example. Or, in the classic “Brains-Money” deal, all losses could initially be allocated to the “money partners,” with subsequent income allocated to them to the same extent as losses were previously, with income then allocated 50% to the “money partners” and 50% to the “brains partners.” This is sometimes called a “flip.” Disproportionate allocations are sometimes called “special allocations.”
Section § 704(b) provides that a partner’s “distributive share of income, gain, loss, and deduction, or credit . . . shall be determined in accordance with the partner’s interest in the partnership . . . if ” the partnership agreement does not provide for how a distributive [i.e., allocable] share will be allocated or if the allocations do not have “substantial economic effect.” Thus, if an allocation does have substantial economic effect, it need not be in accordance with a partner’s interest in the partnership. As I will discuss, a partner’s interest in the partnership is determined under a facts-and-circumstances test. The Regulations provide specific—and at times quite complex—rules as to when allocations have substantial economic effect. The substantial economic effect rules provide a safe harbor. If the partnership agreement complies with the rules, the partnership knows the transaction will be safe. It used to be that practitioners viewed compliance with the substantial economic effect rules as virtually mandatory, but for some time now practitioners have been increasingly drafting agreements to come under the partners’-interest in-the-partnership facts-and-circumstances test. Indeed, in large, complex deals, the latter approach is likely the norm.
The partnership allocations rules have been called “a creation of prodigious complexity . . . essentially impenetrable to all but those with the time, talent, and determination to become thoroughly prepared experts on the subject.” In complex deals, I fully subscribe to this view. But in simpler deals, like the Brains-Money example above, it is an exaggeration. Simpler deals do not require the same kind of command of these rules, and fewer of the rules are relevant. Thus, the burden of the complexity is not the same for all partnerships.
For an allocation to comply with the substantial economic effect rules, the capital accounts must be maintained in accordance with the rules in the Regulations. As the name substantial economic effect suggests, to meet the safe harbor an allocation must have a genuine post-tax, economic effect on the partner to whom the allocation is made. The rules for maintaining the capital accounts help to fulfill this task. Generally, these rules are based on sound economic concepts and are consistent with the applicable financial accounting rules.
As the focus is on the economic rather than tax impact, the rules for keeping partners’ capital accounts are quite different from the rules for computing the bases of partners’ partnership interests.
Under the Regulations, a partner’s capital account is increased by:
A partner’s capital account is decreased by:
Section 752 effectively permits a partner to include her share of partnership liabilities in her basis in her partnership interest. A partner’s capital account, on the other hand, does not include that partner’s share of liabilities. If the partnership has liabilities, a partner’s basis in her partnership interest often will exceed her capital account balance. The amount of basis in the partnership interest is highly important because a partner may receive loss allocations up to her basis in the partnership interest under section 704(d). As the basis in the partnership interest goes down, the capital account can go negative, which the Regulations permit in certain circumstances. For example, if a partner’s basis in the partnership interest is $25,000 and capital account is $15,000, the partner may be allowed to have up to a negative $10,000 capital account balance.
A partnership normally also maintains “book” accounts for the properties it holds. For example, if a partner contributes property with a tax basis of $7,000 and a fair market value of $10,000, the partnership’s tax basis in that property under section 723 is $7,000. However, the partnership’s “book value” (which some tax professors like to call book basis) is the full fair market value of $10,000. Book value, like capital accounts, focuses on the economic value of contributed property. If a partnership makes a distribution of property for which the fair market value differs from its book value, for capital account purposes the partnership recognizes the inherent gain or loss and allocates that gain or loss to the partners’ capital accounts. There may not be any corresponding taxable gain or loss.
As noted above, the Regulations’ substantial economic effect rules (“SEE”) are a safe harbor. An allocation that complies with SEE will be allowed under section 704(b). There are two parts to the SEE test. First, the allocation must have “economic effect.” The Regulations in most instances provide a mechanical test for determining whether or not an allocation has economic effect. Second, because it is often possible to manipulate the economic effect test, the Regulations also provide that the economic effect of an allocation must be “substantial.”
Partnerships have three options under the Regulations to satisfy the economic effect test, the “regular” economic effect test, the “alternative” economic effect test, and the “economic effect equivalence” test, each of which is discussed below.
a. Regular Economic Effect Test. The regular test has three parts:
Example: A and B each contribute $10,000 to the AB partnership and are equal partners. The partnership borrows $40,000 on a recourse basis with only interest due for the first five years of the note. Assume that under section 752, $20,000 of the liability is allocated to each partner. AB purchases equipment for $60,000. The equipment generates depreciation deductions of $10,000 per year. A has a beginning tax basis in the partnership interest of $30,000 and a beginning capital account of $10,000. Now assume that in Year 1 the partnership breaks even except for depreciation deductions on the equipment and allocates the entire $10,000 of that depreciation to A. A’s basis is reduced to $20,000 and her capital account is reduced to zero. In Year 2, the partnership again breaks even on partnership operations except for depreciation, and again allocates $10,000 of depreciation to A. A’s basis is reduced to $10,000, and A’s capital is reduced to a negative ($10,000). If A’s partnership interest is liquidated on January 1 of Year 3 with the partnership relieving A of any obligation on the partnership liabilities, A is required to contribute $10,000 to the partnership to bring her capital account to zero under the DRO rule. Without this requirement, A would be getting more out of the partnership than she put into it, i.e., her tax losses would exceed her economic losses. The contribution of $10,000 ensures that the entire allocation of depreciation indeed has an “economic effect” on her. The contribution increases her partnership interest basis to $20,000. As A’s share of the liability is also $20,000. Under section § 752(d), A’s amount realized on the liquidation of her interest is also $20,000, yielding no gain or loss.
b. Alternative Economic Effect Rules. The difficulty with the regular economic effect rules is that partners are required to have unlimited DROs. This is rarely wise, especially for passive investors. The example I use when teaching this material: Assume the partners form a limited partnership and that all partners have unlimited deficit restoration obligations. An employee of the partnership, while conducting partnership business, runs over and kills a neurosurgeon with eight handicapped children. A large tort liability, in excess of insurance limits, results. The general partner is the only one liable under state partnership law, and he contributes sufficient funds to the partnership to enable it to pay the liability, increasing the general partner’s capital account. The payment results in a large tax loss to the partnership that (if the limited partners had unlimited deficit restoration obligations) may be primarily allocated to the limited partners. The allocation causes the limited partners to have substantial negative capital accounts. Should they have to restore those deficit capital accounts (as might happen if the general partner decided to take this opportune moment to cause the partnership to liquidate), they would in effect be paying the tort liability, something that likely was not contemplated when they entered into the partnership agreement. The bottomless risk that an unlimited deficit obligation poses causes most advisors to recommend that their clients not agree to such a provision.
The Regulations, recognizing this business reality, contain an alternative economic effect test. Under this alternative, an allocation must meet the first two economic effect tests (keep capital accounts according to the rules and upon liquidation, pay to a partner any positive balance in his capital account), but instead of having an unlimited deficit restoration obligation, the third requirement is that the partnership agreement contain a qualified-income-offset provision (discussed below). If this alternative test is met, an allocation will be treated as having economic effect if the allocation does not cause the partner to have a deficit capital account balance or increase an already-existing deficit capital account balance. If allocations cannot drive a capital account negative, how would a deficit capital account arise? Primarily, it would go negative due to distributions, which the Service cannot prevent a partnership from making. Thus, the Service needed a mechanism for eliminating the deficit capital account of a partner who has no obligation to restore it. That mechanism is to require the partnership to allocate income as quickly as possible to the partner to offset any such deficit, i.e., a “qualified income offset.”
Sometimes partners have limited deficit restoration obligations. They will agree to restore a deficit in their capital account up to a certain amount, but not beyond that. Limited deficit restoration obligations are much more common than unlimited DROs. A partner might agree to a limited deficit restoration obligation in order to be allocated more losses. In this circumstance, the partnership will need to comply with the qualified-income-offset rules, and allocations may be made to a partner that create a negative capital account up to the fixed amount that partner is obligated to restore. Thus, if a partner has a $10,000 deficit restoration obligation, he could be given allocations that caused him to have up to a $10,000 negative capital account as long as the partnership otherwise complies with the qualified-income-offset rules.
c. Economic Effect Equivalence. The third alternative provided in the Regulations to meet the economic effect test is the “economic effect equivalence test.” Allocations made to a partner that do not otherwise have economic effect under the rules discussed above can nevertheless be deemed to have economic effect under this test. The economic effect equivalence test is met provided that a liquidation of the partnership at the end of a given year (or at the end of any future year) would produce the same economic results to the partners as would occur if the regular economic effect test were met, regardless of the economic performance of the partnership. In some ways, this test is designed for partnerships which failed to include appropriate economic effect provisions in their agreements, but whose structure is inoffensive. For example, assume A and B contribute $75,000 and $25,000, respectively, to the AB partnership. Assume further that the partnership maintains no capital accounts and the partnership agreement provides that all income, gain, loss, deduction, and credit will be allocated 75% to A and 25% to B and distributions will also be 75-25. A and B are ultimately liable (under a state law right of contribution) for 75% and 25%, respectively, of any recourse debts of the partnership. Although the allocations do not satisfy the requirements of the regular or alternative economic effect rules, the allocations have economic effect under the economic-effect-equivalence test. In principle, this test only applies to the simplest partnerships. In more complex situations, such as when the partners have varying interests over time or varying interests in different items, it will be much more difficult (or impossible) to prove that a liquidation of the partnership at the end of a given year (or at the end of any future year) would produce the same economic results to the partners as would occur if the formal economic effect test were met, regardless of the economic performance of the partnership. Those partnerships usually need to comply with the regular or alternative tests to be safe. That said, partnerships have sometimes used this test to defend target allocations, as I discuss below.
a. General Rules. For all of their complexity, the economic effect rules are not enough to get the job done. They are in the main mechanical rules, and like all mechanical rules can be manipulated inappropriately. Accordingly, the Regulations provide that not only must the allocation have economic effect, but also that economic effect must be substantial. The Regulations provide three independent tests for determining whether the economic effect of an allocation is substantial: (1) “the present value post tax rule,” (2) the shifting tax consequences test, and (3) the transitory allocations test. Because these tests are independent of one another, one must effectively pass all three for the economic effect of an allocation to be substantial, though not all may be relevant. There is essentially no case law on substantiality.
(i) Present Value, Post Tax Test. For the present value post tax test, the Regulations provide that an allocation is not substantial if:
Or, as I like to tell my students, the allocation is not substantial if, on a present-value, post-tax basis, someone is better off, and no one is worse off than would be the case if the allocation had not been made. If no one is worse off, it necessarily means that the allocation only had a tax effect and not a bottom-line economic effect (on a present value post tax basis). For there to be SEE, on a present value post tax basis, if any partner is made better off, then another partner must be worse off and vice versa. The devil is in the details.
In judging whether the economic effect of a given allocation is substantial, one must ask: Compared to what? The comparison would be with how items would be allocated if the allocation in question did not exist. In most of the examples that I will discuss in this Article and that are contained in the Regulations, it is fairly obvious what the comparative allocation would be. But in the real world, determining the comparative allocation can present a major challenge. Generally, the comparison is made to the allocations that would be made in accordance with the partners’ interests in the partnership; but as I will discuss, determining the partners’ interest in the partnership is sometimes no small feat.
For a special allocation, it is likely a given that someone is better off, or the allocation never would have been made. Accordingly, the focus typically is on the second part of the test: is someone worse off? What does it take to have a “strong likelihood” or be “substantially diminished?” The Regulations do not contain definitions of these terms. The Regulations give one example in which the after-tax economic consequences of an allocation was not substantially diminished; it states that $360 is not substantially less than $362.50 (about a .7% difference). But the terminology of the Regulations seems to suggest that to be substantially diminished, it must be very likely that some partner’s present value post-tax outcome will be worse than it would be without the allocation and by a significant amount. How likely does it have to be: 100%, 75%, over 50%? And how much of a detriment does it have to be? Compared to the circumstance when the allocation is not present, does a partner have to be worse off by some absolute dollar amount or by a percentage? I would assume the latter, but then how big of percentage? Apparently, it has to exceed .7%, but by how much? Does it have to be 5%, 10% more? These are all unanswered questions.
In determining whether a partner is better off or worse off, tax consequences that result from the interaction of the allocation with the partner’s tax attributes that are unrelated to the partnership are taken into account. Example: A partner has a substantial net operating loss unrelated to the partnership that would otherwise expire unused. Allocating extra income to the partner would not increase her tax liability to the extent offset by the net operating loss; as a consequence, under the Regulations, the economic effect of the allocation is not substantial.
Another substantiality example: Assume taxpayers A and B are equal partners in the AB partnership. A expects to be in the 50% tax bracket over the next several years. B, on the other hand, expects to be in the 15% tax bracket. Over the next several years the partnership expects to earn approximately equal amounts of tax-exempt interest and taxable dividends. Rather than divide each type of income equally, A and B agree that 80% of the tax-exempt income will be allocated to A and the balance of the tax-exempt income and all of the taxable dividends will be allocated to B. The partners can make this allocation without violating any of the economic effect rules. The economic effect of the allocation is not substantial because on a present-value, post-tax basis, A’s position is enhanced (compared to the situation in which she would if she had been allocated half of each type of income) and B’s position is not diminished (indeed his position is also enhanced, on an after-tax basis, B receives more than he would if he had been allocated half of each type of income).
Unlike the transitory allocation rule discussed below which generally looks ahead no more than five years, the present value post tax rule does not have a time limit. However, the longer an allocation structure takes to come to closure, the greater the risk to the partners, and the more likely that the economic impact of the allocation is genuine. The lack of a formal time limit, therefore, typically is not of great significance.
(ii) Shifting Allocations. The Regulations provide independent substantiality tests for what the Regulations call “shifting” and “transitory” allocations. For these tests, the focus is on capital account balances. Generally, shifting allocations occur within a single tax year, and transitory allocations occur over a period of up to five years. In either case, the economic effect of an allocation will not be substantial if there is a strong likelihood that the capital accounts of the partners would be about the same as they would have been had the allocation not been made, and the allocation results in a net reduction of the partners’ tax liability.
Beginning with shifting allocations, assume our equal AB partnership now owns section 1231 property and capital assets. It expects to sell each type of property in the current tax year and incur a $50,000 section1231 loss and a $50,000 capital loss. The partnership agreement complies with the economic effect rules. Partner A has ordinary income of $300,000 and no section 1231 gains. She can therefore fully use the section 1231 loss but make only limited use of the capital loss. Partner B has $200,000 of ordinary income and $100,000 of section 1231 gains, meaning that he can fully use either type of loss and receive the same tax benefit. The partnership amends the partnership agreement and provides that for the current tax year only, all section 1231 losses will be allocated to A and all capital losses will be allocated to B. While the allocation will have economic effect, the economic effect will not be substantial. There is a strong likelihood (actually, an absolute certainty) that A and B will have the same capital account balances they would have had if the allocation were not contained in the partnership agreement (still a $50,000 loss each, consisting of equal parts of each type of loss). Further, the total taxes of A and B are reduced as a result of the allocation (A’s taxes go down, B’s taxes are unaffected).
(iii) Transitory Allocations. Transitory allocations operate in essentially the same way as shifting allocations, except they occur over a period of years. Under the Regulations, if there is a strong likelihood that: (1) an “original allocation” and a later “offsetting allocation” will leave the capital accounts approximately where they would have been had the allocations not occurred and (2) the tax liability of the partners will be reduced as a result of the allocations, then the economic effect of the allocations will not be substantial. The Regulations provide that if the offset happens and taxes are reduced, it will be presumed that there was a strong likelihood that this would happen unless the taxpayers can present facts and circumstances demonstrating otherwise. However, if there is a strong likelihood that the offsetting allocation will not be made “in large part” (another undefined term) within five years of the original allocation, then the economic effect of the allocation will be substantial.
Expanding on a prior example, assume that our equal AB partnership has predictable, approximately equal amounts of income each year and A has an expiring net operating loss carryforward. To allow A to take greater advantage of the net operating loss carryforward, the partnership allocates all of its income in Year 1 to A. It allocates all of its income in Year 2 to B. Thereafter, it returns to allocating income equally between the partners. The partnership agreement complies with the economic effect rules. The economic effect of the allocation is not substantial because there is a strong likelihood of the offset occurring and the partners’ tax liability is less than it would have been without the allocation (the allocation lowers A’s taxes and, except for modest time-value-of-money considerations, which the Regulations ignore, is neutral as to B). Note that if the offset would occur more than five years after the original allocation (not that B would ever agree to that), the allocation would be allowed.
b. Depreciation–Recapture Gain Chargebacks. The regulatory rules on depreciation and the associated gain chargeback might be the most controversial portion of the substantiality regulations. It is quite common for partnership agreements to allocate depreciation to a subset of partners, typically the money partners, and upon a subsequent sale of the property, to allocate any gain recognized to the same subset of partners up to the amount of depreciation allocated to them. Such a provision is sometimes called a “gain chargeback.” Recognized gain in excess of the amount needed for the gain chargeback might be allocated to all of the partners. If everything goes according to plan, this approach can offer substantial tax savings to the relevant subset of partners. The preferential allocation of depreciation reduces the partners’ ordinary income, and, particularly with real estate, any gain allocated would be favorably taxed as section 1231 gain or long-term capital gain.
One might ask whether there could be a transitory allocation issue, assuming the gain is recognized within five years of the depreciation deductions. The deduction for depreciation allocated in the early years is offset by an income allocation in a later year, leaving the relevant partners’ capital accounts in the same place they would have been had the depreciation allocation never been made. It should at least be possible for this arrangement to violate the transitory allocation rules, but under the Regulations it will not. There cannot be a strong likelihood of the offset occurring because the Regulations assume, for purposes of the substantiality rules, that a property has a fair market value equal to its book value. Book value, in turn, is reduced by depreciation deductions. For example, if book value has been reduced to zero, the Regulations assume for substantiality purposes that the fair market value of the property is zero. Any gain, given the presumption, is a “surprise.” There is no coherent reason for this assumption (“FMV-Book rule”). And in the case of real estate, it is likely to be untrue. Even those critical of Subchapter K reform proposals acknowledge that the regulatory assumption is dubious.
Interestingly, apparently even the drafters of the Regulations could not fully accept this assumption. While the FMV-Book rule is unequivocal, as such, an exception of sorts is carved out for zero book value property that continues to generate net income. To state the obvious, property that continues to generate net income does not have a fair market value equal to zero even if the book value is zero. Yielding to that reality, Example 2 of the Regulations states that property subject to a lease that extends beyond its recovery period can have a “strong likelihood” of producing taxable income notwithstanding the fact that its presumptive value is zero. Thus, while the Regulations sanction a loss allocation offset by gain attributable to the value of future income, they do not sanction a loss allocation offset by income from the property, at least if the property is subject to a long-term lease.
Rather than create a contradictory example, the Regulations should never have promulgated the FMV-Book assumption. Eliminating this assumption can have a large impact in the real estate context. It is common for allocation structures to take advantage of the disparate tax brackets of the partners. Typically, real estate ventures involve shifting depreciation deductions from low-bracket general partners/managing members to high-bracket limited partners/non-managing members. Eliminating the assumption will require these ventures to have a greater economic focus than is currently the case. Of course, if the FMV-Book rule were eliminated, it is not inevitable that the transitory allocation rules will be violated in the case of allocations of depreciation with a subsequent gain chargeback. There still needs to be a strong likelihood of the offset. It depends upon the facts. For any single deal, it may be quite difficult to show this strong likelihood. But, particularly for real estate, investors can be fairly confident that, over a series of investments, on average an investment will prove profitable, and the early depreciation deductions will eventually be offset by favorably taxed gains. Not only does that make the FMV-Book rule difficult to justify, it also raises questions about whether the strong likelihood test is an appropriate standard, as I will discuss in my proposal.
c. Compliance. SEE is undoubtedly complex. While I am not aware of any data on point, I do not doubt that many return preparers “wing it at times,” particularly in light of the low audit risk for partnerships. But it is a bit simplistic to say, as has been suggested, that the partnership world is divided into “elite” and “forgotten” (by which is meant mostly smaller) partnerships, with the latter unable to implement the Regulations. As noted earlier, for many structures, such as the Brains–Money example, one does need to have command of every nook and cranny of the Regulations. For a genuine economic deal, the substantiality rules likely do not meaningfully come into play. It can be a fairly straight-forward matter to keep proper capital accounts and make sure they are properly adjusted. There are other deals where true mastery of SEE is required to be in compliance. An example might be a partnership in which different classes of partners share in different proportions (sometimes called “waterfalls”), with each class having different preferred returns. But these latter partnerships also typically involve much larger dollar amounts, and the partnerships can afford the pricy counsel with the needed expertise.
While the substantiality rules are manageable, I do not mean to suggest that repealing the FMV-Book rule is the only needed improvement. As already noted, it can be very difficult to determine what the alternative allocation should be when analyzing substantiality. This is particularly true in complex deals with many layers of allocations. Does the alternative have to be the least tax efficient or merely moderately so? The “strong likelihood” test can be excessively taxpayer-friendly. For this reason, I propose to reformulate the substantiality test, as discussed below in my proposals for reform.
D. To PIP or not to PIP
Recall that SEE is a safe harbor within the more general rule that an allocation must be in accordance with the partner’s interest in the partnership (“PIP”). But there is only modest guidance on PIP. The baseline rule is that PIP and the partner’s interest in any particular item of partnership income, gain, or loss are generally determined by taking into account all facts and circumstances relating to the economic arrangement of the partners.
The Regulations provide that the following facts and circumstances are ordinarily taken into account for purposes of determining PIP or a partner’s interest in any particular item of income, gain, or loss, though the list is not exclusive:
Although PIP has been an important consideration in determining the partners’ distributive shares for over 45 years, there has been less than universal agreement as to the approach and reliability of PIP. This tension may be illustrated by contrasting the comments of the two major treatises on partnership taxation. Willis, Pennell, Postlewaite, and Lipton conclude that “There is not a conflict between a partner’s interest in the partnership and substantial economic effect. They both rely on the same overriding principle that the tax effects of partnership operations must conform to the economic effect of those operations.” On the other hand, McKee, Nelson, and Whitmire caution that “it is far from clear that identical results would in fact be achieved under both the partners’-interests-in-the-partnership rule and the substantial economic effect safe harbor, and thus drafters of partnership agreements who stray from the safe harbor do so at their peril.”
The Regulations make it reasonably clear that PIP is determined on an item-by-item basis. Thus if a partner has a 50% overall interest but a 90% interest in depreciation, at least the starting point for PIP for depreciation should be the 90% interest and not the 50% overall interest. None of the current PIP cases have actually gotten that far into the weeds, typically involving ill-advised partnerships that either did not have written agreements or made fundamental errors in their written agreements. There have not been many cases that have looked at PIP, as such, but my research indicates that courts are generally getting to the right answer. In the main, the cases have concluded that allocations that fail SEE are allowed under PIP when they are grounded on solid economics and not allowed when they are not. The sample size is small, but caselaw suggests that PIP is a viable standard.
Some have complained that the regulatory definition of PIP is so cursory as to be of little value. Yet, how much more detailed could it really be? There are myriad ways of structuring legitimate economics in partnerships, and PIP Regulations that cover all of them with specificity was never a realistic or achievable objective.
There was a time when complying with the SEE was seen as virtually mandatory. Why rely on a vague PIP standard when a more detailed safe harbor exists? One probably would not want to use PIP for simpler deals, such as the Brains-Money example above. But as partnerships get larger and more complex, complying with SEE can present a major drafting challenge. Further, even if that drafting challenge is met, will the accountants maintaining the books and filing the returns be able to implement it (I speak from experience)? It can be quite difficult to make the capital accounts come out correctly after taking the partners’ underlying economic deal into account. An example would be a partnership with multiple classes of partners, each with different allocation preferences, some of which change over time. Working through all of the allocations and making sure the capital accounts have the correct balances in light of the underlying economic arrangement has made for many a lawyer’s sleepless night. It is easier to focus on where the cash should go. If Partner A invested X dollars, and the partnership succeeds, to what kind of return is A entitled and how much should A receive on liquidation of the partnership? Most lawyers find the latter easier to draft; they can “follow the cash” instead of slugging their way through SEE.
Even if this is true, given PIP’s vague confines, isn’t ignoring SEE too risky? In the early going, the answer was yes, but over time, perhaps encouraged by the low rate of audit for partnership returns, larger partnerships have moved away from SEE and toward “target allocations.” Target allocations do not comply with SEE and rely instead on PIP. While I am not aware of any hard data in this regard, my own casual surveys of American Bar Association (ABA) Tax Section members and general experience in the area indicate that target allocations are the preferred approach for larger partnerships. I discuss target allocations next.
Detailed discussions of target allocations can be found elsewhere, but some discussion is unavoidable. Target allocations arose out of the drafting challenges created by SEE. Target allocations, in a sense, invert SEE. Rather than basing distributions on capital account balances as required by the economic effect rules, it bases capital account balances on intended distributions. The parties agree on how current and liquidating distributions will be made, and income and loss are allocated to the capital accounts to ensure that there is a sufficient balance in the capital account to cover the distribution (especially on liquidation of a partnership interest). Typically, “target capital accounts” are kept following the same rules as are required for SEE. What is missing is the obligation to liquidate in accordance with capital account balances, violating the economic effect test.
For purposes of this discussion (and to avoid unhelpful complexity), I will assume any distribution is in cash. Example: G and L form a limited partnership. G, the general partner, contributes $10,000 to the partnership, and L, the limited partner, contributes $90,000. Upon obtaining a nonrecourse loan from a commercial bank in the amount of $900,000, the partnership purchases a building on leased land for $1.0 million (leaving no working capital in this simple example). Assume Partnership operations result in operating income (rents in the amount of $150,000) being exactly equal to operating expenses (maintenance, repairs, land lease payments, loan interest in the aggregate amount of $150,000), so that net losses equal depreciation deductions for the year. Assume for computational simplicity that straight-line depreciation is allowed over a ten-year recovery period, ignoring the mid-month convention, so that each year’s depreciation is $100,000. The deal between G and L is that losses will be allocated 10% to G and 90% to L, income will be allocated in the same manner until allocations of income equal prior allocations of losses, and thereafter income will be allocated 50-50. Similarly, distributions will be made 10-90 until each partner has recovered his original capital contributions ($10,000 to G and $90,000 to L), and thereafter the partners will share distributions equally. To implement this approach with “target capital accounts,” the partnership agreement might provide that profits will be allocated in the following order of priority: First, to any partners having negative balances in their respective target capital accounts, in proportion to such balances, in amounts sufficient to bring these negative balances to zero. Second, to the partners who are entitled to the distributions under the partnership agreement to the extent necessary so that the target capital account balances of those partners are at least equal to the amounts remaining to be distributed to them under the partnership agreement. If there are not sufficient profits to fully achieve these allocations, the same formula will be followed to the extent of available profits. If profits exceed what is necessary to achieve these allocations, then the excess in this fact pattern would be allocated 50-50. Losses, on the other hand, might be allocated first among those partners whose target capital account balances exceed the balances of those partners’ respective contributions, in proportion to (and to the extent of) their respective excesses, and second, in this fact pattern, 10% to G and 90% to L.
Often, target allocations will generate the same capital accounts balances as SEE. But the beauty of target allocations is that a partnership is not required to liquidate in accordance with capital account balances. A partner might be paid more or less than that balance, though this can generate a taxable gain or (in liquidation) a loss, assuming capital accounts are equal to the partners’ bases in their partnership interests. In more complex deals, target allocations are typically believed to be easier to draft. Part of what makes the drafting easier is that the scrivener has to provide the economic terms of the deal only once in the distribution section of the partnership agreement. The allocations are forced to conform to the distributions to be made, and, at least where there is sufficient gain, cannot be in conflict with them. Indeed, some call target allocations the “forced allocation method.” While SEE often requires the drafter to anticipate the future, target allocations operate after the fact, that is, after the events that give rise to them have occurred. Provided it is clear what the economic deal is among the partners in the event of a hypothetical liquidation, the drafter can focus directly on building up or reducing the balances of the partners’ year-end capital accounts so that they ideally equal the amounts the partners would receive under the hypothetical year-end liquidation.
A more complex example shows the drafting problems that SEE can create: Suppose G and L enter into a real estate venture, under which G and L get their investment back first ($10,000 to G, $90,000 to L) proportionately, from distributions, and thereafter share further distributions 50-50. Times become bad, rents are down, so the partnership cannot satisfy its expenses, including servicing its mortgage. The partners find a savior (Savior 1), who invests $100,000 and become entitled, as a priority, to the return of his investment and an 8% interest equivalent, before G and L will get any distributions (under their earlier arrangement).
Business continues to be disappointing, so in order to save the project from foreclosure, the partners, now G, L, and Savior 1, find a new savior, Savior 2, who negotiates the following deal: Savior 2 gets his investment back first plus a 10% interest equivalent preference, before the existing distribution order among Savior 1, G, and L becomes operative, with the following modification: Savior 2 also becomes entitled to an additional 10% of all other distributions to Savior 1, G, and L, that is, any amounts that go out to those other partners would cause 10% of that amount to be distributed to Savior 2. None of this is particularly unrealistic. I, for one, would not want to have to draft the provisions necessary to implement this agreement under SEE, especially since I would have to provide assurances that the priorities outlined above and understood by the partners will be accomplished. As indicated earlier, when practitioners complain about the complexity of SEE, it tends to be in this context, i.e. implementing complex allocation structures. If I am using target allocations, on the other hand, my job would be easier, I “just” need to make sure the money goes out correctly, and then force the allocations to align the capital accounts correctly.
While more complex deals seem to give rise to target allocations, they can also make sense in some simpler deals. An example: Assume a partnership in which the general partner and the limited partner normally split profits and losses 20%-80%. But if certain performance standards are met, liquidation proceeds are split 30%-70%. Assume that all of the assets of the partnership are sold prior to liquidation. Under the economic effect test, depending on the facts, it may be impossible to allocate enough gain to the general partner to cover his liquidation proceeds. Thus, the current Regulations, if followed rigidly, could prevent the partners from implementing what is otherwise an entirely unobjectionable economic agreement. Target allocations, on the other hand, present no such problem. The distribution rights prevail. While target allocations try to get the capital accounts to match the liquidating distributions, they do not insist on it. Target allocations do not require partners to receive the balance in their capital accounts on liquidation. If the capital account balances do not match the partners’ liquidating distributions, as noted above, the partners will recognize gain or loss to the extent permitted under section 731 on the liquidation distributions. That said, for most simpler deals, SEE works well, is legally safer, and typically is the wiser approach to use.
Target allocations can also backfire, particularly for a service partner who receives a profits interest in the partnership in exchange for services. Normally the service partner does not recognize income on receipt of the profits interest. If that is the case, the capital account of the service partner normally starts at zero. If the intent is that the service partner receives, for example, 20% of the value of the future appreciation of the property held by the partnership, the targeted capital account approach may (depending upon how it is drafted) force allocations of other income to the service partner to build the service partner’s capital account up to the 20% level. This could occur because the appreciation in the partnership property has not yet been recognized, but an event in the partnership agreement requiring the service partner to be at the 20% level has been triggered. In this event, the service partner may receive income before being entitled to receive cash and thus will be required to use his own funds to pay tax on this “phantom” income. Drafter emptor.
Likely, most partnership tax practitioners use target allocations and are apparently comfortable that target allocations represent PIP. Other advisors rely upon the economic effect equivalence test to support target allocations. This reliance is decidedly dubious as it can only be justified if the partnership would have gotten the same outcome had it liquidated in accordance with traditional capital account balances. Yet one of the main reasons for using target allocations is to avoid liquidating in accordance with capital account balances.
There is nothing inherently offensive about target allocations in legitimate economic deals. Target allocations just provide an alternative—some would say superior—method for tracking the partners’ deal. Given how common target allocations have become, there is a good argument for adding a safe harbor for target allocations to the Section 704(b) Regulations. To my knowledge, there are no cases involving target allocations. There are two explanations for this, both of which could be correct: the pathetically low audit rate for partnerships, and the fact that a well-designed, legitimate target allocation should meet the PIP standard, making a Service challenge improbable.
I do not want to leave readers with the impression that target allocations are the promised land. In addition to the fact that target allocations fail SEE and are thus riskier to use, there is a lack of consensus on how to draft target allocations (or even what to call them). The Regulations have special rules for allocating nonrecourse deductions, i.e. deductions attributable to nonrecourse debt. It is not clear how target allocations work in this context. And, of course, there is zero authority supporting (or, admittedly, opposing) target allocations. Given how widely target allocations seem to be used, the Service should issue guidance in this regard.
Before I move on to the heart of this Article, I need to digress briefly to discuss allocations among related parties. The Section 704(b) Regulations are premised on the assumption that partners deal with each other at arm’s length. Related party allocations raise very different issues, which are not directly relevant to my proposals, and have been ably addressed by others. But inasmuch as the Wyden Proposals contains a proposal in this regard, some brief coverage is appropriate.
If the partners are closely related (e.g., two subsidiaries wholly owned by the same parent), they can agree to allocations that make little economic sense but save taxes for the related partners as a group. The Substantiality Regulations in their current form are ill-suited to deal with related-partner allocations. As a result, these regulations can easily be abused by related partners.
Example: Assume taxpayers A and B are equal partners in the AB partnership. A expects to be in the 50% tax bracket over the next several years. B, on the other hand, expects to be in the 15% tax bracket. The partnership earns both taxable and tax-exempt income. But, unlike the substantiality example discussed earlier, income is not predictable. If the partnership allocates the tax-exempt income to A and the taxable income to B, the allocations likely pass the substantiality test because, on a present value, after tax basis, it is likely that one partner will be better off and one worse off than if each received a 50% allocation of each type of income. Given this lack of predictability, normally neither A nor B would agree to this allocation if they were dealing at arm’s length. But what if A and B are related? Perhaps A and B are brother-sister corporations with identical owners. Now, since they are effectively one economic unit, whether A or B comes out better economically is no longer of great relevance, and A and B can focus on making allocations that save them the most in taxes in the aggregate.
Oddly, there is nothing in the Section 704(b) Regulations that definitively addresses this issue. Regulation 1.704-1(b)(1)(iii) states: “[A]n allocation that is respected under [the substantial economic effect rules] nevertheless may be reallocated under other provisions, such as section 482 . . . .” A similar statement is contained in an example in the Regulations. Thus, rather than stating that the allocation lacks substantiality, the Regulations punt the issue to section 482. That is in some ways understandable. Section 482 is designed to address tax issues that arise between related parties and allows the Service to reallocate tax items to prevent evasion of taxes or clearly to reflect the income. If section 482 and its brutally detailed Regulations could be counted on to stop A and B, then it could make sense to not also burden the Section 704(b) Regulations with parallel provisions. But this may not be the case. The lack of authority makes the analysis challenging, but in the view of some the critical issue under section 482 is whether B and C acquired their partnership interests on arm’s length terms; as long as they paid fair value for what they received, which can be resolved with appraisals, they may pass muster under section 482. Under this interpretation of section 482, it could be a straight-forward matter to comply and receive the preferred allocation.
That said, section 482 is broadly written:
…the Secretary may distribute, apportion, or allocate gross income, deductions, credits, or allowances between or among such organizations, trades, or businesses, if he determines that such distribution, apportionment, or allocation is necessary in order to prevent evasion of taxes or clearly to reflect income…
Query whether this language is broad enough to give the Service the authority to simply undo a special allocation and require proportional allocation of the different types of income. If that were the case, perhaps a parallel provision in the Section 704(b) Regulations would not be necessary. The problem is the uncertainty. At a minimum, the Service should provide specific guidance if it wishes to go down this path.
Could the Service choose to challenge the related-party allocation under the substantiality rules? Possibly not under the current Regulations, which in this context look to a “partner’s tax attributes.” The relatedness of the partners is not, as such, a tax attribute, but a business-law attribute. One could perhaps stretch the regulatory language to mean any attribute that can have a tax impact, but it would be just that, a stretch. Further, if the Service could successfully challenge the allocation under the substantiality rules, how would any reallocation occur under PIP? As long as the partners’ separate partner status is respected, it may be difficult to devise a sensible alternative. The PIP rules look to items such as the partners’ relative contributions to the partnership, the interests of the partners in economic profits and losses, the interests of the partners in cash flow and other non-liquidating distributions, and the rights of the partners to distributions of capital upon liquidation, though the Regulations note that there may be other relevant facts and circumstances. At least under the specified facts and circumstances, it might be difficult to reallocate between related partners as their existing allocations may well align with those factors. Note that PIP does not require that partners receive consistent allocations of all types of income and deduction. For the Service to be confident of its ability to challenge related party structures under the substantiality Regulations, Treasury would need to amend the Regulations, and then the Service would still be forced to address the issue on a case-by-case basis.
The Service could challenge the allocations under the Partnership Anti-Abuse Regulations, which give the Service very broad authority to restructure the transaction, including disregarding the partnership, not treating a purported partner as a partner, adjusting the partnership’s method of accounting, and reallocating income and loss. These Regulations have been heavily criticized, and their validity has been questioned. No case has invoked the Partnership Anti-Abuse Regulations, though these Regulations provide that it is a negative factor if substantially all of the partners (measured by number or interests in the partnership) are related (directly or indirectly) to one another. Even if the Partnership Anti-Abuse Rule provided the Service with a valid tool to address related party allocations, the Service would still be required to do so on a case-by-case basis. A better solution is to have a bright-line rule that restricts allocations among related partners.
Two somewhat similar proposals have been made to formally address the issue of related partners. The Wyden Proposals provide that if partners are members of a controlled group (within the meaning of section 267(f)) and together own 50% or more of partnership capital or profits, the partnership must consistently allocate all items based on partner “net contributed capital,” essentially based on relative capital contributions. Professors Cauble and Polsky have made a similar, more detailed proposal (looking at equity value rather than net capital contribution), but without the 50% threshold. The Wyden Proposals are to be commended for addressing the issues of related parties, but with regard to the threshold, the position of Professors Cauble and Polsky is more persuasive. Related-party allocations can be problematic even if the related parties own less than 50% of the partnership. The dollar values may still be large. The non-related partners may care little how allocations are made amongst the related partners. Further, how is the 50% threshold determined? Depending on the allocation structure, operations, and distributions made by a partnership, that might be challenging to reliably calculate. One can debate the other details, which I will leave to others, but what is required is a robust system that requires related parties to use fixed allocations.
That said, it should be noted that related party rules can overreach. For example, assume a father and son, both with substantial personal resources and in high tax brackets, want to each invest $1 million in a venture, with the understanding that the father would have preference as to both loss allocations and distributions. I am told fact patterns such as these are not uncommon. There is nothing inherently offensive about this structure, but it could run afoul of new related party allocation rules. One could well conclude that having related party rules apply to these less problematic structures is a price worth paying to stop the more abusive conduct of related parties. But even if one wanted to address the issue, drafting a formal exception could be challenging and create an unwieldy regulation. A better solution might be rule that would allow related parties to apply to the Service for an exemption in appropriate circumstances.
The many articles that have been written criticizing the section 704(b) regulatory regime are usually plagued by the same problems: a lack of understanding of how the real world operates and, or an apparent lack of understanding of the section 704(b) regulatory structure. This misapprehension can lead to unrealistic hypotheticals that are used to attack the current system and to misguided proposals for “reform.”
Here are some of the problematic examples in the literature (the example numbering is my own):
Example 1: “A and B each invest $50 in AB partnership. The partnership expends and deducts $100 in year one. It expects to earn $106 in year two. The entire $100 loss in year one and the first $100 of income in year two are allocated to A. The remaining $6 income in year two is split evenly between A and B. . . . As explained below, this may be valid under current law (though the regulations are unclear).”
Assuming the partners are acting at arm’s length, B would never agree to allow A to have the losses in year 1 unless B could not use them, and even then it would be unlikely, as commonly the loss would qualify as a net operating loss that B could carry forward indefinitely under section 172. Further, if B could be persuaded to agree because of his personal tax circumstance, the Regulations are clear – the allocation would be disallowed as transitory. One example in the Regulations is almost exactly on point.
Example 2: “AB partnership owns Blackacre, which produces $100 rent per year. A and B are equal partners. Ninety percent of the rental income is specially allocated to B in year one. One hundred percent of the rental income is specially allocated to A in year three. All income allocated is currently distributed.”
As with Example 1, A would normally not agree to this as a 50% partner, unless something else is going at the partner level that would make it rational to do so, and then it would again be a prohibited transitory allocation. The author did not address this issue or explain what happened in year 2 (presumably equal allocation). To be fair, the author did not claim that the example would pass muster, but instead was trying to show how the transaction could be similar to a loan. But why use an invalid allocation for this purpose, and why not address the transitory allocation issue?
Example 3: “ABC is considering an investment of $1,000 to develop a mine which will produce $160 income per year for nine years. ABC has sufficient net operating losses carried over from prior years so that it will not pay taxes for many years. ABC enters into a partnership with DEF, which has a 40% marginal rate, to develop the mine. ABC contributes $850 and DEF contributes $150. Under the partnership agreement, 100% of the mine development expense is allocated to DEF and 100% of the income is allocated to DEF until its capital account is restored to zero. Thereafter, 80% of the income is allocated to ABC and 20% is allocated to DEF. …. The allocations in [this example] are probably valid. They are modeled after example two of the § 704(b) regulations, which holds that a similar plan involving the disproportionate allocation of depreciation deductions to one partner with a gain chargeback does not violate the rule against transitory allocations.”
These allocations likely would violate the transitory allocation rules for the first five years of the deal but for the FMV-Book rule. As already discussed, the FMV-Book rule often lives at some distance from reality and should be eliminated. It is not clear from the example, however, if the author is specifically criticizing this rule. If so, I (and most) would heartily agree. It is a bit of an odd choice of facts, however. Mining most likely involves depletion more than depreciation. Unless percentage depletion were available, depletion deductions are not fixed, but based on how much mining is done. Depletion deductions are thus not as predictable as depreciation deductions and likely would make it harder for the proposed “tax dodge” to work. That said, the FMV-Book rule is not explicitly limited to depreciation and conceivably could be applied in the depletion context. Read generously, the example awkwardly makes a valid point—to wit, that the FMV-Book rule needs to go.
Example 4: “Taxpayer A is an equal member of Partnership ABC which produces $300 of income from the rental of its vacant land. A also possesses a $300 expiring net operating loss from a prior year in a different activity. The partners agree to allocate to A all of the partnership’s income for the year in return for A’s agreement to relinquish his rights to $300 of partnership income in subsequent taxable years.”
The authors rightly complain that this structure is abusive. What they do not appear to realize is that the Regulations agree. It clearly constitutes a transitory allocation and would not be allowed, yet this issue was not addressed with regard to the cited example.
Example 5: This example admittedly is very similar to one in the Regulations. “S and T form a general partnership solely to acquire and lease machinery that is 5-year recovery property under section 168. Each contributes $100,000, and the partnership obtains an $800,000 recourse loan to purchase the machinery . . . . The partnership agreement further provides that (a) partnership net taxable loss will be allocated 90 percent to S and 10 percent to T until such time as there is partnership net taxable income, and thereafter S will be allocated 90 percent of such taxable income until he has been allocated partnership net taxable income equal to the partnership net taxable loss previously allocated to him, (b) all further partnership net taxable income or loss will be allocated equally between S and T, and (c) distributions of operating cash flow will be made equally between S and T. The partnership enters into a 12-year lease with a financially secure corporation under which the partnership expects to have a net taxable loss in each of its first 5 partnership taxable years due to cost recovery deductions with respect to the machinery and net taxable income in each of its following 7 partnership taxable years, in part due to the absence of such cost recovery deductions. There is a strong likelihood that the partnership’s net taxable loss in partnership taxable years 1 through 5 will be $100,000, $90,000, $80,000, $70,000, and $60,000, respectively, and the partnership’s net taxable income in partnership taxable years 6 through 12 will be 40,000, $50,000, $60,000, $70,000, $80,000, $90,000, and $100,000, respectively.”
The author fairly notes that it is hard to fathom a business purpose to this structure. Again, this is an unlikely fact pattern. Why would T agree to a preferential allocation to S if T is putting up half of the money? Normally, T would not, and if T did, it would likely be because of tax considerations taking place at the partner level, which would raise substantiality issues. The Regulations hold that this allocation structure does not violate the transitory allocation rules, but only because the eventual offset happens after year 5. In reality, it would be very difficult to make reliable projections over 5 years out (or over 2 years, for that matter). Even in a longer-term equipment lease with a reliable lessee, where income and depreciation deductions are predictable, it would be difficult to be confident of the economics so far out. The equipment could fail, the lessee likely will have some rights to terminate the lease, etc. That said, at least this structure occurs, though it is uncommon, which may be why the Regulations address it. (An example might be a triple-net lease to a Fortune 500 company.) This example would be less objectionable if it discussed the potential substantiality issue and the improbability of the fact pattern, but those factors did not even get a passing mention beyond questioning the business purpose to the structure. That said, of the examples that I am covering, this might be the most defensible, the high rung of a low ladder.
As this discussion shows, it is difficult to create a real-world example that violates the substantiality rules, makes economic sense, primarily is not tax driven, and involves unrelated parties. That alone points to the viability of SEE.
One of the most common proposals is to require partnerships to have fixed allocation systems, such as those that exist for S corporations. These proposals all live at some distance from the real world. Commonly, deals do not present themselves in a way that would allow such a system to meet the partners’ needs.
A and B form a limited partnership to drill for oil and gas. A is the general partner; B is the limited partner. A contributes $10,000; B invests $990,000. The partnership expects to sustain losses in the early years, but then start to show substantial profits, though there are no such thing as assured profits from an oil well. This is a variation on the Brains-Money partnership. A is the brains partner, and B is the money partner.
Would B agree to invest if he did not get (on these facts) 99% of any early losses, or something similar? Highly unlikely. Would A bother to raise the money if she could only get a long-term interest of 1%? Again, highly unlikely. What kind of deal are the parties likely to strike? On these simple facts, a probable allocation system would be to allocate 99% of the losses to B and 99% of any subsequent income, until income equals prior losses. The other 1%, of course, goes to A. Once income equals prior losses there would be a flip, at which point A would receive a larger interest, perhaps 50%. Who else but the money partner, i.e., the person who funded them, should get the lion’s share of the losses? To allocate the loss to any other partner would give that person a windfall and likely a loss she could not fully use. But A at some point needs to get paid and would be. This allocation structure is unobjectionable both from economic and tax policy points of view.
It has been suggested that the important question is whether the partners would have made the same deal had there been no tax benefits. While perhaps a useful analytical tool, it would be impossible to know if an investor would have made an investment without the tax benefits because that alternative universe did not exist when the investment was made. There are few things in the business or investment world that do not have tax consequences. A preferable perspective is whether the allocations are primarily driven by the economics, as is the case with the example. If so, they should be allowed. Allocations primarily driven by the tax benefits should not be. How does one tell the difference? The Section 704(b) Regulations do so via the substantiality rules, for the most part successfully. But there is ample room for improvement, as I argue below.
It would not be possible to do Example 6 with an S corporation or under proposals for partnership allocation reform that adopt essentially the same system and make fixed allocations based on capital contributions. There is no fixed percentage in Example 6 that either the money or brains partner would agree to at the outset, as noted above. In literally a 2-person deal, it might be possible to live with an S corporation-like allocation structure, by giving the brains partner some kind of option to buy an increased partnership interest downstream. But the grant and, or exercise of the option can have tax consequences (unlike, typically, shifting allocations in a partnership), and as the number of players increases, the use of options becomes impractical in any event. An allocation structure that does not provide flexibility will mean that many legitimate deals will not get done, even when grounded in solid economics. Oddly, this real-world issue is not often discussed in articles that want to move to a fixed allocation structure. Perhaps the best evidence for the importance of flexible allocation structures is the fact that the average S corporation has about 1 ½ shareholders. If the S corporation allocation structure worked well in the business world, that number would be much higher.
This is an article on partnership allocations, not an article on how to restructure business entity taxation, generally. In something of an end-around, one thoughtful article proposes to tax all privately held businesses (including wholly owned corporations, limited liability companies, and unorganized sole proprietorships) at the entity level under a uniform rate schedule, regardless of the form of organization. Unlike many of the reform proposals for Subchapter K, this approach has merit. The income of all privately held businesses would be taxed the same, and the proposal would simplify business income taxation. But note that in this universe, Example 6 would be much less likely to occur. B would be much less interested in participating if losses did not pass through to him. It would significantly increase the relative cost of B’s investment. At a minimum, B would need a larger return, that in turn would make the deal harder to do, and fewer such deals would get done. I will leave for another article the questions of whether the benefits of the proposed system would outweigh these types of detriments.
The Wyden Proposals on partnership allocations are quite brief, so much so that I can quote the relevant portions verbatim:
The partnership tax rules afford tremendous flexibility in the allocation of partnership items among partners. The IRC and regulations provide two sets of rules circumscribing the allocation of partnership items – the “partners interest in the partnership” (PIP) standard and the SEE safe harbor. Both are based on the general principle of economic substance, and both are intended to align tax allocations with the underlying economic arrangement. However, the flexibility of current law has resulted in complexity for taxpayers and the IRS. The following provisions will substantially simplify the administration of partnership allocations and will as a result reduce taxpayer flexibility in this area, thereby curtailing abuse….The concept of SEE was added to the IRC to prevent abuse while preserving flexibility in the allocation of partnership items. However, the SEE regulations contain presumptions that can divorce tax and economics. The regulatory process has been unable to provide administrable rules that prevent tax-motivated allocations under the SEE standard. Moreover, neither the tax policy aims of simplicity nor administrability justify the disconnect between tax and economics. The safeguard itself has been the cause of complexity and proven difficult for the Service to properly audit and administer….The [Wyden Proposals] removes the SEE test for partnership allocations under Subchapter K and….requires that all partnership allocations be made in accordance with the PIP. PIP exists under current law and is based on the facts and circumstances of the economic arrangement (e.g., each partner’s contributions and rights to distributions). It is expected that the Secretary will issue updated and simplified regulations addressing PIP. …The provision will remove optionality of current law, better prevent the shifting of tax attributes between partners, simplify the rules governing partnership allocations, and allow the Service to better focus audit and enforcement efforts.
Were it only this easy. It is not, and if implemented, the proposal likely could prove unenforceable. The proposal seems to suggest that SEE is more flexible than PIP, which is roughly backwards. The more flexible standard is actually PIP, which is why partners rely on it when using target allocations. The SEE rules restrict flexibility; they do not enhance it. They can be highly complex, but as I have noted, the more a deal is based on genuine economics the less likely it is that this complexity will pose a burden.
The Wyden Proposals quoted above state, “the SEE regulations contain presumptions that can divorce tax and economics.” And there is one such presumption, the FMV-Book rule, which admittedly needs to go. But this hardly provides a basis for wholesale change.
As the Wyden Proposals also note,
[t]he regulatory process has been unable to provide administrable rules that prevent tax-motivated allocations under the SEE standard. Moreover, neither the tax policy aims of simplicity nor administrability justify the disconnect between tax and economics. The safeguard itself has been the cause of complexity and proven difficult for the Service to properly audit and administer.
But the SEE rules are, in fact, administrable. They provide a mostly sensible set of rules that in many deals are straight-forward to apply. As the deals get more complex with differing classes of partners and waterfalls, administering the SEE rules admittedly gets much harder for the Service, but that administrative task is not wholly out of reach. What is true is that the Service has not properly administered and audited SEE, not because it cannot, but because it has lacked the resources to do so. When the Service has chosen to litigate allocation issues, it has had a good success rate. Indeed, the Service has chosen to litigate some highly complex cases, something that would hardly occur with rules that are not administrable. What would be fair to say is that the Service cannot properly administer SEE with current staffing levels, but that is an argument for increased staffing, and perhaps having a separate partnership tax division, not for the wholesale elimination of the rules that have been used for decades. Partnership tax advisors usually have a solid understanding of how SEE operates, and countless partnership agreements rely on SEE. That admittedly does not mean that SEE is a perfect set of rules, and I will propose ways they can be improved without throwing the baby out with the bathwater.
The Wyden Proposals state that “[i]t is expected that the Secretary will issue updated and simplified regulations addressing PIP.” But, it is not possible (as I noted above) to definitively define PIP. There are just too many variables. What it is possible to do is to create a safe harbor, currently SEE. But if we eliminated SEE, we would need a new safe harbor. What would it look like? It would probably look a lot like SEE. The Service could perhaps replace SEE with target allocations. But that would not make the Regulations meaningfully simpler. Given how long SEE has been around and on how widely it has been used, it makes more sense to add a safe harbor for target allocations than to eliminate SEE. That step admittedly would make the Regulations more extensive and to that extent more complex, but given how widely target allocations apparently are being used, it would not likely add a significant additional burden for practitioners. And, target allocations may be easier to administer than SEE for the same reason that practitioners often prefer them, perhaps making the addition of a target allocation safe harbor a net win for the Service and taxpayers. The bottom line is that it is naïve to think that any replacement safe harbor would be significantly simpler than SEE. The reason for much of the complexity in SEE is to prevent abuse, hence the existence, for example, of the separate substantiality test. Any replacement safe harbor would also need anti-abuse rules and would have a comparable level of complexity.
The last thing anyone concerned with administrability would want is a PIP standard without a safe harbor. Some practitioners would see that as an opportunity to run with the ball, while others would hate the lack of specific guidance. The Service auditing personnel could be overwhelmed, forced to make case-by-case judgment calls. It could also be a challenge for the courts. It is true that the courts have done a decent job with PIP to date, but PIP cases have not been that common. If PIP became the only rule, there would be far more cases, many in Federal district court where the judges often lack tax expertise. Case law likely would be all over the map. Over time, the courts and the Service likely would create de facto safe harbors to ease decision making. They might even create competing de facto safe harbors. A de jure safe harbor makes infinitely more sense.
Finally, the Wyden Proposals state that “[t]he provision will remove optionality of current law, better prevent the shifting of tax attributes between partners, simplify the rules governing partnership allocations, and allow the Service to better focus audit and enforcement efforts.” In light of the discussion above, it seems unlikely that the proposed changes would achieve any of these objectives.
Why do the Wyden Proposals fall short? I do not have a definitive answer. As I have discussed elsewhere, part of the problem may be that the Wyden Proposals see the world through the prism of large (or even very large) partnerships. These partnerships appear to currently rely mostly on PIP, albeit through the use of target allocations. Perhaps this reliance on PIP caused the Wyden Proposals to focus on PIP as well, though it is perplexing why they would adopt PIP without a target allocation safe harbor. The sophisticated counsel larger partnerships typically have may have made the Wyden Proposals less concerned about any burdens rule changes might have on practitioners. As noted earlier, and as is true for large businesses generally, these large partnerships do own the lion’s share of partnership assets, but smaller partnerships remain important players. But Regulations designed only with large partnerships in mind would do the partnership tax world a large disservice. There is no hard data on how many partnerships use SEE versus target allocations, but I would not be surprised if the vast majority of smaller and mid-sized partnerships still use SEE. Eliminating SEE would be a nightmare for them. It might be possible to grandfather SEE in some fashion but that would make the partnership tax world more complex, not less so. The bottom line is that eliminating SEE makes no sense. What does make sense is to improve it, as I will discuss next.
I have three proposals for reforming the Section 704(b) Regulations:
The present value post tax rule is awkwardly formulated. The economic effect is not substantial if there is a strong likelihood that after-tax economic consequences of no partner will, in present value terms, be substantially diminished. Some see this as a triple negative. “Strong likelihood” lives close to “near certainty” and creates an unnecessary pro-taxpayer bias. We have come to learn the hard way that aggressive planning accompanies liberal substantiality rules. I propose that we reformulate and rein in the present value post tax test as follows:
Where it is probable that an allocation causes a partner’s economic consequences to be significantly enhanced, the economic effect of the allocation is only substantial if it is probable that another partner’s economic consequences will be significantly diminished. For this purpose, the economic consequences shall be determined on a present value, post-tax basis. In making this determination, the proposed allocation shall be compared to such consequences as would apply if the allocation were not contained in the partnership agreement. Any alternative used for this comparison need not be the least tax efficient provided it is driven primarily by economic realities. If a partnership agreement allows for allocations to vary, notice of this fact and a general description of the allocation must be provided to the IRS. In addition to any other penalties that might apply, an additional 10% penalty shall apply to any increased tax that arises when a partnership is found to have made an allocation that lacks substantiality, unless there is substantial authority for the partnership’s position. In defending against the penalty, the burden of proof is on the partnership and its partners.
The proposed language makes the present value post tax more neutral, moderately lowering the bar for finding that the economic effect of an allocation is not substantial. Instead of a strong likelihood, it needs to only be probable that another partner’s economic consequences are not significantly diminished. But this analysis need only be made if it is also probable that the allocation causes another partner’s economic consequences to be significantly enhanced. Thus, an allocation that might be of modest benefit to a partner might not trigger the analysis. But if the partnership is gaming the tax system, it should not be a heavy lift to prove it. There inevitably are still uncertainties. Whether economic consequences are significantly enhanced or diminished is a fact question that would have to addressed on a case-by-case basis, but it is a more workable standard (and I would argue a more understandable standard) than the current definition of substantiality. If an allocation is disallowed, my proposal does not eliminate the difficulty that can arise in searching for an alternative allocation but attempts to make this job easier and more reliable by providing that the alternative need not be the least tax efficient but must be economically driven. But a primary objective of my proposed substantiality provisions is to keep partnerships from gaming the tax system in the first instance. The proposed tightening of the definition of substantiality should reduce taxpayers more aggressive tendencies. Further, by requiring disclosure, applying an additional penalty, and having the burden of proof with regard to the penalty be on the partnership (it normally is on the Service), my hope is that the “intimidation effect” will keep partnerships on the straight (if perhaps not narrow) path, lowering the administrative burden on the Service. In the interest of fairness, however, I do not provide for strict liability, but allow the partnership to avoid the penalty upon a showing of substantial authority for the partnership’s position.
Recall, that the present value post tax test applies even if the shifting or transitory allocation tests also apply. While I would keep both of those latter tests as a backstop, they are likely to be less important in light of the reformulation of the first test. Nonetheless, I would tighten the transitory allocation test. Instead of requiring that the offset happen “in large part” within the first five years—another example of excessively liberal rules, I propose to have it happen in a “significant amount” within five years and would define that as 50% or more of the initial allocation(s). This brighter-line rule is fair and should make administration easier.
A significant problem with section 704(b) allocations is the extent of their application, which is pretty much the waterfront of partnership income and expenses, subject to the substantiality rules. Most infamously, depreciation may be allocated to one partner and be offset by gain allocated to that same partner even within the 5-year transitory limit, due to the FMV-Book rule. Repealing the FMV-Book rule would be a step in the right direction, but an insufficient one. As long as most expenses can be allocated in virtually any way, the temptation will be great to look for ways to game the system. At current partnership audit rates, that temptation is still greater. And even if the Service were able to do audits at a robust level, which most would argue would be a good thing, it would often still be playing catch-up, and would have to dedicate substantial resources to section 704(b) enforcement. I propose, therefore, to amend section 704(b) to provide that a partnership may only allocate partnership taxable income, taxable loss, and net tax-exempt income (in the partnership universe sometimes known as “bottom-line” items). Tax credits may be allocated in the same manner that taxable income or loss is allocated.
It would no longer be possible to specially allocate constituent parts of bottom-line items such as depreciation, and thus the FMV-Book rule and the challenges of dealing with gain chargebacks would be mooted. This approach would significantly simplify the allocation universe while allowing legitimate economic deals to get done. The administrative burden on the Service should be reduced. The more tax driven portions of the sale-leaseback universe would likely see their size reduced, a good thing in light of all of the litigation and challenges this area has produced. At the same time, the Brains-Money deal remains viable.
While relevant hard data and case law are lacking, it stands to reason that the more flexibility section 704(b) provides, the more motivation sophisticated, wealthier partnerships will have to push the envelope. Small and medium-sized businesses are unlikely to engage in these more sophisticated transactions. The dollars involved would tend to make them less necessary, and they likely could not afford the pricy tax counsel necessary to implement them. Accordingly, the biggest impact of the proposed change likely will be the small percentage of large, “elite” partnerships, not a group most would say is worthy of special consideration in the tax code. By making life for these larger partnerships a bit harder, life for the Service audit arm will be made much easier. The large partnerships will have to increase their focus on solid economic deals, hardly a bad thing.
Under my proposal, it would still be possible to have different classes of partners with different preferred returns. Accordingly, some of the drafting challenges that currently exist with SEE could still arise. Target allocations has become the preferred the solution, and there is nothing inherently objectionable about them. Any revised Regulations should contain a target allocation safe harbor, addressing the uncertainties discussed earlier. If we were writing on a clean sheet, there might be an argument for having target allocations as the only safe harbor, but we are not. Too much water has gone over the dam to make wholly repealing SEE a viable option. Crafting a safe harbor for target allocations may prove challenging. As they have, in a sense, been operating in the shadows, different practitioners may approach target allocations differently. It may be no simple matter to achieve wide consensus for any particular safe harbor, but given the apparent prevalence of target allocations, it is a challenge worth undertaking.
Unless we move to an across-the-board entity level tax, Subchapter K is needed and is here to stay. The proposals that would throw the baby out with the bathwater and to a greater or lesser extent repeal Subchapter K are political nonstarters and ignore how deals present themselves in the real world. My proposal, on the other hand, is aligned with how the real world operates, while also seeking to modernize and reform SEE. It would bring target allocations formally into the fold and make SEE less capable of abuse, while permitting legitimate economic deals to take place. I believe its incremental approach makes it politically viable. There is no substitute for a dramatic increase in the Service’s enforcement of Subchapter K, which has been woefully lacking due to egregious underfunding. But under my proposal, the burden of an appropriate level of enforcement should be less than it would be under the current rules. The Wyden Proposals may have fallen a bit short on reforming partnership allocations, but they are to be commended for seeking to reform and bring attention to Subchapter K. Perhaps they will be the catalyst that is needed to bring comprehensive reform to Subchapter K, the necessity of which is hard to dispute.
I would like to thank the following people for their invaluable comments, suggestions, and feedback: Prof. Fred Brown, Prof. Karen Burke, Monte Jackel, Esq., Robert Rombro, Esq., and the members of the Mannes Greenberg Tax Society.
