Menu
Tax Notes logo

The Heart of the BEAT

Posted on Apr. 29, 2019
[Editor's Note:

This article originally appeared in the April 29, 2019, issue of Tax Notes.

]
David P. Hariton
David P. Hariton

David P. Hariton is a partner at Sullivan & Cromwell LLP.

In this report, Hariton analyzes the ambiguities in the base erosion and antiabuse tax statute and explains how Treasury could resolve them.

The base erosion and antiabuse tax was drafted and enacted in less than three months.1 Nevertheless, it is novel and complex in both its structure and its application. And it presents novel questions of interpretation, not only with respect to the concrete meaning of its abstract terminology in specific contexts, but more broadly with respect to what constitutes permissible, or problematic, planning and results.

In general, concerned parties (Treasury, the IRS in its guidance-providing capacity, taxpayers, their internal and external auditors, their inside counsel, their outside counsel rendering advice and opinions, the IRS in its audit capacity, and ultimately the courts) seek to resolve outstanding ambiguities in light of what Congress sought to achieve in enacting the relevant statute. In this case, however, that light is rather dim. The legislative history is brief and vague, and not wholly consistent with the statute as drafted. And Treasury appears to be holding itself on a relatively short leash as regards its capacity to promulgate guidance in a quasi-legislative capacity.2 One of the objectives of this report is to join the chorus of voices encouraging Treasury to be more forthcoming in helping both Congress and taxpayers out. For Congress simply didn’t have time to think the BEAT through very thoroughly.

The Birth of the BEAT

The first public evidence of the incipient BEAT was in that terse framework for tax reform that was offered to the public by the Big Six on September 27, 2017.3 The very last sentence of that document contained what was apparently the BEAT in embryonic form: “The committees will incorporate rules to level the playing field between U.S.-headquartered parent companies and foreign-headquartered parent companies.”

At the time, no one was quite sure what that meant. The implied reference, however, was to the ongoing debate about “inversion transactions,” in which a U.S. parent corporation reorganizes itself as a foreign parent corporation primarily for U.S. tax reasons.4 One of those U.S. tax reasons (apparently not of concern to a Congress that was preparing to enact a territorial system) was to take growing foreign subsidiaries (and their associated foreign operations) out from under the U.S. tax net. The other, however, was to engage in some form of earnings stripping designed to erode the U.S. tax base — for example, to thereafter replace inbound U.S. equity capital with related-party debt, giving rise to outbound related-party interest deductions. Some of the participants in that debate asked, “Well, how does that differ from what any foreign parent of a U.S. corporation can currently do, inversion or no?” and the answer was, “It doesn’t.” Hence the conclusion not that inversions had to be stopped, but more broadly that “the playing field between U.S. and foreign-headquartered corporations had to be leveled.”

Now, of course, at the time the Tax Cuts and Jobs Act was enacted, we already had an earnings-stripping provision — to wit, section 163(j). Section 163(j) was, of course, expanded into a general limitation on tax-deductible leverage (that is, a limit on deductions for interest paid to completely unrelated U.S. parties, including bondholders and banks) applied (inexplicably) even to individuals and partnerships.5 But it continues to limit deductions for interest paid to related foreign parties, and with a lower threshold (such deductions will not be able to exceed 30 percent, rather than 50 percent, of otherwise taxable income).6 It is therefore not clear how much Congress is relying on the BEAT to defend the U.S. tax base from earnings stripping arising from outbound payments of interest.

The ensuing legislative history suggests, however, that Congress was concerned with more than just deductible interest payments: “Foreign corporations often take advantage of deductions from taxable liability in their U.S. affiliates with payments of interest, royalties, management fees, or reinsurance payments. This provision aims to tax payments of this kind. This type of base erosion has corroded taxpayer confidence in the U.S. tax system.”7 But Congress did not have time to focus very sharply on its objectives, and so neither its discussion of other deductible payments in the legislative history nor the statute that it enacted as a result seems to fully parse with respect to them. For example, Congress purported to be particularly concerned with the use of deductions for royalty payments to strip taxable income out of the United States. Yet the statute that it enacted allows deductions for those payments to be capitalized into the cost of goods sold, thereby reducing gross sales revenue, rather than leaving a deduction to which the BEAT can be applied. The statute therefore fails to protect the fisc from this sort of base erosion.

In any case, what legislative history there is focuses primarily on the ability of foreign-headquartered corporations to use related-party payments to erode the U.S. tax base. If, therefore, we are going to interpret the statutory language in a manner that comports with the purported objectives of Congress, we must begin by reminding ourselves of what base erosion is. Base erosion is not something that occurs whenever a U.S. corporation makes a payment to a related foreign affiliate. Base erosion is what occurs when a foreign parent corporation maintains less equity capital in its U.S. affiliates than it would have maintained absent a motivation to reduce U.S. taxable income. That is why the heart of the base erosion concern has always been that foreign-headquartered corporations will capitalize their U.S. subsidiaries with debt, rather than equity, in order to reduce the U.S. taxable income of the latter through the payment of deductible interest expense (hence old section 163(j)). It is true that a foreign corporation can also use royalty payments to erode the U.S. tax base. But the base erosion concern here is not that the U.S. affiliate is paying too much for the use of the foreign parent’s intellectual property. That is a section 482 concern. The base erosion concern is that but for tax considerations, the IP in question would have been developed partly or wholly in the United States, or would otherwise have been contributed to the capital of the U.S. affiliate. And similarly in the case of management services, the base erosion concern is not that such management is overpriced, but rather that it would otherwise have been conducted in the United States.

In other words, there is no sense in which a deduction for a payment made to a related foreign party can be said to be eroding the U.S. tax base if the taxpayer would otherwise have had to make a similar payment to an unrelated party, else the relevant taxable income would never have materialized. This is presumably the basis for the cost of goods sold exception that stands out so prominently in the BEAT.8 If a U.S. affiliate purchases widgets manufactured by its foreign parent for an arm’s-length price of $100 and sells them in the United States for $120, thereby earning a $20 profit, there is no base erosion going on. There is no reason to suppose, for example, that but for tax considerations, the widgets would have been manufactured in the United States at significantly lower cost. And if the $100 purchase price is too high and should have been only $80, that is not a base erosion issue, but rather an arm’s-length pricing issue.

Whatever you think about the efficacy of our transfer pricing rules, the above analysis doesn’t change when we move from the purchase of merchandise for sale to the purchase of depreciable property. If, for example, a U.S. affiliate installs a cappuccino vending machine in the congressional cafeteria, it really doesn’t matter whether it purchases this machine from an unrelated party or from a related foreign affiliate. It seems silly to hypothesize that but for tax considerations, the foreign affiliate would have contributed the vending machine to the capital of the U.S. affiliate. And there is no sense in which the U.S. fisc is more vulnerable to non-arm’s-length pricing here than it is in the case of the widgets.

Nor does the analysis change when we move on to services. It may be possible to hypothesize that a U.S. subsidiary’s management would otherwise have been located in the United States. Apart from this, however, it is not sensible to hypothesize that a foreign parent would otherwise have contributed the services of its foreign employees to the capital of its U.S. affiliates.

Moreover, transactions with foreign affiliates obviously cannot give rise to base erosion unless they serve to reduce the U.S. taxable income of the relevant U.S. affiliates. It seems nonsensical to impose BEAT in respect of related-party transactions that serve to increase the U.S. taxable incomes of relevant U.S. affiliates because U.S. affiliates include as much, or more, in income in respect of their receipts from foreign affiliates as they deduct in respect of their payments to them. In such a case, the relevant worldwide affiliated group merely happens to be engaged in interrelated business operations. That is why old section 163(j) applied (and still applies) only to net interest expense — the excess of interest expense over interest income. Put differently, you can’t have a tax on earnings stripping unless you reference the stripping of earnings.

Yet as Treasury has observed in the preamble to its proposed regulations, “The proposed regulations do not provide a rule permitting netting in any of these circumstances because the BEAT statutory framework is based on including the gross amount of deductible and certain other payments (base erosion payments) in the BEAT’s expanded modified taxable income base without regard to reciprocal obligations or payments that are taken into account in the regular income tax base, but not the BEAT’s modified taxable income base.”9 As we shall shortly see, the BEAT’s inexplicable lack of any sort of netting rule causes it to be both over- and underinclusive in apparently random ways.

Moreover, the BEAT as enacted applies equally to deductions for payments made by U.S.-headquartered corporations to their foreign subsidiaries! The only rationale for this appears to be a brief paragraph referencing the outsourcing of services to foreign jurisdictions where labor is cheaper than it is in the United States.10 Yet the statute seems to belie the notion that preventing outsourcing was one of its objectives, as it expressly excludes from its ambit deductions for the provision of routine services that are eligible for the so-called services cost method.11 And these are precisely the sorts of services that are outsourced by U.S. corporations to foreign affiliates located in countries where labor is cheaper. In any case, this loaded political issue has nothing to do with base erosion and earnings stripping. Many such foreign jurisdictions are high-tax countries, and no one is trying to avoid taxes by hiring employees there. And even if preventing such outsourcing were one of the BEAT’s objectives, that wouldn’t explain why the statute applies to payments of interest by U.S.-headquartered corporations to their foreign subsidiaries.

There is, in fact, a potential rationale for extending the BEAT to the foreign subsidiaries of U.S.-headquartered corporations, even though it isn’t to be found in the legislative history. Congress simultaneously adopted a territorial system, and in a pure territorial system, foreign subsidiaries are equally capable of stripping earnings from their U.S. affiliates, as their earnings will never be subject to U.S. tax. But in point of fact, Congress did not adopt a pure territorial system, and so most of the interest, royalty, or other payments that could conceivably have served to transfer the earnings of U.S.-headquartered corporations to their foreign subsidiaries are included in the former’s U.S. taxable income again as either subpart F income or global intangible low-taxed income. The rationale for extending BEAT to payments to foreign subsidiaries therefore remains something of a mystery.12

Finally, the BEAT does not provide for an alternative foreign tax credit, as the alternative minimum tax does under section 59(a). Rather, the benefit of FTCs may be lost entirely, since they are backed out of regular tax liability for purposes of determining whether alternative BEAT liability applies. This result is obviously not directed at base erosion. Consider, for example, a U.S. corporation that has relatively few deductions for payments to related foreign parties (but enough to be over the 3 percent base erosion percentage threshold described below) and that earns primarily subpart F income subject to an 18 percent rate of foreign tax. After applying the FTC, this corporation’s residual regular U.S. tax liability will be relatively low. It will therefore pay the BEAT, rather than regular tax, at a rate of 10 percent or 11 percent. Thus this aspect of the BEAT appears to be in the nature of a minimum tax on foreign-source income, one that is imposed in addition to the taxes on GILTI and subpart F income, further eviscerating the purported territoriality of our new international tax system. Yet there is nothing in the legislative history that fleshes out the relevant concern in a coherent and comprehensive manner.13 A similar analysis applies to other credits, including credits for energy and low-income housing, although a last-minute deal in Congress provided that regular U.S. tax liability is reduced by only 20 percent of the amount of such credits until 2026.

The Structure of the BEAT

As drafted, the BEAT appears to be an overly broad tax applied (by reference to a novel “base erosion fraction”) to an overly narrow group of taxpayers. The two wrongs may sometimes make a right — but not always.

Where it does apply, the BEAT is structured as an AMT that replaces the old corporate AMT. Indeed, taxpayers sometimes mistake the BEAT’s acronym as standing for “base erosion alternative tax.” It is imposed at a lower rate (currently 10 percent or 11 percent, depending on whether or not the taxpayer’s expanded group has a financial institution) on a broader base (modified taxable income), but the “items of preference” generally include deductions with respect to payments to related foreign parties.

As noted above, the first manifestation of overbreadth is that unlike old section 163(j), the BEAT fails to net offsetting inclusions and deductions, even in cases where they run between the same parties, and even in cases where they arise from the same transaction. Needless to say, the results are illogical, and Congress may well have been assuming that Treasury would provide for netting in appropriate cases. But Congress neglected to say so, and Treasury now claims to lack the authority to do so.14

Many multinational corporations routinely make cross-border payments in both directions between affiliates in the course of managing their cash flows. And financial institutions have a massive amount of inter-affiliate payments running across the border in opposite directions, arising from a broad range of offsetting financial transactions. Current law tends to characterize most of these payments as giving rise to offsetting inclusions and deductions, although they could just as well give rise to neither and could perhaps be restructured to do so. But in any case they do not give rise to any net deduction and therefore do not result in any base erosion. And they doubly don’t result in any base erosion when paid to a foreign subsidiary.

As discussed further below, the statute does contain a second-best solution for financial transactions, in the form of an exception for payments made pursuant to “qualified derivatives.” This appears, however, to have left taxpayers and Treasury to conduct esoteric debates about whether the transactions that give rise to such offsetting cash flows are derivatives per se, a debate that is obviously beside the point as a matter of policy. As old section 163(j) demonstrates, nothing could be more deserving of exemption than offsetting interest payments in respect of offsetting loans running in opposite directions. Obviously, the only interest expense capable of giving rise to base erosion is the net interest expense — that is, the excess of interest deductions over interest income inclusions, as in old section 163(j). And yet offsetting loans are not derivatives, and so the offsetting payments of interest are not exempt. And so far (as discussed further below), Treasury seems to be headed in the direction of cutting back on the derivatives exception in order to defend the inappropriate treatment of offsetting loans, rather than doing the opposite. More on this later.

A similar problem arises in connection with the provision of services. It is true that a foreign parent corporation could seek to erode the U.S. tax base by charging its U.S. affiliates for the provision of home office services that could have been performed in the United States. As noted above, however, if the services do not relate to management and could not have been performed in the United States, there can be no coherent base erosion. And there certainly cannot be any in cases where the U.S. affiliates are receiving as much in payments from their foreign affiliates as they are making to them in respect of the performance of services. Nevertheless, the statute treats any deduction for amounts paid to related foreign parties as an item of preference on a gross basis (unless it qualifies for the costs services exception). Thus, worldwide affiliates that perform services for each other in the ordinary course of running a worldwide business are exposed to the imposition of BEAT under circumstances having nothing to do with base erosion.

Another broad aspect of overinclusion is the fact that the statute treats deductions for the depreciation of property purchased from related foreign parties as items of preference to be added to modified taxable income. As noted above, however, assuming that the relevant purchase price is determined at arm’s length, there can be no meaningful difference between the purchase of such property from related versus unrelated parties. In both cases, the property is purchased because it is anticipated to increase net income through its productivity more than it decreases net income through its depreciation cost. Indeed, a foreign affiliate may be better at buying such property for the right price in its local market and on-selling it to the U.S. affiliate. As likewise noted above, moreover, a concern about arm’s-length pricing is not a base erosion concern. The analysis here does not differ from the one that resulted in the cost of goods sold exception.

At the very least, moreover, there should be a netting provision. For even if one hypothesizes that a foreign parent is somehow failing to increase the capital of its U.S. subsidiaries by failing to contribute depreciable property to them, it cannot be doing so if it is buying as much property from its U.S. affiliates as it is selling to them.

Similarly, the statute does not provide for an alternative net operating loss carryforward, as did the corporate AMT. As a result, the regular alternative NOL carryforward (hair-cutted under the statute by the “base erosion percentage for the year,” but with no indication of which year Congress was referring to) could be deemed to be fully available to offset modified taxable income without regard to whether it has already been used to offset modified taxable income in prior tax years. Rather than helping Congress out by clarifying how the amount of an NOL carryforward is modified for purposes of applying it to modified taxable income,15 Treasury has proposed what seems like an overly broad solution of denying taxpayers the right to use NOL carryforwards altogether to offset the portion of modified taxable income that is attributable to base erosion tax benefits.16

The Base Erosion Percentage

Apparently to make up for some of this overbreadth, the BEAT further provides that it does not apply at all if the taxpayer’s base erosion percentage is less than 3 percent (or 2 percent, in the case of financial institutions). The base erosion percentage, in turn, is defined as all of the taxpayer’s base erosion tax benefits (that is, deductions arising from payments made to related foreign parties) for the tax year divided by all of its deductions for the tax year. The disconnects described above go through the fun house mirror when it comes to applying this fraction.

For one thing, this fraction obviously magnifies the problems arising from the lack of a netting provision, for the treatment of gross payments running in opposite directions as giving rise to base erosion tax benefits can quickly add up to enough deductions in the numerator to take the taxpayer’s base erosion fraction over 3 (or 2) percent. But of course, what is sauce for the numerator is sauce for the denominator: The absence of any netting provision invites U.S. companies to opt out of the BEAT by disaggregating their payment relationships with unrelated parties and thereby including large amounts of deductions in the denominator of the fraction. As a result, the application of BEAT may turn less on the relative significance of the taxpayer’s transactions with related foreign parties than on the degree of its flexibility in structuring transactions with related and unrelated parties respectively. More on this later.

A key interpretive question here is whether Congress viewed the base erosion percentage as (1) a limitation on the amount of deductions arising from payments to related foreign parties that U.S. affiliates may take without paying BEAT, or (2) a rough-justice means of distinguishing U.S. corporations that are suspiciously involved with foreign affiliates from other domestic corporations that don’t deserve to be burdened with all of this added complexity. For if we assume the former, then Treasury should have no problem when taxpayers restructure their affairs or take tax positions designed to reduce their base erosion percentages below the threshold for BEAT imposition. They are merely doing what Congress asked them to do: making sure that (or otherwise demonstrating to Congress and Treasury that) they are not using deductions arising from payments to related foreign parties to lower their U.S. taxable incomes to any significant degree. But if we assume the latter, then perhaps such suspicious U.S. taxpayers should not be allowed to simply turn their backs on a deduction in any given year and walk away from the BEAT. More on this later as well.

In any case, the technical question of what is a base erosion tax benefit is greatly magnified in this context. In effect, it is levered up 33 to 1 (or 50 to 1 in the case of a financial institution) in significance. Indeed, it is obviously the key technical question when it comes to analyzing and applying the BEAT.

The Definition of Base Erosion Tax Benefits

As noted above, the BEAT is effectively imposed on base erosion tax benefits.17 Moreover, the BEAT only applies to begin with if the taxpayer has enough base erosion tax benefits in the numerator of its base erosion fraction. Obviously, then, the application of BEAT turns, as a technical matter, on the existence vel non of base erosion tax benefits, and the heart of the associated uncertainty in the application of the BEAT lies in the definition of that phrase.

And the key point, when it comes to this definition, is that a payment to a related party is not itself a tax benefit. Thus, what is added to modified taxable income (or to the numerator of the base erosion fraction) is not payments, but rather deductions “with respect to” payments (to related foreign persons). We therefore need to know first whether we have a deduction, and second whether that deduction is “with respect to” a payment to a related foreign person.

Yet in this regard, the statute’s definition of a base erosion tax benefit is circular and offers little help. On one hand, section 59A(c)(2) defines a base erosion tax benefit as a deduction “with respect to” any base erosion payment.18 This sends us scurrying off to the definition of a base erosion payment, only to discover that this is defined, in section 59A(d)(1), as an amount paid or accrued to a related foreign person “with respect to which” a deduction is allowable.19 The phrase “with respect to” is itself rather vague. Moreover, while we probably know when cash has been “paid to” a related foreign person, any taxpayer that is potentially subject to the BEAT will invariably be an accrual-basis taxpayer, and it is not clear when or how a taxpayer accrues a deduction “to” a related foreign person.

Let me offer you a few examples to show that I’m not just splitting hairs. In the first example, a U.S. taxpayer, a nonfinancial corporation, invests in 100 shares of IBM stock. The stock goes down in value, and the U.S. taxpayer sells the stock at a loss to a foreign person that is related for purposes of applying the BEAT, but not for purposes of applying section 267. A deduction for the loss is therefore allowed for regular tax purposes. Is this a deduction “with respect to an amount paid or accrued to a related foreign person”? The logical answer would seem to be no. The cash payment is not made to, but rather received from, a related foreign person. And the deduction relates to the realization of a loss that developed while the U.S. taxpayer held the relevant property — that is, it had nothing to do with any relationship between the U.S. taxpayer and the related foreign person. It certainly has nothing to do with earnings stripping. Put differently, the loss would equally have been recognized and deducted had the taxpayer sold the stock to an unrelated person.20

In the second example, a U.S. taxpayer enters into a forward contract to sell a share of IBM stock to a related foreign person one year from now for $100, its current market price. One year later, the U.S. taxpayer purchases a share of IBM stock in the market for $150 (its then current price) and delivers it to the related foreign person. The U.S. taxpayer deducts a $50 loss in respect of the resulting closing of its short position. Is this a “deduction with respect to an amount paid or accrued to a related foreign person?” As in the long position example above, the U.S. taxpayer’s loss does not relate to any base erosion, but rather solely to a change in the value of the referenced property while the taxpayer was exposed to it. Moreover, the only payment here is made to an unrelated person in the market. On the other hand, the payment that gives rise to the loss (that is, the $150 payment to an unrelated person to acquire a share of IBM stock) would never have been made (and so no loss would ever have been deducted) but for the taxpayer’s obligation to deliver a share of IBM stock to the related foreign person. In other words, unlike the long position example, the U.S. taxpayer could here not have deducted the loss by closing out a short position with an unrelated party, for its position with the related foreign party is the only short position it has.

Neither the statute nor the proposed regulations offer any help here. How do you come out on the second example? Take a moment to decide, and then let’s move on to some companion alternatives. The third example is the same as the second, except that before the taxpayer closes out its short position, it marks its short position to market under section 475. This means that it treats the position as sold (presumably to an unrelated person) and immediately thereafter repurchased, which presumably means in the case of a short position that the taxpayer is deemed to have paid someone else $50 to assume its underwater short obligation and then reentered into a new short position at market. The taxpayer therefore deducts a $50 loss. I would imagine that in this case, you might feel a bit more encouraged to take the view that the loss deduction is not “with respect to” an amount paid to a related foreign person. Indeed, you may well have been persuaded by my first example above that losses should never give rise to base erosion tax benefits. They are mere realizations of changes in financial positions that occur in respect of the U.S. taxpayer’s own financial balance sheet, not in respect of any relationship between the U.S. taxpayer and a related foreign person that could result in base erosion or earnings stripping.

But wait. The fourth example is the same as the second, except that the forward contract calls for cash settlement. Thus, at the end of one year, the U.S. taxpayer simply pays $50 to the related foreign person and takes a $50 deduction. Well, this certainly does feel, at first blush, like “a deduction with respect to an amount paid to a related foreign person.” I’d imagine you might have a hard time persuading the IRS otherwise. And yet as a matter of economic substance, this is exactly the same as the other two cases. Perhaps you need to override your first reaction and realize that the deduction is really not “with respect to” an amount paid to a related foreign person, even though it seems to be. It’s really “with respect to” an appreciation in the value of IBM stock. Or perhaps you need to rethink your answer in the other two cases and conclude that the fact that the deduction would not have arisen but for the financial obligation to the related foreign party suffices to turn the deduction into a base erosion tax benefit.21

When Is There a Deduction?

Yet the confusion about “with respect to” seems dwarfed by the question of whether and when there is a deduction to begin with. This problem arises primarily from the failure of the statute to provide for a netting rule. That failure may in some cases serve to convert the BEAT into a tax on gross cross-border payments, rather than on base erosion. But because the tax is still intended as a tax on base erosion, it is still imposed on taxable-income-reducing deductions, rather than on mere payments. And where payments run across borders in opposite directions, one can have offsetting inclusions and deductions, or one can have neither. And the resulting confusion is magnified by the use of the term “deduction” in the denominator of the base erosion fraction, because the point applies equally to transactions with unrelated parties.

Consider, for example, a simple interest rate swap. Under its terms, Party A makes floating rate payments to Party B on a notional principal amount (for example, $1 billion), and Party B makes fixed payments on such notional principal amount in return. These notional payments are likewise netted against each other to produce a single net periodic payment in one direction or the other in any given period. This transaction could equally well be documented as a loan by Party A to Party B of $1 billion in exchange for interest at a fixed rate, and a simultaneous loan of $1 billion by Party B to Party A in exchange for interest at a floating rate. Under the normal income tax (except in rare cases), there is no significant tax difference between the two, as both give rise to the same amount of net income or deduction. But what is meant to happen here in respect of a tax imposed on “gross deductions”?

One possibility is that form alone is respected, and taxpayers are invited to help themselves out of any disadvantageous results. As for the unwary or poorly advised, they will simply have to pay BEAT. Thus, in the case above, a properly advised U.S. affiliate will enter into interest rate swaps with related foreign parties and offsetting loans with unrelated parties. Its deductions for swap payments in the numerator of its base erosion fraction will therefore be dwarfed by deductions for gross interest payments in the denominator, and the BEAT will never apply.22

Indeed, the formalism could go ever further. We could have rulings asserting that if the relevant swap documentation provides that Party A will pay Party B “the excess, if any” of floating over fixed (and Party B will pay Party A the excess, if any, of fixed over floating), then that will result in far fewer deductions (for better or worse, depending on whether or not the counterparty is related, and whether the deduction therefore falls into the numerator or the denominator of the base erosion fraction) than if the documentation simply states that Party A will pay floating and Party B will pay fixed.

The other possibility is that Treasury will seek to apply a “soft doctrine” (for example, economic substance or recharacterization) to produce similar tax results for these economically identical transactions. But what sort of rule would that be? The only rational rule would seem to be a netting rule. That is the rule that would cause the statute to be imposed on base erosion, rather than on gross offsetting payments. And that is also the rule that would prevent taxpayers from blowing up the denominators of their base erosion fractions with offsetting loans. Can you think of another rule that would make any sense?

Another key question in connection with this uncertainty is, what does it mean technically for a deduction to be “allowed”? For the precise statutory definition of a base erosion tax benefit is “any deduction described in subsection (d)(1) which is allowed under this chapter for the taxable year with respect to any base erosion payment” (emphasis added). Common sense might here suggest that what Congress meant is that the deduction in question has to show up on the taxpayer’s return and serve to reduce its otherwise taxable income. For Congress clearly intended the BEAT to operate as a restraint on the taxpayer’s ability to use related-party payments to reduce its taxable income. What concern could Congress have, therefore, with a hypothetical deduction that is not reported on the taxpayer’s return and which therefore does not serve to reduce the taxpayer’s income?

If that is correct, however, it should follow that Congress has no problem when taxpayers restructure their affairs, or take tax positions, in a manner that is designed to eliminate deductions that would otherwise give rise to BEAT liability. Taxpayers might even be able to go as far as to leave tax deductions off their returns (notwithstanding that they are in fact making the relevant payments to related foreign affiliates) if that is what it takes to fall below the 3 percent base erosion threshold and thereby demonstrate to Congress and Treasury that they are not using such deductions to significantly lower their U.S. tax liabilities for the tax year. And if the gross offsetting payments between U.S. and foreign affiliates don’t result in any net base erosion, then the affiliates should feel free to either (1) express this fact by redocumenting their transactions to ensure that there are neither inclusions nor deductions arising from them, or (2) not redocument them, but rather take reasonable positions that the transactions can be reported for tax purposes as not giving rise to either inclusions or deductions (as long as doing so does not serve to lower their regular taxable incomes or otherwise prevent them from clearly reflecting their incomes within the meaning of section 482). Nor should Treasury or the IRS object to any of this. The legislative history of the BEAT is clear that Congress was solely concerned with the use of payments to related foreign parties to lower U.S. taxable income. So why should Treasury or the IRS interpret the BEAT in a manner that causes it to apply in cases where it is clear that U.S. affiliates aren’t doing that?

On the other hand, one might argue that if this is what Congress intended, it wouldn’t have defined base erosion tax benefits to include deductions that are “allowed under this chapter” for the tax year with respect to any base erosion payment. It might rather have referenced deductions that are “taken” on the relevant return, or at least have referenced deductions that are “allowed on the return,” as opposed to “allowed by the internal revenue code.” Section 162 speaks of deductions that are “allowed” for ordinary and necessary business expenses (presumably regardless of whether or not the taxpayer bothers to take them). One might therefore make the technical argument that just as a taxpayer is not allowed to turn her back on income for purposes of applying the income tax, she is not allowed to turn her back on a deduction for purposes of applying the BEAT.

And yet one suspects (given the limited time that Congress had to focus on the BEAT and the various inconsistencies in its application that have been noted above) that Congress didn’t have time to think through the language to this level of detail. It therefore does seem likely that when it said “allowed,” Congress meant “allowed on the taxpayer’s return.” For nothing in the legislative history suggests that Congress was seeking to impose the BEAT in respect of deductions that do not in fact serve to lower the taxpayer’s taxable income.

The Exception for Derivatives

A similar issue arises in connection with securities loans, which generally involve exchanges of financial liquidity between financial institutions (including related financial institutions) for reasons that have nothing to do with tax avoidance. These transactions likewise do not give rise to any base erosion. In this case, however, the statute provides a remedy in the form of an exception for payments “made pursuant to derivatives” (provided the taxpayer is on a mark-to-market method of accounting — that is, is in the business of dealing or trading in securities). Unfortunately, Treasury now seems caught up in a confusing dialogue about whether or not these transactions constitute derivatives and which payments are deemed to be made pursuant to them. In my view, Treasury should be liberal here to better implement the will of Congress.

Let’s start with the technical point: If one affiliate transfers securities to a second affiliate, and the second affiliate immediately sells those securities short on behalf of a customer, the relationship that remains between the two affiliates is necessarily a derivative relationship, as neither party now owns the referenced securities. This fundamental point was recognized by the Supreme Court as early as 1926 in Provost.23 It is reflected in current regulations and rulings that provide, for precisely this reason, that the first affiliate ceases to derive tax-benefited income associated with ownership of the referenced securities (such as dividends eligible for the dividends received deduction or the 20 percent rate of tax, or tax-exempt interest income).24 The “substitute payments” that the first affiliate now receives from the second affiliate are not the real thing, but merely equivalent amounts received pursuant to a derivative relationship. It follows that the transaction in question is a “derivative” as defined by the statute, and the substitute payments of dividends or interest are “payments made pursuant to a derivative” within the meaning of the statute. They are therefore not base erosion payments made by the second affiliate.

Moreover, when the second affiliate transfers the cash proceeds from the sale of the securities back to the first affiliate as “collateral” for its obligation to return substantially similar securities to the first affiliate on demand, the second affiliate is arguably not lending any funds to the first affiliate. For the cash that the second affiliate is now transferring is nothing more or less than the cash that it receives from the sale of the securities that it borrows from the first affiliate. The second affiliate clearly does not part with any use of cash, as it has the same amount of cash immediately after the transaction as it had immediately before it. And the first affiliate has no more cash than it would have had if it had lent the securities in question directly to the customer. The transaction therefore does not give rise to any base erosion. And that is precisely why Congress enacted the derivatives exception. It follows that the rebate fees paid on the cash collateral should also be treated as “payments made pursuant to a derivative” within the meaning of the statute.

In other words, the second affiliate is here merely serving as an intermediary in capital markets in connection with a flow of funds from a customer (or from another securities dealer, in a case where the second affiliate on-lends the securities to an unrelated broker dealer for use by the latter in a short sale) to the first affiliate. It is true that Treasury could nevertheless seek to disaggregate this relationship, asserting that a transfer of securities from the first affiliate to the second affiliate effectively transfers liquidity from the first affiliate to the second affiliate, while the corresponding transfer of cash collateral from the second affiliate to the first affiliate effectively transfers liquidity from the second affiliate to the first affiliate. Thus — Treasury could go on to assert — the transaction can be treated in the same irrational way that the statute insists upon treating two offsetting loans. Yet why would Treasury struggle to do this? It is clear that Congress enacted the derivatives exception precisely to ensure that such offsetting payments running in opposite directions would not give rise to base erosion tax benefits. In other words, Congress enacted the exception precisely to ensure that Treasury and the IRS would not arrive at such an unhelpful result (a result that would be unhelpful precisely because the transaction does not result in any base erosion). Under such an approach, the exception granted by Congress (in exchange, apparently, for a significant decrease in the base erosion threshold and increase in the rate of tax applied to financial institutions) would effectively apply to nothing at all. If anything, Treasury should struggle to do the opposite and aggregate two offsetting loans into a single deemed interest rate swap that can then be treated as a derivative for this purpose. Indeed, for the reasons set out above, it may want and need to do precisely that to ensure that the denominator of the base erosion percentage fraction is not blown up through the disaggregation of interest rate swaps into offsetting loans.

Treasury appears to be caught up in this regard by some confusing “just in case” statutory language apparently added in at the last minute: “This subsection shall not apply to any qualified derivative payment if the payment would be treated as a base erosion payment if it were not made pursuant to a derivative.”25 It’s not clear what this language means, but it cannot be read to apply to payments that wouldn’t exist but for the derivative relationship — otherwise it would eviscerate the entire exception. Suppose that additional language under section 162 read as follows: “No deduction shall be allowed for business expenses that would not be deductible if they weren’t paid.” Whatever the drafter intended by such words, they would have to be interpreted in a manner that still allowed for the deduction of bona fide business expenses.

Thus, the only case that doesn’t involve a derivative is the one where the first affiliate unilaterally borrows $1 billion of cash from the second affiliate. This includes a case where the first affiliate posts securities as collateral for the borrowing, but only if the second affiliate continues to hold the securities in unliquidated form on behalf of first affiliate. As soon as the second affiliate sells the securities, the transaction converts into a securities loan, for the reasons set out above. (Moreover, even the case where the first affiliate unilaterally borrows cash from the second affiliate does not give rise to base erosion if the first affiliate invests the borrowing proceeds in income-producing assets. And of course, that is doubly the case where the first affiliate effectively lends the borrowing proceeds back to the second affiliate in a different transaction.)

Affiliates Providing Services to Each Other

And the same issue arises in connection with the provision of services. As I’ve noted above, there is no base erosion happening when worldwide affiliates provide services to each other in the ordinary course of their businesses. If the pricing is arm’s length, there is no base erosion regardless. But even if it isn’t, there can’t be base erosion unless there is a net deduction in respect of an excess of services received over services rendered.

Thus, suppose that U.S. Affiliate and Foreign Affiliate cooperate to provide advice to their respective customers regarding the business climate in both jurisdictions. U.S. Affiliate has customers in the United States, and Foreign Affiliate has customers outside the United States. Last year U.S. Affiliate received $10 billion from its U.S. customers and paid $5 billion in fees to Foreign Affiliate for its help in providing advice to those customers with respect to the non-U.S. business climate. Likewise, Foreign Affiliate received $10 billion from its non-U.S. customers and paid $5 billion of fees to U.S. Affiliate for its help in providing advice to those customers with respect to the U.S. business climate.

As we all know, the BEAT has no special netting rule. Thus, if the two affiliates use the same documentation this year, U.S. Affiliate will have to include $5 billion of base erosion tax benefits in both the numerator of its base erosion percentage and its modified taxable income. Treasury should not be troubled, however, if these affiliates respond by altering their documentation to eliminate the gross base erosion payments. For there is no base erosion going on, and the alteration will merely allow the affiliates to “prove” this fact to Congress and Treasury. If anything, Treasury should be troubled by the failure of the statute to provide the netting rule that would have eliminated the need to engage in such otherwise pointless restructuring.

Let us now review the options that these affiliates have for altering their documentation to avoid the imposition of BEAT, the better to determine whether any of them should be challenged by Treasury and the IRS. But in doing so, we should bear in mind that any alteration that results in de facto netting for purposes of applying the BEAT is, for the reasons set out above, a good one as a matter of policy.

Let’s start with the simplest idea: U.S. Affiliate and Foreign Affiliate will stop paying each other for their services. Parties in business, whether related or unrelated, scratch each other’s backs all the time. Thus, U.S. Affiliate and Foreign Affiliate will each simply report $10 billion of gross income and no offsetting deduction, thereby cutting out the additional $5 billion of related-party fee income offset by $5 billion of related-party fee deductions. The primary legal issue, of course, is whether the IRS can, and will, seek to “impute” $5 billion of fee income and $5 billion of offsetting fee deductions.

The first thing that comes to mind here is barter exchanges. But it might be possible to distinguish them in this case. For the two affiliates in our example are together conducting a worldwide advisory business, and it therefore might seem reasonable for them to help each other out without paying each other for their respective efforts. A taxable barter exchange generally involves the sort of exchange for which people normally pay each other. Thus, if you are my roommate and I agree to do the dishes provided that you mow the lawn, we do not have a taxable barter exchange. If you are my neighbor, and I agree to maintain the fences provided that you trim the hedges, we do not have a taxable barter exchange. Moreover, a taxable barter exchange normally involves the exchange of one kind of service for a different kind of service. If, for example, I provide you with legal services in exchange for your providing me with dental services, we would appear to have a taxable barter exchange. But if I answer your legal question as a professional courtesy on the understanding that you will answer mine tomorrow, we presumably do not (I hope). Neither do we have one if I agree to take all the kids on the block to school on Monday provided that you do so on Tuesday, or to baby-sit them on Wednesday provided that you do so on Thursday.26 The IRS could conceivably assert the existence of a taxable barter exchange in one of these cases, but how much sense would that make as a matter of policy?

Returning, then, to the relationship between U.S. Affiliate and Foreign Affiliate, the IRS’s primary concern in this context has been to ensure that U.S. Affiliate conducts its activities on an arm’s-length basis and does not use the arrangement as an opportunity to shift taxable income from U.S. Affiliate to Foreign Affiliate. In other words, its concern has been nothing more or less than the objective of Congress in enacting the BEAT. To that end, Treasury has promulgated a detailed set of regulations under section 482 designed to ensure that related parties deal with each other at arm’s length in providing services to each other and therefore do not use those arrangements to erode the U.S. tax base. Let us assume, arguendo, that U.S. Affiliate always receives just as much help from Foreign Affiliate as it provides to Foreign Affiliate and that the dealings between the two parties are therefore at arm’s length (that is, unrelated parties would have been willing to enter into the same relationship). Let us further assume that the IRS is satisfied that U.S. Affiliate has fully complied with all of the requirements of relevant regulations under section 482. Is there any reason for Treasury and the IRS to now seek to recharacterize the transactions between the two affiliates in order to impose BEAT in respect of an arrangement where they have fully satisfied themselves that no base erosion is going on?27 As noted above, all Congress could possibly have been asking for in enacting the BEAT is that U.S. Affiliate not use deductions for payments to related foreign parties to significantly reduce its otherwise U.S. taxable income. What could constitute greater compliance in this regard than not taking any such deductions at all?

In any case, whatever risk there may be of the IRS asserting the existence of deemed additional payments for the provision of services, it would not appear to exist where U.S. Affiliate and Foreign Affiliate are actually sharing the revenue that they derive from their respective customers. Thus suppose that U.S. Affiliate and Foreign Affiliate change their basic economic relationship and agree that each of them will be entitled to half their joint revenue. As a result, unlike last year, if U.S. Affiliate receives $20 billion from its customers next year and Foreign Affiliate receives nothing, U.S. Affiliate will have to remit $10 billion to Foreign Affiliate. As a matter of economic substance, therefore, each party is contributing its services to a common pool of revenue generation. It would seem especially odd for the IRS to assert in this context that each party was nevertheless making deemed arm’s-length payments to the other for services (and was then contributing the services that it had just purchased from the other party to a common pool of revenue generation). The recharacterization doctrine is supposed to combine steps to arrive at their simpler essences, not add additional complexity in order to impose more tax. Moreover, it is a tool for reaching just and rational results. Bear in mind that there isn’t any base erosion going on here.

Is this reasoning any less compelling in a case when U.S. Affiliate contributes capital, and the Foreign Affiliate contributes labor, to the joint generation of revenue? I don’t think so. Suppose that you and I embark on a joint venture for profit. You contribute the brains (for example, the whole thing was your idea) and the sweat (you work on this 24/7), and I contribute the capital, and we split the profits 50/50. Should I be including all the resulting profits in income and then deducting the half that you receive as a fee for brains and sweat? Isn’t it obvious that I should include only half the profits in income and should therefore deduct nothing at all? Does it matter, in this regard, whether you collect the money from the customers and give me my half, or I collect the money from the customers and give you your half? It would seem odd to suggest that the application of BEAT could turn on such a distinction.

Suppose, for example, that I collect $20 million from the customers and give you your half ($10 million). Suppose it is clear, moreover, that the value of your contribution of brains and sweat was $5 million. Is the IRS going to require me to include $12.5 million (rather than $10 million) in income and then posit my arm’s-length deductible payment of $2.5 million to you in respect of my deemed acquisition of 50 percent of your brain and sweat services, which I then contributed to the common pool of brains, sweat, and capital? Or is it going to go even further and require me to include $20 million in income and to then deduct $10 million as a non-arm’s-length overpayment for $2.5 million worth of brain and sweat services, merely because I was the one who collected the money from the customers?

And if the IRS were to insist upon either of these, under what authority would it do so? Section 482? That statute merely allows the commissioner to require related parties to clearly reflect their incomes by dealing with each other on an arm’s-length basis. It doesn’t allow the commissioner to require one of the parties to overstate its income (and then overstate offsetting deductions) so that the commissioner can impose more BEAT. Neither would the soft doctrines of business purpose, economic substance, or substance over form. For the arrangement described above clearly does have business purpose and economic substance, and it already is cast in its simplest, most direct, and most logical form.

This has nothing to do, by the way, with concepts of agency. If U.S. Affiliate happens to be the one that collects the $20 billion of revenue, it doesn’t need to prove that it is collecting $10 billion of it as an agent for Foreign Affiliate. To the contrary, U.S. Affiliate has no right to claim that it is earning all $20 billion for U.S. tax purposes. This goes back to the concept of “attribution of income.”28 The fruits may not be attributed to a different tree than the one on which they grew. A broker-dealer, for example, cannot collect dividends and interest on behalf of one of its customers and then claim that it was the one that earned the income, rather than the customer. Nor can a lender employ formalism to purport to be earning the tax-benefited income that the borrower derives by investing the borrowing proceeds.29 When two parties jointly earn revenue, the revenue belongs to both of them, regardless of who collects it, and regardless of whether that fact helps or hurts either the commissioner or the taxpayer.

Consider in this regard how easy it would be for Foreign Affiliate to sign all of the contracts with customers (including U.S. customers) and collect all the money from them. Foreign Affiliate would then remit to U.S. Affiliate its 50 percent share of the revenue. There would obviously be no means of recharacterizing such an arrangement to produce a deemed deduction for U.S. Affiliate in respect of a deemed payment from U.S. Affiliate to Foreign Affiliate. And yet the application of BEAT to a business arrangement between affiliates cannot turn on this sort of formalism.

This is to be distinguished, however, from the case when Foreign Affiliate remits all the revenue to U.S. Affiliate except for a fixed amount that is in proportion to the amount of services that it performs. An arrangement like that presumably could be recharacterized based on economic substance as an inclusion by U.S. Affiliate of all the revenue (collected on its behalf by Foreign Affiliate) less a deductible payment to Foreign Affiliate for services performed by the latter. This leads us to the most difficult question of all: Is there a safe path for affiliates that are not engaged in net earnings stripping but still want to derive compensation based on the amount of services that they provide (rather than to share revenue)? For it is honestly not clear in the first example above whether U.S. Affiliate and Foreign Affiliate are performing services for each other, or whether they are rather performing services for each other’s customers.

Let us return, then, to the first case above but suppose that the standard contract with the customers of both affiliates is redrafted such that each customer hires both U.S. Affiliate and Foreign Affiliate and pays part of its fee to each of them. As noted above, this has nothing to do with revenue sharing. The amount that each U.S. customer pays Foreign Affiliate is based on the amount of work that Foreign Affiliate performs. Thus, Foreign Affiliate receives nothing from U.S. customers if it doesn’t perform any services. At the end of the day, however, Foreign Affiliate happens to earn $10 billion from both U.S. and non-U.S. customers, and U.S. Affiliate happens to earn $10 billion from both U.S. and non-U.S. customers. Neither pays the other a fee for services because neither needs to. The services in question have been rendered directly to unrelated customers.

It seems pretty clear in this case that U.S. Affiliate has no deductions for base erosion payments. And that is the right policy answer in this particular case, because Foreign Affiliate is not eroding the U.S. tax base. Yet the answer underlines the random aspects of the BEAT, the application of which apparently turns on such formalisms as whether the U.S. customer pays Foreign Affiliate directly or rather pays U.S. Affiliate, which in turn pays Foreign Affiliate. For conversely, there would be little to stop taxpayers from avoiding the BEAT in circumstances when there really was base erosion going on. Suppose, for example, that U.S. Affiliate received nothing from non-U.S. customers because Foreign Affiliate didn’t conduct any business. Suppose further that Foreign Affiliate did next to nothing in exchange for the $5 billion that it received directly from U.S. customers, so the payments from U.S. customers to Foreign Affiliate effectively served to strip $5 billion out of the U.S. tax base (by reducing U.S. affiliate’s income from $15 billion to $10 billion). Section 482 would still be available to challenge the amount of the payments. But the new BEAT wouldn’t appear to help at all. It simply wouldn’t apply.

Suppose now we take matters further. Imagine that it is cumbersome for U.S. Affiliate and Foreign Affiliate to ask their customers to write two different checks to two different affiliates. Each affiliate therefore collects the fixed amount that its customer owes to the other affiliate on behalf of such other affiliate and remits it to the other affiliate in due course. Do we now have a different answer? As noted above, this fact wouldn’t matter at all if the two affiliates were sharing joint revenue. But here each is providing the other (on behalf of its customers) with an amount that is in direct proportion to the amount of services performed. Is U.S. Affiliate here acting as an agent in collecting the fee that its U.S. customers owe Foreign Affiliate and remitting it to Foreign Affiliate? Or are the U.S. customers really paying U.S. Affiliate (so that U.S. Affiliate must include the payment in income), and U.S. Affiliate is then paying Foreign Affiliate (with the resulting deduction giving rise to a base erosion tax benefit)? Does it matter just how aware the U.S. customer is that this is happening? Suppose that the U.S. customer’s agreement to pay fees to Foreign Affiliate is on the 20th page of a boilerplate document that most U.S. customers sign without reading beyond page 2. Does it matter whether employees of Foreign Affiliate speak directly with the relevant U.S. customers (as opposed to merely consulting with employees of U.S. Affiliate who then relay the gist of their advice to the U.S. customers)?

There is a line of cases dealing with agency relationships in the context of wholly owned corporations and their shareholders.30 It generally follows the form of the transaction because there is little economic substance to the distinction between a shareholder and her wholly owned corporation and thus she might otherwise be free to decide post hoc whether to apply tax benefits (such as losses) at the individual or corporate level. This line of cases might have little application, however, in the context of a relationship between a corporation and its completely unrelated customers. Corporations routinely act as agents in receiving income on behalf of their customers, and they aren’t deemed to earn the income themselves just because relevant payers aren’t aware of the fact. But any hard and fast rule here would cut both ways, potentially giving rise to inappropriate results in either direction. What Treasury should do first is net the payments between the two affiliates (regardless of how they are characterized) to determine whether Foreign Affiliate is receiving a net fixed amount for net services performed for either U.S. Affiliate or its customers. Only to that extent should Treasury then consider recharacterizing the transaction so that BEAT can be applied to a resulting deemed deduction for a deemed payment from U.S. Affiliate to Foreign Affiliate.

The Proposed Regulations

As little time as Congress had to think through the BEAT before drafting it, Treasury has so far had even less. Treasury was tasked with getting proposed regulations on the TCJA out in a hurry so that it could meet the 18-month deadline for implementing final regulations with retroactive effect.31 Its first big priority in this regard was to get out the section 965 regulations, and its second big priority was to get out the complex regulations governing new section 163(j). These potentially affect far more taxpayers. Treasury then turned to the new GILTI and foreign-derived intangible income regulations, and the even more complicated FTC package, including rules designed to deal not only with the allocation of expenses, but also with the application of section 163(j) in the foreign context. How much time and energy could have been left for the final big package of regulations dealing with the BEAT? As with Congress, the BEAT came last and had to make do with whatever energy and focus was left for it.

It is little surprise, therefore, that when Treasury was suddenly confronted with a host of practical implementation issues, it deferred to general law. Set out below is Treasury’s general discussion in the preamble to the proposed regulations of some of the issues that I have outlined above:

In general, the treatment of a payment as deductible, or as other than deductible, such as an amount that reduces gross income or is excluded from gross income because it is beneficially owned by another person, generally will have federal income tax consequences that will affect the application of section 59A and will also have consequences for other provisions of the Code. In light of existing tax law dealing with identifying who is the beneficial owner of income, who owns an asset, and the related tax consequences (including under principal-agent principles, reimbursement doctrine, case law conduit principles, assignment of income or other principles of generally applicable tax law), the proposed regulations do not establish any specific rules for purposes of section 59A for determining whether a payment is treated as a deductible payment or, when viewed as part of a series of transactions, should be characterized in a different manner.32

And later in the preamble, Treasury states the following:

The Treasury Department and the IRS are aware that certain reinsurance agreements provide that amounts paid to and from a reinsurer are settled on a net basis or netted under the terms of the agreement. The Treasury Department and the IRS are also aware that other commercial agreements with reciprocal payments may be settled on a net basis or netted under the terms of those agreements. The proposed regulations do not provide a rule permitting netting in any of these circumstances because the BEAT statutory framework is based on including the gross amount of deductible and certain other payments (base erosion payments) in the BEAT’s expanded modified taxable income base without regard to reciprocal obligations or payments that are taken into account in the regular income tax base, but not the BEAT’s modified taxable income base. Generally, the amounts of income and deduction are determined on a gross basis under the Code; however, as discussed in Part III of this Explanation of Provisions section, if there are situations where an application of otherwise generally applicable tax law would provide that a deduction is computed on a net basis (because an item received reduces the item of deduction rather than increasing gross income), the proposed regulations do not change that result.

The trouble, of course, is that the income tax is a tax imposed on net income, not on gross receipts, let alone gross deductions. The common law of the income tax therefore doesn’t have answers to the kinds of questions posed by the BEAT, and what answers it arguably does provide (by way of analogy) are random when applied in this context and certainly do not speak to the purported objectives of the BEAT. Important answers to the various questions posed in this report should not be governed by the outcome of a case that deals, for example, with the reimbursement of routine employee expenditures.

Conclusion

In my view, the statute is simply too ambiguous, and too flawed as drafted, for Treasury to take a cautious approach. It needs to wade into the heart of the issues and resolve them effectively. And if Treasury needs more time to accomplish that, I think it should take it. I think Treasury could go so far as to implement a netting rule for interest payments and service fees so as to better target the statute toward base-eroding relationships. That rule would presumably apply in both the numerator and the denominator of the base erosion percentage fraction.

As for the inconsistencies in the application of the statute to acquisitions of depreciable property from related parties as compared with payments of royalties and acquisitions of property for sale, I’m not sure Treasury can do much unless and until the statute is changed. But it certainly should not seek to expand the application of the BEAT by characterizing various financial transactions as disguised acquisitions of depreciable property in an effort to defend an inconsistent imposition of tax that doesn’t itself seem to make any sense.

FOOTNOTES

1 See infra note 4. The legislation was signed December 22, 2017.

2 Treasury has concluded, for example, that it does not have authority to allocate interest and other expenses away from the global intangible low-taxed income foreign tax credit basket in order to implement what appears to have been congressional intent not to impose U.S. tax on GILTI earnings that are already subject to foreign tax at a rate of 13.125 percent.

3 “Unified Framework for Fixing Our Broken Tax Code” (Sept. 26, 2017), prepared by the Trump administration, the House Ways and Means Committee, and the Senate Finance Committee.

4 See, e.g., the minutes of the Hearing on International Tax Reform before the Finance Committee, 115th Cong. (Oct. 3, 2017) (in which those inversions were prominently discussed).

5 Unlike the corporate-level tax, the individual-level tax is not reduced by interest deductions as compared with equity allocations. Thus, suppose that a closely held partnership earns $100 and also pays $10 to an unrelated finance company. The partners pay tax on $90 regardless of whether the $10 is treated as a deductible payment of interest or a payment out of earnings.

6 Although before 2026, the 30 percent limitation will be on a higher base than that of old section 163(j) (earnings before interest and taxes), the base will be reduced after 2025 to earnings before interest, taxes, depreciation, and amortization, which approximates taxable income.

7 Senate Budget Committee explanation (Nov. 30, 2017).

8 As a technical matter, this is not an exception per se, but rather arises from the structure of the statute. More specifically, amounts paid to purchase goods are not deducted (assuming that the property is not depreciated) but rather included in the cost basis of the property purchased, and so merely reduce gain (or increase loss) from the subsequent sale. There is therefore no deduction available for the BEAT to treat as a base erosion tax benefit.

9 Preamble to REG-104259-18, at 60.

10 “The Committee is also concerned about U.S. and foreign corporations outsourcing their U.S. business operations to foreign jurisdictions at the expense of the American worker. In certain circumstances, this may have the additional effect of reducing the U.S. income tax liability on such companies’ profitable operations in the United States.” Senate Budget Committee explanation, supra note 7.

12 One other possibility is that the purpose of the extension was to make it harder for courts to conclude that the BEAT runs afoul of the non-discrimination clauses in our income-tax treaties with foreign countries. Yet if that was the sole basis for the extension, then it has no policy rationale whatsoever.

13 The Senate Budget Committee explanation, supra note 7, merely states as follows: “The Committee also concluded that this minimum tax should limit the extent to which tax credits permit large, profitable corporations with significant base erosion payments to avoid virtually all tax liability in the reformed corporate tax system. This is to ensure that those corporations with significant gross receipts and deductible foreign related party payments pay an appropriate amount of U.S. income tax on an annual basis.”

14 See preamble to REG-104259-18, at 60.

15 The statute would seem to have authorized Treasury to do this, because modified taxable income for the year must be determined without regard to deductions for base erosion tax benefits, and that would arguably include such deductions in prior tax years. Thus, the current year’s NOL carryforward can be reduced or eliminated (for purposes of determining the current year’s modified taxable income) to the extent that it would have been used up in prior tax years but for the availability in those prior years of base erosion tax benefits to reduce otherwise regular taxable income.

16 See preamble to REG-104259-18, at 48.

17 More specifically, the statute defines modified taxable income as regular taxable income “determined without regard to” base erosion tax benefits. This suggests a recomputation of taxable income without taking base erosion tax benefits into account. However, the proposed regulations would — for the sake of simplicity — revise this to provide that modified taxable income simply equals regular taxable income plus base erosion tax benefits. See preamble to REG-104259-18, at 46-47.

18 The sentence reads: “The term ‘base erosion tax benefit’ means any deduction described in subsection (d)(1) which is allowed under this chapter for the taxable year with respect to any base erosion payment.”

19 The full language reads: “The term ‘base erosion payment’ means any amount paid or accrued by the taxpayer to a foreign person which is a related party of the taxpayer and with respect to which a deduction is allowable under this chapter.”

20 Yet Treasury makes the following statement in the preamble: “In addition, because Section 59A(d)(1) defines the first category of base erosion payment as ‘any amount paid or accrued by the taxpayer to a foreign person which is a related party of the taxpayer and with respect to which a deduction is allowable under this chapter,’ a base erosion payment also includes a payment to a foreign related party resulting in a recognized loss; for example, a loss recognized on the transfer of property to a foreign related party.” Preamble to REG-104259-18, at 22.

21 As discussed further below, if the U.S. taxpayer were a financial institution on a mark-to-market method of accounting, the loss would in any case not be treated as a base erosion tax benefit by reason of the exemption of payments in accordance with a derivative from treatment as a base erosion payment. The question would still be relevant, however, for purposes of determining whether or not the loss had to be excluded from the denominator of the taxpayer’s base erosion percentage fraction (because it was exempt by reason of the derivatives exception).

22 We might consider here the antiabuse rule that Treasury has proposed under reg. section 1.59A-9(b)(3): “A transaction . . ., plan or arrangement that has a principal purpose of increasing the deductions taken into account for purposes of [the denominator of the base erosion fraction] is disregarded for purposes [of determining the base erosion percentage].” But taxpayers would not in this case be entering into interest rate swaps in order to influence the denominators of their base erosion fractions. They would merely be documenting the interest rate swaps that they were going to enter into anyway in a manner that takes account of their tax consequences. Does a relevant “plan or arrangement” extend to the way in which the taxpayer documents a transaction that it was going to enter into regardless of tax? There is presumably no obligation on the part of the taxpayer to document a bona fide business transaction in a manner designed to maximize the taxpayer’s tax liability.

23 Provost v. United States, 269 U.S. 443 (1926).

24 Prop. reg. section 1.1058-1(d); Rev. Rul. 80-135, 1980-1 C.B. 18; and Rev. Rul. 60-177, 1960-1 C.B. 9.

25 Section 59A(h)(3). The full sentence reads: “This subsection shall not apply to any qualified derivative payment if the payment would be treated as a base erosion payment if it were not made pursuant to a derivative, including any interest, royalty, or service payment, or in the case of a contract which has derivative and non-derivative components, the payment is properly allocable to the non-derivative component.”

26 I might further note that taxable barter exchanges have until now been of concern primarily outside the business context, where the relevant receipt of services is taxable, but the offsetting payment for services is not deductible. There hasn’t yet been much litigation about such barter exchanges in the business context because even if the receipt of services is taxable, the offsetting payment of services is generally deductible under section 162, and so net taxable income would not change regardless.

27 To the contrary, the IRS has already demonstrated in a different context that it is willing to look past any potential arguments of technical application that lead to counterintuitive results, provided that related affiliates make a good-faith effort to comply with its arm’s-length dealing rules. Thus, in any case where U.S. Affiliate and Foreign Affiliate are “scratching each other’s backs” (e.g., where they are both engaged in a global securities dealing business and they “pass the dealing book” to each other over the course of a 24-hour day to ensure that there is full coverage at any time), there is the potential argument that some portion of the profits derived on an arm’s-length basis by Foreign Affiliate are nevertheless subject to U.S. tax because the help it receives from U.S. Affiliate causes that income to be “effectively connected with the conduct of a U.S. trade or business” under section 864 and/or earned through a “permanent establishment in the United States” within the meaning of a relevant tax treaty. But proposed regulations under sections 863 and 864 make it clear that Treasury and the IRS do not intend to seek to impose U.S. tax on income that is properly allocated, on an arm’s-length basis, to business activities that are conducted outside the United States. See, e.g., prop. reg. sections 1.863-3(h); 1.864-4(c)(2)(iv), (c)(3)(ii), and (c)(5)(vi)(a); and 1.864-6(d)(3), which were primarily drafted to coordinate with the arm’s-length allocation of the income (under prop. reg. section 1.482-8) arising when taxpayers engage in a global dealing business.

28 Lucas v. Earl, 281 U.S. 111 (1930).

29 Nebraska v. Lowenstein, 513 U.S. 123 (1994).

30 See, e.g., National Carbide Corp. v. Commissioner, 336 U.S. 422 (1949).

32 See preamble to REG-104259-18, at 19.

END FOOTNOTES

Copy RID