Menu
Tax Notes logo

Economic Analysis: Fixing GILTI: Bye-Bye to QBAI

Posted on May 6, 2019

When it comes to the future of international income taxation, all roads lead to a minimum tax.

The form of a minimum tax on foreign profit should depend on its intended purpose. Here is probably the most fundamental issue: Should the minimum tax apply broadly to limit tax competition among nations and therefore impose tax on all low-taxed foreign income? Or should it be narrow in scope — solely a base limitation measure — that imposes tax only on artificially shifted low-tax foreign-source income?

One thing is clear from the outset: Either way, there is no need to retain the clunky exemption from minimum tax for qualified business asset investment (less specified interest). The inclusion of global low-taxed intangible income is the U.S. version of an outbound minimum tax. Unlike subpart F, which applies to each controlled foreign corporation, tax on GILTI is computed at the shareholder level. QBAI is 10 percent of the average of the four end-of-quarter total tax bases of tangible depreciable property used to generate GILTI. Under current law, GILTI in excess of this return on tangible capital is potentially caught in the GILTI minimum tax net. This excess amount is an imperfect indicator of high profitability or the existence of intangibles.

In the case of a minimum tax intended for all low-tax profit, the QBAI mechanism would simply be eliminated. In the case of a minimum tax targeting shifted profits, the QBAI mechanism should be replaced with more fair and accurate profit level indicators. The GILTI rules are a target-rich enviroment. Future articles will discuss other possible fixes.

Made in the USA

A short while ago — when it hurriedly enacted the Tax Cuts and Jobs ActCongress needed to raise corporate revenue to offset revenue losses from a corporate rate cut. To that end, our legislators added two jarring acronyms to the tax lexicon: BEAT (base erosion and antiabuse tax) and GILTI.

The maddeningly complex BEAT effectively imposes tax on the gross receipts of some payments (depending more on their form than substance) to foreign related parties if a U.S. multinational’s U.S. tax is too low or its foreign tax credits (and hence also probably its foreign taxes) are too large. (No, the last phrase was not backwards. The law is.) Its one saving grace is that it applies to perhaps only one or two thousand of the largest businesses in the United States. Undoubtedly, as a matter of policy, it is useful to have a backstop to the arm’s-length method for outbound payments by foreign multinationals operating in the United States. But if we think of the BEAT as a building, it would be better to tear it down and reconstruct from the ground up.

The inclusion of GILTI in U.S. taxable income, on the other hand, is a candidate for rehabilitation. Make no mistake, under the pressures of limited time and money, the drafters of GILTI made several easily avoidable blunders, including no provision for carryforward of FTCs and the disregard of GILTI losses in the computation of aggregate GILTI. But if the regrettable design choices are scraped away, we can see the underlying structure is sound. Moreover, trillion-dollar deficits and the eternal antipathy of the voting public to lightly taxed corporations mean repeal is out of the question. Hopefully, after the 2020 elections, Congress will engage in a second round of U.S. international tax reform. Depending on the election outcome, there could be a tax increase on foreign-source income. At the center of any of these efforts will be a restructuring of the U.S. tax on GILTI.

Dissatisfaction With BEPS

A short while from now — when it is scheduled to issue a final report in 2020 designed to provide consensus-based long-term solutions — the OECD will need to recommend a way to plug the holes in the anti-base-erosion framework it published in 2015. The OECD’s consultation document released February 13 stated: “Certain members of the Inclusive Framework [the group of 120-plus nations participating in the OECD project] consider these [BEPS] measures do not yet provide a comprehensive solution to the risks that continue to arise from structures that shift profits to entities subject to no or very low taxation.”

Two prominent practitioners described the current situation this way: “Apart from this obligation to compensate other entities assisting with control, the BEPS Report seems to require only that the tax-advantaged entity be contractually assigned a risk and perform some modest portion of the control function related to that risk” (Paul Oosterhuis and Joseph Andrus, “Transfer Pricing After BEPS: Where Are We and Where Should We Be Going,” 95 Taxes 89 (Mar. 2017)).

The OECD’s effort to shore up its initial BEPS work was first referred to as pillar 2 in the short policy note released in January 2019. This is in addition to its original mandate by G-20 finance ministers in March 2017 to attribute more taxable profits to the local market sales and marketing (pillar 1).

At the top of almost everybody’s list of possibilities for a pillar 2 anti-base-erosion measure is a GILTI-like “income inclusion” minimum tax on foreign profits. (The other pillar 2 proposal is some yet-to-be-specified tax on outbound payments not subject to significant tax in the jurisdiction where they are received. This is like the BEAT, presumably with as few of the BEAT’s loopholes and traps for the unwary as possible. This proposal is often discussed as a backstop to the minimum tax — yes, a backstop to a backstop.)

Two Types of Minimum Taxes

The OECD received 212 sets of comments on the February public consultation document. A common theme running through many of the comments from practitioners — in addition to the need for international consensus, the need to adhere as much as possible to the arm’s-length standard, the need to slow down the process to consider such enormous changes, the need to allow the 2015 BEPS recommendations to take full effect, the need for dispute resolution mechanisms, and the need to take into account losses in redistribution of taxing rights — is the need for any new proposal to have a clear policy goal and to be grounded intellectually in principle.

In the case of the minimum tax, there seems to be possible primary goals (which we alluded to above): It should be (1) a broad minimum tax on all profits to increase (capital export) neutrality between domestic and foreign investment by domestically headquartered multinationals; or (2) a narrow minimum tax on shifted profits to reduce base erosion.

Under the broad version of the minimum tax, to discourage the outflow of investment to low-tax countries, the home jurisdiction of a multinational business would not allow the effective tax rate on the sum total of foreign-source income of its resident businesses to fall below a designated level. (Assume the new tax applies on an overall, rather than on a country-by-country, basis.) Stated differently, the home jurisdiction might adopt territoriality (capital import neutrality) as its guiding principle rule but would put limits on how far it would go so that it wouldn’t unduly encourage foreign investment by its multinationals. This would be consistent with the long-standing general thrust of U.S. policy as a hybrid of capital import and capital export neutrality. Once the structure is in place, lawmakers could dial up or down the neutrality spectrum by adjusting the designated minimum tax rate. (They could also adjust the angle of the floor on the minimum tax rate, allowing U.S. minimum tax to decline fractionally with any rise in the foreign tax rates. Under GILTI, the U.S. minimum tax declines by 80 cents for every dollar of foreign tax paid because of the 20 percent “haircut” of the FTC under that provision.)

The thinking behind the narrow version of the minimum tax is that pure territoriality is the proper goal of international tax reform. But purity is rare in international taxation. Sometimes — rarely, according to most advocates of this view — the arm’s-length method fails to provide a proper result. So the defining mission of the minimum tax is not to limit capital import neutrality or restrain tax competition but to serve as an antiabuse measure. The implication of this for any minimum tax is that in addition to a low tax rate, something else is needed to trigger the tax. In the case of tax on GILTI, that something else is a high amount of profit relative to net tangible capital.

In the consultation document, the five paragraphs specifically devoted to the income inclusion rule mention that the proposal would draw on aspects of the GILTI rules, but there is no mention of QBAI or anything other than low foreign tax rates triggering the minimum tax. If anything, the document portrays the minimum tax squarely as falling into our concept of a broad minimum tax on all low-tax profits. The income inclusion rule:

reinforces tax sovereignty of all countries to “tax back” profits where other countries have not sufficiently exercised their primary taxing rights. . . . It posits that global action is needed to stop a harmful race to the bottom. . . . The proposal therefore seeks to advance a multilateral framework to achieve a balanced outcome which makes business location decisions less sensitive to tax considerations, limit compliance and administration costs and avoid double taxation. [Emphasis added.]

Skip the Intangibles Talk

It is worthwhile at this point to briefly digress about the use of the word “intangible” in the context of GILTI. Yes, income from intangibles is the Achilles’ heel of the arm’s-length method. Because of their intrinsic uniqueness, it is difficult to find comparables for intangibles unless the taxpayer is licensing rights to a third party. So directing a minimum tax to situations in which intangibles loom large is not a bad something else to add on to an extra tax that targets low-tax income.

But we also know — especially now based on experience with GILTI — that among its many other quirks, the QBAI mechanism is a poor indicator of intangible profits. For decades economists liked the idea of using what was measured as excess returns on tangible capital as a clever way to indirectly measure income from intangible assets. But in reality, business profits — particularly in the short run — move around with all the factors that affect supply and demand and, if the business has transactions with related parties, they can also move around with profit shifting through any adjustments to cost allocations or transfer prices. So in practice, it is imprudent to assume that high profitability is attributable to intangible assets (or that lack of profitability suggests the absence of intangibles).

Does this in and of itself make the QBAI mechanism an inappropriate addition to a minimum tax directed at likely transfer pricing problems? The answer is no. Any excess profitability is an indicator, albeit imperfect, of a potential transfer pricing problem. Significant swings in the location of profits can result from relatively small adjustments to transfer prices of tangible products and of services as well as rights to intangibles.

And you could certainly expect more pressure on the transfer prices of these other related-party transactions if any minimum tax applied just to intangible income. The core problem with QBAI is not that it is a poor indicator of the presence of intangible income but that it is a poor indicator of excess profits. One common example of this shortcoming is the case of a CFC in a service business with lots of employees but little or no tangible assets. Most or all of its profits are GILTI, even though there is no reasoned indication of excess profitability.

Section 482 to the Rescue

So what can be used in place of the QBAI mechanism to demarcate the threshold above which excessive profits exist for purposes of a minimum tax? Of course, in the real world it is impossible to pinpoint transfer pricing errors even under the best of circumstances, much less with a one-size-fits-all formula. But it may be possible to devise a profitability measure, or a set of profitability measures that would vary by industry, that can reasonably flag excess when inappropriate or aggressive profit shifting is likely. It is undoubtedly possible to devise a simple excess profit indicator that is better than QBAI.

To that end, lawmakers should make selective use of the concepts and methods that decades of experience with arm’s-length pricing provides. For example, to determine profits from manufacturing, use the cost-plus method. To determine profits from distribution, use the resale price method. These traditional methods are widely used, relatively easy to apply, and noncontroversial in nature. Just because the arm’s-length method has now been knocked off its pedestal doesn’t mean we should abandon it or simplified variations of it.

This is the wisdom underlying the residual profit-split method of allocating profit. The starting point for application of a residual profit split — in current U.S. regulations and in past and recently updated OECD transfer pricing guidelines — is computation of profits on routine functions calculated under conventional transfer pricing methods. This routine profit can be a substitute for QBAI so that only low-taxed income in excess of routine profit would be subject to minimum tax.

The Case for No QBAI

Would it be such a terribly uncompetitive outcome for U.S. multinationals if the second round of international tax reform adopted a modest minimum tax on all foreign profits? The rate could be in the neighborhood of 10 percent without the fig leaf of territoriality provided for profit below QBAI or some other threshold. Yes, this would be a rejection of pure territoriality. Yes, this would be a floor on the overall rate of foreign tax to limit the incentive to move investment offshore. It could do that with no diminution of its role as an antiabuse measure.

Even for the most stalwart advocate of capital import neutrality on principle, might it not be worthwhile as a matter of simplification to eliminate the QBAI or any potential replacement from future versions of an outbound minimum tax? Everything is approximate; would not a simple low U.S. tax on even low-taxed foreign profits be a good enough approximation of territoriality? Given that any future tax legislation will almost certainly be constrained by budget targets, would not a reduction of the minimum tax rate be a good revenue-neutral trade for jettisoning any exemption amount?

The Case for a QBAI Substitute

If in our reformed minimum tax there must be some semblance of territoriality — and we should take care not to discount the intransigence of some lawmakers on this point — there must be something else, some additional indicator of profit shifting, included as part of the minimum tax. There could be qualitative measures of (an inadequate) foreign tax system that would ultimately end in the creation of white- or blacklists of countries with low taxes that would be subject to the new minimum tax. This could be a diplomatic headache that would involve the State Department.

In any case, the idea probably deserves more study. March 6 comments on the OECD public consultation from Skadden, Arps, Slate, Meagher & Flom LLP discuss how to differentiate between “appropriate” regimes that would not be subject to minimum tax from other regimes: “Arguably, indicators of an inappropriate regime could include the lack of any substantial income tax at all, contrivances of clearly artificial exclusions or allocations of income away from a jurisdiction, and the allowance of permanent deductions in excess of actual costs or value.”

Alternatively, we could stick with quantitative measures — specifically, quantitative measures designed to identify excess profitability. If we stick with this GILTI-like approach, the QBAI mechanism should be replaced with either an arm’s-length determination of routine profit or some rough-justice method that attempts to approximate an arm’s-length markup.

Joining Pillars 1 and 2

Let’s return to pillar 1 of the consultation document, which up until now we have barely mentioned. Pillar 2 seeks to address some of the remaining issues of the 2015 BEPS reports with proposals that backstop the arm’s-length method and keep profits out of tax havens. Pillar 1 is far more ambitious. It would fundamentally transform nexus, source, and transfer pricing rules in order to justify more taxation of multinational profits in market countries.

Within pillar 1 there are three proposals. The user participation proposal would identify specific digital activities in user (market) jurisdictions that would justify allocation to those jurisdictions. The significant economic presence proposal would allocate profits to jurisdictions where a multinational may have no physical presence with allocation determined by formula. There is no practical reason to contemplate details of these proposals because they will almost certainly not be recommended.

By a wide margin, a third approach — the marketing intangibles approach — is the odds-on favorite to be the OECD’s pick. This is a residual profit split with the twist that only residual income from marketing intangibles would be subject to the profit split. How the intangible income from marketing intangibles will be separated from other intangibles is a practical problem of enormous proportions. For our purposes, we want to focus on the easy part of the residual profit split, the determination of routine profit. The simple point we want to make is that the method for determining a residual profit split (a pillar 1 proposal) should be used to determine a narrow minimum tax profit threshold below which minimum tax does not apply (a pillar 2 proposal). The only difference between the two is that the minimum tax threshold can be increased by some percentage to account for some margin in error in the determination of excess profitability due to profit shifting.

The OECD consultation document does not indicate any preferred method for determining routine profit: “There are different ways in which routine profit could be determined for purposes of computing the amount of non-routine income to be subject to the profit split, ranging from a full transfer pricing facts and circumstances analysis to a more mechanical approach.” As noted recently in these pages, proponents of other residual profit-split proposals are not doctrinaire on how routine profit is determined. (Prior analysis: Tax Notes, Apr. 22, 2019, p. 513.)

In a 2009 paper, Reuven S. Avi-Yonah, Kimberly A. Clausing, and Michael C. Durst argued that routine profit could be determined using a 7.5 percent markup over costs. But the authors added: “It is anticipated that the Treasury will by regulation prescribe different markups for geographic locations, or particular industries, in which a markup of 7.5 percent does not constitute a reasonable estimate of a ‘routine’ level of return based on prevailing market conditions” (“Allocating Business Profits for Tax Purposes: A Proposal to Adopt a Formulary Profit Split,” 9 Fla. Tax Rev. 497 (2009)). In a subsequent paper, Avi-Yonah acknowledges that using an arm’s-length markup may be more politically palatable and endorses that approach (Avi-Yonah and Ilan Benshalom, “Formulary Apportionment — Myths and Prospects,” 2 World Tax J. 371 (Oct. 2011)). Conversely, six authors collectively referred to as the Oxford group advocate use of traditional arm’s-length methods to determine routine profit. But they acknowledge:

If over time these transfer pricing disputes proved in fact to be troublesome or costly, the determination of routine profits could be made more formulary, for example, by implementing safe harbors or even mandatory mark-ups on specified costs or rates of return on investment to determine routine returns without reference to specific comparables. [Michael P. Devereux et al., “Residual Profit Allocation by Income,” Saïd Business School, WP No. 19/01 (Mar. 2019).]

The choice of methods for the determination of routine profits, whether for a residual profit split or a minimum tax, will hinge on the trade-off between simplicity of fixed markups versus the accuracy of a full arm’s-length analysis.

The Residual

Now if you’re willing to accept the idea that routine profits can be the same or similar under residual profits and minimum tax proposals, the differences among the leading alternatives now being discussed are mainly about the treatment of residual profits. Those differences are:

  1. Devereux et al. proposal — all residual profits are allocated by gross income.

  2. Avi-Yonah et al. — all residual profits are allocated by sales.

  3. Marketing intangible proposal — residual profits from marketing intangibles are allocated by some apportionment factor (almost certainly sales). The location of all other profits is determined under arm’s-length methods.

  4. Income inclusion/minimum tax proposal targeted at profit shifting — any profits of foreign subsidiary, determined under the arm’s-length method, to be in excess of routine profit and subject to low tax are subject to tax in the multinational’s home country.

Using this characterization of the major proposals, several points are easy to make. The critical issue in all is the treatment of residual profits. Proposals 1 and 2 have a great deal in common, and they would entail the largest change from current law. Proposal 3 is less of a change from current law than proposals 1 and 2 but is undoubtedly much more complex because of the need to separate marketing from other intangibles. All three of those pillar 1-type proposals are far more sweeping than the minimum tax in proposal 4, both in terms of concept and in terms of amounts of revenue up for grabs.

The practitioner community would prefer a conservative approach. From its perspective, if there must be change, let it be along the lines of proposal 4. But the advice embodied in practitioner comments to the OECD may not prevail if budgetary and political pressures force nations to adopt unilateral measures designed to protect their tax base and their domestic businesses.

Copy RID