In his bathrobe, Arthur Dent lies down in front of a bulldozer in a futile attempt to save his home. As he awoke that morning, he was informed that local authorities needed to demolish his domicile to make space for a highway bypass. Later that morning, after a visit to the local pub, Arthur and his alien pal Ford Prefect board a city-block-sized spacecraft from the planet Vogsphere just before it vaporizes Earth. The intergalactic authorities have determined the planet is in the path of a hyperspace bypass. So begins the Douglas Adams classic comedic tale The Hitchhiker’s Guide to the Galaxy. Like Arthur, all we earthlings need to remember is not to sweat the details when larger problems loom.
In the tax universe, some have been waxing philosophical. Do we have a territorial system with a minimum tax on global intangible low-taxed income? Or do we have a multi-rate worldwide system? Does it matter? Most of us, powerless, just do what we can and attend to the details. It is easy to be distracted by the bulldozer of complexity that surrounds section 951A (taxes foreign business profits as they are earned), section 250 (might cut that tax in half), and section 960 (provides tax credits that are supposed to prevent any U.S. taxation of high-tax foreign profits). But overhead looms Vogon Constructor Flagship 861-8 that seeks to allocate expenses to the GILTI basket, implode foreign tax credits, and threaten any semblance of territoriality or rational tax policy.
The expense allocation rules were written for an era when foreign income was taxed by the United States at a single rate (putting aside capital gains for the moment). The interaction of the statutory language of the Tax Cuts and Jobs Act (P.L. 115-97) and existing expense allocation rules results in an overabundance of expenses to the GILTI basket. This is over and above the excess allocation resulting from our lopsided water’s-edge approach (under which a portion of U.S. expenses are allocated to foreign-source income without taking into account any foreign expenses) rather than a symmetric worldwide allocation.
For U.S. multinational corporations that have GILTI and that cannot credit all their foreign income taxes on income in the GILTI basket (that is, excess credit taxpayers), every dollar of U.S. interest, U.S. research, and U.S. stewardship expense allocated to that basket raises their U.S. after-credit tax liability by $0.21, regardless of the foreign rate of tax. In what seems contrary to congressional intent (in the informal sense), many U.S. corporations with average foreign tax rates exceeding 13.125 percent may pay considerable U.S. tax on GILTI.
Moreover, as illustrated, the allocation of expenses to the GILTI basket increases the number of excess credit taxpayers by lowering the threshold level of foreign tax to which the FTC limit applies. The average foreign tax rate of 13.125 percent that presumably would have put corporations in an excess credit position was never that high to begin with — especially when you consider the willingness of foreign governments to implement rules that curtail base erosion and profit shifting. But with expenses allocated to the GILTI basket, the crossover point from “high-tax,” excess credit taxpayer to excess limit taxpayer drops even lower. It is possible that U.S. multinationals with highly leveraged or research-intensive domestic operations and low foreign profits may find they are unable to credit any foreign taxes imposed on GILTI. And to make matters worse, taxpayers cannot credit those taxes in later years because Congress denied carryforwards of tax credits in the GILTI basket.
All that stands between taxpayers and potential vaporization of their FTCs are the regulation writers at the Office of Tax Policy and the IRS. Unfortunately, the canons of statutory interpretation limit how much they can read into the statute and fix obvious problems. The preferred outcome expressed by many affected taxpayers is that Treasury simply not allocate any of those allocable U.S. expenses to the GILTI basket. This is not going to happen. Treasury lawyers have exercised Scalia-like restraint in their interpretation of the law.
In a discussion draft for the November 9 University of Chicago Law School 71st annual Federal Tax Conference, Dana L. Trier, former Trump administration Treasury deputy assistant secretary for tax policy, described several methods of rationalizing FTC rules as they apply to the GILTI basket. Trier explored possibilities in which expense allocations would take into account the reduced effective rate at which GILTI is taxed and enforce consistency in the treatment of expenses for the calculation of direct U.S. tax liability on GILTI and the calculation of foreign-source income for determining the FTC limitation. This is the path Congress should have taken in the fall of 2017. This is the path Congress will take post-technical corrections, when it finally gets around to making substantive changes to the TCJA. But for now, it is all theory. The regulation writers don’t have the flexibility to impose the sort of logical consistency Trier described.
As much as they believe they can, the regulation writers will noodle the edges of the FTC limitation. And perhaps more significantly, they will provide indirect relief by generously clarifying that more low-tax income can flow into the GILTI basket and offset the atrophy of the limit caused by the allocation of expenses. How do we know this? Because another Treasury official said so.
“The solution has to be some combination of interest expense allocation rules [and] look-through rules,” Lafayette G. “Chip” Harter III, deputy assistant secretary for international tax affairs, said in June. “It will not be a thing of conceptual beauty, but we also have to walk a very fine line between . . . avoiding making tax reform a marginal detriment to using the U.S. as a home office jurisdiction on one hand versus opening up the opportunity for cross-crediting of foreign tax credits on the other.” (Prior coverage: Tax Notes Int’l, June 11, 2018, p. 1342.) Harter reiterated Treasury’s inclination for pragmatism at the University of Chicago Law School conference when he quoted Trier’s draft: “The most reasonable course of action at this point may be one that is not conceptually pure.” Taxpayers can hope, but reality dictates that they brace themselves for some significant overallocation of U.S. expenses into the GILTI basket when proposed regulations are released.
To set the stage for analysis of expense allocation on GILTI, we will first look at GILTI from 10,000 feet without any expense allocations. Further, throughout we will set net deemed tangible income return (NDTIR) (equal to 10 percent of qualified business asset investment (QBAI) less interest paid by controlled foreign corporations to unrelated parties). QBAI and CFC interest expense can play an important role in GILTI. In particular, QBAI provides relief that can offset the impact of FTC-reducing expense allocations. But most taxpayers are reporting that the relief provided by this subtraction from tested income is relatively small and the negative effects of expense allocations are much more significant.
When NDTIR does play a role, its effects can vary significantly. For example, in the stylized case used, when QBAI is taken into account, the marginal effective tax rate on investment in the low-tax country varies from 7.3 percent — when the rate of return on investment is 25 percent — to -9.5 percent, when the rate of return is 5 percent. Given the complexity and unsettling variability of the effects of the NDTIR calculation, Congress should consider dropping it from the calculation of the minimum tax base and simultaneously raising the section 250 deduction to whatever rate above the current 50 percent rate maintains revenue neutrality.
Another aspect of reality neglected in the simple model below is profit shifting. Here it is assumed that all related-party transactions are conducted at arm’s length and all profits are economic profits. Since enactment of the TCJA, profit shifting will be more difficult and certainly less lucrative. But by no means has the incentive to shift profit abroad been eliminated. Calculations (not shown here) indicate that profit shifting under current law can significantly reduce already low marginal effective tax rates on foreign investment, and in many cases drive them below zero.
With that background out of the way, now consider two types of U.S. multinational corporations. The first has an average foreign effective tax rate below, and the second has one above, 13.125 percent. We can call the first an excess limit taxpayer and the second an excess credit taxpayer. Also consider that these corporations can make their next investment in either a low-tax country (with an effective tax rate of less than 13.125 percent) or a high-tax country (with a rate exceeding 13.125 percent). Making a whole bunch of simplifying assumptions — including no profit shifting, NDTIR equal to zero, and no U.S. expenses allocated to the GILTI basket — the effective tax rates on marginal investment (and shifted profits) are shown in the boxes of the two-by-two diagram at the top of Figure 1.
G% is the territorial percentage (equal to 50 percent through 2025 in new section 250). F% is the percentage of deemed paid foreign taxes creditable (equal to 80 percent under current section 904(d)). If the excess limit taxpayer invests, its marginal effective tax rate is determined by the combination of foreign tax (tF) plus U.S. GILTI tax ((1 - G%) * tUS) less allowed FTCs (F% * tf). Combined, these effects yield a marginal effective tax rate of (1 - F%) * tF + (1 - G%) * tUS. The excess credit taxpayer also has its marginal effective tax rate determined by the combination of three effects, but this time the FTC limitation is binding, so the third effect is (1 - G%) * tUS. There is no U.S. tax, so the marginal effective rate is the foreign rate: a territorial system.
To animate this discussion a bit, let’s make some plain vanilla assumptions about the characteristics of these taxpayers. Let’s assume that, except for their tax rates, they are identical. They each have $100 of both tangible and intangible investment that generates GILTI and all investment earns a 16 percent rate of return. The excess limit taxpayer has an average foreign tax rate of 5 percent. The excess credit taxpayer has an average foreign tax rate of 30 percent. The low-tax country has an effective rate of 5 percent, and the high-tax country has an effective rate of 30 percent.
Under those assumptions, as shown in the lower portion of Figure 1, the excess limit corporation has an effective foreign tax rate on marginal investment in the low-tax country of 11.5 percent (a combination of the 5 percent foreign rate and the difference between U.S. GILTI tax of 10.5 percent and FTC effect of 4 percent). Especially noteworthy is that the marginal effective tax rate in the high-tax country with a 30 percent rate is only 16.5 percent. This is attributable to cross-crediting: By investing in the high-tax country, the excess limit corporation increases FTCs available to shelter low-tax profits from the U.S. tax on GILTI.
The low rate is the combined effect of foreign tax at 30 percent, U.S. GILTI tax at 10.5 percent, and a reduction in U.S. tax equal to 80 percent of the foreign tax rate of 30 percent. The excess limit taxpayer can cross-credit until there is enough profit in the high-tax country that it is no longer an excess limit taxpayer. The marginal effective tax rate of the excess credit taxpayer not subject to any U.S. tax is simply the foreign tax rate.
Add Expense Allocation
In Figure 1, the height of the horizontal dividing line between the low- and high-taxpayer is set by the 13.125 percent crossover rate between excess limit and excess credit status. Similarly, the position of the vertical line dividing low- from high-tax investment is set by the 13.125 percent rate. But there is nothing sacred about 13 and one-eighth percent. Congress could change the crossover point by adjusting G% or F%. For example, if G% was increased to 70 percent, the crossover rate would drop to 7.875 percent. If F% was reduced to 70 percent (increasing the FTC haircut to 30 percent), the crossover rate would increase to 15 percent.
It is also true that the more expenses are allocated to the GILTI basket, the lower the crossover rate will be. As shown in Figure 2, allocation of expenses to the GILTI basket shifts the horizontal line up and the vertical line to the left. So in addition to increasing U.S. tax on foreign investment for excess credit taxpayers, expense allocation increases the number of excess credit taxpayers. The increase in marginal effective tax rates for the excess credit corporation in our example is from tf to tf + e * tUS.
The expression for the crossover foreign rate between excess credit and excess limit status is:
(tUS/F%) * [(1 - G%) * tUS - e]
where e is domestically incurred expense, expressed as a percentage of foreign economic profits, allocated to the GILTI basket. For excess limit taxpayers, the marginal effective tax rate increases by tus * e.
If these expenses reduce foreign-source profits as computed for the FTC limitation by 30 percent (that is, if e = 0.3), the crossover rate is 5.25 percent. This is shown in the bottom portion of Figure 2. The marginal effective tax rates for the excess credit corporation in our example increase by 6.3 percentage points from 5 percent to 11.3 percent for investment in the low-tax country and from 30 percent to 36.3 percent in the high-tax country.
‘Most Massively’ Pragmatic
The proposed foreign tax regulations submitted for review to the Office of Management and Budget’s Office of Information and Regulatory Affairs are expected to include the long-awaited expense allocation rules for the GILTI basket. (Prior coverage: Tax Notes Int’l, Nov. 12, 2018, p. 764.) So we should have answers any day now. In The Hitchhiker’s Guide, the number 42 is the “Answer to the Ultimate Question of Life, the Universe, and Everything” as calculated by a supercomputer named Deep Thought over a period of 7.5 million years. During the summer and right through the fall, the lights were on late at 1111 Constitution Ave. and 1500 Pennsylvania Avenue. Deep thinkers were at work. Rumor has it they considered and rejected 42 as the solution to the GILTI expense allocation conundrum. They did not throw in the towel.