Tax Notes logo

Pillar 2 and the Corporate AMT

Posted on Aug. 8, 2022
Bret Wells
Bret Wells
Reuven S. Avi-Yonah
Reuven S. Avi-Yonah

Reuven S. Avi-Yonah is the Irwin I. Cohn Professor of Law at the University of Michigan, and Bret Wells is the Law Foundation Professor of Law at the University of Houston Law Center. They thank Nir Fishbien and Heydon Wardell-Burrus for their helpful comments.

In this article, Avi-Yonah and Wells argue that the corporate alternative minimum tax proposed in the Inflation Reduction Act of 2022 would put the United States in a better position than current law, and arguably even better than would a tax reform package that included a conforming global intangible low-taxed income regime but no book-based corporate minimum tax.

Copyright 2022 Reuven S. Avi-Yonah and Bret Wells.
All rights reserved.

The Inflation Reduction Act of 2022, as introduced by Senate Majority Leader Charles E. Schumer, D-N.Y., and Sen. Joe Manchin III, D-W.Va., on July 27 contains none of the changes to the global intangible low-taxed income regime that were included in the Build Back Better legislation passed by the House. Because the Build Back Better changes were designed to make the GILTI regime compatible with pillar 2 of the OECD/G-20 global corporate tax reform framework, the question arises of what consequences the United States may suffer from including the book-based corporate alternative minimum tax (CAMT) in the proposed legislation but not the GILTI changes. Despite the United States’ noncompliance with pillar 2 (contrary to the Biden administration’s agreement with 136 other countries last year to implement a global minimum corporate tax of 15 percent), the CAMT should put the United States in a better position than current law, and arguably even better than would a tax reform that included a conforming GILTI but no CAMT.

Pillar 2 and Current U.S. Law

There are several major problems with how current U.S. law interacts with pillar 2. First, the GILTI tax rate of 10.5 percent is too low compared with the pillar 2 minimum rate of 15 percent.1 Second, pillar 2 requires the application of a conforming income inclusion rule on a country-by-country basis,2 whereas the GILTI regime applies to all controlled foreign corporations and therefore allows the blending of income from high- and low-tax jurisdictions. Third, the GILTI rate applies only to income above a deemed tangible income return. And finally, GILTI is calculated based on tested income, whereas pillar 2 calculations are based on book income.

However, even though the GILTI regime does not conform to pillar 2, it is likely to qualify as a CFC tax regime.3 That means that any tax imposed on the U.S. parent because of GILTI will be allocated to its CFCs and count toward the minimum tax threshold. How the GILTI income inclusion is allocated among tested CFCs remains an important unresolved issue. This will be made more complicated because part of the offset to a GILTI inclusion arises from the use of foreign tax credits. Presumably, jurisdictions that generate significant FTCs would not be allocated a net GILTI inclusion if their tested income also generated sufficient FTCs to offset the GILTI exposure arising from their jurisdiction. But this remains to be worked out.

One would expect that reasonable proxies could be agreed on to make these allocations either by upfront agreement or through a mutual agreement process. Assuming a reasonable allocation of the GILTI income inclusion to the low-taxed entities is made on a CbC basis, the maximum tax imposed on a particular low-taxed CFC by the GILTI regime is 10.5 percent. In that scenario, appropriate credit should be given for the GILTI inclusion so that the maximum top-up tax under pillar 2 would be 15 percent - 10.5 percent = 4.5 percent. This top-up tax should be imposed as either a qualifying undertaxed profits rule or as a qualifying intermediate IIR.

Note that there should be no U.S. FTC relief for top-up taxes paid under either a qualifying UTPR regime or a lower-tier qualifying IIR under GILTI, because if those taxes are creditable for U.S. purposes, there will be a circularity problem. For example, assume a U.S. multinational enterprise has $100 in income from a CFC in Country A, which does not tax it. The income is derived from Country B, which has signed on to pillar 2. The United States will apply GILTI tax at 10.5 percent, and Country B will apply UTPR at 4.5 percent. If the UTPR is not creditable, there is no problem because the two will just apply with the GILTI tax applied first and the UTPR applied thereafter. But if the UTPR is creditable, the U.S. calculation will be GILTI at 10.5 percent - 80 percent of 4.5 percent, which is 3.6 percent = 6.9 percent. But this would seem to require an increase in the UTPR or intermediate IIR to 8.1 percent, and so on. To prevent this issue, Treasury should clarify that top-up taxes under either a pillar 2 UTPR regime or an intermediate IIR regime should not be afforded U.S. FTC relief. Because these are top-up taxes, the disallowance of those credits does not raise double taxation concerns.

In any case, the problem with an unreformed GILTI regime is that from a U.S. perspective, it leaves money on the table because a U.S. MNE will be subject to a 15 percent tax but the revenue will accrue to foreign jurisdictions. Moreover, if the parent of a U.S. MNE has an effective tax rate below 15 percent on domestic income, the UTPR might apply to U.S. domestic income as well (if the MNE is subject to the UTPR because it operates in a jurisdiction that has adopted pillar 2). The same applies to U.S. subsidiaries of large foreign MNEs from countries that apply the IIR.

Importantly, this scenario is quite plausible even if neither the United States nor the EU adopts pillar 2; it is enough if other members of the G-20 in which U.S. MNEs operate adopt pillar 2. The United Kingdom and Canada are about to implement pillar 2, and it is highly likely that France, Germany, Italy, and Japan will do so as well.

Pillar 2 and a Conforming GILTI

If the proposed legislation had included amendments that would have conformed the GILTI regime to a qualifying IIR in lieu of the proposed CAMT, the problem outlined above would be solved because a conforming GILTI regime would turn off the UTPR and intermediate IIR top-up taxes.

However, even a conforming GILTI regime would still leave the parents of U.S. MNEs (and any domestic subsidiaries) potentially subject to the UTPR if the MNE operates in a country that applies the UTPR and if domestic nonrefundable credits or domestic deductions reduce the effective tax rate of the parent below 15 percent. The same problem applies to domestic subsidiaries of large foreign MNEs under the IIR. This has led to criticism of pillar 2 by constituencies worried about particular credits or deductions.

Pillar 2 and the CAMT

The proposed legislation includes the CAMT but does not reform the GILTI regime. Interestingly, this may lead to better results from the U.S. perspective compared with both current law and a reform that includes a reformed GILTI regime but not the CAMT.

The CAMT is based on book income, which means that it includes the income of the U.S. parent and its CFCs.4 Once that income is included, a tentative tax at 15 percent is applied, but it is offset by a credit for regular corporate tax, credits for foreign taxes, and various domestic credits. The resulting tax is then compared with the regular corporate tax, and the higher tax is payable.

For foreign income, it is likely that the CAMT will be a “covered tax” under OECD Model Rule 4.2.1 and thus not be excluded under OECD Model Rule In that case, model rules 4.3.2 and 5.1.1 should allocate the CAMT to various CFCs in each jurisdiction, and if the allocated CAMT results in a 15 percent effective tax rate, it will turn off the application of UTPRs and qualified domestic minimum top-up taxes (QDMTTs) to that income.

The treatment of covered taxes is addressed in section 4.3.2 of the model rules, which states:

Covered Taxes are allocated from one Constituent Entity to another Constituent Entity as follows:

  1. the amount of any Covered Taxes included in the financial accounts of a Constituent Entity with respect to [global anti-base-erosion (GLOBE)] Income or Loss of a Permanent Establishment is allocated to the Permanent Establishment;

  2. the amount of any Covered Taxes included in the financial accounts of a Tax Transparent Entity with respect to [GLOBE] Income or Loss allocated to a Constituent Entity-owner pursuant to Article 3.5.1(b) is allocated to that Constituent Entity-owner;

  3. in the case of a Constituent Entity whose Constituent Entity-owners are subject to a Controlled Foreign Company Tax Regime, the amount of any Covered Taxes included in the financial accounts of its direct or indirect Constituent Entity-owners under a Controlled Foreign Company Tax Regime on their share of the Controlled Foreign Company’s income are allocated to the Constituent Entity;

  4. in the case of a Constituent Entity that is a Hybrid Entity the amount of any Covered Taxes included in the financial accounts of a Constituent Entity-owner on income of the Hybrid Entity is allocated to the Hybrid Entity; and

  5. the amount of any Covered Taxes accrued in the financial accounts of a Constituent Entity’s direct Constituent Entity-owners on distributions from the Constituent Entity during the Fiscal Year are allocated to the distributing Constituent Entity.

The commentary to the model rules provides:

Similarly, subject to the limitations of Article 4.3.3, Covered Taxes arising in connection with an income inclusion under a CFC Tax Regime imposed on another Constituent Entity are allocated to, and included in the Adjusted Covered Taxes of, the Constituent Entity CFC pursuant to Article 4.3.2(c). To the extent Covered Taxes are not allocated because of the operation of Article 4.3.3, such Covered Taxes are included in the Adjusted Covered Taxes of the Constituent Entity-owner.6

The application of these rules to the CAMT on foreign income is unclear. The CAMT is applied directly to the book income of CFCs of a U.S. MNE with FTCs. Assuming that a given CFC is not a permanent establishment, a transparent entity, or a hybrid under the check-the-box rules, the only way the CAMT could be allocated to the CFC is if it is a CFC tax regime. Otherwise, the CAMT on foreign income of a CFC would be allocated to the U.S. parent, but that would produce the inappropriate result that even though the CAMT is imposed on the income of the CFC and brings the tax on that income to 15 percent, the CFC might fall below the 15 percent effective tax rate and be subject to a UTPR or QDMTT, while the tax allocated to the U.S. parent may not have an effect if the parent is already subject to the CAMT at 15 percent before the tax on the CFC is allocated to it. That would cause double taxation of the CFC’s income. That would be an unintended result because the drafters of the model rules did not envisage a situation in which a country would have a CAMT but not a conforming GILTI regime.

To prevent this inappropriate outcome, the model rules should be amended to clarify that CAMT that results from the income of CFCs should be allocated to them. The easiest way to do this is to treat the CAMT as applied to CFCs as a CFC tax regime. Treasury should argue for that result.

For domestic income tax purposes, Treasury should argue that the portion of any CAMT related to low-taxed U.S. domestic income is itself a QDMTT entitled to first-priority status. The OECD model rules define a QDMTT as a minimum tax that is included in the domestic law of a jurisdiction and that (1) determines the excess profits7 of the constituent entities located in the jurisdiction (domestic excess profits) in a manner that is equivalent to the GLOBE rules; (2) operates to increase domestic tax liability for domestic excess profits to the minimum rate for the jurisdiction and constituent entities for a fiscal year; and (3) is implemented and administered in a way that is consistent with the outcomes provided for under the GLOBE rules. A QDMTT may compute excess profits based on an acceptable financial accounting standard permitted by the authorized accounting body or an authorized financial accounting standard rather than the financial accounting standard used in the consolidated financial statements.8 Because the CAMT will be based on applicable financial statements, it will be possible to determine the portion of the CAMT related to the U.S. segment reporting unit.

The GLOBE rules envision more adjustments and more carveouts from a minimum tax than is envisioned by the CAMT.9 The GLOBE rules also contain significant adjustments on deferred tax assets and deferred tax liabilities.10 Even so, the OECD model rules envision that a minimum tax on the income located in the United States that is “equivalent” to the GLOBE rules should be treated as a QDMTT. Given that the segment reporting on the financial statements will make it clear what portion of the CAMT relates to U.S. operations, the variances between the CAMT and the GLOBE rules should be viewed as minor such that the CAMT should be viewed as substantially equivalent to the GLOBE rules, at least in their practical operation vis-à-vis the U.S. jurisdiction. In fact, the CAMT in many respects appears to be a more stringent version of a 15 percent minimum tax than the one contemplated by the GLOBE rules because those rules contain many carveouts that are not in the CAMT.

A further operational complexity will arise in a year when the U.S. taxpayer is subject to regular U.S. taxation and is able to benefit from a credit for previously paid CAMT. In the year when CAMT credits are claimed as a reduction of the U.S. regular tax liability, a QDMTT is technically not paid in that year in the United States. This problem becomes particularly acute if the CAMT credit that is used to reduce the U.S. tax liability on U.S. regular taxable income arises from non-U.S. jurisdictions so that there is a cross-crediting among jurisdictions. In that case, the CAMT credit that is applied to reduce the domestic U.S. regular tax liability could cause the U.S. tax rate to fall below the minimum rate for income generated in the U.S. jurisdiction because part of the CAMT credit may relate to CAMT assessed on book income arising from non-U.S. jurisdictions.

This nuance, and perhaps others, will require coordination on how to handle potential scenarios in which the U.S. taxpayer is not subject to the CAMT in a later year or uses CAMT credits in that future year. One possible solution would be to give taxpayers the flexibility to not claim CAMT credits in a future year and instead carry those credits forward to a still-later year if immediate use would cause the U.S. jurisdictional tax rate to fall below the minimum rate. In another context, Treasury has allowed taxpayers an analogous flexible elective right to not claim benefits if doing so would create a disadvantage under the base erosion and antiabuse tax.11 If U.S. taxpayers were to elect to not cross-credit any resulting CAMT credit against the U.S. regular tax in situations in which that use would cause the U.S. domestic rate to fall below the minimum rate, the affirmative election to not claim that cross-crediting of CAMT credits could mitigate the risk that the U.S. jurisdiction would be considered a low-tax jurisdiction in a year in which CAMT credits are claimed against the regular U.S. tax liability. As a result, an election to not allow for that cross-crediting in whole or in part could prevent the application of a qualifying UTPR or intermediate qualifying IIR for profits generated within the United States.


Overall, the adoption of the CAMT by the United States should result in the same 15 percent being paid by U.S. MNEs that are subject to it (as well as by foreign MNEs investing into the United States), but more of that revenue should be paid to the United States on both foreign and domestic income than under current law. Moreover, the CAMT could result in more revenue to the United States than even a tax reform package that includes a reformed GILTI but no CAMT, because a reformed GILTI covers only foreign income whereas the CAMT should shield both foreign and domestic income from taxation by other countries that adopt pillar 2. However, the ability to then use U.S. CAMT credits in later years to reduce regular taxable income of the U.S. person could in turn cause the U.S. jurisdictional tax to fall below 15 percent in light of cross-crediting use of U.S. CAMT credits against regular taxable income. Treasury should ensure that taxpayers have the flexibility to not be required to claim CAMT credits if doing so would subject them to qualifying UTPR or qualifying IIR regimes.


1 See OECD, “Tax Challenges Arising From the Digitalisation of the Economy — Global Anti-Base Erosion Model Rules (Pillar Two),” at 60 (Dec. 20, 2021) (model rules) (defining a minimum rate under art. 10.1.1).

3 A CFC tax regime is defined in the model rules as “a set of tax rules (other than an IIR) under which a direct or indirect shareholder of a foreign entity (the controlled foreign company or CFC) is subject to current taxation on its share of part or all of the income earned by the CFC, irrespective of whether that income is distributed currently to the shareholder” (emphasis added). Model rules, supra note 1, at 54, art. 10.1.1.

4 See proposed new section 56A(a)(2)(A) and (c)(3)(A) (including the pro rata share of foreign income of CFCs in the CAMT base).

5 See the model rules, supra note 1, at 54, art. 10.1.1. Covered taxes include:

(a) Taxes recorded in the financial accounts of a Constituent Entity with respect to its income or profits or its share of the income or profits of a Constituent Entity in which it owns an Ownership Interest; (b) Taxes on distributed profits, deemed profit distributions, and non-business expenses imposed under an Eligible Distribution Tax System; (c) Taxes imposed in lieu of a generally applicable corporate income tax; and (d) Taxes levied by reference to retained earnings and corporate equity, including a Tax on multiple components based on income and equity.

6 Commentary, supra note 2, at 88, art. 4.1.3(12).

7 Excess profits are defined as GLOBE income minus the substance-based income exclusion. See model rules, supra note 1, at 29, art. 5.2.2.

8 For the definition of a QDMTT, see the model rules, supra note 1, at 64, art. 10.1.1.

9 See OECD model rules 3.2 and 3.3.

10 See OECD Model Rule 4.3.

11 See prop. reg. section 1.59A-3(c)(6).


Copy RID