SIFMA Seeks Changes to Proposed Subpart F High-Rate Exception Regs

September 17, 2020

Internal Revenue Service
CC:PA:LPD:PR (REG-127732-19)
Room 5203, Post Office Box 7604
Ben Franklin Station
Washington, DC 20044

Re: High-tax exclusion for subpart F and GILTI purposes

Ladies and Gentlemen:

This letter provides comments on behalf of the Securities Industry and Financial Markets Association (“SIFMA”)¹ regarding the application of the high-tax exclusion for subpart F and GILTI purposes. Our comments relate to regulations that were issued in proposed and final form on July 23, 2020.² We commend the drafters for their efforts to interpret a complex statutory framework and to harmonize the provisions of the TCJA with elements of pre-reform law that remained in place.

1. Taxpayers should be allowed to allocate expenses by reference to applicable financial statements in respect of all post-TCJA years, and not just in years after the proposed regulations are issued in final form.

The proposed regulations generally provide for the allocation of items of income and expense to a CFC's tested units by reference to their applicable financial statements. By contrast, the final regulations would permit taxpayers to allocate income but not expenses by reference to the tested units' books and records.³ Under the final regulations, the allocation of expenses would be governed by generally applicable U.S. rules, including Treasury regulations §1.861-9.

The use of U.S. tax principles to allocate expenses to tested units, as provided for in the final regulations, has the potential to produce significant distortions, particularly in cases where tested units are treated as separate entities for local tax purposes and in the context of tiered ownership structures.⁴

The approach contemplated by the proposed regulations — a single uniform methodology based on applicable financial statements — is strongly preferable to the inconsistent methods for income and expenses provided in the final regulations. The drafters recognized in the preamble to the proposed regulations that this approach would produce a more effective matching for U.S. and foreign tax purposes.

We encourage Treasury and the IRS to allow taxpayers to rely on applicable financial statements in respect of all periods after the entry into force of the TCJA, and not just for years after the proposed regulations are issued in final form.

2. Foreign taxes that are imposed at an undefined or negative rate should not be relegated automatically to the residual category.

Treasury and the IRS have requested comments concerning a provision of the proposed regulations under which amounts derived by a member of a CFC group will be considered high-taxed if: (i) the amounts are subject to foreign tax at an undefined or negative rate; and (ii) a high-tax election has been made in respect of the group. The preamble says that Treasury and the IRS are considering whether this result is appropriate in all cases.

Under this rule, foreign taxes will be conclusively presumed to be allocable to the residual grouping even if they are paid by a CFC that never expects to have any actual high-taxed income. We believe that this result is inappropriate. We therefore recommend that this presumption be eliminated. Foreign taxes should be deemed to be associated with exempt income only if they are actually imposed with respect to that income, and not otherwise.

Under the new framework for taxing foreign income created by the TCJA, foreign taxes that are allocated to the residual category effectively will be disallowed, and will never produce foreign tax credit benefits. A default rule that automatically relegates foreign taxes to the residual category, without regard to whether the taxes actually relate to high-taxed income, could give rise to significant unfairness. The unfairness could be even more acute in cases where the default rule affects the classification of losses. Under the proposed regulations, the payment of a small amount of foreign tax in a loss year could result in the relegation of losses to the residual category, and make them unavailable to offset other income.⁵

Section 245A(d) appropriately disallows credits for foreign taxes paid or accrued with respect to exempt income. We acknowledge the importance of preventing the use of strategies to generate duplicative benefits by separating exempt foreign income from creditable foreign taxes. But a rule that disallows credits for foreign taxes that can't be easily classified has the potential to create significant mischief.

The TCJA made sweeping changes to the U.S. tax treatment of foreign business income. The new rules can make it more difficult to make effective use of foreign tax credits, and in some cases can result in the disallowance of credits that would have been allowed under prior law.⁶ However, we know of no indication that Congress intended to create some sort of general presumption against foreign tax credits, or to limit the availability of credits in circumstances not specifically contemplated by the statute.

A core purpose of the foreign tax credit system is to prevent double taxation. Under the new rules, taxpayers get only one bite at the apple: credits that are not allowed currently will never be allowable. In view of this potential outcome, Treasury and the IRS should be extremely wary of rules that have the effect of shunting foreign taxes automatically into the disallowed category, or inappropriately dissociating foreign taxes from income. This can require vigilant attention: drafting points that might have been regarded as minor details under prior law can represent on/off switches under the TCJA.⁷

3. Subpart F income should not be subject to the consistency rule.

The proposed regulations would expand the consistency rule to include subpart F income. Under the expanded rule, taxpayers would be required to claim the benefit of the high-tax exclusion on a groupwide basis with respect to all qualifying items of GILTI and subpart F income. Under this all or nothing rule, a U.S. parent company that wishes to invoke the exclusion with respect to a single item of subpart F income derived by a single CFC would be deemed to have elected to exclude every item of high-taxed subpart F income and GILTI derived by every member of its CFC group.

We appreciate the reasons why the drafters thought that the rules governing the high-tax exclusion for subpart F and GILTI purposes should be applied consistently. We thought so too, and said so in our comment letter regarding the prior proposed regulations.⁸ The drafters concluded, however, that conforming the new GILTI exclusion to the longstanding rules governing the subpart F exclusion would afford inappropriate opportunities for taxpayers to get the best of both worlds, by allowing them to claim the benefit of the high-tax exclusion for some income, and to elect not to exclude other high-taxed income so that credits can be applied to eliminate residual U.S. taxes payable on lower-taxed income. The final regulations therefore provide that the benefit of the exclusion must be claimed on an all or nothing basis for GILTI purposes.

We don't think that the all or nothing rule can or should be extended to subpart F income. Section 954(b)(4) provides that foreign base company income “shall not include any item of income received by a controlled foreign corporation if the taxpayer establishes to the satisfaction of the Secretary that such income was subject to an effective rate of income tax imposed by a foreign country greater than 90 percent of the maximum rate of tax specified in section 11. This formulation specifically indicates that taxpayers may claim the benefit of the high-tax exclusion on an item-by-item basis. The statute has been construed for more than 30 years to allow taxpayers to elect to exclude some items of income and not others.

The preamble explains why the drafters believed that the consistency rule should be applied across the board, and not just for GILTI purposes. The drafters thought that, in the absence of an all-encompassing rule, taxpayers would have uneconomic incentives to restructure their foreign business operations to cause them to fail to qualify for a subpart F exception. Maybe so. Some companies may be tempted by these incentives; others may conclude that such restructurings are more difficult, and less rewarding, than they look.⁹

But even if the risk of tax-motivated restructurings is greater than we believe it to be, the statute says what it says. Section 954(b)(4) has been consistently interpreted to be elective for many years.¹⁰ If policymakers believe that taxpayers should no longer be permitted to claim the benefit of the subpart F high-tax exclusion selectively, the appropriate course of action is to amend the statute to say so.

Treasury and the IRS have made extraordinary efforts to provide practical guidance concerning the sweeping and complex changes made by the TCJA in a manner that is consistent with Congressional intent and within the scope of their authority. In this case, however, we respectfully suggest that Treasury and the IRS have overstepped their authority, and that taxpayers should continue to be allowed to claim the benefit of the subpart F high-tax exclusion on an item-by-item (and company-by-company) basis until Congress changes the rules.¹¹

4. The special 24-month limitation period is unnecessary and inappropriate.

The proposed and final regulations would allow taxpayers to make or revoke high-tax elections on amended returns only if those returns are filed within 24 months of the unextended due date of the original return for the year to which the election relates. The preambles say that Treasury and the IRS determined that this special limitation period is necessary to administer the high-tax exclusion and to allow the IRS to timely evaluate refund claims or make additional assessments. They requested comments regarding whether taxpayers should be allowed to make or revoke elections after the 24-month period if they can establish that the election or revocation would not result in time-barred tax deficiencies.

Audits and tax controversies in many foreign countries are more complex and can take significantly longer to resolve than in the United States. The U.S. tax system traditionally has taken account of this by providing special extended carryover periods and statutes of limitations for foreign tax credits. In the context of this longstanding history, we believe that there is no justification for providing a shorter limitation period for adjustments affecting the high-tax exclusion.

The special 24-month period is unnecessary and inappropriate. We encourage Treasury and the IRS to eliminate it. Alternatively, taxpayers could be required to certify that an election or revocation will not result in time-barred deficiencies.

5. A CFC owned by an investment fund should not be considered to be a member of a CFC group solely because an affiliate of the fund sponsor or investment manager is a general partner in the fund.

The proposed regulations provide that a controlled foreign corporation will be a member of a CFC group, and therefore will be subject to the consistency requirement, if a controlling domestic shareholder owns more than 50 percent of the value or voting power of the CFC's stock. If this definition is interpreted consistently with other voting power tests applicable to CFCs, then a foreign corporation that is owned by a private equity fund or similar collective investment vehicle normally should not be included in the CFC group of any partner in the fund, except in very unusual cases.¹²

This result is fair and appropriate, and avoids the potential for conflicts in cases where investors have disparate interests.¹³ In order to eliminate any potential for uncertainty concerning this point, the regulations should include an example confirming this outcome.

* * * *

We appreciate the opportunity to comment on these provisions. Please do not hesitate to contact me at (202) 962-7399 orjsok@sifma.org if you have questions or would like to discuss our comments in more detail.

Respectfully submitted,

Justin Sok
Managing Director, Tax
Securities Industry and Financial Markets Association (SIFMA)
New York, NY

cc:
David J. Kautter
Assistant Secretary for Tax Policy

L.G. “Chip” Harter
Deputy Assistant Secretary (International Tax Affairs)

Doug Poms
International Tax Counsel

Jason Yen
Attorney Advisor (Office of Tax Policy)

Peter Blessing
Associate Chief Counsel (International)

---

Examples

The following examples illustrate the application of the final and proposed regulations to activities conducted by non-U.S. members of a regulated financial services group. The structure and funding of such a group's activities typically is determined by considerations unrelated to U.S. taxes.¹⁴ This is relevant because, in cases where the regulations would produce an inappropriate result, it normally will not be possible to exercise self-help to avoid that result (for example, by changing the group's structure).

In examples 1 and 2, a U.S. parent company (“Parent”) holds its non-U.S. businesses through a foreign intermediate holding company (“Topco”), which is a controlled foreign corporation for U.S. tax purposes. Topco owns two tested units (“Sub₁” and “Sub₂”) that are organized in different jurisdictions. Except to the extent otherwise indicated, Topco is a holding company that does not conduct any activities directly. Example 3 involves a tiered ownership structure in which Parent owns Sub₁ directly (rather than through Topco), Sub₁ owns Sub₂, and both companies are CFCs. In this example, Sub₁ receives regulatory capital funding from Parent, and provides regulatory capital funding to Sub₂.¹⁵ In all three examples, the companies are treated as separate entities for non-U.S. tax purposes.

Sub₁ is a broker-dealer. Its assets are held in the ordinary course of its dealer business; it doesn't own significant hard assets. Except to the extent it qualifies for, and Parent elects to claim, the benefit of the high-tax exclusion, Sub₁'s income will be excluded from foreign personal holding company income under section 954(c)(2)(C), and will be tested income for GILTI purposes.

Sub₂ is engaged in the conduct of a distressed debt business. It isn't a dealer in securities, and hasn't made a mark-to-market election under section 475(f). Except to the extent it qualifies for, and Parent elects to claim, the benefit of the high-tax exclusion, Sub₂'s income will be general category subpart F income.

The final regulations can have unexpected and counterintuitive consequences if expenses are allocated differently, or income and expenses are taken into account at different times, for U.S. and foreign tax purposes. As indicated below, the proposed regulations would alleviate some of these problems, which is why we recommend that they be made available with respect to all post-TCJA years.

The examples are intended to illustrate some of the ways in which the use of inconsistent methodologies to allocate items of income and expense can produce distortions. The drafters of the proposed regulations appear to have appreciated this point, which is why those regulations would introduce an applicable financial statements standard in place of the final regulations' reliance on generally applicable expense allocation rules.

As currently drafted, the proposed regulations would allow taxpayers to allocate expenses by reference to applicable financial statements only in respect of periods beginning after the proposed regulations are issued in final form. We see no reason to defer the use of what clearly represents a more rational standard. Accordingly, we recommend that taxpayers be allowed to use the new methodology in respect of all relevant years. We encourage Treasury and the IRS to provide examples confirming that an applicable financial statements test will produce rational results, and avoid distortions, in the circumstances described below.

Example 1 deals with a case in which one of the two tested units can support significantly more leverage than the other. In Example 2, the tested units have the same relative leverage, but different costs of funds and returns on assets. Example 3 involves a case in which two identically situated CFCs would be treated differently because one owns the other. In each of these fact patterns, the expense allocation methodology prescribed by the final regulations would produce uneconomic results; the proposed regulations would significantly reduce the potential for distortions.

1. Different leverage.

Sub₁ and Sub₂ are tested units. Each of them has gross assets of $50 million, earns an average return of 2% on its assets, and can borrow money at a 1% rate. Sub₁'s business model and regulatory position allows it to support $40 million of debt. Its income is subject to country A tax at a 20% rate. Sub₁'s business model and regulatory position allows it to support $30 million of debt. Its income is subject to country B tax at a 15% rate.

Sub₁ has net income of $600,000 ($1 million gross income minus $400,000 interest expense) and pays country A tax of $120,000 ($600,000 x 20%).

Sub₂ has net income of $700,000 ($1 million gross income minus $300,000 interest expense) and pays country B tax of $105,000 ($700,000 x 15%).

If Sub₁ and Sub₂ had been CFCs, Sub₁'s income would have qualified for the high-tax exclusion and Sub₁'s income would not have qualified. The result would be the same under the proposed regulations, which provide for the allocation of expenses between tested units based on their financial statements.

The result would be different, and worse, under the final regulations. Considered together, Sub₁ and Sub₂ have total assets of $100 million and total interest expense of $700,000. If interest expense is allocated between them under the principles of Treasury regulations §1.861-9, then Sub₁'s interest expense would be reduced by $50,000 (allocated interest expense of $350,000 [$50 million ÷ $100 million x $700,000] is less than book interest expense of $400,000), and Sub₂'s interest expense would be increased by the same amount.

On these facts, neither tested unit would be high-taxed: Sub₁'s effective rate would be reduced from 20% to 18.5% ($120,000 tax ÷ $650,000 net income), or just short of 18.9%. Sub₂'s effective rate would be increased from 15% to 16.2% ($105,000 tax ÷ $650,000 net income).

2. Different funding costs and return on assets.

In this variation, Sub₁ and Sub₂ again are tested units. They have the same relative leverage ($50 million assets ÷ $40 million debt), and are subject to foreign tax at the same rate (20%). However, the business conducted by Sub₂ involves a slightly greater degree of risk and reward, which is reflected in a higher cost of funds (1.5% instead of 1%) and has a higher return on assets (2.5% instead of 2%). Sub₁ has the same results as in example 1. Sub₂ has net income of $625,000 ($1.25 million gross income [$50 million assets x 2.5%] minus $600,000 interest expense [$40 million liabilities x 1.5%]), and pays country B tax of $125,000 ($625,000 x 20%).

If Sub₁ and Sub₂ had been CFCs, both companies would have qualifies for the high-tax exclusion (20% > 18.9%). The result would be the same under the proposed regulations, which provide for the allocation of expenses between tested units based on their financial statements.

Under the final regulations, however, Sub₂'s income would not be high-taxed. Sub₁ and Sub₂ have total assets of $100 million and total interest expense of $1 million. If that expense is allocated between them under the principles of Treasury regulations §1.861-9, then Sub₂'s interest expense would be reduced by $100,000 (allocated interest expense of $500,000 [$50 million ÷ $100 million x $1 million] is less than book interest expense of $600,000), and Sub₁'s interest expense would be increased by the same amount.

Sub₁'s effective rate would be increased from 20% to 24% ($120,000 tax ÷ $500,000 net income). The increase would not produce any benefits for the group: Sub₁'s income would continue to qualify for the high-tax exclusion, and country A taxes would not be allowable as credits. Sub₂'s effective foreign tax rate would be reduced from 20% to 17.2% ($125,000 tax ÷ $725,000 net income), and its income would no longer qualify for the high-tax exclusion.

3. Tiered ownership structures.

In this variation, Sub₁ and Sub₂ are CFCs, and Sub₁ owns Sub₂. They have the same relative leverage ($50 million assets ÷ $40 million debt), and the same return on assets and cost of funds (2% and 1%). Their income is subject to foreign tax at a 20% rate. Each company has net income of $600,000 and pays $120,000 of foreign tax.

Sub₁'s liabilities include a $10 million medium-term loan from its U.S. parent company. Sub₂'s liabilities include a $10 million medium-term loan from Sub₁. In order to qualify for favorable regulatory capital treatment, this funding must be received from a direct parent company.

Under the proposed regulations, both companies would qualify for the high-tax exclusion (20% > 18.9%). Under the final regulations, however, only Sub₂ would be considered high-taxed. The final regulations provide that, for purposes of determining whether Sub₁'s income is high-taxed, interest expense may not be taken into account, and must be added back to net income, to the extent the expense is allocable under U.S. tax principles to assets that give rise to exempt income. If Parent elects to claim the benefit of the high-tax exclusion, then Sub₁'s shares in Sub₂ would fall into this category.

Those shares represent one-fifth of Sub₁'s assets ($10 million ÷ $50 million). If Sub₁ uses an asset method to allocate interest expense, then one-fifth of its interest expense, or $80,000, would be allocated to its shares in Sub₂. Under the final regulations, that amount would not be taken into account in determining its effective tax rate. As a result, even though Sub₁ and Sub₂ are subject to identical tax rules, Sub₂'s income would be considered high-taxed, and Sub₁'s wouldn't.

Sub₁ would be considered to be subject to country A tax at a 17.6% rate ($120,000 tax ÷ $680,000 adjusted income [net income of $600,000 plus 'disallowed' interest expense of $80,000]). Accordingly, its income would not qualify for the benefit of the high-tax exclusion. Moreover, Sub₁'s tested income, and Parent's GILTI, would be increased by $80,000. This will dilute the benefit of credits for foreign taxes paid by Sub₁, and could increase Parent's U.S. tax liability.

We believe that expenses incurred by an upper-tier CFC should be taken into account in determining whether the CFC's income qualifies for the high-tax exclusion, except to the extent those expenses are properly allocable to exempt income received by the CFC itself. The ownership of shares in a lower-tier CFC should not be a basis for effectively disallowing deductions.