Foreign Branch Income Category Is SIFMA’s Focus in FTC Regs

January 28, 2021

Internal Revenue Service
CC:PA:LPD:PR (REG-101657-20)
Room 5203, Post Office Box 7604
Ben Franklin Station
Washington, DC 20044

Re: Proposed foreign tax credit regulations

Ladies and Gentlemen:

The Securities Industry and Financial Markets Association¹ appreciates this opportunity to comment on proposed foreign tax credit regulations (the “proposed regulations”) that were published in the Federal Register on November 12, 2020.²

We commend the Treasury Department (“Treasury”) and the Internal Revenue Service (“IRS”) for their efforts to develop a workable framework for complex new rules, and for their constructive engagement with the issues. The proposed regulations, together with final and temporary regulations (the “2020 final FTC regulations”) issued on the same day,³ address many of the concerns that taxpayers had raised regarding the prior proposed regulations (the “2019 proposed FTC regulations”).⁴

SIFMA welcomes the attention paid to branch category income in the proposed regulations. Banks historically have conducted very substantial activities through foreign branches as well as through foreign subsidiaries. The determination whether to conduct particular activities through a branch or a subsidiary typically is dictated by regulatory and financial considerations unrelated to U.S. taxes. The computation of foreign branch category income therefore is a critical area of interest for our members. Part I of this letter discusses provisions of the proposed regulations dealing with foreign branch category income, including the computation of branch interest expense and the treatment of related-party funding transactions. Part II discusses other aspects of the proposed regulations, including the definition of financial services income, the jurisdictional nexus requirement, the refinements to the rules for determining whether a foreign tax is noncompulsory, and the changes to the treatment of CFC-to-CFC debt for purposes of the CFC netting rule.

I. Foreign branch category income.

A. Background.

Under the proposed regulations, a bank would determine the amount of interest expense that is allocable to branch category income primarily by reference to the borrowing costs actually incurred by its foreign branches. We strongly commend the drafters for having adopted this approach. Determining branch interest expense by reference to actual borrowing costs will produce results that are more closely correlated with the branch's economic income, and will reduce the potential for mismatches between taxable income for U.S. and foreign tax purposes. The proposed new standard clearly represents a significant improvement over the current regulatory framework.

The preamble to the proposed regulations asks for further comments regarding whether:

The proposed regulations should be refined to take appropriate account of related-party funding transactions and differences in relative leverage. The 2019 branch letter provides a comprehensive account of the reasons why these issues are so important to our industry. That letter describes a possible methodology for addressing those issues by making upward or downward adjustments to branch interest expense. The methodology was developed to address the objectives and concerns raised by the government regarding these issues. It is important to emphasize, however, that our suggested methodology is not the only possible means of getting to the right answer.

This letter therefore focuses on the need for a solution, rather than on the advantages and disadvantages of any particular methodology. Getting to the right answer is of paramount importance; the route taken to reach that destination is less significant. If you believe that the approach described in the 2019 branch letter is unduly complex, or is otherwise problematic, we would be happy to work with you to develop fair and workable rules.

B. Related-party funding transactions.

Foreign branches of U.S. banks enter into a large number of interbranch and intercompany funding transactions. Foreign branches derive very substantial net income from such transactions, and pay a correspondingly large amount of foreign taxes — in some cases in the hundreds of millions of dollars annually. The income is attributable to activities conducted in the countries where the branches are located. These intercompany funding transactions are not contrived or artificial; they are part of the ordinary-course conduct of a global banking business. The transfer of funds between business locations (from the place where money is raised to the place where money is needed) is an indispensable tool in managing a global banking business. Interbranch and intercompany funding transactions are routine, ordinary-course transactions.

The preamble to the 2019 proposed FTC regulations takes the view that interbranch funding transactions are essentially comparable to capital contributions and remittances. We don't think this is the right analogy. From the perspective of our global businesses, such interbranch transactions are arm's length financings. That is their economic substance, and how they are treated for foreign regulatory, accounting and tax purposes.

Income from related-party funding transactions can represent a significant proportion of a branch's total income. Foreign countries appropriately subject that income to tax.⁶ From a foreign tax perspective, it typically is a matter of indifference whether items of income and expense are disregarded for U.S. tax purposes (as in the case of interbranch transactions), or are subject to consolidated return matching rules (as in the case of intercompany transactions in which one of the parties is acting through a foreign branch).

It is important for the regulations to eliminate structural obstacles to the availability of foreign tax credits for income derived from these ordinary-course transactions. A failure to prescribe workable rules could give rise to significant distortions, and could make it more difficult for U.S. banks to compete effectively in foreign markets.

We acknowledge the difficulty of this exercise, which will require Treasury and the IRS to address interlocking considerations relating to the measurement, sourcing and allocation of items of income and expense.⁷ But if Treasury and the IRS don't dig into these issues, banks will be left with unpredictable, and potentially very substantial, mismatches between the treatment of branch income for U.S. and foreign tax purposes. We would be happy to work with you to develop streamlined or simplified solutions, and to address any concerns that you may have.

C. Relative leverage.

Banks are required to maintain very substantial equity capital. Regulatory capital requirements were strengthened in the aftermath of the 2008 financial crisis. A bank's capital is available to support all of its activities, without regard to the location of particular assets and liabilities.⁸ If capital is recorded on the books of a bank's U.S. home office, it nevertheless will be available to creditors of the bank's foreign branches. The outcome would be the same if the situation were reversed, and the bank's capital for some reason was reflected solely on the books of its foreign branches.

In view of the importance of capital to our global businesses, we believe that adjustments should be made to ensure that the bank's foreign branches and its U.S. home office (i.e., the foreign branch owner)⁹ have the same liability-to-asset ratio. Under this approach, liabilities and interest expense would not be directly allocated to foreign branches, and instead would be attributed to the home office, to the extent necessary to equalize the home office and foreign branch ratios. By the same token, if a bank's foreign branches are underleveraged, home office liabilities and interest expense would be allocated to the branches. In this regard, the liability-to-asset ratio of a bank's foreign branches, as reflected in the branches' financial statements, has very little practical significance, because the branches don't exist as separate entities — they are just as much a part of the bank as the home office.¹⁰

It is easily possible to envision circumstances in which a foreign branch would have a liability-to-asset ratio that is significantly greater than the bank's overall ratio. If a bank has a 90% liability-to-asset ratio, the financial statements of its foreign branches could reflect a ratio of 100% (matching assets and liabilities of $200 billion) or even 125% (matching assets and liabilities of $250 billion for foreign tax purposes; $50 billion of those assets consist of loans from foreign branches to the U.S. home office that are disregarded for U.S. tax purposes).¹¹ The higher relative leverage would not reflect a judgment that the foreign branch is somehow more creditworthy than the home office — it isn't. On these facts, we think that branch liabilities should be capped at a level determined by multiplying branch assets times the bank's overall liability-to-asset ratio (90% of $200 billion, or $180 billion).¹²

D. Other comments regarding branch category income.

1. Non-interest-bearing deposits should count.

Obligations of foreign branches should qualify for the benefit of the direct allocation and asset adjustment rules without regard to whether the obligations bear interest or not. Thus, if a branch has $50 billion of non-interest-bearing deposits, the value of its assets for purposes of Treasury regulations §1.861-9 (allocation of interest expense in excess of amounts directly allocated under Proposed regulations §1.861-10(g)(1)) should be reduced by the same amount. This seems almost too obvious to bear mentioning; perhaps for this reason, the proposed regulations don't mention it.¹³ In order to avoid any conceivable questions, the regulations should provide explicitly that non-interest-bearing obligations will be taken into account for direct allocation purposes, or should provide an example illustrating that outcome.

2. Effective date.

Banks should be allowed to apply the new methodology for determining branch interest expense retroactively in respect of all years after the entry into force of the TCJA.

II. Other issues.

A. Financial services income.

The proposed regulations resolve virtually all of the concerns that taxpayers had raised concerning the definitions of “financial services entity” and “financial services group” in the 2019 proposed FTC regulations. In their original form, the definitions would have excluded a significant number of genuinely active financial services businesses. We commend Treasury and the IRS for their efforts to address taxpayers' concerns.

We have one further recommendation. As currently drafted, the regulations would treat some related-party payments as financial services income, and other essentially identical payments as passive income, depending on whether the payor and the recipient are section 1504(a) affiliates. We believe that income derived by or received from a related party should not be treated differently solely because the related party is held through a partnership or other flow-through entity.

The problem, and the possible solutions, can best be understood by comparing three fact patterns:

There is no reason to treat amounts received by a financial services entity as passive income solely because the payor is related to the recipient, but is not a section 1504(a)(2) affiliate, and therefore is not a member of the recipient's financial services group. The principal circumstance in which this issue could arise is where a financial services group receives interest from a related entity that it owns through an intermediate holding partnership. Such an ownership structure is not uncommon, and can exist for a variety of reasons.

The classification of amounts as financial services income should not depend on whether a financial services group owns its subsidiaries through holding companies that are corporations or partnerships for U.S. tax purposes. This problem could be remedied either by:

B. Nexus.

The proposed regulations would require foreign income taxes to meet a new jurisdictional nexus requirement in order to be allowable as credits. The preamble says that the proposal is intended to ensure that there is substantial conformity in the principles used to calculate the base of the relevant foreign tax and the base of the U.S. income tax, including with respect to the determination of whether there is a sufficient nexus between the income that is subject to tax and the foreign jurisdiction imposing the tax. The proposal appears to be targeted at controversial new assertions of foreign taxing jurisdiction over digital businesses, such as the digital service taxes recently enacted by France and other members of the European Union, but as currently drafted it is not limited to newly-enacted taxes.

The preamble requests comments on the extent to which these jurisdictional nexus requirements could affect the treatment of other foreign taxes, and on alternative approaches that Treasury and the IRS could deploy to achieve the objectives described above.

We believe that it would be preferable to introduce a new nexus requirement after the successful completion of pending multilateral tax discussions, rather than by unilateral administrative action. If Treasury and the IRS disagree, we encourage them at least to take account of the fact that the nexus requirement could affect the treatment of a wide range of existing foreign taxes, some of which have been claimed as credits for decades, and to endeavor to limit the scope of the requirement to its intended target.

The proposed regulations provide that, in order for a foreign income tax to be creditable, the tax must be “limited to income that is attributable, under reasonable principles, to the nonresident's activities within the foreign country,” including under “rules similar to those for determining effectively connected income under section 864(c).” In this regard, it is important to note that the U.S. tax system cannot easily be described as normative. The system includes some features (including provisions enacted as part of the TCJA) that are inconsistent with the tax systems of most other developed countries, and other features (including some of the rules governing effectively connected income¹⁵) that are badly in need of updating. This is not intended as criticism — the United States is a sovereign country, and Congress can adopt whatever rules that it deems appropriate — but it would be surprising if any foreign country applied exactly the same nexus requirements as the United States.

A number of existing foreign income taxes apply in circumstances that would not attract net income taxation under the principles of section 864(c). For example, a number of countries (including, in some cases, the United States) tax gain deemed to have been realized upon the transfer of an indirect interest in a business conducted in the taxing state.¹⁶ The drafters of the proposed regulations appear to have contemplated that such taxes would continue to be creditable to the same extent as under current law, but it is not clear that they have accomplished that objective. Other countries apply classification or sourcing rules that do not correspond in every respect to U.S. tax principles (for example, by characterizing payments as royalties rather than compensation for services, and vice versa, or by taxing income from the performance of related-party services based on the location of the services recipient rather than the place where the services were performed¹⁷). Accordingly, the new nexus requirement could be construed to disallow credits for a wide range of foreign taxes that traditionally have been claimed as credits.

If Treasury and the IRS believe that a new nexus requirement is necessary, they should limit its application to its intended target. If their concern relates to newly-adopted taxes on digital services, the regulations should say so. The regulations should confirm that the nexus requirement will not affect the creditability of foreign taxes imposed outside that limited factual context.

C. Noncompulsory taxes.

In order for a foreign tax payment to be allowable as a credit, the taxpayer must establish that the tax represents a compulsory exaction. The proposed regulations would expand and clarify the discussion of this longstanding requirement. The changes are sensible and helpful, with the limited exception discussed below.

Under the noncompulsory tax rules, if a foreign tax authority claims that additional taxes are due, the taxpayer cannot simply pay the amount of the asserted liability and then shift the cost to the U.S. Treasury by claiming foreign tax credits. Instead, if the asserted liability is inconsistent with a fair interpretation of foreign law, the taxpayer must contest the assessment in administrative proceedings or in court. In determining whether and how to pursue such a contest, taxpayers may rely on advice received from foreign tax experts regarding the prospects of success, and may make judgments based on a cost-benefit analysis. If the taxpayer reasonably determines that the cost of pursuing a foreign tax contest exceed the potential benefits, it is not required to pursue the contest past the point of good sense.¹⁸

The proposed regulations, however, would create an exception to this rule. Under that exception, taxpayers would not be permitted to take foreign non-income tax costs into account in determining whether the costs of pursuing a foreign tax contest exceed the benefits.¹⁹ This exception is inappropriate and unnecessary. For purposes of the cost-benefit analysis, taxpayers may take a wide range of tangible and intangible costs into account, including legal and accounting fees; the costs associated with extensive discovery; management time; interest and penalties; the implications of a long-running controversy for other pending tax matters or for the taxpayer's relationship with the foreign tax authority; regulatory consequences; and reputational issues. There is no reason to treat noncreditable foreign taxes (for example, VAT, transaction taxes and customs duties) any differently than other costs.

D. CFC netting.

The proposed regulations would revise the CFC netting rules of Treasury regulations §1.861-10(e)(8)(v) to address perceived distortions. As Treasury and the IRS noted when they requested comments regarding the CFC netting rules in a 2018 notice of proposed rulemaking,

Many of the existing expense allocation rules have not been significantly modified since 1988. Furthermore, for taxable years beginning after December 31, 2020, a worldwide affiliated group will be able to elect to allocate and apportion interest expense on a worldwide basis. See section 864(f). The Treasury Department and the IRS expect the implementation of section 864(f) will have a significant impact on the effect of interest expense apportionment and will necessitate a reexamination of the existing expense allocation rules.

Therefore, the Treasury Department and the IRS expect to reexamine the existing approaches for allocating and apportioning expenses, including in particular the apportionment of interest, research and experimentation (“R&E”), stewardship, and general & administrative expenses, as well as to reexamine the “CFC netting rule ” in §1.861-10(e).²⁰

We believe that any changes to the CFC netting rules should be closely coordinated with the development of guidance under section 864(f). We encourage Treasury and the IRS to withdraw the proposed changes for further consideration in connection with that project.²¹

* * * * *

We very much appreciate the opportunity to comment on the proposed regulations. Let me know if you have questions, or would like to discuss our comments in more detail.

Respectfully submitted,

Sincerely,

Justin Sok
Managing Director, Tax
Securities Industry and Financial Markets Association
Washington, DC

cc:
Mark Mazur
Deputy Assistant Secretary, Tax Policy

Jason Yen
Attorney Advisor (Office of Tax Policy)

Barbara Felker
Chief, Branch 3, Office of Associate Chief Counsel (International)