NYSBA Tax Section Report Addresses Proposed PFIC Regs

April 14, 2021

The Honorable Mark J. Mazur
Deputy Assistant Secretary (Tax Policy)
Department of the Treasury
1500 Pennsylvania Avenue, NW
Washington, DC 20220

The Honorable William M. Paul
Chief Counsel (Acting)
Internal Revenue Service
1111 Constitution Avenue, NW
Washington, DC 20224

The Honorable Charles P. Rettig
Commissioner
Internal Revenue Service
1111 Constitution Avenue, NW
Washington, DC 20224

Re: Report No. 1450 — Report on the Proposed “PFIC” Regulations under Sections 1297 and 1298

Dear Messrs. Mazur, Rettig, and Paul:

I am pleased to submit our Report No. 1450 commenting on the proposed regulations published in January 2021 relating to passive foreign investment companies.

We commend the Internal Revenue Service and the Department of the Treasury for the thoughtful guidance issued with respect to passive foreign investment companies in the January 2021 final regulations, as well as the proposed regulations. This Report comments on issues relating to the proposed regulations.

We appreciate your consideration of this Report. If you have any questions or comments, please feel free to contact us and we will be glad to assist in any way.

Respectfully Submitted,

Gordon E. Warnke
Chair
Tax Section
New York State Bar Association

Enclosure

---

Report No. 1450

New York State Bar Association Tax Section

Report on the Proposed “PFIC” Regulations Under Sections 1297 and 1298

April 14, 2021

---

TABLE OF CONTENTS

1. Introduction

2. Summary of Recommendations

3. Reliance on Financial Statements for Asset Test

4. Treatment of Goodwill

5. Treatment of Working Capital

6. Look-through Subsidiary Dividends and Gains

7. Active Banking Exception

---

1. Introduction

This Report¹ comments on the proposed regulations under Sections 1297 and 1298 (the “2021 Proposed Regulations”) published in the Federal Register on January 15, 2021.²

The Department of the Treasury (“Treasury”) and the Internal Revenue Service (“IRS”) previously issued proposed regulations in July 2019 providing guidance related to the tests for classification of a foreign corporation as a passive foreign investment company (“PFIC”) (including the PFIC income test, the PFIC asset test, the subsidiary look-through rule, the active banking exception, the insurance income exception, and the change of business exception) (the “2019 Proposed Regulations”).³ Final regulations, adopting various of these proposed changes with certain modifications, were informally released on the IRS website on December 4, 2020, and published in the Federal Register on January 15, 2021 (the “2021 Final Regulations”).⁴ At the same time, the 2021 Proposed Regulations were issued relating to certain issues that the government felt merited additional consideration and study.⁵

Before the 2019 Proposed Regulations, the only generally applicable guidance for applying the income and assets tests under Section 1297(a) to determine whether a foreign corporation was a PFIC was Notice 88-22, 1988-1 C.B. 489 (“Notice 88-22”), which the 2019 Proposed Regulations and 2021 Final Regulations largely follow. Treasury and the IRS had also issued Notice 89-81, 1989-2 C.B. 399, which provides guidance regarding the application of the income test to foreign banks and securities dealers (the “Active Bank Notice”). In 1995, Treasury and the IRS issued proposed regulations under what is currently section 1297(b)(2)(A), adopting somewhat different rules for the active bank exception (the “1995 Proposed Bank Regulations”).⁶

We previously submitted a report on the 2019 Proposed Regulations (the “2019 PFIC Report”).⁷ Treasury and the IRS presumably gave careful consideration to the various comments submitted and this Report therefore generally will not revisit issues that were resolved by the 2021 Final Regulations. This Report will instead focus on the 2021 Proposed Regulations and issues on which the preamble to the 2021 Proposed Regulations (the “2021 Proposed Preamble”) requested further comments. The Report does not comment on the provisions in the 2021 Proposed Regulations proposing revisions to the qualifying insurance corporation rules or those related to the safe harbors for the domestic subsidiary anti-abuse rule.

2. Summary of Recommendations

The Report's primary recommendations are the following:

A. Reliance on Financial Statements

We welcome the government's willingness to consider allowing in some cases the use of amounts reported on financial statements as a proxy for fair value. As discussed in this Report, whatever the statute and regulatory rules, in reality, few if any tested foreign corporations themselves will undertake asset by asset valuations to determine fair market value for this reason and they cannot be legally compelled to do so. In most cases, U.S. shareholders of the tested foreign corporation will be unable to undertake this exercise even if they were willing to incur the cost and expense because they lack direct access to the relevant information. Accordingly, we believe the use of financial statements for this purpose is essential and should not be limited to non-public companies.

We also recommend that the proposed rule determining when “more reliable” valuation information rather than financial statement values must be used be revised and clarified. We believe a “reason to know” standard, as currently proposed, will create uncertainty that defeats the administrative simplicity benefits of allowing taxpayers to rely on financial statement values.

B. Treatment of Goodwill

Because Treasury and the IRS have determined that some businesses (primarily financing related businesses) that are actively conducted in a colloquial sense may nevertheless be “passive” for PFIC purposes, certain “goodwill” (which, by definition, requires an expectation of future active business conduct) may arguably be considered a passive asset. Accordingly, we understand what motivated the government to require bifurcating goodwill into passive and active components for PFIC asset test purposes. Nevertheless, we think situations in which goodwill is associated with “passive” assets are unusual. Further, we question whether the fact an active business is treated as giving rise to income that is PFIC income by failing to qualify for an applicable safe harbor should necessarily result in viewing the associated goodwill as passive. We believe exceptions that allow treatment of financing business income as active (such as the active banking or active leasing exceptions) are deliberately restrictive because of the difficulties in administering and enforcing a more subjective rule in that context based on how “actively” the business generating the income is conducted. It is difficult to distinguish active financing from passive investment income on that basis. It does not follow, however, that goodwill (which for reasons explained in the Report, necessarily reflects active business conduct) should also be considered a passive asset. The Report therefore requests that Treasury and the IRS reconsider the need for goodwill associated with such businesses to be treated as passive.

If they are unwilling to do so, however, we recommend that final regulations then make clearer that this bifurcation requirement is an exception to a more general presumptive principle that goodwill is generally active when the company conducts an active operating business and is applicable to limited situations involving specified financing businesses that fail to satisfy the particular statutory or regulatory exceptions to foreign personal holding company income treatment. This presumption could be limited in the first instance to a tested foreign corporation that conducts (applying the activity attribution rules already included in the regulations) a substantial operating business (for example, a trade or business within the meaning of Treasury Regulations section 1.367(a)-2(d)(2) that would qualify as actively conducted under Treasury Regulations section 1.367(a)-2(d)(2)). Further, the presumption would not apply to such a tested foreign corporation if it also actively conducts directly or through subsidiaries (applying the activity attribution rules already included in the regulations) as a trade or business (under such principles) any of a set of specified business activities, which could be expanded by future guidance. These could include (whether conducted physically or virtually online) any material lending, financing, commodity-related, real estate-related or insurance business activities or leasing or licensing activities that generate rental or royalty income. In that event, the presumption would not apply and an allocation between passive and active goodwill would be required (under the principles currently in the 2021 Proposed Regulations).

C. Treatment of Working Capital

We applaud the willingness of Treasury and the IRS to consider treating cash assets that represent up to 90 days of expenses as active working capital for PFIC asset test purposes. We agree it is desirable to have bright line rules in this context. However, the 90-day of expenses limit will significantly limit the utility of the exception for certain kinds of businesses with significant capital development costs and a long “tail” period before investments may generate revenue. We recommend that a taxpayer also be permitted to demonstrate by a preponderance of the evidence that (1) a tested foreign corporation's working capital in excess of that 90 day threshold is consistent with the reasonable needs of its business and (2) the amount treated as working capital is not disproportionate to the working capital maintained by most other substantially similar businesses in that industry or sector. It is unlikely that most such businesses will have PFIC sensitivities (because, for example, some competitors may be U.S. companies and even foreign competitors may have few, if any, U.S. shareholders) and this will provide a reasonable benchmark for the amount of working capital a business needs regardless of PFIC considerations.

In addition, as discussed in this Report, we recommend that the 90-day rule, where applicable, be clarified to be a “bright line” test rather than a “cap”.

We also recommend that “active” working capital include capital invested in certain low yielding, high grade and highly liquid, interest-bearing assets, provided that the portfolio designated as comprising the company's working capital for PFIC asset test purposes has an average weighted tenor of less than 90 to 180 days. Failing to permit investment in such interest yielding assets will largely negate the working capital relief provided by the 2021 Proposed Regulations as companies potentially will have to choose between PFIC classification and prudent treasury management. As a practical matter, companies have no incentive to inflate their working capital (so defined) for PFIC purposes as the negative spread between their cost of capital and the yield on high investment grade short term debt securities would make using this exception for “disguised” investment activity an expensive and unsustainable proposition. For reasons discussed in this Report, in our view Treasury and the IRS have the requisite regulatory authority to extend a working capital exception to interest-bearing working capital.

D. Look-through Subsidiary Dividends and Gain

We commend Treasury and the IRS for proposing to remove the requirement that dividends received by a tested foreign corporation from its look-through subsidiaries are eliminated only if the dividend is from earnings previously taken into account under the look-through subsidiary rules. However, we further recommend that the proposed rule requiring basis reduction for dividends not previously taken into account under the look-through subsidiary rule that was proposed in the 2021 Proposed Regulations be eliminated. Requiring U.S. taxpayers to reconstruct a foreign corporation's historic earnings and profits (“E&P”) in the PFIC context in order to make such reductions is impractical and is not administrable for reasons explained in this Report. The rationale for this requirement provided in the 2021 Proposed Preamble — that this rule is necessary to prevent inappropriate elimination of gain on a future disposition of the look-through subsidiary shares — seems uncompelling when compared to the practical burdens this rule will impose. The approach to gain on disposition of look-through subsidiary shares, which is what creates the potential anomaly, itself has a questionable technical and policy basis. In the context of an anti-abuse regime like the PFIC regime, we think this complexity is unnecessary and there is little practical risk of tax planning absent these complex rules. However, to prevent the use of planning that exploits dividends out of pre-PFIC earnings to reduce gain, Treasury and the IRS could adopt an anti-abuse rule which requires a reduction in basis on a sale of look-through subsidiary shares within two years of a prior dividend, by an amount corresponding to the dividend, unless the taxpayer can demonstrate the dividend was out of “taken into account” look-through subsidiary earnings.

E. Active Banking Exception

We recommend that the rules for the active banking exception be synthesized and there be a single regulatory regime rather than requiring taxpayers to apply, alternatively, the proposed regime for traditional banks based on section 954(h) set forth in 2021 Proposed Regulations section 1.1297-1(c)(2) (the “Modified 954(h) Regime”), the 1995 Proposed Bank Regulations or the Active Bank Notice. We see no good policy reason for retaining multiple, similar but somewhat inconsistent tests. The Report provides some detailed recommendations as to how to reconcile the approach of the different rules. It is critical that any active bank regime cover most institutions engaged in bona fide “traditional” banking as that business is actually conducted in today's real-world financial markets and that final regulations therefore adopt the approach most consistent with the purposes of the PFIC regime and the statute while reflecting current market realities to the greatest extent possible, whether that means following the 1995 Proposed Bank Regulations, the Modified 954(h) Regime, the Active Bank Notice, some combination of the three, or none of them.

We agree that section 954(h), insofar as that section relates to traditional banks, is a helpful starting point for integrated rules but think certain modifications are needed in light of the policy considerations peculiar to the PFIC regime, which are not present under Subpart F (on which section 954(h) is focused). Specifically, we believe it is essential to include rules for “qualified non-bank affiliates” within a banking group similar to those under the 1995 Proposed Bank Regulations. We also believe final rules should focus less than section 954(h) does on whether various activities or income are derived in the tested unit's “home” country or taxable there. To prevent abusive planning, such as obtaining a bank license from a light touch regulatory jurisdiction in which the institution has no material activity, we think it makes sense to require some material quantum of “core” traditional banking activity (such as lending) be conducted in the relevant bank licensing country or countries. However, once this level of activity in the licensing jurisdiction is sufficient to demonstrate that the institution is a bona fide licensed bank, we do not think where its activity is otherwise conducted geographically or how it is taxed should matter in determining whether the banking income is “active” for PFIC purposes.

3. Reliance on Financial Statements for the Asset Test

Under section 1297(e), the determination of whether a tested foreign corporation satisfies the asset test either is made on the basis of the fair market value (not tax basis or financial statement values) of the assets of the tested foreign corporation, unless the shares of the tested foreign corporation are not publicly traded and the tested foreign corporation is a controlled foreign corporation (“CFC”) or otherwise elects (in which case the tax basis of assets for determining earnings and profits is used). Comments to the 2019 Proposed Regulations noted that taxpayers frequently use financial statements to determine the value of a tested foreign corporation's assets given the substantial cost and burden of obtaining PFIC specific asset-by-asset appraisals or valuations. The 2021 Proposed Regulations would permit the use of financial statements for the PFIC asset value test determinations in some cases.

The decision to allow the use of financial statements for PFIC asset test valuation purposes is welcome. As emphasized throughout this Report, it is not the tested foreign corporation itself, but its U.S. shareholders, that ultimately are legally required to undertake this exercise. While Congress, Treasury and the IRS have the authority as a matter of U.S. tax law to require asset-specific fair value appraisals for PFIC purposes, they have no jurisdictional power in most cases to compel a foreign corporation to do this. As a practical matter, for many companies, it is simply unrealistic to expect that they will undertake some form of independent valuation or appraisal on an asset-by-asset basis to accommodate what may well be only a small minority of shareholders who are U.S. taxpayers with PFIC sensitivities. Such shareholders, even if they have the resources to engage in such an appraisal in the abstract, may lack the ability to access the information from the company needed to carry this out. Adopting a requirement to determine PFIC status only based upon asset-specific valuations will simply compel many small U.S. shareholders to ignore the law or treat non-PFICS as PFICs. Financial statements are in most cases the best (or only) information available to a shareholder for this purpose.

Financial statement valuations may, in theory, diverge from “true” fair market value and reach results that a different approach to valuation would not. On the other hand, particularly for real operating companies with non-traded assets, it is far from clear that internal valuations, outside appraisals or other non-market valuation metrics of operating assets are likely to be substantially more reliable measures of “true” fair value. Thus, the idea that there exists in that context some more reliable “true” value that is discoverable with sufficient effort is somewhat unrealistic for many assets. Indeed, insofar as the kinds of valuations or appraisals above generally would be prepared specifically with PFIC analysis in mind, there may be a greater risk of PFIC results-oriented valuation than is the case with financial statements that are used for multiple commercial purposes.

Using financial statements as a proxy for fair value is not generally “favorable” to taxpayers. Financial statements, as we understand it, generally are more likely to carry liquid (i.e., potentially “passive”) assets at current market value whereas operating (potentially “active”) assets typically reflect depreciated acquisition cost. In the round, using financial statements as a proxy for value tends to understate the operating (active) assets and overstate the financial (passive) assets of a business although, of course, there may be occasional circumstances in some businesses in which that is not the case.

In practice, allowing shareholders to use financial statements for this purpose reaches similar results to a somewhat analogous safe harbor. FIRPTA regulations similarly require the use of “value” in general when determining whether U.S. real property interests exceed 50% of the applicable denominator such that a U.S. corporation is a U.S. real property holding corporation (“USRPHC”).⁸ Treasury Regulations section 1.897-2(b)(2), however, permits a taxpayer to presume that the test is not met if the test based on financial statement values results in U.S. real estate assets being less than 25% of the applicable denominator using such financial statement values. As discussed above, financial statements generally understate the value of operating assets relative to the value of passive financial assets, which already results in a form of “haircut” analogous to that of the FIRPTA regulation when financial statements are used for PFIC asset test purposes.

Treasury and the IRS have broad regulatory authority under section 1298(g) to regulate to achieve the purposes of the PFIC rules. We believe this grants them abundant regulatory authority to prescribe the methodologies they will accept in determining asset “values” in light of the theoretical constraints and administrative realities mentioned above. The decision to allow the use of financial statement values is a sensible and practical accommodation to commercial reality for reasons of administrative convenience. That emphasis on practicality should also guide the approach to the issues raised by the government in the 2021 Proposed Preamble.

A. Scope of Safe-harbor Relief

Proposed Regulations section 1.1297-1(d)(1)(v)(D) generally allows a taxpayer to rely upon the information in a tested foreign corporation's financial statements in order to determine the value of the corporation's assets for a quarter in which the tested foreign corporation is not publicly traded. Although the 2021 Proposed Preamble states that 2021 Proposed Regulations section 1.1297-1(d)(1)(v)(D) “generally permits a taxpayer to rely upon the information in a tested foreign corporation's financial statements in order to determine the value of the corporation's assets”, in fact it appears this reliance is only available to a tested foreign corporation that is a non-traded, non-CFC (and, perhaps, also to a non-traded look-through subsidiary of a traded company as discussed below).⁹ Notwithstanding the broad preamble language, we assume the above currently is the intended result.

If a look-though subsidiary is not itself traded, arguably for purposes of determining whether that subsidiary company is a PFIC (i.e., when that subsidiary is the “tested foreign corporation”) it may be able to rely on financial statement values under the regulatory language which requires that “the shares of a tested foreign corporation are not publicly traded.” Conversely, however, that language literally may prevent reliance on the valuations so derived in “tiering up” those values when testing the parent. That result seems somewhat contrary to the approach in Treasury Regulations section 1.1297-1(d)(1)(v)(C) which generally seeks consistency for purposes of asset valuation methods insofar as this relates to the use of fair value rather than tax basis.¹⁰ Moreover, inconsistent PFIC results could arise depending on the happenstance of whether the tested parent holds assets directly or through subsidiaries. Regardless of the ultimate resolution of the issues discussed below, this should be clarified.

One rationale for precluding reliance on financial statements when the tested foreign corporation is traded may be that public trading prices permit the determination of a market-based fair value for the gross assets (i.e., by adding back liabilities to the net equity value ascribed to the company by the market). That is obviously true. Indeed, we agree that such fair value serves as an appropriate starting point for applying the asset test. On the other hand, it is still necessary to allocate that gross aggregate fair value among the assets based on their relative fair market values. Effectively, that then poses the same asset-by-asset valuation challenge for a traded company that is faced for a non-traded company. The issue regarding look-through subsidiary assets above is just a special instance of this more general problem, namely, that the individual assets themselves are not traded. Accordingly, the existence of a trading price does not negate the need for taxpayers potentially to rely on financial statements for purposes of asset-specific values in an appropriate way.

In practice, it is much more likely that there will be very small U.S. shareholders who are unable to influence the company to provide the necessary information in a public company context than in the case of a non-traded foreign corporation. For securities law and other reasons, it is unusual for unsophisticated shareholders to invest in non-traded foreign corporations. Typically, a private company investment would be made, for example, by a sophisticated private equity fund acquiring a substantial stake that would at least have the opportunity to negotiate specific PFIC information rights. Accordingly, we question whether this is a good reason to limit reliance on financial statement to non-traded companies.

Another possible rationale for the limited scope of the relief may be that a public company will be incentivized by its need to attract U.S. investors and its U.S. securities law disclosure obligations to undertake the necessary cost and expense of valuing or appraising its individual assets.¹¹ Yet, because of the substantial regulatory burden of U.S. listing, many foreign companies list on non-U.S. exchanges and issue equity to relatively sophisticated investors under Section 144A of the U.S. securities laws. The fact that a company is “traded” does not mean it is necessarily targeting the U.S. capital markets to an extent that would cause it to take on this costly burden. It is our experience that asset specific valuations are almost never obtained for PFIC reasons in that context (even for a U.S. listing) nor is that likely to change regardless of the approach the regulations ultimately adopt to allowing reliance on financial statement values. Moreover, even if a company were willing to undertake this for purposes of achieving its initial listing, it would not undertake the exercise annually for purposes of 20F securities law filings. Any in terrorem effect of precluding reliance on financial statement values likely will merely continue to result in highly qualified disclosure about PFIC status on which small public taxpayers will nevertheless rely in practice not to report the company as a PFIC and to claim qualified dividend treatment for its dividends.

Thus, notwithstanding availability of market information about the fair value of the aggregate gross assets for a public company, one could argue that there may be stronger reasons to permit reliance on financial statements in the traded company context than for a non-traded company. We therefore recommend extending the use of the financial statement safe harbor to public companies. We do not think the fact that a trading price exists undermines the broad authority of Treasury and the IRS under section 1298(g) to extend this administrative relief to public companies given the need for small shareholders otherwise to value individual corporate assets for purposes of the PFIC asset test.

B. Requirements as to Applicable Financial Statements

To the extent reliance on financial statement values is permitted, the 2021 Proposed Preamble notes that the somewhat analogous section 1297(f)(4) rule (under the active insurance company exception) specifically requires the use of information from financial statements prepared under U.S. generally accepted accounting principles (“GAAP”) or international financial reporting standards (“IFRS”). Comments were requested on whether ordering rules similar to those of section 1297(f)(4) and Proposed Regulations section 1.1297-4(f)(1) should apply,¹² and whether other safeguards such as requiring that financial statements to be audited should be imposed.

The 2021 Proposed Preamble also notes that the Treasury Department and the IRS are aware that financial statements may not include values for some types of assets that are important to companies in certain industries, for example self-created intangibles. Proposed Regulations section 1.1297-1(d)(1)(v)(D) provides that if there is reliable information about the value of an asset that differs from its financial statement valuation, that information must be used to determine the value of the asset (the “More Reliable Information Standard”). Specifically, the provision states that “[i]f the tested foreign corporation or one or more shareholders has actual knowledge or reason to know based on readily accessible information that the financial accounting statements do not reflect a reasonable estimate of an asset's value and the information provides a more reasonable estimate of the asset's value, then the information must be used to determine the value of the assets to which it relates.”

Whether valuation information is more reliable than financial statement valuation is based on the facts and circumstances, including the experience and knowledge of the source of the information, whether the information is recent and whether there have been intervening developments that would affect the accuracy of the information, and whether the information specifically addresses the value of the asset in question. The 2021 Proposed Preamble requests comments on whether diverging from financial statements in other circumstances may be warranted, for example, when a tested foreign corporation or look-through entity owns property that is subject to a lease or license that is disregarded under the rules for intercompany obligations between a tested foreign corporation and a look-through entity, and similar fact patterns.

Limiting the rule to financial statements audited by a reputable third-party auditor would limit the risk of abuse by taxpayers. In most cases, in practice, companies will have audited financial statements because of securities laws, demands of minority shareholders or finance providers, regulatory requirements and similar considerations. There may be some small non-public companies with U.S. shareholders that lack audited financial statements but these would invariably arise in a closely-held “family and friends” context (as a third party who does not control the company is otherwise unlikely to invest in the absence of audited financials). There is arguably greater risk in that context that unaudited financials may be an inaccurate proxy for “fair” value. On the other hand, there is also more risk in that context of result-oriented or inaccurate outside appraisals or internal management account valuations being used. While we support requiring that financials used for this purpose be audited financials, on balance, much of the practical protection against abuse also arises from the use of the same values and financial information for purposes other than PFIC analysis (i.e., to provide financial information to stakeholders), where a company's commercial incentives generally favor maximizing asset values generally rather than over-valuing (active) operating assets and undervaluing (potentially passive) other assets. As an alternative to a requirement that all such financials be audited, the government could consider allowing unaudited financials to be used only if these financials statement are regularly and periodically shared with and used by stakeholders other than controlling U.S. shareholders or their family members (e.g., unrelated minority investors or financing parties) for other material commercial purposes and not prepared with a principal purpose of the PFIC analysis.

Although we also would not object to adopting standards like those of section 1297(f)(4) if Treasury and the IRS feel the existence of those standards in the PFIC insurance company exception requires this, we are skeptical whether it really makes sense as a policy matter to develop detailed specific requirements or elaborate objective rules of priority as between different financial statement regimes for more general purposes. Section 1297(f) is somewhat distinguishable in that the relevant test turns on “insurance liabilities”¹³ as a percentage of total assets (unlike the PFIC asset test which is a gross asset test and ignores liabilities). While we are aware of no legislative history explaining specifically the requirements of section 1297(f)(4), there may be insurance-related accounting complexities (such as accounting for contingent insurance loss reserves) which makes the choice of one financial accounting system or another more dispositive in the specific context of section 1297(f). As a more general matter, any financial statement, whether GAAP, IFRS or regulatory accounting-based, is an imperfect proxy for “true” fair market value and, while outside our expertise, we would be surprised if any one regime systematically reaches more or less PFIC-favorable or more or less accurate asset valuation results than any other. The resulting complexity seems unlikely to further any compelling policy purpose.

It seems more important in this context to prevent “cherry picking” of particular asset values among different financial statements if these exist.¹⁴ The 2021 Proposed Regulations already impose the More Reliable Information Standard, which is mandatory not discretionary. If one company has multiple financial statements under different regimes, the burden will be on the taxpayer to show why the use of one rather than the other value is more reliable. However, it would be reasonable to impose an additional consistency requirement of some kind. That is, subject to meeting the More Reliable Information Standard, the tested foreign corporation (or applicable U.S. shareholders) should be required to adopt a reasonable and consistent approach in choosing which financial statement regime to follow for PFIC purposes taking into account the characteristics of different entities relevant to the PFIC analysis and using that set of financial statements consistently for that tested company (e.g., use U.S. GAAP for assets of all U.S. companies and IFRS for all non-U.S. companies that have both GAAP and IFRS statements).

For reasons of practicality similar to those discussed above, we do not suggest developing elaborate rules to address situations such as an asset leased to a look-through subsidiary. Anomalies resulting from differences between the PFIC regime and an applicable financial accounting regime could arise. For example, a look-through subsidiary that owns an asset leased from the tested foreign corporation may not be consolidated for financial accounting purposes even if the lease is disregarded and the entities are, in a conceptual sense, “consolidated” for PFIC classification purposes. In that case, in theory the financial accounting value of the asset might reflect only the residual net value of the asset, and, depending on the financial accounting regime applicable to the look-through subsidiary, the net “in the market” value of the leasehold interest may or may not be reflected. Theoretically, it is possible that this could facilitate the use of structured transactions to take advantage of such discrepancies, but in practice this seems unlikely given all the commercial constraints on entering into transactions with entities that are not closely enough related to be financially consolidated. We are skeptical it will be possible to identify in advance all such possible issues and craft detailed rules. Such anomalies are seldom likely to be determinative to PFIC classification and some theoretical imprecision is inherent in the practical decision to allow the use of financial statement values. Final regulations should, however, specifically note that such potential inconsistencies resulting from inter-company transactions between the tested foreign corporation and its look-through subsidiaries should be taken into account in a reasonable manner when using financial statement values and that appropriate adjustments may be called for when the impact is reasonably likely to be material.

C. Clarification of More Reliable Information Standard

Although we support the adoption of a More Reliable Information Standard in general, we believe the current articulation needs to be revised and clarified if reliance on financial statements is to provide any practical benefit. The 2021 Proposed Preamble explains the requirement as follows:

“The Treasury Department and the IRS are aware that financial statements do not include values for some types of assets that are important to companies in certain industries, for example self-created intangibles. Accordingly, proposed §1.1297-1(d)(1)(v)(D) provides that if a shareholder has reliable information about the value of an asset that differs from its financial statement valuation, that information must be used to determine the value of the asset. Whether valuation information is more reliable than financial statement valuation is based on the facts and circumstances, including the experience and knowledge of the source of the information, whether the information is recent and whether there have been intervening developments that would affect the accuracy of the information, and whether the information specifically addresses the value of the asset in question.” (Emphasis added.)

Allowing taxpayers to value self-created intangibles like goodwill (which are not reflected on financial statements and may represent a significant “active” asset and, in the case of modern businesses, often the most significant such asset related to the operating business) is essential for certain businesses which otherwise might be treated (inappropriately) as PFICs. Generally, financial statements simply do not reflect self-created intangibles. This may represent a conservative and prudent approach as a financial accounting policy matter but necessarily is economically and commercially unrealistic and would reach inappropriate results for PFIC purposes. We therefore welcome the recognition of this practical reality and the decision not to simply bind taxpayers to using financial statements in that respect. At least in the case of traded companies, the existence of a public share trading price from which a gross asset value for the asset may be derived typically provides a good basis for approximation of the value of self-created intangibles as most of the disparity between that value and financial accounting book values in modern businesses is typically attributable to intangible assets (including self-created intangibles like goodwill and uneconomic book depreciation of operating assets).

We also understand why, in principle, if a taxpayer actually has an alternative valuation and knows this is a better measure of value of an asset than the financial statement item the taxpayer should be required to use that more reliable value. Other than in fairly extreme and obvious cases, however, this requirement could raise very challenging factual questions and cause considerable uncertainty. For example, assume that the relevant asset is a building carried on the books at its historic (partly depreciated) value and the property is valued at double that book value for local property tax purposes. Or conversely, assume the property tax assessed value of a recently acquired building is lower than book value. Property tax values in some jurisdictions may bear a questionable relationship to true fair market value but may call into question whether a taxpayer may rely on the financial statement value. The market for some kinds of assets like real estate, while not equivalent to the market for financial securities, may be somewhat more liquid than is typical for operating assets like machinery and equipment. For example, a taxpayer may know that real estate in the locale as a general matter sells for a capitalization rate within a certain range of annual cash flows, which would suggest a different value than that on the financial statements. Does this prevent reliance on the financial statement value? There would be an argument that this does not “specifically address” the value of the asset in question (which the 2021 Proposed Preamble suggests is a relevant factor). On the other hand, that is just a factor mentioned in the 2021 Proposed Preamble so the answer is unclear.

A second source of uncertainty is the extent to which the More Reliable Information Standard imposes a duty of inquiry. The 2021 Proposed Preamble summary suggests that the information must already be available to the taxpayer. However, the actual regulatory language provides that “if the tested foreign corporation or one or more shareholders has actual knowledge or reason to know based on readily accessible information that the financial accounting statements do not reflect a reasonable estimate of an asset's value and the information provides a more reasonable estimate of the asset's value, then the information must be used to determine the value of the assets to which it relates.” (Emphasis added.)

The reason to know standard suggests that taxpayers have an affirmative obligation to seek out information that would imply a value different from the financial statement value. When combined with the factual uncertainties above, there is a risk this could substantially undercut the advantages of a rule allowing financial statement values to be used as a proxy for fair value.

There is a trade-off under the More Reliable Information Standard between (1) preventing taxpayers using values they know to be unreliable relative to readily available alternatives and (2) generating so much uncertainty that this simply defeats the purpose of allowing reliance on financial statement values for reasons of administrative simplicity.

We recognize that trying to balance these competing considerations is difficult. For the reasons stated above, we support the More Reliable Information Standard and can understand the rationale for adopting a “reason to know” standard. However, we believe it is then absolutely critical to make the applicable standards as clear as possible. In particular, we think the final regulations should make clear (including with examples) that this standard is based on the knowledge or “reason to know” of the specific person required to undertake the analysis in their own particular circumstances. For example, to be subject to the More Reliable Information Standard, a very small public shareholder should itself have actual knowledge or reason to know of a better value, actual realistic access to the information and a realistic ability to determine that value themselves. The fact that the tested foreign corporation may have additional valuation-relevant information should not be imputed to the shareholder nor should the standards be applied in a manner that could cause the standard not to be met on the basis that the shareholder failed to undertake some form of diligence exercise vis-à-vis the tested foreign corporation unless there are facts suggesting that such shareholder reasonably could have expected the company to respond favorably.

As an example, consider a tested foreign corporation a substantial stake in which is owned by a private equity fund which has negotiated for PFIC-related information rights but that also has smaller U.S. shareholders who invested later and were not able to negotiate such rights. Absent actual knowledge or a personal reason to know, such a small shareholder should not be expected to surmise that the corporation may be providing valuation information to the large shareholder, imputed knowledge of what those values may be, or imputed some duty to request the information. In practice, given liability concerns (which in our experience lead foreign corporations to resist providing such information unless a shareholder has significant negotiating leverage), there is no reason to think, even if the small shareholder asked, that the tested foreign corporation would agree to provide such information as it already has. Nor would the larger shareholder in the example likely be inclined to discuss its private tax return positions with others.

We recognize that, in theory, this could result in anomalous situations in which some shareholders might know the tested foreign corporation is a PFIC while other small shareholders do not treat it as a PFIC. We do not believe that this is a reason to preclude reliance on financial statements or to adopt an unduly broad More Reliable Information Standard. The problem reflects an inherent flaw in the statutory scheme, which can be complied with fully only upon production of information from a foreign corporation that most foreign corporations do not have, would be unwilling to generate and that neither most U.S. shareholders nor the U.S. tax laws have the power to compel.

4. Treatment of Goodwill

Notice 88-22 provides that, for purposes of the PFIC asset test, goodwill or going concern value must be identified with a specific income-producing activity of the corporation and characterized as a passive or non-passive asset based on the income derived from the activity. It was not entirely clear how goodwill that in theory generated some passive income would be treated under Notice 88-22 (i.e., whether it would be passive or bifurcated as a “dual asset” between active and passive on some basis). Commentators have also criticized the Notice 88-22 approach on the basis that goodwill is per se active.

The 2021 Proposed Preamble notes that the government believes that the approach provided by Notice 88-22 for determining the character of goodwill for purposes of the asset test is a reasonable approach, although other approaches may be more economically accurate for a particular tested foreign corporation. As stated in the 2021 Proposed Preamble, the government agrees that goodwill is attributable to “business” activities but does not agree that goodwill should always be treated entirely as a non-passive asset because the PFIC rules may treat certain business income as passive and it is possible that goodwill would be associated with that income. Further, because companies in different lines of business may be valued as an economic matter under different valuation models, some of which may give more weight to income and others to assets or to other aspects of a business like customer relationships, there is no single basis for allocating goodwill that is likely to be best suited to every company as an economic matter. The 2021 Proposed Preamble requests comments on alternative approaches to addressing the treatment of goodwill for purposes of the asset test.

One could argue that, to the extent the PFIC rules may treat goodwill from an “active” business (in a commercial sense) as generating passive income, it is that characterization that is the real problem and elaborate rules to bifurcate goodwill for purposes of the asset test merely compound the problem. As a general matter, goodwill represents a non-separately identifiable intangible asset that is also non-transferable other than as part of an entire going business. While there is a conceptual distinction between economic and accounting goodwill, fundamentally, goodwill reflects the value of non-specific intangible advantages a business enjoys over its competitors. Liquid financial assets are traded in relatively efficient markets and, by definition, no goodwill would attach to the financial assets and returns on those assets themselves. An asset management services business might attract goodwill (putting aside whether it is plausible in efficient financial markets to routinely outperform other asset managers), but this is related to the services performed and professionals engaged in those services not the financial assets themselves. The ability to earn above-normal returns on a consistent basis is by its nature linked to active business conduct not passive investment or trading. A business may have certain financial assets on its books (e.g., as working capital) but the market almost certainly is not ascribing a value in excess of market value based on its ability to manage its working capital efficiently. Indeed, this is demonstrated by the fact that most purchase contracts in corporate acquisitions merely adjust the purchase price upward or downwards (dollar for dollar) to reflect changes in working capital between signing and closing.

On the other hand, there are a small number of situations in which, for better or worse, based on positions the government has adopted, a clearly active business (in a colloquial sense) that may generate goodwill nevertheless may be treated as a PFIC under the regulations, primarily in the context of financing businesses, because it fails to qualify for a specific “active business” exception like the active banking exception. Where such an active but PFIC business is conducted, a company may reflect goodwill that could in some sense be considered a “passive” asset. We are also aware of situations in which, at times, public markets appear to ascribe a value to a combination of non-operating assets that exceeds the separate value of the assets themselves (for example, recent initial public offerings of special purpose acquisition companies (or “SPACs”)). It is questionable whether this differential should even be characterized as goodwill or is a temporary market anomaly (and in any event, subject to the start-up exception, such companies are likely to qualify as PFICs regardless).

Where a PFIC “active” business is combined with other businesses in a group, it may be necessary to bifurcate the goodwill if one believes that the fact a business fails to qualify for a specific exception to foreign personal holding company income (“FPHCI”) necessarily requires treating the associated goodwill as passive. An example where this would be the case would be a retail group that includes a subsidiary that operates a retail credit financing unit to facilitate customer sales that does not meet the active banking exception because it is not a deposit-taking licensed institution. Another may be a group that has an insurance subsidiary that does not meet the requirements of section 1297(f) to be a qualifying insurance corporation.

We think such circumstances are unusual, however, and it will be even more unusual for a material amount of the business's goodwill to be attributable to such a non-core financing activity. Accordingly, as a practical matter, it is questionable whether the policy concern over attribution of goodwill justifies the resulting complexity, in particular, the difficulty in practice with tracing specific income to a non-separately identifiable asset like goodwill, which may be even more challenging for self-created goodwill.

Moreover, the that reason financing or insurance businesses like that above are even considered PFICs results from the (deliberately restrictive) exceptions to FPHCI treatment that are applied in the case of financing businesses. For reasons of administrability, Congress or Treasury, as applicable, have concluded in such circumstances that it is simply too difficult to administer and enforce the income test (as applied to financing income that otherwise would be FPHCI) using generic factual tests based upon the “activeness” of the underlying financing business. Instead, detailed and deliberately narrow exceptions are provided, which turn on bright-line “proxies” for active business conduct, such as whether a financing business closely resembles traditional licensed, deposit-taking banks, even though this may mean that some very active businesses (that in abstract probably should not be PFICs) could be treated as PFICs. That approach (and the possibility that certain inherently “active” businesses may as a result questionably be classified as PFICs) may be understandable given those administrability and enforcement concerns. It does not follow, however, that goodwill from such businesses (which, unlike the financing income itself, necessarily can only arise from active business conduct) must be treated as passive to prevent inappropriate results. Put differently, it may be difficult for the IRS to determine whether interest income is derived from an active financing business or passive investment. It is not difficult to conclude that any goodwill is necessarily attributable to active business conduct. Passive investment activity would not generate goodwill. We therefore suggest Treasury and the IRS reconsider whether it really is inappropriate to treat goodwill as an active asset merely because the business that generates it fails to qualify for a safe harbor like the active banking, insurance, leasing or licensing exception and financing income of the tested foreign corporation therefore may be deemed to generate “passive” income.

If, however, as appears to be the case, Treasury and the IRS feel strongly that on principle such goodwill should be identified and treated as passive, we would suggest the rules make clearer that this is an exception to a more general presumptive principle that goodwill is generally active when attributable to a company that conducts an active operating business. This presumption could be limited in the first instance to a tested foreign corporation that conducts (applying the activity attribution rules already included in the regulations) a substantial operating business (for example, a trade or business within the meaning of Treasury Regulations section 1.367(a)-2(d)(2) that would qualify as actively conducted under Treasury Regulations section 1.367(a)-2(d)(2)). Further, the presumption would not apply to such a tested foreign corporation if it also actively conducts directly or through subsidiaries (applying the activity attribution rules already included in the regulations) as a trade or business (under such principles)¹⁵ any of a set of specified business activities, which could be expanded by future guidance. These could include (whether conducted physically or virtually online) any material lending, financing, commodity-related, real estate-related or insurance business activities or leasing or licensing activities that generate rental or royalty income. In that event, the presumption would not apply and an allocation between passive and active goodwill would be required (under the principles currently in the 2021 Proposed Regulations). It should be made clear that the tests above apply merely to determine whether a tested foreign corporation can rely on the active goodwill presumption or must characterize and allocate goodwill under the more factually intensive approach that is currently proposed. It creates no general inference that the activity is active or passive, which is determined under the other classification rules of the regulations.

5. Treatment of Working Capital

The 2021 Proposed Regulations provide a limited exception from passive asset treatment for certain cash that is held in a noninterest-bearing account, provided the cash held is not greater than the amount needed to cover 90 days of operating expenses of the entity holding the cash (the “Proposed Working Capital Exception”). Specifically, under Proposed Regulations section 1.1297-1(d)(2), an amount of cash held in a non-interest bearing account that is held for the present needs of an active trade or business and is no greater than the amount reasonably expected to cover 90 days of operating expenses incurred in the ordinary course of the trade or business of the tested foreign corporation (for example, accounts payable for ordinary operating expenses or employee compensation) is not treated as a passive asset.

The 2021 Proposed Preamble explains the prohibition on interest bearing working capital as follows: “[B]ecause the statutory PFIC rules (and FPHCI rules) generally treat an asset held to produce interest as passive, it may not be appropriate to treat an interest-bearing instrument held by an operating company as working capital other than as an asset that produces passive income.” The 2021 Proposed Preamble requests comments on the Proposed Working Capital Exception, including the scope of statutory authority to treat interest-bearing accounts or instruments held as working capital as active assets and the ways in which the exception might be broadened while maintaining appropriate safeguards to avoid uncertainty as to how to determine the amounts and types of instruments properly treated as held for the present needs of a business and to ensure that a business's investments and capital held for future needs continue to be characterized as passive assets.

The recognition that active businesses need working capital and that it is not appropriate as a policy matter to treat all cash as passive regardless of the purpose for which it is held is welcome.

In principle, working capital held for use in an active trade or business should not be treated as a passive asset so long as the amount is reasonable. Positive net working capital is essential to ensure that a business is able to continue its operations by providing sufficient funds to satisfy both anticipated operational expenses and maturing short-term debt. Active businesses need to maintain a prudent level of liquid assets to deal with liquidity problems that can arise through the "normal" operation of the business cycle or unexpectedly. Active businesses also need to accumulate liquid assets for expansion (e.g., building a new plant, beginning or developing a new business) or maintenance. A new or existing business can raise cash to build a new plant, a process that can take several years. Businesses have to raise the cash when it is available, such as when the interest rate environment is favorable, rather than immediately before it is used.

In prior 2010 and 2001 Reports,¹⁶ we suggested using the standard of "reasonable needs" of the business as developed in the accumulated earning tax context as a possible approach. The use of this reasonable needs standard has the benefit that it is a standard that is already developed; it is a test that is based on the particular business being evaluated; and it was designed to prevent companies from accumulating liquid assets rather than paying dividends, which is similar in some respects to the goals of the PFIC regime. We continue to think this is a viable alternative. We recognize that Treasury and the IRS may have concluded, however, that the reasonable needs of the business test as developed for the accumulated earnings tax is too subjective to be administrable and we acknowledge that the standard has been difficult to enforce effectively in practice. Accordingly, we will not reiterate here those earlier recommendations.

However, the working capital exception as currently proposed is overly narrow in certain respects.

First, literally the 90 day of expenses test is not a bright line safe harbor for taxpayers but a “cap”. We agree with the 2021 Proposed Preamble that there are substantial benefits to a bright line test (like 90 days of expenses) in providing certainty for many types of businesses. However, the current formulation does not provide such certainty to businesses. As we read the requirement, the taxpayer also must be able to show that the working capital is not in excess of the present needs of an active trade or business, which is a factual inquiry and seems somewhat similar to the accumulated earnings tax inquiry above, with the related subjectivity and uncertainty. The proposed rule therefore remains uncertain in application but also may be unduly narrow in some cases.

The use of a 90-day limit will significantly limit the utility of the exception for certain kinds of businesses with significant capital development costs and a long “tail” period before investments may generate revenue (for example, the pharmaceutical industry). We understand the difficulty of coming up with a definition of working capital that avoids abuse, is broad enough to capture many different types of businesses and is also clear and administrable. However, we see little benefit to limiting relief in cases like the one above if the standard in any event does not provide a bright-line test for taxpayers.

The government could consider adopting the 90 day limit as a safe harbor and also permitting a taxpayer to demonstrate by a preponderance of the evidence that (1) the amount of working capital is consistent with the reasonable needs of its business and (2) the amount treated as working capital is not disproportionate to the working capital maintained by most other substantially similar businesses in that industry or sector. It is unlikely that most such businesses have PFIC sensitivities (because, for example, some competitors may be U.S. companies and even foreign competitors may have few, if any, U.S. shareholders) and this will provide a reasonable benchmark for the amount of working capital a business needs regardless of tax considerations. Alternatively, the government could adopt such a standard with no 90-day safe harbor. However, we do not believe the 90 days of expenses test should function as a “cap” on what can constitute “active” working capital.

Secondly, the prohibition on interest bearing capital will limit the utility of the exception for many if not most businesses. Given the obligation a company owes to prudently manage its finances, even if all of the ordinary course expenditures are expected to be incurred and payable within 90 days, we would expect it to be unusual for all of those funds to be maintained in non-interest bearing accounts given the availability of liquid short term interest bearing instruments. Possibly, in the current interest environment, the foregone return is small enough that this approach to financial management would be defensible, but, in future interest environments, the requirement may force a company to choose between PFIC status and a prudent approach to its treasury operations. The prohibition on interest bearing working capital is therefore a more fundamental problem. One can infer from the 2021 Proposed Preamble discussion (above) that the primary concern motivating this restriction is a technical rather than a policy concern. The government appears concerned that the statutory definition of “passive” income by reference to FPHCI under section 954(c) precludes rules treating interest-generating assets as active (unless a company meets specific statutory exceptions for active banking, etc.)

This concern is based on a very literal reading of the PFIC asset test defining as passive assets those assets “which produce passive income or are held for the production of passive income.” Even if interest is presumptively passive because it is FPHCI, working capital presents the somewhat anomalous circumstance in which an asset might be said to produce passive income even though clearly it is not “held to produce” passive income. The income earned is purposively incidental. We understand the statutory test literally is disjunctive, but we think such a literal interpretation mistakes the real intent of the language, which is only to ensure that assets held with investment intent (whether or not currently producing passive income) are passive. We do not think it is clear what Congress intended in this somewhat peculiar circumstance.¹⁷ Regardless, Treasury and the IRS have very broad regulatory authority under section 1298(g) to “prescribe such regulations as may be necessary or appropriate to carry out the purposes of this part.” (Emphasis added.)

As a policy matter, the stakes under Subpart F are very different from PFIC. All that would be at issue for Subpart F purposes is whether a U.S. 10% shareholder may be required to pick up as Subpart F income the (generally fairly small) return earned on working capital into current income (possibly with a foreign tax credit for local country tax). In the PFIC context, what is at stake is whether the company is a PFIC at all and whether the punitive PFIC regime applies or does not. Moreover, the effect of treating working capital interest as passive for PFIC purposes is magnified by the asset test, which has no corollary under Subpart F. There would be no practical impact in most cases to treating the interest on working capital as passive solely for PFIC income test purposes absent this knock-on effect. The amount of income is likely small and the kinds of active businesses for which this is an issue generally have considerably less interest income from working capital and any other passive sources than the 75% passive income threshold. It is extremely rare for operating businesses (of the type Congress likely would not have expected to be PFICs) to meet the PFIC income test.¹⁸ As applied to cash and working capital, the PFIC asset test (which has no Subpart F corollary) is what results in no-yielding or low-yielding assets held with no investment intent having a potentially disproportionate effect on PFIC status.

Given the significant policy differences above, the effect of the asset test (which is not relevant for Subpart F purposes) and the broad regulatory authority under section 1298(b) to regulate to achieve the purposes of the PFIC rules, we think there is ample regulatory authority not to treat cash as a per se passive asset and to extend the working capital exception to interest-bearing working capital. To avoid confusion, we think it is preferable to treat working capital (however ultimately defined) and the related income as active for PFIC classification purposes. However, if Treasury and the IRS feel that treating interest income itself as active solely for purposes of the income test is not technically supportable given the direct reference to FPHCI (without redefining FPHCI more generally), as a practical matter treating the working capital as active only for purposes of the asset test (on the basis that the assets are not held to produce passive income) would achieve the appropriate policy result in almost all cases for the reasons above.

We recommend that the rules define working capital (however otherwise defined) to be limited to liquid assets invested in high investment grade, interest-bearing, publicly offered debt securities, money market funds or interest bearing deposits with licensed deposit-taking banking institutions, provided that the portfolio designated as comprising the company's working capital for PFIC asset test purposes has an average weighted tenor of less than 90 to 180 days. The yield on a portfolio of securities that met this requirement would generally be low, and well below the weighted average cost of capital of most operating companies. As practical matter, with one caveat discussed below, most companies have no incentive to inflate their working capital for PFIC purposes as the negative spread between their cost of capital and the yield on high investment grade short term debt securities would make using this exception for “disguised” investment activity an expensive and unsustainable proposition. Theoretically, a business with a very low weighted cost of capital, like a money market fund or high-grade short-term bond fund, could avoid PFIC status if that were the only limitation on qualification as “working capital.” However, if the current restriction that working capital cover only 90 days of expenses were retained, such a fund would fail to avoid PFIC status in practice. Even if a broader or more qualitative definition of working capital is adopted in final regulations, the government could impose additional restrictions, for example, capping the percentage of assets that could be treated as working capital and/or creating a disqualification/adverse presumption if the return on the designated working capital is not substantially lower than (e.g., no more than 25% of) the weighted average cost of capital the company uses during the year for evaluating its investment opportunities.

6. Look-through Subsidiary Dividends and Gains

Section 1297(c) provides that if a tested foreign corporation owns a look-through subsidiary “such foreign corporation shall be treated for PFIC classification purposes as if it — (1) held its proportionate share of the assets of such other corporation, and (2) received directly its proportionate share of the income of such other corporation.”¹⁹ Read literally, this could suggest that an “aggregate” approach applies; that is, for purposes of PFIC classification and the application of the PFIC income and asset tests, dividend income from such a look-through subsidiary are simply ignored and a sale of shares is treated as a sale of the underlying assets. Nevertheless, presumably for policy reasons, Treasury and the IRS have adopted a different and somewhat complex approach, as discussed below.

A. Overview of the Rules for Look-through Subsidiary Dividends and Share Sales

Although the look-through subsidiary rule reasonably could be read to treat a sale of the shares of a look-through subsidiary as a sale of the underlying look-through subsidiary assets, the final regulations treat gain on disposition of a look-through subsidiary shares as a share disposition. However, to prevent double counting, the actual share sale gain is reduced (but not below zero) by unremitted earnings (i.e., earnings taken into account by the tested foreign corporation from the look-through subsidiary for PFIC classification purposes but not yet distributed as a dividend) (the “Residual Gain Approach”). The residual gain is characterized as passive income or non-passive income based on the relative amounts of passive assets and non-passive assets of the disposed look-through subsidiary on the date of the disposition. The relative amounts of passive assets and non-passive assets held by the look-through subsidiary are measured under the same method (value or adjusted bases) used to measure the assets of the tested foreign corporation for purposes of section 1297(a)(2).

Adopting a somewhat analogous approach, the 2019 Proposed Regulations proposed to eliminate dividends from look-through subsidiaries but only to the extent attributable to income of the look-through subsidiary during the period it was a look-through subsidiary that would therefore have been taken into account as tested foreign corporation income in applying the PFIC income test (“Post-LTS Status Earnings”) but not dividends attributable to earnings prior to the time at which the subsidiary became a look-through subsidiary (“Pre-LTS Status Earnings”).²⁰

Importantly, earnings for the period of look-through subsidiary status are treated as Post-LTS Status Earnings regardless of whether there are ultimate U.S. shareholders during the relevant period (and that subsidiary's owner corporation was, practically speaking, a “tested” foreign corporation at the relevant time). For example, if a foreign corporation owns 20% of a subsidiary but acquires an additional 6% beginning in 2014, all at a time when it has no U.S. shareholders, and then in 2021 a U.S. shareholder acquires 1% of the tested foreign corporation, the Post-LTS Status Earnings apparently includes E&P for the period beginning in 2014. Moreover, if a second U.S. shareholder acquires 4% of the tested foreign corporation in 2022, it would also determine the treatment of look-through subsidiary dividends to the tested foreign corporation by reference to the E&P beginning in 2014.

Indeed, that is the only workable way to apply the approach above. The PFIC regime is designed to be an anti-abuse regime. It is not an actual pass-through regime.²¹ The tested foreign corporation never actually “includes” the underlying E&P of the look-through subsidiary prior to a dividend (nor does any U.S. shareholder of the foreign corporation). The tested foreign corporation may merely be deemed to do so, but solely for PFIC classification purposes under the PFIC income test. Such “deemed” inclusion, moreover, would only have actually been relevant for U.S. shareholders who were shareholders at the time. It would make no sense for the “good” E&P to be measured only from the date there is some U.S. ownership (regardless of the size of that ownership stake) as the resulting exclusion of dividends from the perspective of subsequent U.S. owners would be almost random. Conversely, a system that attempted to police look-through subsidiary dividend “elimination” based on each indirect U.S. shareholder's indirect percentage interest in the look-through subsidiary's earnings would be impossibly complex and not administrable.

B. Prior Comments and Government Response

Our prior 2019 PFIC Report recommended that final regulations provide for the elimination of all dividends from look-through subsidiaries (not merely those out of Post-LTS Status Earnings) and made a number of alternative suggestions intended to reduce or eliminate the likelihood that a dividend from a look-through subsidiary would be treated as a dividend from non-accounted-for earnings. Our primary rationale was practical administrability. The test originally proposed effectively would require that a U.S. taxpayer be able to determine a look-through subsidiary's current and historic E&P from its inception in order to demonstrate to what extent a dividend was a dividend in the first instance and then whether it was from E&P that was Post-LTS Status Earnings.

The 2021 Final Regulations did not adopt our recommendation to eliminate all look-through subsidiary dividends whether or not taken into account under the look-through subsidiary regime. The stated concern under the 2021 Proposed Preamble is based on the interaction of such an approach with the regime for gains on sale of look-through subsidiary shares. Eliminating a dividend attributable to Pre-LTS Status Earnings (i.e., of periods before 25% ownership of the subsidiary was acquired) could cause a corresponding amount of the gain otherwise taken into account to be eliminated for PFIC classification purposes. An example illustrates what we understand to be the government's conceptual concern.

Example. An operating company (Foreign TestCo) acquires 100% of another foreign operating company (Foreign TargetCo) at the start of Year 1. Assume that Foreign TestCo is a pharmaceutical company that requires substantial working capital and would be close to meeting the PFIC asset test on a stand-alone basis as a result. Foreign TargetCo has $100 of legacy E&P (applying Subchapter C principles). Economically, the legacy E&P attribute has no impact on the price paid. However, presumably, the purchase price will reflect the assets that are the proceeds of that undistributed legacy E&P. Thus, the price and therefore tax basis of Foreign TestCo in Foreign TargetCo will be $100 higher than would have been the case if that legacy E&P had been distributed before the acquisition. In Year 2, during which no positive E&P is earned, Foreign TargetCo distributes $100. If that $100 is treated as a dividend that is eliminated for purposes of the PFIC income test, the value of Foreign TargetCo and therefore the amount of gain on sale (which is taken into account under the PFIC income test) on a subsequent sale of Foreign TargetCo would be reduced by $100. If, by contrast, Foreign TestCo had made a section 338(g) election, eliminating the legacy E&P of Foreign TargetCo, the distribution would not be income but a return of capital that would reduce both basis and value by $100 and therefore would not decrease future gain.

Importantly, the “disappearing gain” problem above only arises because of the government's decision to adopt the Residual Gain Approach. If the sale of look-through subsidiary shares was viewed as a sale of the underlying assets, the problem would not arise (because distributions would not affect underlying asset basis) and no basis adjustment complexity would be required, although this would be replaced by some complexity of attributing sale value to individual assets.²²

To provide some relief but also address this perceived “disappearing gain” issue, Proposed Regulations sections 1.1297-2(c)(2) and (f) provide rules that would — for purposes of determining a tested foreign corporation's PFIC status under the income test — eliminate a dividend received from a look-through subsidiary that is attributable to Pre-LTS Status Earnings. As a price for this relief, however, the rules require corresponding adjustments to the basis in the look-through subsidiary's stock for purposes of determining gain upon a future disposition of such stock in applying the income test. The 2021 Proposed Preamble invites comments addressing these issues — in particular, on the treatment of pre-acquisition E&P.

C. Comments and Recommendations

We acknowledge the theoretical concern raised above over disappearing gain. We think, however, that the response (and the Residual Gain Approach and approach to look-through dividend elimination generally) elevates theoretical purity over practicality. The PFIC regime is ultimately designed to be an anti-abuse regime. PFIC classification inherently involves “cliff” effects (exacerbated by the “once a PFIC” rule). It is not theoretically pristine, for example, from an abstract tax deferral perspective to penalize a U.S. taxpayer that invests in a foreign corporation 51% of whose assets are passive while exempting a taxpayer investing in one with only 49% passive assets. PFIC classification and the tests for PFIC status are inherently blunt instruments. Designing complex rules to achieve theoretical perfection as to the timing and amount of every item of income to which the PFIC income and asset tests may be applied, particularly when these may be impossible to implement in practice, makes little sense unless there is a practical risk of aggressive planning and abuse.

As stated in our prior report, Pre-LTS Status Earnings are not, in an economic sense, income that has been deferred in the tested foreign corporation (or indeed, by any U.S. shareholder necessarily). As the PFIC rules are intended to prevent a U.S. shareholder from deferring current U.S. tax on passive assets held in an offshore corporation, it makes little policy sense to treat as potential passive income (or income of any kind) for PFIC characterization purposes an item which does not reflect economic accretion during the tested foreign corporation's ownership period for a look-through subsidiary.

Our fundamental point, however, was not based on abstract economic principles but concerns of practicality. Most foreign corporations don't track U.S. principles-based E&P. They may keep track of the ownership history of their subsidiaries but may or may not share that with their U.S. shareholders. The PFIC regime in practice is applicable to U.S. shareholders not the foreign corporation itself. A U.S. shareholder may have no practical way to determine at what point in the past a current look-through subsidiary became a look-through subsidiary, to track the tested foreign corporation's fluctuating ownership percentages in that subsidiary which is now a look-through subsidiary, to reconstruct the lower-tier subsidiary's income during the prior applicable periods or figure out the tested foreign corporation's tax basis in the look-through subsidiary. That income may not even have been reflected on the tested foreign corporations' past consolidated financial statements and stand-alone financial statements for the subsidiary itself may be unobtainable. A PFIC classification rule which is only “workable” for those U.S. shareholders who are able to reconstruct various legacy tax attributes of a look-through subsidiary like those above simply is not an administrable rule in this context.

The Treasury and IRS approach under the 2021 Final Regulations means, for example, that as a condition to eliminating a look-through subsidiary dividend and in order to determine there is no residual gain under the Residual Gain Approach resulting from a distribution subject to section 301(c)(3), a U.S. shareholder must reconstruct the ownership history and the E&P for any look-through subsidiaries of the tested foreign corporation (potentially for many years prior to becoming a U.S. owner) to demonstrate that the dividend is out of E&P that would have been “taken into account” by the tested foreign corporation (albeit notionally) under the look-through subsidiary rules. In many if not most cases, practically that will mean (1) if the burden of proof is on the taxpayer, that there is in fact no elimination of look-through subsidiary dividends or (2) otherwise, taxpayers will in practice just eliminate all such dividends in applying the income test (as we originally proposed) and the IRS on audit will need to reconstruct this E&P. Neither outcome is desirable as a policy matter.

While the government's point that eliminating all dividends in theory allows gain to “disappear” is in some sense correct from a theoretical standpoint, it fails to acknowledge the practical concerns above. More generally, we question whether this concern really makes policy sense when the relevance of the disappearing gain is limited to characterizing the tested foreign corporation as a PFIC or not. For reasons explained above,²³ this is not a situation in which a dollar of gain that is not taken into account is a dollar of gain that therefore escapes U.S. taxation. Moreover, it is questionable what abuse will be made possible under a simpler rule. There may be less residual gain but what impact that may have on the PFIC classification of the tested foreign corporation that owns the look-through subsidiary will depend on look-through subsidiary's ratio of active to passive assets at the time of a future hypothetical sale. In most cases, this is likely to be indeterminate at the time the look-through subsidiary dividend is received. The only abuse case will be the rare circumstance in which the tested foreign corporation intends a sale of the look-through subsidiary (which it knows has Pre-LTS Status Earnings), figures out that the gain (based on the current active-to-passive asset ratio) will be net passive in an amount sufficient to tip the tested foreign corporation into PFIC status and deliberately makes a distribution to cause the gain to disappear to benefit some non-controlling U.S. shareholder.

The proposed solution in the 2021 Proposed Regulations — to allow elimination of the income even if attributable to Pre-LTS Status Earnings at the price of basis reduction — unfortunately does not address the real (practical) problem. Much of the time, a U.S. shareholder may be unable to determine when the look-through subsidiary became a look-through subsidiary. Even if it can determine this, the U.S. shareholder likely cannot reconstruct earnings for the relevant periods. Even if the U.S. shareholder is able to determine the tested foreign corporation's U.S. principles tax basis in the look-through subsidiary (which is doubtful) it must know the entire E&P history to identify which distributions are dividends rather than returns of capital that reduce basis under general principles. If it cannot, presumably the U.S. shareholder may eliminate the distributions under the proposed relief for PFIC income test purposes but must notionally reduce the tested foreign corporation's basis in the subsidiary until there is no more basis, at which point a subsequent distribution would be “eliminated” but also result in gain under section 301(c)(3) that must be analyzed under the Residual Gain Approach.

The proposed solution, unfortunately, therefore partially fixes something that was not the real problem. The proposed solution means (if one assumes that a U.S. shareholder knows the look-through subsidiary stock basis) that the U.S. shareholder is allowed to convert what would have been a dividend today for purposes of the PFIC income test into potential gain at some future time under the income test. The real problem, however, (which neither the original proposal nor the proposed relief addresses) is that in many (likely most) cases a small U.S. shareholder does not know and cannot reasonably determine any or some of the earnings, ownership history, tax basis or other legacy attributes of the look-through subsidiary. Any rule which would require them to know these attributes as a condition to eliminating dividends is a rule which effectively precludes eliminating dividends, contrary to what we believe was intended by section 1297(c). The proposed basis reduction rule may provide some consolation as it allows the U.S. shareholder to defer this problem until they “run out of” look-through subsidiary tax basis but only if they know what that tax basis is.

We do not support constructing a complex basis adjustment regime to address a very narrow problem. It is questionable whether the disappearing gain problem presents any practical risk of abuse even if the result could be viewed theoretically as imperfect. If the government is unwilling to revisit the Residual Gain Regime, and is also determined to prevent disappearing gain as a matter of principle, our recommendation would be to consider converting the basis reduction approach from a general requirement into an anti-abuse provision. For example, the general rule could be that look-though subsidiary dividends are simply eliminated. However, to the extent there is a disposition of look-through subsidiary shares, basis must be reduced by any distributions made by the look-through subsidiary during the two taxable years preceding the sale unless the taxpayer can demonstrate that those distributions were made from Post-LTS Status Earnings for PFIC classification purposes or were not in contemplation of the sale.

7. Active Banking Exception

Our 2019 PFIC Report offered recommendations regarding the extent to which the section 954(c) definition of FPHCI, which is incorporated by cross-reference in defining “passive” income for PFIC purposes by section 1297(b)(1), should incorporate the various financing business exceptions under section 954(h). We expressed support for an exception for active financing businesses not limited to traditional banks. Treasury and the IRS have concluded, for policy and technical reasons based on the legislative history, that (1) the “active banking” exception should not apply to active financing businesses conducted by financial institutions that are not “banks” in a traditional sense even if they conduct similar active direct lending businesses to banks and (2) the reference in section 1297(b)(1) to section 954(c) does not, by its terms, incorporate the section 954(h) exception to section 954(c).

The 2021 Proposed Regulations, however, in Proposed Regulations section 1.1297-1(c)(2), propose to adopt the Modified 954(h) Regime for traditional banks that is largely based on section 954(h) with certain modifications. The 2021 Proposed Preamble suggests that section 954(h) was the basis for the regime because it reflects the most recent expression by Congress of what constitutes active banking and therefore informs what the stand-alone active bank exception in section 1297(b)(2)(A) was meant to cover. Taxpayers may continue to rely on the Active Bank Notice and the 1995 Proposed Bank Regulations pending final regulations.

The 2021 Proposed Preamble requests comments on whether Proposed Regulations section 1.1297-1(c)(2) provides sufficient guidance to foreign banks, such that the Active Bank Notice and 1995 Proposed Bank Regulations can be withdrawn, whether alternatively the 1995 Proposed Bank Regulations should be finalized rather than Proposed Regulations section 1.1297-1(c)(2), whether both sets of proposed regulations should be finalized, or whether a single harmonized set of rules should be provided.²⁴

A. Overview of the Modified 954(h) Regime

To be an active bank under the Modified 954(h) Regime:

(1) Licensing. The institution must be (x) licensed in the United States or (y) licensed or authorized by a bank regulatory authority in the country in which it is chartered or incorporated (or, in the case of a qualified business unit (“QBU”), in the country in which the QBU maintains its principal office) to do business as a bank. The banking license must authorize the institution or such QBU (x) to accept bank deposits from residents of that country and (y) to carry out one or more of the activities listed in section 954(h)(4) (i.e., making loans; purchasing or discounting accounts receivable, notes, or installment obligations; engaging in leasing (including entering into leases and purchasing, servicing, and disposing of leases and leased assets); issuing letters of credit or providing guarantees; providing charge and credit card services; or rendering services or making facilities available in connection with these activities) (“954(h)(4) Specified Banking Activities”).

By contrast, to be an active bank under the 1995 Proposed Bank Regulations the institution must either (1) be licensed by federal or state bank regulatory authorities to do business as a bank in the United States (with the caveat that a limited representative office license that does not include authority to take deposits and make loans does not qualify) or (2) be licensed or authorized to accept deposits from residents of the country in which it is chartered or incorporated and authorized to conduct, in that country, one or more of the “banking activities” described in paragraph (f)(2) of the regulation (“1995 Specified Banking Activities”).²⁵ A foreign corporation does not satisfy these requirements if one of the principal purposes for its obtaining a banking license or authorization was to satisfy the PFIC requirements. A separate QBU banking license does not appear to be taken into account.

(2) Deposit taking. Under the Modified 954(h) Regime, the institution must regularly take bank deposits from unrelated customers in the ordinary course of its banking business. There is no explicit “substantiality” requirement or quantum of deposits required (beyond the “regularity” requirement) and the location of the depositor does not appear to be relevant.

By contrast, under the 1995 Proposed Bank Regulations, the institution must regularly take deposits from residents of its home country and these must be “substantial” (which is a factual, qualitative test referencing other local bank practices). The 1995 Proposed Bank Regulations (unlike the Active Bank Notice and Modified 954(h) Regime) do not distinguish “related” from “unrelated” customer deposits if the depositor is a bona fide “customer”.

(3) Active conduct. The institution must regularly carry on one or more of the 954(h)(4) Specified Banking Activities with unrelated customers in the ordinary course of its banking business.²⁶ Neither the 2021 Proposed Regulations nor section 954(h) imposes the requirement of conducting a lending business if one of the other 954(h)(4) Specified Banking Activities is actively conducted. Under section 954(h)(2)(A)(ii) “substantial activity” must be conducted with respect to that business. What constitutes “substantial activity” is not separately defined. Under 1995 Proposed Bank Regulations, there is no specific requirement that a U.S. licensed foreign bank actively conduct business (although, in practice, it seems highly unlikely a foreign bank would incur the regulatory burden involved if it did not). However, a non-U.S. licensed foreign bank must (in addition to taking deposits as discussed above) “regularly” make loans to home-country customers, although no distinction is drawn between related and unrelated customers as it is in the Active Bank Notice.

(4) Exempted income. Although qualification for the exception is tied to 954(h)(4) Specified Banking Activities, the scope of the exempted income for a bank that meets the requirements to be an active bank is not entirely clear. Qualified banking or financing income under section 954(h) (which is what section 954(h)(1) exempts from FPHCI) is any income of a qualified bank (1) if derived in the active conduct of a banking, financing or similar business, (2) if derived by that corporation or a QBU, (3) if substantially all of the activities to earn the income are conducted directly by the corporation or QBU in its home country and (4) the corporation or QBU is taxed by its home country on the income, with the proviso that (5) qualified banking income does not include income derived from one or more transactions with customers located in a country other than the home country of the eligible CFC or QBU of such corporation unless such corporation or QBU conducts substantial activity with respect to a banking, financing, or similar business in its home country.²⁷ Thus, subject to those conditions, a broad universe of financing income may be exempted (e.g., financial income that is not narrowly traditional banking income) if the company otherwise qualifies as an active bank, even if not tied directly to the activities that allowed it to qualify as a bank, subject to meeting the home country-related and other requirements. On the other hand, Proposed Regulations section 1.1297-1(c)(2) requires the exempted income to be derived in the conduct of the banking business. The extent to which this gloss limits the categories of income conducted by the qualified active bank that may be exempted from FPHCI is not entirely clear.

By contrast, under the 1995 Proposed Bank Regulations and the Active Bank Notice, banking income is limited to the same specified regulatory categories of “banking income” that are also relevant for qualification purposes. However, as noted above,²⁸ the list of income categories is considerably broader for most purposes than under section 954(h)(4).

(5) Treatment of affiliates. The Modified 954(h) Regime applies to the separate tested foreign corporation, and its QBUs, but does not cover income of bank affiliates that are not themselves licensed banks. The 2021 Proposed Preamble states that “[a]s proposed, the exception does not apply to affiliates of a foreign bank that do not independently qualify for the exception, in light of the fact that section 954(h) takes affiliates into account only for purposes of treating the activities of same-country related persons that are CFCs as activities that are conducted directly by an eligible CFC if certain conditions are satisfied. See section 954(h)(3)(E).”

By contrast, the 1995 Proposed Bank Regulations treat income from 1995 Specified Banking Activities earned by a qualified bank affiliate as non-passive for purposes of classifying any member of the “related group” as a PFIC. The related group includes any person related within the meaning of section 954(d)(3) to the tested foreign corporation (i.e., controlling, controlled by or under common control with such person where control is more than 50% of the vote or value of a corporation or more than 50% by value of a non-corporate entity). To be a qualified bank affiliate, at least 60% of the entity's gross income must be banking, securities or active insurance income and at least 30% of the related group's aggregate gross financial services income (as defined in regulations under section 904) must be banking income earned by active banks and at least 70% of the related group's aggregate gross financial services income must be banking, securities or active insurance income (such tests in combination, the “1995 Banking Related Group Requirements”). The Active Bank Notice also has rules for income of qualified affiliates of a bank.

(6) Anti-abuse rules. Section 954(h)(7) contains various anti-abuse rules, which we believe are incorporated by cross-reference into the Modified 954(h) Regime. It disregards (1) any item of income, gain, loss, or deduction with respect to any transaction or series of transactions one of the principal purposes of which is qualifying income or gain for the exclusion under section 954(h), including any transaction or series of transactions a principal purpose of which is the acceleration or deferral of any item in order to claim the benefits and (2) any item of income, gain, loss, or deduction of an entity which is not engaged in regular and continuous transactions with customers which are not related persons. Also disregarded is any item of income, gain, loss, or deduction with respect to any transaction or series of transactions utilizing, or doing business with (1) one or more entities in order to satisfy any home country requirement, or (2) a special purpose entity or arrangement, including a securitization, financing, or similar entity or arrangement, in either case, if one of the principal purposes of such transaction or series of transactions is qualifying income or gain for the exclusion under section 954(h). Finally, a related person, an officer, a director, or an employee with respect to any CFC (or QBU) which would otherwise be treated as a customer is not so treated if a principal purpose of the relevant transaction is to satisfy any requirement of section 954(h). In comparison, the 1995 Proposed Bank Regulations would ignore a license or authorization obtained with a principal purpose of meeting the PFIC exception as discussed above. However, the 1995 Proposed Bank Regulations do not contain other broader anti-abuse principles like those under section 954(h).

B. General Comments and Recommendation.

We believe there should be a single integrated set of rules rather than the current patchwork of tests under the Active Bank Notice, 1995 Proposed Bank Regulations and 2021 Proposed Regulations. It is very burdensome for taxpayers, advisors and the IRS to try to apply three separate sets of rules (which differ in complex and subtle ways) to undertake what is already a complex and fact-intensive analysis. There appears to be no good policy reason for retaining these multiple, somewhat inconsistent alternative tests.

We believe it is critical that any final active bank regime will cover most institutions engaged in bona fide “traditional” banking as that business is actually conducted in today's real-world financial markets. The prior history of the active banking exception is instructive. The Active Bank Notice imposed restrictive tests. This resulted in many situations in which institutions that, intuitively, were clearly bona fide banks were potentially treated as PFICs. The IRS was forced to resort to private letter ruling relief to avoid absurd results²⁹ and in response adopted the more flexible and qualitative approach in the 1995 Proposed Bank Regulations. While it is reasonable to be concerned that the exception should not be available, for example, to, tax haven companies “disguised” as banks solely to avoid PFIC status, prior history also illustrates why it will be self-defeating to let theoretical abuse potential concerns lead to unduly restrictive tests that also disqualify legitimate banking institutions unless the government is willing to devote substantial resources to providing expeditious private ruling relief (which may not even be practicable in contexts in which the issue often arises, like initial public offerings by foreign corporations, due to tight timing constraints). Limiting the active bank exception to licensed deposit-taking banks and their affiliates in practice itself greatly reduces the potential for “disguised” investment entities to qualify as active banks. To the extent that is insufficient, as recommended below, we believe that adoption both of the anti-abuse rule in the 1995 Proposed Bank Regulations and the anti-abuse rules under section 954(h) will better address such concerns than adopting an excessively restrictive definition of banking and banks as a general matter.

The government has concluded that application of section 954(h) in its totality, including its broader application to non-bank financing businesses, is not determinative for PFIC purposes. There is then no compelling reason why section 954(h) must determine the rules under final regulations in other respects. We agree with the 2021 Proposed Preamble that section 954(h) provides useful guideposts that can be applied to interpret section 1297(b)(2)(A) in the absence of final regulations, because the two provisions have similar and complementary purposes. The implication that section 954(h) should be viewed as “more” indicative generally of Congressional intent as to the conditions under which banking income of a foreign entity should be treated as non-passive than section 1297(b)(2)(A) because it is more recent should, however, be treated with caution and applied judiciously. As the 2021 Proposed Preamble notes, the legislative history to section 954(h) explicitly states that the phrase “active conduct of a banking business” under section 954(h) is intended to have the same meaning as under the 1995 Proposed Bank Regulations, implicitly endorsing those rules. There may be aspects of section 954(h) that reflect an updated Congressional understanding of what constitutes a traditional banking business but that does not mean section 954(h) principles should simply be assumed to supplant any prior rules when it is not clear Congress intended a consistent approach in the PFIC and Subpart F contexts.

Final integrated regulations should adopt the approach most consistent with the purposes of the PFIC regime and the statute while reflecting current market realities to the greatest extent possible, whether that means following the 1995 Proposed Bank Regulations, the Modified 954(h) Regime, some combination of the two, or neither. Fortunately, the 1995 Proposed Bank Regulations and the Modified 954(h) Regime overlap in significant respects.

C. Branches and “Unlicensed” Bank Affiliates

The most significant difference between the Modified 954(h) Regime and the 1995 Proposed Bank Regulations lies in the treatment of branches and affiliated entities. The Modified 954(h) Regime takes into account branches that are QBUs (for example, in determining whether the licensing requirement is met) whereas the 1995 Proposed Bank Regulations look only to the country of organization of the legal entity. The Modified 954(h) Regime approach to QBUs better reflects modern realities of multi-national banking in that respect. We suggest that final regulations adopt this approach to QBUs.

As noted above, however, the 2021 Proposed Preamble does not address in detail how the active banking test should apply to income of bank affiliates. It merely mentions that the Modified 954(h) Regime did not incorporate rules for bank affiliates in light of the fact that section 954(h) takes affiliates into account only for certain purposes of treating the activities of same-country related persons that are CFCs in testing related CFC income. It is unclear why that should be dispositive for purposes of final regulations. Treasury and the IRS allowed taxpayers to continue to rely on the 1995 Proposed Bank Regulations which include rules for qualified bank affiliates. They also have clearly concluded that the exceptions of section 954(h) are not technically incorporated by cross-reference under the PFIC statute reference to FPHCI. Otherwise, they would have also adopted an exception for active non-bank financing businesses. Given that technical conclusion, section 954(h) is at most merely instructive for purposes of the PFIC active bank exception. As a recent indication of what Congress considered “active” financing, section 954(h) may indeed be instructive in certain respects. On the other hand, various aspects of section 954(h) reflect policy concerns that are peculiar to Subpart F that should not necessarily be incorporated into the PFIC active banking regime.

Fundamentally, a critical difference between Subpart F and the PFIC rules is that Subpart F generally applies to CFCs on separate company basis while the PFIC rules for many purposes apply a “group” approach (for example, under the look-through subsidiary rule and when applying tests to groups that split ownership and operational activities among different affiliates). There are also other important distinctions. CFC classification turns on U.S. ownership, not the nature of the income and assets of the CFC. Treating income as active banking income may affect what income is picked up as Subpart F income or GILTI (or exempted) but not whether the company is within scope of the regime in the first instance. Under the PFIC regime, by contrast, the treatment of income or assets as “passive” determines whether the entity is subject to the regime at all and not whether a particular “passive” income item must be currently included by a U.S. taxpayer.³⁰ Subpart F as a policy matter is also concerned not merely with whether income is passive but whether it is “mobile” (i.e., easily diverted from the United States to offshore or between different country affiliates). From a “mobility of income” standpoint, the “same country” and “single company” focus of various Subpart F related provisions under section 954(h) makes sense. It makes far less sense as a basis for distinguishing whether income is earned through active business conduct rather than passive investment under the PFIC rules.

As mentioned, Subpart F has no corollary of the 25% look-through rule in section 1297(c), which applies what could be viewed as a form of quasi-consolidation solely for PFIC classification testing purposes. The look-through rule means that treating a separate subsidiary affiliate's income as passive (because the affiliate is not itself the licensed bank and the income is “tested” at the subsidiary level) has the potential to disqualify the bank parent from the exception and cause it to be a PFIC. Given Congressional endorsement of the 1995 Proposed Bank Regulations (which would not have had this result) in the legislative history of section 954(h) noted above, there is no reason to think Congress intended this aspect of section 954(h) to override the 1995 Proposed Bank Regulations for PFIC purposes.

Practically, an approach which requires qualification for the active bank exception on a stand-alone legal entity-by-entity basis will preclude many multi-national banking groups from qualifying for the exception. For regulatory and other non-tax reasons it is common for banks to be in groups that include separately incorporated affiliates that are not licensed banks in order to conduct certain functions and activities. Failing to provide relief for qualified bank affiliates given the 25% look-through rule could disqualify many banks or bank holding company parents of groups that intuitively are clearly bona fide multi-national banks based on the happenstance of group structure. It will also result in anomalous, inconsistent PFIC results for commercially similar groups organized differently to comply with local regulatory regimes.

It is true that extending the active banking exception to qualified bank affiliates may result, for certain purposes, in disparate treatment of activities depending on whether they are conducted within or without a banking group. Inconsistent treatment of the same financing activities in some cases depending on whether these are conducted within or without “banking” groups is, however, an inevitable result of differentiating between bona fide “banks” and other active financing businesses on the basis the government has adopted (i.e., using licensing and deposit taking as proxies to identify bona fide “active” banks). Eligibility for the PFIC active bank exception turns not only on how “actively” lending or other financing activities are conducted but on proxy features like (1) bank regulatory and licensing status and (2) deposit taking activities, which are the primary features that distinguish traditional “banks” from other active financing businesses (i.e., from “shadow banks”). That a similar activity may be treated differently depending on whether it is conducted by a licensed institution or group simply follows from this approach. It would be unwise to seek consistency if the result is treating many bona fide multi-national banking groups as PFICs.

In the abstract, there are four possible approaches to bank affiliates: (1) their financial income and assets could be treated as passive for all purposes unless the affiliate itself is a licensed bank; (2) their financial income and assets could be treated as active for all purposes if there is an affiliated licensed bank owner and the other indicia of active conduct of a financing activity, which would generate active banking income if conducted by a licensed bank, are present at the affiliate level (which would protect not only the bank parent, and shareholders of the bank parent, but any minority shareholder of the affiliate), (3) assuming such active conduct, the affiliate income and assets could be treated as active (a) for purposes of classifying the bank parent, (b) collaterally, protecting the indirect shareholders of the affiliate (i.e., a bank holding company and bank holding company shareholders), and (c) as to the affiliate itself, but not (d) as to any minority shareholder of the affiliate, or (4) the affiliate income and assets could be treated as active for purposes of classifying only the look-through parent/s of a qualified affiliate. For the reasons above, we do not favor approach (1). Having concluded that a non-bank active financing business may be a PFIC, it does not make sense to extend the active treatment to a minority shareholder merely because the other shareholder happens to be a licensed bank. Accordingly, we do not favor approach (2). We see no compelling concerns justifying approach (4), which would frequently result in treating non-bank bank holding company parents of a financial group (a very typical structure) as PFICs. In general, we support approach (3), which is the general approach of the 1995 Bank Proposed Regulations to qualified bank affiliates. Moreover, approach (3) seems most consistent with the approach to attribution of activities among members of a group for purposes of other rules in the 2021 Final Regulations, pursuant to Treasury Regulations section 1.1297-2(e). Accordingly, we recommend that final regulations follow the 1995 Proposed Bank Regulations in this respect and incorporate the rules for qualified bank affiliates. We support including the quantitative tests imposed under the 1995 Banking Related Group Requirements summarized earlier to ensure an affiliate is properly part of an integrated financial group business conducted by the banking group.

D. Country-specific Licensing and Activity Requirements

A second potential material difference between the Modified 954(h) Regime and the 1995 Proposed Bank Regulations is the focus under section 954(h), when defining “qualified banking income,” on whether “home” country activity generates the income or “home” country customers make deposits. The Modified 954(h) Regime treats income as qualified banking income, for example, only if (1) substantially all of the activities to earn the income are conducted directly by the corporation or QBU in its “home” country and (2) the corporation or QBU is taxed by its home country on the income. Qualified banking income similarly does not include income derived from one or more transactions with customers located in a country other than the home country of the eligible CFC or QBU of such corporation unless such corporation or QBU conducts substantial activity with respect to a banking, financing, or similar business in its home country.

As noted above, the focus of PFIC regime is to distinguish between income earned through active business conduct rather than from passive investment. It is important for this purpose, however, to distinguish between two aspects of the PFIC active banking regime: (1) how income and activities are classified for purposes of determining whether an institution is a “bona fide” bank (bank classification) versus (2) what income derived by an institution that has qualified as a bona fide bank is then exempted from FPHCI (scope of active banking income). Clearly “passive” investment income should not count favorably in determining whether institution is a bona fide “bank” rather than some other kind of active financing business, but nor should such income be considered “active” when testing the PFIC status of an entity that otherwise is clearly a bona fide bank. But there may be other categories of income from activities that perhaps should not define a bona fide bank, because they are not “core” activities of a traditional bank, but nevertheless should be considered active when earned by an otherwise bona fide bank. For example, under the government's approach, a stand-alone factoring business should not qualify for the active banking exception (because no matter how “actively conducted” that business may be, only income earned by deposit taking licensed “banks” qualifies). It does not follow, however, that income from factoring earned by a licensed, deposit-taking bank along with its income from core banking activities should be considered passive.

Given the commercial realities of modern multi-national banking, once an institution has been determined to be a “bona fide” bank on an appropriate basis, we see no policy or technical reason for PFIC purposes to limit its FPHCI exempted income to “local” banking income, i.e., based on whether the functions of the active banking business and customers are all concentrated in a single jurisdiction, whether that jurisdiction is the legal “home” country of the institution, or whether the resulting income is taxed in the home country rather than some other country. This is irrelevant to the policy concerns underlying the PFIC regime. In real-world multi-national banking transactions, various personnel in a home office and various QBUs may all be involved in different aspects of a banking transaction. Requiring taxpayers and their advisors to conceptually match up the resulting income between activity, customer, and taxing locations for purposes of the requirements above serves no policy purposes and simply introduces unwarranted complexity. What should matter is the totality of the activity involved in conducting the business (i.e., whether it is active business income rather than passive investment income) and not where it is conducted.

Conversely, taking into account the locus of banking activity may have policy relevance solely for purposes of the bank licensing and authorization requirements, which go to whether the institution is a bona fide bank and whether its bank regulatory status should be respected. The 1995 Proposed Bank Regulations and Modified 954(h) Regime arguably use bank licensing and supervision as a proxy for whether the business is actively conducted as well as whether it is a banking business. The underlying intuition is that “real” bank regulatory regimes typically impose very substantial regulatory burdens on licensed banks, which would deter any institution not planning to engage in a substantial active banking business in the jurisdiction. If the licensing requirements are entirely divorced from the requirements relating to banking activities — i.e., serve as the sole touchstone for whether a business is passive or active — this might create opportunities for abuse. For example, an institution might seek to obtain some kind of limited banking license in a regulatory “light touch” jurisdiction or tax haven in which it conducts minimal real banking business to qualify its income elsewhere as active banking income. We do not believe this is a legitimate concern under the 1995 Proposed Bank Regulations (or any similar framework that requires some minimum quantum of bona fide banking activity on a groupwide basis in the relevant licensing jurisdictions but it could otherwise be of concern. Attempting to solve this problem as a general matter by distinguishing directly between different country bank regulatory regimes (e.g., by comparing them to the U.S. bank regulatory regime, which Congress clearly considered a good proxy for traditional active banking) or based on the nature of the license (i.e., a general banking license versus a limited or international banking license) is not a workable solution. This would require the IRS, taxpayers and tax advisors to make inquiries into foreign bank regulatory regimes and their comparability to the U.S. regime. Such analysis would be subjective and outside the core areas of expertise of the IRS, most taxpayers and their tax advisors. Treasury and the IRS could perhaps consider, however, adopting by ruling a list of jurisdictions that would be treated as equivalent to the United States (i.e., in which the regulatory regime for banks is so clearly burdensome that there is little risk that a corporation would expend the time and resources to become licensed there merely to circumvent PFIC status) (“Equivalent Regulatory Jurisdictions”.)

We would also emphasize that anti-abuse rules, which the 1995 Proposed Bank Regulations and Modified 954(h) Regime include, already constrain taxpayers' ability to abuse the licensing requirement. However, we recognize that anti-abuse rules require a somewhat subjective analysis of taxpayer intent and are not a perfect solution. For example, if bank licensing is legally required in a country and some minimal but non-immaterial local business is actually conducted there, it may be hard to show the license was obtained with a principal purpose of meeting the PFIC rules for purposes of the 1995 Proposed Bank Regulations' anti-abuse rule. The license was obtained because the law required it to be obtained. The IRS would also have to show that the local business was conducted with the proscribed purpose. If the IRS believes this type of abuse is a legitimate concern, then we would recommend that final regulations adopt the broader anti-abuse rules of section 954(h)(7), which would provide a basis for ignoring a low level of business activity motivated by PFIC qualification.

Alternatively, if the government still remains concerned about potential abuse, it could also impose activity restrictions tied to licensing-country activity, provided this is relevant solely for bona fide bank classification purposes and does not also determine the scope of income of a bona fide bank that is active banking income. For example, we would support the use of a test that requires the corporation or QBU to conduct a substantial amount of actual banking business activity in the jurisdiction that granted the banking license on which the company is relying to qualify for the active banking exception (and perhaps also that some material amount of deposits be taken from customers in that jurisdiction). Such activity could be defined by requiring direct lending or, preferably, a somewhat broader but still limited set of “core” banking activities such as the 954(h) Specified Activities. This would back stop the anti-abuse rule by ensuring that the banking license was obtained for bona fide banking business reasons. Other than for that purpose, however, the location of activity, customers or depositors should not be relevant. Such requirements could be made less burdensome if institutions licensed by Equivalent Regulatory Jurisdictions were excused from this requirement.

The categories of income earned by a bona fide bank, its QBUs (or qualified bank affiliates) that should be considered active banking income once the bank classification requirements above are met should be broadly defined (at least as broadly as income derived from the 1995 Specified Banking Activities) as long as the business is earned in the active conduct of the tested foreign corporation's business or that of its QBUs and qualified bank affiliates.

We do not believe adopting rules along the lines above will permit abuse, for example, planning strategies such as embedding an investment portfolio or hedge fund within a banking group in order to avoid PFIC status. The 1995 Specified Banking Activities, for example, do not generally include owning or trading in securities, commodities or derivatives and focus on activities involving “customers”.³¹ Accordingly, income from activities involving securities, commodities or derivatives as a general matter can avoid PFIC “passive” treatment only by qualifying separately under some exception to FPHCI other than the active banking exception, for example on the basis that the assets were dealer property (i.e., inventory). Moreover, the definition of “banking income” in the 1995 Proposed Bank Regulations (which we support) limits banking income to the gross income derived from the active conduct (within the meaning of Treasury Regulations section 1.367(a)-2T(b)(3)) of the 1995 Specified Banking Activities.