SIFMA Suggests Changes to Prevent Harmful Effects of BEAT Regs

February 19, 2019

Internal Revenue Service
CC:PA:LPD:PR (REG-104259-18)
Courier's Desk
1111 Constitution Avenue, N.W.
Washington, DC 20224

Re: IRS REG-104259-18 (Proposed Regulations under Section 59A)

Ladies and Gentlemen:

The Securities Industry and Financial Markets Association (SIFMA)¹ appreciates the opportunity to submit comments on the proposed regulations (the “Proposed Regulations”) under Section 59A of the Internal Revenue Code of 1986, as amended (the “Code”).²

I. INTRODUCTION

We recognize and appreciate the efforts of Treasury to provide guidance on the implementation and application of Section 59A while avoiding undue interference with ordinary course financial markets transactions. However, we believe that several of the provisions in the Proposed Regulations are inconsistent with the relevant provisions of Section 59A. Further, we believe that certain of the provisions in the Proposed Regulations, if adopted in their current form, will have uniquely detrimental effects on financial groups, their customers, and potentially the financial markets. Members of the financial industry serve as counterparties to a large portion of the derivatives entered into in the financial markets, including derivatives that hedge foreign currency risks. Additionally, the efficient functioning of the financial markets often requires the intermediated movement of cash and securities between jurisdictions, which financial institutions typically accomplish through sale-repurchase and securities lending transactions between affiliates. To the extent the Proposed Regulations effectively penalize or discourage either of these types of transactions, the ability of financial institutions to continue to enter into these transactions may be hampered, which could have negative implications for financial markets at large.

Accordingly, in order both to provide guidance that is consistent with the statutory text and legislative intent underlying Section 59A and to prevent the negative effects enumerated above, as well as certain harmful results that are applicable to a wider range of taxpayers, we are recommending that the IRS and the Treasury Department (“Treasury”) make several changes to the Proposed Regulations before they are finalized. Our first set of recommendations addresses issues with respect to the Proposed Regulations that are of particular relevance to the financial industry (“Industry-Specific Recommendations”). Our second set of recommendations includes other recommendations addressing issues with respect to the Proposed Regulations that we believe are more widely applicable (“General Recommendations”). In addition, we also support two sets of recommendations that we understand the Institute of International Bankers (“IIB”) has provided or will provide to Treasury.

Industry-Specific Recommendations

1. Section 988 currency losses (“Section 988 Losses”) should not be excluded from the denominator of the base erosion percentage calculation, except for such losses on transactions with related non-U.S. persons that are also excluded from the numerator. Alternatively, the exception for Section 988 Losses should not apply to positions that are or would be (if entered into with a related non-U.S. person) “derivatives” eligible for the “QDP exception” (as defined below), without regard to the “QDP Reporting Condition” (also defined below, in Recommendation 4(a)).

2. Treasury should eliminate the general exclusion for securities lending transactions from the definition of a “derivative,” although it would be reasonable for Treasury to clarify that interest paid on the “cash leg” (as defined below) of a securities lending transaction does not qualify for the exception for qualified derivative payments within the meaning of Proposed Section 1.59A-6 (“QDPs,” and the exception, the “QDP exception”).

3. Treasury should revise the definition of a “base erosion payment” to exclude a loss recognized by the taxpayer on the sale or exchange of property by the taxpayer to a related non-U.S. person.

4. With respect to the reporting rules for QDPs, (a) Treasury should make explicit that a failure to comply with the Form 8991 reporting requirements by a taxpayer that is not a reporting corporation (within the meaning of Section 1.6038A-1(c)) does not affect whether or not any payments made by the taxpayer are considered QDPs and should clarify the consequences of failing to comply with the Form 8991 QDP reporting requirements for the taxpayer; (b) until the final reporting regulations are effective, all taxpayers should be allowed to report the aggregate amount of their QDPs under a good faith standard; (c) Treasury should clarify that the relevant payments for each derivative transaction should be netted, consistent with the Net Mark-to-Market Method (as defined below), to arrive at the aggregate QDP amount that must be reported on Form 8991; and (d) Treasury should make clear that certain QDPs need not be reported.

5. Treasury should clarify that transfers made pursuant to revenue or profit sharing arrangements in connection with global dealing and similar arrangements are not base erosion payments and should be excluded from the numerator of the base erosion percentage.

6. The Proposed Regulations require that, in the case of positions that are marked to market, all income, deduction, gain or loss on each position for the year must be combined for purposes of the base erosion percentage test (the “Net Mark-to-Market Method”). The Net Mark-to-Market Method should be included as a safe harbor but should not be the exclusive method under the Proposed Regulations for calculating payments and accruals with respect to a position that is marked to market.

General Recommendations

7. Treasury should revise the definition of “base erosion payment” to include an exception for an acquisition of depreciable property from a related non-U.S. person in a nonrecognition transaction.

8. Treasury should (a) increase the threshold percentage interest in a partnership at or above which a partner is required to take into account its distributive share of any partnership amount of base erosion tax benefits from 10% to 25%, (b) eliminate the requirement that a partner's interest in the partnership have a fair market value of less than $25 million in order for a partner not to have to take into account its distributive share of any partnership amount of base erosion tax benefits, and (c) provide that the minimum percentage interest requirement also applies with respect to treating a payment made to a partnership as being made to each partner based on the partner's distributive share.

9. For purposes of determining modified taxable income, the taxpayer's taxable income should be determined by taking into account the taxpayer's entire net operating loss deduction allowed under Section 172 (and thus may be an amount less than zero), rather than only the amount of the deduction that does not reduce the taxpayer's taxable income to an amount less than zero.

Part II of this letter discusses our Industry-Specific Recommendations in greater detail, and Part III discusses our General Recommendations that are not industry-specific. Finally, in Part IV, we discuss two additional recommendations with respect to which we understand IIB has provided or will provide recommendations to Treasury.

II. INDUSTRY-SPECIFIC RECOMMENDATIONS

Recommendation 1: Section 988 Losses should not be excluded from the denominator of the base erosion percentage calculation, except for such losses on transactions with related non-U.S. persons that are also excluded from the numerator. Alternatively, the exception for Section 988 Losses should not apply to positions that are or would be (if entered into with a related non-U.S. person) “derivatives” eligible for the QDP exception, without regard to the QDP Reporting Condition.

Proposed Section 1.59A-3(b)(3)(iv) provides that Section 988 Losses are not base erosion payments and are therefore not included in the numerator of an applicable taxpayer's base erosion percentage. We agree with and support this result. The Proposed Regulations also exclude all Section 988 Losses from the denominator,³ although the preamble indicates that Treasury is specifically requesting comments on whether this wholesale exclusion of Section 988 Losses is appropriate or if the denominator exclusion “should be limited to transactions with a foreign related party.”⁴ For the reasons discussed below, we believe Section 988 Losses with respect to transactions with persons other than related non-U.S. persons should not be excluded from the denominator (such Section 988 Losses, “Non-RP Section 988 Losses”), so that the only Section 988 Losses excluded from the denominator would be Section 988 Losses on transactions with related non-U.S. persons that are also excluded from the numerator.

The preamble states that Section 988 Losses are excluded from the numerator under Proposed Section 1.59A-3(b)(3)(iv) because Treasury has determined that, “these losses do not present the same base erosion concerns as other types of losses that arise in connection with payments to a foreign related party.”⁵ However, the preamble then goes on to state that Section 988 Losses are also excluded from the denominator, even if they are Non-RP Section 988 Losses, without offering any explanation for why these Non-RP Section 988 Losses should be excluded from the denominator. We believe that this exclusion is neither necessary nor appropriate to carry out the provisions of Section 59A.⁶ Moreover, we question Treasury's authority, by regulation, to implement the wholesale elimination of a category of statutorily permitted deductions arising out of transactions with unrelated third parties, given that the statute provides that the denominator (with specified exceptions not relevant here) is “the aggregate amount of deductions . . . allowable to the taxpayer under this chapter for the taxable year.”⁷

There are at least three reasons why the exclusion of Section 988 Losses from the denominator should be eliminated in the final regulations, except with respect to Section 988 Losses on transactions with related non-U.S. persons that are also excluded from the numerator. First, there is no question that the stated reason for the exclusion of Section 988 Losses from the numerator — “that such losses do not present the same base erosion concern” as other types of losses on payments to related non-U.S. persons — has no application to losses with respect to third-party transactions/positions, which would be included only in the denominator. Under Section 59A(c)(4)(A)(i), the numerator of the base erosion fraction includes and reflects the aggregate deductions attributable to certain specified base-eroding payments to related non-U.S. persons, such that it is appropriate for Treasury to take into account the extent to which a given type of payment to a related non-U.S. person actually presents base erosion concerns in crafting exclusions from the numerator. However, under Section 59A(c)(4)(A)(ii), the denominator of the base erosion fraction is based on all of the taxpayer's aggregate deductions (with limited exceptions), without regard to whether those deductions arise out of payments or transactions with related non-U.S. persons or unrelated U.S. or non-U.S. persons — the extent to which a deduction does or does not have base-eroding potential is simply irrelevant to the determination of the denominator. Thus, the stated reason for not including Section 988 Losses in the numerator cannot be the rationale for excluding Non-RP Section 988 Losses from the denominator.

Second, there is no question that Section 988 Losses are real economic losses, that are not fundamentally different from other types of deductions and losses with respect to transactions that are included in the denominator.⁸ Moreover, at least for an accrual-method taxpayer, there generally is not any significant timing mismatch between when a Section 988 Loss arises as an economic matter and when that loss is deductible. There is therefore no reason why a Section 988 Loss should be excluded from the denominator on the ground that it does not represent a real economic loss or that the deduction in respect thereof arises in the “wrong” year or is capable of manipulation. This is in contrast to the only deductions specifically enumerated in the statute that are excluded from the denominator, without regard to whether they are also excluded from the numerator. See Section 59A(c)(4)(B)(i) (excluding from the denominator deductions under Sections 172, 245A and 250); Proposed Section 1.59A-2(e)(3)(ii)(A).

Finally, it is possible that Treasury may believe that Section 988 Losses should be excluded from the denominator simply as a matter of symmetry if Section 988 Losses are excluded from the numerator. We believe the exclusion of all Section 988 Losses overshoots the mark in terms of symmetry. It makes sense from a symmetry perspective to exclude Section 988 Losses on transactions with related non-U.S. persons from the denominator, because those losses are also specifically excluded from the numerator, consistent with the approach taken with respect to other items specifically excluded from the numerator.⁹ However, we see no rationale based on symmetry for excluding Non-RP Section 988 Losses from the denominator. Indeed, excluding them creates an asymmetry, contrary to the statute, between such losses and other real economic losses from transactions with unrelated parties that give rise to deductions that are included in the denominator.

Based on the above, we believe and recommend that all Section 988 Losses should be included in the denominator, except for Section 988 Losses with respect to transactions with related non-U.S. persons that are also excluded from the numerator. We note that this recommendation has the additional benefit of being the simplest approach for both taxpayers and the government to administer.

However, if Treasury concludes that an exclusion from the denominator for Non-RP Section 988 Losses arising from payments or accruals between unrelated parties is appropriate, we urge that any such rule not extend to Non-RP Section 988 Losses that would be QDPs if entered into with related non-U.S. persons, without regard to the QDP Reporting Condition. The reason is that the exclusion of Section 988 Losses with respect to transactions with unrelated parties is especially inappropriate for taxpayers that are marking their currency derivative positions to market. Under the statute and Proposed Regulations, most or all foreign-currency linked derivatives entered into by these taxpayers would in any event be excluded from the taxpayer's numerator under the QDP exception. Thus, while the exclusion of Section 988 Losses from the base erosion percentage calculation may provide these taxpayers with a marginal numerator benefit (e.g., for positions that are not derivatives or are not marked to market), the main effect of the exclusion of Section 988 Losses from the calculation is the exclusion of all Non-RP Section 988 Losses from the denominator. We believe this result creates a baseless asymmetry between the treatment of losses on a potentially very large volume of currency derivative transactions and the taxpayer's transactions with respect to other property. For a taxpayer that marks its currency derivative positions to market, these transactions are no different in any relevant sense from non-currency derivatives. Section 988 Losses incurred by mark-to-market taxpayers are thus different from the perhaps more “idiosyncratic” Section 988 Losses that might be incurred outside of the ordinary course of business by other taxpayers.¹⁰ Even if Treasury believes that the exclusion of similarly idiosyncratic Section 988 Losses on transactions with related non-U.S. persons from the numerator justifies the exclusion of idiosyncratic Section 988 Losses on transactions with unrelated parties from the denominator, that rationale should not extend to Section 988 Losses incurred by mark-to-market taxpayers, which are ordinary course economic losses and therefore are most definitely not idiosyncratic.¹¹

We recognize that the logic of this approach might suggest that Section 988 Losses arising in connection with these same marked-to-market transactions entered into with related non-U.S. persons should not be per se excluded from the numerator, but perhaps should be subjected to the same rules that apply to payments and accruals with respect to all other (non-currency-related) derivatives, i.e., the QDP rules.¹² Accordingly, we would not oppose a “symmetrical” provision that would carve out “derivatives” that are or would be (if entered into with a related non-U.S. person) eligible for the QDP exception (without regard to the QDP Reporting Condition) from the general exclusion of Section 988 Losses in the proposed regulations.¹³ The effect of this proposal would be that payments and accruals with respect to taxpayers' marked-to-market currency derivative positions with related non-U.S. persons would generally be excluded from the numerator as QDPs — subject, as for all other marked-to-market derivative payments and accruals, to the QDP Reporting Condition — and such positions with all other persons would be included in the denominator.¹⁴

Recommendation 2: Treasury should eliminate the general exclusion for securities lending transactions from the definition of a “derivative,” although it would be reasonable for Treasury to clarify that the interest paid on the “cash leg” of a securities lending transaction does not qualify for the QDP exception.

Section 59A(h)(4), with certain exceptions not relevant here, defines a “derivative” as “any contract (including any option, forward contract, futures contract, short position, swap, or similar contract) the value of which, or any payment or other transfer with respect to which, is (directly or indirectly) determined by reference to” any stock, debt, actively traded commodities, currency and other specified items. The general definition of “derivative” in the Proposed Regulations mirrors the definition in the statute.¹⁵ However, Proposed Section 1.59A-6(d)(2)(iii) provides that a derivative contract “does not include any securities lending transaction, sale-repurchase transaction, or substantially similar transaction.”¹⁶

Before we discuss the technical issues with respect to this provision of the Proposed Regulations, it is important that Treasury understand the importance of sale-repurchase and securities lending transactions to the efficient functioning of the financial markets. As Congress¹⁷ and industry experts¹⁸ have recognized over the years, the markets for sale-repurchase transactions and securities lending and borrowing transactions are crucial for the trading of fixed-income securities and equities, including by providing critically needed liquidity in the financial markets, supporting a variety of trading strategies, facilitating trade settlements and supporting general financing techniques. The importance of securities lending transactions is reflected in their volume. The daily size of the global securities lending transactions market is roughly $1.5 trillion in value.¹⁹ In particular, sale-repurchase transactions and securities lending and borrowing transactions are known for providing market participants with a means to obtain specific securities or cash to be used in other transactions. As commentators noted, by improving the ability of taxpayers to settle trades and meet margin requirements, sale-repurchase transactions and securities lending and borrowing transactions support the smooth functioning of derivatives markets and contribute to the resilience of the financial system and the real economy.²⁰

Cross-border sale-repurchase transactions and securities lending and borrowing transactions among affiliates, which move the supply of cash and securities around the world to meet demand for cash and securities, are important to the basic functioning of securities markets. For example, it is very common for a U.S. broker-dealer to require shares of, say, a U.K. company (e.g., to lend to a U.S. customer that is executing a “short sale”), in which event it will typically borrow those shares from the affiliate that holds those shares (or has access to them from its customers), most likely a U.K. broker-dealer affiliate. It may not always be possible for a U.S. broker-dealer to have direct access to third parties, often located in a non-U.S. jurisdiction, who hold those required shares (e.g., a U.S. broker-dealer may not be registered as a broker-dealer in the non-U.S. jurisdiction where the counterparty is based). Additionally, as a business matter, a broker-dealer needs to build a close business relationship with a counterparty in order to access certain hard-to-borrow securities. It would be difficult for a broker-dealer to manage this type of relationship with multiple third-party counterparties in a different jurisdiction and thus, as a practical matter, a broker-dealer generally avoids this added cost by borrowing shares from its affiliate located in the relevant jurisdiction. Penalizing sale-repurchase transactions and securities lending and borrowing transactions by excluding them from the QDP exception altogether may dampen liquidity in the securities markets. From the perspective of the U.S. broker-dealer that is the securities borrower in the first transaction (between it and its U.K. affiliate) and the securities lender in the second transaction (between it and its customer), the securities leg of each transaction is a derivative that it marks to market for U.S. federal income tax purposes. Further, the U.S. broker dealer is essentially a conduit between its U.K. affiliate and the third party, so that it is clear there is no base erosion occurring in these transactions.²¹ And yet, if the U.S. broker-dealer is unable to exclude its “substitute payments” with respect to the underlying securities to its U.K. affiliate from its base erosion percentage (and modified taxable income calculations), the result will be wildly uneconomic, and, to the extent it results in additional tax liability under the BEAT, could adversely affect the ability of U.S. securities dealers to perform this crucial intermediation role in the financial markets.

We turn now to the technical issues with the proposed exclusion of sale-repurchase transactions and securities lending transactions from the definition of a “derivative.” We believe that the decades-long distinction between sale-repurchase transactions and securities lending transactions should be preserved in the context of Section 59A, not only because the distinction is deeply rooted in the tax law, but, for reasons explained below, precisely because preserving this distinction is needed to ensure that sale-repurchase transactions and securities lending transactions — which do bear some similarities, as noted in the preamble to the Proposed Regulations — are treated similarly for BEAT purposes.

As we will discuss further below, sale-repurchase transactions have always been treated as simply collateralized debt instruments for U.S. tax purposes, which do not involve any “derivative” within the meaning of Section 59A, without regard to the Proposed Regulations. On the other hand, a typical cash-collateralized securities lending transaction has for decades (at least) been treated for U.S. federal income tax purposes as consisting of two elements: a “cash leg” consisting of the cash collateral transferred by the securities borrower to the securities lender, and a “securities leg” consisting of the securities transferred by the securities lender to the securities borrower.²² The cash leg is appropriately thought of as indebtedness for U.S. tax purposes, or in any event as generating one or more payments or accruals that we agree, if paid by a U.S. taxpayer to a related non-U.S. person, should not benefit from the QDP exception to the numerator and should be considered a base erosion payment. However, the securities leg of a securities lending transaction is incontrovertibly a derivative within the meaning of Section 59A(h)(4)(A), because it is a contract the value of which (i.e., either the obligation to return a security or the right to receive a security) is determined by reference to a security. This is because, in a securities lending transaction, the securities borrower disposes of the securities it received in such a way that a third party beneficially owns the securities for U.S. federal income tax purposes (e.g., by simply selling them or by on-lending them in another securities lending transaction to another securities borrower, which in turn sells them).²³ The law is clear that the securities lender no longer owns the underlying securities (and the transaction cannot be a collateralized debt instrument) for U.S. federal income tax purposes, because only one taxpayer can own a given security at any given time, and in this case it is clear that the ultimate third-party buyer is the owner, not the original securities lender.²⁴ Because neither the securities lender nor the securities borrower beneficially owns the underlying securities, what exists between them is simply a contract, under which the securities borrower has the obligation to transfer identical securities to the securities lender at the end of the transaction. The value of this contract, to both parties, is determined by reference to the value of the underlying securities. The contract is thus clearly a “derivative” as defined in Section 59A(h)(4).²⁵

We describe in further detail below the various types of transactions that are covered by the proposed exclusion and explain why an exclusion of any portion of each of those types of transactions, other than a portion of a transaction that may properly be treated as debt for U.S. federal income tax purposes, from the definition of “derivative” is inappropriate, both as a technical matter and in order to ensure that sale-repurchase transactions and securities lending transactions are in fact treated similarly, rather than dissimilarly, for purposes of Section 59A. We then note that a limited subcategory of uncollateralized securities borrowing transactions may be sufficiently economically similar to debt financings that those transactions should be excluded from the definition of a “derivative,” and we explain that the most appropriate way for Treasury to address those transactions is through a targeted anti-abuse rule.

Sale-Repurchase Transactions. The first category of transactions covered by the exclusion are sale-repurchase transactions. In these transactions, the nominal securities seller “sells” securities to the nominal securities buyer who, simultaneous with the “purchase,” agrees to “resell” the securities to the nominal securities seller at a predetermined slightly higher price at a later date.²⁶ Throughout the term of the transaction the nominal securities buyer may be the legal owner of the securities, but for U.S. federal income tax purposes the nominal securities seller remains the beneficial owner of the securities.

The preamble states that because sale-repurchase transactions are treated as secured loans for U.S. federal income tax purposes, the Proposed Regulations provide that these are not treated as “derivatives.”²⁷ As noted above, it is clearly correct that sale-repurchase transactions are treated as secured loans for U.S. federal income tax purposes, as is evidenced by longstanding law.²⁸ Section 59A(h)(4)(A) provides that the term “derivative” does not include any evidence of indebtedness, and we therefore believe that a separate exception for sale-repurchase transactions is unnecessary. The interest paid or accrued on a sale-repurchase transaction is unaffected by any “offsetting” payments made by the cash lender — the nominal security buyer — with respect to the collateral securities, because the collateral securities remain beneficially owned for U.S. federal income tax purposes by the nominal securities seller. In other words, as described above, if the nominal securities buyer receives a payment on the securities and then pays it over to the nominal securities seller, that payment (to which we refer as an “underlying payment”) is not a payment between the parties to the transaction for U.S. federal income tax purposes (and thus cannot be a base erosion payment for purposes of Section 59A). Rather, the cash lender receives these underlying payments and passes them over to the nominal securities seller, as beneficial owner of the collateral securities, effectively as an agent or nominee of the nominal seller. Thus, the only deductible payments between the parties that are recognizable for U.S. federal tax purposes are interest payments made by the securities seller (i.e., a portion of the payments in blue in Diagram 1 above), which therefore are not QDPs, even absent regulations.

However, Treasury did not limit the scope of its exclusion to sale-repurchase transactions. After discussing why an exclusion of sale-repurchase transactions is appropriate, the preamble notes that securities lending transactions are economically similar to sale-repurchase transactions, and should be treated similarly.²⁹ The Proposed Regulations accordingly exclude all securities lending and sale-repurchase transactions from the scope of the term “derivative.” This exclusion thus covers both collateralized and uncollateralized securities lendings and borrowings (each as defined below), by U.S. taxpayers to or from a related non-U.S. person. We therefore next address these types of transactions.

Cash-Collateralized Securities Lending Transactions. A cash-collateralized securities lending is (for purposes of this discussion) a transaction in which a U.S. taxpayer lends securities to a related non-U.S. person (rather than merely posting them as collateral) in exchange for the receipt of cash collateral from the related non-U.S. person.³⁰

It is possible that the broad securities lending exclusion in the Proposed Regulations was motivated by a concern about a taxpayer taking the position that the amount it pays or accrues to a related non-U.S. person with respect to the cash that the taxpayer receives is not interest. We agree that this result would not be appropriate, and we believe that, even without the Proposed Regulations, under current law these amounts would not qualify for the QDP exception. This is because we think a cash-collateralized securities lending is properly characterized for U.S. federal income tax purposes as consisting of two distinct transactions: (1) a transfer of the cash collateral coupled with an obligation of the cash transferee to return that cash plus an interest amount (the “cash leg”) — which is likely simply a cash borrowing by the securities transferor — and (2) a transfer of securities coupled with an obligation of the securities transferee to make certain payments to the securities transferor and to return identical securities (the “securities leg”).³¹ (In Diagram 2 above, the cash leg is represented in blue, while the securities leg is in red.) Again, it is perfectly reasonable for regulations to make this clear for Section 59A purposes.

However, the Proposed Regulations also exclude the securities leg of a cash-collateralized securities lending. As noted above, the securities leg of a cash-collateralized securities lending is clearly a derivative under the statute. In the case of a U.S. securities lender and a related non-U.S. securities borrower, the securities leg involves no payments made by the U.S. securities lender to the non-U.S. securities borrower — all of the payments on the securities leg in this transaction go the other way, from the non-U.S. securities borrower to the U.S. securities lender — and is a type of transaction that Congress has explicitly sought to protect from adverse tax consequences through the enactment of Section 1058.³² Thus, we think an exclusion of the securities leg of a cash-collateralized securities lending from the scope of a “derivative” is wholly inappropriate.

Collateralized Securities Borrowing Transactions. The Proposed Regulations also exclude a securities borrowing by a U.S. taxpayer from a related non-U.S. person, including both where the securities borrowing is fully collateralized (a “collateralized securities borrowing”), which we discuss in this section, and where it is not (an “uncollateralized securities borrowing”), which we discuss in the next section. For purposes of simplicity, we refer to cash collateral throughout the section, but the discussion of the “securities leg” as described herein applies equally to collateralized securities borrowings that are not collateralized with cash.³³

Any exclusion of a securities borrowing by a U.S. taxpayer is inconsistent with the plain language used to define a “derivative” in the statute and Proposed Regulations. Both the definition of a “derivative” in the statute and the corresponding definition in the Proposed Regulations explicitly refer to a “short position.” Generally, the only short position that is relevant in the context of a related party transaction that would not be covered by one of the other explicitly enumerated types of contracts is a securities borrowing from the related party. Thus, a wholesale exclusion of securities borrowings from the definition of “derivative” effectively reads the “short position” language out of the statute.

Additionally, because the collateralized securities borrowing is fully collateralized by cash, there is no “net financing,” and, indeed, the “inbound” payments on the collateral made by the non-U.S. securities lender to the U.S. securities borrower will often be greater than the “outbound” payments made by the U.S. securities borrower to the non-U.S. securities lender with respect to the borrowed securities. Indeed, the overall result of a collateralized securities borrowing may well be incremental U.S. net income because the securities borrower may receive more in interest on the cash collateral than it pays on the securities leg of the collateralized securities borrowing. Congress clearly recognizes the need to protect collateralized securities borrowings and lendings from taxes that may be imposed on uncollateralized parallels of such transactions, as is evidenced by Sections 956(c)(2)(I) and (J). Section 956(c)(2)(I) contains an exclusion from the definition of “United States property” (and thereby prevents a Section 956 inclusion) for “deposits of cash or securities made or received on commercial terms in the ordinary course of a United States or foreign person's business as a dealer in securities or in commodities, but only to the extent such deposits are made or received as collateral or margin for . . . [among other transactions,] . . . a securities loan. . . .” By excluding collateral or margin posted for a securities loan, the statute prevents an imposition of tax on a U.S. parent of a CFC with respect to a CFC's cash-collateralized securities lending transaction. Similarly, Section 956(c)(2)(J) excludes from the definition of “United States property” an “obligation of a United States person to the extent the principal amount of the obligation does not exceed the fair market value of readily marketable securities sold or purchased pursuant to a sale and repurchase agreement or otherwise posted or received as collateral for the obligation in the ordinary course of its business by a United States or foreign person which is a dealer in securities or commodities.” This subsection protects a U.S. parent from having a Section 956 inclusion in respect of collateralized securities borrowings by its CFCs. Thus, it is evident that Congress understood that in transactions in which the amount of cash or property that enters the United States may be offset by similar amounts leaving the United States, it is inappropriate to impose tax with respect to the inbound leg.³⁴

We acknowledge that there is similarly no “net financing” in the case of a cash-collateralized securities lending by a U.S. taxpayer to a related non-U.S. person, on which the U.S. taxpayer pays interest on the cash collateral, even though, as stated above, the interest payments on the cash collateral are nonetheless base erosion payments. For reasons of simplicity, consistency or otherwise, the statute simply designates interest expense paid or accrued to a related non-U.S. person as a base erosion payment. However, the fact that Congress created a specific, targeted exception for interest in Section 59A(h)(3) does not justify the expansion of the scope of base erosion payments to include amounts, such as underlying payments on the securities leg of a collateralized securities borrowing, that are squarely within the definition of QDPs and as to which Congress did not create any exception.

Moreover, if a collateralized securities borrowing by a U.S. taxpayer from a related non-U.S. person is not treated as a derivative, with the result that the U.S. taxpayer's payments on the securities leg are treated as base erosion payments, there will be an asymmetry under Section 59A between (i) the treatment of the payments by the U.S. securities borrower to its related non-U.S. person with respect to the borrowed securities in a collateralized securities borrowing (which will not be QDPs and will therefore be base erosion payments) and (ii) the treatment of the payments made by a U.S. securities buyer to a related non-U.S. person with respect to the securities collateral under a sale-repurchase transaction vis-à-vis Section 59A (because, as explained above, the U.S. securities buyer is merely acting as an agent for its related non-U.S. person in holding such securities and making such payments, with the result that such payments are not base erosion payments). Therefore, far from treating these two transactions, under the words of the preamble to the Proposed Regulations, “similarly for purposes of section 59A(h)(4),” they would (if deductions with respect to a securities borrowing were not QDPs) be treated completely differently.

In summary, Treasury should respect the decades-long distinction in tax law between sale repurchase transactions, on the one hand, and collateralized securities borrowings, on the other. However, if Treasury chooses to focus, for purposes of Section 59A, on the economic similarity between collateralized securities borrowings and sale-repurchase transactions, it should recognize that this very similarity should lead Treasury to treat them similarly. And similar treatment is achieved by characterizing the securities borrowing leg as one that gives rise to QDPs. As discussed above, the payments on the security “sold” to the U.S. “buyer” in a sale-repurchase transaction are not treated as payments made by the U.S. “buyer” to the non-U.S. “seller” for U.S. federal income taxes, and therefore are not and cannot be base erosion payments. To ensure similar treatment of a collateralized securities borrowing transaction, where payments on the securities leg are recognized as payments by the U.S. securities borrower to its non-U.S. affiliate, Treasury would need to exclude those payments from the definition of “base erosion payment.” Treating the securities leg of a collateralized securities borrowing transaction as a “derivative” for purposes of Section 59A(h)(4) — which is consistent with the tax law generally and with the words Congress used to define “derivative” in Section 59A(h)(4) — is the most direct, and technically accurate, way to achieve this result.³⁵

Uncollateralized Securities Borrowing Transactions. We acknowledge that Treasury has expressed concerns that an uncollateralized securities borrowing is different from a fully collateralized securities borrowing, and recognize that, in the former, there has been a transfer of value (in the form of the securities) to the securities borrower that is not offset by any transfer of value (in the form of cash, securities or other collateral) to the securities lender. However, for the reasons stated below, we believe that uncollateralized securities borrowings as a general rule still constitute “derivatives” within the meaning of Section 59A(h)(4), with only a limited subcategory of uncollateralized securities borrowings being economically similar to cash borrowings.

As a starting point, we believe that an uncollateralized securities borrowing really is not a financing at all. Even if one accepts the idea that, as suggested in the language in the preamble to the Proposed Regulations quoted above, a transaction with a “significant financing component” should not be treated as a derivative for purposes of Section 59A(h), a securities borrowing lacks any such “significant financing component.” A securities borrowing generally is not indebtedness for U.S. federal income tax purposes; nor is it a financing in the generally accepted understanding of the term. It is, however, “a contract, including a short position or similar contract, the value of which, or any payment or other transfer with respect to which, is (directly or indirectly) determined by reference to” the borrowed securities. The obligations of the securities borrower with respect to the securities leg are as follows: (i) to deliver the securities back to the securities lender, which obligation varies in value based on the value of the underlying securities, (ii) to pay a borrow fee and (iii) to make substitute payments to the securities lender prior to delivery dependent on the payments made on the underlying securities. The first obligation is not in the nature of repayment of principal because the price of the securities can vary significantly in value. This is especially the case in the context of transactions involving equities and lower-grade debt securities, the value of which can be quite volatile. The latter two obligations are not based on any time-value-of-money component with respect to the securities borrower's obligation to return the securities, and thus are not in the nature of “interest.”³⁶

Even if one views (as we do not) an uncollateralized securities borrowing as a net financing, uncollateralized securities borrowings are not similar to cash borrowings as a general matter and so should not be excluded from the definition of a “derivative.” As the cornerstone of a cash borrowing is the obligation to repay principal coupled with the obligation to pay interest, a transaction lacking both of those features should not be considered economically similar to a cash borrowing.

We recognize that a limited subcategory of uncollateralized securities borrowings may be economically similar to cash borrowings. For example, an uncollateralized securities borrowing by a U.S. taxpayer from a related non-U.S. person of a high-grade debt security that is required to be returned to the lender at a date that is extremely close to the maturity date of the debt security may be sufficiently economically similar to a debt financing such that it should be excluded from the “derivative” definition. However, we believe that a targeted anti-abuse rule that excludes transactions that contain specific debt-like features that make such transactions substantially similar to debt financings can better address this narrow subset of transactions.

If Treasury opts not to take the approach outlined above with respect to uncollateralized securities borrowings and adopts a more general exclusion of uncollateralized securities borrowings from the definition of a “derivative,” we recommend that Treasury adopt certain additional rules to mitigate the potential adverse effects and uncertainties that such a rule would create:

First, Treasury should delay the effective date of this exclusion to provide financial institutions and other taxpayers with adequate time to ensure that their systems that track collateralization properly distinguish between collateralized and uncollateralized securities borrowings (as we propose that each be defined below). To the extent that there are any discrepancies between how collateralization is defined for regulatory purposes and how collateralization is defined for purposes of Section 59A, financial institutions would need to build additional systems that account for those differences. As collateralization is often accomplished by umbrella netting arrangements that involve netting various offsetting liabilities, and various complex custodial and margin arrangements are often employed, revising existing systems so that the systems adequately track the extent to which each securities borrowing is collateralized for tax purposes may entail a high degree of complexity. For these reasons, we recommend that any general exclusion of uncollateralized securities borrowings from the definition of a “derivative” should not be effective for any taxable year beginning prior to January 1, 2020.

Second, similar to the rules under Section 956 and the approach Treasury took in the temporary Treasury Regulations for identifying when a notional principal contract was fully collateralized in creating an exception from the bifurcation rule for notional principal contracts with nonperiodic payments, Treasury should provide that a securities borrowing by a U.S. taxpayer from a related non-U.S. person should be treated as fully collateralized for this purpose if the securities borrowing is required to be fully collateralized by a clearance organization or clearing agency, by a federal regulator or pursuant to the terms of the contract.³⁷

Third, because of certain regulatory and other requirements governing collateral posting, we believe that (i) an uncollateralized securities borrowing should be defined for purposes of such an exclusion as including only a securities borrowing where the aggregate value of (A) collateral posted (in the form of cash, readily marketable securities, or (if the securities borrower is able to establish the value of the relevant securities) other securities), (B) offsetting liabilities of the securities lender³⁸ and (C) other amounts that effectively serve as collateral due to the securities borrower's compliance with any specific regulatory regime governing securities borrowing or any other arrangement that has a similar economic effect³⁹ (collectively, “Collateral”), is significantly less than (perhaps 70% or less of) the value of the borrowed securities,⁴⁰ and (ii) a securities borrowing that does not meet the requirement in clause (i) above should be treated as an uncollateralized securities borrowing only to the extent that it is uncollateralized (which would mean for this purpose the extent to which the aggregate value of the Collateral is less than the aggregate value of the borrowed securities), because a securities borrowing is a net financing transaction only to the extent it is uncollateralized.

We also note as a separate but related matter that the proposal to treat losses on transfers of property to related non-U.S. persons as base erosion payments⁴¹ (which we address separately below in Recommendation 3) may significantly worsen the effects of generally excluding sale-repurchase transactions and securities lending and borrowing transactions from the definition of the term “derivative” for purposes of the QDP exception. First, this proposal would likely ensure that a loss incurred by a taxpayer that borrows securities from a related non-U.S. person at the termination of the borrowing (e.g., upon returning securities worth more than what they were worth when they were borrowed) would be considered a base erosion payment. Second, a taxpayer that lends securities to a related non-U.S. person would have similar consequences, potentially both upon the entry into and upon the termination of the lending, if Section 1058 does not apply to the lending. Thus, the rule treating losses on transfers of property to related non-U.S. persons as base erosion payments likely exacerbates the effects of not treating sale-repurchase transactions and securities lending and borrowing transactions as “derivatives” under the QDP rules.

To be clear, eliminating the rule treating losses on transfers of property to related non-U.S. persons as base erosion payments would not ensure that a loss incurred by a U.S. taxpayer at the close of a securities lending or borrowing transaction with a related non-U.S. person is not a base erosion payment, because the loss could nonetheless be considered to be “attributable to” a “payment to” the related person. We therefore believe that eliminating the securities lending exception from the definition of “derivative” is necessary in order to achieve the appropriate treatment of collateralized securities lendings under Section 59A, regardless of whether the rule treating losses on transfers of property to related non-U.S. persons as base erosion payments is eliminated.

The expanded definition of a “base erosion payment” as described in the preamble also captures sales or exchanges of property by a domestic person to a related non-U.S. person on which the domestic person recognizes a loss with respect to the property. This rule is not grounded in the language of either the statute or the Proposed Regulations. The statute and Proposed Regulations provide that a base erosion payment is an amount that is “paid or accrued.”⁴² Additionally, the relevant allowable deduction must be “with respect to” the amount paid or accrued.⁴³ Thus, in order for a recognized loss on a transfer of property to a related non-U.S. person to be considered a base erosion payment, it must be the case that the transfer of property is a payment or accrual, and also that the loss is attributable to the payment or accrual (i.e., to the “amount” or “value” of the property transferred). As we discuss below, (i) only for a limited set of transfers of property to a related non-U.S. person can the transfer be properly considered as an “amount paid or accrued” to the related non-U.S. person (and a sale of property for cash should never be considered an amount paid or accrued), and (ii) in all cases a loss recognized by a taxpayer on the sale or exchange of property to a related non-U.S. person cannot be considered “attributable to” the payment or accrual, even if such payment or accrual exists.⁴⁴

First, a loss on a sale of property to a related person is not an accrual, and in many cases is not a payment. “Accrual” is a technical term pertaining to a method of accounting used by the taxpayer to determine timing of recognition. Generally, cash-method taxpayers do not accrue items of expense until they are paid, with certain statutory exceptions such as the original issue discount rules. Under Section 461 and the general rules for accrual-method taxpayers, those taxpayers generally may not deduct an item of expense prior to the time that all events have occurred fixing the timing and amount of the item, and, in some cases, “economic performance” has occurred. Thus, the concept of accrual relates to the deduction of an item prior to the time it is actually paid, which is a concept that has no relevance or connection to a sale of property — before the relevant transaction in which property is sold or otherwise transferred occurs, the loss from the sale or transfer of the property cannot accrue. As the sale or transfer itself is the event through which the loss is realized, the term “accrual” does not apply meaningfully to a loss from a sale or exchange of property.

Similarly, in many instances — and most importantly, in the case of a sale of property for cash — a transfer of property to a related person in respect of which a loss is recognized is not a “payment,” based on the plain meaning of that word. The Merriam Webster dictionary defines “pay” as (1) “to make due return to for services rendered or property delivered,” (2) “to engage for money,” (3) “to give in return for goods or service,” (4) “to discharge indebtedness for,” (5) “to make a disposal or transfer of (money),” (6) “to give or forfeit in expiation or retribution,” (6) “to make compensation for,” (7) “to requite according to what is deserved,” (8) “to give, offer, or make freely or as fitting,” (9) “to return value or profit to,” or (10) “to bring in as a return.” Although the definition is broad, and would likely cover certain transfers of property (such as an “in-kind” transfer of property in payment for goods or services received, or a transfer of property in satisfaction of a liability),⁴⁵ it does not cover all transfers of property, in particular sales of property in exchange for cash. From the perspective of the property transferor, a sale of property for cash is the opposite of making a payment. The transferor receives a payment in exchange for the transferred property. Under general U.S. tax principles, a taxpayer does not “pay” for U.S. dollars, and thus cannot be said to be making a “payment” of property in exchange for U.S. dollars.

Second, a loss with respect to a sale or exchange of property to a related person is not “attributable” to the “payment” (even if it is a payment, which again we think it clearly is not in many cases), i.e., the amount or value of the property transferred. Rather, the loss is attributable to the transferor's basis in excess of the value of the transferred property — or put another way, the loss is attributable to a difference (including, typically, changes in the difference) between basis and value that arises before the transfer and, more importantly, independent of the transfer; the transfer does not create the loss, as a legal or economic matter, but rather is merely the event that causes the latent loss to be realized, as a matter of timing, for U.S. federal income tax purposes. Thus, even if the transfer of property is treated as an amount paid (however quantified), a loss on the sale or exchange of that property that results from the transfer to a related person is not attributable to that amount paid. A simple observation to make this point clear is that the taxpayer could instead have sold the property to a third party and realized the same loss, and then (although this would be rather unnecessary) delivered the resulting cash to the related non-U.S. person for an equal amount, and had no base erosion payment at all.⁴⁶

Recommendation 4(a): With respect to the reporting rules for QDPs, Treasury should make explicit that a failure to comply with the Form 8991 reporting requirements by a taxpayer that is not a reporting corporation (within the meaning of Section 1.6038A-1(c)) does not affect whether or not any payments made by the taxpayer are considered QDPs and should clarify the consequences of failing to comply with the Form 8991 QDP reporting requirements for the taxpayer.

Proposed Section 1.59A-6(b)(2) provides that, “[n]o payment is a qualified derivative payment . . . for any taxable year unless the taxpayer reports the information required in Section 1.6038A-2(b)(7)(ix) for the taxable year” (the “QDP Reporting Condition”). However, Proposed Section 1.6038A-2(g) provides that before the full reporting rules under Proposed Section 1.6038A-2(b)(7)(ix) become effective, “a taxpayer will be treated as satisfying the reporting requirement described in Section 1.59A-6(b)(2) only to the extent that it reports the aggregate amount of qualified derivative payments on Form 8991.” The reporting requirements in Proposed Section 1.6038A-2(g) apply only to an applicable taxpayer that is a “reporting corporation,” which is generally defined in existing Section 1.6038A-1(c) as either (x) a domestic corporation that is 25% non-U.S.-owned or (y) a non-U.S. corporation that is engaged in a trade or business in the United States. Thus, until the final reporting regulations are effective, in order to satisfy the QDP Reporting Condition, an applicable taxpayer that is a reporting corporation must satisfy the reporting requirements under Proposed Section 1.6038A-2(g).

As discussed above, the reporting requirements in Proposed Section 1.6038A-2(g) apply only to applicable taxpayers that are also reporting corporations. Therefore, it appears that the QDP Reporting Condition does not apply to any taxpayer that is not a reporting corporation, because Proposed Section 1.6038A-2(g) does not impose reporting requirements on such a taxpayer. Treasury should eliminate any ambiguity with respect to this point by explicitly providing that the QDP Reporting Condition applies only to applicable taxpayers that are reporting corporations.

Each applicable taxpayer, regardless of whether it is a reporting corporation, is required to report on Form 8991 its aggregate amount of QDPs for the taxable year. As the preamble to the final regulations makes clear, “[w]hile an applicable taxpayer that is not a reporting corporation would not be subject to monetary penalties and collateral provisions specific to sections 6038A and 6038C, the taxpayer remains subject to BEAT-related reporting obligations, including Form 8991, and applicable consequences for noncompliance.”⁴⁷ However, it is unclear what the consequences of noncompliance with Form 8991 reporting are, and Treasury should therefore clarify the relevant consequences.

Recommendation 4(b): Until the final reporting regulations are effective, all taxpayers should be allowed to report the aggregate amount of their QDPs under a good faith standard.

Prior to the time the final reporting regulations are effective, the consequences to a taxpayer (if any) of incorrectly reporting the aggregate QDP amount should apply only if the taxpayer does not meet a good faith standard with respect to its Form 8991 QDP reporting. Because of the uncertainties associated with the guidance currently provided in the Proposed Regulations (i.e., the uncertainties discussed herein with respect to the Section 988 Loss exception, the Net Mark-to-Market Method, and the repo/securities lending exception), a number of questions arise concerning how taxpayers should calculate the aggregate QDP amount. In particular, for example, many financial institutions do not currently separately track payments with respect to securities lending transactions (e.g., losses resulting from closing out securities borrowings, described above in Part II, Recommendations 2 and 3). Even if Treasury accepts our recommendation that securities lending and borrowing transactions should not be excluded from the definition of a “derivative,” taxpayers will need significant systems build in order to capture and segregate payments arising from securities lending and borrowing transactions in order to report them as QDPs. Additionally, there is no transition period during which taxpayers can build systems appropriately to calculate the aggregate QDP amount. Therefore, a good faith standard for the interim period would be appropriate.

Recommendation 4(c): Treasury should clarify that the relevant payments for each derivative transaction should be netted, consistent with the Net Mark-to-Market Method, to arrive at the aggregate QDP amount that must be reported on Form 8991.

In order properly to report the aggregate amount of QDPs on Form 8991, each applicable taxpayer is required to determine its total QDPs with respect to each of its positions and then combine all such totals in order to arrive at the amount it is required to report. In order to determine the total QDPs with respect to a single position, we assume that payments should be netted, consistent with the Net Mark-to-Market Method. In order for taxpayers to have certainty as to the required method of reporting on Form 8991, and because of the uncertain consequences of a reporting failure, we believe this conclusion should be made explicit in the regulations.

Recommendation 4(d): Treasury should make clear that certain QDPs need not be reported.

Even if, as we recommend in Recommendation 2 above, payments with respect to the securities legs of securities lending and securities borrowing transactions are eligible to be treated as QDPs, it will be difficult or impossible for financial institution taxpayers to capture and quantify certain amounts that might potentially be eligible for this exception. In particular, the amounts of losses relating to “closing” securities borrowings — especially in the case of “matched book” businesses — are not separately accounted for by many financial institutions, and these amounts can be arbitrarily large. The burden of reporting these amounts seems unnecessary, as these amounts (which are often not accounted for under GAAP) are not accounted for in either the numerator or the denominator and thus their actual amounts are irrelevant, so long as it can be shown that they are in fact QDPs. We think it should be made clear that any QDPs with respect to securities lending/borrowing and sale repurchase transactions that are not accounted for under GAAP need not be accounted for in QDP reporting, provided it is established that these amounts are otherwise eligible for QDP treatment, and that they have been disregarded both for numerator and denominator purposes.

Recommendation 5: Treasury should clarify that transfers made pursuant to revenue or profit sharing arrangements in connection with global dealing and similar arrangements are not base erosion payments and are excluded from the numerator of the base erosion percentage.

As we have previously discussed with Treasury, we believe base erosion payments should not include amounts transferred between related parties (on the basis that the transferee, rather than the transferor, is considered the tax owner of such amounts in the first place) pursuant to revenue or profit sharing arrangements with respect to businesses that constitute “global dealing operations” described in proposed global dealing regulations⁴⁸ and other similarly highly integrated businesses of financial institutions. The integrated structure of financial institutions is primarily driven by regulatory and other legal considerations that often require particular entities to face customers, manage risk or engage in other activities. We understand that Treasury has not established any specific rules for purposes of Section 59A with regard to identifying the beneficial owner of income and related tax consequences.⁴⁹ However, should Treasury wish to consider addressing the treatment, for purposes of Section 59A, of amounts transferred in respect of the results of a globally integrated business pursuant to a revenue or profit sharing arrangement, we would readily provide further information for Treasury's consideration. 

We are aware that IIB has provided or will provide Treasury, in its comment letter on the Proposed Regulations, with a recommendation that Treasury clarify that the proposed global dealing regulations reflect the view that income allocated under a revenue or profit sharing methodology permitted by those regulations is treated as owned by the group member to which it is allocated. For the reasons stated by IIB, we support the request of IIB for Treasury to make this clarification.

Recommendation 6: The Net Mark-to-Market Method should be included as a safe harbor but should not be the exclusive method under the Proposed Regulations for calculating payments and accruals with respect to a position that is marked to marked.

SIFMA appreciates that Treasury indicated that the Net Mark-to-Market Method is an acceptable method for calculating payments and accruals with respect to a derivative that is marked to market, but believes it should not be required. It should instead be included as a safe harbor rather than as the exclusive method for calculating payments and accruals on a derivative that is marked to market.

There are several reasons why this method should not be mandatory. First, the Net Mark-to-Market Method is inconsistent with the statutory definition of the denominator as “the aggregate amount of the deductions . . . allowable to the taxpayer under this chapter for the taxable year” plus certain other base erosion tax benefits.⁵⁰ The base erosion percentage is determined by reference to gross deductions and other tax benefits, and not by reference to net amounts, either generally or with respect to specific transactions. Each gross item of deduction or loss with respect to a position that is marked to market reflects a real economic gross loss in the relevant taxable year, and for a taxpayer subject to a mark-to-market method of accounting, there generally is no timing mismatch between when a loss arises as an economic matter and when such loss is deductible. Requiring the taxpayer to offset deductions and losses with income and gain thus is inconsistent with the general approach of Section 59A and is an inappropriate exercise of Treasury's regulatory authority under Section 59A(i).⁵¹

Second, mandating a single method of calculating deductions is contrary to Treasury's stated general approach of relying on existing tax law for determining “whether a payment is treated as a deductible payment, or, when viewed as part of a series of transactions, should be characterized in a different manner,”⁵² including whether such payment is “other than deductible, such as an amount that reduces gross income,”⁵³ rather than providing for specific rules or modifying general methods of accounting that apply only for purposes of Section 59A. In addition, because QDPs are excluded from the numerator, the primary application of the Net Mark-to-Market Method is to understate the amount of deductions includible in the denominator. For reasons of efficiency and because, as we have previously discussed with Treasury, the use of the Net Mark-to-Market Method would only have the result, at worst, of understating a taxpayer's gross deductions for purposes of the denominator of the base erosion percentage, the Net Mark-to-Market Method should be a safe harbor available to taxpayers with respect to derivatives that are marked to market under Section 475, but should not be the exclusive method available.

* * *

III. GENERAL RECOMMENDATIONS

Recommendation 7: Treasury should revise the definition of a “base erosion payment” to include an exception for an acquisition of depreciable property from a related non-U.S. person in a nonrecognition transaction.

The preamble to the Proposed Regulations indicates that Proposed Section 1.59A-3 was purposefully drafted to treat certain nonrecognition transactions in which basis is imported as base erosion payments.⁵⁴ Proposed Section 1.59A-3(b)(2)(i) provides that for purposes of determining what payments are considered “base erosion payments,” an amount paid or accrued includes non-cash consideration, such as property, stock or the assumption of a liability. The preamble makes clear that this is intended to bring nonrecognition provision transactions within the scope of the definition of a “base erosion payment.”⁵⁵ The preamble further notes that the Proposed Regulations “do not include any specific exceptions for these types of transactions even though (a) the transferor of the assets acquired by the domestic corporation may not recognize gain or loss, (b) the acquiring domestic corporation may take a carryover basis in the depreciable or amortizable assets, and (c) the importation of depreciable or amortizable assets into the United States in these transactions may increase the regular income tax base as compared to the non-importation of those assets.”⁵⁶

This expanded definition of a “base erosion payment” captures payments that are clearly not eroding the tax base, and appears overreaching in its scope.⁵⁷ For example, when depreciable property is transferred to a domestic corporation by a related non-U.S. person in exchange for the domestic corporation's stock, (1) there is clearly base accretion rather than base erosion, and (2) in the case of a nonrecognition transaction, the amount paid to the related non-U.S. person (the value of the taxpayer's stock) has no connection to the amount of depreciation deductions that can be taken with respect to the property acquired (which is based on the carryover basis of the property). Moreover, in the case of an “inbound” liquidation of a non-U.S. corporate subsidiary of a domestic corporation that qualifies for nonrecognition treatment under Section 332, it is inappropriate to treat the cancellation of the stock of the non-U.S. corporate subsidiary that occurs by reason of the liquidation as “an amount paid or accrued” to a related non-U.S. person, if only for the reason that the related non-U.S. person never actually receives anything in the transaction — both it and the stock owned by the domestic corporation are eliminated in the transaction.

Recommendation 8: Treasury should (a) increase the threshold percentage interest in a partnership at or above which a partner is required to take into account its distributive share of any partnership amount of base erosion tax benefits from 10% to 25%, (b) eliminate the requirement that a partner's interest in the partnership have a fair market value of less than $25 million in order for a partner not to have to take into account its distributive share of any partnership amount of base erosion tax benefits, and (c) provide that the minimum percentage interest requirement also applies with respect to treating a payment made to a partnership as being made to each partner based on the partner's distributive share.

Proposed Section 1.59A-7(b)(4) provides that a partner must take into account its distributive share of any partnership amount of base erosion tax benefits unless both (i) the partner's interest in the partnership has a fair market value of less than $25 million on the last day of the partner's taxable year (the “$25 Million Threshold”), and (ii) the partner owns less than 10% of the capital and profits of the partnership and less than 10% of each partnership item of income, gain, loss, deduction, and credit for the taxable year (the “10% Threshold”). The preamble explains that Treasury “determined that a threshold of ten percent appropriately balanced the administrative burdens of determining whether deductions allocated to a partner with a small ownership interest in a partnership are base erosion payments with the Treasury Department and IRS's interest in maintaining a consistent aggregate approach to partnerships in applying to the BEAT.” We believe that Treasury should reconsider the $25 Million Threshold and 10% Threshold.

Treasury should substitute the 10% Threshold with a 25% ownership interest threshold (based solely on the partner's percentage interest in the capital or profits of the partnership) and eliminate the $25 Million Threshold due to the compliance burden imposed compared to the anticipated impact. The $25 Million Threshold and 10% Threshold would impose a very substantial compliance burden on an owner of a minority interest in a partnership — who may have very little or no ability to obtain the necessary information from the partnership, especially non-U.S. partnerships which may not maintain U.S. tax books — in circumstances where the impact of including the owner's share of partnership items may have very little actual impact on the owner's Section 59A calculations or liability. Unless a taxpayer owns a sufficiently significant interest in a partnership (so that, for example, it has sufficient negotiating power to require the partnership to put in place the systems needed in order to provide to the taxpayer the information needed for BEAT purposes, which systems the partnership would otherwise be unlikely to put in place), it is very likely that the taxpayer will be unable to obtain the necessary information from the partnership, irrespective of the value of the partner's interest in the partnership. We therefore recommend that Treasury eliminate any value threshold and provide that the compliance burden only apply when a partner owns a 25% interest in a partnership.

Additionally, while the Proposed Regulations provide that the 10% Threshold and $25 Million Threshold apply with respect to the rule treating payments made by a partnership as made by each partner based on the partner's allocable share,⁵⁸ there is no comparable exception to the rule governing payments made to a partnership.⁵⁹ Treasury should thus provide that the minimum 25% percentage interest requirement described above also applies with respect to treating a payment made to a partnership as being made to each partner.

Recommendation 9: For purposes of determining modified taxable income, the taxpayer's taxable income should be determined by taking into account the taxpayer's entire net operating loss deduction allowed under Section 172 (and thus may be an amount less than zero), rather than only the amount of the deduction that does not reduce the taxpayer's taxable income to an amount less than zero.

Under the Proposed Regulations, to determine a taxpayer's modified taxable income, a taxpayer has negative taxable income for the taxable year if a taxpayer has an excess of deductions over gross income, without regard to any net operating loss deduction under Section 172 (an “NOL Deduction”). That negative amount is the starting point for determining the taxpayer's modified taxable income. However, the Proposed Regulations provide that if any NOL Deduction exceeds the amount of positive taxable income before taking into account the NOL Deduction, the excess amount of the NOL Deduction does not reduce taxable income below zero. This appears to alter the operation of existing law.

NOL Deductions should be taken into account for purposes of determining modified taxable income to the same extent that they are taken into account in determining taxable income for general federal income tax purposes. Old Section 172(a) simply provides that the net operating loss deduction is “the aggregate of (1) the net operating loss carryovers to such year, plus (2) the net operating loss carrybacks to such year,” without limiting it to the amount of taxable income for that taxable year. Although current Section 172 limits the net operating loss deduction to 80% of taxable income from that year, that limitation is only effective for losses arising in taxable years beginning after December 31, 2017.

Furthermore, there is no authority in Section 59A for this result. Section 59A(i) provides for regulations as necessary or appropriate to carry out the provisions of the Section. Section 59A(c)(1) defines modified taxable income as “the taxable income of the taxpayer computed under this chapter for the taxable year,” with specified adjustments. It is clear that old Section 172, which is part of Chapter 1, can produce negative income. Limiting the operation of Section 172 in determining taxable income for purposes of the modified taxable income calculation under Section 59A(c) is neither “necessary nor appropriate to carry out the provisions” of that Section. The statutory language clearly provides that the starting point of the calculation is “the taxable income of the taxpayer computed under this chapter for the taxable year” and then enumerates the specific adjustments to be made in arriving at modified taxable income. Accordingly, net operating losses should be taken into account for purposes of determining modified taxable income to the same extent that they are taken into account in determining taxable income for general federal income tax purposes.

We acknowledge that allowing net operating losses to be taken into account for purposes of the modified taxable income calculation to the same extent that they are taken into account in the general taxable income calculation would give rise to certain complex questions concerning NOL carryforward calculations for general federal income tax purposes. Should Treasury wish to explore this recommendation and its collateral implications further, a subset of the SIFMA members involved in preparing these comments would welcome the opportunity to provide further thoughts on these issues.

We are also aware that IIB has provided or will provide Treasury, in its comment letter on the Proposed Regulations, with a recommendation that Treasury reconsider its approach to the NOL Deduction, together with a detailed analysis in support of the recommendation. For the reasons stated by IIB, we support the request of IIB for Treasury to reconsider its approach to the NOL Deduction.

* * *

IV. ADDITIONAL RECOMMENDATIONS THAT SIFMA SUPPORTS

A. TLAC and Other Regulatory Debt

The Financial Stability Board (“FSB”) promulgates international principles and standards for total-loss absorbing capacity (“TLAC”) rules. The TLAC concept as articulated by the FSB is that if losses arise at a bank or other regulated subsidiary, those losses are borne first internally by the affiliate holding debt or equity instruments that qualify as “internal TLAC,” and then externally by third-party investors holding instruments that qualify as “external TLAC.” Each country has, or will have, adopted its own national standards for implementing these rules, which differ in various regards notwithstanding that they are all based on the FSB principles. In the United States, the Federal Reserve promulgate rules that, among other things, govern the amount of internal TLAC securities a financial institution needs to have.

Proposed Section 1.59A-3(b)(3)(v)(A) provides that a portion of the interest paid or accrued on TLAC securities that certain banking organizations are required by the Federal Reserve to issue are not treated as base erosion payments. The amount of interest paid or accrued on TLAC securities that is excluded from treatment as a base erosion payment is limited to the product of the amount paid or accrued to related non-U.S. persons with respect to the TLAC securities and the “scaling ratio.”¹ The scaling ratio generally equals the ratio of the minimum amount of debt that is required under the TLAC regulations to the sum of the adjusted issue prices of all TLAC securities issued and outstanding by the taxpayer.² The preamble explains that “because of the special status of TLAC as part of a global system to address bank solvency and the precise limits that Board regulations place on the terms of TLAC securities and structure of intragroup TLAC funding, it is necessary and appropriate to include an exception to base erosion payment status for interest paid or accrued on TLAC securities required by the Federal Reserve.”³

We are aware that IIB has provided or will provide Treasury, in its comment letter on the Proposed Regulations, with a series recommendations together with detailed analysis concerning an expansion of the TLAC exception. The IIB recommendations include that (i) the “specified minimum amount” of internal TLAC in the form of long-term debt that is eligible for relief from base erosion payment treatment include a maintenance buffer of 1% of risk-weighted assets or 0.50% of total leverage exposure, as appropriate; (ii) the relief for interest paid or accrued on internal TLAC be extended (a) to 2018, and (b) to future transition years (with the relief phased in ratably over the transition period); (iii) the regulations be revised to state that relief is provided for internal TLAC in the form of long-term debt held by a related non-U.S. person up to the specified minimum amount; (iv) the regulations provide relief for intercompany debt that is structured and issued to comply with other U.S. or non-U.S. regulatory or supervisory requirements in addition to those that apply to internal TLAC; and (v) the same principles that apply to intercompany debt incurred by the U.S. operations of a non-U.S. bank apply to intercompany debt incurred at the non-U.S. bank level. For the reasons stated by IIB, we support the proposal of IIB for Treasury to adopt these recommendations.

B. Section 15

Section 59A(b)(1) provides that, with certain exceptions, the “base erosion minimum tax amount” with respect to an applicable taxpayer for any taxable year is the excess of (A) an amount equal to “10 percent (5 percent in the case of taxable years beginning in calendar year 2018)” (the “BEAT Tax Rate”) of the modified taxable income of such taxpayer for the taxable year over (B) the regular tax liability of the taxpayer for the taxable year, with certain adjustments.⁴ With respect to the taxable year “beginning in calendar year 2018” of an applicable taxpayer with a fiscal, rather than a calendar, taxable year, Proposed Section 1.59A-5 takes the position that Section 15 applies with the result that the applicable taxpayer will have a blended BEAT Tax Rate based on the number of days in the taxable year before and after the effective date of the change in rate.

We are aware that IIB has provided or will provide Treasury, in its comment letter on the Proposed Regulations, with a recommendation that Treasury revise Proposed Section 1.59A-5(c)(1)(ii) and (3) to provide that a fiscal year taxpayer is subject to the 5 percent BEAT Tax Rate for the entirety of its fiscal year beginning in calendar year 2018, together with a detailed analysis in support of the recommendation. For the reasons stated by IIB, we support the request of IIB for Treasury to make this revision.

* * *

We appreciate your consideration of our views and concerns, and we would appreciate the opportunity to discuss further the issues in this submission with you and your colleagues. We will be following up with you to schedule a meeting at your convenience.

Please do not hesitate to contact me at (202) 962-7300 or ppeabody@sifma.org, or our outside counsel Michael Farber and Michael Mollerus at Davis Polk & Wardwell LLP. Michael Farber can be reached at (212) 450-4704 or mfarber@dpw.com, and Michael Mollerus can be reached at (212) 450-4471 or mollerus@dpw.com.

Respectfully submitted,

Payson R. Peabody, Esq.
Managing Director & Tax Counsel
SIFMA
New York, NY and Washington, DC

cc:

David J. Kautter
Assistant Secretary for Tax Policy
Department of the Treasury

L.G. “Chip” Harter
Deputy Assistant Secretary (International Tax Affairs)
Department of the Treasury

Douglas Poms
International Tax Counsel
Department of the Treasury

Kevin Nichols
Attorney Advisor
Department of the Treasury

William M. Paul
Acting Chief Counsel and Deputy Chief Counsel (Technical)
Internal Revenue Service

Margaret O'Connor
Acting Associate Chief Counsel (International)
Internal Revenue Service

Sheila Ramaswamy
Attorney-Advisor
Internal Revenue Service

Karen Walny
Attorney
Internal Revenue Service

Julie Wang
Attorney
Internal Revenue Service

John P. Stemwedel
Attorney
Internal Revenue Service