Menu
Tax Notes logo

NYSBA Report Focuses on LOB Provisions of Draft Model Treaty

NOV. 16, 2015

NYSBA Report Focuses on LOB Provisions of Draft Model Treaty

DATED NOV. 16, 2015
DOCUMENT ATTRIBUTES

 

November 16, 2015

 

 

The Honorable Mark Mazur

 

Assistant Secretary

 

(Tax Policy)

 

Department of the Treasury

 

1500 Pennsylvania Avenue, NW

 

Washington, DC 20220

 

 

The Honorable John Koskinen

 

Commissioner

 

Internal Revenue Service

 

1111 Constitution Avenue, NW

 

Washington, DC 20224

 

 

The Honorable William J. Wilkins

 

Chief Counsel

 

Internal Revenue Service

 

1111 Constitution Avenue, NW

 

Washington, DC 20224

 

Re: Report No. 1331 Relating to the Proposed Revisions to the Limitation on Benefits Article of the U.S. Model Tax Convention

 

Dear Messrs. Mazur, Koskinen and Wilkins:

I am pleased to submit the attached report of the Tax Section of the New York State Bar Association (the "Report"). The Report provides comments on the proposed revisions to the Limitation on Benefits article ("LOB Article") of the U.S. Model Tax Convention (the "Proposed Model").

Generally, the LOB Article provides that a resident of a Contracting State will be entitled to treaty benefits only if it is a "qualified person" or satisfies an active business or derivative benefits test. The changes in the Proposed Model, apart from the addition of the derivative benefits provision, restrict entitlement to treaty benefits as compared with the 2006 Model Treaty. We believe that many of the proposed changes go too far in the direction of restricting access to treaties. In our experience, many taxpayers that are not engaged in treaty shopping or other treaty abuse are prevented from qualifying for treaty benefits even under the LOB articles of existing treaties, and we are concerned that the changes in the Proposed Model would, if incorporated into a tax treaty, further exacerbate the existing problem. Accordingly, the Report recommends that the Department of the Treasury ("Treasury") narrow the proposals to address clearly articulated cases of treaty shopping or other treaty abuse, but not deny treaty benefits to companies that employ common non-tax motivated structures. We also make a number of other technical comments.

Our specific recommendations are as follows:

 

1. We recommend that the base erosion test not be added to paragraph 2(d) of the Proposed Model. If, contrary to our recommendation, the base erosion test of paragraph 2(d) is retained, we believe that Treasury should consider not applying it to companies that are wholly or majority owned by a single public company parent.

2. We recommend that the base erosion test not apply to payments of interest to unrelated persons on borrowings that arise from ordinary course capital markets transactions. We further recommend that the exclusion for such interest not be limited to payments to "banks."

3. The Proposed Model treats a payment, even to a qualified person, as a base-eroding payment if the recipient benefits from a "special tax regime" ("STR"). We recommend that if the STR rule in the LOB Article is retained, it should be limited to payments to related persons.

4. We recommend that the exclusion from gross income for "effectively exempt" dividends be removed from the Proposed Model.

5. Proposed paragraph 6(f) defines the term "qualified intermediate owner" (hereafter, "QIO") as an intermediate owner that is a resident of a state that has in effect with the source state a comprehensive tax treaty that includes STR language. We recommend that the STR requirement for QIOs should be eliminated from the Proposed Model at this time, because no existing treaties contain the STR rule.

6. In the context of the derivative benefits provision, we recommend that the "cliff rule" approach to the "as low as" requirement be eliminated in favor of a rule that would apply a blended withholding tax rate.

7. We suggest that Treasury consider not placing a limit on the number of equivalent beneficiaries a company can have, or that it consider increasing the number of equivalent beneficiaries that a tested company can have.

8. With a narrow exception targeted to abuse, we do not support the addition of an "at least as favorable" test where an equivalent beneficiary or QIO seeks to obtain treaty benefits described in Articles 7, 13 and 21. We believe that such a test is too subjective to be applied consistently.

9. We suggest that Treasury consider simplifying and streamlining the various intermediate owner tests set forth in the Model. We question whether the various restrictions upon who can qualify as an intermediate owner or QIO are necessary. In particular, we believe that a subsidiary of a public company that otherwise is a qualified person under paragraph 2(d) should qualify as a "good" intermediate owner. Moreover, we do not think that QIOs should be required to be residents of states, the tax treaties of which confer benefits at least as favorable as those under the treaty being tested.

10. We recommend that an equivalent beneficiary should include a resident of either of the two Contracting States, not only a resident of the State in which the tested company resides.

11. We recommend that the proposed change made to paragraph 3, which would limit the attribution of activities from a connected person to only those cases in which both persons are engaged in the same or complementary lines of business, be eliminated from the Proposed Model.

12. We recommend that the proposed "nexus" condition be removed from paragraph 5 of the Proposed Model, and should be discussed in a Technical Explanation as a relevant factor only.

13. The change to paragraph 1 of the Proposed Model LOB Article, requiring a resident be a qualified person "at the time" when the treaty benefit would be accorded, is unclear. We suggest that a Technical Explanation provide guidance as to the application of the various timing rules in the Proposed Model.

 

We appreciate your consideration of our recommendations. If you have any questions or comments regarding this report, please feel free to contact us and we will be glad to discuss or assist in any way.
Respectfully submitted,

 

 

David R. Sicular

 

Chair

 

New York State Bar Association

 

Tax Section

 

Enclosure

 

 

CCs:

 

Erik H. Corwin

 

Deputy Chief Counsel (Technical)

 

Internal Revenue Service

 

 

Quyen Huynh

 

Associate International Tax Counsel

 

Department of the Treasury

 

 

Henry Louie

 

Deputy International Tax Counsel

 

Department of the Treasury

 

 

Emily S. McMahon

 

Deputy Assistant Secretary (Tax Policy)

 

Department of the Treasury

 

 

Steven Musher

 

Associate Chief Counsel (International)

 

Internal Revenue Service

 

 

Danielle Rolfes

 

International Tax Counsel

 

Department of the Treasury

 

 

Marjorie Rollinson

 

Deputy Associate Chief Counsel

 

(International-Technical)

 

Internal Revenue Service

 

 

Robert Stack

 

Deputy Assistant Secretary

 

(International Tax Affairs)

 

Department of the Treasury

 

 

Thomas C. West, Jr.

 

Tax Legislative Counsel

 

Department of the Treasury

 

* * * * *

 

 

Report No. 1331

 

 

NEW YORK STATE BAR ASSOCIATION

 

 

TAX SECTION

 

 

PROPOSED REVISIONS TO THE LOB ARTICLE OF THE

 

U.S. MODEL TAX CONVENTION

 

 

November 16, 2015

 

 

                           TABLE OF CONTENTS

 

 

 I.   INTRODUCTION

 

 

 II.  SUMMARY OF THE PROPOSED REVISIONS TO THE LOB ARTICLE

 

 

      A. General Architecture

 

 

      B. Specific Changes

 

 

 III. SUMMARY OF THE REPORT'S RECOMMENDATIONS

 

 

 IV.  DISCUSSION

 

 

      A. In General

 

 

      B. The LOB, Treaty Abuse and Double Non-Taxation

 

 

      C. Paragraph 2(d): Public Company Subsidiaries

 

 

         1. In General

 

 

         2. Interest Payments

 

 

         3. STR Payments

 

 

      D. Definitional Issues Related to the Base Erosion Test

 

 

         1. "Effectively Exempt" Dividends

 

 

         2. Tested Groups

 

 

      E. Derivative Benefits -- Paragraph 4

 

 

         1. The STR Requirement

 

 

         2. The Cliff Rule

 

 

         3. Other Issues

 

 

      F. Intermediate Owners and Related Issues

 

 

      G. Active Trade or Business Test

 

 

      H. Discretionary (Competent Authority) Relief

 

 

      I. Time of Qualification

 

REPORT ON PROPOSED REVISIONS TO THE LOB ARTICLE OF THE

 

U.S. MODEL TAX CONVENTION

 

 

I. INTRODUCTION

This report1 of the Tax Section of the New York State Bar Association sets forth our comments on the recently proposed revisions to the Limitation on Benefits article (the "LOB Article") of the U.S. Model Tax Convention (the "Proposed Model"). We previously submitted an earlier report discussing new proposed rules applicable to "special tax regimes" ("STRs") and to subsequent changes in law.2

The report is divided into four parts. Part I is this Introduction. Part II summarizes the proposed revisions to the LOB Article. Part III is a summary of our recommendations. Part IV is a discussion of the issues and our recommendations.

II. SUMMARY OF THE PROPOSED REVISIONS TO THE LOB ARTICLE

 

A. General Architecture

 

The general architecture of the LOB Article in the Proposed Model is similar to that of the existing 2006 Model Treaty (the "2006 Model").3 Paragraph 1 of the LOB Article provides that a resident of a Contracting State will be entitled to treaty benefits only if it is a "qualified person." Paragraph 2 sets out six types of qualified persons: Paragraph 2(a) applies to individuals, while paragraph 2(b) applies to the Contracting States themselves, including their political subdivisions. Paragraph 2(c) applies to public companies and paragraph 2(d) applies to companies that are at least 50% owned by five or fewer public companies (hereafter, the "public company subsidiary rule"). Paragraph 2(e) applies to exempt organizations and pension funds. Finally, paragraph 2(f) contains what is known as the ownership and base erosion test.

Paragraph 3 of the LOB Article contains what is known as the active trade or business test. If that test is satisfied, a resident of a Contracting State can qualify for limited treaty benefits with respect to income derived from the other state in connection with its trade or business, even if it is not a qualified person.

Paragraph 4 of the Proposed Model incorporates a derivative benefits provision that is not found in the 2006 Model. A derivative benefits provision is, however, contained in several current U.S. tax treaties. Unlike the version of this rule found in current U.S. treaties, the proposed derivative benefits provision would apply to "equivalent beneficiaries" from any country that has a treaty with the source state, rather than only to blocs of countries such as the European Union.

Paragraph 5 provides for competent authority relief in certain cases where a resident satisfies none of the other tests of paragraphs 2, 3 or 4. Finally, paragraph 6 contains definitions of certain terms used in the LOB Article.

Apart from the addition of the derivative benefits provision, the LOB Article of the Proposed Model generally restricts entitlement to treaty benefits as compared with the 2006 Model. It does this by adding additional requirements to several categories of "qualified person" in paragraph 2, by narrowing the class of persons entitled to rely upon the active trade or business test of paragraph 3 and by adding an additional limitation to secure competent authority relief under paragraph 5. Unlike the other proposed revisions set forth in the Proposed Model, the proposed revisions to the LOB Article are not accompanied by a draft Technical Explanation.

 

B. Specific Changes

 

The Proposed Model LOB Article provides in paragraph 1 that a resident of a Contracting State must be a qualified person "at the time" the treaty benefit would be accorded. Under the 2006 Model, the determination whether a person is a qualified person is made on a taxable year-by-taxable year basis.

The Proposed Model LOB Article adds a base erosion test to paragraph 2(d). Unlike the base erosion test of paragraph 2(f), the base erosion test contained in paragraph 2(d) is not applicable in testing the company's entitlement to claim treaty benefits with respect to dividends received. This may suggest that the Treasury Department was concerned about conduit structures involving public company subsidiaries, such as back-to-back loans, but not concerned about such companies being the true beneficial owners of dividends paid to them.

This new base erosion test, as well as the other base erosion tests contained in paragraphs 2(f) and 4 of the Proposed Model, is applied twice: once at the level of the company seeking to qualify for treaty benefits and a second time to that company's "tested group," as defined below. In the case of paragraphs 2(d) and 2(f), it must be shown that less than 50% of the company's or tested group's gross income, as further defined below, is paid or accrued, directly or indirectly, in the form of deductible payments either to persons who are not residents of one of the Contracting States and qualified persons described in paragraph 2(a), (b), (c) or (e), or to persons who would be so eligible but for the fact that they benefit from a special tax regime with respect to the payment. The base erosion tests contained in paragraph 4 asks whether less than 50% of the company's or tested group's gross income is paid to persons who are not equivalent beneficiaries or to persons that are equivalent beneficiaries but benefit from a special tax regime. All of the 50% tests exclude arm's length payments in the ordinary course of business for services or tangible property.

Proposed paragraph 6(g) defines "tested group" to include the tested company and any intermediate owner thereof that is both a resident of the same Contracting State and a member of the tested company's tax consolidation regime or similar group regime. No "tested group" can exist if the tested resident has no intermediate owner.

Proposed paragraph 6(h) defines the term "gross income" to mean gross income as determined in the tested person's state of residence, but excluding dividends that are "effectively exempt" from tax and, in the case of a tested group, income received or accrued from another member of the tested group. However, effectively exempt dividends are included in gross income for purpose of testing the tested company's entitlement for benefits under Article 10, relating to dividends received.

The Proposed Model LOB Article adds a derivative benefits provision whereby entitlement to treaty benefits may be granted, under certain conditions discussed below, to companies owned by "equivalent beneficiaries." Paragraph 6(e) defines an "equivalent beneficiary" to encompass two types of persons. The first is a resident of the same Contracting State as the tested company that is a qualified person within the meaning of paragraph 2(a), (b), (c) or (e). The second is a resident eligible for the benefits of a different comprehensive tax treaty with the source state by reason of being a "qualified person" under the counterpart to paragraph 2(a), (b), (c) or (e) of the Proposed Model. In the second case, the equivalent beneficiary must be eligible for benefits under that treaty that are at least as generous as those being claimed.

In order for a company to qualify under the derivative benefits test, seven or fewer equivalent beneficiaries must own, directly or indirectly, at least 95% of the aggregate vote and value of such company's shares and at least 50% of any disproportionate class of shares. In the event of indirect ownership, each intermediate owner must be a "qualifying intermediate owner" (hereafter, "QIO"). Proposed paragraph 6(f) defines the term QIO as an intermediate owner that is a resident of a state that has in effect with the source state a comprehensive tax treaty that includes special tax regime language and entitles such intermediary to benefits that are at least as generous as those being claimed by the tested company. In addition, to meet the derivative benefits test, a company must meet the base erosion test already described above.

For residents of a Contracting State seeking to qualify an item of income for treaty benefits under the active trade or business test, the Proposed Model LOB Article would add a new limitation on the attribution of activities from connected persons. Under the 2006 Model, the active conduct of a business in the residence state by a connected person (essentially a 50% or more affiliate) is attributed to another connected person. Under the Proposed Model LOB Article, a connected person's activities will be deemed to be conducted by the tested resident only if both persons are engaged "in the same or complementary lines of businesses."

Paragraph 5 of the Proposed Model adds a new requirement for persons seeking competent authority relief, pursuant to which that persons must demonstrate that it has a "substantial nontax nexus" to its state of residence. This new requirement incorporates the requirement of Section 3.06(d)(2) of Revenue Procedure 2015-40, 2015-35 IRB 236, which previously applied only for purposes of Treasury's internal procedures.

III. SUMMARY OF THE REPORT'S RECOMMENDATIONS

Following is a summary of this report's recommendations:

 

1. In general, we are concerned that many of the proposed changes set out in the Proposed Model go too far in the direction of restricting access to treaties. While we appreciate that the Treasury Department is concerned about certain practices that may be deemed to constitute treaty abuse, we believe that it is better to address abuses with substantive rules, rather than to deny access to a treaty across the board. In deciding whether to use the LOB Article to address abuses, the Treasury's concerns about abuses should be balanced against the risk that taxpayers with common non-tax motivated structures could be denied all treaty benefits. We are concerned that the Proposed Model LOB Article may not strike the right balance. We believe that the LOB Article should be narrowed to address clearly-articulated cases of treaty shopping or other treaty abuse, such that treaty benefits are not denied to companies that employ common non-tax motivated structures.

2. We recommend that the base erosion prong not be added to paragraph 2(d) of the Proposed Model but be incorporated into specific treaties only where Treasury has identified a specific concern that cannot be addressed in another fashion. If, contrary to our recommendation, the base erosion test of paragraph 2(d) is retained, we believe that Treasury should consider not applying it to companies that are wholly or majority owned by a single public company parent. No base erosion rule applies to public companies themselves, and we believe it would be inappropriate to apply a base erosion test to a wholly or majority owned subsidiary of such a company, because such a rule would limit a public company's ability to operate through subsidiaries.

3. We recommend that the base erosion test not apply to payments of interest to unrelated persons on borrowings that arise from ordinary course capital markets transactions. We further recommend that the exclusion for such interest not be limited to payments to "banks."

4. The Proposed Model treats a payment, even to a qualified person, as a base-eroding payment if the recipient benefits from an STR. We recommend that if the STR rule in the LOB Article is retained, it should be limited to payments to related persons.

5. We recommend that the exclusion from gross income for "effectively exempt" dividends be removed from the Proposed Model.

6. We generally support the concept of applying the base erosion test on a tested group basis. However, we are uncertain why the Proposed Model proposes to test twice, both at the tested group level and at the tested company level. Moreover, we do not understand why the definition of tested group in paragraph 6(g) appears to be limited to the tested company and its higher-tier owners.

7. Proposed paragraph 6(f) defines the term QIO as an intermediate owner that is a resident of a state that has in effect with the source state a comprehensive tax treaty that includes special tax regime language. We recommend that the STR requirement for QIOs should be eliminated from the Proposed Model at this time, because no existing treaties contain the STR rule.

8. In the context of the derivative benefits provision, we recommend that the "cliff rule" approach to the "as low as" requirement be eliminated in favor of a rule that would apply the higher of the two withholding tax rates. We have other technical comments and suggest that guidance on these rules be set forth in a Technical Explanation.

9. We suggest that Treasury consider not placing a limit on the number of equivalent beneficiaries a company can have, or that it consider increasing the number of equivalent beneficiaries that a tested company can have.

10. We do not support the addition of an "at least as favorable" test where an equivalent beneficiary or QIO seeks to obtain treaty benefits described in Articles 7, 13 and 21. We believe that such a test is too subjective to be applied consistently. We would support the application of a narrowly-crafted "at least as favorable" rule to equivalent beneficiaries (but not to QIOs, for the reasons stated below) in certain limited cases where the facts make clear that a person is engaging in treaty shopping, and only to the extent that a Technical Explanation to the Proposed Model sets out the standards to be applied.

11. We suggest that Treasury consider simplifying and streamlining the various intermediate owner tests set forth in the Model. We question whether the various restrictions upon who can qualify as an intermediate owner or QIO are necessary. In particular, we believe that a subsidiary of a public company that otherwise is a qualified person under paragraph 2(d) should qualify as a "good" intermediate owner. Moreover, we do not think that QIOs should be required to be residents of states, the tax treaties of which confer benefits at least as favorable as those under the treaty being tested.

12. We recommend that an equivalent beneficiary should include a resident of either of the two Contracting States, not only a resident of the State in which the tested company resides.

13. We recommend that the proposed change made to paragraph 3, which would limit the attribution of activities from a connected person to only those cases in which both persons are engaged in the same or complementary lines of business, be eliminated from the Proposed Model.

14. We recommend that the proposed "nexus" condition be removed from paragraph 5 of the Proposed Model, and should be discussed in a Technical Explanation as a relevant factor only.

15. Many of our members believe that a denial of treaty benefits pursuant to paragraph 5 should be subject to consultation between the Contracting States, or even to mandatory arbitration.

16. The change to paragraph 1 of the Proposed Model LOB Article, requiring a resident be a qualified person "at the time" when the treaty benefit would be accorded, is unclear. We suggest that a Technical Explanation provide guidance as to the application of the various timing rules in the Proposed Model.

 

IV. DISCUSSION

 

A. In General

 

We applaud Treasury's decision to release the proposed text of the new Proposed Model for public comment prior to its adoption. We believe that the comment process will assist Treasury in developing the new Proposed Model. Given the number of policy and technical issues presented by the proposed changes, including the changes to the LOB Article discussed in this report, we believe it would also be useful for Treasury to publish a Technical Explanation of the LOB Article. This report points out several areas where further guidance under the proposed LOB Article changes would be welcome.

We are aware that in the context of the OECD's BEPS project, Action 6, Treasury has been advocating for the use of an LOB to curtail treaty shopping, and that many other countries participating in the BEPS project continue to prefer a "principal purpose" test, which has in the past been rejected by the United States Senate.4 We understand that some of the changes contained in the Proposed Model have been proposed in light of the Action 6 discussion and in light of the broader BEPS discussions. The OECD has reserved on some elements of the Action 6 Report pending finalization of the Proposed Model.

Many of our concerns and comments relate to the fact that the changes being proposed to the LOB Article would significantly restrict access to treaty benefits even in many cases not involving treaty abuse or treaty shopping. The principal function of the LOB Article is to prevent the practice known as treaty shopping, which occurs when a resident of a third county attempts to gain access to a treaty between two Contracting States by creating an entity in one of those states. The LOB Article solves this problem by limiting treaty benefits to qualified persons and to other persons whose activities and structures take a form that evidences the lack of a purpose to treaty shop. But in our experience, it is already not uncommon for a taxpayer not engaged in treaty shopping or other treaty abuse to be prevented from qualifying for treaty benefits under the LOB articles contained in current treaties. This often occurs because corporate structures put into place for non-tax reasons having nothing to do with treaty shopping may include intermediate owners in third countries and group payments that implicate the base erosion rule.

Companies operating across borders, including but not limited to multinationals, often have complex structures that respond to legal, regulatory and business concerns unrelated to tax treaties. In our experience, for example, most U.S. and other multinationals own stock of foreign subsidiaries through chains of intermediary corporations. In many cases, those intermediary corporations will be incorporated or resident in a third jurisdiction, that is, not in one of the Contracting States that are party to the treaty being examined. There are many reasons for these chains of ownership. One common reason is that the parent has acquired foreign companies and has inherited structures that involve more than a single intermediate owner. While in our experience U.S. multinationals strive to restructure their ownership chains so as to place companies from the same country together, this is not always possible to do, either for legal reasons or because it would entail the payment of large up-front taxes on intercompany gains. It is also common for U.S. and other corporations to employ holding company structures. In such cases, a deductible payment may be made to a company within the group other than the common parent, which company may not be a resident of the country in which the payor or the parent is resident.

The changes in the Proposed Model would, if incorporated into a tax treaty, exacerbate this "square peg round hole" problem. The proposed changes to the LOB Article are designed to, and will, expand the class of persons unable to qualify for treaty benefits, even absent treaty shopping.

To the extent that the main provisions of the LOB Article deny access to treaty benefits, this will place further strain on the competent authority relief provided in paragraph 5. Securing competent authority relief can be time-consuming for both taxpayers and the government and causes uncertainty for taxpayers. For these reasons, we believe that the main provisions of the LOB Article should be designed such that resort to competent authority relief is rare. We therefore believe that the restrictions contained in the LOB Article should be narrowed so that they address clearly articulated cases of treaty shopping or other treaty abuse without unduly denying treaty benefits to companies that employ common non-tax motivated structures.

Another concern raised by the proposed restrictions in the LOB Article is that, to the extent that the United States is hoping to persuade other countries participating in the BEPS project to incorporate an LOB Article into Action 6 in lieu of a "principal purpose" test, these additional restrictions may deter the adoption of the LOB approach. Although most of our members agree that an LOB approach is superior to a principal purpose test, one obvious advantage of a principal purpose test over an LOB approach is that the former does not deny benefits based on the failure to pigeonhole oneself into a complex set of highly technical rules.

To the extent that the proposed restrictions on access to benefits under the LOB Article were motivated by a desire to address the desires of other countries to use tax treaties as a means of addressing abuses unrelated to simple treaty shopping, we are not persuaded that the Proposed Model strikes the correct balance. We comment below on the gulf between the United States' approach and the BEPS approach to these issues as set forth in the final version of the Action 6 report.5 For these and other reasons, we believe that the provisions of the LOB Article should be designed primarily with treaty shopping in mind, and should accord benefits to most non tax-motivated structures that do not implicate treaty shopping.

 

B. The LOB, Treaty Abuse and Double Non-Taxation

 

A number of the changes in the Proposed Model appear to emanate from a desire to use the LOB Article as a tool to combat base erosion and double non-taxation. As discussed below, this is not an entirely new development. However, the result is a complex set of technical rules that expand the LOB Article beyond its original purpose of preventing treaty shopping. While we understand Treasury's motivation for using tax treaties to prevent double non-taxation of income, for the reasons set forth herein we do not believe that the LOB Article is the proper tool to address those concerns. While it may be appropriate to adopt rules that deny particular treaty benefits in targeted abusive cases, such as the STR rules, addressing what may in fact be minor "abuses" by precluding a corporation from accessing a treaty entirely, which is what the LOB Article does, is an extreme remedy. The need to address abuses should be weighed against the harm done by denying treaty benefits, across the board, for what may in many cases be technical foot-faults.

As originally formulated, the LOB Article had little or nothing to do with questions of double non-taxation. The original purpose of the LOB Article was, as stated in the Technical Explanation to the 2006 Model, "to prevent residents of third countries from benefitting from what is intended to be a reciprocal agreement between two countries."6 The Technical Explanation of the 1996 Model Treaty LOB Article speaks in terms only of treaty shopping, stating: "The United States holds strongly to the view that tax treaties should include provisions that specifically prevent misuse of treaties by residents of third countries."7 It went on to point out that the United States relies on domestic law to address other concerns:

 

Article 22 and the anti-abuse provisions of domestic law complement each other, as Article 22 effectively determines whether an entity has a sufficient nexus to the Contracting State to be treated as a resident for treaty purposes, while domestic anti-abuse provisions (e.g., business purpose, substance-over-form, step transaction or conduit principles) determine whether a particular transaction should be recast in accordance with its substance. Thus, internal law principles of the source State may be applied to identify the beneficial owner of an item of income, and Article 22 then will be applied to the beneficial owner to determine if that person is entitled to the benefits of the Convention with respect to such income.8

 

Although an LOB provision was featured in some earlier treaties and in Section 884(e) of the Code,9 the first highly articulated version of the LOB was introduced into the United States' tax treaty with The Netherlands in 1992. That LOB emanated from Treasury's concern that if third-country residents could use Dutch companies to treaty shop, other countries would have little incentive to enter into tax treaties with the United States, thereby harming U.S. businesses. This core concern is summarized in the Action 6 Report as follows: "If . . . a State knows that its residents can indirectly access the benefits of treaties concluded by another State, it may have little interest in granting reciprocal benefits to residents of that other State through the conclusion of a tax treaty."10

The continuing emphasis that the United States places upon treaty-shopping concerns is reflected in other provisions of the LOB Article. For example, the active business test does not incorporate either an ownership or a base erosion test. That is because paragraph 3 is concerned only with the question whether the resident entity was established for good business reasons, as opposed to treaty-shopping reasons. As stated in the Technical Explanation to the 1996 Model:

 

. . . the assumption underlying the active trade or business test under paragraph 3 is that a third country resident that establishes a "substantial" operation in the other State and that derives income from a similar activity in the United States would not do so primarily to avail itself of the benefits of the Treaty; it is presumed in such a case that the investor had a valid business purpose for investing in the other State, and that the link between that trade or business and the U.S. activity that generates the treaty-benefitted income manifests a business purpose for placing the U.S. investments in the entity in the other State. It is considered unlikely that the investor would incur the expense of establishing a substantial trade or business in the other State simply to obtain the benefits of the Convention. A similar rationale underlies the other tests in Article 22. (Emphasis added.)11

 

The same policy is reflected in the public company rule of paragraph 2(c): It is presumed that the myriad owners of a public company are not motivated to form the company by treaty-shopping concerns.12 Thus, under the LOB Article, including the version set forth in the Proposed Model, it is irrelevant whether a company described in paragraph 3 or paragraph 2(c) engages in base erosion or is subject to tax at home on the income it earns.

It is true that the Technical Explanation to previous U.S. Model Treaties states that the base erosion test was intended to ensure that the recipient of a payment is "subject to tax." But this language must be placed in context. A pure "subject to tax" inquiry is too vague and imprecise to be administrable. There are nearly infinite combinations of foreign tax rules that, applied alone or together to different types of income and payments, could lead a tested company to pay tax in its state of residence at a low or zero rate.

Rather, the LOB Article's "subject to tax" test is concerned with whether a person is the true owner of the income in question. Whereas the BEPS project is concerned with base-stripping to low-tax countries, the guiding principle of traditional U.S. policy under the LOB Article has been concerned with base stripping to any non-treaty jurisdiction, even a high-tax one. The question under a U.S.-style LOB Article is not whether a Contracting State's tax base is being eroded, but rather whether the benefits of a treaty are being indirectly accorded to person not otherwise entitled to treaty benefits.

The proper function of the base erosion prong of the ownership and base erosion test is to deter treaty-shopping by backstopping the ownership prong. The example below illustrates one way in which the base erosion prong operates to backstop the ownership rule:

 

Example 1. A, a resident of Country A, which has no tax treaty with the United States, causes corporation X to be formed and resident under the laws of Country B, which does have a tax treaty with the United States. A's cousin B, a resident of Country B, beneficially owns 100% of the stock of corporation X, such that the ownership prong of the test is satisfied. Corporation X is capitalized with 10x of equity from B, and with 100x of debt advanced by A. Substantially all of the income of X is paid to A as interest on the debt owed to A.

 

In Example 1, A is effectively the owner of X from a treaty-shopping perspective. A derives most of the income from the operations of X. B functions much as a nominal owner of X. In this example, the treaty-shopping concern would be the same regardless of whether Country B taxed the interest at a high rate, a low rate or zero.

The base erosion test also serves to discourage treaty shopping by the use of conduit structures, such as back-to-back loans:

 

Example 2. H, a resident of neither Contracting State, lends funds to corporation X, a resident of Country B owned by individual residents of Country B. Corporation X relends the borrowed funds to A, a U.S. person, who makes interest payments to corporation X. The grant of U.S. treaty benefits with respect to the interest paid to corporation X is in effect a grant of treaty benefits to H, who would not be entitled to those benefits directly.

 

In Example 2, the same treaty abuse would be present even if corporation X were subject to tax in Country B on all or a portion of its interest spread, and even if some tax were payable by H. It is the indirect access to the treaty by H, not the taxation of the recipient in its country of residence, that is the concern of the LOB Article.

These examples illustrate the main purpose of the base erosion test, which is to backstop residence. A resident is a person subject to tax. But this kind of "subject to tax" test does not presume to ask whether the resident can lower or reduce its tax rate by base-eroding payments, or whether the country of residence in fact imposes a tax on a particular type of income. Indeed, many countries with which the United States has a tax treaty do not tax dividends received.13 Moreover, the base erosion test applies even if the recipient of the payment is subject to tax at rates equivalent to or even higher than the rates applicable to the tested company.

Although there is no Technical Explanation to the Proposed Model's LOB Article, there is a very detailed explanation of the OECD's version of the LOB Article in the Action 6 Report. A reading of that Report suggests that many of the changes in the Proposed Model may have been designed to answer concerns raised by other countries that the LOB Article, standing alone, is insufficient to address various examples of "treaty abuse."14 The Action 6 Report, like its draft predecessors, makes a distinction between "treaty shopping" and other forms of "treaty abuse." It proposes that a "principal purpose" test be aimed at these other forms of treaty abuse.

Nearly all of the "non-treaty shopping" treaty abuses identified in the Action 6 Report are already policed by provisions of U.S. domestic law or by provisions of tax treaties other than the LOB Article. We presume that, among the reasons that the United States has rejected a principal purpose test is that the United States does not need to rely on restrictions in the LOB Article to address these other abuses. For example, the Action 6 Report lists the following abuses targeted by its principal purpose test:

 

1) dividend transfer and "usufruct" transactions, which are essentially assignment of income problems long addressed by U.S. law;

2) transactions designed to convert capital gains from immovable property into income qualifying for treaty benefits, long addressed by FIRPTA;

3) dual resident company structures, which are addressed extensively by U.S. domestic rules as well as by specific provisions in U.S. tax treaties;

4) the "triangular case" problem involving permanent establishments, which is addressed by the 2006 Model and expanded (we think appropriately) in the Proposed Model.

 

The authors of the Action 6 Report also appear to believe that a principal purpose test is needed to avoid having a tax treaty "trump" a domestic anti-abuse rule. The United States, as a matter of treaty policy, does not share this view. The U.S. position is that domestic law provisions such as anti-conduit and substance over form rules trump treaties. The Technical Explanation of the 2006 Model states, in the discussion of the LOB Article, that substance over form and similar concepts of U.S. law apply first to determine the transaction applicable to tax. Only after that determination is made does a treaty apply to the transaction. Treaty benefits can be denied if domestic law does not respect the form adopted to seek access to treaty benefits.

It appears that the United States and the authors of the Action 6 Report are in fundamental disagreement over what the purpose of Article 22 should be. The BEPS project is divided into three main "pillars": coherence, substance and transparency/certainty. Action 6 was placed under the "substance" pillar, as was Action 7 on permanent establishments, which relates to treaties. The Action 6 Report speaks in terms of "depriving countries of tax revenues," of "letterbox companies" and of the "substance" of a company's operations in a treaty country.

Perhaps one reason that the United States and other countries have had difficulty agreeing on an approach to treaty abuse is that the United States does not generally view treaty abuse as an issue of "substance," in the sense of an entity having a nexus with and reason to be in a particular country. Instead, the traditional U.S. view of treaty abuse has been directed at treaty shopping. Substance and protecting tax revenue was not the traditional function of the LOB Article, and the concerns of the OECD as expressed in the Action 6 Report are not traditional concerns of the United States in the treaty context.

To the extent the changes to the LOB Article in the Proposed Model are seeking to discourage double non-taxation, we think that issue is better addressed directly, and not by tightening the rules for treaty eligibility generally. Other provisions of the Proposed Model, as well as extant U.S. treaties, already deal with, or can be adapted to deal with, double non-taxation issues. Among the existing treaty rules that address double non-taxation concerns are the denial of lower rates of dividend withholding tax to RICs and REITs and other non-taxpaying entities, rules to prevent inappropriate claiming of benefits by exempt permanent establishments, broad FIRPTA rules, rules treating contingent interest as dividends in certain circumstances, a rule that recharacterizes certain capital gains as royalties, rules denying treaty benefits in triangular cases and a savings clause. The Proposed Model would add a special tax regime rule that, if adopted, would operate to address many of the abuses associated with double non-taxation planning. These types of rules are better suited to addressing issues of double non-taxation and base erosion that are rules that would deny treaty benefits across the board.

 

C. Paragraph 2(d): Public Company Subsidiaries

 

1. In General
One of the significant substantive changes in the Proposed Model is the addition of a base erosion prong to paragraph 2(d). We have identified only one significant extant U.S. tax treaty that includes a base erosion prong for public company subsidiaries, that with Luxembourg.15 To date, we believe that the base erosion test has generally applied only to private companies seeking qualified person status under the ownership and base erosion test of paragraph 2(f), and to the derivative benefits clause of those tax treaties that incorporate a derivative benefits clause. We have reservations over the proposal to incorporate a base erosion prong into paragraph 2(d) of the Proposed Model.

As discussed above, the primary and original purpose of the base erosion prong was to backstop the ownership test so as to prevent treaty shopping. Paragraph 2(c), applicable to public companies themselves, contains no ownership test. We believe this is the reason that it also contains no base erosion test. We think that another reason that the base erosion test does not apply to public companies is that public companies cannot be used as nominal owners or conduits in the ways illustrated by Examples 1 and 2. Public companies simply do not present the same type of treaty shopping concerns that private companies do. For reasons having to do with regulation, disclosure and/or reputation, a public company would rarely allow itself to be used as a conduit for treaty shopping by unrelated persons. It is also likely that one reason for not extending a base erosion test to public companies is one of practical necessity; it may have been thought too difficult for large public companies with many different payments to affiliates and third parties to monitor compliance with a base erosion test.

We believe that similar concerns have applied to subsidiaries controlled by one or more public companies, at least in the common case where the subsidiary is wholly or majority owned by a single public company. Many public companies conduct most or even all of their operations through controlled subsidiaries. A public company's disclosure, accounting, regulation, liability and reputational risk extend to its controlled subsidiaries. A controlled subsidiary of a public company is just as unlikely to be used as a nominee for a non-treaty country person to treaty shop as a public company is, and equally unlikely to act as a conduit for third persons. We believe that this is the reason that a base erosion test has not previously been applied to public company subsidiaries; only indirect ownership by the public, through its parent(s), has been considered relevant.

We assume that the reason for adding a base erosion prong to paragraph 2(d) was to prevent a subsidiary of a public company from stripping income out to an affiliate of that same public company not resident in a treaty jurisdiction. Consider the examples below: In Example 3, the subsidiary is owned by a public company resident in the same country, whereas in Example 4 the subsidiary is owned by a public company in the other country party to the treaty.

 

Example 3. P, a publicly-traded U.S. corporation, owns 100% of the shares of S, a corporation that is also a resident of the United States. S derives income from U.K. sources and seeks to qualify for benefits under the U.S.-U.K. tax treaty by relying on paragraph 2(d). However, more than half of S's income is paid in the form of deductible payments to sister company Y, a subsidiary of P resident in a third country (which could be a high-tax country or not). Under proposed paragraph 2(d) of the Model, S would not be a qualified person.

Example 4. K, a publicly-traded U.K. corporation, owns 100% of the shares of S, a U.S. corporation. S derives income from U.K. sources and seeks to qualify for benefits under the U.S.-U.K. tax treaty by relying on paragraph 2(d). However, more than half of S's income is paid in the form of deductible payments to Z, a subsidiary of K resident in a third country. Under proposed paragraph 2(d) of the Model, S would not be a qualified person.

 

Example 4 is similar to an example included in the Action 6 Report discussing the principal purpose test.16 There, TCO, a resident of Country T, had made a loan to SCO, a resident of Country S. Because there was no tax treaty between T and S, TCO assigned the loan to an affiliate, RCO, resident in Country R, which did have a treaty with S and T exempting interest from withholding tax. TCO arranged a back-to-back loan such that SCO paid interest to RCO and RCO paid interest to TCO. If in this example the United States was Country S, the question would be whether RCO was a qualified resident under the LOB Article. Because RCO is not a public company and could not meet the requirements of paragraphs 2(f) given that it is formed in a third country, treaty benefits would be available here only if TCO were an equivalent beneficiary (which it might be) or TCO were a public company and RCO its controlled subsidiary.

The issue presented by these examples could be framed as a double non-taxation issue, or as a conduit issue. If the latter, we would observe that U.S. anti-conduit rules should apply, without the need for adding a base erosion prong.

In Example 3, S's deductible payment is made to Y, a controlled foreign corporation of P. The fact that the subpart F rules apply to this case might suggest that the United States would not desire another country to deny treaty benefits in that example. Further, S would be entitled to treaty benefits if Y were a disregarded entity, because the payment would be treated as having been made to P.17

The Action 6 Report also contains an example involving deductible payments to a permanent establishment, in U.S. parlance a disregarded entity or branch.18 In that example, BCO, a resident of Country B, owns all of the stock of ACO, a resident of Country A, and also owns a branch in Country A. ACO makes a "group contribution" to the branch. The Action 6 Report concludes that this payment should not be treated as a base-eroding payment under the equivalent of Proposed Model paragraph 2(f),19 because the payment is taxable to the branch. The United States, of course, would not view this transaction in the same way. It would deem the payment to be made to BCO. Since BCO is by stipulation entitled to treaty benefits in its own right, no issue would be presented under the LOB Article. We question why the result under the LOB rules should be so radically altered based on whether Y is a controlled foreign corporation or a disregarded entity.

We also note that just as a subsidiary of a public company can make base-eroding payments to affiliates, the public company itself can do so. If one of the purposes of the base erosion test were to prevent double non-taxation by ensuring that the company being tested for qualified person status is "subject to tax," one would expect that the base erosion prong would apply to public companies. However, the Proposed Model, like the 2006 Model and all extant U.S. treaties, does not apply the base erosion prong to such companies. It therefore does not seem appropriate to apply a base erosion to test to the subsidiary while excusing the parent. If Treasury decides to retain the base erosion rule in paragraph 2(d), it might consider exempting from that rule companies that are wholly-owned by, or part of the same tested group as, a public parent, particularly if formed in the same country as the public parent.

We have independent reservations over expanding the classes of persons to which a base erosion test is applied. These arise from the fact that the test can be very difficult to apply in practice. We are aware that the base erosion test is not new to the Proposed Model, but its extension to public company subsidiaries, especially when coupled with the other proposed rules that would further restrict access to treaty benefits, warrants a review of some of the problems encountered in practice in applying the test.

Some of the difficulties encountered in applying the base erosion test were reviewed and discussed in our 2006 Report. These include how to account for gross income and payments made by affiliates, including disregarded entities, as well as timing issues. That report also discussed the difficulty of demonstrating that a given payment is made to a qualified person. For example, a company may make interest payments to a large number of lenders under a syndicated loan facility. It would be impractical in many cases for the borrower to ascertain the qualified person status of all of the lenders in the pool. In our experience, in many cases companies that are subject to a base erosion test are uncertain whether they qualify, because of the practical difficulty of monitoring the status of all of their payees. We separately discuss interest payments in the next section of this Report.

In summary, we believe that extending the base erosion test to subsidiaries of public companies may have the effect of denying treaty benefits in many cases not involving treaty shopping, necessitating resort to competent authority relief which in most if not all cases would be expected to be granted. The risk that the addition of a base erosion test to paragraph 2(d) (especially to subsidiaries controlled by a single public company) would have the effect of denying treaty benefits where no treaty shopping is in fact occurring, as well as the practical difficulties in applying a base erosion test, may outweigh any benefits expected to be achieved by the addition of such a rule. We therefore recommend that the base erosion prong not be added to paragraph 2(d) of the Proposed Model, or at least that it not apply to companies controlled by a single public company.

2. Interest Payments
Treasury has asked for comments regarding the appropriateness of including an exception from the base erosion test for payments in respect of loans made by unrelated banks. We believe such an exception is needed. Absent such an exception, a borrower making payments to lenders would need to determine not only what country the lender was a resident of, but also whether the lender was a "qualified person" within the meaning of the applicable treaty. That inquiry could often entail an inquiry into a number of facts that the borrower would have no reason to know, and that the lender might be unwilling to provide.

We also believe that the exception for payments of interest should be broader than interest paid to "banks," such that it covers payments of interest to unrelated persons who extend credit in ordinary commercial transactions, as well as their assignees. Public companies and their subsidiaries often borrow large amounts in international public markets and pursuant to private debt facilities syndicated to multiple participants in multiple countries. Some of the lenders may not be banks, or at least not banks as defined under U.S. tax principles. We do not see why the status of a lender as a "bank" or not should affect the outcome under the LOB Article. Moreover, originating banks often sell or participate their loans to persons who were not part of the original lending pool. In most cases, the borrower may not be aware of the sale and thus would have no knowledge of the identity of the lenders receiving interest payments.

We believe that many or even most publicly-traded banks do not lend at the level of the public company, but only through subsidiaries. If those subsidiaries can be qualified persons only under a version of paragraph 2(d) that incorporates a base erosion prong, then, particularly in the case of widely-syndicated loans, it would appear to us to be a nearly impossible task for the borrower to determine whether its lenders are qualified persons. To ascertain that fact, the borrower would need to know whether its lenders themselves make base-stripping payments, and to whom. Thus, without an exception for ordinary commercial loans, most borrowers would not be in a position to ascertain whether they meet the base erosion test.

3. STR Payments
We suspect that one of the reasons for adding a base erosion prong to paragraph 2(d) was to treat as base-eroding payments any payments made to persons -- even qualified persons -- that benefit from a special tax regime. But the STR rule in the proposed base erosion prong of the Model is not limited to payments between related persons, as the general STR rule is. As we said in our earlier report,20 we believe that the limitation to payments between related persons is sound, and that it would be difficult and impractical to apply the STR rules as between unrelated persons. We believe that if the STR rule in the LOB Article is retained, it should be so limited. In order to ascertain whether a tested company satisfies the base erosion test, it is necessary to know the qualified person status of the payee. If the payee is not a related party, it may be difficult if not impossible to ascertain its status.

 

D. Definitional Issues Related to the Base Erosion Test

 

1. "Effectively Exempt" Dividends
Among the changes made to the base erosion test, we generally approve of the Model's adoption of a gross income test that looks to income as measured for local tax purposes. Under existing precedents, gross income had generally been measured using the tax principles of the source state. By conforming to local country tax principles, the test would generally be expected to more closely reflect the policy of the base-stripping rules. This is so, whether one conceives of the policy behind the base-stripping rules narrowly as an anti-conduit policy or more broadly as a "subject to tax" test. In our 2006 Report, we addressed the question whether the income and deductions of a subsidiary that is disregarded for U.S. tax purposes should be taken into account as items of the parent -- applying U.S. tax principles and definitions -- or as separate items of the subsidiary -- taking into account only the other country's rules. As stated in our 2006 Report, "[w]hether parent is being used as an intermediary in the manner that base erosion tests aim to prevent is independent of the U.S. tax classification of parent's subsidiaries."21

However, we are concerned about the breadth of the rule that excludes from gross income dividends that are "effectively exempt" from tax in the recipient's residence state. While a gross income definition that looks to local law might or might not exclude exempt dividends, there is a larger policy question implicated by this exclusion. Many countries exempt non-portfolio dividends from tax as a means of avoiding multiple taxation of income in corporate groups. This exclusion is a good example of the tension between an approach to the LOB Article based on treaty shopping and an approach based on a "subject to tax" test unrelated to treaty shopping. Two examples illustrate these approaches:

 

Example 5. Assume that the United States has adopted a participation exemption regime under which all dividends from active subsidiaries are exempt from U.S. tax. U.S. resident individuals A and B form U.S. holding company X to invest in a worldwide business consisting of subsidiaries in France and Germany, F and G respectively. Company X makes a $1 deductible payment to company G, which may or may not be taxable in Germany. If company F pays a dividend to X, treaty benefits would be available, but only because the exception in paragraph 6(h)(i) provides that gross income does not exclude effectively exempt dividends for purposes of determining benefits under Article 10 (dividends).

Example 6. Assume the same facts as in Example 5 except that company F pays a $1 royalty to X. Treaty benefits would be denied with respect to the royalty.

 

Neither of the above examples implicates treaty shopping. Example 5 is clearly anodyne, and we assume that it was this case that Treasury had in mind in limiting the exclusion to payments other than dividends. Example 6 illustrates the conceptual problems in using the LOB Article as a pure "subject to tax" rule.

First, there is no reason to suppose that X would not pay net tax in the United States on the royalty income. The hypothesized U.S. participation exemption regime might contain a rule that allocated a portion of the deductible payment to otherwise exempt dividends, and away from the taxable royalty. In that case, perhaps the deduction allocable to the otherwise exempt dividend would not be counted in the base erosion test.

Second, in Example 6, the deductible payment is made to a person, G, that is not resident in either the United States or France, and thus not an eligible recipient, but is a sister of the paying company and clearly part of a group that has a business purpose for being structured the way it is. We question why any of this should be relevant to whether X should be entitled to treaty benefits, given the absence of any treaty shopping concern.

Third, we think it is confusing for the definition of gross income to operate one way when the question is whether the tested company is entitled to the benefits of Article X, and another way when the question is whether it is entitled to other treaty benefits. The definition of qualified person in paragraph 2 appears to operate in a binary fashion. To say that a person is not a qualified person under a treaty generally -- because the base erosion test is not satisfied -- and yet say that it is entitled to treaty benefits for dividends, seems to add a degree of complexity to the analysis that may not be warranted.

For these reasons, we recommend that the exclusion for "effectively exempt" dividends be removed from the Proposed Model.

2. Tested Groups
The base erosion tests of the Proposed Model incorporate a new provision that would test base erosion twice, once at the level of the tested company itself and a second time to any tested group. Subject to our comments on the appropriateness of base erosion tests in specific circumstances, we generally support the concept of applying the test on a tested group basis.22 In our 2006 Report, we advocated applying a tested group concept in certain cases in which the company claiming treaty benefits owned a subsidiary with which it filed on a consolidated, loss sharing or similar basis under local tax rules. We remarked that "[a] comprehensive analysis of LOB provisions would also examine the question whether the base erosion test, as applied to a subsidiary claiming treaty benefits, should take into account income and deductions of the subsidiary's parent or other affiliates."23

We are uncertain why the Proposed Model proposes to test twice, both at the tested group level and at the tested company level. We believe that it may be sufficient to apply the test only at the tested group level. If Treasury has a concern that this would be insufficient, it would be helpful to understand what that concern is.

We do not understand why the definition of tested group in paragraph 6(g) appears to be limited to the tested company and its higher-tier owners. Base-eroding payments can be made by lower-tier and sister companies that may be included in a consolidated or loss-sharing group. It may be appropriate to tailor the tested group concept on a specific treaty-by-treaty basis, taking into account the types of regimes used in the Contracting States.

There are a number of technical issues presented by the adoption of a tested group mechanic, many of which were addressed in our 2006 Report. We believe that Treasury should publish a Technical Explanation to the Proposed Model describing in some detail how the tested group concept is intended to operate, including examples of how the exclusion from gross income of payments within the group should be taken into account.

 

E. Derivative Benefits -- Paragraph 4

 

The derivative benefits paragraph is new to the Proposed Model, but a version thereof is contained in several existing U.S. tax treaties. We support its inclusion in the Proposed Model. One reason we support its inclusion is that we believe there are many cases in which residents of a treaty country, including residents of the United States, will not be qualified persons under paragraph 2, and may not qualify for the more limited benefits of paragraph 3, even absent treaty shopping. Since we believe that the LOB Article should be capable of addressing the great majority of cases that do not involve treaty shopping, the addition of a derivative benefits article is a welcome development.
1. The STR Requirement
The concept of a QIO is one of the key defined terms used in paragraph 4. It is used in the derivative benefits article to refer to the types of entities through which equivalent beneficiaries may indirectly hold their interest in the tested company. In our experience, it is rare to find that all owners of a company hold their shares directly. For this reason, the definition of QIO will often need to be met in order to qualify for benefits under paragraph 4.

The definition of QIO requires, inter alia, that the intermediate owner be resident of a country that has a treaty with the source state "that includes provisions addressing special tax regimes analogous to the provisions included in this Convention." Of course, no existing tax treaty includes any provisions relating to STRs, because the concept of STRs was first introduced in the Model. If this part of the definition were to be included in the first treaty that the United States enters into following adoption of the Model, it would make the derivative benefits rule a nullity.

The STR language seems designed for the multilateral instrument being worked on by the majority of the countries party to the BEPS project as Action Item 15.24 It does not work well in the context of traditional bilateral treaties, at least until there is a critical mass of treaties containing provisions for STRs. And for the reasons discussed below under "Intermediate Owners," we do not think that the fact that a QIO happens to be resident in a country whose treaty lacks an STR provision has relevance to the purpose of the LOB Article. The QIO is not seeking treaty benefits, and should not be tested as if it were.

We do not think that the STR rule should be part of the Proposed Model at this time. We suggest that any such rule might be included in the Proposed Model only on a "springing" basis, such that it would take effect only after a multilateral treaty were adopted, or until most U.S. treaties incorporated an STR rule.

2. The Cliff Rule
In order to qualify under the derivative benefits paragraph, at least 95% of the interests in a company must be owned by seven or fewer equivalent beneficiaries. The term equivalent beneficiary is defined such that the person must be resident of a country having a treaty with the source state that is "at least as good" as the treaty that the tested company is trying to qualify under. Specifically, if the benefit being sought is with respect to dividends, interest or royalties, the equivalent beneficiary's treaty must provide for withholding rates "at least as low as the rate applicable under this Convention."

This rule is often referred to as a cliff rule. If the rate of withholding on dividends is 10% under the treaty in question, but the rate is 15% under the equivalent beneficiary's treaty, treaty benefits would be denied altogether, increasing the rate to 30% in the case of the United States as source country. This does not appear sound as a policy matter. The better approach would be to apply the higher of the two rates, here 15%. We believe that this is the approach taken under the United States' treaty with the United Kingdom, albeit only pursuant to the Technical Explanation.

Where there is more than one equivalent beneficiary, the highest rate applicable to any of them could be used. Alternatively, one could determine the percentage ownership of each and apply a weighted rate to the company based on the percentage and the rate in effect in their country, but not be lower than the rate contained in the applicable treaty. For example:

 

Example 7. Company C is resident inn Country C, which has a rate of withholding tax on dividends of 10%. Company C is owned by three equivalent beneficiaries, J, K and L. J owns 20% of the stock of C and is resident of a country that has a tax treaty with the United States reducing the U.S. tax on dividends to 15%. K owns 30% of the stock of C and is a resident of a different country having a treaty that reduces the rate of tax on dividends to 10%. L owns 50% of C's stock and is resident of a country whose treaty entitles residents to a 5% rate of U.S. withholding tax. Under the weighted rate approach, C would qualify to receive U.S. source dividends at the rate of 8.5% (20% x 15% + 30% x 10% + 50% x 5%). However, since the rate of withholding tax under the Country C treaty is 10%, the rate applicable in this case would be 10% and not 8.5%.

 

In the event that up to 5% of the tested company is owned by persons who are not equivalent beneficiaries, the source country's default rate (i.e., 30% in the United States) could be used in the formula. We recognize that to apply either of these approaches in lieu of a cliff rule, the company would be required to determine the rate of tax applicable in all of the jurisdictions in which its equivalent beneficiaries reside. However, we think this exercise must be done in any event, given that the company will need to know whether each equivalent beneficiary is a resident of a country and what that country's treaty rate is.

Regardless of the approach used, it will be necessary to know what treaty rate to use where a given treaty provides multiple rates. Many U.S. tax treaties provide one rate -- usually 15% -- for most dividends, and a lower rate, often 5% or even zero, for dividends received by controlling corporations. Special rules may apply to dividends from REITs and other special types of entities. If the tested company would be eligible for a low rate because it controls the payor of the dividend, it should not be disqualified simply because its equivalent beneficiaries are individuals not entitled to that lower rate. The treaty between the United States and Germany addresses this issue by providing as follows:

 

For the purposes of applying paragraph 3 of Article 10 (Dividends) in order to determine whether a person, owning shares, directly or indirectly, in the company claiming the benefits of this Convention, is an equivalent beneficiary, such person shall be deemed to hold the same voting power in the company paying the dividend as the company claiming the benefits holds in such company.25

 

We recommend that the Proposed Model include rules consistent with the foregoing.
3. Other Issues

 

(a) Number of Equivalent Beneficiaries
The limitation of the number of equivalent beneficiaries that must together own 95% of the tested company to seven is arbitrary. We assume this rule is being proposed for reasons of administrative convenience. In our experience, many companies not created for treaty shopping purposes are owned by equivalent beneficiaries from more than seven countries. This is particularly common in Europe, where members of a family or the shareholders of a company may reside in many different countries. Given that the onus will be on the tested company to demonstrate that each of its equivalent beneficiaries qualify under the tests set forth in paragraph 4, we suggest that there is little reason to limit the number of equivalent beneficiaries that a company can have. Treasury should consider at least increasing the number of equivalent beneficiaries that a company can have to some larger number.
(b) Application of the "at least as favorable" test to non-FDAP income
We do not support the addition of an "at least as favorable" test applicable to non-FDAP treaty benefits, because we think the test is too subjective to be applied consistently. Whereas the "as least as low" test can be applied on an objective mathematical basis, the at least as favorable test might take into account any number of differences between treaties, some of which may be important and others not so important. Generally, the benefits of Articles 7 and 13 are the same in most U.S. treaties; presumably it is not required that the treaty with the country in which the equivalent beneficiary resides contain exactly the same words in those Articles. For example, some U.S. tax treaties incorporate the "authorized OECD approach" to measure the income attributable to a permanent establishment, while others do not. We question whether it should be assumed that a taxpayer would treaty shop simply to get the benefits of one approach to attributing income to a permanent establishment over another approach.

We would support the application of a narrowly-crafted "at least as favorable" rule to equivalent beneficiaries (but not to QIOs) in certain limited cases where the facts make clear that a person is engaging in treaty shopping. For example, Article 21 (Other Income) usually operates in a binary fashion -- either Other Income is taxable only in the source state, or only in the residence state. If a foreign company's business is such that a significant part of its U.S. source income is expected to be classified as Other Income, one would expect it would have an incentive to treaty shop into a country that had a treaty with the United States giving only the residence state the right to tax that income. If its equivalent beneficiaries would not be entitled to similar benefits had they invested directly, treaty shopping might be inferred. We suggest that if this rule is retained, it should be narrowed such that it applies only under circumstances in which the particular treaty benefits being sought are clearly benefits that could not be obtained by investing directly; that is, as an anti-abuse rule.

 

F. Intermediate Owners and Related Issues

 

In order to qualify as a qualified person under paragraph 2(d), all of the intermediate owners between an ultimate qualified person and the tested company must be a resident of one of the Contracting States. In order to qualify under paragraph 2(f), each intermediate owner must be a resident of the same Contracting State as the tested company. For purposes of the derivative benefits test, indirect ownership must be through one or more QIOs. A QIO is defined as a resident of any state with which the source state has a treaty that meets the tests described earlier herein.

We appreciate that some of these restrictions on how a qualifying company can be owned indirectly are not new to the Proposed Model. However, we are concerned that these tests, especially in combination with other new additions to the Proposed Model, will result in disqualifying from treaty benefits many companies not engaged in treaty shopping. Moreover, we do not think that these restrictions serve the purposes of the LOB Article.26

In public statements, officials from Treasury have sometimes stated that these restrictive intermediate owner rules are necessary to prevent double non-taxation, for example because a deductible payment may be made to an intermediate owner that is not subject to tax, or subject only to deferred tax, on such payment in its home country. This concern should not arise if such base-stripping payments would cause treaty benefits to be forfeited under any of the subparagraphs of paragraph 2 that incorporate a base erosion test. Moreover, the possibility of making such payments is equally presented where the intermediate owner is a resident of the same country as the tested company; many countries do not tax payments received by a local country entity. Further, any such base-eroding payments may be made to a sister company without implicating the intermediate owner rules.

Statements have also been made that the purpose of restricting intermediate owners is to avoid "parking" of income in a low or no-tax third country. Again, we are not sure why it would be assumed that the recipient is not subject to tax, and why this is relevant to the purpose of the LOB Article. As discussed earlier in this report in connection with Example 3, if the tested company is a foreign subsidiary of a U.S. parent, this "parking" is policed, to the extent Congress deemed appropriate, by the operation of subpart F. If instead the tested company is a U.S. subsidiary of a foreign parent, we think that U.S. anti-conduit rules should generally be able to police any abusive practices. It is only when the payment is made by a non-U.S. company to another non-U.S. company that the issue seems likely to be of concern. Even in that case, we do not think that it should be assumed, a priori, that the payment is not subject to tax (or deferred) in the payee's country of residence.

The only type of payment unique to intermediate owners that cannot be made to a sister company is a dividend, which is not deductible by the payor. If the concern underlying the intermediate owner rules relates to the fact that some countries do not tax intercompany dividends, we would observe that the exclusion of dividends is just as likely to occur where the tested company pays dividends to a parent resident in the same country, because many countries incorporate a dividend exemption. To us, it would seem that the base erosion test as applied to a tested group, as well as general anti-conduit rules should address any of the suggested concerns about the residence of an intermediate owner.

Just as a qualified person described in paragraph 2(d) is not a good equivalent beneficiary, it is also not a good intermediate owner for purposes of applying that subparagraph to the tested company. We believe that this rule may have been an oversight, and should be revisited. We note that the same rule does not appear in the LOB article of the Action 6 Report, because that version includes public company subsidiaries in paragraph 2(c).

 

Example 8. A publicly-traded U.S corporation owns more than 50% of the stock of a Dutch subsidiary through a wholly-owned U.S. subsidiary. The Dutch company cannot qualify under paragraph 2(f) because it has an intermediate owner that is a subsidiary of a public company qualified only by reference to paragraph 2(d).

 

The result in the foregoing example is certainly counter-intuitive. We do not think that the requirements enumerated in the LOB Article should force parent companies to hold their subsidiaries directly. In the great majority of cases, U.S. and other multinationals do not hold their non-U.S. subsidiaries directly, for a variety of reasons have nothing to do with treaty shopping.

We note that a good intermediate owner also excludes a company described in paragraph 2(f). We suspect that the likely concern here is that a tested company might be 50% owned by another company in the same country qualifying only under paragraph 2(f), which in turn might be 50% owned by a third company so qualifying, and etc. up a chain of ownership. In such a structure, although each intermediary company would technically meet the paragraph 2(f) test (assuming no prohibited base erosion at each level), ultimate ownership by qualified persons could be greatly diluted simply by having non-qualified persons own 50% at each level. In this case, there is a rational basis for the exclusion. If the reason that a paragraph 2(d) entity is also not proposed to be treated as a good owner was that such an entity may be as little as 50% owned by up to five public parents, we suggest that this concern be made explicit and that the rule be limited to cases in which the tested company is not wholly owned by one or more public companies or a tested group.

The Proposed Model would require that, to qualify as a QIO, the intermediate owner must be subject to tax at a rate as least as low as that applicable to the tested company under the treaty in question. We respectfully suggest that this requirement is unnecessary and if adopted will make the derivative benefits test virtually unusable in many cases where it should be available. There seems to be no reason to apply such a requirement to a person such as an intermediate owner that is not claiming benefits under the treaty in question. Only the equivalent beneficiary should be tested for treaty equivalency.

Given the above, we question whether the various restrictions on who can qualify as a QIO are more burdensome than is warranted. We suggest that Treasury consider simplifying and streamlining the various intermediate owner tests set forth in the Model.27 Consistent rules, based on easily understood and articulated policies, would make it easier for taxpayers and their advisors to understand and apply those rules. It would lead to more certainty of result, which we believe is essential in the context of an LOB Article and under a treaty generally.

In particular, if it is determined to limit the types of intermediate owners a tested company can have, we suggest that the types of intermediate owners taken into account under paragraphs 2(d) and 2(f) include a simplified and streamlined QIO approach, shorn of the STR and the "at least as good as" requirement. That is, we suggest that for all purposes, a qualifying intermediate owner should include any intermediate owner that is resident of a treaty country. The claim to benefits under paragraphs 2(d) and 2(f) seems equally as strong, if not stronger than, the claim to benefits under paragraph 4. Moreover, we conceive of paragraph 4 as, in effect, another definition of qualified person, albeit one that refers to qualified persons under treaties other than the treaty at issue. It seems peculiar that a company could qualify under the derivative benefits test of paragraph 4 by reason of having intermediate owners in various treaty countries, but be unable to qualify under paragraphs 2(d) or 2(f) because its intermediate owners must be limited as provided in those sections.

Finally, we believe that there is a technical error in the definition of equivalent beneficiary. As defined, it does not include a resident of either of the two Contracting States, but only of the State in which the tested company resides. That is, if the United States is applying the treaty to a foreign company, a resident of the United States cannot qualify as an equivalent beneficiary because, technically, he or she is not entitled to the benefits of an equivalent treaty with the United States. This error has been remedied in several U.S. tax treaties that contain a derivative benefits provision,28 and it has been remedied as well in the Action 6 Report.29

 

G. Active Trade or Business Test

 

We recommend that the proposed change made to paragraph 3, which would limit the attribution of activities from a connected person to only those cases in which both persons are engaged in the same or complementary lines of business, be eliminated from the Proposed Model. We think this change would produce clearly inappropriate results inconsistent with the purpose of the active trade or business exception. The exception is designed to accord treaty benefits to limited types of income earned within groups of connected companies, without regard to their ultimate ownership. Perhaps the classic case in which the exception is designed to apply is where a non-treaty resident owns stock of a parent company resident in Country A, say Spain, which in turns owns 100% of the stock of a subsidiary in Country B, say the United States. If the country A and B entities are part of a worldwide group engaged in the same line of active business, payments between them should be afforded treaty benefits because there is no reason to believe that the parent would be engaged in treaty shopping.

The policy of the active business rule is not limited to simple parent-subsidiary operating groups. We believe it should be irrelevant to the application and policy of the rule whether the parent conducts its business directly, or through a subsidiary located in the same country.

 

Example 9. Parent is a holding company formed and resident in Spain. It conducts the worldwide activity of making widgets. Parent owns 100% of the stock of an active U.S. subsidiary that makes widgets, and owns 100% of the stock of an active Spanish subsidiary that also makes widgets. Under the 2006 Model, the active conduct of a business in Spain is attributed from the Spanish subsidiary to the Spanish parent, such that payments between the U.S. subsidiary and the Spanish parent qualify for benefits under paragraph 3. However, under the Proposed Model, such payments would not qualify, because the parent is not engaged in the same line of business as the U.S. subsidiary -- it is engaged in no business.

 

We see no reason why the fact pattern above should not qualify for treaty benefits as it clearly does under the 2006 Model. We also believe that treaty benefits should be accorded where, instead of acting as a holding company, the parent is engaged in a completely unrelated line of business. The relevant fact is that the group as a whole is engaged in a substantial business the same as or complementary to the business conducted by the U.S. subsidiary. It should not matter whether the parent's local business is conducted directly, or through an affiliate in that country.

In Example 9, it might or might not be possible for the group to be restructured such that payments are made between operating companies, rather than to a holding company parent, in which case the active trade or business exception could apply. However, multinational groups have many non-tax reasons to own their subsidiaries in patterns that cannot easily be rearranged. And in at least one sympathetic case, it would be impossible to rearrange ownership so as to qualify:

 

Example 10. Holding G, a German holding company, is owned by members of a family who are resident in a number of countries. Holding G owns all of the stock of U, a U.S. operating company. U in turn owns all of the stock of Operating G, a company operating in Germany. U and Operating G are engaged in the same line of business.

 

In Example 10, it is not possible for U to pay dividends to Operating G, because Operating G is a subsidiary of U. Yet the group conducts an integrated active business in both Germany and the United States. Under the Proposed Model, payments from U to Holding G would not qualify for treaty benefits, because Holding G is a holding company. This is a perverse result, given that all of the German active business is conducted in a German subsidiary of U, which German subsidiary is a CFC. The United States should welcome the existence of this "sandwich" structure, which is the opposite of an inversion. It should not be denying treaty benefits to Holding G on the ground that Holding G is not an operating company.

Footnote 2 of the Proposed Model states that the derivative benefits test of paragraph 4 "sets forth the appropriate standard for determining whether a holding company or financing entity qualifies for benefits." We do not understand the relevance of the derivative benefits test to the question here presented. The point of paragraph 3 is to allow certain limited treaty benefits regardless of the ownership of the group. In many cases, the tested company will not qualify under the derivative benefits test because it is owned by a parent that in turn is owned by non-treaty country residents or persons that are not equivalent beneficiaries (or more than seven equivalent beneficiaries, or less than 95% owned by equivalent beneficiaries).30

 

H. Discretionary (Competent Authority) Relief

 

Consistent with existing U.S. treaties, paragraph 5 of the Proposed Model LOB Article provides that the competent authorities of a Contracting States may grant all or specific treaty benefits to a resident of the other Contracting State where such resident does not satisfy any of the prior paragraphs of the LOB Article. New language in this paragraph conditions a grant of relief upon the requirement that the resident demonstrate "a substantial nontax nexus to its State of residence."

We acknowledge that the purpose of a LOB Article is to ensure that the person claiming treaty benefits has a nexus to the state of its residence that goes beyond mere formation in that country.31 And we are aware that the condition is contained in Revenue Procedure 2015-40. Nevertheless, for several reasons we think that this condition should be removed from the Proposed Model, and should be discussed in a Technical Explanation as a relevant factor only.

First, we are concerned that "nexus" may be in the eye of the beholder. Perhaps the nexus language is intended to cast doubt on the treaty claims of companies that act as treasury centers, financing companies, or other businesses that are inherently "mobile." But if the nexus condition were included in a treaty, the other party to the treaty could invoke it to deny competent authority relief to a U.S. resident on the grounds that the U.S. resident has insufficient nexus to the United States. The United States taxes corporations on their worldwide income based solely on the fact that they are incorporated in the United States, without any requirement of nexus. It also taxes U.S. citizens on their worldwide income, even if they have no connection to the United States. It thus might be the case more often than not that this condition would operate to deny treaty benefits to sympathetic cases involving U.S. taxpayers.

Second, countries are increasingly asserting taxing authority over taxpayers that lack what is traditionally thought of as physical presence or nexus with the taxing state. To deny a resident access to competent authority relief on the basis of a lack of nexus, while asserting the right to tax its income, seems to us to put taxpayers in a Catch-22 situation.

Third, this is by its terms a discretionary relief provision, and as we understand it, the Contracting States generally reserve the power to grant or withhold treaty benefits in their discretion. And unlike the OECD model treaty, the Proposed Model contains no requirement that the competent authority of a state consult with the competent authority of the other state, even before it denies treaty benefits.32 Thus, if the United States wishes to reserve on grants of competent authority relief in cases where it is not persuaded that the resident has a sufficient nontax nexus to its state of residence, it is empowered to deny the relief, and does not need specific language in the treaty itself.

We have considered whether the reason that Treasury wishes to add a nexus requirement into the Proposed Model is to forestall potential treaty partners from objecting to U.S. denials of treaty benefits on nexus grounds. Treasury may also believe that by incorporating this requirement into a treaty, it will be less likely that a resident of the other country can seek judicial review of such a denial. In Starr International Co. Inc. v. United States,33 the United States argued that its refusal to accord treaty benefits under the competent authority prong of the tax treaty between the United States and Switzerland (the "Swiss Treaty") was not judicially reviewable. The court disagreed, ruling that the United States had failed to overcome the presumption that an agency action is subject to judicial review.

Many of our members believe that a denial of treaty benefits should be subject to consultation between member states, or even to mandatory arbitration. U.S. taxpayers should be protected from arbitrary denials of treaty relief by other countries. Conversely, foreign taxpayers should not suffer denial of treaty benefits without the opportunity for some type of review. This is especially the case if the more restrictive rules of the Proposed Model are incorporated into a treaty, such that discretionary relief is forced to play a greater role in accessing treaties. If an overly restrictive LOB Article has the effect of denying treaty benefits in a large number of routine cases not involving treaty shopping, taxpayers should at least have the right to judicial review of the government's denial of relief under paragraph 5.

We expect that the more restrictive the general rules of the LOB Article become, the more likely it is that the courts would insist upon judicial review of decisions to deny treaty relief under paragraph 5. Conversely, if the LOB Article is drafted narrowly, so that it picks up only abusive cases, the courts may be less likely to assert the right of judicial review. Moreover, in that case the other state should be less likely to insist upon consultation rights, which are present in the OECD Model but lacking in the Proposed Model. Therefore, if Treasury desires to limit review of decisions under paragraph 5, it should reconsider provisions designed to restrict benefits under other paragraphs of Article 22.

 

I. Time of Qualification

 

Paragraph 1 of the Proposed Model LOB Article requires that a resident be a qualified person "at the time when the [treaty] benefit would be accorded." This timing rule is new to the Proposed Model. We think it adds a degree of uncertainty that is not proportionate to any purpose for which it might have been added.

The question of timing has always been an issue under the LOB Article. The ownership and base erosion test of paragraph 2(f) continues to provide, as it has in past U.S. Model Treaties, that the ownership prong must be satisfied "on at least half the days of the taxable period." It is not clear how this retained half-of-the-days test is intended to operate in light of the requirement in paragraph 1 that a resident be a qualified person "at the time" a benefit would be accorded. If paragraph 1 is intended to mean that a resident must be a qualified person on the day that a payment is made, it would appear impossible to know whether that condition is met if the payment is made toward the beginning of a taxable year. On the other hand, it is possible to interpret the new timing language in paragraph 1 as meaning simply that, in respect of the year in question, the resident must be a qualified person. This would be consistent with the way that we believe the base erosion prong operates. Although the base erosion prong does not specify the period during which that test must be satisfied, it scans most naturally as an annual test, and this is the interpretation which we believe has been applied in practice.34

Even a test that looks to a resident's status in the year in which a benefit is accorded will not, as a practical matter, always provide guidance to a putative withholding agent. With respect to a payment made during year 1, a withholding agent cannot know whether the resident will be a qualified person during such year until after the year has closed and the various tests of the treaty can be applied. U.S. withholding practice is to require a foreign payee to certify on a Form W-8 that it is entitled to treaty benefits, and once given, such Form remains valid for three years unless the facts change in such a way that the foreign person no longer qualifies for treaty benefits. Thus, U.S. withholding agents do not normally inquire, and are not required to inquire, as to a person's qualified person status "at the time when the benefit would be accorded."35

It might be useful for the Technical Explanation to the Proposed Model LOB Article to spell out exactly what the timing rules are. For example, Treasury could provide that a party is entitled to rely on its qualification during the full taxable year preceding the year in question, at least absent some fundamental change in status.

 

FOOTNOTES

 

 

1 The principal author of this report is Kimberly Blanchard. Helpful comments were received from Andrew Braiterman, Robert Cassanos, Peter Connors, Timothy Devetski, Kathleen Ferrell, David Hardy, Stephen Land, Deborah Paul, Yaron Reich, Richard Reinhold, Michael Schler, Peter Schuur, David Sicular, Andrew Solomon, Philip Wagman and Diana Wollman. The assistance of Eric Remijan and Daniel Barron is gratefully acknowledged. This report reflects solely the views of the Tax Section of the New York State Bar Association ("NYSBA") and not those of the NYSBA Executive Committee or the House of Delegates.

2See Report No. 1327, New York State Bar Association Tax Section, Report on Certain Proposed Revisions to the U.S. Model Tax Convention (Aug. 19, 2015).

3See United States Model Income Tax Convention (November 15, 2006), Art.22.

4See 145 Cong. Rec. S. 14225 (1999) (reserving approval of the Italian and Slovenian tax treaties on the exclusion of the "main purpose" test from such treaties).

5Preventing the Granting of Treaty Benefits in Inappropriate Circumstances, Action 6: 2015 Final Report (OECD/G20 Base Erosion and Profit Shifting Project, October 2015) (hereafter, the "Action 6 Report").

6 Technical Explanation to Article 22 of the 2006 U.S. Model Treaty.

7 Technical Explanation to Article 22 of the 1996 U.S. Model Treaty.

8 Id.

9 Section 884(e) contains an ownership/base erosion test and a special test for public companies. The anti-treaty shopping purpose of these rules is clear from the legislative history: "Congress was concerned that foreign persons resident in one country would attempt to use another country's income tax treaty with the United States to avoid the branch profits tax. The bill addresses this concern by not allowing income tax treaties to prevail in treaty shopping cases." Committee Report 8841.009 (Nov. 10, 1988). The public company test was based on the "presumption that a corporation whose stock is primarily and regularly traded on a local securities market is more than 50% owned by local residents." Id.

10 6 Report at ¶ 25, Commentary § 3.1.

11 Technical Explanation to Article 22 of the 1996 U.S. Model Treaty.

12See also the legislative history of Section 884(e) cited in note 9, supra.

13 Even today some countries and advisors from those countries continue to insist that U.S. charities and pension funds cannot qualify as residents under what is typically Article 4 of U.S. tax treaty because they are not "subject to tax." This literal reading of the subject to tax test, with which the United States disagrees (see, e.g. the Technical Explanation to Article 4, paragraph 2, of the 2006 Model), does not take into account the proper function of the subject to tax test as a beneficial ownership rule.

14 See, e.g., Action 6 Report at ¶ 20: "The LOB rule . . . only focuses on treaty shopping and does not address other forms of treaty abuses; it also does not address certain forms of treaty shopping such as conduit financing arrangements." The latter half of this statement is incorrect, as demonstrated by Example 2 above.

15 U.S-Luxembourg Treaty Article 24(e).

16 At ¶ 26, Commentary note 8.

17 For further elaboration of this point, see Report No. 1113, New York State Bar Association Tax Section, Report on Limitation on Benefits Provisions and Section 1(h)(11) (June 26, 2006) (the "2006 Report").

18 At ¶ 29.

19 The Action 6 Report does not incorporate a base erosion rule for public company subsidiaries.

20See n. 2, supra.

21 2006 Report, p.8.

22 We are not commenting on Treasury's ability to reach abusive cases under current law and treaties.

23 2006 Report at n. 16.

24 After initially announcing that it would not participate in Action Item 15, the United States recently reconsidered and is exploring the option.

25 A similar rule in included in the Article 6 Report at ¶ 78 in the definition of equivalent beneficiary.

26 Apparently many other countries agree; at several points in the Action 6 Report, it is noted that other countries would not so restrict intermediate owners.

27 The version of the LOB in the Action 6 Report would, for example, eschew the term QIO in favor of requiring that intermediate owners be equivalent beneficiaries. Although the two terms differ in that the latter must be qualified residents under one of the enumerated subparagraphs in paragraph 2, they are in most respects parallel.

28 See, e.g., U.S.-France Treaty Article 30(7)(f)(ii).

29 Action 6, ¶ 78, definition (f)(ii).

30 It is possible that the concern may relate to the scope of what it means for income to be "in connection with or incidental to" the trade or business. If so, this point should be addressed directly by fleshing out the meaning of that language rather than restricting access to the entirety of Article 3.

31 Traditionally, however, the United States has not been particular about nexus where it is the other Contracting State that is being treaty-shopped. For example, in CCA 201343019, the IRS accorded treaty benefits to a company resident in Cyprus, even though that company had no nexus to Cyprus -- it was formed by third country residents in order to gain benefits under the treaty with Cyprus and that third country.

32 We note that in the recent Starr International decision, discussed below, the court implied that it "might" question an OECD-like provision that required the United States to consult with the other state before denying a claim for relief. That decision involved the United States' treaty with Switzerland, which required the United States to consult with Switzerland only in the context of granting relief.

33 116 AFTR 2d 2015-6180 (DC Dist. Col. September 18, 2015).

34 See the 2006 Report.

35 The Action 6 Report contains a timing rule identical to that of paragraph 1 of the Proposed Model. As far as we are aware, no other country operates a withholding tax system along the lines of the one used in the United States, in which withholding agents effectively enforce the rules, including the treaty rules. In most other countries, treaty relief is granted by application to the government itself, such that the timing rule would seem to be easier to deal with in those countries.

 

END OF FOOTNOTES
DOCUMENT ATTRIBUTES
Copy RID