This article was originally published in Limiting Treaty Benefits, May 2010

It may be time to take a fresh look at the Limitation on Benefits article in U.S. income tax treaties. The LOB rules are intended to provide a means of determining whether a particular resident of a treaty partner country has a sufficiently robust connection with that country so that it can be deemed appropriate to grant that person tax benefits in the United States, the country of source, in accordance with the treaty's terms. A variety of tests have been employed for this purpose in U.S. treaties. In addition, there is broad tolerance for treaty partner entities owned by "equivalent beneficiaries." Within these wide confines there are many specific policy choices worthy of being teased out and examined.

In their modern incarnation LOB provisions cover individuals, Contracting States, political subdivisions, local authorities, pension funds, and other exempt organizations. In addition, the competent authority of the source country is authorized, on a discretionary basis, to grant treaty benefits upon a finding that the establishment and maintenance of a person claiming benefits did not have obtaining the benefits as a principal purpose.

The challenge of the LOB provisions pertains not to the foregoing provisions but to those applicable to entities — principally companies, but other taxable entities as well — and the normal commercial situations in which entities operate. Some treaties, such as the recently concluded treaty with Hungary, contain a headquarters company concept, but the main and generally applicable entity provisions are as follows:

  • regular trading of company shares on a recognized stock exchange in the country of asserted residence (the "locally traded" test);
  • regular trading of company shares on any recognized stock exchange if management and control is in the country of asserted residence (the "traded and locally managed" test);
  • ownership of a company, directly or indirectly, to the extent of at least 50 percent by five or fewer companies meeting the locally traded or the traded and locally managed test (the "indirectly traded" test);
  • ownership of an entity, directly or indirectly, to the extent of at least 50 percent on at least half the days of the taxable year, by persons entitled to treaty benefits themselves, provided the entitlement derives from the locally traded test or the traded and locally managed test, or the persons are what might usefully be termed "favored residents": individuals, government entities, pension funds, exempt organizations; and provided further that less than 50 percent of the entity's gross income is paid in deductible form to persons not entitled to benefits of the treaty on these grounds (the "ownership/base erosion" test);
  • maintenance of an active and substantial trade or business in the country of asserted residence, provided the income for which benefits are claimed is associated with that trade or business (the "trade or business" test);
  • ownership of a company, directly or indirectly, to the extent of at least 95 percent by seven or fewer persons who are residents of the United States or the treaty partner and entitled to treaty benefits as favored residents, under the locally traded or the traded and locally managed test, or under the ownership/base erosion test; or by persons who would be entitled to benefits at least as great as those claimed under analogous tests in certain other treaties if the income in question was paid to them in their country of residence; provided, however, there is no erosion of the local tax base of the company claiming benefits (the "equivalent beneficiary" test).

Certain treaties, the one with Hungary being an example, also contain a "triangular" provision covering a permanent establishment in a third country. This provision requires that treaty benefits will be granted only if the profits of the permanent establishment are subject to a combined effective rate of tax in the (treaty partner) country of residence and the third country equal to no less than 60 percent of the general rate of company tax in the country of residence.

This summary does not do justice to any of these tests, but should suffice to stake out the terrain for present purposes. Comments on the tests will be offered below in the discussion of particular problems, issues, and anomalies.

At an early stage of development of the LOB provisions the United States realized that mere stock ownership by local persons would not be sufficient to guarantee a meaningful tax nexus of an entity with the treaty partner country. It would be possible for the local tax base to be depleted through deductible expenses paid to persons having no connection with that jurisdiction, deflecting the "economic ownership" of treaty benefits to a third country. The "base erosion" provisions of modern LOB articles take aim at this practice, impeding the use of a treaty by third-country residents for the purpose of investing in the United States.

It required years for the United States to grasp that the base erosion test was aimed at protecting the tax base in the treaty partner country and that base erosion — when a substantial portion of an entity's gross income is siphoned off through deductible payments to a third country — should therefore be linked to the treaty partner's notion of gross income, not to some hypothetical U.S. view. The test is clear now, though perhaps not always so clear in application. There will always be issues when the United States purports to determine how another country's tax laws work in practice, and it may not be possible to force a country to collect a tax that it does not wish to collect.

The LOB provisions in the treaty with Hungary contain two base erosion provisions, one for the ownership/base erosion test and one for the equivalent beneficiary test. They are not the same, and it is fair to ask why. In the ownership/base erosion test there is a specific exclusion from base erosion for payments in respect of financial obligations to a bank (i.e., interest on a borrowing). This reference is not found in the equivalent beneficiary test. Furthermore, the test in the equivalent beneficiary context is geared to deductible payments to persons other than equivalent beneficiaries, not to all payments to persons not entitled to the benefits of the treaty with Hungary (as in the ownership/base erosion test). Thus, the tax base of the treaty partner resident can be stripped to zero if there is ultimate ownership by equivalent beneficiaries, as long as the stripping is in favor of equivalent beneficiaries.

Another "base erosion" question is what represents tolerable erosion, in the U.S. view. The corporate rate in Hungary is 20 percent. Since the ownership/base erosion test and the equivalent beneficiary test accept a stripping of up to 50 percent of the tax base, the United States seems prepared to live with a 10 percent corporate tax take for entities qualifying under these tests. In the treaty's "triangular" provision, when a Hungarian enterprise has a permanent establishment in a third country, benefits are withheld unless the effective rate of tax is at least 60 percent of the general rate, or 12 percent. In the case of traded companies and the indirect trading test, there is no base erosion test, so the United States apparently accepts a stripping of the entire residence country base in these circumstances. Ten percent, 12 percent, zero — what is guiding these choices? The triangular provision is especially hard to fathom because, even if treaty benefits are allowed there, the provision allows greater taxation than the treaty's general dividends, interest, and royalty articles.

Finally, the base erosion concept applies to amounts paid or accrued "directly or indirectly" to persons not qualifying for treaty benefits. That sounds great. What does it mean? Is this a concern with back-to-back payments, for example of interest, or is it something broader? It is hard to imagine that the United States is going to be concerned if a Hungarian company pays interest or other deductible amounts to an individual resident of Hungary who then pays some other kind of deductible expense to someone outside Hungary. In any event, how would that be enforced? If the aim is merely to forestall abuse, it would seem preferable to include general language to that effect and avoid the facile but confusing "directly or indirectly" concept.

Base erosion provisions are aimed at ensuring that the country of asserted residence collects an amount of tax that is not substantially lower than the normal or expected amount because deductions are used to reduce the local tax base in favor of persons resident elsewhere. What happens, however, if the residence country employs credits, rather than deductions, as a means of reducing its tax? It is well to remember that the LOB article is aimed at countries that lend themselves to conduit use, indeed that invite conduit use, and that the LOB policy is, in situations where it matters most, contrary to the perceived interests of those countries . Take Hungary. The country turned itself into a tax haven of sorts, courting investors from Latin America, Canada, and elsewhere eager to enjoy treaty benefits even though resident in countries lacking treaties, or lacking favorable treaties, with the United States. Is a base erosion provision sufficient here? Suppose Hungary was to enact a law that gave credits to Hungarian companies owned by foreign investors. Would that not fly in the face of the U.S. policy that lies behind the limitation on benefits article? So why does the article, as it appears in the treaty with Hungary, lack any provision aimed at reduction of Hungarian tax through credits? The answer may be simply that we have not yet witnessed any country taking such a step, and we find comfort in the implicit threat to terminate the treaty if such a step was taken. Experience suggests, however, that it is not so easy to terminate a U.S. tax treaty.

The policy choices that have been made with respect to intermediate entities — entities situated between the entity claiming benefits and the ultimate investor — are also mysterious. The treaty with Hungary has three different tests for residence of intermediate entities. For a Hungarian company that claims treaty benefits on the basis of the "indirectly traded" test, treaty benefits are available if it is ultimately owned, to the extent of at least 50 percent, by five or fewer companies that meet either the locally traded or the traded and locally managed test (presumably on a mix and match basis), but each intermediate owner situated between the claimant and the qualifying company or companies must be a resident of either the United States or Hungary. Thus a Hungarian company whose shares are indirectly owned through three tiers of U.S. companies by a top-tier Hungarian resident company whose shares are locally traded qualifies under this test. So would a Hungarian company indirectly owned by a top-tier U.S. company meeting the locally traded test. There is no requirement that qualification of a Hungarian claimant depends on residence of the ultimate investor or investors in Hungary.

For a treaty partner resident claiming benefits on the basis of indirect application of the ownership/base erosion test, the approach is different. If the claimant (including a company) is ultimately owned by qualifying treaty partner owners (ultimate U.S. ownership will not work here), the intermediate owners must all be residents of the treaty partner as well. U.S. residence of intermediates will disqualify. This is a tighter test but, at least to this reader, it is not clear why. Perhaps this is simply a matter of the greater trust afforded to companies. The ownership leg of the ownership/base erosion test can be met by residents other than companies. If the top tier investor is a company (qualifying under the locally traded test or the traded and locally managed test), the only situation that falls outside the indirectly traded test (and therefore must meet the more stringent ownership/base erosion rule for residence of intermediate owners) is one involving a claimant other than a company. Thus, the common denominator for situations falling within the tighter intermediate owner provision of the ownership/base erosion test is that these situations involve either ultimate investors or claimants who are not companies.

When it comes to equivalent beneficiaries, the policy is much more liberal. "Indirect" qualification under the equivalent beneficiary test comes with no limitation on the residence of intermediate owners. A treaty partner company owned by a Cayman company that is, in turn, owned by an equivalent beneficiary is just fine.

In the treaty with Hungary the term "equivalent beneficiary" includes a resident of either Hungary or the United States if that resident meets either the locally traded test or the locally managed and controlled test, or is a favored resident. This provision seems to undermine more restrictive rules in both the indirectly traded test and the indirect ownership/base erosion test, which contain specific requirements for the residence of intermediate owners.

True, there is a base erosion rule in the equivalent beneficiary test, and not in the indirectly traded test, so there is an additional requirement to be met if reliance is placed on the equivalent beneficiary test for treaty qualification. But the upshot is that a traded company (whether traded locally or managed locally, in the treaty partner or in the United States) can invest in the United States through intermediate owners and a treaty partner company in either of two ways — intermediate companies that are all residents of either the United States or the treaty partner, with no need to prove the claimant's lack of base erosion; or intermediate entities resident anywhere, as long as the claimant can show that its tax base has not been eroded. It is worth underscoring that deductible payments to equivalent beneficiaries do not represent base erosion in the equivalent beneficiary version of the base erosion rule. Furthermore, there is no explicit base erosion provision for deductible payments by equivalent beneficiaries. Arguably, an entity whose base is eroded under the treaty of its country of residence would fail to qualify as an equivalent beneficiary (though one can envision the administrative difficulties of ascertaining the pertinent facts). Even if that is the case, however, the rules allow for a "cascade effect" under which a claimant could pay out 49.9 percent of its income as deductible expenses to third-country residents (for example, intermediate owners) and the remainder to equivalent beneficiaries, who would pay out 49.9 percent of their gross income to third-country residents, and so on. The intermediate owners, which need not be residents of treaty partners of the country of source, can of course have a tax base eroded to zero.

Finally, the entire policy of offering U.S. treaty benefits to treaty country entities owned by equivalent beneficiaries stands in contrast to the policies expressed in section 894 of the Internal Revenue Code. The statute focuses on whether a treaty partner views benefited income as income of its resident, an inquiry that determines whether U.S. tax benefits are granted. The idea is that a treaty-based reduction of U.S. tax at source is only justifiable for a recipient who is a being taxed as a resident by a treaty partner. In contrast, the equivalent beneficiary provisions ask whether the U.S. benefits in question could have been claimed if the ultimate investors had invested directly in the United States from their country of their residence. This is a different focus. The equivalent beneficiary test reflects no concern for taxation by the country of the claimant's residence, since base erosion down to zero is tolerated for payments to equivalent beneficiaries. Instead, the equivalent beneficiary test is concerned, exclusively, with whether the United States is being asked to grant more benefits than it would have granted if the investment had been made directly. That is hardly the design of section 894.

The difficulty with intermediate owners and equivalent beneficiaries, as a matter of policy, is that they tend to destroy the coherence of any conceivable U.S. goal expressed in the LOB article. What policy, for example, justifies the granting of treaty benefits to a company owned through intermediate entities by a company whose stock is traded on a recognized stock exchange when the income of the claimant is largely paid out in deductible form to residents of third countries? Even if a case could be made for the locally traded or locally managed treaty partner resident company that ultimately owns the company claiming treaty benefits, what justifies a similar result when the company with ultimate ownership is a resident of the source country? It would appear from Article 22 that traded companies are deemed trustworthy insofar as the stripping of the tax base is concerned. What is the evidence that justifies that conclusion?

The trade or business test also evinces no concern for base erosion. Perhaps this position rests on the proposition that the existence of an active trade or business serves to negate any improper purpose for the acquisition or maintenance of the treaty country entity, serving the same function as payment of a normal residence country tax. Local trading, and local management of a company whose shares are traded elsewhere, fill the same role — indicating a purpose other than simple borrowing, or "abusing," the treaty. These latter tests, however, are undermined by tolerance of intermediate owners and equivalent beneficiaries, particularly when these interlopers are welcomed under differing, but overlapping, tests for acceptance.

The ultimate question, it seems to me, is whether the United States is serious about confining the "use" of its treaties to persons who have a substantial tax-based connection with the country where they claim residence. If so, a rethinking of the convoluted and formulaic LOB concept appears warranted. If not, then perhaps we should stop deluding ourselves and deceiving the rest of the world with largely ineffective provisions of this sort.

This article is designed to give general information on the developments covered, not to serve as legal advice related to specific situations or as a legal opinion. Counsel should be consulted for legal advice.