13 June 2018
On 20 May 2015, the US Treasury Department released proposed revisions to the US Model Income Tax Treaty (the 'US Model'), which (i) introduce 'special tax regime' provisions, (ii) propose changes to the Limitation on Benefits ('LOB') provision; (iii) add new rules on 'expatriated entities'; (iv) include a right for partial treaty termination if a treaty partner makes certain 'subsequent changes' to its domestic law; and (v) target 'exempt permanent establishments'.
The proposals were released as part of the US' bid to influence the OECD's discussions regarding combatting treaty abuse as part of the 'Base Erosion and Profit Shifting' project and represent some significant changes to US treaty policy, focused on limiting base erosion, double non-taxation and treaty qualification. This article summarizes material aspects of the proposals and highlights some of the questions and potential challenges that could arise if they are incorporated into existing US treaties as currently proposed.
'Special tax regimes'
The 'special tax regime' provisions introduce a new concept to the US Model that would deny treaty benefits on interest, royalties and 'other income' if a preferential rate of tax applies to that income in the recipient jurisdiction. While the policy reasons behind the provisions (eg, eliminating double non-taxation) are understandable, the draft provisions raise some significant questions. For example, although the provisions deny treaty benefits on payments to 'related' recipients, there is no definition of 'related' for this purpose so it currently is unclear how closely connected the payor and payee must be.
Further, it is unclear whether the provisions are intended to deny treaty benefits with respect to an item of income regardless of whether that item of income in fact benefits from a special tax regime. A 'special tax regime' is defined as any legislation, regulation or administrative practice (including a ruling practice) that provides a preferential effective rate of tax on interest, royalties or other income, including through reductions in the tax rate or the tax base; notional interest deductions are always considered a special tax regime. However, there could be several reasons why an item of income theoretically could be able to benefit from a special tax regime but would not do so in practice. For example, if a special tax regime is elective, items of income paid to recipients who have not elected into the regime could still be denied treaty benefits if the treaty provisions are intended to apply solely based on the existence of the regime, rather than its application in particular circumstances. Similarly, if the special tax regime is a ruling practice and the related recipient has not applied for a ruling, should treaty benefits still be denied as a matter of policy? This would appear to be unduly harsh.
The special tax regime provisions also could pose significant due diligence burdens, particularly given that the US withholding tax regime relies on withholding agents to enforce the provisions (and those agents may be liable for any under-withholding). It may be difficult or impossible for those withholding agents to determine whether the special tax regime provisions apply without additional certification (for example, on a Form W-8BEN-E) by the recipient upon which the withholding agent could rely. This identification issue potentially is compounded, since under the current proposal, no notification or other public notice is required if a jurisdiction adopts a special tax regime after the treaty has entered into force.
Material changes to the LOB provision include: (i) the addition of a 'derivative benefits' provision; (ii) including a 'tested group' requirement in the 'base erosion' prong of all three 'base erosion' tests (including a new base erosion requirement in the 'subsidiary of a public company' test); and (iii) changing the active trade or business test with the practical effect that holding and finance companies could no longer qualify under this test.
A typical LOB article in existing US treaties contains several possible routes under which resident entities can qualify for treaty benefits, one of which is the 'derivative benefits test'. The proposals would add the derivative benefits test to the US Model for the first time, although with significant differences compared to provisions in existing treaties. Broadly, under the proposed derivative benefits test, a tested resident entity qualifies for certain treaty benefits: (i) it is at least 95%-owned, directly or indirectly, by seven or fewer 'equivalent beneficiaries', each of which is entitled to US treaty benefits at least as advantageous as those claimed by the tested entity; and (ii) less than 50% of the gross income of both the tested entity and its 'tested group' is paid or accrued in deductible payments (other than certain arm's length payments in the ordinary course of business) to persons that are either not equivalent beneficiaries or are equivalent beneficiaries that benefit from a 'special tax regime' with respect to the deductible payment.
This test incorporates a number of important nuances. First, an equivalent beneficiary must qualify for benefits as an individual, publicly-traded company, governmental entity, exempt organization or pension fund under a treaty with the source state (ie the US) that contains a comprehensive LOB provision. Further, the equivalent beneficiary must be entitled under its treaty with the US to a withholding rate on dividends, interest and/or royalties that is 'at least as low' as the rate under the tested entity's treaty and/or to benefits 'at least as favorable' as those of the tested entity under the business profits, gains and other income provisions (depending on what treaty benefits the tested entity is claiming). These tests appear to operate as an 'all or nothing' proposition; for example, if under the respective treaties an equivalent beneficiary is entitled to a 10% rate on dividends and the tested entity is entitled to a 5% rate, no reduced rate is available and full US withholding tax at 30% applies (rather than applying the higher of the two rates, 10% in this case). The policy behind this 'all or nothing' proposition is not entirely clear.
In one liberalizing change, the 'ownership prong' in the proposed test does not impose a geographical limitation on the equivalent beneficiary's residence; the tests in existing treaties require the equivalent beneficiary to be resident either in the same jurisdiction as the tested entity or in a jurisdiction within the tested entity's economic bloc (eg NAFTA, the EEA or the EU). However, the proposed test would add a new requirement that each intermediate owner of the tested entity must be a 'qualifying intermediate owner' or QIO. A QIO is a resident of a jurisdiction that has a treaty with the source state that includes 'special tax regime' provisions; under the current proposal, a QIO cannot be an entity in the source state although the policy behind this is not clear (if in fact it was intended).
The requirement that an intermediate owner's residence jurisdiction has a treaty that includes special tax regime provisions is particularly problematic since (of course) at this stage no US treaty has these provisions. Retaining that requirement, therefore, effectively would preclude any tested entity with intermediate owners from qualifying under the derivative benefits test. Even if the requirement is retained (for example, coming into force once other treaties have adopted special tax regime provisions), should treaty benefits be denied if a tested entity's intermediate owner is resident in a jurisdiction with a special tax regime if the particular item of income does not benefit from the regime? Further, the special tax regime aspect of the QIO test is not limited to payments to related payees (unlike the special tax regime provisions themselves), which may lead to significant practical difficulties in determining whether an entity qualifies for benefits under this test depending on the status of its payees.
The proposed 'base erosion' prong also reflects substantial changes compared to those in existing treaties and would apply to all three base erosion tests under the revised LOB. Under the proposal, both the tested entity and its 'tested group' must meet the base erosion test; a tested group includes the tested entity and any 'intermediate owner' that is resident in the same jurisdiction as, and part of a tax consolidation or similar group with, the tested entity. As the definition requires an intermediate owner, a parent company does not have a tested group although the policy reasons for excluding sister companies and subsidiaries from the tested group definition is not clear. Base eroding payments include deductible payments made to (i) persons that are not equivalent beneficiaries, or (ii) equivalent beneficiaries (related or not) who benefit from a special tax regime with respect to the payment. Further, 'gross income' under the base erosion tests generally excludes dividends that are exempt from tax in the tested entity's state of residence (other than when testing for qualification with respect to the dividends article).
Although eliminating the geographical restriction in the equivalent beneficiary definition is helpful, the new limitations introduced by the QIO requirement and narrower base erosion test likely will make the proposed derivative benefits test more difficult to meet. The addition of a base erosion test to the 'subsidiary of a public company' LOB test also may pose a significant obstacle to treaty qualification for these entities, and it is not clear that there is a sound policy reason for this further limitation.
Another material change to the active trade or business test would prohibit attribution among related entities, unless the resident and related entities were engaged in the same or complementary lines of business. This change effectively would preclude holding or finance companies from qualifying under this test (since the holding company, for example, would not be engaged in the same line of business as its operating subsidiary). Treasury's comments to the proposed active trade or business test indicate its view that the more appropriate LOB qualification route for holding and finance companies is the derivative benefits test but, particularly given the significant changes to that test, this does not seem likely. Treaty qualification of holding and finance companies, therefore, may be much more difficult in practice.
The proposed changes include new provisions on 'expatriated entities' as a backstop to existing (and likely new) domestic rules on 'inversions'. Very broadly, in an inversion, a US corporation with multinational operations combines with a foreign corporation, which becomes the new parent of the group. Post inversion, absent rules to the contrary, the non-US parent is only subject to US federal income tax on its US source income.
The proposed expatriated entities provisions would deny otherwise available reduced treaty withholding rates and impose full US withholding tax at 30% on payments of dividends, interest, royalties and 'other income' if the payor is an 'expatriated entity'. The denial of reduced US withholding rates would apply for the ten-year period following the inversion, beginning on the date the 'acquisition' of the US corporation is completed.
The proposed provisions would supplement existing anti-inversion rules under the US Internal Revenue Code by effectively applying if, post-inversion, the foreign parent was owned more than 60% but less than 80% by the pre-inversion owners of the US company. Critically, the denial of reduced US withholding rates under these proposals is not limited to payments to related parties, although Treasury has suggested this limitation is under consideration; in other words, as currently drafted, reduced treaty withholding rates would be denied if, post-inversion, the US company made covered payments to an unrelated party (for example, interest payments to third party lenders).
Another proposal would add a new article permitting partial treaty termination if, after the treaty is signed, either (i) the general rate of company tax applicable in one of the treaty partners falls below 15%, or (ii) a treaty partner exempts its resident entities from tax on 'substantially all' of the entities' foreign source income (with similar provisions for individuals). The proposal would require 'generally available deductions' to be taken into account in determining the applicable rate of tax. If a treaty partner determines that either condition has been met by the other jurisdiction, the first treaty partner can notify the second through diplomatic channels that it will stop applying the provisions of the dividends, interest, royalties and 'other income' articles to residents of the partner jurisdiction. Generally, the partial termination would take effect 6 months after written notification unless the treaty partners resolve the situation.
It is understandable that a treaty partner would want to re-consider treaty provisions if the partner jurisdiction significantly changed the principles upon which the original treaty was negotiated. However, the provision as drafted raises some concerns. First, it is not clear that a fixed 15% company tax rate is the appropriate measure (particularly since some jurisdictions generally already have lower applicable rates); another option could be to calculate the trigger based on a percentage reduction from the rate in effect when the treaty was signed.
Further, the requirement that 'generally available deductions' or 'other similar mechanisms' be taken into account in determining the effective rate of tax appears to be unduly complex and raises questions such as whether US tax principles or the tax law of the local jurisdiction should apply for purposes of determining what constitutes income or deductions (the latter would be much more sensible).
Any partial termination is reciprocal so residents of the 'non-offending' treaty partner also would be denied treaty benefits. While it is understandable that treaty partners would not be interested in signing a non-reciprocal 'subsequent changes' provision, reciprocity could leave US companies in a difficult position if a treaty partner in which they have operations changed its law and the US company was denied treaty benefits on income from the treaty partner. It presumably also will be necessary to have a publicly-available system identifying partially terminated treaties so withholding agents are able to properly comply.
Exempt Permanent Establishments
Finally, the proposals would add a new section to the general scope article that would address certain income received through a permanent establishment ('PE'). Broadly, if a treaty resident receives income from the other treaty partner that is attributable to a PE outside the residence jurisdiction, treaty benefits will not apply to that income if either (i) the profits from the PE are subject to a combined aggregate effective rate of tax in the jurisdiction of residence and the PE jurisdiction of less than 60% of the generally applicable company tax rate in the residence state, or (ii) the PE is in a jurisdiction that does not have a comprehensive treaty with the source state (unless the treaty resident includes the income in its tax base).
One immediately apparent concern is that the draft technical explanation indicates that the principles of Section 954(b)(4) of the US Internal Revenue Code should apply for purposes of determining the combined aggregate effective tax rate under the first test. Section 954(b)(4) excludes income that is subject to 'high foreign taxes' from 'Subpart F' income under the US 'controlled foreign corporation' rules. However, this rule generally applies US tax principles when calculating the effective tax rate, which likely are different from the principles upon which a treaty partner resident would calculate its effective tax rate under local law. A requirement to determine the 'aggregate effective tax rate' applicable to the profits of the PE under US tax principles seems to be an unnecessarily complex approach (as with the 'subsequent changes' provision).
These proposals represent significant changes to the US Model that ultimately may make treaty qualification more difficult and could significantly limit treaty benefits. Although many of the policy reasons for the proposals are understandable, the proposals as currently drafted raise some significant concerns regarding interpretation, as well as practical hurdles that must be overcome in order for the relevant parties (including treaty claimants, advisors and withholding agents) to accurately identify and comply with their obligations. Treasury requested comments on the proposals and many areas of concern have been identified but the extent to which any of these concerns are addressed in the final provisions and/or the accompanying technical explanations remains to be seen.