Tricky Derivatives
By James J. Tobin, Esq.
Ernst & Young LLP, New York, NY
*The views expressed herein are those of the author and do not necessarily reflect those of Ernst & Young LLP.
Treaty shopping continues to be a hot topic around the world and
one where there seems to be an increasing enforcement focus. We've
seen court cases and rulings involving treaty capital gains issues in
China, India, and Korea, among others, and we have seen increased
audit review activity in many countries around the globe. The focus of
attack in some of these cases has been on the “substance”
of the treaty entity, implying an expectation on the part of some tax
authorities that the treaty jurisdiction entity should perhaps have
some unspecified number of employees and some undefined amount of
independent business activity. Given that treaties are heavily
negotiated bilateral agreements that, to the best of my knowledge,
rarely if ever include any explicit “substance”
requirements apart from requiring that the treaty country recipient be
the beneficial owner of the item of income in question, additional
unstated subjective requirements for eligibility seem to me
unnecessary and inappropriate.
Rather than a general commercial substance requirement, the U.S.
treaty approach is to include in all U.S. treaties a limitation on
benefits (LOB) provision, which requires either ownership or business
connection in the treaty partner country as a prerequisite for
obtaining treaty benefits. An interesting feature of some, but not
all, of the U.S. LOB provisions is a derivative benefits test that
allows treaty access even in the absence of local ownership or local
business activity. While providing a nice objective and reasonable
test for determining treaty qualification, when one spends time
studying these derivative benefits provisions on a treaty-by-treaty
basis, what's striking is the lack of consistency across provisions,
the complexity of the rules, and the inevitable traps for the
unwary.
For those who are not aficionados of the derivative benefits
provision, let's start at the beginning. The concept in principle is
that if the treaty resident company in question is owned directly or
indirectly by residents of another treaty country that would have been
entitled to the same level of U.S. treaty benefits if they had
received the U.S. income directly instead of through the treaty
resident, there would be no potential for treaty shopping abuse and
the taxpayer in question should be entitled to the benefits of its
treaty of residence. In my view, this is a very reasonable and
appropriate approach and it implicitly recognizes that the strict LOB
tests shouldn't be blindly applied because there are many commercial
or foreign tax reasons for the use of third-country business or
investment entities that are in no way motivated by U.S. treaty
shopping objectives.
So what's the U.S. approach for defining what level of equivalence
in terms of treaty treatment is necessary to qualify under the
derivative benefits test as an alternative formulation for
satisfaction of an LOB article? Equivalent or not equivalent--that is
the question. Unfortunately, there is not one answer to this question.
Indeed, there are almost as many answers as there are treaties that
contain a derivative benefits provision. This situation likely is
exacerbated by the fact that there is no derivative benefits provision
in the U.S. model tax treaty. Therefore, it is not
“standard” U.S. treaty policy even though such a provision
is now included in most recently negotiated U.S. treaties. And
apparently the lack of a model has contributed to continued
“tweaking” by our treaty negotiators of the details of a
derivative benefits provision, tweaking that in many cases seems to be
without particular rhyme or reason.
The general framework of the most recent derivative benefits
provisions contains four primary aspects:
• A
specified percentage of direct or indirect ownership (usually
95%);
• That
is concentrated in the hands of a limited number of owners - seven or
fewer;
• That
themselves are equivalent treaty beneficiaries - usually from a
country within a regional trading block;
• Coupled
with satisfaction of a base erosion test.
This conceptual framework seems fine, although one could question
the various thresholds - why such a high ownership percentage
requirement, why such a tight ownership concentration requirement,
etc. However, the bigger issue is the seemingly random variation from
treaty to treaty with respect to one or more of the above criteria. It
creates needless traps and inconsistent results without any logical
tax policy underpinning. The limited list below of examples of some of
these differences illustrates the point. But it would take a lengthy
treatise to do a complete analysis of all of the differences, both
minute and more monumental, which exist from one derivative benefits
provision to the next.
The Specified-Percentage-Ownership and Concentration
Requirement
Ninety-five percent is the somewhat standard threshold requirement
of ownership by equivalent beneficiaries. However, it is interesting
to note that two early LOB provisions took a different approach. The
original derivative benefits provision in the Dutch and Swiss treaties
required at least 30% local country ownership and at least 70%
European or NAFTA owners with equivalent benefits. This forerunner of
the now common derivative benefits provision did not impose a limit on
the number of equivalent owners that together must satisfy the
threshold ownership requirement. Both of these treaties now include
the seven-owners-owning-95%-or-more requirement as an alternative to
the original 30/70 threshold. Interestingly, in the Swiss treaty the
95% derivative benefits rule is included in the Memorandum of
Understanding (MOU) to the treaty and not in the treaty text itself,
and it is styled as a safe harbor for satisfying the Competent
Authority discretionary relief provision of the LOB article. In the
Dutch treaty, the 95% test is part of the treaty LOB article, but the
technical explanation indicates that satisfying the old 30/70 test but
not the 95% threshold will be deemed to satisfy the discretionary
relief standard in that treaty.
The other notable exception to the 95% standard is the recent
U.S.-Canada treaty, which lowers the threshold to 90% for our friendly
northern neighbors.
The Equivalent Beneficiaries Requirement
The biggest differences here relate to the countries in which the
ultimate shareholders must reside in order to be eligible to be
considered equivalent beneficiaries. In this regard, it would help if
one had been paying close attention back in geography class. The
general tendency is to limit the qualifying residence countries to
regional trading blocks, but still those regional trading blocks vary
by treaty. So who can name the countries in the European Union (EU)
versus the European Economic Area (EEA) versus the European Free Trade
Association (EFTA)? Extra credit if you can explain why there are
three different groupings (other than to add excitement to the
derivative benefits determination).
Some of the U.S./European treaty derivative benefits provisions are
limited to EU and NAFTA residents as acceptable owners (the Luxembourg
treaty is an example). Other European treaties also expand acceptable
ownership to encompass countries that are part of EFTA, which includes
Iceland, Norway, and Switzerland (the EFTA also includes Liechtenstein
but obviously there is no U.S. treaty likely there anytime soon so
“equivalent” status won't apply); the Iceland treaty is an
example that gives this favorable treatment to EFTA countries. Some
other treaties, such as the U.K. and German treaties, add the EEA
members as good ultimate residence countries (the EEA includes Norway
and Iceland but doesn't include Switzerland). The Belgium treaty
includes EU plus EEA and adds Switzerland, which I think is
practically the same as saying EFTA.
The most significant expansion in the “good country”
list approach is our recent treaty with Canada. That treaty contains
no limit on the potentially eligible countries and does not pay lip
service to any regional trading block concept, probably because
confining the derivative benefits provision to NAFTA countries only
would be pretty limiting indeed. Rather, any equivalent treaty
resident in the world would qualify, without regard to what
multi-lettered club its country of residence might belong to. This
seems a much more sensible approach given the underlying rationale of
the derivative benefits provision. Let's hope it catches on.
The other aspect of qualifying as an equivalent beneficiary is that
the ultimate owner in question must be eligible for at least as low a
treaty withholding rate on that item of income as the tested treaty
party. Great fun can be had in comparing and contrasting how this test
is interpreted in the various treaty technical explanations,
particularly with respect to how to view the threshold ownership tests
for different levels of U.S. dividend withholding
tax.
The Base Erosion Test
Basically the requirement here is that a company in question will
satisfy this test as long as less than 50% of its gross income is paid
or accrued as a deductible payment to persons who are not equivalent
beneficiaries. But some treaties exclude arm's-length payments for
services, tangible property, and interest to financial institutions
from the test (e.g., the Dutch treaty) and some do not (e.g., the
German treaty). In some treaties the definition of who qualifies as an
equivalent base erosion payment recipient is narrow. For example, the
Dutch treaty qualifies only base erosion payments to residents that
are equivalent beneficiaries under a subset of LOB tests. In others,
it is broad. For example, the Luxembourg treaty includes all EU and
NAFTA residents that are equivalent beneficiaries under any of the LOB
tests. Kind of a big difference (!) with no obvious policy
rationale.
The Whole Treaty or Not
One last aspect I'll note is that in most treaties, satisfying the
derivative benefits test under the LOB provision will entitle a
company to full access to the treaty. However, in some treaties, for
example, the Canadian, Mexican and Luxembourg treaties--the derivative
benefits test applies only for the dividend, interest, and royalty
articles. It seems a bit hard to fathom why other treaty provisions
would be excluded.
Recommendation
I'm not sure how exactly we got to this level of complexity and
inconsistency in the derivative benefits area. But I do think these
provisions are essential to sensible U.S. treaty policy. Cross border
M&A, joint ventures, private equity ownership, etc., are all real
life examples of why a rigid ultimate ownership standard in an LOB
article is too limiting. And while LOB articles also contain other
alternative tests not based on ownership, such as the active trade or
business test and the headquarters company test, these also are
complex, inconsistent, and insufficient. The key policy question is
whether the company is being used to gain U.S. treaty access that
would otherwise be unavailable. Therefore, I would recommend that
Treasury include a standard derivative benefits provision in the U.S.
model and that it have no artificial geographic limits for either
ownership test or the base erosion test. Further, in order to avoid
the need to renegotiate existing treaties that contain these
inconsistent geographic limits, I would suggest a Competent Authority
agreement approach or a notice that satisfying the new standard will
be deemed to satisfy the discretionary relief article of any existing
LOB provision.
Why should Treasury do this one might ask? Well, the main reason is
that it's the right answer and we should not lay traps for taxpayers
who are in no way abusing the system. But perhaps there is a more
self-interested motivation as well. Based on a good number of recent
conversations I have had with colleagues around the Ernst & Young
world, particularly my Canadian colleagues, it seems that foreign tax
authorities may be more focused of late on the fact that treaties are
indeed “bilateral” agreements. Therefore, even though the
LOB article itself and the derivative benefits provision in particular
are U.S. policy-driven and included in treaties at Treasury's
insistence, they do actually apply in both directions.
There are lots of U.S. companies that are foreign-owned and that
own group subsidiaries or that have cross-border businesses or
investments which benefit from foreign treaty relief. While no tax
advisor I know is advocating treaty shopping through the United
States, the LOB provisions create significant questions with respect
to the eligibility of U.S. companies for foreign treaty benefits in
many circumstances. That is an inevitable outcome of a bilateral
agreement. Therefore, a simpler and more expansive and consistent
approach in this area would take away some needless complexity and
cost in both directions.
This commentary also will appear in the November 2009 issue of
the Tax Management International Journal. For more information,
in the Tax Management Portfolios, see Levine and Miller, 936 T.M.,
U.S. Income Tax Treaties -- The Limitation on Benefits Article, and
in Tax Practice Series, see ¶7140, U.S. Income Tax
Treaties.
|
