Ref. Ares(2015)5581988 -
document 8
04/12/2015
ROOM DOCUMENT # 2
Code of Conduct Group (Business Taxation)
4 June 2012
ORIGIN: Commission
WORK PACKAGE 2011 – ANTI-ABUSE
MISMATCHES
1. Introduction
In Roomdoc #2 of 17 April 2012, the Commission had proposed structuring future
discussions on mismatches in line with the approach taken by OECD's work in this area. The
paper subsequently suggested organising the Group's work in two parts: information gathering
and finding / agreeing a solution.
During the discussion of this paper on 17 April, the Group expressed a preference for a
focussed approach avoiding duplication of work undertaken already by the OECD. In order to
assist the Group in determining the proper focus, the Commission was asked to prepare a
summary of the information and examples already gathered by the OECD and on that basis
propose areas of work on which the Group could concentrate.
The work of the OECD on hybrids was led by a focus group of 9 OECD countries under the
heading of the OECD's work on aggressive tax planning in Working Party 8 on the basis of a
project outline agreed in June 2007. The full report by Working Party 8 prepared by this focus
group describes in detail background information and policy issues raised by hybrids, gives
various examples of hybrid arrangements both theoretical and real life cases, gives an
overview of country's response strategies and finally lists several recommendations. That full
2010 report was further discussed in various forums of the OECD and finally led to the public
March 2012 report "Hybrid Mismatch Arrangements: tax policy and compliance issues"
1 Annex 1 to this document consists of a summary of the full 2010 report of the OECD
Working Party 8.
This note contains in paragraph 2 some general considerations to describe the relation
between the work undertaken by the OECD and that of the Code Group. It aims to provide
some guidance for the Code Group for further discussions by identifying areas on which the
Code Group could concentrate. Finally, paragraph 3 of this note gives suggestions for further
steps.
1 The report is available at the website of the OECD: http://www.oecd.org/dataoecd/20/20/49825836.pdf and MS
are recommended to take note of this report.
Page
1 of
17
2. Considerations and suggested approach for Code Group work
a) The OECD work contains substantial information on examples of mismatch arrangements,
describing both the theoretical background and providing concrete examples that have
been detected by tax administrations. The fact that the report is based on experiences of
, does not reduce the value of the report, as most schemes can probably
be copied in many other countries as well.
The OECD work also contains substantial and valuable information about unilateral
response strategies developed by the countries concerned in the report (anti-avoidance
rules) as well as about various measures that have been introduced in bilateral tax
conventions.
Consequently, as far as information gathering and assessing the scope of the problem is
concerned, the Commission at this stage does not believe that a general exercise aimed at
collecting additional information will be necessary in order for the Code Group to discuss
and decide upon further steps. Still, MS remain free to provide additional information
which they consider relevant.
b) The goal of the OECD exercise was mainly assisting tax administrations in taking
effective actions against the effects of mismatches by (1) showing examples of structures,
(2) listing unilateral actions that countries may take or have taken to prevent losses of tax
revenues and (3) suggesting techniques to improve the detection of mismatch schemes. An
additional element – that is only briefly referred to by the OECD – is the harmful tax
competition element: to what extent can mismatches create harmful effects? Thus the
reason for not taking effective action could be twofold:
1. In the first place MS could be tempted not to prevent existing and new investors
from benefitting from a reduced domestic tax base in order to achieve a
competitive advantage compared to other potential investment locations. This
could have damaging effects on competition and on tax revenue.
2. Secondly, negative results could come from MS allowing business to use a
mismatch arrangement involving their MS but effectively eroding the tax base of
other MS (see for example scheme 1e of Annex 1 whereby country B is involved
in but not affected by a hybrid mismatch arrangement).
This tax competition element should be particularly relevant for the Code of Conduct
Group.
2
Page
2 of
17
In other words, where the focus of the OECD is mainly on assisting individual countries
to fight mismatches, the Code Group's focus could be on MSs joint commitments to "stop
being tolerant" to mismatches as well as seeking and implementing commonly agreed
solutions to such mismatches.
c) Closely related to the former point, the policy focus of the OECD is primarily on detection
of mismatch schemes followed by either unilateral anti-avoidance measures or bilateral
measures under bilateral tax treaties. The multilateral approach, for example via uniform
classification rules (harmonisation) or via mutual recognition (coordination), is dismissed
as not being realistic or merely shifting mismatch arrangements to other non-harmonised
or non-coordinated areas. In the Commission's view, it is precisely in this area that the EU
could make progress, taking into account the close relations between EU Member States.
The EU allows for close cooperation at policy level and offers a real opportunity to
explore multilateral solutions (e.g. the agreed solution of PPLs).
d) The OECD report contains examples of country pair tables providing per set of two
countries an overview of the entity and instrument classification per type of entity and
instrument. The Code Group could consider further developing such tables for each
bilateral relation in the EU. However, given the high level of detail and the case-by-case
character, such tables would primarily serve as a detection tool for tax auditors in the field
to effectively apply their unilateral anti-avoidance measures. Moreover, the work related
to developing such tables for all 351 bilateral relations in the EU does not seem
proportionate to the expected benefit.
The Commission would therefore propose not pursuing the further extension or
preparation of bilateral tables within the Code Group and focus on developing more
uniform coordinated solutions instead.
e) The classification of schemes into types of effect as used by the OECD can be very useful
to get an overview of the various planning possibilities available via mismatch
arrangements. Also, in drafting unilateral anti-avoidance measures, countries may very
well focus their legislation on the effect they want to prevent (e.g. no double deduction)
instead of on the arrangement that causes the effect. The Code Group, however, started its
work on mismatches by dealing with an arrangement (the PPL solution, basically
addressing hybrid instruments).
The Commission would therefore propose to primarily use a distinction between
coordinated solutions for hybrid instruments, hybrid entities and hybrid transfers rather
than between types of effects (classification based rather than effects based).
Page
3 of
17
f) By agreeing the solution with respect to Profit Participating Loans, the Code Group in fact
already addressed the topic of hybrid instruments. This means that the remaining topics as
far as covered by the OECD report concern hybrid entities and hybrid transfers. Hybrid
transfers are mainly used in relation to so-called tax credit generators, which seems to be
an issue that is less urgent in an EU context where most MS apply exemption methods
instead of tax credits. The main attention could therefore be given to hybrid entities. This
could be complemented with hybrid PE's, as described in Roomdoc #2 of 17 April 2012.
The Commission suggests that the priority for further work of the Group on mismatches is
given to finding coordinated solutions for hybrid entities and hybrid PE's.
3. Suggestions for further steps
The PPL discussed previously by the Code Group is just one type of hybrid instrument.
Nevertheless, the solution agreed by the Group on PPL's is a solution that fits all hybrid
instruments. Before arriving at this solution for PPL's, the Group took a series of decisions
which could be used as a model in further discussions concerning a solution for hybrid entities
and hybrid PEs. The decisions on PPLs were the following:
1. It was decided that measures ensuring the removal of (the effects of) a mismatch were
preferred over exchanging information and expecting the other MS to address the
issue;
2. It was decided that a mutual recognition approach (one MS follows the treatment
given by the other MS) was preferred over strict harmonisation (EU-wide uniform
classification rules);
3. It was decided that the MS receiving payments under a PPL should adapt to the tax
treatment given by the Source State; and
4. It was decided that the receiving MS does not necessarily have to follow the
classification given by the Source State, as long as it ensures the desired effect (no
exemption in case of deduction);
Similar decisions will need to be taken when developing a solution for hybrid entities. This
certainly applies to decisions 1 and 2. However, for decisions 3 and 4, the situation for hybrid
entities and hybrid PEs is slightly different than for hybrid instruments. For hybrid
instruments (such as PPLs) the typical effect is "deduction/no inclusion". Addressing the
effect (no exemption in case of deduction abroad in case of a hybrid loan) therefore can be
considered equivalent to addressing the hybrid character itself. The effects of a hybrid entity
or hybrid PE are less uniform because they can both lead to "deduction/no inclusion" and to
"double deduction" effects (for examples, see scheme 1a and 2b of Annex 1). As a result, a
unique solution based on the effects as in step 3 and 4 might be more problematic when
discussing possible solutions for hybrid entities and hybrid PEs.
Page
4 of
17
•
MS are invited to comment on the observations and suggestions made by the
Commission under par. 2, letters a) to f) above in combination with the summary of
the full 2010 OECD report on mismatches provided in Annex 1.
•
MS are invited to comment on the suggestion to follow the same approach with
hybrid entities and hybrid PEs as with hybrid loans as outlined in par. 3. In that
context:
o
Do MS agree that measures actually removing the (effects of a) mismatch
involving hybrid entities/PEs are preferred over exchanging information and
expecting the other MS concerned to address the issue?
o
Do MS agree that a mutual recognition approach (one MS follows the
treatment given by the other MS) is to be preferred over strict harmonisation
(EU-wide uniform classification rules)?
o
Do MS agree that under a mutual recognition approach for hybrid entities,
the classification given by the MS under which laws an entity has been
established should be the leading one, with the other MS following the
qualification of the leading State?
o
Do MS agree that under a mutual recognition approach for hybrid PEs, the
position taken by a State concerning the absence/presence of a PE in its
territory or concerning a limited profit allocation to such PE should be
followed by the head office (residence) State?
•
Do MS agree that a classification based approach for these types of hybrids is to be
preferred over an effects based approach in view of the different effects that hybrid
entities and hybrid PEs can have?
Annex 1.
Summary of OECD report on hybrid mismatch arrangements
___________________
Page
5 of
17
Annex 1 to Room document # 2 of 4 June 2012
Summary of the relevant elements of the full 2010 OECD WP8 report
"Report on Hybrid Mismatch Arrangements"3
The OECD paper on hybrids discussed three issues in relation to hybrid structures.
A. Examples of hybrid schemes encountered in the respective countries;
B. Policy responses developed by countries concerned;
C. Detection strategies developed by countries concerned.
The OECD hybrids report is based on input of a focus group consisting of 9 countries, 5 EU
Member States (FR, DE, NL, UK and SE
4) and 4 third countries (USA, Canada, New
Zealand, Australia).
It discusses hybrid mismatch arrangements related to hybrid instruments (debt/equity), hybrid
companies (transparent/opaque) and hybrid transfers (ownership transfer/collateralised loan).
The latter only regards transfers of financial instruments, not tangible property (machines
etc.). Also, dual resident structures are ignored.
The various characteristics of hybrid instruments and hybrid entities and how they work to
reduce the overall tax base in cross border situations are described in detail. Moreover, the
Annexes to the report contain valuable and illustrative background information in relation to
policy responses and the detection of hybrid structures:
• Annex 1 contains country specific examples of debt/equity classification rules (for
R
Australia and SE) of entity classification rules (for Canada and NL) and of ownership rules in
the context of “repos” and similar arrangements (for DE and UK).
• Annex 2 contains illustrative tables showing of entity and instrument classification for
various types of entities and instruments in specific country pairs
clearly showing potential risk areas.
.
• Annex 3 shows legislative responses (anti-avoidance measures) taken by the countries
concerned to hybrid mismatch arrangements. The unilateral measures are organised by
the effect which is achieved with the arrangement.
3 CTPA/CFA/WP8(2010)13/CONF
4 DK also contributed in relation to response strategies.
Page
6 of
17
A. EXAMPLES OF HYBRID SCHEMES ENCOUNTERED;
Comparable to Roomdoc #2 of the 17 April meeting, the examples are categorised by their
effect. For each type of effect several examples are scheduled, generally starting with a basic
theoretical example and subsequently mentioning examples of cases including the countries
concerned:
1) Double deduction:
a) Hybrid entity combined with Group Taxation
b) Hybrid entity combined with Hybrid instrument
c) Chain of (reverse) hybrid entities ("tower")
d) Combinations
e) Third country affected
2) Deduction / no inclusion:
a) Classic hybrid instrument
b) Hybrid entity
c) Hybrid transfer
d) Third country affected
3) Tax credit generators
1a) Hybrid entity combined with Group Taxation – basic example
Hybrid Entity, resident of country B takes out a loan from a third party or another group
company. The loan amount is used to either buy the shares in B Co. (acquisition) or to inject it
as additional equity into the existing subsidiary B Co. Hybrid Entity pays interest on the loan.
As Hybrid Entity is transparent for country A tax purposes, its interest expenses are deducted
directly at A Co level against unrelated income of A Co. In country, B Hybrid Entity is treated
Page
7 of
17
as a corporate tax payer. Its interest expenses can be offset against other income of B Co and
other group companies in country B under country B's group relief regime.
It is recognised in the report that in most cases it is difficult to assess clearly which country's
tax base is at stake. If the structure indeed involves external financing (a bank loan), one can
assume that there is a genuine financing need at group level which justifies that interest due is
deducted at least once: either in country A or in country B (but not in both).
If Hybrid Entity acquires B Co against a loan from a sister company of A Co in country A, the
tax situation in country A does not change (interest deducted at A Co and taxed at A Co's
sister), but country B is suddenly confronted with interest expenses that have been created as
a result of the internal reorganisation. In that case, it is stated that it is country B who loses tax
revenues.
In addition various other considerations are mentioned that require attention to ensure that the
structure "works".
1b) Hybrid entity combined with Hybrid instrument
Hybrid Entity resident in country B borrows money and uses the loan proceeds to buy a
hybrid financial instrument issued by B Co (equity under country A tax rules, debt under
country B tax rules). For country B the situation is neutral: Hybrid Entity earns (accrued)
interest income and has interest expenses. For country A, Hybrid Entity either has no income
Page
8 of
17
(in case of accrual) or earns tax exempt dividend income. However, since Hybrid Entity is
transparent for country A tax purposes, it does have interest expenses in country A which can
be offset against A Co’s unrelated taxable income.
1c) Chain of (reverse) hybrid entities ("tower")
This is a clear example of how complex structures can get. In this structure, B Co deducts the
interest expense (4) in country B. For country B Hybrid Entity is the recipient of the interest
but the interest received is set-off against interest expenses (6). For country A tax purposes,
Reverse Hybrid Entity is resident of country B and pays a dividend to A Co (tax exempt). At
the same time country A allows A Co to claim the interest deduction. An example is given
with Canada as country A and US as country B.
1d) Combinations
Complex examples are given of structures combining all of the elements in 1a, 1b and 1c in
.
Page
9 of
17
1e) Third country affected
This scheme – which is effectively a variant of 1a) – creates a deduction for country A tax
purposes and a deduction for country C tax purposes. Country B tax is not directly affected as
it is basically a conduit company receiving and paying interest.
The important aspect of this scheme is that for country C it will be very hard to detect that the
interest taxed by country B in fact is "compensated" with a double deduction in the country A
– country B relationship.
Again, it is difficult to say whether primarily country A or country B's tax revenues are at risk
as this depends on the structure that parties would ordinarily have chosen (loan by A + direct
debt by A Co to C Co, or loan by A + direct equity relationship between A Co and C Co.
However, in any case country A’s and country C’s combined tax revenue is reduced
compared to both ordinary structures.
Page
10 of
17
2a) Classic hybrid instrument
The concrete examples given in the OECD report include a profit participating loan (with an
unspecified country
a mandatory convertible note (
), an optional convertible note (with an unspecified
), the
forward sale of a mandatory convertible note (
) or capital
contributions under a silent partnership (
). In the last
example, the normal treatment in both countries was debt, but the
n creditor argued that
This basically is the PPL structure that has been discussed previously in the Code Group and
for which the Group already has agreed a solution.
2b) Hybrid entity and group taxation
This type has actually two sub-types. The first type involves the hybrid entity as borrower:
Page
11 of
17
Hybrid Entity borrows from A Co and uses the loan amount to either inject it as equity into B
Co or to acquire shares in B Co (in- or external acquisition). For country B tax purposes, the
interest due by Hybrid Entity can be used to offset other country B group companies’ income
under the country B group relief regime. Any dealings between A Co and Hybrid Entity are
seen as internal dealings by country A. As a result, there is no loan and no interest income for
A Co.
Concrete case examples of this structure are given in relation to
and as a
variant in the relation
. Again, a series of elements are mentioned that will determine
whether the structure actually achieves the result desired by tax planners.
The second type involves the hybrid entity as the lender:
Hybrid entity does not have substance and therefore does not constitute a PE of A Co in B.
For country B tax purposes, B Co can deduct interest expenses whereas the interest income is
directly attributed to A Co (no PE) and will not be taxed in country B. Country A will not tax
Hybrid Entity’s profits either as for country A tax purposes Hybrid Entity is a non-resident
corporation.
Comment [RK9]: As above
Page
12 of
17
2c) Hybrid transfers
A Co enters into a “sale and repurchase” transaction with B Co in country B (related company
or unrelated third party). A Co sells Object Co shares to B Co for a fair market value of €100
and at the same time A Co and B Co agree that after a certain period (e.g. 6 months), A Co
will buy back the shares for €105. Economically this is similar to a loan from B Co to A Co
(principal of €100) repayable after 6 months with interest of €5 and the shares being a
collateral to secure the repayment.
Under Country A’s tax laws, which treat the transaction as a collateralised loan, A Co is
allowed to deduct interest of €5. For country B tax purposes, however, B Co buys the Object
Co shares for €100 and after 6 months sells them at €105, realising a capital gain of €5
exempted under the participation exemption of country B. Because country A allows an
interest deduction of €5 and country B views a tax-exempt capital gain of €5, the scheme
achieves a “deduction / no inclusion” effect.
Page
13 of
17
For country V the participating loan is regarded as supply of equity (“requalification”). The
“interest” on the loan therefore qualifies as income under the participation exemption and is
exempt from country V tax. In
, however, no requalification will take place
according to the
legislation and thus the interest is deductible.
will only be
paras
taxed on a spread (the interest income it receives from Countries X, Y and Z minus the
interest paid to V Co.
Technically there is a deduction in
and no inclusion of the corresponding receipts
in country V. However, since interest income and interest deductions in
offset
each other, the scheme is tax neutral from a
point of view. Because interest expenses
will be deductible for Country X Co, Country Y Co and Country Z Co, economically a
“deduction / no inclusion” effect is achieved between countries X, Y, Z on one side and
country V on the other side even though there is no direct mismatch with “participating loans”
involving countries X, Y, Z.
3) Foreign tax credit generators
These type of arrangements are mostly a variant of various deduction/no-inclusion schemes
(above), but instead of the "no-inclusion" one of the parties claims foreign tax credits for
foreign tax that was effectively not (or not entirely) paid. Most of the schemes involve a
hybrid transfer (1c) or are variants thereof. These type of arrangements seem to be of less
relevance for the MS, which operate mostly exemption regimes.
Page
14 of
17
B.
POLICY RESPONSES DEVELOPED BY COUNTRIES CONCERNED;
The response strategies available for countries to fight mismatch arrangements are divided in
4 different categories.
1. Scrutinising the classification claimed by the tax payer
The most common and basic approach against a mismatch arrangement would be to scrutinise
the position taken by a taxpayer with respect to the tax treatment of a hybrid instrument.
Proper classification of hybrid instruments or entities under domestic law is often a matter of
interpreting a country's tax provisions. This approach therefore may be effective in borderline
cases where tax payers take the position that is most beneficial for them while the proper
classification is actually debatable.
2. General anti-avoidance rules
The application of general anti-avoidance rules or doctrines as "substance over form" or
"business purpose" is considered to a viable instrument against mismatch schemes. This is
certainly true in case of the complete absence of a business purpose for example certain
circular transactions. However, the burden of proof is often with tax administrations in these
cases and the abusive aim of a tax is often difficult to prove in case of real cross border
investments structured via hybrid arrangements.
3. Specific rules targeting mismatch arrangements
Some countries have considered general anti-avoidance provisions insufficient to address the
concerns of mismatches. The report contains an overview of specific measures applied by the
countries in the focus group (including DK). Table 1 below provides a summary overview of
these measures. In addition, a series of arguments are addressed giving potential reasons for
the relatively limited presence of such rules in many countries:
• For some mismatch arrangements, countries may consider that primarily the tax
revenues of other countries are at risk, not theirs;
• Countries may feel they put themselves or their multinationals at a competitive
disadvantage compared to countries that do not disallow hybrid arrangements;
• Arguments relating to fiscal sovereignty (no dependence on foreign tax rules);
• Specific provisions make the legislation more complicated
(administration/compliance);
• Mismatch arrangements do not exist;
• Lack of awareness of mismatch arrangements or a lack of urgency to respond to them;
All these elements may play a role for MS as well, but especially the first two elements
should be relevant in the context of the Code of Conduct.
4. Harmonisation or coordination
The potential for reducing the occurrence of mismatches via harmonisation / coordination of
classification rules, the OECD report refers to the Code solution on PPL's, but warns that this
may only lead to a shift of hybrid arrangements to non-coordinated relations.
Page
15 of
17
5 The 2012 public OECD paper "Hybrid Mismatch Arrangements: tax policy and compliance issues" reveals that also Austria and Italy have this type of anti-avoidance
measure in their legislation.
6 Provision literally not applicable, but arguably could be invoked.
7 Only applies to constructive dividends (hidden profit distributions).
8 Treaty exemption is withdrawn if there is no not taxation abroad. The report contains a separate section of the possibility to address hybrid arrangements in bilateral DTC's.
9 Specific limitations on interest deduction could have this effect but do not specifically target hybrid arrangements.
10 Provision denying foreign tax relief (tax credits) in case of artificial arrangements.
Page
16 of
17
C. DETECTION STRATEGIES DEVELOPED BY COUNTRIES CONCERNED.
Observing that both a discussion about the need for a response strategy and the application of
anti-avoidance measures requires that authorities are aware of the schemes operated by
taxpayers, the final part of the OECD report is dedicated to the detection of mismatch
arrangements. In that respect the report describes experiences of the focus group countries
with the following means of detection:
1. Audits. This can concern routine audits or audits specifically targeted at hybrid
schemes for example using specific hybrid-related checklists.
2. Rulings and other voluntary advance clearance mechanisms whereby taxpayers may
directly or indirectly seek tax administrations view on the effects of an intended
transaction (e.g. the deductibility of certain expenses or the exemption of certain types
of income);
3. Mandatory disclosure mechanisms whereby tax payers are obliged to disclose the
elements of tax planning schemes in advance, whereby non-compliance will invoke
automatic penalties;
4. Public sources for example professional articles in journals or presentations at
seminars;
5. International cooperation. The report explicitly recognises the value of international
cooperation in the detection of mismatch arrangements, precisely because of the cross
border nature of it. The cooperation can take various forms and includes spontaneous
exchange of information, the preparation of country summaries on classification rules
(examples of these are provided in an Annex) and the preparation of country pair
tables which shows the classification rules for the country pair for a series of entities,
instruments and transfers thereby identifying potential risk areas.
D. OBSERVATIONS AND RECOMMENDATIONS
The report ends with a series of observations and recommendations. In essence this is a
summary of the report combined with the recommendation that countries take the necessary
actions to ensure that the contents of the report are shared with their tax administrations and to
follow-up with taking those actions considered appropriate by each country. Also further
international cooperation is encouraged and some suggestions for further work (e.g.
developing shared principles and coordination mechanisms for tax administrations).
______________________
Page
17 of
17