The amendment of the German Foreign Tax Act (FTA) as of
January 1 2008 introduced the concept of the hypothetical
arm's-length test. This is applicable especially when it comes
to document transactions involving intangible value drivers.
The concept of the hypothetical arm's-length test is rather
strange in the international tax landscape. For many
multinationals, it seems not to be in line with OECD transfer
pricing guidelines and not practical. In practice, however,
multinationals' transfer pricing affecting their subsidiaries
in Germany is increasingly challenged because the documentation
fails to comply with the German legal requirements, allowing
tax authorities to impose massive transfer pricing adjustments
and penalties in tax audits.
Legal background in Germany
Section 1 section 3 of the Federal Tax Act (FTA) codifies
the arm's-length principle from a German tax point of view. It
effectively states that, for all types of inter-company
transactions for which the traditional transfer pricing methods
(comparable uncontrolled price method (CUP), resale price
method and cost plus method as well as the TNMM method when it
comes to the remuneration of routine functions) are deemed
inappropriate for testing the arm's-length nature of the
pricing, the hypothetical arm's-length test is to be applied to
analyse the appropriateness of transfer prices.
The hypothetical arm's-length test is based on the German
tax concept of the prudent and diligent business manager. When
pricing an inter-company transaction, the underlying
arm's-length fiction is that the responsible management of both
parties will act as profit maximisers on behalf of their
respective entity. Hence the seller (or manufacturer/service
provider/licensor) will seek to agree on a maximum price beyond
a minimum threshold, while the buyer (or distributor/service
recipient/ licensee) will seek to agree on a minimum price
below a maximum threshold.
The second step of the technical analysis – after
assessing that the hypothetical arm's-length test is to be
applied – is then, based on all economic facts and
circumstances, to determine the minimum price from the
perspective of the seller and the maximum price from the
perspective of the buyer. Given that the test is only
hypothetical because in fact the transacting parties are
related, it is a legal requirement that all the qualitative and
quantitative parameters related to the economic circumstances
are shared in a transparent manner for the analysis (no
information asymmetry as in bargaining situations between
unrelated parties).
The difference between minimum and maximum price form a
potential range of arm's-length prices. In the transactions at
hand, the range will typically be wide. Therefore, the German
law additionally requires a third element of the analysis to
determine the most likely agreement value within the defined
range as the arm's-length price. In absence of a sound economic
analysis justifying a particular value, the German tax
authorities on audit can by law adjust to the arithmetic mean
of the two values (midpoint principle).
Taxpayers who do not meet these requirements give the tax
authorities ample room to impose severe adjustments in favour
of the government.
Application of the hypothetical arm's-length test
From our understanding of the law, we derive that, in
addition to the transfer package rules applicable to one time
value chain restructurings, the hypothetical arm's-length test
is effectively to be applied to the following regular
transaction types:
- Profit split solutions in horizontally integrated value
chains: Both parties conduct the same entrepreneurial
activities and functions. They assume the same risks and
share actual system profits based on some allocation key.
Examples could be industrial groups with a global customer
base served by local entrepreneurs, or multinationals
applying cost contribution solutions to the development of
intellectual property.
- Profit split solutions in vertically integrated value
chains: Both parties have complementary functions, risks and
IP. They share actual system residual profits (after
compensation of routine functions) according to some
allocation key. A typical example could be a product IP
owning manufacturer selling to a high value distribution
entrepreneur holding marketing IP.
- License solutions: The IP owner licenses valuable IP to
an entrepreneurial high value contributing licensee. In its
administrative guidelines, the German tax administration is
crystal clear in that in most cases it will reject
arm's-length documentation of royalty rates as flawed if
based primarily on external comparables (royalty database
benchmarking). This would be justified on the grounds that,
given the uniqueness of IP, the potentially comparable
uncontrolled transactions are in fact effectively not
comparable. Internal CUPs are more likely to be accepted,
provided the transaction circumstances are similar.
- Standard transfer pricing solutions applied between IP
owning entrepreneurial entities: While the taxpayers are
generally free to determine the method through which they
price transactions, the standard methods may still be
inappropriate to assess the arm's-length nature of the
applied transfer pricing between IP owning entrepreneurial
entities for the simple reason that there are no comparable
uncontrolled transactions. For example, take the case where
the entrepreneurial manufacturer sells to the entrepreneurial
distributor at fully loaded costs plus 20%. Residual profit
analysis may lead to the conclusion that this may be split in
a routine manufacturer return component of 5% and an implicit
royalty equivalent of 15% on costs. In line with transaction
type (iii), this implicit royalty component will have to be
assessed under the hypothetical arm's-length test.
The hypothetical arm's-length test implies that,
irrespective of which transaction type above is considered, the
transfer pricing analysis must in fact constitute an effective
profit split analysis in consideration of profit expectations
of both parties and a common understanding of all underlying
fact and circumstances. In case (i) and (ii), the profit split
analysis would be based on actual results, whereas in cases
(iii) and (iv) it would focus on budgeted profit allocation,
given that in the latter the entrepreneurial risks would not be
evenly spread. It is hence worthwhile to understand the
commonalities and differences with the profit split
considerations of the OECD.
German law versus OECD principles
According to the OECD guidelines the transactional profit
split method offers a solution for highly integrated business
operations for which a traditional one-sided transfer pricing
methodology such as comparable uncontrolled price method,
resale price method and cost plus method would be
inappropriate. In particular, the profit split method is often
found suitable in situations where both parties' contribution
to the value chain profit is essential. The profit split
methodology is therefore often considered as most appropriate
to determine transfer prices in a situation where both parties
undertake high value functions using intangible value drivers
and assume entrepreneurial risks.
Under the OECD profit split methodology the combined profits
of both parties to the transaction are split among the parties
on an economically valid basis which should be equal to a split
of profits unrelated-parties would agree upon under same or
similar circumstances. The OECD guidelines thereby acknowledge,
similar to the German regulations on the hypothetical
arm's-length test, that certain situations such as the
licensing of unique technology or the licensing of a corporate
brand name only occur among related-parties. Therefore it might
not be possible to find comparable unrelated data to support
the split of profits among the parties. In such cases, the
arm's-length split ratio must be approximated on a hypothetical
basis.
Comparing this to our overview, it seems a fair judgment to
say that the German hypothetical arm's-length test is primarily
a legal definition of how and when profit split analysis is to
be conducted when a German transacting party is involved. The
German tax authorities are clearly of the opinion that the
German rules in this regard are fully consistent with OECD
principles. The rules and the German administrative guidelines
go beyond the principles insofar as:
- They establish that the analysis must document profit
expectations of both transaction parties and determine
minimum and maximum prices under absolute information
transparency;
- They impose the need to assess the most likely value as
the arm's-length price within a wide range, with the
possibility to impose a midpoint price in the absence of
proper analysis; and
- They impose higher documentation thresholds for accepting
alternative analysis approaches instead (for example, CUP
analyses for royalty rates).
The consequence can be significant income adjustments in
cases where, while the taxpayers deem their analysis to be OECD
consistent, German tax authorities consider local legal
requirements are not met.
Approaches to meet German requirements
Before going down the hypothetical arm's-length test route,
it remains to be checked first whether in the case at hand
there are any proper CUPs to derive directly or indirectly the
value or profit contribution of specific IP embedded in a
transaction. Given the burden of proof for the tax authorities
to demonstrate the analysis of the taxpayer is flawed, the
general perception of tax authorities that there are normally
no CUPs when considering IP-related transactions should not
dissuade taxpayers to conduct such analysis in principle.
This strategy may be most attractive when internal CUPs
exist. Such may be the case, for example, if one of the parties
licenses in IP of similar nature from unrelated licensors (or
vice versa). An analysis of the operating profits of the
licensee in such an uncontrolled transaction before and after
license fee payment could serve as a basis for an appropriate
profit split ratio in the controlled transaction, in that the
licensing of IP between associated enterprises. Internal
comparable data typically provide a more reliable means of
comparison due to the higher transparency regarding the
underlying economics and financial data.
Even when comparable data can be identified, it may in any
case be wise to back up any transfer pricing solution by
following the hypothetical arm's-length test in line with
German law. The first step is to determine minimum and maximum
price from seller/licensor or buyer/licensee perspective. One
pragmatic approach could be that both parties at minimum would
like to get cost coverage for their IP-generating activities.
Another one could be that they aim at a minimum to get to get a
certain minimum profit (such as a certain cost plus return).
The most critical step is then to determine the most likely
value (profit split solution) to bridge the gap between minimum
and maximum price. For this purpose, the analysis can be based
on the general OECD guidelines regarding profit splits. The
guidelines suggest basing the split ratio on the relative value
of functions performed by each of the associated parties of the
transaction, thereby taking into account the risks assumed and
the assets employed.
The contribution analysis is one of two approaches described
in detail by the guidelines. Accordingly, the combined profits
of the transactional parties should be allocated applying an
appropriate allocation key which is identified by reference to
profit expectations of independent enterprises being engaged in
comparable transactions. Typical allocation keys could be
assets or costs.
Asset based or capital based allocation keys require a
strong correlation between tangible and intangible assets
respectively, the capital employed by the parties and the value
generated under the controlled transaction. In the context of
the intercompany provision of IP, assets or capital based
allocation keys can only be applied reliably if all the
intangible assets employed in the controlled transaction can be
identified and their relative value can be determined. In
practice, it proves to be rather difficult to identify and
value intangibles assets as not all of the intangibles are
legally protected. In addition, due to the uniqueness of IP,
comparables often cannot be found and a stand-alone valuation
is often accompanied with a number of uncertainties.
Cost based allocation keys are suitable in case there is a
strong correlation between the relative expenses made and the
relative value created under the controlled transaction. In
relation to the intercompany provision of IP, the allocation
would typically be based on development costs of the parties.
Although cost data is often easily accessible, the application
is in practice often hindered by various obstacles. First of
all, the transacting parties could have development expenses
for different types of intangibles, for example, marketing
expenses for the development of marketing IP, research expenses
for the development of patents and product development expenses
for the development of product know-how. As the nature and the
value generation potential of the various IP categories are
different, it can be considered as inappropriate to measure the
relative contribution of the parties based on different
development expense categories. In addition, there could be a
significant time lag between the incurrence of the expenses.
This makes it difficult if not impossible to determine a common
basis for the allocation.
It follows that cost based allocation keys will normally
only be a justified approach to determine the most likely
profit split in horizontally integrated value chains where all
parties contribute the same type of value added. This may be
the case for R&D intensive business where there is R&D
cost-sharing between the entrepreneurs.
For the other transaction types, no suitable allocation key
may be identified. The alternative approach is to measure or
approximate the market value of the contributions. The OECD
guidelines acknowledge that the approach thereby chosen will
often depend on the facts and circumstances of the individual
case. German guidelines contain a similar approach to determine
the relative contribution of related-parties to a controlled
transaction, the so-called value chain contribution analysis.
This analysis can serve as valuable guidance for the profit
split related contribution analysis.
Value chain contribution analysis – a practical
approach
The value chain contribution analysis starts with an
overview of the entire value creation process of a group of
enterprises, a division or a business unit. Thereby all
relevant business processes or activities must be identified.
In a second step, the relative value of each business process
or activity for the entire value created must be determined.
The final step encompasses an analysis of the relative
contribution of each party to the various business processes
and activities.
To measure the relative contribution of each business
process to the total value-added, objective and subjective
criteria or a mixture of both can be employed. Pure objective
criteria such as assets, capital or costs are characterised by
the same deficiencies as outlined above. Therefore, subjective
criteria such as preferences of management, customers or
shareholders are often employed in practice in connection with
the value chain contribution analysis.
To improve the level of reliability of subjective criteria
used to measure the relative contribution of associated
enterprises being party to the controlled transaction, it is
recommendable to base the measurement on state-of-the-art
principles and tools of valuation theory. One possible approach
could be the conduct of a conjoint analysis, which is a
multi-attribute compositional model to determine the relative
value of certain components to the overall output (value
created) in order to identify a general pattern of preferences
of different stakeholders. The conjoint analysis and an
accordingly designed process ranking model provides a useful
and reliable framework for tracking and documenting the
preferences and decision paths of management in the context of
the value chain contribution analysis.
The value chain contribution analysis can in particular be
useful to determine an appropriate allocation/split of the
profit among the parties as regards the transaction types, if
conventional allocation keys based on assets, capital or cost
can not be identified. The result of the value chain
contribution analysis can serve as sufficient proof and
documentation for a deviation from the otherwise obligatory
midpoint of the range between the minimum price of the
seller/licensor and the maximum price of the buyer/licensee
under the hypothetical arm's-length approach required under
German law.
This approach could enhance the OECD based transfer pricing
analysis and at the same time significantly reduce the risk
that German tax authorities consider local requirements are not
being matched.
Sensible and practical
The value chain contribution analysis as outlined above can
be a sensible and practical approach to bridge the potential
gap between established OECD principles on transfer pricing for
transactions involving significant IP and the German
hypothetical arm's-length test. Existing difficulties for
taxpayers to determine transfer prices for transactions
involving significant IP arising from a lack of international
guidance and substantial differences between German and other
national regulations of various countries can be addressed
pragmatically.
The OECD intangible initiative is aimed at tackling such
difficulties and at establishing international
consensus-orientated guidance to avoid or at least reduce
future complex and monetarily-significant transfer pricing
disputes resulting from intangible-intensive transactions among
associated parties. It should be interesting for taxpayers with
operations in Germany to closely monitor international but also
further German developments in this area.
| Biography |
 |
|
Yves Herve
PwC
Tel: +49 (0)69 9585-6188
Email: yves.herve@de.pwc.com
Yves Hervé is a transfer pricing partner of
PwC and based in Frankfurt, Germany. Hervé is a
member of PwC's global value chain transformation
network and co-leader of its German core team.
Hervé graduated in economics at the
University of Bonn, Germany, in 1993. After his master
studies at the College of Europe, Belgium, he became a
lecturer in economics at the University of
Saarbrücken, Germany. As a member of the European
Institute, he advised several European institutions on
economic issues. After submission of his Ph.D. thesis,
he became a founding member of KPMG's German transfer
pricing practice in 1999. He moved to PwC in 2010.
As a transfer pricing specialist, Hervé has
covered all aspects of transfer pricing advisory work
for multinational clients, from documentation and tax
audit defence to tax planning work covering IP
valuation and migration strategies. He has conducted
several big value chain restructuring projects for
European and US based multinationals, for example,
assisting in the design and implementation of principal
structures in low tax jurisdictions. Next to tax
effective IP structuring, a particular focus of his
work has been on decision-making process implementation
such to document principal substance and prevent PE
creation in other territories.
|
| Biography |
 |
|
Susann van der Ham
PwC
Tel: +49 (0)211 981-7451
Email: susann.van.der.ham@de.pwc.com
Susann van der Ham is a transfer pricing partner at
PwC Düsseldorf.
Van der Ham has more than 10 years of experience in
consulting multinationals in the field of transfer
pricing. Her expertise encompasses transfer pricing
structuring, value chain transformation, system
implementation, documentation and tax audit defence.
She advises large international clients (both German
and foreign headquartered) and has led a variety of
projects, inter alia in the retail and
consumer, automation and automotive industry.
Van der Ham frequently publishes articles in
international and domestic tax and transfer pricing
journals and she is a regular speaker on transfer
pricing events. She holds a degree in economics and she
is a German Certified Tax Advisor
(Steuerberaterin).
|