Transfer Pricing April 2014

Comments on the OECD paper on "Transfer Pricing Comparability Data and Developing Countries" Dear all, WTS is pleased to provide you with comments regarding the OECD paper on "Transfer Pricing Comparability Data and Developing Countries". We appreciate the effort to provide guidance on this important topic. In the following, we would like to comment and assess the four provided possible actions and outline our further ideas for an approach to handle the concerns expressed by developing countries on the quality and availability of information on comparable transactions. Additionally, our ideas should help to achieve a globally agreed and standardized approach for the usage of transfer pricing (TP) comparability data respectively conducting comparability analyses (we have outlined the corresponding paragraph(s) within the OECD paper at the beginning of each paragraph).

A.

Provided possible actions by the OECD

1.

Expanding access to data sources for comparables

(13) We welcome the described measures to expand the access to public data sources. Especially the lack of e.g. gross margin information on a transaction basis makes a reasonable generation and comparison of comparability data complicated if not impossible at all. Hence, developing an internationally standardized methodology for commercial databases how to preferably generate and provide the data could be a good approach, also to reduce deviations in the data collection in different countries. (15-16) Even more could the approach to develop respectively implement standardized accounting standards and disclosure requirements be very helpful, at least across specific regions, to gather sufficient comparability data. Especially a regional approach could be a good way to expand the access to sufficient data sources for comparability data. Also the degree and quality of comparability of the financial data would of course be significantly increased if the data is based on equal accounting standards. (17) We estimate the usage of secret comparables very critical as the data is not available to the tax payer, neither at the date of conducting the transaction respectively concluding the contract by applying the arm's length setting approach nor at the date of testSeite 1 von 7

Transfer Pricing April 2014

ing the transfer prices e.g. at the end of the fiscal year by applying the arm's length outcome-testing approach. Hence, this would contradict the arm's length principle as third parties would also never rely on any undisclosed secret data. Thus, this problem could only be solved by implementing appropriate accounting standards and disclosure requirements. However, we assume that all the above approaches would be long-lasting to implement and do not solve the problem, if no or only very few local comparability data exists. Even though, we would advise to develop this approach in the long run.

2.

More effective use of data sources for comparables

(19-20) We estimate the proposed approach to "...broadening the search for comparables to uncontrolled transactions in the same industry but in other geographic markets..." to be the best way to solve the problem of insufficient local comparability data. Especially in an arm's length analysis based on the transactional net margin method (TNMM), the focus should rather be on the comparability of the transaction respectively the functional profile of the independent party instead of focusing on the geographic market. Thus, an analysis based on a largely homogenous, regional market should also be accepted respectively preferred against a local analysis to obtain a sufficient number of comparables. This is also in line with the view of most international tax administrations and the OECD, which emphasize that functional comparability is more important than product or regional comparability when using the TNMM. (21) We understand the concerns regarding the usage of foreign comparables in a domestic market and thus the discussion regarding the potential need of adjustments. However, the performance of adjustments should only be executed under very limited and strict globally agreed standards which are applied on a mutually agreed basis. Otherwise a high risk of double taxation could be created, e.g. if one country accepts the performed adjustments or even wants to perform adjustments on its own but the other country does not accept any adjustment resulting in different ranges for arm's length transfer prices. Therefore, if adjustments are seen to be necessary, a clear guidance should be developed by the OECD on how to perform such adjustments in detail. Providing examples for such calculations would certainly be very helpful. (22) The approach by "...testing the return earned by the foreign counterparty..." does not seem practicable at all. First, generally only for entities having a routine functional profile a comparability analysis is reasonable as the transaction partner of a routine entity typically is a more complex entity bearing entrepreneurial risks and owning unique immaterial assets. This is also outlined in detail in paragraph 3.18 and especially for the TNMM in paragraph 2.59 of the OECD Guidelines, whereas it is clearly described that the tested party should always be the less complex one. Hence, a testing of the return earned by the Seite 2 von 7

Transfer Pricing April 2014

foreign non-routine counterparty is not possible and against the parameters defined by the OECD for many years now. Second, the disclosure of the results from the foreign counterparty fails mostly due to national non-disclosure rules. Thus, this paragraph / approach should be deleted. Especially, as this approach rather creates the impression to follow a global formulary apportionment instead of being based on the arm's length principle. (23) The approach of "...benchmarking against other industries in the same geographical market..." also does not seem appropriate as the results are not expected to be comparable respectively to be in line with the arm's length principle (e.g. comparing the margins earned in the financial services industry to those in the telecommunication industry). Furthermore, there is no guidance which industries are classified as comparable and which not respectively if and how adjustments between any industry specific results would be necessary. We would also assume that the development of such guidance would not lead to reliable result. Thus, this paragraph / approach should be deleted.

3.

Approaches to reducing reliance on direct comparables

(24) In the case that no direct comparables are available, a formula based approach is considered. However, the OECD also clearly outlines that such a formulary apportionment is "...not a viable way forward;" Hence, we do not understand why this topic is mentioned at all and highly recommend to delete the paragraph as we expect this will be misinterpreted by many tax administrations and would bring the tax payer into many international taxation disputes. (25) Furthermore, the OECD also mentions the usage of the profit split method (PSM) as an alternative method if comparability data is not available. Even though that this approach would be qualitatively realizable, no guidance is given on how a profit split based on an economic or value chain analysis should be carried out in practice, especially if one transaction partner is a routine entity. This is one of the main reasons why most national as well as international TP guidelines in place (especially the OECD Guidelines) prefer for instance the TNMM before the PSM in case one transaction partner is a routine entity. According to paragraph 2.4 and 2.109 of the OECD Guidelines the PSM should only be applied when both parties make unique valuable contributions to a transaction and the PSM can not be used in cases where one party performs only simple functions and does not make any significant unique contributions. Additionally, an arm's length profit split allocation key is much more susceptible for disputes with the tax administrations and moreover could be easier misused as a further possibility for profit shifting. Hence, we would rather encourage the OECD to find a practical solution using some of the approaches discussed in section A.1 and section A.2 instead of moving away from the well established approach of the usage of the most appropriate TP method. Seite 3 von 7

Transfer Pricing April 2014

(26) The application of safe harbours should also be seen critical although in principle this may constitute an appropriate solution especially for standardized routine transactions, e.g. administrative services. Only in case safe harbours are applied on a bilateral or even multilateral basis a potential double taxation can be avoided efficiently. In addition, negotiating bilateral safe harbours is expected to be costly and long-lasting and the transfer prices negotiated would mostly not correspond to arm's length transfer prices. Thus, we would encourage the OECD to develop an approach for the application of global safe harbours based on a clear definition of the scope of possible (routine / low value adding) transactions respectively activities and the methodology for appliance of the safe harbours. This could be based on a global economic analysis as further outlined in step 3 of section B below. Furthermore, the idea of following a regional approach is welcome, not only in case of safe harbours but also in case of searching for comparables as outlined in paragraph 19. (27) The OECD further suggests the usage of the so-called "sixth method". This method resembles the comparable uncontrolled price method (CPM) as for certain commodity products (generally raw materials that are traded on public stock exchanges) exchange prices are also taken as a basis for calculating arm's length transfer prices. However, further pricing elements are not reflected in the sixth method, e.g. the product specific processing stage, the individual transport costs, Incoterms or the connected risks. Hence, third parties would hardly use a price calculated according to the sixth method (i.e. the net-exchange price without any further price elements) and such a price would also not be compliant with the arm's length principle. Thus, this paragraph / approach should be deleted. (28) The appliance of further anti-avoidance approaches is also seen very critical. Such approaches should only be applied on a bilateral or multilateral basis. Furthermore, such an approach is definitely not in line with the arm's principle and this topic does also not correspond to the purpose of this paper. Hence, this paragraph / approach should be deleted. In total, we regard most of the approaches outlined to reduce reliance on direct comparables as non compliant with the arm's length principle. Thus, we would advise the OECD to rather concentrate on the first two possible actions and neglect the described approaches.

4.

Advance pricing agreements and mutual agreement proceedings

(30-32) As fourth possible action the OECD refers to the possible usage of advance pricing agreements and mutual agreement procedures to determine arm's length transfer prices in negotiations between the tax administrations. However, such procedures also require a certain amount of comparable data as basis for the negotiations and therefore do not solve the underlying problem itself. In addition, the negotiations are often affected Seite 4 von 7

Transfer Pricing April 2014

by an asymmetric bargaining power between the tax administrations and do, as a consequence, often not result in arm's length transfer prices but on a fiscal oriented compromise between the tax administrations. Overall, it is expected that such procedures are rather complicated to be implemented respectively executed in most developing countries due to the lack of sufficient trained tax administration staff and organisational procedures. Especially as such procedures are generally long-lasting and resource-intensive. Furthermore, only the fewest developing countries have already introduced programs for advance pricing agreements. Additionally, in the regulations of the tax treaties of most developing countries generally no binding arbitration procedures are envisaged for mutual agreement procedures, meaning that negotiations are not obligated to lead to a mutual consent. Hence, such procedures could be helpful in some cases but should not be seen as the practical solution to solve the problem on short notice. However, in the long run such multinational procedures should be strengthened.

B.

Summary and proposal for further definition of a globally standardized approach

As a result, only some of the approaches within the first two proposed possible actions seem to be reasonable in practice to efficiently solve the problems of insufficient comparability data in developing countries and also to be compliant with a globally consistent application of the arm's length principle. Furthermore, we understand the underlying paper also as opportunity to further define a globally agreed and standardized approach for comparability analyses. Hence, we have summarized in the following our proposal for the next steps respectively for an overall approach to be promoted by the OECD. Step 1: Regarding the developing countries, the OECD should focus on the development respectively implementation of standardized accounting standards and disclosure requirements to ensure and / or improve the availability of comparable and publicly available arm's length data. The proof of the arm's length nature of applied transfer prices with routine entities based on comparability analyses is the best practical and reasonable approach, especially as this approach is already well established in the global TP practice. Step 2: To further facilitate a more effective use of comparability data on a global basis, the OECD should develop a regional approach for performing comparability analyses and move away from a pure local approach. A pure local approach could theoretically be refined without limitation, for example by a distinction between structurally weak regions and metropolitan regions within one country. Hence, a reverse approach by expanding the local market to a regional but still comparable market seems the most practical approach to find a sufficient and statistically more reliable data base for comparability analySeite 5 von 7

Transfer Pricing April 2014

ses and will also still be compliant with the arm's length principle. E.g. Europe as well as North America are in practice already seen as homogeneous markets and New Zealand accepts besides Australia also the "...markets in Europe (in particular the United Kingdom) and North America..." as comparable markets (see paragraph 20 of the OECD paper). Furthermore, an economic analysis of the EU Joint Transfer Pricing Forum (EUJTPF) performed in 2004 has already proven, that a comparability analysis for comparable transactions in a pan-European market provides results with the same quality and validity as country specific results (meaning the pan-European arm's length ranges of different industries do not statistically differ from country specific arm's length ranges of the same industries in almost all cases). Additionally, the results have the clear advantage of a bigger and hence more statistically reliable data basis respectively number of final com1

parables. This should also solve the likely problem that no or only very less local comparability data exists in a specific developing country. Step 3: Eventually we propose that the OECD should initiate an economic analysis (comparable to the EUJTPF analysis) on a global basis to analyse and develop guidance which countries and regions statistically can be summarized as homogenous, geographic markets. Furthermore, if statistically significant deviations between certain countries / regions are identified, an additional guidance could be developed if and how adjustments can be made to adapt regional / foreign arm's length data for local usage. We assume the above outlined steps to be the most efficient and practical approach(es) to solve the problems with comparability data in developing countries but also to further define a globally agreed and standardized approach for the usage of TP comparability data respectively conducting comparability analyses.

1

Vgl. EU Joint Transfer Pricing Forum: "Is Europe one market? A Transfer Pricing Economic Analysis of Pan European Sets", 24. Februar 2004. Seite 6 von 7

Transfer Pricing April 2014

Publisher WTS Steuerberatungsgesellschaft mbH www.wts.de • [email protected]

Contact/Editors Maik Heggmair, T: +49 89 28646-212, [email protected] Kai Schwinger, T: +49 69 1338456-56, [email protected] Michael Wohlfart, T: +49 89 28646-1509, [email protected]

Munich Thomas-Wimmer-Ring 1-3 • 80539 Munich • Germany T +49(0) 89 286 46-0 • F +49 (0) 89 286 46-111 Cologne Lothringer Straße 56 • 50677 Cologne • Germany T +49 (0) 221 34 89 36-0 • F +49 (0) 221 34 89 36-250 Düsseldorf Peter-Müller-Straße 18 • 40468 Düsseldorf • Germany T +49 (0) 211 200 50-5 • F +49 (0) 211 200 50-950 Erlangen Allee am Rötelheimpark 11-15 • 91052 Erlangen • Germany T +49 (0) 9131 97002-11 • F +49 (0) 9131 97002-12 Frankfurt Taunusanlage 19 • 60325 Frankfurt/Main • Germany T +49 (0) 69 133 84 56-0 • F +49 (0) 69 133 84 56-99 Hamburg Neuer Wall 30 • 20354 Hamburg • Germany T +49 (0) 40 320 86 66-0 • F +49 (0) 40 320 86 66-29 Raubling Rosenheimer Straße 33 • 83064 Raubling • Germany T +49 (0) 8035 968-0 • F +49 (0) 8035 968-150

The above information is intended to provide general guidance with respect to the subject matter. This general guidance should not be relied on as a basis for undertaking any transaction or business decision, but rather the advice of a qualified tax consultant should be obtained based on a taxpayer’s individual circumstances. Although our articles are carefully reviewed, we accept no responsibility in the event of any inaccuracy or omission. For further information please refer to the authors. Seite 7 von 7