About The Transfer Pricing Discussion Group

Transcription

1 Selecting The Most Appropriate Method and The Appropriate Roles for Profit Methods th Street, N.W. Washington, D.C (202) January 2008 OECD Transactional Profit Methods Discussion Draft OECD Consultation November 18, 2008 Steven P. Hannes McDermott, Will & Emery Washington, D.C. Representing The Transfer Pricing Discussion Group Boston Brussels Chicago Düsseldorf Houston London Los Angeles Miami Munich New York Orange County Rome San Diego Silicon Valley Washington, D.C. Strategic alliance with MWE China Law Offices (Shanghai) 2008 McDermott Will & Emery LLP. McDermott operates its practice through separate legal entities in each of the countries where it has offices. This communication may be considered attorney advertising. Previous results are not a guarantee of future outcome. About The Transfer Pricing Discussion Group The Transfer Pricing Discussion Group consists of multinational companies based in the United States and other countries. The members are in the auto, chemicals, consumer nondurable goods, durable goods, and pharmaceuticals industries, among others. The Discussion Group serves as an information source for corporate people who specialize in work on transfer pricing and related matters and who are interested in periodically discussing these matters with their corporate peers. Also, the Group prepares comments on government proposals involving transfer pricing and related matters. In a letter from McDermott Will & Emery, dated May 8, 2008, the Group offered comments to the OECD on the 2008 Transactional Profits Methods Discussion Draft. The Group was organized by Steven Hannes of McDermott Will & Emery in Any inquires about the Group should be directed to Mr. Hannes. 2 1

2 Selecting The Most Appropriate Transfer Pricing Method Under The Arm s Length Standard Key OECD Objectives Providing guidance on how to make the selection consistently, and with market place reliability being the key goal, based on the facts and circumstances Removing an unjustified last resort stigma from Transactional Profit Methods Comments The Transfer Pricing Discussion Group strongly supports these goals in general and offers suggestions to help achieve them Discussion of the Alpha, Beta and Gamma selection categories General Observations About Selection Standards for TNMM And Otherwise A critical issue under the arm s length standard in selecting is, or should be, one of reliability, based on what has actually occurred between third parties in the market place In the first instance selection should focus on the comparability of the market place benchmarks and the reliability of their data rather than on the particular method 3 4 2

3 General Observations There may be too much emphasis today on selecting methods rather than on making selections based on the comparability of market place benchmarks and the reliability of their data E.g., paragraphs proposed for the TP Guidelines are worded in terms of finding the most appropriate method and taking into account the respective strengths and weaknesses of each of the OECD recognized methods (emphasis supplied) General Observations A selection rule should take into account that the names of methods are subject to differing interpretations and/or are being misused in practice 5 6 3

4 General Observations A successful appropriate method rule will (1) restore and maintain consistent focus on the comparability of the benchmarks and the reliability of their market place data from actual third party dealings and (2) adjust the emphasis on methods, clarify their names/labels, or somehow assure that these factors do not interfere with the selection rule s consistent and appropriate application General Observations A successful appropriate method selection rule will in the first instance: consistently and naturally choose certain benchmarks (and then methods) over others 7 8 4

5 General Observations In the second instance, a successful rule will encourage making comparability adjustments, even for material differences of various types (functional, financial, data reporting, qualitative, etc.) between the benchmarks in a category and the taxpayer, and emphasize that making such adjustments is generally preferable to abandoning the benchmarks and moving to the next category General Observations E.g., making comparability adjustments, even if imperfect and even for material and subjective differences, such as quality, between one or more benchmarks of certain types and the taxpayer is generally preferable to rejecting these benchmarks and relying instead on another category It may also frequently be preferable, where comparability adjustments cannot be reasonably made, to accept the imperfect comparability of certain benchmarks rather than relying on another category

6 General Observations In the third instance, the selection rule should focus on whether, within an applicable category of benchmarks, more than one method (e.g., gross profit versus operating profit) is available and reliable and, if so, how to select the most reliable one Selection Process: Three Basic Transfer Pricing Categories* A Framework Within Which To Discuss The OECD Goals ALPHA Category BETA Category GAMMA Category The benchmark (comparable) uses Micro market data The benchmark (comparable) uses Macro market data This is largely theorybased 11 * Note that there also may be transfer pricing safe harbors (e.g., for interest rates) or elections (e.g., cost-only for some services or a sharing of costs under specific conditions for activities such as R&D) that, as such, are taxpayer-selected and thus are not categorized as Alpha, Beta, or Gamma 12 6

7 Data Used in The Basic Selection Categories Illustrations of The Basic Selection Categories Financial data from one or more specific market place agreements (the agreements typically are, but do not have to be, in a written contract negotiated between the parties) Financial data from aggregates of market place business activities of a person (which may be a consolidated group of related persons), or, most typically, of a number of such persons Financial data from market place business activities are not relied upon (or perhaps highly generalized or hypothesized market place information is used) 13 Examples: the data source (benchmark) is a contract for the manufacture or purchase of property, a contract for the provision of a service, a license of intellectual property, a contractual sharing of R&D costs. Examples: the data source (benchmark) is the overall financial result of a company and its affiliates or, more typically a group of such companies. Usually the companies are publicly traded and disclose their results without much financial detail and with little or no detail about the terms and conditions of the underlying contracts, licenses, etc. Examples: in theory, two parties engaged in highly integrated transactions, or unique transactions, should split profits 50/50 or 70/30; in theory a party should charge for the cost of stock-based compensation. 14 7

8 Financial Measures in The Three Categories Price, fee, royalty and other similar payments Gross profit Operating (net) profit or margin profit Split of profits Sharing of income or costs (e.g., R&D) Other data from one or more specific market place agreements Gross profit Operating (net) profit or margin profit Other data from aggregates of market place business activities of a group of persons Theory-based split of profits Theory-based sharing of income or costs (e.g., R&D) Theory-based determination or calculation of costs or of other items Theory-based formulas (e.g., worldwide apportionment) OECD Descriptors/Labels for Methods Comparable Uncontrolled Price Method Resale Price Method Cost-Plus Method Traditional Transaction Methods Transactional Profit Split Method Transactional Net Margin Method Transactional Profit Methods

9 Matching The OECD Descriptors/Labels With The Three Categories Traditional Transaction Methods Price, fee, royalty and other similar payments Gross profit Operating (net) profit profit Split of profits Sharing of income or costs (e.g., R&D) Other data from one or more specific market place agreements Gross profit Operating (net) profit profit Other data from aggregates of market place business activities of a group of persons. Theory-based split of profits Theory-based sharing of income or costs (e.g., R&D) Theory-based calculation of costs or of other items Theory-based formulas (e.g., worldwide apportionment) 17 Matching The OECD Descriptors/Labels With The Three Categories Transactional Net Margin Method Price, fee, royalty and other similar payments Gross profit Operating profit profit Split of profits Sharing of income or costs (e.g., R&D) Other data from one or more specific market place agreements Gross profit??? Operating (net) profit profit Other data from aggregates of market place business activities of a group of persons? Theory-based split of profits Theory-based sharing of income or costs (e.g., R&D) Theory-based calculation of costs or of other items Theory-based formulas (e.g., worldwide apportionment) 18 9

10 Matching The OECD Descriptors/Labels With The Three Categories The Other Transactional Profit Method: Profit Split Price, fee, royalty and other similar payments Gross profit Operating (net) profit profit Split of profits Sharing of income or costs (e.g., R&D) Other data from one or more specific market place agreements Gross profit Operating (net) profit profit Other data from aggregates of market place business activities of a group of persons Theory-based split of profits Theory-based sharing of income or costs (e.g., R&D) Theory-based calculation of costs or of other items Theory-based formulas (e.g., worldwide apportionment) 19 Selection Rule Oriented Observations About OECD Descriptors/Labels TNMM encompasses market place benchmarks, with materially different levels of reliability Traditional Transaction Method only encompasses some of the market place benchmarks with the highest level of reliability Transactional Profit Methods encompass market place and theory-based benchmarks with materially different levels of reliability These observations have substantive implications for drafting a successful substantial most appropriate selection rule; the observations are not semantic 20 10

11 Selection Observations OECD selection-related discussions of reliability, and relative reliability among, profit methods should carefully distinguish between Alpha, Beta and Gamma situations Not all profit methods are created equal in terms of reliability For example, an Alpha profit-based application can be highly reliable in market place terms and, in any case, will ordinarily be far more reliable than a Beta application, let alone Gamma Selection Observations The OECD is correct in removing an unjustified last resort stigma from the use of TNMM in the Beta category In practice, the application of TNMM is wide and overwhelmingly, if not entirely, a Beta use (i.e., it is based on the overall financial results of companies) In this sense, and perhaps others, the Transactional reference in TNMM is misleading, or perhaps the method is NOT being applied in practice as was originally intended by the OECD* *The same applies to Transactional Profit Methods and Transactional Profit Split

12 Selection Observations In any case, as part of the OECD Discussion Draft s goal of moving away from the method of last resort stigma for Beta applications of TNMM and any other Beta profit methods, it will be helpful if the OECD selection rule explains and illustrates how to adjust Beta benchmarks and that it is generally far more reliable to adjust them than to select and apply Gamma This will move the OECD rules, appropriately, towards what is occurring in practice: the frequent and appropriate use of Beta TNMM. It will also help improve the reliability and consistency of Beta TNMM and restrict and discourage the use of Gamma (theory-based) profit methods Selection Observations Taxpayers and tax authorities use Gamma by agreement (e.g., an APA or audit settlement), a small but significant percentage of the time. Very few taxpayers choose to use Gamma on their own. U.S. courts have been highly skeptical of Gamma approaches The abandonment of the last resort stigma must clearly distinguish between methods in the three categories that rely on profits to avoid (1) weakening the role of comparability and reliability in selecting market place benchmarks (Alpha, Beta) and (2) increasing the use of theory-based approaches (Gamma)

13 CONCLUSIONS The OECD s most appropriate method and related TNMM goals are commendable, important, and can be achieved A most appropriate selection approach can and should be utilized, and TNMM (in the Beta sense) should no longer bear the last resort stigma A most appropriate selection approach that successfully applies to TNMM and other categories of methods/benchmarks can and should be clear enough so that it will be supportive of consistently selecting the most reliable and comparable benchmarks from those available There is a natural order to selecting the most appropriate benchmarks and then methods Not all profits methods are created equal, and they should be distinguished from one another for purposes of a most appropriate selection approach The OECD should continue to discourage and carefully restrict the use of theory-based approaches, in particular Gamma profit splits 25 The OECD Most Appropriate Method Approach and The Natural Selection Order Among Categories* Price, fee, royalty and other similar payments Gross profit Operating (net) profit profit Split of profits Sharing of income or costs (e.g., R&D) Other data from one or more specific market place agreements Gross profit Operating (net) profit profit Other data from aggregates of market place business activities of a group of persons. *This page does not address any natural order within each category *** Theory-based split of profits Theory-based sharing of income or costs (e.g., R&D) Theory-based calculation of costs or of other items Theory-based formulas (e.g., worldwide apportionment) 26 13