Intellectual Property

Transcription

1 Tax Reference Library No 24 Intellectual Property (4th Edition) Published in association with: The Ballentine Barbera Group Ernst & Young FTI Consulting NERA Economic Consulting

2 Intangibles, market structure and the use of profit split methods 36 By Emmanuel Llinares, NERA Economic Consulting Transfer pricing specialists in many OECD countries (and in the US) tend to favour the transactional net margin method (TNMM - or its US cousin, the comparable profits method (CPM)) to assess the arm s-length nature of intra-group transactions. In practice, traditional methods (comparable uncontrolled price (CUP), resale price method, cost plus method) tend to be relatively difficult to use mainly due to relatively strict comparability requirements. On the other hand, the TNMM and the CPM are easier to develop in many cases and are viewed as relatively inexpensive in the context of preparing transfer pricing documentation. In practice, the majority of transfer pricing documentations seem to be based on either the TNMM or the CPM. Is this practice consistent with the economic profiles of multinational enterprises? Application of the TNMM and comparable searches in Europe Overview of the economic environment of MNEs MNEs decisions to expand international operations are motivated by economic considerations. In particular, the group may seek economies of vertical or horizontal integration. In addition, virtually all multinational enterprises rely on valuable intangibles. In this context, there are several reasons to cast doubts on the applicability of the TNMM given its current application in Europe. Typical comparable company searches in Europe In line with the OECD Transfer Pricing Guidelines, the TNMM as applied in most European countries relies on arm s-length evidence drawn from independent comparable companies. Criteria for independence may vary from one country to the other. A widely accepted view is that a company owned 50% or more by another company is not independent. Another important criterion is functional comparability which requires comparable companies to perform substantially similar functions. Companies engaged in diverse activities are typically not considered as comparable unless financial data are available for business segments. In general, and in accordance with the OECD Guidelines, when applying the TNMM, the tested party is the entity performing the simplest function, assuming lower risks and that does not own any valuable intangibles. In addition, databases of company financials in Europe contain information on a very large number of companies, including privately held companies in some European countries (notably the UK, France, Spain and Italy). In this respect, it is relatively easy to identify companies that perform only one function. On the other hand, the information available on those privately held companies is much more limited than the information available on publicly held companies. Comparable searches and economies of vertical and horizontal integration Due to the limitations of company financial data available in Europe and functional comparability issues, comparable company searches often result in the identification

3 of relatively small companies. A smaller size of company is not an issue to the extent that economies of vertical or horizontal integration do not affect comparability. In practice, this is particularly difficult to show. The first stage would be to identify the reasons for the economies of integration. These could include economies in transaction costs, logistics, brand development, risk management, purchasing synergies, and inventory management costs. The second stage would be to adjust for these differences between deemed comparable companies and tested parties which benefit from economies of integration. This stage is likely to require regression analysis to quantify the impact of those economies on profitability. However, transfer pricing practitioners, when performing comparable searches, rarely obtain sufficient sample size to perform those adjustments. Economies of integration are difficult to deal with in a simple, effective way and therefore are avoided by transfer pricing practitioners. As explained in the section below, using the profit split method generally eliminates the need for adjustments for economies of integration. Purchasing platforms offer a good example of difficulties associated with comparable searches when economies of integration exist. Assuming the purchasing platform yields lower per unit prices of inputs, which entity should earn the profits, if any, attributable with such savings? Should it be the purchasing platform itself, the related parties that purchase the products from the purchasing platform or the head office (where the MNE s management is located)? The allocation of profits, if any, should rely on a value-chain analysis of the MNE and of the industry value drivers. Where the answer is not obvious, it would be incorrect to use the TNMM to benchmark any of the entities that enables these cost savings since none of the deemed comparable companies identified is likely to have similar organization and/or size. In fact, using comparable searches in the presence of economies of vertical or horizontal integration would result in underremunerating (that is, below arm s length) any entity that is being benchmarked. There may also be, in some cases, diseconomies of integration due, for example, to increased compensations in large groups and/or the less direct link between compensation and performance in large groups. The use of comparable searches in this context is likely to result in over-remunerating the entities that are being benchmarked. Comparable company searches and intangibles When entities own or fund the development of intangibles, comparable company searches as they are commonly performed in Europe are unlikely to provide satisfactory results. As discussed above, most comparable company searches in Europe identify relatively small independent companies. In this context and unless it can be shown that these entities also develop valuable intangibles that contribute to the creation of Biography Emmanuel Llinares NERA Economic Consulting 28, avenue Victor Hugo Paris, France Tel: Fax: emmanuel.llinares@nera.com Emmanuel Llinares is a senior consultant at NERA, specializing in intercompany pricing analysis, economic analysis, and the economic valuation of tangible and intangible property. He advises multinational companies on defining and implementing their transfer pricing policies, and assists them with documentation. He specializes in the issues surrounding intangible assets, having assisted groups in the transportation, automotive, biotechnologies, pharmaceutical, telecommunications, business services and consumer electronics industries with respect to the definition of their transfer pricing policies, notably regarding technology and marketing-related intangibles. Llinares has developed innovative techniques relating to the pricing of intra-group financing. He has also managed several advance pricing agreements for multinationals in the automotive, financial services, consumer electronics, and consumer products industries. He has managed the transfer pricing aspects of numerous tax audits. Llinares holds a PhD in economics from the University of Delaware and has taught managerial and international economics at business schools in the US and in France. value in a similar way to that of the tested party, a comparable company search would probably fail to identify suitable comparables. In practice, transfer pricing specialists in Europe rarely use comparable searches to benchmark entities that fund the development of intangibles. Often, the approach is to benchmark entities that perform relatively simple functions. However, even in those circumstances, comparable searches (and the application of the TNMM) may not provide satisfactory results. For instance, ensuring that comparable companies do neither own nor develop valuable intangibles is often not possible with the accounting data available in Europe. Research and development costs or marketing expenses are rarely identifiable, and when they are identified, definitions used are not homogenous. Comparable company searches and complex functions In a number of industries, valuable functions that are integral to the business are performed in a number of jurisdictions and are made available to the group or business unit. Moreover, a large number of functions may not be performed in a few well defined centres but in many jurisdictions with significant 37

4 intercompany transfers of tangible goods, services, and intangibles among them. In those cases, a large number of comparable searches would be required to produce transfer pricing documentation. Relying solely on comparable company searches is then likely to be ineffective and less practical. The use of profit split methods Economic rationale The extent of economies of vertical and horizontal integration, the existence of intangibles on both sides of the transaction, and complex functional and transactional structures limit the use of standard approaches to economic analysis for transfer pricing purposes (that is, performing comparable company searches to apply the TNMM). In those circumstances, the arm s-length nature of transactions can not be verified relying exclusively on comparable searches given insufficient comparable company data in Europe. In such circumstances, profit split methods may be the only applicable method. For a detailed description of the transfer pricing methods, we refer the reader to Harlow Higinbotham, Profit split methods in Robert T. Cole s Practical Guide to US transfer pricing, chapter 10. One of the key differences between TNMM and profit split methods is that while the TNMM generally analyzes financials of one side of the transaction (the tested party, which is often the entity performing the simplest functions, assuming lower risks and that does not own valuable intangibles), profit split methods rely on an analysis of combined profits of the transacting parties instead. The purpose of any profit split analysis is the allocation of combined profits. From a computational aspect, these methods therefore require the preparation of segmented profit and loss and balance sheet statements. Profit split methods enable accounting for the specificities of the MNE and the industry while relying on market evidence, therefore offering an arm s-length answer when there are significant intangibles on both sides of the transaction, economies of integration and when comparables do not exist. Finally, regardless of the economic environment, from a practical point of view, one-sided methods such as the TNMM are more likely to yield absurd results. Since only one side of the transaction is benchmarked, tax authorities from both sides of the transaction may end up disagreeing with the taxpayer. A two-sided method such as the profit split method should generally minimize the risks that those issues arise. Overall profit split method contribution analysis Under this type of analysis, the combined operating profits are split among the participants in the value chain based on the relative value of the functions performed. To justify this split, the OECD Guidelines recommend the use of external market data. In practice, the determination may be based on an analysis of the relative value of each participant s contribution that can either be based on external (such as data 38 from joint ventures for example) or internal data (including financial data). This method is particularly suitable when direct comparables cannot be used. When the split of the profits is based on internal financial data (such as the value of assets or R&D costs) to allocate profits, the method effectively results in a rate of return pricing. One of the advantages of this method is that it can be used when no suitable comparable can be identified. However, as recommended by the OECD, performing a contribution analysis should be based on a detailed analysis of functions, assets, risks, and variables used to divide combined profits among participants should be economically justified. Otherwise, the overall profit split method may be confused with the global apportionment formula. This method can be used in situations where economies of integration differentiate the tested party from the comparables provided these are taken into account in the principle for allocating profits. In this respect, game theory and bargaining theory are valuable tools when applied in a quantifiable way. For the same reason, the method enables accounting for valuable intangibles being developed on both sides of the transaction. The allocation of profit can be calculated based on principles that take into account the contribution of intangibles in the industry s or the group s value creation process. Finally, this method is also well suited to complex transactions where several entities are involved in the same functions and it is not possible to define precisely the scope of functions and responsibilities. While this method has not been widely used, it has been somewhat popular in the financial services industry. For instance, when analyzing global trading activities, trader s actual bonuses have been used as a way to split profits associated with those activities. Comparable profit split method In the comparable profit split method, combined operating profits of participants are split among the participants in the value chain based on data from observed comparable transactions between third parties. Comparable transactions should refer to a division of profits and should satisfy comparability criteria (similar activities and circumstances). The US transfer pricing regulations highlights three main issues relating to the use of this method: comparability requirements, which are similar to those of the Comparable Profit Method (a cousin of the TNMM); similarity of contractual terms in both the controlled and uncontrolled transactions; and in order to apply this method reliably, the combined operating profits (as a percent of the combined assets for example) of the uncontrolled transactions should not vary significantly from that of the controlled taxpayer. This method has been relatively popular in royalty analysis where licensing agreements with third parties available withwww.internationaltaxreview.com

5 in the group are not sufficiently comparable (notably in terms of sales potential for example) to the tested transaction to apply the CUP method reliably. In this context, the method could rely on net present values of operating profits of the licensor and the licensee to define comparable allocations of profits. These comparable allocations of profit are then mapped to forecasted financials of tested transactions to determine an arm s-length royalty rate. This analysis could even be performed in circumstances where only data from the licensee is available (then, the underlying assumption is that the licensor does not incur costs once the intangible is licensed, hence its operating profits are measured by the royalty income received). Residual profit split The residual profit split method relies on a characterization of functions, risks and assets according to which routine functions are distinguished from non routine/entrepreneurial functions. Routine functions are presumably relatively simple functions (that could be sub-contracted, for example). In terms of market structure, these functions are likely to correspond to perfectly competitive market structures or monopolistically competitive market structures, where there are no barriers to entry. These functions should require no valuable intangibles and involve limited risks. In this respect, they fit both with the conditions for the application of the TNMM and the availability of comparable companies financial data in Europe (that is, a large number of relatively small independent companies). Entrepreneurial functions and/or risks on the other hand, involve financing the development of valuable intangibles and/or assuming substantial risks. Entrepreneurial functions fit well with the risk profile of investment centres, that is, profit centres in which responsibility is linked to the continuity of the business. Once the first step of the analysis has been performed and that arm s-length returns have been attributed to routine functions, residual profits are split between affiliates based on one or more split factors. These factors can be similar to that of the contribution analysis. As an example, when the residual profit split method is applied to industries where there is a stable nexus between investments in technology and returns, development costs of those intangibles (R&D costs) have been used to split residual profits between those entities that funded the development of those intangibles. In those cases, implementation will necessitate an analysis of those expenses and in particular the useful life of the corresponding intangible. One of the key features of this method is that it enables some sophistication in terms of apprehending the value drivers. The distinction between routine and entrepreneurial functions, while sometimes too schematic and caricatured, remains a useful reference point which enables a relatively simple characterization of functions, risks and assets. One Table 1: Three profit split methods Profit split method Contribution analysis Comparables profit split Residual profit split Typical transfer pricing issues Integrated and complex function, risks and assets Significant risks on both sides of the transaction Determination of royalties Valuable intangibles developed on both sides of the transaction. Entrepreneurial risks on both sides of the transaction Industries where use is common use Financial services Pharmaceutical industry Consumer electronics Automotive, Consumer electronics, financial services industry issue is that often, the line between routine and non-routine functions risks and assets is thin. In addition, the method, which is not complex to use in many cases, requires determining the relative contribution of each intangible when dealing with more than one category of intangibles. The residual profit split method has probably been the most commonly used profit split method. Industries in which it has been particularly popular include the automotive, consumer electronics and financial services industries. In the latter case, key entrepreneurial activities generally relate to risks rather than intangibles. In industries where intangibles results from significant R&D activities, residual profit allocations are frequently based on capitalized R&D expenditures. Method application summary The table above provides a short summary of the use of the above three profit split methods. Benefits and drawbacks Regardless of the profit split method chosen, one of its key benefits is that it enables taking into account the specificities of the industry and of the group. Notably, profit split methods can account for economies of integration (with the use of game theory for example) and they correctly take into account the returns associated with valuable intangibles. The greater the economic robustness of the principle for the allocation, the greater the explanatory power of the method. Among its drawbacks, taxpayers often cite the need (except perhaps for the comparable profit split approach), to reveal financials from both sides of the transaction. This may be a critical issue in the context of tax audits for example. However, the nature of the transactions at stake and their 39

6 complexity may require such analysis anyway. In addition, applying a profit split method generally requires the taxpayer to reveal financials from the other side of the transactions, thereby providing upfront the information that would be required in a competent authority negotiation. Doing so increases the credibility of a strategy which would involve launching an arbitration procedure. Other drawbacks of the method include its perceived complexity (notably when it is used as a tool to define transfer prices) and its requirements in terms of data (data from multiple related parties may be required). While the analysis required may be sophisticated, the data needed are the same that are used in the context of preparing solid transfer pricing documentation for the parties involved. In fact, an MNE with suitable transfer pricing management resources may prefer profit split methods from a pure resources allocation perspective: since profit split methods are inherently capable of documenting both sides of the transactions, it reduces the need, or the work required, for country-specific documentation and may facilitate the budgeting work. group that are characterized by economies of scale or for which valuable intangibles are being developed in several jurisdictions, or with very complex functions, one may be doubtful about the reliance of transfer pricing policy (and documentation) on only one method, in particular if this method is the TNMM. OECD preference When considering transactional profit-based methods, the OECD Guidelines seem to give their preference to the profit split method rather than the TNMM, notably insisting on comparability issues (section 3.53). In addition to the above, profit split methods, unlike the TNMM, can reliably take into account both sides of the transactions instead of benchmarking only one side (traditionally, the entity performing the simpler functions). This feature of profit split methods should satisfy taxpayers since it is inherently less likely to yield extreme results (profits in one jurisdiction and large losses in the other jurisdiction). This feature should also satisfy tax authorities that are often asking for more transparency. Recently, Maité Gabé from the French tax authorities explained that When the functional analysis is being performed, one should not ignore the nature and the financials of the foreign related party. It is also consistent with the increased transparency promoted by the EU Joint Transfer Pricing Forum. Several factors militate against an increasing use of the profit split methods: The strong reliance of multinational enterprises on the development and the use of intangibles; The existence of economies of integration which are often a key motivation for the global coverage of multinational enterprises and the inability to identify suitable comparables taking into account those factors thus limiting the use of the other methods; and The practical aspects of the methods and the fact it should minimize the risks of obtaining extreme results. A robust transfer pricing policy cannot ignore those aspects. In fact, when considering operations within a multinational 40

7 For more information or to subscribe to International Tax Review please contact: UK Hotline: +44 (0) or US Hotline: