Resale Price and Cost-Plus Methods: The Expected Arm s Length Space of Coefficients

International Alessio Rombolotti and Pietro Schipani* Resale Price and Cost-Plus Methods: The Expected Arm s Length Space of Coefficients In this article, the resale price and cost-plus methods are considered from the perspective of the manufacturer and the seller, and it will be shown that these two perspectives, in fact, lead to one single framework underlying the setting of transfer prices. 1. Introduction This article will focus on two widely used and accepted transfer pricing methods, namely the resale price and costplus methods. These methods will be considered from two different functional perspectives: the manufacturer s perspective, where profits are perceived as a percentage of manufacturing costs, and the seller s perspective, which is more likely to look at its profits as a percentage of its sales volume. Therefore, the two methods respond to different functional roles along the supply chain, and this difference is reflected in the two fundamental aspects of the two methods, namely the discount and markup coefficients. The first approach, related to the resale price method, affects the market price (or resale price), while the second, related to the cost-plus method, affects the manufacturing costs. These methods are designed to produce transfer prices which conform to the arm s length principle, and it would be logical to conclude that their conformity relies on the choice of the discount and markup coefficient levels. In general, the application of a transfer pricing method to several transactions will define the arm s length range, i.e. the range over which transfer prices vary in magnitude. Such range depends on both the method used and the actual transactions being taken into account. When using the resale price and cost-plus methods, the arm s length range is brought about by the chosen coefficients, but nevertheless the arm s length range varies as different transactions are considered. This fact prevents one from knowing what the possible outcomes of the application of the methods can be prior to actually performing the calculations, i.e. benchmarking. Here the aim is to define in advance what those possible outcomes can be by means of what can be called the arm s length space of expectations. It seems reasonable to ask whether it is possible to extract an arm s length range which had the feature of being objective, in the sense that it did not depend on specific transactions; in other words the search is for an expected arm s length range. In the case of the resale price and cost-plus methods, the answer is yes, as what is found is an expected space from which all possible ranges can be drawn. When graphed, such space is three-dimensional, triangular and objective, i.e. apart from being subject to the condition that any party involved has realized a profit, it does not depend on specific transactions. 2. A Review of the Resale Price and Cost-Plus Methods The resale price and cost-plus methods offer an intuitive approach to transfer pricing, and their application is rather simple. The OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (2009) (hereinafter OECD Guidelines) outline the procedures for applying these two methods. With regard to the resale price method, the OECD Guidelines provide as follows: The resale price method begins with the price at which a product that has been purchased from an associated enterprise is resold to an independent enterprise. This price (the resale price) is then reduced by an appropriate gross margin (the resale price margin ) representing the amount out of which the reseller would seek to cover its selling and other operating expenses and, in the light of the functions performed (taking into account assets used and risks assumed), make an appropriate profit. What is left after subtracting the gross margin can be regarded, after adjustment for other costs associated with the purchase of the product (e.g. customs duties), as an arm s length price for the original transfer of property between the associated enterprises. This method is probably most useful where it is applied to marketing operations. 1 An appropriate resale price margin is easiest to determine where the reseller does not add substantially to the value of the product. In contrast, it may be more difficult to use the resale price method to arrive at an arm s length price where, before resale, the goods are further processed or incorporated into a more complicated product so that their identity is lost or transformed (e.g. where components are joined together in finished or semi-finished goods). Another example where the resale price margin requires particular care is where the reseller contributes substantially to the creation or maintenance of intangible property associated with the product (e.g. trademarks or trade names) which are owned by an associated enterprise. In such cases, the contribution of the goods originally transferred to the value of the final product cannot be easily evaluated. 2 With regard to the cost-plus method, the OECD Guidelines provide as follows: The cost plus method begins with the costs incurred by the supplier of property (or services) in a controlled transaction for property transferred or services provided to a related purchaser. An appropriate cost plus mark up is then added to this cost, to make an appropriate profit in light of the functions performed and the market conditions. What is arrived at after adding the cost plus mark up to the above costs may be regarded as an arm s length price of the original controlled transaction. This method probably is most useful where semi finished goods are sold between * Valente Associati GEB Partners, Italy. 1. Para. 2.14 OECD Guidelines (2009). 2. Para. 2.22 OECD Guidelines (2009). 186 INTERNATIONAL TRANSFER PRICING JOURNAL MAY/JUNE 2012 IBFD

Resale Price and Cost-Plus Methods: The Expected Arm s Length Space of Coefficients Figure 1 related parties, where related parties have concluded joint facility agreements or long-term buy-and-supply arrangements, or where the controlled transaction is the provision of services. 3 Pricing methods, among other things, are chosen with regard to functions and assets involved in the tested transactions. In practice, one must deal with many different function/asset configurations within the parties involved and quite frequently there may be some difficulty in obtaining a clear picture. For this purpose, the authors refer to a manufacturer-distributor-retailer structure that does not involve intangibles or special factors that might affect the choices to be made. However, the results found are general and are comprehensive for any function/asset configuration, provided that the resale price and cost-plus methods come into play. The formulation of the concepts expressed above can be written as follows: TP = Market Price (1 discount) = P (1 D) for the resale price method, and TP = Costs (1 + markup) = C (1 + M) for the cost-plus method. In the formulas above TP stands for transfer price, D for discount, C for costs and M for markup. As any transfer price must incorporate a profit for the seller, one can set the intervals over which the discount and markup coefficients may vary: 0 < D < 1 and 0 < M When D drops below zero, the discount is negative and the transfer price becomes greater than the market price; when D is greater than one, the transfer price becomes negative; and when M is less than one, the seller realizes a straight loss. Once the method has been chosen, it can be tested against uncontrolled transactions, and hence one can derive and fix the coefficients to be inserted into the transfer pricing formula. The set of outcomes resulting from the application of the method defines the arm s length range. The OECD Guidelines define the arm s length range as follows: A range of figures that are acceptable for establishing whether the conditions of a controlled transaction are arm s length and that are derived either from applying the same transfer pricing method to multiple comparable data or from applying different transfer pricing methods. 4 Regarding the use of an arm s length range, the OECD Guidelines provide as follows: In some cases it will be possible to apply the arm s length principle to arrive at a single figure (e.g. price or margin) that is the most 3. Para. 2.32 OECD Guidelines (2009). 4. Glossary, OECD Guidelines (2009). reliable to establish whether the conditions of a transaction are arm s length. However, because transfer pricing is not an exact science, there will also be many occasions when the application of the most appropriate method or methods produces a range of figures all of which are relatively equally reliable. In these cases, differences in the figures that comprise the range may be caused by the fact that in general the application of the arm s length principle only produces an approximation of conditions that would have been established between independent enterprises. It is also possible that the different points in a range represent the fact that independent enterprises engaged in comparable transactions under comparable circumstances may not establish exactly the same price for the transaction. However, in some cases, not all comparable transactions examined will have a relatively equal degree of comparability. Therefore, the actual determination of the arm s length price necessarily requires exercising good judgment. 5 3. Derivation of the Coefficient Space The baseline entity for this purpose is a multinational enterprise with three basic functions: manufacturing, distribution and retail. The manufacturing functions are performed at facilities in strategic manufacturing locations around the world, distribution functions are executed at hubs located near marketplace areas and retail functions are performed by local retailers. In compliance with the OECD Guidelines, the manufacturer adopts the cost-plus method, charging C (1 + M) to the distributor. The distributor adopts the resale price method, charging P (1 D) to the retailer that is selling to unrelated parties at market price, P. The transaction flow is illustrated in Figure 1. In the scenario in Figure 1, coefficients D and M are yet to be fixed, although interval conditions, as previously established, still apply. Now, as C (1 + M) is the cost to the distributor, its sales price of P (1 D) can always be expressed in the following formula (Equation 1): P (1 D) = C (1 + M) (1 + S) where S, any number greater than or equal to zero, is the distributor s markup. As the core components of the pricing methods are their coefficients, the focus remains on D, M and S. One can write the manufacturer s markup, M, in terms of the others coefficients as (Equation 2): M = P 1 D 1 C 1 + S In Equation 2, M depends on coefficients D and S, and also depends on the ratio between market price and manufacturer s costs. In this scenario, coefficients behave as variables, and each one of them can be regarded as being a function of the others. On the other hand, the ratio P:C behaves as a constant. In addition: 5. Para. 1.45 OECD Guidelines (2009). IBFD INTERNATIONAL TRANSFER PRICING JOURNAL MAY/JUNE 2012 187

Alessio Rombolotti and Pietro Schipani Valente in practice, the roles assumed by the equation s components are reversed, such that coefficients behave as constants whereas costs and market prices behave as variables; and in Equation 2, P and C are combined in one fractional number; they are not considered separately, as their ratio is the key magnitude of the equation. To conform to traditional notation of mathematical functions, names are changed as follows: D = x S = y M = z and set: P C = Ψ Equation 2 then becomes (Equation 3): z = 1 x 1 + y Equation 3 is the final form of the transfer pricing coefficients equation. As far as the resale price and cost-plus methods are concerned, the graph of the equation illustrates the space with all possible combinations of the coefficient values, provided that they fall into their admissible intervals, i.e. no party is allowed to suffer a loss. See Figure 2. Figure 2 Ψ 1 z = 1 x 1 + y Ψ 1 The space of coefficients is the surface over which the coefficients take on their values, and the line where z = 0 marks the threshold below which the value of z is negative and hence not acceptable based on the assumption that any party involved must realize a profit. The space is shaped as a triangle and uniformly slopes downwards. To be consistent with the previously established interval limits, one may write: 0 < x < 1 0 < y 0 < z However, each coefficient can no longer be treated irrespective of the others; instead they are now all interconnected with each other by the parameter Ψ, which is the one factor shaping the whole surface and fixing the endpoints of the coefficients intervals. The new and actual conditions have changed from the initial set of conditions, and thus, there is now: 0 < x < 1 Ψ 1 0 < y < Ψ 1 0 < z < Ψ 1 The new intervals are now expressed as a function of the parameter Ψ; the y and z upper limits move proportionally to Ψ, whereas the x upper limit is proportional to the reciprocal of Ψ. The parameter Ψ induces the extension magnitude of the whole surface: the greater the Ψ, the greater the surface area, i.e. the greater the ratio between market price and the manufacturer s costs, the greater the number of possible arm s length range choices. Special cases arise when is set, i.e. the parties involved in the transactions are the distributor and the retailer only (in this case, the role of the manufacturer s costs, C, is played by the costs of the distributor); when, i.e. the parties involved in the transactions are the manufacturer and the retailer; when, i.e. the parties involved in the transactions are the manufacturer and the distributor only. These situations represent the limits of the space and are governed by the relations below: when x = 0 z = Ψ (1 + y) 1 1 when y = 0 z = Ψ (1 x) 1 when z = 0 x = 1 (1 + y) Ψ 1 These relations are linear between z and x and between y and x, but hyperbolic between z and y. The space of the resale price and cost-plus coefficients has been identified by expressing z, the markup coefficient, as a function of the others. If one specific point in such space is marked by drawing perpendicular lines onto to the x, y and z axes, one can read the exact value of the other coefficients which define that very point in space. In order to derive the coefficients space, the authors had to reverse the view they usually adopt; in practice, what is variable is here treated as constant, and vice versa what is constant is here treated as variable. Such space is not a statistical magnitude itself, as the relations identified 188 INTERNATIONAL TRANSFER PRICING JOURNAL MAY/JUNE 2012 IBFD

Resale Price and Cost-Plus Methods: The Expected Arm s Length Space of Coefficients Table 1 are determinative; however as it depends on the parameter Ψ, which is a function of costs and prices, it may well be considered, at least to a certain extent, as a stochastic magnitude. It is rather clear that in this model, Ψ is the key factor, as it governs the shape and the amplitude of the surface and hence, it sets the boundaries of any transfer pricing strategy where the resale price and cost-plus methods are applied. 4. A Practical Example Choosing the proper transfer pricing strategy is a multistep process, beginning with the functional analysis of parties involved in controlled transactions. The appropriate transfer pricing method is then selected and applied to uncontrolled comparable (as to functions, risks and assets employed) transactions, and eventually the arm s length range is determined. In this case, the first step would be the one that is usually the last: the derivation of the expected arm s length range, and the point of the argument is that the calculations are performed by estimating the Ψ parameter, without taking into account any actual transaction. Assuming that one is dealing with three groups of associated enterprises operating in unrelated markets, and that the cost-plus method has been chosen, in each case, first make three estimations, one per market sector, of Ψ, which gives the triple: Ψmarket 1 = 1.25 Ψmarket 2 = 2.50 Ψmarket 3 = 5.00 Without loss of generality, assume that the market price in each market sector is 100. These figures allow one to make assessments of (1) values taken on by coefficients in the cost-plus method and (2) the expected arm s length range, as shown in Table 1. Lower and upper markups are set according to the limits imposed by the Ψ parameter in the equation: 0 < z < Ψ 1 Each market condition, i.e. each Ψ estimate, induces its markup range which includes all possible appropriate markup values and is ultimately defined by its endpoints. Naturally, as one may have additional information or the background as to where that range should be located, one may choose just a subset of the whole range by adding an extra constraint. In this example, the three ranges are graphed to compare their amplitude. A set was selected, with regular spacing, of 12 figures from each interval, and their relative transfer price was computed by inserting the markup into the cost-plus formula. See Figure 3. To determine whether the range amplitude is proportional to the Ψ parameter, test the sensitivity of the range amplitude to the Ψ parameter by allowing Ψ to vary from 1.25 to 5.00 and plotting, at each Ψ value, ités corresponding range length. See Figure 4. As shown in Figure 4, the relation is not proportional, i.e. a straight line which as intuition suggests slopes upward, but the rate of growth decreases with Ψ. This information reveals that in situations of very high market price-to-cost ratios, one should expect much more freedom in setting the pricing coefficients than in low ratio cases, but the growth of such freedom decreases as Ψ increases. The function of the expected arm s length ranges is to provide a wider background to the planning, prior to any benchmarking analysis, of the transfer pricing strategy at each Ψ level, for each group of associates. Moreover, IBFD INTERNATIONAL TRANSFER PRICING JOURNAL MAY/JUNE 2012 189

Alessio Rombolotti and Pietro Schipani Valente Figure 3 Figure 4 any point in the expected range may be thought of as a possible position of the associates with respect to diverse functions, risks and assets employed, and it will be possible to locate, along the interval, all benchmarked uncontrolled transactions. 5. Conclusion This article has aimed to show that the arm s length range of the resale price and cost-plus methods can be assessed prior to the benchmarking analysis, and that this fact entails some benefits. This can be done because the pricing method coefficients, which are the core factors of these very methods, depend solely on the parameter Ψ = Market Price Cost Hence, it suffices to estimate Ψ in order to fix all possible coefficients, subject to the constraint that any party involved in controlled transactions must earn a profit. To achieve the final result, it is necessary to move backwards by reversing the view usually adopted in practice, treating variables as constants and vice versa, and taking Ψ as the sole factor from which a three-dimensional surface is drawn. The space thus defined can be referred to as the expected arm s length space, as it contains all possible expectations as to the arm s length range of the resale price and cost-plus methods, and it provides the background and a broader picture of pricing dynamics. In the authors model, a plain manufacturerdistributor-retailer configuration was assumed. However, what if intangibles or other special functions or asset structures come into play? The authors results are valid in general terms, although variations in function/asset configurations will place a constraint on the expected ranges which translates 190 INTERNATIONAL TRANSFER PRICING JOURNAL MAY/JUNE 2012 IBFD

Resale Price and Cost-Plus Methods: The Expected Arm s Length Space of Coefficients into dealing with an appropriate subset of the initial set of coefficients and hence, smaller ranges. The space of the coefficients is itself determinative, i.e. no probability measure was used to establish the results. Nevertheless, it depends upon the Ψ parameter, which may very well be seen as a statistical magnitude. Consideration of the implications of this fact is beyond the scope of this article. The expected arm s length space can be used to plan a transfer pricing strategy over an objective set of values that has not been influenced by previously tested transactions, and can be used, so to speak, to benchmark the benchmarking analysis. For instance consider the example where Ψ = 1.25. If, in a tested transaction, the markup of the manufacturer is identified as being 8%, by inserting the Ψ and the markup of actual values into Equation 3, prior to any benchmarking, it can be determined that no associated distributor can provide a discount greater than 13.6%. On the expected arm s length range, as it provides all possible pricing levels within a specific Ψ-industry sector (i.e. market sector labelled by its specific Ψ), all transactions of any player belonging to that specific market sector will be able to be located, with the expectation that if consistency holds, all transactions referring to similar asset/function configurations will be clustered around an average point. ONLINE New! IBFD Mobile Tax Facts Enjoy the convenience of accessing essential tax facts from your mobile phone IBFD Mobile Tax Facts is a mobile website optimized for smartphone browsing: it loads quickly and is easy to navigate. This cost-effective package enables you to access tax news, global key features and quick reference tables directly from your mobile phone. Three sets are available: Full set: 150 / $ 200 (VAT excl.) Light set: 100 / $ 135 (VAT excl.) VAT set: 75 / $ 100 (VAT excl.) Trial and ordering Go to www.ibfd.org/ibfd-products/ibfd-mobile-tax-facts and click on the Trial button to request your free trial. Alternatively, you can contact Customer Service at info@ibfd.org, or by phone: +31-20-554 0176 IBFD, Your Portal to Cross-Border Tax Expertise IMTF/A01/H IBFD INTERNATIONAL TRANSFER PRICING JOURNAL MAY/JUNE 2012 191