India. Sanjay Tolia, Tarun Arora, Ruhi Mehta and Shikha Gupta, Price Waterhouse & Co., India. Cheil India Private Limited

Transcription

1 India Sanjay Tolia, Tarun Arora, Ruhi Mehta and Shikha Gupta, Price Waterhouse & Co., India Cheil India Private Limited 1.With regard to the Delhi Tribunal ruling in the case of Cheil India on pass-through costs of agents/ intermediaries: a. How would non value-adding/ pass through costs be dealt with for the purpose of computing operating profitability, e.g. costs related to advertisement placement/ obtaining advertising space on behalf of customers incurred by advertising agencies, or payments made to clinical research organisations/ external agencies by pharmaceutical companies, or third party vendor costs incurred for facilitating logistics, etc? b. How would aspects such as risks, level of functions, etc. carried out by the taxpayer in order to conclude which activity/ cost could qualify as non value-adding or pass through in nature, be determined? Non value-adding expenses/ pass-through costs are those expenses/ costs which are incidental or ancillary to the primary business activity of a taxpayer and in respect of which the taxpayer typically neither performs any significant functions nor assume any risks thereof. Accordingly, pass-through costs do not qualify to be considered as a part of operating costs by taxpayers, while computing the operating profitability, since these costs are incidental or ancillary to the primary business activity of a taxpayer. In such situations, dealings with AEs should be tested by using the PLI of OP/ VAE. Under the Indian transfer pricing ( TP ) regulations, the determination of an arm s length price in relation to an international transaction under the TNMM is determined with reference to the net profit margin realised by the enterprise computed in relation to costs incurred or sales effected or assets employed or to be employed by the enterprise or having regard to any other relevant base. Though the Indian TP regulations do not specifically provide any guidance on definition of costs/ relevant base while computing the operating profitability of the taxpayer, the Guidance note on transfer pricing issued by the Institute of Chartered Accountants of India ( ICAI ) specifies certain ratios that may be used to determine the arm s length price while applying the TNMM; viz the Berry ratio being one of them. Thus, OP/ VAE being a variant of the Berry ratio to this extent does find a place in the Indian legislative framework on transfer pricing. The OECD guidelines 2010 contain clear directions on the treatment of pass-through costs and the ruling in the case of Cheil India is the first ruling in India which deals with the issue of pass through/ non-valueadding costs. This ruling extensively relies on the OECD guidelines and establishes the fundamental principle that while applying a cost based remuneration model, a return or mark-up is appropriate only for value-adding services/ costs. The issue is whether a taxpayer entity which is not involved in undertaking a particular activity/ service on its own and does not assume any risks on account of such service/ activity/ cost should be expected to earn a mark-up/ return on such costs. The Delhi Tribunal in the case of Cheil India (which is an advertising agency) noted that Cheil India facilitates the placement of advertisements for its customers/ AE for which it makes payments to third parties for renting advertising space on behalf of customers/ AEs and it recovers all the payments made to the third party vendors/ media agencies from the customer/ AE and also does not assume any risk of non-payment by the customer/ AE. Based on the detailed factual representations made before the Tribunal, the Tribunal noted that the advertising space is let out by the third party vendors directly in the names of customers/ AEs, who only finalise the terms of advertising. Such third party payments do not represent value-adding functions of the Indian taxpayer and therefore should not be considered for the purpose of determining the operating profitability of the Indian taxpayer entity. The Tribunal accepted that in Cheil India s case, a mark-up is to be applied on the costs incurred by the taxpayer in performing the agency functions and not on the gross media expenditure. The Tribunal endorsed the OECD s view that while applying the TNMM, the costs to be considered should be the costs incurred in relation to the valueadding activity, i.e., the costs related to the agency function in Cheil India s case (i.e. advertising services). As per the OECD, in applying the TNMM, the costs to be considered for the purpose of applying a mark-up should be determined on the basis of the value added to the business operations by the taxpayer entity. It is appropriate for a company to pass on the costs of renting advertising space on behalf of its AE (which it would have incurred directly had it been independent) to the AE without a mark-up, and to apply a mark-up only on the costs incurred by the intermediary in performing its agency function. The principle has also been accepted by the renowned economist, the late Dr. Charles H Berry, while formulating the concept of the Berry ratio. In one of his ar- 10/12 Transfer Pricing Forum BNA ISSN

2 ticles, the case of advertising agencies has been deliberated wherein it has been concluded that the cost related to advertisement placement (purchase of advertising space) is a measure of service(s) provided by the media agencies and not by the advertising agency. The Cheil India ruling therefore lays down an important transfer pricing principle that keeping in view the functions assets and risks ( FAR ) profile, a taxpayer should not be expected or entitled to earn a mark-up on pass-through costs, which would have a favourable impact for taxpayers in other industries as well. For example, in the pharmaceutical industry R&D and clinical trial companies typically outsource the actual research/ trial activity to third party clinical research organisations ( CROs ) involving substantial costs. The role of the taxpayer entity is generally limited to providing support/ facilitation services for the clinical research activities conducted by such third party CROs/ institutions, etc. In such cases, the company does not have the capability to perform clinical trials itself but only procures these services from third parties. Since these are essentially non value-adding costs for a pharmaceutical company, such costs may be treated as a pass-through cost and not included in the operating profits of the taxpayer. While evaluating whether a cost qualifies as passthrough or not, the taxpayer s answer lies in distinguishing its FAR vis-à-vis that of the AE in respect of these costs. The differences that are critical to demonstrate include the following: s the taxpayer does not add any value and only plays the role of a facilitator while the other transacting party is the entrepreneurial risk taker in relation to such costs; s the taxpayer simply performs the role of an agent who acts on behalf of and under the control of the other party, who is the real beneficiary and approver of such third party costs; and s the taxpayer assumes no risks in relation to such pass through costs and activities (including results of performance, and success/ failure of such activity) Therefore it is advisable for the taxpayers to maintain robust documentation to substantiate the above. A comprehensive inter-company agreement with the AE along with any other internal correspondence, which clearly brings out the nature of the costs, the precise role of the taxpayer, and the rationale for such an arrangement should be maintained. Further, where possible, market data/ information corroborating the practice followed by the taxpayer to be in line with that followed by other players in the industry should be collated. Moving to question 1.b., while representing the case, the FAR profile of Cheil India was clearly brought out before the Tribunal which was well appreciated by the Tribunal. Some of the critical factual representations made in the matter were: s Cheil India operates as an agent of the customer/ AE and the relationship does not constitute a Principal to Principal relationship. As part of its business operations, Cheil India facilitates the placement of advertisements (for its customers/ AEs) in the print, electronic etc., media. s To this end, it is required to make payments to third parties like advertisement agencies, printing presses, etc., for the renting of advertising space on behalf of its customers. Such payments are fully recovered from the respective customers/ AEs. s Cheil India s business is the provision of advertising and related services and not the sale of advertising slots to customers. s Advertising agencies/ companies simply act as intermediaries between the ultimate customer and the third party vendors in order to facilitate the placement of advertisements. The advertising companies are compensated for their efforts and costs relating to the provision of advertisement services on the basis of an agreed commission fixed as a percentage of the gross media spend for the release of a particular advertisement. s Cheil India submitted all relevant documents like invoices and agreements supporting the above claims. The Tribunal made a note of the contentions and documentation and based on the same decided the case in favour of Cheil India. From the above it can be seen that aspects such as risks, level of functions, etc. carried out by the taxpayer in order to conclude which activity/ cost could qualify as non value-adding or pass-through in nature were demonstrated and evaluated in detail by the taxpayer and considered by the Tribunal in framing their conclusion. It is therefore imperative for taxpayers to demonstrate their level of functions and risks with respect to the costs being claimed to be non valueadding. As an example, in the pharmaceutical industry, the following functions/ risks are generally performed by a company acting as a facilitator for clinical research activities: s the taxpayer acts as a low risk service provider merely performing the function of a facilitator; s such services are typically procured from a third party, central research organisation (CRO); s the taxpayer does not assume responsibility for the quality/ success or failure of the services performed by a third party CRO; s the taxpayer does not assume the risk of drug failure/ adverse side effects, this risk being borne by an overseas AE which is the ultimate sponsor of the trial; s taxpayer does not have the required assets, infrastructure or technical personnel to perform clinical trials itself; and s as per the job description of the employees of the company, they are expected to perform merely a supervisory role. In such a situation: s The taxpayer may be acting as a mere facilitator for the clinical research activities carried out by the third party CROs. s The taxpayer is liable to pay any amount to such third parties when the same is recovered from the relevant customer/ principal and does not assume any risk on this account (for example by taking advance payment). This is a suitable example of the globally accepted term/ concept of pass-through costs. Thus, if the acceptability of application of OP/ VAE as a PLI is challenged, it is imperative for the taxpayer to demonstrate (by way of a robust, well documented 2 10/12 Copyright 2012 by The Bureau of National Affairs, Inc. TP FORUM ISSN

3 transfer pricing policy) a thorough and comprehensive FAR analysis of the taxpayer vis a vis its AEs which takes into account critical considerations like risks assumed and level of functions of the taxpayer. Wrigley India Private Limited 2. With regard to the Delhi Tribunal ruling in the case of Wrigley, which compared licensed manufacturing margins for domestic business with contract manufacturing margins for export business: a. For the purpose of comparability, how would tax authorities of your country deal with the evaluation of different business segments of a taxpayer wherein the FAR profiles and business models of the segments have differences in the level of functions performed and risks assumed by companies in relation to end customer facing businesses, where they act as entrepreneurs (exploiting trade and marketing intangibles), as opposed to when they deal with related parties, say in the capacity of contract manufacturers? b. In the situation referred to in Question 2.a. above, would tax authorities of your country compare the margins of the two business segments by making rule of thumb adjustments for differences in FAR profiles or refrain from comparing the margins of the licensed manufacturing or domestic business with those of the contract manufacturing business? The basic requirement of any transfer pricing ( TP ) analysis is to select uncontrolled transactions after undertaking a detailed FAR analysis. Needless to add, undertaking a comparison of the functional profile of the tested party with that of the comparable companies/ transactions is the bedrock of a benchmarking/ economic analysis. The Indian TP regulations contain guidance on the importance of a FAR analysis while undertaking the TP analysis. Rule 10C(2) of the Income-tax Rules, 1962 (the Act ) provides that in selecting the most appropriate method, factors like the class or classes of associated enterprises entering into the transaction and the functions performed by them, taking into account assets employed or to be employed and risks assumed by such enterprises should be taken into account. However, the guidance provided in the Indian TP legislation is constrained/ limited in sphere and depth. The Delhi Tribunal in the case of Wrigley India has compared the margins earned by the company from its licensed manufacturing domestic business with those of the contract manufacturing margins for export business. However, while doing so the Tribunal did not truly appreciate the differences in the FAR profile of the company while operating in these two segments correctly and thus concluded them to be the same. The issue for deliberation in this case was whether the margins expected to be earned by a company from its export business wherein it deals with its AEs and practically operates as a contract manufacturer can be compared with the margins it earns when it operates as a customer-facing licensed manufacturer assuming related business risks. It may be worthwhile to examine the FAR assumed by companies in relation to end customer-facing businesses, where they acts as entrepreneurs, albeit licensed manufacturers, as opposed to when they deal with related parties, say in the capacity of contract manufacturers in light of the above case, demonstrated in Table 1: In the case of a contract manufacturer, the AEs transact directly with the third parties and therefore the above mentioned functions and risks, in the case of the domestic segment, are actually being undertaken/ assumed by AEs, who are thus entitled to the corresponding returns. For the same reason, in respect of the domestic segment, the above mentioned functions and risks are being undertaken/assumed by the taxpayer, which also earns the resulting profits. Based on the above analysis, the differences in functions and risks between the segments are apparent. Further, owing to the said differences in the FAR, Wrigley India s characterisation itself transformed from a regular risk-taking licensed manufacturer in respect of the domestic segment to a limited risk contract manufacturer in respect of the export segment. Thus, the conclusion regarding the similarity in FAR profiles of both the segments appears to be incorrect. Further, another important dimension which is critical is the advertising and marketing spends usually made by the taxpayer in the case of a licensed manufacturing set-up. The comparison of both segments would not be fair as both segments serve significantly different markets and the cost structures of the segments are different as the taxpayer typically incurs a huge amount of advertising and marketing expenditure in a licensed manufacturing set-up, which is reflected in the operating expenses of the taxpayer and in the case of the export segment, most of the sales are made to an AE and the taxpayer needs not incur any amount on marketing the products. The operating expenses essentially reflect the intensity and Functions performed Risks Table 1 Domestic segment: Licensed manufacturer Caters to third parties Manufacturing, business development, marketing and advertising, inventory management, maintaining a distribution network and relationships, debtor followups Exploits both the trade and marketing intangibles of its AEs Bears credit risk, market risk, contractual risk, price risk etc. Export segment: Contract manufacturer Manufacturing products only for AEs Virtually acts as a contract manufacturer in supplying the products to the owner of the intangibles or licensees of the intangibles for the foreign markets AEs responsible for marketing efforts in their respective territories Not assuming or assuming limited risks such as market risks, contractual risks, credit risks, inventory risks, etc 10/12 Transfer Pricing Forum BNA ISSN

4 nature of functions performed, and gross margins are a measure of whether an enterprise earns enough to cover the costs of its various functions, including advertising and marketing expenses. The Tribunal in this case had not considered the said expenses to be relevant as according to it, they impacted only net margins rather than gross profits. However, this is not true, since gross margins typically include the return on advertising and marketing expenses, just as they include the return on other expenses, with net profits being an arithmetical derivative. Therefore, gross margins and operating expenses, including advertising and marketing expenses, share a high correlation and can certainly not be considered in isolation. Moving to question 2.b., the Indian regulations do provide that reliable and accurate adjustments can be made to account for differences, if any, between the international transaction and the comparable uncontrolled transaction or between the enterprises entering into such transactions. Further, there have been several landmark judicial rulings in India 1 which mandate various comparability adjustments while undertaking the transfer pricing analysis. However, they clearly provide that comparability adjustments can be made provided data is available and adjustments can be quantified and appropriately backed by sound economic and statistical principles and robust documentation. The Tribunal in the above case did allow a rule of thumb relief of four percent to bridge the differences between the segments on account of the differential in the market and credit risk assumed by both the segments. Thus, it can be inferred that the Tribunal in effect admitted some differences in FAR profiles between the two segments, and so somewhat deviated from its own assertion of the FAR profiles of both segments being the same. As mentioned above, the two segments were completely incomparable. Moreover, with such significant diversity in geographic markets, functions performed and risks assumed, even economic adjustments cannot bridge the gap. Therefore, the ad hoc relief granted by the Tribunal with respect to the differences in market and credit risks, besides being unsupported in terms of basis and calculation, seems fairly inconsequential. Agility Logistics Private Limited 3. With regard to the Mumbai Tribunal ruling in the case of Agility, where a split of net revenue (50:50) has been accepted as a CUP: a. Would a sharing of net revenue between two related parties be accepted for the purposes of applying the CUP method on the basis that this was the industry norm? b. Could the CUP method be held to be inapplicable in the circumstances referred to in question 3.a. on the ground that the said method can only apply for comparing absolute prices and not net revenue sharing arrangements? The CUP method prescribed under the Indian transfer pricing regulations for benchmarking international transactions provides for determination of a price of a transaction in an uncontrolled situation which can be compared with the price of a taxpayer s transactions with its AEs. Accordingly, based on a literal reading of the Indian regulations, it is the outcome i.e. the price which can be compared. Further, on the issue of identifying comparables based on a prevalent industry norm, Rule 10 B(2) of the Act provides that comparability of an international transaction shall be judged with reference to conditions prevailing in the markets in which the respective parties to the transaction operate. However, there is limited guidance available in the Indian regulations on what could be considered as conditions prevailing in the market. In general, the factors evaluated under this criterion are aspects such as level of competition in the market, economic development, industry drivers, industry performance/ growth rates, etc. It may be possible to evaluate the pricing arrangements in business/ operational models prevalent between parties in the industry in order to ascertain if the business and pricing arrangements between the taxpayer and its AEs are in line with and comparable with the model and pricing arrangements adopted by unrelated parties in same industry. In the recent one of its kind ruling by the Tribunal in the case of Agility, the issue considered was that, in the case of a logistics company, the revenue received by the taxpayer from the end customers was being split between the taxpayer and the other transacting entity on a 50:50 basis. The premise on which the taxpayer adopted the said arrangement was that this is the typical model/ pricing arrangement followed between parties operating in the logistics industry for sharing revenue between origin and destination counterparts. In the present case, based on the detailed arguments and explanations put forth by the taxpayer, the Tribunal accepted the pricing arrangements prevalent in the industry between related entities to be considered as a comparable since such business is generally carried out between group/ related entities. The Tribunal made a noteworthy observation that in the event of comparable price/ arrangements for the same/ similar activity and FAR profile between unrelated parties not being available in certain situations owing to the nature of the business/ activity, then it may be prudent to consider prevalent arrangements/ prices between related parties, in order to ensure that standards of comparability are adhered to, since comparability of FAR is more critical to any transfer pricing analysis. In arriving at the above decision, the Tribunal accepted the following key arguments put forth by the taxpayer: s the FAR profile of the origin company and destination company is the same; s the taxpayer followed the same 50:50 split of net revenue even in cases where it transacted with unrelated third parties; s the sharing of net revenue in a 50:50 ratio between origin and destination company is an industry norm, which was demonstrated by way of back-up material available in the public domain in this regard; and s the economic conditions prevalent for the companies in the logistics industry are the same in India and neighbouring countries, where the taxpayer had transactions in the present case. 4 10/12 Copyright 2012 by The Bureau of National Affairs, Inc. TP FORUM ISSN

5 Further, based on the detailed FAR analysis presented by the taxpayer, the Tribunal appreciated that the operations in this industry are highly integrated where both the parties (i.e. the companies in the origin and destination countries) provide similar services, undertakie similar functions, employ the same level of assets and assume the same level of risks. Accordingly, the Tribunal in this case has reemphasised the importance of a FAR analysis and the fact that the standard of comparability while applying the CUP method needs to be very high. To this end, it may be prudent even to take into account the industry norm prevalent on the pricing arrangements which the companies typically follow if exact/ precise data for unrelated comparables is not available. The second issue (Question 3.b.) examined by the Tribunal in the ruling was whether, when applying the CUP method a pricing basis, per se can be accepted, instead of the transaction price, as being a comparable when determining whether the basis adopted by the taxpayer was arm s length. According to the Indian transfer pricing regulations, the ALP shall be determined using a CUP, i.e., it stipulates that the price of the transaction be compared/ benchmarked with reference to an uncontrolled transaction. In the Agility ruling, the revenue authorities contended that the pricing mechanism would not be tantamount to a CUP, as, strictly, it falls outside the prevailing legal definition of a CUP. However, the Tribunal has taken a broader view in the present case and accepted pricing mechanism also to be a valid CUP. By way of the ruling, the Tribunal has therefore laid down an important TP principle that if the pricing basis adopted by the taxpayer is itself correct then it cannot be said that the resultant outcome (price or profit) of the taxpayer does not comply with the arm s length standard. Further, in addition to the use of the five transfer pricing methods (in line with the OECD Guidelines), recently, the Central Board of Direct Taxes ( CBDT ) has prescribed the use of a sixth method, the Other method for the determination of the ALP. When compared with the CUP method, the newly prescribed Other method has a wider purview and would cover a price which had been charged or paid or would have been charged or paid. If the Other method is read in context with the decision in the case of Agility, it can be inferred that situations where pricing arrangements are compared could be covered under this method. Serdia and Fulford 4. With regard to the Mumbai Tribunal rulings in the cases of Serdia and Fulford, wherein the implications of the Glaxo Canada ruling were discussed in detail: a. How is the CUP method applied to the import of branded products from related parties for resale, when the products are accompanied by a licence or right to use intellectual property ( IP ), and in particular can the transaction price be compared with the market price of similar unbranded products, which do not involve the licensing of any right to use IP? b. Could the import prices of off-patented or generic APIs imported from related parties for conversion into a full dosage form of medicines, and sold under a trademark or brand licensed from such related parties, be compared with the market prices of generic products of the same or similar formulation? The Mumbai Tribunal rulings in the cases of Serdia and Fulford lay down some notable transfer pricing principles particularly relevant to the pharmaceutical domain. These principles essentially relate to the application of the CUP method to transactions of import of off-patented APIs by the local taxpayer from AEs, which are locally converted into FDF or formulation of medicines and then sold in the open market. These principles have been analysed below by using examples to demonstrate their application. However, before delving into this analysis, it would be worthwhile to first get an overview of a typical drug development process. This would facilitate abetter understanding of value generation in the said process, in turn enabling better comprehension of the functional profile of the AEs vis-a-vis the taxpayer. A typical drug development process, as is represented in Chart 1, begins with the generally longdrawn, complex, time intensive, and intangible creating functions of research (involving numerous stages); evaluation and study of the efficacy of the product (involving several activities); and setting-up of quality norms. These functions lead to the creation of an originally developed API, i.e., an API which is created for the first time by an original developer (i.e., with no one having done so before). An originally developed API is generally patented by the original developer for protection of the IP imbibed in it (i.e., to prevent copying by others ). The originally developed API would therefore be a patented API. However, there is a limited period for which the patent applies, and thereafter this originally developed API goes off-patent and is then referred to as an off-patented API. When an API goes off-patent it can be copied by others to produce a similar formulation. The impact of off-patenting is that the IP which was earlier protected or patented is now generically known and copied by others to produce APIs. Such APIs produced by others, which are copies of off-patented APIs, are typically referred to as generic APIs. Notably, others would leverage on the research, efficacy tests, and quality standards established by the original drug developer. They would, therefore, be able to produce at a lower cost and thus sell at a lower price. Accordingly, in an off-patented scenario even the original developer would be compelled to lower its selling prices in order to remain competitive. In the typical drug development process, once the API is created (as discussed above), it is thereafter converted into formulation. This is the final stage and entails the transformation of the API into FDF, that is the edible form of the API (i.e., making it fit for consumption). For conversion, certain pre-specified steps are required to be followed. As is evident from the discussion above and represented in Chart 1, conversion ranks lower than the other elements in the value chain of a typical drug development cycle. Question 4.b., above, contemplates a situation where the local entity (i.e. the taxpayer) imports offpatented APIs from AEs, converts them into FDF 10/12 Transfer Pricing Forum BNA ISSN

6 Chart 1 Basic Research - Chemical Research Biological Research Pharmaceutical research Numerous stages of Research would be reduced to 45, and the impact on the taxpayer s OP would be as follows: Selling price (market driven) of the taxpayer = 100 Purchase price (based on a comparison with the price of generic APIs) = 45 Operating expenses (remain unchanged) = 20 The resulting OP would be ( ) = 35 (i.e., 35 percent on sales). Clinical Research Phase I (Human Volunteer Studies) Phase II (Initial Human Efficacy Studies) Phase III (Full Scale Clinical Evaluation) Phase IV (Post-Marketing Surveillance) Post Clinical research - Prep of product monographs Develop global manufacturing and quality standards The above stages result in an originally developed API Conversion of API to formulation Testing - efficacy studies, clinical evaluations, surveillance, etc. Quality standards Conversion and then sells the FDF in the open market. Therefore, the taxpayer s role is limited to the last stage of the drug development process. In this situation, the functional profile of the taxpayer, simply speaking, is thus that of a converter and a distributor. Conversion, also commonly referred to as secondary manufacturing, may even be outsourced. Therefore, in effect, the taxpayer is primarily a distributor of FDF along with the added function of conversion, and may thus be best characterised as a value added distributor, who should earn a return for its distribution function, and also for the value addition of conversion (either for converting on its own or for co-ordinating the same with outsourced entities). Having concluded on the functional profile of the taxpayer, let us now assume that the AE compensates the taxpayer with 10 percent on sales, for its distribution function, and also for the value addition of conversion. It should also be presumed that the 10 percent remuneration is at arm s length which has been determined based on a TP analysis and is commensurate with the functions performed by the taxpayer. Let us further assume that since the taxpayer operates in an off-patented scenario, it has been compelled to lower its selling prices in order to remain competitive, and has thus also lowered its purchase price of APIs from AEs. Now, assume that the taxpayer s sale price is 100. Then, as per the arm s length pricing policy agreed with the AE (based on 10 percent profit on sales) the taxpayer should earn an operating profit (OP) of 10 for its distribution and conversion functions. Assuming its operating expenses are a given at 20, then the API purchase price should be 70, being the balancing figure ( ). This situation is depicted in Diagram 1. In the off-patented scenario, others import generic APIs from their respective suppliers at say 45. If the CUP method were to be used, and the purchase price of the taxpayer was to be compared with the price of generic APIs procured by others, then 70 The taxpayer thus ends up with an OP/ Sales of 35 percent as against the arm s length OP/Sales of 10 percent. The application of the CUP method and adoption of the price of generic APIs procured by others as a valid CUP for the purchase price of the taxpayer thus leads to the taxpayer earning a return which is in divergence from the arm s length standard (i.e., far more than is commensurate with its functional profile). Accordingly, to answer question 4.b., an outright comparison of the purchase prices of off-patented APIs imported from AEs cannot be made with prices of generic APIs without considering the functional profile of the taxpayer, and the underlying TP policy. Moreover, a CUP comparison requires extremely high standards of comparability of the product as well as other surrounding facts and circumstances. As regards generic APIs, if the pricing data is derived from customs authorities, then the only known facts about the generic APIs is that they are products with similar chemical composition and properties, which have been purchased by others. However, apart from these known facts, there are several unknowns. For example, the functions of others, who are purchasing the generic APIs, is not precisely known. Others could simply be purchasing and reselling to secondary manufacturers for conversion and subsequent distribution therefore the level at which others operate in the supply chain is not known. Such unknown factors could make it extremely challenging to adhere to the high standards of comparability required for a CUP comparison. Hence, for these reasons too, an outright comparison of purchase prices of offpatented APIs imported from AEs with prices of generic APIs could not be sustained. Moving to question 4.a., it may be noted that, with other facts largely remaining the same as they were in question 4.b., question 4.a. simply introduces an added dimension, i.e., when off-patented APIs imported from AEs are accompanied by a licence or right to use the brand, and the FDF or formulation is sold under the said brand. The situation contemplated in question 4.a. is demonstrated in Diagram 2: The taxpayer continues to function as a value added distributor, and continues its pricing policy with the AE such that it earns an arm s length return of 10 percent on sales commensurate with its functions. As regards the added dimension of a licence or right to use the brand, it may be noted that when a product is accompanied by IP rights (for example brand/ trademark), there is a premium value ascribed to the product. The premium is the return associated with the intangible, which is embedded in the product s market price, and would generally belong to the owner (also presumed to be the developer) of the IP. However, the owner of the IP may not necessarily be the entity which is selling the product in the market. 6 10/12 Copyright 2012 by The Bureau of National Affairs, Inc. TP FORUM ISSN

7 Diagram 1 Open market 100 (in off-patented scenario) Conversion to FDF Taxpayer Purchase of off-patented 70 Supply agreement AE Original drug developer Diagram 2 Open market Sale under X brand Conversion to FDF Taxpayer Purchase of off-patented API Right to use brand X Supply agreement License agreement AE Original drug developer & brand developer and owner In such circumstances, the selling entity can be said to be collecting the premium from the market on behalf of the owner. There must then be a mechanism by which the premium so collected is passed back to its rightful owner. If in the above example, the sale price is 100, then 100 would include the premium associated with the brand (being collected by the taxpayer on behalf of the AE). The premium so collected must be passed on to the AE. If the sale price is 100, then the OP based on the pricing policy of the taxpayer as agreed with the AE would be 10 (10 percent of 100). The operating expenses would be the actual amount spent by the taxpayer (say 20). The selling price, the OP and the operating expenses are thus a given, and hence the total amount that needs to be paid back to the AE can be derived as follows: Sale price operating expenses OP, i.e., = is the balancing figure, and includes the premium associated with the brand, which needs to be passed back to its rightful owner. Since 70 is the cumulative amount that needs to be paid to the AE for import of APIs and for using the brand, it could be broken up into two components, i.e., royalty for use of brand and the purchase price of APIs. If in the above example, the royalty is determined at an arm s length rate of say five percent of sales (5 percent of 100 = 5), then the purchase price of APIs from AEs would be 65 (i.e., 70 5). Notably, the purchase price of APIs from AEs (65 in the above example), should be a derived amount computed after applying the taxpayer s pricing policy agreed with the AE, and after deducting an arm s length price 2 for the right to use the brand. The purchase price of APIs from AEs cannot be compared as it stands with prices of generic APIs (as has been already discussed in the answer to question 4(b) above). Although the CUP method entails a direct comparison of prices (and not profits which are directly linked to the FAR analysis), yet before choosing to apply the CUP method, FAR analysis needs to be undertaken at the very outset, to first ascertain whether application of this method is at all appropriate in the given facts and circumstances. To conclude, a vital observation that may be made from the above examples taken for the purpose of answering questions 4(a) and 4(b) where application of the CUP method has been questioned, is that the above analysis hinges on a critical underlying assumption that the 10 percent remuneration earned by the taxpayer is as per an arm s length TP policy agreed between the taxpayer and the AE, which is consistent with their FAR profile. Therefore, if the applicability of the CUP method is to be challenged under similar facts and circumstances as prevail in the above examples (being that of a value added distributor in the pharmaceutical industry), then a well documented TP policy is an imperative and should form the mainstay of any taxpayer s dispute resolution strategy. Bayer Material 5. With regard to the Mumbai Tribunal ruling in the case of Bayer Material, in which companies having 100 percent controlled transactions were accepted as comparables and also, agency commission rates were applied on an arguably arbitrary basis: a. If a company undertakes both distribution and agency activities for products manufactured by overseas related parties, would the differences in the level of functions, assets and risks assumed by the company for the two categories of activities be taken into account, and different pricing basis be used, e.g. a return on value-adding costs or the Berry Ratio for agency activities, as against a return on sales for distribution activities? b. Could companies having 100 percent controlled transactions with their respective related parties be accepted as comparables for the purposes of testing the results of the taxpayer? c. In the case of marketing agents or procurement service companies, would the resultant return on the value added costs (e.g. operating profit divided by value added cost) in the hands of the taxpayer be taken into account, when applying third party commission rates available from different sources? The answer to Question 5.a. should be a yes, even though that was not the direction in which the Bayer Material ruling proceeded. Nonetheless, if coexistence of distribution and agency functions is assumed as it was in the case of Bayer Material, also contemplated in Question 5.a., and if a distributor s and an agent s business and corresponding remuneration models are examined, then some worthwhile TP principles certainly do emerge, and which have been analysed below. Conceptually speaking, if one were to delve into the spectrum of functions undertaken by a distributor and an agent, two critical differences would generally 10/12 Transfer Pricing Forum BNA ISSN

8 arise. One would be in the context of inventory and the other credit (debtors) - related functions/ risks. Generally speaking, an agent would not be responsible for handling or carrying inventory, and would thus not bear inventory risk. As for credit, if the agent is paid by its principal irrespective of recovery from the final customer, then the agent does not carry significant credit risk. However, if the agent s remuneration is linked to recoveries from customers, then the credit risk would be shared by the agent with its principal. On the other hand, generally speaking, a distributor would be responsible for handling and carrying inventory, and also for collections from customers. It would thus typically bear both inventory and credit risk. However, the distributor could be a limited risk distributor (LRD) taking only flash title of goods, with limited responsibility towards collections. In such a model, the inventory and credit risks would vest largely with the principal. What emerges is that, there could be different permutations and combinations of the inventory and credit related functions/ risks which could be undertaken/ borne by a distributor and an agent. The intensity of these functions/ risks would vary depending upon the arrangement that a distributor and an agent has with its principal, and this would be revealed by means of a detailed examination of the functions, assets and risks of the distributor/ agent and its principal. For TP purposes, distribution and agency activities, both of which are being undertaken by a single taxpayer, could either be segregated or aggregated, depending upon the outcome of the FAR analysis as discussed above. If the functions/ risks of the distribution and agency activities are analogous to each other, then the agency activity could be aggregated with distribution, otherwise not. For instance, where the taxpayer is an LRD, the inventory and credit risks would vest largely with the principal instead of the taxpayer, and the LRD would then in fact be akin to an agent. Therefore, a FAR analysis would be pivotal to the decision of whether to aggregate or segregate. Further, the pricing basis would also vary depending upon whether the activities are aggregated or segregated, and this is discussed in detail below. Pricing basis for a distributor if distribution activity is segregated from the agency function If, based on FAR analysis it is decided to segregate the distribution activity from the agency function, then in respect of its distribution activity, if the taxpayer is considered to be a standalone distributor operating as a regular distributor (or more) undertaking all the typical functions (or more) of a distributor, i.e., marketing, securing orders, procurement, supply chain management, inventory holding and insurance, collection, etc., then the distributor would typically expect a certain return on sales, to cover its costs and to earn a return for the distribution risks undertaken by it. However, if the FAR analysis of the distributor reveals that it is an LRD taking only flash title to goods, with a negligible role to play in functions related to the supply chain, inventory, collection, etc., then the distributor would be operating more like a marketing support service entity and would expect to essentially cover its costs, excluding Cost of Goods Sold ( COGS ), and earn a routine return thereon. Total costs, less COGS, would represent the costs which would embody the value addition by the taxpayer and would essentially be the expenses incurred in rendering the marketing service (or the VAE). Accordingly, the pricing basis would be dependent upon the FAR analysis of the distributor, i.e., the intensity of functions performed and risks undertaken by it. Expressed in terms of PLIs (assuming application of the TNMM), the PLI for a regular distributor who would typically expect a return on sales, would quite evidently be OP/ Sales and the PLI for an LRD whose operations are akin to that of a marketing support service provider could be OP/ VAE or a variant thereof, i.e., the Berry Ratio (Gross Profit/ VAE). In fact the use of the Berry ratio in similar circumstances is supported by the OECD, which states as follows in paragraph of its revised Transfer Pricing Guidelines of 2010: In order for a Berry ratio to be appropriate to test the remuneration of a controlled transaction (e.g. consisting in the distribution of products), it is necessary that: The value of the functions performed in the controlled transaction (taking account of assets used and risks assumed) is not materially affected by the value of the products distributed, i.e. it is not proportional to sales, and (Emphasis supplied) Pricing basis if distribution activity is aggregated with the agency function If, based on a FAR analysis it is decided to aggregate the distribution and agency activities, 3 then the remuneration (pricing) basis cannot be sales or turnover based, as the top-line for a distributor is characteristically different from that of an agent (who typically earns a service fee or a commission), and thus cannot be placed at par. The remuneration (pricing) basis must then be cost based, i.e., the pricing should be based on a return (or OP) on costs, whereby the costs to be considered cannot include COGS, as COGS are specific only to distribution activity (and if the taxpayer is only taking flash title to goods, then COGS would in essence be pass-through and not valueadded costs). Therefore, the only costs which can be considered for determining the remuneration (pricing) basis are the VAE, also a true measure of intensity of functions performed. However, VAE would appropriately measure the intensity of functions performed provided that its composition is also carefully mapped. Pricing basis for an agent if distribution activity is segregated from the agency function If, based on a FAR analysis, it is decided to segregate the distribution activity from the agency function, then applying the same principle as was applied in the case of distribution, the pricing basis would be dependent upon the FAR analysis of the agent, i.e., the intensity of functions performed and risks undertaken by it. If the functions are of routine or low intensity, the risks borne by the agent would also not be significant 8 10/12 Copyright 2012 by The Bureau of National Affairs, Inc. TP FORUM ISSN

9 and in such circumstances the agent would typically like to cover all its costs and earn a routine return thereon (OP/ Total Costs in PLI terms). If, however, the agent assumes significant responsibility for effecting sales and undertakes a gamut of functions for its principal in this regard, then the agent would also (similar to a regular distributor) expect a remuneration linked to sales. As also contemplated in question 5.c., when adopting a remuneration basis which is linked to sales, say a commission on sales, the commission rate could be best determined based on comparable uncontrolled commission rates (using the CUP method). However, a standalone determination of commission rates may not yield bona fide results. A commission rate represents remuneration for the agent s activities, and thus cannot be determined independent of an examination of the nature of and the level at which functions are performed by the agent vis-a-vis comparables. If the price for the transaction of provision of agency services has been set as a commission rate, determined using the CUP method, after adequately taking into account the nature and level at which functions are performed by comparables, then the resultant OP/VAE of the taxpayer may not be a relevant consideration. This is so because the CUP method does not warrant going beyond the price, whereas OP/VAE is a profit measure. Moreover, being a profit measure, OP/VAE may be influenced by several factors extraneous to the transaction per se of provision of agency services, and also to its pricing, i.e., commission. These could be internal business reasons or industry specific factors and/ or peculiarities. Moving to Question 5.b., companies having controlled transactions should ideally not be considered as valid comparables (although they were in the case of Bayer Material), as the assumption inherent to a typical TP analysis is that controlled transactions are not at arm s length. Also, practically speaking, there would be limited or no data available in the public domain to counter this assumption. Where there are no uncontrolled comparable transactions, then industry average data from independent reports/ sources may be used as a reasonable indicator of whether or not the pricing is at arm s length. The industry average data would generally be a combination of uncontrolled and controlled data. Hence, a comparison with such industry average data would certainly not replace a conventional economic analysis undertaken using uncontrolled comparable transactions, but with the given data constraints, it may prove to be useful during dispute resolution. Needless to say, the lack of uncontrolled comparable transactions would need to be sufficiently demonstrated and documented, and the sources for industry average data would have to be reliable. Sanjay Tolia and Tarun Arora are Partners, Ruhi Mehta and Shikha Gupta are Senior Managers in PricewaterhouseCoopers India s National Transfer Pricing Team. They may be contacted by at: sanjay.tolia@in.pwc.com arora.tarun@in.pwc.com ruhi.mehta@in.pwc.com shikha.gupta@in.pwc.com NOTES 1 Mentor Graphics (Noida) Pvt. Ltd. v DCIT [112 TTJ 408], Egain Communication Pvt. Ltd. v Income Tax Officer [109 ITD 101] 2 Independently determined under say a CUP analysis. 3 Which would typically be the case when the taxpayer, with respect to its distribution activity, operates as an LRD. 10/12 Transfer Pricing Forum BNA ISSN