- Purpose of the blog
The purpose of this blog is to address whether the transactional profit split method (TPSM) applies to centralized business models operated by multinational enterprises (MNEs). The assessment will be made in light of the post BEPS transfer pricing guidance, that is, the 2017 OECD Transfer Pricing Guidelines (TPG) and the revised guidance on the TPSM (revised PSM report). Kindly note that the blog will not discuss the impact of the international corporate tax debate triggered by digitalization.
- Centralized business models
A centralized company within an MNE, for example, that is engaged in the business of selling physical goods usually operates with the help of related entities in the value chain (hereinafter “assisting” entities) such as i) research and development (R&D) entities that assist in performing research services ii) procurement service entities that assist in buying raw materials; iii) contract or toll manufactures that provide manufacturing services and iv) limited risk distributors that assist in selling the finished product.
In terms of functions, the centralized company usually takes key decisions associated with the innovation, purchase, manufacture as well as sale of products. In terms of risks, the centralized entity usually controls key risks pertaining to its value chain, that is, risks relevant to innovation, purchasing, processing and selling of the products. In many cases, such entities also own the valuable intellectual property. On the other hand, the other assisting entities usually perform their activities under the supervision and guidance of the centralized entity. They [may] also bear low risks. Moreover, in many circumstances, the “assisting” entities do not own valuable intellectual property.
The current profit allocation principle, that is, the arm’s length principle is based on the philosophy that an entity is entitled to higher returns (profits or losses) when it controls key risks in its value chain. Therefore, if the centralized entity controls key risks through its personnel, then it will be [in principle] entitled to higher returns. This also implies that assisting entities or low risk entities are entitled to returns that commensurate with the lower functions performed and lower risks assumed. Typically, service providers are remunerated on a cost related basis and limited risk distributors on the basis of a certain percentage of their sales. Essentially, one-sided TP methods are used to remunerate low risk entities. It should be noted that some (but not all) MNEs structure their businesses in such a way that “assisting” entities operate in high tax jurisdictions whereas centralized entities are located in low tax jurisdictions.
- Application of the TPSM
The revised PSM report describes that the TPSM may be the most appropriate method where one or more of the following indicators are prevalent: i) each party makes unique and valuable contributions; ii) the business operations are highly integrated such that the contributions of the parties cannot be reliably evaluated in isolation from each other; and iii) the parties share the assumption of economically significant risks or separately assume closely related risks. Taking into consideration this background, the blog explores whether the TPSM applies to transactions between the centralized company and other assisting entities. In other words, should the assisting entities be entitled to higher returns due to the application of the TPSM? We will analyse the issue from the sales side of the value chain and work backwards.
Sales through a limited functional distributor
The existence of unique and valuable contributions is perhaps the clearest indicator that the TPSM may be the most appropriate method. Contributions are “unique and valuable” when they (i) are not comparable to contributions made by uncontrolled parties in comparable circumstances, and (ii) represent a key source of actual or potential economic benefits in the business operations. This concept is discussed in Example 3 and Example 4 of the revised PSM report.
In Example 3, Company A is responsible for the design, development and manufacturing of electronic appliances, whereas related Company B is responsible for the selling of the products. Essentially, Company B develops and executes cutting-edge global marketing activities relating to the new line of products and is responsible for designing the marketing strategy, deciding on the level of marketing expenditure in each country where the products will be released, and validating the impact of the marketing campaigns on a monthly basis. The marketing activities performed by Company B result in a valuable trademark and associated goodwill in the market. The facts indicate that the contributions of both Company A and B are unique and valuable to the potential success of the new line of products. Under these circumstances, the Example indicates that the TPSM might be the most appropriate method, but it does not indicate how it should be applied.
On the other hand, in Example 4, Company B undertakes marketing activities that are rather limited and which do not significantly enhance the goodwill or reputation associated with the trademark. The functional analysis determines that the risks assumed by Company B are not economically significant for the business operations and that Company B does not make any unique and valuable contributions in relation to the controlled transaction. Moreover, its distribution activities are not a particular source of competitive advantage in its industry. Under these circumstances, the revised PSM report argues that the TPSM may not be the most appropriate method as it is likely that the arm’s length compensation for the contribution of Company B can reliably be benchmarked by applying one-sided transfer pricing methods. For instance, the resale price method or transactional net margin method (TNMM).
Manufacturing by a contract manufacturer
Example 8 of the revised PSM report discusses the situation of Company A, the parent company of an MNE group that is engaged in the manufacturing and distribution of electronic devices. Company A decides to subcontract the manufacturing of the electronic devices to related Company B. Under the terms of the contract, Company B will follow the directions of Company A to produce the devices. Once the devices are manufactured, Company B (which seems like a contract manufacturer) will then sell the finished products to Company A, which in turn will market and distribute the products to unrelated customers. The facts state that Company B does not make any unique and valuable contributions in relation to the controlled transactions and the risks assumed by Company B are not economically significant for the business operations of the group. While the operations of Company B are integrated to some degree with those of the parent company and dependent upon the latter, an arm’s length compensation for the contributions of Company B can be reliably benchmarked by reference to one-sided transfer pricing methods. For instance, the cost plus method (CPM) or TNMM. Under these circumstances, the TPSM is unlikely to be the most appropriate method.
Purchasing by a procurement entity
In the context of discussing MNE Group synergies that arise from deliberate concerted actions, the TPG in Chapter 1 puts forward several examples that deal with purchasing entities in the MNE Group. Essentially, Example 3 in Section D.8 discusses a situation wherein Company A purchases goods for resale to other group members. The facts state that if Company A only coordinates the purchasing activity and negotiates volume discounts then that entity should only receive a cost plus return. Example 4 in Section D.8 also discusses a similar situation wherein, instead of taking title to the goods, Company A only provides a procurement service. In this case, the TPG also states that Company A should only be entitled to receive a cost plus return. In light of these examples, it can be argued that the TPSM is unlikely to be the most appropriate method with respect to the transactions that take place between the centralized entity and the purchasing entity of the MNE Group. The CPM or TNMM could be used to benchmark the arm’s length nature of the service fee.
Research by contract research and development entities
At several occasions, the TPG discusses examples wherein an entity in an MNE provides research services to another related entity, in particular, in situations where the latter entity controls the key risks associated with the research activities. For instance, in its annex on intangibles, Example 14 deals with a situation wherein Company S provides research services to its parent entity under the latter entity’s supervision. In that example, Company S is entitled to a service fee for its activities whereas the parent entity is entitled to the returns that would arise from the exploitation of the intangibles. In this situation, it can be argued that the TPSM is unlikely to be the most appropriate method when transactions take place between a centralized entity and a research and development service provider. On the other hand, the CPM or TNMM could be used to determine the arm’s compensation of the service provider.
- The open issue: Highly integrated transactions
In the aforementioned examples, the TPSM seems inapplicable to transactions between a high-risk centralized entity and low-risk assisting entities. Arguably, we arrive at this conclusion by seeing each transaction in isolation, that is, one transaction each between the centralized company with the low risk distributor, contract manufacturer, procurement and research and development service provider. However, the question raised here is whether such transactions can be clubbed together and be seen as “highly integrated transactions” which would then warrant the application of the TPSM?
A highly integrated business transaction is one that features a high degree of inter-connectedness in a way that the functions performed, assets deployed and risks assumed by the parties to the transaction are extremely interlinked that they cannot be reliably evaluated in isolation (Para 2.133). A high degree of integration is typically assumed if the parties perform functions jointly, use assets jointly and/or share assumption of risks to such an extent that it is impossible to evaluate their respective contributions in isolation from those of others. It is also assumed when the integration between the parties takes the form of a high level of inter-dependency (Paras 2.134 -2.135).
A typical example of a highly integrated business transaction may be one in which two related companies jointly perform the same key value-adding functions (for instance, producing and assembling high-quality key-components for an excavator used in the earthmoving business) and jointly use and contribute to the MNE group’s most important assets. Additionally, they transact in a highly integrated manner and operate interdependently (for instance, several process steps are required to be undertaken between the two fully-fledged entities to successfully complete the cutting-edge key-components for an excavator and both manufacturers are highly dependent on each other).
In relation to the question of whether centralized business models can be seen as highly integrated is indeed debatable. The centralisation could involve elements of integration of the activities, such as through the dependency of the assisting entities on the decision-making of the centralized entity (e.g. the assisting entities execute core activities under the control and supervision of the centralized entity which has the capability to bear and control the key strategic risks). However, it could be argued that as long as the functions, risks and assets of the centralized entity are separate and could reliably be evaluated/benchmarked in isolation from the functions, risks and assets from the assisting entities, the TPSM is not likely to be the most appropriate method, even though there is a certain degree of integration and interdependence. As the TPG recognise, most MNEs are integrated to some extent (Para 2.133).
Even if it is argued that the TPSM should apply, the taxpayer could argue that the returns that should be allocated to each assisting entity should be in line with the value that the entity generates. In other words, returns should be allocated on the basis of a proper functional analysis. So, for example, if the tax administration of the State in which the limited functional distributor is set up claims for a profit split (residual or contribution), the distributor should not get more than what the distributor was already being remunerated under a one-sided method.
In our view, transactions between the centralized entity and assisting entities should not [automatically] be considered unique or valuable. Moreover, they do not generally seem to share the assumption of economically significant risks or separately assume closely related risks. However, the definition of highly integrated transactions is still relatively broad and openly formulated with no clear threshold. Thus, some concern still remains that transactions could inadvertently be classified as “highly integrated”, and, consequently, could lead to arbitrary use of the TPSM by tax authorities. This is because the definition of what exactly highly integrated business transactions are as well as the application of the concept is still, to a certain degree, a subjective one, which leaves space for interpretation. Thus, at this stage, an absolutely conclusive answer cannot be provided to the question of whether the TPSM applies to centralized business models that operate in a highly integrated manner. The answer may indeed be found in the accurate delineation of the actual transaction but this still creates tax uncertainty or compliance costs for businesses. Accordingly, it is recommended that policymakers provide further guidance in this area.
The authors would like to thank Mr Johann Muller and Mr Stefaan De Baets for commenting on draft versions of this blog.