The Compensation of DEMPE Control Functions in Post-BEPS Transfer Pricing

Today, one of the most challenging transfer pricing issues facing multinationals is how to reconcile the mapping of their intellectual property (IP) ownership with the mapping of their control for DEMPE functions,¹ responsible for creating that valuable IP. Whether through acquisitions or divestitures, reorganization of supply chains, or evolution of their business over time, many (if not most) multinationals have some divergence between the location of their IP ownership and the location of management and control of the DEMPE functions. How should transfer pricing policy and practices respect IP ownership and the location of DEMPE functions when there are inconsistencies between the two?  

The OECD’s guidance on the matter is vague. The latest version of the OECD Transfer Pricing Guidelines (TPG) (July 2017), expanded to incorporate aspects of the BEPS project, draws a clear distinction between the legal ownership of IP and the entitlement to profit from those intangibles.  Specifically, “a legal owner not performing any relevant [DEMPE function] will, therefore, not be entitled to any portion of such returns related to the DEMPE of the intangible (6.54).” Conversely, if the legal owner of an intangible performs and controls all the DEMPE functions, funds and provides all the assets, and assumes all the risks related to the DEMPE of the intangible, then it will be entitled to all the anticipated returns derived from the exploitation of the intangible. What is the best approach for multinationals that fall somewhere along the spectrum between complete discrepancy and perfect alignment of their IP ownership and DEMPE management and control profiles?  

The OECD TPG anticipates that a profit split approach would be applied to provide the entity managing and controlling significant DEMPE functions with an appropriate share of returns. However, it provides limited guidance on how to calculate this split, where there is a discrepancy between the location of that management and control of DEMPE functions and the location of IP ownership.

Transfer pricing advisors have developed some elaborate methods for addressing this issue; known variously as value chain analysis (VCA) or process contribution analysis (PCA), these methods generally involve extensive interviews with senior management to assess the relative importance of various value-creating processes (e.g., R&D, manufacturing, marketing, distribution, sales, after-sales service, support services, etc.) within the multinational group. Weightings are then assigned to legal entities involved in these processes according to whether they are “responsible” for performing them, “accountable” for them, or merely “consulted” or “informed” (known as “RACI” weightings). The product of the relative importance of the functions and the RACI weightings then determines the share of residual profit to be allocated to each legal entity. 

At first, such an approach might appear scientific, but in practice, the approach is riddled with subjectivity. Who determines the relative importance of each process? Which senior manager thinks their department performs a less important function than others? What if the most senior executive in the workshop speaks first and puts forth a controversial or aggressive position–will subordinates challenge them or fall in line? What relative weights should be assigned to responsibility vs. accountability? A higher weight assigned to “responsible” than to “accountable” seems at odds with the arm’s length principle and the greater importance of management and control (accountability) of DEMPE functions, rather than merely their performance (responsibility). The use of independent comparables is limited in this analysis to determine the returns to routine processes; the residual profit, often most of the multinationals profit, is allocated ultimately according to subjective decisions that have no relationship to parties dealing at arm’s length. Finally, these elaborate exercises are time-consuming and expensive–well beyond the budgets of all but the largest multinationals. Is there a more objective and less expensive means of addressing DEMPE functions in transfer pricing?  

One alternative approach that we have employed in circumstances where control is dispersed involves the use of internal HR data to identify the location of employees, who manage and control DEMPE functions. 

Under this approach, the tax team and their advisors can look at employees above a certain level of seniority, grade, compensation, etc. (i.e., individuals who exercise control over DEMPE functions) and map them by the entity. More senior managers presumably exercise greater control over DEMPE functions, which would be reflected in their grade and their compensation. The decision of what threshold to use (i.e., managers above what level/grade/rank) could be guided by, whether their compensation package includes eligibility for a bonus or participation in a long-term incentive plan, by their participation in certain leadership teams, etc.

The correlation between “rank” or “compensation” and “management and control of DEMPE functions” will never be perfect, and adjustments might have to be made for disparities in average compensation between countries, but this approach would surely provide a firmer, more objective basis for discussions with tax authorities than merely ad hoc weightings based on management’s opinion. This approach could easily include managers involved in sales and marketing, manufacturing or even support processes, to allow for DEMPE functions related to intangibles other than those related to technology.  

It bears emphasising that location of management and control of DEMPE functions must be considered in conjunction with the location of IP ownership. When the two are perfectly aligned or perfectly misaligned, the transfer pricing analysis is relatively straightforward. We are interested here in situations where there is some discrepancy between the two. Where such a discrepancy exists, there would have to be some weighting of the location of value-creating functions determined through analysis of compensation levels relative to the location of legal ownership of IP. There could be some weighting of historical spend on R&D vs. spending on DEMPE functions. It would not be difficult in most situations to test several different approaches and to perform sensitivity analyses on the weightings to place some bounds on the outcome. This approach is not unprecedented; we have observed multinationals in various sectors employ this kind of approach to settle a dispute or as part of an APA. The specific approach and weightings will likely differ depending on the circumstances. 

One of the biggest advantages of this approach for multinationals is that the data they need exists within their organization. Precautions have to be taken to protect confidentiality around that data, but that problem is certainly not insurmountable. Another significant advantage is the reduced reliance on the subjective judgment of management or their advisors. A further advantage is that the approach could be significantly less expensive to implement than a PCA or VCA.

Read Transfer Pricing Times – Second Quarter 2020

_Sources
^1.Functions relating to the development, enhancement, maintenance, protection and exploitation of intangibles.