This paper deploys Transaction Cost Economics (TCE) to elaborate on the shortcomings of ‘mainstream‘ transfer pricing in multinational firms. Departing from the notion that multinationals increasingly (re-)organize their business along multinational value chains irrespective of jurisdictional borders, this paper discusses the nature of the multinational firm and the problem of choosing the right intra-group (transfer) price. The mainstream transfer pricing approach derived from the Arm�s Length Principle (ALP) is deemed inappropriate for globally operating multinational enterprises (MNEs). Referring to the value chain model, the paper suggests that ‘entrepreneurial coordination’ is the key performance feature to be used for valuing business activity and for allocating — for tax transfer pricing purposes — standard mark-ups and residual profits along the value chain. The main findings of this paper are: Neo-classical concepts on marginal pricing may not suffice to establish arm's lengh transfer pricing; the inadequacy between tax-world transfer pricing (getting income allocation right) and business-world transfer pricing (getting management incentives right) might find its explanation in such concepts. MNEs need to be understood as large organizations different from domestic large organizations by the fact that they operate in different jurisdictions and/or institutional environments. Operative business is coordinated along business lines in which value chain processes can e identified. De facto, business-world transfer pricing takes place along such value chains in which tangible and intangible assets are transferred and hence require appropriate pricing from both the tax-world and the business-world perspective. TCE is a worthy candidate for illustrating governance structures and transactional attributes of business between related parties of a multinational group; such features support arguments to establish arm's length transfer pricing. Regularly, a clear cut-off of functional allocation into tax jurisdictions is difficult to achieve because of the high degree of integration into the value chains of the multinational. TCE appears to better distinguish between so-called �routine� and ‘non-routine’ functions. Transactions of the MNE are rarely of an ‘either-or’ feature (either ‘market’ or ‘hierarchy’). Depending upon transactional attributes, the price of such transaction can be assessed by variables describing the institutional and economic context, the transaction-specific contract, the stage of the business process involved, the strategy chosen, and the function pattern (function, risk, assets) Comparable information is rarely found in databases which provide company information. The more non-routine functions and intangibles are involved, the less is the tested function (or business unit) comparable with companies from external databases. Under these data constraints on comparables, the arm�s length tests on transfer pricing will have to resort to internal information if the ALP is intended to remain viable. A next-generation transfer pricing approach may have to make use of patterns of governance to characterize and to value the functional contributions to the overall value chain.
Substitution across profit shifting methods and the impact on thin capitalization rules