INTM463080 - Transfer Pricing: OECD and methodologies: Transactional net margin method ('TNMM')
The transactional net margin method and the profit-split method are described in the OECD Transfer Pricing Guidelines as the transactional profit methods (also called 'other' methods) for establishing the arm's length price. They concentrate on finding the arm's length net profit margin as opposed to the more common gross profit margin sought by the 'traditional' methods of CUP, resale minus and cost plus. The Guidelines consider TNMM at Paras 2.58 - 2.107 (3.26 - 3.48).
TNMM is the OECD method most commonly used for justifying the transfer pricing of a company and it is also (along with cost plus) one of the most misunderstood methods. In some cases the case for a TNMM approach is based on little more than a list of supposedly comparable companies, whose results are within a range that encompasses those of the company under review. It may therefore be argued that its prices are arm's length. This is only a benchmarking method, since the results of the company under review are benchmarked by reference to other companies which are only superficially similar. This approach is in reality a comparable profits method (all it shows is that the profits of the company are comparable to those of other supposedly comparable companies), which is acceptable only to the extent that it is consistent with the OECD Transfer Pricing Guidelines.
The OECD Transfer Pricing Guidelines make it clear that any attempt to use TNMM should begin by comparing the net margin which the tested party makes from a controlled transaction with the net margin it makes from an uncontrolled one. A proper attempt should be made to do this. Where this proves impossible (perhaps because there are no transactions with uncontrolled parties), then the net margin which would have been made by an independent enterprise in a comparable transaction may serve as a guide. Note the strict criteria of an independent enterprise, carrying out a comparable transaction, and the caveat that this will be only a guide. The emphasis is very clearly on finding a comparable transaction. In addition, a functional analysis of both the associated enterprise and the independent enterprise is required to determine if the transactions are comparable. It might of course be possible to adjust results for minor functional differences. To a certain degree however, the more something has to be adjusted to make it resemble something else then the more dissimilar it is. If there is no inherent comparability to begin with, then no amount of adjustment will make something comparable to an extent which allows meaningful conclusions to be drawn.
It is commonly suggested that TNMM permits the use of less rigid criteria for comparability than other methods. This is not the case (see below). This misconception may result from a misunderstanding of the significance of a range of results - the so-called 'arm's length range'. In particular, indiscriminate use of a TNNM will be of little assistance in deciding where to position the tested party within a range of results.
The OECD Transfer Pricing Guidelines acknowledge that there are a number of weaknesses peculiar to the TNMM. The effect of these weaknesses is compounded by the inappropriate application of a TNMM. Amongst these weaknesses is the fact that the net profit indicator of a company can be influenced by a range of factors that either have no effect or a different effect on gross margins or the actual price of a transaction. This makes comparing the tested party with another company very difficult if that company is not affected by the same factors. How will you know if they are? For instance, the supposedly comparable company may have been managed and run very badly, while the tested party may have been run very well. Would it be proper then to draw any conclusions relevant to the transfer prices of the tested party from the fact that it made a 10% operating margin but the company which is said to meet the comparability criteria made an operating margin of 6%? It is easy to see from this simple example that incorrect conclusions can easily be drawn from the use of TNMM.
There is a common misconception that the requirements of comparability necessary for a correct application of the traditional OECD methods can are less rigid in the application of TNMM. In fact, because of the unique difficulty in applying TNMM, comparability standards may have to be even more rigid.
In the light of the difficulties outlined above, it is important to take into account a range of results when using TNMM. A range of results may mitigate unquantifiable differences between the tested party and independent companies carrying out comparable transactions. A range would allow results which would occur under a variety of business conditions. Note however that the range of results has to be constructed from the results of companies carrying out comparable transactions. Acceptance of the existence of a range should not be taken to imply acceptance of the inclusion in that range of companies which are not comparable or which carry out transactions which are not comparable.
