Court:Federal Court of Australia
Citation : SNF (Australia) Pvt. Ltd. Vs. Commissioner of Taxation  FCA 635
Brief : In a recent decision, the Federal Court of Australia (“the Court”), in the case of SNF (Australia) Pty. Ltd. v. Commissioner of Taxation  FCA 635 ,ruled in favour of the taxpayer acceding to the approach of direct transactional (price) comparisons (i.e., use of Comparable Uncontrolled Price Method) adopted by the taxpayer, and rejecting the profit-based analysis (using Transaction Net Margin Method) suggested by the Commissioner of Taxation (“the Commissioner”) for benchmarking the taxpayer’s purchase (of polyacrylamide products or “the products”) transactions with its overseas affiliates.
The Court also held that sustained losses do not necessarily imply that the transfer prices are not at arm’s length. In this context, the Court also observed that parent company support provided under a market penetration strategy might permit a loss-making company to continue operations over an extended period without exiting the market.
After Roche Products Pty Ltd. v. Commissioner of Taxation  70 ATR 703 , this judgement is only the second substantive transfer pricing decision to have been pronounced by an Australian Court or Tribunal.
SNF, a member of the SNF Group, carries on the business of manufacturing and selling certain chemical products (commonly called flocculants and coagulants) in Australia to end-users in the mining, paper and water and sewage treatment industries.
In respect of the products, during the relevant period, i.e., financial years 1998 to 2004, SNF operated as a distributor, and purchased the products from its overseas affiliates (in France, US and China) for resale to end-users. The overseas affiliates were manufacturers of the products, i.e., `SNF manufacturers’ or ‘the suppliers’. The above mentioned relevant period was included in the period of last 13 years, barring two years, during which the taxpayer recorded trading losses(See Note 1 below).
The dispute related to the relevant period, and was regarding the arm’s length nature of the taxpayer’s transactions pertaining to import of the products from the suppliers.
Choice of method
SNF applied the Comparable Uncontrolled Price Method (“CUP Method”) for identified product groups taking independent customers of SNF manufacturers who purchased and on-sold products just like the taxpayer. Supporting invoices and other records were furnished and examined. The taxpayer argued that the average prices it paid were almost always less than the average prices paid for the same product categories by independent purchasers, buying comparable volumes in comparable markets, also selling primarily to end-users and at substantially the same level in the market as the taxpayer. Documentary evidence in the form of annual reports, financial reports, company summaries, certificate of registration, etc. were furnished to substantiate this. In fact, the taxpayer also contended that the transfer prices paid by it were either at or below the costs of production.
The Commissioner objected to the use of CUP Method, as he claimed that the third party transactions relied upon were not truly comparable for the following reasons:
The taxpayer accepted that it incurred trading losses despite evidence of good sales performance and despite significant equity and loans extended by its associated enterprises. Commercial reasons cited by the taxpayer for the trading losses included a combination of intense competition, poor management, defalcations by an employee, excessive stock levels, insufficient sales per person, series of bad debts, etc.
The Commissioner did not accept these reasons, and instead submitted that the test is to determine what consideration an arm’s length party “in the position of the taxpayer” would have given for the products, thereby seeking to assess whether the behavior of the taxpayer was at arms’ length, rather than the pricing of the products. Therefore, the Commissioner argued that an arm’s length purchaser would never have agreed to the prices paid by SNF to the suppliers and that if the taxpayer was independent and dealing wholly independently, it would not have returned losses over the years in dispute. In fact, the Commissioner contended that based on results of application of the Transaction Net Margin Method (“TNMM”) the taxpayer should have earned an average operating margin of 1.7%. The Commissioner argued that despite strong sales performance and significant financial support received from its associated enterprises, the taxpayer failed to record any profit indicating that taxable profits were being shifted outside Australia.
The Commissioner remarked that even if the losses made by SNF were part of a larger market penetration strategy, the same was for the benefit of the parent, who should have therefore borne the costs of establishment and penetration and that should have been reflected in reduced transfer prices, as economic circumstances and business strategies have an impact on comparability4.
Decision of the Court
Choice of method
In the Court’s view, CUP and not TNMM was the correct method for testing the taxpayer’s related party transactions.
Based on the taxpayer’s arguments, the Court rejected the application of TNMM on the ground that it did not provide a proper basis for determining the reasonable ‘consideration’ that an independent purchaser would pay for the products. In the view of the Court, the principal concern with application of TNMM was that it inevitably attributed losses to the pricing of the products in complete disregard of the commercial reasons for losses. Further, when using TNMM there is no mechanism to divide the profit adjustment across the seven year period into each relevant year, which should be considered as separate years for the purpose of Australian taxation assessment. In addition, the Court concurred with the OECD Transfer Pricing Guidelines that the CUP Method is most appropriate where direct transactional data is available, and moreover, the Australian law does not require the same ‘exactness’ for using CUP Method as is described in the OECD Transfer Pricing Guidelines.
The Court did not accept the Commissioner’s argument that the relevant test is to determine what consideration an arm’s length party “in the position of the taxpayer” would have given for the products, and instead stated that the test should be the determination of arm’s length consideration for the acquisition per se. Further, the Court observed that it does not matter what the basis of setting a global price list is, as long as the price paid is one in a truly comparable transaction.
Before accepting CUP Method as the most appropriate method for testing the transaction prices, the Court made the following observations on the comparables and comparability factors:
The Court noted that majority of the CUP transactions occurred outside Australia. In this regard, it agreed with the Commissioner that in relative terms, since comparable transactions in Australia were small in number, they could not, on their own, be relied upon to support a CUP analysis. However, the Court held that comparability was to be viewed in a global context, and there is a global market for the products as viewed from the perspective of the options available to buyers for purchasing the products (France, China or USA). The Court also held that the nature of the market for resale faced by each of the buyers (independent customers of SNF manufactures) in their respective countries was not relevant for the purposes of establishing comparability on the supply side. Similarly, the unique features of the market in which the taxpayer sells are of no importance. Accordingly, the comparable transactions relied upon by the taxpayer satisfy the burden of proof placed upon the taxpayer.
On the flip side, the Court also observed that there was no evidence in relation to how prices were set or negotiated; the nature, size, business and operations of the comparable entities in their respective overseas markets; the profits made by comparable entities of sale of products in their respective overseas markets; and the profits made by SNF manufacturers. However, these were not considered necessary by the Court in the circumstances of the proceedings, for determining the arms’ length consideration.
Another aspect which was not considered necessary by the Court for determining arms’ length consideration, although it appeared true to the Court based on evidence provided by the taxpayer, was the level of market or the level in the distribution chain at which the taxpayer operated vis-à-vis comparable. The Court was of the view that since the comparison was of prices paid to the suppliers by the taxpayer vis-à-vis independent purchasers, it did not matter whether they onward sold to resellers or end-users.
The Court held that since the taxpayer had already satisfied the onus of proof by providing evidence in respect of use of CUP method, as such, the reasons for sustained losses were irrelevant. However, since the issue was raised, the Court nonetheless addressed the same. The Court held that merely because losses have occurred over a long period, it does not necessarily mean that the consideration paid for the goods is not at arm’s length. The consideration paid by the taxpayer for purchase of the products should be compared to the arm’s length consideration, and any special characteristics of the taxpayer (such as sustained losses over a substantial period and the nature of the market in which it resold) should not be determinative against the taxpayer. Genuine losses may occur and may be sustained for many reasons, other than artificial inflation of transfer prices, and including the ones relied upon by the taxpayer in this case, whereby the documentary evidence was not inconsistent with the oral evidence, although not fully documented.
The Court accepted that the taxpayer was being supported by its suppliers by being charged special reduced prices below cost price and less than the price charged to arm’s length purchasers. The Court also accepted that an independent distributor would have exited the marketplace if it had made the quantum of losses that the taxpayer made during the relevant period. However, the Court observed that it was the policy of the SNF Group to support subsidiaries through six years of initial set up periods, and the SNF group globally was engaged in a ‘market penetration’ strategy in Australia which was considered critical to the long-term strategic plans of the SNF Group, which was the reason for the Group’s continued investment in SNF.
The Court held that in view of the reasons cited by the taxpayer for sustained losses and the SNF Group’s business strategy, there was no inconsistency between continuous losses and the arm’s length pricing of the related party transactions.
This judgement assumes significance for its unequivocal stress on testing the transaction prices (supported with appropriate documentary evidence), rather than the financial results of the taxpayer which are impacted by several commercial factors other than transfer prices. In this regard, it has been observed that the Indian tax authorities as well as Indian taxpayers frequently rely on the TNMM for determining the arm’s length nature of the transactions. However, this international precedence is certainly directionally different from established practices being followed not only in India but also internationally. Although, in a practical sense, tax authorities are more than likely to continue to rely upon the TNMM, not just because of its frequent use, but also because its application has been endorsed in several earlier judicial pronouncements (See note 3 below).
The tenets of this judgement could raise a potential price setting/testing dilemma where comparable uncontrolled transactions with direct transactional data do not exist or are inappropriate, or exist but are a one-time phenomenon. To counter such situations, taxpayers could consider putting in place a policy which lays down the basis of setting the transfer prices. However, utmost caution must be exercised in ensuring that this policy also facilitates price testing, and that the price setting mechanism eventually aligns with the price testing methodology. Also, it is advisable that sufficient documentation be extracted and maintained as evidence to support claims regarding direct transactional data being inappropriate or non-existent or a one-time phenomenon. An explanation to this effect should also form part of the policy.
Another theme which is central to this decision is that losses or even sustained losses are certainly a trigger point for an investigation into the pricing of inter-company transactions but are not conclusive of transfer price manipulation by the taxpayer. In addition, the Court recognises that strategic support provided to SNF by virtue of it being part of a multinational group and for the purpose of market penetration, sustained SNF even without organic profits (See note 4below).. .However, at the same time, the Court also acknowledges that independent entities are not expected to bear trading losses over a long period of time. Accordingly, it becomes critical for taxpayers, who suffer losses, particularly on a sustained basis, to adequately explain, document and quantify (to the extent feasible) their losses, and also find suitable alternatives to benchmark their transactions independent of a net margin analysis. This exercise must be undertaken in conjunction with an assessment of the taxpayer’s business model characterization. For example, if based on an analysis of functions, assets, risks, and consequent value chain contribution, a distributor is characterised as a limited risk distributor, then it would typically not be expected to suffer any losses but only low and steady returns.
Having stated all the above, the question of whether this conclusion would have been any different, had direct transactional data not existed and the taxpayer (suffering sustained losses) had not been able to prove arm’s length nature of its transfer prices using the CUP Method, remains.
Further, in summary, a combination of facts and policy surrounding transaction prices, availability of documentary evidence, and business model characterisation of the entities involved in the transaction, would drive the approach adopted in each case.
For the relevant period, i.e., financial years 1998 to 2004, the taxpayer suffered an average operating loss of 11.5%.
Referring to GlaxoSmithKline Inc. v The Queen  TCC 324, and the OECD Transfer Pricing Guidelines paragraphs 1.30, 1.32, and 1.35.
This judgement, which rejects the use of the frequently applied TNMM, has lead to a discussion in Australia on whether the findings can be construed more widely beyond the facts of the case. The Australian Tax Office (“ATO”) has in fact lodged an appeal against the Court’s decision, and has also issued a statement indicating that while the litigation in this case remains in progress, the views expressed in the ATO’s public rulings remain the official ATO view for all purposes, and this position is not expected to change.
The Court is, however, silent on what a reasonable market penetration period would be, although valid commercial reasons were found by the Court to support at least six years of losses.