Commissioner of Taxation v SNF

BETWEEN: COMMISSIONER OF TAXATION OF THE COMMONWEALTH OF AUSTRALIA

JUDGES: RYAN, JESSUP & PERRAM JJ DATE OF ORDER: 1 JUNE 2011 WHERE MADE: MELBOURNE

1. The appeal be dismissed with costs. Note: Settlement and entry of orders is dealt with in Order 36 of the Federal Court Rules. The text of entered orders can be located using Federal Law Search on the Court's website.

IN THE FEDERAL COURT OF AUSTRALIA VICTORIA DISTRICT REGISTRY GENERAL DIVISION VID 731 of 2010

BETWEEN: COMMISSIONER OF TAXATION OF THE COMMONWEALTH OF AUSTRALIA

JUDGES: RYAN, JESSUP & PERRAM JJ DATE: 1 JUNE 2011 PLACE: MELBOURNE

I. Polyacrylamides and transfer pricing 1 In the years 1998 to 2004 the SNF Group of companies was a French based multinational conglomerate manufacturing and distributing industrial chemicals known as polyacrylamides whose principal, but not sole, use was the cleansing of water in an industrial setting. Significant purchasers of polyacrylamides were many but included the pulp and paper milling, the mining and the sewage treatment industries. According to the 21-27 February 2005 edition of European Chemical News, the global capacity for polyacrylamides in 2004 'stood at 910,000 tonne/year' with the SNF Group being 'the main player' in that market. In the same year the price of polyacrylamides ranged from €2.50 to €3.50 per kilogram so that we may surmise that the SNF Group was a considerable enterprise. The ultimate holding company for the group was SNF Société Anonyme ('SNF France') which, apart from some very minor shareholders, was entirely owned by one man: Monsieur Rene Pich, a company director from Saint-Priest-en-Jarez. The activities of, and corporate entities comprising, the SNF Group may be divided between those who manufactured the chemicals - the suppliers if you will - and those who, on the other hand, were its distributors. Amongst the distributors was an Australian company and it is that company which is the respondent to this appeal (the 'taxpayer'). The internal architecture of the group was such that the distributors purchased the chemicals from the suppliers at prices which were determined by those working at the helm of the group within SNF France. We may see at work, therefore, an exemplar of a modern multinational supply and distribution business: worldwide manufacture and distribution of useful goods; disparate corporate entities in multiple jurisdictions with different functions; the processes of commerce alive and extant within the group but the negotiating function clipped by the overriding will of the ultimate parent. 2 A multinational group of that kind has the capacity to move income from one country to another by controlling the prices at which intra-group sales occur. By increasing the sale prices at which members of the group sell to each other, the income of the selling subsidiaries may be augmented and the expenses of the buying subsidiaries correspondingly increased. Through that device an arbitrary amount of income may be transferred within the group. Also, income may just as readily be moved in the opposite direction by the concomitant practice of reducing the prices at which sales occur. The motives which might stir a desire to move income from one jurisdiction to another by such means are multifarious: the rates of taxation may be more favourable in one jurisdiction than in another; there may be accrued losses in one company which can be better utilised against income earned in another; away from the revenue domain, those controlling the group may wish to provide financial support to one subsidiary at the expense of another. This family of practices, which are known together as transfer pricing, have the capacity to erode the collection of revenue even where the avoidance of taxation is not the end sought. There are several forms of transfer pricing. This appeal is principally concerned with only one: the practice under which a local member of a multinational pays increased prices for goods acquired from overseas members of the same group. 3 Division 13 of Part III of the Income Tax Assessment Act 1936 (Cth) (the '1936 Act') is the Parliament's general response to transfer pricing. Its particular response to the practice in question is contained in s 136AD(3) of that division and the main question in this appeal is whether that provision has been engaged. The full text of all the relevant provisions - including s 136AD(3) - appears in Appendix 1 to these reasons. For present purposes it suffices to note that both parties were in agreement that the provision would be engaged only if the prices paid by the taxpayer to the suppliers exceeded the consideration which might reasonably be expected to have been paid if the transactions had occurred 'between independent parties dealing at arm's length' (s 136AD(3)(c), s 136AA). It is convenient to refer to that consideration as 'arm's length consideration'. The Commissioner of Taxation (the 'Commissioner'), who is the present appellant, determined that the prices paid by the taxpayer to its related suppliers for polyacrylamides did exceed arm's length consideration and that the amount of income upon which the taxpayer was to be assessed for tax for the years 1998 to 2004 should be substantially increased beyond that which the taxpayer had initially returned. These supply subsidiaries of SNF France were physically located in France, the United States and China and it will be convenient to refer to them individually as the French, US and Chinese suppliers or collectively as 'the suppliers'. The taxpayer lodged formal objections with the Commissioner for each year of income but to no avail whereon it appealed to this Court. The amounts of tax involved are set out in Appendix 2. They exceed $2 million. 4 Before the trial judge the central legal question was whether s 136AD(3) had been engaged. The parties agreed that it was the taxpayer who bore the legal onus of demonstrating that the prices which it had paid to the suppliers did not exceed arm's length consideration (s 14ZZO Taxation Administration Act 1953 (Cth)). The taxpayer sought to discharge this onus by showing that it had paid less for the polyacrylamides to the suppliers than the suppliers had charged other independent third party purchasers for the same or similar products. The experts called by both parties - a Mr Seve for the taxpayer, a Dr Becker for the Commissioner - agreed that an analysis based on a consideration of such comparable transactions was the preferable approach if such comparable transactions were available. Mr Seve thought that they were available and based his analysis upon them. Dr Becker however disagreed; there were, in his opinion, no suitable comparable transactions available. In that circumstance, he thought it appropriate to fall back to an alternate methodology known as the Transactional Net Margin Method ('TNMM') whose application led him to believe that the prices paid did not represent arm's length consideration. 5 The trial judge concluded (at [146]) that the transactions put forward by the taxpayer were truly comparable and he accepted that the evidence before him revealed that it had generally paid the suppliers less for the same or similar products than comparable independent third party purchasers. He therefore concluded that the prices paid by the taxpayer to the suppliers were not in excess of arm's length consideration, allowed the taxpayer's appeals in each year, set aside the Commissioner's decision to refuse the taxpayer's objections and ordered in lieu thereof that those objections be allowed in full: SNF (Australia) Pty Ltd v Commissioner of Taxation [2010] FCA 635. 6 The contest before the trial judge covered much ground. One matter which emerged at trial, and which was not disputed, was that the taxpayer had persistently made losses in the period in dispute. The Commissioner contended that these losses, which the trial judge found (at [170]) would have forced an independent operator from the market, provided powerful support for his case. The means by which that argument was pursued were both analytic and rhetorical. They were analytic because they formed the foundation of a submission that the transactions which were to be examined for their comparability should be those in which the purchasers of polyacrylamides had - like the taxpayer - made persistent losses in the conduct of their businesses. They were rhetorical, perhaps flamboyantly so, because they suggested that the taxpayer's motive was to make losses in Australia and to move profits to France by transfer pricing. 7 Given that the only question posed by s 136AD(3) was whether the consideration was an arm's length one, this rhetorical use was of little, indeed no, legal consequence. The questions posed by the concept of arm's length consideration 'cannot include the requirement of any investigation or consideration…of motive and purpose': WR Carpenter Holdings Pty Ltd v Federal Commissioner of Taxation (2008) 237 CLR 198 at 212 [38]. In any event, the Commissioner was wholly unsuccessful in persuading the trial judge to the view that the taxpayer's persistent losses were caused by the prices it was paying to its suppliers. Instead, and to the contrary, the trial judge found (at [166]) that the losses were caused by a congeries of factors - unreasonably low sales per salesperson resulting in higher levels of commercial costs expressed as a percentage of sales; competition in the Australian market; excessive stock levels; poor management - none of which included the Commissioner's contention that the losses had arisen from the transfer of 'profits' to the suppliers. 8 This debate before the trial judge was enmeshed, as one might naturally expect, within a web of related, but distinct, factual arguments. One such argument arose from evidence given by M. Pich that the French supplier had made losses on its sales of polyacrylamides to the taxpayer. The trial judge thought that this conclusion meant that no inference could be drawn 'that the prices paid were artificially inflated' (at [71]). In a conceptual field which has excluded from it motives, which focuses - to the exclusion of all of other concepts - on the bare notion of arm's length consideration and which denies, therefore, the propriety of a detour through a multinational's reasons for action, one may reasonably doubt the utility, the significance, of this debate. But whether that be so and whether the trial judge needed to be troubled by the necessity of resolving this issue (a matter which he doubted) the fact is that he accepted M. Pich's evidence and with it the necessary rejection of any suggestion of price inflation, at least, by the French supplier of whose prices M. Pich was speaking. 9 Before this Court the Commissioner pursued three basic contentions: the trial judge had erred in his findings of fact - the evidence which had been led by the taxpayer at trial did not warrant his conclusion that the transactions upon which the taxpayer relied were, in fact, comparable; the trial judge had not attended with precision to the evidence before him and he had failed to appreciate the absences which existed in the record or the significance of evidence to the contrary. The evidence acted upon by the trial judge was, in his submission, either non-existent or at least inadequate. The trial judge had erred a second time in accepting M. Pich's evidence that the French supplier had made losses on its sales to the taxpayer; M. Pich's evidence had been an 'outburst' during cross-examination and had never once been signalled to the Commissioner before the moment of its delivery in the witness box. Finally, the trial judge had erred a third way in law: the concept of 'arm's length consideration' in s 136AD(3) required an examination of transactions 'between independent parties dealing at arm's length' (s 136AA) and those words, viewed in their full context, required, indeed forbade any other consideration but, comparable transactions in which the purchasers shared each and every quality of the taxpayer bearing on price save for the solitary fact of its having been under the control of SNF France. So viewed, the only comparables lawfully permitted by s 136AD(3) were those in which the putative purchasers were themselves making losses; the failure of the taxpayer to search for and recruit comparable purchasers of that kind meant that it had proved nothing of value and that failure, in turn, was fatal where the taxpayer admittedly bore the onus of proof. 10 The following sections deal separately with these issues. In short: the process by which the trial judge approached his findings of fact about the comparable transactions does disclose some errors and this Court's jurisdiction to review the judgment is enlivened and its obligation to reach its own conclusions on the material before it has been engaged. A review of that material does not, however, lead to a different conclusion to that at which the trial judge arrived. The attack on the trial judge's acceptance of the evidence of M. Pich fails, altogether foreclosed by established principles governing appellate review of such findings. The Commissioner's contention that the only comparables which could lawfully be examined under s 136AD(3) were those sharing the same characteristics as the taxpayer (apart from its non-independence from the group) also fails for this is not, to put the matter bluntly, what s 136AD(3) says. A final, and fourth, section deals with a number of minor but disparate arguments arising during the appeal together with some evidentiary matters.

II. Comparable transactions 11 To understand the Commissioner's contentions, one must begin with the manner in which the taxpayer sought to make good its argument that the consideration it had paid was an arm's length one; one must assay the methods of that proof and the arguments arrayed against them and one must compare those methods and arguments with the conclusions at which the trial judge eventually arrived. It is only against that fabric of methodology that the correctness of the Commissioner's arguments can properly be gauged. 12 The taxpayer's case was that it had generally paid less for the polyacrylamides it had purchased from the suppliers than independent third party purchasers had. On that factual foundation were then laid two further contentions, one tacit, the other explicit: tacit, that the third party purchasers were paying an arm's length consideration to the suppliers; explicit, that the taxpayer was paying less. The endpoint of the argument was that it had shown that the consideration which had been paid was less than an arm's length consideration and that in that respect a requirement of s 136AD(3) had therefore not been satisfied. 13 Necessarily, the taxpayer was required to muster transactions which it contended were truly comparable. By the time of the trial it had mustered three such sets. The first set, prepared by the taxpayer itself, consisted of five specified foreign companies (and their subsidiaries); the second set, a group of Australian and New Zealand companies; the third set, prepared by Mr Seve, was a larger group of 21 companies. Each requires separate examination although there is a considerable overlap between the issues arising in relation to each.

(a) the first set of comparables 14 This first set was selected by one of the taxpayer's witnesses, a Mr Karoudjian, who was the customer services manager at SNF France. He explained in his evidence that the products sold by SNF France were divided into product groups. There were at least thirteen such groups and within each group there was usually, although not invariably, more than one product. The product group 'Liquides Polyamines', for example, consisted of the products DP/FL175, FL175, PC/CP2468, FL18-40 and FL28. Mr Karoudjian had access to all of the sales data for SNF France which he examined. He first determined the product groups into which the products purchased by the taxpayer fell. He then identified other independent purchasers buying products in the same groups in similar volumes. Mr Karoudjian also applied a further criterion: he sifted the transactions so as to include only those purchasers whom he believed to be distributors of product. From that process he obtained a list of five companies: Akzo Nobel NV, Betz Laboratories Inc, Hercules Inc, Ashland Inc and Buckman Laboratories International Inc and in his affidavit evidence he explained, principally, but not solely, by reference to its activities as a distributor what he believed each of the companies did. Because they had been expressed in different currencies, Mr Karoudjian converted all of the sales prices into Euros as at each invoice date. There were differences too between the terms on which the supplies occurred: some were CIF (carriage insurance and freight included), some were CBE (carriage not included in the price), others were EXW (which meant that no additional costs were included); there were many other similar such variations in terms but Mr Karoudjian attended to each. In light of those disparate terms, he deducted the transportation costs to ensure, in effect, that the sales data for the companies examined was pitched on a level surface. He then took the average price for each product group across each year for each purchaser. From that process he was able to collate tables of data which showed generally, but not altogether without variation, that the prices paid by the taxpayer were lower than those paid by the five companies he had identified. 15 The Commissioner's attack on this evidence was both procedural and substantive. At the procedural level, counsel for the Commissioner cross-examined Mr Karoudjian and obtained from him useful concessions that he had no direct knowledge about the activities of the five companies so that his evidence on that issue was unmasked as hearsay to which the Commissioner then objected. The trial judge accepted this objection and ruled inadmissible any statements in Mr Karoudjian's evidence concerning 'the status of customers, as to whether or not they were end users or distributors of similar products' (at [108]). With the taxpayer apparently then bereft of direct evidence about the nature of the businesses conducted by the five companies, the argument mounted by the Commissioner was that the transactions had not been shown to be comparable because the businesses were not shown to belong to the same genus as the taxpayer. At the level of substance, the Commissioner advanced the opinion of Dr Becker who denied that it was correct, or at any rate useful, to examine the products by way of groups: the averaging of prices concealed significant variations within the groups and those variations, once revealed, diminished the capacity of the five companies to serve as comparables. 16 The taxpayer's responses, like the Commissioner's initial criticisms, were in terms both procedural and substantive. Procedurally, the taxpayer parried the Commissioner's objections to Mr Karoudjian's evidence by tendering annual reports for the five companies to establish that they were, indeed, distributors rather than end-users of the products; substantively, more evidence was called up to meet Dr Becker's objections. That evidence took the form of an opinion from Mr Seve, an accountant apparently expert in matters pertaining to transfer pricing. One may see in Mr Seve's evidence the serving of distinct aims: on the one hand, the shoring up of Mr Karoudjian's evidence by endorsement of his five companies as appropriate comparable purchasers; on the other, the assertion of fresh comparable purchasers based this time, not on product groups, as Mr Karoudjian's evidence had been, but instead on references to individual products purchased by the taxpayer and the comparables. 17 Before this Court the taxpayer submitted that the evidence showed that the five companies were distributors rather than end-users. As to all five companies - Akzo Nobel NV, Betz Laboratories Inc ('Betz'), Hercules Inc ('Hercules'), Ashland Inc ('Ashland') and Buckman Laboratories International Inc ('Buckman')- the taxpayer's witness, Mr Schlag, gave evidence that he had selected them (in a corresponding comparables exercise in relation to the US supplier) because 'they all purchased at least some identical products and they were distributors of these products in their respective marketplaces'. Another of the taxpayer's witnesses, Mr Schroeter, gave evidence that Buckman and Akzo Nobel NV were distributing flocculants (a sub-class of polyacrylamides) in the Australian market. The Commissioner sought to overcome this evidence in two ways: Mr Schlag's evidence was not to be understood as evidence of the comparability of the functions of the five companies but was to be seen merely as the reason why he had selected them in the first place; moreover, the evidence of Mr Schlag and Mr Schroeter was inadmissible hearsay (although he accepted that no objection had been taken to it at trial). 18 The taxpayer made other points too. The Commissioner's complaints about Ashland and Hercules were said by the taxpayer to be unfounded because the Commissioner's witness, Dr Becker, had himself accepted that both were distributors. Away from Dr Becker, the taxpayer submitted that account should be taken of the nature of polyacrylamides, their purpose in the industrial process and their distinct quality of not being an ingredient of some other product. The users of polyacrylamides were said by the taxpayer - it bears repetition - to be the mining industry, those involved in the treatment of sewage and the pulp and paper milling industry. So understood, and with that contextual background in mind, the annual reports of the five companies would reveal them, both in terms and in substance, as distributors of polyacrylamides for self-evidently they were not participants in any of those three industries. 19 The annual report for Buckman in 2005 showed that it was a seller of chemical products: 'Sales to the Company's three target markets (pulp & paper, water, and leather) accounted for 88.6% of total sales'; the annual report of Akzo Nobel NV in 2000 exposed its nature as a multinational engaged in three industries - pharmaceuticals, coatings (i.e. car finishings and the like) and 'chemicals' - with the last industry including a significant division supplying chemicals for the pulp and paper industries; the Form 10-K lodged with the United States Securities and Exchange Commission for BetzDearborn Inc for the 1997 year revealed that it was also called Betz Laboratories Inc and that it and its subsidiaries were 'engaged in the engineered specialty chemical treatment of water and industrial process systems operating in a wide variety of industrial and commercial applications with particular emphasis on the chemical, petroleum refining, paper, food processing, automotive, steel and power industries' and that it 'develops, produces and markets a wide range of specialty chemical products, and provides the technical expertise necessary to utilize the products effectively'. A similar extract for Hercules indicated the existence of a division called 'process chemicals and services' which were 'designed to enhance the manufacturing processes, reduce the operating costs or improve the quality of the end products of our customers' and that 'important raw materials for that division 'are cationic and anionic polyacrylamides'. The Form 10-k for Ashland for the year 2000 showed it to be a diversified multinational with an industrial division providing 'process chemicals and technical services to the pulp and paper and mining industries'. 20 Having identified the nature of those businesses what was clear, said the taxpayer, was that they were not engaged in the conduct of mining, pulp and paper milling or sewage treatment wherefrom it must follow that the five companies must have been - just like the taxpayer - engaged in the distribution of polyacrylamides or, most favourably to the Commissioner, doing no more than providing services to those industries. Counsel for the Commissioner accepted in reply that there was no evidence that polyacrylamides were used as constituent elements in other products, an important concession. 21 The taxpayer did not rest on that evidence alone. It recalled the evidence of M. Pich who in his affidavit had indicated that it was a key strategy of the SNF Group 'to derive income from the wholesale sale of large volumes of flocculants to distributors and re-sellers, who then on-sell to the end users of the product' and it called in aid of the deduction which might reasonably be drawn from that evidence that it was likely that the entities buying from the group bore that character. 22 The trial judge's treatment of this issue was brief. He quoted from the evidence of Mr Schlag and Mr Schroeter and recorded that the taxpayer 'also tendered a number of annual reports of various companies which demonstrated that the customers carried on selling operations and were not end users of the products' (at [85]) although he did not identify which companies were involved. That documentary evidence, he concluded, 'supported the evidence which was not objected to of Mr Schlag and Mr Schroeter'. He went on to say: 'On the available evidence, I make the following observations and findings as to the comparable transactions:…(b) The taxpayer sold the large majority of products to end-users, although there were exceptional circumstances where this did not occur' (at [144]). He then recorded his reasons for that (and other findings) at [145]: I make these findings on the bases of the admissible data produced by the witnesses called by the taxpayer, the documentation tendered by the taxpayer relating to the customers and purchasers, and affidavits of Mr Pich, Mr Schroeter, Mr Karoudjian and Mr Schlag to the extent not objected to by the Commissioner. It will be apparent that the matters sought to be proved by the evidence objected to by the Commissioner have been proved by other evidence not otherwise objected to. 23 Before this Court the Commissioner pursued the argument that there was no, or no sufficient, evidence for this conclusion. That there was evidence that the five companies were not end-users is plain and the Commissioner's submission to the contrary is untenable. It is true that the effect of the trial judge's rulings on evidence was to exclude Mr Karoudjian's evidence that four of the five companies were distributors but his evidence was far from isolated (the reference to four of the five companies requires explanation: for reasons given in Section IV, infra, Mr Karoudjian's evidence about one of the companies, Buckman, was in fact admitted into evidence). There was evidence to the same effect from Mr Schlag and, in the case of Akzo Nobel NV and Buckman, from Mr Schroeter: 'Other companies that were distributing flocculants and coagulants in Australia included…Akzo-Nobel, Buckman Laboratories…'. As to that evidence one should accept neither the Commissioner's submission that Mr Schlag was not giving evidence that the five companies were distributors but merely explaining why he selected them, nor his contention that Mr Schlag's evidence and the evidence of Mr Schroeter was, in any event, inadmissible hearsay. 24 Neither of those arguments is sound. It is true that Mr Schlag was explaining the reasons for his selection of the five companies but the reason he selected them was because he thought they had the characteristics he described. The Commissioner's argument impermissibly treats those two concepts as mutually exclusive whereas they are, in fact, coincident. As to the second argument, there are two problems. The first is that an examination of Mr Schlag's evidence about his reasons for choosing these five companies- 'The Customers were selected as they all purchased at least some identical products and they were distributors of these products in their respective marketplaces' - does not reveal it to be hearsay. It is possible that cross-examination of Mr Schlag might have revealed that he did not have direct knowledge of these matters and that what appeared to be direct testimony was, in truth, hearsay but Mr Schlag was not cross-examined and this was not demonstrated. The Commissioner's argument that this evidence was inadmissible hearsay rests therefore upon the unarticulated premise that, had Mr Schlag been asked how he knew of the matters to which he deposed, he would have answered that he had been told them by someone else or had gleaned it from some other document. But, it need hardly be said, he may equally have explained that he used to work for one or more of those businesses, or that he had made it an area of special study, or that his duties included some which might ordinarily be expected to equip him with such knowledge. That being so it is simply not appropriate to conclude that Mr Schlag's evidence was, as the Commissioner submits, hearsay. 25 Even if it were shown that Mr Schlag's evidence was hearsay and assuming that Mr Schroeter's was too, it would not follow that the trial judge was disabled from acting upon it once it was admitted into evidence, as it was, without objection. There is, in the Commissioner's favour, some academic support for the proposition that inadmissible hearsay, even when admitted, continues to have no value. The best arguments were marshalled by Professor W N Harrison in 'Hearsay admitted without objection' (1955) 7 Res Jud 58. Reference is made in that article (at 69-70) to Walker v Walker (1937) 57 CLR 630 which appears to show that a hearsay letter, once placed in evidence, is a legitimate source of probative material. Only Starke J arrived at the contrary conclusion. Of the four judges in the majority on this point, Evatt J (at 638) perhaps put it most clearly: 'I deny the proposition that, merely because the document was 'hearsay' and therefore inadmissible, it is necessarily deprived of probative value'. Professor Harrison sought to argue that Walker was better understood as a decision dealing with the principle which requires a cross-examiner who calls for a document to place it into evidence but this cannot be correct. Walker deals with both arguments and deals with them separately. So much is clear from the dissenting judgment of Starke J who agreed with the other judges as to the operation of the cross-examination rule (at 635) but disagreed with them about what the status of the document was once it was placed in evidence. That being so, Walker binds this Court to reject the Commissioner's contention. 26 In any event, later commentators have differed from Professor Harrison. Professor Heydon, as his Honour then was, commented that a contrary view to Professor Harrison's was preferable 'if only to prevent a large jump in the number of possible appeals based on the hearsay rule': 'Current Trends in the Law of Evidence' (1977) Sydney Law Review 305 at 307 fn 7; an observation having a certain resonance with the present appeal. The same author in the eighth edition of Cross on Evidence observed that 'the modern authorities appear to have moved beyond the conclusion for which Harrison contended': at [1680] p 133. In that regard, it is useful to note the conclusion reached by Spigelman CJ in Seltsam Pty Ltd v McGuinness (2000) 49 NSWLR 262 at 287 [149] that the words 'not admissible' in the Evidence Act 1995 (NSW) means 'not admissible over objection'. No principled reason for approaching the meaning of the same words in the statutory embodiment of the hearsay rule (s 59 of the Evidence Act 1995 (Cth)) presents itself; no justification for taking some other course. 27 Nor ought one to accept the burden of the Commissioner's secondary argument that the trial judge should not, in any event, have accepted this evidence as going to the question of weight. Before this Court that argument was revealed to have two facets: the hearsay nature of the evidence degraded its probative value and an examination of the scope of the five companies revealed that they were not comparable even if they were distributors. But hearsay is not inherently unreliable; each case turns on the hearsay in question and the merits of the matter will be driven by an assessment of its reliability. In this case, those merits slant decidedly against the Commissioner for it is not to be forgotten who these witnesses were. Mr Schag was the corporate controller of the US supplier and Mr Schroeter, the managing director of the taxpayer. Located in such senior positions within the SNF Group, it cannot be correct to suggest that the information internally supplied to them about the nature of the customers to whom the SNF Group sold its product would be unreliable. Quite the opposite is the case: situated where they were within the institutional framework of the SNF Group the more natural inference is not that the information they received about such customers would be unreliable but, entirely to the contrary, that it would be cogent. 28 We also reject the Commissioner's argument that the businesses of the five companies were different to the taxpayer's and hence their status as distributors, even if otherwise established, was to be set at nought. Here the point was that the taxpayer was small and the five companies large. Save for the matter of customer rebates, however, there was nothing to which this Court was taken by the Commissioner to suggest that the size of comparable distributors had any effect on the prices paid by them. And, insofar as the rebates were concerned, the evidence of Mr Karoudjian was that these had been taken into account in adjusting the prices for the comparison process. Mr Karoudjian gave precise evidence at paragraphs 58-76 of his supplementary affidavit, to which no objection was taken, about the pricing discounts which were given to various customers. Discounts or rebates were given to a number of companies; the only relevant rebate Mr Karoudjian found in his analysis was to Akzo Nobel NV. In those circumstances, no reason appears for declining to accept the five companies as distributors performing a similar function to the taxpayer; still less when the evidence of M. Pich that the 'usual model SNF Group subsidiaries is to sell primarily to re-sellers' is brought to account. It need hardly be added that M. Pich's evidence of that matter needs to be seen in a context which includes both the failure by the Commissioner to suggest to him that this evidence was wrong - an essential aspect of fairness - and the trial judge's conclusion - not directly confronted in this Court - that M. Pich was a witness of credit. 29 The lay and documentary evidence before the trial judge therefore abundantly supported the notion that the five companies were equivalent distributors. But the taxpayer's evidence went, in fact, beyond this for its expert witness, Mr Seve, also opined that the five companies were appropriate comparables. The Commissioner attacked this evidence too and here his point was that the taxpayer had failed to prove the assumptions on which Mr Seve's opinion rested, rendering it irrelevant. Mr Seve's evidence was that he exposed Mr Karoudjian's five companies to a comparability analysis which he drew from an Organization for Economic Co-operation and Development ('OECD') publication, Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations. Each comparable was to be considered against a register of five comparability factors. The factors were: the characteristics of the property in question; the functions (i.e. activities) of the proposed comparables; the contractual terms and conditions on which the comparable transactions took place; the economic circumstances of the markets in which the transactions occurred; and the business strategies of the respective parties. 30 Although Dr Becker had sounded a warning about Mr Karoudjian's use of product groups rather than precise products, the Commissioner before this Court abandoned any suggestion that the products purchased by the five companies were any different to those purchased by the taxpayer. But debate persisted about the remaining factors. The Commissioner submitted that the taxpayer had failed to prove that the five companies were functionally equivalent to the taxpayer. It will be apparent from what has been said already that this contention is essentially unsustainable: it was established by the evidence of Mr Schlag, Mr Schroeter, the annual reports set out above and the evidence from M. Pich that the SNF Group sold to distributors and by Mr Karoudjian's evidence that Buckman was a distributor. So far as Mr Seve's evidence is concerned, the Commissioner mounted a somewhat different variant of the previous complaints made about the status of the five companies: it was not shown, so the Commissioner submitted, that the five companies occupied similar places in the supply chain; those further along the chain, closer, as it were, to the end-users might be charged higher amounts (in perhaps the same way that retailers are charged more than wholesalers). But this cannot be accepted for the fact is that functional comparability was shown. The five companies were each distributors and, as the trial judge correctly observed at [155], they were distributors purchasing from the French supplier. One does not begin to discern from the Commissioner's submissions the reasons why these five companies, large multinational chemical distributors as they were, might be charged different tariffs due to their propinquity to the ultimate consumer; one does not start to see the kinds of factors - economies of scale, increased levels of customer service and so on - which explain alike the frequent disparity seen between retail and wholesale pricing. One sees instead only their size, their function and their location only one link down the chain of supply from the French supplier. Functional comparability was established. 31 Nor is there substance in the Commissioner's argument, pursued in writing and again orally, that there was no evidence supporting Mr Seve's treatment of the terms of the contracts governing the transactions. Mr Seve had before him the complete set of the French suppliers' sales to all of its customers. These were provided to him in separate Microsoft Excel files, one for each year. Those data contained a field setting out the delivery terms. Just as Mr Karoudjian had, Mr Seve then adjusted the prices in the data so that they were all ex-works prices and it was those adjusted figures he used. The full set of that data was placed in evidence before the trial judge as Exhibit DK13 to the affidavit of Mr Karoudjian of 13 November 2008. Resort to that exhibit shows the delivery terms, the products concerned and the purchasers. Further, of the payment terms Mr Seve himself noted, at paragraph 11.15 of his report, that the taxpayer paid on delivery whereas the five companies were generally on 14 to 90 day terms. He concluded that matter was immaterial. The evidence, therefore, of the contractual terms was extensive and the Commissioner's 'no evidence' ground correspondingly unmeritorious. The trial judge's conclusions about this evidence were short: 'The essential terms of the comparable transactions were the same or similar to those of the taxpayer including the size of the orders, the terms of payment and delivery, and taking into account currency conversions and isolated rebates' (at [144(b)]). Error may be discerned in that conclusion insofar as it suggests that the terms of the contracts were the same; they were not and neither Mr Karoudjian nor Mr Seve suggested they were. Error may be discerned too in the statement that the terms of payment were the same; they were not as Mr Seve explained. It falls then to this Court to form its own views about those two matters on the material before it. The correct finding about those two matters, in light of the above evidence, was that Mr Kaorudjian and Mr Seve had rendered the sales data effectively neutral on delivery terms by normalising it and that the payment terms were different but in an immaterial way. Those conclusions provide no warrant, however, for a departure from the balance of the trial judge's reasoning process to which they are not material. 32 According to Mr Seve, the next of his comparability factors - the economic circumstances - required a consideration of 'whether the markets in which the independent and associated enterprises operate are comparable, and in particular differences that exist do not have a material effect on the price relevant to the potential comparable transaction [sic]'. One ought to dispense at once with the Commissioner's initial argument that 'no evidence was led about the economic circumstances of those parties and in particular whether they purchased and sold at a profit'. That submission misses altogether the end to which this factor was plainly bent. It was not concerned with the economic circumstances of the parties but of the markets in which the transactions were taking place. Nor, it should be noted for the sake of completeness, did the Commissioner submit to us that Mr Seve's conclusion that the markets were 'broadly comparable to Australia from a macroeconomic perspective' was unsupported by evidence. 33 Yet it was under the rubric of this factor that a real debate did exist. In his report Mr Seve opined that, in addition to the general economic factors, a 'key issue that is directly relevant to an analysis of this comparability factor is the relative stability of pricing structures for [the French supplier] between different markets. All other things being equal, the more consistent the pricing between economically similar markets, the more support for the use of alternative market data.' 34 This 'key' issue necessarily required an examination of the pricing arrangements of the comparables. Mr Seve thought that it was important 'that many of the [comparables] are part of global company groups that, I am informed by [the taxpayer], negotiate global pricing arrangements for multiple markets with the SNF Group'. But it was not just the comparables that needed to be examined; symmetry just as much required scrutiny of the French supplier. 'Further, [the taxpayer] has represented that the pricing structure that is generally applied by [the French supplier] between companies in different markets references a universal price list'. What was the point of this universal price list? Was it intended that it should be used to establish a universal global fixed price for polyacrylamides? The answer was no: 'As is normal commercial practice, prices can vary as a direct result of customer buying/negotiation power however putting this to one side, unless the market is structurally different to the extent that the list price is not economic, the price list is referenced as a guide to stable pricing'. What should be thoroughly grasped at the outset is that Mr Seve was at pains to accept that the prices actually paid might well fluctuate as between the comparable purchasers reflecting their respective buying strengths and individual negotiating postures. Mr Seve's point was not to suggest that these variations did not exist but, in contradistinction, to show that those admittedly existent variations could be seen as divergences within a global market. The evidence was not about the existence of a single global price but, instead, about the existence of a single global market in which price dispersions were to be expected. The comparability of the transactions was to be seen as arising, not from the fact that the prices were the same, but instead from the fact that there was a single market. 35 The taxpayer led a good deal of evidence about this. M. Pich gave evidence which assumed the existence of such a market and this was noted by the trial judge (at [138]): it was his aim for the SNF Group 'to hold 50% market share of the flocculent market globally'; 'The market share of the SNF Group globally at this time is approximately 38%'; 'The prices for purchases by [the taxpayer] from members of the SNF Group were set…always having regard to the global market for the products'. M. Pich was also cross-examined as to whether the SNF Group negotiated its prices globally. He was asked whether he was 'aware of the practice whereby SNF globally negotiates global pricing arrangements with … customers?' to which he responded '[t]here are a few cases, but they exist'. Mr Karoudjian also gave evidence that the French supplier 'generally charged each subsidiary of a single Customer the same price for a product, regardless of the location of the subsidiary and the delivery terms'. Mr Schroeter gave evidence about another multinational purchaser, Nalco, and the prices paid by it (Nalco was not one of Mr Karoudjian's five companies but was part of Mr Seve's later analysis and it is nevertheless useful at this stage to consider its role in the global market issue). Mr Schroeter's evidence was that '[t]he ex-works price charged to Nalco was consistent regardless of where they were in the world'. He gave similar evidence about the prices charged to Betz: 'Well, from the information I have seen, it was very consistent'. The product profile in European Chemical News for 21-27 February 2005 on 'Polyacrylamide' (referred to in the first paragraph of this judgment) spoke in terms of a market plainly assumed to be global: 'Global capacity for PAM stood at 910,000 tonne/year'; 'World prices have increased in 2004 by 15%, mainly due to the increased raw material costs'; 'Global demand growth is forecast at 5.4%/year to 2010'. 36 There was tendered also a number of documents which supported the notion that prices were negotiated on a worldwide rather than local basis: a price list for Ashland for various products of 15 September 1999; a letter from SNF Group to Hercules of 4 December 2002 enclosing a revised price list for anionic emulsions with the list itself being prominently headed 'Revised Global Prices' and setting out prices for various products; a similar letter to GE Betz of 26 November 2002 attaching a list of 'global prices' for powders, emulsions and solutions with the products listed in groups; another revised price list for GE Betz expressed to take effect from 1 July 2003 and setting out prices for similar products; a combined price list for Betz, Hercules and EF Houghton of 1 January 2001 setting out identical prices for a number of products and in particular fixing a price of $US0.8221 for a product called AN 923 PWG; a price list issued by the US Supplier for Ashland showing, in particular the same product being sold for US$0.77; another price list issued to GE Betz effective from 1 October 2003 showing the products grouped together (as Mr Karoudjian had done) and featuring largely similar pricing. 37 All of this evidence, detailed and particular, pointed to the existence of a global market. Standing back from the evidence that conclusion should hardly be surprising: the products in question were high volume industrial chemicals used in worldwide industries and inherently transportable. It is difficult to see how the market could not be a global one. The taxpayer also relied on extracts from annual reports for various multinational groups to make good the proposition that their operations were global too. It is unnecessary to set them out; they show, as might be expected, that the operations of these large multinationals were global. 38 The trial judge's treatment of the topic was brief. He set out the evidence in which M. Pich had spoken in terms of a global market and mentioned tangentially the existence of the global price lists: 'Whilst the evidence is unclear as to how the global price lists were referenced in terms of the pricing in different countries or geographic regions, what is clear on the evidence is the actual prices paid by the taxpayer and the customers' (at [139]). Beyond that material he did not go. His conclusion? 'All the evidence and the surrounding circumstances point to a global market' (at [140]). The Commissioner was critical of this reasoning. In his written reply he submitted that the 'other evidence' to which the trial judge referred was 'never set out or even described by his Honour. One does not know what it is'. Consequently, the Commissioner submitted that the trial judge had erred by failing adequately to set out the basis for the finding. This contention was not advanced as a ground of appeal and may be put to one side although it may be observed that it is not altogether without merit. 39 The finding that there was a global market was a pivotal part of the trial judge's reasoning process deployed elsewhere. He dismissed Dr Becker's contention that foreign companies should not be used as comparables: 'I have already indicated that the relevant market is a global market. This finding overcomes the submissions of the Commissioner which related to the comparability of foreign markets and the Australian market' (at [152]). And he dispatched any concerns that there was no evidence of comparable Chinese sales both because the prices 'were in the same range as those paid to the other suppliers' and because 'of the view I take that the acquisition of the products by the taxpayer and the customers and other purchasers occurred in the context of a global market' (at [78]). 40 The Commissioner's submission was that there was no evidence that there was a global market and that the evidence instead showed that the prices charged differed between various markets. He submitted that the evidence of M. Pich and Mr Schroeter referred to above was not evidence of a global market. But this cannot be right: plainly it was. The Commissioner submitted that the taxpayer had not established the size of the market, the manner in which it operated or the identity of its participants. But this too is simply incorrect. The evidence showed the global market was about 910,000 tonnes per year; that the largest participant was the SNF Group and that other participants were, at least, the five companies identified by Mr Karoudjian or - and it will be necessary to return to this later - the much larger number of companies identified by Mr Seve. The Commissioner submitted that the extract from the 21-27 February 2005 edition of European Chemical News from which the market size was derived had been written by a Ms Elaine Burridge and of her nothing was known: 'Moreover, the one page article, on any fair reading, does not support the existence of a global market price for polyacrylamides'. But, of course, the taxpayer was trying to establish not a global market price but a global market. Nor, where it was the Commissioner who introduced Ms Burridge's extract into evidence, is the denunciation of her expertise before this Court a submission redolent of merit particularly as it had been Dr Becker who had put the page forward on the Commissioner's behalf in the first place as a footnote in his first report. In any event, one could reasonably expect that a journal entitled European Chemical News might be written by persons not altogether unfamiliar with that topic. 41 In each of these points one may see, more importantly, a failure on the Commissioner's part to grasp the nub of Mr Seve's evidence. The suggested global market was not put forward by Mr Seve in his report to prove the existence of a universal price but as a global starting point from which negotiations might then lead to variations. The Commissioner submitted that significant aspects of the evidence suggested a substantial price dispersion from the lists. And, of course, it did. M. Pich had accepted that differences existed between the highly competitive Australian market and other markets such as Germany, Spain, Europe and the US. Mr Seve accepted, under cross-examination, that there were dispersions away from the price list and he conceded in answer to a question of whether it was not the market which set the price but 'the negotiation of the parties?' that this was '[c]orrect. Which is actually the critical point'. This led then to this question: 'So the markets are irrelevant?' to which he replied '[y]es, actually'. The Commissioner made much of this but to little effect. Comparability did not arise because the prices charged in different countries were the same; it arose because there was a single market. If Mr Seve's evidence had been that the five companies were comparable because the prices charged in each country were the same, the Commissioner's argument would have followed. But Mr Seve made no such suggestion. 42 One must reject therefore the Commissioner's written submission in reply that M. Pich's evidence fell 'far short of what is needed to prove the existence of a global market and, more significantly, the existence of a global market price'. None of the taxpayer's witnesses was trying to establish a global market price: it was not Mr Seve's evidence and it most assuredly was not M. Pich's evidence. Revealed in the submission instead is the Commissioner's failure to grasp the point being put against him; to understand the difference between comparability arising from the existence of a single market and comparability arising from an identity of price. 43 This appeal, it is true, is not concerned with a process of valuation but there are nevertheless certain parallels between the determination of arm's length consideration and the valuation of property. One of those parallels is the use, in some circumstances, of a methodology of comparables. Where, as here, that methodology is used one may utilise comparables from the same market - say the market in the eastern suburbs of Sydney for two bedroom terraces - but it would be eccentric indeed to say that a variation in the prices for such terraces ('price dispersion') signalled the non-existence of the market. We should rather take such variations to prove the existence of competitive advantages and disadvantages within the market. So here, the fact that there was price dispersion in the polyacrylamide market did not disprove the existence of that market so long as the dispersion related to competitive rather than local effects. 44 The taxpayer made this point with especial force by pointing to evidence given by Mr Seve under cross-examination. Counsel for the Commissioner asked what independent verification he had undertaken with respect to the sales data to which Mr Seve replied: 'I used the evidence of price dispersion within markets to corroborate what was told to me in relation to price dispersion not being related to geographic markets but related to individual customers or individual clients'. In that circumstance, no reason appears to interfere with the trial judge's conclusions on this matter, densely expressed although they may have been. 45 The last of Mr Seve's five comparability factors was the business strategies of the parties. Mr Seve explained that these strategies were 'generally less likely to impact a CUP analysis of this type than other comparability factors' but thought that they could be relevant ('CUP analysis' stood for comparable uncontrolled price analysis; uncontrolled denoting, in substance, independence). He instanced, as an example, the situation where a taxpayer seeks to penetrate a market or to increase its market share by agreeing 'to purchase a product for a price that leads to an unprofitable outcome for that product in order to widen its range and in so doing, provide a better service level to existing and potential customers'. Mr Seve thought data about such strategies difficult to obtain and he acted on the assumption that there were no significant strategic issues 'relevant from a comparability perspective' with respect to the five companies. 46 The trial judge did not deal with this topic at all, making no findings about it and this, it is convenient to assume, was an error on his part enlivening an obligation in this Court to consider what the correct finding about it should have been. The Commissioner submitted that the taxpayer had not led any evidence about the five companies' business strategies so that this assumption on which Mr Seve's report rested had not been made good. But why was that so? One must surely start with the assumption that the business of the five companies was conducted on an ordinary profit making basis: the taxpayer's burden of proof did not require it to tender proof positive of every fact no matter how mundane or pedestrian; no burden fell on it to prove that the comparable purchasers had been properly incorporated or, more importantly, that they did in fact carry on business. Matters of that class - the ordinary affairs of commerce, the matters of everyday regularity - could safely be assumed unless they were contradicted by evidence. Into that class of facts may be put the proposition that the proposed comparable purchasers were selling their products to make a profit. Evidence was not required from the taxpayer to prove that matter because the onus lay on those who wished to contradict it. No doubt the Commissioner could have sought to prove that, for example, Akzo Nobel NV, had engaged throughout the period in aggressive price cutting in order to obtain market share but it would be curious indeed if the taxpayer had been required, in the absence of any suggestion to that effect, to prove that it had not. Because the trial judge did not deal with this matter, it is appropriate to make the findings which should have been made: there was no reason to think that the businesses of the five companies were conducted on other than the ordinary basis of seeking to make profits from their distribution activities or that they were pursuing any out-of-the-ordinary strategy which might have affected the prices paid by them. 47 It follows that the Commissioner's contention that the first set of comparables were not sufficiently proved is incorrect and must be rejected. Those comparables were supported by direct testimony from Mr Schlag (as to all five companies), Mr Schroeter (as to two), Mr Karoudjian (as to one), M. Pich (as to all), substantial documentation in the form of annual reports, the evidence of Mr Seve as to comparability (which was admissible) and the evidence of a global market. It was, by itself, a powerful case. A similar process was undertaken by Mr Schroeter for the US supplier to that which Mr Karoudjian had undertaken for the French supplier. He used the same five companies and adjusted the prices in the same way to reflect the differing terms. Those prices were generally more than those paid to the US supplier by the taxpayer. Mr Seve did not analyse the US sales data and so expressed no view on Mr Schroeter's analysis. But that analysis used the same product groups as Mr Karoudjian's and the same process of normalizing the sales data in relation to terms of sale. The economic circumstances were the same because the market was global and there was no reason to think that the business strategies of the comparables differed. It is difficult, therefore, to see any reason why Mr Schroeter's analysis did not provide additional evidence that the taxpayer was paying less than arm's length consideration. 48 The Commissioner launched a final volley against Mr Karoudjian's and Mr Schoeter's comparables: only 112 of the 35,000 sales involved occurred in Australia; only 64 of those were direct sales; 99.68% of them occurred in other geographic markets; functional comparability was (again) not established; product group sales did not compare like with like. These points are of no substance. The first three all fail when one accepts, as one should, that foreign markets were comparable. The fourth is wrong for reasons already given. The last is inconsistent with the Commissioner's concession in this Court that the product equivalence was not in issue for Mr Seve's report. In any event, there is no reason to accept that Dr Becker's detection of price dispersion amongst the group members signified anything other than ordinary competitive effects between purchasers.

(b) the second set of comparables 49 The Commissioner's expert, Dr Becker, was critical of the taxpayer's use of the five companies comprising the first comparables. One of his criticisms, set out at paragraph 10 of his second report of 23 March 2009, was that, of the five companies identified for the first comparables, only one, Betz, had purchased any of the products in Australia. The other proposed comparable companies 'only purchased product (from SNF manufacturers) in other countries with potentially different market conditions and potentially different supply chains that did not involve [the taxpayer]'. Dr Becker then analysed 52,000 pages of invoices to conclude that the US Supplier had in fact sold products to Australian and New Zealand companies none of which was included amongst the five companies used in the first comparables analysis. Pausing there, the gravamen of those observations was, presumably, both to suggest the desirability, where possible, of using local comparables and to emphasise the fact that the taxpayer's failure to consider local comparables had not had its source in the headwaters of necessity. 50 Dr Becker's evidence elicited a further affidavit from Mr Schroeter of 9 April 2009 in which he identified the Australian and New Zealand sales the existence of which Dr Becker had previously detected. Cutting out a good deal of detail, it appeared that seven named products had been purchased by three companies. The companies were Ichem Ltd, Fernz NZ and Nalco, through its subsidiary, Nalco Australia (and some later purchasers of those businesses). One of the seven products ('AN934SH') was also purchased by the taxpayer; the rest were either products closely related to products purchased by the taxpayer (for example, premium as opposed to standard products) or were identically priced. These companies were then put forward as relevant comparables. The pricing information showed that the taxpayer almost always paid less for its products than these entities. 51 The trial judge did not think these sales, at least on their own, were sufficient: 'In relative terms, the comparable transactions that occurred in Australia were not great. I accept that on their own the transactions relied upon that took place in Australia would not support the CUP analysis undertaken by the taxpayer. However, particularly in view of the fact that there is a global market, putting together all the comparable transactions relied upon by the taxpayer the burden placed upon the taxpayer is satisfied' (at [146]). There are problems with this. The comparables were not just Australian but included New Zealand companies. More importantly, Mr Seve did not carry out a comparable analysis for these transactions limiting himself to the initial five companies and his own selection. On the other hand, Mr Schroeter, in completing the Australian and New Zealand transactional analysis, carried out the same process on the pricing data that Mr Karoudjian had used for the same five companies, normalising it to remove freight costs and currency issues. 52 Before this Court, the Commissioner submitted that the overlap between the product codes purchased by these customers and those purchased by the taxpayer was around 3%. Thus, so the argument ran, they were insufficient to found a proper basis for comparability. In a sense, it is this submission that the trial judge accepted although it appears he overlooked the New Zealand purchasers. But this made no difference to the trial judge's reasoning which was based on the premise that the existence of the global market allowed one to aggregate all transactions together regardless of jurisdiction. That may well be sufficient to overcome the effects of overlooking the New Zealand purchasers but it encounters the difficulties thrown up by the Commissioner's second argument: the taxpayer did not establish that Fernz NZ, Orica Chemnet or Ichem Ltd were distributors although it did establish that the fourth, Nalco, was a distributor (as to which, see below). It followed from Mr Seve's view that there needed to be functional comparability that evidence about the sales to those entities could not assist the taxpayer. This was a serious objection to this set of comparables which the trial judge appears to have overlooked. In its submissions, the taxpayer did not take this Court to any evidence which suggested that the companies were distributors and it follows that functional comparability was not established. In those circumstances, the trial judge should not have relied on this second set of comparables even in the limited way he did.

(c) the third set of comparables 53 Another criticism made by Dr Becker in his first report related to the fact that the first set of comparables was based on product groups rather than individual products. He observed that the invoices for the US supplier showed that there were price fluctuations of up to 100% within the product groups. That criticism appears to have elicited from the taxpayer a third set of comparables. It was prepared by Mr Seve. Using sales data provided by the French supplier (but not the US supplier) Mr Seve identified the top twenty products by volume and price which had been purchased by the taxpayer. These were signified by an alphanumeric code (for example, AA30A00). These codes were referred to in the appeal as 'code arts'. Mr Seve then identified all of other sales of those products by the French supplier to independent companies. From this information he then identified a number of companies who might be potential comparables by examining the volumes of purchases and markets involved. He then exposed these companies to an analysis using the five comparability factors to which reference has already been made to determine whether they were truly comparable and concluded that they were. Having drawn up a list of comparable transactions, his conclusion was that the prices paid by the taxpayer for these precisely identified products were generally less than the prices paid by the comparable purchasers he identified. 54 The Commissioner submitted that this evidence should have been rejected because the assumptions on which Mr Seve's report rested were not proven. Those assumptions related to his five comparability factors. The first - the character of the property - was not in dispute. The third - the terms of the contracts - was in dispute but it is apparent that Mr Seve had been provided with the raw sales data which contained those terms. His report referred to those terms at paragraph 6.21 (including 'delivery term' and 'payment terms') and he took into account the fact that the taxpayer paid on delivery whereas the comparable purchasers typically were on 14 to 90 day terms. He also adjusted the prices so that they were ex-works prices. All of that material was in the sales data provided to him and placed into evidence (as Exhibit DK 13). The fourth and fifth factors - the economic circumstances of the market and the business strategies - have been dealt with above. 55 The Commissioner's principal submission on this part of the case was that the second factor - functional comparability - was not made good. He submitted that the evidence did not establish comparable functionality for 18 of the entities. The trial judge did not deal with these contentions. It is necessary therefore to deal with each separately. At Appendix 3 to these reasons there is a table which has been extracted from Mr Seve's report and which sets out the comparable purchasers identified by him, the particular products purchased and the prices paid by both the comparable purchasers and the taxpayer.

Bayer 56 Bayer was a purchaser of CB30B40 which was also purchased by the taxpayer. The Commissioner submitted that the only testimonial evidence about the status of Bayer was given by Mr Schroeter and Mr Karoudjian and it was rejected as inadmissible by the trial judge. This submission should be accepted. The taxpayer submitted in response that Dr Becker had accepted that Bayer was a distributor. Now it is true that Dr Becker had selected 260 companies 'with similar activities as [the taxpayer]' and that his initial list included Bayer. But he had gone on further to reduce that list to ensure that the companies 'had similar business operations as [the taxpayer]' and his final list of seven companies did not include Bayer. Read in its full context, Dr Becker's report does not provide satisfactory evidence that Bayer was a distributor. Bayer's annual report for 1998, however, was also placed in evidence. It showed at p 10 that Bayer had several divisions including a chemicals division which included a specialty products group. That group had 'been enlarged to include the paper industry…'. Given the uses to which polyacrylamides may be put it can be inferred that Bayer supplied those chemicals to that industry. The Commissioner submitted that a passage on p 18 of that report ('[t]ogether with Uerdingen, Germany and Baytown, Texas, the Belgian plant is one of Bayer's three major production sites for Makrolon') made good the proposition that Bayer was involved in manufacture. But of what? The section in question was concerned with polymers; indeed, the product Makrolon, as the same page shows, is a polyurethane. In that circumstance, the evidence supports the conclusion that Bayer was a distributor of CB30B40.

Nalco 57 Nalco, through Nalco IT and Nalco MB, purchased AA20A00, BI55A00, BI15B00 and BI45A00 all of which were purchased by the taxpayer. Mr Schroeter gave direct evidence that Nalco was a distributor ('Nalco is a reseller of chemical products to end-users both in Australia and around the globe'). There was evidence in its consolidated financial statements for 2003 that Nalco 'is engaged in the worldwide manufacture and sale of highly specialized service chemical programs. This includes production and service related to the sale and application of chemicals and technology used in water treatment, pollution control, energy conservation, oil production and refining, steelmaking, papermaking, mining, and other industrial processes'. The Commissioner submitted that this did not mean it was a distributor. Given that polyacrylamides are not, as the Commissioner accepted, the constituent ingredients for other products, it is difficult to conceive what Nalco was doing with the products if it was not selling them. The appropriate conclusion is therefore that Nalco was a reseller of AA20A00, BI55A00, BI15B00 and BI45A00.

BTC 58 BTC purchased AA00A00 and BI25A00 which were also purchased by the taxpayer. The only testimonial evidence about its activities was given by Mr Karoudjian and Mr Schroeter but this evidence was held inadmissible and excluded. The only documentary evidence consisted of a document entitled 'Certificate' issued on 9 January 2009 by an organisation known as 'DQS' which was, according to the certificate, a member of 'INET', the international certification network. The certificate was issued in respect of the BTC Group and, in particular, ten organisational units/sites all having as part of their names the expression 'BTC Specialty Chemical Distribution' and being incorporated in one or other of Germany, the United Kingdom, Spain, France, Portugal, Slovakia, Belgium, Denmark or Italy. It certified that BTC and the subsidiaries had implemented and maintained a quality management system in relation to the 'marketing and sales of performance chemicals and functional polymers'. The Commissioner's submission was that this merely showed that BTC had achieved certain standards. That submission should be rejected. Two features of the certificate suggested that BTC was a distributor: the names of the entities and the fact that the business audited was a business involved in selling. The taxpayer emphasised that this latter feature was inconsistent with BTC being an end user of polyacrylamides, a submission which should be accepted. It is appropriate to conclude therefore that BTC distributed AA00A00 and BI25A00.

BASF Fin 59 BASF Fin purchased MY00Z61 which was also purchased by the taxpayer. The testimonial evidence about BASF Fin came again from Mr Schroeter and Mr Karoudjian. Mr Karoudjian's evidence was rejected but Mr Schroeter's evidence that BASF Fin distributed in Australia was admitted. The only documentary evidence was its financial statement for 2003. That report revealed BASF Fin to have a chemicals division matters but beyond that are not clear. What is clear is that BASF Fin was a chemicals company and not a participant in the pulp and paper milling, sewage treatment or mining industries. One cannot be certain what it was that BASF Fin was doing with the MY00Z61 which it purchased but it is unlikely that it was using it for its own purposes since it was not involved in the industries which used polyacrylamides. Whilst the evidence is not altogether satisfactory, it is a sufficient basis for concluding that BASF Fin was not an end-user of the product.

Chemipol 60 Chemipol purchased BI25A00 which was also purchased by the taxpayer. The testimony of Mr Schroeter and Mr Karoudjian about the activities of Chemipol was ruled inadmissible by the trial judge. There was, however, a document tendered from a website known as 'OneSource' which described its business. OneSource appears to be a corporate information service. The Commissioner submitted that the extract showed that it was in the chemical manufacturing industry and, indeed, it did recite: 'Industry: Chemical Manufacturing'. The difficulty is that it also said this: 'Business description: Wholesale of chemical products'. Given the use to which polyacrylamides may be put, it is difficult to see that it could have been using BI25A00 other than for distribution purposes. The appropriate conclusion is that Chemipol distributed BI25A00.

Acideka 61 Acideka purchased BI55A00 which was also purchased by the taxpayer. Mr Schroeter's and Mr Karoudjian's evidence about its activities was not admitted. A company extract for Acideka downloaded from 'OneSource' described Acideka's business as '[p]roduction and distribution of chemical products'. The Commissioner submitted that this was evidence that it was a manufacturer but it is hard to accept that submission given the word 'distribution'. Further, given the uses to which polyacrylamides may be put it is difficult to see that Acideka could have been doing anything with BI55A00 other than distributing it. The appropriate conclusion is that Acideka distributed BI55A00.

Betz, Ashland, Hercules and Buckman 62 The conclusion has already been drawn above that these four companies, which comprised four of the five companies in the first set of comparables, were distributors.

Kemira 63 Kemira purchased BI45A00 which was also purchased by the taxpayer. Mr Schroeter and Mr Karoudjian gave evidence about its activities but this was rejected by the trial judge. Its annual report for 1998 contained this statement: 'Kemira Chemicals is a major producer of chemicals for the pulp and paper industry and a manufacturer of other industrial chemicals. It is also the leading European manufacturer of water treatment chemicals. It has production facilities in 24 countries'. The Commissioner submitted that this showed it was a manufacturer and this would appear to be correct. It is difficult to understand why Kemira, which, from the appearance of its activities, was likely to be a producer of polyacrylamides, would be purchasing polyacrylamides from the SNF Group. On the other hand, it is difficult to imagine, given their use, any purpose for which it could have been using them itself. In the circumstances, which include the absence of any direct evidence of distributing activities, the matter is too unclear to permit an affirmative finding. The appropriate conclusion is that it has not been shown that Kemira was a distributor of BI45A00.

Nerolan, Chatzopo, Nordman, Clarflok, Bioconsult, Henke, Ribco, Kempro, Acquaplan 64 The only testimonial evidence of the activities of each of these nine companies came from Mr Schroeter and Mr Karoudjian and this evidence was rejected as inadmissible. The Commissioner submitted that there was no other evidence about these companies' activities and the taxpayer did not take the Court to any evidence to contradict that proposition. In those circumstances, it was not established that any of these were distributors.

Nobel 65 No submission was made by either party about Nobel. It is not clear if it is to be equated with Akzo Nobel NV. It is appropriate, therefore, to proceed on the basis that it has not been shown that Nobel was a distributor. 66 The table prepared by Mr Seve which is at Appendix 3 thus included a number of businesses which were not proved at trial to be comparable because their status as distributors had not been established. Nevertheless, a significant number were shown to be comparable: Bayer, Nalco, BTC, BASF Fin, Chemipol, Acideka, Betz, Ashland, Hercules and Buckman. If Mr Seve's table is modified to remove those purchasers who were not shown to be on-sellers it still shows that each of the remaining entities paid more than the taxpayer. Using the average prices paid across the seven years in question, the table at Appendix 3 shows the taxpayer paying consistently lower prices for its products than the other independent distributers as this summary table (prepared from Appendix 3) demonstrates: Product Distributor and average price paid Taxpayer's average price paid CB30B40 BAYER €1.59 €0.99 AA20A00 NALCO IT €1.96 €1.37 BETZ FR €1.77 €1.37 AA00A00 BTC €1.82 €1.42

• MY00Z61 BASF FIN €0.48 €0.49 BI55A00 ACIDEKA €2.24 €2.20 NALCO MB €3.09 €2.20 BI25A00 BTC €2.29 €2.21 CHEMIPOL €2.49 €2.21 BI15B00 NALCO MB €2.69 €2.30 BI45A00 NALCO MB €2.89 €2.24 AA20A00 GE BETZ €1.81 €1.37 DJ60B47 GE BETZ €1.72 €1.56 BI45A00 GE BETZ €2.76 €2.24 DJ30B42 ASHLAND €1.63 €1.33 BI55B00 ASHLAND €2.82 €2.26 BI10B00 BUCKMAN €2.77 €2.10 HERCULES €2.36 €2.10 BI20B00 P GE BETZ €2.57 €2.16 BI10A00 BUCKMAN €2.86 €2.04 BI40B00 BUCKMAN €2.92 €2.27 BI20A00 BUCKMAN €2.91 €2.28

It will be seen that only in the case of MY00Z61 and BASF Fin did the taxpayer pay a higher average price and then only by 1 Euro cent. 67 This was evidence capable of supporting the taxpayer's case that it was paying less for the same products than other independent distributors. The Commissioner nevertheless denied that this was so: Mr Seve's analysis of the individual chemicals covered only twelve products which comprised only 35-40% of the total dollar value of the taxpayer's purchases from the French supplier. This submission is unpersuasive. The comparables prepared by Mr Seve were the only comparables available for the whole of the sales data which had been examined and it was these twenty products (not twelve as the Commissioner submitted) which alone had been bought by the taxpayer and the other purchasers in sufficient volumes to warrant consideration. The evidence suggested that, for those products, the taxpayer paid less. Without some suggestion that the method of selection adopted by Mr Seve had somehow slanted the process in the taxpayer's favour - an allegation not made by the Commissioner - this material provided a proper basis for concluding that the taxpayer had paid lower prices. 68 Although the reasons given by the trial judge for reaching the same conclusion were inadequate in that they failed to deal with any of the evidence, the conclusion reached by him that the evidence established that the taxpayer had paid prices 'which were lower than the larger majority of prices paid by the purchasers in comparable transactions over a similar period of time' (at [144(c)]) is essentially correct.

Conclusions on comparables 69 The evidence before the trial judge consisted of three different sets of comparables. Although his Honour distinguished between them in an early part of his judgment, the critical parts of his reasoning did not draw the distinction. This is to be regretted: the Commissioner's objections were detailed and whilst some, perhaps many, lacked merit they still called for a reasoned analysis. It is possible to perceive deficiencies in the trial judge's approach to all three sets of comparables. They were not claimed to have the same terms and the trial judge overlooked the evidence given by Mr Karoudjian, Mr Schroeter and Mr Seve of the process of normalization. The trial judge does not appear to have grappled with the consequences of Mr Seve not having expressed a conclusion about the second set of comparables. More importantly, his Honour did not deal with the significant issues as to whether any, and if so which, of the purchasers in the second or third set of comparables were distributors. The global statement that the evidence was that they 'were not end users of the products but, like the taxpayer, on-sold such products' (at [153]) did not accurately reflect the evidence before him which was much more complex. Nor is this a case where one is confronted, as often happens, with minor gaps in what is otherwise a thorough analysis. Errors are shown in the trial judge's approach and a duty is therefore cast upon this Court to review the material for itself. 70 That review, however, shows that the first and third set of comparables were - in the ways outlined above - appropriate. About the first set there can be little doubt. As to the third set, the reduction in the number of available comparables does not affect the acceptance of Mr Seve's opinion. On the basis of that material, it is clear that the taxpayer paid less for the same or similar products in comparable markets than arm's length purchasers did. The trial judge's conclusion was correct. 71 There are two final footnotes to that conclusion. First, the evidence presented did not touch upon the position of the Chinese supplier. As the trial judge pointed out, however, the relevance of that lacuna was diminished when the market was recognised to be global. Secondly, so long as appropriate comparable transactions were available, an analysis of relevant comparables was appropriate. Dr Becker thought the comparables approach was inapplicable because proper comparables were not available. It was that paucity of comparables which drove him to use the TNMM. It follows from our acceptance that the existence of an extensive range of comparables has been established that the premise on which Dr Becker's use of the TNMM rested is not made good. The Commissioner's submission that the trial judge should have accepted the TNMM is, therefore, incorrect.

III. The evidence of M. Pich 72 The taxpayer's case at trial was that the prices paid by the taxpayer to the suppliers were less than those being paid by independent third parties. It accepted that it had been incurring losses for a number of years but claimed that it had been pursuing a strategy of increasing its market share. 73 In the midst of the debate about that part of the case the taxpayer called M. Pich. He was cross-examined by counsel for the Commissioner. During the course of that cross-examination, he gave evidence that he desired to make the French supplier cash flow positive. The following exchange then occurred: Mr Steward: Do I take it then that you would agree with me that the reference there to France, 'cash flow positive' is a reference to an immediate goal? The witness: In France, during the years 1997-2002, we lost 3.7 million euro on shipping product to Australia, which means around $6 million during these years. We sold under cost during all the period. And after it was - we wanted to become - the goal of the company is to be about five percent positive in France and five per cent positive for its subsidiaries, this is a goal - original goal. [sic] 74 The revelation that the French supplier had been making a loss on the products it sold to the taxpayer was an important factual development for it tended to support the taxpayer's other evidence that it was paying less than independent third party distributors. 75 The cross-examiner sought at once to counter this evidence as the following exchange shows: Mr Steward: Mr Pich, where in your affidavit do you refer to this loss of 3.7 million euro? Take me to the paragraph, please. The witness: Well, it is not inside. Mr Steward: It's not inside the ---. The witness: No. Mr Steward: It's not in your affidavit? The witness: No, no. We have calculated. And is difficult to calculate because we have only the costs per year and shipment per quarter. But the average is about 3.7 million euro Mr Steward: And when did you do this calculation, Mr Pich? The witness: We have done sometimes ago, two, three weeks ago. Mr Steward: Two or three weeks ago? The witness: Yes. Mr Steward: You made this calculation? The witness: Yes. Mr Steward: Who asked you to do it? The witness: Is Mr Karoudjian. Mr Steward: Mr Karoudjian asked you to do it, did he? The witness: Yes. Mr Steward: And why did he ask you to do it? The witness: Because I wanted to know how much losses were made in France during this period. [sic] 76 Unsurprisingly, counsel for the taxpayer re-examined M. Pich about this evidence: Mr De Wijn: Yes. Now, can I ask you this question. If SNF Australia was not a subsidiary of SNF France, would you have sold product to it at the prices that you sold to it over the period 1997 to 2003? The witness: No. Consistently we were under the price we sell to other parties. Mr De Wijn: If your Honour pleases. His Honour: Well, did you want to object to that question, Mr Steward. I don't --- Mr Steward: Well, we'll make a submission about that later, your Honour. Mr De Wijn: Well, my learned friend asked a whole lot of questions about the strategy on the basis of them being related companies. So, with respect, the question is a proper question. Mr Steward: Well, with respect, it's not, it's a hypothetical question, not a question about facts. But, anyway, your Honour---- His Honour: Well, if that's the only objection you have, then we can deal with that in a submission. Mr Steward: That's what I was planning to do. His Honour: I was more concerned that you were concerned about it coming at a particular stage in the proceedings, but that's not a problem you're taking, so don't worry about it. Anything further, Mr De Wijn? Mr De Wijn: No, your Honour. 77 There are three aspects of this evidence which deserve emphasis. First, the witness gave evidence three times that the prices charged to the Australian distributor were less than those charged to other parties. That evidence was, for reasons already given, essentially correct. Secondly, his Honour was astute to avoid any potential unfairness attending the late leading of the evidence. Thirdly, the Commissioner's only objection was to its hypothetical nature. 78 The trial judge began his treatment of this evidence by recording that no objection had been taken to it but that the Commissioner had submitted that it should not be accepted in many respects. The trial judge continued: 134 I say at the outset that subject to one qualification I accept the evidence of Mr Pich. As I say, his credit was not attacked, and I consider the evidence he gave logical and in conformity with commercial reality, once one accepts (as I have) that the taxpayer was being supported at all times by its parent in pursuit of Mr Pich's world wide objective. 79 His Honour's one qualification related precisely to the evidence to which we have just referred: 135 The one qualification relates to evidence given in re-examination. In re-examination Mr Pich did make a reference to losses made by SNF France, based upon calculations made some two or three weeks prior to giving evidence. Those calculations were not produced to the Court, nor were they called for by the Commissioner. No reference was made to these losses in Mr Pich's sworn affidavit. There was some lack of clarity in the exact amount and period of the losses, and I am not prepared to accept the figures given by Mr Pich. However, I am prepared to accept, although I regard it as irrelevant to the main task, that losses were made in the relevant period by SNF France which Mr Pich regarded as creating a 'catastrophic situation' in France. I do not see any inconsistency in this oral evidence with any of the documentation before the Court. 80 The trial judge was therefore alive to the risks posed by M. Pich's presentation of the evidence so late in the piece. But, being aware of that risk, his Honour nevertheless went on to conclude that sales had occurred at loss-making prices for the French supplier. His Honour rounded out his treatment of this topic at [168]-[169] in these terms: 168 The suppliers supported the taxpayer by charging special reduced prices for the taxpayer's purchases of polyacrylamide that were destined for particular large customers, and also by bearing that currency risk on sales to the taxpayer. The evidence establishes that the reason the suppliers tolerated sales below cost price, and less than the price charged to arm's length purchasers, was that the SNF group was supporting the taxpayer. 169 Further, as I stated previously, the outcome of this price support was that in addition to the losses incurred by the taxpayer in Australia, the SNF group lost some money on its sales to the taxpayer as it was selling to the taxpayer at below the cost of production. 81 The Commissioner submitted that these findings involved error because the trial judge had failed to test closely M. Pich's oral testimony and to assess its weight 'particularly given its consequence, for the first time, as an outburst in cross-examination'. Reliance was placed upon the statement of Fullagar J in Pascoe v Commissioner of Taxation (1956) 30 ALJ 402 at 403: Where a person's purpose or object or other state of mind in relation to a given transaction is in issue, the statements of that person in the witness box provide, in a sense, the 'best' evidence, but, for obvious reasons, they must, as Cussen J observed in Cox v Smail ([1912] VLR 274, at p 283), 'be tested most closely, and received with the greatest caution'. 82 The difficulty with this submission is that the trial judge was aware of the need for the evidence to be closely scrutinised: he rejected the evidence insofar as it dealt with precise figures; the exchange which occurred between his Honour and counsel for the Commissioner during the re-examination shows that the trial judge was well aware of the difficulties posed by the late emergence of the evidence. It may be doubted whether Fullagar J in Pascoe advanced anything more than the caution - customarily applied by trial judges - that the evidence of witnesses who have an interest in the outcome of litigation needs to be approached critically. 83 The Commissioner submitted that the 'veracity' of M. Pich's evidence needed to be 'tested' by this Court against three matters. The first of these was said to be the fact the consolidated financial statements from the SNF Group confirmed that it had enjoyed an average operating profit of about 7.5% in the years in question. But this is simply illogical: there is nothing untoward in the notion of a group as a whole being profitable but individual components of it running at a loss. 84 The second point was that M. Pich's evidence took the form of an 'unprompted exclamation'; that it 'was itself made in an inconsistent manner'. As to the 'outburst' point, the only record this Court has are the words on the pages of the transcript. Nothing in that transcript provides a basis for concluding that there was an outburst. Nor do we perceive to the requisite standard the inconsistency to which the Commissioner refers. It is true that M. Pich gave evidence under cross-examination that the French supplier had lost AUD$6 million shipping product to the taxpayer during 1997-2002. It is also true that shortly afterwards he gave evidence that 'we have lost 12 million during this period in a year. We have lost 5 million in France - 5/6 million in France, which it was a catastrophic situation for us. It was - we have lost 17 million in fact during the period.' But this evidence is not contradictory - in both cases, losses made by the French supplier in selling to the taxpayer were around AUD$6 million. It is true that in the first part of the answer he refers to the loss having been suffered 'during this period in a year' but in the latter part he refers to 'the period'. It is quite possible that he was intending to refer to the period 1997-2002 and that the first reference to 'year' was a slip. It is difficult for an appellate court to assess such a matter. Correspondingly, it is almost impossible for this Court to say on the strength of such material that M. Pich concocted his evidence about the losses. Other much more innocent explanations are available which, as it happens, are consistent with the trial judge's conclusion that M. Pich was a witness of credit. The material provides no basis for appellate interference. 85 In effect, the Commissioner invites this Court to depart from the trial judge's finding that M. Pich was a witness of credit, to find that his evidence about loss making was false and to do so because this Court is meant to discern from the transcript an 'outburst' and apparently inconsistent evidence involving the word 'year'. Not only does such an invitation fly in the face of repeated instructions from the High Court (Fox v Percy (2003) 214 CLR 118 at 127 [26] per Gleeson CJ, Gummow and Kirby JJ) but it does so with material which does not begin to pass the threshold for seeking to have this Court conclude, in effect, that M. Pich was lying. 86 The third point was that M. Pich's evidence was contradicted by another witness, Mr Schroeter. He was said to have agreed in cross-examination that various suppliers were earning a profit on the products sold to the taxpayer. In order to understand this point, it is necessary to digress slightly. Mr Schroeter was the managing director of the taxpayer and took up that position in November 2002. M. Pich told Mr Schroeter that he expected the taxpayer to be self-funding for all growth and future capital expenditure from 2005. Mr Schroeter gave evidence at paragraph [58] of his affidavit that, for his own benefit, he had analysed the transfer prices in late 2002 or early 2003 'to ensure that I was comfortable that the transfer prices were reasonable'. He prepared a document entitled 'Full Absorbed Cost by Plant for Typical Production' which was exhibited as RHS 9 to his affidavit. The Commissioner objected to this evidence and it was rejected. The trial judge said (at [109]). I also indicate that the 'costing model' presented by Mr Schroeter (which was also objected to by the Commissioner on the basis of relevance) is not relied upon by me in my consideration, and in my view, is irrelevant. 87 Despite its rejection, the Commissioner nevertheless cross-examined Mr Schroeter on the document. The relevant part of the cross-examination was follows: And the final column represents in your mind the price at which France or another SNF manufacturer would sell to SNF Australia? --- No, the final column is the actual price. The actual price. Thank you. And if you could go to the third row, under the heading, Powders. Do you see that? --- Yes According to you - no, I beg your pardon. The fourth row. Do you see the figure 2.02? --- I do, yes That represents for that product the full absorption cost for the relevant SNF manufacturer, does it? --- That's my estimate, yes. And then they sell in Australia at 2.28 euros? --- Yes And would you agree with me, would you not, that based upon those figures there, the seller into Australia is making a profit on the sale of that product, being the difference between 2.28 and 2.02 euros? --- Less freight and handling to get it to Australia. 88 The Commissioner relies upon this as one of two items of evidence to support a submission that 'Mr Schroeter agreed in cross-examination that various suppliers were earning a profit on the products'. 89 The second item of evidence was another extract from the cross-examination of Mr Schroeter which was in the following terms: So if we go back to the fourth column, I take it that, even after taking into account freight, your figures there a profit is made by the seller of the SNF product there mentioned to Australia? --- A profit of --- Just yes or not will do? --- Yes. We can do the mathematics. Thank you. Similarly, in the next column, for the product FO4440SH, you agree with me, do you not, that the seller of that product, based on your figures, makes a profit for the sale to SNF Australia? ---They do. Could you go to the next page? You see the fourth row, and you also see the cost of freight is here estimated to be 0.17? --- Yes. This is all done in dollars now, not in euros - US dollars. Do you agree with me, do you not, that, based on your numbers, a seller of that product in the fourth row makes a profit from the sale of that product to SNF Australia? --- It does. Go the next page, 589. This is again expressed in US dollars. Do you see that? --- Yes. Cost of freight is again 0.13? --- It is. And do you agree with me that in relation to each of the sales therein referred, based upon your number, the seller of the product to SNF Australia makes a profit? --- They do. 90 Because the Commissioner successfully objected to the tender of the document to which this cross-examination relates it is difficult to follow this evidence. An attempt was made to place the document before us by the taxpayer but we do not think it should be examined since it does not form part of the trial record. The best that one can say of the first piece of evidence is that Mr Schroeter agreed that one of the suppliers was selling something called 'powders' to the taxpayer at a profit of €2.28 less freight and handling. Was this a single transaction? When was it? How much was involved? Which supplier? For how long? Without the answers to these questions one does not really know what this evidence means. And without knowing what it means no sufficient material appears to interfere with the trial judge's acceptance of M. Pich's evidence. So too, the second piece of evidence only shows that a supplier was selling FO4440SH at a profit to the taxpayer. The same difficulty arises. 91 In those circumstances, the three matters relied upon by the Commissioner do not warrant this Court's interfering with the trial judge's acceptance of the evidence of M. Pich. This Court has been taken to nothing which casts any doubt on M. Pich's credibility or honesty.

IV. The requirements of s 136AD(3) 92 The Commissioner submitted that ss 136AD(3) and 136AA(3)(d) together required the positing of a hypothesis. The hypothesis was to consist of an arm's length purchaser in identical circumstances to the taxpayer; in practice, a comparator with all of the qualities of the taxpayer relevant to price save its membership of the SNF Group. On the particular facts of the case, this confined comparable purchasers to those who had a similar history of losses to the taxpayer. 93 The learned trial judge rejected this argument (at [44]): I do not accept the Commissioner's submission 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. The essential task is to determine the arm's length consideration in respect of the acquisition. One way to do this is to find truly comparable transactions involving the acquisition of the same or sufficiently similar products in the same or similar circumstances, where those transactions are undertaken at arm's length, or if not taken at arm's length, where suitable adjustment can be made to determine the arm's length consideration that would have taken place if the acquisition was at arm's length. Just as in a valuation, the focus is not on the subjective or special factors of the parties involved in the transaction (eg whether they were financially sound or not), but is on the transaction itself and the consideration paid. In this sense, the task is not dissimilar to that undertaken in a valuation - see eg Boland v Yates Property Corporation Pty Ltd (1999) 167 ALR 575 at [82] to [83] and Spencer v The Commonwealth (1907) 5 CLR 418. 94 One should reject at the outset the Commissioner's submission that the trial judge 'impermissibly substituted the inquiry mandated by s 136AD…with an inquiry as to valuation.' The trial judge said no such thing. This reference to valuation was in the nature of an analogy. The trial judge did not reason that s 136AD required a valuation approach but construed it as requiring the circumstances of the taxpayer to be disregarded. That conclusion his Honour regarded as similar to the approach in valuation cases. The words '[i]n this sense, the task is not dissimilar to that undertaken in a valuation' are reconcilable with no other interpretation. 95 Apart from that unpromising false start, the Commissioner began his submissions on this point uncontroversially. The Court was to begin, he submitted, with Division 13 and the text of s 136AA(3)(d) and, in particular, with its stipulation that arm's length consideration was taken to be 'the consideration that might reasonably be expected to have been given or agreed to be given in respect of the acquisition if the property had been acquired under an agreement between independent parties dealing at arm's length with each other in relation to the acquisition' (emphasis added). Next, the Commissioner submitted that the provision required the erection of a statutory hypothesis based on applicable facts; in effect, a statutory deeming provision. The hypothesis was to consist of all of the facts in the real world with one fact changed by the deeming provision, namely, the fact that the taxpayer and its parent were not independent of each other. This required one to assume a set of facts in which the purchaser was an entity with all of the qualities of the taxpayer except its relationship to the parent manufacturers. Support for this reading of the provision was to be found, so the Commissioner submitted, in the principle that statutory fictions erected by deeming provisions were to be strictly construed (Federal Commissioner of Taxation v Comber (1986) 10 FCR 88 at 96 per Fisher J) and the allied principle that one should not travel beyond the hypothesis created by the statute (Estee Lauder Pty Ltd v Federal Commissioner of Taxation (1988) 19 ATR 1228 at 1236). 96 The Commissioner's argument hinges, as he submitted in reply, on the proposition that one of the independent parties referred to in the definition of 'arm's length consideration' in s 136AA(3)(d) was the taxpayer. This was the case, so he submitted, because the definition provision had to 'be read in the context of s 136AD, the operative provision' and that provision commenced 'its inquiry with the 'taxpayer' and it is the 'taxpayer' which is the subject of the hypothetical in s 136AA.' The three propositions for which the Commissioner contends are, therefore: (a) the definition provision must be read in the context of the operative provision; and (b) the operative provision - s 136AD(3) - begins its inquiry with the taxpayer; (c) therefore the hypothesis required by the definitive provision - s 136AA(3)(d) - must relate to the taxpayer so that 'arm's length' in that provision means 'arm's length from the taxpayer'. 97 The critical words of s 136AA(3)(d) in question are 'between independent parties dealing at arm's length'. It is the burden of the Commissioner's argument that one of the independent parties to whom the provision refers must necessarily be the taxpayer. That observation lays bare an ambiguity in the expression 'independent parties at arm's length'. In the context of ss 136AD(3) and s 136AA(3)(d) those words could equally cover any of the following three situations: (a) a purchase by the taxpayer from a hypothetical arm's length supplier; (b) a purchase by a hypothetical purchaser from the taxpayer's actual supplier; (c) a purchase by a hypothetical purchaser from a hypothetical arm's length supplier. 98 That ambiguity underscores the fact that the description of a transaction as being at arm's length is a statement about the independence of two parties from each other. The connexion thus disclosed is a relative one. Generally speaking a statement that two parties have a relative connexion of a particular kind does not carry with it any information about their absolute status. A requirement, for example, that two businesses be more than 20 km apart says nothing about where either business is situated. If one were to look at the definition provision in s 136AA(3)(d) in isolation it would be unsound to read it as requiring any more than that the two parties in question should be independent of each other; that is, the ordinary meaning is not as the Commissioner contends. 99 The question then is whether the ordinary meaning is somehow displaced or modified by the fact that the definition provision feeds into an operative provision - s 136AD(3) - which in turn utilises it to assess the position of the taxpayer. There is no doubt that s 136AD(3) is, as the Commissioner submits, about the taxpayer; that it requires a comparison between that which was actually paid by the taxpayer and an arm's length consideration; and, that, in appropriate circumstances, it then substitutes one for the other. However, it does not follow from acceptance of all those features that arm's length consideration - which does not, in general, refer to the actual position of either party - must be treated as overlaid by a further requirement that the consideration not only be at arm's length but that the arm in question be attached to the taxpayer. 100 The Commissioner sought to buttress his position by reference to the Explanatory Memorandum which accompanied the passage of the Bill introducing Division 13. It contained a statement that the provision would require the Commissioner to redetermine the taxpayer's assessable income: … basically by using the internationally accepted 'arms length' principle, a principle relevant under existing section 136. The arm's length principle is also at the base of provisions in each of Australia's comprehensive double taxation agreements that enable the determination of profits attributable to business activities in one or other of the countries concerned. 101 The internationally recognised arm's length principle required, so the Commissioner submitted, an analysis involving an inquiry into what a purchaser in identical circumstances to those of the taxpayer would have paid, but for its membership of the group. Where was this principle to be found? The Commissioner contended that resort could be had to the OECD Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations (the 'guidelines'). At [1.6] these stated that the arm's length principle seeks 'to adjust profits by reference to the conditions which would have obtained between independent enterprises in comparable transactions and comparable circumstances'. 102 These words demanded, so submitted the Commissioner, that '[o]ne simply removes the fact of interdependence and non-arm's length dealing, but otherwise the exercise involves taking into account all the circumstances which bear on the price'. The deeply impractical nature of this submission is manifest from the outset. It seeks to discern the presence of a strict norm of operation inflexibly requiring one kind of comparable and forbidding all others and it refuses to admit the possibility of making adjustments for differences. That very degree of inflexibility underlay the Commissioner's submission that the trial judge's acceptance of any of the three sets of comparables which had been placed before him under this part of the argument were to be seen as not only counselled against by Dr Becker but forbidden by the very terms of s 136AD(3) (and s 136AA(3)(d)) as understood in light of the OECD guidelines. But what is to occur, one may ask, if no such comparables are available; what if there exists no other business sharing all the same features of the taxpayer bearing on price so that the crystalline perfection the Commissioner submits is demanded by s 136AA(3)(d) cannot be achieved? The Commissioner's submission necessarily means that a taxpayer, who bears the onus in tax appeals, can never succeed in such a case for the bar will be set at an unattainable height. 103 It is highly unlikely that the guideline's reference to comparable circumstances was intended to bring about such eccentric outcomes. Indeed the guidelines upon which the Commissioner seeks to rely make clear that such outcomes were not contemplated at all by the OECD. For example, it did not intend that only such rigidly identical comparables should be brought to bear: 'To be comparable means that none of the differences (if any) between the situations being compared could materially affect the condition being examined in the methodology (eg price or margin), or that reasonably accurate adjustments can be made to eliminate the effect of any such differences' (guidelines at [1.15]). It followed that 'when making the comparisons entailed by application of the arm's length principle, tax administrations should also take these differences into account'. Indeed, it was precisely these guidelines on which Mr Seve relied for his five comparability factors. Paragraph 1.17 was as follows: As noted above, in making these comparisons, material differences between the compared transactions or enterprises should be taken into account. In order to establish the degree of actual comparability and then to make appropriate adjustments to establish arm's length conditions (or a range thereof), it is necessary to compare attributes of the transactions or enterprises that would affect conditions in arm's length dealings. Attributes that may be important include the characteristics of the property or services transferred, the functions performed by the parties (taking into account assets used and risks assumed), the contractual terms, the economic circumstances of the parties, and the business strategies pursued by the parties. These factors are discussed in more detail below. 104 Thus, far from forbidding Mr Seve's approach, the guidelines recommended it. The Commissioner also placed reliance upon the decision of the Special Commissioners in DSG Retail Limited v Commissioners for her Majesty's Revenue and Customs (2009) UKFTT 31 (TC) 1. That case was concerned with the construction of s 770 of the Taxes Act 1988 (UK) which used the expression 'the price which it might have been expected to fetch if the parties to the transaction had been independent persons dealing at arm's length'. The Special Commissioners said (at [78]): Third, s 770 requires one to determine the price which would have been paid if the parties 'had been independent parties dealing at arm's length'. It seems clear to us that those words do not require any adjustment to be made in setting the price to the actual characteristics of the parties other than their independence. The actual assets, business and attributes of each party remain constant and may be relevant to the determination of the arm's length price. The language of para 1(2)(a) is different: 'differs from the provision which would have been made between independent enterprises'. It is at first sight possible that those 'independent enterprises' may not be enterprises which do not share the same attributes as the actual parties to the provision. But is clear to us that that interpretation is not consistent with the OECD model (see in particular the emphasis on comparability in the extracts below which presupposes looking at the actual characteristics of the enterprises between which provision has been made) and therefore that para 1(2)(a) should be interpreted as requiring consideration of what provision independent enterprises sharing the characteristics of the actual enterprises would have made. 105 On its face this appears to provide some support for the Commissioner's position. But that appearance is not borne out on closer analysis. The Special Commissioners went on to explain that, when there were material differences between the taxpayer and any proposed comparable, such differences should, where possible, result in adjustments and not the exclusion of the comparable. Thus, at [94] the Special Commissioners observed that the OECD guidelines stressed 'the importance of looking for comparable transactions so long as they are comparable, the need to strive to make adjustments to create comparability if at all possible, and that more than one method may be considered'. They then set out paragraphs 1.15 and 1.17 of the OECD guidelines, the relevant portions of which are set out above. 106 What that reveals is not that the OECD guidelines required that the only comparables which might be considered were those in identical circumstances to the taxpayer but rather that those differences which were material should be taken into account through a process of adjustment. For making the adjustments the OECD guidelines suggested the use of a methodology to take account of material differences. In that last regard it is worth recalling the process Mr Seve undertook to reduce all prices to ex-works prices. It was that kind of process which the OECD guidelines contemplated. On no reading did they support the Commissioner's submission that one was required to examine only putative purchasers who were in the same circumstances as the taxpayer. 107 In any event, the argument not only lacks merit on its own terms but is also legally unsound. It rests upon the assumption that the expression 'arm's length consideration' is to be interpreted in light of the guidelines. That requires one to identify what the guidelines are, which, in turn, requires a retracing of steps back to transfer pricing. The problem of transfer pricing is part of a larger family of problems relating to double taxation between nations. Australia is a member of the OECD. In 1977 the OECD promulgated its 'Model Taxation Convention on Income and Capital' (the 'Model Convention'). It has been the longstanding practice of the OECD Council to recommend to its member States that, when concluding bilateral tax conventions on income and capital, they should do so in the form of the Model Convention. It has also been its longstanding recommendation that member States follow the commentaries which have been prepared on the articles of the Model Convention as modified from time to time: see Recommendation of the OECD Council concerning the Model Tax Convention on Income and on Capital, C(97)195/Final (adopted by the Council on 23 October 1997). 108 Article 9(1) of the Model Convention is entitled 'Associated Enterprises'. It provides: Where: a) an enterprise of a Contracting State participates directly or indirectly in the management, control or capital of an enterprise of the other Contracting State, or b) the same persons participate directly or indirectly in the management, control or capital of an enterprise of a Contracting State and an enterprise of the other Contracting State, and in either case conditions are made or imposed between the two enterprises in their commercial or financial relations which differ from those which would be made between independent enterprises, then any profits which would, but for those conditions, have accrued to one of the enterprises, but, by reason of those conditions, have not so accrued, may be included in the profits of that enterprise and taxed accordingly. 109 Article 9(1) attempts to address at a high level of generality the problems thrown up by transfer pricing by providing for pricing as if the transactions had been between independent parties. The precise words of Article 9(2) are not directly relevant to this appeal but, we note, they ensure that a determination by one jurisdiction that profits should be included in a taxpayer's income in that jurisdiction pursuant to Article 9(1) results in a corresponding omission of income in the other jurisdiction, thereby avoiding double taxation. 110 The suppliers in this case were located in France, the US and China. Australia has concluded bilateral tax treaties with each of the three countries and France and the United States are both members of the OECD. In the relevant years, the treaties were: (a) The agreement between the Government of Australia and the Government of the French Republic for the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income done at Canberra on 13 April 1976 as amended by the protocol, done at Paris on 19 June 1989; (b) The agreement between the Government of Australia and the Government of the People's Republic of China for the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income done at Canberra on 17 November 1988; and (c) The agreement between the Government of Australia and the Government of the United States of America for the avoidance of double taxation and the prevention of fiscal evasion with respect to taxes on income done at Sydney on 6 August 1982 as amended by the protocol, done at Canberra on 27 September 2001. 111 Each convention contained an article in precisely the same terms as Article 9 of the Model Convention and we, therefore, do not, need to set them out individually. 112 In Australia, each of the above conventions was and is a separate schedule to the International Tax Agreements Act 1953 (Cth) (the US and Chinese agreements are contained in Schedules 2 and 28 respectively; the French agreement was in Schedule 11). In the case of the Chinese agreement, the International Tax Agreements Act provides that it is to 'to have the force of law according to its tenor' (see, in the relevant year, s 11S). In the case of the French and United States agreements, it is provided that each are to have the force of law in relation to withholding tax and other tax in the French case (see s 9A) and 'tax in respect of: ….(b) income…' in the US case (see s 6). Those expressions are not relevantly different in their operation, at least in relation to this appeal. Both are legislative instructions to treat the relevant schedule as a domestic law in relation to income tax. 113 There is no doubt that as a matter of customary and public international law each of the three treaties is be interpreted in accordance with Article 31 of the Vienna Convention on the Law of Treaties done at Vienna on 23 May 1969. In the case of Australia and China, this conclusion is straightforward for both have acceded to that treaty. In the case of France and the United States, whose accession is not complete, the same conclusion flows because that treaty is 'no more than an indorsement or confirmation of existing practice' (Thiel v Federal Commissioner of Taxation (1990) 171 CLR 338 at 349 per Dawson J) or 'because the interpretation provisions of the Vienna Convention reflect the customary rules for the interpretation of treaties' (Thiel at 356 per McHugh J). Article 31 provides by cll (2) and (3): 2. The context for the purpose of the interpretation of a treaty shall comprise, in addition to the text, including its preamble and annexes: (a) Any agreement relating to the treaty which was made between all the parties in connexion with the conclusion of the treaty. (b) Any instrument which was made by one or more parties in connexion with the conclusion of the treaty and accepted by the other parties as an instrument related to the treaty. 3. There shall be taken into account, together with the context: (a) Any subsequent agreement between the partiers regarding the interpretation of the treaty or the application of its provisions; (b) Any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation; (c) Any relevant rules of international law applicable in the relations between the parties. 114 The context referred to in Article 31(2) includes Article 9 of the Model Law and the commentary which accompanies it: Thiel at 349 per Dawson J and 356-357 per McHugh J, with whom Mason CJ, Brennan and Gaudron JJ agreed at 344; Commissioner of Taxation v Lamesa Holdings BV (1997) 77 FCR 597 at 604. It may also be that they are admissible under Article 32: cf Thiel at 350 per Dawson J. Subject to one matter, however, the official commentary on Article 9 throws no light on the issue at hand. The one exception may be the following passage. This Article deals with adjustments to profits that may be made for tax purposes where transactions have been entered into between associated enterprises (parent and subsidiary companies and companies under common control) on other than arm's length terms. The committee has spent considerable time and effort (and continues to do so) examining the conditions for the application of this Article, its consequences and the various methodologies which may be applied to adjust profits where transactions have been entered into on other than arm's length terms. Its conclusions are set out in the report entitled Transfer Pricing Guidelines for Multinational Enterprises and Tax Administrations, which is periodically updated to reflect the progress of the work of the Committee in this area. That report represents internationally agreed principles and provides guidelines for the application of the arm's length principle of which the Article is the authoritative statement. 115 The 'Committee', for completeness, is the OECD Committee on Fiscal Affairs. The guidelines referred to in this paragraph are the guidelines relied upon by the Commissioner (being the guidelines in place in the relevant years). Paragraph 16 of the preface to the guidelines says: OECD Member countries are encouraged to follow these Guidelines in their domestic transfer pricing practices, and taxpayers are encouraged to follow these Guidelines in evaluating for tax purposes whether their transfer pricing complies with the arm's length principle. Tax administrations are encouraged to take into account the taxpayer's commercial judgement about the application of the arm's length principle in their examination practices and to undertake their analyses of transfer pricing from that perspective. 116 It is plain from that statement, however, that the guidelines are just that - guidelines. Under Article 31(3) they can be examined only if they reflect the subsequent agreement of the States in question or, under Article 31(3)(b), 'any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation'. In order to permit recourse to the guidelines it would therefore be necessary to show that Australia and each of China, the US and France had either agreed to apply the portion of the guidelines relied upon in their performance of the equivalent of Article 9 of each agreement or that it was their practice to do so. 117 There was no evidence that any of the States in question had adopted the practice of applying the guidelines to any of the circumstances in which Article 9 of the Model Law might obtain in their jurisdictions. It must therefore follow that they may not be examined. This conclusion, it should be emphasised, should not foreclose any future attempt to demonstrate that the guidelines do, in fact, evidence State practice. 118 The guidelines then are not a legitimate aid to the construction of the double taxation treaties. The next step in the Commissioner's argument was that they could be used to assist in the interpretation of ss 136AA(3)(d) and 136AD(3) because those sections could be illuminated by the meaning of treaties. The premise for this argument is not made good, not only because the guidelines fail to support his contention, but they are, in any event, not permissible materials for interpreting the treaties. Had it arisen it would have been appropriate to accept in principle that materials that could illuminate Article 9 of the Model Law could also throw light on ss 136AA(3)(d) and 136AD(3). Division 13 of Part III of the 1936 Act is the domestic implementation of Australia's various undertakings embodied in Article 9 of the Model Law. The fact that the provisions are the domestic embodiment of Article 9 of the Model Law does not alter their nature as Australian law and it is in that capacity that this Court is required to interpret them. There is no principle of statutory interpretation which requires domestic legislation of the present kind to be read as if it were itself an international agreement: cf. Russell v Commissioner of Taxation (2011) 190 FCR 449 at 455-456 [25]-[31]. 119 Where, however, the domestic provisions are obscure or unclear then it would be unsound not to attempt to resolve that ambiguity in a way which is consistent with Australia's obligations. There are two routes to that conclusion. One is to say that Division 13 of Part III should be interpreted consistently with the Schedules to the International Tax Agreements Act. This is not a difficult conclusion to draw since the two statutes are to be interpreted as one: see International Tax Agreements Act s 4. The other route is to observe that, since Division 13 is evidently intended to give effect to the equivalent of Article 9 of the various double taxation treaties, it is appropriate to interpret it consistently with Australia's obligations under those agreements. In that last regard, it is crucial to observe that the whole text of each treaty has been given domestic effect. In cases where the exact text of a whole treaty has been given effect by domestic legislation it would be surprising if it were interpreted without keeping that fact in mind. It should be noted that these taxation treaties stand in a very different position to, for example, the Refugee Conventions whose text is not given the force of law. Where Parliament implements a treaty using its expressions and its provisions then naturally enough one must begin with the words Parliament has used. But where Parliament expressly decides to incorporate the whole text of a treaty in domestic law and makes it plain, as here, that it is doing so, then it is appropriate to construe the provisions in accordance with the ordinary principles governing the interpretation of treaties. This is because the Parliament's use of the treaty shows its intention to fulfil its international obligations. This has been accepted by the High Court in respect of the double taxation treaties: Thiel v Federal Commissioner of Taxation (1990) 171 CLR 338. 120 This conclusion is unsurprising. The double taxation treaties are designed to ensure that the taxing regimes of two jurisdictions do not result in double taxation. If they were to be interpreted in a manner which would permit or foster conflicting outcomes between the two States in question their whole point would be frustrated. It is true, as Dowsett J has observed in Russell (at 455-456 [26]-[29]), that the High Court has indicated in the context of the Refugee Conventions that domestic courts must recall that their task is to interpret the Migration Act 1958 (Cth) and not the Conventions. But, unlike the present legislation, that Act does not adopt and apply the whole text of a treaty. When a State decides to implement, as it has in this case, a Model Law, it would be quite inappropriate to disregard that fact when construing the resultant statute. 121 For these reasons, the Commissioner's arguments based on the guidelines should be rejected because they fail at every level. The trial judge described his approach to the question of arm's length consideration this way (at [44]): The essential task is to determine the arm's length consideration in respect of the acquisition. One way to do this is to find truly comparable transactions involving the acquisition of the same or sufficiently similar products in the same or similar circumstances, where those transactions are undertaken at arm's length, or if not taken at arm's length, where suitable adjustment can be made to determine the arm's length consideration that would have taken place if the acquisition was at arm's length. 122 This is a correct statement. It is consistent with the Model Law, the commentary and, if it mattered, the guidelines. 123 The Commissioner also contended that his construction was supported by two other overseas decisions concerning transfer pricing. Each was a decision of the Federal Court of Appeal of Canada: GlaxosmithKline Inc v The Queen [2010] FCA 201; and R v General Electric Capital Canada Inc [2010] FCA 344. Neither case is of assistance, however, because the text of the relevant provision - s 69(2) of the Income Tax Act RSC 1985, c 1 (5th supp) - was materially different. The hypothesis which that provision required expressly involved the taxpayer ('the amount … that would have been reasonable in the circumstances if the non-resident person and the taxpayer had been dealing at arm's length').

V. Miscellaneous issues 124 The Commissioner contended that the taxpayer was obliged to specify a particular figure which was the arm's length consideration for it was only by so doing that it would be able to prove that it had paid less than that consideration. Section 136AD(3) will be enlivened when it is shown, amongst other things, that the taxpayer has paid a price which 'exceeded the arm's length consideration'. Where that is the case, the provision has the effect of creating a statutory fiction to be applied in the realm of income tax. The fiction is thus: the price paid by the taxpayer is taken to be not the price which was paid but instead 'the arm's length consideration'. The 1936 Act accepts the practical reality that this fiction - easy to state and, in principle, easy to grasp - may prove difficult to apply in circumstances where the assessment of the price may be obscure or elusive. It thus provides a further provision - s 136AD(4) - which permits the Commissioner to determine a specified figure to be the arm's length consideration where 'it is not possible or not practicable for the Commissioner to ascertain the arm's length consideration'. That provision was utilised by the Commissioner in relation to this taxpayer. 125 Recourse to a determination of the arm's length consideration under s 136AD(4) is neither a necessary nor inevitable consequence of the operation of the transfer pricing provisions: when what is involved is something regularly traded in a market between arm's length traders there may be little difficulty in ascertaining an arm's length consideration as the Full Court of this Court noted in WR Carpenter Holdings Pty Ltd v Federal Commissioner of Taxation (2007) 161 FCR 1 at 9 [31]. One might say in such cases that a market value may plainly be established so that one can at least identify that value as the arm's length consideration. But that does not mean, more generally, that there is only one arm's length consideration. Often enough, for example, goods will change hands at prices which are different to the market value for perfectly legitimate reasons such as a need to secure long term or large volume arrangements or with traded securities, a premium for control and so on. No doubt, it was for similar reasons that Dr Becker, the Commissioner's own witness, in response to the question '[a]nd you accept that, typically, there's not one arm's length price for a particular product?' answered '[t]hat's correct, yes'. That evidence and these observations demonstrate that, whilst there are parallels between an assessment of market value and an assessment of arm's length consideration, the limitations inherent in those parallels need to be kept steadily in mind. It is likely that a price negotiated at market value will be an arm's length price but it does not follow that prices which are not at market are, simply thereby, not at arm's length. Nor does it follow that there is only one such price. 126 Those observations are, of course, theoretical whereas the present question is the statutory one, whether ss 136AD(3) or 136AD(4) require the identification of a single price. In this case a determination was made under s 136AD(4). That provision has been accepted as operating as an averment: per the Full Federal Court in WR Carpenter at 9[32]. But does this require the taxpayer in review proceedings to identify a particular arm's length consideration? The terms of s 14ZZO of the Taxation Administration Act tend to suggest that the taxpayer's burden is only to show that the assessment is excessive. In WR Carpenter, however, the Full Federal Court said this (at 10 [37]): The only difference may be this: in order to show that an assessment made in reliance on determinations made under para (d) of s 136AD(1) or s 136AD(2) and s 136AD(4) is excessive, it would be necessary for the applicants to show that the arm's length consideration is both ascertainable and less than the deemed amount; that, in itself, would seem to require the applicants to prove the actual amount of the arm's length consideration. In the course of doing so they would necessarily establish that the arm's length consideration was ascertainable. Whereas, in order to show an assessment made in reliance on a determination made under para (d) of s 136AD(1) or s 136AD(2) when the Commissioner has ascertained the arm's length consideration otherwise than in reliance on a determination under s 136AD(4), it may only be necessary for the applicants to prove that the correct arm's length consideration is less than the Commissioner's ascertainment of it; it may not be necessary to prove the actual amount of the arm's length consideration, that is, that it is ascertainable. 127 This is an obiter dictum. It would, for the reasons just given, produce grave difficulties to proceed on the basis that there was a single arm's length consideration. That should only be done if the text of ss 136AA(3)(d), 136AD(3) and (4) intractably requires it. Is the text that inflexible? The references to 'arm's length consideration' in s 136AD all lead back to s 136AA(3)(d) and its definition of that expression as meaning 'the consideration that might reasonably be expected to have been given or agreed to be given in respect of the acquisition if the property had been acquired under an agreement between independent parties dealing at arm's length with each other in relation to the acquisition'. It is true that the word 'the' gives some support for the idea that there is but one arm's length consideration. Whatever force that may have is, however, overcome by the content of the concept itself which is incompatible with such a confined reading. 128 Of course, it is true that s 136AD operates to engage the statutory fiction that the consideration paid was 'the arm's length consideration' and that the existence of more than one such value may give rise to practical problems of administration. But it is precisely to situations of that kind that the Commissioner's power conferred by s 136AD(4), to determine the arm's length consideration, is apposite: 'where, for any reason (including an insufficiency of information available to the Commissioner), it is not possible or not practicable for the Commissioner to ascertain the arm's length consideration… [it] shall be deemed to be such amount as the Commissioner determines'. Where there is more than one arm's length price (as often there will be), the Commissioner may determine which he will apply. Correspondingly, in review proceedings the taxpayer will be entitled to succeed if it shows that the prices paid by it were arm's length prices or less than arm's length prices. If it pursues the latter course there is no need for it to establish a particular price as the arm's length price. It will be sufficient to show that it paid less than an arm's length price. 129 In this case, that is what the taxpayer did. It proved that the prices paid by it were less than the prices paid by independent comparable purchasers. Those prices were arm's length; the taxpayer's prices did not therefore exceed arm's length consideration. 130 On another topic, the Commissioner submitted that the taxpayer was able to continue to trade, not because of the alleged price support, but because of an injection over the period of $31.2 million in share capital from its parent. It was said that an independent distributer would not have continued to trade at a loss for 13 years to further a market penetration strategy for the benefit of an unrelated supplier. It is unclear, legally, where this submission fits in the appeal. But that is of no moment for it exposes a fundamental problem in the Commissioner's case. The unstated conclusion is that the losses were being generated by the excessive prices paid by the taxpayer for its polyacrylamides. But the trial judge found as a fact that this was not so; and rather that the losses had been caused by unreasonably low levels of sales, by competition in the Australian market, by excessive stock losses and by poor management. Ground 6(d) of the notice of appeal attacked that conclusion but no arguments in its favour were advanced by the Commissioner either in writing or orally. The taxpayer noted that failure in its submissions. The ground was not pressed in this Court. 131 Finally, it is necessary to deal with some rulings on evidence. The trial judge held in his reasons for judgment that certain parts of the evidence of Mr Karoudjian and Mr Schroeter were inadmissible hearsay. Those parts comprised: (a) evidence of the prices charged by various companies (other than Nalco in the case of Mr Schroeter) to which objection has been taken; (b) any statements which had specifically been made on the basis of information supplied by another person; (c) any conclusionary statements as to comparative markets, the independence of customers, similarity of product or description of market; and (d) any statements concerning the status of customers, as to whether or not they were end users or distributors of similar products. 132 The trial judge did not identify which specific paragraphs of the evidence were affected by that ruling. The parties are in disagreement as to the effect of the ruling on a number of paragraphs of evidence. The paragraphs, the objection and the appropriate ruling are as follows: Paragraph Objection Ruling Mr Schroeter's affidavit of 11 July 2008 18 'for the water-soluble polymer market' Conclusionary hearsay Admitted; words not objected to at trial 22, 28, 31, 32 Conclusionary hearsay Admitted; words not objected to at trial 37 Conclusionary hearsay Rejected 41 Conclusionary hearsay Admitted; witness giving direct evidence 42 Hearsay Admitted; direct evidence of SNF prices 82 Hearsay Admitted

Paragraph Objection Ruling Mr Schroeter's affidavit of 9 April 2009 23 Hearsay Admitted

Paragraph Objection Ruling Mr Karoudjian's affidavit of 22 July 2008 7 (Chapeau) Hearsay Rejected 7E Hearsay Admitted; not objected to at trial 8 Hearsay Reject the words 'and they were distributors… …during the audit period'

133 It should be added that the method of dealing with these objections which the trial judge adopted left the status of the evidentiary record unclear. Generally speaking, this is not an appropriate way to deal with objections. One of the important functions of a trial judge is to superintend the evidence. This facilitates the resolution of appeals. As a general rule, an appellate court should not be required to deal with rulings on evidence. In some circumstances the approach adopted by the trial judge may be appropriate. A case in which one of the Commissioner's chief strategies was the vigorous pursuit of a 'no evidence' contention was, however, not such a circumstance.

VI. Disposition 134 The appeal should be dismissed with costs. I certify that the preceding one hundred and thirty-four (134) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justices Ryan, Jessup & Perram.

Associate: Dated: 1 June 2011 Appendix 1 Income Tax Assessment Act 1936 (Cth) 136AA Interpretation … (3) In this Division, unless the contrary intention appears: … (d) a reference to the arm's length consideration in respect of the acquisition of property is a reference to the consideration that might reasonably be expected to have been given or agreed to be given in respect of the acquisition if the property had been acquired under an agreement between independent parties dealing at arm's length with each other in relation to the acquisition; and … 136AD Arm's length consideration deemed to be received or given .. (3) Where: (a) a taxpayer has acquired property under an international agreement; (b) the Commissioner, having regard to any connection between any 2 or more of the parties to the agreement or to any other relevant circumstances, is satisfied that the parties to the agreement, or any 2 or more of those parties, were not dealing at arm's length with each other in relation to the acquisition; (c) the taxpayer gave or agreed to give consideration in respect of the acquisition and the amount of that consideration exceeded the arm's length consideration in respect of the acquisition; and (d) the Commissioner determines that this subsection should apply in relation to the taxpayer in relation to the acquisition; then, for all purposes of the application of this Act in relation to the taxpayer, consideration equal to the arm's length consideration in respect of the acquisition shall be deemed to be the consideration given or agreed to be given by the taxpayer in respect of the acquisition. (4) For the purposes of this section, where, for any reason (including an insufficiency of information available to the Commissioner), it is not possible or not practicable for the Commissioner to ascertain the arm's length consideration in respect of the supply or acquisition of property, the arm's length consideration in respect of the supply or acquisition shall be deemed to be such amount as the Commissioner determines. Taxation Administration Act 1953 (Cth) 14ZZO Grounds of objection and burden of proof In proceedings on an appeal under section 14ZZ to the Federal Court against an objection decision: (a) the appellant is, unless the Court orders otherwise, limited to the grounds stated in the taxation objection to which the decision relates; and (b) the appellant has the burden of proving that: (i) if the taxation decision concerned is an assessment (other than a franking assessment) - the assessment is excessive; or (ii) if the taxation decision concerned is a franking assessment - the assessment is incorrect; or (iii) in any other case - the taxation decision should not have been made or should have been made differently. Appendix 2 Year Tax Penalty Interest Total 1998 16,304.04 4,076.01 21,534.78 41,914.83 1999 132,096.24 33,024.06 158,350.94 323,471.24 2000 551,364.84 137,841.21 517,456.66 1,206,862.71 2001 No assessment issued 2002 17,436.30 4,359.05 21,795.35 2003 120,930.90 30,232.70 151,163.60 2004 226,869.00 56,717.25 283,586.25 TOTAL $2,028,793.98

Appendix 3 Unit Price (€)1 Code Art Client 1997 1998 1999 2000 2001 2002 2003 Avg AA30A00 SNF AUST 1.39 1.22 1.48 1.38 1.37 1.43 1.49 1.40 RIBCO 1.79 1.78 1.74 1.75 1.88 1.85 1.91 1.81 CB30B40 SNF AUST 0.90 0.91 0.99 1.04 1.03 1.02 0.99 BAYER AL 1.54 1.59 1.59 1.59 1.62 1.62 1.59 BI25B00 SNF AUST 2.13 2.05 2.26 2.39 2.34 2.35 2.15 2.17 RIBCO 2.78 2.57 2.55 2.55 2.53 2.53 2.56 2.57 KEMPRO 2.66 2.61 2.58 2.09 2.20 2.21 2.24 2.31 AA20A00 SNF AUST 1.39 1.22 1.34 1.37 1.39 1.44 1.36 1.37 NALCO IT 2.35 2.35 1.98 1.98 1.97 1.88 1.65 1.96 BETZ FR 1.82 1.75 1.79 1.76 1.86 1.77 AA00A00 SNF AUST 1.79 1.30 1.46 1.45 1.39 1.48 1.99 1.42 BTC 1.75 1.71 1.75 1.86 1.85 1.82 MY00Z61 SNF AUST 0.46 0.49 0.49 BASF FIN 0.49 0.48 BI55A00 SNF AUST 2.08 2.07 2.28 2.32 2.27 2.14 2.18 2.20 AQUAPLAN 2.77 2.78 2.75 2.77 2.82 2.80 2.77 2.78 ACIDEKA 3.34 3.29 3.28 3.28 2.29 2.18 2.39 2.24 NALCO MB 2.91 2.81 2.64 2.44 2.85 2.67 2.53 3.09 NEROLAN 2.52 2.52 2.68 BI25A00 SNF AUST 2.11 2.07 2.33 2.33 2.26 2.28 2.22 2.21 BTC 2.37 2.35 2.25 2.27 2.38 2.46 2.17 2.29 CHATZOPO 2.35 2.35 2.35 2.35 2.34 2.34 2.21 2.34 CHEMIPOL 2.48 2.53 2.52 2.39 2.39 2.33 2.49 NORDMAN 1.96 2.01 1.98 2.21 2.02 AA30B00 SNF AUST 1.39 1.25 1.42 1.47 1.39 1.49 1.87 1.41 CLARFLOK 1.89 1.89 1.87 1.88 1.97 1.87 1.89 BI15B00 SNF AUST 3.04 2.94 2.76 2.94 2.28 2.35 1.99 2.30 NALCO MB 2.56 2.29 2.69 BI45A00 SNF AUST 2.13 1.95 2.25 2.35 2.27 2.31 2.15 2.24 BIOCONSU 3.24 2.90 0.64 1.67 1.67 2.51 2.38 1.68 NALCO MB 2.87 3.17 3.02 3.09 2.61 2.43 2.16 2.89 NEROLAN 2.61 2.79 2.26 2.49 2.45 2.99 2.99 2.50 AQUAPLAN 2.69 2.66 2.64 2.81 2.99 2.37 2.56 2.70 HENKE RF 2.63 2.47 2.50 2.50 2.07 1.72 2.48 KEMIR RF 2.48 2.48 1.51 2.30 AA20A00 SNF AUST 1.39 1.22 1.34 1.37 1.39 1.44 1.36 1.37 GE BETZ 1.86 1.82 1.80 1.76 1.85 1.81 DJ60B47 SNF AUST 1.81 1.71 1.91 1.61 1.41 1.46 1.40 1.56 GE BETZ 1.79 1.77 1.69 1.77 1.71 1.57 1.72 BI45A00 SNF AUST 2.13 1.95 2.25 2.35 2.27 2.31 2.15 2.24 GE BETZ 2.89 2.91 2.81 2.76 2.83 2.75 2.57 2.76 DJ30B42 SNF AUST 1.43 1.58 1.28 1.24 1.30 1.54 1.33 ASHLAND 1.65 1.65 1.65 1.65 1.65 1.63 BI55B00 SNF AUST 2.13 2.20 2.32 2.33 2.33 2.30 2.24 2.26 ASHLAND 2.93 2.86 2.78 2.78 2.91 2.97 2.57 2.82 BI10B00 SNF AUST 2.13 1.80 2.23 2.27 2.38 2.12 2.10 BUCKMAN2 2.30 3.09 2.77 2.29 2.14 2.08 2.77 HERCULES 2.14 2.11 2.36 NOBEL 2.14 BI20B00 P SNF AUST 2.13 2.19 2.54 2.50 2.64 2.52 2.41 2.16 GE BETZ 2.63 2.63 2.57 AA10A00 SNF AUST 1.39 1.31 1.49 1.47 1.48 1.49 2.16 1.46 NOBEL 2.23 2.21 2.21 BI10A00 SNF AUST 2.13 1.85 2.11 2.12 2.10 2.07 2.05 2.04 BUCKMAN2 2.80 3.04 3.15 2.86 BI40B00 SNF AUST 2.13 2.19 2.29 2.38 2.30 2.27 BUCKMAN2 3.26 2.91 3.14 2.92 BI20A00 SNF AUST 2.17 2.21 2.31 2.34 2.22 2.42 2.28 BUCKMAN2 3.26 3.26 2.91 2.91 BI20B00 SNF AUST 2.27 2.19 2.14 2.20 NOBEL 2.67 2.57 2.64

1 Represents average ex-works unit price by number of transactions 2 Unit prices for Buckman have been adjusted to include off invoice rebates provided. These rebates were as follows: 1999: 10.47%, 2000: 3.56%, 2001: 5.25%, 2002: 2.23%, 2003: 3.75% The maximum rebate of 10.47% has been applied to calculate the average unit price for the 1997 to 2003 period