Transfer Pricing Emerges From the Shadows

Printer Friendly Version

Over the last 15 years there has been a noticeable discrepancy between word and deed.  On the one hand, the Australian Taxation Office (“the ATO”) has devoted enormous amounts of time and effort to transfer pricing – releasing Rulings and booklets, delivering speeches, making announcements and generally trying to get taxpayers’ attention.  The work in Australia was matched by an equally voluminous output from the OECD, not to mention the extensive work of the tax authorities of other countries, especially the US.

On the other hand, very few transfer pricing cases have made it to the Australian courts, and those that have been litigated recently – Carpenter, San Remo, Syngenta or Daihatsu for example – have all involved procedural skirmishes.  The decision of Justice Downes in Roche Products in the Administrative Appeals Tribunal is, finally, a more visible and more important result for all this ATO effort.

The Roche Products decision is significant because it is the first Australian case in 45 years on the substantive issue which lies at the heart of all transfer pricing disputes – the appropriate method (or methods) to be used to determine arm’s length prices.  But the judgment also demonstrates why so few transfer pricing cases are litigated – the Tribunal had a great deal of difficulty in resolving that issue in the face of the competing opinions of four experts, each of whom produced a range of prices using slightly different information, assumptions and adjustments.

Evidence in the case suggests that the matter came to be litigated because of the ATO’s impression that the Australian subsidiary of the Swiss Roche pharmaceutical group was insufficiently profitable – that the group’s profit margin worldwide from the sale of branded prescription drugs was 75%, but in Australia it was only 36%. Not surprisingly, the ATO decided to look harder.

Facts

The Commissioner assessed Roche Products, the Australian subsidiary of the Roche group, to further tax for 11 years.  (There is no discussion of penalties in the case.)  The amended assessments were based on the ATO’s view that Roche Products had overpaid other Roche group companies located in Switzerland and Singapore for purchases of:

• branded prescription drugs – for example, Valium, Mogadon and Rohypnol.  Most were patented although some patents had expired.  These drugs were usually imported already manufactured but were packaged in Australia;
• non-prescription medicines – for example, Aspro and Berocca; and
• medical equipment and consumables used in pathology and other diagnostic procedures.

The judge found the ATO’s amended assessments for purchases of non-prescription medicines and medical equipment excessive, but increased the taxpayer’s assessable income from the sale of prescription drugs by over $58.7 million.  The taxpayer may, however, feel that it largely won the case. On many issues its expert witnesses were clearly preferred by the judge over the ATO’s experts and it won in all areas in dispute but one.  Moreover, while the adjustment may seem large, it relates to 11 income years and is less than $5.5 million a year on average, although penalties and interest will likely be substantial. The $58.7m figure which the judge arrived at – it was not taken from the reports of any of the experts – just happened to about half of the Commissioner’s amended assessment.  This rather unsatisfying split-the-difference outcome is often seen in US transfer pricing cases.

Argument and reasons

It was conceded that Roche Products was not purchasing at arm’s length from the offshore Roche group companies (in the sense that the parties were not independent) and so the only issue in the case was the determination of the arm’s length price.  Much of the evidence in the case (and much of the judgment) revolved around the differing approaches and opinions of the US economists engaged by the parties.

In trying to determine the arm’s length price, the Tribunal considered the meaning, applicability and consequences of the pricing methods discussed in the OECD’s Transfer Pricing Guidelines: the price in an uncontrolled transaction, subtracting an appropriate margin from the uncontrolled resale price, adding an appropriate mark-up to the seller’s input costs and profit methods.  These methods form the basis of the ATO’s 70-page Ruling on transfer pricing methods, TR 97/20, and were used by the various experts on either side.

With regard to the branded and usually patented prescription drugs, all of the experts and the Tribunal faced the common problem which the judge described at the outset of his opinion, that ‘pharmaceutical companies rarely sell their products through third parties [so that] there is generally no free market in which the products in question are sold [nor even a market in] potentially comparable products.’

Nevertheless, the judge observed that ‘the way to evaluate transfer prices is to look at comparable prices at arm’s length, rather than comparing different aspects of the subject’s business to a range of other businesses,’ the approach which had been taken by one of the Commissioner’s experts.  Because of this, the judge largely rejected the opinion of one of the Commissioner’s experts who had undertaken an analysis of the profits of companies which undertook similar activities to Roche Products – a profit-based comparison – rather than an analysis of sales by the Roche group of comparable products in uncontrolled transactions. (It seems that the expert had proceeded in this fashion because her report was prepared at a time before any comparable sales had been identified.)

So, much of the judgment considers evidence of actual transactions undertaken by members of the Roche group, trying to isolate those sales that could be regarded as sufficiently ‘comparable:’

• the company argued that the comparison should include sales by the Roche group of generic drugs to independent drug wholesalers in Australia; the Commissioner argued that the only comparable transaction was sales by the Roche group of just one newly-patented drug to an unrelated drug company in Australia;
• the Commissioner argued that the comparison should exclude the sale of one drug because the sales volume was low; the company argued it should not be excluded. (The evidence showed that Roche Products was actually paying more for this drug than one of the independent drug wholesalers in Australia!)

Similarly, other parts of the judgment considered what adjustments should be made to the prices of actual transactions to make them more ‘comparable:’

• there was evidence that the Roche group’s terms of trade with subsidiaries differed from the terms for sales to third parties;
• generic drugs were sold by the kilo while patented drugs were sold in blister packs.

The judge took the view that all sales should be considered unless the sale was unrepresentative, and appeared generally content with the way that two of the experts made adjustments to account for marketing differences.  In the end, he increased the taxpayer’s income from the sale of prescription drugs by $58.7 m, on the basis that Roche Australia should have reported its taxable income based on a gross profit margin of 40%, rather than the 31-40% it had been reporting over the period.

With regard to the sale of non-prescription medicines and medical equipment, both sides agreed that there were no comparable transactions and so a profit based method was to be used.  But the judge did not disturb the taxpayer’s prices in either case.  He observed that the low profit appeared largely due to operating costs, particularly the significant (and in some cases unproductive) advertising expenses incurred in Australia, rather than the price charged by the offshore Roche companies – according to the judge, one drug ‘would still have been a disaster if it had been given to Roche Australia.’

What is important about the case?

The case may appear long and difficult but by the standards of overseas transfer pricing cases, it is fairly standard or even below average.  The last significant Australian transfer pricing decision in 1963 on the oil industry was considerably longer.  Along the way and amidst all the detail, several important points emerge.

The first is the strong and clear preference that the judge expressed for using a comparable uncontrolled price to determine the appropriate transfer price.  As the judge put it, the task is ‘identifying the comparable sales and making appropriate adjustments.’ It is noteworthy that he referred to general valuation principles to be applied in determining market prices as well as the specialist and voluminous material relating to transfer pricing. This is a healthy corrective to the view that there is something particularly special and arcane about transfer pricing in the tax area.

Secondly, the range of comparables should extend widely.  As the judge put it, ‘all arm’s length transactions are included’ unless they are ‘anomalous or atypical …’ The judge was unwilling to exclude either the sales of generic drugs (as the Commissioner had wanted) or the single patented drug (as the company had wanted).

Thirdly, the judge was unpersuaded by the mere fact of the apparently low profitability of the Australian subsidiary compared to the position of the worldwide group.  The judge commented that it was the Swiss parent which owned, and was entitled to a profit from, the intellectual property in the various drugs.

Fourthly, the analysis of the sales of non-prescription medicines and medical equipment shows that the judge was not prepared to allow a simple and uncritical application of profit methodologies. He focussed on the significance of the operating costs in reducing the profit of the Australian subsidiary and cautioned that ‘one of the problems of profit based methodology is that … it inevitably attributes any loss to the pricing’ a conclusion he was not prepared to accept.  Of the use of the profit methods by one expert he said it required ‘multiple subjective determinations which admit of error at every step.’

Fifthly, it seems clear that the judge was slightly frustrated by the case.  The various experts each had access to different information, they elaborated their initial positions when a competing report appeared from the other side and, according to the judge, ‘their approach to the issues before me must have been coloured by their United States experience.’  For example, the US experts constantly used the inter-quartile range (which eliminates the data in the bottom 25% and top 25% of the sample as a means of removing outliers) which is sanctioned under US regulations but which has no equivalent in Australian law. The judge also appeared concerned by the parties treating the OECD’s Transfer Pricing Guidelines as if they were the law to be applied rather than the provisions of the Income Tax Assessment Act and the two tax treaties involved. The judge remarks on more than one occasion that Australian law requires (in this instance) ascertaining an appropriate purchase price for property and this should be the focus of the inquiry (and not, as one of the experts asserted, whether an arm’s length owner would consider the Australian division as a whole sufficiently profitable).  And the judge was clearly unimpressed when the experts resorted to anecdotes and their understanding of the lore of the pharmaceutical industry to justify particular opinions.

Finally, the judge referred to the old chestnut (from an Australian viewpoint) about whether tax treaties confer an independent power to raise transfer pricing adjustments. He expressed the tentative view that they did not but, as just noted, he commented on how little interest the parties took in the law. It is not clear why the ATO (and taxpayers) regard this issue as so important. There are differences at the margins between the provisions in tax treaties and the domestic law on transfer pricing where the issue may be important, but in most cases, as in this one, at the end of the day it is unlikely to matter. He did note that, as the Tribunal stood in the shoes of the Commissioner, he did have access to a specific power in the tax legislation to do the best possible where the evidence available makes it difficult actually to determine the arm’s length price.

The future

Although each side had a partial win before the Tribunal, it remains to be seen whether this decision will be the end of this case.  The reasons handed down on 2 April were tentative – the judge referred to them as published ‘on a preliminary basis.’ We understand that both parties have some reservations and may yet bring the matter back to the Tribunal for further argument before the final decision is issued.

Even if the final decision emerges in the same form, an appeal may well follow.  The hearing before the Tribunal lasted 12 days and involved two senior counsel, two juniors and two experts on each side, as well as 20 others assisting them.  Now that all the evidence has been collated, prepared and led, an appeal on issues of law may be a modest further investment of time and money for either side.

On the other hand, an appeal only lies from a decision of the Administrative Appeals tribunal on a question of law and, as the judge notes more than once, little law was involved. Taxpayers go to the Tribunal in cases of this kind precisely because it has all the powers of the Commissioner and the issues are essentially factual – the same avenue was chosen in the last case 45 years ago, and with three senior counsel involved. Losing in the battle of experts is unlikely to involve questions of law.

More broadly, the major issue is whether this decision is indicative of how transfer pricing will play out in courts in Australia. In the US where there has been much more litigation, the IRS has largely lost many of the big cases. It is often reported in the US that, as a result, the IRS usually settles for a relatively modest proportion of the original assessment though of course there are exceptions like the recent US$3.4 billion settlement in another pharmaceutical case. It may be that the most visible legacy of this case will be found in the impact it has in the way the ATO makes its original transfer pricing adjustments and then in its settlement practices.


For further information, please contact:

Sydney

Andrew Mills
(61 2) 9225 5965
Andrew_Mills@gf.com.au

Melbourne

Richard Buchanan
(61 3) 9288 1903
Richard_Buchanan@gf.com.au