Case note
When the tax office may rewrite the bargain, and when it may only reprice it
Glencore Investment Pty Ltd v Commissioner of Taxation [2019] FCA 1432; Commissioner of Taxation v Glencore Investment Pty Ltd (2019) 272 FCR 30; (2020) 281 FCR 219; [2020] FCAFC 187
The central question in transfer pricing is whether the tax office may price the bargain the parties actually made, or may substitute a different bargain. This is the case usually cited for the proposition that it may not. The Full Court held the opposite on the point of principle, and the taxpayer still won on the facts.
- Court
- Federal Court of Australia, then the Full Court
- Bench
- Davies J at first instance; Middleton, Steward and Thawley JJ on appeal
- Decided
- 3 September 2019; 6 November 2020
- Outcome
- Assessments held excessive at first instance. On appeal, allowed in part: the Commissioner won the point of principle and the 2009 freight issue, and lost the case
The facts
The CSA mine near Cobar in New South Wales is a high grade underground copper mine and, by world standards, a small and expensive one: in 2006 it produced about 0.2 per cent of world copper concentrate and sat near the ninetieth percentile of direct cash costs. Its cost base made it vulnerable to swings in revenue: [17] to [19].
The Australian mining company sold all of its concentrate to a Swiss group company under life of mine offtake agreements. A whole of production offtake guarantees the miner a home for everything it produces and transfers risk to the buyer, which the Commissioner's own expert accepted matters: a producer without a home for its production has a serious problem: [70], [71].
Copper concentrate has no terminal market of its own, so it is priced off the refined copper price, less treatment and refining charges. Those charges can be set against an annual industry benchmark, against the volatile spot market, or by price sharing, which fixes them as a percentage of the copper price for a term of years. Price sharing is not a different rate but a different mechanism: it removes the risk that the charges and the copper price move in opposite directions: [72], [73], [321].
In February 2007 the parties amended their agreement for three years. The charges became a flat 23 per cent of the copper price, replacing a half benchmark and half spot arrangement; the buyer gained wider choice over the pricing period; the required copper content was reduced; and, for 2009 only, the freight allowance was fixed at US$60 per tonne, which was the rate for shipping to India and considerably more than the rates to China, Korea or Japan: [28] to [33]. Because the formula was fixed for three years, the buyer also lost its right to terminate during that period: [30].
Division 13 of the Income Tax Assessment Act 1936 (Cth) and Subdivision 815-A of the Income Tax Assessment Act 1997 (Cth), read with the associated enterprises article of the Australia and Switzerland agreement. Both have since been replaced by Subdivision 815-B.
What was in dispute
The Commissioner's case was that the February 2007 amendment was not one independent parties would have made, and that the dealings should be priced as though the parties had made a market-related contract instead. The question of principle was whether he was entitled to do that, or whether he was confined to pricing the contract in front of him.
What the Court decided
At first instance Davies J held he was confined. Reading the authorities through the 1995 OECD Guidelines, her Honour held that the comparison must ordinarily be based on the transaction as the parties structured it, subject to two narrow exceptions, being a case where economic substance differs from legal form and a case where the structure is commercially irrational and also practically impedes the tax administration from determining a transfer price: [40], [315]. Neither applied. Price sharing was a recognised and legitimate way of pricing copper concentrate, not merely a component of a price, and there was no suggestion that tax rather than ordinary commercial considerations had shaped the terms: [319] to [321]. The conclusion is the sentence the case is known for: there was no warrant to restructure the agreement from a price sharing contract into a market-related one: [322].
The Full Court disagreed with that reasoning, and this is the part most often left out.
Where parties do not specify an actual price, but rather agree that the consideration payable is to be determined by a formula or some other methodology, the Commissioner, in our view, is also permitted — in the sense of having a legal capacity — to substitute a different formula or a different methodology which he considers will result in the ascertainment of the arm's length consideration.
Middleton and Steward JJ at [154].
There is thus no power or authority to substitute different terms of a contract where those terms are not seen as defining the consideration received, relevantly, for the supply of goods. Whether a term of a contract is to be characterised as such a term will need to be decided on a case by case basis.
Middleton and Steward JJ at [155].
So there is a power to substitute, and it has a boundary. The boundary is textual: because the provision works by deeming, and a deeming provision is read strictly, the word "consideration" directs attention to the clauses that define the price. Clauses that merely bear on price indirectly are outside it, and whether the Commissioner may reframe those must, their Honours said, be very seriously doubted: [155]. On these facts the price sharing clause and the pricing period clause were part of the price formula, so the power was available: [158], [159]. Their Honours were also frank about the Guidelines, describing the language of the exceptions as highly generalised and frustratingly opaque, and the Guidelines themselves as only a guide: [153].
Thawley J agreed in the result and rejected the primary judge's rule more squarely still. The constraint, in his Honour's view, is not a closed list of exceptions but reliability: the further the hypothetical agreement departs from the actual one, the less it proves about the arm's length consideration for what was supplied. It is one thing to substitute terms independent parties would not have agreed, and another to ignore terms they would have: [262].
Having won the principle, the Commissioner lost the case. The Full Court held he had asked the wrong question. His case compared the old contract with the new one and asked whether an independent party would have agreed to the amendments. The statutory question is whether the consideration actually received in the relevant years was less than an arm's length consideration for the concentrate in fact supplied, and that question did not feature in his analysis: [166], [188].
On the merits, risk management was held to be part of price rather than separate from it, and an arm's length deal is not always perfect and does not always favour one side to the maximum extent: [162]. A more conservative miner might rationally trade forecast earnings for certainty, which here included three years' immunity from the buyer's right to terminate: [195] to [197]. The 23 per cent rate sat within an unchallenged industry range and the Commissioner led no evidence from anyone as to what the rate should have been: [194].
The Commissioner did win one issue. For the 2009 year the freight allowance had been set at the India rate although almost no shipments had gone to India in the two preceding years, and the taxpayer led no evidence explaining why, or that US$60 was an arm's length figure. On that issue it had not discharged its onus: [234] to [237].
Why the decision matters
The case is the leading Australian authority on the limits of reconstruction, and it is routinely cited for the proposition that the tax office must take the transaction as it finds it. Read as a whole it does not establish that. What it establishes is that where the parties have agreed a formula rather than a figure, the Commissioner may substitute a different formula, and that the limit on the power lies in identifying which clauses define the price.
The reason the taxpayer nonetheless won is a lesson in itself, and it is about the question rather than the answer. A case built on whether an amendment was one an independent party would have agreed to is not the statutory question. The statutory question is about the consideration received for what was supplied.
Two propositions are worth carrying beyond the case. Risk allocation is part of price, so a party that gives up forecast upside to remove volatility has not necessarily struck a bargain outside the arm's length range. And transfer pricing is not an exact science: the Court warned against making compliance an impossible burden merely because a revenue authority can find someone, years later, who would have priced the deal differently: [203], [213].
The freight finding is the counterweight. Where a term is priced by reference to a basis that the facts do not support, and the taxpayer offers no explanation and no expert evidence, the onus decides it. The same taxpayer won the large issues on evidence and lost the small one for want of it.
Where it stands
The decisions stand. Nothing in the citator material read reverses, overrules or doubts them, and they have been applied repeatedly since, including in Singapore Telecom Australia Investments Pty Ltd v Commissioner of Taxation (2024) 302 FCR 192.
It was the Commissioner who sought special leave to appeal to the High Court, not the taxpayer. The material read for this note does not record the outcome of that application in either of the citators consulted, and no statement about it is made here.
The statutory position is the most important qualification and it cuts against the taxpayer. The primary judge's two exceptions came from the 1995 OECD Guidelines, and the second of them required that the structure practically impede the tax administration from determining a transfer price. Section 815-130 of the Income Tax Assessment Act 1997 (Cth), which now contains the reconstruction power, contains no such requirement and no element of tax purpose. The Full Court's own limit, confining substitution to clauses that define the consideration, was drawn from the text of the repealed Division and has no evident textual home in the current provision either.
The consequence is that this case is a weaker constraint on reconstruction than its reputation suggests, and weaker again under the provisions now in force. It should not be relied on as a shield without checking the current text.
A case note by Boettcher Law. It is general information about Australian law and a summary of two published decisions, and it is not legal advice on any particular arrangement. Paragraph references are to the Full Court's reasons unless the text says otherwise.