Rio Tinto Doesn't Set Its Prices. The Whole Company Is a Bet on One Number It Can't Control.

In 2024, Rio Tinto's iron ore business threw off $16.2 billion in underlying EBITDA - more than its copper and aluminium arms combined, several times over.² The company shipped 328 million tonnes of Pilbara ore² and posted net earnings of $11.6 billion, up 15% on the year.¹ A triumph of operational mastery? Read one line further and the story changes: that 15% jump came in a year when the iron ore price fell 11%.¹ Earnings rose while the only thing that mattered fell. The reason is not management. It is the gap between what Rio digs up and what it gets paid for it - a gap that opens and closes on a number set thousands of miles from any boardroom.

The official story is that Rio Tinto is one of the world's great mining operators, and that its profits reflect the discipline of running the lowest-cost ore in the business. That's half true and entirely beside the point. Rio is supremely good at getting ore out of the ground cheaply. But it has almost no say over what that ore sells for - and the selling price, not the cost, is what turns a year into a triumph or a loss.

The price is decided in Qingdao, not the boardroom

Here is the trap a commodity producer cannot dig its way out of. Rio mines a substance that is, by definition, interchangeable - a tonne of 62% iron fines is a tonne of 62% iron fines, whoever's name is on the ship. So no buyer pays Rio a premium for being Rio. The price is the spot index, and the spot index is set by global supply against Chinese steel demand. In Q4 2024 that benchmark - 62% Fe fines, delivered to Qingdao - averaged $103.40 a tonne, down from $128.25 a year earlier, with analysts forecasting oversupply to keep it pressed down through 2025.⁸ Rio can control how cheaply it produces a tonne. It cannot control the single number that decides what the tonne is worth. That is what it means to be a price taker: you bring the mine, the world brings the price, and the world is in a worse mood some years than others.

The concentration is the danger. A diversified miner can let a weak metal lean on a strong one. Rio is diversified on paper - copper, aluminium, the rest - but in the numbers it is an iron ore company with side businesses. When the iron ore price is high, the whole company glows; when it falls, there is nothing big enough to catch it. The 2024 results show the buffer working in the producer's favour. The next exhibit shows what happens when it runs out.

2015: the year the buffer ran out

By its own account, Rio watched the iron ore price drop below US$40 a tonne in 2015 - what its results statement called 'an 80% fall from the peak of the market in 2011.'⁴ That is the price-taker's nightmare made literal: the product unchanged, the mines as efficient as ever, and four-fifths of the headline value simply gone, decided by demand that moderated and supply that didn't. Iron ore net earnings fell 51%, a $4.2 billion drop, to just under $4 billion.⁴ The whole company tipped into a net loss of US$866 million.⁴ In its half-year deck, Rio said the quiet part plainly.⁵

Notice what that sentence does not say. It does not say 'we ran the mines badly' or 'our costs blew out.' It says prices are under pressure - the passive voice of a company describing weather it cannot change. There is a subtlety worth keeping honest, though. The 80% figure is the spot index, peak to trough. Rio's own realised price held up better than that, because its contracts lag the market: in 2014 it achieved an average of $84.3 per wet metric tonne, with about a quarter of sales priced off the prior quarter's index lagged a month rather than the live spot.³ So the contract structure and rising volumes softened the blow. The collapse Rio actually felt was real but less violent than the index headline. The price taker has a little shock absorber. It does not have a brake.

The same trap, wearing aluminium's clothes

If the cycle only bit through iron ore, you could call it bad luck. But Rio walked into the same trap with a different metal, and paid far more for the lesson. In 2007 it bought Alcan for $38.1 billion in equity - roughly $44 billion including assumed debt - near the top of the aluminium cycle.⁷ By January 2013 the cumulative writedowns on that deal had reached around $30 billion, close to 80% of the purchase price.⁷ On 17 January 2013 the company announced a single non-cash impairment of about $14 billion post-tax - some $10 to $11 billion of it on the aluminium assets, mostly Rio Tinto Alcan, plus roughly $3 billion on a Mozambique coal business and smaller write-downs.⁶ The CEO, Tom Albanese, stepped down the same day.⁶ Same disease, larger cheque: buy a cyclical asset when the cycle has lied to you about how much it's worth, then spend years admitting the truth.

Isn't this just blaming the weather?

The fair objection is that this read lets management off the hook entirely - if everything is the price, then nothing is the executives' fault, and that can't be right. It isn't, and the Alcan story is the proof. The writedowns weren't purely a cycle-timing error. The primary filing and contemporaneous reporting tie them to structural aluminium oversupply, high energy costs, and strong regional currencies - the deal was a worse business than the cycle alone explains, which complicates the tidy 'bought at the top' narrative.⁶ So leadership owns the decision to buy, the price it paid, and the cost discipline that determines who survives the trough. The honest line is narrower than 'it's all the price' and sharper than 'it's all skill': Rio controls whether it is a low-cost survivor of the cycle, but not whether the cycle turns up or down. And the difference between an $11.6 billion year and an $866 million loss¹⁴ lives almost entirely in the part it doesn't control.

Rio Tinto is one of the most efficient diggers of ore on the planet, and that matters enormously - in a price trough, low cost is the difference between a bad year and a fatal one. But efficiency is the floor, not the ceiling. The ceiling is set in a market Rio cannot move, on a number it can only react to, and the 2024 boom, the 2015 loss, and the $30 billion Alcan unwind are three readings of the same instrument. The company's real product was never iron ore. It was a leveraged bet on a price - placed, settled, and re-placed across decades, by people who knew the house always picks the number.