The most informative number in the copper market this year is zero, and it is not the price.
Copper is expensive. On 11 August 2026, Trading Economics had it at 6.61 US dollars per pound, up 47.33 per cent on the year and 6.06 per cent on the month. A price that has risen by nearly half in twelve months tells you the market is tight. It does not tell you where the tightness sits, and for anyone selling into that chain, where matters more than how much.
What a treatment charge is, and why it is the better signal
Copper does not leave a mine as copper. It leaves as concentrate, typically a powder running somewhere between a fifth and a third metal by weight, with the rest being rock, sulphur and whatever else the deposit contained. Turning that into cathode requires a smelter and a refinery, and the miner pays for the service.
The payment is structured as a deduction. The smelter buys the concentrate at a value derived from the exchange price of the contained metal, then subtracts a treatment charge, quoted in dollars per dry metric tonne of concentrate, and a refining charge, quoted in cents per pound of payable metal. Together they are known as TC/RCs, and they are the smelter's gross margin. They are negotiated annually between the largest miners and the largest smelters, and the settlement becomes the benchmark that most of the rest of the market prices around.
The number moves inversely with concentrate scarcity. When mines are producing more concentrate than smelters can process, smelters are the scarce party and TC/RCs rise. When smelters are competing for insufficient concentrate, they bid the charge down to win tonnage.
The 2026 benchmark is reported to have settled at approximately zero dollars per tonne.
The settlement is consistently reported. Antofagasta and a Chinese smelter agreed zero dollars per tonne and zero cents per pound for 2026, against 21.25 dollars and 2.125 cents for 2025. It is the lowest benchmark on record, and it means smelters gave away their entire treatment margin for the year in order to secure feed. A smelter that processes concentrate for nothing is not running a business. It is defending a plant, keeping furnaces hot and workers employed against the day feed returns, and accepting that the year will be paid for out of by-products and acid rather than out of copper.
That is a more precise statement about the market than any price. It locates the shortage at the concentrate stage specifically, not in refined metal, not in demand, not in inventory. Concentrate is the scarce thing.
The price, and the trouble with comparing prices
The exchange has been behaving accordingly. LME three month copper reached a record 14,527.50 dollars per tonne on 29 January 2026, in what Benchmark Mineral Intelligence recorded as the largest single day rise since 2008.
A word of caution on reading that alongside the spot figure above. The 6.61 dollars per pound quoted by Trading Economics and the LME three month contract are different instruments on different exchanges with different delivery terms, and converting one into the other to declare a new high or a retreat is a sleight of hand that appears constantly in commentary. What can be said without arithmetic games is that copper made a record in January 2026 and remains, in August, at or near those levels and far above where it sat a year earlier.
Where the scarcity came from
Two supply events are consistently cited, and their dates matter more than the citations usually admit. The Democratic Republic of Congo imposed a ban on copper concentrate exports in early August 2026, which places it after the January record rather than among its causes. Codelco suspended expansion works at El Teniente on seismic grounds, also in early August. A duration of as long as two years has been quoted for that suspension, but by a union representative rather than by Codelco.
On the demand side, the drivers named are the buildout of AI datacentres and of the electricity grids required to supply them. Copper is the material a datacentre is wired with and the material a transmission upgrade is built from, and both of those programmes were committed to before the metal repriced.
We would attach medium confidence to the individual details in that paragraph, particularly the duration of the El Teniente suspension, which is the kind of figure that is revised. The overall narrative, a concentrate market short of feed while structural demand builds, is well corroborated and is what the treatment charge independently reports.
The case against
Presenting the deficit as consensus would be dishonest, and it is not consensus.
Goldman Sachs publicly forecasts copper prices declining from their record highs during 2026. That is a serious house with a serious research function taking the other side, and any seller building a contract around the assumption of a continued rise should know that it exists.
The two positions are less contradictory than they look, which is the useful part. A concentrate shortage and a falling refined price can coexist. Zero treatment charges mean smelters are short of feed. They say nothing about whether refined metal is short at the consumer, and a high price attracts scrap, releases inventory and eventually rations demand. A seller of concentrate and a holder of cathode are exposed to two different markets that happen to share a headline number.
What follows practically is that a seller should be more confident about the direction of treatment terms than about the direction of the price, because those are the two questions the available evidence answers with different degrees of certainty.
What this means from Karachi or the Gulf
Pakistan's copper trade is, at present, a single relationship. Roughly 95 per cent of the country's copper exports and 98 per cent of its copper mattes went to China in 2024, according to CSIS in April 2026.
Concentration of that order changes what a strong market is worth. When 95 per cent of your material has one destination, the market price sets a ceiling on what you can be paid and the buyer's alternatives set the floor, and a seller with one buyer has no floor. A global concentrate shortage improves a seller's position only to the extent that a second credible buyer exists to be quoted against the first. Otherwise the shortage is captured downstream and the miner reads about it.
The practical work this points at is unglamorous. Qualify a second smelter, in a second jurisdiction, on a small tonnage, before you need it. The cost of doing that is a shipment sold at a slightly worse net back and a set of assays and site visits. The value of it is that every subsequent negotiation is conducted against a real alternative rather than a hypothetical one, and the buyer knows it.
The second is to read your own contract for who captures the treatment charge. In a year when the benchmark may be at or near zero, a concentrate sale priced on terms fixed when treatment charges were meaningful is transferring a large sum to the counterparty for a service the market has stopped valuing. That is not a complaint about the buyer. It is the mechanical consequence of a term that was reasonable when it was written and has since been overtaken.
Prices are reported everywhere and are the first thing anyone looks at. Treatment charges are reported in trade publications, settle once a year, and describe with far more precision which participant in the chain is actually short. A seller who tracks only the first is watching the loudest number rather than the most useful one.