Why Are Copper Prices Rising Again?

Copper just did something that should make both AI investors and air-conditioner makers uncomfortable. Three-month copper on the London Metal Exchange touched a record $14,858.50 per tonne on September 10, taking its 2026 rise to about 19% and its one-year gain to roughly 48%. 

Hours later, copper futures fell more than 4% after Reuters reported that the White House had not decided whether to extend tariffs to refined copper. That sudden reversal is not a side story. It reveals the main point of this rally: copper has a genuine long-term supply problem, but its latest price spike also contains a large and fragile tariff premium.

Let's break down why copper prices are rising, how much of the rally is driven by real demand, what could happen next, and how investors can get copper exposure through US-listed Copper ETFs.

Copper Price Today: What Exactly Happened?

The benchmark LME price reached $14,858.50 per tonne in early trading on September 10, above the previous day's settlement of $14,767.50. COMEX copper in New York also touched a record $6.894 per pound. At those levels, copper was up about 19% in 2026 and nearly 48% over one year, according to Bloomberg reporting carried by Mint.

Then the policy premium cracked. Reuters reported that the White House was reconsidering a possible tariff on refined copper because of concerns about higher manufacturing costs. Copper futures fell more than 4%, while Freeport-McMoRan shares dropped more than 7% during the session. The administration had not abandoned its efforts to expand domestic copper production, but the report made clear that a tariff was not guaranteed.

The cleanest interpretation is that copper is trading as two assets at once. It is an industrial metal with tight mine supply, and it is a geopolitical asset whose location has become unusually valuable.

Why Are Copper Prices Rising? The Five Main Drivers

1. Copper mines are producing less than expected

The latest rally is not being driven mainly by booming copper consumption in 2026. The more immediate problem is that mines are underperforming.

Global copper mine production fell 1.1% in the first half of 2026, while concentrate production declined 2.6%, according to preliminary International Copper Study Group data reported by The Wall Street Journal. Concentrate is the copper-rich material that mines send to smelters. If concentrate output falls, smelters can exist on paper but still have too little raw material to process.

Chile, Indonesia and the Democratic Republic of Congo recorded notable declines. Peru and Mongolia added output, but not enough to offset the weakness elsewhere. Chile, the world's largest copper-producing country, now expects production to fall 2.6% in 2026 to about 5.3 million tonnes. Ageing mines, lower ore grades, maintenance problems and project delays are all limiting supply.

Lower ore grade simply means miners must dig, crush and process more rock to recover the same tonne of copper. Think of squeezing oranges that contain less juice every year. The factory can work harder, but the output may still disappoint while electricity, water and labour costs rise.

2. A sulphuric acid shortage is threatening mine output

Some copper is extracted through a process called solvent extraction and electrowinning, or SX-EW. In simple terms, sulphuric acid is used to dissolve copper from lower-grade ore before the metal is recovered. This route accounts for roughly 17% of global copper supply, according to Reuters.

The 2026 disruption to shipping through the Strait of Hormuz reduced the movement of sulphur, a key input used to make sulphuric acid. China then restricted sulphuric acid exports from May. Chile and the DRC are especially exposed because their mines rely heavily on acid-based processing.

Goldman Sachs estimated in April that prolonged disruption could put about 125,000 tonnes of DRC output and 200,000 tonnes of Chilean output at risk in 2026. Those are risk estimates, not confirmed losses, but together they equal 325,000 tonnes, or roughly 1.4% of 2025 global mined supply. In a market with thin available inventories, even a partial disruption matters.

3. US tariff fears moved copper into the wrong warehouses

This is the most important short-term driver.

In July 2025, the US imposed a 50% tariff on semi-finished copper products and copper-intensive derivative products. Refined copper cathode, concentrate, ore and scrap were not included. However, the White House proclamation asked the Commerce Department to review whether a 15% tariff on refined copper should begin in 2027 and rise to 30% in 2028.

Traders responded by sending refined copper to the US before any possible duty. By early September, COMEX warehouses held about 695,600 tonnes, while Shanghai Futures Exchange stocks had fallen to about 63,000 tonnes. LME warehouses held about 234,000 tonnes on September 4, and a large share had already been marked for withdrawal. The three exchanges together still held nearly one million tonnes, but most of the growth was trapped inside the US

This is the copper market's airport-luggage problem. The world may have enough bags in total, but that does not help if most of them arrive in New York while passengers in London and Shanghai are waiting at empty carousels.

The September 10 reversal makes this distinction even more important. If the refined-copper tariff is delayed or dropped, some US inventory can eventually return to international markets. That would reduce the location premium even if the longer-term mine problem remains.

4. The world has too many smelters chasing too little concentrate

A copper mine produces ore and concentrate. A smelter turns concentrate into metal, and a refinery produces high-purity copper cathode. These are different parts of the supply chain, and capacity has not grown evenly.

China has expanded smelting capacity much faster than the world's mines have expanded concentrate supply. S&P Global estimates that China accounts for 12 million tonnes of the world's 29 million tonnes of smelting capacity. The result is aggressive competition among smelters for raw material.

Treatment and refining charges, commonly called TC/RCs, are the fees miners normally pay smelters to process concentrate. When concentrate is abundant, smelters can charge more. When it is scarce, the fee collapses. The 2026 annual treatment charge benchmark settled at $0 per tonne, an extraordinary sign that the balance of power had shifted towards miners.

This does not automatically mean the world is out of refined copper today. It means the pipeline is under pressure at its first stage. A refinery can report rising output by using concentrate and scrap accumulated earlier, even while the future feedstock position is getting worse.

5. The weaker dollar and investor flows amplified the move

Copper is priced in US dollars. When the dollar weakens, the metal becomes cheaper for many non-US purchasers, which can support demand. Commodity funds and trend-following traders can then amplify an existing move as prices break previous records.

This financial demand is not imaginary, but it can reverse much faster than mine supply or grid investment. It is another reason a strong long-term copper case can still experience violent short-term declines.

Is There Really a Copper Shortage?

Yes at the mine level and in some regions, but not yet as a clean global shortage of refined metal. That difference is essential.

An older ICSG forecast from October 2025 expected a 150,000-tonne refined-copper deficit in 2026. The group's more recent April 2026 forecast reversed that view and projected a 96,000-tonne surplus, citing slower demand growth and more secondary production from scrap. Preliminary first-half data later showed an apparent refined surplus of about 131,000 tonnes.

The annual surplus is equal to only about one to two days of global consumption. It is therefore too small to provide much comfort when inventory is concentrated in one country, mine output is falling and buyers are paying extra for immediately available metal.

Our view is that the phrase “copper shortage” is directionally right but often used too loosely. The more accurate description is a three-layer squeeze: less concentrate is being mined, available metal outside the US is tight, and the market is pricing a much larger shortage later in the decade.

Is AI Driving Copper Prices Higher?

AI is a real copper-demand driver, but it is not the main reason copper jumped to a record this week.

Data centers need copper in busbars, power cables, backup systems, cooling equipment and the grid that connects the facility to power generation. S&P Global estimates that copper demand linked to data centers will rise from 1.1 million tonnes in 2025 to 2.5 million tonnes by 2040. AI training facilities are expected to account for 58% of data-center copper demand by 2030.

However, traditional demand is still larger. Buildings, machinery, appliances, electricity networks and transport consume far more copper today. AI matters because it adds a new source of demand on top of an already crowded stack, not because GPUs suddenly absorbed the global copper market in 2026.

Source: S&P Global, Copper in the Age of AI.

S&P Global's broader projection is striking. Total copper demand could rise 50%, from 28 million tonnes in 2025 to 42 million tonnes in 2040. Without meaningful expansion, the supply gap could reach 10 million tonnes. The International Energy Agency separately estimates that announced projects could leave primary copper supply about 25% short of requirements by 2035.

These long-term forecasts are not promises. Higher prices encourage recycling, mine expansion and substitution. They can also reduce demand. But the response is slow. S&P estimates that a new copper mine takes an average of 17 years to move from discovery to production. A price spike this quarter cannot create a new mine next quarter.

How Higher Copper Prices Affect AI, Semiconductors and the Global Economy

Copper does not hit the economy like oil. Consumers see an oil shock quickly through petrol, freight and airline costs. Copper moves more slowly through construction budgets, electrical equipment, vehicles, appliances and utility investment.

That makes copper a form of capital-cost inflation. It raises the cost of building the assets needed for future economic growth.

The copper cost pass-through test

A simple way to estimate the pressure is:

Additional raw-copper cost = copper used × change in copper price

The table below compares the September 10 record of $14,858.50 per tonne with a $10,000 reference price. It shows gross exposure before contracts, hedges, fabrication costs, recycling, substitution or supplier margins.

The vehicle and building intensities come from S&P Global. The data-center estimate uses S&P's 30 to 40 tonnes per megawatt range for non-crypto data centers and its example of a large 2 GW hyperscale facility. A project would procure copper over time, not at one day's spot price, so this is a sensitivity test rather than a construction invoice.

The calculation gives investors a useful filter. A company is more exposed when copper represents a large share of its product cost, pricing is fixed for customers, contracts are long, and hedging is limited. It is less exposed when it can pass metal costs through quickly, use recycled copper, substitute aluminium in suitable applications or lock prices in advance.

Who benefits and who faces pressure?

The semiconductor point is often misunderstood. Copper is critical inside chips and fabrication facilities, but the raw metal inside a high-value processor is a tiny part of that processor's price. The bigger economic effect is outside the chip, in power delivery, data-center wiring, cooling systems, transformers and new electricity generation.

For the US, a refined-copper tariff would help some domestic miners and smelters, but it would also raise costs for electrical equipment, autos, construction products and industrial machinery. The US imports roughly half of its copper needs and has only two operating copper smelters, according to Reuters. A tariff can change where copper is stored within months, but a new smelter or mine takes years.

For China, the picture is mixed. It dominates global smelting capacity and remains the largest copper consumer, but poor smelter economics and expensive concentrate create pressure. For Chile, Peru and the DRC, higher copper prices support export earnings and government revenue, although operational problems can prevent producers from fully capturing the price.

For India, expensive copper raises costs for power equipment, cables, consumer durables, construction and electric mobility. S&P Global expects India's electricity consumption to grow 4.2% annually through 2040, faster than the US or China. That makes secure copper supply a growth issue, not just a commodity-market issue. Some Indian appliance makers are already increasing the use of aluminium where product design and safety permit, but substitution cannot cover every application.

Copper Price Forecast: How Much Longer Can the Rally Last?

The long-term case is stronger than the near-term price setup.

At nearly $14,900 per tonne, copper had moved above most 2026 base-case forecasts. Deutsche Bank had forecast a 2026 average of $12,125 and a Q2 peak of $13,000. In April, Goldman Sachs maintained a 2026 average forecast of $12,650. Citi's May scenario work used $12,000 as its base case, $15,000 as a higher case and $10,000 as a lower case.

The market has already moved close to Citi's higher case and 17% to 23% above the other base estimates. That does not prove copper must fall, but it shows that investors at current levels are paying for several favourable conditions to persist at once.

Our central view is straightforward. The old $7,000 to $8,000 copper range is unlikely to be a reliable long-term normal because grids, EVs, data centers and defence are adding demand while mines take longer to build. But nearly $14,900 is difficult to explain through the 2026 refined balance alone. It includes a meaningful premium for tariff uncertainty, scarce non-US inventory and the risk of further mine disruption.

That creates a likely pattern of a higher long-term floor but sharp corrections around it.

These are analytical zones, not point forecasts. Copper can move outside them because currency, war, Chinese demand and investor positioning can change quickly.

What Investors Should Track Next

The copper price itself is the least useful signal after a large move. These indicators tell us whether the rally is becoming more fundamental or more fragile.

Backwardation means copper for immediate delivery costs more than copper promised for later. It is the market's version of paying extra for same-day delivery. A wide backwardation usually signals that metal is needed now. Contango is the opposite and is normally less alarming.

The single best confirmation would be tight mine and concentrate data alongside falling total exchange inventories. If prices remain high while global stocks rise and backwardation fades, the rally is increasingly dependent on policy and investor flows.

How to Hedge Against Rising Copper Prices or Invest in Copper

The right instrument depends on the exposure.

A manufacturer that consumes copper has a business risk. Its goal is to reduce uncertainty in future input costs, usually by matching futures, options or supplier contracts to the quantity and timing of physical purchases. A copper ETF may not be a precise commercial hedge because the fund can track COMEX futures or mining shares rather than the company's exact grade, location and delivery month.

An investor has a return objective. US-listed ETFs offer two broad routes: copper futures, which are closer to the metal price, and mining equities, which add company, country, cost and stock-market risk.

Expense ratios and asset values are based on issuer data and available fund pages around September 2026. Assets change with prices and flows. CPER is structured as a commodity pool and uses futures, while the other three funds own mining equities. Investors should check the latest prospectus, liquidity, spread, tax treatment and India-to-US remittance rules before allocating.

The most important choice is not which ticker had the strongest recent return. It is whether the thesis is about copper itself or the profit leverage of miners.

CPER is the cleaner expression of the metal-price thesis, but futures funds can differ from spot copper because contracts must be rolled from one expiry to another. COPX and ICOP can rise faster than copper when miners expand margins, but they can also lag if wages, energy costs, taxes or operational failures absorb the higher metal price. COPJ adds the possibility of future discoveries and project approvals, along with much greater financing and execution risk.

After a near-50% one-year rise, a staggered allocation and disciplined position size are more sensible than treating copper as a one-way trade. The September 10 reversal showed how quickly a policy premium can disappear.

The Bottom Line

Copper prices are rising because several real constraints have collided. Mine output fell in the first half of 2026, sulphuric acid became harder to source, Chinese smelters competed for limited concentrate, and US tariff fears pulled a record quantity of metal into American warehouses. The longer-term demand case from grids, EVs, data centers and defence then gave investors a powerful story to attach to that squeeze.

But the market is not facing a simple global refined-copper deficit today. The latest ICSG forecast still points to a small 2026 surplus, and the US tariff decision can release part of the inventory distortion. That is why copper can have a strong decade ahead and still be vulnerable after touching almost $14,900.

The most useful mental model is to separate the copper price into four layers: economic demand, structural electrification demand, the location premium and the event premium. The first two can last for years. The last two can disappear on a headline.

For investors, that distinction matters more than any single price target.