A copper crisis is coming. Almost no one is ready.
Copper is the physical bottleneck under both the energy transition and the AI build-out. Demand is compounding across the grid, renewables, electric vehicles and data centers all at once, while a new mine takes roughly 17 years to build and ore grades keep falling. The result is not that the world runs out of copper. It is that supply cannot grow fast enough, and on the IEA’s own numbers the market breaks from sixty years of balance into a widening gap in the late 2020s. Here is the full research, with every claim sourced.
- Copper is a supply-rate problem, not a shortage of copper. Reserves are large. The binding constraint is how fast new supply can be permitted and built against demand that has grown almost every year for a century.
- Demand reached 28.2 million tonnes in 2025 and keeps compounding. More than half of all the copper ever mined has been used just since 2000, and the IEA projects demand near 33 million tonnes by 2035.
- A new mine takes about 17 years from discovery to full production. Grades have fallen 40% since 1991, and IEA mined supply peaks in the late 2020s then falls below 19 million tonnes by 2035, counting every announced project.
- Price alone is not unlocking supply. Freeport's Bagdad expansion has cleared its own $4/lb go-ahead price for years, yet across four straight investor updates the decision slipped backward while copper rose.
- The balance breaks in the late 2020s. On IEA figures the market moves from sixty years of balance to a gap of roughly 7 million tonnes a year by 2035. The world pays in higher prices and cancelled projects, not by running out.
Six decades in balance. Then the lines part.
For as long as records exist, copper supply and demand have tracked within about 2% of each other. On the IEA’s own figures, that balance breaks in the late 2020s, opening a gap of roughly 7 million tonnes a year by 2035.
Demand has grown almost every year for a century.
Through recessions, oil shocks and the financial crisis, refined copper usage has fallen more than 1% in a single year only once in the past quarter-century. The next leg of growth is broad, not just green: ordinary development is the largest single driver.
A century of compounding demand. Most of it since 2000.
Refined copper usage has grown almost every single year for over sixty years, on one consistent measure. More than half of all the copper ever mined has been used just since the year 2000.
Of the roughly 7.4 million tonnes of demand growth the IEA projects to 2040, traditional uses and electric vehicles together are about three-quarters of it.
Electrification is copper-hungry by design.
Copper carries more current per unit than any affordable metal, so cleaner power uses far more of it. An offshore wind plant needs about seven times the copper per megawatt of a gas plant. Substitutes displace only about 1.3% of copper use a year; aluminum, the main one, needs about 1.6 times the cross-section and corrodes more.
Not scarcity. The cost and the clock.
The world is not short of copper in the ground. It is short of new mines, because finding and building them has become slower and far more expensive, and only four world-class projects are due online in the next five years, only one of them on track.
Freeport’s Bagdad mine in Arizona has run for a century, with roughly 80 years of copper still underneath and a plan to double output. That expansion needs a $4/lb copper price to go ahead. Copper has traded above $4 in five of the last six years, and sits well above it today. Yet across four consecutive investor presentations, the approval date moved backward, from end-2025 to the second half of 2026, while the price kept rising. If the highest price in a generation cannot pull the most obvious expansion forward, the constraint was never price.
The market clears through price, not depletion.
There is no copper deficit today. The problem starts later this decade, when the gap opens. A market that cannot add supply fast enough does not simply run dry. It rations, through higher prices that make some projects uneconomic and destroy some demand at the margin.
That is the auction underneath the energy transition and the AI build-out. Both are priced today as a story about chips and capital. Underneath, they are competing for the same tonnes of a metal the world cannot mine any faster, and the bill comes due in copper.
Get the sourced version of every video.
Every WireSift video ships with a page like this one: the charts you can explore, and every claim traced to its source. Get the next one in your inbox.