BNEF: copper deficit from 2028 needs $332B of new mines, while most other metals ease
The battery-metal panic is fading in BNEF's numbers. Copper is the one metal where the long-term gap is getting harder to close, not easier.
What happened
BloombergNEF released its Transition Metals Outlook 2026 on 20 October. It forecasts a copper supply deficit beginning in 2028 and widening to 20 million metric tons by 2050, equal to 44% of projected global demand, and says addressing the shortfall would require $332 billion of investment in new mines. BNEF expects copper demand to rise 61% between 2026 and 2050, to almost 46 million tons.
Elsewhere the picture improved. BNEF says nine of the 10 materials it studies are on track for sufficient supply capacity through 2050, compared with three projected deficits last year, and graphite, platinum and steel have shifted from deficits to surpluses. Lithium faces a temporary deficit that should ease as new capacity comes online in 2027, according to the report.
Our take
This lines up with the near-term picture from the ICSG: plenty of refined copper for the next two years, then a mine-supply wall. BNEF says new mines take an average of 17 years to come online, so the money has to be committed while the market still looks comfortable. That timing mismatch is the real risk.
For battery metals, the message is the opposite. BNEF expects long-term downward price pressure on cobalt, lithium, manganese, nickel and palladium, which supports our view that the risk for those metals sits in refining concentration, not in tonnes. BNEF notes eight of the 10 materials have more than half of refining capacity in a single country, seven of them in China.
Why it matters for AI infrastructure
BNEF's report does not single out data centres, but copper is the metal AI build-outs lean on hardest, from grid connections and substations to busbars and cabling inside the hall. A structural deficit from 2028 lands just as many announced campuses are scheduled to come online.
Expect more developers and hyperscalers to look at long-term copper offtakes, aluminium substitution where it is technically possible, and direct stakes in miners, a theme we flagged last week.
What we're watching
- Final investment decisions on large greenfield copper mines
- Whether near-term refined copper surpluses delay new mine funding
- Copper substitution (aluminium) in grids and data-centre power systems
- Refining concentration: BNEF counts seven of 10 materials concentrated in China
Get takes like this every Sunday.
One free email with the week's export controls, list changes and price moves, and what they mean for AI infrastructure.
See a sample issue