Copper Tightness / Grid Demand Pull

Tracks whether copper's inventory drawdown and price extreme are the physical confirmation that electrification and data-center grid buildout are converting into metal demand, separate from the monetary metals and from the equipment order books.

September 16 update: the price leg gave back the extreme. Copper fell 4.53% in five sessions to $6.50, bigger than all but 24 of the last 120 such moves, a week after one-year-high closes of $6.85 and $6.81, and nothing new arrived on warehouse stocks, equipment lead times, or input-cost language, so the 42-day London drawdown through mid-August is still the entire case. The metal is still above the $6.37 line whose break would restore the demand-gap read, and it fell in the same window that marked down chips, small caps, and miners, which points at the rates repricing rather than at physical demand turning. Next check: stocks continuing to draw while the price stabilises above $6.37, and turbine, transformer, or cable makers reporting longer lead times rather than only larger order books. September 12 update: the price leg gave some back and nothing new arrived underneath it. Copper fell 1.78% to $6.47 after closing $6.85 and $6.81 on September 8 and 9, the highest in a year, and no fresh evidence landed on warehouse stocks or equipment waiting times, so the 42-day London drawdown through mid-August is still the entire inventory case and the grid-construction attribution still rests on the demand mix rather than on capex data. Next check: stocks continuing to draw while the price holds, and turbine, transformer, or cable makers reporting longer lead times rather than only larger order books; inventories rebuilding with a break back below $6.37 restores the demand-gap read. September 9 update: the candidate opens on an inventory series and a price extreme rather than on a forecast. Copper closed $6.85 on September 8 and $6.81 on September 9, the highest closes in a year, with London warehouse stocks drawing for 42 consecutive days through mid-August, while electric-car sales grew about 2% and solar installations are forecast to fall, so the consumer and renewables end-markets are not the source of the pull, which leaves grid and data-center construction. No standing row holds this: [[grid-hardware-bottleneck]] scores the equipment order book at GE Vernova and its peers, the AI rows score chips and financing, and the retired industrial-metal-demand-gap row was built on the opposite claim, that copper was contradicting the buildout, this is that row's premise inverted, not its reopen condition. What is unproven is transmission: a drawn-down warehouse and a high price are not yet a cost in anyone's accounts. Next check: whether stocks keep drawing and the tightness appears in manufacturers' input costs or in turbine and transformer lead times rather than only in the exchange price.