Gold Fell3.6% While Treasuries Bought the Bid: October 30 Was a Funding Squeeze, Not a Safe-Haven Moment

On Oct. 30, spot gold settled at $3,875.84, down 3.62%, its weakest level since early October 2025, and silver settled at $48.90, down 6.22% and below $50 for the first time since November 2025. Over the same three sessions the 10-year Treasury yield slid from 5.05% to 4.94%. In a normal risk-off, gold and Treasuries rise together; here Treasuries bid while bullion was sold into a violent technology selloff, with Intel down 28% and the Nasdaq down more than 3%. The safe-haven bid skipped bullion entirely, landing in duration and in the corporate credit that the AI buildout keeps printing.

Three facts separate this from an ordinary macro selloff. First, the Fed held rates at 3.75%-4.00% on Oct. 28 with two policymakers dissenting in favour of a hike, so the hawkish repricing was already in the price before the equities broke. Second, oil still rose, with WTI up 10.9% over the month, which left inflation expectations rising rather than falling. Third, and decisively, the marginal dollar that would normally have bought gold went into long-dated Treasuries and into investment-grade credit issued to fund data centers. This was a rotation of the hedge, not an abandonment of it.

This is where the correlation regime investors relied on for a decade actually breaks. For a year, gold and bonds have traded as opposites: the bond market selloff is what dragged gold 28% from its January peak, per Morningstar, and on Oct. 30 the same trade simply continued with a different velocity. Gold was the funding leg of the AI trade; Treasuries were its financing leg. When the AI equity broke, the collateral got sold and the financing got bid, which is why gold fell 3.6% on a day the 10-year rallied 11 basis points over three sessions.

Silver's problem is not that the physical market loosened in October; it is that its industrial base is being engineered smaller at the same time as its financial holders were forced to sell. The World Silver Survey 2026 puts photovoltaic silver use at about 151 Moz in 2026, a 19% decline after a 6% fall in 2025, because silver loadings have climbed from roughly 8% to more than 20% of cell cost. Mainstream cell loadings are expected below 5 milligrams per watt by 2027, and copper substitution could remove about 90% of the silver in a module. The metal's industrial tell now reads as demand destruction, and industrial use of 657.4 Moz was 58% of total silver demand, the largest single category.

Mining equities fell harder than the metal they sell, which is the operational signature of a flow event: Hecla Mining dropped 6% and is60% below its 52-week high, First Majestic Silver fell 6%, Coeur Mining fell 6.2% and Pan American Silver fell 6.3%, against gold down 3.6%. Equity multiples on peak-cycle earnings are the thing that breaks first, because the earnings are marked to a spot price that has moved 31% since January.

The supply chain did not move as one piece. Producers of bullion lost roughly a tenth of their equity value in a session because their revenue line is marked to a falling spot price against costs that reset over quarters. The companies that consume the metal had the opposite experience: the same move that broke Newmont and Pan American Silver widened their margins. Cheaper silver and cheaper gold are a cost windfall for solar cells and jewellery, and only a revenue problem for miners — which is the opposite of the usual narrative that metal rallies lift the whole chain.

Note what the hawkish quote does and does not explain. It explains why gold entered Oct. 30 at a one-year low, because higher-for-longer real rates had already compressed the metal 31% from its January peak. It does not explain why gold fell 3.6% on a session when the 10-year yield fell, or why silver fell twice as far. The hawkish regime set the starting line; the squeeze decided the direction.