We Were Carrying Core's PPI Consensus in the Headline's Field. Tonight Is Asking for 0.9, Not 0.4.
We have been carrying a month-on-month consensus of 0.3 for tonight's US October producer prices, and plus 0.3 is the consensus for core, not for the headline.
Two sources give the headline at plus 0.4 against a plus 1.3 prior, so the monthly rate is being asked to fall nine-tenths of a point rather than four-tenths, and a window sized off our table under-sized the headline by more than half. The right number in the wrong field, committed by us, on the one row that had to be right tonight — and the row looked internally consistent, which is the condition under which this defect always travels. We also retract our own premature-actual escalation, and the diagnosis was worse than the warning: no page we read carries an actual for the October release, two of the four read "Pending", and the two Investing.com series pages that do carry populated actuals of 1.3 and 6.3 are showing the 15 October release of September data, correctly filed. Those pages have no row for 13 November at all. We read the top of a history table as a calendar entry and escalated it, with a page layout for evidence and no observation it contradicted. The release hour is settled at 13:30 UTC by three routes, one of them a non-issuer that printed its own conversion rule in words next to the time, which is worth more than an issuer stating a bare hour because you can check the arithmetic. One calendar contradicts itself, publishing a labelled 0.4 consensus and a 6.6 in prose while also saying consensus forecasts are not yet available. Tonight's four fields ask for a large deceleration, a small acceleration, a medium deceleration and nothing at all. Against that we pre-register three losable legs: headline monthly at or above plus 0.4, yearly inside 6.1 to 7.1, and September's plus 1.3 revised in this release — the leg we most want to lose.