Start with the prints, because they are not close calls. PPI month-on-month came in at 1.3% against a 0.5% forecast — more than double. Retail sales rose 0.8% against a 0.3% forecast — also more than double. Headline CPI year-on-year rose to 3.7% against 3.6% expected, and up from 3.4% the prior month. Core CPI year-on-year rose to 2.6% against 2.5% expected, up from 2.4% prior. Existing home sales beat too, at 3.99M against 3.95M expected and 3.93M prior. Every one of the five releases in the data block landed on the hot side of the consensus line, and three of the five came with month-on-month or year-on-year acceleration, not just a beat.

The 10-year yield did not treat any of this as reason to rise. It fell 1.95% on the day and 2.71% over the week, closing at 5.089%. The 5-year told the same story at shorter duration: down 1.86% on the day, down 1.84% on the week, to 4.899%. The 2s10s spread sits at +0.33 — a normal, unstressed curve shape, not one bracing for a policy surprise. Every other stress gauge in the regime read agrees: VIX at 16 (calm), the financial stress index at -2.14 (benign), high-yield spreads at 325bp (tight). The regime classifier calls this pre-vote risk-on with 78% of weighted votes — a market positioned for calm into a vote, not for a fight over an inflation surprise.

The vote in question is the Fed's October 28-29 decision, and odds markets price an 86% probability of no change, on $43.5M of volume, a contract that does not resolve until October 29. That number is the hinge. A bond market that is 86% certain of the near-term policy outcome has less reason to reprice sharply on four data points that land before the decision than it would reprice to new information about what happens after it. One widely-read blog, writing off the Fed's own September 15-16 minutes, argued the Fed looked poised to tighten further on mixed inflation signals (capitalspectator.com) — the general shape of today's complaint, just two prints earlier. Today's prints are hotter than the ones that prompted that read, and the rate move went the other way.

Two readings of the same gap are both defensible, and they point to different places. If the yield decline reflects confidence that the October decision is settled regardless of what inflation does between now and the meeting, then the test is the data due after the vote — starting with Building Permits on October 20 (expected 1.40M, prior 1.403M) — which would need to show the acceleration holding into November before the curve moves. If instead the decline reflects demand for duration unrelated to the inflation calendar — funding conditions are flat, with SOFR-EFFR at 0bp and the Treasury General Account down $71bn over four weeks to $920bn, freeing cash into the system — then the next inflation print should matter more than this one did, not less.

Either way, the honest statement is that four hot data points and a falling 10-year yield happened on the same day, and the facts here do not say which explanation is doing the work.

The data sweep goes beyond the headline CPI and PPI numbers. Core CPI month-on-month held at 0.3%, matching the prior month and running above the 0.2% forecast — a smaller miss than PPI or retail sales, but a miss in the same direction. Existing home sales rose to 3.99M, above both the 3.95M forecast and the 3.93M prior reading, the second straight increase in the series as given.

One widely-read blog ties the housing data to affordability math rather than volume: the median Miami household earns about $62,000 a year against mortgage-linked housing costs near $2,900 a month, more than half of income annualized; the median Los Angeles household earns about $82,000 against $3,500 a month, 51% of income (zerohedge.com). Existing home sales rising against that backdrop says demand is clearing at those prices, not that the prices have become easier to carry.

The macro stress composite reads 4.3, labeled Low. That composite sits uneasily next to a PPI print that beat forecast by more than a full point and a retail sales print that beat by a factor of two-plus — a reminder that a low stress score describes financial-market plumbing (funding, credit spreads, volatility), not the inflation data arriving alongside it.

The next scheduled test is Building Permits Preliminary on October 20 at 8:30 ET, expected at 1.40M against a prior 1.403M — a number that, if it also beats, would make today's clean sweep three-for-three across housing, prices, and spending.

The rest of today's commentary is on Pro and Full Access.

Those sections are absent from this page, not hidden.

See the plans · Sign in