Headline prices rose 0.4% on the month and held at 3.4% over the year, both in line. The one number that missed was core, at 0.3% against 0.2% expected, and the bond market's two ends read it differently.
The August consumer price index, released Friday at 8:30 a.m. Eastern, delivered three numbers that matched expectations and one that did not.
Headline prices rose 0.4% from July on a seasonally adjusted basis, in line with the 0.4% consensus but a fourfold acceleration from July's 0.1%. On a twelve-month basis the headline rate held at 3.4%, unchanged from July and matching expectations. Core prices, excluding food and energy, rose 2.4% over twelve months, down from 2.5% in July and in line with forecasts. The consumer price index itself came in at 334.98, above the 334.85 consensus and up from 333.92.
The miss was in core on a monthly basis: prices excluding food and energy rose 0.3%, against a 0.2% consensus and up from 0.2% in July. Real average earnings rose 0.2% on the month, up from 0.1%.
The annual rate and the monthly rate are pointing opposite directions
That divergence is the entire story of this print, and it is the reason the reaction was not clean.
Annual core inflation decelerating to 2.4% is the number that supports patience. It is the lowest core reading of the recent sequence, it moved in the right direction, and it is the measure most often cited as the underlying trend.
Monthly core accelerating to 0.3% is the number that argues the opposite. Annual rates are arithmetic built partly from what happened a year ago, so a falling twelve-month figure can coexist with deteriorating current momentum. A 0.3% monthly pace, sustained, annualizes to roughly 3.7%, well above the Federal Reserve's target. One month is not a run rate, but it is the first month in this sequence where core came in above expectations rather than at them.
The headline acceleration from 0.1% to 0.4% arrives against a backdrop that makes its likely source obvious: Brent crude has held above $100 for weeks and U.S. diesel prices crossed $6 a gallon for the first time this week. What the component tables will show about how much of that reached core categories is the detail worth waiting for, because that is the mechanism by which a supply shock stops being something a central bank can look through.
The front end and the long end disagreed within minutes
The 10-year Treasury yield spiked to 4.984% immediately after the release, a fresh 52-week high, then retreated to 4.956% by roughly 8:35 a.m. Eastern, holding a small gain against Thursday's 4.945% close.
The 30-year did something different. It touched 5.382%, also a 52-week high, before reversing to 5.349%, below Thursday's 5.362% close. The long bond finished the first five minutes lower in yield than it started.
That split is informative. A market that reads this print as raising near-term Fed risk while lowering long-run inflation risk produces exactly this shape: pressure at the 10-year point, relief at the 30-year. The core monthly overshoot argues for the first, and the annual core deceleration argues for the second, and the curve appears to have priced both at once rather than choosing.
Equity futures chose. S&P 500 e-mini futures traded as low as 7,594.25 before the release and stood at 7,643.5 minutes after it, up 45 points or 0.59%. Nasdaq 100 futures rose 192.25 points, or 0.66%, to 29,327.5. The CBOE Volatility Index fell 4.82% to 16.98 from Thursday's 17.84 close.
Gold fell $35.80, or 0.81%, to $4,371.50, extending a decline it has run through a week of nominally supportive news. The dollar firmed, with the euro down 0.27% to 1.15822. West Texas Intermediate fell 3.25% to $99.15 and Brent fell 3.15% to $104.24, both continuing a reversal that began before the print.
What next week's meeting now turns on
The Federal Open Market Committee meets September 15 and 16. Interest-rate futures and prediction-market pricing had put the probability of an increase between roughly 60% and 70% before this release, after a hotter-than-expected August producer price report on September 10.
This print does not obviously resolve that. A committee focused on the twelve-month core rate has a number that improved. A committee focused on monthly momentum has a number that deteriorated, alongside an energy shock still working through the pipeline. Both readings are defensible from the same release, which is the least helpful outcome for anyone trying to price the meeting.
The next scheduled input is the University of Michigan consumer sentiment survey at 10:00 a.m. Eastern Friday, which carries one-year and five-year inflation expectations readings. Expectations were running at 4.0% and 3.3% respectively in the prior survey. If those move up alongside a hot core monthly print, the case for looking through the energy shock gets materially harder to make.