August CPI at 3.4% Hands [Kevin Warsh] a 90%-Priced Hike and a Curve That Already Broke

Bureau of Labor Statistics reported August headline CPI at +0.4% month-on-month and +3.4% year-on-year — both in line with the Bloomberg consensus and unchanged on the year. On a headline basis, the release was almost a non-event. Underneath, the composition broke hawkish: core CPI accelerated to +0.3% MoM from +0.2% in July, the largest monthly print since April, even as the year-on-year rate eased to 2.4% from 2.5%. A 3.4% headline paired with a 0.3% core print tells the Fed the underlying trend is firming just as the goods side should be rolling over.

Shelter ticked back up to +0.3% MoM from +0.1% in July — a small re-acceleration in the largest core component — while transportation services rose 0.5% MoM and used vehicles added another 0.4% MoM. The only meaningful disinflation in the report was motor vehicle insurance, which fell 0.8% MoM. Inside the core basket, Reuters flagged a 5.9% jump in mobile-phone service costs and a 2.7% rise in airline fares on jet-fuel pass-through. The Piedmont Crescent summary called it \"no better than expected\" — and the bond market agreed, just not in the way economists had hoped.

Gasoline rose 3.9% MoM in August and27.4% YoY, accounting for more than a third of the monthly index gain on its own, per BLS and The Guardian. Fuel oil is up 52% YoY; the energy index as a whole is up 16.3% YoY, the highest annual reading since May. The supply shock is the Iran war, which has kept the U.S. retail gas average above $4 a gallon all month. August 2026 is now the most expensive August for gasoline ever recorded, beating 2022.

That distinction matters for the rate-path math. Demand-driven inflation is what rate hikes can fix; energy-driven inflation is what rate hikes can deepen. Friday's PPI release made that ambiguity worse: headline PPI rose 5.4% YoY (above the 5.3% consensus) and core PPI climbed to 4.6% YoY, the highest since June, per CNBC and Reuters. Producer prices feed through to core PCE — the Fed's preferred gauge — and economists now expect August core PCE to print +0.3% MoM after +0.2% in July.

The Treasury curve bull-flattened on the print — short rates moved, long rates barely budged. That is the textbook signal of a market repricing the policy path, not the growth outlook. The 2-year's move to 4.628% pulled real rates higher at the front end of the curve, while the 30-year held at 5.356% — a73 bps gap that looks generous until you remember mortgage spreads sit on top of it.

The 25-point swing in implied probability across three sessions is what makes this a credibility event, not just a data event. The market has effectively told [Kevin Warsh] what it expects him to do. Defying it means breaking the front end of the curve.

Kevin Warsh was sworn in as the 17th Fed chair on May 22, 2026, after a 54–45 Senate confirmation that was the most narrowly partisan in the institution's history. He replaced Jerome Powell, who remains on the Board of Governors with a term running through January 2028. The timing was deliberate — President Trump nominated Warsh on March 4, 2026, after a year of public attacks on Powell for not cutting rates fast enough.

What Warsh gave the White House instead was hawkishness. At Jackson Hole on August 28, he said the Fed \"has work to do\" if inflation does not return to 2%, pledged to re-establish the central bank's inflation-fighting credibility, and indicated rates could rise further — language Reuters and AP read as a clear regime change. June's Summary of Economic Projections — still Warsh's first official dot plot — already showed a median 2026 funds rate of 3.8% and 3.6% for 2027, per RSM. That is one hike baked in, with optionality on more.

The 30-year fixed mortgage rate hit 6.71% the week of September 3, Realtor.com and CNN reported — a 2026 high and within striking distance of 7%. A confirmed September hike would push that benchmark toward 7.0%–7.1% on most lender rate sheets, because mortgage spreads to the 10-year have already widened through the front-end repricing. D.R. Horton disclosed20% order cancellations in mid-2026 as mortgage rates climbed, per Yahoo Finance, and existing-home inventory has stayed frozen for the same reason. The transmission from the Fed's policy rate to a30-year fixed is mechanical — and that machinery is already running.

What is not mechanical is the Fed's optionality. A September hold does not unwind the energy shock; it just punts the credibility question to October and December, when the Core PCE prints for August (Sep 30) and September (Oct 30) will have absorbed any Iran-oil pass-through. October also brings the next CPI release on October 14 and the next FOMC on October 28. The path of least resistance for a chair who campaigned on credibility is to deliver the hike the market has already priced.

Over days to weeks, the catalyst is unambiguous: the September 16–17 FOMC decision, with the SEP and dot plot released the same afternoon. A 25 bps hike is now the consensus trade; an unexpected hold would trigger the largest front-end repricing of the year (a 25–30 bps rally in the 2-year) and put the spotlight on the new chair's first deviation from market pricing. The August core PCE print on September 30 is the secondary near-term catalyst — if it lands at +0.3% MoM as expected, it confirms the hike narrative; if it cools to +0.2%, the case for an October pause strengthens.

Over one to three years, the structural question is whether [Kevin Warsh] can deliver on his credibility pledge without breaking the housing market or the small-cap debt complex. The June dot plot still showed a median 3.6% funds rate by end-2027 — implying roughly one more hike after September and a hold through 2027. A second hike in October or December followed by a 2027 hold would be the textbook landing: inflation back near 2.5%, fed funds at 3.75%–4.00%, 10-year stabilizing in the high-4s, and the 30-year fixed easing back toward 6.3%–6.5% as mortgage spreads normalize. The risk to that path is the Iran energy shock feeding through to wages — August's average hourly earnings, inflation-adjusted, fell 0.3% YoY, but nominal wage growth is still running above4% in the establishment survey.