What the Composition Reveals

Beneath the headline, July showed a specific divergence. Goods prices fell 0.1 percent month-over-month, driven by a 2.7 percent decline in gasoline and a 0.9 percent decline in furnishings. Services prices rose 0.3 percent, pushed higher by a 1.2 percent increase in financial services and insurance and a 0.3 percent gain in housing. Year-over-year, goods prices are up 1.3 percent while services are up 2.5 percent. Disinflation has largely come from goods; services remain sticky.

The Fed cares about services more than goods because services inflation reflects domestic labour costs and pricing power rather than global supply dynamics. Services inflation well above target with a labour market still growing wages at 4 percent-plus is a structural problem the Fed cannot easily solve without slowing the economy further.

The Distance to Target

Headline PCE at 3.7 percent is 1.7 percentage points above the Fed's 2 percent target. Core PCE at 3.3 percent is 1.3 percentage points above. Both readings have been essentially flat for four months. This is what Warsh described as "concerning" at Jackson Hole. The Fed has spent two years assuming current policy is restrictive enough to eventually bring inflation to target. The July print, following June, May, and April, undermines that assumption.

What Analysts Should Actually Watch

Three data points now determine whether the July stubbornness was noise or signal. The August CPI on 10 September, which typically leads PCE by two to three weeks and provides an early read. The August PCE on 30 September, released after the September FOMC. And the specific services categories in both reports, particularly housing, financial services, and healthcare.