Underlying US inflation on the Fed’s preferred measure was uncomfortably warm for a third month in February, delayed data released on Thursday showed.
Core PCE printed 0.4% MoM in the release. Unrounded, the index rose 0.367% from the prior month.
Although economists expected another warm read, confirmation of a wrong-way trajectory for core inflation on the metric monetary policymakers consult was incrementally concerning.
Put differently, yes, we already knew things were moving in the wrong direction even ahead of March’s energy shock, but that doesn’t completely take the sting out of Thursday’s readout. It’s a wince moment.
The core gauge rose 3% versus the same period a year ago. So, underlying inflation’s stuck a full percentage point above target.
Headline PCE likewise posted a 0.4% MoM gain. YoY, it rose 2.8%, as expected.
Notably, personal spending missed both in nominal terms (+0.5% versus +0.6% seen) and adjusted for prices (+0.1% against +0.2% expected). Personal income fell, posting a 0.1% decline, below every estimate.
The only good news from Thursday’s BEA release covering February came courtesy of so-called “supercore” inflation — i.e., services price growth excluding housing. That metric rose just 0.2% MoM and 3.2% YoY. Those are tolerable prints if you’re the Fed.
Meanwhile — and this is an afterthought, but it’s worth a mention. The final read on Q4 GDP, also released on Thursday, showed the US economy expanded at a 0.5% clip during 2025’s final quarter. That constituted another downward revision.
Here’s how that print progressed. Economists expected 3% ahead of the advance read, which printed less than half of that at 1.4%. That 1.4% was itself revised in half to just 0.7% in the second estimate. As of Thursday, it’s 0.5%. Thank God there’s not a fourth estimate.
On the bright side, GDI showed a 2.6% gain. That rescued the GDP-GDI average the NBER uses in its official cycle-dating process. The average of the two for Q4 2025 was 1.5%, down sharply from the prior quarter, but certainly not indicative of recession.
Praise Jesus there is no fourth revision
So correct me if I’m wrong. The US can release the first “close to consensus” GDP estimate, and then revise it down significantly 2 times 3 months later, which by then no one cares because it’s an afterthought. I reckon the same pattern for NFP, which tends to be upbeat in the first months of the year (2026’s February was an outliner), only to get revised down significantly 6 months later when no one cares anymore. And here I thought the Chinese statisticians are the only geniuses at math. All’s well as long as the stocks are up, I suppose.