Uh-oh.

A key measure of underlying demand surprised to the downside in Friday’s update on the American consumer.

The so-called “control group” in the Commerce Department’s monthly retail sales report dropped 0.4% in July, undershooting every estimate. Consensus expected a 0.3% advance.

Recall that the control group, which economists use to refine GDP forecasts, was resilient during four months of war, which is to say despite a sharp rise in gas prices, Americans still found a way to fund discretionary spending.

One month doesn’t make (or necessarily break) a trend, but the disappointing read suggests the previously bulletproof American consumer began Q3 on shakier-than-expected footing. The three-month average annualized pace of control group sales, which was still running north of 8% headed into Friday’s release, is now just 5.6%.

The headline retail sales print showed a 0.6% drop. That too was a big miss, and it constituted the largest MoM decline since the blowback from “Liberation Day.”

This is, of course, nominal spending. The decline on the headline was in part attributable to falling pump prices. Gas station receipts fell 0.9% after a near 6% decline the prior month.

All caveats aside, and even considering only five major categories showed an actual decline, this report was weak. The nonstore line showed a pronounced 2.2% drop, and the only meaningful gain was at clothing stores. The food services and drinking places line, the only services-sector category in the release, reflected a halfway decent, 0.5% advance.

Suffice to say this is yet another piece of evidence to support the contention that Kevin Warsh’s Fed won’t be raising rates next month.

We now have a very poor jobs report, consecutive cool inflation readouts (CPI and PPI) and now a lackluster read on nominal spending.