BREAKING: The US Consumer Refuses to Break and That Settles the Fed Debate

August advance retail sales beat across the board, and every layer beneath the headline came in stronger than the headline itself. With a 25bp hike already 92% priced going into today's FOMC, the report does not change September but it makes the case for stopping there considerably harder to argue.

Key numbers:

• Headline: +1.2% m/m vs +0.8% expected; prior -0.6% revised to -0.5%
• Ex-autos: +1.4% vs +0.5% expected, a near-triple beat
• Control group: +1.4% vs +0.4% expected, after -0.4% in July; this is the line that feeds PCE goods and GDP tracking
• Ex-autos and gas: +1.2% vs -0.3% prior, the cleanest read of underlying demand
• Y/Y nominal: 6.0% vs 5.01% prior
• Category detail: gasoline +3.1%, nonstore +2.6%, furniture +1.9%, electronics and appliances +1.6%, restaurants and bars +1.2%, sporting goods +1.2%, building materials -0.2%
• Market reaction: minimal. Hike odds unmoved at ~92%, two full hikes still priced through year-end, US10Y above 5%

The gasoline objection does not hold

Retail sales are nominal, and August was the month the oil shock hit the pump. Gasoline stations at +3.1% m/m is price, not volume. Brent rose roughly 18% on the month. That is a real distortion of the headline

It is also already stripped out, and the ex-autos-and-gas measure still printed +1.2%. This is the whole story. Higher fuel costs normally act as a tax on the household budget: consumers pay more at the pump and pull back elsewhere. In August they did not. Restaurants and bars, the first line households cut when squeezed, rose 1.2%. Only building materials fell, and barely. The US consumer absorbed a significant energy shock without reallocating away from anything.

Why that matters for the Fed

Central banks are supposed to look through supply-side energy shocks, and the case for doing so rests on one assumption: that the shock destroys demand on its own. This report falsifies that assumption. When spending is strong enough to swallow a fuel price spike whole, the conditions for second-round effects are in place, firms facing higher input and freight costs will discover they can pass them on.

It also kills half of the equity market's stagflation narrative. The S&P 500 has slid to six-week lows on the assumption that an energy shock plus a tightening Fed produces a growth accident. This data says there is no "stag" right now, only the "flation." That is a worse mix for bonds than for stocks.US500 is up today after hitting the lowest level since August 3. However, in first minutes after the data release, US500 lost about 7 points. Source: XTB

Three caveats, none of them dovish

This is a rebound from a negative month, so the two-month average is far more pedestrian than the headline implies. Some of the strength in furniture and electronics has the signature of pull-forward buying ahead of expected price increases, which borrows from future quarters, but is a symptom of unanchored inflation expectations and therefore an argument for hiking, not against. And energy-driven crowding-out usually works with a one- to two-month lag, so its absence in August says nothing about October.

The configuration the Fed faces is uncomfortably simple: a supply-side inflation impulse arriving while the demand side shows no sign of cooling. That is when central banks hike, and keep hiking.