Growth

Retail sales fell six tenths of a percent in July. And more store categories sold more than sold less — seven up, four down. So how does spending fall in a month when most of retail went up?

The headline: seven hundred sixty-three point six billion, the biggest monthly drop since May of last year. Prices barely moved in July, so volume says the same thing — down about seven tenths. In May and June, dollars and volume disagreed, and that gap was the story. Not this month.

So which is it — a real drop, or June's spike coming back? Here's the check. Average June and July together, compare to May, and you get flat. The two months cancel. Then run the trend underneath: spending after inflation was growing half a percent a month earlier this year. Now it's two tenths. Nothing broke in July. Over two months, it stopped moving.

Now the split. Car dealers fell almost two percent. Online sellers fell more — down two point two percent, their worst month since January last year. Everywhere else, mostly up. Clothing rose almost two percent. Restaurants rose for a third straight month.

Put the dollars on it. Those two took three quarters of a point off the total. The other eleven added two tenths. Last month we asked whether June's narrow gain becomes weak. It didn't — it reversed. The two categories that made June gave it back. And June's online gain has since been revised down by roughly half.

One layer down. Through June, households saved two point seven percent of their income — five straight monthly declines, the lowest since mid twenty twenty-two. That's the cushion this spending has been running on.

So what did the bond market do with a weaker consumer? The two-year Treasury fell seven basis points across this week's two inflation reports. Friday, on this one, it rose two. This week we said the case for tightening hadn't gotten stronger. Demand came in soft, and the two-year still moved the other way.

The watch is September sixteenth. The next retail report prints at eight thirty that morning. The Fed announces their rate decision at two. On this data, a move in either direction isn't justified.

Here’s a number that landed right where it started. Second-quarter growth: one and a half percent, same as a month ago. The fine print calls it a downward revision of less than a tenth. The pieces underneath moved more.

That was the second estimate; each quarter gets published three times as more data comes in. The official line: an upward revision to consumer spending, partly offset by an upward revision to imports. Two moves, opposite directions.

Here’s the arithmetic. The consumer went from adding two point one two points to two point three one. The subtracting lines — inventories, trade, government — went from taking off one point eight two to two point oh two. Demand up two tenths. Drags up two tenths.

The consumer revision has a shape. Services up, led by health care — hospitals and doctor visits — from a survey that hadn’t landed in July. Goods down: recreational goods, computing equipment, gasoline, once retail and energy data got revised.

Now the offset. Imports subtract from growth by arithmetic, not by weakness. But this one wasn’t Americans buying more from abroad. It was led by a territorial adjustment for Puerto Rico — which these accounts treat as the rest of the world.

Which is why the same release carries a second measure: final sales to private domestic purchasers. Demand inside the country, inventories and trade stripped out. Revised up three tenths, to four point two percent — the line our July read called what the Fed watches.

That read closed on two questions. Whether inventories and trade kept subtracting — they did, by more. And whether the consumer could keep spending — that got revised up.

One more thing moved. The inflation gauge inside these same accounts — consumer prices excluding food and energy — was revised up two tenths, to three point six percent annualized.

So a release that left the headline alone moved both things the Fed weighs, demand and prices, the same way. Up. On the morning, the two-year moved two basis points. Our claims read said the case for easing wasn’t building. This didn’t build it.

The catch: a second estimate is still an estimate. The third lands September thirtieth, with the annual update that rewrites history. But the Fed meets the sixteenth. That’s the read they’ll have.