Growth
Retail Sales — July 2026
Released August 14, 2026 · 8:30 AM ET · Source: U.S. Census Bureau
The read · narrated
The read
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.
The numbers
| Measure | Latest | Trend |
|---|---|---|
| Retail sales (dollars) | −0.6% m/m | ▼ biggest drop since May 2025 — autos and online did it |
| Real retail sales (volume) | −0.7% m/m | ▼ weakest in twelve months — June and July average flat vs May |
Advance Monthly Sales for Retail and Food Services, U.S. Census Bureau, July 2026 release CB26-131 (August 14, 2026): retail & food services total (category 44X72), ex–motor-vehicle-&-parts (44Y72), ex–motor-vehicle-&-parts-&-gasoline (44W72), and the category cuts — motor vehicle & parts dealers (441), nonstore retailers (454), gasoline stations (447), electronics & appliance (443), health & personal care (446), clothing (448), food services & drinking places (722) — pulled seasonally adjusted from the Census API with the release’s aggregation identities verified in code. Displayed headline, ex-auto and category figures are the official published changes from release Table 2; the contribution arithmetic is computed from the seasonally adjusted dollar levels. *Advance-estimate caveat (release text): the ex-auto (−0.3%) and ex-auto-&-gas (−0.2%) cuts each carry a 90% confidence interval that includes zero — the Census Bureau notes there is insufficient statistical evidence to conclude those changes differ from zero. The headline −0.6% (±0.4) does not include zero. Revisions: this release left the June monthly change unrevised at +0.2% while lowering its level from $768,553M to $768,072M, and revised May from +1.0% to +0.9% ($766,876M → $766,192M). Two figures our July 16 read aired have since moved: June nonstore retailers, reported then as up nearly 2%, is now +0.9%; June motor vehicle & parts, also reported as up nearly 2%, is now +2.4%. Breadth is counted across the thirteen major retail groups Census publishes (441–454 and 722) on the released one-decimal figures: July was seven up, four down and two unchanged (food & beverage and sporting goods each publish as 0.0); June was seven up, five down and one unchanged. Counting instead on unrounded levels would move two categories off zero and is not what the release shows. Concentration: motor vehicle & parts (−$2,560M) and nonstore retailers (−$3,148M) together subtracted 0.74 percentage points from the month, against a total decline of 0.58 points; the remaining eleven categories added 0.16 points. “Volume” is real (inflation-adjusted) retail & food services sales — the Census series deflated by CPI, published via FRED (RRSFS); it is not part of the advance release. Its −0.7% July change, its standing as the weakest of the last twelve months, and the three-month average shift (+0.5% a month across February–April to +0.2% across May–July) are computed on the current vintage. The two-month check — June and July averaged, $765,837M, against May’s $766,192M, a difference of 0.05% — is computed from the same seasonally adjusted headline levels. Consumer prices rose 0.1% m/m in July (BLS, August 12), which is why the dollar and volume figures land within a tenth of each other this month; in May and June they differed by more than two tenths. Control group: the media-convention control group (total less autos, gasoline, building materials and food services) fell 0.5% in July. Nonstore retailers are 32.9% of that aggregate and contributed −0.75 points to it; excluding nonstore, the control group rose 0.4%. The control group is not a breadth measure — its decline is accounted for by e-commerce rather than being evidence of weakness beyond it. Formula and components are stated here because the control group is not a Census-published series. The saving rate is the BEA personal saving rate (via FRED): 2.7% in June 2026, a fifth consecutive monthly decline from January’s 4.4% and the lowest reading since June 2022; it lags retail sales by one month. The 2-year Treasury yield is the constant-maturity series (Treasury.gov daily par yields, cross-checked via FRED DGS2): 4.25% Monday August 10 → 4.20% after Wednesday’s CPI → 4.15% after Thursday’s PPI and jobless claims → 4.17% at Friday’s August 14 close — down 7 basis points across the two inflation prints, up 2 on this one, and down 8 on the week. The tightening reference is our August 13 PPI read. Next Retail Sales release: September 16, 2026, 8:30 AM ET — the same morning as the FOMC rate decision at 2:00 PM ET.