Business Inventories +0.8% vs +0.3% Est — Stock Built Three Times Faster Than It Sold

Fundamentals · 2026-09-16

Business inventories rose 0.8% m/m in July to $2,764.7 billion against a 0.3% consensus — nearly triple the forecast — while sales rose just 0.3%, so stock built almost three times faster than it moved. June was revised to +0.1%, a change the Census Bureau flags as not statistically distinguishable from zero. On the year inventories are +3.8% against sales +8.9%, which pulls the inventories-to-sales ratio down to 1.30 from 1.37 a year ago. Merchant wholesalers are the pressure point: inventories +1.3% m/m and +5.7% y/y, the fastest of the three sectors. Retail sales fell 0.7% in July while retail inventories rose 0.8%. Manufacturers +0.4% m/m. The auto inventories-to-sales ratio rose to 1.91 from 1.85, and building materials to 2.15. July data, a seven-week lag. Next release 15 October.

What It Changes

• It confirms the inventory overhang this desk has been tracking, and locates it one layer back from the shelf. Wholesale inventories are up 5.7% on the year against retail at 3.9% and manufacturers at 2.0%. The stock is piling up between the factory and the store.
• The monthly split is the uncomfortable part. Inventories +0.8% against sales +0.3% is a build outpacing demand by nearly three to one, and inside it retail stock rose 0.8% while retail sales fell 0.7%. That is the shape of an involuntary build.
• But the annual picture says lean, not glutted, and the ratio is the reason. At 1.30 the inventories-to-sales ratio is below last July's 1.37 and comfortably inside its historical range. Nothing here looks like 2022.
• Read the ratio with care, because both halves are nominal. Sales are up 8.9% on the year in dollars, and a large part of that is energy prices rather than volume — wholesalers alone are +13.0%. A nominal ratio falls when prices rise faster than quantities, so some of this "leanness" is an artifact of the price surge, not a genuine drawdown.
• It changes nothing today and is stale besides. This is July data released seven weeks later, on an FOMC afternoon, and we already know what August did — retail sales rebounded 1.2% this morning. The involuntary build it implies has probably already been absorbed.

Impact

• USD — Neutral — business inventories has no history of moving the dollar, it is second-tier, it is seven weeks old, and it is landing four hours before a rate decision that is 92% to 93% priced.
• The one channel that matters is arithmetic rather than sentiment: inventory investment is a direct line in GDP, and a 0.8% build against a 0.3% sales gain adds to third-quarter growth even when it reflects goods nobody bought.
• US Indices (ES / NQ / YM / RTY) — Slight bearish — an involuntary build is a margin problem before it is a growth problem, and it shows up in discounting rather than in this week's tape.
• RTY and the consumer-discretionary complex carry it. The categories with rising inventories-to-sales ratios are autos at 1.91, building materials at 2.15 and furniture and appliances at 1.60 — all rate-sensitive, all domestically exposed, and all the small-cap end of the market rather than the mega-cap end.
• Futures came into the session +0.2% to +0.4% across Dow, S&P and Nasdaq-100 in relief after the S&P's fifth decline in six sessions, and nothing in this release will disturb that before 14:00.
• Gold (GC) — Neutral — no transmission mechanism from this release to gold that is worth writing down.

Inside The Number

Manufacturers' and trade inventories totaled $2,764.7 billion at the end of July, up 0.8% (±0.1%) from June and up 3.8% (±0.5%) from July 2025. Consensus was 0.3%. Sales — the combined value of distributive trade sales and manufacturers' shipments — came in at $2,120.7 billion, up 0.3% (±0.2%) on the month and 8.9% (±0.3%) on the year.

That gap is the release. Inventories grew at nearly three times the rate of sales in a single month. When stock builds faster than it sells, one of two things is happening: businesses are stocking deliberately ahead of expected demand, or demand disappointed and the goods stayed put. The sector detail says the second, at least in retail.

Retail is where the involuntary build shows. Retail sales fell 0.7% in July while retail inventories rose 0.8%. Nobody orders more stock into a falling month on purpose; that is the gap between what was ordered weeks earlier and what actually sold. It is also consistent with what we already knew — this morning's advance report put July retail and food services sales at -0.5%, revised up from -0.6%, and the retail-only measure excluding restaurants at -0.7%.

Wholesalers are the sector actually accumulating. Merchant wholesaler inventories rose 1.3% on the month and 5.7% on the year — the fastest of the three sectors on both horizons — against retail at 0.8% and 3.9%, and manufacturers at 0.4% and 2.0%. This is the overhang the desk flagged after the wholesale report, now confirmed in the combined data: the stock is not on the shelf and it is not in the factory, it is in the warehouse between them.

The inventories-to-sales ratio says the opposite, and both readings are true. At 1.30, the total business ratio is unchanged from June and down from 1.37 in July 2025. By sector: manufacturers 1.47 from 1.56, wholesalers 1.20 from 1.28, retailers 1.27 from 1.28. Every one of them is leaner than a year ago. On that measure there is no glut at all.

Here is why the two readings can coexist, and it is worth being precise about. The ratio is nominal over nominal. Sales are up 8.9% in dollars, and a substantial share of that is price rather than volume — merchant wholesaler sales are up 13.0% on the year, and the wholesale layer is where petroleum and energy products sit. Import fuel prices are up 26.8% over the year and import natural gas has more than doubled. When the denominator inflates faster than the numerator, the ratio falls without a single extra unit leaving a warehouse. So treat 1.30 as flattered. The honest statement is that inventories look lean in dollars and less lean in goods, and this release cannot separate the two because Census explicitly does not adjust it for price changes.

Inside retail, the ratio detail is where the real signal is, because it is category-specific. Three categories have a higher inventories-to-sales ratio than a year ago: motor vehicles and parts at 1.91 against 1.85, building materials at 2.15 against 2.10, and furniture, home furnishings, electronics and appliances at 1.60 against 1.54. Auto retail inventories are +5.2% y/y while auto sales managed just 2.1%; building materials inventories are +8.1% y/y against sales of 5.1%. Those are the two clearest accumulations in the report, and both sit in the most rate-sensitive corners of the consumer economy — which is exactly where you would expect stock to strand with the 10-year at 2007 levels.

The categories going the other way are the discretionary ones. Clothing inventories were flat on the month and the ratio fell to 2.11 from 2.16. General merchandise inventories fell 0.1%, with the ratio down to 1.24 from 1.27. Department stores sit at 2.63 against 2.68. Where households have actually been spending — and this morning's advance report had restaurants up 1.2% and nonstore up 2.6% in August — the shelves are clearing.

One caveat on the vintage, and it matters more than usual. This is July data published on 16 September, a seven-week lag, and it is the most backward-looking release the feed covers. We already know what happened next: August retail sales rose 1.2% and core rose 1.4%, both well above consensus, published four and a half hours before this one. A July build followed by an August sales surge is a build that gets absorbed. The alarming reading of this report was very likely true for about three weeks.

And one on the data itself. The Census Bureau notes in this release that, "due to schedule and resource constraints," the revised historical data and seasonal adjustment models for the Manufacturers' Shipments, Inventories and Orders survey will remain unchanged for the remainder of 2026. Frozen seasonal factors get progressively less accurate as the year runs on, so treat the manufacturing component with a little extra skepticism through December.

The Internals

The three sectors, seasonally adjusted:

Sector · Inventories m/m · Inventories y/y · Sales m/m · Sales y/y · Read

Total business · +0.8% · +3.8% · +0.3% · +8.9% · Build outpaced sales nearly 3 to 1

Merchant wholesalers · +1.3% · +5.7% · +0.8% · +13.0% · Fastest accumulation of the three

Retailers · +0.8% · +3.9% · -0.7% · +5.0% · Stock rose while sales fell

Manufacturers · +0.4% · +2.0% · +0.8% · +8.3% · The leanest sector on both horizons

Inventories-to-sales ratios, and every one is lower than a year ago:

Sector · Jul 2026 · Jun 2026 · Jul 2025 · Read

Total business · 1.30 · 1.30 · 1.37 · Flat on the month, leaner on the year

Manufacturers · 1.47 · 1.47 · 1.56 · The largest annual improvement

Retailers · 1.27 · 1.25 · 1.28 · Ticked up on the month

Merchant wholesalers · 1.20 · 1.19 · 1.28 · Lowest of the three, and rising again

Dollar levels, seasonally adjusted, in billions:

Measure · Jul 2026 · Jun 2026 · Jul 2025

Total business inventories · 2,764.7 · 2,742.5 · 2,663.0

Total business sales · 2,120.7 · 2,113.7 · 1,946.6

Manufacturers inventories · 966.9 · 963.4 · 948.2

Merchant wholesaler inventories · 958.9 · 946.7 · 906.9

Retailer inventories · 839.0 · 832.5 · 807.9

Retail inventories by category, with the ratio that tells you whether it is a problem:

Category · Inventories m/m · Inventories y/y · I/S ratio now · I/S ratio a year ago · Direction

Building materials and garden equipment · +0.7% · +8.1% · 2.15 · 2.10 · Worsening, and the fastest annual build

Motor vehicle and parts dealers · +0.8% · +5.2% · 1.91 · 1.85 · Worsening

Furniture, home furnishings, electronics, appliances · +0.4% · +3.9% · 1.60 · 1.54 · Worsening

Food and beverage stores · +0.4% · +3.4% · 0.77 · 0.75 · Marginally worse, lowest ratio in retail

Retail trade, total · +0.8% · +3.9% · 1.27 · 1.28 · Broadly flat

Retail excluding motor vehicles and parts · +0.8% · +3.2% · 1.09 · 1.12 · Improving

Clothing and accessories stores · 0.0% · +2.3% · 2.11 · 2.16 · Improving

General merchandise stores · -0.1% · +1.4% · 1.24 · 1.27 · Improving

Department stores · +0.7% · +0.4% · 2.63 · 2.68 · Improving, highest ratio in retail

Where The Stock Is Piling Up

The warehouse layer is the answer, and it has been for months. Merchant wholesaler inventories rose 1.3% in July and are up 5.7% on the year — faster than retail at 3.9% and more than double manufacturers at 2.0%. Wholesalers now hold $958.9 billion, within $8 billion of the manufacturing sector's entire stock and growing nearly three times as fast. Goods are leaving factories, reaching distributors, and stopping there.

Autos are the single clearest accumulation. Retail motor vehicle and parts inventories are up 5.2% on the year while the same dealers' sales rose just 2.1%, and the inventories-to-sales ratio climbed to 1.91 from 1.85. At 1.91 months, autos carry the heaviest stock burden of any major retail line except building materials, clothing and department stores — and unlike those three, the auto ratio is going the wrong way. Import prices for automotive vehicles and parts were flat in August and up only 0.8% on the year, so this is not a price story; it is unsold metal.

Building materials is the fastest build in retail. Inventories +8.1% on the year against sales of 5.1%, with the ratio at 2.15, the highest of any major retail category and up from 2.10. That line also fell 0.2% in this morning's August advance report — the only category of thirteen that declined. A housing market frozen by a 5% 10-year leaves lumber and fixtures sitting in yards.

Furniture and appliances is the third. Inventories +3.9% on the year, ratio up to 1.60 from 1.54. The same rate-sensitivity, the same mechanism.

Where stock is clearing is just as informative. General merchandise inventories actually fell 0.1% on the month, with the ratio down to 1.24 from 1.27. Clothing was flat and its ratio improved to 2.11 from 2.16. Department stores improved to 2.63 from 2.68. Retail excluding autos has a ratio of 1.09 against 1.12 a year ago. Strip the two or three rate-sensitive categories out of retail and there is no inventory problem at all.

Manufacturers are the cleanest sector in the report. Inventories up just 0.4% on the month and 2.0% on the year, the slowest of the three, with the ratio improving to 1.47 from 1.56 — the largest annual improvement of any sector. Factory shipments rose 0.8%, twice the rate of the inventory build. Whatever is going wrong in this report, it is not going wrong at the factory.

And a note on what the sales column is really measuring. Total business sales +8.9% year over year sounds like a boom. Merchant wholesaler sales at +13.0% are doing most of that work, and the wholesale layer is where petroleum, fuel and energy products are distributed. Census states plainly that none of this is adjusted for price changes. A year in which import fuel prices rose 26.8% and import natural gas more than doubled will inflate wholesale dollar sales without a single additional barrel moving.

Against This Morning's Open

• This is the third release missing from The Open's calendar today — The Open tabled NAHB at 10:00, the FOMC at 14:00 and the press conference at 14:30, omitting both 08:30 prints and this 10:00 one.
• The 10:00 omission is the more instructive of the three. The slot was not empty on the calendar — NAHB was there — so this was not a missing time, it was a missing second entry at a time already covered. One row at a time reads as complete when it is not.
• None of the three would have been the session's fulcrum, and The Open was right that the dot plot and the vote count decide today. The issue is completeness of the list, not the ranking inside it.
• Where this release does connect to The Open's read is the 10-year at 4.996%, having touched 5.04% intraday. The three retail categories accumulating stock — autos, building materials, furniture — are precisely the rate-sensitive ones, so that yield level and this inventory build are the same story told from two ends.

What This Sets Up

• Next release — Thursday 15 October, covering August, the same morning as the next advance retail report and the next Empire State survey.
• That August report should show the build being absorbed. Retail sales rose 1.2% in August with 12 of 13 categories higher. If retail inventories still rise faster than sales in the August data, the involuntary-build reading survives its best test.
• Two benchmark revisions land before the picture settles. Revised retail estimates are scheduled for 28 September, feeding into the 15 October report; revised wholesale estimates for 26 October, feeding into the 17 November report. Both incorporate the 2023 and 2024 Annual Integrated Economic Survey, and both can move the level and the trend at once.
• Watch the auto ratio specifically. At 1.91 and rising against a year ago, it is the one category where stock and sales are diverging on both horizons. A third consecutive monthly increase would make it the clearest inventory problem in the US consumer economy.
• Watch whether wholesale keeps outpacing retail. 5.7% against 3.9% on the year is the overhang. If that gap widens again in August, the goods are backing up rather than clearing.

What Is This?

• What it is: Manufacturing and Trade Inventories and Sales, published monthly by the Census Bureau, is the combined inventory and sales picture for the whole goods economy. It stitches together three separate surveys — the Monthly Retail Trade Survey, the Monthly Wholesale Trade Survey and the Manufacturers' Shipments, Inventories and Orders survey — into one set of totals. The wholesale and manufacturing components are unrevised from their own earlier reports, while the retail component is revised and expanded from the advance report. It is the last of the month's inventory releases and the most complete, which is also why it carries a seven-week lag: July's figures publish in mid-September.
• Why it matters: Inventory investment is a direct component of GDP, so a build adds to measured growth whether or not anyone wanted the goods. More usefully, the inventories-to-sales ratio is one of the oldest cycle indicators there is: it says how many months of sales a business is holding in stock. A rising ratio means goods are accumulating faster than they move, which historically precedes production cuts, discounting and margin pressure. A falling ratio means the opposite. The current reading matters because the desk has been tracking an overhang forming at the wholesale layer, and this release is where all three layers can be compared directly.
• How to read it: Four cautions, and the first is the big one. Nothing here is adjusted for price changes — Census says so explicitly — so both halves of the inventories-to-sales ratio inflate with prices, and in an energy shock the ratio can fall while real stock is flat or rising. A build is ambiguous by construction: deliberate stocking and unsold goods look identical in the data, and only the sales column beside it tells you which. The ratio's level is more informative than its change, and it varies enormously by category — grocery runs near 0.77 because food spoils, department stores near 2.63 because clothing does not. And the lag is severe: by the time this prints, a newer month of retail sales is usually already public, so check what happened next before drawing a conclusion.

Want the full explainer? Economic News Events, Explained breaks down this release and every other one we cover, in plain English.

_For informational purposes only. Not investment advice._