Q4 2025 real GDP grew 0.7% annualized, down from 4.4% in Q3, and was revised down 50% from the 1.4% advance estimate. Private demand carried the quarter. Consumer spending added 1.33 percentage points and private investment added 0.57, while government subtracted 1.03 and net exports subtracted 0.22. Even that private strength was concentrated. Services provided 1.25 of the 1.33 consumer boost, with health care adding 0.33 and housing and utilities 0.25. On the investment side, information processing equipment added 0.66 and inventories 0.28, so a big share of the quarter leaned on tech spending and stockbuilding rather than broad expansion.

What the deeper cross checks say

• Inventories helped, but they were not the real engine. They added 0.28 points, about 40% of the headline, yet real final sales to private domestic purchasers still grew 1.9%. Census inventory data also do not show a broad overproduction problem, with total business inventories up 0.1% in December while the inventories to sales ratio fell to 1.36 from 1.39 a year earlier. Wholesale showed the same pattern, with inventories up 0.2% and the ratio down to 1.27 from 1.30.

• The consumer kept spending, but with less cushion. Real disposable personal income rose only 0.2% in Q4, real per capita disposable income slipped, the saving rate fell to 4.0% from 4.4%, and personal interest payments rose to 591.0 billion from 585.9 billion. That looks less like fresh real income strength and more like spending through thinner savings and higher debt service.

• Autos were a huge drag, but not the whole problem. Motor vehicle output subtracted 1.37 percentage points and GDP excluding motor vehicles was 2.1%, but structures still fell 7.1%, residential investment fell 0.5%, exports fell 3.3%, and federal government spending fell 16.7%. This was not just an auto story. It was a broader loss of cyclical breadth.

Investment was narrow and prices looked firmer than the volume side

Investment was mostly a tech equipment and IP story, not broad based strength. Information processing equipment added 0.66 points, intellectual property added 0.31, and inventories added 0.28, while structures subtracted 0.21, transportation equipment subtracted 0.38, and residential investment subtracted 0.02. Census showed private construction in 2025 down 2.9% from 2024, with residential down 2.6% and nonresidential down 3.1%. Core capital goods orders were flat in January, which fits the idea that business investment outside a narrow tech pocket was soft.

The price side also looked steadier than the real volume side. Gross domestic purchases prices rose 3.8%, PCE rose 2.9%, and core PCE rose 2.7%, but market based core PCE was cooler at 2.4%. BEA also had to impute October CPI during the shutdown, which adds extra model risk. So this was not a clean deflation print, but it was a weak real growth print under still sticky price indexes.

Shutdown distortion mattered but it did not explain everything

The shutdown knocked about 1 percentage point off Q4 GDP, so adding it back only gets growth to roughly 1.7%, still far below Q3. Later cross checks stayed soft. Industrial production rose 0.7% in January and manufacturing rose 0.6%, but capacity utilization was only 76.2%, well below its long run average. The February Philly Fed shipments index was barely positive at 0.3. February payrolls fell 92,000 and federal employment was down 330,000 from its October 2024 peak.

Bottom line

The headline stayed above zero because a handful of categories did outsized work while housing, structures, exports, autos, and public demand all weakened. One major cross check is still missing because Q4 GDI was not included and will not arrive until the April 9 release with the third estimate and corporate profits. In a quarter like this, that missing income side view matters more than usual.