US second-quarter growth revised up to 2.2%, fueled by consumer spending, AI investment, and stock-market wealth effects

Consumer spending is about 70% of US economic activity, so a 3.8% annual pace adds roughly 2.7 percentage points to growth ^[5][15]. By itself, consumer spending more than covers the 2.2% headline ^[1][15]. The first-quarter pace was 0.7%, so spending growth accelerated by 3.1 points from one quarter to the next ^[5][14]. Fortune's report credits part of that to a strong stock market, one that reflects enthusiasm over artificial intelligence and enriches wealthy investors who then have more to spend ^[6].

Economists had expected the final estimate to leave the earlier 1.5% figure roughly unchanged. It added 0.7 point ^[2][12]. "The economy is increasingly reliant on AI gains and the corresponding wealth effects boosting higher-income households' spending power to fuel recent growth," said Michael Pearce, chief U.S. economist at Oxford Economics ^[10]. "The economy remains sensitive to a sudden reversal of optimism on AI," he said ^[13]. He was describing a risk, and spending in the quarter was still accelerating ^[5][13]. Neither Fortune's account nor Pearce puts a figure on the wealth effect or splits AI equipment between imported and domestic ^[6][10].

Chips push the other way in the accounts. GDP counts only domestic production, so imports are subtracted ^[7]. A foreign-made chip installed in a US data center is counted once in the 9% business-investment line and taken out again in the 1.7-point import drag, so its net contribution to output is close to zero ^[7][8][16]. The 4.6% underlying rate measures demand before trade is netted out, and imports met part of that demand ^[9][7].

A cooling in AI could reach these figures by more than one route. If companies cut chip orders, investment and imports fall together, and the headline takes a smaller hit than the investment line alone ^[16]. If AI stocks fall, the households Fortune describes lose the gains behind their spending, and they sit inside a component worth about 70% of the economy with no import line to offset the drop ^[5][6]. If neither happens, housing offers a second support: residential investment rose 2.8%, its first increase since the end of 2024, after high mortgage rates depressed the market ^[11].

I think the exposure runs mainly through stock prices and the consumer, and the chip surge is roughly neutral for measured output ^[15][16]. The counter-case is that much of the 9% investment pace is domestic spending that no import line cancels. If so, a capex pullback costs more output than the netting suggests ^[8][16]. The view is wrong if consumer spending holds its pace through a quarter in which AI-linked stocks fall.