Big Tech’s AI spend is shifting from a growth story to a cash-flow test. The “Magnificent Seven” plus Broadcom and Oracle have lost roughly $2.7 trillion in value in June, as investors take a harder look at the companies paying for and supplying the AI build-out.
The bill comes into view
The drop started with the biggest tech platforms, then widened to Broadcom and Oracle, two companies tied closely to AI infrastructure. The companies in focus sit on both sides of the trade: Nvidia and Broadcom are linked to the hardware boom, while Microsoft, Alphabet, Amazon, Meta, and Oracle are tied to the spending boom. Apple and Tesla remain part of the megacap growth group investors have treated as AI-adjacent.
The practical issue is cash. Data centers, chips, power, networking gear, servers, and cloud infrastructure are becoming the cost of staying competitive in AI. Nomura’s Charlie McElligott described the big cloud platforms as “the funding shorts” behind bottleneck trades in areas such as memory, chips, optical equipment, networks, servers, and power infrastructure. The same companies spending heavily on AI are also the revenue source for many suppliers investors have been chasing.
That makes free cash flow, the money left after capital spending, a central metric. Yahoo Finance cited Nomura projections that hyperscaler free cash flow is expected to fall sharply as AI spending gets more expensive. That cash is what companies can use for buybacks, acquisitions, dividends, and future investments, so a smaller cushion changes the financial story investors have long attached to the biggest tech platforms.
The open question is whether the largest AI spenders can keep funding data centers, chips, power, networking gear, and cloud infrastructure while preserving the cash flow investors have come to expect.
