Related: The AI IPO Race Heats Up as Anthropic and OpenAI Eye Wall Street
The market is starting to hand Big Tech a bill for the artificial intelligence boom, and the number is staggering. The so called Magnificent Seven, joined by chipmaker Broadcom and software giant Oracle, have shed roughly 2.7 trillion dollars in combined market value over the course of June, according to a Yahoo Finance analysis, as investors take a harder look at the enormous cost of building out AI.
The selloff began as a narrower reset of the Magnificent Seven, the cluster of megacaps that has carried the market for two years, which alone had given up around 2 trillion dollars earlier in the month. It has since widened to pull in Broadcom and Oracle, two companies tied tightly to the AI infrastructure build-out, turning a wobble in a handful of stocks into a broader repricing of the entire trade. The mood is a sharp turn from earlier in the year, when investors eagerly absorbed record deals like SpaceX’s blockbuster public debut.
The reset cuts across both sides of the AI complex. On the hardware side sit Nvidia and Broadcom, the companies selling the chips and components that make the boom physically possible. On the spending side sit Microsoft, Alphabet, Amazon, Meta, and Oracle, the hyperscalers pouring hundreds of billions of dollars into data centers, while Apple and Tesla round out the group as megacaps investors have treated as AI adjacent proxies. When sentiment turns on the cost of AI, both ends of that chain feel it at once.
What changed is not the technology but the math. Nomura cross asset strategist Charlie McElligott described the hyperscalers as the funding shorts behind the AI bottleneck trades, meaning the companies doing the heavy spending are also the revenue source for the chip, memory, optical, networking, and power suppliers whose shares investors have been chasing all year. In that arrangement the spenders sit at the base of the whole structure, and when they wobble, everything built on top of them wobbles too.
The specific pressure point is free cash flow. Data centers, chips, power, and networking gear have become the entry fee for staying in the AI race, and that spending is climbing fast enough that hyperscaler free cash flow, the money left after the build-out is paid for, is projected to fall sharply. That cash is what funds the buybacks, dividends, acquisitions, and general financial cushion investors have come to expect from the largest technology platforms, and watching it shrink is what has the market nervous, because it strikes at the very thing that made these companies feel safe to own.
The deeper shift is in how the group is valued. For two years investors rewarded the promise of AI revenue still to come, content to look past the bills as long as the growth story held. Now they are pricing the cost of getting there, and the biggest AI names are no longer trading solely on future revenue but on what it takes to chase it. Whether June turns out to be a healthy correction or the start of something larger will depend on a single question, which is whether Big Tech can keep spending at this pace without losing the cash flow story that made it the most valuable corner of the market in the first place.
Related on Entrelligence: Goldman Sachs’ $110M bet on AI in finance, and AI cutting movie production costs.