America's largest tech corporations are pouring hundreds of billions into artificial intelligence, with projected spending reaching $765 billion this year and nearly $1.2 trillion by 2027. This massive capital outlay comes even as several industry giants report negative cash flows, demonstrating the speculative nature of the current investment frenzy.
Amazon, the largest of the four "hyperscalers," has increased its capital spending forecast to $220 billion for the year. The company reported a negative free cash flow of $7.6 billion for the trailing 12 months. Just two days earlier, Meta disclosed a 91% drop in cash generation compared to the previous year. Alphabet, another tech titan, announced within the past week that its cash flow had turned negative for the first time on record. Alphabet finance chief Anat Ashkenazi informed analysts that free cash flow would remain "under pressure" as the company pursues the "AI opportunity." These earnings reports confirm that AI investments are distorting corporate balance sheets, even as executives continue to promote the future returns from new data centers and the specialized chips required to run them.
Capital's Gamble
A primary driver of these escalating costs is the global memory crunch. This shortage is fueled by intense demand for AI processors, which rely on memory supplied by a concentrated group of vendors. Tesla CEO Elon Musk described memory pricing as "insane" on his company's earnings call within the past week. Amazon CEO Andy Jassy echoed this sentiment, stating that the "inflated price" of memory chips directly pushed his company's capital expenditure guidance higher. Investors have reacted with sharp divergence. Tesla and Alphabet shares both sank within the past week after reporting negative cash flow and projecting increased spending. Meta's stock fell two days ago following a weak forecast and ongoing uncertainty regarding its AI monetization strategy.
The Cost Shift
Apple, while spending far less than its peers, faces significant exposure to this memory crisis. Memory technology is a critical component in every consumer device the company produces. Apple has already raised prices on its Macs and iPads, and analysts anticipate further iPhone price hikes later this year. Apple CEO Tim Cook, who will step down on September 1, attributed a weaker-than-expected forecast issued one day ago to "supply constraints." Cook stated, "If you look beyond September, we see the market pricing for memory continuing to increase, which could drive an increasing impact on our business. And we're continuing to evaluate this." He added that the problem isn't expected to ease this year. For Apple, the memory shortage translates into a revenue problem, preparing the company for potentially weaker consumer demand as prices rise. For the hyperscalers, it represents a major cost hurdle for the memory-intensive AI systems they acquire from Nvidia.
Market's Judgment
Despite the widespread capital expenditure, the market's reaction has been mixed. Microsoft, however, experienced its best day on the market since 2008 after reporting better-than-expected results and higher capital expenditure guidance. Wells Fargo analysts, who recommend buying Microsoft shares, wrote that "MSFT has room to meaningfully re-rate." This rally reduced Microsoft's stock drop for the year to approximately 7%. Apple shares slid one day ago after its Q3 earnings, with the memory shortage weighing heavily on its outlook. In contrast, Amazon's surging cloud growth propelled its stock higher. Mark Mahaney, an analyst at Evercore ISI, noted one day ago that Amazon Web Services' "revenue growth [is] dramatic, but the profitability is rising." He called it "just the breakout that the stock needed." Wedbush analysts, in a note issued today, described Amazon's report as the "cleanest beat" among the hyperscalers, praising management's explanation of how it plans to achieve returns on its capital expenditure. They highlighted this clarity as a factor in the differing share reactions between Google and Amazon, despite what they viewed as "similarly strong fundamental prints with raises in capex." Across the megacap sector, few stocks, excluding Micron, are experiencing breakout years, despite healthy revenue growth. This muted market response reflects growing skepticism over whether the massive AI buildout, increasingly financed by debt, will ultimately yield the expected returns for capital. Chinese AI labs are releasing new, cheaper "open-weight models" that are narrowing the performance gap with market leaders OpenAI and Anthropic. These open models can be downloaded, modified, and hosted on user-chosen infrastructure, posing a direct threat to the nearly $1 trillion valuations of OpenAI and Anthropic in the private market. Dana Harlap, investment strategist at JPMorgan Chase, questioned in a report within the past week, "Is it all one big AI trade?" Harlap observed that the market is "more critical — and more discriminating — across hyperscalers as investors try to separate AI winners from losers." The long-term success or failure of these hyperscalers to generate an "acceptable return on investment" on their heavy capital expenditures will likely dictate the returns of the entire AI ecosystem, underscoring capital's relentless pursuit of profit.