AI Infrastructure Stocks Slide as Wall Street Weighs Data Center Buildout Risks

GE Vernova fell almost 9%, Caterpillar dropped more than 4%, Vertiv fell close to 8% and Oracle slipped almost 4%, according to CNBC. Amodei wrote that a slowdown in AI development does not mean 'halting model training or technical progress.'

RBC Capital Markets equity analyst Rishi Jaluria told CNBC that a slowdown in model development and training is likely a headwind for Oracle's cloud infrastructure business, which has driven the stock. Oracle has spent the past 18 months laying off employees in slower-growth areas and pouring billions into compute and data centers. Industrial companies such as GE Vernova, Caterpillar and Vertiv, neoclouds including Nebius and CoreWeave, and server makers Dell and Hewlett Packard Enterprise all depend on continued demand. An unnamed tech investor told CNBC, 'Any significant delay could cost them.'

MIT Technology Review reported that the spending is historic. Hyperscalers will spend about $750 billion this year on data centers, and total AI capital investments from Alphabet, Microsoft, Amazon, Meta and Oracle could exceed $5 trillion over the next four years. Total AI revenues, however, will be around $150 billion to $200 billion this year, former SEC chair and MIT Sloan professor Gary Gensler told MIT Technology Review. 'The challenge is that the spending does not have commensurate revenues yet,' he said. 'And then the question is, is that an investment that will be paid off in the future?'

Jessica Wachter, a finance professor at Wharton and former SEC chief economist, and a collaborator estimated cumulative expenditures will reach nearly $1.1 trillion through 2027. To break even by 2030 after accounting for capital costs, a 15% return and depreciation, the companies would need to raise their productivity by a factor of 2.7, MIT Technology Review reported. Wachter said that would require the kind of economic growth seen during the U.S. IT boom starting in the mid-1990s, but compressed into a few years. If they cannot meet such profit goals, she said, 'Then they will fall behind on their interest payments, and that risks bankruptcy.' The research paper concludes that if a productivity boom fails to materialize, 'the current buildout will be the largest misallocation of capital in history.'

Cash flow is already under strain. Alphabet reported nearly $120 billion in revenue in its latest quarter, but AI infrastructure spending consumed it and left a free cash deficit of about $5.9 billion, its first shortfall since Google went public in 2004, according to MIT Technology Review. Free cash flow for the group is expected to dip into negative territory. The investments could soon reach around 3% of U.S. GDP, and the risks are spreading through financial mechanisms as loans are passed along.

Debt markets are adding pressure. CNBC reported that data centers face increasing pressure, including a nationwide backlash, as a rush to secure AI debt is expected over the next six weeks. Amazon raised roughly £4.25 billion, nearly $6 billion, last week, and Alphabet tested the European market in May with about $10 billion in euro bonds. One source told CNBC that any new debt deals announced this fall will be priced at a significantly higher rate than previous issuances. 'Fixed income investors are demanding more reward,' the person said.