Data centre and cloud shares slide amid AI gloom

Dell, HPE, NetApp, Oracle, and CoreWeave all saw their shares stumble this week, amid concerns that the AI infrastructure buildout is poorly timed and the significant investment may not pay off.

The selloff follows concerns that data centre and cloud companies may be overvalued and that enterprise spending on AI isn't enough to cover the costs of the trillion-dollar infrastructure buildout. At the same time, Gartner warned that enterprise and consumer demand is failing to prop up that spending.

Dell, HPE and NetApp’s shares all declined yesterday — 11%, 8% and 7%, respectively — following a damaging report from analysts at Barclays, which downgraded the companies ratings saying they "don't see a meaningful AI inflection yet," and raised concerns about "enterprise AI timing."

Though the decline caught attention, all three firms have seen their shares climb dramatically over the course of the last year, with Dell's share price up more than 300%.

Still, the share slide spread to others in the industry, with Oracle down 9% and CoreWeave slipping 15%.

The Barclays analysts suggested that businesses are spending just 5% of IT budgets on AI, adding that the strong profits seen at data centre-adjacent companies may reflect high prices caused by component shortages rather than an industry shift to AI.

"There is growing sentiment about enterprise AI adoption creating a hardware refresh," the analysts reportedly said. "We believe the strong enterprise revenues are more of a function of memory costs than a move to enterprise AI at this time."

Not enough AI demand?

Major hyperscalers and AI leaders are expected to spend more than a trillion dollars rolling out AI infrastructure such as data centres this year alone, via a network of deals that sees cloud providers, hardware makers and AI developers investing in each other.

AI developers continue to struggle to pay for the high costs, however, with reports that OpenAI's annualized revenue is on track to be $50bn — up significantly from the year before but $20bn behind previous estimates. Anthropic's annualized run rate exceeded $65bn over the summer, and it's set to top $100bn by the end of 2026, according to Reuters reporting.

"In 2026, AI providers are expected to spend nearly $1 trillion on AI infrastructure, and are forecast to invest an additional $6 trillion between 2027 and 2030," said Will Sommer, senior director analyst at Gartner. "However, our research indicates that enterprise and consumer AI revenue will be insufficient to support profitable returns on all planned AI capital expenditures during this period."

Earning its keep

Previous research by Bain & Company suggested that the AI industry is going to need to be worth $6trn within the next few years — with $4trn in entirely new applications and services — in order to pay for the ongoing investment.

Gartner's latest analysis suggested AI data centre owners will need to generate $8.5 trillion in revenue to cover already announced capital spending, financing costs and a 12% return on invested capital. The analyst firm thinks this won’t happen until 2032 at the earliest, leading to trillions of dollars in write-downs.

"However, total enterprise and consumer AI revenue that can be attributed to AI data centers during this period is forecast to reach no more than $3.6 trillion," said Sommer. “Because most AI-native providers operate asset-light business models, we predict that less than half of AI revenue will be earned by providers that own significant AI infrastructure."

Neocloud providers — those who build specifically for AI workloads — could be the most at risk, Gartner added, as hyperscalers have more diversified revenue sources to keep them afloat. The analyst predicted that 80% of neocloud providers will "cease independent operations" by 2030 because of such pressures.