Meta shares drop after report suggests blueprint for a cloud infrastructure push

Meta Platforms shares slid sharply Thursday morning, with the stock down as much as about 4% early in the session, after investor attention turned to a possible expansion beyond social media into cloud infrastructure and AI-related compute services.

The renewed discussion was tied to a report that Meta may be developing “blueprints” for a cloud infrastructure business. The idea described is to create a platform that could sell Meta’s excess computing power and give outside customers greater access to Meta’s popular artificial intelligence models.

Analysts cited in coverage framed the potential as a double-edged sword. On one hand, selling compute could improve earnings, with one set of estimates suggesting Meta’s earnings per share (EPS) might increase by about 20% for each gigawatt (GW) of compute power the company sells. A gigawatt is a measure of electrical power capacity, and the projection implies the business could scale quickly if demand materializes.

On the other hand, the same coverage pointed to a cost curve that worried investors. Wolfe Research estimates referenced in the article said Meta’s capital expenditures, or capex (spending on long-term assets like data centers and equipment), could rise to about $200 billion in 2027, up from a previous estimate of roughly $160 billion.

The capex estimate matters because it would likely require a financing plan. The coverage said the higher spending would probably require a capital raise, adding another layer of uncertainty on top of the competitive and execution risks that come with entering or expanding in large-scale cloud services.

The plan, if pursued, would put Meta in direct competition with entrenched cloud infrastructure providers, including Amazon Web Services, Microsoft Azure, and Alphabet’s Google Cloud, while also positioning it against newer “neocloud” operators such as CoreWeave and Nebius Group. Even if Meta’s starting point is excess capacity, building a customer-facing platform at scale can be capital intensive and operationally complex.

Meta did not provide any details in the material cited here about the timing, size, or structure of a cloud initiative, and the discussion appears to be based on reporting and investor interpretation of what Meta could be planning. That gap, together with the prospect of higher capex, appears to have driven the fast market reaction.

For the near term, investors are likely to watch for any follow-up from Meta that clarifies whether the company is studying the business line, how it would monetize compute, and whether it would change guidance or financing plans tied to capex and AI infrastructure.