CoreWeave beats on revenue and lifts guidance, easing AI slowdown concerns

CoreWeave reported third-quarter revenue of $3.66 billion, up from $1.37 billion a year earlier and above the $3.55 billion average analyst estimate compiled by LSEG, according to Reuters. The Nvidia-backed AI cloud provider also lifted its annual revenue guidance to between $13.0 billion and $13.3 billion, from a previous range of $12.4 billion to $13.2 billion. Shares rose more than 7% in extended trading after a sharp slump last month that saw the stock crash more than 40% as investors fretted about the company's heavy borrowing to fund purchases of chips, land and energy for AI data centers.

CEO Mike Intrator said in a statement that "customer contracts come first, and investment follows," adding that because CoreWeave's investments are tied to committed demand, lenders have something concrete to lend against and customers typically help fund the buildout through prepayments. The company has never turned a profit, according to LSEG data, and had more than $50 billion in debt as of June. Yields on its 2030 bonds are currently at about 15.2%, while its five-year credit-default swaps trade at 1,135 points.

The selloff triggered a broader crisis of confidence in AI stocks, amplified after the Democrats' sweeping win in the US midterms as investors assess potential for tighter regulation in America's data-center build-out. Recent earnings from hyperscalers also hinted at delays in data-center projects and a slower ramp in spending. CoreWeave operates about 52 data centers globally supported by advanced Nvidia chips with a global active capacity of over 1.7 gigawatts — enough power to supply more than 1.2 million US homes. Its customers include Microsoft, Meta, OpenAI and Anthropic. Microsoft has tightened oversight of some data-center projects from neocloud companies Nebius, IREN and Nscale as it focuses on expanding its own data centers, Bloomberg News reported last week. Brokerage Truist said market sentiment had become disconnected from improving economics at AI cloud providers, citing higher pricing and increasing prepayments from customers that will reduce dependence on external capital. Intrator said in a memo last month that the firm's deal backlog is large enough to more than cover its debt obligations.