CoreWeave Has Fallen 49% From Its 52-Week High. Is the Beaten-Down AI Stock a Bargain or a Value Trap?

Few stocks capture the AI infrastructure boom -- and its risks -- quite like CoreWeave (NASDAQ: CRWV). The company rents out the high-end computing power that trains and runs AI models, and demand for it has been ferocious. Yet as of this writing, shares sit near $77 -- about 49% below the 52-week high of $153.20.

The latest leg down has a specific cause. Earlier this month, reports surfaced that Meta Platforms plans to build its own AI cloud business and sell excess capacity to outside customers. Meta happens to be one of CoreWeave's largest customers, so the news raised an uncomfortable possibility: one of the company's biggest buyers may be about to become a competitor.

Shares have fallen for four straight sessions since. For dip buyers, a decline like this is tempting. But a lower price only helps if the business underneath it can support the stock. So, which is this, a bargain or a value trap?

Image source: The Motley Fool.

Staggering growth

The top line leaves no doubt about demand. In the first quarter of 2026, CoreWeave's revenue more than doubled, rising 112% year over year to $2.1 billion. That followed 168% growth for full-year 2025, so even as the rate cools, the company is still expanding at a pace almost no business its size can match.

The backlog is just as eye-catching. CoreWeave signed more than $40 billion of new contracts during the quarter, lifting its revenue backlog to $99.4 billion. That figure dwarfs the roughly $12.5 billion in revenue it expects to generate this year, and on paper it offers years of visibility.

Management still guides for $12 billion to $13 billion in revenue this year, with the exit rate climbing toward $18 billion to $19 billion annualized. Few companies grow into their promises this fast.

The physical footprint is scaling to match. The company now holds more than 3.5 gigawatts of contracted power and recently surpassed 1 gigawatt of actual capacity, a milestone only a handful of cloud operators have ever reached.

The trouble is what all of this costs. CoreWeave is borrowing heavily to buy graphics processing units, lease data centers, and secure power, and the bills are climbing faster than sales.

Its first-quarter net loss more than doubled to $740 million, from $315 million a year earlier, and it widened from a $452 million loss in the prior quarter. Net interest expense alone more than doubled year over year, to $536 million, as the debt load grew.

The spending, meanwhile, is only accelerating. Management expects capital expenditures of $31 billion to $35 billion this year, against that same roughly $12.5 billion in revenue. The demand is not in doubt. The economics are.

The Meta problem and the price

The Meta news sharpens the risk considerably. CoreWeave holds a roughly $21 billion agreement with Meta that runs through 2032, so one of its largest customers is reportedly building the very capability CoreWeave sells.

To be fair, that agreement still binds Meta as a paying customer for now, which limits the near-term damage. CoreWeave's customer base is broadening, too, with recent deals signed alongside AI labs such as Anthropic and Cohere.

But those customers share a trait -- they are deep-pocketed enough to build their own capacity over time, exactly as Meta is now doing. When one of your biggest buyers decides it can do the job itself, the long-term pricing power of the whole industry arguably starts to look shakier.

Then there's the valuation. CoreWeave isn't profitable, so there's no price-to-earnings ratio to lean on. Measured against sales, its roughly $42 billion market capitalization works out to about 3.3 times this year's expected revenue.

That might look reasonable for a fast-growing software company. But CoreWeave isn't software. It's a capital-intensive, heavily indebted infrastructure business with no profits in sight and a customer list that now includes its newest rival.

So is the sell-off an opportunity or a warning? To me, it's a warning. CoreWeave is executing an ambitious plan in a booming market, and its top-line growth is hard to fault. But the road to durable profits runs through tens of billions in spending, a mountain of debt, and pricing power that its own customers are working to erode. That is more uncertainty than I want to underwrite. I'd stay on the sidelines and look for AI exposure where the path to profitability is clearer.

Daniel Sparks and his clients do not have positions in any of the stocks mentioned. The Motley Fool has positions in and recommends Meta Platforms. The Motley Fool has a disclosure policy.