CoreWeave: A Cloud with Storm Clouds

CoreWeave. The name sounded promising enough. A cloud provider, they called themselves. But in this business, clouds often have silver linings and leaden undersides. The stock, recently, had taken a tumble. A quarter of its value gone in a week. A sharp drop, even for a market accustomed to altitude sickness.

They’re good at taking in money, that much is clear. Revenue surged. A hundred and ten percent year over year. Sixty-six point eight billion in contracted revenue. Numbers that would make a lesser company preen. But numbers, I’ve learned, can lie like a politician on election night. The backlog is impressive, yes, but it’s a promise, not a paycheck.

The problem wasn’t growth. It was the cost of it. They were burning cash faster than a cheap motel room catches fire. Operating losses were widening. The spending plans for ’26? Aggressive is a polite word. It felt more like a gamble with someone else’s money.

The Price of Altitude

They’re building something big, no doubt. A billion and six in revenue last quarter. Up from seven hundred and forty-seven million the year before. That kind of velocity is rare. The CEO, Intrator, called it a defining year. Five billion in annual revenue. Fast. Too fast, perhaps. Like a car with a stripped engine, it could sputter out at any moment.

They’re projecting twelve to thirteen billion for ’26. Another hundred and forty percent jump. Ambitious. The kind of ambition that usually ends in a faceplant. They need that revenue, and then some. Because the profit profile is…unflattering. Operating margins contracted. From positive fifteen point one percent to negative five point seven. The net loss? Four hundred and fifty-two million. A chasm, really.

Revenue was up a hundred and ten percent. Operating expenses? A hundred and sixty-two. The math wasn’t pretty. Building data centers isn’t cheap. It’s a capital-intensive business. They’re reliant on debt. A lot of it. And every dollar earned seems to vanish into interest payments. A treadmill to nowhere.

They announced capital expenditures of thirty to thirty-five billion for ’26. A staggering sum. They claim it’s to support that backlog. A long-term play. But long-term doesn’t pay the bills today. Free cash flow? Negative seven point three billion in ’25. A black hole disguised as a business model.

Risk and Valuation

The assets they’re building are worth something, certainly. Even with the debt load. But the market capitalization of thirty-eight billion? Excessive. The price-to-sales multiple is around seven. The market seemed to be pricing in perfection. A flawless transition from cash-burning builder to profitable software platform. A fairy tale, in my book.

There are structural constraints. Cloud computing is a brutal business. Constant upgrades. Relentless competition. Any macroeconomic slowdown could hit them hard. Customers pulling back on spending. Financing drying up. A domino effect. The whole thing felt…precarious.

The fundamental narrative of insatiable demand for computing power is still intact, I’ll grant them that. But CoreWeave looks risky. Worsening losses. A heavy debt load. The spending might pay off. Maybe. But at this price? It’s a gamble, not an investment. Even if they hit those ambitious targets, there’s no guarantee the margins will justify the price.

Loading widget...

I’ve seen a lot of clouds in my time. Some deliver rain, some deliver promises. CoreWeave feels like a storm brewing. A beautiful, dangerous storm. And I, for one, am staying on the ground.

Read More

2026-03-04 06:42