<
Welcome to OKRVV B2B Platform.
HOME > NEWS > Market experts

The Neocloud Doom Cycle Is a Refinancing Problem, Not a Demand Problem

Core Tip:The market has decided that CoreWeave and its neocloud peers are in trouble. It is right about the risk and wrong about the diagnosis. What is breaking is not demand for AI computing. It is the balance sheets that finance that computing. That distinction

The market has decided that CoreWeave and its neocloud peers are in trouble. It is right about the risk and wrong about the diagnosis. What is breaking is not demand for AI computing. It is the balance sheets that finance that computing. That distinction matters enormously for the B2B buyers who lease GPU capacity from these companies, because it defines what they should do about it.

The numbers behind the doom cycle label, coined by Paul Meeks of Freedom Capital Markets on CNBC, are stark. Yields on some CoreWeave bonds maturing in 2031 rose above 16 percent, and the spread on its five-year credit default swaps, which insure against default, climbed above 1,240 basis points, up 50 percent in a month. Neocloud share prices have been cut roughly in half over the past month, and last Thursday's slide contributed to the Nasdaq's steepest one-day drop in 19 months.

The balance sheet explains the nerves. At the end of June, CoreWeave held 5.5 billion dollars in cash against 35 billion dollars of debt, including 1.3 billion dollars owed at a 15 percent interest rate on a facility due in March 2028, according to its latest earnings statement. Its second-quarter filing showed three customers generating 72 percent of revenue, and a 103.7 billion dollar backlog of remaining performance obligations of which only about 41 percent is expected to convert to revenue within 24 months.

Yet the demand side keeps delivering. Analysts expect CoreWeave to report roughly 3.55 billion dollars of revenue for the latest quarter, up about 160 percent from a year earlier, according to LSEG data cited by CNBC, with Nebius expected to post sixfold growth to about 908 million dollars. Meeks, who coined the doom cycle phrase, nonetheless models CoreWeave revenue reaching 42 billion dollars in 2028 with EBITDA margins near 70 percent that year. IREN offered a preview of what good execution looks like in this environment: its AI cloud revenue more than doubled from the prior period, and the stock jumped 20 percent the day after results.

The bears' strongest argument is not about demand at all. Partners Group chief investment strategist Anastasia Amoroso told CNBC that neoclouds will be the most impacted part of the AI complex because, unlike the hyperscalers, the leading ones are unprofitable and heavily dependent on a small number of customers. Morningstar analyst Luke Yang puts the near-term bankruptcy risk low and the funding gap high: the company has the money to deliver existing contracts, but five years of growth require new capital at a moment when its beaten-down stock makes equity expensive and its debt already carries double-digit rates. The IPO window tells the same story, with Nscale's September filing still lacking an updated prospectus, SB Energy stalled, and Australia's Firmus scrapping its listing plans outright.

Politics compounds the financing problem. Voters rejected data centers in 40 of 41 ballot measures in last week's midterms, and Democrats won governors' races in Georgia, Iowa, Nevada, Ohio and Texas. CoreWeave chief executive Michael Intrator wrote on LinkedIn that the company does not anticipate any impact on existing customer commitments or its backlog from shifts in the political landscape. The credit market is pricing something less reassuring.

This is what a doom cycle actually is. Creditors worried about repayment charge more, which raises the cost of the very growth that would reassure them, which makes the next round of financing harder still. IREN co-chief executive Daniel Roberts described the shift precisely on his earnings call: capital markets have become more selective in the last few weeks, and the questions have moved from how big the contract is to whether it can be delivered, financed, and run at a profit.

For B2B buyers of GPU capacity, this converts a market question into a procurement question. Neocloud credit risk is counterparty risk. A five-year take-or-pay contract is only as good as the supplier's ability to finance the data center behind it, and the buyers who treat supplier balance sheets as part of vendor due diligence will be better positioned than the ones who treat compute as a commodity. In practice that means reading customer concentration, debt maturities and backlog conversion schedules as carefully as price per GPU hour, watching CDS spreads and bond yields as supplier health metrics, and avoiding single-supplier dependence on any one leveraged operator.

None of this means the neoclouds are finished. The correction is sorting the sector by balance sheet, which is what corrections are for, and the strongest operators are still signing the largest contracts in the industry's history. But the era in which compute buyers could ignore how their suppliers are financed is over. The doom cycle may spin further or it may not. Either way, the burden of proof has moved from the order book to the liability side of the ledger, and everyone who rents AI capacity should adjust their vendor files accordingly.

Disclaimers:
Some contents of this website are uploaded and reproduced by netizens spontaneously, which does not mean that this website agrees with its views;If it involves content, copyright and other issues, please contact within 30 days, and we will delete the content at the first time!

🔗 Shares 4 💬 Comment 0 💰 Award 0 📚 online submission
More>Related Comments
No comments yet. Why not share yours?