Neocloud Stock Collapse: Buying Opportunity or Run for the Hills?

Neocloud stocks, including CoreWeave (NASDAQ: CRWV), Nebius Group (NASDAQ: NBIS), IREN (NASDAQ: IREN), and Hut 8 (NASDAQ: HUT), took a major hit after technology research outfit Gartner predicted that 80% of companies in the space would cease to exist by 2030. Data center real estate operators, like Applied Digital (NASDAQ: APLD), also took a hit on the report.

Neocloud operators are next-generation cloud computing companies whose infrastructure is designed specifically to run artificial intelligence (AI) workloads, like large language model (LLM) training and inference. They primarily lease computing capacity powered by Nvidia graphics processing units (GPUs), while typically pairing it with high-speed networking and storage.

Image source: Getty Images.

While traditional cloud computing operators, like Amazon Web Services, Microsoft Azure, or Alphabet's Google Cloud, offer a broad range of services, neoclouds are solely focused on AI workloads. Customers can rent out GPU capacity by the hour or lease dedicated capacity through long-term contracts. Some operators also provide managed software services to help customers develop and deploy AI models.

In June, Gartner predicted that neocloud providers would capture 20% of the AI cloud market by 2030. It projected that this market would grow to $267 billion and that traditional cloud computing providers would struggle to keep pace. At the time, Gartner analyst Enrique Castera said:

"Neoclouds are offering differentiation via superior performance on AI workloads, flexible deployment models, and a strong commitment to data sovereignty, often at a more competitive price point. The AI cloud market is entering a new phase where sovereignty, performance, and infrastructure specialization are becoming primary decision factors for enterprises. As demand for GPU-intensive workloads accelerates and traditional cloud models struggle to keep pace, it is creating the conditions for a new class of providers purpose-built to deliver AI infrastructure at scale."

A few months later, Gartner's opinion of the space has completely flipped. The technology advisory firm projected that there will be nearly $1 trillion spent on AI infrastructure this year, with an additional $6 trillion spent between 2027 and 2030. However, it does not believe that consumer and AI revenue will be large enough to support this spending.

Senior director analyst Will Sommer said that to get a 12% return on invested capital (ROIC), AI data center owners would need to generate a combined $8.5 trillion in revenue to cover capital expenditures and financing costs. However, Gartner currently only projects cumulative AI revenue to reach $3.6 trillion by 2030.

That gap, in its opinion, will lead to massive write-offs in the AI data center space. Gartner believes that neoclouds will be the companies most hurt in this scenario. The biggest reason for this is that neoclouds need to see immediate returns on their infrastructure investments. Hyperscalers have other core businesses that generate ample operating cash flow that will allow them to sustain their AI investments.

Gartner said: "By 2030, Gartner predicts that cost pressure and a lack of a differentiating business model will force over 80% of neocloud providers to cease independent operations.​"

A buying opportunity or a sector to avoid?

Neoclouds are spending a massive amount building out their infrastructure and financing most of it with debt. Even with creative financing solutions, like GPU-collateralized special purpose vehicle structures, customer prepayments, and vendor financing, neoclouds still face high debt service burdens and have rapidly depreciating assets. (GPUs have economic useful lives of around six years.)

This makes investing in the space risky. Unlike the big three cloud computing providers, neoclouds don't have strong core businesses to help fund their build-outs. Cloud computing is a scale business where size matters, and if these companies can't reach a level where they can cover their fixed costs, they will be in trouble given the debt they are taking on to fund their data center investments. While I do think the business model could potentially work, there are too many companies chasing it, likely leaving a lot of losers along the way.

Neoclouds are currently in a high-leverage race to reach scale and profitability, and produce strong operating cash flow to cover their expenses. It can be done, as the cloud providers have shown, and a few top players with anchor customers may survive and thrive. However, the majority will likely fail, and potentially all of them will.

As such, I would avoid investing in the space or trying to pick any potential winners, as there is no guarantee there even will be any.

Geoffrey Seiler has positions in Alphabet and Amazon. The Motley Fool has positions in and recommends Alphabet, Amazon, Microsoft, and Nvidia. The Motley Fool recommends Gartner. The Motley Fool has a disclosure policy.