Why CoreWeave, Nebius, and other neocloud stocks are being pounded: Goldman Sachs explains

Neocloud stocks such as CoreWeave and Nebius are ending the week under heavy pressure, extending what has been a rough stretch for a corner of the AI trade that had previously attracted intense investor enthusiasm.

Both CoreWeave and Nebius have fallen roughly 15% over the past five days, a sharp sell-off that has raised fresh questions about the financial durability of companies helping power the AI infrastructure boom.

At first glance, the weakness has been tied to concerns specific to CoreWeave, particularly around its financial position. But the broader market message appears to be bigger than any single company. Investors are increasingly reexamining the economics of the AI buildout itself, especially in an environment where interest rates are staying higher for longer.

That dynamic was laid out clearly in a new Goldman Sachs note, which argued that pressure is building across AI-linked credit markets, not just in public equities.

According to Goldman, spreads on U.S. dollar high-yield AI-related debt widened by 25 basis points between Oct. 13 and 14, hitting 360 basis points over Treasurys. That move stood out because non-AI high-yield debt widened by only two basis points over the same period and still remains below its March peak.

In other words, the market is not broadly panicking about all risky corporate debt. It is becoming more selective, and AI-related borrowers are receiving special scrutiny.

Goldman said the widening has been most pronounced among AI-related corporate bonds and data center project finance. That matters because neocloud companies sit close to the center of the AI infrastructure spending wave. Their businesses often depend on costly access to chips, power, servers, and financing, all while investors expect rapid growth.

Amanda Lynam, a strategist at Goldman Sachs, said some softening in AI sentiment can quickly put pressure on data center debt, particularly when investors begin reassessing the credit quality of the companies buying or leasing that capacity.

She added that bonds tied to neoclouds, as well as companies with major exposure to frontier AI labs, are now “at the epicenter of the funding cost debate.” That is a critical point for equity investors: when debt markets demand a higher premium, the cost of growth rises, and richly valued AI stocks can come under pressure fast.

By contrast, Goldman said AI-adjacent sectors such as energy and utilities have held up better. Those businesses tend to have more diversified revenue streams and more defensive earnings profiles, making them less vulnerable to swings in enthusiasm around generative AI.

The underlying concern is straightforward. Building the AI economy is extremely capital-intensive. Companies need massive spending on data centers, networking gear, GPUs, power capacity, and real estate. If borrowing costs stay elevated and energy costs remain a risk, lower-rated companies become more exposed. That makes investors rethink how profitable or scalable some of these business models really are.

For neocloud companies in particular, that is a serious issue. Many are seen as key enablers of the AI boom, but they also operate in a brutally competitive, high-spending environment. If capital becomes more expensive, the market may no longer be willing to pay premium valuations simply for AI exposure.

Goldman does not appear to be calling for a full-blown collapse. Lynam said there could be “some near-term reversal” in the recent sell-off in AI-linked high-yield debt. But she also cautioned that differences between issuers will likely remain substantial.

That suggests investors are entering a new phase of the AI trade—one where the market is no longer rewarding every company connected to the theme equally. Instead, balance sheet strength, funding flexibility, customer quality, and exposure to energy and financing pressures may matter much more than the AI label alone.

For CoreWeave, Nebius, and other neocloud names, the message from Wall Street is becoming clearer: this is no longer just a story about AI demand. It is a story about whether companies can finance that demand profitably in a tougher credit environment.

And right now, that question is hitting their stocks hard.