CoreWeave and Nebius Have Cratered 25%-35%. Dip Buy or Credit Trap?

The selloff is large enough to matter - and still early enough to punish both FOMO and panic

This looks more like a setup to watch than a verdict to chase. Over the past month, CoreWeave has fallen about 36% and Nebius about 43%; other recent summaries show CoreWeave down roughly 35% from its 2026 high and Nebius down nearly 25%. That is enough to trigger both FOMO and panic, but it still looks more like a financing-risk repricing than proof that either business model has broken.

What changed first: financing stress, then customer-focus risk

The initial trigger was credit markets, not earnings. CoreWeave's CDS topped roughly 855 basis points, which on a widely used model implies a 50% five-year default probability. At the same time, customer concentration became the problem investors could price fastest. Both companies rely on major AI clients such as Meta and Microsoft, and concerns that Meta may expand into cloud computing weakened the narrative that had supported these stocks.

Why the debate still matters

The bull case is that this may be an overreaction in a sector where AI demand remains strong. The bear case is harsher but credible: if financing stays expensive and hyperscalers internalize more workload, these businesses could lose both margin of safety and multiple.

That tension is the opportunity. A 25% to 40%-plus move can start a cleaner valuation baseline - or the first honest reckoning. Right now, the market is still deciding which one.

Meta's strategy is reshaping the neocloud demand story

Why investors changed their framework

The selloff was less about weaker AI demand than about a faster reset in how investors value demand tied to a customer's spare capacity.

For a while, the market used a simple shortcut: AI compute demand is hot, so rented capacity must be sticky. That logic started to crack as Meta sounded more self-reliant. Meta had previously said it would only enter cloud if it had excess computing capacity, and as of early June, he claimed Meta did not have that spare capacity. That helped reassure investors that external renters such as CoreWeave and Nebius would remain relevant. When reports emerged that Meta was moving toward a cloud offering, the market repriced that comfort quickly.

Demand may still exist, but bargaining power is shifting

The key point is not that customers have stopped needing capacity. Reporting still shows Meta using external leases to gain additional computing power while it expands its own footprint. That suggests the thesis is shifting from pure demand weakness to a question about how sticky that demand really is if a major customer becomes more self-sufficient.

What bulls and bears are actually arguing about

Bulls focus on growth potential. Bears focus on customer mix. Even if demand stays strong, investors now have to ask whether external capacity is a durable advantage or mainly bridge demand while hyperscalers build more infrastructure.

My read is that the bear case won the first round because investors finally separated "AI demand" from "sticky revenue from the right customer." What is needed now is proof that Meta's internal build-out meaningfully reduces net purchases, not just changes the story around them.

CoreWeave's financing win matters, but it does not settle the debate

Why the latest funding deal matters

From here, the equity story is second-order. The first-order question is whether CoreWeave can keep accessing capital before fear overwhelms the AI buildout narrative.

That is why the latest funding win matters. CoreWeave closed a $3.1 billion delayed draw term loan facility. Just as important, the proceeds support infrastructure dedicated to two customer contracts, which can make the structure easier for lenders to underwrite because the capital is tied to defined deployments rather than serving as a blank check.

Why equity bulls and credit traders may still disagree

The tension is that financing success and credit stress can both be true at the same time. CoreWeave's CDS topped roughly 855 basis points, and on a widely used model that implies a 50% five-year default probability. That helps explain the equity selloff: not a final call on AI demand, but real anxiety about whether a leveraged grower can keep funding expansion at workable terms.

Equity buyers tend to focus on growth and contract-linked assets. Credit traders focus more on leverage, cash flow, and refinancing risk. When those views collide, the stock can swing far enough to create opportunity - or far enough to signal that the market is moving from narrative reset to balance-sheet reckoning.

What would confirm a dip buy, and what would break the thesis

The market has already punished financing fear. The next signals need to show whether that fear was temporary or structural.

Signals that would support buying the dip

A more constructive setup likely needs evidence that demand and financing are reinforcing each other, not just that market confidence is holding up. Watch for:

• More capital coming through customer-contract-backed financing, especially structures tied to infrastructure dedicated to two customer contracts.
• Signs that major clients are still using external capacity as additional computing power rather than treating it as temporary bridge spending.
• No further deterioration in credit sentiment after the recent credit-default-swap costs surging.

Signals that would break the dip-buy case

Do not confuse a pause in selling with a healed thesis. The bear case remains intact if:

• financing conditions tighten again after the recent transaction, suggesting investor appetite was narrow or tactical rather than durable;
• customers lean harder on internal infrastructure, strengthening the view that Meta plans to launch its own cloud computing service;
• the market starts treating these deals less as long-term AI capacity and more as flexible, short-duration demand.

The practical takeaway is simple: be more constructive only if financing keeps getting anchored to contracts and customer usage. If the next signals point mainly to tighter funding and more customer self-reliance, this may have been less of a dip to buy and more of a reset in how the market values the quality of that demand.