Market Misread? The Real Reason CoreWeave Shares Dropped

CoreWeave shares fell after a CAPEX revision and a delayed data-center project, but its $55.6B backlog and deep AI partnerships signal strong long-term fundamentals.

👁️ 0
CoreWeave
Photo: CoreWeave

CoreWeave has become one of the most closely watched names in the U.S. AI infrastructure space this year — a company that major players like Meta* and OpenAI turn to when they need thousands of GPUs deployed at speed. After going public at $40 in March and surging to $187 in June, the stock has now settled near $80.

For many investors, the volatility looked alarming. But viewed through an AI-infrastructure lens, the company’s fundamentals tell a very different story.

AI-Optimized Data Centers Remain the Core Advantage

CoreWeave’s architecture — much like the newer neocloud models emerging on the West Coast — is built specifically for AI workloads rather than general-purpose enterprise hosting. Engineers often describe it as “purpose-designed compute fabric,” which essentially means faster training times, lower latency and better GPU utilization compared with traditional hyperscalers.

That’s precisely why Meta*, OpenAI and several fast-scaling AI labs continue to rely on CoreWeave as they expand training clusters and inference fleets.

Why the Stock Dropped: A Market Misunderstanding

The selloff traces back to two developments:

  • A broken deal with Core Scientific, which created a burst of headline pressure.
  • A CAPEX revision for 2025 — lowered from $20–23 billion to $12–14 billion.

At first glance, the CAPEX adjustment triggered fears of weakening demand. But the company clarified that the change was driven by a contractor delay. The physical facility for a new data-center buildout simply wasn’t delivered on time. The customer stayed. The contract terms didn’t change.

In other words, this was a logistical slowdown — not an AI-demand slowdown.

Growth Momentum Still Building

CoreWeave’s operational performance remains striking. Third-quarter revenue increased 134% year over year, and the company now holds a $55.6 billion backlog. Roughly half of that is expected to convert into revenue within the next two years, giving the company one of the strongest monetization runways in the AI-compute sector.

At the same time, CoreWeave continues to add high-profile agreements with Meta*, OpenAI and other large-scale AI developers, expanding its foothold in the neocloud segment — a market that smaller, highly specialized providers are capturing faster than legacy tech giants.

A More Rational Valuation

After the selloff, CoreWeave’s valuation now reflects a more grounded market view. The company trades at a P/S below 9x, compared with roughly 70x for Nebius — a spread that highlights how much room investors have to reassess the relative value of AI-native infrastructure providers.

The sector still carries infrastructure and execution risks, but demand for GPU-dense compute remains exceptionally strong, and CoreWeave’s client base continues to scale at a pace few competitors can match.