Despite a more than 10% drop in its stock price at Friday's open following a second-quarter revenue forecast that missed expectations, analysts remain broadly positive on AI cloud computing infrastructure firm CoreWeave, Inc. (CRWV.US). The company's Q2 revenue guidance of $2.45 to $2.6 billion, with a midpoint of $2.53 billion, fell short of the consensus estimate of approximately $2.7 billion. Its adjusted operating profit guidance of $30 to $90 million was also significantly below market expectations of around $154 million.
Wall Street remains optimistic about the company's long-term order book and capacity expansion. Wells Fargo reiterated its Overweight rating and raised its price target on the stock from $135 to $155. The analyst team, led by Michael Turrin, noted, "While Q1 was strong, CoreWeave's below-consensus Q2 guide and reiterated annual guide make for a complicated setup for the rest of the year—heavily weighting performance to the back half. However, execution continues, with nearly $100 billion in backlog and over 1GW of active capacity."
Jefferies maintained its Buy rating and $160 price target. Analysts led by Brent Thill stated, "CoreWeave delivered a strong Q1, signing over $40 billion in new committed orders with revenue 5% above consensus. While concerns remain about the steep ramp in adjusted EBIT for the second half of FY2026 ($81M in H1 vs. $919M in H2) and the ~$500M increase in the midpoint of 2026 capex guidance due to component pricing pressure, we see high visibility from the signed backlog (avg. 5-year term) with cost pass-through supporting long-term margins. The current $59B market cap compares to $99B in remaining performance obligations (RPO) and a multi-trillion dollar AI infrastructure build."
Analysts highlighted that the revenue backlog reached $99.4 billion, up from $66.8 billion last quarter. While this was only slightly above some expectations, they pointed out that the increase in RPO was better than expected due to CoreWeave signing over $40 billion in new committed orders (with RPO converging toward revenue backlog) as Nvidia's deal share was replaced by other customers.
The analysts added that the backlog is becoming more diversified, with nearly $10 billion now from the financial services sector. The company also noted it now has 10 customers with commitments exceeding $1 billion each. Furthermore, pricing across product generations remains firm. Demand for Nvidia's Blackwell, Hopper, and Ampere architectures is expanding alongside growing inference needs. CoreWeave indicated that average pricing for A100, H100, H200, and L40 instances increased sequentially, with near-term fleet capacity remaining sold out.
CoreWeave's rise is essentially a case study in deep integration with the Nvidia ecosystem. During the previous period of acute AI computing shortage, the company rapidly secured its position through forward-looking, large-scale stockpiling (especially of H100 chips) and a close partnership with Nvidia, becoming a lifeline for many tech companies eager to train large models but unable to secure sufficient allocations from general-purpose cloud giants. This focused "pure GPU cloud" strategy has given it a differentiated edge over comprehensive cloud providers like AWS and Azure in specific areas, offering customers extreme performance with less service bundling.
The nearly $100 billion revenue backlog is the most direct proof of this niche value. With an average contract length of five years and a client list now featuring 10 heavyweight players each committed to over $1 billion, its services have penetrated from startups to large financial institutions (where backlog is nearing $10 billion). This constructs a "moat" of near-term visibility.
However, deep integration is a double-edged sword. Its business model heavily relies on Nvidia's chip supply roadmap and pricing strategy. As the industry shifts from training to inference, dispersing demand across Blackwell, Hopper, and even older Ampere architectures, can CoreWeave maintain its procurement advantage and cost pass-through ability? Should Nvidia adjust its partnership strategy or increase its own direct cloud service investments, CoreWeave's barriers could be directly challenged.
Finally, there is the "timing game" of revenue recognition. The $99 billion in RPO is not immediate revenue; its conversion into reported income will take years. The market's concern is whether five-year contracts in the fast-iterating AI field hide the risk of clients demanding renegotiations or even cancellations due to changes in technology roadmaps. This risk is particularly pronounced if a portion of orders is highly concentrated with a few leading AI companies.
The core pain point behind CoreWeave's stock decline also lies in its extremely uneven annual performance distribution. Jefferies pointed to a stark expectation: adjusted EBIT for the first half of 2026 is only $81 million, but it is required to surge to $919 million in the second half. This near-exponential growth is highly dependent on the concentrated delivery and deployment of several large orders in the latter half of the year. The uncertainty around flawless execution explains why the market reacted with a sharp sell-off to a slightly weaker Q2 guide—investor tolerance for execution risk is diminishing.
Adding to this is the heavy pressure from capital expenditures. Due to rising component costs, the company has raised the midpoint of its 2026 capex guidance by approximately $500 million. While management emphasizes the ability to pass these costs to customers through contract terms, the limits of this pass-through ability in a competitive environment remain a question. Sustained high capital expenditures and potential financing needs will test the company's financial resilience in a high-interest-rate environment.
Comments