CoreWeave can rally on updates to OpenAI, Perplexity deals, Wells Fargo says

Wells Fargo raised CoreWeave’s 12-month price target to $135 from $125, implying 28% upside, and said the company could benefit from investor focus on more than $63.9 billion of deals secured in Q1 2026 and an estimated backlog above $90 billion. The bank expects constructive commentary on OpenAI delivery timelines and unit economics, with first-quarter revenue seen slightly above the high end of $1.9-$2.0 billion guidance. CoreWeave has risen 41% in the past month and 146% over the past year.

Analysis

The key second-order issue is not just that CoreWeave has more backlog, but that the market will start underwriting the quality of that backlog rather than the headline size. If management can show conversion into visible delivery schedules and cash collection, CRWV can de-risk like a contracted infrastructure provider; if not, the stock can quickly re-rate from “AI capacity winner” to “capital-intensive promise.” The balance-sheet angle matters because investors will increasingly model whether growth is self-funding or still dependent on aggressive financing to bridge GPU procurement and customer ramp.

A less obvious beneficiary is Nvidia. Every incremental sign that hyperscale AI demand is broadening beyond a few flagship labs supports GPU demand durability and weakens the bear case that supply is about to outpace end demand. That said, if CoreWeave’s economics look too attractive, the market may also infer that the real bottleneck is access to scarce GPU supply, which could push pricing power further upstream to NVDA and away from the cloud rental layer.

The setup is more vulnerable over the next 1-4 weeks than over the next 12 months. The stock has already moved hard enough that a merely decent print may be treated as a disappointment if guidance does not show accelerating backlog monetization or improving unit economics. The main reversal risk is an increase in reported costs per deployed GPU, slower customer delivery milestones, or any hint that large RPOs require more equity-like funding than investors currently assume.

Consensus appears to be underpricing the possibility that the best trade is not CRWV outright, but the spread between the AI infrastructure enablers and the most levered model-dependent name. If CoreWeave confirms durable demand with acceptable economics, that helps NVDA and keeps META/MSFT spending narratives intact; if not, CRWV bears the brunt because the market has already rewarded the growth story much faster than the cash-flow story.

Key Decisions for Investors

• Long CRWV into earnings only as a tactical trade, preferably on a 2-5% pullback pre-print; target is a 10-15% post-earnings upside move if backlog conversion and delivery timing are better than expected, but trim aggressively if management avoids quantitative unit-economics detail.
• Pair trade: long NVDA / short CRWV for the next 1-3 months. This isolates sustained AI capex demand while fading the risk that cloud-rental economics get squeezed by financing and depreciation assumptions; use CRWV as the leg with higher fundamental execution risk.
• Buy CRWV downside protection via put spreads for the 1-2 month window around earnings. A 10-15% downside scenario is plausible if guidance is merely in line and investors focus on balance-sheet intensity rather than backlog headline size.
• If you want exposure to the AI infrastructure theme with less execution risk, shift incremental capital toward META or MSFT rather than CRWV. They are indirect beneficiaries of the same capacity buildout but with far better balance-sheet resilience and lower financing sensitivity.
• Set a post-earnings decision rule: if CRWV fails to show explicit conversion metrics for the backlog, reduce longs regardless of headline revenue growth, because the next leg higher requires monetization visibility, not just order accumulation.