DA Davidson reiterates buy on CoreWeave stock after strong quarter By Investing.com

DA Davidson reiterated a Buy rating and $175 price target on CoreWeave after Q1 fiscal 2026 results showed strong backlog growth and active power rising above 1 gigawatt, with 400 megawatts of additional contracted power added in the quarter. Revenue beat estimates at $2.08 billion versus $1.97 billion expected, though EPS missed at -$1.40 versus -$0.91 consensus. The stock has already surged 134% over the past year and nearly 80% year-to-date, suggesting the report is supportive but likely incremental for shares.

Analysis

The market is likely underpricing the optionality embedded in CoreWeave’s contracted power ramp: once a GPU provider crosses the gigawatt threshold, the business starts behaving less like a single-vendor AI server rental and more like a scarce infrastructure platform with multi-generation monetization. That matters because capacity additions are now the gating factor for revenue acceleration, while customer signings on newer architectures imply the company is moving up the stack from opportunistic compute supply to a preferred deployment partner for frontier workloads. The second-order winner is the broader AI datacenter ecosystem — power, cooling, electrical gear, and land banks — because every incremental megawatt effectively pulls through a bundle of capex-heavy vendors before it turns into revenue.

The deeper issue is not demand, but duration mismatch. Revenue can outrun profitability for several quarters if depreciation, financing costs, and power commitments scale faster than utilization, so the equity story becomes highly sensitive to slippage in ramp timing or a pause in hyperscaler/AI lab spend. That creates a reflexive risk: if customers delay model rollouts or shift workloads back to in-house clusters, the backlog headline remains intact while realized margins compress, which is exactly the kind of setup that can punish the stock months before any top-line deceleration shows up.

Consensus seems focused on backlog growth and misses how quickly sentiment can turn when a capital-intensive growth narrative trades at a premium while earnings remain negative. The overdone part is assuming every new system generation will monetize similarly; in practice, newer GPU cycles often pressure pricing before they improve mix, and the market may be extrapolating a straight line from contracted capacity to free cash flow. The underdone part is the benefit to suppliers with real bottlenecks — not semis alone, but power distribution, liquid cooling, and grid interconnect names — because CoreWeave’s expansion suggests the constraint has shifted from chip availability to physical deployment speed.

Key Decisions for Investors

• Stay tactically long CRWV only on pullbacks or post-breadth confirmations; treat it as a momentum trade, not a valuation anchor. Upside can persist for 1-3 months if backlog-to-revenue conversion stays intact, but trim aggressively on any sign of utilization slippage or margin compression.
• Pair trade: long AMAT or MRVL vs short CRWV over the next 1-2 quarters. Benefit from the infrastructure buildout without taking direct balance-sheet and financing risk; the pair expresses AI capex expansion with lower earnings fragility.
• Add exposure to power infrastructure beneficiaries like VRT, ETN, or PWR on weakness. These names capture the second-order spend from AI capacity buildouts and have more visible cash generation than the compute lessor model.
• Consider a bearish hedge in CRWV via put spreads six to nine months out if the stock continues to re-rate faster than backlog converts. The risk/reward improves if implied volatility stays elevated while profitability remains distant.
• If you want a cleaner relative value expression, go long a diversified AI infrastructure basket and short CRWV. That isolates the market’s willingness to pay for scarcity while reducing single-name execution and financing risk.