Why CoreWeave Still Looks Undervalued: A Practical Long Trade Plan

Hook / Thesis
CoreWeave remains one of the clearest pure-plays on the GPU compute layer underpinning generative AI. The market has punished capital-intensive infrastructure names on headline capex and near-term margin pressure, but that punishment creates an asymmetric opportunity: if demand for large-scale GPU capacity stays strong, CoreWeave should see utilization and pricing power rise faster than the market currently expects.

In short: buy a disciplined long with a tight stop and a multi-month horizon. The setup favors patient capital because the company is asset-heavy and quarter-to-quarter volatility in margins and capex will persist. But the secular driver - enterprises and AI cloud-native firms needing specialized GPU pools - is durable. I outline a trade entry, stop and target below and explain the logic and key risks.

What CoreWeave Does and Why the Market Should Care
CoreWeave operates large-scale GPU compute infrastructure optimized for AI training and inference. Unlike general-purpose cloud providers, CoreWeave targets GPU-heavy workloads and often provides custom compute clusters, managed services and software integrations tailored to model builders. The business benefits directly from two structural trends:

• Rapid growth in generative AI model training and inference demand, which consumes GPU capacity disproportionately to traditional CPU workloads.
• Tight supply dynamics in high-performance GPUs and the logistical complexity of deploying and maintaining GPU pools, which favors specialized operators with relationships for procurement and efficient utilization.

For investors that want exposure to AI infrastructure without owning large-cap hyperscalers, CoreWeave is one of the few public names that is concentrated in the GPU stack. That concentration makes its revenue and margin trajectory highly levered to utilization and pricing of high-end accelerators - a risk, but also the source of upside.

Supporting Arguments - Why Now?
There are three practical reasons the stock looks undervalued today:

• Demand-outstrips-visible-supply: Customers building large language models and fine-tuning at scale have immediate capacity needs that are difficult to meet with self-provisioned data centers. That makes third-party GPU providers, especially those that can rapidly scale and offer flexible contracts, more valuable than headline cloud price comparisons suggest.
• Operational leverage on utilization: Once hardware is deployed and utilization rises, incremental revenue flows through at a higher rate, improving gross margins materially. The market has focused on near-term capex, but the follow-on margin expansion is where the return profile lies.
• Specialization premium potential: CoreWeave’s stack and customer integrations create switching costs versus moving entirely to general-purpose clouds for certain workloads - this can support higher realized prices per GPU-hour for differentiated services.

Valuation Framing
Current public market pricing has baked in skepticism about capital intensity and the timing of margin inflection. That skepticism is understandable, but valuation should be considered against the expected path of utilization and the multiple the market assigns once investors can see durable margin improvement.

Think of the valuation in two pieces: the replacement value of deployed GPU capacity and the growth premium for ongoing software, services and differentiated customer relationships. If CoreWeave continues to grow revenue and utilization at a healthy clip, the multiple compression seen during periods of capex focus should reverse as EBITDA improves. This trade assumes the market will re-rate the company as a higher-margin, predictable revenue business over the next several quarters if utilization holds up.

Catalysts (2-5)

• Large multi-quarter customer commitments announced or disclosed on earnings that reveal multi-million-dollar ARR-like contracts.
• Public commentary or data showing sustained high utilization rates across deployed GPU fleets.
• Improvements in gross margin driven by higher utilization and better pricing per GPU-hour.
• Strategic partnerships with model providers or software vendors that embed CoreWeave as the preferred compute partner.

Trade Plan - Concrete Entry, Targets and Time Horizon
This is a directional long trade sized for a swing-to-position horizon. The plan assumes some near-term noise from capex headlines and quarter-to-quarter margin swings, so the position should be sized accordingly within a diversified portfolio.

Rationale: The $25 entry reflects a point where the downside is limited if utilization falls and the broader market continues to de-rate asset-heavy names. The $18 stop caps losses if clearly-deteriorating demand or a major execution failure emerges. The $40 target is reachable if utilization trends higher and gross margin improvement becomes visible to the market, prompting a re-rating over the 46-180 trading day window.

Position Management
Plan to review the trade after each quarterly report. If the company delivers several quarters of rising utilization and improving gross margins, gradually scale toward target. If near-term earnings reveal both deteriorating demand and a need for materially higher capital expenditure without corresponding revenue visibility, exit to the stop. Tight stops matter here because of the name's operational leverage.

Risks and Counterarguments
Below are the primary risks and at least one counterargument to the bullish thesis.

• Capital intensity and margin pressure: Building and maintaining GPU farms is expensive. If GPU price declines or the company must accelerate capex to keep up with hyperscaler competition, near-term margins could compress further.
• Customer concentration: If a few large customers represent a meaningful share of revenue, the loss or downsizing of a major account could materially impact utilization and short-term cash flows.
• Competitive pressure from hyperscalers: Large clouds could undercut pricing or offer bundled solutions that lure away some customers, especially if they improve GPU procurement or product flexibility.
• Macro and AI cycle risk: A slowdown in enterprise AI spending or a postponement of model training projects would reduce demand for third-party GPU capacity.
• Execution risk: Scaling GPU farms at low-cost and high utilization requires tight operations and supply-chain execution; missteps would delay margin improvement and lengthen the path to profitability.

Counterargument: It is possible the market is correctly cautious. If demand for specialized external GPU capacity plateaus because enterprises internalize more of their AI infrastructure or hyperscalers provide equivalent integrated services at lower cost, CoreWeave could struggle to keep utilization high enough to justify its deployed capacity. Under that scenario, the stock’s current valuation could be appropriate.

That counterargument is why the trade uses a strict stop and a multi-month time horizon: the thesis depends on visible utilization and margin improvements, not just narrative about AI demand.

What Would Change My Mind
I would reduce or reverse the bullish stance if any of the following occur:

• Quarterly reports show declining utilization sequentially or materially lower average realized pricing per GPU-hour without offsetting cost declines.
• New disclosures indicate a major customer loss or meaningful revenue concentration risk crystallizing into lost revenue.
• Management signals the need for significantly higher-than-expected capital to stay competitive, with limited visibility to when that capex converts to utilization and cash flow.

Conclusion
CoreWeave presents a tactical opportunity: the market's discount for capital intensity and near-term margin noise looks excessive relative to the upside if utilization and pricing for GPU capacity remain strong. The proposed trade is not a flip; it is a measured, long-term (180 trading days) position designed to capture a re-rating as operational leverage is realized. The risk-reward is favorable so long as the company demonstrates sustained demand, visible margin improvement and sensible capital allocation. Strict stops and active monitoring are essential because this is a high-conviction, technically-capitalized business with real execution risks.

Key Monitoring Items
Watch quarterly utilization, average realized price per GPU-hour, and any disclosures of multi-quarter customer commitments. Those metrics will determine whether CoreWeave is on the path to the margin expansion the market needs to re-rate the stock.