104 Billion Reasons to Buy CoreWeave Stock

CoreWeave reported Q2 revenue up 112.5% YoY to $2.6B, alongside a sharply higher backlog of $104B (up 245.5% YoY and 4.6% sequentially), which helped the stock surge ~20% post-results. Losses remain large with operating loss of $49M and net loss per share of $1.14 (worse than $0.60 a year ago), though both improved sequentially. The article frames strong backlog visibility and rapid AI infrastructure demand as supportive, tempered by risks like customer concentration (67% of FY2025 revenue from one client).

Analysis

The market is likely to keep rewarding the most levered AI infrastructure names for a few sessions, but the cleaner expression of this spending cycle remains the suppliers with pricing power and balance-sheet flexibility. CoreWeave’s backlog is supportive for near-term revenue visibility, yet it does not solve the two things the market will eventually re-rate on: customer concentration and the cash conversion gap created by heavy compute build-out. In other words, the stock can stay strong on momentum, but the multiple is vulnerable once investors shift from "can they fill capacity?" to "can they fund it efficiently?"

Second-order, the real beneficiaries are the diversified AI enablers: NVDA for accelerator demand, and MSFT/AMZN/GOOGL for cloud capex translation, though hyperscalers may see some FCF and margin pressure before the revenue offset shows up. The more interesting risk is that if hyperscalers keep expanding their own AI fleets, third-party GPU cloud providers may face eventual pricing compression or lower renewal leverage, which would cap CRWV’s terminal valuation even if growth stays rapid for several quarters. The stock looks more like a high-beta growth commodity than a durable software multiple.

Contrarian view: the consensus may be underestimating how much of the backlog is already "spoken for" by one large customer and overestimating how much future upside is left after a 20% post-earnings move. The thesis breaks if management can show sequential improvement in gross margin, utilization, and customer diversification over the next 1-2 quarters; it weakens sharply if capex growth at hyperscalers slows or if guidance implies backlog quality is less sticky than the headline suggests. Absent that, this is a relative-value story, not a must-own outright long.

Key Decisions for Investors

• Prefer a relative long in NVDA versus CRWV over the next 1-3 months: it captures the same AI capex wave with less balance-sheet and customer-concentration risk. Risk/reward: if AI spend stays hot, NVDA participates; if growth normalizes, CRWV’s multiple should compress faster.
• Use strength to trim or hedge CRWV rather than chase it outright; if the stock remains >10-15% above the pre-earnings level, consider a 1-3 month put spread as a tactical hedge against multiple giveback. Falsifier: another quarter of sequential margin improvement and broadening customer mix.
• Stay long MSFT/AMZN/GOOGL as diversified capex beneficiaries, but pair against the most expensive AI infrastructure beta only if you want relative exposure. The market is likely to reward these names for demand visibility without pricing in the same financing risk as CRWV.
• Watch for any sign that hyperscaler capex re-acceleration slows in the next earnings cycle; that is the key 1-3 month catalyst that would unwind the entire CRWV bull case. If capex holds, the trade remains a rotation out of pure-play infrastructure and into platform winners.