Back to News
Market Impact: 0.64

CoreWeave (CRWV) Q1 2026 Earnings Transcript

CRWVMETANVDAMCONFLX
Artificial IntelligenceTechnology & InnovationCorporate EarningsCorporate Guidance & OutlookCompany FundamentalsCredit & Bond MarketsBanking & LiquidityInfrastructure & Defense

CoreWeave posted Q1 revenue of $2.1 billion, up 112% year over year, with $1.2 billion of adjusted EBITDA and a 56% margin, while backlog surged to $99.4 billion after more than $40 billion of new bookings. Management raised the floor for 2026 exit annualized run-rate revenue to $18 billion and reaffirmed $30 billion-plus for 2027, while also increasing full-year CapEx guidance to $31 billion-$35 billion on higher component pricing. The company closed an $8.5 billion investment-grade DDTL facility at an implied cost below 6% and said more than 75% of 2027 revenue is already contracted, supporting continued hypergrowth despite a wider net loss and heavier near-term margin pressure.

Analysis

The key second-order read is that CoreWeave is starting to look less like a pure AI infrastructure story and more like a financing-and-allocation platform with contracted demand as the collateral. That matters because the new debt structure lowers the marginal cost of capital exactly when the company is able to reprice capacity upward; if that persists, equity value should compound faster than headline EBITDA suggests. The market is likely underappreciating how much of the current squeeze is not on demand, but on bottlenecks in power delivery, labor, memory, and deployment sequencing — which means the true constraint is operational throughput, not customer appetite. The strongest near-term winner is NVDA, but not for the obvious reason. This quarter implies longer effective useful life for prior-gen GPUs and more fungible demand across architectures, which extends the monetization window for Hopper/Ampere inventory while preserving pricing power on newer cycles; that is supportive for NVIDIA’s mix and for suppliers that can serve retrofit, networking, and storage layers. META benefits indirectly from de-risked access to scarce capacity, but the deeper implication is that large AI buyers may increasingly prefer contracted infrastructure partners with balance-sheet scale over building all capacity internally. The contrarian risk is that consensus is extrapolating a linear conversion of backlog into earnings when the real path is lumpy and capital intensive. If component inflation persists or power interconnects slip, the company can still look operationally strong while equity holders absorb more dilution, more debt, and slower cash conversion than the backlog headline implies. The stock can work from here, but the easiest upside is probably over the next 2-3 quarters as margins inflect; beyond that, the debate shifts from demand scarcity to returns on incremental capital deployed. The setup argues for owning the supplier of the picks-and-shovels rather than chasing the most levered AI infra beta. However, because the equity already embeds a lot of execution success, this is also a stock where any pause in backlog conversion or financing market hiccup could re-rate the multiple quickly. The cleanest trade is to express bullish AI capex through the highest-quality financing beneficiary while hedging against a broader de-rating of high-multiple infrastructure names.