Why Nebius Stock Jumped Today

Nebius Group posted first-quarter revenue of $399 million, up 684% year over year, and EBITDA improved to $129.5 million from a $53.7 million loss. The company also raised its contracted power target to more than 4 gigawatts by end-2026 and said it is on track for $7 billion to $9 billion in annual recurring revenue by year-end. Recent acquisitions of Tavily and Eigen AI expand its push into inference and agentic AI workloads, reinforcing the growth narrative.

Analysis

NBIS is becoming a barometer for the AI infrastructure bottleneck, but the more important takeaway is that scarcity pricing is still intact. When a supplier can keep re-accelerating guidance while adding power and adjacent software layers, it suggests the market is not yet in a normal competitive equilibrium; the near-term beneficiaries are still the scarce-capacity owners, not the model/application layer. That said, the move into inference and agentic workloads is strategically important because it shifts NBIS from a pure capacity seller toward a higher-retention, harder-to-displace platform with better gross margin durability.

The second-order effect is on capital allocation across the AI stack. If NBIS can lock multi-gigawatt power and monetize it efficiently, then hyperscalers and GPU buyers may be forced to sign longer-duration contracts or prepay to secure supply, which supports demand visibility for NVDA but can also raise the cost of competition for smaller AI cloud entrants. The winners are likely the infrastructure names with access to power, land, and financing; the losers are undifferentiated compute resellers and any operator relying on spot capacity. Intel is only tangentially exposed here, but the broader implication is that the GPU-led ecosystem still dictates procurement priorities, leaving non-NVIDIA silicon a tougher sell in constrained environments.

The contrarian risk is that the market is extrapolating linear growth from a capacity-constrained environment that can normalize faster than expected once new data centers come online. Power buildouts and acquisition integration are multi-quarter execution risks, and any delay in permitting, grid interconnection, or customer ramp would compress the narrative quickly. The stock may be pricing in an unusually clean path from bookings to revenue conversion; if utilization or pricing softens even modestly, the multiple can de-rate sharply because the current setup is story-driven rather than cash-flow mature.

Key Decisions for Investors

• Long NBIS on 3-6 month horizon only if using a tight risk budget; best entry is on post-gap consolidation, with upside tied to continued contract announcements and downside if power/interconnection headlines slip.
• Pair trade: long NBIS / short a basket of smaller AI infra aspirants with no power visibility or software adjacency; the spread should widen over the next 1-2 quarters as capital migrates to scarce-capacity platforms.
• Stay structurally long NVDA on any NBIS pullback; the tighter the compute bottleneck remains, the more pricing power accrues to the dominant GPU supplier over the next 6-12 months.
• Use call spreads rather than outright longs in NBIS if initiating after a strong run; the core risk is multiple compression from any execution delay, while upside is still meaningful if ARR guidance is reaffirmed.
• Avoid chasing INTC as a secondary AI beneficiary here; this is a capacity-and-ecosystem story, not a broad semiconductor beta trade.