Prediction: Nebius Group Stock Will Skyrocket to $600 in 3 Years

Nebius reported Q1 2026 revenue of $399 million, up 7.8x year over year, and adjusted EBITDA of $129.5 million versus a $53.7 million loss a year earlier, with margin at 32%. Management guided to 2026 revenue of $3.0 billion-$3.4 billion, end-2026 annualized run-rate revenue of $7 billion-$9 billion, and 4 GW of contracted power capacity, supporting a backlog of more than $46 billion. The article is highly constructive on Nebius’ long-term AI infrastructure growth and suggests substantial upside if execution continues.

Analysis

NBIS is transitioning from a pure capacity story to a capital-allocation and utilization story. The market is still mostly underwriting the backlog as if it converts linearly, but the real second-order lever is that software layers should lift realized gross margin and reduce the probability that incremental MW are commoditized into low-return rental business. That matters because the next phase of the move is less about revenue surprises and more about whether the company can keep ROIC ahead of the cost of capital while scaling power commitments faster than competitors can replicate them.

The near-term winners are the hyperscalers that need external capacity without waiting on their own build cycles, but the more interesting beneficiary set is the upstream ecosystem: power developers, electrical equipment, and liquid-cooling vendors should see a multi-year demand tailwind if NBIS keeps converting contracted power into active MW. The likely losers are smaller neoclouds that lack either balance-sheet depth or software differentiation; they will struggle to match price, latency, and customer stickiness once NBIS’s platform becomes embedded in inference and agentic workloads. This also creates a subtle read-through for META/MSFT: outsourcing some AI infrastructure is a speed advantage now, but it may also signal that internal capex alone cannot meet demand, which could support broader AI spending intensity across the sector.

The main risk is timing mismatch. Equity is discounting a smooth ramp, but the stock is vulnerable to any slippage in power delivery, permitting, grid interconnects, or customer acceptance tests; those are month-to-quarter issues, not years, and they can produce abrupt multiple compression even if the long-term thesis remains intact. Another underappreciated risk is concentration: when a meaningful share of future value is tied to a handful of counterparties, any renegotiation, delay, or in-house build decision from those customers can reset backlog quality faster than headline revenue growth would suggest.

Consensus is probably underestimating how much of the upside is already in the stock. The setup is strong, but after a 5x run the easy re-rating is likely behind it; the next leg needs execution beats, not just narrative momentum. That makes the trade asymmetrical for holders: upside remains large over 12-36 months if utilization and margins track, but the stock can still give back a meaningful chunk on any quarter where bookings are fine but capacity conversion slips.

Key Decisions for Investors

• Maintain a tactical long NBIS, but size it as an execution trade rather than a core compounder; use a 3-6 month horizon and cut exposure on any delay in capacity conversion or guidance reset, since the stock is now sensitive to timeline risk more than story risk.
• For investors wanting AI infrastructure exposure with less single-name operational risk, pair long NBIS / short a weaker or less differentiated neocloud peer basket on a 6-12 month horizon; the thesis is that capital access and software attach rates will separate winners from capacity landlords.
• Buy META and MSFT as indirect beneficiaries of external AI capacity expansion; if NBIS proves that third-party supply is still constrained, these names should retain pricing power on AI monetization while avoiding the full infrastructure build burden.
• Use call spreads instead of outright longs on NBIS into the next 1-2 quarters; the expected upside is still meaningful, but the implied volatility after a 5x move likely overstates near-term incremental return unless another backlog or margin surprise lands.
• Monitor power- and data-center-equipment suppliers for a delayed second-order trade; if NBIS execution stays clean for 2-4 quarters, the best risk/reward may migrate from NBIS itself to the industrial beneficiaries of its buildout.