Nebius Stock Had a Record Month in May. Is It Too Late to Buy?

Nebius posted Q2 2025 sales of $105 million but raised its annual revenue run-rate outlook to as high as $1.1 billion and ultimately reached $1.25 billion, while now guiding to more than $3 billion in 2026 revenue. The company said its record pipeline more than quadrupled in Q1 and expects contracted capacity to exceed 4 GW by end-2026, underscoring strong AI infrastructure demand. The article is constructive on fundamentals but cautions that the stock has already priced in a lot of growth after a 67.2% monthly rally.

Analysis

NBIS is now a classic scarcity-to-capacity monetization story, but the market is increasingly discounting not just utilization, but flawless execution on power, network, and customer onboarding over the next 12-18 months. The second-order issue is that once capacity is effectively spoken for, incremental upside shifts from revenue surprise to margin conversion and financing discipline; any slippage in deployment cadence can hit the multiple harder than the fundamentals. In other words, the stock has likely moved from “scarce GPU access” to “show me the ramp,” which is a much less forgiving regime.

The beneficiary set is broader than NBIS. NVDA is an indirect winner because demand for inference capacity extends the life of the AI infrastructure spend cycle, but the value capture is likely to migrate toward operators that can secure cheap power and near-term deployable sites rather than pure chip suppliers. INTC is a peripheral beneficiary only if AI infrastructure buildout broadens into edge and networking demand; otherwise it remains a weak read-through. The real competitive threat is to smaller cloud and colocation operators that lack contracted power or balance-sheet flexibility, as NBIS’s ability to lock up capacity may tighten available supply and raise replacement costs across the sector.

The contrarian risk is that the market is treating contracted capacity as synonymous with monetizable revenue, when the gap between signed power and realized cash flow can be 6-12 months or longer. If AI inference pricing compresses faster than utilization rises, gross margin expansion could disappoint even as top-line growth remains strong. A second risk is capital intensity: accelerated capacity builds may force equity dilution or expensive debt at exactly the point when sentiment is most euphoric.

Near term, this looks stretched rather than broken: the setup supports momentum for weeks, but not necessarily clean upside from here without another positive guide. The more attractive trade is to own the AI infrastructure “picks-and-shovels” with better cash flow visibility while fading the highest-expectation operator names. If the stock consolidates for 1-2 quarters while capacity comes online, that may reset the entry point; if it gaps on another guide-up without corresponding margin commentary, that’s usually the point to harvest gains.

Key Decisions for Investors

• Trim/hedge NBIS into strength over the next 2-4 weeks; upside still exists, but the risk/reward has shifted from asymmetric to tactical momentum with a high downside multiple if execution wobbles.
• Pair trade: long NVDA / short NBIS for 3-6 months if the goal is to own the AI capex beneficiary with stronger monetization and lower balance-sheet risk; this reduces exposure to capacity-execution disappointment.
• If long NBIS, finance it with an upside-call spread rather than outright stock ownership: 6-12 month call spreads can capture another guide-up while limiting damage from a 20-30% mean reversion.
• Watch for any primary equity or convert issuance over the next 6-9 months; if announced, reduce exposure immediately because financing risk will likely re-rate the stock before the capital is deployed.
• For more conservative AI exposure, rotate incremental capital toward NVDA on dips rather than adding to NBIS chase levels; the payoff is lower beta to execution and better liquidity if sentiment turns.