Meet the Incredible AI Stock That's Crushing Nvidia This Year

Nebius said annual recurring revenue reached $1.25 billion at the end of 2025 and expects $7 billion to $9 billion by the end of 2026, implying roughly 20x revenue growth over two years. Wall Street is even more aggressive, projecting 523% revenue growth in 2026 and 206% in 2027, while Nvidia's investment is presented as a strong vote of confidence. The article is broadly bullish on Nebius, though it is commentary rather than a direct corporate update.

Analysis

NBIS is becoming the clearest public-market expression of AI infrastructure scarcity, but the important second-order read-through is not just “growth is fast” — it is that capital is beginning to concentrate around a handful of vendors that can reliably deliver GPU capacity, networking, and deployment speed. That favors the pure-play infrastructure layer over application software in the near term because customers are paying up for immediacy and throughput, not feature breadth. It also suggests a winner-take-most dynamic where scale begets more scale: better utilization and larger contracted demand should improve financing access and lower unit economics versus smaller neocloud competitors.

The biggest underappreciated signal is Nvidia’s ownership angle. When the market leader invests downstream, it can be read as both validation and a hedging mechanism against customer concentration risk; if one “partner” starts compounding faster than the parent, that tells you where scarcity rents may be accruing. But this also creates a subtle overhang for NBIS: the market may be extrapolating straight-line revenue while underestimating the execution burden of translating backlog into actual delivered capacity, especially if GPU supply, power, or data-center buildout becomes the bottleneck rather than demand.

The contrarian risk is that consensus is pricing an infrastructure super-cycle without enough attention to balance sheet stress and dilution risk. Hypergrowth names can look optically cheap on revenue multiples while being expensive on incremental capital required to support the next dollar of ARR; if financing conditions tighten, the equity can de-rate quickly even if top-line targets stay intact. Near term, the catalyst set is quarterly backlog conversion and forward guidance, but over 6–12 months the key reversal trigger is any sign that demand is normalizing faster than capacity is expanding.

NVDA is a quieter beneficiary: if NBIS and peers keep ordering aggressively, Nvidia’s upstream pricing power and attach rates stay elevated, but the bigger opportunity may be in suppliers to the AI buildout with less headline risk. NFLX is not directly in play here, and the market should avoid force-fitting every AI flow into beneficiaries that do not own the constraint. The trade is less about chasing momentum and more about owning the bottleneck while hedging the part of the story most exposed to financing and delivery risk.

Key Decisions for Investors

• Go long NBIS on pullbacks only, with a 3-6 month horizon; thesis works if management keeps proving capacity conversion, but size modestly because the stock is vulnerable to any missed delivery milestone or financing overhang.
• Pair trade: long NBIS / short a basket of higher-level AI software names with weaker direct monetization, to express the view that infrastructure scarcity captures value first over the next 2-4 quarters.
• Add NVDA on weakness as a lower-beta beneficiary of downstream buildout demand; this is the cleaner way to express AI capex persistence with less single-name execution risk than NBIS.
• Use call spreads instead of outright longs in NBIS if implied volatility is elevated; risk/reward is best for a 6-9 month window where upside from guidance compounding can outpace time decay.
• Set a hard risk limit on NBIS if capital intensity or dilution language worsens in the next earnings cycle; that would be the first sign the market is confusing revenue growth with durable equity value creation.