Already Up Around 240% This Year, Is It Too Late to Buy Nebius Stock?

Nebius reported $399 million in first-quarter revenue, up 684% year over year, but the stock has already surged about 240% this year and now trades at nearly 90x trailing revenue. Nvidia's $2 billion strategic investment and partnership underscore strong confidence in Nebius' AI cloud platform, yet the article argues the valuation leaves limited upside and meaningful downside risk. The piece is primarily a valuation-focused commentary rather than a new operational catalyst.

Analysis

The market is rewarding Nebius less for current fundamentals than for perceived optionality on becoming a scaled inference/training utility. That creates a classic winner-takes-most setup where the real beneficiary may still be Nvidia: if Nebius continues to expand, it likely translates into sustained accelerator demand, networking spend, and software ecosystem lock-in, while smaller AI infrastructure peers get forced into a capital-intensity arms race. The competitive risk is not just valuation compression; it is customer concentration and procurement discipline, because hyperscalers can pause incremental outsourced AI spend much faster than Nebius can flex costs.

The second-order issue is that the stock is now trading like a momentum vehicle, not a business with smoothing power. At this multiple, even a strong quarter can disappoint if revenue growth decelerates a few hundred basis points or if operating losses fail to improve on schedule; the market will punish any sign that growth is being bought rather than scaled. That makes the next 1-2 earnings prints the key catalyst window, with downside likely to show up first through multiple compression rather than a fundamental collapse.

The contrarian read is that consensus is anchoring on top-line growth while underweighting path-to-profitability and financing flexibility. In AI infrastructure, the first phase of growth is often front-loaded by customer test-and-train spend, but the durable phase depends on repeat workloads and utilization, which are harder to prove and easier to miss. If utilization or gross margin trends wobble, the stock can de-rate sharply even if absolute revenue keeps rising, because expectations are already priced for near-flawless execution.

Key Decisions for Investors

• Avoid initiating a fresh long NBIS at current levels; if already long, trim 25-50% into strength and retain only with a tight trailing stop, as the near-term risk/reward is skewed by multiple compression over the next 1-2 earnings cycles.
• For investors wanting AI infrastructure exposure, prefer long NVDA vs. long NBIS on a relative basis over the next 3-6 months; NVDA benefits from broader capex diffusion across the ecosystem while carrying less single-name execution risk.
• Consider a put spread on NBIS into the next earnings date (e.g., 1-2 months out) to express a valuation-reset view with defined risk; target a 30-50% potential payoff if the market reprices to a lower sales multiple on any deceleration.
• Watch CRWV as the closest public comp for read-through on AI cloud monetization; if CRWV trades materially below NBIS on forward sales, that divergence is a signal to fade NBIS momentum rather than chase it.
• If NBIS pulls back 20-30% on no fundamental deterioration, reassess for a tactical long entry only after the market reanchors to a more realistic revenue-growth-to-margin conversion path.