Nebius Group N.V. (NBIS) Presents at Citi's 2026 Global TMT Conference Transcript

Nebius CEO Arkady Volozh said the AI-infrastructure startup has raised $40 billion and secured 5 gigawatts of contracted capacity since launching in 2024 with an initial 10MW data center and $2.5 billion of capital. The company is positioning itself as an AI cloud-infrastructure builder, leveraging prior hyperscale experience to meet accelerating demand for computing capacity. The update signals rapid expansion in funding, contracted power and commercial operations, though no revenue or profitability figures were disclosed.

Analysis

NBIS is being valued primarily on contracted capacity optionality rather than demonstrated returns on that capacity. The key underwriting question is not whether AI demand exists, but whether Nebius can convert signed power commitments into funded, energized clusters while preserving GPU utilization and pricing; a 5GW buildout can require well over $100B of cumulative capex depending on configuration. That creates a potentially steep financing curve, with equity dilution, project debt covenants, and vendor-payment obligations likely to matter more to the stock than near-term revenue growth.

The competitive asymmetry is against NBIS: MSFT, AMZN, GOOGL and ORCL can absorb temporary underutilization and finance data-center construction at materially lower cost of capital. Conversely, NBIS could earn a scarcity premium if it has secured deliverable power in constrained European markets, where grid interconnection—not GPUs—is the binding constraint; this would also benefit equipment suppliers such as VRT and ETN, though their exposure is diversified and the incremental contribution from one customer is unlikely to move estimates.

Over the next days, conference-driven enthusiasm can support NBIS given a thin public operating history, but a 1-3 month rerating requires independently auditable disclosures: energized MW, customer concentration, contracted versus reserved capacity, capex per MW, and gross-margin or utilization targets. The 6-18 month risk is that AI infrastructure shifts from capacity scarcity to customer-specific optimization, favoring proprietary hyperscaler stacks and compressing independent GPU-cloud rates before NBIS's full buildout is online. The contrarian view is that the market may be treating announced capacity and financing access as equivalent to revenue visibility; they are not until power delivery, hardware procurement, and customer take-or-pay terms are disclosed.

C is not a clean read-through despite hosting the event. Unless Citi identifies a financing, advisory, hedging, or capital-markets mandate tied to Nebius's expansion, the earnings sensitivity is immaterial.

Key Decisions for Investors

• Do not chase NBIS on conference commentary. Establish a watch position only after disclosure of energized capacity and binding customer economics; require evidence that incremental capacity is funded without a dilutive equity raise before treating the buildout as earnings-accretive.
• For a tactical long, use a defined-risk NBIS call spread 3-6 months out only if management provides audited or contractually specific capacity milestones. Target a 2:1 payoff structure; exit on a capex increase without matching committed revenue, or evidence that utilization is below management's implied break-even level.
• Prefer a relative long VRT or ETN versus NBIS for AI-infrastructure exposure over the next 6-12 months: these companies monetize broad industry power-and-cooling bottlenecks without bearing NBIS's single-platform financing and utilization risk. Reassess if hyperscaler capex guidance rolls over or data-center order/backlog conversion weakens.
• Monitor NBIS financing terms, not just total capital raised: a secured-debt package, prepayment requirements, or equity issuance at a discount would be a near-term short catalyst; a fully funded, non-recourse project structure backed by investment-grade customer commitments would falsify the bearish balance-sheet thesis.