Nebius Group N.V. [NBIS] · Equity Underwriting Memo

Catalysts

Nebius Group N.V. [NBIS] — Catalyst Calendar

⚠️ SUPERSEDED IN PART — 2026-07-29

The position verdict in this document is retired. Under the current framework (references/criteria.md, 2026-07-29) the memo outputs an analysis, not a position. Whether an analysis justifies a position is a question about a particular book, and two books answer it differently.

The Gate block and the Gate 4 expected-return arithmetic below are also superseded, by the named Criteria (each with a type: BINDING or MEASURED, returning PASS / FAIL / INDETERMINATE), the reverse-DCF implied-path test, and a separate 12-month target.

→ Current analysis: Nebius_Criteria_and_Valuation_2026-07-29.md

Everything else here — the research, the evidence, the mechanism work — stands. Residual references to "Watchlist" in the prose below are the historical record of the 2026-07-27 assessment and are left intact deliberately.

Task 6 of the Equity Underwriting Memo · Built 2026-07-27

Position verdict retired 2026-07-29 — the memo outputs an analysis, not a position. The triggers below are retained as monitoring items under Catalyst Criteria (MEASURED), which is what they always measured: "upgrade" means moves toward converting the watchlist entry to a Long; "downgrade" means moves toward removing it from the watchlist entirely. Levels reference Task 5: conversion trigger $155, Base target $175, invalidation $132, 200-day MA $140.29 (rising).

1. Dated catalysts

# Date / window Catalyst Linked to Upgrade trigger Downgrade trigger
C-1 ~2026-08-06 (see date note below) Q2 2026 results Thesis Pillar 1 (capacity conversion); Bear-case probability Q2 ARR ≥ $3.0bn (Base path implies ~$2.9–3.1bn); FY26 revenue guidance maintained at ≥$3.3bn; connected-power guidance of 800MW–1GW reaffirmed Q2 ARR < $2.8bn, or FY26 revenue guidance cut below $3.0bn, or connected-power guidance for YE26 reduced. Any of these is the Bear case running.
C-2 Rolling, next likely Q3 2026 Second and third asset-backed financings Gate 4 (p_bear); the dilution assumption ≥$2bn raised at ≤SOFR+300, backed by contracted cash flows, with no ATM equity issuance. This is the single highest-value upgrade event available — it validates the July 17 template and cuts p_bear toward 20%, lifting E[R] above 15%. Any ATM equity issuance at or below $180, or an asset-backed deal priced wider than SOFR+450. Either says the self-funding thesis has failed and the equity is funding the buildout.
C-3 Q3 2026 (Jul–Sep), reported Nov 2026 The Q3 capacity step-up — management said explicitly "we expect capacity added in Q3 to significantly expand our footprint" Risk R-5 (execution on the ramp) Connected power materially higher QoQ and on track for the 800MW–1GW year-end target; Q3 revenue ≥ $900m Connected power flat or the year-end target walked back. Grid interconnect and transformer delays are the industry norm; a slip here pushes 2027 revenue down and the whole DCF with it.
C-4 H2 2026 NVIDIA Vera Rubin NVL72 first deployments (Nebius among the first globally) Risk R-3 (GPU obsolescence vs 5-year depreciation) Deployed on schedule and prior-generation (Blackwell/GB300) fleet pricing holds — evidence the 5-year life is defensible Prior-generation GPU rental pricing falls materially as Rubin arrives. That would make the 4→5 year life extension look aggressive and force D&A up ~$2.2bn in the 2031 model (≈ −$20/share on Base).
C-5 ~Nov 2026 Q3 2026 results All pillars; the second real ARR datapoint of H2 Exit-ARR guidance narrowed to the upper half of $7–9bn Exit-ARR guidance narrowed to the lower half, or withdrawn
C-6 ~Feb 2027 FY2026 results and the exit-ARR print — THE anchor catalyst Gate 3; the 2027 base from which the terminal value is extrapolated Exit ARR ≥ $8.5bn and FY2027 guidance ≥ $10bn. This is the first event that meaningfully recalibrates the terminal, and it is why the stated horizon is 18 months rather than tied to the next print. Exit ARR < $7bn, or FY2027 guidance below $8bn. Bear case confirmed; remove from watchlist.
C-7 Early 2027 Meta $12bn tranche commences Revenue visibility; Risk R-1 Commences on schedule at contracted terms; and Nebius exercises the optional $15bn tranche by selling that capacity to its own customers at market rates above the Meta pre-agreed price — that would be direct evidence the on-demand market is tight Any delay, renegotiation, or restructuring of the tranche. Also: if Nebius sells the $15bn tranche to Meta rather than to its own book, that implies market pricing is below the pre-agreed level — a quiet bearish tell most commentary will miss.
C-8 2027 (ongoing) Missouri (1.2 GW) and Alabama become operational Owned-capacity thesis; capital efficiency On schedule and inside budget; owned-site unit economics disclosed Slippage into 2028, or capex-per-MW materially above plan
C-9 2027, phased Pennsylvania 1.2 GW AI factory phased delivery Long-run capacity; terminal value Phase 1 energised on schedule Permitting or interconnect delay
C-10 Any time Meta cloud-services announcement Risk R-1 — the single most important risk Meta explicitly rules out third-party cloud resale, or its 2027 commitment is increased Meta formally launches a cloud offering monetising excess capacity. On 2026-07-01 the report alone took 17% off NBIS in one session; the confirmation would be materially worse and would force a re-underwrite of the entire neocloud terminal multiple.
C-11 Ongoing NVIDIA relationship (9.3% shareholder, $2bn invested, supplier, technology partner) Risk R-7; allocation priority Deeper technical/commercial commitments; continued Exemplar Cloud status on new generations Any reduction in the stake, or a comparable strategic investment by NVIDIA in a direct competitor at similar scale

Date note on C-1 — flagged, not resolved

Sources disagree on the Q2 2026 reporting date. MarketBeat lists 2026-08-06, which is consistent with the company's own recent cadence (Q2'25 reported 2025-08-07; Q3'25 2025-11-11; Q4'25 2026-02-12; Q1'26 2026-05-13). Catacal lists 2026-07-28 (sourced from TipRanks, marked "verified"). No company press release confirming either date was found on EDGAR as of 2026-07-27. Treat the window as 2026-07-28 to 2026-08-10 and check the IR calendar (group.nebius.com/investor-hub) before acting on anything date-dependent. This ambiguity is one more reason no catalyst-dated option position is recommended.

2. Undated but monitorable

Item Why it matters What to watch
Customer-concentration disclosure Risk R-4 is currently un-quantified. Microsoft + Meta are inferred to be most of the $33.6bn RPO but no concentration table is disclosed. The FY2026 20-F (due ~April 2027) should carry a customer-concentration note. Top-two below ~60% of RPO would be a genuine upgrade input (it is condition (iv) of the Task 5 conversion test).
Convertible conversion prices ~50m shares (≈20% of the base) of if-converted dilution caps the Bull case. Terms of the Q1'26 $4.3bn convertible issuance in the FY2026 20-F. If the strikes are near $250–300, a rally through them triggers both dilution and convert-arb short covering — an unusual, two-sided technical setup.
Insider buying Zero open-market purchases in 23 Form 4 filings over twelve months. Per Cohen/Malloy/Pomorski, buying is the informative signal and it is simply absent. Any clustered, non-10b5-1 purchase by Volozh or the CFO would be a real (if small) upgrade input. Continued selling is not a downgrade — it is noise.
Short interest composition 29.1% of float, of which most is inferred convert-arb hedging. FINRA bi-monthly settlements. A rise in short interest without new convertible issuance would change the reading from "hedging artefact" to "directional bears" and would be a genuine downgrade input.
Peer read-across CoreWeave −41% and Oracle −50% over twelve months while NBIS is +254%. The group is de-rating; NBIS is the outlier. CRWV's next print and its FY2027 consensus. If CoreWeave's 2027 consensus is cut, NBIS's 4.53x FY27 EV/revenue premium becomes much harder to defend.
AI capex cycle The macro factor that dominates the Bear case. Hyperscaler capex guidance from MSFT/META/GOOGL/AMZN. A synchronised guide-down is the Bear scenario arriving from outside the company.

3. Monitoring cadence and the specific checks

Frequency Check Threshold
Daily (passive) Price vs the $155 conversion trigger and the 200-day MA Alert at $160; act at $155
Weekly Weekly close vs the 200dma ($140.29, rising) and the 50dma ($226.47, falling) Close below the 200dma → remove from watchlist. Close above the 50dma → Gate 6 mitigation satisfied.
Per earnings Full PEAD check: surprise vs pre-print consensus, revision direction, and whether the stance is with or against the expected drift Per references/incremental-refresh.md
Bi-monthly FINRA short-interest settlement Rise without new convert issuance = reading change
Per 6-K Every financing announcement — size, spread, structure, and whether any equity was issued C-2 triggers
Quarterly Book-level review: cluster weight vs any MU or other AI-capex position 25% sector cap; NBIS+MU treated as one position

4. Handoff for ongoing maintenance

Do not re-run a full initiation for routine updates on this name. Going forward:

The next scheduled touch point is C-1, the Q2 2026 print, in the 2026-07-28 to 2026-08-10 window.