AI Consensus Investment Thesis
Nebius Group (NBIS) Stock Forecast and AI Rating
Forecast targets and rating
Final recommendation
No Action - Keep Monitoring
One horizon is weak; wait for clearer confirmation.
1-Year
PARTIALLY SELL$165
-14.7%5-Year
NEUTRAL$257
+32.8%Latest flagship insight
Why This Hyper-Scaling AI Infrastructure Asset Faces An Imminent Valuation Reset
There is high consensus that explosive top-line scaling driven by sovereign AI mandates is temporarily masked by extreme capital intensity. However, sharp divergence exists regarding long-term pricing power as major hyperscalers transition from customers to competitors, threatening severe multiple compression for this heavily leveraged infrastructure asset.
This is the latest published deep-analysis batch. Audit previous forecasts in full transparency
Warren Buffett (Value Purist), Superintelligence (Anthropologist), Ray Dalio (Strategist), Machiavelli (Insider), Elon Musk (Visionary), Michael Burry (Vulture), J.P. Morgan (Titan), Sherlock Holmes (Whistleblower). Some archetypes run in multiple modes, resulting in 12 advisors total.
Research support only. We don't give financial advice.
Gemini
Claude
ChatGPT
GrokExecutive Summary
Dotted terms open concise definitions. Browse technical terms
Interactive forecast chart
The base-case narrative for this AI infrastructureai infrastructureThe compute, networking, storage, power, cooling, software, and facilities used to develop and operate AI systems.View full glossary entry asset centers on a classic reflexivity loopreflexivity loopA feedback cycle in which market perceptions influence fundamentals or behavior, which then changes market perceptions.View full glossary entry where explosive short-term revenue growth masks structurally challenging owner economics. While a multi-billion-dollar hyperscaler backlogbacklogThe total value of signed contracts yet to be fulfilled, representing future revenue visibility.View full glossary entry and sovereign AI mandatessovereign ai mandatesGovernment requirements governing domestic control, location, access, or procurement of AI data, models, and infrastructure.View full glossary entry guarantee near-term capacity utilization, the asset operates as a capital-intensive hardware utility rather than a high-margin software platform. Over the medium term, relentless capital expenditurecapital expenditureCapital expenditure (CAPEX) is spending on long-lived assets or major improvements expected to support operations over multiple periods.View full glossary entry requirements and rapid GPU obsolescence cycles will pressure free cash flowfree cash flowFree cash flow (FCF) is cash generated after operating expenses and capital expenditures that remains available for debt reduction, reinvestment, or shareholder returns.View full glossary entry, while hyperscaler insourcing threats will compress valuation multiples. Professional quants must focus on the transition from peak-cycle scarcity pricing to commoditized compute utility economics.
Key insights
- High-density GPU cluster deployment drives near-term top-line compounding, supported by strategic supplier relationships that guarantee priority hardware access.
- Extreme capital intensitycapital intensityA business condition where large upfront spending on assets and infrastructure is needed to operate and grow.View full glossary entry and a short three-year hardware depreciation cycle structurally limit long-term return on invested capitalreturn on invested capitalReturn on invested capital (ROIC) measures how efficiently a business generates after-tax operating profit from the capital invested in its operations.View full glossary entry and free cash flowfree cash flowFree cash flow (FCF) is cash generated after operating expenses and capital expenditures that remains available for debt reduction, reinvestment, or shareholder returns.View full glossary entry.
- Hyperscaler customer concentration exposes the asset to severe monopsony pricing pressure as key clients scale proprietary custom siliconcustom siliconSpecialized chips designed for specific workloads like AI inference to optimize power and performance.View full glossary entry.
- Sovereign AI mandatessovereign ai mandatesGovernment requirements governing domestic control, location, access, or procurement of AI data, models, and infrastructure.View full glossary entry in Europe provide a localized demand floor, partially insulating the asset from global compute price wars.
- The restrictive macroeconomic regimemacroeconomic regimeThe prevailing set of economic conditions including interest rates and inflation that influence asset pricing.View full glossary entry of elevated interest rates increases refinancing risks and capital costs for debt-heavy infrastructure builders.
- Valuation multiples are projected to compress from speculative technology premiums toward traditional hardware-leasing utility benchmarks as growth normalizes.
Deep Dive
Alpha Gap & Repricing Catalysts
Where does the current market narrative diverge from our AI Forecasts—and what could close the gap?
Market Narrative
What does the market currently expect? The dominant market narrative frames the asset as a flawless, pure-play winner of the artificial intelligenceartificial intelligenceComputer systems designed to perform tasks that ordinarily require human perception, reasoning, learning, language, or decision-making.View full glossary entry revolution, serving as an indispensable provider of picks and shovels. Mesmerized by triple-digit revenue growth and multi-billion-dollar backlogsbacklogThe total value of signed contracts yet to be fulfilled, representing future revenue visibility.View full glossary entry from technology giants, the crowd treats these capacity reservations as guaranteed, risk-free annuities. Retail and sell-side analysts widely assume that the compute deficit will persist indefinitely, allowing the company to maintain software-like margins. This optimistic consensus treats the extremely high price-to-sales multiple as a non-issue, viewing the massive capital expenditurecapital expenditureCapital expenditure (CAPEX) is spending on long-lived assets or major improvements expected to support operations over multiple periods.View full glossary entry plans as a bullish indicator of insatiable demand rather than a heavy operational burden.
Alpha Gap
What is the biggest difference between market expectations and our AI forecasts? The core analytical blind spot lies in the fundamental misclassification of the asset's economic model. While the crowd prices the company as a high-margin, asset-light software platform, it is actually a capital-intensive hardware utility subject to rapid depreciation and physical constraints. The market incorrectly treats the massive backlogbacklogThe total value of signed contracts yet to be fulfilled, representing future revenue visibility.View full glossary entry as a permanent moat, failing to recognize that major technology clients are using this capacity merely as a temporary bridge while they construct their own data centers and scale custom siliconcustom siliconSpecialized chips designed for specific workloads like AI inference to optimize power and performance.View full glossary entry. When these clients eventually insource their workloads, the company's peak margins will compress severely. Furthermore, the market ignores the massive cash drain of the capital expenditurecapital expenditureCapital expenditure (CAPEX) is spending on long-lived assets or major improvements expected to support operations over multiple periods.View full glossary entry treadmill, where hardware must be replaced every three years, preventing the business from generating sustainable free cash flowfree cash flowFree cash flow (FCF) is cash generated after operating expenses and capital expenditures that remains available for debt reduction, reinvestment, or shareholder returns.View full glossary entry.
Repricing Catalyst
What could make the market recognize and close that gap? The convergence of the market narrative toward reality will likely be triggered by upcoming earnings cycles in late 2026 or 2027. Key catalysts include a major technology client formally declining to exercise optional capacity tranches or announcing a shift of workloads to internal custom siliconcustom siliconSpecialized chips designed for specific workloads like AI inference to optimize power and performance.View full glossary entry. Additionally, the massive capital expenditurecapital expenditureCapital expenditure (CAPEX) is spending on long-lived assets or major improvements expected to support operations over multiple periods.View full glossary entry pipeline will begin hitting the income statement as heavy depreciation charges, causing a sequential drop in operating marginsoperating marginsOperating profit as a percentage of revenue after operating expenses, before interest and taxes.View full glossary entry. These operational reality checks, combined with the liquidity drainliquidity drainA reduction in money or financing available to markets, which can tighten credit and pressure valuations.View full glossary entry from high-profile technology initial public offerings, will force analysts to revise growth models downward, making the multiple compressionmultiple compressionA decline in valuation ratios such as price-to-earnings or enterprise-value-to-sales.View full glossary entry impossible to ignore.
Sentiment and Timing
What do sentiment, volatility, and the market-recognition cycle suggest about the thesis timing?
- Greed / Fear
- Greed
- Volatility
- High Erratic
- Cycle position
- Overshoot
While there is strong consensus that the asset is currently in an overshoot phase characterized by high erratic volatility, sentiment remains split. Some perspectives highlight extreme greed driven by short-term revenue momentum, while others point to growing fear regarding capital intensitycapital intensityA business condition where large upfront spending on assets and infrastructure is needed to operate and grow.View full glossary entry and the sustainability of hyperscaler contracts.
Macro Regime Fit
Does the current market environment support the thesis? The macroeconomic backdrop presents a highly challenging, mixed setup for this asset. On one hand, the global de-globalization trend and sovereign AI mandatessovereign ai mandatesGovernment requirements governing domestic control, location, access, or procurement of AI data, models, and infrastructure.View full glossary entry act as powerful geopolitical tailwinds, driving localized demand for secure compute infrastructure. On the other hand, the restrictive monetary regimemonetary regimeThe prevailing framework of central-bank policy, interest rates, money creation, and exchange-rate management.View full glossary entry of sticky inflation and higher-for-longer interest rates is a severe headwind. For an asset-heavy business requiring tens of billions in capital expenditurescapital expenditureCapital expenditure (CAPEX) is spending on long-lived assets or major improvements expected to support operations over multiple periods.View full glossary entry, elevated borrowing costs destroy free cash flowfree cash flowFree cash flow (FCF) is cash generated after operating expenses and capital expenditures that remains available for debt reduction, reinvestment, or shareholder returns.View full glossary entry and increase refinancing risks. Furthermore, energy grid constraintsgrid constraintsLimits in electricity generation, transmission, distribution, or interconnection capacity that restrict new supply or demand.View full glossary entry and power cost inflation directly squeeze operating marginsoperating marginsOperating profit as a percentage of revenue after operating expenses, before interest and taxes.View full glossary entry, creating a hostile environment for heavy hardware operators.
Advisor Disagreement
What do our AI Advisors disagree about most? The primary disagreement across reports centers on the long-term viability of the company's business model and the durability of its customer contracts. Some perspectives view the multi-billion-dollar backlogbacklogThe total value of signed contracts yet to be fulfilled, representing future revenue visibility.View full glossary entry as a secure financial bridge that will successfully fund the transition to a dominant sovereign utility. Conversely, more cautious views argue that these contracts are fragile, temporary arrangements that will be aggressively renegotiated or canceled as hyperscalershyperscalersVery large cloud or internet-platform operators that run computing infrastructure at global scale.View full glossary entry scale their own custom siliconcustom siliconSpecialized chips designed for specific workloads like AI inference to optimize power and performance.View full glossary entry. Additionally, there is disagreement regarding the timing and severity of multiple compressionmultiple compressionA decline in valuation ratios such as price-to-earnings or enterprise-value-to-sales.View full glossary entry, with estimates of the eventual valuation resetvaluation resetA material adjustment in price or valuation multiples after expectations or required returns change.View full glossary entry ranging from a moderate correction to a severe drawdown.
Base-Case Forces
Near-certain positive forces
Top Drivers / Tailwinds
Near-certain forces that support the investment thesis. These forces are treated as part of the base case (more than 60% probability of occurrence). Impact columns are specific to this asset class.
Scroll to view all columns
| Driver / Tailwind | Category | Est. stock-price impact | Est. earnings impact | Why it matters |
|---|---|---|---|---|
| Sovereign AIsovereign aiA national strategy to develop or control domestic AI models, data, computing capacity, skills, and governance.View full glossary entry Compute Mandates | Regulatory | +25% | +45% | Geopolitical fragmentationgeopolitical fragmentationStructural shifts in global trade and policy impacting international business operations and market access.View full glossary entry and regulatory shifts toward sovereign AIsovereign aiA national strategy to develop or control domestic AI models, data, computing capacity, skills, and governance.View full glossary entry hard-fencing compel European governments and enterprises to onshore compute capabilities. This forced localization restricts foreign competition and drives highly inelastic, state-backed demand for localized infrastructure. This regulatory tailwind guarantees high capacity utilization and provides a structural baseline of revenue that is insulated from broader macroeconomic fluctuations and global market efficiency. |
| Hyperscaler Compute BacklogbacklogThe total value of signed contracts yet to be fulfilled, representing future revenue visibility.View full glossary entry | Sector And Industry | +35% | +50% | Multi-billion-dollar capacity reservations from technology giants provide immense near-term revenue visibilityrevenue visibilityHow clearly future revenue can be estimated from contracts, subscriptions, or recurring demand patterns.View full glossary entry and cash flow. These long-term contracts act as a financial bridge, allowing the company to secure construction financing and power allocations. As long as the global compute deficit persists, this contracted backlogbacklogThe total value of signed contracts yet to be fulfilled, representing future revenue visibility.View full glossary entry de-risks the aggressive capital expenditure cyclecapex cycleA recurring period of rising and falling capital investment across an industry or economy.View full glossary entry and shields the company from short-term demand shocks, buying time to build its infrastructure empire. |
| Nvidia Strategic Partnership | Competitive Positioning | +20% | +25% | A deep strategic partnership and direct equity investment from the leading semiconductor manufacturer secure priority access to next-generation GPU architectures. This preferential supply chainsupply chainThe network of suppliers, producers, logistics providers, distributors, and customers involved in creating and delivering a product or service.View full glossary entry alignment creates a formidable moat against smaller cloud providers, guaranteeing hardware availability in a highly constrained market. This structural advantage allows the company to command premium pricing and attract top-tier developers who refuse to train on older hardware. |
| Agentic AIagentic aiAutonomous artificial intelligence systems capable of performing complex tasks and decision-making without constant human intervention.View full glossary entry Token-Burn Acceleration | Innovation And Product | +15% | +30% | The transition of artificial intelligenceartificial intelligenceComputer systems designed to perform tasks that ordinarily require human perception, reasoning, learning, language, or decision-making.View full glossary entry models from simple chatbots to autonomous, multi-step agentic loops exponentially increases compute requirements. Executing complex workflows requires massive token throughput, keeping GPU utilization rates pinned at theoretical maximums. This structural shift transforms compute from a cyclical commodity into high-margin, baseload digital energy, supporting long-term revenue compounding and protecting the business from near-term macro demand shocks. |
Near-certain negative forces
Top Frictions / Headwinds
Near-certain forces that could slow, cap, or damage the thesis. These forces are treated as part of the base case (more than 60% probability of occurrence). Impact columns are specific to this asset class.
Scroll to view all columns
| Friction / Headwind | Category | Est. stock-price impact | Est. earnings impact | Why it matters |
|---|---|---|---|---|
| Extreme Capital Intensitycapital intensityA business condition where large upfront spending on assets and infrastructure is needed to operate and grow.View full glossary entry | Capital Allocation | -45% | -60% | The business model requires staggering capital expenditurescapital expenditureCapital expenditure (CAPEX) is spending on long-lived assets or major improvements expected to support operations over multiple periods.View full glossary entry to build out data center capacity, guaranteeing deeply negative free cash flownegative free cash flowA condition where cash outflows for operations and investment exceed cash generated in the period.View full glossary entry for the foreseeable future. Funding this aggressive expansion treadmill requires continuous external financing, forcing the company into highly dilutive equitydilutive equityIssuance of new shares that reduces the ownership percentage and earnings per share for existing shareholders.View full glossary entry offerings or expensive debt accumulation. This structural capital drain severely drags on intrinsic per-share value and increases vulnerability to credit market freezes. |
| Hyperscaler Insourcing and Competition | Competitive Positioning | -35% | -40% | Major technology clients are actively developing proprietary custom siliconcustom siliconSpecialized chips designed for specific workloads like AI inference to optimize power and performance.View full glossary entry and scaling internal cloud capacity. As these clients achieve compute self-sufficiency, their reliance on third-party providers will evaporate, threatening massive contract attrition and a demand cliff. This transition from anchor customers to direct, margin-crushing competitors structurally impairs long-term pricing powerpricing powerThe ability of a company to raise prices without losing significant customer demand.View full glossary entry and leaves the company with highly vulnerable stranded assetsstranded assetsAssets that lose economic value earlier than expected because of market, technology, policy, environmental, or physical changes.View full glossary entry. |
| Hardware Obsolescence and Depreciation | Operational Efficiency | -30% | -50% | Cutting-edge GPU clusters are subject to rapid technological obsolescencetechnological obsolescenceThe risk that a product, asset, or capability loses economic usefulness because superior technology emerges.View full glossary entry, losing competitive relevance within three years. This short lifecycle triggers a massive wave of depreciation expenses that systematically crushes GAAP net income. To maintain parity, the company must continuously reinvest capital, trapping it on a perpetual treadmill that suppresses return on invested capitalreturn on invested capitalReturn on invested capital (ROIC) measures how efficiently a business generates after-tax operating profit from the capital invested in its operations.View full glossary entry and prevents the business from ever maturing into a cash-generating fortress. |
| Energy Grid and Power Constraints | Macroeconomic And Macrofinancial | -15% | -20% | Operating gigawatt-scale data centers requires massive electrical power, exposing the company to global energy shocksenergy shockA sudden disruption or price change in energy markets that affects inflation, demand, and operating costs.View full glossary entry and local grid bottlenecks. Surging electricity costs directly compress operating marginsoperating marginsOperating profit as a percentage of revenue after operating expenses, before interest and taxes.View full glossary entry, while slow regulatory approvals for grid interconnects delay deployment schedules. These physical constraints push revenue recognitionrevenue recognitionRevenue recognition is the accounting process for recording sales when performance obligations are satisfied and economic benefits are reasonably measurable.View full glossary entry to the right, puncture the hyper-growth narrative, and remind investors that physical atoms move much slower than digital bits. |
What Could Break or Accelerate the Thesis
Plausible downside scenarios
Tail Risks
Tail yet plausible downside scenarios selected for their highest potential impact.
Scroll to view all columns
| Tail scenario | Chance of Occurring | Stock Price Impact | Exposure category | Why plausible / what changes |
|---|---|---|---|---|
| Enterprise AI ROIenterprise ai roiThe measurable financial return on investment generated by corporate adoption of artificial intelligence technologies.View full glossary entry Collapse | 35% | -60% | Regulatory | If corporate deployments of generative AIgenerative aiArtificial intelligence that produces new text, images, audio, code, video, or other content from learned patterns and user input.View full glossary entry fail to deliver measurable productivity gains, enterprise IT budgets will contract sharply. This demand shock would cause a cascading collapse in compute rental rates, leaving the company with billions in stranded, rapidly depreciating hardware assets and triggering a severe liquidity crisisliquidity crisisA condition in which an institution or market cannot obtain cash or funding needed to meet near-term obligations without severe losses.View full glossary entry that would permanently shatter the growth narrative. |
| US Export Control Expansion | 25% | -60% | Political And Geopolitical | To maintain technological supremacy, the US government could expand export restrictions, blocking the delivery of next-generation silicon to foreign-headquartered entities. Denied access to frontier chips, the company's capital expenditurecapital expenditureCapital expenditure (CAPEX) is spending on long-lived assets or major improvements expected to support operations over multiple periods.View full glossary entry pipeline would freeze, rendering existing facilities obsolete and forcing customers to migrate to compliant US providers, causing a catastrophic collapse in enterprise valueenterprise valueA measure of total business value, commonly calculated as equity value plus debt and preferred claims, less cash and cash equivalents.View full glossary entry. |
Plausible upside scenarios
Tail Opportunities
Tail yet plausible upside scenarios selected for their highest potential impact.
Scroll to view all columns
| Tail scenario | Chance of Occurring | Stock Price Impact | Exposure category | Why plausible / what changes |
|---|---|---|---|---|
| Pan-European Sovereign Cloudsovereign cloudCloud infrastructure designed to keep data, operations, and legal control within specified national or jurisdictional requirements.View full glossary entry Treaty | 35% | +60% | Regulatory | The European Union could formally mandate the company as the preferred, state-backed sovereign infrastructuresovereign infrastructureSovereign infrastructure refers to strategic physical or digital systems maintained under national control for security, resilience, and policy autonomy.View full glossary entry provider for defense, healthcare, and governmental deployments. This regulatory captureregulatory captureWhen regulatory agencies act in the interest of the industry they are supposed to oversee.View full glossary entry would eliminate foreign competition by fiat, converting the commercial enterprise into a protected institutional monopoly with locked-in, high-margin state revenue. This would guarantee a decade of high-margin revenue and justify a permanent premium. |
| Strategic Hyperscaler Buyout | 20% | +45% | Competitive Positioning | Facing intense antitrust scrutinyantitrust scrutinyAntitrust scrutiny is regulatory examination of conduct, pricing, mergers, or market power that may reduce competition or harm consumers.View full glossary entry and regulatory barriers in Europe, a major US technology giant could acquire the company outright to instantly secure compliant local data centers. This acquisition would command a massive control premium, valuing the company on the strategic scarcity of its physical footprint rather than traditional cash flow metrics, delivering an immediate liquidity event to shareholders. |
Company Financial Analysis
Nebius Group Earnings and Financials Analysis by AI
Financial figures available as of Jul 05, 2026. Only filings and source records available by this analysis date are included.
Earnings and financials
The company is experiencing a massive surge in sales, with recent quarterly revenue growing by hundreds of percent compared to last year. This explosive growth is driven by huge contracts to rent out computer chips for artificial intelligence. However, behind these impressive sales numbers, the company is actually losing a lot of money on an operating basis. Most reports agree that the positive net income shown recently is a temporary accounting trick from selling off older assets, rather than healthy day-to-day profits. Experts disagree on how long this sales boom can last. Cautious analysts warn that once the current shortage of computer chips ends, rental prices will drop, and the company's massive costs will overwhelm its sales. More optimistic views believe that the high demand for AI will keep sales growing fast enough to eventually make the company highly profitable.
Revenue, earnings, and cash flow
The table compares up to five fiscal years of revenue, net income, and free cash flow available to this analysis.
| Fiscal year | Revenue | Net income | Free cash flow |
|---|---|---|---|
| 2025 | USD 529.8M | USD 101.7M | RUB -3.7B |
| 2024 | USD 117.5M | USD -641.4M | RUB -562.1M |
| 2023 | USD 20.9M | USD 241.3M | RUB 746.4M |
| 2022 | USD 13.5M | USD 745.6M | RUB 682.4M |
| 2021 | RUB 4.8B | RUB -196M | RUB -473.7M |
Valuation context: historical P/E
The table compares up to five fiscal years of point-in-time valuation evidence available to this analysis.
| Fiscal year | P/E | Earnings basis | Currency basis | Ticker / reporting |
|---|---|---|---|---|
| 2025 | >100x | TTM | Native | USD / USD |
| 2024 | -- | TTM | Loss-making | USD / USD |
| 2023 | -- | TTM | Unavailable | USD / USD |
| 2022 | -- | Annual | Unavailable | USD / USD |
| 2021 | -- | Annual | Loss-making | USD / RUB |
P/E uses historical market capitalization and earnings known at each period. Cross-currency observations are normalized to USD using point-in-time FX rates.
Profitability and margins
The company currently shows very strong gross margins, meaning it makes a high profit on the basic rental of its computer chips due to temporary shortages. However, its actual operating margins are deeply negative because of the massive costs of building out its network and the heavy depreciation of its hardware. Analysts agree that the current high profit margins are likely at their peak and will face downward pressure. As more computer chips become available and big tech companies build their own systems, rental prices will fall. Additionally, rising energy costs to run these power-hungry data centers will squeeze profits. While some optimistic views suggest that the company will eventually achieve high, stable profits as it reaches a larger scale, cautious analysts warn that the stock's extremely high price is based on unrealistic expectations of permanent, software-like profit margins that a hardware utility simply cannot sustain.
The table compares up to five fiscal years of operating income and reported profitability margins.
| Fiscal year | Operating income | Operating margin | Net margin |
|---|---|---|---|
| 2025 | USD -596.2M | -112.5% | 19.2% |
| 2024 | USD -440.7M | -375.1% | -545.9% |
| 2023 | USD -327.5M | -1567.0% | 1154.5% |
| 2022 | USD -158M | -1170.4% | 5523.0% |
| 2021 | RUB -177.4M | -3.7% | -4.1% |
Balance sheet and leverage
At first glance, the company's balance sheet looks very strong, with over nine billion dollars in cash. This cash pile was recently boosted by selling new stock and issuing convertible debt. This immediate cash gives the company the safety net it needs to fund its near-term growth. However, the long-term outlook is much more fragile. Because the company has committed to spending twenty-five billion dollars on data centers, its cash will be used up very quickly. This will force the company to take on a massive amount of new debt. Analysts agree that the company's financial safety depends entirely on its ability to keep collecting upfront payments from its largest customers. If those customers pull back, the company's debt load could quickly become dangerous. There is disagreement over whether the company's physical assets, like its computer chips, provide enough safety for investors if things go wrong.
The table compares up to five fiscal years of debt, liquidity, net cash or debt, and current-ratio evidence.
| Fiscal year | Total debt | Cash + short-term investments | Net cash / (debt) | Current ratio |
|---|---|---|---|---|
| 2025 | USD 5B | USD 3.7B | USD 1.3B net debt | 3.08x |
| 2024 | USD 49.7M | USD 2.4B | USD 2.4B net cash | 9.60x |
| 2023 | RUB 584.8M | RUB 121.2M | RUB 468.7M net debt | 0.89x |
| 2022 | RUB 1.4B | RUB 1.1B | RUB 276.6M net debt | 1.28x |
| 2021 | RUB 1.9B | RUB 1.4B | RUB 824.5M net debt | 1.80x |
Net debt below zero is displayed as net cash. Current ratio is current assets divided by current liabilities.
Capex and investment intensity
The company is spending an astronomical amount of money—up to twenty-five billion dollars—to build new data centers and buy the latest computer chips. This heavy spending is aimed at securing a dominant position in the AI market before competitors can catch up. However, this aggressive strategy is a double-edged sword. On one hand, it allows the company to build a massive physical network that is hard for rivals to copy. On the other hand, computer chips become obsolete very quickly, usually in about three years. This means the company is trapped on a treadmill where it must constantly spend billions just to keep its technology up to date. Analysts agree that this spending will keep actual cash savings negative for years. They disagree on whether this massive investment will successfully pay off or if it will leave the company with outdated, useless equipment if customers decide to go elsewhere.
The table compares up to five fiscal years of capital expenditure and research-and-development investment.
| Fiscal year | Capital expenditure | R&D spend |
|---|---|---|
| 2025 | RUB 4.1B | USD 177.3M |
| 2024 | RUB 807.7M | USD 129.7M |
| 2023 | RUB 83.4M | USD 112.3M |
| 2022 | RUB 14.6M | USD 58.3M |
| 2021 | RUB 598.4M | RUB 647.5M |
Quarterly Forecast Scenarios
Nebius Group Quarterly Forecast Scenarios
One row per forecast quarter. Asset scenario targets are shown in USD; benchmark values are shown in USD.
Scroll to view all columns
| Timeline | Bear Case Stock Price (USD) | Base Case Stock Price (USD) | Bull Case Stock Price (USD) | S&P 500 benchmark (USD) |
|---|---|---|---|---|
| $183.28 | $197.47 | $245.81 | $736.09 | |
| $146.62 | $190.88 | $270.39 | $724.14 | |
| $109.97 | $171.47 | $248.76 | $713.57 | |
| $93.47 | $164.74 | $278.61 | $723.38 | |
| $79.45 | $164.02 | $328.76 | $716.61 | |
| $71.51 | $163.12 | $309.03 | $738.34 | |
| $64.35 | $170.82 | $355.39 | $746.91 | |
| $57.92 | $177.42 | $426.46 | $771.93 | |
| $55.02 | $179.41 | $447.79 | $782.73 | |
| $52.27 | $173.33 | $403.01 | $807.31 | |
| $49.66 | $177.65 | $451.37 | $822.72 | |
| $49.66 | $190.90 | $550.67 | $846.10 | |
| $52.14 | $199.27 | $594.72 | $852.53 | |
| $54.75 | $203.44 | $654.20 | $882.91 | |
| $57.49 | $204.93 | $621.49 | $900.46 | |
| $57.49 | $218.85 | $714.71 | $922.67 | |
| $57.49 | $224.66 | $771.89 | $935.91 | |
| $57.49 | $224.69 | $741.01 | $952.30 | |
| $57.49 | $237.05 | $844.75 | $976.28 | |
| $57.49 | $256.59 | $996.81 | $1,002 |
Research Provenance
References & Context
This Nebius Group consensus analysis combines structured market evidence with independent AI-agent forecasts. External references below are limited to sources recorded by the researcher agents for this forecast batch.
Primary analysis inputs
- iPulse AI Multi-Agent Forecasts — independent analyst personas, model outputs, and consensus synthesis.
- iPulse AI Global Events Context — macroeconomic, geopolitical, regulatory, and industry-event context.
- Structured market history — prices, distributions, volatility, identifiers, and listing metadata.
- Company earnings and financial statements — revenue, profitability, balance-sheet, cash-flow, and investment trends.
Independent AI Advisor panel
- AI Advisors
- 12
- AI Researchers
- 7
- AI Thinkers
- 5
Context retained with this Consensus
The same public-safe market, global-event, and fundamental context supplied to the AI Advisor panel.
Global context snapshot
2025 Full-Year Global Market and World-Events Context
Download Archived SnapshotCoverage 2025-01-01 to 2025-12-31 · Knowledge cutoff 2025-12-31
- File size
- 90.8K bytes
- Words
- 12.8K words
- Characters
- 90.8K characters
This full-year context package covers the principal geopolitical, economic, monetary-policy, technology, trade, energy, and institutional developments that shaped global markets during 2025. It gives the forecasting model a chronological account of major world events together with their likely transmission into growth, inflation, interest rates, supply chains, commodities, currencies, public markets, and sector-level investment conditions.
The package also includes monthly and quarterly macroeconomic and cross-asset reference tables spanning US and international growth, central-bank policy, sovereign yields, major equity indices, foreign exchange, energy, industrial and precious metals, and digital assets. Quarterly and full-year high-impact summaries are integrated; monthly quantitative series remain working values pending final audit, and that qualification is part of the preserved context.
| Top 3 Market Shifts From File | Date | Status |
|---|---|---|
| DeepSeek shock and AI economics reset | 2025-01-27 | OPEN ENDED TREND |
| US tariff regime escalation and trade-system rupture | 2025-02-01 | ACTIVE POLICY REGIME |
| Federal Reserve easing cycle after a prolonged hold | 2025-09-17 | ACTIVE POLICY REGIME |
Representative Sources of the Context File
And more sources from the retained context package.
2026 Year-to-Date Global Market Context through 2026-05-31
Download Archived SnapshotCoverage 2026-01-01 to 2026-05-31 · Knowledge cutoff 2026-05-31
- File size
- 78K bytes
- Words
- 10.9K words
- Characters
- 78K characters
This year-to-date package described the geopolitical, macroeconomic, monetary-policy, technology, trade, energy, and cross-asset developments available through the batch knowledge cutoff of 2026-05-31.
It supplied dated market and policy context, including rates, sovereign yields, equities, foreign exchange, energy, metals, and digital assets, for the forecast generation workflow.
| Top 3 Market Shifts From File | Date | Status |
|---|---|---|
| The Iran and Strait of Hormuz conflict shocked energy markets | 2026-02-28 | STARTED AND ONGOING |
| U.S. monetary policy entered the Warsh transition | 2026-01-30 | STARTED AND ACTIVE POLICY TRANSITION |
| Agentic AI and infrastructure spending kept expanding | 2026-01-01 | OPEN ENDED |
Representative Sources of the Context File
And more sources from the retained context package.
Fundamental context
Income statement
34 fieldscostOfRevenue · currency_symbol · date · depreciationAndAmortization · +30 more fields
Balance sheet
64 fieldsaccountsPayable · accumulatedAmortization · accumulatedDepreciation · accumulatedOtherComprehensiveIncome · +60 more fields
Cash flow
32 fieldsbeginPeriodCashFlow · capitalExpenditures · cashAndCashEquivalentsChanges · cashFlowsOtherOperating · +28 more fields
Outstanding shares
4 fieldsdate · dateFormatted · shares · sharesMln
annual: 2010-01-01–2026-01-01, 17 periods; quarterly: 2023-06-30–2026-03-31, 12 periods
Currencies cited: USD (quote USD; primary reporting USD; converted/valuation USD).