Overview
Revenue beat consensus by roughly 1.7%. The stock rose 34.14%. The gap between those two numbers is where this analysis has to start.
Nebius Group reported second quarter 2026 results before the US open on August 12 and closed at $259.20, up 34.14%. Per
Benzinga's coverage, revenue surged 454% year over year to $582.3 million against a Street estimate of $572.75 million, adjusted loss per share came in at 12 cents against a consensus loss of 62 cents, and the company said it has more than $40 billion in customer commitments.
A $9.5 million revenue beat plainly cannot drive a move of that size. Three things of different character did: a large beat on the earnings line, upgrades to the order and capacity metrics, and a technical factor most coverage skipped, namely that roughly 24% of the free float was held short. Separating those three layers is the only way to judge how much of this rally reflects a fundamental re-rating and how much reflects positioning.
Key Takeaways
Second quarter revenue was $582.3 million, up 454% year over year and 46% sequentially, beating the $572.75 million consensus by roughly 1.7%.
Adjusted loss per share was 12 cents against a 62-cent consensus loss, with adjusted net loss of $33.2 million versus $91.5 million a year earlier, while GAAP net loss from continuing operations was $190.4 million.
Nebius AI cloud revenue was $574.9 million, up 514% year over year and about 98% of group revenue, generating $285.7 million of adjusted EBITDA at a 49.7% margin.
Group adjusted EBITDA reached $236.2 million against an adjusted EBITDA loss of $21 million a year earlier, with annualized run-rate revenue hitting $3 billion at the end of June, up roughly 56% from $1.9 billion at the end of March.
Customer commitments exceed $40 billion, remaining performance obligations stand near $37 billion, and the company expects more than $9 billion in customer prepayments this year.
The quarter included four AI cloud deals averaging more than $1 billion each, with total contract value nearly quadrupling and new-customer contracts up more than ninefold.
The 2026 contracted power target was raised to 5 gigawatts from more than 4 gigawatts, with plans to deploy over 1 GW annually starting in 2027.
Second quarter capital expenditure was roughly $5.66 billion to $5.7 billion, above the $4.7 billion analysts expected, with cash and equivalents of $8.04 billion at quarter end.
The Revenue Beat Was Small. Something Else Wasn't.
$9.5 Million Against a 34% Move
Start with an honest comparison. Revenue of $582.3 million against a $572.75 million consensus is a beat of roughly $9.5 million, or about 1.7%. For a company with quarterly revenue under $600 million, that is close to in line.
The real beat came on earnings. An adjusted loss of 12 cents per share against a consensus 62-cent loss means the shortfall was a fifth of what was expected. Group adjusted EBITDA reached $236.2 million against an adjusted EBITDA loss of $21 million a year earlier, and that reversal says more about operating efficiency than the revenue growth rate does.
Segment data sharpens the picture. AI cloud revenue of $574.9 million grew 514% year over year and accounted for about 98% of group revenue, with $285.7 million of adjusted EBITDA at a 49.7% margin. A segment margin approaching half is unusual for a company still building capacity at scale.
Two Loss Measures Need Separating
One easily missed detail is the size of the gap between adjusted and GAAP results. Per
TradingKey's summary, second quarter GAAP net loss from continuing operations was $190.4 million, affected by higher interest expenses and non-operating headwinds.
Against an adjusted net loss of $33.2 million, that is a difference of roughly $157 million, arising mainly from interest, share-based compensation and other non-operating items. Interest cost in particular persists and grows with the debt load. Investors assessing the profitability improvement need to watch both lines rather than the adjusted measure alone.
Where the $40 Billion in Commitments Comes From
Four Billion-Dollar Deals in One Quarter
Order data forms the second pillar of this rally. Per
Reuters coverage published via Global Banking & Finance, the company signed four AI cloud deals averaging over $1 billion each during the quarter, total contract value nearly quadrupled, new-customer contracts jumped more than ninefold, customer commitments surpassed $40 billion, and the company expects more than $9 billion in customer prepayments this year.
One quantifiable detail concerns unit pricing. The same report notes that AI cloud deals signed in the quarter carried annual contract values above $20 million per megawatt, with that capacity expected to come online in late fourth quarter. That figure provides a conversion basis for assessing the revenue potential of the contracted power target.
Definitional differences matter here.
Data cited on Yahoo Finance shows remaining performance obligations near $37.0 billion against customer commitments above $40 billion, a gap of roughly $3 billion. Remaining performance obligations correspond to contractual obligations recognised under accounting standards, while the commitments figure uses a broader definition, and investors should confirm which one they are comparing.
Commitments Against Current Revenue
The ratio is what makes those figures meaningful. Against 2026 revenue guidance of $3.0 billion to $3.4 billion, $40 billion of commitments is roughly 12 times. Against $3 billion of annualized run-rate revenue at the end of June, it is roughly 13 times.
That says the order book has run far ahead of delivery capacity, so what determines the share price is no longer demand but the pace of fulfilment. The company's own targets confirm this: management reaffirmed an exiting-2026 annualized run-rate revenue goal of $7 billion to $9 billion, meaning more than a doubling from $3 billion within six months.
Capital Intensity Exceeds Peers
The cost of building capacity shows in the spending. Per Reuters, second quarter capital expenditure totaled about $5.7 billion, topping analysts' expectations of $4.7 billion, as the company continued investing aggressively in GPUs and data centre expansion.
The full-year figure conveys the intensity better. Per the
earnings call summary published via Yahoo Finance, management reaffirmed 2026 targets including revenue of $3 billion to $3.4 billion, an adjusted EBITDA margin of roughly 40%, capital expenditure of $20 billion to $25 billion, and the $7 billion to $9 billion annualized run-rate revenue goal.
At midpoints, $22.5 billion of capital expenditure is roughly seven times a $3.2 billion revenue midpoint. For comparison, peer CoreWeave's full-year capital expenditure runs roughly 2.9 times its revenue guidance. That gap indicates Nebius sits at an earlier and steeper point on the build curve, with a correspondingly longer verification period for capital efficiency.
Financing Is Diversifying
Multiple channels underwrite that intensity. The earnings call summary shows second quarter operating cash flow of $2.3 billion and quarter-end cash and equivalents of $8 billion, alongside the issuance of 12.7 million Class A shares through the at-the-market equity program at a weighted average price of $224, raising about $2.8 billion in gross proceeds, with 12.3 million shares remaining available as of June 30.
In July the company announced a $775 million asset-backed debt facility secured by contracted cash flows, priced at SOFR plus 250 basis points and backed by deployed GPU infrastructure and an investment-grade customer contract. Management cited the more than $40 billion committed backlog as capable of supporting further asset-backed financing.
There is a trade-off worth noting. At-the-market issuance means continuous dilution during a rising share price, while asset-backed financing monetises future contracted cash flows early. Both solve the funding problem, but they affect per-share value in different directions.
For investors following both AI compute and crypto, the competition between the two for power and high-end chips is worth tracking. On venues offering crypto trading alongside stock futures, such as
MEXC, relative performance between the two often signals shifts in capital preference early.
How Much of the 34% Came From Short Covering?
This is the most practical layer in understanding the move, and the part most coverage skips.
Per
Yahoo Finance's market analysis, Nebius's high short interest also appeared to magnify the move, with short positions representing approximately 24% of the company's free float, increasing the potential for gains as bearish positions were unwound. The same analysis noted that positive sentiment toward the wider neocloud sector contributed as well, with CoreWeave's second quarter results released after the close on August 11 helping trigger buying across other AI cloud infrastructure stocks.
The intraday path fits that pattern. Summaries show the stock opening near $226 with an intraday low around $216 before closing at $259.20, advancing in stair-step fashion with dips bought quickly. Roughly $33 between the open and the close indicates most buying occurred during the session rather than at the opening auction, consistent with progressive short covering.
To be explicit, the share of the move attributable to short covering cannot be quantified without trade-level and securities-lending data. But with 24% of the float held short, attributing the full 34% to a fundamental re-rating overstates how much the market actually repriced on this report.
Two other events contributed over the same period. The same analysis noted reports that Nvidia was directing an approximately $500 billion infrastructure funding initiative toward AI cloud specialists with Nebius prominently involved. That is reported information without official confirmation and should be treated cautiously. Separately, the company announced a sale of computing power to Reflection AI worth over $1 billion through 2029, which lifted the shares more than 4% in premarket trading.
Risks and Scenarios
The first risk is conversion speed. Annualized run-rate revenue must move from $3 billion at the end of June to $7 billion to $9 billion by year end, more than a doubling in six months. Management said deals signed in the quarter are expected to come online in late fourth quarter, making second-half delivery progress the decisive variable.
The second is capital intensity and financing. Full-year capital expenditure at midpoint runs roughly seven times the revenue midpoint, with the gap covered by prepayments, equity issuance and asset-backed debt. TradingKey's summary listed heavy cash deployment and concerns over future capital raises among the watch items, and noted DA Davidson cutting its price target from $250 to $175 with a Neutral rating.
The third is external constraint. Summaries note a one-year moratorium on new hyperscale data centres in New York, alongside industry-wide bottlenecks in power, transformers and grid equipment. These act directly on how fast capacity comes online.
The fourth is valuation and volatility. Summaries show the stock moving from a close near $182 on July 20 to $259.20 on August 12, with multiple double-digit daily swings along the way. With 24% of the float short, moves in both directions get amplified.
On scenarios, the base case is new capacity coming online as planned in the fourth quarter with annualized run-rate revenue approaching but not necessarily reaching the lower bound of guidance, shifting focus toward 2027 delivery capability. A second is capacity arriving faster while unit pricing holds above $20 million per megawatt, allowing the exit target to be reached early and shifting the valuation logic from revenue growth to earnings delivery. A third is power or equipment bottlenecks delaying delivery, where, given the sharp multi-week advance and with short-covering momentum already spent, the drawdown could exceed what any operational shortfall alone would justify.
Status distinctions: revenue, losses, commitment figures, capacity targets and guidance come from the company's disclosures and public call records. The Nvidia funding initiative comes from media reporting without official confirmation. Analyst price targets are third-party judgements. Short interest and intraday price data come from third-party sources and change constantly.
Exclusive View from James Mitchell
What deserves recording about this rally is that revenue beat by roughly 1.7% while the stock rose 34.14%. That ratio establishes that the market was not pricing this quarter's revenue. Decomposed, the move contains at least three components: an adjusted loss per share narrowing from an expected 62 cents to 12 cents, which is a re-rating of the path to profitability; $40 billion of commitments and a 5 gigawatt contracted power target, which is a re-rating of capacity visibility; and roughly 24% of the free float held short, which is a release of positioning. The three are entirely different in character. The first two can persist. The third happens once.
Three misreadings look likely. The first is treating $40 billion of commitments as committed revenue. Remaining performance obligations sit near $37 billion against commitments above $40 billion, a definitional gap of roughly $3 billion, and either way realisation depends on capacity arriving on schedule, with management stating that deals signed in the quarter come online in late fourth quarter. The second is underestimating capital intensity. Full-year capital expenditure at midpoint runs roughly seven times revenue, far above peer CoreWeave's roughly 2.9 times, indicating Nebius sits earlier on the build curve with a longer verification window, while interest cost already shows in the difference between a $190.4 million GAAP net loss and a $33.2 million adjusted net loss. The third is ignoring what the short structure contributed. With 24% of the float short, reading the full 34% as fundamental re-rating produces inflated expectations for further upside.
Three verifiable metrics deserve tracking. First, the quarterly trajectory of annualized run-rate revenue, since moving from $3 billion at the end of June to a $7 billion to $9 billion year-end target requires more than a doubling in six months, and the third quarter print alone will show whether that slope is achievable. Second, whether annual contract value per megawatt holds above $20 million, the most direct expression of unit economics, since a decline in newly signed pricing would indicate competition eroding pricing power, which reveals more about industry structure than headline contract totals do. Third, the pace at which the at-the-market program is used, with 12.3 million shares remaining available as of June 30, since heavy use at elevated prices would offset part of the per-share improvement.
For cross-asset investors, the lesson is that heavily shorted names react to good news out of proportion to the underlying improvement. This is not market failure but positioning dominating pricing at a specific moment, and crypto markets know the pattern well: instruments with deeply negative funding rates rebound far beyond what contemporaneous fundamental change can explain. The method for handling this is to decompose the price reaction into three parts, expectation revision, narrative re-rating and positioning release, of which only the first two are durable. From a risk management standpoint, chasing a short-covering rally carries more risk than establishing a position in a name with comparable fundamental improvement and normal short interest, because the force driving the price does not recur once it completes.
This analysis rests on company disclosures, public call records and credible reporting available now. Delivery progress, the financing environment and capacity constraints could each change the conclusion, and no single scenario should be treated as a fixed expectation.
FAQ
Why did NBIS stock rise 34% in one session?
Three factors compounded. First, a large earnings beat, with adjusted loss per share of 12 cents against a 62-cent consensus loss and group adjusted EBITDA swinging to $236.2 million from a $21 million loss a year earlier. Second, upgraded order and capacity metrics, with customer commitments above $40 billion and the 2026 contracted power target raised to 5 gigawatts from more than 4. Third, positioning, with roughly 24% of the free float held short and covering amplifying the move. Revenue itself beat by only about 1.7%.
Is the $40 billion in customer commitments committed revenue?
Not entirely. Data shows remaining performance obligations near $37.0 billion against commitments above $40 billion, so the definitions differ and investors should confirm which figure they are using. More importantly, revenue recognition depends on capacity being delivered on time, and management said AI cloud deals signed during the quarter are expected to come online in late fourth quarter. Against annual revenue guidance of $3.0 billion to $3.4 billion, commitments are roughly 12 times.
How profitable is the AI cloud business?
Considerably so at segment level. AI cloud revenue was $574.9 million in the second quarter, up 514% year over year and about 98% of group revenue, generating $285.7 million of adjusted EBITDA at a 49.7% margin. Group adjusted EBITDA was $236.2 million, near a 41% margin. Note that GAAP net loss from continuing operations remained $190.4 million, affected by higher interest expense and non-operating headwinds.
How large is the capital expenditure?
Second quarter capital expenditure was roughly $5.7 billion, above the $4.7 billion analysts expected. Full-year guidance is $20 billion to $25 billion against revenue guidance of $3.0 billion to $3.4 billion, roughly seven times revenue at midpoints. For comparison, peer CoreWeave's full-year capital expenditure runs roughly 2.9 times its revenue guidance, indicating Nebius sits at an earlier and steeper stage of capacity construction.
How is the company funding all this?
Through several channels. Second quarter operating cash flow was $2.3 billion with quarter-end cash and equivalents of $8 billion; the company sold 12.7 million Class A shares through its at-the-market program at a weighted average price of $224, raising about $2.8 billion; and in July it announced a $775 million asset-backed debt facility priced at SOFR plus 250 basis points, backed by deployed GPU infrastructure and an investment-grade customer contract. It also expects more than $9 billion in customer prepayments this year.
How much did short covering contribute to the move?
It cannot be precisely quantified without trade-level and securities-lending data. What is established is that analysis identified short positions at approximately 24% of the free float, increasing the potential for gains as bearish positions were unwound. The intraday path supports that reading: the stock opened near $226 with an intraday low around $216 before closing at $259.20, meaning most buying occurred during the session. Attributing the full 34% to fundamental re-rating is therefore not safe.
Can the year-end run-rate target be met?
It is demanding but backed by capacity. Annualized run-rate revenue reached $3 billion at the end of June, up roughly 56% from $1.9 billion at the end of March, against a reaffirmed year-end target of $7 billion to $9 billion, requiring more than a doubling in six months. The key variable is when deals signed during the quarter come online, which the company expects in late fourth quarter. The third quarter print will be the first real test of that slope.
How does this compare with CoreWeave?
Both are neoclouds at different stages. CoreWeave reported second quarter revenue of $2.575 billion with a $104.2 billion backlog and full-year capital expenditure running roughly 2.9 times revenue. Nebius reported $582.3 million of revenue with more than $40 billion of commitments and full-year capital expenditure running roughly seven times revenue. Nebius is smaller, growing faster and far more capital intensive, giving it a longer verification window on capital efficiency. CoreWeave's August 11 results did help trigger buying across the sector.
Disclaimer
This article is provided for informational and research purposes only and does not constitute investment advice, financial advice, legal advice, tax advice or any recommendation to transact. The financial figures, commitment data and guidance referenced here come from company disclosures, public earnings call records and credible reporting; all figures covering the remainder of 2026 are guidance rather than confirmed results and actual outcomes may differ materially; statements about third-party funding initiatives come from media reporting without official confirmation; analyst price targets and views are those institutions' judgements rather than statements of fact; short interest and intraday price data come from third-party sources and change constantly; and every ratio calculation in this article is a simple arithmetic illustration. Prices of crypto assets, equities and other related financial instruments can move sharply over short periods, and heavily shorted, highly valued names are especially prone to violent two-way moves around earnings windows, with the possibility of total loss of principal. Futures contracts referencing a stock carry leverage risk that magnifies losses proportionally, holding such contracts is not equity ownership and confers no shareholder rights, and the risk profile differs fundamentally from holding shares directly. Historical performance, technical indicators and on-chain data cannot guarantee future outcomes. Readers should conduct their own independent research, verify official information directly, and evaluate any decision against their own financial circumstances, investment objectives, experience and risk tolerance, consulting a qualified professional adviser where appropriate. The MEXC Crypto Pulse team accepts no liability for any direct or indirect loss arising from use of or reliance on the information in this article.
About the Author
James Mitchell specializes in technical analysis, market trends, and trading strategies for both Bitcoin and altcoins. Based in London, he has over 10 years of experience in financial markets. Before joining MEXC Learn, James worked as a senior analyst at a leading European investment firm, where he developed expertise in risk management and quantitative trading. His transition to cryptocurrency markets began in 2017, and he has since become recognized for his data-driven approach. He holds a Master's degree in Financial Economics from the London School of Economics. His analytical approach combines traditional technical analysis with on-chain metrics to provide readers with actionable insights.
Areas of Expertise:
Research References