2026 Q1 -tulosraportti
63 päivää sitten
‧51 min
Tarjoustasot
Ei dataa
Viimeisimmät kaupat
| Aika | Hinta | Määrä | Ostaja | Myyjä |
|---|---|---|---|---|
| - | - | - | - |
Huomioi, että vaikka osakkeisiin säästäminen on pitkällä aikavälillä tuottanut hyvin, tulevasta tuotosta ei ole takeita. On olemassa riski, että et saa sijoittamiasi varoja takaisin.
Välittäjätilasto
Dataa ei löytynyt
Yhtiötapahtumat
Datan lähde: FactSet, Quartr| Seuraava tapahtuma | |
|---|---|
2026 Q2 -tulosraportti 28.7. |
| Menneet tapahtumat | ||
|---|---|---|
2026 Q1 -tulosraportti 13.5. | ||
2025 Q4 -tulosraportti 12.2. | ||
2025 Q3 -tulosraportti 11.11.2025 | ||
2025 Q2 -tulosraportti 7.8.2025 | ||
2025 Q1 -tulosraportti 20.5.2025 |
Asiakkaat katsoivat myös
Foorumi
Liity keskusteluun Nordnet Socialissa
Kirjaudu
·3 t sittenEventful day: GPU spot prices in free fall (Kalshi: B200 -31 % in one month), NBIS breaks below $200 on cash burn fear — and Nebius responds the same day with asset-light partner model Four things hit simultaneously. (1) Kalshi has for the first time made GPU rental prices transparent and tradable — and the curves show steeply falling spot prices. (2) The market has shifted focus: large contract headlines no longer lift the stock — investors demand proof that capex ($20–25 billion for 2026) translates into revenue, utilization, and cash conversion. (3) Nebius' defense is contract coverage and owned infrastructure: >75 % of contracted power is now at company-owned facilities. (4) And at 2:30 PM today, Nebius launched an asset-light partner model that directly addresses both the cash burn fear and the commoditization threat from Kalshi transparency. Kalshi Compute Forward Curves — launched today 15.07.2026: market-based forward curves for B200/H200/A100. B200: peak $6.11/t (30.05.2026) → $4.22/t (21.06.2026) = -31 % in under one month. Mansour's own figures: hyperscalers spend >$700 billion on compute this year, and the market is expected to reach $7–10 trillion by 2030 — a liquid derivatives market could be 10–20x larger than the spot market. His analogy: compute is where oil was before NYMEX — only OTC trading. Next step: futures and perps. - Reflection AI contract (announced 14.07.2026): >$1 billion compute agreement — Reflection AI (American startup, open models + autonomous code agents) gains access to Nvidia GB300 via Nebius infrastructure until 2029. Distributed over the period: probably <$300M/year. - Capacity (Q1 report + July): contracted power >2 GW (end 2025) → >3.5 GW (Q1 2026) → target ≥4 GW in 2026. New Pennsylvania facility up to 1.2 GW — Nebius' 2nd GW-campus in USA (after Independence, MO). Pipeline generation +3.5x sequentially in Q1. GPU demand still exceeds capacity. - CapEx/financing: 2026 guidance raised to $20–25 billion (from $16–20 billion). Media coverage of Alabama lawsuit and "Meta cloud threat" as share price pressure at Q1. - Asset-light partner model — launched today 15.07.2026 at 2:30 PM (press release, EQS/Nebius IR): Partners (data center developers, infrastructure investors, regional operators, national AI projects) finance, own, and operate the facilities — Nebius provides system architecture, supply chain access, hardware design, software/service stack, and brings capacity to market via its own sales organization. Revenue models: revenue sharing, licensing fees, commissions, and committed capacity. Volozh: "much better margins than conventional wholesale bare-metal contracts". First agreements already entered — but no names or amounts disclosed. High-margin growth with "minimal incremental capital requirements"". Price context: $194.09 (-7.8 %) at closing time 14.07 — below the $200 level. Down despite contract news; the market prices cash burn and execution, not order intake. (From +28 % last 30 days to correction; this year's earlier rally provides a high expectation base.) Possible impact: Neutral to slightly bullish — the partner model changed today's picture. - Bear: Steep spot price fall + price transparency gives customers negotiating power; capex increase without corresponding EBITDA proof increases financing risk (cf. $4.3 billion convertible); sentiment shift = contracts alone no longer lift the price. Partner model: no figures yet, revenue shared with partner, SLA responsibility for facilities Nebius does not physically control, and the $20–25 billion in owned capex remains unchanged. - Bull: The asset-light model directly addresses the market's two main objections — capex-free growth with higher margins, and positioning as a platform/distribution instead of pure capacity rental just as Kalshi commoditizes the GPU hour. The contract book grows faster than capacity; >75 % owned infrastructure supports margins; Kalshi futures can let Nebius hedge unsold capacity.
- 3 t sittenNebius introduces business model to scale AI cloud globally through infrastructure partnerships Pairs Nebius's systems architecture, software stack and customers with partner capacity Gives data center developers, infrastructure investors, regional operators and national AI projects worldwide route to tap fast-growing AI cloud market Generates high-margin revenue stream with minimal incremental capital requirements Amsterdam, July 15, 2026 --- Nebius (Nasdaq: NBIS), the AI cloud company, today announced a new business model that lets infrastructure partners deploy Nebius's full-stack AI cloud platform in their own AI data centers. The model brings additional capacity to Nebius customers, and expands the availability of value-added AI compute globally at a time when demand continues to outstrip supply. Under the model, partners finance and own the infrastructure and hardware, and operate the data centers. Nebius supplies its systems architecture and supply-chain access; deploys and maintains its hardware design and software and services stack on the partner infrastructure; and takes the resulting capacity to market through its global sales organization. Partners get fully-owned AI infrastructure assets, designed to Nebius standards, and a fast route to serve the AI cloud market. Nebius's architecture and platform transform a partner's raw capacity into a production-ready AI cloud, which Nebius then connects to customers. Because Nebius brings the demand, partners can begin generating a return as soon as the capacity goes live. For Nebius, this asset-light approach expands the capacity it can offer its customers, such as AI natives and enterprises, with minimal incremental capital requirements. Partners' data centers will join the Nebius capacity pool, adding incremental capacity to that coming online from Nebius's owned data centers and colocations. Arkady Volozh, founder and CEO of Nebius, said: "Our new asset-light model gives infrastructure partners a flexible way to benefit from the explosive growth of AI. Our software allows partners to reach a much wider customer base with much better margins than conventional wholesale bare-metal contracts. We're inviting data center investors, regional partners and others with capacity or capital to contribute to join us in serving this demand -- combining their assets and local strengths with Nebius's technology, platform, operational expertise and customer demand." Nebius anticipates pursuing a variety of economic arrangements under this partnership model, including revenue-sharing agreements, licensing fees and commissions, as well as committed capacity arrangements that would provide Nebius with access to additional compute to be sold to customers. The company has already entered into initial arrangements under this asset-light model. As part of the partnership agreements, Nebius will equip partner teams to run the site and will remain responsible for the cloud software and service levels, while the partner manages the facility and hardware. Customers receive the same standard of service whether they run on Nebius's own infrastructure or a partner's.You're hitting the core — and the answer is that it's neither of the two pure interpretations, but a franchise model (think Marriott). Marriott doesn't own the hotel buildings; real estate investors do. Marriott provides the brand, booking system, and guest flow — and takes its share. Nebius does the same: the partner provides capital, buildings, and power; Nebius provides the software-stack, Nvidia-allocation (supply chain access), operational standards — and most importantly "the customers", via its own sales organization. Regarding your objection to 1 — «why share profits when demand is sky-high?» Because the bottleneck is not capital. A regional data center investor with $2 billion and a power plant lacks three things he cannot buy with money: GB300-allocation from Nvidia (goes to established relationships), a production-ready cloud-stack (takes years to build — orchestration, multi-tenancy, InfiniBand, SLAs), and AI-native customers who sign billion-dollar contracts. His alternative is not «keeping all AI-cloud profits himself» — it's "bare-metal wholesale at thin margins", or empty capacity. Volozh explicitly states: partners get «much better margins than conventional wholesale bare-metal contracts». A share of value-added cloud revenue beats 100 % of a commodity lease. The Kalshi curves clearly show what happens with commodity leases. Regarding interpretation 2 — selling the stack to Meta/CoreWeave/IREN: No, and that's deliberate. CoreWeave has its own stack, Meta is a customer (not an operator-buyer), and IREN is building its own cloud. More importantly: if Nebius sold the software detached (à la VMware), they would arm competitors and lose the customer relationship. The model keeps "software + demand bundled" — that's the moat itself. The partner never gets the stack without Nebius also owning the customer relationship. The target group is therefore non-competitors: infrastructure funds, data center developers, national AI projects (sovereign AI is probably the big win here — countries that want their own AI infrastructure but lack operator expertise). So you understand it correctly in direction 1 — but with one clarification that changes the calculation: it's not «rent out your GPUs to us», it's «your capital + our entire business on top». Your scenario 2 would have been more explosive for the stock, yes — but it would also have been selling the family silver. The franchise variant is the capital-light compromise that scales without creating tomorrow's competitor.
·10 t sittenSomething I haven't thought much about but which Will have a Gigantic impact on the Earnings margins for Cloud providers is the development and replacement from previous A100 and H100/H200 chips over to Blackwell B200/GB300 and vera rubin. Cost pr. Token falls by 35% between H100 and Blackwell and then it falls by an additional 90% FROM Blackwell to Vera rubin! This means that the price pr. 1 token run on H100 will fall on Vers rubin down to 6.5% of the price the Providers (Nebius, coreweave, oracle) pay for the størm that runs through the chip at the current time. Bullish long on AI infrastructure as a whole.
Hey, I also listen quite a bit to Andreas Steno - where did you hear what you're referring to? :-)- ·10 t sittenNew York's statewide temporary halt ("moratorium") for new data centers (≥50 MW) for up to one year has no direct impact on Nebius. There are two data centers with regulatory risk. Birmingham (Alabama) introduced a 180-day moratorium on "hyperscale" data centers in March 2026. Nebius was exempt (applications were submitted in January 2026; the first building permits were granted in April, with construction work for over $855 million under contract), and the moratorium expired in June, only to be replaced by a stricter regulation for data centers. The conflict, however, ended up in the legal system: A legal document from May claims that the city helped developers circumvent the moratorium. On July 11, a judge ruled that the neighbors' lawsuit in Oxmoor Valley can proceed, and imposed restrictions on construction times following a motion for sanctions for violating court orders. A hearing on a temporary construction halt is ongoing July 14–16 and could completely stop the approximately 300 MW project pending trial. A separate lawsuit from the Greater Birmingham Humane Society, filed July 10, disputes the permits. The situation in Vineland (NJ) has evolved from a simple "deferred permit" to something more technical and with more at stake. NJDEP (the environmental authorities in New Jersey) has sent notices of technical deficiencies related to the question of "common control": whether the Vineland facility should be approved as an independent unit, or merged with Corning's nearby industrial facility and counted as a single source of pollution. If NJDEP approves the merger, Nebius will be subject to the stricter federal approval regime "Title V/PSD" – which involves oversight from EPA and mandatory public hearings, potentially extending the timeline by 1–2 years. This is a far more serious consequence than an administrative delay. No public decision from NJDEP was available as of mid-July 2026. Nebius/DataOne management describes this as "routine" and publicly maintains the goal of delivery in November 2026. Local politics are favorable to them – the Vineland city council approved a five-year tax agreement (PILOT) in January 2026 – but construction continues without final air and water discharge permits from DEP, without an independent environmental impact assessment, and with limited public insight. The Vineland timeline has not been officially shifted, and management maintains the November forecast. PS! I am holding all my shares in Nebius
Yllä olevat kommentit ovat peräisin Nordnetin sosiaalisen verkoston Nordnet Socialin käyttäjiltä, eikä niitä ole muokattu eikä Nordnet ole tarkastanut niitä etukäteen. Ne eivät tarkoita, että Nordnet tarjoaisi sijoitusneuvoja tai sijoitussuosituksia. Nordnet ei ota vastuuta kommenteista.
Uutiset
Tämän sivun uutiset ja/tai sijoitussuositukset tai otteet niistä sekä niihin liittyvät linkit ovat mainitun tahon tuottamia ja toimittamia. Nordnet ei ole osallistunut materiaalin laatimiseen, eikä ole tarkistanut sen sisältöä tai tehnyt sisältöön muutoksia. Lue lisää sijoitussuosituksista.
2026 Q1 -tulosraportti
63 päivää sitten
‧51 min
Uutiset
Tämän sivun uutiset ja/tai sijoitussuositukset tai otteet niistä sekä niihin liittyvät linkit ovat mainitun tahon tuottamia ja toimittamia. Nordnet ei ole osallistunut materiaalin laatimiseen, eikä ole tarkistanut sen sisältöä tai tehnyt sisältöön muutoksia. Lue lisää sijoitussuosituksista.
Foorumi
Liity keskusteluun Nordnet Socialissa
Kirjaudu
- ·3 t sittenEventful day: GPU spot prices in free fall (Kalshi: B200 -31 % in one month), NBIS breaks below $200 on cash burn fear — and Nebius responds the same day with asset-light partner model Four things hit simultaneously. (1) Kalshi has for the first time made GPU rental prices transparent and tradable — and the curves show steeply falling spot prices. (2) The market has shifted focus: large contract headlines no longer lift the stock — investors demand proof that capex ($20–25 billion for 2026) translates into revenue, utilization, and cash conversion. (3) Nebius' defense is contract coverage and owned infrastructure: >75 % of contracted power is now at company-owned facilities. (4) And at 2:30 PM today, Nebius launched an asset-light partner model that directly addresses both the cash burn fear and the commoditization threat from Kalshi transparency. Kalshi Compute Forward Curves — launched today 15.07.2026: market-based forward curves for B200/H200/A100. B200: peak $6.11/t (30.05.2026) → $4.22/t (21.06.2026) = -31 % in under one month. Mansour's own figures: hyperscalers spend >$700 billion on compute this year, and the market is expected to reach $7–10 trillion by 2030 — a liquid derivatives market could be 10–20x larger than the spot market. His analogy: compute is where oil was before NYMEX — only OTC trading. Next step: futures and perps. - Reflection AI contract (announced 14.07.2026): >$1 billion compute agreement — Reflection AI (American startup, open models + autonomous code agents) gains access to Nvidia GB300 via Nebius infrastructure until 2029. Distributed over the period: probably <$300M/year. - Capacity (Q1 report + July): contracted power >2 GW (end 2025) → >3.5 GW (Q1 2026) → target ≥4 GW in 2026. New Pennsylvania facility up to 1.2 GW — Nebius' 2nd GW-campus in USA (after Independence, MO). Pipeline generation +3.5x sequentially in Q1. GPU demand still exceeds capacity. - CapEx/financing: 2026 guidance raised to $20–25 billion (from $16–20 billion). Media coverage of Alabama lawsuit and "Meta cloud threat" as share price pressure at Q1. - Asset-light partner model — launched today 15.07.2026 at 2:30 PM (press release, EQS/Nebius IR): Partners (data center developers, infrastructure investors, regional operators, national AI projects) finance, own, and operate the facilities — Nebius provides system architecture, supply chain access, hardware design, software/service stack, and brings capacity to market via its own sales organization. Revenue models: revenue sharing, licensing fees, commissions, and committed capacity. Volozh: "much better margins than conventional wholesale bare-metal contracts". First agreements already entered — but no names or amounts disclosed. High-margin growth with "minimal incremental capital requirements"". Price context: $194.09 (-7.8 %) at closing time 14.07 — below the $200 level. Down despite contract news; the market prices cash burn and execution, not order intake. (From +28 % last 30 days to correction; this year's earlier rally provides a high expectation base.) Possible impact: Neutral to slightly bullish — the partner model changed today's picture. - Bear: Steep spot price fall + price transparency gives customers negotiating power; capex increase without corresponding EBITDA proof increases financing risk (cf. $4.3 billion convertible); sentiment shift = contracts alone no longer lift the price. Partner model: no figures yet, revenue shared with partner, SLA responsibility for facilities Nebius does not physically control, and the $20–25 billion in owned capex remains unchanged. - Bull: The asset-light model directly addresses the market's two main objections — capex-free growth with higher margins, and positioning as a platform/distribution instead of pure capacity rental just as Kalshi commoditizes the GPU hour. The contract book grows faster than capacity; >75 % owned infrastructure supports margins; Kalshi futures can let Nebius hedge unsold capacity.
- 3 t sittenNebius introduces business model to scale AI cloud globally through infrastructure partnerships Pairs Nebius's systems architecture, software stack and customers with partner capacity Gives data center developers, infrastructure investors, regional operators and national AI projects worldwide route to tap fast-growing AI cloud market Generates high-margin revenue stream with minimal incremental capital requirements Amsterdam, July 15, 2026 --- Nebius (Nasdaq: NBIS), the AI cloud company, today announced a new business model that lets infrastructure partners deploy Nebius's full-stack AI cloud platform in their own AI data centers. The model brings additional capacity to Nebius customers, and expands the availability of value-added AI compute globally at a time when demand continues to outstrip supply. Under the model, partners finance and own the infrastructure and hardware, and operate the data centers. Nebius supplies its systems architecture and supply-chain access; deploys and maintains its hardware design and software and services stack on the partner infrastructure; and takes the resulting capacity to market through its global sales organization. Partners get fully-owned AI infrastructure assets, designed to Nebius standards, and a fast route to serve the AI cloud market. Nebius's architecture and platform transform a partner's raw capacity into a production-ready AI cloud, which Nebius then connects to customers. Because Nebius brings the demand, partners can begin generating a return as soon as the capacity goes live. For Nebius, this asset-light approach expands the capacity it can offer its customers, such as AI natives and enterprises, with minimal incremental capital requirements. Partners' data centers will join the Nebius capacity pool, adding incremental capacity to that coming online from Nebius's owned data centers and colocations. Arkady Volozh, founder and CEO of Nebius, said: "Our new asset-light model gives infrastructure partners a flexible way to benefit from the explosive growth of AI. Our software allows partners to reach a much wider customer base with much better margins than conventional wholesale bare-metal contracts. We're inviting data center investors, regional partners and others with capacity or capital to contribute to join us in serving this demand -- combining their assets and local strengths with Nebius's technology, platform, operational expertise and customer demand." Nebius anticipates pursuing a variety of economic arrangements under this partnership model, including revenue-sharing agreements, licensing fees and commissions, as well as committed capacity arrangements that would provide Nebius with access to additional compute to be sold to customers. The company has already entered into initial arrangements under this asset-light model. As part of the partnership agreements, Nebius will equip partner teams to run the site and will remain responsible for the cloud software and service levels, while the partner manages the facility and hardware. Customers receive the same standard of service whether they run on Nebius's own infrastructure or a partner's.You're hitting the core — and the answer is that it's neither of the two pure interpretations, but a franchise model (think Marriott). Marriott doesn't own the hotel buildings; real estate investors do. Marriott provides the brand, booking system, and guest flow — and takes its share. Nebius does the same: the partner provides capital, buildings, and power; Nebius provides the software-stack, Nvidia-allocation (supply chain access), operational standards — and most importantly "the customers", via its own sales organization. Regarding your objection to 1 — «why share profits when demand is sky-high?» Because the bottleneck is not capital. A regional data center investor with $2 billion and a power plant lacks three things he cannot buy with money: GB300-allocation from Nvidia (goes to established relationships), a production-ready cloud-stack (takes years to build — orchestration, multi-tenancy, InfiniBand, SLAs), and AI-native customers who sign billion-dollar contracts. His alternative is not «keeping all AI-cloud profits himself» — it's "bare-metal wholesale at thin margins", or empty capacity. Volozh explicitly states: partners get «much better margins than conventional wholesale bare-metal contracts». A share of value-added cloud revenue beats 100 % of a commodity lease. The Kalshi curves clearly show what happens with commodity leases. Regarding interpretation 2 — selling the stack to Meta/CoreWeave/IREN: No, and that's deliberate. CoreWeave has its own stack, Meta is a customer (not an operator-buyer), and IREN is building its own cloud. More importantly: if Nebius sold the software detached (à la VMware), they would arm competitors and lose the customer relationship. The model keeps "software + demand bundled" — that's the moat itself. The partner never gets the stack without Nebius also owning the customer relationship. The target group is therefore non-competitors: infrastructure funds, data center developers, national AI projects (sovereign AI is probably the big win here — countries that want their own AI infrastructure but lack operator expertise). So you understand it correctly in direction 1 — but with one clarification that changes the calculation: it's not «rent out your GPUs to us», it's «your capital + our entire business on top». Your scenario 2 would have been more explosive for the stock, yes — but it would also have been selling the family silver. The franchise variant is the capital-light compromise that scales without creating tomorrow's competitor.
- ·10 t sittenSomething I haven't thought much about but which Will have a Gigantic impact on the Earnings margins for Cloud providers is the development and replacement from previous A100 and H100/H200 chips over to Blackwell B200/GB300 and vera rubin. Cost pr. Token falls by 35% between H100 and Blackwell and then it falls by an additional 90% FROM Blackwell to Vera rubin! This means that the price pr. 1 token run on H100 will fall on Vers rubin down to 6.5% of the price the Providers (Nebius, coreweave, oracle) pay for the størm that runs through the chip at the current time. Bullish long on AI infrastructure as a whole.Hey, I also listen quite a bit to Andreas Steno - where did you hear what you're referring to? :-)
- ·10 t sittenNew York's statewide temporary halt ("moratorium") for new data centers (≥50 MW) for up to one year has no direct impact on Nebius. There are two data centers with regulatory risk. Birmingham (Alabama) introduced a 180-day moratorium on "hyperscale" data centers in March 2026. Nebius was exempt (applications were submitted in January 2026; the first building permits were granted in April, with construction work for over $855 million under contract), and the moratorium expired in June, only to be replaced by a stricter regulation for data centers. The conflict, however, ended up in the legal system: A legal document from May claims that the city helped developers circumvent the moratorium. On July 11, a judge ruled that the neighbors' lawsuit in Oxmoor Valley can proceed, and imposed restrictions on construction times following a motion for sanctions for violating court orders. A hearing on a temporary construction halt is ongoing July 14–16 and could completely stop the approximately 300 MW project pending trial. A separate lawsuit from the Greater Birmingham Humane Society, filed July 10, disputes the permits. The situation in Vineland (NJ) has evolved from a simple "deferred permit" to something more technical and with more at stake. NJDEP (the environmental authorities in New Jersey) has sent notices of technical deficiencies related to the question of "common control": whether the Vineland facility should be approved as an independent unit, or merged with Corning's nearby industrial facility and counted as a single source of pollution. If NJDEP approves the merger, Nebius will be subject to the stricter federal approval regime "Title V/PSD" – which involves oversight from EPA and mandatory public hearings, potentially extending the timeline by 1–2 years. This is a far more serious consequence than an administrative delay. No public decision from NJDEP was available as of mid-July 2026. Nebius/DataOne management describes this as "routine" and publicly maintains the goal of delivery in November 2026. Local politics are favorable to them – the Vineland city council approved a five-year tax agreement (PILOT) in January 2026 – but construction continues without final air and water discharge permits from DEP, without an independent environmental impact assessment, and with limited public insight. The Vineland timeline has not been officially shifted, and management maintains the November forecast. PS! I am holding all my shares in Nebius
Yllä olevat kommentit ovat peräisin Nordnetin sosiaalisen verkoston Nordnet Socialin käyttäjiltä, eikä niitä ole muokattu eikä Nordnet ole tarkastanut niitä etukäteen. Ne eivät tarkoita, että Nordnet tarjoaisi sijoitusneuvoja tai sijoitussuosituksia. Nordnet ei ota vastuuta kommenteista.
Tarjoustasot
Ei dataa
Viimeisimmät kaupat
| Aika | Hinta | Määrä | Ostaja | Myyjä |
|---|---|---|---|---|
| - | - | - | - |
Huomioi, että vaikka osakkeisiin säästäminen on pitkällä aikavälillä tuottanut hyvin, tulevasta tuotosta ei ole takeita. On olemassa riski, että et saa sijoittamiasi varoja takaisin.
Välittäjätilasto
Dataa ei löytynyt
Asiakkaat katsoivat myös
Yhtiötapahtumat
Datan lähde: FactSet, Quartr| Seuraava tapahtuma | |
|---|---|
2026 Q2 -tulosraportti 28.7. |
| Menneet tapahtumat | ||
|---|---|---|
2026 Q1 -tulosraportti 13.5. | ||
2025 Q4 -tulosraportti 12.2. | ||
2025 Q3 -tulosraportti 11.11.2025 | ||
2025 Q2 -tulosraportti 7.8.2025 | ||
2025 Q1 -tulosraportti 20.5.2025 |
2026 Q1 -tulosraportti
63 päivää sitten
‧51 min
Uutiset
Tämän sivun uutiset ja/tai sijoitussuositukset tai otteet niistä sekä niihin liittyvät linkit ovat mainitun tahon tuottamia ja toimittamia. Nordnet ei ole osallistunut materiaalin laatimiseen, eikä ole tarkistanut sen sisältöä tai tehnyt sisältöön muutoksia. Lue lisää sijoitussuosituksista.
Yhtiötapahtumat
Datan lähde: FactSet, Quartr| Seuraava tapahtuma | |
|---|---|
2026 Q2 -tulosraportti 28.7. |
| Menneet tapahtumat | ||
|---|---|---|
2026 Q1 -tulosraportti 13.5. | ||
2025 Q4 -tulosraportti 12.2. | ||
2025 Q3 -tulosraportti 11.11.2025 | ||
2025 Q2 -tulosraportti 7.8.2025 | ||
2025 Q1 -tulosraportti 20.5.2025 |
Foorumi
Liity keskusteluun Nordnet Socialissa
Kirjaudu
- ·3 t sittenEventful day: GPU spot prices in free fall (Kalshi: B200 -31 % in one month), NBIS breaks below $200 on cash burn fear — and Nebius responds the same day with asset-light partner model Four things hit simultaneously. (1) Kalshi has for the first time made GPU rental prices transparent and tradable — and the curves show steeply falling spot prices. (2) The market has shifted focus: large contract headlines no longer lift the stock — investors demand proof that capex ($20–25 billion for 2026) translates into revenue, utilization, and cash conversion. (3) Nebius' defense is contract coverage and owned infrastructure: >75 % of contracted power is now at company-owned facilities. (4) And at 2:30 PM today, Nebius launched an asset-light partner model that directly addresses both the cash burn fear and the commoditization threat from Kalshi transparency. Kalshi Compute Forward Curves — launched today 15.07.2026: market-based forward curves for B200/H200/A100. B200: peak $6.11/t (30.05.2026) → $4.22/t (21.06.2026) = -31 % in under one month. Mansour's own figures: hyperscalers spend >$700 billion on compute this year, and the market is expected to reach $7–10 trillion by 2030 — a liquid derivatives market could be 10–20x larger than the spot market. His analogy: compute is where oil was before NYMEX — only OTC trading. Next step: futures and perps. - Reflection AI contract (announced 14.07.2026): >$1 billion compute agreement — Reflection AI (American startup, open models + autonomous code agents) gains access to Nvidia GB300 via Nebius infrastructure until 2029. Distributed over the period: probably <$300M/year. - Capacity (Q1 report + July): contracted power >2 GW (end 2025) → >3.5 GW (Q1 2026) → target ≥4 GW in 2026. New Pennsylvania facility up to 1.2 GW — Nebius' 2nd GW-campus in USA (after Independence, MO). Pipeline generation +3.5x sequentially in Q1. GPU demand still exceeds capacity. - CapEx/financing: 2026 guidance raised to $20–25 billion (from $16–20 billion). Media coverage of Alabama lawsuit and "Meta cloud threat" as share price pressure at Q1. - Asset-light partner model — launched today 15.07.2026 at 2:30 PM (press release, EQS/Nebius IR): Partners (data center developers, infrastructure investors, regional operators, national AI projects) finance, own, and operate the facilities — Nebius provides system architecture, supply chain access, hardware design, software/service stack, and brings capacity to market via its own sales organization. Revenue models: revenue sharing, licensing fees, commissions, and committed capacity. Volozh: "much better margins than conventional wholesale bare-metal contracts". First agreements already entered — but no names or amounts disclosed. High-margin growth with "minimal incremental capital requirements"". Price context: $194.09 (-7.8 %) at closing time 14.07 — below the $200 level. Down despite contract news; the market prices cash burn and execution, not order intake. (From +28 % last 30 days to correction; this year's earlier rally provides a high expectation base.) Possible impact: Neutral to slightly bullish — the partner model changed today's picture. - Bear: Steep spot price fall + price transparency gives customers negotiating power; capex increase without corresponding EBITDA proof increases financing risk (cf. $4.3 billion convertible); sentiment shift = contracts alone no longer lift the price. Partner model: no figures yet, revenue shared with partner, SLA responsibility for facilities Nebius does not physically control, and the $20–25 billion in owned capex remains unchanged. - Bull: The asset-light model directly addresses the market's two main objections — capex-free growth with higher margins, and positioning as a platform/distribution instead of pure capacity rental just as Kalshi commoditizes the GPU hour. The contract book grows faster than capacity; >75 % owned infrastructure supports margins; Kalshi futures can let Nebius hedge unsold capacity.
- 3 t sittenNebius introduces business model to scale AI cloud globally through infrastructure partnerships Pairs Nebius's systems architecture, software stack and customers with partner capacity Gives data center developers, infrastructure investors, regional operators and national AI projects worldwide route to tap fast-growing AI cloud market Generates high-margin revenue stream with minimal incremental capital requirements Amsterdam, July 15, 2026 --- Nebius (Nasdaq: NBIS), the AI cloud company, today announced a new business model that lets infrastructure partners deploy Nebius's full-stack AI cloud platform in their own AI data centers. The model brings additional capacity to Nebius customers, and expands the availability of value-added AI compute globally at a time when demand continues to outstrip supply. Under the model, partners finance and own the infrastructure and hardware, and operate the data centers. Nebius supplies its systems architecture and supply-chain access; deploys and maintains its hardware design and software and services stack on the partner infrastructure; and takes the resulting capacity to market through its global sales organization. Partners get fully-owned AI infrastructure assets, designed to Nebius standards, and a fast route to serve the AI cloud market. Nebius's architecture and platform transform a partner's raw capacity into a production-ready AI cloud, which Nebius then connects to customers. Because Nebius brings the demand, partners can begin generating a return as soon as the capacity goes live. For Nebius, this asset-light approach expands the capacity it can offer its customers, such as AI natives and enterprises, with minimal incremental capital requirements. Partners' data centers will join the Nebius capacity pool, adding incremental capacity to that coming online from Nebius's owned data centers and colocations. Arkady Volozh, founder and CEO of Nebius, said: "Our new asset-light model gives infrastructure partners a flexible way to benefit from the explosive growth of AI. Our software allows partners to reach a much wider customer base with much better margins than conventional wholesale bare-metal contracts. We're inviting data center investors, regional partners and others with capacity or capital to contribute to join us in serving this demand -- combining their assets and local strengths with Nebius's technology, platform, operational expertise and customer demand." Nebius anticipates pursuing a variety of economic arrangements under this partnership model, including revenue-sharing agreements, licensing fees and commissions, as well as committed capacity arrangements that would provide Nebius with access to additional compute to be sold to customers. The company has already entered into initial arrangements under this asset-light model. As part of the partnership agreements, Nebius will equip partner teams to run the site and will remain responsible for the cloud software and service levels, while the partner manages the facility and hardware. Customers receive the same standard of service whether they run on Nebius's own infrastructure or a partner's.You're hitting the core — and the answer is that it's neither of the two pure interpretations, but a franchise model (think Marriott). Marriott doesn't own the hotel buildings; real estate investors do. Marriott provides the brand, booking system, and guest flow — and takes its share. Nebius does the same: the partner provides capital, buildings, and power; Nebius provides the software-stack, Nvidia-allocation (supply chain access), operational standards — and most importantly "the customers", via its own sales organization. Regarding your objection to 1 — «why share profits when demand is sky-high?» Because the bottleneck is not capital. A regional data center investor with $2 billion and a power plant lacks three things he cannot buy with money: GB300-allocation from Nvidia (goes to established relationships), a production-ready cloud-stack (takes years to build — orchestration, multi-tenancy, InfiniBand, SLAs), and AI-native customers who sign billion-dollar contracts. His alternative is not «keeping all AI-cloud profits himself» — it's "bare-metal wholesale at thin margins", or empty capacity. Volozh explicitly states: partners get «much better margins than conventional wholesale bare-metal contracts». A share of value-added cloud revenue beats 100 % of a commodity lease. The Kalshi curves clearly show what happens with commodity leases. Regarding interpretation 2 — selling the stack to Meta/CoreWeave/IREN: No, and that's deliberate. CoreWeave has its own stack, Meta is a customer (not an operator-buyer), and IREN is building its own cloud. More importantly: if Nebius sold the software detached (à la VMware), they would arm competitors and lose the customer relationship. The model keeps "software + demand bundled" — that's the moat itself. The partner never gets the stack without Nebius also owning the customer relationship. The target group is therefore non-competitors: infrastructure funds, data center developers, national AI projects (sovereign AI is probably the big win here — countries that want their own AI infrastructure but lack operator expertise). So you understand it correctly in direction 1 — but with one clarification that changes the calculation: it's not «rent out your GPUs to us», it's «your capital + our entire business on top». Your scenario 2 would have been more explosive for the stock, yes — but it would also have been selling the family silver. The franchise variant is the capital-light compromise that scales without creating tomorrow's competitor.
- ·10 t sittenSomething I haven't thought much about but which Will have a Gigantic impact on the Earnings margins for Cloud providers is the development and replacement from previous A100 and H100/H200 chips over to Blackwell B200/GB300 and vera rubin. Cost pr. Token falls by 35% between H100 and Blackwell and then it falls by an additional 90% FROM Blackwell to Vera rubin! This means that the price pr. 1 token run on H100 will fall on Vers rubin down to 6.5% of the price the Providers (Nebius, coreweave, oracle) pay for the størm that runs through the chip at the current time. Bullish long on AI infrastructure as a whole.Hey, I also listen quite a bit to Andreas Steno - where did you hear what you're referring to? :-)
- ·10 t sittenNew York's statewide temporary halt ("moratorium") for new data centers (≥50 MW) for up to one year has no direct impact on Nebius. There are two data centers with regulatory risk. Birmingham (Alabama) introduced a 180-day moratorium on "hyperscale" data centers in March 2026. Nebius was exempt (applications were submitted in January 2026; the first building permits were granted in April, with construction work for over $855 million under contract), and the moratorium expired in June, only to be replaced by a stricter regulation for data centers. The conflict, however, ended up in the legal system: A legal document from May claims that the city helped developers circumvent the moratorium. On July 11, a judge ruled that the neighbors' lawsuit in Oxmoor Valley can proceed, and imposed restrictions on construction times following a motion for sanctions for violating court orders. A hearing on a temporary construction halt is ongoing July 14–16 and could completely stop the approximately 300 MW project pending trial. A separate lawsuit from the Greater Birmingham Humane Society, filed July 10, disputes the permits. The situation in Vineland (NJ) has evolved from a simple "deferred permit" to something more technical and with more at stake. NJDEP (the environmental authorities in New Jersey) has sent notices of technical deficiencies related to the question of "common control": whether the Vineland facility should be approved as an independent unit, or merged with Corning's nearby industrial facility and counted as a single source of pollution. If NJDEP approves the merger, Nebius will be subject to the stricter federal approval regime "Title V/PSD" – which involves oversight from EPA and mandatory public hearings, potentially extending the timeline by 1–2 years. This is a far more serious consequence than an administrative delay. No public decision from NJDEP was available as of mid-July 2026. Nebius/DataOne management describes this as "routine" and publicly maintains the goal of delivery in November 2026. Local politics are favorable to them – the Vineland city council approved a five-year tax agreement (PILOT) in January 2026 – but construction continues without final air and water discharge permits from DEP, without an independent environmental impact assessment, and with limited public insight. The Vineland timeline has not been officially shifted, and management maintains the November forecast. PS! I am holding all my shares in Nebius
Yllä olevat kommentit ovat peräisin Nordnetin sosiaalisen verkoston Nordnet Socialin käyttäjiltä, eikä niitä ole muokattu eikä Nordnet ole tarkastanut niitä etukäteen. Ne eivät tarkoita, että Nordnet tarjoaisi sijoitusneuvoja tai sijoitussuosituksia. Nordnet ei ota vastuuta kommenteista.
Tarjoustasot
Ei dataa
Viimeisimmät kaupat
| Aika | Hinta | Määrä | Ostaja | Myyjä |
|---|---|---|---|---|
| - | - | - | - |
Huomioi, että vaikka osakkeisiin säästäminen on pitkällä aikavälillä tuottanut hyvin, tulevasta tuotosta ei ole takeita. On olemassa riski, että et saa sijoittamiasi varoja takaisin.
Välittäjätilasto
Dataa ei löytynyt