>>> CRWV — SKEPTICAL TAKE


CRWV — SKEPTICAL TAKE
  • 3Q annualized Adj. EBITDA ≈ $3.4 B vs annualized interest ≈ $1.2 B.
  • Estimated GPU base ≈ $20 B. Even using a 10-yr depreciation (~$2 B/yr), profitability remains razor-thin (EBITDA $3.4 B – Dep $2 B – Int $1.2 B ≈ $0.2 B net).
  • Stronger than OpenAI, but still a reminder that scale ≠ free cash flow — disruption this large demands heavy upfront burn.

MACRO / RISK
  • Core risk: U.S. power generation. Data-centre load rising > 25% YoY vs grid expansion < 5%. Structural energy inflation incoming.
  • Power & cooling costs already up 30–40% YoY across Tier-1 DC markets. This will squeeze gross margins and become a political flashpoint for the new Trump administration as utility bills surge.
  • No bubble yet — but expect policy & fiscal intervention before valuations re-rate higher.

NVDA — ROADMAP / SIGNAL
  • Jensen Huang’s recent keynote outlined a clear implementation path for AI infrastructure across every sector.
  • NVIDIA forecasts data-centre TAM approaching $1 trillion by 2028, anchored by its new Blackwell architecture (up to 68× faster than Hopper).
  • Message was clear: “Every company becomes an AI factory.” Tokenization and AI compute integration are now mainstream, not theoretical.

CAPEX VS BUYBACKS
  • In this environment, capex > buybacks. Investing in compute, cooling, and grid capacity creates tangible future cash flow; financial engineering does not.
  • Capital deployed into productive infrastructure compounds value and secures leadership — a more credible long-term shareholder return strategy than incremental repurchases.

BOTTOM LINE
  • CRWV’s fundamentals aren’t broken, but the margin for error is tight.
  • AI infrastructure remains a capital-intensive, power-constrained trade — leaders will win by spending wisely, not by cutting reinvestment.

**** Vol Dispersion **** (BAYN, ENR, SIE, ASML, LVMH, ADS, BNP, DPB, ADYEN, OR)

  • Biggest IV-RV spread premiums:
    • Bayer (YTD: +42.7%; RSI: 53); IV 57.8 vs RV 23.6 with vol dispersion of 32.1; IV in the 98th percentile
    • Siemens Energy (YTD: +113.4%; RSI: 56); IV 60.4 vs RV 43.3 with vol dispersion of 15; IV in the 78th percentile
    • Siemens (YTD: +34.3%; RSI: 57); IV 35.7 vs RV 23.9 with vol dispersion of 9.8; IV in the 89th percentile
    • ASML (YTD: +31.9%; RSI: 53); IV 35.2 vs RV 25.7 with vol dispersion of 7.5; IV in the 42nd percentile
    • LVMH (YTD: +1.1%; RSI: 68); IV 28 vs RV 18.8 with vol dispersion of 7.1; IV in the 38th percentile
  • Biggest IV-RV spread discounts:
    • Adidas (YTD: -30.8%; RSI: 35); IV 27.8 vs RV 47 with vol dispersion of -21.2; IV in the 38th percentile
    • BNP Paribas (YTD: +25.5%; RSI: 40); IV 25 vs RV 38.6 with vol dispersion of -15.6; IV in the 42nd percentile
    • Deutsche Post (YTD: +37.3%; RSI: 77); IV 22.4 vs RV 34.3 with vol dispersion of -14; IV in the 25th percentile
    • Adyen (YTD: +2.1%; RSI: 54); IV 36.4 vs RV 45.6 with vol dispersion of -11.2; IV in the 46th percentile
    • L’Oreal (YTD: +6.2%; RSI: 38); IV 21.4 vs RV 30.3 with vol dispersion of -10.9; IV in the 22nd percentile

>>> US After Hours Summary: AMD +3.7% higher on analyst day strategy; UPXI +19%,

After Hours Summary: AMD +3.7% higher on analyst day strategy; UPXI +19%, CAE +6.2% and ALC +4.7% higher on earnings; OKLO -1.6% modestly lower on earnings

After Hours Gainers:

Companies trading higher in after hours in reaction to earnings/guidance: UPXI +19%, PAL +8.7%, CAE +6.2% (also announces organizational changes to simplify structure), LPTH +6%, ALC +4.7%, DOX +0.7% (also increases dividend; also announces several expanded and extended managed services agreements),

Companies trading higher in after hours in reaction to news: PROF +4.5% (distribution and supply agreement for TULSA-PRO and Sonalleve technologies), AMD +3.7% (unveils strategy to lead compute market and accelerate next phase of growth), FIVN +2.7% ($50 mln accelerated share repurchase agreement), BBCP +1.8% (subsidiary completes acquisition of C.G.A. Concrete Plumbing), IONS +1.4% (convertible notes offering), PSN +0.5% (awarded position on Pacific Deterrence Initiative Multiple Award Construction Contract), BRC +0.1% (name and ticker change), APAM +0.1% (reports October AUM), GTN +0.1% (expanded partnership with Memphis Grizzlies),

After Hours Losers:

Companies trading lower in after hours in reaction to earnings/guidance: STXS -7.1%, OKLO -1.6% (also expands collaboration with Idaho National Laboratory),

Companies trading lower in after hours in reaction to news: BHVN -7.4% (proposed public offering of $150 mln of common shares), ELDN -3.3% (proposed underwritten public offering of common stock), CNTA -3% (public offering of American Depositary Shares), AB -1% (reports October AUM), LCID -0.9% (convertible notes offering), DDD -0.4% (several new products it will showcase at Formnext 2025), BAX -0.2% (decreases dividend), CSR -0.2% (confirms Board of Trustees initiated review of strategic alternatives),

SCMP : China’s Moonshot claims to build models with fewer high-end AI chips than

China’s Moonshot claims to build models with fewer high-end AI chips than US rivals use
In a Reddit forum, a Moonshot AI executive says start-up is ‘outnumbered’ by US rivals in terms of ‘high-end GPUs’ used for training models

Chinese artificial intelligence firm Moonshot AI continues to develop AI models with fewer high-end graphics processing units (GPUs) than what its US rivals use, according to the Beijing-based start-up’s executives.
In a three-hour-long “ask me anything” session on Reddit on Monday evening, a Moonshot AI representative with the handle “ppwwyyxx” – the same moniker used by co-founder Wu Yuxin on X – said the company was “outnumbered” by rival US firms in terms of “high-end GPUs” used for AI model development.

He also confirmed that Kimi K2 Thinking, a new reasoning variant of its open-source Kimi K2 model, was trained on Nvidia’s older H800 GPUs, which were banned for export to China in late 2023.

That reflected how Chinese AI companies have been making the most of available resources on the mainland to create cutting-edge models, despite stringent US tech export restrictions.

With Kimi K2 Thinking’s release last week, Moonshot AI – a unicorn valued at US$3.3 billion and backed by Chinese tech giants like Alibaba Group Holding and Tencent Holdings – ignited fresh debate about another “DeepSeek moment” in the global AI industry, while raising questions about recent efforts by OpenAI and its CEO Sam Altman to secure more than US$1.4 trillion in infrastructure deals with the likes of Nvidia, Broadcom and Oracle.
Alibaba owns the South China Morning Post.

In the Reddit discussion, another Moonshot AI representative identified as “ComfortableAsk4494” – the online handle of founder Yang Zhilin – made a direct reference to OpenAI’s massive data centre buildout when asked when the Chinese firm planned to release its next-generation foundational model, the K3. Yang replied: “Before Sam’s trillion-dollar data centre is built.”
Asked why OpenAI was “burning so much money”, Moonshot AI representative “zxytim” – the same handle used by co-founder Zhou Xinyu on X – said: “Only Sam knows, we’ve got our own way and our own pace.”

Those comments cast light again on the cost-efficient strategy behind the development of Kimi K2 Thinking, which was trained at a mere cost of US$4.6 million, according to a CNBC report.

Moonshot AI’s Yang, however, dismissed the reported figure. “This is not an official number,” he wrote on the Reddit forum. “It is hard to quantify the training cost because a major part was research and experiments.”

Even without factoring in its costs, the latest model has impressed the AI community. Thomas Wolf, co-founder of open-source developer platform Hugging Face, posted on X that Kimi K2 Thinking was another case of an open-source model achieving industry-leading performance.

Last month, Li Zixuan – Zhipu AI’s head of global operations – questioned Facebook owner Meta Platforms’ AI talent poaching spree that reportedly involved US$100 million in signing bonuses.

“We don’t believe it should cost you a hundred million dollars to hire these people,” Li said on the Manifold podcast.

Both OpenAI and Meta have committed to huge investments in data centres and advanced GPUs, which they said were needed to train powerful AI systems that could far surpass human intelligence. Meanwhile, Chinese AI start-ups continued to be restricted from accessing those chips under US export controls.

(ZH) Trump Admin To Lend "Hundreds Of Billions" To Build Nuclear Power Plants

Trump Admin To Lend "Hundreds Of Billions" To Build Nuclear Power Plants

While the market is finally starting to grapple with the most unpleasant question of who will plug the funding gap needed to build out all the data centers required to make the AI dream a reality, a gap which Morgan Stanley recently calculated would be as large as $2.9 trillion in capex funding needs, of which at least $1 trillion will come in the form of debt (and mostly private debt)...
... there is another, just as critical question: who will fund the energy buildout that powers these data centers?
Recall, last December Morgan Stanley calculated that the US would need at least 36GW in new power to be brought online by 2028 to energize all the (yet to be built) data centers, a number which one year later is surely far higher.

And at a cost of $50-60BN per GW of power, we can quickly add several more trillion dollar that will be needed in the next several years: money, which as this Bloomberg article makes painfully clear, are simply not available right now.
So where will this money come from?

Why the US government of course.
By now it should be clear to all but the most purple-haired libs that the US will need nuclear power plants - both conventional and modular - and lots of them, to have even a remote chance of ever catching up to the ever growing energy needs... and as we said recently, one can print money, but one can't print energy.
But they sure can try.... and according to Secretary Chris Wright, nuclear power will receive most of the money from the Energy Department's Loan Programs Office (LPO) as the Trump administration pushes to quickly break ground on new reactors. According to Reuters, The LPO has hundreds of billions of dollars in financing aid, including loan guarantees for projects that struggle to get bank loans.
"We have significant lending authority at the loan program office," the Secretary of Energy said at a conference hosted by the American Nuclear Society in Washington D.C. "By far the biggest use of those dollars will be for nuclear power plants, to get those first plants built. The U.S. currently has no commercial nuclear reactors being built, though several intend to reverse their permanent shutdown status and open again, and there are other plans to build new large and small reactors."
During Trump's first term in the White House, the only use he made of the LPO was for financing reactors at the Vogtle nuclear power plant in Georgia. Expect a flood of debt deals in the coming weeks and months as the clock is ticking for the US to catch up to China, which currently has 29 reactors under construction to America's zero.
As we reported at the time, President Trump signed an executive order in May that called for the US to break ground on 10 large nuclear reactors by 2030, while Alphabet, Amazon, Meta Platforms and Microsoft have been investing billions of dollars to restart old nuclear plants, upgrade existing ones, and deploy new reactor technology to meet the electricity demand from artificial intelligence data centers. Others are also joining the fray, but as Sam Altman made clear, everyone is expecting the US government to the be the "insurer of last resort."
Wright said he expects electricity demand from AI to attract billions of dollars in equity capital to build new nuclear capacity from "very creditworthy providers." The Energy Department could match those private dollars by as much as four to one with low cost debt financing from the loan office, he said.
"When we leave office three years and three months from now, I want to see hopefully dozens of nuclear plants under construction," Wright said. Make that all hyperscaler CEOs too because without those dozens of nuclear plants, guess what: the AI bubble is going to burst in the most spectacular fashion.
Cooling towers at the Three Mile Island nuclear power plant in Middletown, Pennsylvania, Oct. 30, 2024
The Trump administration tipped its cards last month when it struck a deal with the owners of Westinghouse to invest $80 billion to build nuclear plants across the US. Westinghouse is owned by uranium miner Cameco and Brookfield Asset Management (both Canadian companies).
Westinghouse has designed a modern reactor called the AP1000 that can power more than 750,000 homes. CEO Dan Sumner said in July that Westinghouse would meet Trump's call to build large new plants with the AP1000 design. But Westinghouse has struggled in the past to build the AP1000 on time and on budget. And as a stark indication of just how capital intensive building NPPs can be, Westinghouse went bankrupt in 2017 from cost overruns at big nuclear projects in Georgia and South Carolina.
Which, naturally, opens the door wide open to far cheaper projects from small modular reactor developers such as Oklo and Nano Nuclear.
Cameco Chief Operating Officer Grant Isaac said last week that the U.S. government has a number of options available to facilitate the financing of Westinghouse reactors, including the Energy Department's loan office.
"We're assured that there is a lot of interest in investing this minimum $80 billion in order to begin the process," Isaac told investors on Cameco's third-quarter earnings call.
Under the terms of the October deal, Westinghouse could spin out as a separate, publicly-traded company with the U.S. government as a shareholder.

>>> US Gapping down

Gapping down
In reaction to earnings/guidance:
  • OM -27.1%, IHRT -9.6%, CRWV -9.6%, GEMI -8.9%, LIF -6.8% (also to acquire Nativo), BYND -6%, TDW -4.2%, HROW -4%, SE -3.9%, RGTI -3.3%, OCS -2.8% (also, stock offering), FRMI -2.1%, EGY -1.5%, WULF -1.3%, MLYS -1.2% (also, files mixed securities shelf offering), RPAY -1.1%
Other news:
  • VOR -32.1% (prices offering of 10.0 mln shares of common stock at $10.00 per share)
  • CMTL -15.3% (rebranding of its Terrestrial & Wireless Networks segment)
  • DTIL -8.4% (prices offering consisting of common stock and warrants)
  • AMBC -4.6% (rebrands to Octave Specialty Group)
  • CLSK -4.3% (prices offering of $1.15 bln of 0.00% Convertible Senior Notes due 2032)
  • INO -3.7% (proposes underwritten public offering)
  • ADCT -2.6% (stock offering by selling shareholders)
  • EHC -1.9% (opening of Rehabilitation Hospital of Amarillo)
  • ROL -1.8% (prices secondary offering 17,391,305 shares of its common stock by LOR, Inc. and Rollins Holding Company at $57.50 per share)
  • PROF -1.3% (regains distribution rights for TULSA-PRO)
  • MLYS -1.2% (files mixed securities shelf offering)
  • CODI -1.2% (to delay 10-q filing)
  • COGT -1% (public offerings of convertible notes and common stock)