Stefano retweeted
Michael Burry was right about the AI bubble:
• Palantir $PLTR: -25%
• Alphabet $GOOGL: -30%
• Nvidia $NVDA: -36%
• Micron $MU: -42%
• ServiceNow $NOW: -61%
• Nebius $NBIS: -63%
• Oracle $ORCL: -65%
• Sandisk $SNDK: -67%
• CoreWeave $CRWV: -74%
• IREN $IREN: -80%
All while the S&P 500 is only 6% off its high
Stefano retweeted
$SNDK en $753, desde un máximo de $2350.
$MU en $681, desde un máximo de $1260. 👀
Fui de los pocos que se mantuvieron firmes en llamarlo una burbuja en estas dos empresas y explicar las razones. Al mínimo indicio de desaceleración (slowdown), la burbuja iba a estallar.
Y eso es precisamente lo que estamos viendo.
Stefano retweeted
“No me extraña que te pasen cosas, chaval”
Le dice este impresentable a @Jon_Echeverria_.
El escaño del PP en Guipúzcoa es más necesario que nunca.
Stefano retweeted
These permabulls look like two kids who just got called down to the principal’s office… I’m dying 🤣🤣🤣 $NVDA
Stefano retweeted
Los que viven del mercado casi nunca son los que más ves en redes
Pocos seguidores
Cero reels
Nunca explican su método en público
Están ocupados: la plataforma abierta, las posiciones activas, el diario del día anterior para revisar
Antes de seguir a alguien, fijate de qué vive: del mercado o de contarte el mercado
Stefano retweeted
Edwards de Societe Generale advierte que la liquidez global está bajando fuertemente y que ese es uno de los problemas por los que la tecnología está cayendo. Hartnett también insiste en eso. Fuera de ellos dos, pocos se están dando cuenta del problema.
Stefano retweeted
"Since going public, the three $CRWV founders have sold, combined, over 3 billion in stock. Let me say that again."
"How much stock have the three founders bought back? Zero." -Steve Eisman
I sleep well at night knowing CoreWeave's $CRWV senior management has total confidence in their company.
Así va a caer toda la AI
NVIDIA’s RTX 5090 and RTX 5080 graphics cards have suddenly become cheaper in Japan, with price drops of up to 22%.
The RTX 5090 dropped from ¥1.28 million to under ¥1 million, while the RTX 5080 went from ¥359,800 to ¥299,800.
The price drops happened right after Jensen Huang’s reported resignation, but the two events don’t appear to be connected.
The price drops appear to be driven by excess stock and improved supply
24
Stefano retweeted
Algunos llevamos más de un año advirtiendo que toda la locura CAPEX alrededor de la IA no es más que ingeniería financiera basada en contratos circulares financiados por Nvidia.
Oracle fue el canario en la mina que destapó este asunto en septiembre de 2025.
Los CDS no mienten.
NVIDIA reportedly helped AI startups buy its own GPUs instead of lowering prices and selling its excess stock cheaper.
According to a new report, former CEO Jensen Huang used NVIDIA’s money to fund AI startups and data centers so they could keep buying its chips.
The problem is that NVIDIA had ordered far more chips and memory from TSMC than the market could actually use.
CFO Colette Kress pushed back against these deals, even cutting a planned $250 billion commitment to OpenAI down to $125 billion.
If true, NVIDIA may have been helping keep GPU prices high instead of letting excess supply make them cheaper.
Stefano retweeted
NVIDIA Allegedly Engineered Scarcity By Funding AI Startups To Buy Its GPUs tpu.me/kdn4
Steve Eisman literally explained why Anthropic's IPO story has a hole in it.
> His answer: "there's stuff in the Anthropic prospectus that is extremely negative."
" The Big Short investor sat on CNBC two weeks ago, after most people had read the filing and called it all good news. "
- The filing stops at the second quarter. No third quarter numbers. (01:34)
- He says there is enough in it to see them anyway. And his read is blunt: revenue growth "slowed dramatically." (02:45)
- His numbers: sequential growth of 147% in Q2. Then 45% in Q3. (04:41)
- Then the part that bothers him more. Divide revenue by compute spending. In Q2 the ratio was 2.08. In Q3 it fell to 1.7. Each dollar of compute started bringing back less. (05:57)
> His name for the cause is "the end of token maxing."
He is honest about the gap. The Q3 figures came from people outside the filing. "I can't confirm that it's 100% true," he says, though he calls them pretty reliable.
In January at Davos, Dario Amodei showed his own curve, zero to ten billion dollars in three years, and admitted it would be crazy if it kept going.
Eisman thinks the crazy part is already over.
The clip runs six minutes on CNBC's YouTube. Almost nobody stays for the efficiency ratio.
Stefano retweeted
Me acabo de ver las tres entrevistas de The Wild Project, y lo primero que me parece increíble es que Abascal haya perdido esta oportunidad de demostrar sus capacidades.
Sigo a @JordiWild desde hace muchos años y me parecía un entrevistador increíble, y además siempre lleva a personas muy interesantes a su podcast, pero creo que se ha coronado, le ha dicho en todo momento a los tres candidatos lo que le hubiera dicho yo o cualquiera de nosotros si los hubiéramos tenido delante, no se ha cortado un pelo.
Estas son las entrevistas que se echan de menos y no el lavado de cara que suelen hacer ciertos canales a los candidatos.
Un tío abiertamente de centro izquierda, que ha dado caña a todos por igual, más Jordi Wilds y menos séptimas.



































