If you are an AI investor, do not miss this. The 7 biggest AI companies have committed $4.5 trillion to future data center leases and purchase deals that do not show up as debt on their balance sheets. Only about 15% of their total obligations are actually on the books, according to J.P. Morgan. Look at the gap: Google: $113B on the books, $1.08T off Anthropic: $8B on, $865B off Meta: $110B on, $810B off Microsoft: $110B on, $575B off Nvidia: $37B on, $422B off Amazon: $223B on, $419B off Oracle: $157B on, $322B off Anthropic's off book commitments are over 100 times bigger than its recorded debt. The credit market has started to notice. The cost to insure the debt of these AI companies against default is rising fast. Hedge funds have also cut their exposure by 8%, the biggest pullback since 2016. And after the midterms, up to 37.5 GW of data centers planned for 2028 could be paused by new state rules. These companies signed trillions of dollars in future bills, betting AI demand keeps growing. What happens if it doesn't?

Key events for next week. -Monday: Constellation Energy reports before the open, offering insight into political uncertainty surrounding power infrastructure for AI data centers. - Tuesday: CPI inflation data will be closely watched for signs of easing price pressures and potential implications for Fed policy. CoreWeave reports after the close, with investors expecting details on how the heavily indebted company plans to finance expansion and manage borrowing costs. - Wednesday: Corteva reports, as investors assess the factors behind its recent stock weakness. - Thursday: Nebius reports before the open, with attention on new orders, completed data centers, project pipelines and potential construction delays or cost overruns. Cisco and Applied Materials also report after the close, while Kraft Heinz holds an investor meeting. - Friday: Producer Price Index (PPI) data provides another key reading on inflation and the potential direction of monetary policy. $CRWV $NBIS $CEG $ON $SPX

SanDisk has experienced a larger decline, likely because it’s a storage device manufacturer, whereas Micron, as a chipmaker, has fallen more slowly—but it’s unclear where it might find support. Historically, when SOXL dropped, it fell by one-tenth. If that were to happen again, it would require another 30% decline from current levels. But previously, it rose 20-fold before correcting, whereas this time it’s risen 50-fold—making a simple proportional comparison difficult. AI is a battle for dominance—will the U.S. really back down? Even if U.S.-based data center construction halts, I suspect they’ll still proceed, possibly by seeking alternative spaces—whether in the ocean or space.

$VRT I'm adding some Verdiv today. Why? - Seems like capitulation to me. - It cools the AI data centers - Orders +100% YoY, $15B backlog. - Revenue +61% last quarter, guidance raised 3x. - 42% of liquid cooling patents + Nvidia partner. - Down 45% on political noise = rare 20x 2027 P/E.

🚀 $CIFR DAILY CHART SCREAMING OVERSOLD + MASSIVE CATALYSTS = HUGE LONG-TERM OPPORTUNITY! 🔥 The daily chart is looking absolutely brutal in the best possible way for patient bulls 📉➡️📈. $CIFR has been hammered down near 52-week lows (~$6.50 area after the latest flush), trading way below its major moving averages with RSI deeply oversold in the high-20s. Classic “throw the baby out with the bathwater” setup on a high-conviction AI infrastructure name. Recent catalysts are screaming execution:
✅ First HPC/AI data center lease revenue recognized at Black Pearl ✅ Barber Lake lease extended to 20 years — contracted revenue jumped to ~$9.1B on that site alone ✅ Total contracted revenue across campuses now ~$16.6B with projected ~$864M average annualized NOI ✅ 5.9+ GW development pipeline + conditional 1 GW base load at Colchis + lateral gas pipelines for up to 2.5 GW on-site power Cipher is successfully pivoting from Bitcoin mining into pure-play AI/HPC data centers with long-term, high-quality contracted cash flows. The short-term noise (revenue transition, dilution/debt concerns, Texas headlines) has created an extremely oversold entry on the daily chart while the long-term thesis is getting stronger, not weaker. Analysts still see massive upside (consensus targets in the mid-to-high $20s). This is the kind of deep-value, multi-bagger setup that rewards those who can zoom out. 💎🙌 Not financial advice — but the risk/reward on this oversold AI data center pure-play looks incredibly asymmetric for the long term. 🚀 #CIFR #CipherDigital #AIStocks #DataCenters #HPC #Oversold #AIInfrastructure #LongTermInvesting

OpenAI isn't slowing down its AI infrastructure spending despite growing public opposition to data centers. CFO Sarah Friar says the company will keep raising capital and investing to meet demand as it moves toward a potential 2027 IPO. $MSFT $NVDA $ORCL $AMZN

In August, the upward movement in stocks signaled a positive divergence. This was also accompanied by a correction timing that began on the AI side. If the unexpected move was intended for AI stocks to undergo downward consolidation, it may have been aimed at capturing the margin range. I’m not certain, but it’s possible.

Here’s what the Democratic congressman said—this has been widely condemned in the U.S.: “As long as we stop developing large models, China’s large models will also stop progressing.” Abdul El-Sayed, who has just been elected as the first Muslim U.S. Senator in American history, claimed that the United States should fully suspend cutting-edge AI research and need not worry too much about China catching up. Sayed asserted that China’s AI is merely a copycat of American AI, likening it to someone being towed by a speedboat. He argued that when America moves fast, China progresses quickly; and if America halts development of frontier large models, China will automatically stop progressing too.