AI spend is being treated as a capital problem, not a chip-supply problem: AMD and AVGO say money is the constraint, META got hit on $2.1 billion negative free cash flow and huge capex, and the “neo cloud” trade is called oversold after a brutal selloff. NVDA, MU, and Hynix are framed as buyable on fresh signals, while BTC, SOL, and TSLA are the cleaner long ideas; ETH and ETH ETFs are the laggards, and MSTR’s stock/STRC financing gets mocked as a bad setup. The market backdrop is still hostile — Brent at 111.63, U.S. bonds just under/at 5%, VIX at 22.37, S&P 500 at highs with a fresh sell signal — but the video’s core claim is that the best entries are appearing in the most punished AI and crypto names.
Scorecard showed 9 hits/partials versus 6 misses; ETH was only a 1% miss, while AMD, SK Hynix, Broadcom, and Micron were said to have been hammered, and Solana was called correct even though the buttoning showed it as a miss.
U.S. government bonds were said to have fallen to just under 5%, then to 5%, framed as good news; Brent at 111.63 was framed as inflationary and rate-bearish because it hammers everything.
S&P 500 was at all-time highs while macro was called a mess; a new all-time-high sell signal was said to have fired one and two weeks ago, with QQQ also showing a sell signal at the top and an ugly downtrend tied to AI problems.
The macro AI recap was inflation and rate fears, AI debt fears, a vicious neo-cloud selloff, a massive memory slump, and capital constraints; AMD and AVGO were used as examples of management saying money, not chips or supply, is the constraint.
META was sold off on $2.1 billion negative free cash flow and huge capex guidance; LRCX was down 9% on supply constraints; TSLA was rewarded after positive free cash flow and a $5 billion capex cut.
VIX spiked to 22.37, called a good place to short, but it remained above its 200-day moving average around 18.
BTC was chopping near $82,000 support and around 400 bucks, with ETF flows described as very positive for the last three days; the chart was said to be boxed in after a daily candle from 80K to 87K, with resistance at 87K and sideways action for about a month.
BTC was said to be tracking gold for months; the BTC/gold ratio was over 20 ounces of gold per Bitcoin, off a bottom around 12, and was called evidence that BTC is smoking gold.
SOL had good support around 104 last week, which worked; price was 105.62, and SOL was said to be outperforming ETH over the last two months as predicted.
ETH was described as unloved: ETH ETFs were selling off while BTC ETFs and SOL ETFs were positive, leading to the question whether TradFi thinks ETH is too slow to survive the future.
MSTR had nerves around MSCI getting cut; the stock fell from the 160s to 133 and then bounced, while Michael Saylor was blasted for ATMing stock to hoard cash to buy back STRC, with STRC at 99.11.
The NeoCloud trade was called completely oversold after a top at 15.30 and a near-50% drop; a small green shadow was treated as nearly a buy signal.
CRCL was given a buy signal down at $8 as a good risk/reward, while a Schwab options read showed heavy selling of $10 calls, buying of out-of-the-money strikes, and 1,300 contracts at the $10 strike; a $7 strike a year out was mentioned for long-term gains.
TSLA had a key technical line at 380, which had been resistance for 2-3 months and had to hold for a move toward 400; the stock also got a covered-call setup example from 395 to 364 back to 389 in hours.
TSLA fundamentals were framed as strong: capex cut by $5 billion, positive free cash flow called nuts, cybercabs doing 90,000 miles a day at about $2.00-$2.50 a mile, a 2,000-cybercab year-end target, version 15 coming, and humanoid robots in production at 20,000 a month with version 3 complete.
TSLA megapacks were said to be sold out for 2027, implying $11 billion of profit and a potential $1.1 trillion valuation on a 100x multiple; SpaceX was tied to a $2 trillion-to-$6 trillion path, with data centers in space still two or three years away.
TSLA/SpaceX ratio trading used 2.54 as the current level versus a 2.4 threshold, with a target of around three SpaceX shares per Tesla share if a merger happened; the ratio was said to imply 2.54 to 2.55 SpaceX shares now.
PLTR was at 180.998 with a comment about heading back above 200; ALAB had a beautiful sell signal at 396 and a request for level four.
NVDA had a clear buy signal at 212, now around 213, after a clear sell signal in the 240s on October 6; the stock was called the granddaddy of the space.
SK Hynix was said to have a first buy signal in quite a long time at a floor of 1.2 million one, with memory trading at a PE of 7.
MRVL had a sell signal at 230, a floor around 2223, and no buy signal yet, though a buy signal was expected in 6 to 8 hours with a possible rebound toward 300-330.
AMD had its all-time high at 658.66 and was down to 514 in a nasty downdraft; the constraint was said to be money, not materials, chips, or supply, and the trading model was said to have caught the turn 100% of the time over the last 119 trades, producing 200,000% of profit.
GOOGL/GOOG was hit by capex fears and described as not juicy; a level-five bounce scenario around 310-320-315 was mentioned, with PE 16, limited revenue growth, and 260% earnings growth tied to SpaceX capital gains.
AVGO had floors at 338-333 before breaking lower; 322 was called the floor for both AVGO and GOOGL, with 322 also described as a good level to add more AVGO and 308 seen as unlikely.
MU was said to be overblown on China fears because China does not make HBM, only DRAM; the claim was that there is no Micron memory for the next 1.5-3 years, yet MU was still selling off, with a desire to see it back in the 700s.
HOOD had a 5-year path from 34 to 155, then 60, a turn at 70, back to 125, and now below 100; the chart showed a sell signal, PE 43, flat revenue since Q3 2025, down income, and a possible tiny flutter at 95 with risk to 87 or lower.
UMAC was called the kill zone after a 24-standard-deviation mean reversion move to 8.24; the floor was around 16, with confirmation at 17, and the same level was said to have been hit exactly in July 2026 at 15.95 and again now at 16.