AI and semiconductor names led a sharp V-shaped rebound after a “crazy AI crash,” with CRWV’s 135 billion backlog, MU’s 12.29% pop, and broad strength across MRVL, ALAB, AMD, AVGO, TSM, TSLA, and NVDA. The setup is framed as mean reversion across equities, rates, oil, VIX, and crypto: the 10-year moved from nearly 5.4 to 4.8, VIX faded from above 28-29 to 17.48, WTI bounced off an $85-$86 area, and BTC/SOL/ETH were all described as rebounding from key moving-average and level breaks. Crypto remains the main weak spot relative to AI: ETH weakness is tied to ETH ETF selling and SOL is still below its 200-day around $89-$90, while MSTR/MCI, STRC, and BTC are all treated as dependent on a BTC rebound. The video ends with a broad buy-the-dip framework: use leverage sparingly, lean into mean-reversion buy signals, and favor NVDA, MU, MRVL, PLTR, TSLA, SpaceX, and selected metals when they hit the named levels.
AI stocks saw the biggest dollar-value crash since COVID March 2020, then a fast V-shaped recovery; CRWV lit the fuse with better-than-expected earnings and 135 billion in backlog.
Semis ripped on the rebound: MU +12.29%, MRVL +11%, ALAB +9%, SKH +11%, AMD +7%, AVGO +6%, TSM +4%, TSLA +4%, NVDA +4%; PLTR and SOFI were each +2%, MSFT +2%, ETH +1.4%, BTC +0.61%, while SpaceX was flat to down 0.3%.
Week-over-week calls were marked right on AMD, AVGO, MRVL, ALAB, SpaceX, MU, PLTR, Google, and Lumentum; SKINX was flat week over week but down during the period before rebounding fast.
BTC and SOL were the misses: BTC failed to reach level three around 63K, SOL pierced level five at 77K and broke the 200-day without bouncing there; BTC later turned up from 73K-74K on the 4-hour chart.
NVDA’s prior “under 200” call was wrong after a spy scandal hit the stock and it fell out of an upward channel, dropping about $52, or 23.98%; after-hours at 174 was expected to get back to 200 quickly because CUDA still works and Vera Rubin chips are sold out for years.
The 10-year yield moved from nearly 5.4 down to 4.8, then rebounded fast after Fed comments about possibly increasing rates twice before year-end; the VIX spiked above 28-29, then fell to 17.48, with 14.46 support and ~18 as the 200-day.
WTI crude printed a very clear double top at 106, then slid back toward $86-$85; the view is that $85 should rebound unless peace in Iran changes the setup.
BTC was described as not reaching level three around 63K but piercing level five at 77K; the 4-hour chart showed a turn from 73K-74K, and BTC vs gold was hammered with BTC down $5,000-$6,000 while gold held up better.
SOL was called ready to turn back up from 83 toward its 200-day near $89-$90, with 200-300 cited as the upside if crypto regains legs; SOL has outperformed ETH by about 7% since August.
ETH weakness was tied to huge ETH ETF sell-offs and Tom Lee losing $6 billion on his ETH trade; ETH was around $2,000, with a rejection of $25,000 calls and an explicit buy setup under the 200-day toward 2730, a 35% move.
MCI/MSTR risk hinges on a report due at 11:00 p.m. Eastern tonight: some sell pressure is possible, but not lethal because the company holds over 850,000 BTC; if it avoids MSCI removal, it could reclaim the 200-day and move from 122 back above 132, but only if BTC rebounds.
STRC was dismissed despite being back at 99.99 every week: the flywheel is cooked, only one buyer is left, and the preferred trade is BTC instead.
SpaceX dipped to 120 on Friday, then mean-reverted to 147; the move was framed as a 2.31-standard-deviation setup with a 70%-75% backtest win rate, and the AI-data-center-in-space thesis was backed by 12,000 satellites already up and solar powered.
TSLA was framed as a no-brainer buy at 350 or below after a crash low near 335 and a move back to 391; the self-driving anecdote was 225 miles with zero interventions, and the Cybertruck hit 11,000 self-driving miles with 99% of miles in driving mode.
Astera broke through level 3, bounced off level 2 at 189, traded around 233, and is expected to go to 300 and then $500; MRVL got a huge buy signal off level 2 at 129 and is tied to capex spend, Momentum, and photonic connectivity with Lumentum.
AMD got a mean-reversion buy signal at 1.52, then ran from 403 to 562 in a $153 move in a couple of days, later straight up through 510, 536, and 570; the mean-reversion sequence was described as spike, sell signal, then trend turn a day or so later.
AVGO had a 330 kill zone that worked for three weeks, then crashed to 290 and recovered to 345; the earlier drop was tied to midterms, capex spend, and the AI selloff, with the Friday rebound described as running into the burning building.
MU is still expected to go back to $1,000, then $1,200, then $1,500-$1,600 within a year or less; Optimus needs more memory, but Micron supply constraints keep doubling that memory out of reach right now.
PLTR broke through level five, shot to 152 during the AI crash, and is expected back toward 200 after a huge mean reversion; the performance math was described as minus two over 2.5, the worst since March 2026.
Gold had no weekly buy signal but showed a daily 90.24% win-rate buy setup around 3920s-3960s, with October 2023 as the lower mean-reversion reference; silver had a clear sell at 121 and no buy signal at 50, the cheapest since October 2025.
Copper on COMEX sold off with gold, but 6.19 was called a buy with a 6.90 target because it was deeply oversold, two standard deviations below where it should be and outside the channel.
Risk management stayed explicit: use margin sparingly, keep no more than 20% of margin in play, and hold 80% as a safety mattress because it takes a lifetime to make and a second to lose.