September setup: a Labor Day closure, a historically weak first half of the month, and a 15th-18th volatility window are framed as the next near-term risk, while AI and AGI are treated as the structural trade that overwhelms rate noise. The core portfolio message is to keep dry powder, use limit orders into a 4.7% or larger dip, and concentrate in the tiny set of names that can compound as AI monetization accelerates. Tesla is the central long: a 2026 AGI callout, Cybercab/Optimus math, and a 2027-2028 Tesla-SpaceX merger thesis are used to argue for extreme upside, while Nvidia is treated as the benchmark toll-booth winner with far richer economics than peers. The second segment shifts to execution: ignore 24-hour stock-chart signals for now, rely on regular session volume, and treat after-hours as a trap until markets, algos, and agents are actually running books 24/7.
Labor Day weekend means U.S. markets are closed Monday, and the video is still scheduled to publish that day.
Since 1928, September is described as the worst month on average, down 1.1%, with the first half of the month usually doing the damage and midterm years even worse.
A volatility window is flagged for Sept. 15-18, roughly 9-12 days away in the notes, with a forecast for a 25 bp FOMC hike at 66%-70% odds and Japan also expected to raise rates on Sept. 17-18.
A 25 bp hike is framed as hurting risk assets but not really AI assets, because AI names are said to be printing money and largely immune to rates unless they are heavily levered like ORCL.
The plan is to keep dry powder, layer bids, and use limit orders into a 4.7% selloff or worse; a 5% index drop is the hoped-for setup to buy high-beta names down 10% or more.
Q4 is described as strong, with October possibly weak in the first week depending on macro, then November and December as moon months.
AGI is defined as an AI that can learn, reason, and do any intellectual task a human can do, and GPT6 Astra is framed as having hit AGI this week, matching Elon Musk’s 2026 AGI forecast.
TSLA is modeled at 3.95 billion shares; at 8,000 a share the market cap is calculated at 31.6 trillion, with market cap treated as future profit value rather than global GDP.
NVDA is said to have made about 198 billion in profit over the last 12 months, implying roughly 0.22 trillion in earnings; at a 5.6 trillion market cap that works out to about 28x earnings, down from about 70x a couple years ago.
Amazon is cited as the long-run analogue: earnings growth eventually grew into valuation over 22-24 years and produced huge returns.
The portfolio rule is a concentrated core: hold an 80% core bag and trade around it with 20%, with only about 0.3% of assets expected to matter enough to win.
Buffett’s 50% drawdown rule is cited: if a 50% decline is intolerable, the asset is not for you; if money is needed in 6-12 months, volatile assets are not fine.
TSLA’s robotaxi and Optimus model is built on $40 trillion of opportunity, split as $30 trillion Optimus and $10 trillion CyberCab/robotaxi, with TSLA assumed to capture 25% of that and generate $10 trillion of revenue at 25% net margin.
A separate digital-AI claim puts a $10 trillion opportunity around SpaceX, says digital AI could create 20%-30% of world economy, and says humanoid robots could grow the economy 10x from a $126.3 trillion GDP baseline to $1,260 trillion.
TSLA Cybercabs are said to be headed to Japan, where autonomous cabs are wanted in four major cities, and a Tesla-SpaceX merger is planned for 2027-2028 with the combined asset correlation going to 1 and a big acquisition premium on TSLA.
Cybercab ownership is framed as possible for $30,000 or less before 2027, reportedly confirmed by Elon Musk, with rollout constrained by local regulation; Texas, Florida, and Nevada are expected to embrace it while New Jersey and New York may try to block it to protect medallions and taxis.
The robo-taxi age rules are explicit: Waymo alone at 18+, 16+ in Arizona with parental authority, under 18 banned in San Francisco; TSLA Model Y robo taxi forbids kids under 8 and requires an adult for ages 8-17, while Cybercab forbids under 13 and requires an adult for ages 13-17.
The toll-booth basket is ETON, H, 6755, and 6727-style physical bottlenecks: ETN for medium-voltage switchgear, Hitachi Energy for large transformers, Ajinomoto for co-wash packaging film, and Nitto for lowCT glass cloth, all described as capacity-capped and already priced in through backlog and lead times.
Margin math favors NVDA over toll-booth names: NVDA is said to keep nearly 80% of GPU revenue as profit, while ETN is given 9.6% net income, Ajinomoto 8%-12%, Hitachi 12%, and Nitto 18%, making ETN the most interesting of that group by margin.
The execution warning on charts is sharp: 24-hour stock sessions create thin-volume, after-hours noise and a '24-hour trap,' so buy/sell signals should be trained on regular market hours until algos and agents are actually running books 24/7.
Buffett-style P/E and capex ratios are called less useful in the AI era because the businesses are growing too fast and making too much money for old valuation tools; for the S&P 500, the old ratios still matter more than for fast-moving AI winners.
The work-on-the-winners rule is to 'excise the 0.3% of assets and hold them,' let compounding do the work, and ignore legacy technical-analysis chatter that assumes the world is stable.