The setup is bullish over the next year but expects several sharp sell-offs first: volatility, not whether AI or stocks are dead, is the real risk, and the biggest gains were framed as coming from buying the March and end-of-July lows rather than avoiding the tape. The core case leans on NASDAQ history, a still-intact breakout in QQQ/"Triple Q" around 746, easing inflation/rates after the September hike, and huge AI capex that keeps flowing through hyperscalers into chips, data centers, cybersecurity, and power. The near-term bear case is concentrated into dated event risk — CPI, the October 28 Fed meeting, November 3 midterms, mid/late-November NVDA earnings, and a December 9-11 cluster — but those were presented as opportunities if levels hold. Longer term, debt monetization, inflation, and post-midterm seasonality were framed as constructive for stocks, with October 28 singled out as the nearest flush date.
By this time next year, many viewers were framed as being "way, way richer," but only if they survive several sell-offs and avoid emotional mistakes.
NASDAQ doom content on AI was framed as repetitive: investors have supposedly heard for years that they won’t make money from AI even as NASDAQ kept hitting all-time highs.
The dot-com bubble was cited as proof that euphoria can last for years and make people rich before the final dump, so selling the whole move on a bubble call can miss almost all upside.
Greenspan’s December 1996 bubble warning came when NASDAQ was about 1,300; it topped above 5,000 in March 2000, so exiting then missed most of the run.
In the "Triple Q" pattern, a March Iran-war correction cut the fund 12% from the high, with the close at 558 on March 30, before a record close of 746 on June 2, up 34% in 2 months.
A second summer correction began with nearly a 5% one-day drop on June 5 on almost 100 million shares; the fund then ended about 11% off the June high, with the low printed July 29.
From the July 29 Fed-day low, QQQ jumped more than 3% on August 4 and ran to 732 by mid-August, then pulled back to a higher low around 705 in mid-September, versus 662 at the prior floor.
The July 29 low and the September 15-16 low both came on Fed decision days, and from the 705 low QQQ gained almost 3% on September 21 on nearly 47 million shares, closing Friday at 749.58, more than 6% higher in 12 trading days.
The July 29 low to the September 22 record close took just under 8 weeks, erasing an 11% correction in less than two months.
NASDAQ 100 was described as having come back from every correction, bear market, and total meltdown in its history, with the 20-day around 728 and the 50-day around 717, both rising and price above both.
Friday volume was about 34.5 million shares versus 57-67 million at the July lows, and the close was only about $3.5 above the prior June record, so the move was not yet a confirmed breakout.
746 was set as the key confirmation line: a close and hold above it would convert old resistance into support; using the 746-to-662 base implies a measured move toward about 830, roughly 11% above current levels.
Downside checkpoints were 717 at the 50-day, then 705 at the higher low; a close below 705 would break the setup and put 662 back in play, about 12% below current levels.
The September 16 Fed hike was the first since 2023, a quarter point to 3.75%-4% on a unanimous 12-0 vote because of inflation, while August CPI ran 3.4% y/y with more than a third of the monthly jump from gasoline tied to the Iran war.
Since 1988 there have been seven hike cycles; after the first hike the S&P fell about 4% over the next six weeks, then recovered over the following five or six weeks, and was up 9% on average 12 months later, except in 2022.
After the September jobs report, the economy added 29,000 jobs versus 90,000 expected, unemployment had already risen to 4.2% in July, August was revised down by another 60,000, and odds of no hike at the October meeting jumped to 77%; NASDAQ hit a new high that same day.
Washington and Tehran were tied to the November 3 midterm date circle, with the expectation that both sides want something announced before then; if conflict ends, oil falls, inflation falls, and the case for further hikes weakens.
Negotiations over the Strait of Hormuz and Iran were framed as eventually turning into a deal because money talks, which would open the Strait, cut energy prices sharply, and push Fed focus back to employment.
A Time interview last week had Trump saying, "Certain levels of inflation will also pay off that debt very rapidly," on a $40 trillion national debt; the claim was that the government will inflate its way out of debt, with debt service exceeding $1 trillion in a single fiscal year for the first time in August and debt now above the size of the economy for the first time since World War II.
The CBO was cited projecting debt at 101% of GDP today and 120% in 10 years; historical context given was 106% in 1946, 23% by 1974, and a 2.5% Treasury-yield cap from 1942-51 while inflation ran as high as 14% in 1947.
Massive inflation was framed as bearish short term if the Fed fights it, but bullish long term for stocks because real assets and companies get repriced in fewer dollars while the Fed eventually pivots back toward unemployment.
AI capex was framed as enormous: big spenders are on track for about $940 billion in 2026, with final numbers over $1 trillion, after selling $225 billion in bonds in the first half of 2026, about 10x the pace a year earlier; 2027 capex was estimated at $1.3 trillion and "way higher."
The "doomers" were countered with ownership data: the top 10% of households own about 87% of stocks, the top 1% own just over half, and the bottom half owns about 1%, so the funding base is sovereign wealth funds, mega-cap balance sheets, private credit, and very high net worth buyers rather than 1999 credit-card day traders.
Midterms on November 3 were framed as the worst year of the four-year presidential cycle for stocks, with the S&P averaging under 3% in the 12 months before midterms, a 17.5% average peak-to-trough midterm-year drop since 1950, and a 31.7% average gain one year after the midterm low.
The NASDAQ has already had two corrections of more than 10% in 2026, and October 28 was singled out as the nearest single-date flush point because the July 29 and September 15-16 Fed-day lows showed where bottoms have formed.
The explicit dump-date map for the next 12 months was October 14 for September CPI at 8:30 a.m., October 28 for the Fed decision, November 3 midterms, mid-to-late November for NVDA earnings, December 9-11 for a Fed decision, CPI, and government funding expiration, January 26-27 2027 for the next Fed meeting plus Q4 earnings, and March 16-17 2027 for another Fed meeting with a DOT plot.
NVDA mid-to-late November earnings were framed as the most important report in the entire AI trade, with the stock needing not just a beat but continued blowout guidance so the market does not think the run is already priced in; buybacks and prior earnings management were cited as reasons for a high chance of another beat.
December 9-11 was labeled an ugly three-trap-door window because it stacks a Fed decision, CPI, and government funding running out; January 26-27 2027 was framed as the next chance for hyperscalers’ Q4 earnings to reveal enormous 2027 capex and revive bubble headlines while still feeding preferred AI infrastructure names.
Three preparation checks were given for each dump-date: chart levels, dry powder at those levels, and whether leverage forces liquidation before cash can be deployed; leverage in this market was called asking for trouble, with most people advised to use none.
The buyers who made the most money in 2025 and 2026 were framed as those buying the March breathing-cycle lows and the end-of-July lows when it felt worst.
Three sectors were singled out to buy "rain or shine": cybersecurity, data centers/chips/networking/cooling, and energy tied to AI.
Cybersecurity was backed by Gartner global security spending of about $245-$250 billion this year, up about 13%, with cloud security growing close to 29%, making it a budget item less likely to be cut in a slowdown.
Data-center buildout was framed as funded for years because hyperscalers have borrowing capacity and cash-generating businesses, while the race with China and national-security concerns mean the government won’t let it stall for lack of funding.
Energy demand was backed with electricity shares: data centers used under 4% of U.S. electricity in 2023, McKenzie projects almost 12% by 2030, and a utility research group sees as high as 17%, implying winners in generation, transmission, and intermediate equipment.