Nvidia Falls 23% From Its High as Board Investigates Jensen Huang

Nvidia's board opened an independent investigation into founder and CEO Jensen Huang over the weekend, and the stock has not revisited its high since.

The issue, as first framed by macro investor Bryan Rich of Logic Fund Management, is Huang's relationship with an executive at one of Nvidia's customers — a woman U.S. intelligence believes is a Chinese intelligence asset. The story ran in late October. On October 28, the day before publication, Nvidia traded as high as $220 per share. It has not returned to that level. At the lows on the day of Rich's publication — November 9, 2026 — the stock sat 23% below its October 28 peak.

That single fact now hangs over the entire AI trade. Huang is not a interchangeable chief executive; he is the founder-leader of the company that supplies the compute underpinning the AI buildout. Any question mark attached to him is a question mark attached to the sector's central narrative.

How bad has the selloff been?

The numbers are stark, and they extend well beyond Nvidia.

• The deepest leg of the AI selloff began the very day the Huang story ran.
• That same day, Alphabet had its worst trading session since its 2004 initial public offering.
• The next day, Intel fell 29%.
• The day after the election, AI suppliers sold off again.
• Last Thursday, chip stocks fell 12% in a single session, the Nasdaq dropped 5.5%, and Nvidia fell 9%.

A 29% single-day decline in Intel and a 12% one-day drop across chip stocks are the kind of moves that typically mark a break in a market regime, not routine volatility. Whether that break becomes a top is the question Rich's November 9 analysis sets out to answer.

His framing, drawn from a longer piece originally published on November 9, 2026, is blunt in its two-part thesis: the Huang investigation adds genuine risk, but loose monetary policy suggests the AI boom is not over.

What does the Huang investigation actually change?

The board's investigation is independent, which matters. This is not a management-led review that can be quietly closed; an independent probe carries procedural weight and a timeline the company does not fully control.

The alleged facts are unusual for a corporate governance story. The counterpart in the relationship is an executive at a customer of Nvidia, not an employee, which complicates any conflict-of-interest analysis. The more serious element is the intelligence dimension: U.S. intelligence agencies reportedly assess the woman to be a Chinese intelligence asset. If that assessment is accurate and the relationship involved the disclosure of sensitive information — commercial or otherwise — the exposure would run beyond a personal-conduct matter into national-security territory.

What is not yet known: the duration of the relationship, the identity of the customer involved, whether any Nvidia information changed hands, and how the board's independent investigators were selected. The gap between what is alleged and what is established is wide, and markets have repriced Nvidia 23% off its high on headline risk alone, before any findings exist.

Does Fed policy say the boom is over?

Rich's answer is no, and his reasoning rests on a comparison with the last great technology bubble.

In 2000, at the dot-com peak, Federal Reserve policy was genuinely restrictive — the central bank had ratcheted rates high enough to choke the cycle. Today's setup is different. The Fed has delivered only one hike in this cycle, and real rates (interest rates adjusted for inflation) sit near the inflation rate itself. By Rich's read, that makes current policy less restrictive than at the dot-com top.

The distinction matters because liquidity conditions, in Rich's analytical framework, are one of the three forces that determine where a technology cycle stands. He tracks the policy path — rates, liquidity, industrial strategy — alongside elite capital flows and the technology cycle in compute, data, and energy. His daily briefing, Pro Perspectives, reaches more than 25,000 investors, and his positioning approach runs through two subscription strategies at Logic Fund Management: the Billionaires Portfolio for event-driven value and the AI-Innovation Portfolio for infrastructure plays across data centers, networking, software, and robotics.

A caveat belongs here, and Rich's own framework supplies it. Loose policy is a necessary condition for a cycle to continue, not a guarantee that it will. The 2000 comparison cuts both ways: the dot-com boom also ended in a 78% Nasdaq drawdown once it did end, and rates were only part of that story. Policy that is less restrictive today tells investors about today, not about how long the tail risk stays contained.

What does the power market say about AI demand?

The strongest counter-signal to the top-of-the-boom thesis comes from electricity, not equities.

Constellation Energy has secured more than 8,500 megawatts in long-term power deals, including agreements with Google and Amazon. The company also raised its earnings outlook. In plain terms, hyperscalers are signing multi-year contracts for physical electricity at a scale that only makes sense if they expect to be running AI data centers for years to come.

Power purchase agreements are not sentiment trades. They are capital commitments with delivery schedules, and they are difficult to unwind without penalty. When Google and Amazon lock in thousands of megawatts, they are revealing their internal demand forecasts more honestly than any earnings-call commentary does.

Rich draws the connection explicitly: persistent, contracted demand for AI infrastructure is the real-economy check on whether the equity selloff reflects deteriorating fundamentals or a repricing of risk premiums. The Huang headline hits the supply side of the narrative; the 8,500 megawatts hit the demand side, and they point in opposite directions.

Is this the top of the AI boom?

Rich's published answer is that the boom is probably not over, but the risk profile has changed.

The bull case in his telling rests on two legs: monetary policy that remains loose by historical bubble standards, and contracted physical demand for AI compute that continues to show up in the utility sector. Constellation's raised earnings outlook is corporate guidance, not speculation, and it arrived during the very weeks when chip stocks were falling double digits.

The bear case is equally concrete. Nvidia, the largest and most systemically important company in the trade, is down 23% from its October 28 high of $220. Its founder-CEO is under an independent board investigation with a national-security dimension. Alphabet just logged its worst day since going public in 2004, Intel fell 29% in a single session, and chip stocks dropped 12% last Thursday alone. Selloffs of that breadth do not require new information to keep going; they generate their own.

What the source does not settle is the timeline. An independent investigation can run for months and can end in exoneration as easily as in sanction — but while it runs, every interim headline about Huang, the customer executive, or the intelligence assessment is a fresh volatility event for Nvidia and, by extension, for the AI trade.

The takeaway Rich leaves investors with is a two-sided one: policy and power demand argue the cycle has further to run, while the Huang investigation demonstrates how quickly concentrated leadership risk can unwind a quarter's worth of gains — and the next catalyst in this story is likely to come from the investigators, not the market.