Two days ago I did the slow, careful math on how long it would take Jensen Huang to sell his 870 million Nvidia shares. This one is the fast version. Forget the polite founder who trims a little every quarter. Say he wakes up Monday, November 16, and decides he wants out of all of it. How fast can he actually do it, what stops him, and what does it do to a stock that already lost almost $1.9 trillion in market value in six weeks?
Here’s my answer up front: he can’t sell a share on Monday, he could legally be out by early December, and if he really goes for it, I think Nvidia loses another trillion dollars in market cap within two weeks. That’s a prediction, not a fact, and I’ll show you exactly where the math ends and my opinion starts.
How many Nvidia shares does Jensen Huang have right now?
Roughly 870 million shares. Nvidia’s 2026 proxy statement listed his beneficial ownership at about 870.6 million shares (3.58% of the company) as of March 23, 2026. Since then his Form 4 filings show small changes, not a sell-off: about 45,700 shares withheld for taxes in June, the same again in September, and a gift of 438,000 shares on September 17. I couldn’t find a single Form 4 or Form 144 filing from him in November.
Not all of that pile is his to cash out like a bank account. Some of it sits in trusts and partnerships, and some belongs to his family foundation. But for a worst-case question, let’s use the full 870 million. At Friday’s close of $162.85 that’s about $141.8 billion. If he dumped it and the price slid on the way out, he’d realistically walk away with something like $130 billion before taxes. For the backstory on the man himself, here’s a plain profile of the engineer who started Nvidia in a Denny’s booth.
Why he can’t start selling on Monday
Three things are in the way, and none of them is Nvidia’s permission. Nvidia can’t sell his shares and can’t stop him from selling them. The law and the calendar can.
- Earnings are Tuesday, November 17. Huang ran the company until Thursday. He almost certainly knows how the quarter that ended in late October went. Selling before those numbers are public is the textbook definition of insider trading. Company trading windows usually stay shut until results are out, and even after that they normally open only after a full trading day or so.
- He’s still an affiliate. An executive who just left is usually treated as an affiliate for at least 90 days. That means every sale goes through Rule 144: he files a Form 144 notice with the SEC when he places the order, it shows up on EDGAR, and the whole world sees it.
- A new trading plan can’t sell right away. His 2025 10b5-1 plan was for 6 million shares. A brand-new plan comes with a mandatory cooling-off period (30 days for most people, 90 days or more for officers and directors), so a plan signed Monday doesn’t sell Monday.
So the realistic first day he could sell is Wednesday, November 18 or Thursday, November 19, after the earnings release, through a broker, out in the open.
The legal speed limit is bigger than his whole stake
Here’s the part that surprised me. Rule 144 caps an affiliate’s sales in any three-month window at the greater of two numbers: 1% of the shares outstanding (about 243 million NVDA shares) or the average weekly trading volume over the four calendar weeks before he files.
Those four weeks were insane. NVDA traded about 471 million shares the week of October 19, 515 million the week after, 847 million the week of November 2, and a monster 2.34 billion shares the week he resigned. That averages out to roughly 1.04 billion shares a week. His entire 870 million stake fits under that cap. If volume stays this crazy, the law lets him sell all of it in one three-month window. The only real speed limit left is how much pain the market can take.
How fast he could get out, starting this week
Over the last 20 trading days NVDA averaged about 209 million shares a day. Here’s how long 870 million shares takes at different paces, counting Thanksgiving and the holidays:
| Pace | Shares per day | Trading days | Done by (starting Nov 19) |
|---|---|---|---|
| 10% of daily volume (the careful way) | ~21 million | 42 | ~January 21, 2027 |
| 25% of daily volume (aggressive) | ~52 million | 17 | ~December 14, 2026 |
| 42% of daily volume (“two weeks”) | ~88 million | 10 | ~December 3, 2026 |
| Block trades / a bought deal | Hundreds of millions at once | 2 to 5 | Late November, at a discount to big funds |
If he’d been free to start on Monday the 16th, the two-week version would wrap up on Monday, November 30. That’s the scenario everybody on my feed is really asking about. It’s legally possible after earnings, and it would be the most violent founder exit Wall Street has ever seen.
Why $130 billion of selling can erase a trillion
This is where most people’s mental model breaks. Market cap isn’t money sitting in a vault. It’s just the last trade price multiplied by every share in existence. Nvidia has about 24.3 billion shares, so every $1 move in the price moves the market cap by about $24 billion, even if only a small slice of shares actually traded hands at that price.
Look at Anthropic for a clean example. According to its IPO filing, it had raised $141.1 billion in total and held about $132 billion in cash and short-term investments as of September 30. The IPO talk puts its value at around $1.3 to $1.4 trillion. So roughly $130 billion of real money supports a price tag about ten times bigger. That isn’t fraud. That’s just how stock prices work: the price on the screen is set by the few shares changing hands, not by cash in the bank. If you’re following that one, here’s the ANT listing date and price range everyone is watching this week.
It works in reverse too. Economists Xavier Gabaix and Ralph Koijen estimated that for the stock market as a whole, every $1 that flows in or out moves total market value by about $5. One stock usually reacts less, because other funds step in to buy the dip. A standard trading-cost model (the “square-root law”) says that with Nvidia’s current volatility, 870 million shares would push the price down about 3% to 7% from the selling alone. That’s $135 to $270 billion of market cap, or one to two times the size of his sale.
Getting from there to a trillion takes a 25% drop, from $162.85 to roughly $122. The selling itself won’t do that. Panic would. Picture a Form 144 for hundreds of millions of shares landing on EDGAR the morning after a shaky earnings call. Funds that were holding on because “Jensen still owns 3.6%” would see the founder heading for the exits, and they wouldn’t wait around to be the last ones out. That’s how a $130 billion sale turns into a trillion-dollar move.
My prediction
If Huang files a Form 144 after earnings and starts selling in size, I think Nvidia is worth about $3 trillion or less by the end of November, down from $3.96 trillion today. That’s another trillion gone in about two weeks, on top of the $1.85 trillion it already lost since the October 6 peak close.
To be fair to the other side: there’s still no filing showing he’s selling, he’d owe tens of billions in tax, and he’d be wrecking the value of every share he hasn’t sold yet. If Tuesday’s numbers are strong and no Form 144 shows up, this whole scenario stays on paper. Watch EDGAR, not Twitter. For the TV version of the bull case, here’s what Jim Cramer has been telling viewers about Nvidia lately, and for how the server makers are holding up, Super Micro’s latest quarter is a good read on Nvidia-dependent hardware.
Whatever happens to the stock, the chips keep shipping and every one of them needs power and cooling. If you’re actually building, price out a real GPU cluster, chips, power and cooling included, before Tuesday’s numbers land.
FAQ
How many Nvidia shares does Jensen Huang own?
About 870 million. Nvidia’s 2026 proxy listed roughly 870.6 million shares (3.58%) as of March 23, 2026. Since then his filings show small tax withholdings and a 438,000-share gift on September 17, and no November sales. At the November 13, 2026 close of $162.85 the stake was worth about $141.8 billion.
Can Jensen Huang sell his Nvidia shares on Monday, November 16?
Realistically, no. Nvidia reports earnings on November 17 and he very likely knows the results, so selling before they are public would risk insider-trading charges. The earliest realistic sale date is November 18 or 19, through a broker, with a public Form 144 notice.
How fast could Jensen Huang sell all his Nvidia shares?
At about 42% of Nvidia’s recent average volume of 209 million shares a day, he could sell everything in 10 trading days. At 25% it takes about 17 trading days, and at 10% about 42. Rule 144 caps affiliate sales at the greater of 1% of shares outstanding or average weekly volume, which is currently about 1.04 billion shares, more than his whole stake.
Could Jensen Huang selling wipe $1 trillion off Nvidia’s market cap?
The selling alone would likely move the price about 3% to 7%, or $135 to $270 billion of market cap. A $1 trillion drop needs a fall of about 25%, to around $122, which would take panic selling by other investors on top of his sales. That’s the author’s prediction if he sells in size, not a certainty.
Sources: Nvidia 2026 proxy statement (via secondary reports); SEC Form 4 filings for Jensen Huang (June 17 and September 16 to 17, 2026); CNBC and ABC News (November 12 to 13, 2026); Anthropic IPO filing as reported November 10 to 11, 2026; SEC Rule 144 and Rule 10b5-1; Gabaix and Koijen, “In Search of the Origins of Financial Fluctuations: The Inelastic Markets Hypothesis”; NVDA daily price and volume through the November 13, 2026 close. Timelines, price-impact ranges and the trillion-dollar call are my own estimates and opinion, not financial, legal or tax advice.
Chart © 2026 Kevin Katovic, all rights reserved. To license it, email support@kevinkatovic.com.