“To hell with Jensen Huang and his spyin’ mistress. Who needs ’em, anyway?”
So shouted semi stocks, including Nvidia, on Wednesday, when the SOX was on track for a second barnburner session in three days. (Or a third in four if you count November 6’s 4.2% advance as a “barnburner.”)
If Wednesday’s intraday move held, the chip gauge would boast a near 20% rally from the November 5 close. The rolling four-session gain was on pace to be the largest since the erratic trade around the original COVID panic.
There’s the chart. Before 2020, you have to go all the way back to 2002 to find a comparable four-session rally.
As I mentioned in a comment Wednesday, the grab for downside hedges into last week’s semi panic set up a combustible situation: Bought puts could be dry kindling in the event of a rebound.
If the panic subsided and dip-buyers emerged, causing semis to run away higher (i.e., away from downside strikes), dealer buying to unwind short hedges against those puts would amplify the upside move in the underlying.
Meanwhile, the leveraged ETF rebalancing flows do matter, particularly given that the dynamic’s by now so well-socialized that market participants are inclined to front-run it.
The figure above gives you a sense of how pronounced the estimated EOD rebal “buy impulse” from levered semi ETFs was on Monday — and how pronounced it was poised to be on Wednesday.
“There’s $26.5 billion of leveraged ETF rebal EOD,” Nomura’s Charlie McElligott remarked, just after noon ET. “This is ‘the reversal of the reversal’ as the rally sees tech lead again.”
That $26.5 billion figure represents a snapshot in time across a universe of levered products which includes the “juiced” semi ETFs. The figure below’s a snapshot I took using my universe and my model around the same time.
So, a similar picture: About $23.5 billion delta to trade. That’s a lot, and as you can see, it was concentrated in chip stocks and tech from a “theme” perspective. If you break it down by underlying constituents, allocating broad-index ETF flows across their member stocks, the Mag7 buy impulse came to about $3.8 billion. That captures the Mag7 exposure embedded in products tracking the Nasdaq 100 and S&P 500, which my “theme” view keeps in their broader index buckets.
Anyway, Wednesday’s “We are so back!” trade in semis was helped along meaningfully by CoreWeave, which did a lot to allay pervasive concerns about — and I’ll be polite here — the company’s operating outlook.
Not only did the most widely-discussed neocloud manage a beat and raise in Q3 results released late Tuesday, CEO Michael Intrator told CNBC Wednesday that he’s “sidelining” his 10b5 stock sale plan.
Recall that The Information (and they weren’t alone) sneered at Intrator last month, when he described concerns about the viability of the company’s debt-fueled expansion as “irrational.”
Intrator, Martin Peers wrote on October 14, “could win some brownie points by pledging to stop selling his stock for a while.” That’s just what Intrator did Wednesday.
CoreWeave was on track to notch one of its best five days as a public company.
The final two charts in this article are indicative of what’ll be in the “Heisenberg Report Analytics” .pdf. More info here — > https://heisenbergreport.com/2026/11/10/lets-get-mechanical-about-this/
Just for some added color: using SMH as a proxy, semis are up about 14% at today’s close from their nadir on the 5th. The SOXL (leveraged 3x) is up about 65% over that same period. Both charts are still in fairly steep bear patterns — lower highs and lower lows — and both still sit below their 200-day moving averages, (especially the SOXL).
Yeah, I mean, the article title’s deliberately sardonic. This isn’t a piece aimed at making a new fundamental bull case for semis although honestly, I’m not bearish.
I (only) suspected the title was an exaggeration, but I was still impressed by how much the SOXL actually moved. Leveraged ETFs are not my cup of tea, but I thought the chart comparison was an interesting one.
Read this comment, then also read the two after it –> https://heisenbergreport.com/2026/11/05/um-excuse-me-but-your-chips-are-bleeding/comment-page-1/#comment-94098
Long story short, the math on those leveraged products doesn’t work like most people think it does. “dayjob” did a good job of pointing that out. And my reply to him is how I suggest most people should think about leveraged stock positions.
Thanks. Other than using a “cheater bar” to loosen the lug nuts on my car, I have never employed leverage of any kind (nor do I ever intend to.) I’ll try to stay in my lane a little more next time.
I’m still bullish on AI, but wouldn’t it be something if the AI trade unraveled because Huang couldn’t keep it in his pants? This guy’s affairs could swing trillions of dollars worth of wealth.
And what does this mean for the vol of the vol of the vol?
I’m surprised H didn’t brag about his buy call. Actually I’m not because H knows it can turn on a dime tomorrow. I gambled with you on this and plan to hold until to you publish the I told you so piece three years later.
https://heisenbergreport.com/2026/11/05/um-excuse-me-but-your-chips-are-bleeding/comment-page-1/#comment-94098
It is hard for me to imagine an outcome to this situation that doesn’t still have Jensen Huang as the CEO.
There is a long and fascinating historical list of corporate espionage. Evidently, the oldest documented case of corporate espionage is attributed to Father Francois Xavier d’Entrecolles, who in 1712 revealed the manufacturing methods of Chinese porcelain to Europe.
Wave Dash: He did (rightfully) brag about it, but in the comments of the “your chips are bleeding article.” And also posted a funny gif lol