On Wednesday, the 22nd of October, Jensen Huang went on TV and told us that the neoclouds were a good investment.
This was the third such appearance he made in a week. On October 15, he said that Nvidia cards weren’t in need of a backstop. Then, on October 21, he said that a GPU easily lasts 8 years, if not more. Of course, on the 22nd, he said that the companies that borrow to buy his GPUs and rent them out were a good investment.
If this sounds like the barber telling you that you need a haircut, it’s not far off.
A neocloud is a company that borrows money to buy Nvidia chips and puts them in a data center and rents them by the hour to the AI labs.
On October 15th, CoreWeave (a neocloud)’s bonds yielded 15.6%, up significantly from the 9% they yielded when they were sold in April this year.
Since then, CoreWeave is down 22%, and Nvidia itself is down 4%. CoreWeave is also down 68% from its 2025 high.
On one hand, you have the CEO of the most valuable company in the world saying that his customers are a good investment.
On the other hand, you have the people who lent those customers money saying that they want 15% a year to continue doing it.
One of these two is wrong. If history teaches us anything, it is that usually the bond market is right and tech CEOs are wrong, because the bond market needs to be paid back.
And that spells trouble for Nvidia, not because it owns $2 billion worth of CoreWeave relative to Nvidia’s $5 trillion valuation (that is quite irrelevant), but because Nvidia has moved towards a position where it finances its customers.
You’ve heard this part of the argument a million times.
What you haven’t had a good answer to is:
When does it become a problem for NVDA?
What should we expect the stock price to do?
And what does this mean for the broader stock market?
This report lays out the exact mechanisms which matter for NVDA and which inform the short position which we have initiated in the stock.
NOTE: This report is the most important report you will read in November. Countless hours have gone into preparing it, on the back of months of observation and analysis.
It is yours, absolutely free of charge, as long as you hand over your email.
I consider this report to be so important for the public benefit that I will not charge for it.
I will ask, however, for your email. This helps me more than you imagine, and I believe it’s a fair ask.
You give an email, I give you hours and hours of professional research for free.