Odd Lots · · 43 min
How Wall Street Is Bracing For the Post-Jensen Huang Nvidia Era
With Stacy Rasgon
Summary
Stacy Rasgon of Bernstein walks through how Wall Street is preparing for Nvidia without Jensen Huang, arguing that Huang's flat, personally driven management, handshake supplier relationships and long-horizon bets like CUDA were central to the company's dominance. He downgraded Nvidia and roughly halved his price target, citing key-man risk, a cheap-but-headline-prone stock and expected investigations, yet he sees near-term data center demand as intact, with 2027 guidance implying about 70% growth. The discussion also covers competition from Broadcom, AMD and Qualcomm, the risk that AI-assisted code porting erodes CUDA over time, the roadmap decisions facing a successor, and the size of Nvidia's supply and financing commitments. He further argues that China export policy should allow older chips while keeping Blackwell and Rubin restricted, and that China revenue exposure is smaller than some estimates suggest.
Key takeaways
- Stacy Rasgon (bullish on NVIDIA): Nvidia's near-term data center business is not at much risk from the CEO exit; customers are not cutting orders.
- Stacy Rasgon (bearish on NVIDIA): Investigations and leaks will keep coming, and whatever they find will be spun negatively against Nvidia.
- Stacy Rasgon (bearish on NVIDIA): Downgraded Nvidia despite cheap valuation; governance and investigation risk makes a buy rating hard to defend.
- Stacy Rasgon (bearish on NVIDIA): Losing Jensen Huang will be hard for Nvidia; new CEO inherits relationships built on his personal handshake deals.
- Stacy Rasgon (uncertain on NVIDIA): Nvidia's next roadmap calls, such as memory architecture after Feynman, are big risks without Huang.
What was said, by company and theme
11 views from 1 speaker. Every quote is checked against the episode's transcript (how).
NVIDIA
Across the shows: 147 speakers in other episodes, 135 bullish and 106 bearish calls, past 30 days
Where everyone stands on NVIDIA
Stacy Rasgon · Bernstein ResearchBullish
Nvidia's near-term data center business is not at much risk from the CEO exit; customers are not cutting orders.
Nvidia has not cut its numbers after the CEO exit. Rasgon does not think the near-term trajectory is at much risk and does not expect customers to buy less data center product next year because of the turmoil. Nvidia's 2027 guidance implies revenue growth of at least 70%, which would put the company near $700 billion. He expects the upcoming earnings to show no material change in the near-term trajectory, and says that a change would be very problematic.
“I actually don't think near term trajectory for their business is really at a ton of risk. I don't think customers at this point are buying less data center stuff from them next year because of this.”at 8:22
Stacy Rasgon · Bernstein ResearchBearish
Investigations and leaks will keep coming, and whatever they find will be spun negatively against Nvidia.
Rasgon sees it as clear that elements in the US government want Huang gone and that China hawks will push for more investigations. He expects a steady stream of news from inside and outside the company. If investigators subpoena texts and emails, they will likely find something, and it will be spun in the worst light. He sees this as the reason he cannot comfortably advocate the stock. He does expect that the board and Nvidia will eventually see that letting Huang go was a mistake, but says Huang made it impossible to stay.
“I think you're going to see more and more news flow and tape bombs coming up, both from inside the company as well as externally.”at 8:42
Stacy Rasgon · Bernstein ResearchBearish
Downgraded Nvidia despite cheap valuation; governance and investigation risk makes a buy rating hard to defend.
Rasgon's firm downgraded Nvidia from outperform and cut its price target roughly in half after Huang's departure. The case for keeping the buy is that the stock is cheap, was already cheap, and has become cheaper, and that the valuation prices in a lot of bad news. But he could not comfortably advocate owning the stock to clients given the expected flow of investigations, leaks and headlines about the company and its former CEO, plus uncertainty about the next CEO. He says the downgrade was hard to make, though the writing itself was quick. He still sees the stock as a potential buy on the dip for others.
“It's hard for me to maintain the buy on it. That's why we downgraded it.”at 10:12
Stacy Rasgon · Bernstein ResearchNeutral
Direct China revenue is small, about 8% of sales, with almost no data center shipments into China.
Nvidia now reports revenue by customer headquarters rather than shipment location, which makes China exposure clearer. Direct sales to Chinese customers, including Hong Kong, were roughly 8% of sales last quarter, and that includes much of its gaming and auto revenue. Singapore and Malaysia are about 1%, and US customers are over 60%, with Taiwan in the low 20s. Nvidia has licenses to ship older data center parts to China, but Beijing is not allowing purchases, so data center shipments into China are effectively nil. Rasgon thinks any smuggling would come through Taiwanese customers, but he doubts it approaches the 20% some estimates suggest.
“The direct sales to Chinese customers, so China including Hong Kong, I think last quarter was roughly 8% of their sales.”at 13:33
Stacy Rasgon · Bernstein ResearchBearish
Losing Jensen Huang will be hard for Nvidia; new CEO inherits relationships built on his personal handshake deals.
Huang ran the company with a flat structure: most key people report to him directly and he makes the key decisions, which let Nvidia move very fast. Customers and suppliers have added capacity because they trust Huang personally, and many supply-chain decisions rest on his handshake relationships. Rasgon is not worried about day-to-day operations, since the company has strong technologists, but he expects the loss of Huang's direct control to cause problems. He also flags talent retention, since many employees have very large net worths and may have stayed mainly because Huang was there.
“customers and suppliers that are presumably adding capacity because they trust Jensen. And Jensen's convinced them to take that leap of faith and build it. And I think those are going to be very difficult things for a new CEO to come in and fulfill.”at 21:12
Stacy Rasgon · Bernstein ResearchUncertain
Nvidia's next roadmap calls, such as memory architecture after Feynman, are big risks without Huang.
Rasgon is less worried about next year's revenue than about the longer roadmap. Rubin is shipping and Rubin Ultra looks baked, but Feynman and what follows still need architecture decisions, including memory architecture and how much HBM to use. These are major inflections that have to be called. Huang has a strong record of seeing five to ten years out, and his investment in CUDA about fifteen years ago, when many doubted it, created the current dominance. Rasgon does not know where a successor would see the puck going.
“I'm not worried about so much the revenue next year... But I mean, think about the roadmap... These are big, major inflections that will have to be called.”at 22:14
Stacy Rasgon · Bernstein ResearchUncertain
CUDA is still a real moat, but customers are multi-sourcing and AI-assisted porting could erode it over time.
Many large customers multi-source: OpenAI has deals with Nvidia, Broadcom and AMD, Meta has similar arrangements, and Google is selling TPUs and buying GPUs. Broadcom, AMD and Qualcomm are all growing their AI chip revenue, so competitors are doing fine. Rasgon thinks CUDA is still a moat, but AI is getting good at writing software, and companies are experimenting with using it to port code off CUDA. Customers may need to think harder about diversifying their footprint, not next year or the year after, but over the longer term. Keeping CUDA dominant depends on software decisions Huang made and that still have to be made.
“everybody talks about NVIDIA CUDA as a moat. CUDA is their software ecosystem. I think it is a moat.”at 33:17
Stacy Rasgon · Bernstein ResearchUncertain
Nvidia's vendor financing of AI buildout depends on Huang's willingness to commit capital; a new CEO may be less bullish.
Nvidia has a reputation as the bank of AI, and Rasgon thinks much of that was driven by Huang's willingness to commit company capital, including backstops and supply commitments. Reports suggest the CFO was not fully on board with the size of these commitments. Huang was clearly the most bullish on committing capital. It is unclear what appetite a permanent CEO will have, and that is one of the unknowns Rasgon flags. The CFO is Kress, who is now interim CEO.
“they do have this reputation of being the bank of AI. And again, I think a lot of this is... being driven by the force of Jensen's will.”at 34:45
Stacy Rasgon · Bernstein ResearchNeutral
Most of Nvidia's huge cash outlay goes to securing supply, not to backstops or financing.
Rasgon argues that the hundreds of billions of dollars Nvidia is said to be spending are mostly physical cash outlays to lock in supply, including memory, wafers and co-ops, rather than backstops and financing guarantees. He ties this to Jensen Huang's own partnerships and relationships, and says Huang was clearly the more bullish of the executives on committing company capital. The point is a correction to how the spending is being described, not a call on the stock.
“You have to remember most of the actual physical cash outlay that's gone out from [Nvidia] for stuff... not backstops and financing. It's to secure supply. It's to secure memory and wafers and co-ops and that kind of stuff.”at 35:31
US–China AI & chip race
Where everyone stands on US–China AI & chip race
Stacy Rasgon · Bernstein Research
Selling older Nvidia chips to China would have been wise; restricting them now is too late.
Rasgon agrees with Huang's view that the US should let China buy older chips such as H200s and H20s, which are several generations old. Lutnick has said the goal was to get China addicted to US technology so it could be monitored. Restricting sales only pushes China to build its own ecosystem, which is what is happening now, since China is blocking Nvidia purchases to favour local players. He would not sell Blackwell or Rubin chips to China. He also thinks the government's planned crackdown on Chinese firms renting cloud compute from outside China is a separate debate, and a hole worth plugging if Blackwell and Rubin access is large. Allowing sales of lower-end chips is a different question.
“So I'm on Jensen's side on the China policy. I don't think we'd be selling Blackwells or [Rubins] to China, but older stuff, I think it would have been a good idea. I think it's too late now, though.”at 26:56
AI infrastructure & data centers
Where everyone stands on AI infrastructure & data centers
Stacy Rasgon · Bernstein Research
The AI accelerator market is large enough for several suppliers, including Nvidia, to keep growing.
Rasgon's main question about competition is whether the opportunity is big. Nvidia has held most of this market, and its share will likely fall, but if the pie is large there is room for everybody. Nvidia's guidance implies roughly 70% growth. Broadcom has suggested its data center revenue will double to about $115 billion in 2027 and double again in 2028. AMD's GPU sales look set to almost triple, toward the low $40 billions, and Qualcomm plans about $5 billion. Everybody is doing fine right now.
“If it's big, there's probably room for everybody. And if not, everybody would be screwed, right?”at 31:53
Across the shows
What other episodes Chorus covered said about the same companies in the 30 days up to this one.
- NVIDIA147 speakers · 135 bullish, 106 bearish calls
Topics
- AI accelerator market size
- CEO succession and key-man risk
- CUDA moat and competition
- China chip export controls
- China revenue exposure
- legal and investigation risk
- near-term data center demand
- product roadmap execution
- stock rating and valuation
- supply commitments and capital spending
- vendor financing of AI buildout
More from Odd Lots
- The Impossible Trilemma That Hangs Over the Nvidia Crisis · November 16, 2026
- Why a $2 Billion Law Firm Suddenly Collapsed · November 13, 2026
- What Nvidia's Megaspeed Scandal Is Really All About · November 12, 2026
- The Perfect Storm That Clobbered Hedge Funds · November 11, 2026
- The Macro Implications of the Blue Tsunami · November 6, 2026
Ask your AI about this episode
Chorus gives Claude, ChatGPT and other assistants every view from this episode and every other show Chorus follows. Ask who agrees, who changed their mind, and what else was said about the same companies.
Chorus reports what speakers said. It isn't investment advice. How Chorus works · About Chorus