Could Chipmaker Nvidia Be the Tip of the AI Bubble?

1. As of the date of this article, officers, contractors, shareholders, and/or employees of Streetwise Reports LLC (including members of their household) own securities of Apple Inc. and Tesla Inc.
2. Steve Sobek wrote this article for Streetwise Reports LLC and provides services to Streetwise Reports as an employee.
3. This article does not constitute investment advice and is not a solicitation for any investment. Streetwise Reports does not render general or specific investment advice and the information on Streetwise Reports should not be considered a recommendation to buy or sell any security. Each reader is encouraged to consult with his or her personal financial adviser and perform their own comprehensive investment research. By opening this page, each reader accepts and agrees to Streetwise Reports' terms of use and full legal disclaimer. Streetwise Reports does not endorse or recommend the business, products, services or securities of any company. 

For additional disclosures, please click here.

On November 25, the week after its big earnings test, Chris Reilly of RiskHedge Research sent an email to subscribers noting five "awesome facts about Nvidia Corp.'s (NVDA:NASDAQ) (Chief Executive Officer) Jensen Huang."

"Nvidia tumbled 3% today," Reilly wrote about the Artificial Intelligence (AI) giant. "If you own it, don't stress the day-to-day moves . . . Obviously, NVDA is one of the greatest growth stocks of our generation. It soared 1,243% over the past five years."

At the top of the list was the fact that the Nvidia founder once cleaned toilets at Denny's. "He founded Nvidia with two buddies at a booth in that same Denny's 30 years ago," Reilly wrote. "Nvidia is now the world's most valuable company, and Huang is worth over US$160 billion. The American dream!" Other facts about the CEO include that he has a tattoo of Nvidia on his arm, is known for his leather jacket look (similar to Steve Jobs' turtlenecks), and competitor AMD's CEO, Lisa Su, is his first cousin once removed. Finally, Huang is the longest-serving tech CEO. He's been in that role since 1993. In the early days, he mortgaged his house to keep Nvidia afloat. Nvidia went public in 1999, at the height of the dotcom bubble, and it's up 466,000% since then.

"I'd say he's done a decent job," the writer said.

Creating 'Thousandaires' on Robinhood

According to a Yahoo! Finance "Morning Brief" email sent to subscribers on November 26, Nvidia has become a cultural phenomenon, turning Huang into a celebrity and creating new "thousandaires" on Robinhood. The company's name has become synonymous with the foundation of the AI surge.

However, this prominence has also dragged Nvidia into various AI bubble discussions it would prefer to avoid, the update noted. The company has been actively working to distance itself from these narratives.

Barron's reported late Monday that Nvidia sent a memo to analysts stating that it "does not resemble historical accounting frauds," addressing concerns raised by some analysts and commentators, the Yahoo! Finance update reported.

Then, on Tuesday, according to the email update, after The Information reported that Meta Platforms Inc. (META:NASDAQ) was in talks to purchase Google chips — news that caused Nvidia's stock to drop — the company took to X to express its delight in Google's success, while emphasizing that Nvidia's chips are "a generation ahead of the industry." They also highlighted that Nvidia offers "the only platform that runs every AI model and does it everywhere computing is done."

What Went Down?

Nvidia shares declined alongside the broader market on November 20, despite the chipmaker reporting better-than-expected third-quarter earnings and offering a strong forecast for the current quarter just a day earlier, as reported by Daniel Howley for Yahoo! Finance on the same day.

Nvidia's stock ended the trading session down 3.15% after initially rising at market open. Other members of the "Magnificent Seven" also saw declines. For the fourth quarter, Nvidia anticipates revenue of US$65 billion, plus or minus 2%, surpassing Wall Street's expectation of US$62 billion.

"Blackwell sales are off the charts, and cloud GPUs are sold out," Huang said, according to Howley's report. "We've entered the virtuous cycle of AI. The AI ecosystem is scaling fast — with more new foundation model makers, more AI startups, across more industries, and in more countries. AI is going everywhere, doing everything, all at once."

In Q3, Nvidia reported earnings per share (EPS) of US$1.30 on revenue of US$57.01 billion, exceeding analysts' predictions of EPS of US$1.26 on revenue of US$55.2 billion, according to Bloomberg consensus data. In the same period last year, the company reported EPS and revenue of US$0.81 and US$35.1 billion, respectively. The AI leader's data center business generated US$51.2 billion, beating estimates of US$49.3 billion. Nvidia's gaming revenue was US$4.3 billion, slightly below the US$4.4 billion forecast.

"Blackwell Ultra is now our leading architecture across all customer categories, while our prior Blackwell architecture saw continued strong demand," Chief Financial Officer Colette Kress noted in Howley's report. Revenue from its China-specific H20 chip was "insignificant," she added.

Nvidia's report follows the company's market capitalization, briefly surpassing US$5 trillion last month. Before the earnings release, Peter Thiel's hedge fund sold its entire approximately US$100 million stake in Nvidia, the Yahoo! Finance report said.

Despite recent gains, the AI sector faced criticism recently from investor Michael Burry, known for his successful bet against the housing market before the 2008 financial crisis, Howley wrote. In a post on X, Burry alleged that companies like Meta and Oracle Corp. (ORCL:NYSE) are inflating their earnings by downplaying the depreciation of data center equipment. During Nvidia's earnings call, CFO Colette Kress seemed to counter this assertion, stating that Nvidia's accelerators have a long lifespan due to the company's CUDA software, and that its six-year-old A100 GPUs continue to operate effectively because of ongoing CUDA updates.

Nvidia's Huang: No Bubble Here

In another report on the company for Yahoo! Finance on November 20, Howley noted that during Nvidia's (NVDA) Q3 earnings call, the company addressed its critics head-on, with Huang and Kress directly tackling concerns about a potential AI bubble.

Huang wasted no time in his prepared remarks, explaining why he believes Nvidia is uniquely positioned to maintain its leadership in the AI market. "There's been a lot of talk about an AI bubble," Huang stated, according to the report. "From our vantage point, we see something very different. As a reminder, Nvidia is unlike any other accelerator. We excel at every phase of AI."

Huang also addressed concerns about circular investments involving Nvidia and the broader AI sector, Howley said. These include arrangements like the one between Nvidia, Microsoft Corp. (MSFT:NASDAQ), and Anthropic PBC, where Anthropic will spend US$30 billion on Microsoft's Azure cloud services. In return, Nvidia and Microsoft have committed to investing US$10 billion and US$5 billion, respectively, into Anthropic.

Nvidia has been at the center of several such deals, including similar partnerships with OpenAI and other companies. This has raised investor concerns about the possibility of these agreements creating artificial demand for GPUs, fueling fears of an AI bubble, Howley wrote.

However, Huang explained that Nvidia's strategy of funding companies that subsequently purchase Nvidia chips is part of the company's broader effort to expand the reach of its CUDA ecosystem.

Crossroads: Tipping Point or Bubble?

According to a report by Arsheeya Bajwa and Stephen Nellis for Reuters on November 20, the company is at a crossroads for either a tipping point or a bubble. Huang believes in an AI transformation, dismissing the notion of an AI bubble. He said he envisions the type of computing Nvidia excels in becoming integral to everything from coding to operating fleets of robots in daily life, boosting investor confidence and driving Nvidia shares up over 5% in early Thursday trading.

However, a growing number of market skeptics worry that the only direction from a tipping point is downward. Despite the results and forecasts, there are long-term apprehensions that Nvidia's growth could be hindered by factors beyond the control of even the most valuable publicly traded company in history.

In a regulatory filing, Nvidia revealed that the majority of its booming business depends on four unnamed customers, the Reuters article said. In the third quarter, 61% of its US$57 billion in revenue came from these customers, up from a 56% concentration among four customers in the previous quarter.

Nvidia's high reliance on just a few customers it is entangled with, and the circular nature of some of its deals has raised concerns, particularly as none of the entities have reported massive profits from AI yet.

"A lot of this growth is coming from loss-making startups or loss-making projects, so most likely the cycle ends badly unless all these companies agree to stop spending together and let profits shine through, which is a near impossibility," said Chaim Siegel, an analyst at research firm Elazar Advisors, according to the Reuters article.

'Most Dangerous Bubble World Has Ever Seen'

But in a November 24 contributed opinion piece for Streetwise Reports, Michael Ballanger of GGM Advisory Inc. noted that "the emergence since 2000 of artificial intelligence has increased demand for the microchips that are the monopolistic domain of a company trading at 52 times earnings and 43.2 times sales while sporting a market cap of in excess of US$4 trillion."

Nvidia is not only using its monopolistic position to raise barriers for potential competitors but is also heavily dependent on the "LLM" (Large Language Model) technology for its survival — a technology that "may or may not work," according to Ballanger.

Ballanger quoted from Julien Garran's recent paper on the LLM AI "bubble."

"Make no mistake, I think that this is the biggest and most dangerous bubble the world has ever seen," Ballanger reported he said. "The misallocation of capital in the U.S., which also includes housing, VC, and crypto, is already 17 times the dot-com bubble and four times the 2008 real estate bubble. And as it unwinds, it will not just threaten substantial economic malaise, it will threaten to overturn the entire globalist agenda, and the new financial order that developed with the advent of Thatcher and Reagan from 1979 and 1982, accelerated with the fall of the Berlin Wall in 1988, and sped up again with China's accession into the WTO in 2002. The LLM AI bust will be as important a backdrop to the next phase in global macro as the savings and loan crisis was enforcing the pace of US reflation from 1991 that saw the dramatic rise of the Southeast Asian economies at markets of South Korea and Taiwan in the early '90s. Except this time, unlike 1994, the train has no brakes. We are increasingly likely to see reflation, inflation, and the rise of the Indian middle class along with the fall of AI emerge as the dominant macro themes of the second half of the 2020s."

"The extensively documented theory that LLM AI is a massive bubble aligns with everything I've been hinting at but have refrained from stating outright, mainly because it might overshadow my microeconomic perspectives on gold, silver, and copper," Ballanger wrote. "All 'bubbles' originate from the environment of easy money and central bank interventions. The self-proclaimed visionaries leading Nvidia, Microsoft, Apple Inc. (AAPL:NASDAQ), and the infamous Tesla Inc. (TSLA:NASDAQ) are part of the same close-knit group of stock promoters who have mastered the art of using the right jargon to push major investors worldwide 'over the wall,' fostering self-perpetuating, cult-like behaviors in areas such as LLM AI, crypto, and more."

The key takeaway from the interview is: "You're going to see emerging markets and resources winning," Ballanger said.

Not the Only Option

Nvidia is not the only option in the sector, of course. On November 25, GGM sent an email update to clients featuring an article on the recently announced Alphabet Inc. Class A (GOOGL:NASDAQ) tensor processing unit (TPU), which is being touted as "less expensive" than the Nvidia graphics processing unit (GPU) chip.

Tyler Durden, writing for UncoverAlpha on ZeroHedge.com on November 25, quoted a former Google Cloud employee as saying, "The amount of performance per dollar that a TPU can generate from a new generation versus the old generation is a much significant jump than Nvidia."

"The elimination of Nvidia Corp.'s monopoly is the obvious result of Alphabet's development of the 'TPU chip,'" Ballanger said. "If this AI bubble forces companies like Alphabet and Microsoft to start using chips less expensive than those being sold at ridiculous prices by NVidia, then the competitive advantage leading to such enormous growth disappears."

An intriguing paired trade could be "Long GOOG/ short NVDA," aligning with Buffett and betting against the bubble, the advisory noted.

"If we experience a year-end rally (which I highly doubt), GOOG is likely to outperform NVDA. However, for speculators like myself, I anticipate a swift, sharp drop to the 200-day moving average, which could elevate the NVDA December US$170 puts to approximately US$17.00 from their current level of US$7.65 . . . If we get a year-end rally (which I strongly doubt), GOOG will outperform NVDA," the update said.

'Flashes of Dotcom Tulips'

In a research note on November 23, Chaim Siegel of Elazar Advisors LLC downgraded the stock to a Neutral rating and revised the firm's EPS from US$11.64 to US$9.62, a projection that assumes a 25% sequential growth in the Datacenter segment through 2026. This growth was implied when CEO Jensen Huang mentioned a US$500 billion backlog through 2026. Achieving this would require Datacenter to grow sequentially by 25%.

"I see bull-case sell-siders moving to that US$11 range, Siegel wrote. "I would believe that hedge funds were there already on the CEO's mid-quarter comments. So, my numbers, while above the mass of sell-siders who have updated, are now below the bull-case group, which everybody is probably following and hoping for."

The analyst continued, "My hair is gone, and my beard is almost fully gray, so I do have some flashes of dotcom tulips," referring to the collapse of the Dutch tulip-buying bubble in 1637 and the dotcom bubble.

"Nvidia's sell-side and hedge fund bull cases likely need a smooth path, no bumps," Siegel said. "Our numbers now assume a touch less than that, so we move to Neutral after a big run. If you ask us if the stock is still cheap on our numbers, I'd agree. But still, tech doesn't like slowdowns. This is an acceleration but now factored into the Street. But the reality, within a year or so, much like FSD (full self-driving) has needed to prove itself, AI I think will need to prove itself with a path to a profitable business model to hit those numbers."

Analyst Remains 'Negative' on Nvidia

Seaport Research Partners Analyst Jay Goldberg said the firm remains "negative on Nvidia as signs of competition increase" in an updated research note on November 30.

"We see Nvidia facing growing competitive pressure," wrote the analyst, who rated the stock Sell with a US$140 per share price target, compared to the stock's US$177 per share price tag at the time of writing. "To address this, the company has been leaning on a variety of sales mechanisms to adapt. These measures are not fully reflected in financials, but they are already material and look likely to grow significantly next year."

Goldberg said Nvidia has entered into US$26 billion worth of cloud computing service contracts. The company asserts that these will be utilized for research and development and its DGX product line. Goldberg said he viewed these as a type of rebate which, if accounted for, would reduce gross margins by 400 basis points next year, or at least US$0.30.

However, Google has unexpectedly succeeded in encouraging third-party adoption of its internally developed TPUs, Goldberg wrote. While TPUs are not suitable for everyone, they can surpass Nvidia systems in many performance measures.

The company invested US$6 billion this year in private enterprises. It has additional commitments totaling US$17 billion (including US$5 billion to Intel). The agreement with OpenAI is not finalized, but could potentially add another US$100 billion to the commitments.

"Working capital is up significantly this quarter," the analyst said. "The company reasonably maintains this is a bullish indicator signaling demand. We also see it as a function of the company smoothing out working capital needs among its ODM customers."

Expert: Nvidia Can Go to US$10 Trillion

But not everyone is as dour in their predictions. In an interview with growth stock expert Chris Wood published on November 24, Reilly of RiskHedge wrote about "the next phase of the AI boom" and how it will affect Nvidia.

"First, some perspective. Nvidia hit the US$5 trillion milestone less than three months after hitting US$4 trillion," Wood said. "Berkshire Hathaway (BRK.B) is worth about US$1 trillion today. It took super-investor Warren Buffett about 60 years to build Berkshire into a trillion-dollar company. Nvidia added that much value in months."

Reilly asked, how much higher, realistically, the stock could go.

"Nvidia can become a US$10 trillion company before 2030, possibly as early as 2028," Wood answered. "Right now, it's sitting at around US$4.3 trillion after the recent dip. Nvidia is sitting at the center of one of the biggest megatrends of our lifetime."

Moving from “box-level computing," these industrial AI systems create a ripple effect across the entire ecosystem, Wood said.

"Every part of the rack, from energy delivery and cooling to networking and memo Bandwidth, must scale up," Wood told Reilly. "They're the bottlenecks shaping the AI boom. And that's incredibly bullish for the companies providing these solutions. The ones powering, cooling, and connecting the AI engines of the future."

Ownership and Share Structure¹

About 5% of the company is held by insiders and management and holding companies, and about 69% by institutions. The rest is retail.

The company's market cap is US$4.3 trillion with 24.3 billion shares outstanding. It trades in a 52-week range of US$86.62 and US$212.19.