Michael Burry earned a whopping $800 million by shorting the U.S. housing market ahead of the 2008 financial crisis. Whether the famed investor has made comparable money since then is far less clear. Still, his reputation endures. Investors continue to closely track his high-profile bets, hoping to ride his coattails to similar gains.

Burry’s X handle, @michaeljburry, has long been known for its pattern of posting and deleting, a tactic that adds to the mystique around his trades and, more recently, funnels attention to his new newsletter, “Cassandra Unchained.” Launched on Substack after he closed his hedge fund, Scion Asset Management, in November, the newsletter lays out the technical logic behind his bets while allowing Burry to sidestep the SEC disclosure requirements that apply to institutional investors managing more than $100 million.

His celebrity status was cemented by the 2015 film The Big Short, which turned Burry into a household name. That visibility has granted him a level of credibility few investors retain for so long, even when their predictions miss the mark.

“People like superstars, and they love to listen to folks who they think are smart and successful,” Tom Sosnoff, founder of investment media network Tastylive, told Observer. “He is a personality and a contrarian. He is interesting and pretty famous in the world of finance. Love him or not, people listen to him.”

While Burry’s early success is well documented, his performance since then is harder to evaluate. As a hedge fund manager, he is only required to disclose limited information through quarterly filings such as 13Fs, which reveal long equity positions but not short positions, derivatives or overall performance. As a result, the full picture of his gains and losses remains largely opaque.

There have been claims that Burry has made more than $1 billion in total trading profits, but those figures have never been independently verified. “He has never been audited before. It’s his word against others,” Sosnoff said.

Nvidia and Palantir in the crosshairs

Despite the uncertainty around his track record, Burry’s words still move markets. His recent bearish bets against Nvidia and Palantir have drawn particular attention, with Burry arguing that both sit at the center of an A.I.-driven market bubble.

On Nov. 3, regulatory filings revealed that Scion had placed roughly $1.1 billion in bearish options positions tied to those companies. The structure of the trade—largely long-dated put options—gives him time for the thesis to play out rather than requiring an immediate downturn.

“His timing was very good,” said Sosnoff. “He pretty much got short Nvidia near the top (around $200), and it’s now down 10 percent to 15 percent. It’s a good call.”

Palantir, which represents Burry’s largest short at roughly $912 million, has not fallen as sharply. The stock is down about 7.8 percent from its Nov. 3 level. Still, because the position is structured with options expiring in 2027, some analysts say it’s far too early to judge.

“His logic is extremely good, and he has over a year to be right,” David Trainer, CEO of A.I.-driven investment research firm New Constructs, told Observer.

Trainer, a former hedge fund manager, also backed Burry’s broader critique of A.I. hyperscalers, arguing that companies such as Oracle and Microsoft are using aggressive accounting practices, particularly around GPU depreciation, to flatter earnings.

“These companies are definitely using questionable billing and receivables to make their earnings look better,” said Trainer. “I can’t say if Burry has been right or wrong in previous trades, but I think he has made some money. “This time [with the A.I. Bubble], he seems right.”

The cult of the contrarian

Not everyone is convinced. Matthew Tuttle, CEO of Tuttle Capital Management and a frequent contrarian himself, said Burry’s post-2008 track record is far less impressive than his reputation suggests.

“When you look at the calls Burry has made since 2008, they have not been good,” he told Observer. “He has said ‘this is going to crash and that is going to crash’ many times since, and he hasn’t been right.”

Still, big bearish bets tend to attract attention precisely because they go against the grain.

“Any time someone makes a major down call, there’s a fascination with it as long [bullish] calls are always okay because the market always goes up,” said Tuttle.

That dynamic helps explain why hedge fund stars can remain influential long after their best trades are behind them.

“If I’m the main character in a movie and in a book like Burry and have been right in a big way, that buys me a lot of getting things wrong,” added Tuttle.

The same dynamic applies to other market personalities such as Robert Kiyosaki, Peter Schiff and CNBC’s Jim Cramer, whose reputations often outlast their accuracy.

“Robert Kiyosaki is constantly calling a bear market, and he is wrong, and Peter Schiff has been calling gold up for a long time,” said Tuttle. In Schiff’s case, it eventually worked—but more because of timing and luck than brilliance.

“When you say gold is going to go up every year, and one year it does well, does that make you a genius? I would argue it doesn’t,” he added.

Fame as financial fuel

Wall Street is full of one-hit wonders whose early success grants them enduring influence.

“Most of the time, they don’t risk their money,” said Sosnoff. “If they have one big win one year, they’re set. Their reputation is made.”

John Paulson, who famously made $15 billion betting against subprime mortgages, fits that mold, as do figures like Ralph Acampora, who called the 1990s bull market, and Paul Tudor Jones, who predicted the 1987 crash.

Other famous short sellers have stumbled. Jim Chanos, known for shorting Enron, closed his Kynikos fund in late 2023 after his Tesla bet went wrong. Bill Ackman lost roughly $1 billion betting against Herbalife in 2018, despite previously scoring a massive win betting against mortgage insurers during the financial crisis.

Ultimately, fame often matters more than accuracy.

“We live in a world where celebrities (movie, social media) have megaphones, and Michael is a celebrity because of the movie,” NYU Stern professor Aswath Damodaran told Observer. “Put simply, I will wager that most people who follow his advice (good or bad) are doing so because they liked the movie, think he is Christian Bale or like Batman, rather than because they read his treatises on Nvidia or Palantir. “

That doesn’t mean Burry lacks insight. “Michael actually is a good macro thinker and often willing to break away from the herd,” Damodaran added. “But so are many other smart investors who never get noticed.”

“Short at your peril.” That’s what some Tesla investors are saying in response to famed short seller Michael Burry’s latest warning, in which he slammed the EV giant as “ridiculously overvalued.” Burry took aim at the upcoming share dilution tied to CEO Elon Musk’s massive pay package, arguing that Tesla’s management hasn’t adequately informed shareholders about its impact.

Musk’s $1 trillion stock-based compensation plan, overwhelmingly approved by Tesla shareholders in a November vote, will dilute existing investors’ ownership by 3.6 percent each year, Burry said in his newly launched newsletter.

“Tesla’s market capitalization is ridiculously overvalued today and has been for a good long time,” Burry wrote. “With recent news of Elon Musk’s $1 trillion dollar pay package, dilution is certain to continue.”

Dilution affects a stock’s price by mechanically reducing a company’s profitability and value on a per-share basis. When new shares are issued for stock-based compensation, the company’s market value is spread across a larger number of shares, pushing down per-share metrics like earnings per share. If management doesn’t counterbalance this with buybacks or rapid growth, selling pressure can build.

Tesla bulls, however, are betting that the company’s future growth will outweigh the impact of dilution. Musk’s compensation package is structured as 12 tranches totaling 425 million shares. The first tranche of 35.3 million shares unlocks once Tesla’s valuation hits $2 trillion (up from the current $1.4 trillion). Some investors believe Tesla could reach this milestone as early as next year, if its robotaxi service takes off.

“There will be some dilution, but not much,” Noah Hamman, founder and CEO of EFT investor AdvisorShares, told Observer, adding that “the resulting valuation you will get from that news will be huge.”

Musk’s new pay package is structured similarly to the one approved in 2018. That earlier plan didn’t dilute the stock thanks to Tesla’s meteoric expansion (from a $54 billion valuation in 2018 to $650 billion in 2023), Hamman noted.

“There was no concern about dilution,” Hamman said. “I don’t know how anyone could have been unhappy with the stock.”

Matthew Tuttle, CEO of Tuttle Capital Management, known for issuing leveraged ETFs, backed some of Burry’s views.

“Burry’s core point is right: stock-based comp is not ‘free’ — it’s equity issuance used to pay wages, and the bill shows up as dilution (or as cash buybacks later, if the company chooses to offset it),” he told Observer. “When investors fixate on ‘adjusted’ earnings that add stock-based compensation back [later down the road], they can end up valuing a business on a pre-dilution fantasy.”

But Tuttle disagreed with Burry’s criticism of Tesla’s valuation. He argued that Tesla’s broad technological footprint, including EVs, autonomous vehicles, robotaxis, energy storage and software, means it can’t be valued as a pure-play automotive company.

If one of those options “hits at scale, the stock can stay expensive on conventional metrics for a long time,” he said.

What a difference a day can make. Nvidia shares surged on Wednesday after the chip giant reported better-than-expected earnings and guidance, defying growing talk of an A.I. bubble and high-profile exits from investors like Peter Thiel. But just 24 hours later, Wall Street’s A.I. darling nosedived more than 3 percent yesterday (Nov. 20) as hopes for another interest rate cut this year faded following a hotter-than-expected September jobs report.

Those renewed rate concerns, combined with lingering fears that A.I. stocks are dangerously overpriced, triggered a sharp market reversal. The Dow erased a 700-point gain to finish down 386 points yesterday, while the Nasdaq fell 486 points and the S&P 500 dropped 103.

“NOT a bubble”

Nvidia reported total revenue of $57 billion for the August-October quarter, beating analysts’ expectations of $55 billion, with EPS of $1.30 versus the predicted $1.26. The company predicted that sales for the current quarter ending January should reach about $65 billion, above Wall Street’s consensus of $61.7 billion.

Wedbush analyst Dan Ives praised the results, arguing they show the A.I. boom remains solid. He pointed to Nvidia’s $2 billion data-center revenue beat and strong demand for its upcoming Blackwell and Rubin chips. He reiterated his view that the A.I. build-out is still in its early stages, with tech giants—or “hyperscalers” such as Google, Meta and Microsoft—expected to spend between $550 billion and $600 billion next year to accelerate A.I. adoption.

“The pure Nvidia numbers/guidance and strategic vision show the A.I. Revolution is NOT a Bubble…instead it’s Year 3 of a 10-year build out of this 4th Industrial Revolution in our view,” Ives said in a client note this week.

Why notable investors are dumping Nvidia stock

Still, after Thiel sold his $100 million stake in Nvidia—joining other major sellers like SoftBank, which recently unwound $5.8 billion—some investors began to wonder whether Nvidia and the broader A.I. trade may be approaching a plateau.

Concerns over lofty valuations have grown louder, particularly from famed short-seller Michael Burry, who has a $187 million short position against Nvidia. Burry has accused Oracle, Meta, Amazon and Google of understating depreciation for their A.I. infrastructure to inflate earnings through 2028.

Joseph Schuster, CEO and founder of ETF operator IPOX Schuster, said recent selling looks more like profit-taking than a sign that an A.I. bubble is ready to burst.

“The market is rotating, consolidating and digesting huge gains,” he told Observer, noting that investors are in a “risk-off market” for the rest of the year. “It doesn’t mean that something will burst or fall apart. It’s a normal part of the cycle.”

Schuster doesn’t expect a dot-com-style correction and remains optimistic about the A.I.’s long-term fundamentals, but he said Nvidia and other big tech names could see more downside in the weeks ahead. Nvidia, specifically, could slip to the $165–$170 range, he said.

Jane Edmondson, head of thematic strategy at ETF research firm VettaFI, agreed. “I have been in bubbles before, and it does seem different this time,” she told Observer. “I do feel like we are still very early days and only just now starting to see [A.I.] use cases that could be groundbreaking.”

“If you look at Nvidia’s order books, they are very strong and Huang expects lots of business to come, and I agree with him,” she added.

She rejected views that Nvidia’s price-to-sales (P/S) ratio of 30x is unsustainable, saying, “I have never seen companies achieve this level of growth. It’s exponential.”

As the market matures, Edmondson said, the winners and losers will become clearer. She expects firms with their own growth capital—such as Google, Meta and Alphabet—to emerge stronger than those that rely heavily on leverage.

“If someone can do it, it’s JB Straubel.” That’s how Edward Sanchez, a senior analyst at automotive consultancy TechInsights, describes the strategic vision of Tesla’s co-founder and former CTO. Straubel has set out to build the world’s largest electric vehicle (EV) battery recycling company. Founded in 2017, his startup Redwood Materials aims to create a circular supply chain for lithium-ion batteries, sourcing them from automakers, dealers, parts suppliers, dismantlers and waste management firms.

The company is currently recycling 20 GWh of batteries annually—enough to power 250,000 EVs—and is working to increase that capacity substantially over the coming years. To get there, Redwood is building a $3.5 billion battery materials and recycling facility in Ridgeville, S.C., which will produce cathode and anode components for future vehicles and employ 1,500 people in the U.S. “Battery Belt.”

Straubel has raised alarms about A.I.’s soaring energy demands, warning they could strain future power supplies. In June, he launched Redwood Energy to help tackle the issue by repurposing second-life battery packs—still with half their useful life left—to power data centers and stabilize the electric grid.

To jumpstart that effort, Redwood Energy struck a deal with Crusoe, a company focused on sustainable energy for A.I., to power a new data center in Abilene, Texas. The project will run on a 63-MWh microgrid designed to operate independently from the main electricity network.

​”We are developing techniques that use A.I. and robotics to conduct these tests in an automated fashion because one of the challenges we have with lithium-ion recycling is that there are an enormous amount of batteries of different [power] cell types and shapes,” Harper told Observer. “By using computer vision systems, we can reduce a great deal of labor.”

Once a battery is unpacked down to the individual cell level, A.I. can help identify the so-called “black mass”—the material inside the cathode and anode. The cathode contains the most valuable charge metals, including cobalt, nickel, manganese and lithium.

Redwood uses a heating method known as “reductive calcination” to recover those metals and use them to build new cathodes for fresh cells. ​”An A.I. model can automatically detect the cathode’s materials and optimize the reductive calcination process,” Ali Khazaeli, an independent battery data scientist, told Observer. “For instance, if you have a nickel manganese cobalt (NMC) oxide cathode or a nickel manganese cobalt aluminum (NCA), you will need to use different operating temperatures and conditions.” Newer Teslas use NMC while older models use NCA.

A Redwood spokesperson declined to say whether the company is applying A.I. to grow its business. Analysts, however, believe it’s very likely under consideration—if not already happening behind the scenes.

​Cellphones and laptops could be a “gold mine”

Sanchez agreed that A.I.-driven automation could help Straubel’s company scale, especially as it looks to incorporate consumer-electronics batteries into its operations. EV battery packs are harder to come by, since cars typically last 10 to 15 years. Early models like the Tesla Model S are just now reaching the point of repurposing, while most others—at least for Tesla—won’t be widely available until 2028 to 2030.

“JB has said the next big gold mine is in people’s desk drawers—cellphones, iPods and laptops that no longer have a use for consumers but whose batteries are still valuable,” Sanchez said. “It will be a more manual process as they will be dealing with smaller batteries instead of huge packs coming from cars,” he added, noting the logistical challenges this shift would create.

Redwood has not announced partnerships with consumer electronics giants such as Apple or Samsung, but Sanchez said such deals “will make sense in the future.”

​”They have good OEM partnerships such as with Toyota, GM, VW and BMW, so from that standpoint, they are fairly well positioned [to meet their growth targets],” he said. “The challenge for them will be to pursue small-scale battery harvesting from consumer electronics while procuring materials from second-life packs. They are taking on tremendous logistical operations, so the ramp-up to 5 million cars is going to be challenging for sure.”

Apple, meanwhile, has already stepped up its own recycling efforts. The company plans to use 10 percent recycled cobalt in all Apple-designed batteries by 2025 and has a partnership with Li-Cycle, the bankrupt recycler recently acquired by Glencore.

​”A lot of companies rushed into battery recycling to get a first-mover advantage, but then realized there weren’t enough good materials to recycle,” Harper mused. “As more EVs hit the market, however, I don’t think this will be a problem in the future.”

Correction: A previous version of this article contained an outdated forecast of Redwood Materials’ recycling capacity. 

A German-style fix?