The ultimate foundation of AI is electricity, and behind electricity is copper! "The Big Short" bets on the resource gap behind the AI computing power frenzy
Burry separately assesses the investment value of leading companies in AI infrastructure and the resource demand brought by AI construction—he firmly questions the investment returns of certain AI stocks, while being optimistic about the long-term demand for copper resources required for the expansion of data centers, power grids, and distribution systems.
According to Zhitong Finance APP, Michael Burry, famously known as "The Big Short," continues to steer clear of the overcrowded and leveraged AI compute theme trades and searches for value in other sectors, positioning in globally renowned copper miner Ero, QXO led by Brad Jacobs, as well as three significantly discounted stocks focused on home consumption, food retail, and animal health assets.
Michael Burry has long held a skeptical stance toward the AI compute expansion and the frenzy of AI infrastructure investment. The key logic in Burry's latest portfolio moves, which retail investors should pay attention to, is that he prices "AI company valuations and profit quality" separately from "the physical procurement demand generated by AI construction" — meaning Burry evaluates both the compute chain-driven investment value along the AI infrastructure supply chain and the resource needs brought about by AI buildout. While he remains firmly skeptical of the investment returns from certain AI stocks, he is optimistic about the long-term copper resource demand required for data centers, power grids, and distribution system expansion.
Burry’s recent skepticism toward the global AI compute stock investment frenzy centers around the extreme valuation of hot AI-related stocks, the relationship between capex returns, equipment depreciation, and future payment commitments. In his position disclosure on September 9, he reduced portfolio risk by selling Nvidia and Palantir puts expiring December 2026 without rolling them, while still holding short positions in both companies’ stocks, as well as holding Palantir and QQQ ETF puts expiring 2027 (QQQ tracks the Nasdaq 100 Index, regarded as a barometer for AI and tech stocks). At that time, his top three stock short positions were Oracle, Palantir, and AI cloud leader Nebius.
On September 19, he further expanded his focus to Amazon, Meta, Alphabet, Microsoft, and Oracle, estimating that these five companies' off-balance-sheet commitments around yet-to-start leasing, procurement, and guarantees exceed an astonishing $3 trillion. He worries there is a “catastrophically severe mismatch” between long-term fixed expenditures and the rapidly iterative AI hardware and still-unrealized commercial returns, warning that “when the music stops, these off-balance-sheet commitments can very quickly become real liabilities.” These latest market developments and public information reinforce his bearish stance on AI compute industry leaders, while he actively adjusts his shorting tools and position durations.
Away from the Cheers of the AI Party, Seeking Overlooked Value—Burry Ignores AI’s “Cheerleaders” and Finds Opportunity in Copper, QXO, and Three Other Stocks
Although staying away from the overheated, leveraged AI compute stocks, Burry is indirectly betting on AI via ERO, stating that data center construction will boost copper demand, while it takes 18 years to bring new mines online.
“The room is packed, and today is pushing AI-related stocks higher, but I basically ignore those ‘cheers’,” the “Big Short” investor wrote in a subscriber post on Substack.
Burry named five companies: Ero Copper (ERO.US), QXO (QXO.US), Temple & Webster, Sprouts Farmers Market (SFM.US), and Zoetis (ZTS.US). He also disclosed holding both QXO common shares and its mandatory convertible preferred stock.
Burry is indirectly betting on AI through Ero, a copper and gold producer focused on Brazil. “Everyone in the room is going to need a lot of copper,” he said.
On Monday, COMEX copper futures settled at $6.6865 per pound, about 2% below the record high on September 9. Driven by disruptions in mine production, low inventory levels in China, and inventory restocking ahead of China’s National Day, the international copper benchmark—the LME copper futures price—has risen 19% so far this year and 46% over the past 12 months. Its annual performance exceeds SPDR S&P 500 ETF Trust (SPY.US), Invesco QQQ Trust (QQQ.US), and SPDR Dow Jones Industrial Average ETF Trust (DIA.US), with the latter three returning about 13%, 18%, and 9%, respectively.
Burry acknowledges the recent rally partly reflects temporary supply weakness. His long-term investment thesis is based on a severe mismatch between copper demand and the time required to develop new mines. Quoting Apollo Chief Economist Torsten Slok, Burry notes that the number of large copper discoveries of over 500,000 tonnes has dropped from double digits per year in the 1990s and early 2000s to only one or two per year recently—and none expected in 2025.
It takes 18 years to bring a new deposit online, while AI data center demand for copper can materialize within two to three years. S&P Global forecasts global copper consumption will rise from 28 million tonnes in 2025 to 42 million tonnes by 2040.
Burry believes Ero is a higher-cost copper producer whose profit margins could expand significantly if copper prices remain high. “Ero common shares are just right for me,” he wrote in his subscriber post, calling it a mid-sized position.
Ero maintains its 2026 production guidance at 67,500–77,500 tonnes, with C1 cash costs between $2.15 and $2.35 per pound. Burry also praised the management team that started in early 2025, saying they delivered “three rather difficult commitments” and adding that a forensic review of the company’s accounting “found nothing amiss.”
Despite ERO shares having doubled in the past year, Burry still sees significant valuation discounts compared to larger copper miners. On Stocktwits, retail sentiment on ERO fluctuated between “neutral” and “bearish” in the past week, even as thread volume doubled and the number of watchlists increased by 5%, indicating rising retail interest.
Burry also said he strongly supports Brad Jacobs’ M&A integration moves with QXO. He holds QXO common stock as well as the B series mandatory convertible preferred stock with a stated dividend rate of 5.5%. “The share price has already fallen a lot, and I see an opportunity to participate,” he said.
QXO, led by United Rentals and XPO founder Brad Jacobs, is consolidating a fragmented building materials distribution market through serial acquisitions. Burry believes scaling up and strong execution may drive a re-rating, though the stock is still exposed to “shocks in the housing market, interest rates, not to mention diesel prices.”
Burry likes QXO’s preferred shares, noting their current yield is 7.4%, and dividends are cumulative. Unless converted early, the securities will automatically convert to common stock and be settled around May 15, 2028. “The attractiveness of both the common and preferred shares is broadly similar, but I like the dividend’s downside protection,” he said.
On Stocktwits, retail sentiment for QXO has risen from “bullish” a week ago to “strongly bullish” and thread volume increased 467% month-on-month, with a 4% increase in watchlists.
Burry increased holdings of Australia-based online furniture retailer Temple & Webster, traded OTC under TPLWF, calling it “a fairly big position.” Fears over the housing and consumer spending caused the stock to fall about 82% in the past year. For fiscal 2026, revenue grew 11% to A$665 million but net profit dropped 62% to A$4.3 million. Burry said he plans to disclose the full investment rationale soon.
The short-focused investor also made “modest” increases to Sprouts Farmers Market and Zoetis, completing his buying spree. Sprouts shares have fallen about 43% in the past year, and Zoetis shares have halved amid weak companion animal demand and downgraded performance guidance. Burry said Zoetis has “always impressed me,” but is now “extremely out of favor,” lacking both pandemic-driven tailwinds and any connection to AI, and is not on the market’s radar currently.
On Stocktwits, retail sentiment for Zoetis was still “bearish” last week with thread volume down 75%, but the watchlist count has grown 9% in the past month. Meanwhile, retail sentiment for Sprouts improved from “bearish” a day earlier to “neutral,” still below the “bullish” tone of a week ago; thread volume fell 67% while watchlists held steady, up just 0.1% in the last month.
The “Big Short’s” Contrarian Investment Playbook: Shorting AI Valuations but Betting the End Game for AI is Power—and Power Needs Copper
As “The Big Short,” Michael Burry questions the valuations and profit quality of popular AI compute stocks while being optimistic about the long-term demand for power grids and distribution systems amid the AI data center construction boom. Burry sees investment opportunities in the strong copper demand driven by grid and distribution system expansion, and is positioning accordingly.
Burry’s central bullish thesis is slow supply response: It can take up to 18 years from copper mine discovery to production while a data center can go operational within two or three years, resulting in a clear time mismatch between new procurement demand and mine supply. S&P Global predicts global copper demand will grow from 28 million tonnes in 2025 to 42 million tonnes in 2040, an increase of about 50%, driven by AI, electrification, and other economic activities. So, Burry’s copper investment targets resource links where buildout demand is realizable but supply is hard to expand quickly.
Regarding Ero Copper, he’s betting on the profit elasticity created by rising copper prices and operational improvements. The company reaffirmed its 2026 copper production guidance of 67,500–77,500 tonnes, C1 cash costs of $2.15–$2.35 per pound, with expectations for higher output and lower costs due to operational improvements in the second half.
According to Burry’s latest view, Ero still holds a valuation discount versus large copper miners. Its higher costs mean slim initial margins, so with controlled output and costs, copper price increases would drive substantial profit growth. Delivery of operational commitments by the management team adds a company-level support to this price elasticity—combining “resource supply and demand, operational improvement, and valuation discount” as the trio of investment factors.
Copper miners with operating resources, feasible expansion projects, and stable supply networks are well-positioned to benefit from the scale-up of AI inferencing. As inference calls, agent tasks, and online services scale, compute demand will transmit to server capacity, power supply, and distribution construction.
IEA (International Energy Agency) now forecasts that global data center power consumption will nearly double from 485 TWh in 2025 to 950 TWh in 2030. Copper is used in power cables, transformer windings, distribution equipment, and busbars in equipment racks, with its conductivity helping to control resistance loss in high-power devices; short-distance high-speed rack interconnects also use copper cables and related components—for example, Nvidia’s GB200 NVL72 uses copper cable boxes for intra-rack connections.
Power infrastructure makes up the main demand for copper, while high-speed copper interconnects reflect more of AI’s added technological value in signal integrity and the manufacturing of data center internal high-speed connectors and cables. At the same time, longer-distance links in data centers are speeding up adoption of optical interconnects and CPO technology. For company earnings, miners primarily benefit from resource scarcity, sales, and copper price-driven profit elasticity; traders mainly gain from supply organization, logistics, financing, and arbitrage; high-speed cable manufacturers gain from product upgrades.
The other four stocks reflect a logic of value recovery and capital allocation. QXO’s appeal comes from Brad Jacobs’ experience in consolidating fragmented construction materials distribution markets and the potential for efficiency improvements and re-rating post-scaling. Burry holds both the common and B series mandatory convertible preferred stock, with the 7.4% yield based on then-current prices. Temple & Webster, Sprouts Farmers Market, and Zoetis, meanwhile, represent contrarian strategies in home consumption, food retail, and animal health, respectively.
Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.
You may also like
US Dollar Index pushes higher on the case for another Fed increase

SHIB Pay Expands Shibarium’s Payment Push
XRP rebounds 9% to $1.54, analysts eye resistance after “Uptober” narrative returns


