Doug Casey Calls AI a Super Bubble, Bets on Energy, Gold Miners and Grains

Key Points

• Doug Casey argues today's AI infrastructure buildout mirrors history's biggest speculative manias, and warns that even profitable "picks and shovels" suppliers could get swept down if the broader data center boom unravels.
• He sees deep value in energy, naming Ecopetrol and Petrobras for their high single-digit dividend yields, alongside uranium and coal as the unglamorous power sources he expects to fuel AI long after any bubble bursts.
• Casey's most contrarian picks are small-cap gold and silver miners—where all-in sustaining costs near $1,700 against $4,000 gold are finally generating real margin—and agricultural commodity ETFs tied to corn, wheat, and soybeans he sees sitting at cyclical lows.

AI bubble fears keep resurfacing, and depending on who you ask, the story is either just getting started or already cracking at the edges. Doug Casey, founder of International Man and a self-described technophile who has invested across more than 50 years and 155 countries, falls firmly in the second camp. He thinks the AI trade isn't just a bubble. He thinks it's a historic mania, and he's putting his money in three places most investors aren't looking: energy, mining, and farm commodities.

Casey doesn't dispute that artificial intelligence will reshape the world. What he disputes is whether the companies building it out right now have any real path to earning money from it. That's the tension running through his entire pitch.

A Super Bubble, Not Just a Bubble

Casey's read on the market is blunt. He believes today's AI spending could eventually be compared to historical manias like the Mississippi Bubble or the South Sea Bubble, and possibly dwarf the 1929 stock market crash. Margin debt has surged roughly 50% over the past year, by his estimate, and he sees retail investors pouring money into companies with little revenue and no earnings.

He uses SpaceX NASDAQ: SPCX as his case study. Casey's concern—that most of the capital Elon Musk has raised is flowing into data centers and AI rather than the core rocket business—is now playing out in public markets. SpaceX completed its IPO in June and carries a market cap above $2 trillion, even as it posts steep GAAP losses tied to its AI and infrastructure buildout.

His broader point: a company can be technologically dazzling and still be a poor investment if the price already assumes a future that hasn't arrived.

That skepticism extends to the picks-and-shovels trade as well. Memory chips, cooling systems, and power suppliers feeding the data center boom do generate real earnings today. But Casey calls the whole setup a daisy chain. If the data center buildout gets recognized as a massive misallocation of capital, he expects the suppliers to get pulled down with it.

Why Energy Still Looks Cheap

The first place he'd put money has nothing to do with AI: Old-fashioned energy—and not just oil and gas. He's also positioned in uranium and coal, which he considers the unglamorous fuel sources that will keep the lights on regardless of what happens to the AI trade.

Energy stocks made up about 20% of the S&P 500 back in 1980. Today, that figure has shrunk to roughly 4%, even as oil and gas remain just as critical to the global economy. With West Texas Intermediate crude trading around $70 a barrel, Casey sees a sector the market has simply stopped paying attention to.

He's looking outside the U.S. for the best entry points. He favors Ecopetrol NYSE: EC, Colombia's national oil company, and Petroleo Brasileiro S.A. - Petrobras NYSE: PBR, Brazil's equivalent, both of which offer high single-digit dividend yields.

He also likes Meren Energy TSE: MER, a smaller offshore African oil producer with a roughly $1 billion market cap and a similar payout, plus unexplored concessions he believes give it real upside beyond current oil prices. For investors wary of emerging-market exposure, he notes Alberta, Canada, is home to small oil and gas names yielding 5% to 7%.

On power, Casey is unambiguous. Nuclear, in his view, is the safest, cheapest, and cleanest form of mass power generation, and coal works in the near term as well. His core argument: even if the AI trade collapses, the demand for electricity that AI created isn't going away.

Nuclear stocks were the hottest trade in the market as recently as late 2024. The fact that nearly every name in the sector has since sold off is precisely what makes the entry point interesting to him.

The Case for Small-Cap Gold Miners

Mining is an industry he calls a terrible business, but one he's owned stocks in for most of his investing life. With gold trading near $4,000 an ounce, he isn't buying the metal itself as a speculation. What he sees as undervalued are the companies that mine it.

The math is what excites him. Industry-wide, the all-in sustaining cost of producing an ounce of gold runs around $1,700. With gold prices roughly double that figure, miners are generating real margin for the first time in years, yet mining stocks represent only about 2% of the S&P 500. Casey expects that gap to close and sees potential for tenfold returns across the sector, with some smaller names capable of going much further.

These are mostly nanocap companies, often run by founding entrepreneurs, and they're prone to volatility, fraud, and outright failure—Casey references Mark Twain's famous line about a gold mine being a hole in the ground with a liar at the entrance.

He won't name specific stocks publicly, given how thinly traded they are. What he will share is his screening framework: a set of nine criteria he calls the Nine Ps, covering factors like management track record, geological quality, access to capital, and jurisdictional stability. His point is that volatility and risk aren't the same thing, and at current prices, he believes the odds tilt toward investors who do their homework.

Corn, Soybeans, and a Fertilizer Shortage

The third area doesn't involve stocks at all. Agricultural commodities—specifically corn, soybeans, wheat, and rice—supply roughly 60% of the calories consumed worldwide, and right now, prices for all of them are sitting at or below breakeven for farmers. A cyclical commodity bull market, in his view, is setting up from those depressed levels.

A looming fertilizer shortage adds urgency to his case. Disruptions in the Strait of Hormuz have cut off significant flows of sulfur and urea, both byproducts of natural gas and critical inputs for crop production.

He expects food prices to rise over the next several years regardless of what happens to AI stocks.

For most investors, he recommends commodity ETFs over futures contracts. He specifically points to the Teucrium Corn Fund NYSEARCA: CORN, noting that similar vehicles exist for wheat and soybeans.

The reasoning circles back to his opening: dollars are losing value, bonds carry interest rate, credit, and currency risk all at once, and tech stocks are priced for a future that may not arrive on schedule. Raw materials, in his view, are where safety and upside happen to overlap right now.

The Contrarian Case

The AI story isn't going away—Casey freely acknowledges that. But he'd argue that's different from saying the stocks are worth owning at any price. Keep an eye on energy dividends and grain prices. Those are the signals he's watching.