Larry McDonald — SpaceX Could Crash the Market; Buying Commodities Instead

Agnico Eagle is a big, top-quality gold miner he calls the best-managed in the world. It's down 40% — not because the business broke but because "tourist" money that piled into gold last year got flushed out when the Iran war pushed up diesel costs and rate-hike fears, and cash-strapped emerging-market central banks dumped gold. The company still produces $6–7 billion a year of free cash (money left after running the mines) and is buying back $2 billion of its own stock.

That's the exact setup hedge-fund manager David Einhorn taught him to hunt: cheap, big free cash flow, real buyback, washed-out sentiment — while the broad market sits at highs. With gold he thinks heads to $6,500 in ~18 months, he calls it "a beautiful risk-reward situation."

First Quantum is a copper miner whose flagship is a Panama mine he calls "a dream" — the copper sits near the surface (cheap to dig) and is five minutes from the ocean (cheap to ship) — yet it was shut down on environmental rules. That's his whole copper thesis in one mine: the best deposits on Earth are being suppressed by regulation just as demand explodes (100 million robots, rebuilding war zones, a $2 trillion US grid rebuild, thousands of data centers). Suppressed supply + soaring demand = own the copper miners.

BHP is the world's largest diversified miner (copper, iron ore and more). "You should own the BHPs, the Rio Tintos" — he expects these hard-asset giants to take a far bigger slice of the S&P 500 over the next 5–10 years as the index migrates away from being half technology. A cheap, asset-rich way to own the copper and metals the AI/grid buildout consumes.

Rio Tinto is the other giant diversified miner he names alongside BHP — same idea: a core, asset-heavy copper/metals holding that benefits from the supply-suppressed, demand-rich commodity regime and from money rotating out of crowded tech into hard assets.

Alcoa is a big aluminum producer. He's owned it three years, trimmed a little, but stays long because "aluminum is going to be a bedrock of the data centers and of the power grid rebuild." The US grid is like a 75-year-old man being asked to lift weights — it needs a ~$2 trillion rebuild, and that takes enormous amounts of aluminum.

Newmont is the only gold miner left in the entire S&P 500 — a fact he finds crazy given the gold bull case. He groups it with Agnico and Barrick as the "quality" miners: they generate cash, buy back stock and don't dilute shareholders, the opposite of the sketchy junior miners that constantly issue new shares.

Barrick is one of the world's largest gold miners. He names it with Agnico and Newmont as a quality operator that returns cash to shareholders instead of diluting them — his answer to a viewer worried about insider selling and dilution across the gold-mining sector: stick to the quality names, avoid the juniors.

Tourmaline is Canada's largest natural-gas producer, sitting on "trapped" gas — cheap because there aren't enough pipelines to ship it. He owns it and is "waiting for it to do something." His catalyst: military strikes that wound Middle-East gas production hand "a huge fat check" to North-American producers, because buyers who got burned now want gas from safe jurisdictions. Layer on the AI angle — move data centers to the cheap stranded gas — and he calls it "the best AI play out there and a play on the war."

Antero is a big US natural-gas producer — "the Anteros in the United States" — the American leg of the same trade as Tourmaline: cheap gas that becomes more valuable as the Middle East is wounded and as data centers need power near the gas.

Range Resources is another US natural-gas name he lists with Antero. Same logic: low-cost American gas, strong cash flow, and a structural bid from the AI-power buildout plus a war that makes North-American supply safer than Middle-East supply.

Energy Transfer owns a vast US pipeline network — "the OG hard asset" that mainstream TV won't cover. It's in his core portfolio. The pitch: those pipeline "tentacles" are exactly what's needed to move gas to the data centers, so it's a quiet AI trade. At 12× earnings with a 7% dividend and strong free cash flow, he expects the next five years to repeat the last (roughly 7% a year in dividends plus ~30% price appreciation).

Schlumberger (now "SLB") is the world's biggest oil-services company — it does the drilling and well work for oil producers. It's in his core portfolio and up ~40% in 2026 (vs the Nasdaq +13%), barely scratched even though oil fell 30% from its April peak. Why so resilient? There's less than $3 trillion of oil-and-gas stock versus $41 trillion in the NASDAQ-100, so even a trickle of money rotating out of tech overwhelms it.

The kicker is AI: he calls SLB "the Google of oil services" because it sits on enormous offshore-drilling and ocean-floor reserve data that AI can monetize — "a sleepy, sexy AI play." Trading near its 200-week moving average signals, to him, the start of a multi-year bull market.

Intuitive Surgical makes the da Vinci surgical robots. It's down ~20% on a couple of missed quarters, and value investors won't call it cheap. But the insight from billionaire AI-medical family offices is that it owns the world's best surgical data — every robotic operation adds to a dataset nobody else has, and robots already let surgeons operate across borders. Feed AI that data over the next 10 years and it becomes hugely more profitable. He's buying the priceless data moat, not the current multiple — while everyone else crowds into chips.

This trust holds physical uranium — you own the metal, not a mining company. He'd "much rather own SRUUF" than the producers right now, having lightened Cameco and NexGen. Reasoning: a US national-security uranium reserve is coming and data centers need ~5 more gigawatts of nuclear power, while new mine supply keeps slipping. Since uranium has no real spot market, when contract buyers finally step up the price has to do the adjusting. He frames it as ~20–25% downside versus 200–300% upside. (Canadian investors can buy it as U.U / U.UN on the TSX.)

FCG is a basket of natural-gas equities — a prior pick that's up 22% since September, and he "still likes it a lot." His one-liner: it's "a very under-loved part of the S&P 500," and natural gas is the cheap, overlooked fuel for the AI-power buildout while everyone fixates on chips.

Core Natural Resources is a coal miner — his coal pick that worked, up ~50% (the broad coal ETF up ~30%). The theme is "power for AI": data centers need electricity now, and coal is one of the fastest sources to ramp. "Everyone's in the chips; no one's in the power." He still loves it.

Diageo owns Hall-of-Fame liquor brands (Johnnie Walker, Guinness, Smirnoff) and trades below its 200-week moving average — deeply washed out. The young-people-drink-less story is real and well-known, which makes it look like a "value trap" (like Nike). But he argues the bigger force is a quant game: funds go long high-momentum semiconductors and short low-momentum staples, mechanically suppressing names like Diageo. When the quarter-end rebalance and an inflation bounce flip that seesaw, the staples snap back.

General Mills is a packaged-food staple. He lumps it with Kraft Heinz, ConAgra and Campbell's as names where the bearish story (Ozempic, people eating less packaged food) is fully known and over-discounted. Consumer staples are "the most offsides we've ever seen" versus the S&P; they do well when the consumer weakens, and the June-30 quarter-end rebalance plus an inflation bounce should flip the seesaw back in their favor for six months.

ConAgra is another packaged-food maker in the same oversold-staples basket — momentum-suppressed and over-bearish, a mean-reversion candidate into the quarter-end unwind.

Kraft Heinz is the staple the viewer specifically asked about. His take: the reasons Wall Street hates it (Ozempic, packaged-food decline) are exactly the consensus capitulation he likes to buy against, with a rebalance and inflation tailwind ahead.

Campbell's rounds out the washed-out staples group — same view: the structural-decline narrative is priced in, and a momentum unwind plus consumer weakening favors it.

Vale is Brazil's giant iron-ore/metals miner. He loves commodity producers in a commodity boom, and Brazilian stocks are depressed because the market fears another term for the socialist president Lula. His pattern recognition: the media pumps up the establishment candidate before an election, then a market-friendly surprise (a Bolsonaro-style win, as happened with Milei in Argentina) sends stocks booming.

EIDO holds Indonesian stocks. His "capitulation model" scores how washed-out a market is, like rating a hurricane category one to five — and Indonesia just went through a category-five (an oil-price shock plus political risk), trading near a record low. That's "a screaming buy" for a one-to-two-year trade — the same pattern as Argentina before the Milei election, which he bought and made money on, only even more beaten down here.

He doesn't own Nvidia and thinks being long it is "the dumbest trade." His core objection: its "forward earnings are baloney" because they depend on a handful of big customers, not the millions of small ones Microsoft or Google have. At ~$4.5 trillion it would have to reach $8 trillion just to double — poor odds. Channeling Peter Lynch ("run from the hottest stock in the hottest sector"), he says you can't make money on a stock with 50 analyst buys; uranium offers far better risk/reward.

Cameco is the biggest Western uranium miner. He's owned it five years but has "lightened up," preferring the physical uranium trust. The reason: producers like Cameco chronically overpromise on how fast new production will come (the Elon-robotaxi analogy), and Canadian weather and environmental rules slow things further. Those misses are bullish for the uranium price — but a reason to be cautious on the miner's stock right now.

NexGen is developing a big Canadian uranium mine the world's supply forecasts lean on. He calls its Saskatchewan project "a mess" that won't come online on time — management embellishes the timeline. He's lightened his position. The delay is a supply problem that helps the uranium price, but it's a reason to be wary of the stock itself.

Capital One is "the biggest subprime lender," down about 25%. He uses it as the warning light for the "K-shaped" economy: behind the record-high bank indices, the wounded financials exposed to auto-loan defaults and credit-card delinquencies are telling you ordinary households are in trouble. It's evidence for his stagflation thesis, not a buy recommendation.

SpaceX came public at roughly $2 trillion — worth more than Amazon by day four — which he sees as the classic late-cycle hubris moment (like the AOL–Time Warner or RJR Nabisco deals). The danger isn't just the price: bankers project $1.3 trillion of revenue over five years, and if that's real it would have to come out of the same Mag-7 tech giants who've already committed $5 trillion to building data centers on Earth. With $800 billion of off-balance-sheet financing propping up those data centers, a sudden close of the "funding window" — people simply running out of money for the next raise — could, he says, "create a credit crisis."