McDonald’s Crash Thesis: Concentration, Not All Equities

Larry McDonald’s central argument is that a technology-heavy S&P 500 conceals substantial damage elsewhere—and exposes passive investors to an increasingly narrow AI bet. He predicts a 30% index decline over the following year, followed by potentially eight flat years. Yet his position is selectively bullish: battered brands, long-duration bonds and hard-asset businesses could outperform. The distinction matters: his thesis is rotation, not abandoning every stock.

• Passive ownership can disguise concentrated risk. McDonald characterizes roughly half the S&P as technology/AI exposure and argues that continuing passive inflows reinforce the largest holdings. He also portrays prospective SpaceX, Anthropic and OpenAI IPOs as an eventual transfer of expensive private holdings into retirement portfolios through index inclusion. These are his valuation and future-inclusion claims, not established outcomes. His proposed alternatives include equal weighting or an advisor-built portfolio with more hard-asset exposure.
• The consumer squeeze could become the recession catalyst. Higher energy prices and interest rates have already delivered what he estimates is a $300 billion consumer shock. Large declines in consumer-facing brands are, to him, evidence that much of the market has already “crashed.” Further yield increases could worsen banking and commercial-property refinancing stress before ultimately forcing a reversal. He explicitly acknowledges uncertainty over whether an AI collapse would drag otherwise attractive stocks lower.
• Bonds offer a conditional, tactical opportunity. He sees depressed long-duration prices as potentially offering equity-like returns if recession pushes yields down. His illustration is an Apple bond reportedly trading near 47 cents on the dollar, potentially recovering to 72–73 if the 30-year Treasury yield falls to 4%. Importantly, he calls this a bear-market rally, not a durable bond bull market, and allows for yields rising first. Debt-ceiling-related reductions in net Treasury issuance are another possible catalyst, not a certainty.
• Forced selling can create stock-specific opportunities. Nike illustrates his search for capitulation in established brands. He argues that fund reporting pressures and individual tax-loss selling can temporarily overwhelm fundamentals, setting up a January rebound. His outlook for beaten-down stocks and bonds runs into the following June–July; the large returns he discusses are speculative scenarios, not promises.
• AI users with proprietary data may beat AI spenders. He criticizes technology companies for diverting profitable businesses’ cash into costly, potentially commoditized AI products with uncertain payoffs. Slowing capital expenditure is his proposed trigger for a mega-cap reset. Conversely, he highlights Baxter, Intuitive Surgical and, in a later passage, Schlumberger as businesses whose specialized data could help monetize AI.
• Hard-asset equities are his longer-term preference. Persistent inflation and higher discount rates, he argues, diminish distant software cash flows while supporting resource ownership. He favors energy, materials and industrials, including cash-generative natural-gas producers. The counterargument is real: slower AI construction would weaken commodity demand. His response is that low relative allocations and valuations could still support a broader rotation.
• Gold and stablecoins fit his debt-financing thesis. He links central-bank gold diversification to sanctions risk. Separately, he predicts stablecoins could reach $1–1.5 trillion within five years, increasing Treasury-bill demand and helping suppress financing costs. This is a proposed financial-repression mechanism, not enacted compulsory adoption. The host notes that stablecoins could strengthen dollar demand; he agrees reserve-currency loss is unlikely within his or his children’s lifetimes.

Caveats: These are McDonald’s opinions, estimates and conditional forecasts. Apple bond and Nike metrics were not visually verified; several source names and figures are unresolved or inconsistent. All forecast horizons are relative to the interview, not today.

Full transcript