If a Stock Market Crash Is Coming, Warren Buffett Says Investors Should Make This Important Move

The S&P 500 and Nasdaq Composite have risen roughly 80% and 91%, respectively, over five years, while market caution is growing over AI-driven valuations and the durability of the investment boom. Capital Economics expects the AI bubble to begin bursting in 2027; global AI-related capital spending is projected to reach $1 trillion in 2026, and the CAPE ratio is 41.3 versus its 17.4 historical average. The article recommends Buffett-style long-term investing in fairly valued businesses rather than trying to time a crash, noting Berkshire Hathaway held $365.5 billion in liquidity.

Analysis

The key risk is not simply that AI expectations disappoint; it is that supplier earnings depend on customers continuing to convert large capex budgets into workloads and monetizable returns. If that conversion slows, hyperscalers can defer orders before end-user AI demand visibly weakens, transmitting the shock quickly to accelerator and memory suppliers. A second-order spillover would reach networking, foundry, and data-center power suppliers; the initial equity repricing could therefore precede evidence of weaker AI adoption.

For the next 1–3 months, elevated market sensitivity to rates and capex guidance matters more than a forecasted bubble date. Watch hyperscaler capex plans, order lead times, and evidence that AI revenue is keeping pace with infrastructure spending. Over 6–18 months, the differentiator is likely to be returns on deployed capital, not AI exposure alone. A broad market decline is not assured: durable demand and falling yields could sustain earnings and valuations despite high aggregate valuation measures.

Berkshire Hathaway’s liquidity provides optionality, not a guaranteed near-term catalyst; cash yields can support earnings while also making large acquisitions less urgent. The contrarian risk is overgeneralizing from the dot-com analogy: today’s AI suppliers may have real earnings, so indiscriminate shorting can be costly. Without current valuation, positioning, and options data, avoid treating the article’s historical comparisons as a timing signal.

Key Decisions for Investors

• Keep NVDA and MU exposure sized to tolerate a capex-driven drawdown; do not add solely on secular AI demand. Reassess after major hyperscaler capex updates and supplier order commentary.
• Consider a modest relative-risk reduction in AI hardware versus BRK.A for investors seeking less direct capex sensitivity, while recognizing Berkshire is a diversified equity, not a pure hedge. Stage any shift rather than making a binary market-timing call.
• Alert: evaluate put spreads on NVDA or MU only if implied volatility and spread cost are acceptable. Missing inputs are current option pricing, portfolio exposure, and liquidity; do not initiate based on the article alone.
• Falsify the cautious thesis if hyperscalers maintain or raise spending while demonstrating improving AI monetization, and NVDA/MU order or earnings guidance remains resilient. Escalate caution if capex plans are cut, supplier lead times or orders weaken, or yields rise materially.