How Much Would an AI Crash Destroy?

Summary

Key Points

• Micron and SK Hynix generated about 17% of the global stock market’s May return while representing only 1% of the MSCI All Country World Index, highlighting the outsized impact of a few AI‑related chip stocks. 0:00
• Diversification should involve assets that do not move together, not just accumulating many growth stocks, to protect against concentration risk. 2:35
• AI spending has spread beyond tech giants to utilities, real estate, construction and everyday services, tying many sectors to AI‑related capital. 4:04
• Wealth created by AI companies is quickly flowing into the broader economy through luxury purchases, housing demand and services, making many households indirectly exposed to AI. 5:32
• If AI valuations revert to long‑run averages, the market could lose $20‑33 trillion, dwarfing the dot‑com crash, and a massive equity correction would directly affect middle‑class spending. 12:02
• Off‑balance‑sheet AI‑related capital spending is estimated at $3 trillion, and private‑credit funds have lent $2‑3 trillion to AI and software firms, raising default risk in a downturn. 20:33
• Historical parallels such as railway mania and the dot‑com bubble show that even successful technologies can leave investors who overpay with prolonged losses. 23:00
• European equities are currently cheap, offering lower volatility and dividend yields around 3%, providing a defensive ballast if AI expectations falter. 31:58
• Value investors see upside in Europe as companies cut costs and improve, contrary to the stereotype of stagnant economies. 32:36
• Allocating a modest portion of portfolios to stable, dividend‑paying sectors like pharmaceuticals, utilities, banks and industrials can reduce overall risk while maintaining exposure to US markets. 34:43