U.S. stock market now in full run-up to Great Depression levels

U.S. equity valuations have risen above pre-1929 levels, with the S&P 500 Shiller CAPE at 39-41 in late May 2026, more than 2x its historical average of about 17 and only exceeded during the 2000 dot-com peak. The rally is concentrated in AI-linked mega-cap names, leaving the market vulnerable if AI spending disappoints or if higher rates, persistent inflation, private credit stress, and fiscal pressures push bond yields higher. While the article highlights significant downside risk, it also notes that strong earnings, AI productivity gains, and resilient balance sheets could still support equities.

Analysis

The key second-order risk is not “stocks are expensive,” but that passive and systematic flows have made valuation discipline less relevant right up until it matters. When a narrow group of mega-cap AI leaders dominates index returns, any disappointment in capex ROI, pricing power, or forward guidance can trigger simultaneous de-risking across index funds, vol-targeting, and retail momentum accounts. That creates a gap-risk setup: the downside is likely to arrive faster than fundamentals deteriorate, because positioning is crowded in the same names and the same factor exposures.

The market’s real vulnerability is a rates shock combined with earnings breadth weakness. If inflation re-accelerates or long-end yields back up, the discount-rate effect hits long-duration equities first, but the follow-through is broader: higher debt-service costs squeeze buybacks, private credit stress tightens lending, and softer hiring hits consumer-linked sectors with a lag of 1-3 quarters. In that scenario, defensives and cash-generative value can hold up better than the index, not because they are “cheap,” but because their equity duration is shorter and their financing sensitivity is lower.

The contrarian miss on this tape is that elevated valuations do not necessarily mean immediate downside; they often precede a long period of mediocre returns rather than a crash. That argues for a time-horizon distinction: near-term the market can grind higher on earnings and AI enthusiasm, but over 6-18 months the distribution of outcomes is skewed toward multiple compression if breadth does not improve. The catalyst to watch is not a recession headline—it is any sign that AI spending growth is slowing while yields remain sticky, which would undermine the narrative that justifies the premium.

Key Decisions for Investors

• Buy 3-6 month downside protection on QQQ via put spreads, targeting a 10-15% drawdown scenario; structure for limited premium outlay because the risk is a fast de-rating rather than a slow grind.
• Pair trade: long XLU or XLP vs short QQQ for a 2-4 month window if 10Y yields stay elevated; this expresses lower equity duration and better earnings visibility if multiple compression starts.
• Trim or hedge crowded AI-beta exposure via short calls against NVDA/MSFT or a basket hedge against SMH; the risk/reward favors reducing convex downside from any capex disappointment.
• Long TLT only if growth data weakens while inflation rolls over; otherwise avoid catching falling knives in duration-sensitive assets because the immediate driver is yield volatility, not recession probability.
• If the tape continues to narrow, rotate into equal-weight exposure via RSP over SPY for 6-12 months; this reduces concentration risk and benefits if leadership broadens beyond mega-cap tech.