Stock Markets Are Doing Something They've Only Done Once Since the 1870s. Should You Be Worried?

The S&P 500 has posted a third consecutive double-digit annual gain — +24.23% in 2023, +23.31% in 2024 and roughly +16.4% last year — driven by strong corporate earnings, AI enthusiasm and easier Fed policy, and a $10,000 investment three years ago would be worth about $18,600 today. However, the cyclically adjusted price-to-earnings (CAPE) ratio climbed to about 40 in 2025, its second-highest reading since records began in 1871 (the 1999 peak near 44 preceded the 2000 crash), signaling rich valuations and the need for selective, high-quality stock selection rather than broad market exuberance.

Analysis

Market structure: The rally is concentrated — AI infrastructure leaders (NVDA), cloud/datacenter vendors, and market operators (NDAQ) are primary beneficiaries as passive and thematic flows chase earnings and AI narratives. Broad indices look expensive (CAPE ≈40 vs 44 in 2000), so marginal demand is shifting from breadth to a handful of mega‑caps; small caps, cyclical industrials, and rate‑sensitive financials are the likely losers as money rotates into quality growth.

Risk assessment: Key tail risks are a Fed pause/no‑cut surprise or sticky CPI that keeps real rates >2% and forces multiple compression (historical drawdowns of 20–40% in high‑CAPE regimes); regulatory action on AI or a NVDA supply/guide miss could produce 30–50% idiosyncratic shocks. In days–weeks, positioning risk around Fed/CPI prints and January earnings is dominant; over months–years, concentration risk and buyback/ETF flows (second‑order liquidity effects) determine realized returns.

Trade implications: Favor idiosyncratic long exposure to AI infra on disciplined entries (buy on >=8–12% pullbacks), pair with broad‑market hedges rather than naked long SPY. Implement 3–6 month SPY put spreads sized to cost 0.5–1% portfolio risk, sell covered calls to monetize low IV on large caps, and reduce small‑cap/cyclical weight by ~50% into high‑quality software and exchange operators (NDAQ).

Contrarian angles: The consensus CAPE‑doom view understates the role of lower real rates and secular revenue growth from AI — multiples can expand further if Fed cuts materialize; conversely, crowding into a few names makes index risk largely idiosyncratic. History (1999 vs today) differs on earnings validity and cashflows, so avoid binary “CAPE = crash” plays and prefer hedged, size‑constrained exposure.

Key Decisions for Investors

• Establish a 1–2% initial long position in NVDA on any pullback of >=8% from current levels; add up to 2% more on outperformance at next quarterly print. Use a stop loss or hedge if NVDA falls 25% from entry.
• Trim S&P exposure (SPY) by ~25% of current equity sleeve within 30 days and redeploy proceeds: 50% to high‑quality AI/infra names (e.g., NVDA, enterprise software) and 50% to defensive growth/exchange operator exposure (NDAQ 1–2% position) to capture market structure fees.
• Buy a 3–6 month SPY 5%–10% OTM put spread sized so maximum loss equals 0.5–1% of portfolio to protect against a 15–30% CAPE‑driven drawdown or policy shock around next CPI/Fed cycle.
• Implement a pair trade: long NVDA (1–2%) vs short SPY (dollar‑neutral) to express idiosyncratic AI upside while hedging beta; reassess after 3 months or if NVDA out/underperforms by 20%.
• Reduce Russell 2000/IWM exposure by ~50% within two weeks and reallocate into SaaS/recurring‑revenue names and exchange operators, avoiding cyclical commodity plays until real rates fall below ~1.5%.