Back to News
Market Impact: 0.35

Will the Market Keep Going Up in 2026? This Is What History Says.

AMZNNVDAMETAGOOGLNFLXNDAQ
Artificial IntelligenceInterest Rates & YieldsMonetary PolicyInflationTechnology & InnovationMarket Technicals & FlowsInvestor Sentiment & PositioningCorporate Earnings
Will the Market Keep Going Up in 2026? This Is What History Says.

The S&P 500 completed a three-year rally that delivered a 78% gain through 2025, driven largely by generative AI enthusiasm and big-cap winners (Amazon +175%, Nvidia +1,200% from 2022 lows). Falling interest rates and continued consumer spending underpin the rally, but valuation metrics are rich — the S&P 500 CAPE is up ~40% to about 39, a level last seen in 2000 — prompting warnings of elevated bubble risk. The piece advises portfolios remain diversified with defensive holdings and cash reserves while retaining exposure to high-quality AI-linked growth names given their profitability and cash buffers.

Analysis

Market structure: Gains are concentrated in AI hyperscalers and their supply chain — NVDA (+1,200% since 2022) is the primary hardware bottleneck, while AMZN, GOOGL, and META capture cloud and AI spend. That concentration inflates pricing power and narrows breadth (S&P up 78% over three years; CAPE ~39), elevating systemic risk from idiosyncratic shocks to a few names. Cross-asset effects: falling rates support duration instruments and equity multiples, compressing equity vol and strengthening risk-on flows into tech while pressuring USD and boosting EM FX/cyclicals unevenly. Risk assessment: Key tail risks are (1) swift regulatory action (antitrust, AI export controls) within 6–18 months, (2) NVDA supply disruption or price correction that removes GPU runway, and (3) a Fed surprise back to tightening if CPI re-accelerates (>0.3% monthly). Near-term (days–weeks) expect vol compression around earnings; medium (3–6 months) earnings/monetization disappointment could trigger 15–30% drawdowns in crowded names; long-term (1–3 years) depends on durable AI margin capture vs. competitive erosion. Hidden dependencies include hyperscalers’ reliance on third-party datacenter capacity and GPU supply chains, creating nonlinear downside if either breaks. Trade implications: Favor concentrated, hedged exposure to AI leaders while adding defensive duration and option protection. Implement size limits (single-name exposure ≤2% of portfolio) and buy tail hedges (3-month SPX or QQQ put spreads costing ~0.3–0.8% of portfolio). Use covered-call overlays to monetize low IV on NVDA/AMZN and use pair trades to peel valuation risk off momentum names into cyclical/value exposures over 6–12 months. Contrarian angles: Consensus underestimates regulatory and monetization lag — high CAPE implies mean reversion risk that history (2000) shows can take 2–3 years to play out even with intermittent rebounds. The market may be overpricing perpetual AI profits; look for mispricings in non-AI cyclicals, financials (NDAQ exposure to fee growth), and industrials where earnings are discounted. Unintended consequence: crowding increases liquidity fragility — a 20–30% shock to NVDA/AMZN could cascade into broader tech multiple compression.