Back to News
Market Impact: 0.35

America’s world turned upside down

Artificial IntelligenceTechnology & InnovationGeopolitics & WarTrade Policy & Supply ChainTax & TariffsSanctions & Export ControlsAntitrust & CompetitionElections & Domestic Politics

2025 was dominated by an AI investment boom—estimated to account for roughly 40% of U.S. economic growth via spending on data centres and infrastructure—paired with signs of speculative excess and heavy tech-sector borrowing that raise bubble risk. Geopolitical shifts under a second Trump term include sustained U.S. tariffs (effective average ~15%), Chinese retaliation with triple‑digit tariffs and export controls on critical minerals, and limited battlefield change in Ukraine despite heavy Russian casualties (est. ≥100,000 dead) while the EU committed €90 billion to support Ukraine for 1–2 more years of operations. The combined effect is accelerating de‑Americanisation of trade and supply chains, elevated policy and regulatory uncertainty (including antitrust and export‑control dynamics), and heightened political risk ahead of the November 2026 U.S. midterms—factors likely to favor defensive positioning and careful risk‑adjusted exposure to AI names and geopolitically sensitive sectors.

Analysis

Market structure: AI capex winners are hyperscalers and infra providers (NVIDIA, ASML, EQIX, DLR, LRCX) capturing pricing power for GPUs, fabs and colo with estimated 30–50% revenue upside in a scenario where AI capex stays elevated for 12–24 months. Losses concentrate in high-valuation, high-debt pure-play AI software platforms (SNOW, AI/C3.ai) whose earnings require sustained customer ROI to justify current multiples; if AI productivity gains disappoint, revenue growth could rerate by 30–60% over 6–18 months. Risk assessment: Tail risks include harsh antitrust/regulatory action (10–30% downside to mega-cap techs in 6–18 months), a sudden liquidity crunch from over-levered AI investment (contagion to credit markets, widening BBB spreads by 100–300bp), or a geopolitics shock (Russia/China escalation) that spikes oil +15% and safe-haven demand for USD/Gold. Hidden dependency: infra REITs and chipmakers rely on continuous capex; a 20% capex pullback would collapse utilization and REIT rents within two quarters. Trade implications: Immediate (days–weeks) hedge equity exposure with 3-month VIX call spreads and 1–2% allocation to long gold (GLD); short high-debt AI names (SNOW, AI) over 3–9 months while being long NVDA and EQIX for 6–18 months. Cross-asset: buy 6–12 month long-dated NVDA LEAPs (25% OTM) for asymmetric upside; overweight critical-minerals exposure (LIT) for 6–24 months as export controls and de‑Americanisation lift prices. Contrarian angles: Consensus assumes AI productivity is immediate; evidence suggests front-loaded infrastructure spending with delayed real GDP impact — price-in a 30–50% pullback in non-infra AI valuations over 6–12 months. Historical parallel: 1890s rail/1900s oil — overbuilt assets survive and consolidate; concentrate long exposure in durable infra owners (EQIX, ASML) and be ready to rotate into software later post-clearance of regulation or after demonstrated broad productivity gains.