Barclays favors U.S. equities as energy shock meets AI-driven earnings

Barclays expects U.S. corporate earnings to rise about 15% in 2026 and recommends staying overweight U.S. equities, citing AI-driven capex and margin support despite the S&P 500 being down ~3% year-to-date. The firm says energy-price spikes from the U.S.–Iran conflict have hurt stocks and bonds but that worst-case scenarios are unlikely, and Big Tech forward multiples have fallen to their lowest levels since early 2025. Barclays highlights a policy asymmetry—Fed can look through energy-driven inflation more than the ECB—reinforcing a U.S./Europe equity gap and advising investors to lean into structural winners while hedging near-term macro and policy risks.

Analysis

The policy divergence between central banks will act less like a single macro shock and more like a directional flow generator: expect FX and cross-border equity flows to amplify a U.S. equity bid while compressing European risk premia. That creates a persistent bid for USD assets and for capital-goods names exposed to AI-driven corporate reinvestment — think multi-quarter revenue visibility for equipment suppliers rather than the narrow handful of AI software winners.

Second-order supply-chain winners are in power delivery and industrial electrification: increased data-center buildouts lift demand for high-voltage transformers, UPS systems and specialty copper — suppliers with limited incremental capacity can enjoy multi-year pricing power. Conversely, exporters/EM importers with large energy import bills and leveraged balance sheets will see cashflow strain sooner than headline GDP moves suggest, pressuring local banks and credit spreads within 3–9 months.

Tail risk is a policy feedback loop: if energy inflation persists beyond two quarters, the Fed’s tolerance window narrows and the implied equity risk premium re-rates higher quickly. Geopolitical escalation that disrupts maritime flows or LNG shipments would be a near-term shock; a slower, structural reversal would come from AI capex plateauing if enterprise adoption stalls or chip supply materially outpaces demand — that’s a 12–24 month risk.

The cheapest behavioural inefficiency now is conviction concentration: investors price AI as a handful-of-names phenomenon, understating breadth in component suppliers and infrastructure beneficiaries. Positioning should therefore favor exposure to the capex chain and policy-armed U.S. growth while using time-limited, cheap hedges against energy or policy re-pricing events.

Key Decisions for Investors

• Long Lam Research (LRCX) 6–12 month horizon — reason: exposure to multi-year wafer fab transitions and less idiosyncratic concentration than fabless chips. Position size: 2–3% NAV. Risk: 25–30% downside if AI spend disappoints; reward: 30–60% upside if capex normalizes.
• Pair trade: Long Digital Realty (DLR) / Short Vanguard FTSE Europe ETF (VGK) for 3–9 months — play infrastructure demand + policy differential. Target relative outperformance 4–8%. Use 50/50 notional to keep beta neutral; stop-loss if EURUSD moves >4% against USD in 2 weeks.
• Energy hedged directional: Buy 3-month XLE call spread (bull call spread) sized to 1% NAV with simultaneous short 1-month Brent call to finance part of the premium. This limits cost while capturing energy spikes from geopolitical blips. Risk: capped loss = premium; reward: 3–5x premium if a short-term supply shock hits.
• Credit/EM guard: Buy protection via short-dated CDS or overweight USD IG cash and underweight EM sovereigns for 3–6 months to insulate vs import-driven sovereign stress. Expect 1–2% drag on carry but materially reduces tail-loss from FX-devaluation cycles.