UBS: AI investment is ‘lone buffer’ for emerging markets as energy costs soar

Emerging markets have moved sharply into risk-off mode amid the third week of Middle East hostilities, with UBS noting EM is now one of the worst‑performing asset classes since the conflict began. The primary risk is energy: historical EM drawdowns occur when oil rises above ~$90/bbl, removing valuation support and reversing recent EM gains. AI exposure (the largest driver of EM returns and earnings upgrades over the past 15 months) is the main potential buffer if U.S. hyperscaler CAPEX remains immune to the conflict.

Analysis

The immediate market reaction is largely liquidity- and positioning-driven rather than a clean rerating of secular EM growth; forced ETF outflows and cross-asset de-risking will amplify moves in illiquid mid/small-cap EM names and local-currency credit, creating dispersion you can trade. Expect an acute window (days–weeks) of outsized price moves as managers de-lever, followed by a medium-term (1–6 month) selection phase where earnings resilience — not headline beta — determines winners.

AI-related capex acts as a structural partial floor for a narrow subset of Asia-exposed exporters: foundry, advanced packaging and capital equipment vendors with direct hyperscaler revenue will see demand visibility survive geopolitical noise, while consumer-facing, energy-intensive sectors will re-price margins and capex plans faster. A second-order logistic effect that is underappreciated is insurance and rerouting premia on shipping; longer transit times will raise working capital needs for manufacturers with tight inventory turns, pressuring EM SMEs disproportionately.

Tail scenarios are binary and time-sensitive: a rapid de-escalation and visible hyperscaler capex confirmations should produce sharp mean reversion (weeks), whereas maritime disruptions or sanctions that raise input-cost inflation will force a longer multiple reset (quarters). Tactical opportunities lie in short-term volatility structures and concentrated pairs that separate AI-capex beneficiaries from broad commodity/currency-exposed EM beta.

Key Decisions for Investors

• Pair trade (1–3 months): Short broad EM equity exposure via EEM (size 1–2% NAV) and go long NVDA (size 0.5–1% NAV) to isolate AI-capex resilience vs EM beta. Use EEM 3-month 1x put spread to limit cost (buy 1x 10% OTM put / sell 1x 20% OTM put) and size NVDA as outright equity exposure or buy a 3-month call to define risk. R/R: asymmetric — limited downside on hedged EEM vs uncovered upside if AI narrative holds; stop-loss on pair if NVDA/EEM spread reverses >20% intraday.
• Selective EM semiconductor exposure (6–12 months): Long TSM (add 0.75–1.5% NAV) with a protective 6–9 month collar (buy 1x put ~15% OTM, finance with covered calls ~10% OTM) to capture durable fab demand while capping geopolitical drawdown. R/R: captures multi-quarter revenue visibility from hyperscalers; hedge cost limits max loss to ~10–12%.
• Tactical commodity/energy hedge (1–6 months): Buy OTM call spreads on US E&P names (e.g., PXD or DVN) or an oil-call spread (WTI 3-month 1x call spread) sized to offset projected EM profit compression risk. R/R: protects portfolio margin risk if energy premia widen; cap premium outlay with spreads and trim at first signs of diplomatic progress.
• Credit/volatility hedge (1–3 months): Purchase EMB 3-month puts or pay up for a short-dated EM sovereign CDS proxy to protect against funding stress in carry trades and FX. R/R: relatively low cost insurance that pays if flows reverse sharply; unwind on stable primary issuance or visible central bank FX intervention.