AI threatens to widen inequality among states: UN

The UNDP Asia-Pacific report “The Next Great Divergence” warns that AI could widen inequality between developed and developing states, reversing decades of convergence and creating security and migration spillovers if unaddressed. While the region could see AI raise annual GDP growth by about 2 percentage points and ASEAN economies could gain nearly $1 trillion over the next decade, early gains are concentrated in advanced economies (Singapore, Japan, China) and millions of jobs—especially for women and youth—face high automation exposure, prompting a call for urgent, coordinated policy action.

Analysis

Market structure: Winners are large cloud and AI-stack providers (AMZN, MSFT, GOOGL) and GPU/AI chipmakers (NVDA) that capture outsized pricing power for compute; expect gross-margin tailwinds of +200–500 bps for top cloud providers if AI services scale over 12–24 months. Losers are low-connectivity emerging markets, labour-intensive service exporters and small EM banks whose NIMs and fee pools face compression as automation substitutes low-skilled labor. Compute scarcity (GPUs/GPUs-in-data-centers) will keep short-term pricing tight — expect enterprise AI spend growth of 30–50% YoY in 2025 vs. base IT spend growth of ~5–8%.

Risk assessment: Tail risks include rapid regulatory fragmentation (data localization/export controls) that could cut accessible TAM for US cloud/AI by ~10–25% within 12–24 months and geopolitical divestment that fractures supply chains (NVDA revenue shock scenario of -15–30% in a hard split). Immediate market moves will be driven by headlines (days–weeks); medium-term (3–12 months) by policy and capital investment cycles; long-term (3–10 years) by education/infrastructure investment rates in EMs. Hidden dependencies: grid reliability, tertiary education capacity, and telecom capex; catalyst watchlist: export-control decisions and major public AI procurement/funding in next 30–180 days.

Trade implications: Direct plays: overweight NVDA and AMZN for 6–18 months to capture compute and cloud AI monetization; hedge regulatory tail via long-dated out-of-the-money put protection sized to 10–20% of position. Pair trade: long AMZN (cloud AI) vs short INFY (Indian IT services exposed to automation) for 6–12 months as pricing power shifts to platforms. Options: buy 6–9 month call spreads on NVDA (caps cost) and buy 3–6 month straddles around major AWS/GCP product announcements on AMZN/GOOGL. Rotate from EM consumer/outsourcing to AI-exposed large caps and Asian digital infrastructure (Singapore, Korea ETFs).

Contrarian angles: Consensus assumes permanent divergence; history (Industrial Revolution) shows catch-up over decades if EMs get targeted capex/education — this implies select EM infra investments (edge compute, local cloud) are underpriced. Market may be overpricing GPU scarcity into small-cap AI software winners while underpricing regulatory fragmentation risk for hyperscalers. Unintended consequence: protectionism could create local monopolies in EMs — overweight regional telcos/edge providers if local cloud mandates appear; monitor policy signals closely over next 90 days.

Key Decisions for Investors

• Establish a 2–3% portfolio long position in NVDA over the next 2–4 weeks via 6–9 month call spreads (buy 60–70 delta calls, sell 90–120 delta calls) sized to target 40–100% upside; cap cost and set a trim at +60% realized gain.
• Add a 2% long position in AMZN for AWS/AI exposure, funded by selling 1–2% covered calls 3 months out at ~+10% strikes; increase to 4% if AWS announces major native AI customer wins in next 90 days.
• Implement a pair trade: long AMZN (2%) / short INFY (1.5%) for 6–12 months to express platform monetization vs. outsourcing margin erosion; rebalance if INFY underperforms by >15% or AMZN outperforms by >25%.
• Reduce EM sovereign bond exposure by 1–3% and buy 6–12 month protection: purchase puts on EMB-sized ETF or equivalent CDS protection if EM sovereign spreads widen >50 bps or EM FX depreciates >5% vs USD within 60 days.
• Overweight Asian digital-infrastructure equities: initiate 1–2% positions in Singapore ETF (EWS) and South Korea ETF (EWY) to play regional AI adoption; review after 6 months or sooner if regional capex announcements exceed $10bn aggregate.