Cathie Wood says AI already boosting labor productivity By Investing.com

Cathie Wood says non‑farm productivity is currently +2.8% y/y and could accelerate to as much as 6% annually as new AI tools scale. She cites Anthropic’s annualized revenue at $19B and OpenAI expanding from ~$20B to ~$25B, and forecasts $10–12 trillion in AI-driven revenue over the next 5–10 years; Nvidia’s Jensen Huang projects ≥$1 trillion in sales through 2027. Headline risk: oil is trading above $100/barrel on Iran supply fears, posing a near-term inflationary/geopolitical downside even as the AI narrative supports a bullish outlook for tech equities.

Analysis

The immediate market tilt toward “AI winners” is correctly focused on compute-layer capture, but the larger, less-discussed leverage is in the adjacent supply chain: power delivery (PSUs, VRMs), high-end DRAM/HBM, advanced PCBs and colo real-estate. Those nodes create sticky bottlenecks — a supplier with 6–12 month lead times can extract margin and create multi-quarter outperformance even if end-demand normalizes. Conversely, software-first vendors that monetize AI through user engagement face much shorter lead-times and therefore faster mean reversion if advertising or CPI-driven consumer weakness reappears.

Regulatory and competitive tail risks dominate the path from enthusiasm to realized profit. Export controls, EU/US regulation on model use, or large hyperscalers pushing bespoke silicon can compress incumbent pricing power over 6–24 months; inventory digestion cycles can flip near-term EBIT momentum inside 1–3 quarters. Near-term catalysts are earnings and guidance cadence from key infra buyers (quarterly), while true secular payoff for platform players is a 3–5 year story tied to enterprise replatforming and hybrid-cloud capex rhythms.

Positioning should reflect asymmetric payoff: concentrated long-dated optionality on dominant accelerators, tactical levered exposure to system integrators during visible rack orders, and selective compression trades on high-multiple software/advertising plays that look priced for flawless execution. Market crowding in the largest names raises volatility risk — use defined-risk option structures to participate in upside without line-item tail exposure. Finally, monitor downstream KPIs (rack shipments, colo utilization, HBM pricing) as higher-fidelity leading indicators of revenue realization rather than headline AI narratives.

Key Decisions for Investors

• Long NVDA via 24–36 month call spread (allocate 3% NAV): buy 2028 Jan LEAP calls and sell a higher strike to fund ~50–60% of premium. R/r: asymmetric upside (2x–4x if NVDA retains share and pricing) with max loss = premium; trim on 40–60% paper gain or if guidance weakens.
• Tactical long SMCI directional for 3–6 months (allocate 1–1.5% NAV): buy near-term calls to capture rack/order catalysts, take profits at +40% and stop at -25%. Rationale: more levered exposure to system refresh windows; risk of component shortages and order cancellations is highest in next quarter.
• Pair trade: long NVDA (notional) / short APP (notional ~50% of NVDA size) for 9–12 months to express hardware-led AI outperformance vs ad/monetization IP. R/r: reduces macro-ad cyclicality; target a 2:1 upside skew where NVDA +50% / APP -20% produces >=2.5x return on net capital.
• Income/volatility play on APP: sell 30–45 day covered calls or cash-secured puts around upcoming guidance releases to harvest elevated IV (allocate excess cash notional 0.5–1% NAV). R/r: collect premium (target annualized yield 15–30%); risk is assignment if stock gaps up on positive surprise.