Billionaire Bill Ackman Has 25% of His Hedge Fund in 2 Brilliant AI Stocks (Hint: Not Nvidia)

Pershing Square held 25% of its portfolio in two AI-related names as of December—14% in Amazon and 11% in Meta—reflecting high conviction in AI-driven monetization. Amazon's thesis cites AI-driven retail efficiency and AWS momentum (cloud revenue +24% in Q4; operating margin +1.5 percentage points ex-items); the stock trades at ~28x earnings with a $285 median target implying ~37% upside from $208. Meta's thesis centers on adtech leadership and custom AI chips (Street EPS CAGR ~22% next three years); it trades at ~26x earnings with a $855 median target implying ~41% upside from $606.

Analysis

The most important second-order beneficiary of accelerated AI adoption will be the operational stack that links cloud models to physical execution — fleet orchestration, edge inference software, and integrators that retrofit existing fulfillment networks. Those vendors (and the foundry/packaging capacity that serves their custom silicon) will see demand profiles that are lumpy but sticky: one successful pilot can convert into multi-year contracts with >30%+ gross margins and high switching costs, creating durable oligopolies beneath the marquee cloud and ad platforms.

Near-term catalysts are concentrated and fast-moving: quarterly ad metrics and Cloud product launch cadence will re-price multiples within 1–3 quarters, while capital intensity and chip supply dynamics play out over 12–36 months. Tail risks include regulatory intervention on targeted advertising or data flows, an inflection in advertiser ROI that compresses CPMs by mid-single digits, and a bout of margin disappointment if AI-driven automation takes longer to convert into per-order cost savings than investors assume.

Consensus is underestimating optionality and over-indexing to headline capex: if AI-enabled yield improvements to advertising and fulfillment each deliver low-single-digit margin lifts, the re-rating could be >30% over 18–36 months; conversely, if model inferencing costs rise faster than monetization, downside is similar in magnitude. Watch three signposts to arbitrate: sequential ad price-per-impression and ARPU trendlines, multi-quarter growth in custom-chip revenue and utilization, and incremental gross margin per unit of fulfillment driven by automation pilots.

Key Decisions for Investors

• Buy AMZN Jan-2028 call spread (buy 1x ATM LEAP, sell 1x +30% strike) — 18–30 month horizon. Rationale: defined-cost exposure to multi-year margin optionality without funding open-ended premium; target 35–45% gross return if automation and Cloud monetization accelerate. Risk: total premium loss if adoption lags — size to 1–1.5% NAV.
• Buy META Jan-2028 LEAPs or buy equity with 6‑month protective puts — 12–36 month horizon. Rationale: asymmetric payoff from ad yield improvements and AR hardware optionality; target ~40% upside vs current levels. Risk: cyclical ad slowdown or regulatory shocks could compress multiples; cap position at 1–2% NAV and hedge with short-dated puts sized to limit drawdown to ~15%.
• Long NVDA 6–12 month calls (near- to mid-term) — tactical 6–12 month play. Rationale: capture continued GPU demand for model training and inference; upside >50% if data-center orders reaccelerate. Risk: product cycle inventory or architecture shift; keep position small (0.5–1% NAV) and take profits on 30–40% intraday advances.
• Pair trade: long AMZN / short INTC equal-dollar 12–24 months. Rationale: isolate AI software & services capture vs incumbent silicon risk — expresses convective flow of value up the stack. Risk/reward: aim for net 20–30% tail return if AMZN captures ad + automation economics while INTC fails to regain competitive silicon share; mark-to-market volatility likely, size at 1–2% NAV net exposure and set a stop at 12–15% adverse move.