Billionaire Bill Ackman Has 10.5% of His $19.5 Billion Portfolio in Amazon. What's the Thesis Behind Pershing Square's Bet?

Pershing Square still holds nearly 8.6M Amazon shares (~$2B, 10.5% of its portfolio) after trimming, signaling sustained conviction despite the capex ramp. The article cites Amazon’s planned capex of ~ $220B in 2026 (vs ~ $132B in 2025) to fund AI data center build-outs, supporting AWS growth (from 20% revenue in 2024–25 to 30%+ in 2026) and retail momentum (15% unit volume growth in Q1). Ackman expects Amazon to compound total earnings at 20%+ pace, reinforcing a risk-on stance toward AI-driven AWS and retail share gains.

Analysis

The key market mechanism is not “capex is high,” it’s whether Amazon can turn a multi-year capacity build into an operating leverage story faster than the market’s depreciation model catches up. If AWS is really sustaining >30% growth while retail share gains persist, the company’s competitive moat widens because rivals must either match infrastructure spend or accept slower product velocity and weaker service levels. That said, the near-term loser is Amazon’s free cash flow optics, which can suppress the multiple even if earnings power improves; the market often punishes the bridge before it rewards the destination.

Second-order winners are the AI supply chain and data-center ecosystem: GPUs, networking, power, cooling, and construction vendors see a longer runway if hyperscalers keep spending. The more interesting competitive effect is on mid-tier cloud and e-commerce players: a stronger Amazon usually forces smaller platforms to spend more on fulfillment and compute just to hold share, which can compress margins across e-commerce and cloud-adjacent names.

Catalyst path matters. Over the next 1-3 months, the stock likely trades on confidence in AWS monetization and the sustainability of capex, not on last quarter’s growth print. Over 6-18 months, the thesis breaks if capex expands faster than revenue/operating income or if AWS growth slips back toward the low-20s, because then the market will reprice the build-out as inflationary rather than strategic. The consensus may be underestimating how durable Amazon’s retail share gains are, but overestimating how quickly the market will grant credit for AI infrastructure spend.

Key Decisions for Investors

• Stay structurally long AMZN, but add only on pullbacks tied to capex/Fcf نگرvity; 6-12 month horizon, with thesis invalidation if AWS growth falls back below the mid-20s or margin expansion stalls.
• Express the share-gain thesis with a pair: long AMZN / short XRT for 3-6 months. Risk/reward improves if consumer spending softens while Amazon keeps taking online retail share.
• Treat AMZN capex confirmation as a second-order bullish signal for AI infrastructure exposure; if future guidance validates continued spend and AWS growth holds above 30%, buy NVDA on weakness as a downstream beneficiary.
• Watch for any sign that capex is failing to lift operating income leverage. If the next earnings cycle shows higher depreciation without corresponding AWS acceleration, trim AMZN and expect multiple compression.