Warren Buffett vs. Bill Ackman: One Piled Into Amazon While the Other Sold

Amazon is framed as an AI and cloud beneficiary, with AWS cited as a key profit engine and demand for AI services remaining strong. In Q4, Warren Buffett cut his Amazon stake by 77% to about 0.1% of his portfolio, while Bill Ackman increased his position by nearly 65% to 14% of his portfolio. The article is mainly a comparative commentary on billionaire investing styles rather than new company-specific fundamentals, so near-term market impact should be limited.

Analysis

AMZN is the cleaner expression of the AI capex cycle than the usual semiconductor beta trade: the upside is not just model demand, but the monetization of inference, storage, and workflow integration across an installed enterprise base. The second-order effect is that every incremental AI workload reinforces AWS switching costs, which should support margin durability even if headline capex growth moderates. That makes AMZN more resilient than pure AI beneficiaries whose revenue is still more directly tied to one-time hardware spend.

The positioning signal matters: a high-conviction allocator adding materially while a long-term value holder trims often marks a regime where the market is still underpricing duration of growth. The risk is that investors extrapolate AI demand too far ahead of actual utilization; if utilization rates lag spend, AWS could see a multiple compression event even while revenue grows. Near term, the stock likely trades on commentary around enterprise demand and capex discipline; over months, the key variable is whether AI-related workloads become a larger share of cloud mix without margin dilution.

The contrarian miss is that Amazon’s AI exposure may be less about “winning AI” and more about being the default toll road for everyone else’s AI adoption. That creates a more defensive growth profile than the market typically assigns to large-cap tech, especially if customers diversify model providers but keep infrastructure anchored in AWS. Relative to the broader mega-cap complex, AMZN looks like a better risk-adjusted way to express AI infrastructure demand without taking single-vendor model risk.

BRK.B’s trim should not be read as a negative fundamental signal for Amazon so much as portfolio reallocation after a strong run; the bigger takeaway is that the stock has moved from “cheap optionality” to “quality compounder,” which narrows forward return expectations but does not break the thesis. If AI spending stays rational, AMZN can rerate modestly higher from here; if spending turns into capex bloat, the stock probably de-risks before the business does. That asymmetric setup favors buying weakness rather than chasing strength.

Key Decisions for Investors

• Go long AMZN on 3-5% pullbacks over the next 2-6 weeks; target a 12-month upside in the mid-teens if AWS AI demand remains durable, with a stop if management commentary shifts toward margin pressure from overbuild.
• Pair trade: long AMZN / short a basket of higher-beta AI hardware names over 1-3 months; thesis is that software-infrastructure monetization is more durable than one-time capex beneficiaries if AI spend normalizes.
• Use AMZN Jan-2027 call spreads to express upside with defined risk; attractive if you expect multiple expansion from sustained AI inference demand but want to avoid paying full equity downside.
• Trim or avoid chasing BRK.B as a “better Amazon” proxy; the relative signal suggests Amazon has already moved from mispriced to fairly recognized, reducing near-term alpha from fresh entry.
• Watch for any AWS commentary on utilization versus capex growth; if capex accelerates without corresponding usage metrics, fade the rally and consider a tactical short against XLK or QQQ into earnings.