Greg Abel Just Dumped Amazon. David Tepper Nearly Doubled His Stake. One of Them Is About to Look Very Smart.

Berkshire Hathaway's Greg Abel likely exited Amazon in Q1 2026 as part of a cleanup of positions formerly managed by Todd Combs, while David Tepper nearly doubled Appaloosa's Amazon stake after its February sell-off. The article frames Tepper's move as a bet on Amazon's AI infrastructure spending and AWS leadership, with the stock already up more than 20% since the end of Q1. The piece is largely interpretive commentary rather than fresh market-moving news.

Analysis

The key signal is not the Berkshire exit itself, but the governance reset it implies: Abel is likely clearing legacy, manager-specific exposures and steering the portfolio toward simpler, more liquid, higher-conviction holdings. That tends to reduce idiosyncratic alpha from “unknown” smaller positions, but it also lowers the risk of dead money and frees capital for larger repurchases or future public-market dislocations. In other words, the sell is more about portfolio architecture than a negative fundamental view on the asset.

Amazon’s weakness into the quarter looks like a classic setup where the market confused capex intensity with deteriorating economics. If AWS capacity build-out is being accelerated for AI workloads, the near-term P&L hit should matter less than the optionality on future inference demand and the strategic lock-in created by scarce cloud capacity; the second-order beneficiary is AMZN’s ecosystem, not just cloud margins. The more subtle loser is any AI infrastructure competitor forced to match spend without Amazon’s balance-sheet breadth and consumer cash-flow engine.

Tepper’s move is more interesting as a timing signal than a valuation call. If he added on the drawdown, he is effectively expressing a 6-18 month view that the market will re-rate capex-heavy AI platforms once capacity starts monetizing; that argues for buying weakness, not chasing strength after a 20% rebound. The contrarian miss is that the upside may be capped until investors see evidence that AI-related spend is converting into operating leverage rather than just revenue growth.

Near term, the biggest risk to the bullish case is a second wave of sentiment de-rating around spending discipline, especially if cloud growth decelerates or if broader AI trade leadership rotates away from mega-cap platforms. But on a 12-month horizon, the setup still favors Amazon because incremental capacity can become a scarcity asset if demand remains tight. That makes the stock more attractive as a volatility-managed long than as a momentum trade.

Key Decisions for Investors

• Go long AMZN on any 3-5% pullback over the next 2-4 weeks; target a 6-12 month hold, with a risk/reward skew favoring another multiple expansion if AI infrastructure monetization starts to show through in AWS margins.
• Buy AMZN Jan-2027 calls or call spreads to express the thesis that capex becomes a monetization catalyst over 12-18 months; downside is capped premium, upside benefits from re-rating if AI demand stays sticky.
• Pair trade: long AMZN / short a basket of less-capitalized cloud or AI infrastructure names with weaker balance sheets over 3-6 months; thesis is that capacity-rich platforms can fund the AI buildout without diluting returns.
• Avoid treating BRK.B’s sale as a bearish signal on Amazon fundamentals; for traders, fade any knee-jerk post-13F weakness in AMZN because the flow is more likely portfolio cleanup than fundamental deterioration.
• If AMZN fails to hold recent gains and gives back the post-Q1 rebound, reduce exposure quickly — that would suggest the market is re-pricing AI capex as pure expense again, which is the main reversal risk.