This Dominant Hedge Fund's Top 3 Holdings Are Genius Buys Now

Amazon, Micron, and Alphabet are highlighted as top holdings of David Tepper’s Appaloosa, with the article arguing all three remain attractive amid the AI buildout. Amazon is investing $200 billion in AWS this year and AWS growth hit its fastest pace in nearly four years in Q1; Micron is benefiting from a memory-chip shortage and trades at 16x forward earnings; Alphabet’s Google Cloud revenue rose 63% in Q1. The piece is largely bullish commentary rather than new company-specific news, so the immediate market impact is limited.

Analysis

The common thread is not “AI winners” in the abstract, but the capex translation layer: cloud infrastructure spend is becoming the primary monetization path for AI demand, which structurally favors hyperscalers that can fund buildout internally and amortize it over software-like margins. That creates a second-order winner set in power, networking, and data-center component suppliers, while pressuring smaller cloud providers that cannot match the pace of capacity expansion or offer similar bundled services.

Micron’s setup is more than a cyclical upturn; memory shortages tend to create an inventory reflex in which OEMs and hyperscalers over-order for 1-2 quarters, extending pricing power beyond the initial spot tightness. The key risk is that when supply normalizes, earnings can mean-revert much faster than the market expects, so the current multiple is only cheap if the demand curve remains stretched into the next capex cycle. That makes MU attractive tactically, but fragile if pricing starts to plateau before consensus models have fully reset.

Alphabet’s cloud acceleration is the cleaner long-duration signal because it monetizes AI indirectly through infrastructure utilization and directly through TPU demand, which raises switching costs for enterprise customers. The market may still be underestimating how much of the upside is in operating leverage rather than headline growth: a few points of incremental cloud mix can expand earnings faster than revenue, especially if paid search remains resilient. The contrarian risk is that this becomes a crowded “quality growth” trade, so any moderation in cloud growth could trigger multiple compression even if fundamentals stay healthy.

Against that backdrop, the consensus may be overpaying for narrative certainty in AMZN while underappreciating the durability of GOOGL’s optionality and the cyclical torque in MU. The best setup is to lean into names with visible capacity constraints and near-term pricing power, but avoid assuming all AI capex is equally durable; the market can rerate quickly once spending shifts from scarcity-driven urgency to capacity optimization.

Key Decisions for Investors

• Long AMZN on a 3-6 month horizon via common stock or call spreads: target upside from AWS capex monetization and margin leverage; trim if cloud growth decelerates or capex intensity stops expanding.
• Long MU for 1-2 quarters as a cyclical momentum trade: attractive if memory pricing remains tight, but size smaller than a structural compounder because earnings can mean-revert quickly once supply catches up.
• Pair trade: long GOOGL / short a basket of lower-quality cloud infrastructure beneficiaries over 6-12 months, betting that TPU-driven workload share and operating leverage matter more than pure growth optics.
• Buy NVDA or AI infrastructure leaders only on pullbacks, not strength; use them as a hedge against the theme, but recognize hyperscaler capex can eventually normalize and compress multiple expansion.
• Set a risk-control level on MU and AMZN positions: reduce exposure if 2 consecutive monthly industry checks show memory price stabilization or cloud capex guidance flattens, as both trades depend on continued scarcity and spend acceleration.