Better Stock to Buy Right Now: Amazon vs. Costco

Amazon reported Q4 2025 net sales of $213.4B (+14% YoY) and earnings of $21.2B (+6%), but flagged a planned ~$200B in 2026 capex largely for AI/AWS infrastructure; forward P/E ~26x. Costco posted fiscal Q2 2026 net sales of $68.2B (+9.1% YoY) and net income of $2.04B (+13.8%), with U.S./Canada membership renewal at 92.1%; forward P/E ~48x. The author prefers Amazon due to its cheaper valuation and AI-driven growth tailwind, while noting Costco's defensive pricing power amid inflation and tariffs.

Analysis

AWS-scale AI investment creates a multi-layered beneficiary set beyond the obvious GPU makers: data-center power vendors, substation/electrical-equipment suppliers, and systems integrators will see multi-year order visibility, which tightens supply and forces smaller clouds to either partner or consolidate. That flow-through favors firms with deep tooling and proprietary silicon roadmaps (advantage: vertically integrated hyperscalers and foundry-aligned chip vendors) and squeezes commodity x86 suppliers unless they pivot to differentiated accelerators.

Costco’s model — high recurring membership cashflow plus pricing discipline — makes it structurally defensive if consumers retrench, but it also sets up second-order margin risk: maintaining low prices during cost shocks requires either supplier concessions or accelerated private-label penetration, which shifts bargaining power and could compress supplier margins across categories. Tariff and freight volatility will therefore manifest as either a temporary earnings hit or a strategic acceleration of supplier consolidation and vertical integration.

Primary near-term catalysts are execution on AI monetization (12–36 months) and consumer-income elasticity under tighter credit conditions (0–12 months); both can rapidly re-rate these names. The consensus is underweighting the lag between capex and profitable AI services — if AWS converts capacity into high-margin, recurring agentic services, valuation dispersion will compress quickly in favor of Amazon, but an execution miss or regulation could cause a material multi-quarter downside for capital-intense players.

Key Decisions for Investors

• Buy AMZN Jan-2028 LEAPS calls (or equivalent 18–36 month call exposure). Rationale: asymmetric payoff to multi-year AI monetization with max loss = premium paid; target 50–100% return if AI services ramp within 24 months; reduce position by half on a 60% gain or on clear signs of monetization slowdown.
• Pair trade — long AMZN / short COST (equal-dollar notional) for 6–24 months to isolate an AI re-rating vs retail multiple compression. Risk management: stop the pair if COST outperforms AMZN by >15% in any 90-day window; target spread capture = AMZN outperformance of 25–40%.
• Long NVDA (cash or 6–12 month calls) to play AI infra supply tightness and pricing power. Expect elevated cyclicality; set a trailing stop at 30% drawdown and takeprofit at 50%+.
• Buy a 6–12 month put spread on COST (e.g., buy a nearer-dated put and sell a lower strike) to hedge funded retail exposure or to express concern about multiple compression. This limits max loss to the net premium and profits if Costco’s premium multiple contracts by >10–15%.