Here's How Much You'd Have Today if You Invested $5,000 Each in Amazon and Costco 3 Years Ago. The Gains May Surprise You.

The article argues Amazon should outperform Costco over the next three years, supported by AWS’s $128.7B 2025 revenue and an expected global cloud market reaching $2T by 2030, with 10%-15% tied to generative AI. It cites AWS’s 28% share vs Microsoft Azure at 21% and notes Amazon’s forward P/E of 29.2 vs Costco’s 41.1. Using $5,000 invested on July 14, 2023, the values shown are ~$9,186 for Amazon vs ~$8,756 for Costco, and the piece frames Amazon’s AI infrastructure/capex as the key swing factor.

Analysis

The market is still treating Amazon as a retail proxy with AI upside, but the real re-rating lever is whether its cloud spend converts into durable operating leverage. On a 6-18 month horizon, AMZN has a better path to multiple expansion than COST because the incremental dollar of growth can come from a much higher-margin engine; COST’s quality is already well owned and priced as a bond substitute, which limits upside if macro stays benign.

The second-order winners are the AI infrastructure suppliers and adjacent software stack: if AWS keeps leaning into proprietary silicon and capacity buildout, demand should stay supported for NVDA/AVGO-style picks-and-shovels, but only if hyperscaler capex remains disciplined enough to lift free cash flow. The main loser is not COST’s business model, which remains resilient, but its relative multiple: as investors rotate toward duration and AI monetization, defensive compounders can underperform even without fundamental damage.

Near term, the key risk is that AI capex keeps outrunning monetization, leading to a stock-level de-rate in AMZN despite solid top-line growth. That would be especially likely if AWS growth merely tracks the market rather than re-accelerating versus MSFT Azure over the next 1-2 quarters. The contrarian point is that the consensus may be too focused on TAM and not enough on ROI; if management shows margin expansion from custom chips and higher utilization, the stock can re-rate quickly, but if not, the AI narrative becomes a cash burn story rather than a growth story.

Key Decisions for Investors

• Long AMZN / short COST as a 6-12 month relative-value trade: AMZN has more upside to operating leverage and a lower starting multiple; risk is a continued multiple premium for COST if defensives stay crowded.
• If AMZN pulls back on any capex-related skepticism, use that weakness to initiate a 3-6 month call spread rather than outright stock; thesis is that the market will pay for evidence of AI monetization before it pays for the full TAM story.
• Watch MSFT vs AMZN on cloud share prints over the next 1-2 quarters; if Azure re-accelerates faster than AWS, reduce AMZN exposure because the stock’s AI premium becomes harder to justify.
• For lower-risk exposure, pair long AMZN with a smaller short in a consumer-staples ETF basket or COST only if valuation stays >35x forward earnings; if COST derates, cover quickly because the downside is limited by defensive flows.