Amazon's Stock Just Soared 15%. Here's Why Now Is Just the Beginning.

Amazon shares jumped ~15% after Q2 earnings as AWS strength reinforced management’s AI infrastructure narrative. AWS sales rose 37% YoY with operating income up 64% YoY, while Amazon said it plans $220B in data center capex this year to support AI model training and ongoing compute usage. The piece argues AWS could reach $1T in revenue (from ~$148B TTM) and maintain ~39% operating margin, implying ~$391B operating profit for the cloud segment and materially boosting the stock’s long-term upside.

Analysis

AWS is increasingly behaving like the margin engine that determines Amazon’s multiple, not the retail business. The market will likely keep rewarding AMZN as long as it can prove that incremental capex is converting into sustained utilization and operating leverage; that matters more than headline revenue because cloud is the scarce asset with pricing power. Second-order, the biggest beneficiaries are the semiconductor and infrastructure stack tied to AI buildout, but the capex mix matters: the more Amazon shifts to proprietary silicon and internal systems, the less of each dollar flows through to NVDA.

Near term, the stock can keep drifting higher for weeks if management keeps reinforcing demand visibility, but the first real test is whether revenue growth outpaces depreciation and data-center overhead over the next 1-3 quarters. If free cash flow lags while capex stays elevated, the market can easily flip from “AI platform premium” to “cash burn skepticism,” especially after a large post-earnings rerate. That would hit not just AMZN, but also high-multiple cloud names as investors question whether hyperscale expansion is becoming an arms race instead of a profitable moat.

The consensus may be underestimating the durability of AWS demand, but it may also be overestimating the immediate benefit to NVDA. The structural bull case for AMZN is 6-18 months out: every incremental workload moved onto AWS deepens switching costs and raises lifetime value, yet the path there can be volatile if pricing competition from Azure/GCP or in-house chip substitution compresses take-rate. Falsifier: if AWS growth decelerates materially or margins flatten despite continued capex, the thesis of accelerating per-share value should be cut quickly.

Key Decisions for Investors

• Buy AMZN on 3-5% pullbacks over the next 2-4 weeks; use the post-earnings momentum only if volume stays supportive. Risk/reward is attractive as long as AWS revenue growth and operating margin do not roll over in the next two quarters.
• Relative-value pair: long AMZN / short NVDA over 3-6 months. Thesis: the market is likely to over-attribute AWS capex to NVDA, while Amazon retains the cleaner earnings rerating if cloud utilization stays high. Falsifier: NVDA re-accelerates on supply-tight AI server demand or Amazon signals a higher external silicon mix.
• Use an AMZN call spread 3-6 months out rather than outright calls if implied vol remains elevated; target upside continuation while limiting exposure to a capex-driven FCF disappointment. Cut if AMZN gives back the post-earnings gap and fails to reclaim it within 2-3 weeks.
• Watch for a short opportunity in cloud-infrastructure beneficiaries if AWS monetization weakens: if management guides to heavier-than-expected depreciation or weaker free cash flow, consider shorting a basket of high-multiple AI infrastructure names on any sympathy rally.