Alphabet vs. Amazon: Both AI Stocks Have Been Hammered, but One Looks Like a Better Buy Now

Google Cloud revenue rose 48% year over year to $17.7B in Q4 2025 (now >$70B annual run rate), while Amazon Web Services grew 24% y/y to $35.6B in Q4. Alphabet reported consolidated Q4 revenue of $113.8B (+18% y/y) and Amazon reported net sales of $213.4B (+14% y/y). Both firms plan massive 2026 capex — Amazon ~ $200B and Alphabet $175–185B — to build AI infrastructure. The author favors Alphabet due to faster cloud growth and a cheaper P/E (~25 vs ~28 for Amazon).

Analysis

Alphabet’s cloud momentum is less about one quarter of outperformance and more about asymmetry in monetization: its AI layer directly lifts high-margin search and ads, compressing the payback horizon on incremental data-center spend. Expect visible margin expansion to show up in multi-quarter cadence — not overnight — as higher ARPUs compound with lower incremental sales/marketing per incremental compute sale, likely crystallizing within 4–8 quarters.

Amazon’s play remains broader and higher-capex, which amplifies execution and power-cost risk: every incremental $10–20bn of annual build adds nonlinear exposure to hardware sourcing (GPUs), power contracts, and real-estate timing. That raises a realistic scenario where AWS stabilizes share but ROIC trails Alphabet’s cloud for 12–36 months unless Amazon extracts new, high-margin software revenue streams from its infrastructure investments.

Second-order winners include GPU and data-center supply-chain beneficiaries — Nvidia’s pricing power, fab-constrained alternatives, and data-center power/colocation providers — while potential losers are smaller cloud-native vendors that face renewed price competition. Key catalysts to watch are large enterprise multi-year contract disclosures, GPU supply cadence, and any regulatory or power-grid constraints; conversely, rapid open-source AI stack adoption or macro ad-spend deterioration are clear reversal risks within 3–12 months.

Key Decisions for Investors

• Pair trade (6–12 months): Long GOOG/GOOGL 2% portfolio weight vs short AMZN 1.8% weight (dollar-neutral). R/R: target relative outperformance of 20–30% if Alphabet sustains higher ARPU gains; stop-loss if GOOG underperforms AMZN by 12–15%.
• Options play on GPU tailwinds (3–9 months): Buy NVDA call spreads (buy 1x nearer-dated ATM call, sell 1x higher strike ~10–15% OTM) to capture continued cloud GPU demand while limiting theta decay. Max loss = premium paid, target 2–4x return if supply tightness persists.
• Event-driven long/hedge (12–24 months): Accumulate INTC via long-dated calls (LEAPS) as a tactical hedge for a scenario where hyperscalers diversify away from Nvidia CPUs/accelerators. Payoff is binary but caps premium outlay versus outright long equity.
• Risk-off trigger: if Amazon announces >10% YoY 4-quarter capex overshoot or sequential guidance miss, convert pair into outright short AMZN (add 1% portfolio) and take profits on GOOG-sized longs — this trade flips asymmetry in 1–3 quarters.