Amazon and Alphabet Are Each Spending At Least $200 Billion on Data Centers This Year. Here's Which Stock Outperforms Through 2028.

Amazon plans about $220 billion in 2026 data-center capex, versus Alphabet's roughly $200 billion. In Q2, AWS generated $42.2 billion in revenue at a 39% operating margin and contributed about 60% of Amazon's operating profit; Google Cloud grew faster, up 82% year over year versus AWS's 37%, but had a 36% margin and a $514 billion backlog compared with AWS's $496 billion. The article argues AWS's greater importance to Amazon's profits makes Amazon's investment more consequential and predicts Amazon will outperform Alphabet through 2028.

Analysis

The key asymmetry is earnings sensitivity, not which cloud unit is growing faster: AWS gives Amazon more operating leverage, but also makes Amazon’s profit outlook more exposed to cloud execution. That cuts both ways. If added capacity is delayed, underutilized, or priced competitively, depreciation and operating costs can absorb cash before revenue converts to profit. Google Cloud’s faster growth could still support Alphabet’s multiple even if its near-term consolidated earnings contribution is less pronounced.

Treat backlog as an indicator of demand, not as booked near-term revenue or guaranteed returns. The investment case depends on delivery schedules, cancellation terms, utilization, pricing, and incremental returns on the buildout—none are established here. Power availability, equipment lead times, and customer workload migration could also shift monetization beyond the market’s expected window. Near term, elevated capex may pressure free cash flow and make both stocks sensitive to any guidance suggesting slower capacity deployment. Over 1–3 months, watch cloud growth, operating margins, capex guidance, and backlog conversion commentary; over 6–18 months, utilization and returns on invested capital matter more than headline spend.

The contrarian risk to the bullish Amazon comparison is concentration: AWS’s larger earnings role amplifies disappointment as much as acceleration. A relative-value position is more defensible than an outright call absent valuation and consensus data. The thesis weakens if AWS growth or margins decelerate while capex remains elevated, or if Google Cloud sustains faster growth with improving margins and conversion.

Key Decisions for Investors

• Consider a modest long AMZN / short GOOG relative-value position only if valuation, beta, and portfolio exposures support it; stage entry around earnings rather than chase the article’s bullish framing. The payoff is AWS-driven profit leverage, while the hedge reduces broad mega-cap tech exposure.
• Treat the trade as a 1–3 month catalyst position and review after each earnings report: track AWS and Google Cloud growth, operating margins, capex guidance, backlog conversion commentary, and free-cash-flow trends. Do not infer revenue timing from backlog size alone.
• Scale down or exit if AWS growth or margins weaken while spending stays elevated, if capacity constraints delay customer delivery, or if management indicates deteriorating returns on new infrastructure. Reassess the short leg if Google Cloud’s growth advantage persists alongside margin expansion.
• For a 6–18 month supply-chain screen, monitor data-center power, cooling, networking, and semiconductor suppliers as potential beneficiaries of sustained buildout; avoid a directional supplier trade until deployment timing, orders, and customer concentration are verified.