Amazon's AWS Backlog Just Hit $496 Billion: Is the Stock a Screaming Buy Under $270?

AWS backlog reached $496B in Q2, implying nearly 12 quarters (~3 years) of revenue at the current $42.2B run-rate, signaling persistent demand. Amazon is planning $220B in 2024 capital expenditures, largely for data centers, and CEO Andy Jassy cautioned it may not fulfill cloud demand in 2026 and expects similar constraints in 2027. Overall, the article frames a multi-year AI/data-center growth tailwind supporting a bullish valuation outlook.

Analysis

The signal here is not “AWS has a big backlog”; it is that hyperscale cloud is now capacity-constrained for multiple planning cycles. In that setup, the bottleneck shifts from demand generation to allocation of scarce compute, which usually favors the ecosystem supplying GPUs, networking, power, and cooling more than the platform owner itself. Near term, backlog can cap upside because revenue recognition is gated by buildout pace and lead times, not just demand.

The less appreciated drag is capital intensity. When a cloud leader has to pre-fund capacity, free cash flow and reported margins can lag demand for several quarters as depreciation ramps before utilization catches up, creating a window where the stock can look operationally strong but still underperform on multiple expansion. If customers can’t get AWS capacity, incremental spend leaks to Azure, GCP, and OCI, which improves competitive parity more than it changes aggregate cloud demand.

Consensus is reading backlog as linear upside, but the real question is conversion rate and price per unit of compute. If AI workloads shift from training to lower-footprint inference, or if Amazon keeps accelerating capex without a corresponding utilization inflection, the narrative can flip from scarcity to overbuild quickly. Key falsifiers are a backlog growth stall, a capex guide cut, or one to two quarters of AWS growth decelerating despite the spend surge.

Key Decisions for Investors

• Add to AMZN only on 3-5% pullbacks, or express via a 12-18 month call spread; upside is backlog conversion and operating leverage, while the main risk is free-cash-flow compression from sustained capex.
• Long NVDA / short AMZN as a relative-value trade over the next 1-3 months if you think the market underestimates that constrained cloud capacity still monetizes first through chips, networking, and data-center gear before it shows up in hyperscaler margins.
• Set a downside alert on AMZN if AWS revenue growth slows for a quarter while capex remains elevated; that would be the cleanest signal to fade the story with reduced exposure or a short-dated call spread.
• Watch MSFT and GOOGL as secondary beneficiaries of AWS supply constraints; if enterprise demand is truly outstripping AWS buildout, relative share gains may accrue to whichever hyperscaler has the most available capacity, not necessarily the one with the largest backlog.