Should You Buy Amazon Stock For The Demand AWS Cannot Yet Serve?

Should You Buy Amazon Stock For The Demand AWS Cannot Yet Serve?

Wed, September 16, 2026 at 6:25 PM GMT+3 3 min read

Amazon.com (AMZN) stock has returned 8.9% over the past twelve months and trades about 13% below its 52-week high. Its cloud arm spent those months doing something the share price has not reflected. AWS growth sped up again in Q2 2026, and the contracts queued up behind it kept building. The upside case for the stock rests almost entirely on turning that queue into revenue.

Why Is AWS Accelerating At This Size?

AWS revenue grew 36.7% year over year in Q2 2026, up from 28% in Q1 2026, and management counts it as the fifth straight quarter of acceleration. That direction is rare on a base this large. Management's explanation is mechanical. Agentic AI leans on ordinary processors as much as on accelerators, because post-training reinforcement learning and agent tool use mostly run on CPUs.

That CPU work favors Graviton, Amazon's own processor, now used by 98% of AWS's top 1,000 EC2 customers. At the same time, Amazon Bedrock aggregates leading foundation models while custom Trainium silicon powers dedicated AI compute—backed by multi-year, multi-gigawatt commitments from Anthropic and OpenAI. Because an AI workload pulls traditional cloud compute and storage along with it, AI expansion and core cloud consumption reinforce each other.

How Much Cloud Business Has Amazon Already Signed?

The backlog of committed cloud contracts stood at $496 billion in Q2 2026, growing at triple-digit rates year over year. That is not one year's revenue waiting to be booked. Most AI capacity is contracted for at least five-year terms, so it converts to revenue over years. What the backlog does say is that demand has stopped being the constraint.

AWS is a $169 billion annualized business. Management says even the capacity it is building cannot meet all the demand it has in 2026. The bulk of 2027 capacity is already reserved, and a good deal of 2028 is too. Revenue arrives only as that reserved capacity is delivered.

What Should You Watch As The Data Centers Open?

The common worry was that AI workloads would arrive at worse margins. So far they have not. AWS operating margin rose 520 basis points year over year in Q2 2026, excluding a gain from derivative accounting on energy contracts. The CFO credited disciplined efficiency gains and capacity optimization for the improvement, though, and repeated that AWS margins will fluctuate. Management says AI economics are tracking what core cloud did at the same stage, slightly ahead of it.

The cost of the build-out is the honest counterweight. Amazon now expects roughly $220 billion of cash capital expenditure in 2026, raised from about $200 billion because memory prices climbed. That higher figure is the one to watch, and it moved for a reason Amazon does not control. Free cash flow stays under pressure until the data centers open, start billing, and the servers are a few years into use.

Amazon stock has gained more than 30% inside two months on 14 separate occasions since 2010. The move is there when the operating news is good enough. If the capacity is landing, it shows up first in what management is willing to guide, and that signal is worth reading across every large company raising its outlook.

How Much Of Your Wealth Should One Opportunity Carry?

Spotting an opportunity is the enjoyable half of investing, and deciding how much of your net worth rides on it is the half that compounds. Concentration tends to arrive by accident rather than by decision. What your largest position would do to your net worth is exactly what the Trefis Wealth team computes, with the same rules-based systematic discipline that runs our High Quality Portfolio. Request a free vulnerability audit of your biggest positions.