Amazon's 12% Jump Is an AWS Bet: $300 Billion Repricing on AI Revenue

AWS, not Amazon as a whole, drove the rally

This was mainly a repricing of AWS. Amazon shares jumped more than 12% before the bell, putting the company on track to add about $300 billion in market value. Investors looked past a 10% increase in planned capital expenditure to $220 billion and focused on AWS, which now looks more clearly tied to near-term AI monetization.

AWS revenue rose 37% to $42.2 billion, beating the consensus growth rate of 31.21%. That suggests demand is running ahead of expectations even as investors scrutinize enterprise AI spending more closely. For bulls, that is the payoff investors wanted to see after heavy capex: stronger cloud growth and more evidence that AI workloads are turning investment into revenue.

A more cautious read is still possible. Large spending plans, weaker short-term cash flow, and AI investment pressure across the sector mean the long-term payback still has to be maintained. But the immediate market message was clear: investors rewarded Amazon for showing that its AI spending is already helping AWS.

Why one strong quarter still matters

The key question is no longer whether AWS simply had a good quarter. It is whether AWS is becoming a larger part of Amazon's overall earning power. That is a valuation shift. A business can earn a higher multiple when one unit stops looking like a supporting growth line item and starts looking central to future profits.

AWS now has the strongest cloud growth in over four years, and management said capacity was still too tight to fully satisfy customers even after raising spending. That supports the view that demand is outpacing supply rather than simply absorbing more investment.

Backlog and OpenAI give the demand story more visibility

Amazon is sitting on a $200 billion backlog, TD Cowen said, along with several unannounced deals not included in that figure. Backlog is not the same as recognized revenue, but it can improve visibility if AWS continues converting that pipeline as capacity comes online.

The OpenAI relationship is the clearest single example. TD Cowen highlighted a multi-year arrangement that includes a $38 billion commitment from OpenAI for AWS compute. Reuters said AWS has benefited from major infrastructure and chip-supply partnerships this year, including deals with OpenAI and other leading AI companies. That makes the case less abstract: large customers are committing capacity, and that can help support utilization as Amazon keeps building.

Trainium matters if mix improves

There is another piece investors are watching: cost structure. Management has said AWS' annual AI revenue run rate has surpassed $15 billion, while TD Cowen noted that the Trainium business has reached a multibillion-dollar run rate and was growing more than 150% quarter over quarter in the third quarter of 2025. If customers use more of AWS' own silicon over time, the business could protect margins better than a simple volume story would imply.

What keeps the rerating alive-and what breaks it

The first credit has been given for showing that AWS can turn AI spending into visible demand. What matters next is whether that payoff remains visible.

The market already signaled what it wants. Amazon was rewarded for strongest cloud growth in over four years even with larger spending plans, in a reaction that contrasted with peers that appeared to have longer paths to returns. That leaves the next few quarters as the real test.

What supports the bull case

• AWS keeps posting growth that stands out relative to the sector.
• Backlog and unannounced deals help translate into revenue as capacity expands.
• AI revenue remains meaningful, with management saying it has surpassed $15 billion at an annual run rate.

What could weaken the move

• Spending rises faster than revenue.
• Backlog does not convert into recognized revenue at the pace investors expect.
• AI demand proves less durable than the current quarter suggests.

For now, this still looks like an early credit to Amazon rather than a fully verified long-term outcome. If AWS keeps showing demand ahead of capacity, the rerating can hold. If monetization slows, the market may stop paying up for the capex story.