Amazon Web Services reported a 37% year-over-year increase in revenue in the second quarter of 2026, the fastest rate in eighteen quarters. Because AWS had been slowing down for some time and analysts had begun to question whether the cloud industry was settling into something more established and predictable, the figure attracted attention. For the time being at least, the acceleration provided an answer to that question. It also gave Wall Street firms, who had previously been largely optimistic, newfound faith in Amazon’s long-term investment outlook.
Morgan Stanley designated Amazon as a top pick and increased its price target to $350. Twelve months ago, the company’s projections of AWS growth rates of 47% in 2027 and 42% in 2028 would have seemed ambitious, but given the acceleration already in place, they now seem reasonable. AWS has an annualized revenue pace of $169 billion, a $496 billion contract backlog, and operating margins that are significantly higher than 39%. Currently, about 60% of Amazon’s total operating income comes from that 21% of total revenue. The calculation of what happens as that share increases over time is simple: if AWS continues to grow more quickly than commerce, total margins will increase and the stock will follow suit.
Analysts seem to have been waiting for precisely this kind of acceleration to validate what the bull thesis has always demanded: that capital expenditures would result in long-term, high-margin revenue rather than disappearing into a competitive environment where Google Cloud and Azure matched every dollar spent. There is still competition. The competition to supply AI infrastructure at scale is fiercely competitive, and Microsoft and Alphabet are not standing still. Nevertheless, a level of revenue visibility that is rarely provided by pure growth projections is provided by AWS’s contract backlog and the specificity of customer commitments—Anthropic is scheduled to secure up to 5 GW of Trainium capacity, while OpenAI committed about 2 GW.
Contrary to some headlines, it is more difficult to ignore the risk aspect of the long-term investment outlook for Amazon. Depending on estimates, Amazon directed between $130 and $220 billion in capital expenditures, an aggressive buildout that included robotics, data centers, custom chips, and satellite infrastructure. As capital expenditures increased by 68%, free cash flow became negative at about -$7.6 billion.
The amount of long-term debt has increased to about $119 billion. These are actual figures, not accounting jargon, and they show a business placing huge bets on a return that hasn’t yet materialized. Customers have already reserved the majority of 2027 capacity, according to CEO Andy Jassy, which is somewhat consoling, but it’s important to consider the difference between recognized revenue and reserved capacity.
It’s difficult to ignore how the business mix has changed in such a short time. What started out as an online retailer that investors primarily valued based on gross merchandise volume has evolved into a business where enterprise cloud contracts, custom silicon, and AI infrastructure are the main topics of discussion. Ten years ago, Amazon’s advertising division was hardly mentioned in public discourse.
Last quarter, it increased by 26% to $19.81 billion. This type of revenue compounds in ways that are easy to underestimate because it is high-margin, scalable, and built upon an existing logistics infrastructure. Although only 40 to 45 percent of U.S. fulfillment centers are highly automated, Jefferies has separately identified warehouse automation as another margin lever, pointing out that there is substantial potential to reduce unit economics over time.
The stock is currently trading at about $262, which is about 11% less than its 52-week high of $287.20 and roughly 24 times forward earnings. It is less expensive than Microsoft at 25 times, and it has more room to grow its 13.7 percent operating margin than most competitors. Under what the analysts referred to as conservative assumptions regarding AWS growth rates and operating margins, a Motley Fool analysis predicted a path to $600 per share by 2030. That would suggest a $6.5 trillion market capitalization, which may seem high until you keep in mind that the framework that gets there is contracted cloud infrastructure with known clients paying known rates for known capacity rather than speculative technology.
It’s still unclear if the pressure on free cash flow is transient, a known consequence of building ahead of demand, or if it persists longer than management anticipates. This type of wager has previously been made by Amazon, which spent years constructing warehouse infrastructure that appeared to be a waste of money before it suddenly appeared to be a competitive moat.
Investors who have followed the company long enough have become accustomed to the pattern and are able to put up with the discomfort. Every price target, analyst recommendation, and earnings call between now and the end of the decade is based on the central question of whether this specific cycle follows the same arc at a scale that dwarfs anything Amazon has attempted before.
