Just a few months ago, Amazon’s resurgence seemed unstoppable – until it actually was. Amazon stock closed at an all-time high of $284 on August 3, after receiving a huge boost after the earnings call. That’s up nearly 45% from its 52-week low in February. The $300+ seemed like just a formality. But since then, Amazon has fallen more than 7%, making last month’s rise-and-fall chart look more like seismic activity than a stock trade. So what exactly happened? And what will it take to get stocks back on track? In a note this week, Morgan Stanley analysts made the case for why they added Amazon to their list of top picks and raised their price target from $335 to $350, offering some food for thought in a note suggesting a 33% upside from Monday’s closing price. Before we get into what’s going to happen with Amazon in the coming weeks and months, and why Morgan Stanley is so positive, let’s take a quick look back at how we got here. How we got here It had a strong last quarter, with the company’s Amazon Web Services (AWS) cloud business smashing revenue and operating profit expectations. But what really brightened up the stock were CEO Andy Jassy’s comments about the 30-year lifespan of data centers and the revenue dynamics of the hardware used in them. “The useful life of our servers is now at least five to six years, and most of our AI capabilities these days are at least five-year contracts, which means we’re generating significant free cash flow on our network of servers and equipment within two to three years of breakeven,” he said on the second-quarter conference call. In an earnings analysis at the time, the club’s portfolio director Geoff Marks wrote that Jassy “delivered exactly what investors like us wanted to hear: a clear explanation of how our massive investment in artificial intelligence will generate returns.” The CEO’s remarks on the July 30 post-earnings conference call were so market-moving and fundamentally important that he devoted half of the next day’s homestretch to them. The stock price soared more than 15% on July 31st, and rose another 4.5% on August 3rd, hitting the aforementioned all-time high closing price. AMZN YTD Mountain Amazon YTD Jassy’s words resonate well beyond Amazon and are central to our broader view of the market today. We believe the high-quality, well-worn portion of the AI hardware trade presents a buying opportunity. Jim Cramer made his case in Sunday’s column. Jim’s key line is: “Nvidia’s chips are going to hold up. Their value could be maintained for 10 years.” Amazon uses a lot of its own custom chips, but it also has a lot of Nvidia chips in its cloud. That remains the same, Jassy said on the second-quarter conference call. “We also continue our strong partnership with Nvidia, and will continue to make AWS the best place to run Nvidia chips, as we have customers who will run on Nvidia for as long as they can predict. We strongly believe that customers want choice. Choice is good for them, good for competition, and it also helps reduce inference costs, which is something customers care about very much.” Where are we going? Ahead of Amazon’s third-quarter earnings next week, scheduled after the closing bell on October 29, we’re taking some time to think about our big position and what Amazon must do to get its stock price moving again. As of Tuesday, Amazon was our largest weighted holding at 5.3%. The company has a buy rating and a price target of $300 per share. At last week’s October monthly meeting, Jim said he has great confidence in Jassy and CFO Brian Olsavsky, adding that investors here who don’t own Amazon should buy. Jim added that Amazon’s third-quarter conference call will be key, arguing that if Jassy can prove that AI investments are paying off and that capital spending growth is starting to slow, the stock could really move higher. Digging into Morgan Stanley’s bullish view, analysts laid out a five-point checklist of what they think investors need to buy into the stock. (1) solid revenue growth for AWS, (2) better-than-expected operating margins and profit performance from AWS, (3) further proof of the importance and strategic value of Bedrock, (4) proof that Amazon can win with agent AI, and (5) investor confidence that Amazon stands to benefit in the age of physical AI (robots). Let’s unravel them one at a time and give us your opinion. AWS revenue growth: In the short term, it will be all about AWS, as the cloud is what makes artificial intelligence possible. “AWS’s growth and profit runway is undervalued,” Morgan Stanley said. In our view, the demand for computing is undeniable as more companies learn to leverage AI, realize its benefits, and as a result demand more. So the only question on our minds is: Can Amazon continue to win new business at the pace that Wall Street demands? We think it’s possible, and would remind doubters that Jassy may not have been the founder of Amazon, but he led the launch of AWS. That was four years before Microsoft’s Azure became a real competitor in the market. So when it comes to acquiring new clients in the cloud, Jassy is probably the last person you’d consider betting on. Even if we support sales growth, it’s important to note that we value the stock based on expected future earnings per share (EPS), preferably rather than sales. AWS Profitability: Although AWS primarily benefits from the demand for artificial intelligence, it is also a source of AI-related investments. When weighing up its AI investments and its ability to monetize that spending, AWS’s operating margins are even more important than revenue growth, and could be the main factor determining what the stock’s price will be in the short term. “Our bottom-up AWS spend construction gives us confidence that AWS can operate with incremental EBIT margins in the high 40% range in 27/28,” Morgan Stanley said. In other words, analysts believe Amazon could generate operating margins of more than 40% on new AWS business generated in 2027 and 2028. According to FactSet consensus estimates, AWS’s overall operating margin is expected to be 37.5% in 2027. We believe the more important consideration for investors is whether to believe the available lifecycle commentary by Jassy. Otherwise, your position will actually depend on the next quarter. If you do it like we do, you need to understand that profitability may not be as great as the market demands from quarter to quarter. However, in the long run, this cost will increase as the fixed cost of standing up a data center shell (the building the chips go into) decreases and the composition of data center costs shifts toward variable chip costs driven by short-term graphics processing unit (GPU) demand. Amazon simply buys more as it sees demand and builds customer contracts to ensure profitability. This goes back to the heart of the matter. We don’t see any change in the dynamics that Jassy commented on in the last earnings call, so AWS’s profitability is expected to remain strong in coming quarters. Bedrock: Generative and agentic AI solutions are a key driver for AWS, and we think Amazon is approaching this in the right way. Rather than pouring billions into a frontier-model arms race, Amazon has focused on AI solutions that power its products with proven profitability. “We continue to believe that AMZN’s ability to demonstrate Bedrock’s strength through this period of speculation (and its importance in driving GenAI adoption across the economy) is undervalued and could result in multiple expansions from current 15x 28-year EPS levels,” Morgan Stanley asserted. According to FactSet, the consensus forecast is for 2028 earnings per share to be a multiple of 18.4. For consumers, it’s Amazon’s all-in-one Alexa for Shopping Assistant. On the enterprise side, it’s Bedrock. Bedrock is Amazon’s “platform for building generative AI applications and agents at production scale.” The platform provides everything organizations need to build, but is model agnostic, offering hundreds of models for customers to choose from based on their unique needs. We think this is the right choice for companies that rent more computing than anywhere else in the world. No one cares which model the client uses, as long as the client consumes AI tokens (unit price for computing power) and demands the compute that Amazon sells. Indeed, to support our view that this is the right strategy for Amazon, we would like to call out Microsoft’s recent actions. Microsoft has responded positively to its revamped Copilot strategy, focusing more on an integrated user experience than being the world’s leading developer of AI models. Agentic AI: Amazon’s model-agnostic strategy with Bedrock looks solid. But the company’s retail arm is a bit more of a question mark, given that the company has taken steps to block market access for agent AI solutions such as Meta Platforms’ Muse. With major competitors like Walmart, Wayfair, and Best Buy choosing to work with Muse, it seems like we’ll see more in the future regarding Amazon’s plans to adapt to changing consumer shopping habits. After all, Amazon’s unwavering focus on consumer experience is what made it the online retail leader it is today. Morgan Stanley said, “While we don’t know exactly how agent deployment will protect retail media businesses, we expect we can learn more from other connectivity templates. …Potential mitigations include moving from Muse to Alexa for Maybe it’s a handover to Shopping (with AMZN managing the user experience/inventory), maybe it’s a take rate adjustment, maybe it’s revenue sharing…but we expect AMZN to continue leveraging its major e-commerce infrastructure and loyal user base.” Monetize. ” Physical AI (robots): In the long run, the benefits of AI to Amazon’s retail operations become a little easier to see, at least on the logistics side. We’re talking about physical AI here. From warehouse robots to self-driving cars, physical AI could have a huge positive impact on how packages arrive at our doorsteps. Morgan Stanley pointed to a February 2025 study that said every 10% of the world’s goods that pass through Amazon’s robot-enabled fulfillment centers could generate recurring savings of $2 billion to $4 billion annually. Analysts now believe the potential savings are increasing due to efficiency gains. They believe the increased use of robots in Amazon fulfillment centers will be a real catalyst next year. “The extent to which AMZN’s robotics fulfillment centers will increase (new builds and refurbishments) over the course of 2027 could show how AMZN is also a physical AI winner…technology provides more revenue power for reinvestment and outflow.” Conclusion Amazon needs to achieve AWS revenue growth and profitability if it wants to see a strong end to 2026. The holiday season presents an opportunity to better understand how Amazon adapts to agent-AI-driven consumer shopping habits. We like the Bedrock strategy and think it can certainly help Amazon win in the AI space in the long term. While robotics is a little further down the road, it’s hard to think of a business whose cost profile would be more positively impacted by the introduction of robotics than Amazon. As we look to both Amazon and the broader market beyond the end of the year, we should continue to see evidence that the hardware currently in operation actually has the longevity that Jassy talked about less than three months ago. We believe it will, and that’s why we’re adamant about our AI infrastructure deal with Amazon. (Jim Cramer’s charitable trusts are long: AMZN, NVDA, MSFT, META. See here for a complete list of stocks.) As a subscriber to Jim Cramer’s CNBC Investment Club, you will receive trade alerts before Jim makes a trade. After Jim sends a trade alert, he waits 45 minutes before buying or selling stocks in his charitable trust’s portfolio. If Jim talks about a stock on CNBC TV, he will issue a trade alert and then wait 72 hours before executing the trade. The above investment club information is subject to our Terms of Use and Privacy Policy, along with our disclaimer. No fiduciary duties or obligations exist or arise from your receipt of information provided in connection with the Investment Club. No specific results or benefits are guaranteed.
