Only a couple of months ago, Amazon’s comeback rally looked unstoppable — until it was. With a huge post-earnings boost, Amazon shares closed at a record high of $284 each on Aug. 3 — an increase of nearly 45% from its 52-week low back in February. It looked like $300 and beyond was just a formality. Since then, however, Amazon has dropped more than 7%, with an up-and-down chart over the past month looking more like seismic activity than stock trading. So, what the heck happened? And what’s it going to take to get shares going again? Analysts at Morgan Stanley gave us some food for thought in a note this week, making their case for why they added Amazon to their top picks list and boosted their price target to $350 from $335, implying 33% upside from Monday’s close. Before getting into what we need to see from Amazon in the weeks and months ahead and Morgan Stanley’s reasons for being so positive, let’s do a quick recap of how we got here. How we got here Last quarter was strong, with the company’s Amazon Web Services (AWS) cloud business crushing sales and operating income expectations. However, what really lit up the stock was CEO Andy Jassy’s commentary on the three-decade usable life of a data center and the profit dynamics of the hardware that goes into them. He said on the second-quarter call: “The servers currently have a useful life of at least five to six years, and most of our AI capacity these days is being contracted for at least five-year terms. That means that we’re driving significant free cash flow on the servers and network of equipment in the two to three years after we break even.” In our earnings analysis at the time, Club portfolio director Jeff Marks wrote that Jassy “delivered exactly what investors like us wanted to hear: a clear explanation of how its massive investment in artificial intelligence will generate returns.” The CEO’s remarks on the post-earnings July 30 conference call were so market-moving and fundamentally important that we dedicated half of the next day’s Homestretch to it. The stock soared more than 15% on July 31 and another 4.5% on Aug. 3 to the aforementioned record-high close. AMZN YTD mountain Amazon YTD Jassy’s words reverberated far beyond just Amazon and are very much at the heart of our broader view on the current market. We think that the high-quality, beaten-down parts of the AI hardware trade are presenting buying opportunities. Jim Cramer made his argument in his Sunday column . Here’s a key line from Jim: “Nvidia’s chips hold up. They may hold their value for 10 years.” Amazon uses a lot of its own custom chips, but it also has a lot of Nvidia chips in its cloud. Jassy said during the Q2 call that won’t change. “We also continue to have a deep partnership with Nvidia and will continue making AWS the best place to run Nvidia chips, as we have customers who will run on Nvidia for as long as we can foresee, and we believe strongly that customers want choice. Choice is good for customers, competition, and driving the cost of inference down, which customers care deeply about.” Where we need to go Ahead of Amazon’s third-quarter earnings next week, set for after the closing bell on Oct. 29, we’re taking a moment to reflect on our big position and what we think Amazon must do to get shares moving again. As of Tuesday, Amazon is our largest-weighted holding at 5.3%. We have a buy-equivalent 1 rating and a $300-per-share price target on the stock. During last week’s October Monthly Meeting , Jim said he has tremendous confidence in Jassy and CFO Brian Olsavsky, adding that investors who don’t own any Amazon here should buy some. Jim added that Amazon’s Q3 conference call will be key, arguing that if Jassy can demonstrate that AI investments are paying off and that capital expenditure growth is beginning to moderate, shares could really take off. Digging in on Morgan Stanley’s bullish note, the analysts laid out a five-point checklist on what they think is needed for investors to bid up the stock: (1) Strong AWS revenue growth, (2) better-than-expected operating margin and profit results from AWS, (3) further proof of “Bedrock’s importance and strategic value,” (4) proof that Amazon can win in agentic AI, and (5) investor confidence that Amazon stands to benefit in the coming age of physical AI (robotics). Let’s unpack those one at a time and provide our views. AWS revenue growth: In the near-term, it’s all about AWS because the cloud is where we find the artificial intelligence. “AWS’ growth and profit runway is under-appreciated,” Morgan Stanley said. In our view, there is no denying that demand is out there for compute, as more companies learn to leverage AI, realize the benefits, and as a result, demand even more. So, the only question in our mind is: Can Amazon continue to win new business at the pace demanded by Wall Street? We think it can and would remind doubters that Jassy may not be an Amazon founder, but he headed the launch of AWS — four years before it would see a true competitor in the market in Microsoft’s Azure. So, when it comes to winning new clients in the cloud, Jassy is probably the last person to consider betting against. Even if you buy into the revenue growth, you have to keep in mind that we value stocks, preferably, on forward earnings per share (EPS) estimates, not sales. AWS profitability: AWS is the primary revenue beneficiary of artificial intelligence demand, but it’s also the source of AI-related investments. Weighing AI investments versus the ability to monetize that spending, AWS’ operating margin, even more than revenue growth, is likely the primary determinant of where shares go in the near-term. “Our bottom-up AWS expense build gives us confidence AWS can operate at high 40s% incremental EBIT margins in ’27/’28,” Morgan Stanley said. In other words, the analysts think that Amazon can generate operating margins of over 40% on new AWS business generated in 2027 and 2028. According to the FactSet consensus estimate, overall operating margin at AWS in 2027 is expected to be 37.5%. We think the more important consideration for investors is whether they believe in the usable lifecycle commentary from Jassy. If you don’t, then your position really is reliant on the next quarter. If you do, as we do, then you must understand that profitability may not be as superb as the Street demands every quarter. Longer-term, however, it will grow as the fixed costs of standing up the data center shell (the buildings the chips go into) wind down and the mix of data center expenses shifts more to variable chip costs driven by near-term graphics processing unit (GPU) demand, with Amazon only buying more as they see the demand and structuring customer contracts to ensure profitability. This gets back to the heart of the matter: We see no change in the dynamics Jassy commented on during the last earnings call and, therefore, expect AWS profitability to be strong in the quarters to come. Bedrock: Generative and agentic AI solutions represent key drivers for AWS, and we think Amazon is going about it the right way. Rather than piling billions of dollars into the frontier model arms race, Amazon’s focus has been on AI solutions that enhance its own, proven, money-making offerings. “We continue to believe AMZN’s ability to showcase the strength of Bedrock through this age of inference (and its importance to driving GenAI adoption across the economy) is under-appreciated and a potential multiple expander from current 15X ’28 EPS levels,” Morgan Stanley argued. According to FactSet, the consensus forecast puts a multiple of 18.4 times on 2028 earnings per share estimates. On the consumer front, it is 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.” While the platform provides everything an organization needs to build, it is model agnostic, offering up hundreds of models for the customer to choose from based on their unique needs. We think that the right move for a company that rents out more compute than anyone else in the world. Who cares which model clients use, so long as they burn AI tokens (the unit costs of computing power) and demand the compute Amazon is selling. In fact, in support of our view that this is the right strategy for Amazon, we would call out recent action at Microsoft, which saw a positive reaction to the company’s revamped Copilot strategy, focusing more on the integrated user experience rather than being the developer of the world’s leading AI models. Agentic AI: Amazon’s model-agnostic strategy with Bedrock seems solid. However, the company’s retail segment is a bit more of a question mark, given the company has taken steps to block agentic AI solutions, such as Meta Platforms ‘ Muse, from accessing the marketplace. Since major competitors including Walmart , Wayfair , Best Buy , and others have opted to play ball with Muse, we suspect there is more to come regarding Amazon’s plans to adapt to changing consumer shopping habits. After all, a relentless focus on the consumer experience is how Amazon became the online retail leader it is today. Morgan Stanley theorized, “While we don’t know exactly how the retail media business will be protected through agentic adoption, we expect to learn more from templates of other connections. … Some potential mitigants may be a hand-off from Muse to Alexa for Shopping (with AMZN controlling user experience/inventory); perhaps it’s a take rate adjustment, perhaps it’s a revenue share, … but we expect AMZN to leverage its leading e-commerce infrastructure and loyal user base to continue to monetize.” Physical AI (robots): On a longer-term time horizon, the benefits of AI to Amazon’s retail operation become a bit easier to see, at least on the logistics front . Here, we’re talking about physical AI. From warehouse robotics to self-driving vehicles, physical AI is likely to have a massively positive impact on how packages get to our door. Morgan Stanley pointed to its study from February 2025, in which they said that every 10% of global units moving through Amazon’s robotics-enabled fulfillment centers could generate $2 billion to $4 billion of annual recurring savings. Today, the analysts think the potential savings are increasing, thanks to efficiency gains. They think that ramping up usage of robotics in Amazon fulfillment centers will be a material catalyst next year. “The extent to which we see AMZN’s robotics fulfillment centers ramp (new builds and retrofits) throughout ’27 could showcase how AMZN is a physical AI winner as well … with the technology delivering more earning power to re-invest or flow through.” Bottom line If we want to see a strong close to 2026, Amazon needs to deliver on AWS revenue growth and profitability. The holiday season brings an opportunity to better understand how Amazon will adapt to agentic AI-oriented consumer shopping habits. We like the Bedrock strategy and think it will indeed help Amazon win on AI long-term. While robotics are a bit further off, we struggle to think of a business that could see a more positive impact on its cost profile from the implementation of robotics than Amazon. Looking beyond year-end, both for Amazon and the market more broadly, we need to continue seeing evidence that the hardware being stood up now really does have that long lifespan that Jassy spoke about less than three months ago. We believe we will, which is why we’re sticking with Amazon and the AI infrastructure trade. (Jim Cramer’s Charitable Trust is long AMZN, NVDA, MSFT, META. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
