Microsoft

Microsoft's fourth-quarter results topped the high end of guidance on key items. Revenue increased 17% year over year in constant currency to $90.0 billion, versus the high end of guidance of $87.8 billion, while operating margin was 45.1%, compared with the high end of guidance implied at 44.7%.

Why it matters: Results are good overall, as Azure growth and all three segments beat the high end of guidance. Further, the outlook is slightly ahead of our model for the first quarter. Critically, we see strength in Azure, in both traditional and artificial intelligence workloads, which is pulling along other AI solutions. Near-term demand indicators remain robust. Commercial bookings grew 18% year over year in constant currency, excluding OpenAI, and grew 11%, including OpenAI. Remaining performance obligation was up 84% to $678 billion, about 30% of which will be recognized in the next 12 months. Demand for Azure AI services is surging, which is a clear long-term positive. While Azure remains capacity-constrained, both traditional and AI workloads were strong. Azure growth was 43% in constant currency for the quarter, surpassing guidance of 39.5%, while capital expenditure grew 110%.

The bottom line: We keep our fair value estimate for wide-moat Microsoft at $600 per share, while raising our growth forecast and offsetting it with a margin decrease due to higher Azure capital expenditure. The stock remains one of our top picks.

Coming up: First-quarter guidance is slightly ahead of both FactSet consensus and our own estimates, and includes $90.4 billion in revenue, 48.5% implied operating margin, and $4.70 in implied EPS at the midpoints. Capital expenditure is guided to $50 billion.

Big picture: Results are consistent with our long-term thesis, which centers on the expansion of hybrid cloud environments, the proliferation of AI, and Azure. We center our growth estimates mainly around Azure, Microsoft 365 Copilot traction, and uptake of other AI solutions.

Guidance is positive overall. We think the Azure acceleration will be well received by investors, as will the margins, which indicate that the economics of building this massive fleet of data centers will produce attractive returns for the company. Importantly, management is changing the useful life of buildings from 15 years to 25 years, which will provide a minimal boost to margins in 2027. In the future, some leases will shift from finance to operating as a result of this, which should drive capex for calendar 2026 to $175 billion, from $190 billion previously. We do not view this as controversial. Microsoft seemed to thread the needle on capex, as capex acceleration is a double-edged sword for investors. Invest too much, and investors question future returns, but invest too little, and they question the longer-term competitive position. Microsoft is investing aggressively but has not changed its capex plans following Alphabet’s announcement last week. Importantly, CFO Amy Hood said the company will be free cash flow positive in fiscal 2027.

Intelligent cloud performance remains the key pillar of the long-term investment case, in our view, and it performed well this quarter. Overall, the segment accelerated to 31% year-over-year growth in constant currency. Microsoft cloud revenue increased 27% in constant currency to $59.3 billion. Azure remains the primary driver, growing 43% in constant currency, compared with guidance of 39%-40%. Azure growth accelerated both sequentially and year-over-year, and management expects further acceleration in the coming quarters, including 45% growth in the first quarter of fiscal 2027. Both AI and traditional workloads were good, with large customers continuing to sign large deals. Azure remains capacity-constrained, which we think will continue at least through fiscal 2027.

Management continues to describe AI usage, including various copilots throughout Microsoft’s portfolio, as enjoying strong uptake. The company provided some worthwhile data points, such as 30 million paid Copilot seats at attractive pricing (up from 20 million last quarter) and high Copilot usage. Copilot is seeing the same engagement as Outlook and Teams, which is very positive; and customers are also building and using agents within Foundry. Based on usage, growing demand, and surging forward-looking metrics like bookings, Microsoft continues to invest aggressively in AI. This remains the best strategy at this point, in our view. We have no immediate cause for concern, but we are definitely monitoring the $250 billion OpenAI commitment through 2032 as somewhat risky. To that end, management noted none of the dollars added to RPO in the quarter came from the AI labs. RPO was up 84% year over year to $678 billion, with $51 billion added during the quarter. The company stresses fungible data centers that can perform any function around the world as helping to mitigate risk if demand falls off quickly, which seems sensible. We still see Microsoft as well-positioned in AI, given the evolving OpenAI partnership, recently released proprietary models, internally developed silicon, and public cloud leadership.

In PBP, Microsoft commercial cloud delivered strong results, with overall segment trends similar to recent quarters. Segment revenue was up 14% year over year in constant currency. Commercial cloud grew 16% year over year on an adjusted basis in constant currency, while Dynamics 365 was up 12%, and LinkedIn up 10%. Overall, small and medium businesses, along with front-line workers, performed well but were adversely affected by pricing, while Copilot add-ons helped support per-seat pricing. M365 Copilot surpassed 30 million paid seats, indicating strong traction.

MPC remains under some pressure, even though this quarter's results were better than expected. Segment revenue was down 5% year over year in constant currency. Search, news, and Windows were solid, while gaming lagged, given a strong period last year from Microsoft-branded game launches. Computer original equipment manufacturer partners are building inventory based on rising component prices, which is helping, while year-ago periods are challenging due to the end of life for Windows 10. Gaming continues to struggle after the last round of price increases from last year, so the recent price cuts are helping to spur engagement again. Management stated gaming should grow in fiscal 2027.

Microsoft’s ability to deliver on the bottom line despite swelling Azure investments and other factors is impressive. We think this bodes well for margins as Azure capacity continues to come online, where we see some initial gross margin pressure. Although there surely is investor angst around the steady-state capex and Azure growth trade-off. Tight Azure capacity should help ensure the drag on margins is not too onerous, which it is obviously not thus far. GAAP operating margin was 45.1%, compared with 44.9% last year and the midpoint of guidance, implied at 44.1%, with upside driven by revenue strength, continued careful cost management, timing of outlays, internal efficiencies from AI use, and favorable product mix. A voluntary early retirement program and additional actions in gaming should provide margin levers in the face of gross margin pressure.

Our fair value estimate for Microsoft is $600 per share, which implies a fiscal 2027 enterprise value/sales multiple of 11 times and an adjusted price/earnings multiple of 30 times.

We model a five-year compound annual growth rate for revenue of approximately 16% inclusive of the Activision acquisition. We envision stronger revenue growth ahead as Microsoft's prior decade was bogged down by the downturn in 2008, the complete evaporation of mobile handset revenue from the disposal of the Nokia handset business, as well as the onset of the model transition to subscriptions (which initially results in slower revenue growth). However, we believe macro and currency factors will pressure revenue in the near term. We believe revenue growth will be driven by Azure, Office 365, Dynamics 365, LinkedIn, and emerging AI adoption. Azure, in particular, is the single most critical revenue driver over the next 10 years, in our view, as hybrid environments (where Microsoft excels) drive mass cloud adoption. We believe the combination of Azure, DBMS, Dynamics 365, and Office 365 will drive above-market growth as CIOs continue to consolidate vendors. We believe More Personal Computing will grow modestly above GDP over the next 10 years.

We also model operating margins to remain approximately steady near 47% in fiscal 2026 (actual) over the next five years driven by gross margin pressure from a mix shift to Azure offset by some operating leverage. We expect some interim pressure on both gross margin and operating margin in fiscal 2025 from an accounting change, Activision pressure, and investment in Azure capacity.

For Microsoft overall, we assign a wide economic moat, arising primarily from switching costs, with network effects and cost advantages as secondary moat sources. Based on the company’s segments, we believe the productivity and business processes and intelligent cloud segments have earned wide moats, and the more personal computing unit warrants a narrow moat. We believe Microsoft’s moat will probably allow the company to earn returns in excess of its cost of capital over the next 20 years.

We believe customers value Microsoft’s products as stand-alone solutions and for the company’s immense product breath, and these applications are tightly integrated with one another. In our opinion, the strength of these products is crucial but should not overshadow the importance of all the solutions being offered under one umbrella by Microsoft as customers are usually looking to consolidate vendors. These factors combine to reinforce our wide moat. As Microsoft offers a wider set of related and compelling solutions, we believe it becomes more deeply entrenched in its customers as they adopt multiple products.

Switching costs for software are often driven by several factors, in our view. The more critical the function and the more touch points across an organization a software vendor has, the higher the switching costs. There is also the direct time and expense of implementing a new software package for the customer while maintaining the existing platform and retraining employees on a new system. Additionally, there is operational risk of changing software vendors, including business process re-engineering, loss of data during the changeover, and overall project execution. A major implementation is likely to involve a system integrator and can take in excess of a year in bad cases. Also, lost productivity is likely to be an issue as customers move up a learning curve on the new system along with the distraction of users involved in the function where the change is occurring.

Retention metrics typically help inform investors on both the presence and the durability of a moat. These come in two types: gross, which describes what percentage of the customer base remains customers after a given period; and net, which highlights what percentage of the customer spending is retained by the software provider after a given period. Microsoft does not provide retention metrics because the user base has a large percentage of small business and consumer users, and the expansive breadth of the product portfolio makes an overarching retention rate less relevant. Software firms regularly see lower retention rates for small business users than for enterprise customers, so retention is weighed down by a large proportion of small business customers. We believe Microsoft’s customer retention for enterprise customers would be in the elite category if it was provided.

Productivity and Business Processes

We rate Microsoft’s productivity and business processes segment as having a wide moat based on switching costs and network effects. PBP represents approximately 30%=35% of total revenue and consists of Office 365, Dynamics 365, and LinkedIn.

The business world rotates around a variety of Microsoft products. Chief among them is the Office productivity suite, which completely dominates the market as a quasi-monopoly. Countless third-party applications have Excel or Office 365 plug-ins to ensure their software works with the popular suite. Given how embedded Office is into business users, we think it would be very difficult to switch providers. Further, even if a customer decided it wanted to switch to a different vendor, there is effectively only one meaningful competitor, which is Google’s Workplace. On this front, we think Workplace’s mandates have been relegated to universities, public schools, and startups. Even on the startup area, many Workplace users graduate to Office as those companies mature.

While Excel may be the export vehicle of choice for manipulation and analysis purposes, that financial data is exported from an ERP system, and Microsoft Dynamics is one such system. An ERP is the core system of record that midmarket businesses rely on to conduct business and monitor operations. Changing the workflow and business processes of any organization is difficult task, but changing an ERP system is more challenging still. In fact, we tend to view these foundational software packages as being the hardest to convert to another vendor, and therefore typically see the highest switching costs within software. We see the life of an ERP system as being a minimum of 10 years, often with a yearlong sales process and an implementation period that is likely to last months or even a full year at significant cost and disruption.

Office 365’s moat is supported by a network effect. A large installed base attracts software developers to create products specifically for Office. These add-ins make the office suite more compelling and therefore draw in additional subscribers. For example, in the financial community, a wide variety of add-ins for Excel, designed to smoothly integrate popular platforms such as PitchBook (a subsidiary of Morningstar), FactSet, Bloomberg, and CapitalIQ, have been created. These solutions increase the utility of Office 365, which attracts more subscribers, which in-turn attract more developers in a classic fly wheel. Similarly, LinkedIn, being the largest professional networking platform, also benefits from a network effect, where the large member base attracts more users. Here again, the market lacks a second viable competitor to LinkedIn, further insulating the company.

Intelligent Cloud

Microsoft's Intelligent Cloud segment includes Azure, OpenAI, Nuance, GitHub, Visual Studio, Microsoft Intelligent Data Platform, Microsoft Fabric, and Windows Server, SQL Data Base Management System. We assign the segment a wide moat rating based on high switching costs, network effects, and cost advantages. IC represents approximately 40% to 45% of total company revenue, with Azure representing 25% to 30% of total company revenue, or two thirds of the IC segment.

Azure represents the core of the Microsoft at this point in our view and builds upon the company’s overall Software as a Service model by offering Infrastructure as a Service and Platform as a Service. In an IaaS model, the provider offers the necessary hardware, virtualization, networking, and storage as a computing service delivered over the internet. Other basic software-level functionality can be layered in and still have the offering be considered IaaS. However, as software is added, IaaS quickly becomes PaaS. In PaaS model, the provider also offers and hosts operating systems, middleware, and core IT applications (notably database). We believe there are two public cloud leaders in Microsoft Azure and Amazon AWS, and a distant third but viable competitor in Google.

The secular trend for the past 15 years has been increasingly to outsource internal IT systems to the public cloud model, which is initially clearly cheaper than purchasing the equipment and software and paying staff to manage IT operations. Outsourcing of IT infrastructure is a strategic decision that is not undertaken with haste, and we think customer relationships will be long tailed with high retention. Indeed, enterprise customers often migrate one workload at a time, so even getting to a critical mass in the cloud is a yearslong process for them.

As with most software elements, we think customer retention will be higher for larger customers and lower for the smallest companies. However, once small businesses reach some level of scale, we think the retention rates increase meaningfully, because as these companies continue to grow, they benefit the most from outsourcing large portions of their IT infrastructure. It is simply cheaper and easier for small and medium-size companies to outsource.

Over the long term, especially for larger customers it is not likely to be cheaper to outsource all workloads to the cloud. In these cases, we expect larger customers to employ a hybrid model, with some workloads in the cloud while others reside in the corporate data center. Here again Microsoft shines, as they have been serving enterprise customers for decades already and it has an unrivaled structural advantage in its ability to move clients from an on-premises Microsoft environment to a Microsoft environment in the cloud via Azure.

Also, the early lead and substantial market share inherent in Azure and the underlying core services has led to a wide variety of developers joining the ecosystem bringing in applications, middleware, and development tools. The huge and increasing variety of supported software available then helps attract new users, which continues to draw in more developers in a classic flywheel. The sheer size of this ecosystem has helped Azure and Microsoft Server become the path of least resistance for CIOs and IT managers.

More Personal Computing

Microsoft's More Personal Computing" segment includes Windows, Gaming, Devices, and Search. We assign the segment a narrow moat rating based on switching costs and network effects. MPC represents approximately 25% of total revenue.

Microsoft Windows is the product that put Microsoft on the map and clearly has high switching costs and a network effect associated with it to this day. Entire businesses are built around the Windows environment, which itself was just a more usable version of Microsoft’s text-based operating system, DOS. Microsoft basically has a 50-year history of powering the computer revolution with its usability, reliability, ecosystem, software support, and product road map. While changing a given application, painful as it might be, is possible, it is not feasible to change operating systems from a practical standpoint. Software applications are overwhelmingly created for Windows and its quasi monopoly in the personal computer operating system market. This could possibly be the largest network effect within software where the influx of new users, especially at the dawn of the computer age, ushered in a plethora of software developers, which in-turn made PCs more usable. Outside of Apple, there is no credible operating system for personal computers.

We think the acquisition of Activision finally establishes a narrow moat for Microsoft’s gaming segment. High switching costs are explicit in the hardware platform, while acquisitions have built a game development empire, with strong intellectual property that underscores an intangible asset moat source. Activision bolsters Microsoft’s already strong gaming division with an iconic library that includes traditional console game developer Activision, PC developer Blizzard, and mobile developer King. Microsoft had minimal exposure to mobile, the largest gaming platform, so we think King and its Candy Crush franchise instantly provided mobile chops. Activision has some of the most popular games of all time under its umbrella, including the Call of Duty franchise, Diablo, and World of Warcraft, which changed the gaming industry in 2004 with not only its multiplayer game, but also its monthly subscription model.

We expect Microsoft to rapidly add Activision content to Game Pass, which we estimate has about 25 million subscribers. Activision has approximately 360 million monthly active users. We tend to think the purpose of deals like this is to gather exclusive content. However, part of the value of Activision is that it develops games across platforms and consoles, so we expect a relatively light slate of exclusive Xbox content over the next several years.

We do not believe the Microsoft Surface enjoys a moat of any kind. We do not believe that there is much, if any, long-lived differentiation between a Microsoft Surface and generally competitive products such as a Windows or Mac laptop, an Apple iPad, or a Google Chromebook. This is generally reflected in low or lower product gross margins. Other devices within this category include computer peripherals (keyboards, mice, and so on) and the HoloLens. On the HoloLens in particular, we remain skeptical on virtual and augmented reality within the next several years. Microsoft has not enjoyed much success in noncore devices in recent years, with the high-profile Zune and large acquisition of Nokia serving as reminders that hardware is a challenging business. Finally, we also do not believe there is a durable moat in Microsoft’s Bing search engine.

Public cloud remains the present and future of enterprise computing, and Azure is a leading full stack service for AI and more traditional public cloud workloads.

Microsoft 365 continues to benefit from upselling into higher-priced stock-keeping units as customers are willing to pay up for better security and Teams Phone, which should continue over the next several years.

Microsoft has monopoly like positions in various areas (OS, Office) that serve as cash cows to help drive Azure growth.

Momentum has slowed in the ongoing shift to subscriptions, particularly in Office, which is generally considered a mature product.

Investments in new data center builds based on AI-related demand are massive and may drag down returns over time.

Microsoft lacks a meaningful mobile presence.