Microsoft: Something Doesn’t Add Up

Microsoft Corporation (MSFT) reported a solid fiscal first quarter in 2026, beating both top and bottom line expectations by 3% and 12%, respectively. But, despite the strong momentum, Microsoft’s stock has dropped 15% since the earnings release. This logic does not add up. Most of the setback has stemmed from fears about AI spending and sustainability, since Microsoft is one of those that are betting heavily on AI infrastructure.

Even so, the fundamentals have not changed, and there are still several factors that point to continued momentum in 2026. In the most recent quarter, Microsoft’s revenue rose 18% to $77.7 billion, driven by its Intelligent Cloud segment, which contributed $30.9 billion alone to the top line. On that subject, this segment’s revenue increased 28% compared to the same period last year.

Within that sphere, Azure grew 40% in constant currency, with AI services delivering 18% points to that growth. In terms of profitability, gross profit rose 18% to $53.6 billion, while gross margin stayed relatively flat at 69%. That is impressive considering that Microsoft is building data centers at an unusually high pace. Meanwhile, operating income climbed 24% to $37.9 billion, and net income rose 12% to $27.7 billion, or $3.73 per share.

That said, the bottom line figures are a tad lower as Microsoft took a $3.1 billion charge related to its OpenAI investment.

Looking at the balance sheet, Microsoft ended the quarter with $102 billion in cash and short-term investments against a long-term debt of $35.4 billion. Zooming out, Microsoft’s total assets reached $636.3 billion with total liabilities of $270 billion. That is a healthy financial position. Speaking of cash, Microsoft’s cash flow paints an even brighter picture.

Operating cash flow for the quarter increased 32% to $45 billion, while free cash flow jumped 33% to $25.7 billion. So despite record spending on infrastructure, cash generation is actually accelerating. For second quarter, management is guiding revenue to come in between $79.5 and $80.6 billion, up between 14% and 16% year-on-year. Also, Intelligent Cloud revenue is expected to come in between $32.25 and $32.55 billion, or 27% higher year-over-year.

Meanwhile, Azure is expected to grow 37% in constant currency. If any of this materializes, it should more than help clear up uncertainties around AI computing demand. Of course, CapEx is also expected to increase from here, exceeding that of fiscal 2025 growth rates, and more importantly, computing capacity remains limited through at least the end of the fiscal year.

Microsoft stock currently trades at $460 per share.

Over the past six months, MSFT stock is down roughly 9%, while it is up 8% on a one-year scale. Zoom out five years, and the stock is up 113%, mainly driven by its cloud infrastructure business. In terms of valuation, Microsoft trades at a FWD price-to-book ratio of roughly 8x, which is about 60% higher than the sector median. Microsoft’s real value comes from its relationships with practically every single Fortune 500 company and its dominant position in the AI market, so this premium makes sense.

However, Microsoft’s FWD P/E ratio (GAAP) sits at 29x, which is around 8% lower than the sector median—despite being a notably dominant player in AI and cloud computing. At this point, Microsoft is the cheapest in the Magnificent 7 lineup, considering traditional valuation metrics. This 8% discount reflects short-term worries around AI spending, not the fundamentals.

There is really nothing to worry about Microsoft’s current standing. Moreover, Microsoft has been building AI infrastructure at an incredible pace. Once these data centers fill up, there will be a clear path to margin expansion. On that subject, Microsoft’s current levels might be a good entry point for investors with a five-year or longer horizon. In the short term, it is certainly possible Microsoft could continue to decline as the market reacts to AI spending news or Bubble Bursting rumors.

But again, this is very short term.

Several factors could boost Microsoft’s trajectory over the next 12 months, each with the potential to bring in billions in revenue, though they will play out at different times. Azure AI has grown by 40% compared to the same period last year, but this figure actually understates demand. Microsoft’s commercial backlog increased 51% to $392 billion, these are signed contracts with customers who have committed to spending on Azure.

On average, the contract duration is about two years, meaning this backlog will turn into actual revenue in a relatively short period of time. Management’s response to this strong demand is doubling its data center footprint over the next two years. One of these projects is the Fairwater campus in Wisconsin, which is expected to have 2GW of power capacity by itself.

Across everything Microsoft is building, the company plans to add more than 80% AI capacity this fiscal year alone, and the total installed base is doubling by fiscal 2028. Azure is growing at 40% now, but that pace will eventually slow as the law of large numbers kicks in. A more reasonable rate might be 25% annually as the business matures further.

At 25% annual growth from today’s levels, reaching $1.54 trillion in revenue would take about 6 to 7 years.

There is potentially a path to 5x net income by 2031, assuming Microsoft keeps its margins steady as the infrastructure investments pay off. Another potential growth driver is the capacity expansion itself. In the first fiscal quarter alone, capital spending reached $34.9 billion, which puts Microsoft on track for nearly $140 billion in annual spending.

All that money is going towards bringing new data centers online quarter after quarter. Azure revenue would be higher if Microsoft just had enough capacity to serve everyone in line. These constraints are holding back Azure growth by a notable margin. A modest 5 to 8% bottleneck would translate to $4 to $6 billion in quarterly revenue sitting on the table because Microsoft literally cannot cater to everyone.

These figures are pretty conservative considering how much Microsoft is building to address the demand. Throughout fiscal 2026 and into 2027, new data centers are expected to come online, so all that pent-up demand should start converting into actual revenue. More importantly, Nvidia’s GB300 GPU clusters will be deployed in these new data centers, and this will be Microsoft’s first large-scale deployment of the said GPUs.

These new chips deliver about 30% more performance per chip than older models, which means Microsoft can serve more customer work for every dollar spent in these data centers.

If everything goes right, this expansion could easily add 2 to 3% to annual revenue growth through 2026 and beyond as supply limitations ease. From 2027 to 2028, operating margins may expand by up to 2% per year from these investments alone and contribute significantly to the $138.5 billion quarterly net income target. Another potential growth opportunity for Microsoft is its restructured deal with OpenAI, it might even be its most important strategic move in the AI business.

For starters, Microsoft owns 27% of OpenAI at a valuation of $500 billion. That is nearly a 10x return on the $13.8 billion that Microsoft invested in OpenAI. Now, OpenAI has committed to buying $250 billion worth of Azure services on top of what they are already spending. The agreement gives Microsoft exclusive commercial licensing rights to OpenAI’s technology through 2032 or at least until AGI, or Artificial General Intelligence, is verified.

AGI is basically AI that can do intellectual tasks as well as a human can. That said, AGI is still theoretical. Microsoft’s exclusivity means that any company that wants to build apps using OpenAI models must run them on Azure. So, Microsoft pretty much controls the distribution channel for the world’s top AI models so far. That $250 billion commitment will likely be spread over 5 to 7 years, working out to about $35 to $50 billion in extra revenue from Azure once it is fully ramped up.

Azure’s current gross margins stand at 70%, so that would translate to $24.5 to $35 billion in annual gross profit.

As for the bottom line, that would be $50 billion in added Azure revenue per year if that $250 billion spreads over 5 years. And, at 68% gross margins, that is roughly $34 billion in extra gross profit. Microsoft is at a turning point where short-term challenges have overshadowed a strong long-term opportunity. Current fears around AI spending and sustainability, Microsoft’s capacity constraints, and limited movement in gross margins have all contributed to the stocks’ 15% decline since earnings, but most of these are fixable issues or non-fundamental setbacks.

In 2020, Microsoft might be known as a business software giant, powering most of the world’s PCs and building backroom technology and tools. But Microsoft had also spent more than $10 billion buying development studios behind some of the most popular video games in the world. To Microsoft, it is all about the future of software. Microsoft has made big bets on the video game industry, buying “Minecraft” maker Mojang for $2.5 billion in 2014.

Then Microsoft bought five more studios in 2018, including role-playing game maker Obsidian, known for the space adventure “The Outer Worlds” and the well-received “South Park: The Stick of Truth”. In 2019, Microsoft bought Double Fine, maker of adventure game “Psychonauts”. Interactive entertainment will be a key technology in the next 10 years and that gamers who use Microsoft products expect the company to make titles like those made by the studios it has bought.

Microsoft announced its $7.5 billion cash purchase of ZeniMax Media, which owns several industry-leading game developers, including Bethesda Softworks and iD Software.

The idea of having content was so Microsoft could reach larger communities. That was why Microsoft would consider buying even more video game companies, and why the company continued to invest in its Xbox Game Pass subscription service. Content was just the incredible ingredient to Microsoft’s platform that the company continued to invest in. That doubled the size of Microsoft’s creative organization.

Microsoft’s purchases marked the most dramatic ways the company was looking to build up its Xbox brand. The purchases also gave Microsoft more games to field in what was increasingly become a hits-driven business. Competitor Sony was known for a range of homegrown series, from the post-apocalyptic thriller “The Last of Us” to the treasure-hunting game “Uncharted” to the action series “God of War”.

The Xbox team, meanwhile, was best known for its “Halo” and “Gears of War” space war series, and its “Forza” racing games. Microsoft did not need another major content/IP acquisition. While Microsoft might be home to some big hit titles, many gamers had said the company needed to up its game, particularly when compared with Sony. Gamers saw a noticeable difference with Bethesda.

Bethesda’s games was offered on Microsoft’s Xbox Game Pass subscription service the same time they hit store shelves.

The games was also available through Microsoft’s xCloud video game streaming service, which allowed people to play games over the internet similar to the way they streamed movies from Netflix. That was a huge investment in games that people were going to get to play. Bethesda ran semi-independently, in an effort to keep the company building the games that brought it success in the first place.

It was about the culture of the teams. It was not about Bethesda becoming Microsoft. It was also why the game companies would continue to be on Microsoft’s radar. Microsoft always looked for places where there was that commonality of purpose, mission, and culture. The Xbox team worked with ZeniMax companies including Bethesda since the first Xbox was released in 2001.

Microsoft always looked to grow inorganically where it makes sense. Bethesda was not the only consumer product acquisition Microsoft had been involved in over the past. Microsoft also made a bid to buy TikTok, the rapidly growing China-based social media network built on memes and short videos. TikTok, which had been downloaded more than 2 billion times around the world and counts more than 100 million users in the United States, was at the center of a political battle between the Chinese government and President Donald Trump.

TikTok’s parent company, ByteDance, approached Microsoft to attempt to create a company that would be fully owned by the company.

Over the course of the few weeks, Microsoft was part of a shifting public acquisition process, dictated at times by Trump, who threatened to ban TikTok from United States app stores over undefined national security concerns. Trump said he would not ban the app if it is purchased by a United States company, and for a while it seemed Microsoft would be the one. Though, Trump announced that he has approved a deal in which it appeared business software maker Oracle and retailer Walmart would take stakes in a new company, called TikTok Global, though some uncertainty remains.

If Microsoft had won the acquisition, it would have been part of the company’s efforts to expand its software. The company finally confirmed the existence of the Xbox Series S on September 6th 2020. One day later on September 7th 2020, the company announced the hardware specifications of its smaller, cheaper next-generation console. At $299, the Xbox Series S had a great value proposition for people who did not care about physical games or 4K resolution.

The Xbox Series S was built to play all the same next-gen games as its more powerful sibling, the Series X, except at a lower target resolution of 1440p rather than 4K. Both consoles could play games at frame rates of up to 120 frames per second. And the Xbox Series S still delivered the same visual fidelity as the Series X. The Xbox Series S was capable of hardware-accelerated real-time ray tracing, variable-rate shading, and mesh shaders.

By designing two consoles in parallel from the very beginning, Microsoft could deliver the same core gaming experience while also making it as easy as possible for developers to scale their games across both consoles with minimal effort.

That meant that the Xbox Series S delivered the same incredible next-generation experience and features as the Series X, just at a reduced rendering resolution. The Xbox Series S delivered 4 teraflops of graphics performance, which meant it was 33% less powerful than the One X and 67% less powerful than the Series X. But the Xbox Series S managed to deliver the same performance as the Series X because of the way Microsoft built it.

In terms of the specifications, the Xbox Series S contained the same eight-core Zen 2 CPU architecture as the Series X. Except down-clocked slightly to run at a constant frequency of 3.6 GHz instead of 3.8 GHz. With simultaneous multithreading enabled, the Xbox Series S CPU ran at 3.4 GHz instead of the Series X’s 3.6 GHz. The Xbox Series S GPU contained far fewer compute units, 20 instead of 52, and ran at a slower frequency of 1.565 GHz compared to the Series X’s 1.825 GHz.

The lower raw numbers would not hold back the Series S too much because of the virtual memory multipliers provided by the Xbox Velocity Architecture. The Xbox Series S system memory was scaled in line with the target resolution of 1440p. The Xbox Series S boasted 10 GB of GDDR6 RAM instead of the Series X’s 16 GB, with less memory bandwidth. Another major difference was the internal storage.

While both consoles delivered the same storage bandwidth, the read speed is 2.4 Gbps for uncompressed data and 4.8 Gbps for compressed data.

The SSD in the Xbox Series S was half the size of the one in the Series X, 512 GB instead of 1 TB. That could be a sticking point for the all-digital Xbox Series S, with storage requirements for many modern games creeping up near the 100 GB mark, if not topping it.