Microsoft Earnings Analysis: FY26 Q1 is All About a Fungible, Planet-Scale Fleet for the AI Bubble ⭐

On Wednesday, Microsoft reported that it earned a net income of $30.8 billion on revenues of $77.7 billion. But its spending on AI, $34.9 billion, was dramatically higher than ever before and it outpaced the company’s profits by a wide margin. And so when I reviewed the post-earnings conference call and studied the company’s earnings call slides, I quickly realized that this analysis would go in an entirely different direction.

That is, instead of looking at Microsoft’s individual businesses, or just focusing on the consumer businesses you and I care about the most, I’m going to look instead at what I feel is the single biggest issue today with Big Tech.

No, it’s not the massive market power that these firms wield, per se. But rather that once these companies achieve a certain scale and become more financially powerful than most countries, their priorities change. And the Microsoft of today is not Satya Nadella’s Microsoft, it’s Amy Hood’s. This is a Microsoft whose primary business is manipulating money. And it is incredibly successful at that, all while harming the products and services it builds, the partners it does business with, and the customers who use those offerings. This is, in short, the enshittification of the U.S. corporation and the U.S. financial system.

You know. A fun topic for a casual read. Sorry.

An utter lack of transparency

One might argue that this is Microsoft not reading the room: I have been complaining about the ongoing transparency reductions in its disclosures to shareholders, investors, and Wall Street every quarter for many years, and this past week, The Wall Street Journal finally took notice, albeit only with regard to the company’s partnership with OpenAI. But that’s not what this is. Instead, we should see this as just the next obvious step in Microsoft seeing what it can get away with in a world in which Securities and Exchange Commission (SEC) oversight is literally nonexistent and all of Wall Street is complicit in not calling on Microsoft and other Big Tech firms to meet their legal requirements.

The changes will seem subtle if you don’t follow Microsoft’s earnings as closely as I do. Microsoft updated the template it uses for its PowerPoint-based earnings call slides, which arguably does look more professional but also distracts just enough to trigger a bit of confusion.

At first, I thought I was missing something. But then I realized what it was. Data. The presentation is the same length, 22 pages, as before. It has the same four top-level sections. The fonts are smaller, so there are some screens that actually have more information—clutter, really—on them in a classic example of poor presentation skills. But then I saw it. The Business Highlights page that lists the top-level financial points for each of Microsoft’s three business units is gone. And the useful, non-clutter information provided for each business unit is down.

For example, we previously got nine data points for Productivity and Business Processes, and now there are just five. Missing in this quarter’s rundown are overall Microsoft 365 Commercial products and cloud services revenue (which was just a percentage change anyway, like most Microsoft data points), Microsoft 365 Consumer products and cloud services revenues (ditto), LinkedIn sessions (ditto; a year ago, this was “record engagement”), and Dynamics products and cloud services revenue (with a breakdown of cloud vs. on-premises products). And that’s just one slide.

It’s possible and maybe even likely that this data can be found somewhere else, but who has time to page through a complex Excel spreadsheet? And what would one find even if that data were there? Microsoft has never reported the actual revenues generated by Microsoft 365 products and cloud services in Commercial or Consumer and it never will. A percentage change just vaguely shows momentum since we never had a hard number to begin with. These useful numbers are hidden, and on purpose.

What does Wall Street ask Microsoft?

It’s not just the SEC that’s not doing its job. Wall Street analysts have turned our financial system into state-sponsored gambling, and they’re playing with the money Americans hope to use in retirement. So instead of hard questions about the data Microsoft omits, we get a lot of sucking up.

Let’s step through this mess.

Why isn’t Microsoft’s stock price even higher?

“Congratulations on another outstanding quarter,” Morgan Stanley’s Keith Weiss said, kicking off the Q&A in standard form. “If I’m looking at Microsoft, this is two quarters in a row, we’re really seeing results that are well ahead of anybody’s expectations when we were thinking about this company a year ago or five years ago. 111% commercial bookings growth was not on anybody’s bingo card, if you will. Yet, the stock is underperforming the broader market.”

Astonishing. So this guy’s only criticism is that Microsoft’s stock price, which soared Wednesday thanks to these earnings, briefly catapulting the software giant’s market cap above $4 trillion, is somehow underperforming. And to be clear, he’s not criticizing Microsoft here, he’s criticizing Wall Street for not rewarding Microsoft enough.

And I know what you’re wondering. Surely this guy asks a question at some point. And he does, in a rambling way. He asks whether AGI—artificial general intelligence, the ultimate goal of OpenAI and, presumably, other AI firms—is what’s “weighing” on Microsoft’s stock price right now ($529 a share, up from $432 exactly one year prior). And whether Microsoft achieving AGI might be that thing that triggers the next explosive growth in the stock price.

Seriously.

Satya Nadella replied that this guy’s question “touches on something that’s pretty important,” where I might argue that it instead escaped from the gravity of anything important and went spinning off into space. But then, his full answer was even less clear and more rambling than the question that inspired it.

“Excellent,” Weiss responded after Nadella stopped babbling and finally said that AGI wouldn’t happen anytime soon. “That’s super helpful.”

These are the moments that try a man’s soul. At least my soul. I didn’t find any of that helpful at all.

But let’s move on.

Will Microsoft ever collect the money customers have allegedly promised it?

“On the bookings blowout, many are somewhat concerned about concentration risk,” Jefferies’ Brent Hill asked next. “Amy noted a number of $100 million contracts. And not to go into a lot of detail, but can you just give a sense of what you’re seeing in that 51% RPO and 110%-plus bookings growth that gives you confidence about what you’re seeing in terms of the breadth and extent of some of these deals on a global basis.”

Two points here.

One, he literally just asked the CFO of Microsoft to “not go into a lot of detail” when replying to a question about Microsoft’s finances.

Two, RPO is Remaining Performance Obligation, a measure of revenues a company expects to book in the future from existing customer contracts that have not yet been completely fulfilled.

Two, he’s referring to a comment that Hood made earlier in the call about Microsoft now having $400 billion worth of commercial RPO, an increase of 50 percent YOY, and with a “weighted average duration of only two years.” Because of this growth in promised spending by its customers, Microsoft accelerated its spending on datacenter CPUs and GPUs in the quarter, and it now expects its capital expenditure spending in FY26 to be higher than it was in FY25. RPO is one of several “demand signals” that Microsoft uses to determine what it needs to spend on building out its AI datacenters, basically.

That $400 billion in expected revenues, which will occur over a period of longer than two years, spans “numerous products” and involves companies of all sizes, including, go figure, OpenAI. “These are contracts being signed by customers who intend to use it in relatively short order,” she finally says. Which is not really two years, but is also apparently not several years or more either.

Are we in an AI bubble?

“Congratulations on the quarter,” Bernstein’s Mark Moerdler says, getting the suckling back on track. “It’s pretty amazing what you guys are doing.”

To his credit, Moerdler asks a real question: “How much confidence do you have that the software, even the consumer internet business can monetize all the investments we’re seeing globally, or, frankly, are we in a bubble?”

If you had “bubble” on your Microsoft earnings Bingo card, you can check that one off.

Anyway, Amy Hood jumps right in before Nadella can say anything she’ll regret and downplays the risks to Microsoft’s mammoth AI spending. Microsoft has 15 to 20-year leases on datacenter sites. Demand is increasing, and in many places. New products launch and customers use them. “When people see real value, they actually commit real usage.” “When you see these demand signals … we do need to spend.” In short, she feels good about it.

Finally given a moment to provide his own answer, the CEO of Microsoft tells Moerdler that only two things matter: Microsoft’s “planet scale token factory” and the continual modernization of Microsoft’s datacenter “fleet.”

And if those seem like nonsense terms to you, welcome to the dance. But these terms came up repeatedly in this call, I believe for the first time. Nadella had said “planet scale” twice in his prepared remarks, and he mentioned the fleet three times. But this question caused him to say “fleet” seven more times. That fleet is “fungible,” he says. It’s been modernized and made more efficient. And for the love of God, we’re just going to beat that term to death, I guess.

But Nadella casually says something very concerning during what turns out to be a very long-winded answer that, like Hood’s, never addresses the term “bubble.” After noting that the cloud was “so much more expansive” than the server market and stating that AI would be the same when compared to the cloud, he glosses over tools (coding), security, and health, and … then this.

“And in consumer, one of the things is it’s not just about ads, it’s ads plus subscriptions. That also opens up opportunity for us.”

So there you go. There’s no AI bubble.

“That was pretty amazing,” Moerdler responds. “I really appreciate all the details.”

I agree with the first sentiment: That was pretty amazing.

But seriously, how much money have you flushed down the OpenAI toilet?

The next question, from Karl Keirstead of UBS, gets off to a pretty amazing start. He literally prefaces his question by saying that it’s “for Amy.” And then he, too, asks the CFO of Microsoft to not “go down too complex an accounting path with this question.” Yeah, don’t expect details from the person in charge of the company’s finances during an earnings conference call.

“The investment in OpenAI that sits at another income at $4.1 billion is so large that I think the audience listening in could benefit from a little bit more color about what that is,” he asks. “It feels like it’s so much larger than you were running through other income in prior quarters that it mustn’t just be your share of the OpenAI losses. Could you just describe that and what we can expect in subsequent quarters, and whether this signals any kind of accounting change? Thanks so much.”

I couldn’t find a reference to $4.1 billion, but that’s OK. Amy never answers the question.

“The Q1 number was not impacted at all by the new agreement that was put in place,” she says. “Secondly, that increased loss was all due to our percentage of losses and OpenAI equity method, just to be very clear. There is not anything there that is not the increased losses from OpenAI.”

So I guess the losses from OpenAI are … increasing? Is that why Microsoft is building out more datacenter capacity to the tune of $35 billion in this quarter alone? Is this a “signal”? Is there a single signal that points to Microsoft recouping its investments in OpenAI and AI in general anytime soon, if ever? No one will ever answer those questions.

Properly admonished, Keirstead responds, “Understood. Thank you.”

What is the economic theory called when companies with no money have promised to spend a lot of money?

Mark Murphy from JP Morgan is next. And to his credit, he does talk around an important question in a convoluted way that I will restate in plain English: A small number of “AI natives” (companies) are committing to spend many hundreds of billions of dollars that none of them actually have. How does Microsoft determine that it will ever actually get the money these companies have promised to spend?

I love that Nadella stepped on Hood’s toes to get a word in edgewise first. I do not love that his answer is nonsense.

“It goes back to what I said about building the asset itself such that it’s most fungible,” it starts off. Microsoft has great first- and third-party businesses. There’s a balance there. These AI-first companies will be on the leading edge of the curve, the enterprise will be on the rear end (my words) of that curve. In short, Microsoft has confidence that its “fleet” will be utilized. If not by those companies, then by … someone.

Hood says that Microsoft has a large, flexible fleet that can be used for anyone and for any purpose, so she repeats what Nadella just said.

“And so, you’re right,” she finally says, though Murphy asked a question instead of stating an opinion. “Some of these large contracts have delivery dates over time, so you get a lot of lead time in being able to say, oh, what’s the status? And so, I think we’re pretty thoughtful around what’s always gone in our RPO balance, and been considerate of that. There’s always been that taken into account when we publish that bookings number and publish the RPO balance.”

I don’t mean to put words in anyone’s mouth, but I feel like no one asked the question. Microsoft may be “thoughtful” when it evaluates the ability of a company that has no money but commits to spending billions. But how will it actually collect that money? From someone else, apparently.

Why doesn’t Azure have enough capacity?

Brad Zelnick of Deutsche Bank, sensing that no one has sucked up to Microsoft in several minutes, gets things back on track by “echoing [his] congrats on an amazing start to the year.” And his concern isn’t where the money is coming from, why Microsoft is spending so much money on AI, or even why Wall Street doesn’t value the company’s stock enough. It’s how Azure’s lack of capacity will impact revenues going forward. If only Microsoft could spend more on capacity!

This is “a great question,” Hood claims.

Microsoft has “worked very hard” to allocate resources correctly, she said and mitigate “it” as best it can, but “we have been short in Azure … I think it’s probably hard for me to give an exact number. But it is safe to say that the number could be higher.”

Is there anyone more qualified or knowledgeable about this topic than the CFO of Microsoft? No. But that was the answer. Which Zelnick thought was a “great” response.

The elephant in the room. No, not that elephant

There was only time for one more question, so Kash Rangan of Goldman Sachs used his time wisely to get right to the heart of the matter.

“Amy,” he started, “I just wanted to congratulate you. I think you said before that it is possible to accelerate Azure growth while getting efficient margins, and you’ve done it. Congrats on that.”

And then he had a question. For Nadella. And this was about “the elephant in the room,” which I assumed meant Wall Street’s complicity in the lack of transparency we get from Microsoft’s earnings.

It was not.

“There’s talk that another hyperscaler came in and took away the business that was rightfully Microsoft’s,” he claims and, no, I have no idea what he is referring to there. “I’m wondering if you could offer some perspective on your criteria to, is it about a certain volume of business that you wish to execute on the Microsoft paper, or something broader than that, that I don’t think maybe people fully appreciate the terminal value that Microsoft will have on its balance sheet at the end of these contracts, which I think is probably being underestimated, as you have the full stack and you’ve got the multiple vectors to monetize, be it databases, Foundry, and to your point, that you are a platform company, not just a hyperscaler? Maybe that’s what it is all about, or maybe there’s another story about you letting the other hyperscaler company come in from nowhere, and claiming a big piece of that four-to-five-year puzzle. Thank you so much once again, really appreciate it, and congratulations.”

For the love of God. But don’t worry, Nadella gets it back on track.

“It just always goes back to, I think, the core principle, which is to build a fleet that is fungible across the planet and works for third-party and first-party and research,” he says. “That’s essentially what we have done.”

Look, I understand there are talking points, that these executives are staring at a list of those points during this call and are trained to drive home whatever messaging as often as possible. But if the main talking point this quarter is literally “a fungible planet-scale fleet” and not “THIS is WHY we are spending so much money,” I am at a loss for words.

I guess no one can just say that OpenAI is using non-Microsoft clouds now, and that Microsoft no longer has the right of first refusal on that. But Nadella talked around his answer to a question that likewise talked around specifics.

“We say no to some of the demand that may be something that we could serve, but it’s not in our long-term interest,” he finally says. “And so, that’s the decision making we’ve done, and we feel very, very good about the decisions. In some sense, I feel even each time we say no, the day after, I feel better.”

Perhaps Microsoft could say no to some of this spending? I think we would all feel better the next day if that happened.

Or as Rangan replied, “That’s fantastic, very clear.”

These are the people who determine whether Microsoft’s stock is worth buying. And those are the questions they asked. So I feel good about the future.

But I kid.

Notable notes from the conference call

Since I neglected to actually write about any Microsoft products or services here, I will at least run down a few data points that were provided during the call. I mean, if you got through all that nonsense above, you deserve something in return.

“Microsoft Cloud revenue surpassed $49 billion, up 26% year over year.”

There is no such thing as Microsoft Cloud, and the definition of what constitutes this non-business can and does change quarter-to-quarter.

“Already, we have roughly 10Xed our investment [in OpenAI].”

This is spending, not a return on the investment.

OpenAI has contracted an incremental $250 billion of Azure services;

OpenAI does not have that money.

“We will increase our total AI capacity by over 80% this year, and roughly double our total datacenter footprint over the next two years, reflecting the demand signals we see.”

There are no words.

“We are building a fungible fleet that is being continuously modernized and spans all stages of the AI lifecycle, from pre-training to post-training, to synthetic data generation and inference – and it also goes beyond genAI workloads to recommendation engines, databases, and streaming. We are optimizing this fleet across silicon, systems, and software to maximize performance and efficiency.”

This is how the Empire described its fleet of Star Destroyers in “Star Wars: The Empire Strikes Back.”

“We have 80,000 customers, including 80% of the Fortune 500.”

How many are actually using AI actively day-to-day?

“Once again this quarter, Azure took share.”

From AWS, presumably. So what’s the share?

“We now have 900 million monthly active users of our AI features across our products.”

I inadvertently launched Copilot over 12 times while writing this post. Can I assume that explains much of this usage?

“And our first party family of Copilots now has surpassed 150 million monthly active users across information work, coding, security, science, health, and consumer.”

Wait, wait. You have over 900 million MAUs and yet only 150 million of them are using your first-party Copilot products? Please explain. (They do not.)

“When it comes to coding, GitHub Copilot is the most popular AI pair programmer, now with over 26 million users.”

Oh, there you go. A hard number. Was that so hard? Heh.

“All up, GitHub is now home to over 180 million developers, and the platform is growing at the fastest rate in its history – adding a developer every second.”

Most are not paying, but sure. Another hard number.

“Every Windows 11 PC now is an AI PC.”

It was as easy as waving a wand.

“Edge now has taken share for 18 consecutive quarters.”

Um. Looking just at desktop browser usage share, which puts Edge in the best light, Edge has gone from 8 percent to 10.37 percent, so, yes, a gain. But Chrome, the market leader, has grown more, from 67.55 percent to 73.81 percent.

“We took share again in search [with Bing].”

Don’t make me look. Do not make me look.

OK, I looked. And sure, the chart of web search usage share is looking good for Bing. Google has a 90 percent share, the same as one year ago. And Bing has a 4.1 percent share, up from 3.96 percent one year ago. They’re killing DuckDuckGo Search, perhaps. #winning

“One billion monthly active users of Entra.”

This one is literally interesting because it probably constitutes a huge percentage of commercial Windows users. If there are 1.4 or 1.5 billion Windows users overall, that’s most of them.

“In gaming, we expanded our reach across every endpoint, focused on our higher margin content & services.

In plain English, Microsoft laid off thousands of employees in gaming, closed studios, canceled games, and is forcing Xbox to achieve an impossible 30 percent profit margin.

“Reaching 155 million monthly active users of Minecraft, an all time high.”

Nice.

“Set a new record for overall content and services revenue for the quarter.”

Thank you, Activision.

“We also saw a great response to the Xbox Ally launch two weeks ago.”

Really?

“Set a new records for players on PC.”

It’s all because it’s an Xbox! Also, most of them are on Steam.

“In closing, our planet-scale cloud and AI factory, together with Copilots across high value domains, is driving broad diffusion and real-world impact.”

Those are nonsense words. Don’t believe me? Just ask Copilot.

“Paid M365 commercial seats grew 6% year-over-year with installed base expansion across all customer segments, though primarily in our small and medium business and frontline worker offerings.

I see. What is the number of commercial Microsoft 365 seats now?

/crickets

“M365 consumer cloud revenue increased 26% and 25% in constant currency, again driven by ARPU growth. M365 consumer subscriptions grew 7% to over 90 million.”

I love me the hard numbers.

“Windows OEM and Devices revenue increased 6% year-over-year, significantly ahead of expectations, driven by strong demand ahead of Windows 10 end of support as well as a benefit from inventory levels that remained elevated.”

One wonders what PC sales growth was like when previous big-bang Windows releases like Windows XP and Windows 7 hit end-of-life. But one is too lazy to look. I will bet real money it was much higher in both cases, however.

“In Gaming, revenue decreased 2% and 3% in constant currency. Against a strong prior year comparable, Xbox content and services revenue increased 1% and was relatively unchanged in constant currency, driven by better-than-expected performance from third-party content.

One now wonders how profitable this business would be if it weren’t for hardware. Also, “strong prior year comparable”? What happened in the quarter a year ago that was good for gaming?

“The combination of OpenAI’s conversion to a public benefit corp and the ongoing nature of our partnership will result in increased volatility. Therefore, going forward, we will provide our outlook excluding any impact from our investments in OpenAI. On that basis, in Q2, other income and expense is estimated to be roughly $100 million as interest income will more than offset interest expense.”

I also find that problems go away when I ignore them. And this is a solid accounting practice that Intel also employs for its Foundry business.

“Next, capital expenditures. With accelerating demand and a growing RPO balance, we’re increasing our spend on GPUs and CPUs. Therefore, total spend will increase sequentially, and we now expect the FY26 growth rate to be higher than FY25. As a reminder, there can be quarterly spend variability from cloud infrastructure buildouts and the timing of delivery of finance leases.”

OK, so the AI spending will be bigger and spending will accelerate. I don’t have the skill to do the math on this, but if the figure was about $35 billion this quarter, and if the spending rate is higher than it was last year, what does that mean for Q2? $40 billion?

Does this ever end?

“[In the current quarter,] Windows OEM and Devices revenue should decline in the mid-single digits. We expect continued momentum from Windows 10 end of support.

Those are contradictory.

“Devices revenue should decline year-over-year.”

This is the only way in which Surface is mentioned these days.

“And in Xbox content and services, we expect revenue to decline in the low to mid-single digits against a prior year comparable that benefited from strong first-party performance, partially offset by growth in subscriptions. And hardware revenue should decline year-over-year.

One wonders for about the 20th time how long Xbox hardware revenues can decline.

“We remain focused on delivering real value to our customers that results in durable revenue growth for the long term.”

I feel the value. Do you?