Microsoft: Why Now Is A Good Time To Invest In The Shares

Summary:

• Microsoft reported its earnings and provided guidance for the coming quarter, slightly below expectations, leading some analysts to question  the impact of generative AI on the company’s growth.
• The market meltdown and subsequent recovery affected stock prices, with Microsoft experiencing a decrease in share value before a bounce back which still leaves it down 10.5% since 7/5.
• Microsoft’s capex growth is substantial, with a focus on expanding cloud and AI capacity, leading to questions about the impact on margins and the business model.
• GenAI is providing the company with a noticeable bookings tailwind outside of Azure, with positive trends being seen in GitHub and Dynamics.
• Microsoft is seeing noticeable market share gains which can be attributed to the totality of its genAI portfolio.

Interpreting the Microsoft quarter: Will generative AI be the business driver that has been expected?

Microsoft reported its fiscal Q4 results on July 30^th and provided guidance for the coming quarter and year. The company’s overall results were considered to be uninspiring; its guidance for fiscal Q1 was also marginally short of prior expectations. Does that mean that the AI revolution is a chimera? And why does Microsoft need to grow its capex so substantially?

Before starting the analysis, I think a comment about the market meltdown and subsequent recovery is in order here. The meltdown had its genesis in two unrelated factors. One of those were the succession of poor metrics with regards to employment and construction activity. The July employment report was a shock to many market participants of the bad kind. The other factor, apparently, was the unwinding of the Japanese carry trade. This is not a place to discuss that, but the combination of weak macro data and the unwinding of the carry trade proved to be toxic for stocks. After excessive volatility in Japanese markets, the Vice Governor of the Bank of Japan has rolled back expectations for future rate increases and this has calmed markets resulting in a significant bounce. And in addition, after a fairly decent ISM non-manufacturing survey, a decent retails sales report, and declines in 1st time unemployment claims, expectations for a recession have been rolled back by a bit.

A couple of weeks ago strategists and hedge funds were focusing their investment bets on cyclical stocks and even small caps. Now, investors seem to feel software is a decent investment based on some decent earnings and a less dire macro outlook. IT stocks were valued below historical averages before this latest valuation implosion according to the latest Meritech analysis to which I have linked here. Do I expect a recession will impact the growth rate of software companies? Perhaps, although I make the case in this article that Microsoft, in particular, is already starting to see a positive growth inflection from GenAI tailwinds.

This is an article about Microsoft and specifically the reality of its genAI tailwinds which I believe are underappreciated by some. This last quarter had a number of positive metrics regarding the progress of the genAI solution and how that revolution is and will be impacting Microsoft growth. The reality belies much of what has recently been written, particularly by some click-seeking financial journalists who simply are unaware of the opportunities that are already developing in this space.

I have a significant position in Microsoft dating back 8 years now, and my take on their latest earnings report was substantially different than the consensus. I haven’t sold a share and I have no intention of doing so in the near term. If my overall weighting in Microsoft wasn’t quite elevated, I would already have bought more shares.

In writing about high growth IT companies these days, I find it important to note a caveat. Over the last 4 months, these shares have under-performed “tech” to a greater or lesser extent. This isn’t due to any lack of growth, or deterioration of the outlook for these companies. In fact a series of earnings reports from 8/6 were consistently better than feared.

That said, for several months, investors wanted their AI fix straight, and pivoted noticeably to what I might call AI hardware shares and the so-called Magnificent 7. And then, when those shares had been priced to perfection, the market rotated to small cap and cyclical non-tech shares. That put enormous pressure on valuation.

Microsoft is of course one of those Magnificent 7 and like many of its brethren in that select group, it hit a high for the year on July 5th. The high that day was $468/share. And then the valuation fell victim to the 2^nd pivot and the shares had fallen by about 10% over the last 3+ weeks before its marginal fall on the day after its earnings release on what was a strong day for most tech shares. The shares fell by another 7% over the last 6 trading days before the bounce being seen on Tuesday, August 6 and over the following 10 days. So, at this point the shares are down by about 10.5% from their high. Based on my revenue and margin forecasts which are not materially different from the consensus-at least in the short term-the shares have an EV/S based on 12 months forward revenue of about 9X, a 3 year CAGR of 18%, and a free cash flow margin for the next 12 months of 29%. It is that 3 year CAGR estimate that really is key assumption undergirding my share purchase recommendation.

The shares do trade as analogs of the other Mag 7 even if each one of these has quite different business drivers. I find Microsoft’s metrics to be attractive, particularly as I believe that the CAGR estimate is likely to see upward revisions as genAI bends the demand curve beyond what is currently in my forecast.

I certainly do not have the temerity to suggest that I have second sight and thus have an ability to know when the next market pivot might come and who will be favored when it does come. Personally, I think the concept of buying what are described as blue chips or cyclical non tech is more than a bit illogical-but I have seen plenty of illogic that can persist for months in my years of participating in equity markets. The fact is that the just ended spike in the valuation of AI hardware ran several months beyond its sell-by date, based, at least in part, or so it now has become apparent, on the Japanese carry trade. This article is about the long-term prospects of Microsoft and not how it will trade relative to other Mag 7 shares, or what group might be in favor over the coming weeks. Simple answer is that I don’t know and have no idea as to where to obtain such knowledge.

The combination of these factors have led the shares lower although the actual daily pull back was 1.1% on the day of the earnings release. But the real picture is that demand signals for AI are accelerating. In fact my math, based on what the company has disclosed regarding the generative AI contribution to revenues, is that growth was probably more than 20% sequential last quarter within Azure. Azure AI customers grew 60% year over year; and these customers are coming back for more seats and greater usage.

AI customers in other parts of Microsoft’s business have grown even more rapidly-the company called out that its GitHub co-pilot was already generating more revenue than GitHub had generated when Microsoft acquired the company (about $250 million +). The CEO called out its Copilot for Dynamics transforming that business. It is difficult to quantify just how much co-pilot means to Dynamics 365 in terms of revenue growth: what is hard to dispute is that Dynamics 365, with constant currency revenue growth of 20%, is growing at least 4X more rapidly than the ERP/CRM spaces.

This article will not try to analyze every nook and cranny of Microsoft’s business; there are far too many and most of them are not particularly meaningful in analyzing the company’s future growth prospects. While almost all Wall St. analysts rate Microsoft shares a buy or a strong buy, the shares are a bit more controversial amongst SA authors with 7 buy recommendations and 4 hold recommendations that have been published over the past month.

The issue for those authors publishing hold ratings is generally one of valuation and the valuation concern is based on growth concerns. And these growth concerns are mainly based on a belief that AI is a chimera, at least in part, and a bubble, and that Microsoft will not achieve the growth inflection that justifies its valuation.

Now, as previously mentioned, I am a Microsoft holder and have been for 8 years now. It ranks amongst my top holdings and has done so partially based on appreciation. I share that to be transparent and to provide a context for my comments. I personally have a strong belief that AI/gen AI will be one of the more seminal events of my lifetime-which at this point extends to a time when telephones had rotary dials that clicked and were made of bakelite. Over time, and without hyperbole, genAI will change, and for the most part enhance, most phases of the human experience.

I have, in one way or the other, been a stake holder in the IT space for more than 50 years. The other day I read a blog post from the folks at GitLab about software that wrote itself. The reality is that the headline was a bit of hyperbole-I can’t quite ask the GitLab product to create a holistic supply chain management application tuned for my company. But it is now on the horizon-5 year’s maybe. But to the point of Microsoft, its GitHub product with co-pilot is close to similarly enabled-that is why GitHub growth has inflected so much higher.

I think it is hard to convey just how revolutionary that development is. And then to relate it back to Microsoft, and look at all of its gen AI products. There are some investors who are of the “show me the money” camp. Microsoft’s latest quarter didn’t do that, quite. But Microsoft’s latest quarter did suggest that a growth inflection based on AI was on the horizon.

I do not believe that Microsoft’s current valuation, and even most published price targets, really reflect the AI opportunity. So, this article will try to look at what that opportunity is and why I believe that Microsoft shares, at their current price undervalue those prospects. I recommend Microsoft shares as a fresh money buy, at this time and at this price.

Zeroing in on what Microsoft reported and how it guided

While my enthusiasm for the growth opportunity of gen AI for Microsoft and for much of the software industry is substantial, that enthusiasm is ultimately grounded on specific financial attainment. Last quarter’s headline financial attainment was a bit meh, so to speak.

Microsoft reports it revenues in several categories. The largest of these is what Microsoft calls the Intelligent Cloud, which is about 44% of the total, and which grew by 20% in cc last quarter, marginally greater than the prior forecast. This is where revenues from Azure are reported, and where much of the current contribution from AI is found.

The second largest segment reported by Microsoft is that of Productivity and Business Processes This segment, about 31% of total revenues grew by 12% in cc. This segment includes the Office products, LinkedIn and Dynamics, the Microsoft ERP/CRM solutions.

Finally, there is the More Personal Computing segment. This segment, which is where Microsoft reports gaming revenue was 25% of the total and grew by 15% in cc. This segment includes search and news advertising as well as Windows. Almost all of the revenue beat was concentrated in this segment, probably not what most investors prefer, as this segment has more limited growth opportunities than the other 2 segments.

Overall, Microsoft’s beat on revenues was about 1%, smaller than most prior revenue beats for the company. The company’s EPS was $2.95, an upside of $0.01 and which was also smaller than most prior quarters.

The company provided what was seen as a muted outlook. Specifically, the guidance works out to be for EPS of $3.08 for this quarter vs. the prior consensus of $3.19. The revenue forecast for the quarter has gone from $66.3 billion to $64.8 billion. By the end of the year, the implied revenue and EPS forecasts have been revised upward.

Some key points around these forecast: CapEx growth is starting to rise-dramatically, perhaps, as it grew by 78% last quarter with the company CFO forecasting continuing growth this current fiscal year. Commercial bookings growth of 19% cc, was quite a bit above prior management expectations. The company is signing larger and longer contracts for Azure which will ultimately show up in revenue growth. The company’s opex forecast is for single digit growth; the company is practicing renewed discipline on opex growth in order to partially offset the growth of the cost of goods expense due to the capex needed to support Azure and genAI workloads.

Where and when is the inflection from generative AI?

For the most part it hasn’t quite happened yet, or perhaps better said, it hasn’t happened in the way wished for by investors. And that left some cohort of investors impatient. High growth investors are rarely anything other than impatient in my experience, but in terms of AI they seem more impatient than in the past when similar paradigm shifts were being born. Generative AI is software that dramatically enhances the productivity and the experience of users.

When it first burst into the consciousness of investors, the immediate reaction was to buy anything related to the technology. Some commentators decried this and called it a bubble mentality. Other commentators started to say that the technology wouldn’t/couldn’t be monetized and started to write about spurious data comparing the investment in GPUs and the revenue contribution from gen AI.

I think, perhaps, a typical genAI use case might provide readers at this point with how the technology works and why the leadership at Microsoft has been willing to make a huge bet on providing infrastructure so that users can run their newly created genAI on Azure.

I think most readers will be very familiar with call centers and how they have worked. Many call centers these days are powered by bots, the kind of bots that have proliferated in recent years and which often come from UiPath (PATH). The problem with bots is they lack intelligence by design, and are thus notoriously inflexible and that often leads to frustration on the part of customers when trying to deal with an issue.

With genAI, life can be very different. Sam’s Club is an early user of genAI in some of its call center applications and where it has been partially deployed according to a story relayed to me by a colleague. One real life story concerns an adventure with a kind of healthy oat product. Those readers who have any familiarity with this author know it isn’t me-I am neither a Sam’s Club member nor a consumer of healthy oat products. But I have friends who are, actually.

In a this particular use case, an order for a healthy oat product had gone astray. The call center genAI app can determine that the warehouse app didn’t correctly allocate the product and that it hasn’t shipped to the customer. The consumer is told what happened, and when, and is asked if the app can resubmit the order for the customer and make sure it reaches the customer on a certain date. Sometimes, the app will provide the customer with a promotion on additional products. The result is a customer whose order problem is resolved promptly, without frustration and without any human intervention. These kinds of genAI applications are proliferating rapidly; a similar one that has become a reference for Microsoft is that of a German retail chain.

Most software executives, and that includes leadership at Microsoft are excited about genAI technology because of the use cases it can animate. Very typically software has been sold to enterprises on their need to stay competitive. If one call center works better than another, it will inevitably result in market share gains. Needless to say, in time, essentially all call center apps for almost all kinds of consumer facing interaction, will be rewritten to incorporate this technology. I recognize that investors want to see the money from this in current operating reports. There apparently is a temptation to believe that genAI is just a myth or another tech bubble. The results from Microsoft, specifically suggest that this is not the case.

Most 3^rd party analysts who had written on the subject had always believed that 2025 would be the year in which the GenAI growth inflection became visible. That really hasn’t changed and the fact is that it hasn’t changed for Microsoft either. But investors seem to believe that something not foreseen is retarding the growth of GenAI, both broadly and at Microsoft, clearly a pioneer and share gainer in this space.

This past quarter, Microsoft reported that the cc growth in Azure was 30% compared to a forecast of 30%-31% growth. The prior quarter, growth had been 31%. This quarter, 8 points of the 30% growth has been attributed to AI compared to 7 points of growth last quarter. As mentioned earlier in this article, this implies that AI revenue showed a sequential growth of a bit more than 20%, basically an annualized triple digit growth rate. That isn’t bad performance for an infant technology whose deployment is complex, costly, and lengthy. In addition to AI on Azure, and AI as part of GitHub, Microsoft has what it calls co-pilots that are genAI add-ons in many other components of its offering. Copilots exist for Microsoft365, for Dynamics, for finance, for Windows, in healthcare, in security and AI capabilities have been added to Teams and to LinkedIn.

I would point out two things. One is that total revenues from AI are quite a bit greater than the specific calculation of AI revenue ($5 billion in total) as reported for Azure. But more than that, it is Microsoft’s genAI capabilities that are driving much of the growth in the intelligent cloud and in productivity and businesses processes. For example, Microsoft has a product it calls Fabric. Fabric isn’t considered as part of Microsoft’s AI revenues; it is a unified data and analytics platform that includes data movement, processing, ingestion, transformation routing and event routing. Fabric has seen mass adoption-there are currently 11,000 users. But the need for Fabric is being driven substantially by users deploying genAI applications on Microsoft infrastructure. Another example of this drag is seen in the results of Microsoft 365 where ARPU is rising due to the purchase of co-pilots to go along with the MS 365 suite. This phenomenon is basically seen across every segment of Microsoft’s software offerings.

Microsoft’s total bookings rose by more than 31% in constant currency. Bookings to be recognized as revenue in the next year rose by 20%. The backlog has increased to $269 billion, up by 14% sequentially and has grown as a percentage of revenues. Those numbers were above expectations and reflected Microsoft customers making larger, multi-year commitments for Azure as part of their establishment of their own infrastructure to support their AI applications.

One of the more unique features of Microsoft’s current financial performance is that it has indicated that it is been experiencing capacity constraints that have limited its ability to grow Azure revenues to match actual demand.

Azure growth included 8 points from AI services where demand remained higher than our available capacity Therefore, our Azure consumption business continues to grow faster than total Azure.

And in H2, we expect Azure growth to accelerate as our capital investments create an increase in available AI capacity to serve more of the growing demand.

There was no indication of the precise amount of revenue that was foregone last quarter because of capacity constraints. Similarly, there was no indication of the precise growth ramp to be expected in 2H of this current fiscal year. Interestingly, the revenue consensus does not at all reflect this expectation for accelerating Azure growth. In fact the 1^st call consensus shows a peculiar pattern in looking at percentage revenue growth in the next several quarters with an acceleration next quarter, a sharp deceleration in the December quarter and then an acceleration in the last 2 quarters of the fiscal year. Some of this reflects the anniversary of the Activision acquisition. Expectations for FY’ 2026 show no growth acceleration from percentage revenue growth expectation for FY 2025. That seems improbable, specifically given the company’s conference call comments. The company is not increasing its capex by 78% without a very specific line of sight with regards to demand for genAI/Azure workloads and usage.

Microsoft is made up of many moving pieces of which Azure is but one. There are more than a few moving pieces in the Intelligent Cloud segment, although Azure is now the principle component of this segment which was 44% of revenue last quarter and which is likely to be 46%-47% of total revenues this year. If indeed, Azure growth accelerates as capacity becomes available, and as the strength in current bookings translates into reported growth, than there should be a noticeable acceleration in revenue growth to greater than the current forecast level of 14%.

Microsoft Capex-a pretty substantial bet that seem very controversial

Last quarter Microsoft’s capex was $19 billion. This compares to capex of $10.7 billion in the year earlier period, an increase of 78%. Almost all of the capex has been to grow the company’s cloud and AI capacity. Half of the capex is for data centers and the balance is for servers, both CPUs and GPUs. Microsoft hasn’t projected specific levels of capex; its statement is simply that the company expects capex to be higher in the current fiscal year. than its most recent levels

Ultimately, capex impacts Microsoft’s cost of goods sold through higher depreciation expenses. So, the company is indicating it will tightly manage opex in order to manage the company’s operating margin performance.

There has been some comment that the investment required to support the growth of genAI revenues will pressure margins and ultimately impact the company’s business model. About half of the company’s capex is a product of building new data centers. Data centers have extended useful lives; Microsoft typically keeps datacenters in operation for 15 years before they are physically replaced. The balance of capital expenditures are

“dependent on demand signals and adoption of our services that will be managed throughout the year.”

Basically, Microsoft gets orders for Azure and AI services, and then in turn places orders for CPUs and GPUs. In that regard, capex is an excellent leading indicator for future growth of Azure and AI

So I would say – and obviously, the Azure AI growth, that’s the first place we look at. That then drives bulk of the CapEx spend, basically, that’s the demand signal because you got to remember, even in the capital spend, there is land and there is data center build, but 60-plus percent is the kit, that only will be bought for inferencing and everything else if there is demand signal, right? So that’s, I think, the key way to think about capital cycle even.

So if you look at it, we have both the landing of the seats itself quarter-over-quarter that is growing 60%, right? That’s a pretty good healthy sign.

This issue, i.e that of demand signals being received vs. the growth of capex was one that occupied much of the conference call and vexed some analysts. Personally, it is hard for me to understand investor angst about genAI demand. To paraphrase an ad that was popular 40 years ago-the beef is in the productivity. The CEO called out call centers as an almost universal use case for genAI. It is almost impossible for me to imagine that over the next 5 years call centers will not entirely be migrated to applications based on genAI. Microsoft suggested that it is already saving several hundred million a year-in its own call centers-based on the use of genAI applications.

Of course developing a genAI call center for a call center is a process requiring lots of iteration, lots of model training, lots of inferencing, lots of capex, and a fairly intense deployment process. These factors can and will result in some variability and ups and downs in the cadence of actual revenue growth for genAI for Microsoft specifically. But Microsoft, which obviously has a better specific line of sight when it comes to understanding the demand signals it is seeing, is making a very substantial bet that its capex will be needed because current demand is, and has been exceeding supply, for Azure and for genAI.

Microsoft Valuation

Is Microsoft overvalued? Of course I don’t think so and I have quite a bit riding on that evaluation. The real issue is the whether or not genAI is or will precipitate a significant growth inflection and if so, by just how much. I am certainly not a fortune teller.

I don’t believe that the results from this past quarter are at all indicative of what genAI will mean to Microsoft in terms of growth. There are bits and pieces of a glimpse of the impact that genAI is having and can have. Some of that shows up in the spike in commercial bookings which to repeat rose 20% year on year and by 14% sequentially. Some of it shows up in the growth in seats shown above in the commentary of CEO Nadella.

The company talked about accelerating growth from Azure and AI in 2H of its fiscal year. That is not really reflected in the models of most analysts, at least as those reported to 1^st call. And growth of 14% projected for what will be FY’26 does not show any growth inflection either, after adjusting for the impact of the Activision acquisition on the company’s reported growth this past year.

Of course, not all of Microsoft is AI and Azure, but as noted it is, these days 45%+ of total revenue and that percentage continues to increase. During the course of the conference call, the CEO and the CFO reiterated a theme that this transition is similar to the transition from 8 years ago as the company moved its offerings to the cloud and saw a substantial increase in its growth rate.

There is a difference in this transition in that most of what is happening is based on new applications, new workloads, and significant market share gains that are being seen in many of the company’s existing markets.

I don’t want to pretend that I have some special formula to quantify the spike that I expect in terms of revenue growth due to the advent of genAI. It is clearly significantly greater than the numbers being reported in terms of the genAI percentage growth on Azure. And given just how difficult drag revenue can be to identify and forecast that is likely to remain the case. And at this point the company hasn’t provided specific guidance as to the acceleration in revenue growth as capacity becomes available to support Azure/AI consumption demand in 2H.

I acknowledge that I had to make some assumptions in modelling what will happen when capacity does become available. Overall, I think the combination of capacity additions and genAI demand growth will lift the growth in the Intelligent Cloud segment from 19% to 27% as a run rate by the end of the current fiscal year, i.e. a year from now. I think the productivity segment of the company business may continue to grow at 11% while the more personal computing segment will grow at 12%, which eliminates the last impact from the Activision acquisition. This brings expected revenue growth to about 19% as my base case assumption rather than 14% which appears to be the current consensus forecast for that metric by the end of the current fiscal year. But to be conservative, in my valuation model, I used a revenue growth assumption for the next 12 months of 16.5%.

Overall, my model includes a 3 year CAGR assumption of 18%, a free cash flow margin assumption of 29% which accounts for the growth in capex necessary to support higher revenue growth and a current EV/S estimate of 9X. This combination yields a valuation of about 20% below average for the company’s growth cohort. It is likely, in my opinion, that Microsoft shares will ultimate trade at a premium to average for the company’s growth because of the conservative nature of company projections and its strong line-of-sight with regards to future demand.

Risks to the thesis

Of course the major risk to this thesis is my view as to the potential growth rate for genAI. I don’t pretend that my forecast is set in stone, or is in any way part of any specific company guidance. As with any new technology, and especially one like this which is so revolutionary, quantification is at least as much of a guess as anything else. I have tried to present the view that genAI will become a necessity which is embraced by almost all enterprises with use cases as yet to be developed, or even imagined. But simply looking at the use cases which are visible today such as the enhancement of call centers yields an exceptional opportunity.

The other risk to the thesis is the ultimate ratio of required capex to revenue for genAI. This is still very early in the evolution. Initially capex growth will exceed the growth in dollars of genAI revenue. Just when and to what extent that crosses is still really unknown. Obviously the spike in capex last quarter to $19 billion, up 78% year over year, would, if continued, challenge free cash flow margin assumptions.

Estimating future capex growth is another guess, and thus another risk to the thesis.

Finally, it is possible that macro-economic conditions deteriorate far beyond current consensus expectations. If they do, the genAI revolution may be delayed, but certainly not denied.

Wrapping Up: Reiterating the positive investment thesis for Microsoft shares .

Simply put, I believe that investors are under-estimating the growth inflection that genAI is and will bring to Microsoft. Last quarter revenue growth was slightly less than forecast. Most of this can be laid at the door of Azure growth-but Azure growth was capacity constrained. GenAI is already having a material impact on Microsoft revenues-outside of Azure. In particular, growth of GitHub, and growth of Dynamics are already seeing acceleration because of genAI. Other segments of Microsoft’s business will benefit as well.

I think the visible and identifiable use cases for GenAI such as its ability to improve both the productivity and the user experience in call centers is far greater than generally realized.

This has led me to boost my 3 year CAGR estimate for Microsoft about 200-300 basis point to 18%. That is obviously significant for a company of this scale. With that change, and even after adjusting for capex growth needed to support that kind of revenue growth, Microsoft’s valuation is noticeably below average for my estimate of its growth cohort. Thus, I reiterate my buy recommendation for the shares and expect them to produce significant alpha over the coming 12 months and beyond

---

Analyst’s Disclosure: I/we have a beneficial long position in the shares of MSFT either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Seeking Alpha’s Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

---