Microsoft Stock Is Getting Crushed. When To Buy?

0:00 Everyone and welcome back to another undervalued or not video. We're going to talk about Microsoft in this video. We're talking a lot about meta, about Google. By the way, meta today was up 4.7% new spark 1.1 came out. Zuck started tweeting again, so all is good. Still extremely undervalued though. Microsoft. Microsoft is also one of those names where okay, it's a given that it's up there, but then again, when you look at the stocks performance hasn't really been great. And it's not like this is an expensive name.

0:34 Still is very profitable. So, what's going on actually with Microsoft? Let's first look at the heat map for today. So, as you can see bright green for meta back at $631, which is a key level of four the name. But if you look at Microsoft Microsoft is a $2.86 trillion company trading at around $384. If we go and look at the stock itself over the past 5 years, over the past 5 years the stock is up 38.2% It is experiencing a drawdown right now.

1:07 Let's put the last 12 months or so. It is experiencing a drawdown of around 29%. It's down 23% over the past 12 months, but from its peak it's down 29%. It was down 35% not that long ago. So, what's the deal with Microsoft? Is this a company worth buying right now? Is this another one of those too cheap to ignore? Is the market just not looking at the right things? Is the market only interested in an entropic, a SpaceX, AI and open AI, etc., etc. and not really interested in the likes of a Microsoft a meta could say Amazon. Yes, Amazon is a bit more pricey than a Microsoft but still very, very good companies. Even an Nvidia, right? An Nvidia is yes, it's worth $4.9 trillion, still up 24.3% over the past year, down 14%. That's the drawdown right now, but it's still a company that's trading at okay, 31 times trading PE, but only 20 times forward PE, a PEG ratio of 0.5. I mean, it's quite interesting to see that certain companies in the market are getting loved and super high premium, while others, which are usually the ones that are still doing extremely well today, are not really getting that much attention. Now, it might be because It might be because of this. If we look at CapEx, the CapEx growth. This is over the last 12 months for Microsoft. I mean, you can clearly see what has happened over the past 5 years. It has gone from around $20.6 billion to now $97.2 billion. It definitely will reach $100 billion plus by next quarter. Again, last 12 months.

2:58 So, okay, we we know we know about the CapEx fears. We know. We've heard it a thousand times already. But a Microsoft, unlike many other companies, a Microsoft is still generating here a ton of free cash flow. In this case, over the last 12 months, that's over $70 billion in free cash flow. But we've seen that also with a Google. Free cash flow there is not going to go negative. And that's that's a huge, huge competitive advantage right now. Because if if let's say monetization isn't as great as we thought, okay, then then those that aren't profitable are probably going to go out of business, but those that have an insane amount of cash and that generate an insane amount of cash, well, they can go and shop around. Like, "Oh, I like this company. I like that product. I like that feature. I'll buy it. I'll integrate it and my ecosystem becomes better and better.

3:54 So, again, what's the problem with the market here? Well, before we try and answer that, from a purely technical analysis point of view, we can see here, this is on the weekly, back in June, early June, it tried to go above the 50 on the weekly, didn't happen, big rejection there, went all the way back down under the 200-day, which by the way, if we go and look at the daily right now, as you can see, we are in no-man's land.

4:21 And the daily was the 200-day moving average at the start of June, on the weekly, it was the 50, which are key areas. Rejections, now we are a bit in no-man's land. If I would be a betting man, I would assume that the next earnings report is when this thing starts to go up because it will show yet again Copilot adoption, Azure growth, talk a bit more about, well, being model agnostic, which is going to be the trend more and more. And so, right now, the market is saying one, CapEx is outpacing revenue growth, right? We have around $190 billion planned for 2026, which is more than AWS' total annual revenue. AWS is still the biggest one out there.

5:08 Copilot adoption is unproven at scale, roughly 20 million paid seats against over 450 million commercial Microsoft 365 users, that's a low conversion rate, the OpenAI dependency risk, 45% of Microsoft's $627 billion backlog ties to that one customer. Okay, but we have seen Microsoft try to diversify away from OpenAI. Then there is, I don't think that's just a big issue, the Xbox restructuring. I do think that the restructuring is needed, and honestly, with this whole Sony scandal, right?

5:47 Where all physical discs don't mean much. I think there is a huge opportunity here for Xbox. And I'm someone that never had an Xbox in my life. I'm a Sony guy, but hey, if Sony is going to screw us over, maybe maybe it's time to move. And so, let's let's try and understand why the market is wrong here. Microsoft is of course still being viewed as a software name, right? It's the biggest software player on the planet.

6:15 But Microsoft is, as we've said, Microsoft is model agnostic, moving towards usage-based pricing and deeply embedded in enterprise workflows most competitors can't replicate. And you will see why this makes sense. And we've talked about that before, but I'll gladly repeat myself. Then, with regards to Copilot, Copilot seats grew from 15 million to over 20 million in one quarter, up 250% year-over-year. Weekly engagements now matches Outlook. And Nadella calls it becoming a daily habit. Of course, me personally, as you know, that that little key on my keyboard here, that is Copilot, I never use it.

6:56 But everybody that I talk with, that works for a big company, big enterprises in tech, they all have to use Copilot. So, I guess yes, you have to go to the enterprise in order to make quite a lot of money in AI. As for AI analyzed revenue run rate, that sits at $37 billion. That's up 123% year-over-year. Productivity and business process revenue is at an all-time high of $135 billion for a trailing 12 months, compounding 14% annually.

7:27 As for the $190 billion capex, this is something that, yeah, it it it's happening across the board. We don't know what the true ROI is, right? We can put a number there. But what we can see is one. Commercial RPO is $627 billion. That's up 99% year over year. So, clearly there is quite a lot of demand. Second of all, operating margin just hit an all-time high of 46.8% this quarter or last quarter, up from 41.6% 5 years ago. So, the core business is more profitable than ever even during this build-out. And then as for the biggest question here, right? The dependency on OpenAI. So, the non-OpenAI portion of Azure is still growing 28% year over year. Azure overall is up 40% last quarter. So, clearly there is growth ex-OpenAI. And we've seen them, right? Approach Anthropic more and more.

8:27 We we also know that they want OpenAI to go public. So, yeah, they can maybe exit the position. They they can cash in. You name it. Then more recently, there was a report that showed that Microsoft is already swapping out OpenAI and Anthropic models for its own in-house AI in parts of its stack, actively reducing single vendor dependency. And of course, if they feel comfortable doing this in-house, this means that in-house the AI models are quite good. Otherwise, I don't think they would be making this move. Moving on to the next big thing is why the move to cheaper AI actually helps Microsoft. So, you have the AI labs, and then you have Azure Microsoft.

9:08 So, the AI labs, they build a model. Margin gets squeezed as workloads cheaper open models, right? The Those have been the scary headlines. Then you have Azure Microsoft. They own the road. They charge the same infrastructure margin no matter which model runs. So, enterprises are shifting from always use the best model to routing cheap tasks to cheap models, hard tasks to frontier models. You are hearing this from Meta. You're hearing this from Microsoft. Heck, we've been hearing this from Palantir for a long, long time. This doesn't reduce the total AI usage, it explodes it. It makes sense. Cheaper tokens means companies can run agents in loops, read entire code bases, do things that were too expensive before.

9:53 And if you look at the token volume for Microsoft here, over a hundred trillion tokens Microsoft has processed in a single quarter, that's up 5x year-over-year. Over 300 Foundry customers on pace for a trillion tokens or more each this year. The estimated rate AI inference cost is falling per unit of capabilities is approximately 10x a year. And the typical hyperscaler infrastructure operating margin is around 35 to 38% sticky regardless of which model runs.

10:28 And so, cheaper tokens historically haven't meant less cloud revenue for hyperscalers, they've meant dramatically more total usage. Falling price per token and rising total spend can happen at the same time. And then we move on to the Azure AI Foundry, the orchestration layer, and yes, it is very similar to what Palantir does. Azure AI Foundry lets enterprises run any model, OpenAI, Anthropic, Microsoft's own Open Wave through one managed layer with security, very important, compliance, and billing already built in. And then comes the big deal, especially the big advantage for these hyperscalers is switching models will happen, we know that. Switching cloud providers, yes, that there are some that do it, but most big, big companies anyways use all three or two of the three, but there are only three out there, right? So, you're not going to switch clouds every other day. So, for example, Amazon AWS, they have Amazon Bedrock, right? They already show this pattern playing out.

11:30 Multi-billion dollar annualized run rate, customer spend growing 60% quarter over quarter across over 100,000 customers. Azure Foundry is built on the exact same logic. And actually, we have I think yesterday read a sort of a report that said that maybe Meta is going to do the same thing for cloud. Now, all that said, is the stock actually cheap? Well, if we look at the pricing multiples, we're going to look at the DCF in a bit. We look at the trailing 12 months forward P. We can see that the trailing 12 months P is 30% lower than the average, the 5-year average. The forward one, 31%. Price to sales, 15% lower than the average. Forward price to sales, 24% lower than the average. And then the forward EV to EBITDA is 37% lower than the average. Now, of course, previously you might say, "Yeah, okay, but previously maybe it was trading at the premium." And I would say, "You know what? Okay, correct." But currently, if you look at those numbers across the board here, I I don't think it is trading at a premium. Now, of course, the one metric across the board, we've seen this, the one metric that is more expensive, that is elevated, is of course the free cash flow one. Which makes sense. A huge increase in CapEx.

12:48 So, the free cash flow multiples are coming down. Which we've seen across the board, but then again, they're still generating a ton of free cash flow every single year. Now, with regards to the DCF, well, we have a bull, a base, and a bear case. And as you know, this is available to all of you for free in the Google Drive along with all the other DCFs out there. And so, according to my assumptions here, the weighted intrinsic value of Microsoft sits around $508, which is 32.5% upside from where we're at right now. The base case, 518. The bull case, $671, the bear case, $326, which is around 15% lower than where we're at right now. And these are basically my assumptions here for revenue growth for the next 10 years. These are the assumptions for EBIT margin also for the next 10 years.

13:43 You can play around with those numbers, of course, on your own, make your own story for this to make sense and for this to, of course, reach these types of conclusions, right? Revenue to reach around $907 billion by year 10 in the base case scenario. For the bull case, that's $976 billion. For the bear case, $775 billion. Free cash flow to the firm, $264 billion for the bear case, $425.5 billion for the bull, and then $362 billion for the base case all in year 10. Does this make sense to you? Does this not make sense to you? That's up to you. That's why you do these types of exercises. You can make up your own story. Of course, you can make up all of the numbers to fit the number that you want, right? To fit the narrative, but the story needs to make sense, right?

14:34 You cannot say, "Oh, it's going to grow 40% every single year for the next 10 years." No, it doesn't make that much sense. And so, to me, it's quite clear Microsoft clearly undervalued, and I'm not the only one clearly analysts think this is also an undervalued name. The consensus 12-month target here sits at $560, which is 46% upside from where we're at right now. Now, you might be asking yourself, "Okay, so why aren't you buying the stock?"

15:04 Well, that's a good question. If I had a lot of cash, if I didn't already have plenty of positions in my portfolio, or you could say you could add it to the retirement portfolio, although, like I said previously, I'd rather maybe put Nvidia in there than a Microsoft, but who knows? That might change. But to me, right now, if I did own it, I would probably be buying more at this point in time. But, I don't own it, so I don't need to open up a position right now. I'd rather just continue to invest in Meta because I do think that Meta is more of a no-brainer and more undervalued than Microsoft.

15:42 But, then again, Microsoft, if you look at the company, if you look at the business right now, if you look at the growth, it is still there. It is still extremely, extremely attractive. It's just that in this game right now, just like we've seen with Meta, unless you come out with some fancy schmancy model and benchmarks and market the, you know what, out of it, the market will not care. Right? Meta today is at $630. Is Meta today suddenly so much of a better company than it was a week or so ago? No. They just happened to release Musepark 1.1, and people like the benchmarks or the the market likes it.

16:20 Okay. But, a week ago, we knew that Musepark was getting better and better. We knew that Meta were going to offer new products and services. We know that eventually they're going to charge money for the subscriptions. Millions of people will pay the subscriptions. We know they're going to be making a lot of money. We also know that the core business is already making quite a lot of money. But, the headlines, the headlines and sentiment, that's what moves the stock in the short term. Especially, they move the stock from going extremely undervalued to back to fairly valued.

16:56 That's what we're waiting to see for Meta. What we're waiting to see for a Microsoft as well. Because the numbers are there. It's not that the last couple of quarters this company has not done well. this company has done well. Remember two quarters ago, Azure growth was, I think, what was it? 38% instead of 39%? Stock dropped. I mean, come on. Is this a joke or what? So, the market can be, of course, irrational at times, especially in the short term.

17:27 For the long run, I don't think Microsoft is going anywhere. I think they're going to become a huge winner here as well in this whole AI story. What do you think? Let me know down in the comment section below. Check out all the links below as well. We'll see each other in the next one. Bye-bye. >>