AI Capital Expenditure (Capex) Is Off the Charts: Who Stands to Lose?

In this podcast, Motley Fool analyst Tim Beyers and contributors Jason Hall and Travis Hoium discuss:

• Fallout from quarterly reports from Kyndryl and Monday.com, and what may be next for both.
• Why the capex spending won't slow anytime soon.
• Whether the debt-fueled growth at CoreWeave and Oracle is sustainable over the long term.

Don't wait! Be sure to get to your favorite bookseller and pick up a copy of David Gardner's new book -- Rule Breaker Investing: How to Pick the Best Stocks of the Future and Build Lasting Wealth.

To catch full episodes of all The Motley Fool's free podcasts, check out our podcast center. When you're ready to invest, check out this top 10 list of stocks to buy.

A full transcript is below.

This podcast was recorded on Feb. 09, 2026.

Tim Beyers: Big Tech is spending big. Who could lose amid all the winning? You're listening to Motley Fool Money. Welcome, Fools. I'm your host, Tim Byers, and with me are two top Fool contributors and recurring guests. It's Travis Hoium. Jason Hall. Thanks for being here, Fools. How are you feeling?

Travis Hoium: Good. This is a wild market. This is one of the crazier markets that I've seen. We're near all-time highs, and yet we've got some stocks that are just getting absolutely crushed. Kind of crazy.

Jason Hall: There's a lot of investors trying to figure out which part of the market is right.

Tim Beyers: To that point, Jason, let's talk about two that are on the radar this morning as we are taping. This is on Monday late morning Mountain Time. Both Monday.com, ticker MNDY, and Kyndryl, ticker KD, are getting absolutely slaughtered. Monday has been down more than 20% through various parts of the day. Kyndryl down more than 50%. These two are a little different, though. We'll talk about Monday in a second, but just the story on Kyndryl as the CFO and the General Counsel both appear to have been dismissed. Kyndryl is going to delay its quarterly filings because it cannot verify that its internal controls over financial reporting are sound, and they need to report on this for lack of a better term and explain what the heck is going on here. I'm sure, Vasoas of the market, that sounds like fraud or chicanery or just incompetence. But either way, it's terrible. This is not one of those deep value plays that I think you spend a lot of time putting money behind. Monday might be different. I know you had a question about this one, Jason, because it did report. It beat estimates in its most recent report. The outlook, though, came in a little short on the revenue side and on the margin side. I'm guessing that that codes to the market as this company is getting disrupted. I'm not sure that's true.

Jason Hall: Tim, the company comes and reports mid-20s revenue growth, and you see a 20% haircut, you're like, what's going on here? You look at the existing results, and we saw operating income and margins fell. It really looks like in the second half of the year, they're guiding for substantially slowing growth and continued deterioration of the margins. To me, I think that's really the story. Sure, the revenue decline is slowing growth is that, but revenue growth is slowing, and you're losing. You're supposed to be getting more operating leverage for a business like Monday.com. That's not happening.

Tim Beyers: Well, it's not, but there's a couple of mitigating factors here. I will point you those of you who are listening who are members of the Motley Fool. First of all, we really appreciate you. We love our entire audience, but we appreciate you, those who are tuning in as members. There is guidance on the site. Please go to your services. You will see some guidance that we have issued, a little bit of analysis on these results from my co-captain on Super Nova Odyssey, David Meyer, and also Alicia Alfieri on the Rule Breakers team, she's also issued a bit of guidance here. But sum it up, Jason, super quickly, there are some foreign exchange effects here. That does have a drag on margins. Also, really early in the deployment of AI agents for monday.com. It's only right now at 1 million in trailing 12-month annual recurring revenue because they were just launched in October, but that's scaled really quickly. Now, over time, do I think that's going to be much higher margin revenue as that scales up across the entire customer base? Yeah, I do.

Jason Hall: There's a lot of expense that's coming along with that first group of customers.

Tim Beyers: You're going to invest in that to build it. But the customers are taking them up on it and putting AI agents into their Monday.com environment. That's a good sign. I don't see a lot to be really worried about here. But as I've said many times, it's always good to take a breath, pause. You don't have to buy more right now. You don't have to sell right now. It's a good time to let everybody else light their hair on fire. You do not need to join them. It's painful to light your hair on fire anyway. You really don't want to do that. But let's talk about lighting hair on fire. Boy, has there been a lot of spending, Travis? Like, huge amounts of spending, just a metric ton of cash on CapEx. That did seem to be the story of last week's earnings. I did a short analysis here. I looked at Gemini. I did not fully check the numbers, but I eyeballed it, and it looked right. At 400 billion over the last four reported quarters. This is among the top six companies. In terms of net capital expenditures. This is Amazon, Alphabet, Microsoft, Meta, Apple, and Nvidia. That's a huge amount of money. Here's what I want to know from both of you, starting with you, Travis. Is this worth it? Like, is this really worth it? Because this is one of the stories. Do we expect as investors to get something out of all this?

Travis Hoium: It really depends. It depends who you look at. I think it's the answer there. Let's start with Alphabet, because I think they were the talk of the market in 2025. The question for Alphabet was, are they going to be disrupted by OpenAI and by all these chatbots and stuff like that? The one way to not be disrupted is to outspend all of these companies so that your infrastructure is better so that your models are better. Alphabet, this is a scorched earth story. They are ensuring that no one is going to disrupt them. The momentum that they have with Gemini, with putting our AI tools into search, they are foreclosing, I think, that opportunity for disruption. It reminds me a little bit of Facebook [Meta's]. Was it about 10 years ago when Snap had a ton of momentum, and they said, You know what? We're going to copy exactly what you're doing, and their growth just stalled out. We're starting to see the Cary and the coal mine with that with OpenAI. Their growth is slowing, and Alphabet is actually starting to pull away some of those users. Actually, I thought a really effective ad at the Super Bowl last night. But that's the story for them. The other companies are a little bit different. Meta is going to have ROI on their own business. Mark Zuckerberg gets to do whatever he wants spending in AI because he controls the company. I don't know if we're necessarily going to get a payoff for them. Amazon, I don't know. They're late to the party, but you got to spend the money to play. Then Apple is not putting a huge amount toward this in the first place. When you look at those four companies and add Microsoft in there, too. They need to be in this game. They need to grow Azure. They are tied to OpenAI. It's all very rational. By the way, they may say in 2027, we're going to pull this back a little bit. Instead of spending $600 billion, we're only going to spend $400 billion. That could also be another rational move because those competitive threats by then may not be what they were six months ago and what they are today.

Tim Beyers: Jason, are we going to bigfoot this market? Is this Travis right here? Is that the strategy?

Jason Hall: I think so, but thinking about it from the next second-order impact. What we're seeing, with the exception of Amazon, really, is these companies are substantially changing the way they have managed their balance sheets. For years and decades before taking on the substantial amount of capital, a lot of it we're seeing is starting to be funded by debt, which is new for a lot of these companies to be using debt in this way in these large quantities. Now, they mostly have the money to do it. They have the cash flows to do it. But Meta, for example, has gone from a company that had massive amounts of net cash to, if they're not there now, they certainly will be within a few more quarters of having substantially more debt than cash. With this kind of capital investment, it's a hamster wheel, where you're constantly having to reinvest, to refresh, to stay ahead. It is an all-out sprint for these businesses. It is a land war right now. There's no doubt about it. Everybody's trying to capture market share. If AI is even remotely close to what we're starting to think that it could be, this probably will be a multiple-winners space because so much of what so many of these businesses do is providing the access to the compute, where they don't necessarily have to have the best AI, but you have to have good enough AI to sell to people, to monetize it for the ones that are doing AI that they're building themselves. In a way, I think Apple's strategy for what they do is really compelling to me because they are going to be positioned to do what they've always done really well, and that's have great hardware, a great ecosystem and interface that they build that they own, and then highly curate the things that have access to it. They could actually still be a winner in this, even though they're clearly way behind on building AI themselves, and they've walked away from it because they've been so bad at it. What I don't know is when inevitably we do outrun the demand because right now, let's remember that. Every single CEO that we've heard from has been clear. This is a demand issue. There is still far more demand than they can meet because they can't even get enough electricians and permits in place to connect these data centers to the grid. As long as that's the case, I think the inevitable outrunning of demand and overbuilding is probably farther into the future than maybe we realize.

Travis Hoium: Let's also keep in mind that the economics of this is probably going to be very different than tech has typically been in the past.

Jason Hall: Absolutely.

Travis Hoium: I think the analogy is going to be something either the airline industry or the oil industry, where if you buy an Nvidia GPU, you spend $100 billion on CapEx. If you're not running that at 100% utilization, then how do you get more demand? Do you lower your prices? Do you just run it under 100% utilization? The economics start to get a little bit more hairy when your demand does not quite hit supply. That's why you get all these games with businesses like airlines. How do you maximize the return? How do you keep the flight full, maximize the dollars that you're getting from each one of those flights? That's not necessarily something they're going to be able to do quite as easily. That's why I thought oil is the other thing where it's just a commodity. If tokens are just a commodity in the future, is that the business that they want to be in? That's why I think the first thing we need to look at is, what was the threat, and have they eliminated the threat? If you do that, then you can look at, are we spending this money wisely? In 2026, maybe not.

Tim Beyers: There's OK, let me make sure I heard you both correctly here. We are in the scorched earth stage of the market, and we have not yet got to the stage where we're fixing the land, starting to water the plants, and actually get a harvest out of this thing. We're still just burning crops, baby. This is what we're doing.

Jason Hall: I think that's right.

Tim Beyers: We're going to talk about in our next segment a couple of companies that are doing exactly that, and whether or not these are winners or are they losers? You're going to want to stay tuned. You're listening to Motley Fool Money.

ADVERTISEMENT: Could AI help you do more of what you love? Workday is the next-gen ERP powered by AI that actually knows your business. We help you handle the have-to-dos so you can focus on the can't-wait-to-dos. It's a new workday.

Tim Beyers: Fools, let's talk about a couple of companies that are subject. I picked these two, Jason and Travis, because I think they are especially sensitive to how much spending is going on in the AI sector. That is CoreWeave, ticker CRWV, and Oracle, ticker ORCL. Now, full disclosure here, guys. I have said publicly, and I am sticking to this, that I think that CoreWeave is overstretched and might go bankrupt. I've said this before, so I'm going to start with them, and I'm going to kick it to you to start here, Jason. I want to give you some data here. Feel free to tell me I am wrong about this because they have absolutely crushed it. CoreWeave has beaten the market by over 136% over the past year. That is absolutely outstanding. If you had this stock for a while, you have been a winner. But this is a company that has about $14 billion in debt, pays over 300 million in interest in recent quarters. They are levering up massively. They do have high-interest rate debt. They have some that is double-digit interest rate debt. The thesis here is that they have more than $50 billion in backlog that they cannot yet realize. They're building. It's what Travis said. It's scorched earth. We're building, we're building, we're building, and someday we're going to be able to harvest. Do you believe it?

Jason Hall: I don't think believing it is really the right question to answer. I think what it comes down to is, are the things that they're bringing to the table, and probably the biggest one is what makes CoreWeave a little bit different from the hyperscalers that are selling bandwidth is it's faster. It is faster. Like OpenAI, for example, is going to be able to speed up their process. They're going to be able to develop AI works faster. It's going to happen faster. Technically, that should lead to them being able to advance the product faster, make it better, get to a point where their ROI is better more quickly. The question is twofold. Number one, how long is that 10-15% advantage in speed? You can say speed. It's not exactly speed, but it's latency more than anything. Like, how long is that going to be something that's worth more money? Is the pause the stop and take your breath from all of the other hyperscalers, is that going to happen before they do get to scale? Here's the other biggest question. Is OpenAI going to be able to make their commitments, that's tied to the big CapEx and the debt that's there? Lastly, to make it even more complicated. Are partners like Nvidia going to continue to have the purse strings connected to the CoreWeaves of the world as providers of capital at need? How long is that going to continue to be the case before a Jensomong says, now we're at the point where it's good money after bad, and we have to just walk away? I don't know the answer to all of those complexities, but this is what I do know. I do know that it means that CoreWeave is heavily reliant on a lot of things that they have absolutely no control over and minimal influence to survive any sort of a slowdown in this sprint. If there is a slowdown, they get in trouble faster than anybody.

Tim Beyers: I'm going to give you a stat, Travis. I know you want to get in here 'cause it's important, but I ought to give you a stat. 2026 CAPEX is estimated to come in at about 220% of revenue. That seems absolutely bonkers to me.

Travis Hoium: It is. What I think it is is a confidence game. If you remember, I know Jason and I were covering a company called Sun Edison about a decade ago. The idea with Sun Edison was they could build wind and solar projects basically to infinity by simply selling some debt here, selling some equity there. As long as your cost of capital, your combination of the cost of those two funding sources, were lower than what your return was going to be on the asset that you were building, you were good to go. You could just do this over and over and over and over again. The problem becomes, when does the department of the market say no? When does the market say, You know what? You already mentioned the debt. CoreWeave doesn't exactly have cheap debt. You're not doing solar projects where the debt was 4%, 5% interest rates. This is higher-cost debt. The debt market is already telling us what it thinks. They're already saying, Hey, we don't quite believe this. If the equity market goes the same direction, then the whole thing topples over. That's really the risk that you have to look at with these companies. It's a very, very different calculation if you're Sundar Pichai, and you're looking at Google's cash flow and balance sheet, and you're just saying, why don't we spend everything that we possibly can to make sure that we're not disrupted by OpenAI? Because at the end of the day, if we need to pull that back, that's fine. We still got more cash coming in the next year. CoreWeave is in the same position. There is no next year if they don't get that funding, if they don't get the next project. It's a very, very different risk profile, and you start adding to debt. Everybody who has studied bubbles in the past will tell you the two signs that it's surely a bubble is when people start using debt and when nobody thinks it's a bubble anymore. That is terrifying. I will say, in CoreWeave's defense, what they say the CapEx that they are committing to is success-based, so they will not commit the CapEx until they have a contract that would allow them to apply the CapEx they're buying.

Jason Hall: But a contract doesn't matter if the counterparty papers and meets theirs.

Travis Hoium: It's paper. I have some doubts about that. Quickly, because we're running short on time here. Let's talk about Oracle. Oracle has a $523 billion backlog. They also have quite a bit of debt, but they also generate quite a bit of free cash flow here, so they just did, Jason, a $50 billion capital raise, half through debt, half through equity. They are just raising money hand over fist. It does seem to me that of all of the companies that are going scorched earth, this company is scorching the most earth relative to what it can bring to bear in terms of resources. Tell me, both of you quickly, do you think that this strategy for Oracle in particular, which has an interesting place, by the way, in the market, playing Switzerland in the data center market with these Exadata machines, it makes sense. Yet, I mean, this is a heck of a bet, Jason.

Jason Hall: I think in a lot of ways, it's a necessary bet, though.

Tim Beyers: We're back to it again. You must go scorched earth. There is no choice.

Jason Hall: Oracle's future would look vastly different if the company doesn't make an attempt to be a substantial part of the AI infrastructure, where AI lives, and having it connected to its own data because of its enterprise customers. It's become table stakes for a business like Oracle. I think the saving grace, though, is that if things don't turn out, if we don't have a panacea, Oracle still has a pretty good business that's there. It's going to be heavily encumbered. I don't think it goes away. Like, I don't think it goes to zero, and then somebody's buying the pieces off the scrap heap, like you could see with our CoreWeave. But I do think that they're taking a necessary risk. I guess if we were to ask the question between Oracle and CoreWeave, which of those do I think that there's a better risk profile? I think it's Oracle because the floor is higher, and the business is more stable. But I want to point out that at 7.5 times sales, before the past couple of years, this is the most expensive Oracle has ever been. As much as the stock has fallen, it's still not cheap.

Tim Beyers: Travis, where are you on this?

Travis Hoium: One correction for you. They are operating cash flow positive. As of three quarters of the growth, they are not free cash flow positive. Because of all the CapEx. But I want to clarify that because they have $100 billion worth of debt, and they're only adding to that debt load. They are leveraging up the company, betting on it. By the way, you mentioned the RPO, the remaining performance obligations. That was when the stock jumped in the fall. Who is providing all of that backlog? Most of it is OpenAI. Again, it's a confidence game. Does OpenAI get to be the thing that you think it could be in the future and live up to all these high expectations? Maybe, but we talked about the one problem with that early on. Alphabet is spending $180 billion making sure that doesn't happen. That's what's got to worry you if you're Oracle.

Jason Hall: Well, I mean, look at how quickly things change, though. I mean, we were talking about Gemini and Alphabet, how much it had moved from a really troubled business to a big leader. The past six weeks, the story has been clawed.

Tim Beyers: By the way, I will tell you, I have access to full Claude, and it's good. It's very, very good.

Jason Hall: We're going to be talking different names this summer. That's my point.

Tim Beyers: I'm sure that we will be. Fools, we're going to preview tomorrow's show. Stay tuned. You're listening to Motley Fool Money.

ADVERTISEMENT: The world moves fast. You workday, even faster, pitching products, drafting reports, analyzing data. Microsoft 365 Copilot is your AI assistant for work built into Word, Excel, PowerPoint, and other Microsoft 365 apps you use, helping you quickly write, analyze, create, and summarize. You can cut through clutter and clear a path to your best work. Learn more at microsoft.com/365Copilot.

Tim Beyers: Coming up tomorrow, more interesting stuff. We got a lot more earnings coming. Emily Flipping is going to be doing a Rule Breakers earnings roundup. She's going to be welcoming Toby Bordelon, and guess who else is going to be back. Jason Hall will be back. They will be talking, Datadog, Ferrari, and Spotify, all of which will be reporting on Tuesday morning. Please make sure to tune in and leave a comment for us if you have thoughts about that show, and leave a comment for us today so you can tell us what you think about this show. I know we went through quite a lot here, and we will take your questions and maybe do some follow-ups on them. As always, people on the program may have interest in the stocks they talk about, and the Motley Fool may have formal recommendations for or against, so don't buy or sell stocks based solely on what you hear. All personal finance content follows Motley Fool editorial standards and is not approved by advertisers. Advertisements are sponsored content and provided for informational purposes only. To see our full advertising disclosure, please check out our show notes for Jason Hall and Travis Hoium, our producer is Anan Choco Balu. Our engineer, as always, is the incomparable Dan Boyd. Fools, I am your host, Tim Byers. See you again soon, Fools. Fool on.

Jason Hall has no position in any of the stocks mentioned. Tim Beyers has positions in Alphabet. Travis Hoium has positions in Alphabet. The Motley Fool has positions in and recommends Alphabet, Kyndryl, Meta Platforms, Monday.com, and Oracle. The Motley Fool has a disclosure policy.