Jensen Huang's AI Capex Pulse Check

0:01Matt Frankel:NVIDIA CEO just startled investors. Motley Fool Hidden Gems Investing starts now.

0:09Matt Frankel:Welcome to Motley Fool Hidden Gems Investing. I'm your host today, Jon Quast, and I'm joined by guests Jason Hall and Matt Frankel, all subbing in for the regulars today. But we want to go ahead and quickly get to the biggest news of the week, and that was NVIDIA, a more than$5 trillion company reporting its financial results on Wednesday afternoon. And for me, this was as much as a macroeconomic pulse check of as much as anything. NVIDIA CEO Jensen Wong coming out and saying that CapEx spending for AI is expected to continue to go up. If you look at 2025 in the top five hyperscalers, So these are big businesses such as Google and Meta.

0:56Matt Frankel:These companies spending roughly$500 billion in CapEx in 2025. For this year, looking at around$800 billion, and some of these companies are starting to go free cash flow negative. And so you start to think maybe we're reaching a peak with AI CapEx. But Jensen Wong saying$1.3 trillion is what he expects to be spent next year just by the top five. That's a 60 % year-over-year jump if we take these assumptions. Matt, you're kind of pointing out here that NVIDIA, if anyone has a pulse on what is happening, I mean, it's really kind of coordinating everything. Yeah.

1:36Jon Quast:Jensen Wang, one of my favorite things about him as a CEO is he's not trying to deliver the best quarterly results. They are delivering the best quarterly results, but that's not his primary focus. He really wants to shepherd the AI build out. What I mean by that is, think of all, and we've talked on other shows about the circular deals and things like that going on in AI. NVIDIA is really at the center of it all. They're investing in all the frontier AI Labs. They have financing partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman, KKR to raise over$500 billion of third-party capital to really just invest in all of the little bits and pieces of what's going on.

2:21Jon Quast:There is some circular financing, and I have an issue with how that's being reported as sales growth. If I pay Jason$100 to teach me something and he gives me the$100 to teach him something. Did we each really make $100? No. And according to Gap Accounting, that would be$200 in revenue by those combined entities, even though it was just$100 getting passed back and forth. And that's what we're seeing in the AI space right now. But at the same time, that does help both of us establish our business, do the research we need. And there is some tangible benefit to that. So NVIDIA is really leading all that.

3:01Jon Quast:And it's a really interesting dynamic, but it's not just about this quarter. They're really driving that$1.3 trillion build out all by them, not all by themselves, but they're helping to drive that. I think it's important to note that even somebody like Jensen Wong probably doesn't really know exactly how this cycle is going to play out. They're going to get information sooner. They know what their sales rates are. They know what the orders looking like are coming in. And they know how quickly their partners like Taiwan Semi can actually do the manufacturing. But I think Wong is really leaning into the optimism and the numbers back it up, right?

3:42Even if there is a certain degree of hype. But the part of the story that may not be getting enough attention isn't that Alphabet and Meta recently and Oracle before that have flipped over to generating negative free cash. It's that they're doing it even as they're core businesses just continue to pump out gobs and gobs of positive operating cash. Here's a crazy number. Over the past four quarters, Meta, Alphabet, Amazon, Microsoft, and Oracle, those are the top five hyperscalers, they've generated almost$700 billion in operating cash flow. Maybe with that context, that big CapEx number we're talking about isn't really as scary as it seems.

4:22Matt Frankel:Of course, the difference between the operating cash flow and the free cash flow. The operating cash flow is what the business is producing, and the free cash flow reflects what it is investing in infrastructure, what we're talking about right here. So that's really the delta that we're highlighting. But I just want to talk a little bit here. It seems like investors have kind of lowered expectations here because you look at a business of NVIDIA scale growing 100 % year over year, trading at only 24 times its forward earnings. Matt, what are what are investors a little bit pessimistic about here, perhaps?

4:57Jon Quast:Well, it's not that they're pessimistic. It's at some point, the numbers just get too big. It's the same reason why Warren Buffett said Berkshire Hathaway's next 50 years aren't going to be as good as its first 50 years. It's because the math just doesn't work for 20 % annualized returns for that long from a$1 trillion base. And the same thing kind of applies here. So either growth, pricing power, or both will have to give at some point. 106 % growth year over year in this latest quarter. They're projecting 70 % growth next year. That was a big positive surprise. But 70 % from 106 is still a deceleration, it's worth pointing out.

5:39Jon Quast:The market is pricing that in. It's not going to grow by 70 % next year, then another 70 % next year, and so on and so on. at some point it would exceed US GDP within a few years at that rate. So it can't happen forever. And that's what the market's really pricing in. So it's really tough to evaluate NVIDIA on traditional metrics like forward PE ratios, because at some point it's going to have to hit a threshold.

6:05Matt Frankel:Well, let's turn a little bit now to another semiconductor company that, by the way, speaking of Jensen Wong, he's spoken very highly about this company called Marvell, ticker symbol MRVL. This is a roughly$200 billion company. Wong says it could be worth a trillion someday, but this stock reporting, and the big thing here is not that it reported 37 % growth for the quarter, even though it did. For next year, it's looking to grow. It raised its guidance from 45 % growth to 50 % growth. So an acceleration into the rest of this year and into next year, that would seem to corroborate a little bit here what Jensen Wong is saying, that CapEx spending is going to pick up even more, but the market doesn't seem to like this because Marvell stock is down a little bit today.

6:50Yeah, I've spent actually most of the early part of this week doing a deep dive into Marvell's business. And it's stunning how this seemingly niche company, and it is relatively niche, has just positioned itself for a massive, massive opportunity. Matt Murphy is the CEO. He's done an extraordinary job over the past decade of turning the company around and just pointing it right at the sweet spot of both what it's really good at and making it indispensable for some of its most important customers, which kind of happened to be these hyperscalers, and maybe most importantly, either developing or acquiring really critical technology that can keep up with the insanely fast pace of data volume and speed growth in the data center.

7:33Now, let's zoom out here. If NVIDIA's server clusters are the brains or other CPUs and GPUs are the brains. Marvell's technology is like the central nervous system of the data center. That means it connects the brain to every part of the body, no matter how near or far from the brain that it is. Also interconnects distributed sites together. So you have data centers that are hundreds of miles apart. Their technology is important there. But as to the specifics of the opportunity, last year they did an investor day around AI. the short version of what they think is attainable is about a$220 billion market for accelerated compute by 2028.

8:15They think they can get 25 % of that. That's a$55.4 billion revenue number. For context, this year, the company's saying they're probably going to do about$12 billion. I think they're going to do more than that, but let's just say they do that. We're talking about increasing revenue fourfold in about three years. Now, if it can maintain operating margins of 35%, I think they can probably do better than that. But let's just go with a baseline of 35%. What's the math look like? A trillion-dollar valuation would mean about 50 times operating income. Now, that's rich. But if growth does keep accelerating from there, it's really not outlandish, especially the stock right now trades for more than double that same multiple.

9:03Jon Quast:Jensen Huang did not give a timeframe when he thinks it's going to hit a trillion-dollar valuation. That's one thing to definitely point out. I wanted to point out that we mentioned the circular deals just a minute ago. Marvell has one with Google, where they pledged to give Google warrants to buy up to$12 billion of the company's stock, making them one of the largest shareholders, but only if Google is spending money with them. For all of those warrants to invest, Google has to spend$120 billion cumulatively through 2033. If that is the first of several deals, then$55 billion in revenue could just be a starting point, honestly.

9:42Jon Quast:If they get deals like that with the other hyperscalers, they're cutting into Broadcom's chip business there. There could be a lot more. It's not an outlandish prediction. I don't know how long we're going to see these giant valuation multiples because in 2033, I have to imagine the AI build-out is going slower than it is now. Like I said, the numbers are just going to get too big. I have reservations when it comes to, we're going to have some margin compression in the overall industry between now and five, six years from now. But it's still a pretty amazing business, as you said, for essentially a niche company to be making these deals and to, you know, they're putting their money where their mouth is when it comes to that market opportunity.

10:29Matt Frankel:Well, with AI infrastructure spending continuing to go up by tens of billion dollars a year, even if it's a decelerating rate, you better believe we're going to be talking about it on Motley Fool Hidden Gems Investing. But when we come back, we're going to be talking about something else. It's going to be called the mythos moment in cybersecurity. This is Motley Fool Hidden Gems investing. As an investor, I'm buried in data and making sense of it all is hard. That's where Claude helps me every day. I regularly give Claude a company's financial statements going back a few years and ask it to flag anything that looks like an outlier.

11:00Line items moving in a way that didn't match the trend around them. It surfaced a lot of things I probably have skimmed past before. Things like expenses growing faster than revenue or margins quietly improving while the headline numbers look flat. Claude doesn't tell me what to think. It helps me see just where to look closer. That saves me time, helps me find more opportunities to invest, and more risks to avoid. Claude is the AI for minds that don't stop at good enough. It's the collaborator that actually understands your entire workflow and thinks with you. Whether you're debugging code at midnight or strategizing your next business move, Claude extends your thinking to tackle the problems that matter.

11:36And with Co-Work, Claude actually builds real spreadsheets with working formulas, not just a CSV dump that leaves me with 90 % of the work undone. For problems worth solving, get started with Claude at Claude.ai slash fool. That's Claude.ai slash fool. And check out Claude Pro, which includes access to all of the features mentioned in today's episode.

11:56Matt Frankel:Claude.ai slash fool. Welcome back to Motley Fool Hidden Gems Investing. So Anthropic, this is one of the leading AI application companies for consumers and businesses. In April, it released something called the Mythos AI model. And the thing about this was it could quickly find and exploit vulnerabilities in software and it could exploit them faster than humans could respond. A few months later, actually, the US government asked it to pause Mythos for a little while. And so this was kind of a big deal. And it kind of caused cybersecurity investors to panic thinking, oh, no, the threats are getting much worse.

12:35Matt Frankel:But for CrowdStrike, it was saying that Mythos was actually great for its business. And accordingly, this week, it reported numbers. And Jason, this was actually a really great quarter for CrowdStrike. Yeah, it was extraordinary. And the thing is, the quarter was a good quarter, beat expectations, but it's really the guidance of the reacceleration of growth in the business. And it's another example, too, of a stock that is widely considered extremely overvalued, can still go higher when the business reports great results that beat even those highest of expectations. As we're recording this, shares have given back some of those gains are down a good bit late in the morning of the 28th.

13:16But CrowdStrike shares are still up like 10 % for the week and they're up 81 % for the year. The big driver is those expectations for accelerating growth. The company's calling for annual recurring revenue, so that's ARR, to grow about 41 % from where it was a year ago by the end of next quarter and then keep growing from there. This is a dominant, dominant business, keeps expanding its share of the market and also signing its customers up for more and more tools. At last count, more than half of its customers used six or more of the almost three dozen modules that the company offers. Now, it does trade for around 140 times my estimates for what their full-year free cash flow is going to be.

14:00But if growth keeps accelerating, I think free cash flow margin will explode higher. and it's a stock that could get a lot cheaper really quickly without the stock price falling just based on the operating leverage that they would get and their profits exploding.

14:18Jon Quast:Yeah. CrowdStrike's earnings were a blowout, even in the context of all the beaten raises that we've seen from the industry this quarter. The big number that I focused on, net new annual recurring revenue, that growth rate was 51%, meaning that the new annual recurring revenue they added this quarter was 51 % greater than what they added in this quarter last year. They've never done that before, even when they were in the really early stages of their growth. That's the highest net new growth rate ever. But I'm going to push back on Jason a little bit because if the 2020 to 2021 timeframe taught me anything, and Jason and I were very active in investing in that time, it's that valuation always matters at least a little bit.

14:58Jon Quast:Based on CrowdStrike's own internal goals for their long-term growth rate they feel they can sustain, the stock's trading about seven times the sales it will produce in a decade from now. So revenue acceleration is impressive, but I still have a really tough time wrapping my head around this one valuation-wise. Yeah, there's no pushback or argument from me on that. It is extremely, extremely richly valued. As long as it keeps delivering, that's going to be the case. But we all learned with the outage a couple of summers ago, So one speed bump and a lot of value gets washed out.

15:34Matt Frankel:I want to circle back to this mythos moment because as Matt pointed out, this record net new annualized recurring revenue for CrowdStrike, but it doesn't seem to be a tailwind for all cybersecurity companies equally. So we got reports this week from SentinelOne, Okta, Rubrik, and specifically with SentinelOne, You know, looking at 21 % growth and only about 20 % growth for the year, so slightly decelerating, especially compared to 22 % growth last year. I guess as I zoom out, I'm just asking myself, why is this a tailwind for CrowdStrike, but SentinelOne doesn't seem to be seeing the same uplift here, Matt?

16:15Jon Quast:Yeah, I mean, so the winners like CrowdStrike, they're already profitable. They're already funding their AI build-out through their expanding free cash flow. SentinelOne is, one of the things that stood out to me is that they recently cut 8 % of their workforce, if you remember that news. And they specifically said they were going to get the cost savings from that to invest in their AI security. So it's kind of like CrowdStrike's at a position of strength here compared to SentinelOne, first of all. And CrowdStrike's newer products like Flex, Falcon Flex is helping them win bigger, longer term deals than competitors.

16:52Jon Quast:Their customers want better outcomes at lower cost and CrowdStrike's delivering that better. The ARR coming from Falcon Flex grew by 101 % year over year in the latest quarter, talking about a blowout number. So it's really, they have a position of strength. They had a first mover advantage. They're an AI native platform. They were part of the original Mythos team that got early access and they've done a great job of capitalizing on that.

17:17Matt Frankel:Jason, is there any reason to hope here beneath the surface that Sentinel One is actually doing a little bit better than it looks on the headline number? Yeah, a couple of things. The stock's up 42 % this year. The business is growing very well. And the thing is that the context of comparing it to CrowdStrike, which you should because they're competitors, like direct competitors over the same customers, makes it hard. It's a giant shadow CrowdStrike cast. That new ARR number that we're talking about from CrowdStrike that is an incredible number. it's bigger than Sentinel One's entire business, right?

17:49Just the new business they're acquiring every quarter is bigger than Sentinel One's entire business. So that, I mean, that should really contextualize it. But I think the thing that matters a lot is if you look at, you know, kind of peel back the layers, pop open the hood for Sentinel One, where it's growing is really, really compelling. CEO, founder, Tomer Weingarten sat down with me and fellow fool, 10 buyers about a year and a half ago. And he told us, he's like, look, guys, AI is the most important, biggest threat to the enterprise and the biggest opportunity that we have in front of us by far.

18:21And you look at where they're growing non-endpoint. So again, thinking about endpoint, that's like a core offering for their business and for CrowdStrike. Non-endpoint offerings now make up more than half of Sentinel One's ARR. So even as CrowdStrike is dominating there, Sentinel One's growth is accelerating. Its AI security business grew by triple digits. Cloud and data are accelerating growth. So I've been saying for a while that I believe broadly there's going to be a lot of winners in cybersecurity. And I do think that just the space is big enough for companies like Sentinel One and to a lesser degree, Okta, in a different business because there's different needs to win share in this massive tailwind of opportunity.

19:05Matt Frankel:Yeah, when it comes to trends to pay attention to, I can think of few as important as cybersecurity. When we come back, Dick's Sporting Goods had its worst trading day in years. You're listening to Motley Fool Hidden Gems Investing.

19:19As a podcaster, my voice is heard by thousands of people. And now with Vanguard Investor Choice, I can be heard by the companies I invest in too. Vanguard Investor Choice makes it easy for eligible Vanguard fund investors to have a say in how their funds vote at company shareholder meetings. With just a few clicks, you can set your proxy voting preference and make your voice heard on topics like executive pay, board director elections, and more. Investor participation is the heartbeat of a healthy corporate governance ecosystem. You have a voice. Let it be heard. Visit Vanguard.com slash Investor Choice to learn more.

19:57Vanguard investors own shares of Vanguard index funds and those funds own shares of the companies they invest in. Available for Vanguard index funds that participate in investor choice. Vanguard Marketing Corporation Distributor.

20:11Matt Frankel:Welcome back to Motley Fool Hidden Gems Investing. You know I am a fan of obscure holidays and today is National Cherry Turnover Day. Not just any turnover, so don't go get an Apple turnover today. It is cherry turnover today. And I thought for the show we could celebrate a little bit. You know, cherry red, like a down stock chart. Turnover kind of sounds like turn around. So let's talk about some stocks that are down and in need of a turnaround on National Cherry Turnover Day. And I want to start here. I'm basically going to throw you a stock. You're going to make your case whether the company can turn it around or not.

20:50Matt Frankel:So let's start with PayPal. You know, PayPal, this is a digital financial platform that allows people and businesses to send money, check out on websites, even behind the scenes in some cases. It owns Venmo. PayPal is down today after Stripe reportedly pulling its bid for the company. It's down for the year, and it's down more than 80 % from its all-time high way back in 2021. Matt, you're going to go first here. Is this a candidate for a turnaround?

21:22Jon Quast:Now, PayPal is not going to retake its 2021 high anytime soon. But I can make the case that this is the strongest turnaround story of the three that you're going to mention here. So, I mean, the new CEO, Enrique Lores, is putting in some cost reduction measures. He's focusing on the highest potential areas of the business, like Venmo, like buy now, pay later. And it's starting to pay off. In the second quarter, EPS beat the estimate, the revenue rose 5 % year-over-year, which honestly, given PayPal's last few years, is pretty strong. Pay with Venmo did really well, Buy Now Pay Later did really well.

22:03Jon Quast:The long-term savings, they're targeting$1.5 billion eventually in annual run rate savings, they're already at$400 million in that. Free cash flow is strong, they're buying back stock hand over fist. They said the branded checkout product is stabilized, which was a big concern of investors. So even without the Stripe deal of possibility now, I think this is the strongest turnaround case here. Yeah, I think PayPal, like, I don't even, I think maybe you could almost say the turnaround is already happening. The numbers that Matt talked about, this is a business that has always generated tons and tons of cashflow.

22:39It's incredibly cash generative. I mean, it's a cash cow. I think really what's happened is investors have turned around their expectations of the business from a company that should be growing at much higher rates, taking share, expanding its margins. And, you know, there's a lot of us that have just been slapped in the face to the reality that it is a grind. It is a tough, extremely, extremely competitive business with a lot of big players that will fight over over their market share. And I've talked about this before on the podcast, as much as I didn't like the way things went down with Enrique Loras moving from the board chair to the CEO seat, and the abrupt way that it happened with Alex Chris, who he replaced, the business did need maybe a CEO whose background was more about being just a price taker in a tough commodity-driven business where you can't really create huge, moats.

23:37You just have to be a really disciplined operator, good blocking and tackling just the fundamentals of your business and do smart things like when your share price is down, buy more shares. Take that extra free cash flow that you generate and create value that way instead of trying to buy market share. And I do think that with the right expectations, investors can do perfectly fine in PayPal and they don't really have to turn the business around. They just have to let it be what it is.