We had one of our investors, I was on a call recently and he was asking, we're kind of having a similar conversation and I said, "I think it's going to be different this time." He basically almost wanted to end the call and said, "I'm going to send you a t-shirt that says, 'I think it's going to be different this time.'" Because that's a very dangerous thing to say. That's what I think it's going to be. I always think of the hardware side. Everything from Micron, Nvidia, the data center builder, energy, all that collectively is the brain of AI. The inference part really hasn't started. We've seen some pieces of it. A lot of people are trying things out. Agents starting to go, but for the most part, we haven't seen much on the value that's been created besides experimentation, but the brain just keeps getting bigger and bigger.
All right, welcome to a very special edition of the RiskReversal podcast because I have two of my favorite market prognosticators on this. That would be my co-conspirator here in crime. That's Guy Adami. Guy, how are you?
All right, I have a guy that you know, you've known him for a very long time. He's been coming on CNBC's Fast Money for over 10 years. This would be Gene Munster. He is the managing partner at Deepwater Asset Management. He's also been a massive, massive contributor to our podcast over the last five years. Gene, welcome back to the podcast.
You know, I say it every time. It's a thrill to be here and I appreciate you having me along for your magic carpet.
When do you announce the ownership of a Minnesota sports franchise?
Sounds like somebody's leaked some news to you.
Nobody leaked anything to me. I'm just curious.
Wait, is that all that SpaceX money there? You've been talking about this for years. He's been a huge SpaceX bull, by the way. Okay, just want to be clear about that. So, are you and Elon going to buy the Vikings or something like that?
All I can say is this is two things. Number one is that Minnesota hasn't had a championship team. I'm 55 years old since I was born. Actually two Twins teams have. But for the most part, we're basically owing my lifetime. That's number one. And number two, I don't even have a chance to even consider buying a team here. I would need many, many SpaceXs before I get there. That'd be fun. I think there's still a lot left in the AI trade, but for now, we're just going to observe these teams.
Noted. They'd be fortunate to have you as part of an ownership group regardless of the sports franchise. It is great to see you, Gene, by the way.
So, by the way, this is special because you're here obviously, Gene, but this is going to be the last podcast that Guy and I do in this current studio. We've been over here for I think three years plus. And our great friends and sponsor of the pod, Current, that would be Stuart Sop, CEO, co-founder, and obviously Trevor Marshall. They've done a bunch of pods with us. They are moving. And you know who's just engulfing this whole building? Palantir.
It's Palantir. And they're engulfing the building as Palantir is down nearly 50% from its all-time highs earlier this year. So maybe we'll talk a little bit about that. We got to talk about SpaceX. I know you've had a little quiet period. We're recording this Thursday into the close and man there is a crazy split screen going on right now. Apple is down nearly 5%. It was down 6.5% earlier. All the hyperscalers are getting smacked. But we have Micron up 20% after its blowout quarter and guidance. So Gene, let's start with Micron here a little bit because you know you have wavered a little bit I want to say in some of the hyperscalers. When I say wavered, you've gotten increasingly less bullish in them. I want to talk about what that trajectory has been for you, but you've also been a staunch bull on the semi trade, which is something that I can speak for Guy right here. He and I have been skeptical of the memory and the storage and this kind of last gasp of semis. So give us the 411 on Micron and what you think it means for like let's say the broader semis in storage and memory.
Well, I'll start with kind of like the most important point which is the number of SCAs they have. They've announced these are strategic customer agreements, SCAs, and again there's just a ton of numbers around any given quarter around the guidance but I think this is really the substance of why the stock is doing what it's doing today. So last quarter they announced their first customer that was a five-year customer. Now traditionally in the memory space it didn't make sense for customers like Apple to do a deal with somebody like Micron because the pricing environment is boom and bust. There's no reason to lock in pricing for 5 years. That's ridiculous because you wouldn't want to do that if the price is high now and lock in high prices in five years from now. But the biggest takeaway was that the number of SCAs went from one announced last quarter to 16 in total. So there's 15 incremental. Now some of those are not five-year customers. Some of those, probably six of those were automotive customers that they referred to as being small and three years strategically important long-term but three-year deals. But that does leave about six new five-year deals with large customers. And so effectively, if you do the math, it says that any or most of Micron's large customers have decided to go on this five-year journey with them. And why that's important is because when we think about the question about how long they're going to be out of supply demand equilibrium, they push that back a year. It's probably going to be somewhere in 28 or even 29. But how when that happens, the fact that we're seeing the customers, these are the people that are going to be putting the orders and these are the people that are seeing the kind of or they know what their plans are in terms of building these products and the data centers around the need for this memory. They're basically signaling to Micron that we're not near an end. If they thought we were near an end, that's the last thing you want to do is send a five-year deal. And so Guy, to kind of bring it in, I'll just say yes as two questions. One, like what's the key takeaway from Micron is I think the key takeaway is that their customers are seeing no end in what their appetite is to buy these chips, this memory. And then separately kind of the broader AI trade. You can kind of extrapolate that to we're still very early and or we're still much room to go in the broader AI trade. The market's not feeling that across the board today, but I think that that's a reasonable takeaway from the results.
All right, Gene, against the backdrop of I miss this entirely. Let me ask the following question. So, you just basically said we're taking the historic cyclicality out. We're not saying that it's not going to be cyclical five or eight, 10 years from now, but you know, in the short term, that's gone. So, if that's the case, the fact that they're going to go from about $75 of earnings to about $150 of earnings suggests that it's way too cheap. I mean, if this cyclicality is out of the stock, this stock should be trading somewhere north of $1,500 a share, if not higher than that based on everything you just said.
The issue here is that's all that math is correct. And if you think about I don't even know what the stock is up in the past year, like 1,200% or last I checked it was up 1,000% over the past year. is that the multiple has been unchanged over the past year. So that means that earnings have gone up basically a thousand percent. Earnings are going up in line with what the stock is going up. But Guy, to your point is that the muscle memory from investors around investing in memory companies is so powerful. It's still hard despite everything I just said about the SCAs. It's still hard for investors to give a bid because you know you just fast forward there's two narratives that are playing. One is that one side of the narrative is that yes things are going great but eventually all the law of large numbers impacts every company. So that's a negative. And then the second is that this again in the context of the stock being up 20% today but the second is that not only does all companies have to face the law of large numbers but also when it comes to memory there's just so much scar tissue about boom and bust. And so that's been the like I don't think Micron's going to get a like a respectable multiple probably ever. But I do think that the earnings power will continue to drive this higher and you know there's just not very many companies. It's really them and SK Hynix that are kind of running the show here and it takes so many years to get up to speed for other manufacturers. So big picture they're not going to get the multiple but I think the stock can still do well.
So, Gene, help me out with this. Okay, so the SCAs and we have, you know, RPOs and we have all these like three-letter acronyms that have popped up over the last, you know, kind of couple years or so. And in those sort of things that obviously investors, you know, they'll get behind if you tell them that there's a remaining purchase obligation of $300 billion and you know, that sort of thing and that you know, this was the Oracle story, right, going back to September. But why is it that investors didn't buy that? Okay. And did it have to do with because they're building this infrastructure that is not too dissimilar than what you know their far more deep-pocketed competitors that would be Google and Amazon and Microsoft have? Or when you think about this is it that investors in tech have gotten used to the fact that we have these cycles and really very few of them have turned out to be five or six-year sort of cycles without some sort of pullback in demand, some sort of pullback as far as investor appetite. Right. And when I hear this about these longer-term contracts now, is this a take or pay? Meaning like, you know, if you sign this contract, if you try to get out of it, you have to pay the full thing. You know, like these are some of the things that obviously if you go back 25 years from now, you thought that this was money good, but they ended up not being money good, you know. So I'm just curious is like why are investors willing to value these contracts right now right here at the sort of valuations that they're willing to do. I don't mean the stock valuation. I mean like what they're willing to assign to it.
I mean it all comes down I mean the central we're going to maybe name this pod. It's is it's is probably do you believe and the at the kind of the core of all these roads they lead to this question is do you believe that the and we've talked about it Dan before is do you believe that the substance of AI will exceed the hype and I think that the market is willing to believe in these long-term contracts whether it's what Oracle talked about last fall like you said is these huge numbers off of very shaky company you know shaky in quotes with OpenAI just being a relatively new company kind of all that growth coming from them. Commentary from OpenAI shortly after about them having some concerns that they may be able to meet that obligation with Oracle. But at the center of all this is investors. Yes. A question like what's going on is that there still is this underlying belief that this is going to be longer and be more transformative. Now that underlying belief if you're going to kind of think about the layers of this that's I think the foundation but above it there's this layer that is really concerned and that's why we see these kind of quick these movements in the market is that people get spooked, space goes down for a couple days therefore the AI trade is off kind of a thing.
So, you brought up Oracle and you don't need to opine on the stock, but it's pretty clear to me that Larry Ellison has pushed all his chips in the middle of the table and has said that Oracle is going to be either the epicenter or close to the epicenter of this entire AI thing. And he's made a pretty big bet that right now he's not being rewarded for. Does that make sense? Am I sort of framing it correctly?
You're framing it right that he wants to be at the epicenter. I think the challenge is that there are different people showing up at this table and we've recently kind of referred of the adult table and then the card table that the kids have been at. Oracle has been at the kids table. So has Intel and I think that when it comes to I talked about kind of the underlying theme and this question do you believe that the substance will exceed the hype is bringing these companies like Oracle and Intel to the big table. I think there's just still is again muscle memory around you know do they have the chops to do it especially when you put it in the context of what the tech companies who have been at the adult table for a long time the type of resources that they're putting into it and the success that they have had it's not just success in terms of what we expect to have happen if we look at you know Google's cloud business for example I mean going from 25% growth to what is it 50, 60% growth guiding to 56, 6% growth. I mean, you've just seen these huge changes, but with companies with real substance and real scale. And so, I think that that's kind of the issue. Yes, they're all in. I think at the end of the day, it's just again back to the belief question. I believe that we're still very early. I'd love to kind of go through my checklist of how I still think we're in the second inning. It sounds like it's intellectually dishonest to say that, but I think we're still that early. But I think that Oracle is going to be a beneficiary. I don't know if the stock is necessarily a good buy here, but their business will clearly benefit because I think that the infrastructure is just not going to be there if we kind of look across whether it's the adult table or the kids table to support what's happening.
Yeah. And I guess you know when you think about the innings okay you could still be in the second inning of the technology and its application and adoption and all that sort of stuff but it doesn't mean that the stock market or the names in which are doing this capex build and are meant to get the leverage and you know the return on these sorts of investments it doesn't mean that they're like in the second inning right? And I think if you look at the hyperscalers aren't they telling you something different Gene? Like when you think of Micron and you think of an 80 plus percent margin. They had negative margins in 2023. The company was losing money and now they're expected to have $100 in earnings. You know, their sales are expected to 10x this year from 2023. And I guess my point is is like when you think about what all of these server makers, right, like all of this is getting passed down the chain. And so when I think about the stock market and how it's being reflected right now, this AI trade, it really is only just memory storage because they have the pricing power. So like their margin is really at the expense of all their customers right now. And it seems to be investors are focused on that.
I think the central question comes down to demand. There's the margin question. There's another piece of this conversation which is understandable. I'm 55 years old. I was an analyst working at Piper as a tech analyst from '95 and became a senior analyst in 2000. Saw everything that happened. There's so much around what I saw there and just how spectacular that blow up was. But we look at what has happened with these companies. You talked about Micron and their margins and kind of look in the rearview mirror and of course we think about Nvidia. you know, the quarter I think it was their fiscal '21 there was it down 6% the revenue year-over-year. The and I'm not suggesting you're in this but I think that the trap is that to kind of look at what's happened and to say this is how the business has always been. The question isn't what has the business been in the past. The question for me is what's the demand outlook going in the future and who is there to meet that demand and the complexities around meeting this demand. This is not easy. This is not something that startups can do. When we think about on the infrastructure side, whether it's on the chip side, the memory side, building data centers, the energy component to it, it's just really it's tough to do. So if you're in the camp that you think that inference that thinking at scale is going to fundamentally change humanity. I mean I'm using these fluffy terms because I believe that this is going to happen. I get more comfortable at just saying I'm not going to look at the rearview mirror. Again the phrase that haunts me. I had one of our investors, I was on a call recently and he was asking, we're kind of having a similar conversation and I said, "I think it's going to be different this time." And he basically almost wanted to end the call and said, "I'm going to send you a t-shirt that says, 'I think it's going to be different this time.'" Because that's a very dangerous thing to say. But that's what I think it's going to be. I think that what we see over the next few years I'm just kind of add one point to it like try to put some substance around this is that one I always think of the hardware side everything from Micron Nvidia the data center build out energy all that collectively is the brain of AI the inference part really hasn't started we've seen some pieces of it a lot of people are trying things out agents starting to go but for the most part we haven't seen much on the value that's been created besides experimentation but the brain just keeps getting bigger and bigger and of course one way to measure the brain is to look at what the capex from the hyperscalers are. That's 50% of Micron's revenue is the hyperscalers. And for next year, as of last night, I haven't checked it today because I'm sure it's gone up, but the street is looking for 22% growth in hyperscalers capex. Now that compares to about 80% growth in calendar '26. So a meaningful step down. Nvidia is supposed to grow at 40%. That's what Jensen is basically suggesting his comments. So that's where the street's at for calendar '27. I think that the capex for next year is probably going to grow somewhere between 40 and 60%. I mean meaningfully higher.
Not consensus. I think kind of like the optimistic consensus might be 30%. But I mean it just it seems like reckless to even have that but if again I think these use cases start to present themselves the demand will be there. The need to continue to build will be there. I think the biggest rate governor on this whole thing is going to be around energy. But the so that's how I guess get some peace at being so bullish about what's going to happen is just the sense that you know the capex the brain will keep getting bigger and if you build the brain eventually the application I think we're actually at AGI now. We can talk more about that, but I think we've just made some spectacular progress and I think you both would agree that AI is going to be impactful and change a lot, but maybe to the degree that I believe is a little bit different.
Well, that's why we're fortunate to be able to speak with you, Gene. I mean that sincerely. So, you know, taking I'll take you at your word that we're in early innings and you know, I could push back, but it's not for me to push back. Here's my question. Understanding we're in early innings and understanding some winners will still emerge. In your opinion, who's done the best job at leveraging AI? I have some thoughts. But who's gotten the best return on their invested capital so far on the AI front in your opinion?
I mean, it's Google and I mean, kind of the irony if we were going to play the tape back and go back to a pod that we did back in April of 2025, I was declaring the beginning of the end of Google when Eddie Cue under when he was deposed related to the Apple Google relationship talked about the amount of Google search on Safari went down for the first time and thinking that basically chat bots are going to start to take away some of that energy around search. So I think Google is probably the best the company that has I mean at scale I mean basically they've done two things. Number one is they've leveraged their infrastructure. They've built out Google Cloud significantly. They've had this spectacular growth in Google Cloud. I talked about the change in the growth numbers before but search has gone from a 9, 10, 11% growing business. I think it was 18 or 19 last quarter. Like that's amazing. And so I think Google's a great example of the hardware side they've done right. They've also done the application side and the use case side with search. Now we still own Google. It's a company that we're still I'm not trying to send mixed messages here. We still think it's one to own. We don't and this wasn't your question. I don't think that Google is where the upside exists, the big upside. To get big upside, you got to find the companies that get a narrative change. That's what Google had over the past year and a half. And so I think to answer your question, I think that's probably been the company that has had just the most impact. And there's just so many layers, so many different levers that they can pull around AI in the future.
