Dan Niles: People buying Micron and SK Hynix are Missing THIS major thing coming

Dan Niles as our first guest here to get started. He was recently with us on our SpaceX live stream and he's a guest that's been calling tech cycles for decades and this year he's really been telling anybody that will listen that the smart money belongs in the semiconductors actually building AI not the hyperscaler spending on it. We've talked about this a little bit before founder and portfolio manager at Niles Investment Management and he's been comparing 2026 to some of those late 90s internet buildout all year. So really want to talk about this. Welcome to the show, Dan Niles. Let's bring him on now. Dan, great to see you again. Thanks for joining us. It's been a couple weeks here really just since the SpaceX IPO, but we're back again with another trillion dollar IPO here. >> Yeah, they just seem to keep coming, don't they? So, we'll uh it'll be interesting to see how this goes ahead and plays out given the interesting times we're in. >> Yeah, year of the trillion dollar IPO, one might call it. We still have OpenAI, Anthropic, and others to come. and they don't seem to want to settle for any less. This is a super fascinating industry. I think people might be even more interested in this memory head-on specific industry than they are even in SpaceX to be honest with how hot the trade's been all year. What do you make of SKH Highix, you know, coming to the US markets and it's a bit unique because this is a company that obviously has already seen massive growth in international markets but now coming to NASDAQ. >> Well, I mean, I think it speaks to the fact that you have multiple things going on. Obviously, memory or at least obviously to me, memory is a bottleneck. I think most people agree with that. The question is how long does it last? And I think SKH coming into the US market just speaks to the amount of capital needed to build out the infrastructure to support AI. So I always say it's never the data, it's how people react to the data, you know. So do you look at the fact that SKHix is raising high high 20s, billions of dollars is a good thing because it shows that oh my god the demand is still so strong they need to raise this money or do you view this as a bad thing because oh my god they're raising this money to expand. So I think for right now, you know, the stuff that's going on with memory and obviously you've seen, you know, violent swings in the Korean stock market and the semiconductor index over the last month or so, you know, every day. And I think the Korean stock market has had what, three trading halts, I think. Um, >> yeah, >> recently. So, you know, I'm more if you're asking longerish term, I just look at a very big picture and say you've had Agentic AI come out on January 30th. That requires, you know, and I'm using obviously the open claw formalization on January 30th as sort of the starting point. And then that requires 10 to 100 times more tokens um for those types of requests versus just chatbased AI. You've also got the Vera Rubin platform coming out that requires at least three and a half times more memory in the minimum configuration of that relative to the Blackwell generation. And so from a longer term perspective, I see memory still being pretty strong at least for another year as we go through this AI buildout. So, you know, it's a complicated answer, but that's kind of how I'm thinking through this right now. And in general, I'm bullish on memory over the longer term. In the short term, I think there's going to be a lot more uh, you know, in terms of stock prices being driven by what is mattering more, the token minimization as you see companies trying to control their AI spending bills or kind of the architectural shift underneath? >> Yeah, that makes sense to me. And I think people are really starting to pay attention as the need for memory heats up. I want to talk about this graphic that you had shared with me about China ultimately may have at least a 20 30% share. The United States really leading in drram industry market share from 1975 and that has majorly reduced as Korea's become such a large player. Looks like as of 2025 it was 28% US, 5% Japan, 62% Korea, 5% China. What do you make of this massive rise for Korea as well as this big decrease for US and Japan as well? >> Well, I what I'm trying to point out is that I see a lot of arguments that people are making these days of, oh, you know, don't worry about China at all because they're trailing edge and it really doesn't make a difference. And what I tried to show with that chart is and I, you know, I started in the industry. I'm an electrical engineer by training and you know worked at digital equipment corporation in the 1980s and if you go back to that period of time you know Japan when they started in semiconductors they were also trailing edge and you can see that the US had 90% market share most of that was Intel for those viewers who who don't know the history of the market they had almost all of that share and then Japan started to ramp up and Intel almost filed for bankruptcy in the mid1 1980s and they shut down producing memory and gone into this business called microprocessors. And then I was, you know, on Wall Street in the '9s and I remember looking and seeing how the Koreans were getting into the market and they were also trailing edge. But you can see by what's happened with the share, they went ahead and then they put the Japanese pretty much out of business. Now what you're seeing is China ramping up and people are like, "Oh, don't worry about the Chinese. They're several generations behind." But you have to remember China from an end consumption basis, they buy 20 to 30% of all PCs made, all smartphones made, etc. And so it'd be crazy for people not to think, especially when you know that Apple is out there trying to qualify CXMT. And by the way, they're going public next week on Thursday and they'll be raising close to $4 billion or so. It's crazy not to think that they're going to keep ramping that up, especially with the support of the Chinese government because when you're absorbing that much of the end market for electronics, you're probably going to want to get self-sufficient on that. And so longer term, they're going to be a problem. And especially because obviously China is a massive country. It's going to be stateup supported and that's going to cause an issue. And so that's more of a I wouldn't the the way I think about it is that affects your terminal multiple and how you should be thinking about it. And I just wanted to bring that up because I see a lot of you know people talking about oh it's different this time and I'm like it's it's not different this time. It's the exact same thing. It's obviously not going to matter near term, but from a longer term perspective, it China is definitely going to matter. >> Yeah. As I typicallying seeing them going from, you know, 0% to 5% just after so long not having been a big part of this market. Uh, good points. We're taking that graphic. Thank you. All right. So, kind of going from there, you talked about one year here at least seeing continuous pieces. So when we talk about skhinik IPO incere is the valuation attractive to you where it's coming out and then what are you comparing it to and how are you going about playing this memory theme? >> Sure. So yeah I mean it trades at a discount to micron and so you can say okay it it's trading at you know 20% to 30% cheaper depending on how you want to look at it. You know, the way I think about it is they they are the leader in high bandwidth memory, right? Samsung obviously has number one market share globally for memory, but SKHX really got HBM up and running earlier and then Samsung sort of was playing catch-up to SKX and Micron. So to me, the discount makes it actually on the margin probably more attractive than Micron is to some degree. They're all the same trade. Now, Samsung obviously has other things in there like cell phone division, etc., which kind of complicates this. And so, when you go kind of pure play, it's really SKH Heinix and Micron. And so, I can see that valuation gap closing over time. Now, whether that's because Micron comes down or SK highix goes up in the near term, we'll have to see how that plays out. But from a stock price perspective, if you're a US investor, you really only had one choice up until today or sorry, two if you want to throw in SanDisk, obviously. So you had two choices. Now you have a third and they're a really good company. And so I can certainly see money coming out of SanDisk and Micron and people putting some of that into SKH Heinix. So you have three names in your portfolio as opposed to just two. For me, if a space is doing well, I typically try to say, okay, I don't want as much company specific risk if I have a big picture theme, which in this case is memory is going to continue to be important. I rather have diversified exposure to the space unless there's some really super compelling reason not to. >> Besides for valuation, what's the biggest risk you think for investors as they potentially invest into this new player in the memory space that's coming to US markets? I think it's next Thursday X CXMT goes public and that if nothing else will get people thinking huh China's in DRAM and there's something public and they're raising money because I don't think because there is you know CXMT is private right now right and YC which is you know the NAND version u memory producer they're also private but CXMT coming public may actually have people who have longer memories going, hm, I remember the Japanese in the 1980s. I remember the Koreans in the 2000s. You know, what is the right terminal multiple? Because that's what everybody's debating. And so, I think that's going to be the very I'm highly curious to see how the markets take that. Obviously, they're listing in China, but them being public and you being able to see what their stock price is doing, I think will be very interesting because I think that's the long-term threat to the um to all the players, whether it's in Korea, Japan, or here. >> Do you think a variety of other players in the space, we could see them going public over the next year as there's just increased demand and people need to raise capital for these buildouts? >> Yeah, I mean, there's only one other major one, right? It's Yangzey in in China. And so I think yeah I given given the amount of capital needed I mean this is a very capital intensive business right and so while the money's available you should go and get it because as you've seen you've now got the biggest companies in the world whether it's Amazon or Google with that $85 billion offering. Amazon just did a bond offering just a couple of days ago. you know, the AI buildout is so expensive and the money right now seems infinite, right? Like there's enough money to fund everything, but money is not infinite. And so at a certain point, it's going to get more problematic for the longer you wait to raise the capital because the capital's already been allocated. So, um, it's going to be it's going to be fascinating year as this as this plays out. >> What do you make of these kind of graphics that show the money's just kind of moving around in a circle a little bit through the ecosystem and when people are saying, "Oh, all this money is being generated and revenue is coming from this when it's it's really just kind of players passing money along." Nvidia, SK, Highix, Micron, right? Do you really see actual value creation that is equal to what people are talking about in terms of the money that's being invested? Well, here's the interesting part, right? So, if you go back to So, you have Open Clock come out on January 30th. In March, you have people at Meta talking about their leaderboards for token maxing and everybody is cranking up and focused on, hey, how do we get our employees to use more tokens? Then you have Uber come out in April and say, "Hey, we blew through our entire AI budget in four months." And then you fast forward to June and you have the Coinbase CEO come out and put a post out on June 26th saying, "Yeah, we've gone ahead and added, you know, done five different things, but one of them is we now route queries to the cheapest models where we can." and we've managed to cut even though we're generating more tokens, we've cut our revenues um or the AI bill spend which ultimately turns into revenues for the next set of companies. We've cut that AI bill by almost 50% 50. So >> then you go, "Oh my god." Now for for those who don't kind of understand, >> if you go ahead and ask anthropic what 3 plus 3 is, that's that's stupid, right? But that's what you go back, you know, six months ago, that's probably what people were doing, right? You can route that query to a very very very simple model and pay a tenth of the price of sending it to Anthropic. And so what people are doing now is saying all right for the really simple queries like summarize my emails you don't need anthropic for that you can send that to a really cheap open source model. Now if you're trying to you know figure out some nuclear physicist stuff then yeah okay then I might want to use the anthrop anthropic model for that. And so that's a quick way very simple example but quick way to get your AI bills down. So that's what I'm really watching to see what happens when these companies guide for the September quarter. The June quarters are going to be just astronomically good. The question is what happens for guidance for September and now that the interesting development over just the last couple of days is obviously Meta coming out and saying hey we're going to sell Metamuse Spark at 1/4th of the cost of the anthropic models or the chat GPT models and then Grock also came out and had a big step up based on their work with cursor in terms of their model performance. So yeah, now you've got more even more lowcost alternatives. And so >> that revenue generation, >> it's going to get very interesting what that looks like for the next 3 to six months. And that's why back in June, I posted about the fact that I was worried about a speed bump um coming because you're switching from let me spend all the money I can on tokens to oh my god, I'm blowing through my entire budget in four months. I need to cut my my spend back. And so what does that do to the next three months in terms of the revenues being generated from AI as people go through that token minimization process? I do want to definitely double tap into that. One thing I wanted to point out here as well is that SKHYV is now live and trading. Um that is >> that is opened here and it is trading at 169 170.03 at this moment up about 14% from the 149 initial pricing. >> Interesting. Yeah, I mean it's going to be fascinating just to see the interplay between that Micron SanDisk as the day wears on and you know how is the money moving and flowing because obviously they just raised I think it was like 28 billion or so I don't know what the final number ended up being >> but somewhere and if you're a portfolio manager what you probably did was say okay you know I own SanDisk and Micron let me sell a little bit of that and put it in deskhine But you know there's a lot of other stuff going on. Obviously meta today is having a huge move based on what they've talked about um with regards to their presence in AI and then you know what does that mean for capex and so this is going to be really super interesting between now and then also next week when you have CXMT come public in China. >> Yeah, a lot happening all at once. I did want to talk specifically about well two topics here. one. Oh, first off, we also have Evan on here as well. Evan, you want to throw in a question? >> Yeah, I've been Can you hear me? >> Yeah, we got you. >> Wow, this is the greatest day of my life. I've been trying to talk for so long. Shout out Dan, I always love getting here, Dan. I was screaming like SKI needs to start trading. It's live. It's live. We're going here. How are you doing? Honestly, I just want to say hello at this point. I don't even have something. I've been so focused on trying to talk. I always love when Dan joins the the show. I am very interested to see what ends up happening on the Chinese companies end up coming in. Obviously, the US is always going to have some thoughts of not wanting to use the China product. At least this administration will not want to use some of the chips, but that doesn't mean it won't satiate demand around the world and then US can uh have some of that supply can be unlocked and and prices maybe come down. But my thought >> Well, there there's there's a nuance to that though, Evan, right? Because let's for example, right, the second biggest market for them is China. Yeah. >> So, is the US going to be fine with Apple buying Chinese memory for phones only sold in China? Because don't forget, we have midterm elections coming up in November, right? And so, the last thing that the US administration wants to see right now is prices for iPhones going up. Now, prices for iPhones weren't raised. prices for everything else that Apple sold was raised pretty much. So, I'm sure there's a lot of negotiation going on in the back room and I I would be stunned because I think there's a lot of people that think, oh well, you know, given security concerns, etc., Apple will never be allowed to buy Chinese memory. I actually think that's going to happen. And the day that happens, my guess is that you're going to see the US memory stocks getting absolutely clobbered because I think there are a lot of people out there that don't think that's going to happen. But Tim Cook is probably the best CEO bar none in terms of being politically savvy. I mean, he was one of the first to call President Trump when he was elected during his first term. And don't forget, they got exemptions from tariffs during Trump's first term, and then they got some exemptions this second term. So, if anybody can end up using Chinese memory, it's going to be Apple. And from a bigger picture standpoint, the Trump administration does not want to see inflation going up into elections. That's a huge political problem. So, I wouldn't be too >> I think that's going to occur. >> Do you view uh Apple raising the prices that they did, choosing not to do the iPhone? It came out right after that. They're trying to get the memory from CXMT, a Chinese company. I think that was actually kind of just a negotiating tactic there. The thing that care about is the iPhone. Once they raise those prices, that's coming in. Xbox also raised their prices on the same day. I think just one question uh here on this. If I'm looking at this from the AI trade side of this, this has become a bottleneck. Prices are accelerating the AI data center. You're seeing these increased AI capex spends. If memory were to come down in prices, I imagine these companies would not pull back their capex spend. They would just spend more efficiently. Uh and maybe that would loosen the demand a little. They loosen the market a little bit more. Is that environment where Chinese memory come in comes into the market a good thing for the AI trade as a whole? Maybe not these memory names in a pocket, but uh the AI trade as a whole. Maybe these hyperscalers. >> Yeah, I mean it obviously depends what which which names you're talking about. If you're talk talking about the guys spending the money, it's nothing but good for them, right? Because the biggest issue with all of these stocks is what's your return on invested capital. And so if you see capex numbers going up and you see EPS estimates not going up or revenues not going up or going down, god forbid, because all of a sudden your depreciation charges are, you know, they're outweighing any benefit you're getting from deploying AI, that's a huge issue. So, if you can come out on your call and say, "Hey, we're cutting we we're able to spend less on capex because memory prices have come down. Pick your number." Um, or they're not going up as much as we had originally thought. That's great for them. Now, the semiconductor space, that's going to be a totally different issue, right? Because people are going to turn around and go, "Oh my god, like the prices are coming down or the the guys at the back end are taking down the numbers." the first guy who's ahead and and there's a there's a sequencing to this too because don't forget right now everybody sort of has this view of and I'll I'll quote Larry Page of Google on this. I'd rather go bankrupt than lose this race. So nobody's willing to blink. The first guy that goes ahead and says, "Hey, I can go ahead and spend a little bit less because I'm doing it more efficiently." then I think you'll see everybody else kind of doing the same thing at which point a different spot obviously on the trade where you could see a lot of these semiconductor names reversing hard. Now I don't to be clear I don't see that happening. If anything with some of the things that have come out from Meta over the last couple of days, it seems like, you know, their spend is going to be huge next year because now they can go ahead and monetize excess capacity. But yes, in a vacuum, if pricing doesn't go up as much, that's great for the guys actually spending the money, which if you look at the Magnificent 7 this year, they're up, you know, 3%. Right? the semi-index is up 80. So you're going to see some of that 80% move go into the only 3% move in the magnificent 7 that have been hurt by the capex spend which has benefited% move in the semiconductor. >> That makes sense. Uh to go I have to say I apologize for coming in. Uh historically the memory trade has been considered to be a very cyclical uh industry and the definition of cyclical breaking is what we have seen right here. This is an Nvidia type couple quarters here for these companies. My question is what do you think comes after this? We know what the we know it takes two to three years for memory to come online. We know all these things. Do we end up settling settling out at a much higher place in this? Is this the like what do you think the future of the memor memory industry looks like? is that kind of cyclicality at a just materially higher level. Um I'm just curious. >> I I I don't think so because I I look at this very simply. The part and I lived through this in the 90s with the Koreans entering um the the market and you could argue an SRAMS with the Taiwanese entering that market back in the mid 90s. And everybody always makes these arguments and I was one of them way back then because I was a lot less experienced going, "Oh, well, you know, yeah, Korea's entering, but they've got trailing edge memory and so they're not going to be as impactful." And now look at it. The the two, you know, two of the three biggest are Korean, right? And they're leading edge. He is the leading manufacturer of high bandwidth memory. So I I look at this and I go, China has been put on notice multiple times during President Trump's first administration, through the Biden administration, and then now currently with the restriction on Nvidia chips that they need their own source of AI supply. they are going to ramp up high, you know, they're going to ramp up DRAM and ultimately they're going to figure out high bandwidth memory just like the Koreans figured it out, just like the Japanese figured out um you know back in the uh 80s. So for me, I don't view this industry any less cyclical than it was before because you have a brand new entrant that has incredibly deep pockets that's going to be funded by their own government that has an absolute need as much as nuclear weapons or aircraft carriers to have their own semiconductor industry. Ultimately, I don't view this space as any less cyclical than it was before. I know people are going to make that argument in the near term and in the near term it's right because of all the reasons you brought up. It takes a couple of years to get capacity ramped up but I also think about if you're getting more efficient with your routing of AI queries then that also introduces a wrinkle. That's why I've sort of said speed bump in the sense that you could have these stocks getting whacked um in the near term. But I do think they ultimately the semi-index goes to an all-time record high from these levels. But to answer your question directly, which is it is it any less structurally cyclical, I would argue it's about to get more cyclical very shortly because you're going to have China starting to ramp things up into the end of this decade. >> Great questions, Evan. I had a question off of the Apple piece. We saw Apple come out and raise the prices on a variety of their devices because of these increasing costs. Do you think that that's something that will continue to happen across the board kind of these downstream effects of these increasing costs of chips and memory and pieces like that? >> Well, that's the interesting part, right? So, this gets back to the thing of will Apple be allowed to use Chinese memory. the the bigger question is will all US companies be allowed to use Chinese memory that is being sold into the China market right so could that extend to Dell HP whoever and and that's going to be I think the interesting part because to your question which is which is a really good one if that isn't allowed to happen you have two choices as a company you either eat the margin miss your earnings estimates forecast and get your stock clobbered or you raise prices. I know what these companies are going to do. They're going to do exactly what Apple did. And to some extent, Apple and Microsoft has given everybody else political cover to go ahead and do that, right? The last thing that the US administration wants to see is prices going up. That's why President Trump had that tweet about how, you know, he he made Walmart cut prices supposedly. And so that's the headlines that the US government wants to see is prices going down. They do not want people going into the midterms seeing that prices are going up regardless of whether it's, you know, their fault or not. They don't want consumers seeing that the prices of things that they love to buy are going up, right? And so this all comes back to what Evan and I were talking about earlier, which is all right. Well, will Apple be allowed to buy Chinese memory? And I think the answer is going to be yes. And we'll have to wait and see. >> Yeah, it seems like it's probably the logical conclusion here because they don't want to see those rapidly escalating prices. And to be honest, you say people things things people love to buy. Uh pretty much everything that people loves to buy these days has a chip in it. And so it is yeah it's pretty pretty unanimous across the board. And then just one other question in terms of the token prices because you talked about routing to open models to basically reduce those prices. We were kind of sold this idea that hey the token costs are going to come down right over time. Like they're more expensive now they'll come down over time and it's been exactly the opposite of that. as these companies have gone towards IPOs and realizing that they need to put real revenue on the books are actually increasing those token costs especially on these higher quality models like Fable and others along those lines. Do you think that that is just going to kind of be the set pathway and we're just going to continue to see increase on these higher level models and that we really won't see that decrease that we were sold the idea of? >> Sure. >> Hope not. Our AI developer promised us those token costs are coming down. Well, it it depends on how you're defining it, right? Because is so is the cost to produce each token going down? Yes. But if what you're trying to do is something super complicated and you have models that can now do it, the costs are going to are going up for those cases. So that's why I get back to this blended case of let's say that you know and I'm I'm making up numbers here but just let's say that your your cost to produce a token went from 10 bucks to 20 at the high end but you've also got a model now that you can route to at two bucks. And so you put 90% of your volume through that model at two bucks. your average cost for tokens is going to come down, right? Because now you're being smart about that 3 plus three question going to the model that cost you two bucks and not to the one that cost you 10. And so the Fable and Mythos and those models as they can do more amazing things, yeah, it makes sense that the prices on those will be higher, but they'll come down over time too as the next version of Fable and Mythos comes out, etc. The one thing that's kind of different about tokens relative to other other businesses is because you're having to generate a new token each time. The cost of each incremental token in a query doesn't kind of come down, right? And so to some degree that scales linearly with the number of tokens. And so there's no added advantage to more tokens being produced in the sense that there was with some of the internet models where hey the more data you get the you know the the the lower cost it becomes going forward because you have all of that prior um data to train on to answer those questions. You don't necessarily have that with queries and you could say to some degree maybe you do if people are asking the same question on the World Cup and so you can cache that and serve it and so you don't have to generate a whole brand new set of tokens every single time. But you know there's also this thing where it's not quite as linear as when the internet was ramping in terms of cost coming down to you know what Evan just said. So yeah, there's a lot of moving pieces at the same time and um yeah, it's going to be fascinating to see how this plays out. So I'm I am tracking token production, but I'm also tracking cost per million tokens and that has been coming down really since um you know June. So for the last four weeks or so that's been kind of dropping and while the token production had been ramping and now that actually looks like it's flattening out too, which I have to admit kind of surprised me. And so we'll have to see because flattening units combined with dropping prices is not good. And that's not something I expected to see right now in terms of I expected the prices to come down but I didn't expect the token production to kind of start to flatten out which is kind of surprising. >> I will say >> that is a yeah and and Evan last year we do have to switch. We have a guest coming in right now. So Evan yeah and I get Dan a chance for final thought. Maxin thing that was popular for a while. Maybe some some people subscribe to that a little bit less. I've been seeing a little bit more. So, I wonder if that's also a part of it. But I always love having Dan on. Awesome guy. Very smart person. I always get uh more educated when you're on here, sir. Is there anything you want to leave the people with? >> No, nothing. I The one thing I would say is go read the Coinbase CEO's post on June 26th. I think that'll it's a great one and it gets into everything we've been kind of talking here about in terms of pricing versus cost versus um token production. So I think it it was a great it was a great post. So June 26 Brian. So anyway, take care guys. Thanks for having me on. Talk to you later. >> Thank you Dan. Appreciate you. Before we go into that, you can see on the screen right now for those that are looking to get exposure to SKH Heinix from a trading perspective, right? So, you have people that have talked on here today about, hey, I'd like to wait for this to come down a little bit. Well, if you're someone that thinks it's going to come down, here is a potential way to play it, right? SKHZ, that's going to allow you to have 1x short exposure. If you're on the other side saying, "Hey, this is undervalued in comparison to Micron already. I want to get my exposure, but this isn't necessarily a long-term hold for me. I want to trade around it." SKHX gives you that 2x exposure to the long side. Not long-term products, but great trading products from our friends over at Leverage Shares. So big thank you to them. Want to give a shout out to them. And you'll see these products go live on Monday, Tuesday area.