Why Bubble Talk is Totally Wrong TCAF 238

listened to my first Taiwan semi earnings call today. you did? It was very strange. thought I'm sure you listened to million of them. That the analysts ask questions and then the CFO repeats the question that has been asked for every single question. It's bizarre. think don't know what's happened there because he he literally he's like, "All right, so the question was I'm like, that's weird." And then he just did it the entire call. So, it used to be that there was like actually an audience and so people would ask in the audience and they were on stage. that was that was how it used to be. And but you heard the call. Like heard the question. don't know. But that that's possible explanation. Maybe it's cultural, you know, thing. thought it was odd that their their deck looks like it was made in the 90s, like their investment deck. What is that? don't know. Well, mean, it's it's only like trillion dollar company, so you got to give them some time. But it looks like it was made on MS DOS like honestly like pulled one chart but it just looks like it hasn't been updated since 1984. How do how do you listen to conference calls? Do you like you dial in or do you have you discovered quarter yet? Yeah, I've have Quarter but we use something called ERA. okay. What's that? And so it's like it's similar to Quarter. You basically go through and you just click click through into into call. Makes it really easy. So we're investors in Quarter but also like power users. I'm obsessed with it. don't know how ever used to because like seeing the words get illuminated as they're being spoken. Yeah. So, you're reading it and hearing it at the same time and feel like it sinks in better. So, it's interesting. talked to one of your sales guys. Yeah. From from Quarter. Okay. And told them that think forget exactly what the pricing was, but Era was giving us better pricing. Okay. know some people there. So, yeah. Well, think they actually use Quarter on the back end. really? Yeah. You should white labeling it. Okay. All right. How many how many calls I'm supposed to be? Yes. How many calls are you listening to each quarter? Like do you have do you have like people listening for you on some or do you get to everything yourself? no have we have team, right? and yeah, so we have team that that listens to everything and then selectively will, you know, I'll I'll go through and and listen to certain calls. Okay. What are the ones that you personally won't miss? mean definitively the hyperscalers. TSM did miss TSM today because had just personal appointment. don't worry, the eye story is intact. That's what they said. Great. read your notes. yeah, so don't know. It depends on which quarter it is. I'll I'll try to listen to like Walmart or Costco to understand what's happening on that side of the consumer, an industrial company, healthcare company. So, just smattering because there's no way that as portfolio manager is where we're going to be able to get through the entire portfolio. It's too Yeah. It's it's too many. It's like and and they're all together and it's like fire hose. Yeah. You know, it's great service. There's company there's subset called the transcript that subscribe to where they pull from like the most important quotes from all different companies that of course you can't listen to all of them. So that gives you good sense of like don't obviously don't listen to they're doing that with they're doing that with AI. don't know man. This guy listens to lot of calls but yeah I'm sure I'm sure. mean alternatively you can just go into perplexity and I'll I'll give them the the the lineup for the day and but no quarter can do that. They can all do it. Like give me the give me the best quotes. What did they say about this? It's it's great. all right this is going to be fun. So this is your first my understanding your first podcast. No. Yes. no. It's my No, he says no. What was Did Was you do Joe Rogan last week or something? What was What was your don't know. What was it, Scott? What What else? Morning Star. Morning Star. All right, fine. But we're going to pretend this is your first one. I'm sure it was great. Shout to Morning Star. I'm sure it was terrific. All right. th this this one's going to this is going to be lot of fun. And I'm so appreciative of you coming and doing this. Are you okay? Yeah. These are kind of uncomfortable. think you have them on backwards. do. think Why didn't you tell me that before? told you. just noticed now. So, this should be on the left your left side. Is that better? We got it. Yeah, that's better. My hair okay? You look great. Okay, great. Daniel, blame you. feel like you're not you're not on you're not on top of these things. So, all right, guys. How how how we looking? Our guest has nine more conference calls to listen to today. We're looking pretty good. Let me come in and search you up. All right. All righty. That's our theme song. So, usually the guest will deliver freestyle rap over this. Are you Do you have something prepared? No, do not. All right. We'll do it next talent. Sorry. Episode 238. Adding income has its advantages. Invesco's income advantage ETFs are designed to provide consistent monthly income and maintain growth potential, all with less volatility and downside risk mitigation. Access some of the world's best known stock indexes now with added income with Invesco's income advantage ETFs. Invesco, let's rethink possibility. There are risks when investing in ETFs, including possible loss of money. ETF's risks are similar to those of stocks. Investments in the tech sector are subject to greater risk and more volatility than more diversified investments. Before investing, consider the fund's investment objectives, risks, charges, and expenses. Visit invesco.com for perspectus containing this information. Read it carefully before investing. Invesco Distributors, Inc. My god. 238. That's your lucky number. It really is. Ladies and gentlemen, welcome to the 238th edition of the Compound and Friends, your favorite investing podcast. Arguably, and many would say the best podcast in all of investing. We appreciate you guys listening each week. We are doing, John. We're doing record numbers this year. Record numbers. All right. our our guest today has never ever before been on podcast. have it on high authority. she is Dr. Ancher Crawford, the executive vice president and portfolio manager of the Alger Capital Appreciation, Alger Focus Equity, and Alger Spectra large Cap Growth Equity Funds and more. One more. The Alger concentrated equity ETF. just counted four. You're doing four? Yeah. Okay, that's amazing. Congratulations. The ticker on that last one is CNQ and that was launched in 2024. Anchor holds several patents and was awarded fellowship by the National Academy of Sciences. She came to Wall Street 22 years ago and trained at Alger as semiconductor analyst before becoming head of the technology team and began her career as portfolio manager in 2012. in high school. She was the No, I'm just kidding. All right. Well, welcome to the show, Ank. Are you excited? You fired up for this? I'm excited. totally am. All right. So, we're going to have some fun. We're going to talk big picture. Encore. I'm I'm We're told it is pronounced encore. Like like encore. All right. apologize. Let's do it that way instead. Are you sure, though? Because my like the way that did it. All right. Encore. want to want to start with like the big picture stuff that everyone is talking about today, but then we're going to go much more in depth on semiconductors and your like sort of your expertise. Michael and have so many questions that we want to ask you and some of them will be like Wall Street person trying to understand the chip business better, but some of them might even be insightful questions and I'm certain by the end of this recording all of our audience will be much more knowledgeable about that area of the market. Sounds good. Yeah. hope can answer all your questions. Okay. Question one. What is the fastest land mammal? No, I'm just kidding. cheetah. All right. Can we talk about your approach to investing just generally speaking? How do you manage your funds and how do you think about your role as portfolio manager? Yeah, mean look, Alger is growth manager. We've been growth manager since the 60s. the idea of growth was really incepted by Fred Alger. And is that right? Yeah. Like he was the one of the first growth managers think ever. And so when when he incepted what growth is it it really came in two different forms like it it the essence of it was where is the change because often times where there's change there's unrecognized opportunity and if you can recognize that change before the market then you make lot of money right so you know that is the core of what we look for is an incredible amount of change or just change in general and that comes in two buckets one what we call high unit volume growth, which is typical growth company that is having high revenue growth and they're, you know, disrupting the market in some way. Typical, you know, they may or may not have lot of cash flow right now, but they're growing into it. This is what will attract your attention to stock, right? So, it's tight topline growth. and the other side of the equation makes us little bit different in our philosophy relative to most growth managers because it's what we call life cycle change. And these are companies that have already been through their growth initiative. They've probably saturated out their markets and then they have to figure out who they are, right? They've grown up and they're like, "Do have more growth left in me or do like what do do from here?" There are some great examples of that. Great example an example of that. Apple's an example of that. Microsoft is an example of that. You know, Western Dig right now is an example of that. the IP Nvidia video is an example of that. That's right. Right. So, it's it's as these companies try and reinvent themselves. Danaher was an example of this. Right. So, it's almost as if, you know, they they have to figure out how they're going to grow again or not grow, right? but we're looking for the ones who are inflecting positively on growth. And what's interesting about that leg of the the philosophy is that often times those are incredibly low multiple stocks. And people will say,"Wh are you buying value?" And what say in response is, "No, this isn't value. This is unrecognized growth." Okay, like that. The pitfall there, lot of companies try to reinvent themselves and can't. That's right. And so part of that thesis is, you know, what kind of management team do you have and are they capable of actually turning around the ship? mean, so think about Microsoft when Satia came on. you kind of have to hear what he's saying and then see what he's doing. Mhm. in order to execute on the next decade of growth. So that's really what we're looking at. The management team is crucial for executing that kind of strategy. Okay. And then away from the company itself, every once in while in the markets, we have moment where an innovative new technology emerges and the entire world changes right before our eyes. So, obviously we we're talking about AI today and I'd just love to hear your take on where you think we are in the AI story and are you still as excited about it now as you might have been 3 years ago when that change first came along and we got you know the original LLMs and people started talking about it for the first time. Yeah. think my excitement and my awe of the capabilities has only increased over the last three years. mean, you know, remember three years ago, we wrote paper called AI and the declining cost to create. And in order to to write that paper, there was lot of research that that needed to be done. And remember talking to lot of kind of AI natives, you know, people who were in the AI supply chain, talking to them about the capability of of what was possible. what is happening today is happening what thought would happen two or three years from now but you know the timelines have been pulled in and in part because the capability is just exponential something specific that's happening today or you're just saying like the pace of it think that like agents agents today and and what they're capable of today thought were going to happen you know in okay and that capability would come in 28 not necessarily today and so would would you oftentimes as as analysts we think linearly and the market thinks linearly but we are in time of exponential growth and so it's really hard to get your arms around what exponential actually means and what the end point is. Okay. and so for for me it's been mean it's continuous learning process. have never slept less and and worked more because there's just every day there's there's new updates that are happening and there's just something new to learn. What's like the most jaw dropdropping example of an agentic product or something that you've gotten glimpse at where you just said, my god, we really are living in the future." I'll tell you the weekend that OpenClaw came out. Okay. didn't sleep. Not long not not that far. Not that that was February. Yeah. and had taken my my my son's on the ski team and like had to wake up and like get him ready and like stayed up the entire night doing doing what? Playing with it or reading about it? was reading about it and at the time was at was you know had phone was at someone's ski house and so didn't really have access to to the capability of of what it could do but and and tried to launch one at home to be completely honest and you know it was kind of mess. was little bit afraid of the security aspect of it and but then then Perplexity Computer came and that was probably month ago. Perplexity Computer. Perplexity Computer. What What is that? Perplexity Computer is almost like OpenClaw, but it can't it doesn't have all of the functionality because it won't go log in and do things for you after having logged in, right? So, you know, it will again go do things like built an app and in that app was having really hard time publishing it and kind of putting or putting payments into it and Perplexity Computer completed the task for me. Okay. So, you have this thing paying bills for you on the internet. Well, this thing mean as as was telling Mike before you know have daughter who's going to college and the entire college process was so horrible. thought that no one should ever say more. No, no one should ever have to go through that. know. And so we we we hired these college counselors. So just recreated it using an AI bot. The what? The common app, filling out the common app. It's the common app, the colleges, and it basically just guides your child through the entire application process. And you feel comfortable enough that it's doing exactly what traditional college adviser would do? do. Really? mean we worked with three different college adviserss would say you know this your first one going yes okay it's lot of guts that you have that because you haven't been through the process before okay but now that saw it it was like and and it was actually my daughter's idea she's like you know there's lot of kids that can't get counselor you know why don't we democratize this process that's true and so you know on one of the snowy weekends we couldn't do anything so started building this app okay so it's filling It's filling out the common app with your kid and she's interacting with it and it's not telling her what schools to apply to, but like helping her organize all the things she has to do for each school. Well, it goes through and it it asks you questions like it asks you 12 pretty deep questions. think all of us should answer these questions because it tells you lot about yourself. Okay. and then it uses those questions. You tell it you put on it put in all your stats. You tell it what colleges you want to apply to. It takes the questions and just organizes it for you. So, it has version control. You can send you can send it to your parents or your counselor for comments. Okay. but it has something called strategic intelligence that allows you to allows the the kid to say like how am supposed to answer this question? Yeah. You know, given my background, given my stats, given what I've been involved with, how is the best way like what should really be saying? as explaining this. I'm starting to think like the colleges should almost force the kids applying to apply via AI interaction to prove that they're going to be ready for the world that's coming or that's already here. Quite frankly, it like it it almost seems like this should be part of the process. Yeah. mean, think it should be. built it. So, you know, think the point is like not necessarily the app, and and how it will, you know, can really change the experience of applying to college. But the fact that like I've never coded before. Yeah. And built this in probably it took me like 60 hours to build it. That's that's long time actually. know it was she's never coded before. So without it, it would have taken you 60 years. That's it would it would have been it would have been infinite because would never have done it. You just wouldn't know where to start. So based on what you said earlier about the type of companies that you buy, these companies are experiencing exponential growth. There's nowhere near saturation. mean they just started based on guess let's use like the most recent round anthropic raised at $800 billion. Would this be interesting to you or is it like too early in their growth phase? So interestingly enough, we recently participated in the last round and that last round 3 weeks ago was at 380 billion. My bad. Open AAI was 800 and open AI was 8 850 think. So okay so so we we did actually participate in that round. and in part what's happening is mean we have never before in our in in the course of the market seen companies that are staying private till they're almost trillion dollars. And so there's lot of value that our clients aren't really capable of touching. and so we're actually going and and finding really great companies that we believe in that we think will grow in the private markets that have road to IPO over the next two to three years. Which vehicle is this in? it's actually across all our funds. And so as part of the the guess the 40 act funds, we're we're able to take 15% of the fund and put into private into private. And is this the first time in the history of Elgra that you guys have done this? No, we've done it before. Like we were early investors in Palunteer back in think 2012 13 something like that. and we were investors in Chime. So it's not something that we've done often but we are seeing really interesting pipeline of companies. Now, if you if you have company like if you have companies going to $500 billion before they even file an S1 and you're growth investor, on the one hand, it's sad that they're not public and more people can't get access. But part of me feels like this helps make the case for active management, at least for growth funds. Like growth funds should be doing this as differentiator versus, you know, their their whatever their benchmark is, right? well, mean, we should be doing it in part because there's so much value for our clients. Yeah. That is residual and and we're not able to to capture in the public markets and then and then these things come public and you have to another decision to make. That's right. We are up lot. Do we stick around? mean, you probably have lock up or whatever it is, but yeah. And look, we haven't had to contend with that quite yet. Okay. and we'll see. It'll be really interesting as Anthropic and Data Bricks and OpenAI come public what happens. Well, yeah, because you don't know how the public market will react to it. Like if the public market vomits it, you might buy more. That's right. So, what sort of access of financial information do you have as investors in these companies? we get depending depending on which company it is, it's it's quarterly. We get quarterly statements. now you know not all of them are like have full they won't have conference calls and but we have access to the management teams or someone at the company neither is Berkshire Hathaway so don't don't worry want to want to talk big picture you you say that we are short compute globally short every everyone is short compute everyone's short compute for how long do you think that persists because one of the things if you're invested in lot of the stocks in the One of the things is like, am going to be like the last buyer? Am like am going to buy right before all of sudden there's like glut and everyone overbuilt and you know the capex plans start to moderate and that's everyone's biggest fear arguably that's kept people out of this for three years. some people but how do you think about that compute shortage and how much more runway there might be? So think of it in many different aspects. First of all, I'm going to go back to that first thing said that we think linearly. We've been trained to think linearly. You give someone two data points, what do they do? They connect them between here and there. That's right. And and it's also they extrapolate it. Right. So but what happens if in fact that growth is exponential and again like and and that's this is why people are struggling with the capex numbers and the growth numbers because we really haven't experienced exponential growth and so you know that that's one aspect do believe that this is exponential kind of growth in in the amount of compute that we need. and that is predicated on the fact that intelligence is when you start to build neural network it is an exponential problem especially my intelligence like it's usually exponential problem have always said this right in other words the am because when you talk about neural and you talk about like it's the amount of connections that's right and the amount of connections is infinite it's big number multiplied by another big number multip and for every layer beyond it's an exponential problem right so for every every connection here has you know an exponential kind of more connections beyond it right so think that when you think about how how we can use intelligence and as we try to recreate intelligence that demand is it's it's difficult to to fathom but it is very real so so Think about agents. You know, year ago we were worried about you know, are we are we in bubble? Yeah. Then we got agents and right now we're all stopped out on on the agents that we can use and the tokens we can use and you know, people are trying to token max and their their bills are getting out of control. Why is that? Yeah. Because once we can ask someone to do or something to do something for us that is productive, we will deploy it. The CTO at Uber said they already blew their entire 2026 budget on compute. It's and it's like April. Did you see Did you see that news? did see that. What What do you make of that? Is that Do you think that they just have the guts to say it and there are lot of companies in that same position because of the cost of all this comput? Yes. think think there are lot lot of companies that you know want to use more of the compute and now that it's becoming accessible. Yeah. and democratized because you don't have to be coder to use it and anyone can use it. mean, usage do think will go through the roof if they have the tokens. We're going to do some market stuff. but to segue there, does having the level of conviction that you do in the exponential opportunity enable you to live through the periodic draw downs not just for tech and AI, but just like marketwide? Is that big part of your portfolio manager persona? that's don't know if have portfolio manager persona, per se, but Well, it seems like you're very It seems like you have this like this innate bullishness about the the possibilities. And so maybe that would enable you to endure more volatility than somebody who is either skeptical or doesn't have that same level of conviction. Well, let let me give you an example. you know, in February got pretty worried about the market and it was kind of like guttural instinct of, you know, something doesn't feel right. We're we're starting to put ships into into the seas near Iran. Like, and instead of just sitting there despite my bullishness, you know, raised 7% cash in our in our concentrated equity fund. Yeah. Right. Not by selling anthropic. Not by selling anthropic. Right. and so it was kind of like more of like take you take everything across the portfolio just to raise the cash to to buffer yourself for volatility. that doesn't mean necessarily that you know you're you're incredibly negative on the AI trade. In fact, through the course of that period of time, I've only become more bullish. So, that doesn't mean there won't be volatility with that trade because not everyone believes the same thing do. Well, you can raise cash because you have the impression that others are going to sell and that that's going to open up opportunities. That's right. And that doesn't make you bearish. It just makes you, think, cognizant of the fact that other people are going to create those opportunities for you. And you have to have cash to take advantage of that. That's right. And and you know what? If you can sell something for profit and buy it back 30, 40% cheaper, you you do that any day. have dumb question that know Josh is thinking, but he's afraid to ask you, so I'll do it. when when you and everybody says we're short compute because everybody's saying the same thing. What literally does that mean? Like where are the bottlenecks and what does that mean for investors? I'll handle this. Electricity is very big bottleneck. Stop. Stop. Stop. It's enough of you. No, go ahead. well, mean the unit of measure that we're talking about is tokens, right? Okay. So, for Michael's benefit and some of the people in our audience, could you please answer the question? No. No. I'm Josh really wants Josh really wants to answer it. So you should let him. Why do we why do we measure this in tokens? What do does our audience is like 90% people that are either investing for other people or for themselves and just 10% absolute lunatics? What is the audience need to understand about the way that you're thinking about the compute shortage and why tokens matter to this conversation? And are all tokens created equal? That's really great question. one of the best questions I've ever heard. so tokens are unit of intelligence and think of it as mean technically token is two or three letters. So you send out question and it's 100 letters long, that's 30 tokens, right? that's technically what like your input token would be. But really you should think of tokens as being units of compute. And the more com the more thinking or the units of thinking and units of intelligence. And the more intelligence you need, the more tokens you use. It's as simple as that. Cuz not just your your prompt, but then the response coming back to you. That's right. And the tokens come from where? You buy them from these companies. The tokens is just think about as like if your chip is flywheel or if it's factory, you know, this is why Ry Jensen calls it an AI factory. So it's factory where you're saying I'm putting in three tokens. I'm getting out lot of intelligence or I'm putting in this many tokens and need lot of intelligence out. It has to grind through it and that grinding is compute, right? The computer literally has to carry out its calculations. That's right. And then reply back. So guess the question is it's like all right so pay monthly subscription to these companies to these LLMs for the privilege of using the compute. And the bottleneck is where? Like they can't get enough because Nvidia can't get them enough chips because Taiwan semi can't manufacture enough of them. like where where is it getting caught? right now it's you don't have enough chips, right? you you still don't have enough chips. Although there's other parts of the supply chain now that are get that are becoming short. So DRAM is you see what's happening in the DRM market. DM pricing's you know grown think 100% year-over-year and what that is unusual for commodity. Was that foreseeable for you? Did you know that was going to happen? You know what? Early on like two two years ago was like, "God, this is going to really pull on DRAM." But the memory market was horrible. So quite honestly, thought was wrong. was like, "My gosh, I'm not seeing response in the memory. You were so ahead of your time that nobody else had gotten to that conclusion." That's so interesting. The market was giving wrong signal and you just thought it was so obvious that that it would have priced it in your So you thought you were wrong when the market was wrong? Yeah. thought thought was wrong and then you know it basically just took off at the September of last year. So Micron was at $85 in August or August of last year. Where is it now? $450, right? So So it wasn't until just now that the market was like, my god, you know, we we we have this massive DRAM shortage." Because the the pricing had just started to take off. Mic Micron was like $175 billion market cap. think it's like 400 now or something, maybe even more. Is it really? only know the share price. don't know what the the market cap is anymore. But these things move so fast. It's hard to keep track of the market caps. You know what? You know what's the thing with the memory stocks? try to explain this to try to explain this to somebody that doesn't know finance or stocks at all. 515. Sorry to interrupt. And market cap. Is that nuts? It's half trillion. No big deal. No big deal. try to explain this to somebody who is like not in our industry because they're like, "Should buy SanDisk or whatever?" I'm like, "Yeah, don't care what you do." He's like, "No, no, like explain, explain it, explain it, explain it." So, rather than try to like go into like memory chips and shortages and it's cyclical, but maybe this time it's not cyclical. explained it. explained to him in language that he would understand. said, "Think about like restaurants in New York, right?" So, we have restaurant called The Corner Store. It's the hottest restaurant in the world. You literally cannot get in. I've been there twice, not to brag. people line up at at 3:00 in the afternoon on Thursday and like hold places for people. And that's just to put their name down so they could sit at the bar. Okay. You like cannot go in this place. It's like Taylor Swift's Diner. Okay. Was it good? No, it's great. It's It's unbelievable. But the the point is, mean, it's nothing. It's not like the It's very good. It's roast beef sandwiches. Like it's not like you're not going to fall out of your chair. And that's the point. There are 500 other restaurants in the same neighborhood, but people aren't lined up out the door of any of these other ones or maybe one or two of these other ones. And so my my point was like right now so many people recognize the opportunity in this small handful of stocks. there's like line to buy them and they sell off and then they recover just as fast because just everyone wants to be there. It's not that you can't buy another stock. It's not that you can't go to another restaurant. There's only one corner store. And if you open five corner stores on the same street, all next door to each other, each one of them would have thousand people waiting online because right now everyone just believes these stocks are going to keep doing what they've been doing or that restaurant is like the only place that they want to be. But it it changes over time. So if you're about to buy it today, don't look at what the chart has done over the last year because it's not going to repeat. or if it does repeat it would be extremely anomalous. So the the one contention would have with with that analogy is that the corner store stupid bastard. No, please give me is is that you know you're saying there could be many other corner stores around. Yeah. Yeah. but in reality because these are technology plays there's actually for each technology segment there's only handful of but you and are old enough to remember when when memory was the ultimate semiconductor related commodity. Okay. But let's talk through that life cycle. Yeah. So when started you know my career there was like 12 different memory companies. Three were in Taiwan. They were funded by the Taiwanese government. Yeah, there was like three in Korea funded. There was Micron. I've even forgotten all of the all of the memory companies now over time because of the nature of how difficult it is to and this is actually story of the semiconductor industry because it was so hard to actually produce things at smaller and smaller node. They were all forced to consolidate. Yeah. And so what happened? We basically came to this point in time as consolidated industry where you basically have three players providing DRAM. Sandis got bought by somebody and then came back. That's right. It got by Western DEG and then spun back out. These stocks were 8 10 times earnings when there was chip when there was glut in memory. You could not give these stocks away. mean there's eight times earnings in August. Eight nine times earnings in August. And so so the the the thing is like there there's only three companies you can buy to play memory, right? To play DRM, right? And so there is no alternative. So and only three suppliers to the industry. Forget about the stocks like the commod the the product itself. That's right. There's only three of them. Yeah. So there is actually the the reason there's line out the door is because if you want to play memory, there is actually no alternative. When you say play, you mean from the investor's point of view? from an investor's point of view. So, Western Dig in January 25 had market cap of 15 billion and it's now 125 almost. Is this bubble? No. And in part what's happened is like think about the data that is produced. So, this this is very interesting the problem is that Michael can only think linearly. You have to think exponentially. That's why love him. But so so think about all the data that spring produced. there's video, there's, you know, these agents that are producing things for you that need to now be stored. And so all of that data has to go somewhere. Now, the hard disk drive industry is very similar to the memory industry. There's two players, Seagate and Western Ditch, and they're not adding capacity, right? They're not actually adding physical units. They're increasing the density of their hard disk drives, which is how they add capacity, but they're not doing it at rate that is going to cause supply demand imbalance. Right. And that's another business nobody wanted to be in 5 years ago. That's right. You know, love hearing you hearing you say it's not bubble cuz these aren't your biggest positions. So, it would be very easy for you to say like, "Yeah, that is really stupid." Yeah. Well, do think that the bubble talk is really stupid. Go. But but we can we can get to that name names. No, I'm just kidding. Why is it stupid though? Because to take the the contra side at certain point like these stocks will not be rising at the same rate that they have been. So it could be that it's not bubble for the products and the chips, but it might be bubble for the investments people are making or and or it could be bubble in demand for compute because there won't be revenue at the end of the part. part that don't believe right at least at least through 26 27 and into 28 based on the capex that we expect don't think that we'll be in supply demand balance for compute we have that kind of visibility don't know if if 2030 is going to be that way right depending on how much capex we put in the ground it's almost like asking me in whether or not we're going to be in bubble in 2000 well mean don't know you tell me where the you tell me where the valuations are you can't predict what people will do right you can't predict what humans will do with their money. It's impossible to know what they'll do 3 years from now, but based on the fundamentals of the announced capex over the next couple of years, you think the market is still in an imbalance benefiting the sellers? That's right. Okay. That's right. And so, but but the the reason that we have to reconsider all this bubble talk is yes, the stocks have gone up. Yes, it's gone from 15 to 125 billion. But look what the numbers have done. mean just in the last three months Western Digit's numbers have gone from 10 to 25. Wow. Yeah. Right. So which numbers the earnings numbers for 26 or 27 one of those like 10 to $25. Yeah. It's like different company. It's completely different company and this is this is classic life cycle change for for us right where where the industry is changing because of you know the market changing. So so when like Nvidia for long time people be like isn't this bubble? Isn't it over for Nvidia? Well, tell me what the right numbers are for Nvidia. Tell me what the right numbers are for Micron and Western Digge. And I'll tell you if we're in bubble or not. And one of the things people have trouble with though is pattern matching and their overeagerness to say this thing looks like that thing. And won't even go into like.com stuff cuz it's played out. But just in the last few years, the solar stocks came and went. the electric vehicle stocks came and went. We had 20 electric vehicle related IPOs in 2021, 2022. and lot of them literally went to zero. Most of them are single digits even today. And that's despite their fundamentals having having improved. So people say, "Here we go again. They're bidding these stocks up like crazy. They think ABC. If and don't come through, these stocks are all going to crash." Okay. So, so, so let me let me comment on that in that the first thing how do you differentiate is is the the question. so, so let me just about the bubble and like people who often talk about this being in bubble right now. what have found is that sometimes it's almost easier if you don't understand something be like, it must be bubble." agree with that 100%. So there's lot of like portfolio managers that that have talked to and met with and CIOS that and it's interesting because the the PMs are often like they came from healthcare or consumer and getting your arms around what is happening with that kind of background is really hard because part of your advantage as growth manager is native to this sector. Personally, think so because it seems so obvious to me like what is about I've had I've never had this much clarity in my career and in part because it's it's pretty clear what's about to happen. Okay. So, you think you think that it's people that cuz we say bubble talk is code for missed out or I'm so smart and these people just made all this money in XYZ. it's impossible, therefore it must be bubble. Like that's how people think. know, but that's kind of crazy. agree. But Right. So, so, so, so that's on the bubble part. that thought was important to highlight because do think when hear people talk about bubble, always ask like, tell me, tell me where you think it's bubble. Is it just the capex numbers? Okay, let me just Why don't you justify? Why don't justify to you that capex number? And and can. Yeah. so first of all, can you can in second. Let me get to solar question because the solar question is really important one. Solar was massive market and in 2000 when was it 2000 when when TR when we had this company called Trina Solar and JK Solar there was all these all the Chinese panel makers the Canadian panel right right and all of those guys CSIQ Canadian solar they came they came to market and you know was particularly like graduated from Berkeley you know loved alternative energy thought like oil was evil and you know, would go to these meetings and the guy was like, "Yeah, we we basically built the solar company in my garage." And was like, "That's so weird. You you built you built solar company in your garage and you're IPOing for $2.4 billion, which now seems like nothing." But I'm like, "How does that work?" And was just young analyst. was like, "There's something wrong with this." Like, if can go I'm material scientist. can just go build spin up photovoltaic cell right now. That's right. And so like why why do these guys get to win? And so what happened there was the techn technological hurdle was so low that as if as long as there was any profit to be had it would basically get deprecated. It became manufacturing challenge to just make them the cheapest and of course China will always win. But then there was also this big government subsidy component to it which you know would come and go. but just like people lost lot of people lost lot of money. So, all right. So, in this case, the dollars being spent on this are real and growing. The profitability of it is high from the outset. It's not like one of these things where everybody has to lose money in order to subsidize it. few companies are willing to do that, but lot of these companies are coming out of cash flow in their capex. And that makes it very different. And and and think what you have to remember is if you look across the entire supply chain because from 2000 to 2020 we have had the entire supply chain has been consolidated. So think about that right you had you know semicap equipment consolidated over the last 20 years. the hard the box makers the box makers consolidated you know the chip makers consolidated the mean even like the networking guys consolidated and so you're coming at this problem from and and why did they consolidate not because they had to it's just the technology became so complicated that in order to go from like 6-in wafer to an 8 in wafer you almost had to join forces because the R&D budgets were out of Okay. Right. And and that happened for the chip guys too for for to do tape out which is basically spin up new chip. The the the cost of doing tape out was going up you know 3 four 5x for every single day to fit all the transistors do all the testing right so the way to do it is you consolidated the market so you just had more scale so the R&D budgets were advertised over more revenue. was going to say can't remember the last like chip IPO like we astero when and when was that? 2024. But that's the point like remember an era where there was semiconductor IPO every week, right? And it's just we don't have that. And now there are like just handful of semiconductor companies, you know, there used to be 45 all using the same three foundaries. That's right. And really one that matters. So can ask you this? think the biggest question for investors if we could see into the future that would put this debate to bed is the transition from the hyperscalers from asset light to asset heavy spending all of this money and see you're getting ready to cook and can't wait for for you to jump in like Amazon which which is one of your biggest positions. All right. So, they're spending whatever they're spending on capex. Are they going to get return on their investment that will allow for this to keep going? So, what would say is can't 100% say that they will get the same return that they expect. And and going into this year, this was my number one like back in before Meta put out their big capex number. remember talking to our team and saying I'm really worried that you know the hyperscalers are going to zero cash flow and that handoff of we we we value things on cash flow. You know as growth managers we we want to see the cash flow and they're reinvesting all of their cash flow. They're reinvesting all all of them are investing all of their cash flow. And that handoff can be tricky one because whenever you have valuation regime handoff to going from PE to cash flow or cash flow to PE from EV to sales to to earnings there's always period of tricky and so and we're living through it right now which is why they haven't done anything for while. That's why they derated. We used to prize how asset light they were and how high their cash flow was and now the story is different. Now they're in the most massive investments life cycle we've ever seen and there is no there is no more free cash because they're plowing it right back into data centers. That's right. And and what would liken it to is when they started building out the cloud. Right. So in 2015 11 12 13 14. Right. So so Amazon started building their cloud probably little bit earlier than that because they were doing it for internal use. they were spending lot. We just didn't know because they it was kind of inside of their It was Amazon. They weren't even telling people like right time and even Microsoft if you look at how they spent they were able to kind of build this cloud although they spent lot of money but it wasn't like this sort of spend was it? It wasn't this kind of spend but only because the cloud didn't come as fast as AI has come. So if we basically had said let let me give you the capability of the cloud with the you know with the 5G network and you know give it to you immediately the spend would have been significant and that's what's happening with AI like the the time the time horizons have have compressed so much and the capability is so it's exponentially growing. So what is Amazon spending the money on and where is the return going to come from? well they're spending the data centers chips cooling you know the entire shebang of you know how you have to build up data center what's the Amazon relationship with anthropic in your mind as an investor like how key is that it's like because remember 3 years ago Amazon was talking about bedrock AI and saying we're basically going to be bring your own model like anybody can use any AI model they want at AWS and the street like that and then Anthropic just took off for the enterprise and all of sudden it started being more like well you know if you want to play Anthropic the way to do it is via Amazon because of that relationship. So is that is that part of why you're so bullish on Amazon or not necessarily? No, look think think Amazon if you look at the valuation for Amazon today on PE on GAP PE basis. Yeah. and for forget exactly what it is, but it is absolutely not egregious. Yeah. you know, it's at like teens. Yeah. For on on an earnings, which is maybe the lowest valuation you've ever been able to buy the stock for since it's been published. That's right. mean, compare it to Costco or Walmart. It's think 10 points lower. Yeah. Than either of those. So, is that is that really kind of the right valuation for grocery store? It is grocery store. It is grocery store. That's right. But it's it's very efficient grocery store. The partnership with Anthropic and the investment like how important is that to your thesis for why AWS could could be one of the biggest beneficiaries of AI or is it not that important? mean it's important because it drives the it drives the top line for for AWS. Right. Right. So mean if AWS is going to grow in the mid to high 30s this year and part of that is because of anthropic. Okay. So it is important in that aspect but really for the duration because anthropic is going to work with other companies as well right they do and they will all right right and so it can't be the only crux so Amazon look they have tranium like just like Google has their TPU Amazon has tranium and so it is important to own the stack okay because it does lower your cost of compute Andy Jasse was talking about tranium and some of the other chips that Amazon is making and saying if we were to be valued or if people were to think about them as though we were selling these chips to third parties, it would be like $50 billion business. Do you think that do you think that the street is starting to don't want to say rerate because it hasn't really but do you think the street is starting to feel little bit more bullish about Amazon's ability to produce chips but more more importantly potential willingness to become big chip seller given the shortage of compute like could that be whole new leg to why the stock should go higher and own the stock so want you to consider my question to That's an interest. so what would say to that is that look, it's an important aspect of the technology. They've been working on tranium for long time. think they are behind the curve actually and they are catching up. So trrenium 4 is still delete all that. It's still an okay chip. It's not fantastic chip. think TPU is fantastic chip. Yeah. That competes with Nvidia. It's the tensor processing unit from Alphabet. That's right. and mind you they they they've been working on that for 5 years longer than Amazon's been working on their tranium products. Okay. So Amazon will catch up. think what is exciting about Amazon on that aspect not that they're going to sell their chips is that they can effectively make compute layer that is you know cheaper. So you can do compute on tranium that will you know be cheaper for you to run potentially. so more supply for supply constrained market. So well think it's just more more so to think about the stack. Okay. Right. So if if you go and say I'm going to I'm going to use AWS with tranium chip to do certain workload. You know might it be cheaper if done on tranium? It may be versus very broad powerful GPU that maybe is not necessary overkill. That's right. For that one task and it goes to your question about tokens actually like are all tokens created equal? We didn't answer that question did we? And the answer is no. Okay. So all all tokens are not created equal, right? fundamentally believe that this market is going to there's going to be tokens that you pay lot for and there's going to be tokens that you know when ask GPT what the weather is going to be right versus helping create an atomic bomb like very different. And so why should we be paying the same dollar value for each of those tokens and not quite the atomic bomb? don't ask it to do that. All right, get my kid into Princeton. got it. Same idea. No, think that's really important part of the conversation is that, you know, the initial build is like, all right, it's GPUs. These GPUs are insanely powerful and probably over it's like sending putting lion in your house to catch mouse. If your daily use of claude is like, what should have for lunch today? So, right. So the the market will figure out what to use for what or maybe already has. You're saying, right? It hasn't figured it out quite yet because all tokens are still kind of the same token value. Yeah. and so you know do think drug discovery tokens will have completely different value in the market than you know your educational tokens. So that won't impact me as user of these LLMs, but that will impact the way the enterprise version is being priced for corporate customers, right? And and also think it might be it might impact what model you use, right? So well, so people are already figuring that out. People are already saying, use this for this, use this for that. Regular people, not professionals. would imagine professionals, they go way more in depth on what which of the company's operations are we carrying out here versus there, right? Or or an app that you build, right? If you're going to build an app for drug discovery, you might, you know, you may you may price the app at level that it allows you to use the most advanced tokens or or the most kind of valuable tokens. Okay. All right. So, the products will be in part priced on how or or which tokens are being utilized. That's right. So, these companies keep saying that not spending enough is an existential threat. They have to spend and they're able to say that because we're in bull market and even though some of these stocks haven't done amazing in the very short term, like investors are still giving them the benefit of the doubt. is there it sounds like there it sounds like there's you don't think there's risk that they might say on call, hey, you know what, we're actually going to take our foot off the gas pedal, but like is there chance that the market and know we're speculating here that maybe the market could force their hand like if Microsoft didn't stop going down and the stock is down 45%. At what point does Satia say, "All right, guys, like maybe we should chill out with all this." The metaverse example, the like the stock market enforced the end of metaverse spending. think we all would agree on that. like that. He probably would have kept going two more quarters had the stock not been in 70% freef fall. Now Zuckerberg controls all the votes. So it really had to just be his acknowledgement that all of these investors who are selling the stock can't all be wrong, right? That's the that's the concern there. think think again it's very different than the metaverse because the metaverse the promise of the metaverse know believed in it by the way. Credit was actually actually believe in the metaverse still but you need an AI. land did you buy in the Did you buy beach house next to Snoop Dog's virtual beach house? just played Roblox with my kids. So, but you know, the metaverse is actually do think it can be reality. It just needed the capability of AI to get there. And so, in 2020, like remember doing doing like little thing for for Alger on on the metaverse and said, "This is 2030 event, right? It's 10 years out." So he was trying to invest in 2020 or 20 when was that 21 2020 2021. Yeah. So change the name of the company for right to to basically get to 2030. That was long period of time by also all by himself. All by himself with kind of deprecated compute, right? So he didn't have GPUs then that were going to drive this AI revolution. So this is different. This is not that. So this is different and and in part because you know everyone's so worried about the ROI like will will we get the ROI back and what would say right now at this moment in time they're basically making their money back in 18 months. Wait, how? What? So the cost of compute is actually rising right now. If you look at the cost of an H100, it has been rising. It hasn't been going down. And so you look at the compute stack and and how the the per hour pricing of chip inside of one of these data centers, the payback is 18 months. Meaning the meaning because the price of the chips is going up. That's right. An investment that you might have made 18 months ago, now you're in the money. But wait, is that the right way to measure it? For now, it is. But the point is, what is the ROI? Right? So if you get paid back your capital in 18 months, then do you really have to justify your ROI? Now the other thing think that people like miss in this is that they can turn as soon as they see the demand start to decline or get curbed. Guess what? They'll they'll turn off their capex and we're just we're nowhere we're nowhere near we're nowhere near that. But the point is it's not like it's not like the telecom, you know, bubble. Let's go next. So this is the other thing the bubble people will say. Look at these deal announcements. Every one of them is circular. Every one of them involves company vendor financing some sort of an investment or paying customer to pay them back to either buy compute or chips or use of data center or lease something. and I've heard Jensen I've heard Jensen deflect that. It's like, of course, it's circular. We're all in the same ecosystem doing business with each other. How could it not be? So, you know, and I'm sympathetic to that, but like what would you say to people that that say, remember the late 1990s when one company need to make needed to make earnings this quarter, so they would do deal and then the next quarter the other guy had to make earnings, so they would do the reverse deal." understand this is not the same thing, but how do you answer that? answer that by look, every deal has been different. think that Amazon metad deal worries me more because you're giving away part of your company to say use my chips and I'm going to give you 10% of my company. What's the Amazon Meta deal? not Amazon AMD metadata deal. okay. Right. So Meta Meta gets 10% of AMD if they get to certain amount of revenue and AMD chips. Right. So effectively they're giving it for 0% gross margin. Yeah. you know that to me is little bit more quote unquote circular open and and meta open and meta both are both had deals with with AMD that that look like this. Nvidia basically seating the market to enable the market is less worrisome because mean Nvidia's revenues on on AI like the AI based revenues is going to be hundreds of billions of dollars. Yeah. So the two billion that they gave to Cororeweave is like are they really enabling their own revenue? Not really. Well, they're seating partner who if that partner is successful, it means years and years and years of future chip sales. That's right. Why is that bad? Why is that bad? don't think it's bad. It's not bad. think it's actually common sense. It's It's not Well, we see examples of this in every single industry is another thing to point out. So, you see, Pepsi gives restaurant all kinds of free advertising and banners and things to decorate their bar and therefore sells more Pepsi into that bar for years and years to come. This is just high-tech version of it with billions of dollars, not thousands of dollars, but it's like fairly common thing. Yeah, that's great example. you know, do think you probably see it. Millerite Miller Light says, "Here's billiard's table. put it right in the middle of your bar and the lamp hanging over it is going to say Miller Light and the rest the the tavern owner is like, "Okay, cool. We have pool table." they're buying Miller Light forever and the bar does better. And like this is obvious stuff. Yeah. We're just talking about it on much greater scale here. Yeah. And and that's what scares people. think again if you if you think about the magnitude of what's happening, all my metaphors have to do with eating and drinking. don't know if you're picking up on this. So the audience definitely is. so so the you know the the the magnitude of what's happening here is just it's hard to fathom. Again if we go back to that like it's really hard to think how big and how massive this change is and that's why think people are just struggling with it. Okay. So give the audience give the audience three stocks that will double before the end of the year. no. I'm just kidding. Can we talk about like some of the actual investments that you're making in the public markets? Sure. Okay. all right. So, we're looking at some of your holdings. Wait, can can we go back to one thing? Yeah, shut up, Josh. So, so one of the things wanted to go back to is that this fact like in in 2000, one of the issues with what happened in 2000 in that 2000 bubble is the capex that was spent spent was longived assets. So, you put it into the ground, you think about like fiber optic cable, you put it in the ground, you spend billions of dollars and they were effectively longived assets. like nobody needed them for 5 years. No one needed them for 5 years and then demand fell apart. But you also didn't have the technology to take to take to make use of those cables. There was no web video. There was no mean do you remember using the internet in 2000? Yeah. It was horrible experience. Yeah. It was like chat rooms where you're typing. It was there's no social media yet. There was no YouTube. There was no there mean you could barely like download web page without it you know glitching or take like 10 minutes to download song from Napster like that was the internet right mean 2004 2005 like downloading movie would take two hours yeah it would be it would be horrible experience and so we had dreamed the dream in 2000 but we actually didn't have the technology was timing mismatch right before the demand for all that stuff they invested in And what's really different now is actually all the technology is in place. And it's like I've likened this to when I've I've spoken to advisor just to make it little bit more tangible. I've likened it to you know if you were an F1 car engineer. Thought you were going to say Burger King. She's not using food analogies. Sorry. can't think of good food analogy, but if you if you were like, you know, an F1 engineer and you dreamt up this amazing car and you're like, "This is the way it's going to work and think I'm going to win because, you know, I've I've dreamt this amazing car up and you get to the race and you have four wheels." Yeah. Right. This time you've actually built the car and the only thing that you're waiting for is gas. Okay. Right. or you know electricity to charge your battery. But like the point is that now the only thing you need is the compute. You have all the other place. The tech is here. So therefore the capex is going directly into something that is currently working. That's right. But speaking of the capex, there's another doctor investor who's on the other side. Dr. Bur says that there's accounting shenanigans which with which how they're advertising these expenses. Yeah. don't think that's right. So he he's basically saying, they're using 5year depreciation, but they're six-year depreciation and they're actually four years, which is absolutely just incorrect." So why is he why is he saying that? don't know. You should ask him. He's saying that the So you're saying these chips last really long time. They might not be the newest. They might not be the newest versions, but the A100's are still working. And those are from 8 years, seven years ago. Those are from 20 what? 26. They're from 2021. Sixy old chips are still That's right. very actively being used. mean, the idea that these computes are just or that the the the chips are going to die in 3 years. don't don't really understand it. All right, let's do some tickers. So, some of your holdings and sort of know little bit about some of these. but why are you bullish on? Let's we'll just go like few. tell us about Nebius. wow. What that's like big long story which is really fascinating. will try to shorten it for you guys. Yeah, keep it to like 90 minutes. understand that Nebius my my understanding is this was originally part of Yandex. which was the Russian Google. It wasn't part of Yandex. it's what Yandex became? No. So, so Yandex was the the Russian Google. Yeah. headed by man named Arati. Yeah. And when the Ukraine in Russia, Google searches you. That's right. All right. Go ahead. When when the Ukraine war happened, you know, he had just disagreement with the the politics. Okay. And you know, he was one of the quoteunquote oligarchs that you know was sananctioned, but he was also on on the outs with Putin. He ends up leaving. 1500 of his engineers end up leaving. think Arotti paid for all of them to leave. And then they he sat there. was like, "Okay, have $2 billion. What are we going to do?" And they said, "The future is really AI. We know how to build data centers. Let's build AI data centers because we think because we think we can do it better than everyone else." Okay? And so, you know, remember they were Yandex, they got delisted. So, it's in Europe. Now, it's in Belgium or it's in the Netherlands. Netherlands. They're based in in the Netherlands. But one day they get this call from the New York Stock Exchange being like, "You're going to get listed on Monday." Right? And they're like, "What do you mean?" Like, "We're like private company right now." And it was Thursday. So they worked all weekend. They get listed and all of sudden there was this like weird ticker on the on the exchange NBIS. Yeah. And didn't know what it was either. Yeah. And so like was like, "Is it wart?" Like what is this thing? People were trading it though. People may have And but you know what? Initially when people traded it, it was it was all the people that were stuck in the index, right? Right. That were like, "Thank gosh, could get out now." And and so they're like, "Get me out. I've been stuck here for years." And then and but they didn't have it. No one knew what the story was. So they put out this 110 page deck on online. And don't know. remembering them being like, need meeting with you." And they had no time. Yeah. Yeah. So, hunted them down in Davos and as one does. as one does and it was and just got to we got to know the company really really very well. Okay. So, you you look at this as cuz know they're in autonomous vehicles and they have lot of stuff going on, but like you look at this as data center play that's doiciled in Europe but doing business all over the world. What would say is that this is actually potentially the next hyperscaler, the next AI native hyperscaler. Really? Yes. Okay. tell us about tell us about QXO. QXO. QXO is fantastic company. It's not it's not typical tech company, right? but we all know Brad Jacobs. Brad Jacobs was the, serial entrepreneur. he put together URI, XPO, left XPO and began QXO. QXO is building products company. He started it, probably about year and half ago. bought Beacon Roofing and he's basically going to consolidate the industry. Okay. What's really exciting about this is that it's Brad Jacobs. we really like you talked about the management team. we know what he is. That guy's amazing. He is Brad Jacobs might be one of the most impressive people. Why is he in roofing right now? because he's consolidating the home the the building products market because it's incredibly fragmented. Yeah. So, what he's seeing is that well, there's all these mom and pops that are in building products. Tile and like like all kinds of materials. He will he just bought Kodiak. Yeah. And he's starting to get into the rest of it. So, he started with roofing, then he's going to get into everything else. He's going to take pricing, add efficiency, add technology. So, you're growth investor. Like, how much potential do you need there to be in story like this to include in your portfolio? Do you think this is like if it all know we don't know that what's going to work but if it really works you're not looking for like 20% return that's not how you invest. No. Okay. So let me give you just like the framework for how he thinks. Beacon is about billion dollars in IBIDA. He thinks that he's going to increase the IBIDA of this company to 5 billion in 29 to30. That's big deal. That's big deal in quiet sleepy industry. That's right. And he's done this over and over. This guy He's done it over and over again. never heard of this guy. think maybe four years ago he was on Patrick Oonis' podcast. He's been on think think again. what an incredible story. incred you know what he's also an incredible human being if you if you see how he takes care of the people that work for him. he's he's workaholic. he truly takes care of them. Like he is like good boss people. He's great leader. So he has all the everything you want to see in fantastic CEO. just listened to his book. think it was called how to make billion dollars. Yeah. The first one or the second one. There's two of them. How to make How to make few billion dollars. What's what's the investment case for Apploving in world where we think the apps are going to write themselves. tell us about that. You you own that in your app and some of your funds. Yeah. So Apploven is it's not really an app creating tool. It's an advertising tool. The loving it's the it's the love. It's the loving part that matters. and what's exciting about AppL, and it used to be, but this stock did get caught up in the software. It did. And in part because it's part of the IGV and this is this is part and we have to talk about software at some point, but yeah. we'll go have dinner. know. need another hour with you guys. Okay. so with App Leven, you know, they're they're basically an advertising company and what they realized is that they have lot of data. They have probably 50% share of, you know, served ads in mobile games. And what they realized is, you know what, we're pretty good at this. We're using an AI engine to serve up these ads and we're improving kind of the the return for our clients. Why don't we start putting in lipstick or you know, actual goods and monetize our ads differently. Okay. And so that whole story of we're not only going to show you games to be downloaded inside of games. So you're so you're playing Candy Crush and an ad pops up for Scrabble, like digital, like an app. Okay, so they got very good at that. If people play this game, they'll probably play that game. Okay, fine. They know how to price it. The the game app wants wants more exposure. Great business. You're saying they're now taking this business in another direction. That's right. So all of sudden they're entering the e-commerce market to sell to sell goods and it helps them monetize their inventory significantly. Okay. And so really interesting. mean, Adam, another look, one of the the keys for us is the management team. Adam is founder CEO, hungry. He's very involved with with who gets hired, how they get hired, the headcount. Like, he's keeping his headcount flat. Okay. Regardless of the the topline growth. So, it's just cash generation. Like, you seem to be on firstname basis with lot of the founders of the businesses that we're talking about. How important is access to management for you as portfolio manager? If there's company like Nebius that says, "We're not going to talk to you. don't care if you come to Davos. We don't do like are does that cool you off in terms of wanting to be an investor or is it not that important?" No, it's important because they end up being the stewards of your capital, of our clients capital, right? We're betting mean it is an important leg of of any story is like who is leading company. Yeah. You know what is their integrity right? You know it's it's really important to know to know the management teams. Obviously like you get too close to them though where you like them personally and they they screw over shareholders or they're more inepth than you thought. Like is that the the downside to that? well it's not like I'm friends with them per se. You just need to be able to ask them questions. That's right. And need to need to understand like are they are they good are are they are they good stewards of capital. mean it's important to me for me to understand that Brad is great leader. Okay. But that that is really like he is he has incredible humility. That is important for me to understand. have two more things we have to get to or the audience will kill me. One of them is software. But before we go there said two days ago think Intel is like the stock of the year so far. Yeah. The comeback in Intel is extraordinary for lot of reasons. Number one, it's not every day that former Dow component, fallen angel, formerly one of the biggest market caps in the country, drops almost to single digits, has new CEO come in after multiple failed new CEOs and actually strikes chord with growth investors again and reinvents itself. So that thing that we talked about early in the show about that reinvention, this looks like that on steroids, but don't know as much about semis as as you do obviously. So I'd love to get your take on like is this buy here or did we miss the whole thing or investors getting too excited? I'll be honest, don't know. What do know is Lipu is is the new CEO. Amazing. Lipu Lipu. Yeah, Lipu is the new CEO of Intel. he is again an incredible he's visionary leader. He is visionary and mean his story with Cadence mean someone should do deep dive on on Cadence and see how he turned Cadence around. He took it from almost going bankrupt to the the company that it is today. Okay. And that was under Lipu's direction. So if there's anyone that could take this asset and turn it around, it's Lipu. The foundry is now being seen as strategically important asset. Whereas they were being ridiculed as recently as three years ago, these idiots thought they could build Taiwan Semi in the United States. Yeah. And and what would say that's where I'm skeptical. You don't So you don't believe necessarily that that's great asset? And Taiwan Semi is doing it too in Arizona, aren't they? Taiwan Semi is building it building their fabs in Arizona. but if we're short compute and Intel is company that can help bring more compute to the market, right? But you have to it's not easy, right? So, so if you if you've been inside of fab and built chip, it's it's difficult to build chip. would which is why like you only have one, right? You only have one foundry now. You have TSM. So, you're not you're not convinced that this foundry is going to be as successful as the bulls in don't know, right? Because it is it is bet on the execution of the engineers inside the foundry. Lipu is not an engineer, you know, so hopefully he will hire the right people to make this happen. Okay. But don't know. Okay. So, you're not in the you're not currently in Intel? No. Have you looked at it recently? We have looked at it recently. just it's it's an existential question like can you be foundry or can you not be foundry? Okay. SAS apocalypse. Jensen has said the bears have the story completely wrong. Agent agents and Agentic AI is actually going to utilize all of the these CRM and and corporate software to help people. They're not going to spin up their own versions. Therefore, these companies do still have Moes and they are relevant for the age of agentic AI. How do you feel about that? don't agree respectfully. You think Jensen's being too nice? think think Jensen can't really say what what is happening because Okay, so that's what think. And so why do you think that? think what you think like no way like there's no there's no way all of these companies can remain relevant if if Anthropic is going to continue to come on at the rate that it's been coming on. So that doesn't mean think they're all cells, but you know, think it's really hard to know obviously, but right now they're all being treated like cells. Well, because the growth story, Salesforce cannot do what they've been doing for the last 15 years where you reup and the contract goes up, right? And that's huge part of the story, right? Raise prices every year and no one's going to push back. How could that be possible? But and of course, we can't know is the market overdoing it, okay? With the selling. So if you look at let's take Salesforce Salesforce as of yesterday was trading at 10 times EV to free cash flow. if you include stockbased compensation right if you right no if if you exclude stockbased compensation including stockbased compensation it's like 14 and half. That's not that compelling. It's it's kind of trading like broken tech with little option value. That said, think there's other interesting things out there right now that, you know, I'm definitive will cross this chasm. So, you know, the issue for me on the entire software space is that we only use 15 to 20% of the software of any of these packaged software. Yeah. They're selling it per head. And people might say, you know what, I'll pay you per use. I'm done with this game. That's right. And so so it it's more competitive market, right? People can write their own software and maybe you know maybe and there are lot of people that will have will will say that's not right because you have the security issues or right and it's valid but again why am paying you more if don't use the software? If one of my employees comes to me and says wrote this sick program. run wealth management firm. I'm not going to be like cool plug that into our system. Let's see what happens. don't care how much money I'm saving. will destroy potentially destroy my business. So that's the vibe coding thing. But the the there's middle ground where the IT department at company can do it themselves, has makes concerted effort to actually replicate something that they're paying third party for. And they show it to the board of directors and they say like, "Guys, we really can do this and we can save the company $80 million over the next 5 years." And the company says, "All right, that's what we're going to do." Like that's the real threat. pipeline is coming down for these companies. That's right. And so think the best case scenario is they don't get you know uplift as you were saying on that DBNR right which is or is the the basically the expansion the expansion revenue so they no longer can grow on seatby seat basis how much they're selling and that's big part of the growth story. know that if we think AI if we think that AI is threatening like employment in general then higher headcounts are going to be harder to come by period even if you can get company to sign another package doesn't mean they're going to have the same amount of employees. That's right. And so that that dream of that 40% operating margin that you know was why these SAS businesses traded at the multiple that they did. think that's different terminal value. don't know if it's 30. don't know if it's 20. My my gut is that given the landscape it should these should be 20% operating margin businesses. Is cyber security immune or not immune? think parts of cyber will be immune parts of it will not be right. So you know Apollo also or crowd strike more likely to be immune because they're platforms and they can build on that platform just like Microsoft. Yeah. But point solutions across liked like identity or Okay. point solutions across all of software are going to be at risk because they'll just be usurped by either you know new players or people who will like you said write code inside of the company and or by bigger players that basically just add that on. think the market has this right when look at constellation. So, Constellation is like collection of thousand tiny niche software businesses where it's like this is dry cleaning software. So, don't know this is 20,000 dry cleaners in America and they all standardize on this one tiny piece of software that seems super susceptible in way that Crowd Strike's Falcon platform is not susceptible. And if you look at the dispersion even within software, let's say they're all down. The ones that are the ones that seem more replaceable, those are down more. Like it doesn't look like just wholesale paddic. It looks fairly reasoned. Yeah. mean like cadence isn't down as much as team atassian. Right. Right. Atalassian has zero moat. Yeah. And that's why it's trading down to don't know where it is anymore, but that Okay. Are there any software stocks that you think it's already overdone and the market really has it wrong and are buys? you know, do think Apploven was put into that bucket. Okay. And just because it's in the index and they just because it's in the index and there were some short reports on it and you know, people are worried about advertising and agentic advertising and how does how does AppLovven get involved with that? you know, think companies like Palunteer, they're def they're like the shepherd that is basically having everyone cross the chasm with them. Do you own Palanteer? We do own Palanteer, not in the the ETF, but across the other That one didn't come down until the end, right? Like that one held up better than the rest of SAS. Well, it it kind of sold it. It actually don't know. It went from 210 to 160 before like the SAS apocalypse and then it went, you know, later. Okay. But so it has it has pulled in quite bit. is Adobe dead? They're acting like there's nothing special about Adobe. And meanwhile, see OpenAI shut down Sora and don't know who's using all these other tools, but is anyone paying for anything? don't know. They are paying for Adobe. just started to use Higsfield Higfield AI, which is basically video generator. It's pretty amazing. What are you using it for? What sort of videos you generate? I'm honestly just play with all these tools to figure out what what they can do. And with Higsfield, was trying to figure out if can make an advertisement for my app, right? And to see if like if wanted to to have Tik Tok video, how might make it, right? and it was pretty good. mean, it wasn't great, but it was pretty good. don't know if don't know if anybody told you this, but am the creator of something called Halo. Have you heard this term? was told you were on scale of 1 to 10, how clever is that? Would you say it's clever? would say No, that's good right there. No, no, no. Wait, wait. Okay. When when we did the when when when we first thought about this like AI and the declining cost to create the correlary to this was not mean Halo is like really awesome way. She's trying to say she came up with Halo. Listen to this. No, didn't. But was that any tangible asset would get multiple expansion and and and here we have it. mean basically so you you your your packaging of it is beautiful. Yeah. Gorgeous. Well, all right. So the reason I'm asking you this question though could we overdo it? Could we overdo which part? I'm sick of it. I'm sick of hearing about Michael's going to throw himself out the window for years. all right. looking at the Mag 7, if you had to rank all of them from most halo to least Halo and said think so my personal opinion Apple and Amazon are the two most Halo stocks in the Mag 7. Meaning the physicality of what they have and do. Okay, they don't own data centers. Apple, but like the my rectangle is my rectangle and it's not going anywhere and it's central to how the user is interacting with all of this AI. think it's very halo. What do you think about that? okay. If Apple doesn't act soon, that impact that that Halo designation is going away because over time, do believe that AI will allow us to extract ourselves from our screens. Okay? And into what though? Glasses. It could be glasses. It could be mean you've seen you've seen the open AI you've seen the open button. It doesn't matter. It does maybe not that but some form something will come along. It could be it could be bracelet. It could be spectacle. It could be something right. It could be something that that hangs with you that acts as acts what Apple should be doing is becoming that personal agent. can't believe they're not doing it already. don't don't understand why. Where is gentic Siri? don't know. Why does Why doesn't Siri book my plane tickets? don't know. Okay, agree with you. And so, but they are not doing it. And maybe they can they can't do think they can't do it. Well, think they can do it, but they are. So, they fired the AI guy recently. think they don't have the horses. They They didn't don't don't know if they fired him or if he left. Sure. But think the issue is that they're so concerned with security as their whole being has been we are going to keep your data. We're the safe tech. And so how do you do this and still maintain the story that you're basically the safest tech? So you know think there's there's probably bit of you think that's conundrum. You think that's why they announced they announced new and improved Siri and AI capabilities two full years ago? Have not delivered. Multiple iPhone models have come and gone and they don't have it. It's cra I've never seen anything like it. I'll tell you it's incredibly frustrating because going off there. Are they certainly hope so? certainly hope so because don't know if they're nervous. think that to some extent they've, you know, milked this device enough. personally hate carrying around an iPhone. Okay. wish didn't have to. love it. take it in the shower. No, if there were new form factor that everyone was doing, would do if it's an amulet that you were around your neck and you talk into it and people just got used to that, would roll with it. But there just isn't. But there will be. So, so our job is to figure out what what the possibilities are in 3 years. Okay. Right. Is it possible for there to be different device that is screenless? Is there is it possible for there to be different device that you interact with differently so you're not you know opening this 150 times day? Okay. Right. And think there is you know but if that goes away what happens to meta? Yeah. So so there this is discussion that we've had internally of as you go to more screenless world what happens to other business models? That's the Instagram right? If don't if don't have my rectangle and I'm not doing this, then where does Tik Tok and Instagram get revenue from? Zero. Wouldn't it be wonderful to not have Tik Tok and Instagram? You know how bad it is for our brains? It's like really good for my ego, though. At least Instagram. All right, get it. mean, but they said, listen, remember Google Glass was already like 12 years ago. It was bust. The Snapchat spectacles didn't go anywhere. Apple's own stupid helmet that only Michael bought. No one else on earth. He sold one. he returned it to his credit. Like, I'm waiting, but until think the rectangle is the rectangle, and don't think that anything can get in between us and our rectangles. Yeah, think for now, you're right. Yeah. But the longer they don't have solution, the more at risk their model becomes. Their their capex went down year-over-year. They seem super comfortable on their la. Well, don't know if it's rest. actually think this part is brilliant. They're like, "You guys spend the capex. We're just going to be the layer on top. Take the 30% in the app store. We're just going to we're just going to sit on top and use your compute." It's like you guys go spend all the capex. think it's it's either Apple or Amazon are the best of the Mag 7 year to date. for different reasons. Well, is Apple still the best business in the world? think the biggest size and the highest gross margins probably. No, Apple's gross margins are not the highest gross margins. Nvidia's gross margins are higher. Nvidia. Yeah, Nvidia would be the best. Even Google Google's gross margins are higher. Is that right? Okay. Yeah. Apple's Yeah. 40% 30 38 40% is Apple's. That's it. Yeah. What piece of And Cory, did you have fun on the show today? It was great. All right. So, we're going to take brief intermission and then we'll No, I'm kidding. have to say this has been so illuminating for me. I've learned so much just talking to you today and think you are amazing on the show and thank you so much for doing this. Thanks for having me. You guys are fun. Would you come back? would love to. What are you doing next week? Not next week. This guy this guy just already shaking his head now. All right. want to tell people how they can learn more about Alger and your funds. So, is it alure.com? Is it that simple? Yeah. Okay. and do you do you not doing social media really, are you? little bit. No. No. Good for you. so respect that the most think. Maybe I'll have maybe I'll write bot to do it for me. we want to have you back to talk about robotics and also know look at your face. Look how excited you are. All right. Well, so we're going to we have lot to talk about. All right. So, we'll do maybe we'll do that later this year. just want to say thank you so much for joining us and guys, thank you for listening. Compound appreciates you. If you have not yet hit the like button on this episode on whatever platform you are watching or listening to us, now would be pretty good time to do that. What do you think, John? Right now. 100%. 100% he says. All right, guys. Thanks again. We'll see you soon. All right. Amazing. Did you like how that came out? Yeah. All right.