Episode 20: John Rotonti on the AI Buildout and "Multiple Worlds"

Hey guys and welcome to episode 20 of Shun the Bull Pod with Drowsey and Bear. As always, you can reach us on Twitter @investing_bear or @DrowseyInvestor. You can hit us up by email @Shun the Bull Pod 2025 at gmail.com. And without further ado, we want to welcome back our first three-peat guest, John Rautanti, is a portfolio manager at Bash and Fuduciary. He is one of our absolute favorite guests to have on and just an unbelievable investing resource. So John, welcome back for number three. He also much for having me. Honestly, when you texted me a few weeks ago to invite me, I was just kind of thinking like, "Man, I'm miss having a good conversation with Drowsey and Bear." And then the text popped up. So I always look forward to these conversations. Well, for now. To start the conversation, we, John, we had you on on episode four, which was in August of last year. And in episode four, you absolutely blew our minds by introducing us to AI infrastructure companies like Eaton and Vertiv and other of the companies that are actually building infrastructure to allow AI to develop. In episode nine, which took place in November, again, too long ago, but you talked about where we were in the process and telling us about the power usage. And really, ahead of the curve on where we were focused and really where the investing community now is the common wisdom of the bottlenecks you had talked about several months ago, such as energy. We've talked about memory. We've done a lot. So in this episode, we wanted to kind of hit on two facets. The first, and this will be the opening question is kind of where we are, but I want to give you some nuance there. And the second is the valuation of where we are. So to start, without giving us, we're in the first inning because I think that is literally the most abused phrase in the financial community right now. And because we're all baseball guys here, I want you to give me as descriptive of a, as descriptive of a baseball game as you can to where we are in this build out, but you can't say first inning. So give me the bullpen, give me the roster, give me everything. Okay. Descriptive where we are. So I think, I think it's important to just first say, this is the most uncertain environment I can remember since the global financial crisis in 2008, 2009. Wow. Easily, and I've been doing this for 25 years, literally spending every day just following markets and economies and stocks and businesses. And if you just think about AI is creating massive uncertainty, but then also the conflict in Iran, massive government deficits and debt around the world, a new, a new Fed chair coming in that wants to dramatically change things to some degree, you know, some things he wants pretty dramatic change. And so there's a lot of uncertainty out there. And so this is just my, you know, sort of base case, best guest at this time. But I reserve the right to change my mind often and frequently. I think we are, I think we're really early innings. I'm not even, you know, sort of base case. I'm not confident the game has even really started. At least one version of the future, you know, we're really early. Hyper scalers came out with mega cat X numbers yesterday. You know, they're going to spend over $700 billion just this year. They're very likely going to spend over $800 billion just next year. As we've talked about on the show before, you know, Jensen is sort of the architect and general contractor of the AI global infrastructure build out. He is setting the cadence and setting the pace of the race. And you know, he's still saying three or four trillion through 2030. And then on top of that, another one to 1.5 trillion of electrical grid upgrades and build out through 2030. And so that's where I'm at today because of all the uncertainty, what I try to do, but drowsy and bare, is I try to latch on to the things that I think are most certain based on what I know and believe today. Obviously, this could all change. One of those things that is certain is the AI technological revolution is directly and closely tied to and dependent on physical infrastructure. We talked about that last time, you know, tech companies these days, they're trying to get their hands on GPUs or other accelerators, gigawatts of power and then all of the electrical power and mechanical components that they need to outfit these data centers and, you know, the grid, either grid or behind the meter. Number two, you know, the US under invested in its electrical grid for too long. The US grid is like 40 years old on average and it's already stressed to the limits with increasing blackouts and it's in dire need of upgrading and hardening to make it more resilient to extreme weather events, to support decarbonization, the re-industrialization in America and the electrification of everything, including AI. And so if an industrial renaissance and AI supremacy are the end goals of our country, an electricity across, you know, the whole chain, generation, transmission, distribution is the primary limiting factor, I think, in the bottleneck that must be dealt with and I think that will require a decades plural sustained investment into the grid. You know, so that's number two. Number three is AI is an age-defining technology. You know, it has the potential to fundamentally change the way the world works and lives. On the same scale of or even more transformatively so than the internet, mobile smartphones and the cloud. Number four, and importantly, there's a massive effort by the White House to support advanced domestic manufacturing and re-industrialization. And importantly, this is one of the only policy issues that I can think of that is bipartisan support. You know, under the Biden administration, they pass the inflation reduction act. They pass the CHIPS Act. And so re-industrialization, ensuring of critical national security, important manufacturing, it's bipartisan. And so I think that's important because there's almost nothing that is bipartisan these days. And then just really quickly, five and six and a five is, you know, I think we're in a cold tech war with China. And I think both the US and China see this as existential. And then, you know, six is these hyper-scalers. They're searching for the God particle. Some of the founders have said that. I don't think they're going to let up in that race. And, you know, one last thing I'll say about this and then throw it back to you. So in the in the barren's tech roundtable a few months ago, this was the biggest takeaway from me from that interview with a panel of tech analysts and investors. So Brent Thill probably all know him. He's an analyst at Jeffries. He said that AI will be globally adopted by every organization board, the board of directors. And then he said, I want to read this. Actually, he said, we didn't see boards moved to the cloud. We didn't see boards think about Bitcoin, the move to Web 3.0 or many of these other tectonic shifts. It's AI multiple board members I've spoken to will have said that they're all in and they're reorienting their companies across every industry. He says all the fieldwork we do points to absolute global adoption starting at the board level, not at the IT level where we usually saw past tech trends start. So I thought Brent Thill, I thought that was a really important point that I was not aware of. I thought that was a really good point. We eat a lot of questions on this. Next era energies, the ones that are really driving much of the holistic change, those companies and the industry looks incredibly hot right now and incredibly fairly valued and it has even doubled and tripled over the last couple of years. So if you're an investor and I buy what you're selling completely, I think the way you're explaining everything, matches. Yeah, with the caveat of well, but there's a lot of amazing resources out there saying exactly what you're saying. I could be wrong. I could be wrong. Don't buy it to sell. Combine what you're selling and the narrative you're telling. But I'm also looking at these prices and if I'm thinking towards I want to play this trend, Well, this trend is out of the bag. The cat is out of the bag. So How would you think about these valuations as you're right on the trend, but maybe everyone else is now watching it as well. You're doing Japanese futures like there. You're trading oil. - Turn off my alert really quickly, y'all. Give me one second. Okay. So, no, but I'm getting text messages. And Apple allows you to see your text messages on your computer. So latitude investment management in February of 2026, they put out a note on valuation, not specific to power and electrical infrastructure builders, but just to the S&P 500 in general. And they did some work and they said, when you pay a PE of 25 times, and if you assume an 8% discount rate, 70% of the value resides in the cash flows beyond your 10. Okay. And most analysts and investors are modeling anywhere from three to 10 years out. And so that's the projection period. It's 10 years pretty much as the max for most models. And so if you're paying a PE of 25, 70% of the value resides in cash flows beyond your model period, beyond your projection period. And so when you pay high PEs, you were making a bet on the duration of growth more than the rate of current growth. Because rage growth, is that based on some estimated growth rate, 'cause I'm just trying to follow the math? Yeah. So they basically, they just reverse DCF, the market with I think. Oh, like GDP growth or just the average growth for the maybe from the average earnings per share growth of the S&P 500 over time, which is like six to eight percent over time. Over a long period. Now these days, by the way, it's much higher. So analysts are expecting the S&P 500 to grow, I think earnings per share, earnings per share at the S&P 500 to go 17% in 2026. And I think now 15% in 2027. Growth rates are elevated right now relative to history. And that's one of the reasons you're seeing higher PE ratios, even at the index level, which is 22. And then that's-- For anybody trying to follow along, if those growth rates stay at 15% or higher, for even a couple of years, that kind of, it would actually not be as backloaded into that-- You post-forward some in the value. That's exactly right. That's interesting. It's like we are rated for the growth that we're seeing, and it's just a matter of, this is very familiar to me from back in 2020, when SES companies had the same problem as it sounds like the whole market is having now, where it's like how long will we grow at these heightened rates? That's right. That's right. So when you're paying high multiples, you're really making a bet on the duration of growth. And I'm trying-- So the portfolio that I manage, the average PE of the portfolio is probably 30, around 30, and that's high. I think it's justified based on three or four things. If you look at the justified PE ratio formula, and Mobusin is written about this, PEs are-- They're just condensed version of a DCF, and they're driven by three primary factors, return on incremental invested capital. So returns on invested capital going forward. That's number one. Number two is no-pat growth, they're operating income growth. And then number three is the cost of-- is risk, which flows through that formula in the terms of the cost of capital. And companies that have lower cost of capital that are less risky typically have less cyclical earnings growth, so more stability in their margins, more stability in their returns, and that are thought to have long duration earnings per share growth. So that all flows through the cost of capital. So I-- And the companies that I own are high return on invested capital. I think they have modes and manage the teams that will keep returns on invested capital high going forward. So I think they have high incremental returns on invested capital. Very strong balance sheets, really well run. And most importantly, given their high PEs, I think they have a very long duration runway of growth. And so that's where I focus a lot of my time. It's really trying to get comfortable with the duration of growth, especially when the multiples are high. And so if you look at-- if you look at the grid build out, you know, that's a 20 or 30 year project. I think a trillion or 1.5 trillion will be spent on the grid in the US in the next five years. But to update the grid to where it needs to be, we're talking trillions of dollars. To put some more big numbers on it, Jensen, Wang, and Annie LaNmoss, have talked about a $50 trillion market opportunity in physical AI, not just robotics, but just anything physical, industrial, factory-related. And so it's $50 trillion. Well, that's not going to happen overnight. You know, we're talking about really, really, really long-term trends. And so, yeah, the market is highly valued. My portfolio on a PE basis is higher valued than the market. But I believe that the quality of the businesses I own and the duration of growth is better than the market. And so I think it's justified. However, the PE ratio of my portfolio is definitely-- every portfolio has risk. If you showed me a portfolio where the average PE was seven, well, the risk there is that you own a bunch of shit codes that could go out of business. Right. So one of the risks in my portfolio-- there's more than one. But one is definitely the elevated multiple. Do you think about it on a company by company-- I know you think about the portfolio as a whole. But can you give an example of a company that you would either consider lowering your percentage on or maybe even just selling out of because you look at the expected growth rate over whatever term you feel comfortable with? And you say, I don't see any chance it's high enough to justify this price. I mean, yeah, I trade around positions. I wouldn't say actively. But if a position is up a lot, and by a lot, I mean way more than the growth of intrinsic value and way more than I will trim it back. But it doesn't mean I want to get rid of it. So we own Vertif, ticker VRT in the portfolio. At one point, it was up-- Yeah. Close to 300%. All about that for a long time. In the last 12 months. I can give you-- Congrats, by the way. And that's pretty awesome. Thank you. Can't complain too much. Yeah, thank you. Thank you. And not just furtive, but some of the other ones you mentioned on our last episode that you were with us. We-- Yeah, we have a lot that are-- Having a good year. Thank you. Yeah, we have a lot-- not a lot. But we have several that have performed really well recently. But Vertif was up 300% in 12 months. And so I took some off the table. That was strictly risk mitigation. But can you sort of like walk us through-- I mean, you don't have to give exact specifics of what you trimmed or ever. But what was the thought process? So I just noticed that they grew almost 30% year over year last year, revenue. This year, it looks like it's actually expected to accelerate. I'm seeing 35% for analyst expectations. You may expect even more. But how do you handicap that or map that out for the next several years? They're actually-- yeah, so they're actually expecting-- they're expecting this year to grow earnings per share at like 50 over 50%. And it's a-- That was just really rare, haven't it? Yeah, yeah, yeah. No, I know. But there's a lot of operating leverage. So the company generates 30% to 35% incremental operating margins. It's a manufacturer. It does its own manufacturing. It owns its own plants. And so as it drives more volume through its fixed asset base, it spreads those fixed costs over more units, they get a lot of operating leverage that way. Number two, importantly, with Vertif. And several of my other companies have similar operating systems. But Vertif has the Vertif operating system or VOS. And this is a system of how they operate that's become part of the culture. And it really permeates every level of the business from how they run their factories lean and Six Sigma to how they manage their working capital, to everything that the safety of their employees, everything that they do is based on this operating system. Well, this operating system was implemented by Dave Cody, who's an industrial titan and the former-- the former CEO of Honeywell and he's the executive chairman of Vertiv. He's very involved in this company. Not only did he, like I said, implement this Vertiv operating system and sort of nurture it and he's constantly monitoring and tweaking it. But, you know, he's on every earnings call. He's a very active executive chairman. And so it's the operating, it's the fixed nature of the cost structure and the operating leverage. It's the Vertiv operating system on top of that. It's pricing power. So let me just tell you all just really quickly about pricing power. So, would McKenzie put out a report this week, this Tuesday, April 28, 2026, would McKenzie is a, you know, energy basically research firm and consultant. And they're highly regarded. And I use a lot of their research. So the report says that, you know, the US has expected electrical equipment specifically for data centers, specifically for data centers. So not for the grid in the US is expected to more than triple from 20 billion and 2025 to over 65 billion in 2030. And, you know, the report says things like the scale of electrical and power equipment that is needed is staggering. The report says, you know, there's a quote unquote structural transformation that will define the next decade of the electrical equipment market. Right. And so because of this, there's this massive supply demand imbalance. And there's a two to three year wait time lead time for a lot of the key electrical equipment. And because of that supply demand imbalance, Invertive, you know, fits squarely into this. Okay. So because of that supply demand imbalance, they have incredible pricing power. So the report says that and this, this got me to like double take electrical equipment manufacturers are going back to hyper scalers with year old purchase orders and increasing prices on those year old purchase orders by 20%. Now think about the cajones on that company. I love it by 20% and not just to anyone to hyper scalers to the most powerful. Yeah. So it's not just vertive. This is this is kind of across the industry, right? This is the electrical equipment manufacturers and they operate a little oligopoly. So there's not a lot of them. Depends on, you know, the parts that you're selling and where you're selling them to along the supply chain, but companies like vertive and eaten and GE vernova, which we own, all three. And then, you know, outside of that, you look at companies like outside the US like Siemens, Schneider, Electric, Hubble. There's a smaller market cap company called Powell. So there's there's not there's not a lot. The the the technical complexity. So if you just think about vertive, so vertive is the market share leader of both thermal, so cooling and power equipment that goes inside the data center. So market share leader on both. And typically, last three years, let's say, when we really started to build out these AI factories, the builders and the operators of the of the data centers were buying, you know, discrete parts from different vendors based on price and based on when they could get the parts. So like availability and and price. Well, now it's become the technology has become so complex and the the way that that thermal cooling and power interrelate. So like thermal and power have to work together more and more in the future. And so what's happening is companies like vertive are building customized all in one converged systems that do thermal and power. And eaten is doing the same thing. And G. E. Vernova is doing the same thing for the grid, but not thermal and power, they're doing power and electrical equipment. And so they're building these converged systems that are interdependent. And because they're interdependent, they're much more efficient, which means less energy usage, which means that the data centers themselves data center operators can grow revenue can can manufacture tokens and grow revenue at lower cost because they're using less energy. On how we view like the political movement or the social movement against AI kind of AI. If you look at any of the polling AI is incredibly unpopular, whether it's like not in my backyard, growing data centers or the recent attacks on Sam Altman, which definitely have been quite scary. I wanted to kind of post you on how you view the potential for either legislation to kind of slow down this train or a societal sort of backlash against some of this. And if that weighs in at all because right now it just seems like it's full steam ahead for decades. So I wanted to see if you incorporate that in your thinking at all. Yeah, so I actually, so legislative pushback is one of several bottlenecks. Right, we talked about the grid bottleneck. So that same report y'all that I just mentioned from Wood McKenzie. This this this this also almost haven't been followed by chair. So that report says that 183 gigawatts. I know 183 gigawatts of plan data centers in the US have signed contracts with utilities for power supply. Doesn't mean they've started building, right? They've just signed contracts that they will have a hook up. But there are still 600 gigawatts of announced or planned projects for data centers that are still waiting in line to sign a contract for a power hook up with utility. So 600 gigawatts 600 gigawatts it cost 50 billion to build a gig. That's $30 trillion. Now I'm not saying all 600 of those will be built. But if they are, that's $30 trillion. And that 600 gigawatts waiting in line by the way, these lines are four to five years just to get a contract signed saying you will have a power hookup. Four to five years just for that. So so there's a massive backlog in in power part. There's a massive back. Wait list. There's a massive shortage of workers. And I know y'all have heard of this. It's it's it's it's electricians. It's high voltage electricians. It's line workers that get up on those high high poles to build long distance transmission lines. It's welders, it's plumbers because of all the liquid cooling and all the flushing of those lines that you need to do. There's a massive shortage work like not like a little shortage like 500,000 shortage of workers in the US. So much so that that meta just announced they're giving a free four week course to learn how to be a fiber optic electrician and technician free. So so yeah, you've got legislative and you've got neighborhood pushback. Not in my backyard. You know, you've got shortage of power. You've got shortage of labor. You've got a lot of different bottlenecks all along the line. And I actually think that's a good thing because if it weren't, we would be building even faster and more violently than we are today. And the bubble wherever it is would be would be, you know, inflating faster and more violently than it is today. But I think that these, you know, these bottlenecks, whatever you want to call them, are just going to extend the duration of the build out. And if you know anything about me, my investing philosophy is based on duration of growth, not rate of growth. And one of the things I'll just say about all of the legislative pushback, you know, I think we all heard like 50 per half of planned data centers got canceled this year or something like that, right? And so that's like a big headline that's scary. Correct. But what did what did alphabet do last night? Right? Like they said that they're going to, I don't know if the word was significantly or substantially increased their catbacks in 2027. Meta increased their catbacks for this year. You know, Amazon's going to want to be the biggest catback spender next year. So catbacks is going up. So if we take a step back from like the headline, you know, breaking news, half of data centers are canceled. I like how I like how you think together how you explain to you. So the actual quote, and again, this is time stamping the episode. We're taping this on Thursday morning right after hyperscaler earning. Paloosa last night, Wednesday after the market. Google and just to add in what Google said yesterday. So Google said they are significantly increasing their catbacks compared to 2026. Exactly what Dr. J. Roge has said. But I like how you're explaining even if there is pushback, which I think I mean, I've said on the show a number of times I'm long AI investing in short humanity. I'm kind of an AI skeptic as it grows and grows. But as you're saying, it's almost these pushbacks are creating efficiencies really. Rather than these inefficiencies and this kind of chaotic attempt to grab any of the resources you need all at once, the kind of slowing down and waiting is actually smoothing out what is going to be a multi-decade process regardless? - Well, yeah, it allows us to catch up with the supply chain because I said that, most electrical equipment orders have a wait list of like one to three years. Well, if it's transformers, it's four to five years, right? If it's a combine cycle gas turbine, it's four to five years. And so yeah, it allows you to catch up the supply chain. It extends the duration of growth and it just, I think, I think, I hope at least, it prevents the bubble from getting too big, too fast. And so, and by the way, it's also why they're talking about orbital data centers because they don't wanna deal with the legislation and then not in my backyard. And so that's enough. That's one reason I think why orbital data centers and space data centers are gaining so much popularity right now. I just, I wanna quickly, quickly, quickly, just go back to what Bear said about the valuation 'cause I don't wanna doubt it's a question. So what I do is I look at the total shareholder return formula which says that mathematically, precisely to the decimal, you know, a stocks return is equal to the dividend yield plus the earnings per share growth plus more minus the change in the PE multiple, right? And so if we just assume really quickly no dividend yield, right? If a PE is 30 and over the next 10 years, the PE contracts from 30 down to 25, well that's a 2% annualized headwind to returns over 10 years. So if, and my sort of minimum hurdle to put something in the portfolio is something that I think can generate 12% annualized returns. That's sort of the minimum hurdle rate, right? And so if I want something that can generate 12% annualized returns and I think the PE is gonna attract contract five points over 10 years, that's a 2% annual headwind. That means I need to find a company that I think can grow earnings per share 14% annualized. 14 minus the 2% headwind gets me to 12, right? Well, the current PE on the market is 22. What if that PE that I'm investing in goes from 30 down to 22 to just a market PE? Well, that's a 3% annualized headwind. What if the PE goes from 30 to 20 over 10 years? That's a 4% annualized headwind. So if there's a 4% annualized headwind, I need to find something that I think can grow earnings per share by at least 16%. And so I'm constantly every day, 'cause bear actually what it looks like. I'm constantly every day, every day, looking at the stock prices in my portfolio, what I think they're gonna be worth in five years. And once again, it's a range of values. This is not precision. And just what I think the earnings per share growth is going to be and constantly balancing. What I think the earnings per share growth is gonna be versus where the PE is today and how much I think it's gonna contract. And I make little trades around there. I don't really bust out of anything that I love, but I do make trades. And doing so allows me to also manage my cash position, which is large right now. I like to have cash. And so that's what I'm doing. And I'm constantly thinking in the back of my head, no, 15% annualized is a five year double. 20% annualized is a four year double. 7% annualized is a 10 year double. And so I'm constantly throwing these numbers throughout my head throughout the day. Does that answer your question about valuation? - Absolutely, yeah, that's exactly the way I think about it. And without all that precision. And of course, the hardest part of that is what you think the growth is gonna be, not just for this year or next year, but the durability of that growth is something we've, well, something I've always struggled with going back to the the Heddy Sass days of 70% growth. It's like, where do you go with that? How do you even begin to? - It's hard. - Where that goes. - It's really hard. You know, another thing I've been thinking about recently, y'all is like, 'cause we started with where we are, what inning and you want it specifics like. - Right. - What? - Pitch count. - Well, that's just one version, like one version of the future. And it's kind of like my base case right now. But so another question I've been asking is, can we say the demand for intelligence is cyclical at all? Right? Is there some natural endpoint where we say we don't want any more intelligence? And I think the hyperscalos are answering that question. They're saying no, there's no cycle to this. There's no natural endpoint to intelligence. What does MetaWon2 do? MetaWon2's to put, you know, super intelligence, AGI and everyone's pocket, right? And into a billion people's pockets. And, you know, and so I think that's the mentality. - Well, first off, I'm sure it's humanity. - And I think they're gonna continue to use that mentality until something like-- - You are, I'm not sure I agree that it's best to have an EGI pocket. And hyperscalos have all the reason the world return on investment, lots of-- - I didn't say I agree with that. - To make it work. So I'm with you and I'm not gonna pause on that. But I wanted to ask if we're breaking from the cyclical cycle and there isn't an end in this base world. And again, with the caveat, we could be completely wrong. But if this is just gonna go on and on and Meta's seeking AGI and I don't know, Amazon seeking a similar end game and they're all kind of combative, does that breaking of the cyclical nature filter down to all these companies that make up the AI infrastructure? For instance, Baron, I've talked a bunch on micron. Boring micron from 10 years ago. Cyclical, we didn't give it a time of day. It wasn't our thing. It comes and goes, who cares about memory. Now it's completely broken out. Now is the cyclicality of micron and other base parts of this AI build out gone. And this is just gonna be up into the right along with the hyper scalers and all these massive companies. - Man, it's such a good question. I don't have a good answer, but I will say that you know, AI is lifting a lot of boats, right? And if you, if, and this is not, this is not, I don't think this has happened to this extent with past technological revolutions. All of the boats that are being brought along, if you look at what Nvidia is doing and how it's building its ecosystem, it's got hundreds of suppliers and it's trying to bring as many of them along for the ride as it can and it's trying to give them financing. It's trying to keep them alive. It's doing a lot of things. And it's, like I said, it's lifting a lot of boats. And so what you're seeing is that a lot of things that we're old are now new again, right? And so like everyone's talking about memory, which was like you said, it was seen as hyper cyclical, violently cyclical, very short product life cycles. Erratic margins, just really hard to predict. Well now memory is really important. CPUs are all the rage right now, right? I mean, Amazon put out an article recently explaining that CPUs are reclaiming relevance. And what else is an old lagging analog chips because you can't run a data center without things that manage power and voltage. And so like everything that was old, memory, CPUs, analog chips is new again and is getting like really like a fresh life, a new life. And so I don't know if all cycles have been removed across the supply chain. I don't know, but I do think that a lot of things are getting a resurgence. And I don't know how long that will, that resurgence will last. Yeah, I mean, it's really a hard to answer question of just what will change and what won't change as we go throughout this super cycle. Or it's not even really a cycle as we've just talked about, how do you think about memory? And I mean, one of the things with memory, and I think this is going to be the case with other bottlenecks that we have to blow through is the pricing. The pricing, for anything where there's not real, differentiation or you know, you call it a commodity. The pricing power really isn't there long term. And this brings up another, I know I'm throwing a lot of word salad at the wall here, but like the other thing this brings up is like, what about the breakthroughs that might come? We all remember the deep seat moment of when we're everyone worried, oh man, maybe we don't need as much compute as we thought. Ha ha, right? That has not played out at all. We still are grasping for as much compute as we can possibly get. But I do see breakthroughs every now and then, like the Google announcement of, you know, this is obviously in development or whatever, but they've come up with a way to make the need for RAM or memory much less. They're more efficient or whatever. So do these efficiencies and potential efficiencies play into your thinking? And is there any way to predict that at all? And does that, and I guess sub question, you know, what are the different bottlenecks that that question sort of addresses? So we talked a little bit about memory. Is there other things where, It might be possible to see what's coming or when it might show up. Yeah, and I agree with everything that you just said. And because it is so uncertain and unpredictable, how ideal with it is, I don't touch it. I mean, I don't. So the way that I build this portfolio and the way that I invest in general is, I never wake up and say, "How can I get AI exposure today?" Or. That's my cloud code from the. What's the hottest AI stock? Or what's the best way I can make the most money on AI? That's just never a question that goes through. All right. You know, rather, my questions are, what are the highest quality, most superbly managed, industrial infrastructure and technology companies I can find that have a long history of proving they are adaptable and resilient and that have adaptability and agility built in to their sort of culture and their corporate history, right? And that have proven that they use that adaptability to sort of position themselves, you know, to be agile and to position themselves to benefit from technological revolutions, no matter what the technological revolution is. And so that's what I'm looking for. I'm looking for companies that, like I said, are super high quality, have long duration growth or superbly managed, but most importantly have a history of adaptability and adapting in a way such that they position themselves either organically through R&D or through acquisitions, but they adapt in such a way that they position themselves to thrive no matter what the technological changes. And so I think I have companies in my portfolio that have proven that. Now, once again, I could, I could be proven wrong, but that's how I'm investing in this trend. And so, you know, I never chase something like memory or optics or other things that made that I don't understand as well, that I can't predict, like you said, Bear, that just highly uncertain, highly unpredictable, I just avoid them. Now, I get some like, like, you know, we own an Nvidia in the portfolio, right? And VIDE is doing, right, they're doing everything, I mean, everything, right? And so like, you know, at GTC, they said they're putting out a rack of just GPUs, just CPUs. And so like, do I need to chase Intel? Look, whoever was in Intel early, like literally props, incredible, yeah, $30 billion, right? Like, we own it. Incredible, incredible foresight, incredible investment. Following lip-boot tan into the, into the company is a great move. I've known, I've not known, but I've followed lip-boot for over a decade because I've owned cadence design systems that long. Or sorry, I've followed it that long, I've not owned it that long. By the way, that's another company we own in the portfolio is cadence design. And so I don't feel like I don't ever feel like I need to chase something because I feel like I own companies that position themselves to benefit through their agility. And by the way, that is how I think you build terminal value. Like, you know, if you, if you have a proven history of positioning yourself to benefit no matter what the technological breakthrough is, and you're just going to be there no matter what because your products and your services are so mission critical, I think that's how you really build terminal value and, and, and, and back to the valuation companies that have high terminal values traded high multiples. Because like I said, 70% of the value for the typical, you say perfectly, um, and I've been getting those out years, almost endless amounts of gawking and flat for our last episode where I revealed quote unquote, but I said that I have a plus 40% position in video now, which we'll do our quick portfolio reviews in a little bit, but it's up to 44% now. And insert meme of job drop and all this, but the way I view it and you eloquently just said it is, Nvidia has a hand in all of this. Nvidia not only is kind of driving and he's and gentsons at the helm of the AI revolution and some of the buzz phrases, I'm with you on that, but they're investing in products that are throughout the ecosystem. They're investing in companies, whether it's the neoclouds or other startups that are across the ecosystem. So as the entire AI trend wins, Nvidia is going to win whether the GPU count lessons a little bit or whether trainium in the next few years takes parts of kind of the GPU build out. Nvidia is going to win along the way and they've also through the circular investing are part of all these other companies as well. So rather than picking, this is how I viewed it or at least justified in my head a bit, rather than picking the next, I'm going to talk a stare labs here, credo and some of these other smaller names that are niche parts of the AI infrastructure build out that are going bananas as and rightly so, why not buy the biggest company that has little bits of all of these companies built in and the success of all those companies are built on Nvidia success. So that's sort of how I've justified having such a long position or such not only long but large position and I know, okay, insert hate mail on 40% and questioning my sanity. I'm with you guys. I'm with you, but that's how I'm thinking of it a bit. I agree completely and you mentioned Nvidia, but there are other examples of that. Like if you want, if you want VC exposure, right? And this is another way that I build the portfolio bear is what Trowsey just said. If you want VC exposure as a public market investor, you know, you can get it through companies like Nvidia, which we own in the portfolio, through companies like Alphabet. I mean, look, Alphabet owns a hundred billion dollar stake in SpaceX and massive stakes in Anthropic. It has a stake in Anderreal and has a stake in AST space mobile, right? I mean, I'm just, I'm just, those are just the names coming to mind right now, right? Obviously, Amazon has a multi billion dollar stake in OpenAI now, massive massive stake in Anthropic. It owns a stake in Energy, which is a business working on manufacturing, small modular nuclear reactors that just IPO, I think this week, Amazon owns a stake in Ioncube for quantum computing, right? So, so both are investing heavily in quantum autonomous driving. Both of them are investing in that alphabet as clear leader there with Waymel. Both are investing in drones. Both are investing in space robotics. Amazon has a clear leader there. It's got over a million robots in operation. By the way, if you've ever a million robots in operation, you're going to be able to train more robots on AI than anyone else. Amazon is going to be a clear robot winner, I think. And then, you know, not necessarily VC venture capital, but cybersecurity is very topical right now as well, right? Well, Apple, alphabet and not Apple, but Alphabet in Amazon, clear cyber platforms, cyber leaders on their cloud platforms. I mean, Alphabet owns Mandient and Wiz now. And so I agree completely, Drowsy, that you can get broad exposure and even, and, um, through these criticism that we've heard, you know, once in a lifetime, not a email yet, but just, I mean, they're huge companies. Google is 2 trillion. As we're taping, it's 15 bucks today after earnings, Amazon's down a bit. We need our fact checker, fact check. Google's more than two trillion, right? So completely underplayed it there. I think it's over. It's over. It's over. It's over. I've got Google it's almost 4.5. Yeah. And it's actually one, one thing that folks have said is just how much more room is there? But if you're taking, you have a 10% stake in the hottest IPOs and not just IPOs, but the hottest companies and the hottest industries. And that's not even counting your core, waymo, moonshots and not even your core search, which is increasing. I mean, this is just an incredible way to play so many different things. And you know, and Baron, I've talked about this a bunch too. A $1 trillion company today was a $200 million or $200 billion market cap company a few years ago. And we really don't know the upper limits of what a company can get. The SMP indexer like myself is just eating this up, the segment on Google, Amazon and video. I still think maybe a little diversification is in order for you, Drowsy. But I agree with everything we're saying. And, you know, Google and Amazon have always been sort of my favorites. I've diversified into the SMP 500 long sense, but I had positions of both of those. Man, I wish I just held those as well. Yeah. Yeah. So we own alphabet and Amazon and Nvidia and the portfolio. So regarding on how big they can get, you know, I think, I think much bigger. And you know, and the reason is because I think, I think GDP is going to get much bigger. And I think the SMP 500 is going to get much bigger. And so they're percent, they can double hypothetically, you know, to yes, you know, eight to $10 trillion companies. But maintain their sort of percentage of the market cap of the S&P or their percentage of GDP because I think all of this is going to grow. I think token usage is exploding right now. That's not a thing. It's exploding right now. Blackstone President John Gray told TBPN a couple days ago that Blackstone owns 270 companies. And their LLM usage, their LLM spend. I'm sorry, it was up 15x and Q1 of this year. These testimonies are crazy. And so what's happening? Right. And so anthropic is throttling you. Open AI is giving up on Sora. You're seeing tech layoff so they can free up room for token budgets. And you're seeing in the CapEx numbers, these companies want to manufacture more tokens, which means building more AI factories. The economist reported that tokens on open route are quadrupled between January and March of this year. And so what's going to happen is I think these tools, oh, and, oh, yeah, slot, Torsten Slock at Apollo said, showed that new business formation, new businesses are forming specifically because of AI. He shadowed the graph that shows it, new business formation specifically because of AI. And so what I think is going to happen, y'all, is I think, yeah, I think some jobs are going to be lost. I do. I mean, they are in tech that's already happening. But I think you're going to have an explosion of entrepreneurs, you know, the single startup that can achieve a billion dollars in revenue. I think we've already had one. It was in the news two, three weeks ago, you know, a single person generated a billion dollars in revenue just using AI tools. So I think you're going to see an explosion of entrepreneurship. And, you know, entrepreneurship is one of the drivers of GDP. I mean, it's really, you know, productivity and number of laborers, but entrepreneurship flows through that. And so, you know, if, if entrepreneurship explodes to the upside, I think GDP will grow. And so I think we can support much larger evaluations for some of the mega-cap tech companies. These are Lynchpin, let me be, you know, I know where I'm talking about, like, what I'm trying to do with the portfolio is I'm trying to buy Lynchpin companies. I'm trying to buy companies that are a Lynchpin, not only to an ecosystem and supply chain that they build and rely on. And I'm talking, like, you know, I'm talking about hundreds or thousands of components that they're getting from different companies, thousands different companies. Yeah. So what I really like and what's interesting to me about your approach is I would think the companies that you focus on, some of them maybe even more so than Nvidia are, well, they're certainly immune to a certain type of the sort of technological innovation that I was talking about that, you know, that certainly the AI models can be, you know, thrown off by or the memory chips can be thrown off by if there's efficiencies, if there's innovations that come along. And Nvidia is even potentially prone to that since they get so much of their business as GPUs, right? So, TPUs become more of a thing or if there's some other kind of innovation that's possible. The other stuff in your portfolio, I totally get, it's really hard to see how, you know, like the ground movers and the equipment that does all that stuff and builds the data centers. It's really hard to see how that doesn't persist for as long as we're building data centers, which as you said, maybe forever. So I totally get that sort of way of way of looking at it. But just because of, you know, my co-hosts, I think we have to go, I want to go back to Nvidia for one moment and just say, you know, well, let's sort of play Devils ad. Let's play Devils ad advocate for the markets side. Do you John agree with me that there, it does sort of make sense that the multiple has started to come in a bit on Nvidia. And you know, do you think the market's still super underestimating the out years, the future years, what they can do or do you think that that's, you know, they're kind of getting as much revenue or, you know, maybe it grows by 15 or 20% from now. But from the world as they can logically get in a year or that will be, you know, coming up on some kind of flattening growth rate soon. I mean, Nvidia historically was pretty cyclical historically. And so yeah, I think the market is worried about the sustainability of the revenue growth. But also the margins as they get more competition from TPUs and from training them and maybe from AMD. We can at least say the margins probably not going to expand much from here, right? Exactly. Exactly. Yeah. And so yeah, I think I think the market is worried to some degree on, like I said, the sustainability of growth and the margins. Growth in the near term is pretty secure. They keep, they keep reiterating more than $500 billion, right of incremental business by the end of 2027. I think it's 2028 and 2029 that the market is really, you know, debating. That's the bull bear debate. And then like I said, the sustainability of those extremely high margins. And sort of to your point, well, I was just, I want to hear where you land, but I was just going to sort of explain for the listener, sort of see your point, John, from earlier. It's not just that, you know, they're going to have, let's just say 500 million one year, billion. These numbers are too large. Yeah, I know, I know. Me too. I do that all the time. So half a trillion, let's just say of revenue in one year very soon, if it's not 2027, you know, but the problem, I think for a lot of people is they don't realize that or people don't think about the fact that as you said before, there's going to be a growth rate on that and the earnings that go along with that for you to really, you know, be able to count, when you're doing a discounted cash flow to discount 2028 and 2029, if they just do 500 billion, if they just do that every year from now until Kingdom come, that's not, that, that plays unfavorably into a DCF. Definitely, definitely. Even though that's half a trillion dollars a year, which sounds pretty good. Yeah. So where do I stand is I haven't sold any, you know, I think it's a three and a half, three and a half percent position in the portfolio. And you know, I have, I like it, I like it, waited where it is. I like the holding. My conviction has not changed. But, you know, I just want to keep going back with these times are so uncertain, right? Like so uncertain. And so I have more conviction, as Bear said, in the sort of picks and shovels plays, you know, like a company like, Amphanol, which we own, a company like GE, Vernova, a company like Quanta, a company like Train, Eton, these companies in my opinion are going to be fine without AI, without AI. Like if AI was, you know, what's exactly if AI was to disappear hypothetically, you know, I think their P multiples will come in, but like these are incredible businesses, stand alone, absent of AI. And you know, AI is probably like 30% of their business on average. It's not as much at Quanta, but you know, just on average across the sort of picks and shovels group. But they will, you know, stocks will fall hypothetically if AI disappears, stocks will fall, multiples will come in, growth rates will slow, but then they will, you know, fill it in with other just industrial work. And so, you know, I have more conviction in those names, but I'm not low conviction on Nvidia. I mean, it's in the portfolio I like where it's at. I think Jensen is, you know, he's the architect of all of this. On paper, it's the highest quality business you will ever see. I mean, if you look at the growth rates, the margins, the returns on investment capital are above 100%. It's an, it's just an incredible business. Yeah, that's, that's a great point. That's where I could be sort of too low. I don't want, I want Josie to hear what you said, because I think, I think what you said is important is that there are other higher conviction names for you. And I would agree that there are other higher conviction names for me than Nvidia. However, where I could be wrong is what you just said, which is, this is a pretty dang impressive business right now. What they did last quarter is recently his last quarter and it made do even better this quarter. Growing 60%. I mean, it's incredible. Yeah. Yeah, growth. I mean, like it's, it's the obvious thing to say, but it's still true. And, you know, until that slows to what we all think it will eventually have to slow to, I'm just going to be wrong. So we shall see what the, what the next few corners bring. And then just the optionality, like you can't model Jensen, right? a one of one. You know, you can't. Yeah. And also, none of us are saying that it's not worth $5 trillion plus that it's already, you know. So it's like, exactly. This is this is hardly a bear case. It's it's more just a, you know, how high can the trees actually grow? It's a case for, yeah. It's a wonderful, wonderful, wonderful business. And yeah, I think, you know, I'll say this just to kind of invert it. Um, that low PE, that low PE on, you know, two, two and three year out numbers, it gets really low. I mean, you could, you're looking at like a 12 to 15 PE on two or three year out numbers, depending on what they end up printing. Um, that low PE could, could be an indicator that the market has it wrong. Okay. Or it could be an indicator that the market thinks earnings have peaked and are about to fall off a cliff. We don't we don't know yet. Um, I'm of the right. I'm making a pinnacle. You're making a really good point about the margins too. I think the market, the market is buying revenue growth at least for a couple more years. Um, the market may be pricing in too much margin contraction. So that's another way I could be wrong. Yeah. Um, I'm of the opinion right now that it is, uh, that it is, that it is a undervalued to reasonably valued position. I mean, I, I don't think that it's overvalued. Um, and I think it's, I think it's possibly undervalued. I don't think they're at peak earnings right now. I could be wrong. And then like I said, there, there's just optionality that with Jensen that you just don't know and you can't, you can't imagine, you can't model completely agree with all of that, including the optionality. I think, I think, I think what what I heard was fairly value to undervalued. And I mean, I, I think what we are both saying that I'm not sure if drowsy shares or not is that the conviction is just not there, there are other companies where it's easier to see, uh, or it's more obvious that they are under on the undervalued side of that. Um, I hear you to, to, to, to a free land of value. Yeah. Instead of for me, for me, the way that I look at it. And if you believe, but some analysts, again, uh, I've already said about 10 ways I could do that. I was reading through 20 trillion market cap by 2030. So, oh, no, no, and I'm not saying that's in my anything modeling and, and it's that, I don't know, I would be, I don't know, that's any plus in estimates like four years in the industry, which compared to like credo or a lab or any of these other companies is actually really underperforming potentially or micron up, you know, 500% or sandisk of 1,300% in a year. But it does point towards fairly valued or even slightly undervalued, um, to some extent. And, and no, when you, when you hear numbers like 20 trillion, um, you know, I think, I think the immediate gut reaction, um, is that's outrageous, right? But, but like I said, like, what if, what if AI is the technology that ushers in a fourth industrial revolution, like what if and and how much are we going to be spending on this stuff once AI start to discover things like really discover things, you know, cures for diseases, um, you know, and you know, planets with with with life on that or something like it could you imagine how much is going to be plowed into this if we discover life form another planet. I'm not saying we will, but we don't know what's going to be discovered yet. We don't know what new industries and new products will be completely invented that we're not thinking of right now. Um, and you know, like, like I said, so Jensen and and Elon Musk have both quoted 50 trillion dollar tam for physical AI, because they say they all factories, all physical AI needs to be replaced or infused with AI. I mean, physical infrastructure needs to be replaced or infused with AI. Um, so if if that number is right, and like I have no idea if it's even in the right galaxy, but if that 50 trillion dollar physical AI number is right, um, and if token budgets continue to explode up, and if this does usher in a fourth industrial revolution, remember, revolutions last 50 to 100 years. Well, if you start thinking like that, then maybe a 20 trillion dollar valuation on one company isn't like so gut rich, you know, wrenchingly outrageous. Once again, I'm, I'm not thinking that Nvidia is going to become a 20 trillion dollar company, but I do want to keep an open mind right now just because AI is like nothing we've ever seen before. And if you look at the spending from these hyper scalers, um, that's like nothing we've ever seen before. And everything is flipping on a Ted right now. Like what do what do what do like investors like to see a lot of times it's like, um, balanced and like stable capital allocation philosophies, you know, balance meaning they get, you know, balance means they get, you know, they pay 30 percent of their earnings out as a dividend. So there's 30 percent payout ratio. And then they, you know, maybe make some acquisitions. Um, they, they keep their debt ratio within a certain target level, um, debt to EBITDA, which means if it starts approaching the top of that end, they have to pay down some debt. And then they use the rest for acquisitions and buybacks right now that's going to be completely inverted. And mega cap companies are not doing anything. They're just spending on capex. They're just spending on catbacks. There's no more room really for big dividend increases. There's no more room for buybacks. Um, and, and, and we don't know what all of this catbacks is going to create, but it's a massive, massive, massive amount of money. And if you just look at space, like one possible version of the future, right, is, is that we start to build out space infrastructure. Um, I mean, like right now, right, like we have mining operations on earth. In the future, we could have mining operations in space. Right now, we have factories on earth. In the future, we could have factories in space, not just AI token factories, but maybe factories making parts, right, to repair and, and replace other stuff that's out in space, other infrastructure we put out in space. Right now, we have telecommunications infrastructure on earth. And we're actively working to build that out with SpaceX and with AST space mobile and with Project Leo and Amazon. Um, you know, right now we have transportation networks on earth. We could one day have transportation networks in space if you talk about launch pads on the moon and Mars. So like metaphorically speaking, you know, this is like before railroads, like we're, we're, we're on earth. You know, metaphorically speaking, we're about to railroad space hypothetically in one version of the future, right? And it's not the galaxy with railroads. But, but metaphorically speaking, right? Like digital roads, like AI railroads in space. And even if none of this comes to, I agree that it's all possible. I agree that 50 trillion is, you know, there could be an understatement. I agree that, you know, there could be a 20 trillion dollar company sometimes soon. I just, I, and, and you know what, even if there will be a 20 trillion dollar company at some point, not too long from now, even if this doesn't work out in the way that we're all talking about. Because if you think about it, I think it was 2017 or 2018, there was no one trillion dollar companies. Right. And most, I may be off a couple of years, but, you know, and, and the ones, do you remember Fang? You know, Netflix was in there, you know, so like, I do remember Fang. It's just, it's funny. The, obviously Apple, you know, Microsoft, Microsoft wasn't even in Fang, right? It was right. Right. Right. So, so some of the ones that were obviously the other three besides Netflix have all, you know, benefited greatly. But it is interesting that other companies that we weren't expecting have come along and even surpassed some of them. And certainly a lot of companies have surpassed Netflix in size. So, so certainly, you know, the future's come in. It's going to be, it's going to be wild. We, we, you know, I think really, you know, just to remind everybody, if all of this comes from your fruition or even if something completely different plays out, the, the Mag Sevener is going to do really well. The S&P 500 are going to do really well. I mean, I think so. I, it's just, it's so unpredictable. And so, like, I wake up every day telling myself, like, I know nothing. I have, like, because it's so uncertain, it's so uncertain. This technology is literally game changing. It's game changing. And we don't know where this is going to end up. I just want to keep an open mind that things could get really, really, really big. But also, that may not happen in my lifetime, because, you know, industrial revolution, like I said, and that's what Bezos calls it. It's an industrial revolution. An industrial bubble is what Bezos calls it. These things take 50 or 100 years. And so, you know, I'm talking about things that may happen, you know, point to transition into doing a portfolio review. So, Vereniah, as folks know, we compete to have the worst, I mean, best portfolios right now. So, we'll do just a quick review. And John, definitely welcome commentary, but no need to review your portfolio. And then I think that sort of the unpredictability, first off, great episode title, but also would be a great point to begin our next conversation with you, hopefully, in a little bit as well. So, Veren, do you want to go first or second? I think I think you went first last time. So, why don't I go first this time? And let's, we'll just kind of breeze through. So, after a brutal April, or sorry, a brutal, brutal march. My April, my April wasn't quite as bad. So I was up 12% this month. Again, I don't think it was anything special that I did. I think this was completely a bounce back from just being decimated earlier on. AI trade was back on for some of it. In video, I finally caught some tailwinds as well. So I, on the year as of right now, and this isn't including today, which Mag 7 is taking a hit, and Nvidia is down big, but I'm down 4.8% on the year, coming in after being down 15% plus last month or as of last month. So this has been a good, a huge comeback for those rooting for Team Drowzi. This is a big win, hopefully, and it continues. I know it's probably not good enough to overtake the champion. Let me see, I didn't, I didn't really make any moves this month. In fact, I did not make any moves. I didn't buy or sell anything in this portfolio. I just held on and really it's going to come down to pretty pivotal in terms of beating the market and being bare. He's riding high in the S&P. Obviously, a bunch of my portfolio is leading the S&P like Nvidia and Amazon and Google and Microsoft. All top 10 holdings for me drive my gain so that would also even out what bear is doing, but it really is going to come down to Axon and CrowdStrike. Axon reports next week and to put in a teaser for listeners, we're going to have an Axon heavy episode on our next shoe in the bull pod. So definitely send Axon questions and thoughts are ways. But it's going to come down to those earnings. And I think if which I hope in plan for having a good earnings week next week, I think that's where I have a potential to kind of kick in to high gear beyond kind of S&P riding. So down 4.8% this year to date, not bad for getting better. - Yeah, very nice April and April was good to me too. So after being down whatever it was, double digits at the end of March, I am now down less than 2%. It's minus like 1.6% right now. You're coming back though. You're closing the gap and yeah, I mean, April was nice to see things kind of come back. It's interesting to me the things that really surged and the things that either just kind of went up single digits or very low double digits. And unfortunately that's a lot of my holdings. You know, like I, Rubrik is still number one and I'm not really happy about that and having been happy about that for a while. It's only 6.4% for a single. - Did you see that? - Some dude, Claude Coded, security in his company and it deleted all the data in like 20 seconds. But I just could not stop thinking about Rubrik and that should be on every commercial for Rubrik going forward. I was like this, cement why Rubrik should exist. - Yeah, get your data back. But yeah, I mean, Rubrik is one of those that's only ended there right now. It's only up 8% in April, whereas like Credos up 81% in April. So it's, you know, things, things have, some things have bounced back more than others. Obviously as Credos come back, I've trimmed it some. But Rubrik, Hingehelt, Applovin and Credo and even Sofai now, there are all over 4%. I added a little bit to Sofai these last couple days after it reported and that's a great quarter or not, not great enough. Oh, and Reddit, I've gotten Reddit back to 4% plus. So then you got Mercado Libre, see limited ethos life, which I need to talk about on the podcast sometimes soon. Figure and your beloved axon. So those are all the 2% plus positions, everything else is tiny. So, but yeah, happy to be, you know, almost back to even. My way, Joe would be if John did his portfolio, he'd probably have us both be, but you know, keep hearing about people who are positive this year. And it's, it's, it's very impressive. And it's, it's like, this, this was definitely the year, not to be in software and to actually sort of focus on the, the stuff that's part of what's growing the most in the economy. And so I'm glad I have Credo and a couple others in there like that that are, that are benefiting from that because a lot of, a lot of the things that aren't are somewhat, say rightly, being left behind, you know, we didn't talk about this, we could have a whole another hour on this if we wanted to, but, you know, slowly, slowing growth or even any slowing growth is not necessarily the recipe when there's so many things that are experiencing the exact opposite. If, if multiples are high, when growth slows, it just becomes violent to the outside. Totally. And just absolutely, absolutely violent. Yeah, absolutely violent to the downside. I guess we're due for another one at some point, but I hope this build out continues on for, for as long as we hope in at least one version and one world within the J-Roe universe. So I think that's a good place to wrap up kind of episode 20. John, as you know, we're unbelievably huge fans of yours. We always learn so much having you on the show. We love learning from you and it introduces us to a whole new investing world. Again, another J-Roe world, but a whole new world. So really, thank you so much for coming on another episode and I'm sure we'll have you back very, very soon. And also just again, congrats. I just want to highlight how many really good picks you have made that you, you know, we now have the receipts. You know, you mentioned them on our last podcast. So just absolutely phenomenal job. Thank you all very much for having me. Always love being on the show and I wish you the best of luck.