This Could Become the Biggest Bull Market in Years

One of the reason why the market keeps getting this wrong is cuz analysts continue to get the whole capex trade wrong. >> What's up everybody? It's LG here and welcome to Milk Road AI, the daily AI show that should have put in that bid for SpaceX but just didn't believe in Elon enough even though he's done it for like three companies now. Today's August 5th, 2026 recording on the fourth. Let me ask you something. Which hyperscaler are you betting on for the next few years? Are you picking just one or rotating between them as their stocks bounce following big events like earnings? This is our Wednesday rollup episode where I sit down with our three leading AI analysts, Melvin, Vincent, and Kyle. And on today's show, we get everyone's emotional reaction to the market's massive implosion and then rebound this week. And also dive into SpaceX and what it would take for them to actually start buying the stock. Plus, we look at which hyperscaler might be the best to bet on and whether all of AI is just the same trade. This is more of a portfolio strategy episode. And finally, the boys give us the stocks they're picking this month as it's payday in Milkroad Pro, which means each analyst gets an extra $500 added to their portfolio. And you can see what they're buying this month for just a dollar at the link below. A reminder today that our podcast is free and it wouldn't be possible without our partners at Securitize, the regulated rails for tokenization, and BitGet stocks 2.0 with real liquidity, real dividends. Keep an ear out later in the show for a message from both of them. All right, guys. We're back and before we go further, we are in the presence of a celebrity today. Avin, somebody who just, no, not Melvin, not even Vincent, somebody who was just on Fox Business News or whatever it is, Fox Business yesterday. Kyle Reedhead is gracing us with his presence today after being viewed by millions and millions of Americans yesterday to discuss the everything bull market. Welcome, Kyle, to our show. Thank you for coming on. Glad to be back. I'm here all the time. What are we talking about? >> Thanks for making time for the peasants now that you're you're you're a huge star. But dude, how was let's let's kick off with that though, Kyle. How was your appear appearance on like Fox Business? >> Yeah, it was fun. They uh they're fans of Milk Road. They love our research. They love our analysis. Uh Charles Tay, who's the host there, runs the show, he actually reached out and was like, "Hey, I love what you guys are doing." Uh and so yeah, he reached out. So anyway, went on and talked about the everything bull market, which I think you guys listening to this probably already know. Um, but you know, over the last week, we've just seen we had the ISM come out earlier this week, which is absolutely crushing. It was like 56, I don't know, two or.5 or something. So, the economy is expanding. Earnings have just been absolute record earnings for all of Q2 so far. We're at higher earnings than we've been in like 25 years. So, the economy is absolutely booming. And then at the same time, we've got, in my opinion, inflation going lower. It was a little blip because of the war, which who knows, may or may end. I feel like I say this every single week, but looks like it's ending this day. I'm sure it'll be different next week, but anyway, and so oil is back down to like, I don't know, 75 or even lower. Um, and so at this point, and then AI adoption is accelerating. And so I think, you know, the AI infrastructs that got sold off last week. I'm sure we'll talk about that. We were all calling for the last two weeks that they're going to rip back. So like, bye bye bye. And now the application layer, I think, is taking off, too, because everyone's seeing how good the earnings are. If anyone saw Palunteers's earnings from yesterday, holy it was unbelievable. So, we literally are going into what I call the everything bull market where everything around AI, whether you're integrating it or you're building it, is just absolutely booming. I don't think that's going to stop anytime soon. >> And just a reminder, our last couple episodes of this rollup edition were called I think even the headline thumbnail on YouTube was buy the dip and the next leg. Uh, and looks like you guys have been right so far. Vincent, how are you feeling today about the market now that you're in the presence of a of a mega celebrity? >> Yeah, I'm I'm a bit nervous now. >> It has been a roller coaster of a ride the last couple of days. I mean, last week we had the Leopold blow up, right? And uh it kind of explained why everything sold off that aggressively. Um then we had like Kyle mentioned crazy earnings. We saw all the the cloud providers seeing massive backlogs and and and the market actually seeing through copex sentiment changing on copex I think uh that was really bullish. We saw the infra providers also today the the chemicals players like unity that I cover really good earnings as well and then on the AI adoption side of things we had pollanter we had hood we have Lily tomorrow reporting really good earnings. Um, so I think we s we see that that the bull the bull market is is fundamentally intact and I think there are a lot of great entry opportunities still in names that um yeah, basically rallied a couple of percentage points during the last couple of days. >> Wow, Melvin, how about you, man? >> I feel great, man. I feel like my portfolio has doubled basically in the last, you know, a week or so. So, um, so we're, you know, slowly climbing our way back to all-time highs again, or S&P has actually reached all-time highs today. Um, and like everyone mentioned, we had the big big, uh, selloff last week, uh, with Liupold. Actually, funny enough, Citadel is the ones who, you know, pretty much caused the, um, selloff because of market fears about rate hikes and whatnot. And actually, Citadel came out today saying that, um, you know, this is actually really good for the market. were structurally going higher. Uh, which is funny because they were able to manage to get like 10 or 20 billion dollars in assets from Liupole. Now all that's about to run up. Um, so structurally like everyone said the bull market's still intact. We had great earnings and SpaceX is actually reporting today and I have a strong feeling that SpaceX was probably like likely to raise their capex and that's going to be a huge huge deal for rest of the market. um because we've seen time and time again all the capex numbers are just getting absolutely erased and it's going to continue to go higher. So I remain extremely bullish than ever before. Can can I ask a question to you guys? How was this week kind of emotionally for you? Because if you think about it how we felt last week doing this rollup and how we're feeling this week, right? It's at least it's to me it's completely different. But then again, it shows if you have conviction in names like Micron, like Bloom Energy, Nebas, uh, or or Palunteer, buying the dip proved once again to be the best thing you could do as an as an retail investor, right? >> 100%. The thing is, emotionally, I feel very similar to I was last week. I feel like we were on the podcast last week laughing, just saying, "Hey, this is crazy. We should buy this stuff." like maybe there was a little bit more nervousness of like what is going on right now? Why are we down so bad? >> But at the same time, I think we were all pretty convicted that this was going to come back and you know, we were we were still laughing and having a good time and not really worried. We were saying the bull market's going to continue. And look, we're not even a week out. Well, I guess we're a week out from when we were chatting before. Um, but it was, you know, we chat on Tuesday. Wednesday is when the market really nuked and fell apart. And I don't know, I still feel just as convicted, just as excited. If anything, you know, when things get really green and too green and everyone's saying that they're excited, that's when I start to get more nervous. Uh, now I'm not there yet. I think it's still somewhat early days, but um, you know, you wish you bought more and and those crazy moments, you know, >> as ever, >> of course. >> Just going to pause there for a second to point out that the market is showing signs of something kind of different happening and our analysts at Milk Pro are all over it. They spent the last couple weeks making a lot of trades, getting out of some positions and then getting into a lot of new ones, getting ready for the next wave of robotics space or even kind of picking some different AI winners. If you want to see what they have in their portfolios, what positions they're opening, it's just a dollar in Milkroad Pro at the link below. It's always those scariest days. And even Kyle, we did that episode last week that was, you know, the what really stood out for me from that 5 days later now is that it was the biggest retail sell-off day ever, right? And that alone it's like what are the chances of that just happening multiple days in a row or multiple weeks in a row like probably pretty slim. So it's like that in itself is like such an it's so always so obvious in retrospect for people that it's like man you had a historic sell-off day that's a signal buy always right but but it's so hard to do for people in general. Melvin, how how did you feel man because you were you're you know you're leading the portfolios here and we've seen a huge swing in yours man in the last couple week last two months has been wild. I mean, it was tough to watch, but I didn't give no shits. Like, I knew this was going to come back up, to be honest. Honest to God, I feel like I need to sell one of the my kidneys at this point to buy these freaking gifts cuz I am running out of money. I basically have went through my entire savings. I'm like asking Kyle for a raise here. I'm like, please. >> Fun fact, I think I have I should screenshot this and show it on the podcast next time, but there was multiple times where Melvin's like, "Bro, can you can we please get more money for the portfolio? Can we please get more money for >> and I even asked Kyle if I can have my paycheck like over two weeks earlier. This is how bad like it has gotten cuz honest Yeah, go ahead. >> But Melvin, you know why the Leopold funds blew up, right? It was because of leverage. So don't do that >> No, no, I don't do leverage. I I I don't do leverage. I only do a little bit of options and just you know uh buying um stocks like quality stocks and I the quality of stocks that I'm have a deep conviction on like I want to be able to hold and buy more like Bloom Energy I added a position like Vince said you were deep in it. It's up 30%. You know so stuff like that you can't pass up these buying opportunities man because market only gives you so many before we go next higher up. So by now you you're 10 5 years two years you will thank yourself. >> It's it's it's crazy if you think about the last year or so or maybe even like let's say 18 months. There was the deepseek moment last year where everybody freaked out. It was best buy the dip ever. Uh then we had April this year I think as well and then we had last week. All of them like those really sharp selloffs just time and time keep proving that they're great buying opportunities because the long-term structural kind of AI buildout remains intact. Um and we were right once again buying the dip. >> You know what's interesting though and I'm actually just going to find this tweet I had. Yeah, here it is. Let me share a screen cuz the market and one of the reason why the market keeps getting this wrong is because analysts continue to get the whole capex trade wrong for some reason. So this chart was pretty funny. Um back in uh at the start of 2026 so in January the consensus from analysts is that uh we would have 546 billion in capex from hyperscalers. Start of Q2 earnings. This was like two three weeks ago I guess three weeks ago now. they had already moved it all the way up to 757. Okay, for the total of 2026. Two weeks later, they're already moving from 757 to 793. So, it's very clear that these analysts have absolutely no idea what they're talking about. They every single quarter they think that it's going to start to decelerate and every quarter it continues to accelerate further and for some reason these analysts just cannot figure it out. So even like their numbers for 2027 were 630 at the beginning of this year. 630 billion. We already know I mean I think Morgan Stanley put one out you know 1.3 or 1.4 trillion. So like more than double what they were estimating at the beginning of this year. So I mean the analysts here on Milk Road have been telling us this whole time like capex is going way higher than what analysts are suggesting. There's no way it's going to be this low. And we keep saying that every quarter. And if you if you were the one that like understood that and you knew that what the analysts are saying is wrong, this game is a lot easier, right? You've been able to buy Micron earlier, Nibius much earlier, and all these things that everyone's Bloom Energy, all this. So, the analysts just keep getting it wrong. And I I don't really know why. It's like they can't quite figure it out, but that that's the biggest thing that I think is holding people back and why everyone thinks this is a bubble or they can't seem to buy these dips, you know? But as long as you believe in this and this keeps going up, then the buying the dips is a lot easier. And to add one point is that this is just hyperscalers. These are just like four to five companies. Um there's so many so many different companies outside of this that are spending massive amounts of money like Frontier Labs that are not being included in here. Um a lot of different smaller companies. So this is >> SpaceX and Tesla. Yeah, there's a bunch missing. But I think the the earnings of the hyperscalers were really important this time around for this the numbers you're showing on the chart to go up because what the hyperscalers were proving is that a they're seeing the demand but that they can al also convert that demand into earnings that basically proving that they're able to keep the pricing power and earn crazy margins on that compute. And because the the analysts are seeing that that they're expecting or they're also and and and that's also why the the the sentiment on on the whole capex narrative is changing because they're understanding it. They're investing into something that is really valuable in the future, right? That's also why they want them to increase capex. >> Look at this chart, man. Like their their cloud revenues are So what do we got here? Google's 25 billion, Amazon's 42 billion, and Microsoft is 60 billion a quarter. And it's accelerating. Like, you don't get products that are, you know, 100 billion plus per year in revenue that are accelerating. Like, these are supposed to be mature businesses at this point, and they're actually like getting more and more. Like, the growth is getting more and more. It's it's actually crazy. Um and so and then the backlogs are like 2.3 trillion in total between the four hyperscalers. 2.4 trillion. Yeah. So like I think the market finally is like okay yeah not only are they going to spend more but they're actually making a lot more too. And you know right now I think they said it was about maybe it was Amazon said this or or Microsoft but three years it takes for that capex to turn into to profit. And I think that'll likely shrink right as they get better at this and um and they get more efficient at it. Um, I think that'll probably shrink. So, um, it's it's a crazy business they've got going on right now. >> How do you think about the legacy business of hyperscalers and how important is that to the stock price or do you think it's all about the cloud business and that's the only thing investors care about? >> Uh, no, I I think if they can continue to grow their legacy businesses and especially if they can use AI to make it work even better, like Meta improving their ad business because of AI, that's huge. Google improving their ad business because of AI. That's really huge. Even like their search business is sure getting a little bit um you know obviously competition from LLMs, but they're finding ways to kind of rejig that like the AI overviews, the AI summaries. >> Search business grew 70% of Google, >> but that grew probably because they had the AI overviews, right? Like that's part of that. So they're finding ways to put AI into their existing products to get them better. Amazon again with their ads business getting better as a result. Microsoft has tons of opportunities that they can do with their businesses by implementing AI. So I think both are going to grow cloud and their their businesses and that's the key and especially if they can both accelerate um it's it's unbelievable these companies. >> Yeah. And like if you're not using AI to accelerate your legacy businesses, you're essentially screwed. Market is not going to reward you. You know those kind of businesses can be replicated using AI, right? like this is why we have the great AI selloff or a great software selloff in the last like month or two. So if you're not using it, your business is essentially not going to survive in the long run. >> Yeah. >> Yeah. >> I mean what what's pretty clear is that we do not see this cannibalization that was expected right for instance with Google search and AI. What I'm trying to figure out is what is the most attractive hyperscaler to buy um and then kind of differentiate between the cloud business on the one side and the legacy business on the other side and kind of how important those things are. Here's what I think it is their business. So the the moat in AI is just compute and all four of these things just buying as much comput as possible. Everything else is noise. And so what I think is every time there's an earnings where the market hates it and they sell it off, you just buy it. It doesn't matter which one of the four. So like Meta right now I think is probably the cheapest because they got sold off a ton on their earnings. Okay, so you buy that. Google two weeks ago it sold off. Okay, so you buy that. You know, like then Microsoft and Amazon pumped recently. Okay, so maybe you just hold off a minute. They'll sell off at some point. It's just like you just keep adding to these things because they're they're buying up the most important thing in this world. Yeah, but that's only covering the the cloud business, which they're more or less all the same, right? >> But it's not it's not because everything we just talked about, everything is AI going into their existing products. And so that compute is either used for cloud or it's going to be used into into their own other products. What I'm trying to figure out is which of the legacy business businesses is benefiting most of AI and which can scale and increase margins and earnings per share the most with AI and that's not yet fully clear to me. Is it is it Meta with increasing ads? Is it Google because they can adopt Gemini across their Google suit? Is it Amazon because they can enhance uh whatever sales conversion in the retail and then eventually robotics in the logistics? That's kind of where where my focus is now on trying to figure out what's the most attractive hyperscaler to buy here. >> Buy some Oracle. It's been beat down. Like I think it's around $146. I'm not sure. They went as low as 110. They have a backlog of $600 billion. Six to 700 billion. Um, and they have a great legacy business as well. So, I think I think if you look at if you just look at the one of the investment thesis I always look for is great founders. Who is a great founder and Larry Ellison is a great founder. If you look at he essentially owns all of media. He has tons of other businesses. He's an incredible founder. Um, he's best friends with Elon. Elon doesn't hang out with losers. So, like he is the guy. No, I'm serious. He's the guy that you want to take a bet on because he's betting everything his entire house on AI and that's why I remain bullish on Oracle. >> I mean where I prefer hyperscalers is kind of on the on the risk perspective. I agree Oracle probably has the higher up up upside than than hyperscalers but their legacy business just derisks the entire trade right and if you have a lot of kind of beta names in your portfolio I'm not sure want to add further beta with a name like Oracle >> on the kind of compute trade >> I just think the riskreward in these hyperscalers exoracle is so good and their price to earnings are like under 20 like Google's 18 not even 17 right right now like what it's just crazy right so I just think you just have such a good riskreward and all these companies these four companies are going to go to you know 10 trillion in the next couple years um and the downside is so limited because their businesses are so good so um just such an easy buy you guys think um Nvidia is overvalued undervalued sorry it's been at the same price it seems like for the last year or two after its big runup businesses has only gotten better fundamentals only has gotten better. Like this is a this is a company that's absolutely printed money and and it's essentially the king maker of AI, right? But yet the stock price has not reflected that. Why do you guys think that is? >> So I can give mine. I got two reasons. So one just semis in general um are still pretty undervalued, right? Um and they pulled back a ton after everything that happened over the last four weeks. and they haven't they're just starting to get a bit of a bounce. Like Micron's up 8% today. Even Nvidia is up, you know, 3% today. So like they're starting to finally come back, but a bunch of the others jump back a lot faster than them. And I think it's because there's so many businesses now that are trying to build their own chips, right? Um and what I think the market is still missing is how many chips are actually needed going into the future we're going into. I don't think the market fully gets that as we go into an agentic world and a robotics world. And so I think that the market is one scared that Nvidia is going to get all this competition which it really is not like like even Google and all these others that are building chips. They are but it's edge cases and it's so small it's not going to cut into what they're building. And then the second reason I think the market is is um holding back Nvidia is the circular financing scare. So, they think that they're kind of leveraging up or getting too risky um by, you know, giving prepayments to companies and and backstopping OpenAI and Nibius and all these other companies. Um, I think that's nonsense. I think they're building more revenue streams, which is very smart for something they already know is going to happen. Um, so I think the it's the risk profile. Um, and I think the market's going to figure that out pretty soon. >> Yeah. So, I agree on the competition side that markets get that wrong. um because we will need so many chips that it doesn't matter whether hyperscalers and others produce their own chips. We're going to need every chip where they are producing essentially right the deeper point to me though is I think Palanteer and Nvidia are basically the same story. They were stocks of the prior cycle. they performed really well and then capital shifted into memory into behind the meter uh energy into compute nebas all those names and those names provided that good of returns that there was no need to cycle back into into uh a name like Nvidia that's the core reason I mean if you look at Palanteer today it's up 30% because they had just I mean their earnings yesterday >> unbelievable earnings we'll talk about those in just a second because it was the next level. >> I think Nvidia, I mean, they're not going up 30% in a day. That's that's pretty clear. But I think they will perform pretty well until the end of the year because the earnings will just once again prove that they are the king in the entire AI trade. >> Their earnings are going to be lights out. Guarantee us. >> Yeah, I I I do think that if you don't own enough Nvidia, now is a good time to definitely pick it up because there's some interesting data that came out. Um there's about 20 million AI chips that are running in um world's data centers as of right now. And that count is essentially doubling every 9 months. So which basically means that you're on the pace for 200 million chips by end of 2028. Where is all this going to come from? Like there's not going to be enough GPUs to support this amount of growth. Sure, that growth curve may um come down, but they can't even enough produce enough GPUs right now. So this is why I think you know Nvidia's margin's going to go up. This is going to be the same story as you know why memory makers are going up right because there's there's essentially a huge need for it and Nvidia controls the supply of GPUs as of right now. >> The market keeps thinking that the margins are going to come down from all these semi stocks. That's the like big fear in the market and I just I don't think we're getting there yet. Which by the way if those margins do come down that's super bullish for the hyperscalers cuz now the hyperscalers are buying cheaper products, right? Uh, so they're going to get more for for their for their money. So you kind of own just both. Yeah, I don't think it's happening either anytime soon. >> Just buy everything at this point. >> Just the everything bull market. I lost. >> Wow. >> Come all the way back around. >> Palanteer or SpaceX. Which one? >> Just just one more point on this everything bull market. I mean if you think about it from the portfolio construction perspective you can buy chips Nvidia you can buy compute whatever hyperscalers or neocloud you can buy memory as kind of an infra play you can buy energy with bloom you can buy the AI adopters so just also from a kind of a a diversification perspective you can think about adding an Nvidia or a hyperscaler to bring the risk down >> none of that is diversified so this is the thing where People are like, I got energy here, I got GPUs here, I got memory here. It's all the same trade. Don't >> memory. Memory. Memory and energy. Yes. Memory and energy. Yes. But hyperscalers and >> hyperscalers are different. Agree. Hyperscalers are different. Nvidia knows part of the semis. That's part of the infrade. There's there's the way I think about it. You have three categories. You have the infrastructs. They all move together. Now, some move more than others at times, right? But that's that's energy. Every single semi which includes memory. That's your neoclouds. All that is in one trade. Then you have the models. That's just your enthropic open AI. Uh and then you have your application layer, right? And that is your palunteers, your service nows, your, you know, Eli Liies, whatever. Those are the three categories. There's a bit of diversification there, you know, because like when infrastructs come down, application layer goes up. Although right now we're in the everything bull market and so they all just go up together. >> Uh so that's the good news. Except crypto, which I the everything bull market has to do with stocks. That's what I have to keep explaining. >> I mean, crypto is at least not going down a lot over the last couple of weeks. >> Oh my god, guys. Honestly, it's usually Vincent who brings up crypto, but now Melvin's bringing up crypto, even as a joke. My god, Kyle's doing everything he can to not do crypto shows. And here we're like, let's Hey, Kyle, can you tell us about crypto for just one second? Just tell us a bit more about crypto. >> No, we're not talking about that right now. >> Real world assets like funds, treasuries, and private credit are still running on rails built decades ago. gated, paperwork heavy, slow to settle. Everyone's talking about tokenizing them, but far fewer can actually do it and do it without cutting regulatory corners. Securitize can. It's the SEC regulated infrastructure bringing real world assets on chain. 9 years in native tokenization, not wrapped, backed by Black Rockck, Morgan Stanley, and Kathy Woods Arc Invest, and chosen by the New York Stock Exchange, Van, BNY, and Apollo to do it at scale. It's the regulated bridge between traditional finance and crypto. Tokenize the world at milkroad.com/securitize. >> Everyone's tokenizing stocks these days, but almost nobody's doing it right. Thin liquidity, prices that drift from the real thing, dividends that just vanish. Bit stocks 2.0 is different. Real NASDAQ and New York Stock Exchange depth through licensed brokers, prices mapped one to one, dividends paid to your account in real time. Plus, you get the lowest fees in the market at just 0.04%. And you can trade them like any other crypto as margin in earn in grid trading. Tokenized stocks finally done right. Head to milkroad.com/bitgget to get started. I want to talk about SpaceX. I think that's something we spent so much we dedicated so much time a month or two ago whenever the IPO was to SpaceX and we made all these predictions. It fell to like the bottom was like 107 a week or two ago. Uh earnings are tonight, right? You guys think they're going to announce some more capex. So by the time people listen to this, we will know what the deal is. But already the the stock is bouncing back, I think, at like 123 as we're talking now. So, is this is this is this something? Did this thing bottom? Is that Are you guys piling in? Melvin, you're going to DCA your way down? >> No, I'm not DCA. >> I'm not DCAing here because I do think Well, I think DCA will be after what Elon, you know, after the whole um um the presentation today and how that call goes. But there is something interesting though because the like the whole IPO like unlock is basically what happened to Meta back in 2012 2012 uh when they IPOed their stock actually dropped 50% uh by the time the first lockup uh expired 91 days later in August 2012 then it dropped 48% when the next batch of shares unlocked in October by 180day uh the 20 in 2012 um the stock was down 41 31% roughly from the IPO. Then by December, it was down 29%. So after a full year of IPO, it was basically down 36%. And anyone who saw those numbers, panic sold one of the greatest long-term investments of like at at the absolute worst time possible, right? And SpaceX looks very similar. Um 2 days after today, 20% of shares will unlock. That is roughly 911 millions uh shares a float. um which is larger than the SpaceX entire current float right now. Then another 7% locks and another seven another seven another seven until all the way until December 9th. That is when like all of the unlocks will finally happen. Um so the question that Elon has to answer today for to prevent this you know IPO drop or the stock from dropping is is lockup psychology right like does the management or the print give investors enough um you know confidence to hold all these lockup period um you have to Elon have to have to be able to sell these stories um and he did a great job before the IPO now he has to you know do the same thing over today now the second thing is uh starship progress I V13 flights um succeeded uh like last week or two weeks ago um there was like a blue booster flash down they launched 20 V3 satellites which you know which came down 20 minutes after so Elon has to you know tell the investors or show investors that there is progress on the starships continued progress on the starships as well um and then it's going to be all about the AI compute buildout right space SpaceX is roughly on track to build two to four gigabytes of comput capacity by year end and uh Google and Anthropic has already like locked in um you know most of that and Eli has to you know tell investors basically like we're going to continue to build out this is a very profitable business and these are long-term um um you know contracts and finally the the cursor you know we all know Elon right like bought cursor couple months ago um they're the Elon has to show that cursor is actually starting to bring in revenue um not just you know it's not just going to be like a fluke um like some of his other other buys or maybe not other buys but he has to show investors that he's making money from cursor. I think that's going to be the key key questions uh during this earnings call. >> So interestingly we just talked about the three categories of the market right which was the info info side the model side and the application side. SpaceX like the hyperscalers fits all three of these. It's got a compute business. It's got its own model business with Grock and Cursor. It's got the application side with, you know, Starlink, uh, with, um, uh, with X itself, right, and ads and all that. So, it's got it's got kind of all three, which is is quite interesting. And so, I don't think the market fully understands it yet. But either way, its numbers are not going to come in great. Like, it's going to be like, I don't know, 6.9 billion in revenue, maybe. Uh, but it's going to be like 10 point something billion in capex that is spent that quarter. So, like, is that worth 1.5 trillion? Like, definitely not. So, could this go much lower? Especially when you consider all the unlocks coming like for sure it can. What I think they need more than anything is I I think of SpaceX as a compute company today. the space does not matter right now. Like it it really does not matter because that's still so many years out from actually driving any reasonable um like significant revenues to the company. What can make a difference now, like literally by the end of this year, is the amount of compute that they can build and sell because SpaceX is actually the best company in the world at building out compute. They are so fast. It's unbelievable what they have been able to do. And if they say that, hey, we're going to do that again. There was um a report that came out from I think they're called like Fund AI or Funday or something like that. And they're predicting that SpaceX is planning to launch uh or to build out eight 8 gigawatts. If they build 8 gigawatts, right now they're monetizing one gawatt at like close to $50 billion a year. So imagine they did eight. Now that wouldn't launch like in the next couple months obviously, but imagine they could do that, I don't know, in 2027 or whatever. It completely changes the this business completely. So, I don't know if that's actually if that's true. But if they sort of guide that and some of that's going to come sooner than later, then maybe SpaceX is going to have a big move. But otherwise, I think it's probably a better chance that we go down. And at that point, I'm going to be buying more. I have bought some SpaceX. Uh, and I would be buying more if it does go down. Short-term, I agree compute is the is the most important part of the story. I still bet on the Tesla SpaceX merger eventually down the road. uh probably in 2027 is my guess, but that's just a guess. Don't know. Um and it's it's it's just overvalued. It's it's I mean to me I explained this in in when SpaceX IPOed I think the majority of the business is 2030 and even later. And I learned my lesson with Tesla and kind of physical AI robo taxes, right? This is similar to Tesla in 2018 2019 when they started to talk more and more about robot taxes, right? It's 2026 and we're still not there. So yeah, that's that's my take. Not not buying even if the stock will go down further. The other number that I think is going to matter is um Starlink is obviously the like cash cow. This is what's bringing in a lot of the revenue for the business. Uh especially previous to to the compute business that they only started like was that like three months ago. Um and so Starlink still brings kind of the bulk of the revenue. Yes. Now over the last 3 years they've had to lower their um what they're charging. So their prices. So I think it went from like $99 down to like $66 average revenue per customer. And I think the market will want to see that stop going down if that can increase or even just stay flat. I think the market's going to like that. Especially if they have some meaningful growth, which I'm I'm pretty sure they will. There's also a big business in what Starlink is building. like to go internet directly to device and not have to go to to to satellite just first like it's huge and they can get this globally to a billions of people that don't have access to the internet um and then you know provide way better service and faster service to even people that do already have internet at better costs. So it's a it's a potential massive product as well. And then, you know, the thing we haven't even talked about is they also have like one of the biggest social media applications in the world that Mark Zuckerberg just just joined and Jensen Wong just joined. And like you name the people that are using this thing. If they can really monetize that well with their subscriptions and with um with uh you know advertising, which I think they're going to show good numbers here. I don't know, but my guess is they are. It's been kind of down only since they acquired it, but I think that's going to we're going to see a turn in that. That's also a pretty big business and that can continue to grow as well. So there's just so many businesses within this one company that is unbelievable. And then when they merge with Tesla, it's like what is this company now? It's insane. So oh, and then there's Terraab. Right now it's still far away, but they're going to be building their own chips and this is going to be huge, right? At a scale that no one's done. So it's just there's so much to hear from this. But that's many years out. That's many years. >> Exactly. That's But the picture you're painting here is really attractive. That's kind of the long-term vision of SpaceX, right? But there's opportunity cost to capital now. And I think there's just too many other good buying opportunities at the moment after last week to to to put the money elsewhere versus SpaceX. >> Can I highlight one thing that um Kyle just mentioned? Um going back to fun AI is if if SpaceX builds out 8 gawatt of you know compute one gawatt roughly cost 100 billion dollars to build out. So this is the reason why why they're getting like all the analysts are getting the numbers wrong because the number essentially like it used to be 50 gawatt like four months ago. Now everything has doubled in price like the memory and um optical transceivers and everything. So now those buildup has gotten so much higher in price which means we're likely going to move the capex up even higher. So which means all the infrastruct stocks and everything that got hammered last week like Vincent was saying is going to go move higher. >> Can we talk about a stock that Kyle should have bought more but he did not? >> Yeah. Which one? >> All of them. >> Yeah. All of them but Palunteer mainly. Ah, >> yeah. man. I was I So, just before you go into it, I I was doing a bunch of research on it yesterday because I was going to cover it because it had its earnings and so I was like, "Oh, This is looking like kind of nice." And I was like, "You know what? Maybe I'll just put a bid in before earnings just for fun." And then I had to one, I was on a a broadcast and then I had to get a couple things figured out for the live show, so I didn't do it. And I was like, "Okay, damn it. Market's closed. Can't buy it." And then I'm live on the show and I couldn't even see the earnings because I was like talking and whatever, but all of a sudden I'm just seeing the price go up and up and up. I was like, "Fuck, I guess they had a really good earnings." >> Oh my god, man. That's hilarious. Melvin, go ahead. >> To give some readers context, Ky says this all the time. Don't listen. >> Oh, I >> He's like, "I missed out Nebas. I missed out my never mind." No, I'm just >> I didn't miss out on nothing. >> No, no. I watched the live stream. He was being serious. So, >> no, but in all seriousness, this was once again the best earnings of Palunteer. They're just proving that this is one of the best ways to play the AI adoption side of things because to them it doesn't matter which is the best model. to them it's it it it it's really important that the companies remain in charge of their own data which all of the companies want right and they're providing the product or the the software layer on top that enables the companies to do that. So and also if you look at some of the numbers I mean 149% US commercial revenue growth year-over-year is just telling you that all the companies in the US are standing in line to adopt this because they themselves are not able to use AI properly right um and then if you look at one final number 157% of net dollar retention rate so this is telling you that the existing customer base. So once you're once you're a customer of Palunteer and you spent $1 last year, this year you're spending 100 or or $1.57, right? So once Palunteer is embedded in a company, they keep using more of Palunteer because they un they get value out of AI versus models just where companies spending tokens and increasing the budgets. This is just uh and and and and and the score of 40, operating margin at 62% telling you they're growing the business while being profitable like no one else out there. So yeah, uh mic drop. >> Yeah. And this is everyone said that software was going to be cooked, Palanteer was going to be cooked. Uh and now you saw like Service Now two weeks ago had their earnings and it was great. Not Palunteer level great, but it was still great. and they have a similar type business which is basically they go in they work with companies and they help them kind of build out agents and work streams and automate and all this kind of stuff. Palanteer I think does it on a whole different level. Um and you know they're just making their customers so damn profitable that it's one of those businesses that are so sticky because you're like my business is completely different when I use Palanteer versus when I don't use it, you know? Um and Alex Karp talks about this all the Like I didn't want to just be this thing where it's like a software that you pay and you just are kind of stuck with us. I want to actually make a real difference to their business. And man are they ever they they went over a bunch of case studies and it's crazy what they're doing. So their CTO actually made some really interesting points during their call yesterday uh which basically she basically I'm just paraphrasing it but it which basically uh goes like this. The public AI benchmarks are essentially theater companies that optimize their models to look you know companies you know optimize their models to look good on generic you know tests but that does not translate into real business value what actually you know matters is how a model performance performs on a specific like proprietary data and task you know like palenteers and the example that struck with me yesterday was Nvidia's neatron ultra model with zero cost post training beat Frontier models on five real customer production tasks within 24 hours just being plugged into Palunteer's platform and that implication is huge right because as AI models get cheaper and more commoditized um that is actually really good for Palunteer because customer can swap the underlying platform or the model for Palunteer's you know uh software essentially >> this is the flywheel for the everything bull market too Right? Because in order for this all to work, companies need to integrate and use AI. And it needs to actually work for them and make their businesses more profitable so that they can spend more on tokens and build more with AI. And as long as that keeps happening, they will keep spending more and more and more and more and more. And as long as they make more money from it, it just keeps going. And that means we need more buildout, more buildout, more compute. And so we what we want is Palunteer, Service Now, whoever else is doing this. There's there's more companies, Snowflake, etc. It's like, keep pushing them into other businesses, get them more and more profitable, and that means we need more and more buildout, you know what I mean? We need more chips, we need more memory, we need more compute, and as long as that just flywheel keeps happening, and then robotics steps in, it's like this bullark bull market doesn't end. Now, can Palanteer touch that many companies and make that big of a difference and Service Now and the others? Like, I don't know. We'll have to see. But if they can just keep doing this, like this is the flywheel right here. >> Yeah. But the thing is also think about your company in in any sector and your major peer starts to adopt Palunteer for various different use cases. The competitiveness of that peer is increasing massively because they can reduce costs, they can enhance efficiency, whatever the the case is, right? You as as as the as the peer, you need to start adopt AI and you can probably not do it fast enough on your own. So you're probably also reaching out to Palanteer. So it's kind of this reinforcing loop that is really bullish for for Palanteer moving forward. >> And if you don't are not working with Palanteer yet because also don't forget I don't think they can just scale this out to a billion clients right away. Like they also have capacity as well. >> The biggest issue issues. >> Yeah. So what are companies doing that aren't working with Palanteer or the other companies is they're just like hey everyone go spend as many tokens as you possibly can. Yan Van was on the podcast yesterday. John interviewed him. what he asked him straight up. He said, "What are you guys doing in your company right now? Are you, you know, using like an open router type where you're going for different models? Are you cutting back because it's not working? Like what are you doing?" He goes, "Look, we'd love to do that, but we we don't have the the ability to build our own kind of like routing module yet." So, he's like, "Right now, we're token maxing. I'm just telling everyone spend as much as you possibly can." He's like, "We're probably going to do this for two to three more quarters at least because we just we need to keep unlocking new and new things." And so we're just spending everything we can on AI right now because we know it keeps unlocking returns. That's literally what he said. Uh and this is the the founder CEO of of VANC, you know, a huge financial company. And so I think this is what every company's working towards right now if they are not working with a Palanteer and and and other tech companies. So it's um the adoption right now is just insane. And we haven't touched consumer agentic yet. That's still to come at the end of this year, next quarter. So where enterprises are being loaded up into AI, consumers are still just chatting. Wait until Siri AI or maybe whatever Google's going to do comes in and now consumers start going ham on agents and man everything below market. >> And one one way to play this in my opinion is Robin Hood and they had the earnings last week and Flat the CEO came out saying two things. a over 100,000 um agents are already trading on their platform. That is within a couple of weeks of them launching uh that option for customers. And then B and that's the really bullish part. He said that Agentic AI will drive the majority of volume on the platform eventually. I think this is something that investors today in Robin Hood that are mainly focused on how are we in a crypto b bull market or bare market. Are we are we having a lot of people trading on the platform or not? It all doesn't matter because they have 27 million customers and each of and each and every one of them will have many agents on the platform spending stuff trading. This will drive the volume on the platform like crazy. And it's one of the kind of teases that that people are not understanding yet. And it makes me really bullish on the name. >> Guys, I I want to wrap this up soon. It is payday week this week in Milkroad Pro, which Melvin, this is this is kind of your wish answered, but not exactly. Uh every every analyst gets an extra 500 bucks to add to their portfolio at the start of every month. Uh and you guys have been posting in Pro about how you're spending it. We have an AMA tomorrow morning. By the time people have listened to this, it will have happened. But uh with with Melvin and Martin and John where they kind of talk through their choices and kind of discuss uh the their market views. But Vincent, uh you were just kind of pitching us on Robin Hood again on that next phase of AI. How are you are you dumping that 500 bucks into Robin Hood, which I think is still like was like under 90 bucks or under 100 bucks last time I checked. But you tell me what you're thinking. >> Uh no, because Robin Hood is I think the second largest position in the Melro portfolio already. So, so that's fine for me. I'm uh I'm I'm staying in cash for for those 500 bucks. I bought the dip uh last week already on on a couple of names. I think there will be more buying the dip opportunities down the road and I'm waiting for them mainly in in in memory because I want to build a bigger position there. Um, and then once I figured out which of the hyperscalers are the most interesting one from kind of a legacy business perspective and how they benefit from AI, then I will probably start building a a position there. Uh, but yeah, not not 100% clear on that yet. >> Melvin, what are you doing? >> I am going to buy more memory. Um, specifically more S new position is going to be Samsung. So, this is something I wanted to buy for a really long time. Um, but every time I, you know, try to buy it, Cosby always freaking closes, so I can't freaking buy it. Um, or like the timing is like off. So, I'm going to I'm going to buy it later tonight. Um, so yeah, Samsung, I think, is massively undervalued compared to um SKH High. Samsung is actually projected to take the HBM um lead in in share market by next year. Um, and I think there's a their HPM4 is already confirmed. Um, there's we all know there's a multi multi-year shortage in terms of memory. Um, and I think Samsung will greatly benefit from that. And I I think I want a little bit of exposure to Korean market as well. Um, so I think I think Koreans and I align on on our values a little bit. >> Nice. >> Just leverage gambling. >> Funny story. Yeah. Was it yesterday Melvin? Me, Melvin, and Martin were all on a call and we were talking about what we're going to do with Payday and all three of us said Samsung and we're like, "What the heck?" Uh, but anyway, I um I made a buy today, but I'm not going to tell anyone on this podcast what it is. We've given up too much alpha. If you want to see it, go sign up for a dollar and you can see what I bought this morning. >> Your buy uh already fueled one some of the questions that we talked about today on the show. So, if you guys want to see that is just a dollar at the link below, already a few trades today uh and already some thesises on your payday allocations, guys. So, great to see that so far. Uh, and great and and thank you for another great episode, guys. Great to chat. Uh, a nice long one, but there's so much to talk about and I feel like next week, uh, we'll have even more. So, thank you guys. >> Thanks, guys. >> Thank you. >> Want to stay ahead of the biggest technological shift in history? Subscribe now to get insights straight from the sharpest minds in tech and finance. Quickly, you'll note this show is for educational purposes only. Nothing here is financial advice. Investing always carries risk. Never invest more than you can afford to lose. Thanks for tuning in. See you in the next one.