SPEAKER_01 0:00

You are listening to the AI Investor Podcast from 24-7 Wall Street. On today's episode, we're going to look at AI in the news, the hardware trade taking a step back. Meanwhile, software stocks have their day in the sun. There's a big conversation around sovereign data, and we're going to look at where the market might be misreading Broadcom's potential. All that and more is next. Eric, welcome back from your international trip. You went to Investicon, you flew over the big pond. I saw the presentation you put together. Incredible, very impressive. I know our listeners, um, I'm just I'm just gonna tease it a little bit. It's fantastic. Once it's released in the world, you got you got to set aside some time and go go enjoy that. But Eric, how was how was the trip? How was Investicon? How was Emmett? You saw our old boss, uh David Gardner, and some of our old friends. I think I saw that Bill Mann was there.

SPEAKER_00 0:46

Yeah, yeah. And number one, if if anyone here hasn't been able to go to a great investing conference, it's really it's like giving you oxygen, you know. At InvestiCon, I think there's 20 to 30 stock ideas along. You know, one of the wonderful things, Austin, is as you know, between us two, we we have different strategies and and we have a lot of ways that we invest similar. But the beauty of investing is it's an infinite game, and there's there's no one way to win, right? There's there's a lot of ways to invest that are uh specific to how much risk someone wants to have, what industries you're interested in, what kinds of returns you want. So going to a conference like this, you're able to just hear from so many experts and and hear amongst all of, you know, I'm gonna dive really deep into AI. And that's what I did in my presentation. Other people are going to talk about more simple on-the-surface ideas, but ones that, you know, are a little bit more complex around uh quality companies, right? Something like looking for only 15 quality companies you want to hold forever is a simple idea, but it's incredibly complex. You know, Warren Buffett famously would do all the research he did, think about stocks for 90 plus hours a week, and it's all in service of a few ideas a year. So I thought I thought the conference was outstanding. I know most of our listeners are in the United States. So going to something like Investicon, it's a long trek, but I thought it was really wonderful. And Austin, second, I did tell some of the people who were there, I wanted to just give them a shout-out on today's podcast. So Lucas, Tim, Brandon, Bernard, Andreas, Finn, Dennis, David, Mike, Kevin from Belgium. I think I'm probably missing a few people here, but we had more than a dozen people who came up to me and said, I, you know, I'm a huge fan of the podcast. I just wanted to say hi. We had a couple people who came from South Africa just because, you know, they they they wanted to come up. Uh, we we grabbed a beer and chatted afterwards. It was really wonderful. So it was also, you know, a moment. We talked about how we would love for this podcast to be a place that has ideas and hopefully people like the kinds of knowledge that we're sharing, but also something of a community where people can come together around investing in this theme, we believe is going to be the dominant one across the next decade. So I I thought it was really fantastic. But on the other side, I do need to note I am trying to get back. I I I um I had a long day traveling back on Monday. I'm kind of getting back to work. So I think we'll keep this to a little bit tighter of an episode. And and and we're not just saying that we have to do it because we have a hard cutoff, but I'm going to really dive in in this episode to a to a few earnings. And what I think is interesting about the earnings that we're going to talk about, which have all happened in the past few days, is what they say about where the market stands today. So we're going to cover Snowflake, we're going to cover Broadcom, and we're going to a little more lightly cover Credo. But while we're going to talk about these earnings, the lessons we're going to discuss are something that's high level, whether or not you own these companies, and are going to apply to where the market is today and what the market looks like across the next 12 months. So while it's a little bit more compact episode, I think what we're discussing is a really big idea today.

SPEAKER_01 4:11

Thank you for that, Eric. Since you said it's a short episode, I'm going to buckle up for an hour and a half here because that is the way it goes. But you know, there's one name you didn't mention. You know, we get it, you're popular, you know everybody in this space, but you didn't mention Jorge, who was your co-host on our last episode. Um, I loved seeing him. Fantastic episode where you discussed Marvell, really got in the weeds on the technical. So uh fantastic video. If anybody hasn't got a chance to listen to that episode or watch on YouTube, you should. And I mentioned that because we've always wanted this to be a community. And I think Jorge is a wonderful example of somebody who was a listener, actually ended up doing some writing for us, still does on 24-7 Wall Street, and then did a research project for you. So we really do like to bring our listeners in, and it's not just lip service. And as one more example of that, a little bit of a housekeeping item, uh, we are on the hunt for a tax and financial planning expert. So it's one of the topics that I think we know a lot of our readers have questions about on 24-7 Wall Street. So if you have a background in publishing, writing, and a strong interest in taxes or certification, please reach out. You can email me and Eric at editorial at 247wallstreet.com. Uh, we're going to be looking to bring on someone like Jorge, right? A listener who maybe became a writer, um, specifically to cover some of our tax and financial planning content on 24-7 Wall Street. But Eric, people are not here for taxes. People are here for AI investing. So let's talk about what's going on with AI news. We always start with the news, sort of table, you know, set the table, what's going on in the industry. We're just a couple weeks out from Anthropic's IPO, and that's resulting in some, shall we, shall we say, window dressing, gamesmanship, some jockeying for you know, pole position here between open AI and Anthropic. Talk to me about what you're seeing, and I'd love to hear about Claude's uh 5.1 release in the context of Anthropic's upcoming IPO.

SPEAKER_00 5:55

Yeah, and just broadly touching on AI news before we're going to dive into the earnings, I thought it was just worth mentioning. We we are having some releases happening. Uh, as you mentioned, there's gonna be some gamesmanship. Both these companies want to look as strong as possible before their IPO. We've seen some things like cost cutting to get some migration towards open AI models versus anthropic. Uh, I think there's even some talk recently of um meta uh in in some internal town hall saying they wanted to reduce their anthropic usage because it might put anthropic in a little bit worse of a light before it's IPO. So you can see the games that are afoot from these companies. But Anthropic released uh Fable 5.1 and OpenAI is readying the release of their Astra models. Uh, you're seeing Sam Altman, who's the CEO of OpenAI, he's going on a tour. One thing that's really caught my eye from these, which is interesting, is there's some reports. You're always looking at how these models are built as a leading indicator of what the future downstream impacts are going to be for companies in the portfolio. And OpenAI is reportedly using for their Astro models a new technology called recurrent depth, or you call it looped transformers. At the beginning of 2025, uh, we had talked about memory as a top trend. And the reason why was because we had seen how um memory-intensive the next breakthrough in models was going to be, which was reasoning, or we called it chain of thought. And it is just what it sounds like, being able to watch a chain of thought as a model is effectively working its way towards its answer, and it's especially um effective and verifiable domains like coding. So, you know, this reasoning breakthrough is awesome what led to number one, um, a lot of the advancements we saw the next year across coding. And number two, a lot of the increase in the need for memory from this. So, what's happening with loop transformers is it's really interesting. It's it's effectively an efficiency advancement that changes a method for how to reuse layers. And the benefit of this is you can double the size of a model, but this won't 2x the usage of RAM or storage. You know, it's not going to 2x your memory component. It's going to 2x the amount of compute needed instead. So there's a couple downsides to this. It's maybe going to make this reasoning chain a little more opaque. But Austin, the bottom line is this shifts huge parts of the workload from being limited by memory to needing uh a larger focus on compute, which is a uh, you know, would be a downstream win to companies like NVIDIA and Broadcom. So I wanted to talk about this just because we are going to see two companies, OpenAI and um Anthropic as well. They're going to be IPOing soon. We're going to see a lot of new models coming from SpaceX itself. We're going to see Meta making a major push. I saw Meta was up 4% today on the announcement of some new models coming out to the market. And that's going to give Meta a chance to change its narrative around. And what we're going to be watching, in a sense, is what is the architecture of these shifts and how it's going to benefit companies downstream. Because as we'll talk about today as we film this, I think Broadcom's off 6%. We'll talk about that more. But um there are some things that aren't reflected in financials that are flowing downstream to benefit a company like that. So I thought that was a little bit interesting and uh kind of leads a bit into what we'll talk about with earnings today.

SPEAKER_01 9:42

Um, wonderful context there. And yeah, this is also early. So of course, I don't think our listeners should read into this that there's like suddenly the memory trade doesn't make sense anymore, right? These are just developments in the industry. And as we have talked about in the past, the high price of memory and the limited supply causes people to have every motivation on earth to try and find work rounds to make you know use of the memory that they have available. But the mem, you know, memory is still the precious resource in this industry by far, you know, memory and chips broadly. But you had talked about Broadcom selling down. Now, Broadcom is one of your largest positions, it's one of the biggest positions in the portfolio, and also one of the best positioned in this industry, particularly as we've seen like the increase in ASICs and uh the important role that custom chips are playing and how many companies want to build custom chips that are more energy efficient than off-the-shelf, super powerful NVIDIA chips. But me, but we're also, it feels like we've been in like a two or three month cycle of no earnings are good enough, and and the market keeps misreading these things. I know you and I were texting around NVIDIA's earnings. They actually sold off 3% after earnings. They ended up rebounding the next day, but um pre-market, they were off 3% on just some misreads there. So I'm curious what you're seeing out of the most recent earnings that have come down the pike and what story it tells us about the strength of the AI trade or this this seeming migration we're having where some people are, you know, hopping into the software trade.

SPEAKER_00 11:06

Yeah, and I think earnings this week was kind of a tale told in three parts, if you will, that we had three different companies. And this is actually great to frame up NVIDIA. It's going to dovetail off what NVIDIA had said last week, but three companies all delivering earnings and getting three wildly different reactions from the market. So, you know, we have company one, Credo. This was a company that has seen a brutal sell-off after earnings that would normally be considered outstanding. Uh, shares are down, I think, more than 30% across the past week. Company two is Broadcom, and they issued massive 2028 guidance as we had predicted on uh when we had last recommended them. But as of this morning, it looks like the stock is down roughly 6%. And then we have Snowflake, which is up 24% this morning, as I see them. These are all recommendations in the portfolio, I should add, following three different paths. Now, what's interesting here is Credo delivered 115% growth. Broadcom, they're forecasting doubling AI revenue next year and another double in 2020. And we'll talk about how in excess of Wall Street's expectations that was. And then you have Snowflake, their growth is accelerating, but compared to these companies, a relatively meager 36%. So the question is, and what we're going to deconstruct today, why are these companies that are growing at such outsized rates seeing such brutal short-term reactions? And what does that say about the state of kind of the AI market in general? And what does it say about what qualities investors are looking for right now? Because Austin, if we go and look at present valuations for these companies, this is what gets really important to look at. Because a lot of people will just look at a trailing P and say, well, Broadcom's very expensive.

SPEAKER_01 13:07

Well, but when you have high growth rates, your trailing price to earnings multiple is becomes far less relevant, right?

SPEAKER_00 13:15

And press is very quickly. Right.

SPEAKER_01 13:17

And people on this podcast will have heard you talk about forward price to earnings multiple quite a bit. And that's why, right? If you're a steady eddy uh consumer goods staple bank stock, and well, not bank stocks for different reasons where earnings is a different metric, but if you're a steadier grower, the price to earnings multiple can make a lot more sense because what you grow at today is somewhat, you know, reflects with what you'll grow at in the future, consumer goods in particular. But with tech, if you're growing 100% a year, you know, the current valuation will almost certainly, especially in this industry in this moment right now, look unreasonable, which is why you always look to the forward multiple.

SPEAKER_00 13:53

And so for Credo, they are now at 14.6 calendar 2020 earnings, once again, coming off 115% growth in the latest quarter, and a guidance for the fiscal year they're currently in of 80% growth, also predicting a rapid acceleration into the back half of this year, which is gonna be November to April. Broadcom, they're now trading for 10.3 times their fiscal 2028 earnings as of this morning. And that's with when we get to 2028, a hundred percent growth. So 10 times earnings, 100% growth. Snowflake, as a comp to these two companies, is now trading at 93 times 2028 earnings. Their growth is accelerating, but as of right now, Wall Street has, I think, 2028, like 28% growth. So the the company 93x at 28% growth, you know, the real expectations are gonna be higher. We've got 10 times earnings at 100% forecasted growth, and we have another company at 14.6, um, growing 80% this year. So the question, right, is why? And it all comes back to the big picture, which, as you mentioned earlier, the market's essentially flat since early June. Coming through May, massive run-up. We had a really sharp correction right at the beginning of June. And right past the point we have that correction, we've essentially been flat. In that period, AI infrastructure, the companies building the actual chips selling into data centers have been out of favor. Software has seen an incredibly positive um amount of price action with people flooding into that area. So, what is the market concerned about? Well, I think number one, peak earnings. We've talked over and over again on this show that 2027 is largely accounted for. When a company, when we look at estimates for 2027, those are largely locked in. Could companies still disappoint? Yeah. Mostly because you would have areas like a lack of power to be able to build. But that also probably means you push those earnings into 2028, right? Which is kind of we call it the zone of uncertainty. It's when a lot less of the demand is as certain as it would be in 2027. The second thing, Austin, that people are concerned about is that the economics when you reach peak earnings get worse. So you maybe have less growth moving beyond that, but you're not going to be able to command the margins you see today because we've seen memory go up past 80% gross margins. But a lot of these other companies too, they're commanding substantially better gross margins than they did prior to the AI boom. A company like Credo is up to, I believe, 68% gross margins. Um, uh Broadcom has been able to claim much higher margins, like an NVIDIA at 75% margins. If we go to the other side where demand stalls, well, demand may stall, but earnings may fall because NVIDIA may be only able to command 65% gross margins, maybe less when we don't have this market where people are taking whatever available compute that we have. And the counter to this is software is running. Why is that? Well, because we're suddenly entering a time that the market consensus is that there is no limit to what growth could be. Um, we're increasingly seeing from the market itself, and I want to clarify where where I'm talking versus the market, a long period of rising revenue. You look at a company like Snowflake, we'll talk more about this. They're a consumption-based company. If the belief is that usage of AI is exploding and going to continue exploding, that could potentially give them a ramp for a decade or more. That allows the market to price in kind of this unknown future of benefits, right? This is this is kind of where a lot of the hardware, the infrastructure plays were in 2024, right? You knew this was going to be big, you knew they were going to grow, but you weren't able to see an endpoint. Once there's an endpoint for the stocks, the market generally reprices down. Another thing that is kind of moving away, the threat of disruption, whether fair or not, is kind of, you know, the market's the market's not pricing this in. And Austin, the the analog we get to, and I've talked about this so many times on the show, is smartphones. In the middle of last decade, smartphone growth, which we expected to be over a decade, it was so fast and so furious of uh that it essentially all happened, what was expected to happen from 2010 to 2020, ended up mostly happening in like 2012 to 2015. And this left a period where Apple was trading at a PE of around 10, even though it was the dominant company. And the question is, are they going to be able to deliver? Meanwhile, its competitor, Google, who, you know, doesn't get their revenue from selling phones, but mostly on consumption, right? Consumption of search was able to just keep growing at a steady rate. And they never saw this existential dread that um Apple did. The way it worked out was Apple was a generational buy at 10 times PE because it turned out they were able to grow because the trend was so solid that they maybe, maybe the total units didn't grow, but but the market continued maturing and growing. The question is, are we going to see the situation play out again? The analog for today being Apple is more equivalent to an NVIDIA. Apple would be equivalent to a Broadcom. So, you know, that that is the big question as we kind of break down each of these earnings one by one. So, Austin, I'll talk about Broadcom first, but I also want to pause here just in case, you know, there's anything I need to slow down or any questions that kind of need double clicking into.

SPEAKER_01 20:20

Um, not slow down, a fantastic explanation. And I want to I I want to stay on this point though, because this is sort of it, this reflects the bear case here, which you know we will sort of address. But let's try and steel man it a little bit, right? Let's let's try and give the bear case here some credit and then unpack where we would stand on it. So the argument is that cyclicality still exists, right? And these are cyclic, these are historically cyclical sectors, particularly memory, where we've seen cyclicality in the 80s and in the 90s, and people look at those stock charts and know what happened. So the bear case here is that cyclicality still exists. It's just at a fundamentally different magnitude now, right? The the industry has stepped up because of all this demand, and now we're going to see cyclicality. So you shouldn't be buying at these peak earnings, which is when it looks cheap. Um, and there's good reason to believe that, right? The as the saying in economics goes, the solution to high prices, or in this case with memory and optics, high margins, is high prices, right? Because when you have these persistently high prices and high margins, it attracts investment, it attracts competition, supply corrects, and then pricing corrects, and nobody wants to be left holding the bag. So that's kind of the steel man argument, right? We're we're going to see cyclicality in these sectors. It's just now at a fundamentally higher level. So people are trying to get out before, you know, what while they believe they're at peak earnings. But then I have to ask myself uh, is a is what we're seeing now with artificial intelligence like those prior cycles, which were hardware based, where people would buy memory or buy hardware. It has some useful life of another couple of years, and then they upgrade, and that result. And these like prior, you know, cyclicality that we you see in industrials, you see in technology. Or is AI a fundamentally different industry and layer where you actually just get persistent growing demand? Right. You don't have this big capex cycle where you buy a bunch of stuff, you amortize it over five years, and then you upgrade your servers, your your caterpillar equipment, whatever. And I think it's fundamentally different because what we've seen is AI usage continues to go up. It's not going up in some big temporary boom and then it falls off, especially for inference. Inference is a daily task. And when you add more agents, it compounds on itself. So what we're seeing now is this is a persistent growing steady demand, not this huge pull-forward demand where people make an investment and then that investment's good for a couple of years. This is a fundamentally different experience. So that's the argument I would make going against the steel man. I'm not sure how you feel about that or what you would say to that in turn, but that's why the the pri the argument about buying these things at peak earnings just doesn't map for me right now because this is a much different capex cycle. This is a daily, persistent, growing capex cycle.

SPEAKER_00 23:08

Yeah, it's I I think the way you framed it there is, you know, that's the game. That is that is the trillion dollar question at hand today. And it will be the single question that probably defines investing success across the next decade. One thing that had been asked about in Investicon, I shared stock ideas, but they had said, what if you needed to buy and hold three stocks for the next 10 years, what would they be? And in my mind, that's largely a question of, well, if I believe the next 10 years is framed by an explosion of nearly infinite intelligence at cheap prices, what are the companies that I feel like have the highest probability of maintaining their economic modes through that time? And I said, I would, I would probably select something like an Amazon because of their dominance and being able to serve the intelligence and what they can add with their custom units and how they benefit from trends like robotics and their core business. I said Taiwan Semiconductor because I think they would be the company that ultimately is one of the hardest to disrupt and is ultimately serving the means for providing that intelligence. And I said probably something like a Quanta Services, because if we are arriving at this future, we we reach a point of needing a significantly larger investment across our entire grid to reorient our economy towards this. So I think I, of course, you know, we've talked so much about Austin, and we continue to land on this concept that when you go back to 2010 through 2020, so much of the growth of AI was it was fundamentally in many ways costless once once you had created the game was essentially getting users on a platform because once you had them on that, there's zero marginal cost and you are able to collect significant revenue. If the differentiator of this next decade plus is going to be intelligence, does that create a meaningful step change and the value equation where it used to only be collected by the big companies who had consumers on their platforms towards intelligence? And if it is, we will likely see a dramatically different cycle for these for these hardware plays. So I would lean more towards kind of that steel man situation you had built there. But I would also say as well, you know, environments like this are why we diversify, right? The AI portfolio is not supposed to be 100% of what someone would invest. This is this is a specific theme. However, this theme is increasingly a large percent of the market itself, right? So this is why we've tried including um a stock like Amazon in the past and buying in an outside way. This is why we have bought infrastructure, but it's also why we're increasingly diversifying um into different areas of infrastructure, such as how we have had people like John Rotanti on talking more about industrials recently. And it's also why we have tried buying these best of breed software stocks. So on a day like today, when we have Broadcom down, well, we also own Snowflake. So we're gonna be able to mitigate this to some extent, that we can see a couple of different dispersions of value and we want the best of breed, regardless of what the market's going to decide when. So um, yeah, I would say to your point though, uh, and we'll we'll talk more about this with Broadcom. It does seem that I believe when the record will be read in five years, this current moment, where the best companies, the most dominant in terms of competitive advantage and revenue growth in the market, like NVIDIA and Broadcom, are trading, for close to those Apple rates, I believe we will have closer to a repeat of what happened to Apple last decade than we will have these companies getting wiped out.

SPEAKER_01 27:20

Well, let's talk about those two names and the divergence here. No, no story right now illustrates it better than the Broadcom versus Snowflake story. Broadcom, all in all, is strong print. You know, it it technically is light on guidance, but at a at a completely irrelevant level, in my opinion. Meanwhile, they're down, I think, five and a half percent as we're filming this. You've got Snowflake up 22%. Now they're both in the portfolio. We like both companies, but I do think that this illustrates the hardware software dynamic pretty well. So talk to me about what you're seeing in both of those earnings and what you believe it shows about the AI industry right now and where people are putting capital. Happy to be invested in both, but important for us to understand where other investors are because it will actually create opportunities, right? Maybe the opportunity in Snowflake is now smaller than it was when you initially bought it, and the opportunity in Broadcom is better.

SPEAKER_00 28:12

Yeah, and you mentioned where people are focusing this morning is Broadcom itself. A little light guidance for next quarter and for next year. And I'll I'll give the specifics on those. So for next quarter, they had guided towards 30, I think it was 34.8 billion in revenue. Wall Street was somewhere around 35 billion. The stock initially dropped five per 5 or 6% on that guidance. And what we've said over and over again about Broadcom, and we said this when we issued that recent recommendation, was what's going to happen is we're going to have a coming conference call. And the CEO, Hawk Tan, he's going to come on and he's going to issue a whopper of 2028 guidance. He's clearly set up for this. It's going to be significantly in excess of Wall Street. And once that happens, people are going to see that where Broadcom is trading today is essentially at 10 times their 2028 earnings, which will likely be too low relative to this company's, you know, their opportunities. So what Broadcom said was AI revenues for this year are going to be about 58 billion. They said next year it's going to be 115 billion. And then the year after, it's going to be 230 billion. They said they have clear line of sight. They could actually do more, but that's what they have today in supply. So when we look at 2028, before they report yesterday, Wall Street believed they would book about 180 billion in AI revenue in 2028. So that 20, 30 billion figure, that's $50 billion over what Wall Street expected. They also said they were going to do $30 plus in EPS in fiscal 2028. If you change that to calendar, the January through December period, that's closer to $35 per share, which again, it's trading for less than $350 as we film. So that gets below 10 times earnings for 2028. And the consensus for EPS in uh fiscal 2028, we said that was gonna be over $30 was their guidance, is over $26. So Austin, again, this is just generally a huge beat. And the thing about companies issuing guidance for the future, it's almost always conservative, right? They will they will say what they have they feel in the bag. And that doesn't mean a hundred percent chance of beating, but it is normally something they have a high confidence in beating. So the reality is they're probably going to deliver better than these numbers. So this is this is all very interesting. We we almost exactly predicted the situation, right? We said on the next earnings call they would beat earnings dramatically, or they would guide, I should say, 2028. We we specifically said 2028, um, it would be in excess of what Wall Street expected. And our expectation was this would lead to a re-rang of the stock. We're seeing, we're seeing a difference from Wall Street here, where where they delivered this blowout guidance, and Wall Street's essentially meeting it with a yawn. So the question is, what is what is causing this, right? This is a very similar situation to NVIDIA, as you brought up earlier. NVIDIA said next year we're going to deliver 70% growth. Wall Street was expecting 44%, and initially this was met with a yawn. So a few things, Austin. I I would say, why are they down today? Uh, well, number one, their margins are going to begin compressing a little bit. So this is that story I've talked about that people fear a part of the story of reaching peak earnings is margin compression. The the thing with more uh sorry, I should say Broadcom is as they scale XPUs, which is their processors, it's known that their margins are going to compress. Second, they're losing some market share with Google. We talked about this as one of the risks when re-recommending the company that recent podcast. Well, where what is going on here? Well, they're essentially selling Google's processors, TPUs, to Anthropic, which is going to become their largest customer. So there is some mix shift here. Um, this the other thing I would say related to Marvell or sorry, Broadcom, and why they're dropping right now, is that um their guidance for next year is a little bit light. So I think we have Wall Street saying, well, what we know is the certain thing, which is 2027, is lighter and what has less certainty in 2028 is beating. But we're going to look at the kind of sure thing and we're going to focus on that. You'll notice that their reaction, they were initially down 5 or 6%. They mentioned the 2028 number, it bounced back to green, and then the morning after it bounced back to the 5 or 6% level. So it was essentially Wall Street kind of discounting this future. Um, so let's let's just set talk about the bottom line on this for Broadcom. I think the bottom line is this is a company. I'm surprised by the reaction here, but a company that continues to be relatively dominant, continues to be diversifying its customer base, and I believe it is cheap relative to their opportunities in the years ahead. This is a company that, again, if they miss, if they miss some numbers, it's probably because there's not enough available power. There's not enough other things, which is going to push revenue into later periods. Um, if there is one fear about them, it is largely that they're having to use kind of financing techniques um to be able to facilitate some of these sales to companies like OpenAI and Tropic. But I would say if if you're someone who isn't interested in a Broadcom at these levels, well, NVIDIA is trading for similar rates, right? In 2028. So I would just end this by saying if Broadcom isn't the right stock for you, I do think something like NVIDIA might be a fill-in. And these are companies that are gonna trade under relatively similar dynamics. But at the end of the day, it does seem like we are at a moment that they're trading for something that's gonna be looked back at as a mistake unless we see a very strong unwind where where it does end up 2020 as kind of a peak. And as we talked about earlier, there are reasons we don't believe this will be the case.

SPEAKER_01 34:49

How does that map to something like Credo, still on the hardware side of the trade, had a had a brutal day, revenue beat of you know almost uh $500 million, or revenue of almost $500 million, not a beat of almost $500 million. Um, all in all, really impressive, yet the company sells off hard. Is this just an extension of the hardware um jitters, right? And people just trying to get out before that peak, what they believe to be the peak 2028 build-out?

SPEAKER_00 35:15

Yeah, I think it's very related in that credo, their guidance for the next quarter was maybe a little um disappointing relative to elevated expectations. And they're talking about an acceleration happening um in the back half of their fiscal year. And again, there's there's just with the fears around what's going on with these infrastructure stocks, people are ready to get out of them, right? So, as I've talked about a lot, the risk right now is if you're holding, if you are holding a company that is trading for 25 times or 30 times their 2028 revenue, and it's it's it's a hardware company, you just need to be aware of the risk that's going to re-rate down. But the lower and lower these companies go towards this, you know, 10 times level, the more that's decreasing the relative risk that you're holding. So I think Credo is gonna be one that's really interesting. It's it's a victim to the same exact story as Broadcom. You know, we're we're talking about individual stocks here, Austin, but 75% of what's going on right now is this broader market narrative that different stocks fall into the buckets for. And for a company like Credo, you know, in the near term, they're gonna continue selling their bread and butter products for the majority of their revenue, which is kind of these cables that we've talked about that make copper more efficient. But they've got a lot of diversification and they've got some inroads into what Jorge talked about last week, which is kind of this revenue opportunity from a lot of new memory technology. So I think that's one we'll continue looking at. And the big question is just how much do we want to allocate towards hardware ideas compared to maybe some software ideas and continuing diversifying the portfolio ahead.

SPEAKER_01 37:07

Let's contrast what we're seeing on the hardware side with what we're seeing on the software side. So Snowflake's up 24%, or maybe about 20% at the time of this filming. It's also dragging fellow uh you know data sovereignty uh play Palantir up about 7% today. So this idea, right, that you your data is yours, you want to have a company like Snowflake or Palantir sit in between you and the model provider so they're not training on your data. This is something Alex Carp has talked about. Is the growth that we're seeing here in Snowflake up over 20% today, Palantir up 7%, the result of that thesis? Is it people rotating from hardware to software as we've talked about? Because one of the problems that I have is if hey, I'm I'm glad any position in the portfolio is doing well. But I like Gavin Baker talked about a while ago, how can it be mutually true that Snowflake and Palantir are well positioned for this multi-year run of like token management and agentic operations? And Broadcom and Credo are at this like cyclical build high. They can't both mutually be true, right? That like because one feeds the other. So what's going on with Snowflake's earnings? Is this data is this data sovereignty, the Alex Carp um argument? Is it just a really, really good print that nobody expected? Is it the rotation from hardware into software? Some combination of all of them.

SPEAKER_00 38:28

Yeah, and I'll try and keep this uh quick. Uh as I noted earlier, we do have a hard stop here. So I'll just say what Snowflake. They they beat by about 5%. Um half of their revenue acceleration is due to AI. The market really liked that. Their acceleration has increased from 25.5% Q1 2026 to 36% in the most recent quarter. When we had first done our software sleeve at the beginning of 2026, we had talked about, we wanted companies we felt good about the ability for AI to provide this accelerant because it was going to cause a re-rating. And that is what we're seeing. Um, you know, when we compare to something like a snowflake to some of these hardware infrastructure companies, what else does Wall Street like? Well, number one, they're very diversified. Instead of just a handful of customers providing their revenue, like you'll see often from a Credo or a Broadcom, it's going to be extremely broad-based. They're levered to consumption growth, which is a big thing. And you talked about sovereign AI. Well, that is driving massive usage of their AI tools. Cortex Code, uh, which I think they call Coco now, uh, that jumped 28% sequentially. Their AI intelligence tool jumped 12% uh sequentially. So, Austin, I think, I think the bottom line, I'll just end the podcast here, is we can only go from where we are at this moment. I think on the hardware side, we have a lot of stocks now with some significant revenue expectations, priced to the cheapest levels we've seen. And we need to decide what fits in the portfolio, what's best of breed, what's positioned. And we'll likely make some recommendations there when we look at how compelling many of these levels are. When we look at software, as you mentioned earlier, a snowflake is less attractive today than it was because of how furiously it's risen up. But we made a recommendation, something like Service Now, that benefits from many of the same trends. And you do have a lot of SaaS companies that will see the same tailwinds as Snowflake, but maybe they're not trading at a hundred times 2028 earnings. And that's what we need to sort through. I think we're going to be sorting through what are the infrastructure deals, what do we want to selectively add to so we don't get too concentrated, but we take advantage of the opportunity on software. We want to look for some more companies that diversify us there, but maybe aren't trading for quite as wild of rates, but still have the same tailwinds. I think that's going to be one of the key themes as we enter kind of the last quarter of the year.

SPEAKER_01 41:07

Wonderful. Well, thank you for that uh wrap-up. As you mentioned, we do have a hard stop. So let's just run furiously into it. Eric, thank you for your thoughts and commentary. I recommend all of our listeners go check out the last episode with Jorge. Thank you very much for the good work you did at InvestiCon representing 24-7 Wall Street and the AI opportunity broadly. And just a final note to our listeners please comment, share any feedback on Spotify, YouTube. And if you are interested in the tax and personal finance world that we had mentioned earlier, please email us at editorial at 247wallstreet.com. Eric, until next time, thanks a bunch. The AI Investor Podcast is for educational purposes only and should not be considered investment advice.