Dialogue. S&P 500 Valuation, Intuit & CSU Earnings, Sea and Meli Update

(upbeat music) Welcome to the Synopsis, a business and investing podcast. I am joined by none other than the Drew Cohen of Drew Cohen money. How are we doing over there, Drew? I know again, you are on another four week sockets to replace me, but backed by popular demand. And by that, I mean, one comment who said they missed me. I mean, it's overwhelming. I'm back. So, you know, I get where you are coming from, but it can't be done Drew. And it's true, it's true. We did get that one comment coming through. And since Alex isn't on Twitter, it was very quick to message that to him and it usually ignores my text, but, you know, for some reason, he replied to down pretty quickly. I mean, nothing makes me stop like I'm doing to, you know, recognize my fan. So I appreciate you out there. That's right, this is Daniela Ron. He's got a big fan base. And today we have, again, I keep, I was just talking to Drew offline about this where I was like, Drew, let's do, he had such a great, I wanna call it wealth management, investment management video that just laid everything out so well and I was like, let's do that. And he's like, no, why don't we talk about software companies and their impact on AI for the fourth time in a row? He goes, the people love it. And I go, here we are again. He is pigeon holding me into talking about Intuit, Axon, Consolation Software, everything that is gonna bring up our AI software discussion and what side of the fence we're on and whether we're gonna be wrong or not. But here we are again. So let's do it. The people that just want the hits, you know, you try to push out new content and they're not that interested. They just want you to play that same old song. Yeah, play the hits, I get that. No, I'm with you. And now the hits keep evolving, they keep getting interesting. You know, before we get into that though, I thought you put out a pretty interesting video which some could argue. It was a macro video. I don't know. It was an interesting video about you talking about the S&P 500 concentration kind of to where historically that's been kind of, you know, the average multiple it's trading at, even Oswath puts in his kind of implied equity risk premium. A bunch of indicators kind of demonstrating that obviously earnings need to grow pretty quickly from here or historically, you know, these companies are at all time high multiples and things of that nature. Of course, this time could be different where in the AI world, and these companies could be structurally better than they've ever been historically, which of course was the argument, you know, going back to kind of the big tech and the S&P 500's historic multiple being higher for a very long period of time, because you know, arguably, I think Google does deserve a higher multiple than probably the old world companies that were in the top, you know, 10 of the S&P in the early 2000s, late '90s, obviously differently structured companies. So there's an argument to be made there, but Drew, what kind of was your takeaway from that, you know, research you did and that post you put out? - Yeah, so one kind of key point is that I don't believe the index concentration in and of itself is the key risk. I think it's the corresponding high valuations that tend to come alongside that concentration, that's the risk. And so if you're looking at the top 10 companies, they're, you know, 40% of the S&P 500 roughly speaking and that's almost double what it is historically. But the reason why that is is because they have all grown in valuation and market caps at the same time to then become a larger portion of the market cap weighted index. And so if you wanna ask what is the real risk, it's in my opinion less of the concentration and it of itself and more the actual fact that valuations are elevated versus history. We're looking at, you know, a four times 21 times multiple for the S&P 500 and that's about one and a half standard deviations where it is historically. And there's all sorts of different data on what that means for future expected returns. You know, there's one JP Morgan report that says the return is usually under 5% looking for it for the next five years or 10 years at that multiple. And so that's less than, you know, the historical 9% percent that we usually talk about. And there's some, you know, figures where it's even much lower. And so that's kind of the real risk, the concentration risk, yeah, it's true when you're buying S&P 500, you know, you wanna buy this big diversified basket of stocks. And okay, you kinda didn't think you were really buying 40% into 10 stocks. Having said that, these businesses are more diversified than most businesses are, right? You know, you pick Google for example, you have GCP, you have search, you have Waymo, you have YouTube, that's a lot of different kind of business engines, it's not really just one end market that is the risk there. Even if, you know, you pick Tesla, well, maybe not the best example 'cause they're mostly producing cars and less of them right now. But, you know, you could say, - No, Drew, what are you talking about? They're not a car company, they're robotics company. - And they're robotics. - That Tesla robot is gonna be in your house. And you're gonna be, you know, just thinking about the days you thought they were a car company. - And an energy company? - And I hope I'm right for all Tesla investors. I hope I'm right. - Yeah, but, you know, AWS and Amazon retail. So anyway, you get the point, these are more diversified businesses than most kind of historical businesses were. So that does give a little bit of a different kind of layer of geography, even if it is the individual business. But the real point and take away here is the actual valuation multiples. And in the last video I did, you know, I kind of talked about a lot of them. I was actually surprised how low some of them were, particularly in video, it's only at 26 times forward earnings. Google though was 31, Apple 32, Microsoft 24, Amazon 32, Broadcom 31, meta 19, Berkshire halfway 20, and Tesla drumroll please 180. So the real risk, the point here is it's the valuation multiples and not all of those are high and some of them can justify it with higher growth, which is another kind of a conclusion. The video is that the S&P is growing earnings at a faster rate than it has historically as well. And so that could help sustain, you know, a higher valuation level. And so if you were concerned about this, kind of one of the things I talked about was you could own an equal weight index. And that over certain periods of time, especially when the market weight index is doing better, it tends to outperform. And so that's just one potential option if you are kind of concerned about this. - Yeah, I think it was a good overview. And what I kind of liked as your segue or one of your conclusions from this analysis, which I think is true really in any period in market history is that of course the argument for owning the S&P is that a very small percentage of the companies within the S&P derive all the outsize or excess return over bonds, right? So not owning those is detrimental. If you don't own, I think it's 4% of companies derive like 96% of excess return over bonds for the S&P. So-- - Yeah, that's your best immersed research. - Best immersed research, yeah, thanks for sending me their address. - 5.5% I'm just pronounced that, yeah. - I like it, we'll take it. It wasn't a fake source, I swear. And so again, if you don't own those companies, then you're really taking on all this equity risk and you're probably, you could have a very high potential of missing those. But the flip side of that is it's very difficult to determine, well, what is the forward return of the S&P going to be for the next 10 years? I think it's a much more, it's an easier question to answer is, floor and decor gonna hit these X number of stores, maybe not because it's been a little shaky for them. But of course, over a basket of maybe 20 to 25 companies that you know exceptionally well that you can underwrite their return, what is the expected value in your confidence in that portfolio versus these 500 companies? And again, over time, obviously, in this period, if you're in 2026 and you didn't own the memory chip stocks, like you got crushed, it is what it is. I mean, if you don't own oil and gas and March or April or whatever, monthly, probably got crushed too. So the point is over short term, very difficult to, I think I'll perform the S&P, but if you are confident in the basket of companies you have, I think in a long run, you have more confidence in what I think the expected return of the portfolio will be, which you're kind of at the mercy of the S&P, which is I think a much more difficult return to project. Yeah, and it's interesting though, because I think if you're looking at the big constituents of the S&P 500 and how much they're growing, you can make the argument some of them are expensive, but taking the whole 10 as a whole, it's not necessarily egregiously expensive, not by a long shot like it was during the last tech bubble, but we are still seeing a lot of bubble behavior pop up in the market and some of that's in the index, but it's not, at least as far as I could tell, all concentrated in the top 10 stocks, which I think could be kind of one of the differences. I mean, we'll see whether or not Nvidia ends up being cyclical. I would imagine on a long enough time frame, you do see some cyclicality in that business, but at least as it stands today, it doesn't seem to be to the same extent, kind of the levels you were saying during tech bubble with Cisco and all of that. And the other thing that I wanted to point out again and really emphasize was that earnings estimates for the S&P 500 are 11 to 13%, versus 67% historically. So that is faster, and then it is a question how long can they really maintain that? There are real productivity gains from AI, a lot of companies seem to be laying off employees that could lead to more profitability, and a lot of these gains are kind of concentrated in these big companies. And so where that all shakes out, I kind of feel like I don't want to make a very strong opinion on it either way, and say I'm not going to own any sort of market cap weighted index as a result of this, but I do think it makes sense to have a balance of the two. And it does also make sense to temper your expectations for what returns could be here in the future. And the issue too is with I guess the equal weighted index is that you're not getting that benefit of, there's sometimes being these small companies or relatively small in the S&P 500 that go on these crazy tears, and you'll miss that in the equal weighted because the S&P market cap weighted, it does have a little bit of kind of a momentum play and that the more of stock goes up, the more of it it will own. And so it's kind of helped in some periods and sometimes the companies as we can see can actually grow into their valuations as we saw within video. I don't know how long ago it was, but in video it was trading at a very, very high double digit multiple not so long ago, I don't know 80 times something like that. And you could see it kind of grew into that. That is not something I personally would have ever been comfortable assuming that level of growth and understanding that. But as a market cap investor in the S&P 500, you still get some benefit as a result of all of that. It just is kind of covering your bases. But to go back to your point, I agree it makes sense too when you can actually look at what the assumptions are in an individual stock because I don't really know how long the entire S&P 500 is going to grow earnings at the rate it's growing or whether or not these valuation levels are over undervalued. It's too hard to say because there's too many components. Yeah. And I go back to, I think 2021 is a good example of this. And I don't want to insinuate this as a bubble or anything of that nature because I think that a lot of the software SaaS companies, I'm thinking the snowflakes, the crowd strikes, all these software companies that had a huge fervor around them when interest rates went up and these are companies that were growth companies. So you had to kind of reset that discount rate associated with them and they kind of took a big hit. A lot of people were kind of considering, oh, this was the SPAC software bubble IPO frenzy. And again, it was kind of concentrated to that portion of the market. The broader S&P was actually pretty reasonably value, you know, had a pretty reasonable valuation at that time. So I feel we're entering into another kind of period and again, not saying AI is a bubble. I think like you mentioned, NVIDIA investors, I mean, you know, when it was 80 times earnings, you could be, oh, that's way too high. And again, they grew into it. So, you know, each investor has to make their own call. Obviously, there's a compelling narrative to the memory stocks, AI companies, SpaceX, at one and a half trillion. Somebody thinks there's a compelling narrative there. Yeah, no, but I will say I do think there's a lot of, I'll say, bubble-like behavior in the market, memory in particular. There, anytime you see a stock go up, what is it, 1,000 percent? We're more than that. And the course of 18 months, something like that. And I could understand, you know, the demand for memory. And this is a bottleneck. And there's only a couple players and it'll take them a long time to build out that capacity. But remember, if you are a long-term investor, you are discounting the entire cash flows of the entire life of the business, not just the next couple years. And so, whenever I hear people talk about this, they're always just talking about the next couple years. It's very hard to build this out and all of that. And right now, it seems like people are very focused on that. But if you're looking out decades, I don't think most investors would actually want to own this business outright and be forced to hold it for the next 20 years. And have their entire returns just be off the cash flows and could generate at some of these valuations. Because competition will exist. You can't charge huge premiums for your products indefinitely and capacity will eventually build out. And I do think that we will see that these memory companies are going to be cyclical and this time is not different. That doesn't mean it can't take two years, three years to play out. But the market can decide whether or not to look ahead to that. You know, it could all set and decide in a couple of months from now, oh, we're actually going to start to worry about what earnings are going to be a couple years out because we're seeing some signs that capacity is starting to expand. Yeah. And my whole thing is is I don't like calling, I don't like being a naysayer. I don't like being a Debbie Downer because I think markets historically, it's very difficult to be a successful naysayer. I think that my take on it is each company has its risks and takes with it at a certain valuation. And you know, again, very high valuations mean that very large growth and very long-term sustainable advantages and people can make that call. It doesn't mean I'm going to make that call, but again, I wouldn't be betting against any of this. I feel like there's the perpetual bears who just sit in their corner and everyone kind of sneers at them while they make money and the Tesla famous short shorts of this valuations crazy. And again, Elon Musk continues to do the impossible. So I know it took a little jab at the $1.5 trillion SpaceX valuation, but I mean, he's about to be the world's first trillionaire and continues to do things that I don't know, defy what anyone thinks is feasible. So there's some premium to that. And people are obviously willing to pay that, which again, I don't think it's completely irrational behavior. It's just a personal decision, I suppose, about how you want to allocate capital. I'd necessarily, you know, don't want it that on Elon Musk doing, you know, data centers in space, but he might execute on it and whom I'd say he won't. I just think there's, you know, other places I'd like to allocate capital, but I don't like calling this a bubble. It's all coming crashing down and everyone, the music's going to stop. I don't know. I mean, I don't know that to be true. Yeah, I think that's a fair take and I think if, you know, we were looking at Nvidia a couple of years ago, I never said, you know, oh, that's a bubble or something like that. I just knew I didn't understand it. So I didn't participate in that. And I think that that's, you know, probably the most intelligent thing to do is not participate in things that you don't understand and stay clear of them. And, you know, my comments on memory, this is just based off of history. I've seen that these are cyclical businesses. And usually when you're betting against what happens in history, it's not the best. But, you know, more power to you if it ends up that way. It doesn't hurt me if these do end up doing another 5x from here, whatever it is. And Micron is, you know, multi-trilling dollar company. That's okay with me. I don't want to make that bet and those assumptions to me seem a little lofty. But I hear you that, you know, a lot of times, especially if you're looking at history, people who try to call tops or call bubbles, usually it doesn't age so well. Having said that, I do feel like as an investor, you do have to have some sense of what's happening in the market. And when you do see some of these behaviors and very, very fast price move-ups and you're trying to understand how to allocate capital, I think it's context you should take into account. And no, I think to your point what I'm trying to say is there are, I look at Jeremy Grantham, you know, has a permabare. He's kind of like a, I don't, he's probably rebrand himself as a quality kind of investor. But I think, I don't think so. You don't think so. He doesn't really say he's a deep value guy, but, you know, I don't know. He obviously talks like that. I think that people get really bitter, especially when you can have someone who's like doing quote unquote, deep value investing and doing these, you know, five, six times earnings companies and they're just sitting there and just getting there, you know, kind of, just kind of getting rocked year in a year out. And they just look at everyone else like with a bitter, bitter lens and they're just like these idiots don't know what they're doing. And like, yeah, well, they made a lot of money. Like, whatever you've done. You know, so I don't like, I don't really like having that narrative of, oh, these people are all going to lose their shirts. I think that you just have to be confident in who you are, which is if you kind of are taking a more, I don't know, measured approach to a risk return scenario, then yeah, you're probably not going to hit a thousand X, you know, a thousand percent micron or you might not hit a core weave, but you can just got to be okay with that. And if you're not okay with that and you're kind of, you don't need to like poke at other people who got it right. I mean, that's kind of my take on it. I'm not a bitter person when it comes to people succeeding in the market. However, that may come. Yeah, no, I agree with that sentiment because I think the problem is it also colors your investing. And if you're, I mean, a lot of the times that I don't know, the people are buying of something times five and six times earnings. There's probably a reason why it should be trading at that. It does still happen that there's like undervalued companies and just no one cares about them and they stay undervalued. But that's also, you know, the reason why Buffett originally moved from, you know, more to value investing to quality investing. Because if you're buying a mediocre business just because it's cheap, time is against you because you need it to very quickly realize to its market value and order for you to sell it and make your return. Otherwise, if you're stuck in it, you're stuck in a business that's investing company at a low ROIC. And so the longer it takes in order to realize that return, your washer investment is going to be. So there's a speculative element in buying a cheap stock too, which I think most people miss. And I don't want to characterize exactly the way he invests because I don't really know. But I will say, you know, when you have a stock that's, you know, a well-known cyclical business that goes from 70 billion to 750 billion in about a year, that's a cause for a little bit more caution, I would say. And you wonder then to what other areas of the market are kind of exhibiting the same behavior and whether or not there could be any knock-up effects too because it's not just, you know, some speculative, you know, pockets in the market. These are actual cash flows that are moving and could have downstream impacts on other businesses. And so you just kind of want to understand all of that. But I, you know, if you called AI a couple years ago and you were invested properly, you know, more power to you, it doesn't hurt me. And I know in a million years, I never would have bought a micron at 60 dollars a share saying that I think this is going to be a 10X. There's no version of history where I would have been able to think that that's what would have happened. So I'm fine with that. And there's lots of different ways you can make money. So you just got to be, you just got to be confident until you are, you know, it's easily, I think there's two kind of ways people react to this. And I'm trying to do a third, which is you're either very kind of FOMO and then you start riding the wave. I think that's a dangerous thing to do, which is if you don't understand these things and you went about micron or nevitia, you know, eight years ago, I don't know why you would, all the, and again, maybe there's incremental information, but there's the FOMO aspect. I don't think that's a great way to follow it. And then I think there's a living in resentment and bitterness that you're not a part of it and just like praying on its downfall and all these feet, you know, and I think that, again, that clouds your judgment as well. I think you need to look at a third, which is you need to stay in your zone. And if your zone's doing well and you're kind of know what you're about, then just stay sure to stay sure to that, you know, you don't need to so be so worried about what the cool kids are doing every month, right? Having said that, I do a little bit at the second two. I mean, you don't want a financial market to work where it's just, you know, you're able to just bet on something that is some hot stock at the time and all of a sudden make a million dollars. I feel like that doesn't create good incentives for other people's investing habits. Well, and of course, this is. is him being bitter or resentful, but we all know historically that that usually doesn't last very long. So that's all I'm saying. Again, maybe this time's different. Maybe some of this works out, but again, historically, I don't think this time is different, but okay. Quick money and easy money usually doesn't end well. Okay. That's enough about, yeah, maybe this time. That's enough about the macro. Let's get into it. We did pretty good because we started this podcast where Alex said, I don't want any topics and I just want to talk. And we're ready 24 minutes in. So pretty good on you. There you go. Now let's get into my favorite topic, which is talking about how old software companies are going to interact with AI. No, it is interesting. I mean, you've spent a lot of, you have Adobe coming out, which you obviously had a great YouTube video about and then you also are putting out the Speedwell Research Report on that in the coming days, I believe. So that'll be exciting. It'll be out probably by the time this is released. So I'm sure you're going to flesh out a lot of your, and again, specific to Adobe, but I'm sure a lot more of your kind of overall thesis about legacy software and their interactions with AI. And I'm sure a little bit of Adobe. Yeah, give a little, yeah, give a little, what do you got going with Adobe? So it was interesting going more into kind of the risk of AI with Adobe because I was able to identify eight separate risk in Adobe and AI specifically. And was the problem is almost none of them are really kill shots to Adobe. A couple of them are, I don't think they're very likely. But outside of the kill shots, all of these other risks, they're all things that could like knock off a little bit of tam, a couple points of growth rate. But you don't exactly know for sure because they never disclose their customer mix of enterprise versus consumer. And there's very different competitive dynamics between the two. But you could kind of see a lot of these risks just weighing a little bit on growth and then maybe over time. You know, growth kind of deteriorates and even though it's a very low multiple stock, it doesn't look quite as attractive as maybe you saw it to be. On the other hand, I could walk away from this reporter and purposely being a little vague. We'll talk more about this maybe another time. You could walk away from that saying that they're definitely going to be growing, you know, at least GDP rates. And if you could do that, then it's pretty easy to get, you know, a 10% plus return, if not much higher as an investor. And if they continue to grow, you know, even just high single digits, you could be looking at a pretty good return. So we can say more about Adobe on a separate podcast, but it was kind of interesting because there was a lot of risk. But when you really kind of parse them out, the individual risk or do not seem that strong. And then it's kind of like an issue where you have so many things at once. It's like hard to consider all of them. And I feel like that's probably a lot of what, you know, scares investors off a little bit on the name and, you know, maybe rightfully so you shouldn't have to spend, you know, months thinking about a stock and doing a deep dive. You know, there is something to say if something's easy, then you should do it. And if it takes, you know, a long time to kind of get comfortable with it, then, you know, maybe there's something to be said about that. Or you could say on the other side, that's where the opportunity lies. And so anyway, I'll kind of leave it there on Adobe, but some of the things we learn will be talking about kind of in this conversation on AI software. Yeah. And it's just a good time. And again, their reports are evergreen, but I think it's going to be a pretty compelling discussion. And I know I'm excited to read it. Why don't we get into a little bit about a software company that you feel is kind of well-positioned in this kind of AI transition here, which is into it, which we've talked to a little bit about. I know they just had their earnings released. And I know we also put out kind of a feeler for some questions that we might be able to address and into it earnings was one of them. So I think we're going to do a very deep dive into it at some point, but why don't we get into a little bit about what kind of the incremental updates are we're having here? Yeah. So to clarify what you say, I wouldn't say they were very well-positioned for AI. I would say they were positioned for AI. So what happened with earnings is they basically tanked 20%. And what kind of scared a lot of people was that TurboTax, the amount of users using it, it fell. And the free-meam users, the free users also fell from about 8 million to 7 million. And I think a lot of people saw that headline, and they instantly connected that to the belief people are going to LLMs to doing their taxes. They're going to AI. This is evidence of disruption. For the first time, we're saying evidence that a software company is being disrupted on the consumer side. And obviously AI is going to be able to do taxes in the future. And that is going to be the end of the day. Can you connect the dot of this TurboTax? Why free-meam is even relevant to it? Because also the IRS put out their own free-meam service for simple returns. So why does Intuit care about these users? They to upsell them eventually thinking more complex returns? That's right. Yeah, that's right. So the way it works is that you could usually file federal taxes for free, simple tax returns for free. And then as you start adding in other things, they charge you for that. There's been a lot of other services for a long time that have competed on price and are cheaper than TurboTax. And TurboTax sometimes, and in the past, would be more competitive against that, giving offers, discounts, and this time they said they kind of stepped back from the market doing less of that. So that was one of the reasons why that was kind of down. The second reason is there's just fewer tax filings this year. This year around 30 basis points less, which was like, I think they said that amounted to roughly 2 million filings and they have like 60% market share. So that's a little over a million less filings for them that you'd expect. So that kind of explains a lot of the drop in the free users. The risk though is not that these are very valuable users, but that this is evidence of erosion in the business model. And then people would think that that's because of AI. Do you defend this risk? Because I'm taking the exact opposite. I don't understand how anyone would download the proper tax form, upload this to an LLM. The LLM would properly query it, then it would fill it out, and then it would then have to send it to the eye. It just seems like such a worse deal. Sure. So that's not what's happening. What's happening is their old competitor set, who's existed for a long time, got more aggressive on pricing and took a little bit of market share. They set on the call as one point of market share. But this is only amongst the least valuable users. So this is people who, with under $50,000 a year, that file pretty simple returns. And so they basically got less competitive in that market, didn't have offers. At the same time, their competitors got more competitive. They lost one point of market share. So they set on the call explicitly that AI was not a factor in that. Now, how exactly they would know that? I'm not entirely sure to be fair, but it makes sense to me that if you go to one of these LLMs, the experience now, it just doesn't work. It's going to lie. It's going to fabricate something. It's going to make up some numbers. So that to me seems credible that you're not seeing any actual threat today from AI doing people's taxes. That it was just the result of these other competitors being more competitive on pricing. So that was on the turbo set tax side. I think then that kind of combined and linked to people's fears that AI is only going to continue to get better and they're going to continue to move up market over time and get even into more complex returns. And if you're already seeing it at the low end, eventually it's going to move up to the high end. This whole business is a zero. And I think that's kind of why the market reacted the way it did. Now, the turbo tax business itself, though, is only 25% of revenues today and about a similar amount of profits. And so the stock was down 20% today. So it almost wiped off the entire kind of value of this business today. If we could think of turbo tax and the rest of the business is as kind of holding a similar multiple of being similar quality, which maybe you disagree with, but kind of simple math. You could see that a huge chunk of value was taken out of the market today as a result of these fears on turbo tax. This is a weird one. I still can't name another. I'm a CPA. I can't name another online tax service. I can't. Again, I haven't looked for any, but I just feel like the market share, it's still so strong. And again, if you're having someone who's searching to save, it's as a percentage of the headache saved, the turbo tax into it 100, 200, it's just, it's not a big amount of money for the amount of headache it saves. So for me, this is kind of a weird reaction that it was as visceral. I do find it to be a little strange. And again, you and I, I think our conclusion through many hours of this AI discussion, which is until the AI agent becomes literally so good that I say, go into all of my records logging into everywhere and file my taxes. And it does literally all of that. And you know, and again, at that point, that is like, I don't know who has jobs at that point because that's a pretty difficult task to do. I don't know how, and again, I think that's the, that's the general narrative, which is that's where we're going. And if that's where we're going, yeah, I don't think into it's going to do well. If I have a literally an AI agent who is better than any tax filer in the world who has access to everything I've ever had and who can just do everything, then yeah, I don't think into it's going to be worth a lot of money. But I don't know who's paying that agent. Again, I don't know maybe I'm paying clawed in that instance. I'm not sure, right? This is where we keep getting into these discussions. So let me steal man the bear case a little bit more. So you have, you do have a lot of these other companies because into it only is about 60% market share there. The other 40, there's a lot of these small companies. They're called like direct free filing, like file free USA. There's a bunch of them. Those are roughly the names. I'm kind of forgetting them, but there's a bunch of them and people go on them because they're cheap, right? They'll advertise all over the place, especially during tax season. You go on them. It's similar, you know, in flow and all that. You upload everything and the ideas as AI gets better, then they'll be able to also accommodate more complex tax returns as well. And the problem is you can't go directly to an LLM because LLMs are probabilistic and so you need basically business logic added on top of that. Something that's going through and is able to read all of the tax laws and turn it into a deterministic logic where it's not going to fabricate an itemized deduction, something like that. So into it has that, other people could build that as well. There are some fear that maybe the LL models themselves will build that business logic layer. I don't really see that happening unless they do it just because they need to gain more like consumer usage because people aren't using their apps. I can't see Google doing that, maybe Chatchy PT does it to try to stoke some usage, but even then there's kind of a lot of liability with that. I know XAI was asking to get people's tax returns, but they've always kind of cut certain things a little looser in experimentation. As it exists today, you can't do it. You just can't reliably get your tax returns done from AI. Now, we all know AI is going to get better and all of that, but you're still going to need that deterministic business logic layer. Yes, it's not impossible to build that. It's not trivially easy either, and then there's that liability aspect to it as well. The real thing though is just existing competitors using AI to then move up market more in undercutting them on price. I think what we've seen though is that into it's basically been leaning more into complexity and more into a human layer. Their TurboTax Live business that uses utilizes humans and human advice and all of that is grown. It was 36% year-of-a-year. That now is more than half of the entire TurboTax business is using a live expert. These are more complicated tax returns that are just going to be harder to do over time. At the end of the day, you're talking about your taxes. It's not that much money. If you get this wrong, it's really bad. You could be personally liable for penalties. Why wouldn't you pay if it's a complex return a few hundred bucks to make sure that this is done correctly and into it is liable in case there's a mistake? Again, my version of this is AI agent who has reached such a level of accuracy that again, so many of these businesses are obsolete. That's a narrative you could take that bet. What do you think about that? I think that's an outcome that's not crazy. I don't know if that outcomes in 10 years from now or maybe in never coming up. I could see someone who's filing their own taxes or something and maybe they would usually use an account and instead they're just going in LLM asking questions. Maybe they're going back a couple times to confirm it's correct, understand the rules and that kind of ways on them. At the same time, it is hard for me to see how AI doesn't eventually get good enough to be able to do something like taxes. But it's just not there. It's just not there. I don't know that the trust is going to be there either. To me, it's just, and this is, I feel like this is where we land on and every time. It's just difficult, difficult to call. That's where you need to go. You talk about how the cash flow of business is over the lifetime of that business is into its lifetime value impaired possibly. I think the only way it is is that if, for me, this is the only world I can imagine, is you have an agent layer, I don't know who controls that agent, but they are so good at everything. They are embedded into every aspect of your life that you say, "Go file my taxes," and it does it. Then into its introuble. So even then, kind of interesting. They're doing a lot of instant refunds. You file your tax return. If you get a refund, they'll basically finance that refund for you. Now you're getting involved in actual money layer. That's something that, less likely, like an AI company is going to want to do that. They give money to a customer immediately based off of the returns it calculates. Right. Yeah, that's an interesting thing too. Again, it just comes down to, I don't know, this is where it all is price expectation, what's my return and my willing to take it. It seems like the price has been cut down quite a bit to where it might be worth taking that bet. I think that, if you're looking at all the different software companies, I did this video where I ranked a bunch of them, but I probably would now think about it a little bit more, say that TurboTax, I could see some risk for it, but it's still only at that lower-end DIY segment. They're still growing more into the human segment, and now they're opening up stores too, which is twofold. One, so people can bring in their tax returns and get help in person. Two, it also is basically advertising because they weren't showing up on Google Maps and people just searching accountant on Google Maps. That'll help them. It seems like they're just pushing more into a human, a trust layer, higher complexity, and all of that. Even again, that's still a quarter of their business. Of course, that's the main you want it to be secondarily pressured, but they still have the QuickBooks franchise, which is a very different sort of, I think, the risk that they're facing there. I agree. I completely, completely different narrative on the QuickBooks because, and this was where we ultimately came to our service now discussion, which is why I felt kind of strongly that service now is a difficult business to disrupt because, like, if you're a fine CFO or a finance or you want to do any type of books, maybe the AI can actually go through and categorize everything I suppose, but again, it gets it wrong. Right now, QuickBooks has AI that suggests things and it's just wrong or you don't like the category it's in and maybe it can learn that over time, but I still have to go in there and look at financial statements. I mean, assuming I still have a job or anyone has a job, again, I make that caveat, which we could be in the Wally world where we're just floating around on chairs, drinking big gulps. That could be the world, but assuming we're not in that world, you need to log in and do something somewhere. And I don't think the AI agent is going to completely obfuscate that portion of the business. I can see the tax return portion of the business being disrupted more. But again, you got to pay for that price and is that happening next year? Is that happening in 10 years? Is that, I mean, 20 years? I mean, how much cash flow do you really need from that portion of the business to justify the valuation? And maybe you put it higher discount rate on that or so lower growth rate on that or maybe even a negative growth rate on that at some point and see if it makes sense. I have nothing to add. Good. That means I did a good job. So let's move on from into it and we get on to another software company, which I think this podcast has been quite a bit of time on. We did our deep dive of constellation, Chris Meyer on where he spoke about constellation. They only meet once a year, essentially, is when we get to hear from the great oracles of constellation at their AGM meeting. What and we can do a deep dive on it just happened. So it's a little fresh. But what was there? AI take. I think you and Chris did a really good job of fleshing out the whole AI situation with constellation, which I kind of go back to my quick books thing, which is again, if you have chicken coop software or bowling alley software, this all needs to live somewhere. You know, an AI agent might make it better or maybe make it harder to code easier to code, that was really never the barrier to entry anyway, because these are already very simple systems. And you still need to have some interaction, assuming, again, humans like to do things like each chickens or go bowling. I don't know. Again, maybe that changes. I'm not taking that bet, especially across the thousands of companies they have. If bowling goes out tomorrow, they'll be fine. So what was their narrative about it? What was management saying on the AI? I think they basically just confirmed a lot of the things we expected, which was that constellation didn't look at AI and say, hey, that's not relevant to us. Let's just keep doing what we're doing and not incorporate any of that. Instead, they gave a lot of examples of how AI was making product development faster, allowing them to give users the products that they requested a lot quicker. In one example, they said we're building features not in six months, but in six hours. And so they're utilizing AI, which is what you would have expected them to do. You know, they're in the software business. And now they have an easier way to make software. And as we've talked about at LibDOM, the risk to this business in barriers to entry was never the creation of the product. Right? That's not the same as winning business and distributing the product. And once a customer's on the product, very sticky and hard to leave. And so as long as they are able to provide a product and experience that is as good, even marginally potentially worse than an alternative, even at a little higher price, they're still not going to leave. You know, that's just because of the switching cost involved in all of this. And so I, you know, it's good to, it's not like I necessarily learned anything, but it was good to confirm that all of that was the case that they do have a lot of examples of companies going out there building new things, new features and all of that that they're selling and doing well. We didn't see that show up though in the organic revenue numbers. That thing has really still been stubborn at around 2%. And so my, you know, base cases, it's kind of the same, you know, theory is before. You don't really expect too much organic growth from constellation, but excess cash is going to be allocated very well into acquisitions. And so it's just continuation in the same story. I think there's some more upside opportunity that AI does accelerate their product roadmap and allow them to charge more in the future. But, you know, we haven't really seen that yet. But I think that's more likely now than at any point in the past. So that's kind of there is an upside opportunity. But, you know, the tricky part with that though is that, you know, a lot of their businesses, they do still go out of business. They said one of the most common reasons that someone switches the software is a business is acquired, combined or goes out of business. And so the last time they gave turn statistics, their turn was something like 7%, which is, you know, a little bit on the higher side. But that's the reason for the majority of it. I mean, that makes sense if you think about it. It's like a business lifetime of like 14-ish years, you know, which is probably the case for a lot of these small businesses. I just, a constellation again, this year or like on a trailing 12 month basis down 50% or 45%. This is 45% ago or a year ago. The AI narrative was very alive and well and the video was ripping and everyone in the world was taking over. I don't quite understand what I mean in the numbers has anything materially changed with constellation that has it's just a just a narrative shift that's happened with constellation. And again, I'm just not I guess maybe the reinvestment opportunity and you talked with Chris Marabell about this is maybe the biggest risk that perhaps there the ability for the deploy capital at this rate is going to decline which I think has always been the risk for constellation. And I think that's actually less of a risk now as they've done more of these big deals and they've done the permanent equity, minority interest and so there's even less of a risk of that now because they've been deploying so much capital and they deployed a lot of capital to last quarter in year to date. So that's less of a risk now as well. As you get bigger that numbers get harder but they've already shown their ability to kind of step it up to the first level. So in terms of what actually precipitated this sell off, I think this is a very educational moment for investors who believe that stock prices are driven by actual business developments. Sometimes even threats, narratives and all that because as you point out the AI narrative was there for a while of what really caused this decline as people decided to start caring more and fearing about it more and it all happens together at the same time. It's almost like this high-digger idea of the guist or the moment in time. I'm probably butchering his philosophy there but for some reason for somehow certain ideas just take a hold of a crowd and it's very hard to predict that and when they do it just can lead to very strong behaviors, heard-like behaviors and people who wouldn't have really questioned software so much even if they saw something from Anthropic come out a year ago, all of a sudden now are pessimistic on the entire space which is fine. People can make their decisions and all that but it certainly is odd when nothing else really changes or develops that all of a sudden people decide that this is something I'm going to care about now and yeah, Anthropics and putting out more demos and all of that and maybe it feels a little bit more salient so there's a little saliency by us and all that but a lot of this existed a year ago but it didn't react at the time. Yeah, it's an interesting one and again one one will continue to monitor. I think Constellation has a again I think it was you know always reassuring for the AGM because you're always kind of in the dark and it's been obviously a big change in the stock price. Not a lot of big change not a big change in the business necessarily from you know financial standpoint but interesting to hear straight from the horse's mouth what I think a lot of investors were insinuating. Yeah and I'm not complaining I don't mind it when you can buy you know a great company at a better price. That's what you should want and you should hope that these sort of things continue to happen. You know growing up when I would read a lot of these investing books I just kind of figured that everyone would read these and the market would just become more efficient over time and it's interesting how that has not happened and that fear was very misplaced. Yeah, I mean I think it's just the interpretation. I do think you know most companies most stocks they just don't go down for I mean it's not like there's how to put in no bearcares here right I mean I think there's a bear case and I again people think there's more credence to it or not and at a certain price you take the better you don't. I think it's like a lot of times when there's a bear case it seems scarier before you analyze it more and I like you know I give this example a lot but very much with meta right there's a lot of scary things happening at one time with meta and then you would analyze it more and you realize the real risk is like whether or not they're going to be able to increase ad pricing and everything else really wasn't a risk and then whether or not they'd be able to do that is a question of whether or not all of the spend, tech spend everything they were doing at the time all of you know the smartest minds working to improve ad targeting whether or not that would yield any sort of result and if you thought it did then the outcome of that just mechanically would be ad prices increased and so you know you it seems scary but then you get into it and you're like yeah it's kind of hard to think that like the ad prices are never going to go up and they're never going to get better at ad targeting ever and so sometimes it seems scarier and then you zoom in and it's not quite as scary and I think that's kind of the case with constellation you see Claude whip up you know software really quickly and you're like what the hell like what why doesn't someone just use that like they could just type and then you know you kind of go into all these other factors we've talked about before and you see that it's not that simple it's not that simple yeah um no I think it's a good point I mean you know knowledge is power right and so let's get on to the I guess we'll say final AI discussion plus software which is or something about Rick Smith and axon who he loves the latest buzzwords of the day you know he was all about the cloud everything's in the cloud now we're all about AI I don't know how many times per earnings call he says the word AI I'm going to guess over under 50 I'm going I'm taking the over on 50 personally what do you think about that I don't like that you get to set the odds all right you're come on you know what do you so what's axon I mean it seems like axons thriving the AI world they are applying AI they are integrating it with physical hardware they are continuing to just win more and more wallow chair demonstrate efficiency to the you know police stations and then also just security apparatuses just past their traditional police departments seems like they're firing on all cylinders what what's your thoughts here firing on all cylinders is absolutely accurate they continue to post quarters where they're growing more than 30% a year and almost in every respect revenue growth has been stronger than what I would have kind of expected a few years ago when we wrote that report and AI has been a big beneficiary of them in ways that you know we weren't really thinking at that time either because you know they have the body cam business and that's just recording video footage now they could use AI to translate all of that in right actual police reports it's called wonderhaft and police spend a lot of time doing paperwork and now they're using AI to basically have the police people have their own paperwork written for them and so that's not an issue and they're saving time that way and you know they have the taser business where they really don't have competitors they also just started gaining more prominence international with that and then the body cam business yeah there's some competitors but the evidence.com platform the axon platform and all that they're dominant in that and it continues to give them more of an advantage too because of all the software and AI features they're adding on whereas competitors are really far behind there so everything that's happened it's just like kind of competitively advantaged them even more and now the new latest thing is drones which could be another big thing for them because there's no real good defense system that police have against drones so they made an acquisition of didron and they now are going to basically help enable police to use drones in policing all of the video involved in that plus also then doing like anti counter drone surveillance and all of that if someone you know has a drone up there you know god forbid maybe there's a bomb under something how to address that threat and so in every respect this business has kind of had more and more growth legs you know from you know the drones thing that was there but now it's really started taking off international took off even more their latest taser the taser 10 has been very successful all of these AI additions to their entire platform all of it is really gone really well for them and you know the stock though got really expensive it got really up there I think at one time it was close to 30 times sales it's down about 50% and now trades you know around 10 times sales which you know it's still expensive for stock and especially because they have a really weird cash conversion cycle for some reason where very little earnings actually converts to cash which I'm not a fan of predictably you know the last 12 months their cash flow was like basically 20 million dollars and that's before we take into account they have 630 million dollars in stock base comp so of course you back that out and it's actually negative and you know I understand mature margins and all that and I teach that and all that but I don't quite understand why the disconnect is that much you know I get you know an R&D expense of you know 720 million which is what it is LTM I don't quite understand you know the SGNA of over a billion though I don't understand why there really hasn't been much leverage on that line item so you know it's some thoughts there the business has been really strong it's a very competitively advantaged business AI is a clear beneficiary for them it doesn't make their business weaker in any respect they have very few competitors they're very dominant and they're very still expensive and the cash conversion is not great so that's kind of my my couple sentences on axon yeah and again it's another thing where business fire and all cylinders down 50% from its all time high again still it's still not cheap by most metrics but I think yeah I mean you know the argument is then you know if there really are that dominant and no one is threatening them and they're growing this quickly they're gonna have her own way to grow into that valuation yeah I mean it's it's a good counter argument and again another interesting company that got kind of caught up in this narrative but let's move on from the AI software and let's get into kind of the real world here into the physical world the world of atoms which is the e-commerce players the international e-commerce players specifically Mellie and C focusing both on kind of the developing world specifically Mellie Arshantina South America and C in Southeast Asia again these were kind of I think these were the darlings of the 2021 kind of I don't want to say bubble whatever you want to call that they were the 2021 NVIDIA and micron they were really excited you know everyone was I think one of those is a compliment and the other one it's unclear Yeah, time will tell whether both of those were compliments or not. And so what's your, I know you've been doing a lot of work on them. What are your just thoughts? They just released earnings? I mean, what's going on with these businesses? I mean, I feel like they've kind of lost the attention of not completely lost, but I mean, I don't think people are that excited about them anymore, you know? Yeah, and they've all been accelerating growth and re-enbesting back into the business. If we're taking Mercado Libre first, they'll say, you know, on growth metrics, everything was great. The one thing that we can kind of nitpick on is basically, as they're investing back into the business, it is kind of a suggestion that maybe they were over-earning in some of their businesses, particularly in Brazilian commerce, in certain categories, because they said they were going to take rates down in order to be more competitive, right? And increase the sellers on the platform, and, you know, that way we can have more selection there. And so you don't do that unless, you know, it's in response to competition, and unless you were so expensive that it was restricting sellers. And so that, to me, kind of suggests that, you know, this was a market that they were competing against Amazon, but Amazon was pretty distracted for some time. And so maybe they were able to, you know, get a little high on some of the monetization, the seller commissions and all that, and then once see who owned Shopee, who came into Brazil, now is the leader by order volume, but not order value, they've basically had to be much more competitive. So I think Shopee basically made Brazil a little bit worse of an e-commerce market for Mellie. As a result, you know, they've had to take down their take rates. They've also done a lot of other initiatives, including offering free shipping on, you know, very low order values, I believe it's 49 reels, which is very, very cheap, you're getting free shipping. And they're leaning more into kind of the lowest end of commerce to kind of not get disrupted by Shopee, who's been really been able to take a hold there. At the same time, they've rolled out a slow shipping network, which allows them to more profitably ship these items, because part of like the innovator's dilemma, if you will, is that when you build up your logistics network for speed, it's not optimized for cost. And so then it becomes an issue when someone is coming in trying to sell, you know, $6 items. You can't do that and offer free shipping profitably. And so they've kind of pulled that bandaid off. They still went into the lower market, but then they opened up this slow shipping network, which allows them to more cost efficiently deliver these items. And then also as they build more volume, it should just make the overall logistics operation more unit profitable. And so it all makes sense. I think that, you know, it's a little bit of a net negative that you have more competition, but such as life. And the other thing I want to point out here, though, is that they have this figure called Nimble, net interest margin after loss. And we also saw that that went down a little bit. And that's basically for a few reasons. One of them is they're going a lot more into credit cards, credit cards, lower interest rates, worse economics, basically, than like the merchant loans, personal loans they were doing before. Then they also extended duration a little bit. A couple months, I believe, it was from five to eight months or six to eight months, something like that. And whenever you have a longer duration loan, you have to provision more basically against that because that's kind of what Cecil says. So the two takeaways I have from this, though, is that yes, they're growing very quickly and finance it, and that's great. This is also if anything goes wrong and you ask me like what went wrong with me, a few years from now, I would just guess credit risk. 100% that's what I would guess because anytime you have someone growing very quickly in credit, you have to wonder are credit standards loosening. I'm sure they would say no, and I'm sure they would say that, you know, they're doing some experiments, but it's very small dollar amounts. And it's worth it for the customer acquisition cost, even though we, you know, model out X number people are going to default. It's still going to be net worth it for us to acquire these customers because if they become or call apogo customers, then they're also going to spend more on e-commerce and it's a stickier customer. And they're also, you know, profitable using all the financial services. So all of that makes sense to me. I'm just saying like that is where the risk is going to lie is in that aspect of the business. And so that's kind of big picture, but they are, you know, don't want to under emphasize. They're growing a lot more. They re-accelerated growth. And I do think that, you know, it shows their flywheel is really working between the financial services in e-commerce segment. It's a pretty close loop. And then on top of that, they have the Mally membership program, which offers, you know, value to a lot of different video streaming services. More than, you know, you could get from prime, they partner with HBO, Disney, maybe Paramount, a couple others, Netflix. And so it's a pretty good value. And you could package basically all the streaming services together, pay a higher fee for that, like a premium fee. But then that's also meaning that you're getting everything kind of bondled. So, you know, it's a good service for sure. And they're, you know, they're growing a lot more as a result of all that. So I'll pause there. That's kind of the takeaways from Mally, the credit risk, the lower, you know, take rates. Those are kind of the main things competition and credit risk. Yeah. Mally has become quite an interesting business. And I know Steve's doing the same thing with these kind of micro loans at these very high rates. And they're kind of, it's obfuscating the finances. And it's difficult because now you're looking at a bank and are mingled with e-commerce and are mingled. So I think it's becoming a little hairier and more difficult to understand. And then again, I know you alluded to the net interest margin, which is exceptionally high, which, you know, again, an investor has to determine whether that's sustainable. And it might be because I mean, obviously nobody is really in a position to lend to these people at this scale, right? So. Oh, sorry. I have to call out that I said something on the podcast with Rose, where I said, I don't know, you know, how these high interest rates are that sustainable when you do have this much competition. And not the full reason, but part of the reason why the interest rates came down a little bit in the nimble was competition. So, you know, whether they stabilize here continue to fall. I'm not sure. But to me, such a, you know, very profitable business, it seems unlikely it's going to state that profitable, you know, at least forever. But, you know, it's still an evergreen market. They're still growing. But we do see a little bit of pressure there. That doesn't mean they can't still be, you know, a very good unit economic business. And it could still make sense for them to acquire these customers. Just the economics are kind of unusually high. Yeah. And, you know, always have to be very concerned when you have a exceedingly high margin business and really think through how they're going to defend them in the long term. And, you know, I mean, Mel, he's got a unique touch point with these consumers. But of course, there's other large players in each of these markets. And obviously, I'm sure, you know, this is all public information. So they're looking at what they're making. And I'm sure. Very interested in getting in on that. So, why don't we get a little bit real quick and to see. It's just such a cool company. See, I mean, what a very scrappy start from free fire into the e-commerce player. I mean, they are quite a compelling narrative. And I like keeping up with them. They're fun to kind of watch. So what's been going on with them if you want to just give a brief update before we maybe do a deep dive on them at some point. Yeah, very brief. I would just say, you know, last quarter, everything look good, you know, growth was very strong. The upside surprise was that Granada, their gaming division did a lot better than people expected. It had its best quarter in a couple years. So, since actually basically since COVID when everyone was locked down and playing games. So that was one of the reasons why the stock kind of popped on that news. And so, otherwise, I don't know that I have too much to add. You know, the same sort of credit risk thing, you know, that critique still exists there too with C money and that growing really quickly within shoppy. And then you have another risk with TikTok shops. That's kind of their key competitor right now. Don't know that we can really saw anything that was like concerning though. And that regard and so we'll continue to monitor that because that's kind of the key to risk. I guess for that businesses, you know, you have the credit business risk. And then you also have that, you know, is TikTok shops going to continue to weigh on potentially their ability to grow. I really think so. I think they're both going to grow. I mean, would shoppy have grown more in a market where TikTok shops didn't exist probably. But is it, you know, a direct threat? I don't really think so. I still think it's more tangential. But we'll have to continue to watch that. Yeah. And again, I think we'll see whether, you know, Speedwell eventually does a deep dive on them. But again, I think you have two great YouTube videos on them and on both Malian and see it. Maybe we'll do an update. You'll have an update video in store for us at some point. So until then I have one kind of final really pressing question that was asked, which I thought was interesting. And is is Drew Cohen money or Speedwell or any of your subsidiaries hiring help or do you just want to do it all. So I do get a good number of requests for interns and help and I'm honored by all of them. I don't know how to use an intern in a way that would allow me to work faster and not just teach basically. And so I have not figured that one out yet. Ideally, it would be cool to expand and cover more companies and all of that. I don't know how to do that though because I'm also a little bit of I want to say control freak, but I don't trust other people's analysis necessarily. I'm not a control freak, but I want everything done exactly how I wanted and not a second different, not a little bit different, which I get. I mean, I think that you know, I mean, it's very difficult. Obviously, if there was a ability for everyone to have this level of insight on all these companies, then it wouldn't be a very unique compelling value proposition, which obviously many people do find this to be unique inside. I know I do. So I it's not easily replicated. And look, I wish I could take over some of the workload every time I've tried to use it to actually analyze a business. It gets like it's good at summary and it could say like the things people are talking about, but it is just like, I don't know, it just misses on analysis. It doesn't actually understand how like different things impact other things. - Sometimes I can get it for that. - Yeah. (laughing) - Yeah. So it's a lot better, I think, at search than, like, regular search was before. It could definitely pull in facts that you wouldn't find otherwise than you got to confirm it didn't lie about them. But that's kind of where we're at with there. So, you know, AI, I don't know that it's coming for taxes this year, but maybe my job is not too far behind. You know, I swear if, like, in 10 years from now, they still are like, we're just around the corner from AGI. I'm going to just be upset because they've just been over-hyping this now, where I'm just, like, expecting humanity to end any day now, just expecting it. It's definitely not going to be like autonomous vehicles that were just around the corner for many years. I guess they're here now, but yeah, it'll be interesting if this time is actually different yet to be proven. So, yeah, no, I think for now Drew, you're still in a job and AI has not taken your analytics quite yet. So, you know, hang on to the next three to four months while you still can. And, you know, I know it'll take even longer for them to copy your humor. So, there you go. Exactly. I mean, I feel like I'm in a pretty good place, honestly. I don't know about you. No, I mean, you're great. You're the right price, too. Yeah, the right price. There we go. Perfect. Zero dollars. And on that note, we'll see you next time. Until next time. (upbeat music)