AI's Hidden Risk: When Every Quant Owns the Same Stocks | Kai Wu

 Sign up to The Lead-Lag Report on Substack and get 30% off the annual subscription today by visiting http://theleadlag.report/leadlaglive.

Transcript

Live Setup And Welcome

SPEAKER_00 0:06

And uh already one of those Fridays, it is Friday, right? I gotta make sure we're streaming live, gotta make sure it's Friday too.

SPEAKER_01 0:12

Yeah, last I checked, it's Friday.

SPEAKER_00 0:14

I don't even I don't even know what what day and time, I don't know who I am at this point. I uh I spoke already to like four different people. Uh and uh and apparently uh Kai was waiting in this waiting room forever and I'm in the wrong link, so it's like uh my fault on that. Uh those that are watching this live, let me just make sure that we are indeed streaming. And as I keep referencing, Lead Lag Live is back to being live, which means if you are watching this, feel free to engage. We can see your comments, we can see your posts across LinkedIn, YouTube, and X. Uh, let's make this as interactive as possible as I make sure everything's set up here. Uh and as all my agentic AIs are running nonstop in the background, I literally have 40 tabs open across my three screens, all of which are perplexity, all of which are different agenti AIs, just so I can see and monitor what the heck is going on. Uh so it's it's been quite an interesting dynamic the last several days here. Uh one second, folks, and we'll get going. Perfect. Okay, so welcome everybody. Uh my name is Michael Guy and I'm the publisher of the Lead Lag Report and founder of Lead Lag Media. Uh, this is another episode of Lead Lag Live with Mr. Kaiwoo of Spark Line. We talk about a lot of things. I keep trying to think through how do I make these podcasts uh more engaging and more relevant, obviously, to the current market environment, beyond just the sort of standard you know philosophy from an investment perspective. So I've got a number of questions for Kai here, uh, which perplexity is helping me with. Uh so I appreciate the AI side uh from that standpoint. All right, so so Kai, I know you're you're very well uh versed and very deep into longer-term research. Uh let's face it, the reality is people love to talk short-term. So let's talk short-term for a second

Oil Up Stocks Up What Gives

SPEAKER_00 1:44

here. And I want to get your reaction to uh what do you think is going on in markets? Um, it seems perplexing, not because of perplexity, that um you got oil at uh at base in new highs and the market at new highs. Uh and I thought that wasn't supposed to be a thing. We've had a hell of a rip the last four or five weeks. Um I don't know. Have you ever seen anything quite like this? And did do you do you attribute anything unusual to the behavior?

SPEAKER_01 2:08

I mean, I think clearly the market's looking past what's going on in uh Iran. And um, you know, more it is returning to fundamentals. Um, if you look at the this uh a quarterly earnings season, um, you know, it's been really strong, right? I mean, uh just what it just a couple days ago, the uh the big tech companies um you know reported earnings and and revenues, all of which are up. Not all the stocks are up because of uh Cabec's guidance, which I'm sure we can discuss as well. Um, but yeah, I mean it's been a strong earnings season and ultimately fundamentals are what matter. I think investors are kind of resetting to that, you know, knowing that with the Chiopo politics, it's so unpredictable and it's so easy to get whipsawed. Um and so people are more or less looking past that and focused on you know, are these companies actually making money? Um, which I think is a healthy thing.

Redefining Fundamentals For Intangibles

SPEAKER_00 2:57

All right, I want to focus on the word uh fundamentals for a second here because uh it used to be the case that fundamentals were about, you know, low price to book and price to earnings and all these things. Uh and it doesn't seem like the market really rewards those fundamentals of old. Um how should investors think about that word fundamentals in what looks like a totally different environment than anybody's ever seen?

SPEAKER_01 3:18

Yeah, so I think the key to recognize is that the economy has changed dramatically since you know the days of, say, the old school value investors like Ben Graham 100 years ago and and you know, when accounting principles were kind of first outlined, right? So most balance sheets will recognize uh tangible capital investments as being a um source of book value, whereas intangible assets like brands or human capital, IP, network effects, those things are not uh capitalized on the balance sheet. And hence, uh, you know, firms that are very intangible rich have have generally been kind of dismissed by uh more kind of traditional security evaluation metrics. I think that that's been a uh important um you know distortion and important driver of a lot of these businesses over time. And you know, it's been um challenging for uh kind of the most traditional value investors to recognize that. Um so I think that's a really important thing to think about. Um, you know, in particular, you know, we the the the best performing stocks over the past 10, 15 years, obviously the Magnificent 7 hyperscalers, um, these are names that on traditional, you know, backwards looking earnings, you know, folks were just not appreciating the extent to which these companies were investing in um these intangible assets that that were you know would eventually become MOATs. So the best example might be like Amazon, right, which had basically no net income to shareholders for a long time, because what they were doing was reinvesting in the business, building out Amazon Web Services, um, selling goods, you know, quite frankly, at not a huge margin with the idea of building customer goodwill and then building such such scale that they now network effects. If you want to go out and buy a good, like obviously you know, Amazon will be your first stop and you kind of want to even think twice to be a prime. So I think that's that's been a you know important component. Uh traditional value metrics, of course, would would generally have said, oh, these are you know expensive stocks, I would never want to have bought those, which would be a huge mistake. You know, over the past uh 10, 15 years, the Mag 7 have compounded at over twice the rate of the market, creating, I want to say $23 trillion of value for shareholders, or a tremendous amount. Um, and what's interesting is that when I look at things using this intangible value lens, which is to augment traditional book value with also intangible assets like those mentioned, actually these stocks, you know, for a long time appeared quite cheap and quite attractive. And that might be changing, but you know, at least over the past 10, 15 years, um, you know, the distinction between you know a traditional value and then looking at more intangible augmented value would have made a big difference with regards to you know being able to, as a value investor, look, you know, um with a straight face and say these stocks are value stocks.

SPEAKER_00 6:01

All right, so I think this is this is um interesting to kind of think through from a thought experiment perspective.

AI Threatens Software Moats Or Not

SPEAKER_00 6:06

Okay, so you've got intellectual property, brand brand equity, human capital, and you've got network effects. In a world of AI where you just have this SAS pocalypse scare and maybe that's over for now. Um of those four, where does a company have a moat in an ink where AI can compete like that?

SPEAKER_01 6:26

Yeah, I mean, I think this is a really important theme. So SESPocalypse, just um, you know, for viewers, basically, software stocks uh have historically been, say for the past several years, kind of the darlings of the market trading as you know, elevated multiples, a lot of private equity uh money has gone into this in addition to the public markets, and they've kind of been, you know, the the I guess the go-to um sector and have performed really well from a stock performance standpoint. Um, since roughly October of last year, so say for the past six or so months, um, these stocks are in a huge drawdown. If you look at the IGV index, which is the software index, it's down like 50 plus percent, and some of these names are down even more. Um, and and it it's pretty easy to trace back to why that, you know, what investors are scared about, they're they're worried about the idea of um agentic AI um disrupting these software moats. Uh as Michael alluded to earlier on, it's it's pretty easy now for anyone, even non-technical people, to kind of vibe code software. Um, and that that can be quite quite good, actually. Um, and so to the extent that coding agents are um you know just just getting started and ramping up, to what extent do these software companies still have a moat? Now, so I think if you step back and think about this intangible value framework, right, there's these four pillars of intangible value, the first being intellectual property, of which like innovation and coding is a subset. And then there are the other three brand equity, human capital, network effects. So I think the first thing I would say is that for a lot of these companies, um, you know, the example I always returned to is Salesforce, since that's kind of the one of the largest uh companies in the index. Um, the moat has never been its code, right? So Salesforce's moat has never been its code. Even before AI, um, you know, they never had to produce UI. There are plenty of Silicon Valley startups that could build a a shinier product. Right. The reason why Salesforce has city in this is because you know they have an in with all the enterprise companies and our system of record, they're kind of embedded deep in the organizations of these big companies. So for a company to you know rip them out and swap out either a vibe-coded alternative, which would be just not really an option, or even a a competit a competitor is is so painful. And so the switching costs are kind of part of the network effects of the business. There's of course also the brand distribution component where they have access to all these customers and customer support, which is of course, you know, the hardest part of software is not so much building the initial software, it's maintaining and um providing support around that product. You know, there's also the brand reputation component, right? You you can get fired for trying to be too aggressive, but you're not going to get fired for buying the incumbent, whether it's Salesforce, SAP, Oracle, which these kind of old legacy companies that have stuck around for a long time. So, like the moats of these of many of these software businesses are not its code, right? It in fact, if you think about it that way, then potentially these companies tend to benefit from AI to the extent that their largest cost is actually the software developers, which can now, you know, if they are 10 times more productive, um, that cost can go down. Um, and and by the way, a lot of these companies themselves are investing in AI. Um, whether or not they will be the best almost doesn't matter because they already have pre-built distribution um in terms of like the um you know logged in customers. And again, like not every company is like Salesforce. They're probably one of the more extreme examples because they serve the largest companies. I would say that you know, as you go down the stack and and look for software companies that are more primarily serving the long tail, like small and medium businesses, those guys have a have less of a moat in general. Because, you know, for example, for myself, I run a smaller company, you know, I I might actually, if you know, if if compelled, uh switch um out a say CRM. Um and and you know, so I think it it there's a lot of variance across uh companies and in software. So I don't think it's I don't think it's fair to say that everybody is oversold. Um, but you know, I've actually done this, I'm in the midst of doing this work now, um, studying the history of disruption. You know, when you go back to Kodak and Blockbuster and all these companies, and when you look at these historical uh uh technological shifts, one thing you do find is that the market is quite bad at identifying winners and losers in real time. Like oftentimes, like the front runners in a given um in a given era actually don't end up winning, right? Like the you know, AOLs of the internet era. Um, so I think you know that you have to be quite thoughtful about that. And also, you know, likewise, like the incumbents who people expect to be disrupted, in many cases, they actually end up sort of surviving and then ultimately thriving, um, perhaps pivoting their business model, or or it turns out that their complementary assets, say the brand that they had, were actually the limiting factor for success, not so much the innovation itself, which in many cases may end up being commoditized. So, yeah, just but but to answer your question, yeah, I think software stocks, you know, obviously they're down big. I think there will be many of the companies in that category that have no moat and will go away or be acquired uh for pennies. Um, but there are obviously some some gems and some quality companies in that in that mix that, you know, with an intangible lens and understanding the complementary assets that go around just, you know, not just I I IP and so and and and um code, um, you know, one one can uh can find some some nice some nice value.

Liability Regulation And Trust As Moats

SPEAKER_00 11:27

On that brand reputation point, I've been making the point myself that, you know, oddly enough, I think there's also moat in just having liability. Right? Like if you're a Fortune 500 company, you don't want to vibe code your own sales force because you want to have somebody sue, right? If something goes wrong with that underlying data. So oddly enough, having liability, which is part of the brand reputation, right, you can argue, um, is maybe some way of protecting at least the larger firms from getting vibe coded away.

SPEAKER_01 11:52

Yeah, I mean, I think I think that applies more more generally, the kind of regulatory aspect, right? Like you think about like doctors, right? People all all said, oh yeah, you know, these these radiologists are gonna be disintermediated by AI because you know, image classification is kind of one of the canonical use cases for AI. And as it turned out, that was wrong. Actually, radiologists get paid more than ever, and there's now a shortage, right? Because you know, what ended up happening is at the end of the day, you know, you need a medical license. As a patient, I'm not gonna, you know, there's no way for an AI, at least currently, to get a medical license. So someone needs to always be supervising these things if they can use tools, um, which they which have always existed, by the way. Technology has always existed. That's that's owing to the benefit in this case of the the MD, right? The doctor who has that license. Same in our industry, right? Like, you know, I'm a financial advisor, I'm an RIA, I have a C R D number, blah, blah. Right. Like that, that is kind of a a bit of a moat for better or worse. And same in the legal industry, too, right? Like, one thing that we learned, um, perhaps the hard way, is that um conversations with ChatGPT are not privileged. Um so, you know, I I think the regulatory aspect is is is actually quite important. And and, you know, um, for better or worse, um, that that is going to be a moat for many, uh, many companies, um, at least for the time being.

SPEAKER_00 13:07

It's funny, I just had this conversation with uh an advisor in the Lead Like Network who said that um advisors don't realize that the AC SEC uh in their next exam, uh, one of the questions is going to be um if they're using note takers and how well they've vetted note takers because of that that concern around sort of the the privacy and privilege privilege information. And all this is evolving and the regulatory bodies don't know can't seemingly move fast enough. And they themselves don't even know exactly how to go about it. So it's still wild west from that that perspective.

SPEAKER_01 13:37

Absolutely. I mean, you know, I was uh well that's a training article came out, which I'm sure most people, since they're probably on Twitter right now, um, have read, you know, kind of the doomsday scenario. One of the points I was making is that, you know, really the the state of the world where we have like this, you know, massive shock to the economy that we can't recover from is one where this thing this all happens very fast. So ultimately there's a lot of path dependency in the rollout of AI. If AI like um transformation happens very gradually over a long period of time, then I think we'll be okay because the you know, labor market will have to adjust, we can all you know kind of reskill and and and gradually kind of get up to speed, even though even the people who are the furthest behind with AI, the regulatory institutions can kind of you know make make adjustments too. I think the state of the world where we have massive labor market and economic disruption in a bad way is one where things just happen too fast. Um, and so there's some governors on the speed that which through which AI can diffuse. One is obviously just the compute constraints, right? We can only build power plants so fast. Um, and then the second uh constraint is institutional, right? The regulatory, you know, labor unions, even which, you know, maybe you could argue that there's not, you know, some cons with with um those types of things. But at least in this case, I actually think there's some kind of underappreciated positives in in all of this, the regulation, which I guess is generally bad, but at least the the one so underlining is that it you know we'll kind of slow out, slow down the rollout of AI to some extent, um, which of course, which I think on a macro level is actually helpful. Um, and also creates opportunity. So I think one of the things you know that you and I have discussed is you know the extent to which in the long run AI will become table stakes. I think that's that's just kind of obvious. At some point, say 50 years from now, just pick a long-term horizon, we'll all be using AI, and there will be no kind of mode or no differentiation because everyone will be doing it. But there's going to be a window of time, um, which I think will be attenuated due to you know regulation and other factors, that you know, it that there'll be a separation between winners and losers, right? In the same way in historical uh technological revolutions, like the internet, there is you know the early adopters and the laggards. And so I think there's an opportunity for folks, regardless of what your industry is, to you know, lean into AI like you seem to be doing, Michael, um, and kind of like leverage it to the extent that you know you you have some advantage over uh over the rest of the field.

SPEAKER_00 15:53

I I think I think the implication there also is that from a longer-term perspective, human capital um will maybe matter more from the standpoint that that human capital, which cannot the element of the human capital can't that can't be replaced by AI is the human-to-human connection, the charisma, the ability to communicate, right, the soft skills, which you could argue is the only thing that may be an actual moat for individuals, you know. It's like AI will never be able to, I would argue, close a sale like uh a person talking to a person would.

unknown 16:22

Right.

SPEAKER_01 16:22

And I think that go that goes hand in hand with the trust issue, right? Like the trust liability point goes hand in hand with the idea that, you know, at the end of the day, I want to, you know, uh be sold by a human. I want to, you know, provide, give my life savings to a human financial advisor, right? I want like a a human to be the person who's ultimately responsible for the medical care, even if they use AI tools. Um, so I think I think that's right. And I think what's going to happen is is with with uh jobs is you know, they always say jobs are a bundle of tasks, right? What we do on a day-to-day basis is like you know, some combination of 50 different things and in in some proportion, some of those tasks are better done by AI, and that is what it is. And those things will eventually be kind of outsourced. Like we will no longer sit there and like hand hand do long division, right? That's outsourced to Excel, our calculator. And then we'll instead fill our time with um the tasks that are more human, um, where we have humans that have comparative advantages, which means like talking to other humans and and and things like that. And so I I think that, yeah, I mean, there's going to be some change um on the horizon, but not necessarily bad.

Mag 7 CapEx And Index Concentration

SPEAKER_00 17:24

Does this um intangible value framework, which applies to your funds, which we'll talk about in a second, does that result in diversifying away from the concentration risk in the S ⁇ P, the MAG 7s, or does that uh actually result in more concentration because there's a lot of intangible value in the large cap tech side?

SPEAKER_01 17:42

Yeah, I mean, I I can give you the answer now and and and talk about how it's changed through time. So if you go back, like start start, go back to like go back five years ago when the fund was first incepted, you know, we were actually overweight AI infrastructure, right? So at the time this was a 2021-ish, you know, we had NVIDIA, we had all these other names because that was the best way to express the AI trade. Now, fast forward five, six years, and that's been the best performing theme, right? If all you did was just buy AI infrastructure stocks, you would have outperformed by a lot. And that includes the Mag 7, but it also includes other names like you know, Oracle and uh Broadcom and folks like that. But what's happened is is two things. Well, I guess, yeah, two things. One is that um the the Magnificent Seven, who are kind of the paragons of that uh category, they've they've basically shifted their business model from asset light and are becoming increasingly asset heavy. What I mean by that is say the CapEx that they're spending relative to sales has gone from like 4% to like 14%. The free cash flows of companies in the Mag 7 have dwindled from you know high numbers and you know to very low numbers because they are you know spending so much money, trillions, by some measures over the next several years, investing in AI data centers, which is a very capital intensive business. Unfortunately, look at the history of capital cycles, you look at you know past scenarios, you know, asset heavy businesses, um, firms that invest heavily in um in building that infrastructure, um, have generally not performed well. Right. So I think that all else equal makes us more concerned about the the infrastructure trade. And the second piece is just valuations, right? Obviously, the stuff that's done really well, and as a result, the multiples have expanded, um, which you know you do need to wonder when when when will things be played out, right? I think we started to see some um some vibe shift, I guess, in the market starting in October last year, when up until that point, whenever a firm would announce more CapEx on AI, their stock would go up. Investors were you know cheering on Oracle when they announced a big deal with OpenAI. Now it's almost the opposite, right? You had Meta come out with earnings recently, and they um you know had a good quarter. They they they they beat, but like they're um but they they also got it higher on on CapEx. And you know, folks are a little bit skeptical, as they probably should be, given the metaverse situation, um, with regards to how efficiently ZootSuck will be spending this money and whether or not they'll actually be ROI generated. So I think there's been a bit a bit of a shift now in terms of infrastructure. You go back to the history of technological revolutions, and it's always been the case that in the very early phase, as it should be, um infrastructure stocks are kind of the best reforming, right? Because you need to build out the say fiber up with the cables in order to run the internet. You need to build out the electrical lines or the railroad rail tracks in order to, you know, then run trains on top of them. So of course infrastructure does best in the beginning, but never has it been the long-term winner, right? Never over the history of these past um revolutions has it been the case that the long-term winner has also been the builder, right? So in the dot-com boom, it was Netflix, Google, and say um, you know, Meta that ultimately uh took all the profits. And the the telecoms, you know, they thought their stocks went down 92% in the bust um and and never recovered, right? Global crossing and these folks went bankrupt. Um, so I think it it it does, it is of concern. And then to the final part of your question around index concentration, yeah, I think that that's one of the big reasons why I think active management is increasingly relevant today, right? His over the past 10 years, all you had to do was buy SP 500 or QQQ, and you know, it was basically a concentrated bet on Mag 7. As Mag 7 did better, the weights in the index increased because by definition market cap weighted, and that worked out well. It was a kind of momentum trade, basically. Now the problem is that if you look at the index today, say the SP, um, 33% of the index is in these seven stocks. And then if you add, as I mentioned, some of the other names, you get to 50%. So half your money is in AI infrastructure trade. And we're kind of I I don't think, I don't know if we're late in the cycle, but we're definitely, you know, somewhere it we're well into the we're we're well into the the build out at this point. Let me put it that way. Um and um, you know, to the extent where investors in the SP are like, oh, well, I'm gonna pass the index fund and like pretty diversified, you know, that's I think a little bit deceptive to the extent that a lot of I mean basically half of your investments are tied to a single trade. And if that works out, that's great. But you know, it's not. Obvious that there are definitely risks to the AI build out, you know, and the investors in particular in that in that build-out.

SPEAKER_00 22:07

You'd alluded to one of your funds. Let's

Sparkline ETFs And The Early Adopters

SPEAKER_00 22:11

talk about the two ETFs that you've got. You've gotten some good momentum recently, some nice AUM growth. And let's talk about sort of intangible value from an international perspective versus domestic perspective.

SPEAKER_01 22:22

Yeah. So we have, yeah, we have two funds. I mentioned the original US-focused fund, you know, and kind of how their exposure to AI infrastructure versus the flip side, which is early adopters, has changed. Well, actually, let me talk more about early adopters. Right. So the extent to which we've um kind of lightened our exposure to the AI trade with respect to infrastructure, the flip side is that it's not like we've said, all right, well, now we're just going to hold like nothing, right? We're not going to like go underway AI, because I actually do potentially personally believe, and the models also agree, that AI is potentially transformative. It's just a question of how do you want to express your exposure to that. So if you're not going to do infrastructure, then the the obvious um counterweight to that is the adopter adopters, right? The the application layer, the folks who don't build out the models, but instead use the models in order to create value at the at their enterprise. So you can look at any given vertical, whether it's financials or or or software, which we we covered already, um, you know, biotech, pharma, healthcare. Um, and what you find is that there's a pretty significant spread between the folks who are, I would say, cutting edge, who are like really leaning into AI and incorporating it in organizations, and those that are not. What's really interesting is that from a valuation standpoint, the market doesn't seem to be tracing this. In other words, on average, companies that are in the former category versus the latter are trading at the same price to earnings ratio. So you can almost think of it as like a free call option on an AI, you know, a kind of bullish case on AI. Because there's there's really not a state of the world, I don't think, where AI is indeed transformative, where the value doesn't, where there doesn't create value also for the folks using it, right? If it doesn't create value for the folks using it, then they'll stop using it and then the infrastructure stocks will crash. Right. So you kind of those things are almost tied together. But the nice thing about the early adopters, right, the folks in any given industry who are you know leaning into AI is that if AI isn't true, indeed transformative, they benefits. But if not, then they're not like really spending that much money on it, right? Like they're not spending trillions of dollars billing it out. They're just you know making API calls, let's say. Um and you know, maybe they have to lay off a few AI AI engineers. That's not going to be a huge cost um and their valuations and the multiples are not pricing in any kind of market share gain or in growth uh due to AI. So I think that's a that that's an interesting, um, that's kind of how we're expressing this. This is the kind of the second order beneficiaries of of AI. Um so if you look at, say, the you know, US-based fund, you know, we've kind of rotated into you know away from just kind of pure tech exposure and hardware exposure into um other sectors in what you would call that the old economy, right? So industrials and and healthcare and such. And then in um in the case of software, you know, more recently, you know, this is this is only kind of something that we're now kind of pushing into, you know, given the sell-off that we discussed, um, looking for um folks with moats in the software sector um that you know we think you know might actually be in the oversold category. And then if you go abroad, one of the interesting things is this, which is, you know, people over the past 10, 15 years, obviously, we know that um the international stocks have underperformed US stocks. And that's largely because of the success of the Mag 7 and um, you know, who are the biggest parts of the index and the absence thereof in the non-US space. So on the AI front, when it comes to innovation, um, it's it's been pretty clear that, yeah, you know, there's not that much innovation, um, much less innovation, let's say, outside the US than in the US. And that's been a large component of why, you know, an investment in real US dollar terms and say EVA index has basically been dead money for 10 years. Um, and yeah, there's of course like pockets of companies that are part of the infrastructure trade outside the US, and that would include like ASML and CSMC, but for the for the kind of um, you know, in general, right, that they've kind of been missing out on the category. So that's why we're pretty excited about um the opportunity in early adopters, because early adopters, not only do they exist across more industries than just say hardware, they also exist across more countries, right? So, like, you know, whether you're a Japanese industrial or a Swiss pharmaceutical company, like you stand to benefit from using AI, you know, in the same way that a company that just happens to be domiciled in the US would also, right? You know, most of these companies now compete on a global scale. And by the way, these companies are trading at, say, just look at their PE ratios, like a 50% discount to US stocks, just because of the the you know, investors have generally now, because of their underperformance, shunned um the international indices. So you have this interesting scenario where um, you know, you have uh cheapness and evaluation, yet the potential upside, if they get it right, of incorporating you know some of these innovations on the early adopter side um in in the non-US markets in Europe and Japan. Um you have to know where to look, right? Because the index itself, if you look at the IFA index, 50% industrials and banks, right? Banks are you know a third of the index. Um, not saying all European and non-US banks are a legacy, but many of them are. Um and so that I think the key is looking at the subsets of non-US stocks, the pockets of value that are truly modern, both in terms of the um the IP side, the innovation AI exposure, but also the other intangible assets like you know, human capital, brand, network effects. Um when you look at those things collectively, what you do find is that that subset of companies outside the US have actually historically done quite quite well. They've done just fine. They're kind of more US-like when it comes to their um growth rates and their and their profit margins. Um so that's not only been not only, I think, how our position moving forward, but also if you just go past that look at the past 10 years, those are the guys within the non-US sectors that have actually outperformed and you know, we think will continue to outperform.

Valuing Intangibles Beyond DCF

SPEAKER_00 27:53

There used to be this uh this thing called discounted cash flow analysis. Remember DCF? I feel like that's no very uh it's like it's like you see those memes of like, you know, I guess it's like the the gram and dod fundamental investor, intelligent investor getting thrown on the in the garbage. Um the reason I'm mentioning that is um when you think about valuing uh intangible assets, it it doesn't sound like it's interest rate driven in the way DCF would be, right? You're not discounting cash flow in the same way that comes from intangible value. So how do you think about value the actual valuation of the intangible side?

SPEAKER_01 28:23

Yeah, so a couple things here. So so one interesting component of the of intangible assets is that they are less interest rate sensitive than tangible assets. And it has to do with the fact that, you know, there it's harder to collateralize, say IP, right? You're not gonna go to a bank and say, oh, yeah, maybe a loan against this patent, or you know, it's harder to do that. Um and so interest rates do tend to matter less in in this case. Um, you know, where intangible assets are really useful is in figuring out like companies where companies are investing, right? Like, because some value investors say, you know what, like price earnings, price of the book doesn't seem to work that well anymore. What if I look at free cash flow, as you mentioned, you know, and and kind of DCF that? Um the challenge is like trying to know whether a company with good price to free cash flow or or such, where they're actually investing that money, right? And there's of course this distinction between maintenance and um investment, right? Whether you're not whether you're just kind of trying to replace um depreciating assets or amortizing assets, I guess in the case of intangibles, or you're actually building something new that will show up in earnings down the line. Right. So one of the things I I I've studied is the extent to which intangible investment feeds into growth, into future margin expansion. Right. And what you can what you can do, you can do this globally in the US, across you know different sectors across different countries, is look at like the companies on any given time period that invest most in intangible assets that have the highest intangible value. And then ask the question over the next one, two, three, up to 10 years, how that plays out in say ROE, so margins or profitability, or um earnings per share, so earnings growth. And on both metrics, you see this J curve where like companies that invest heavily in intangible assets. Let's imagine you're doing like a lot of RD or you're doing a marketing campaign or hewing capital um formation. In these cases, it actually costs money initially, right? Because you're not you're you're spending money to make this investment. Um, but then over time, you you find say years two, three, four, you're not seeing a positive return on that investment. And then say 10 years out, you're seeing a pretty significant, like, I think it's like a 10 percentage point increase in in margins and such. Right. So obviously companies wouldn't be making these investments if they didn't see um, you know, see ROI on them, right? So, you know, I I think that the the intangibles are are are really important. And, you know, as a it's the case, it's an unfortunate case that traditional metrics like price to books, prices earnings tend to omit or at least be biased against these metrics in a way that you know has caused the factor just not be that what be that good. I think there's one important um you know compounding factor moving forward, which is this AI disruption. One thing I found when I study the history is that um, you know, traditional price of the book, let's say, that tends to um work just fine in protected industries. This is like the things that are not facing disruption. But as soon as you go to industries where there's a technological paradigm shift, let's say e-commerce um disrupting brick and mortar retail, that's when you start ending up with like um what you find is that the factor actually underperforms. So all the underperformance in the uh traditional value factor over the past 15 years is from trying to apply it into sectors that are facing technological disruption. In other words, by using price to book, you're basically buying all the following knives. You're buying all the value traps, right? You're saying, hey, this stock, you know, Blockbuster looks really cheap because its price fell and its earnings have yet to collapse. Although we all know it will collapse, or at least we know with hindsight it will collapse. And I think that gives me a lot of pause when it comes to you know people trying to bottom fish in, say, the software sector, right? The traditional price to book, price to earnings, they're all backward looking. Um, and so you know, I think there's a lot of risk that you know you say, oh, well, this stock X is like super cheap because it has a single digit PE ratio. I'm just gonna buy that. You know, may actually not work because it hasn't worked historically. If you add the intangibles back in, this is kind of the thing I find, it actually tends to fix the problem, right? So intangible value tends to work in both um non-disrupted and disruptive industries, which is why you know I have a bit more confidence as we kind of try to do the bottom phishing in the self with the SAS stocks, um, that looking in for firms which also have a good uh brand and high switching costs and network effects um are likely more likely to survive um the AI disruption than some of their

Where To Learn More And Closing

SPEAKER_01 32:24

peers.

SPEAKER_00 32:24

I think that that makes a ton of sense. Um, for those who want to learn more about uh the funds, Kai, um, where would you point them to and talk about your accessibility? Because one of the things that I like about you is that you're available and you're happy to talk to people, you know, because hey, you're an entrepreneur trying to try to build your own funds, you got a unique perspective on markets.

SPEAKER_01 32:41

Yeah, so I I have uh the ETFs. I mean, you can just Google um our ETFs, but if you go to etf.sparklinecapital.com, etf.sparklinecapital.com, you can learn more with the funds. We also have a separate advisor website, just sparklinecapital.com, where we post all our research. Much of the things I discussed today um can you know can be found in in much greater depth in in these pieces. Um and yeah, you feel free to send me an email. It's pretty obvious to find my email, or you know, also on like Twitter and LinkedIn at uh the handle C K A I W U.

SPEAKER_00 33:11

Appreciate those that watch this. This will be an edited podcast on the YouTube channel soon enough. And uh as always, uh thank you to Kai uh and everything that he does. Uh, I'm always a big fan of the way he thinks about markets, and I think the intangible value framework is definitely a unique and distinctive one. So uh thank you, Kai, appreciate it. Thanks, Michael. Cheers, everybody. Give us seconds.