Yet Another Value Podcast - Advanced Portfolio Management: A Quant's Guide for Fundamental Investors (Fintwit Book Club)

Episode Date: March 5, 2025

Welcome to the latest edition of Yet Another Value Podcast's Book Club. Once a month, Andrew and co-host, Byrne Hobart, will discuss their thoughts on the book, "Advanced Portfolio Management: A Quant's Guide for Fundamental Investors" by Giuseppe A. Paleologo.See Byrne's writing at: https://www.thediff.co/"Advanced Portfolio Management: A Quant's Guide for Fundamental Investors" on Amazon: https://www.amazon.com/Advanced-Portfolio-Management-Fundamental-Investors/dp/1119789796Chapters:[0:00] Introduction + Episode sponsor: Fintool[2:24] First thoughts and overall impressions of "Advanced Portfolio Management"[5:54] Which pieces do you use to implement into your investing process that Byrne picked up from this book / how they implement "stop-losses"[19:00] It's not enough to have great ideas / "it only takes one" vs. concentrated portfolio / making good calls over time[29:17] Is there anything to buying things that are classified wrong in order to generate alpha?[32:54] How much are these factor and pod-shop models gameable[40:23] How much does Byrne think matching uncorrelated data is going to be taken over by AI vs. fundamental investors going forward[48:25] Beating the bots / final thoughtsToday's sponsor: FintoolFintool is ChatGPT for SEC Filings and earnings calls. Are you still doing keyword searches and going to the individual filing and using control F? That’s the old way of doing things before AI. With Fintool, you can ask any question and it’s going to automatically generate the best answer. So they may pull from a portion of an earnings call, or a 10k, whatever it may be and then answer your question. The best part- every portion of the answer is cited with the source document.Now- if you’ve tried to do any of this in ChatGPT you may know that the answers are often wrong or hallucinations. The way Fintool is able to outperform ChatGPT is their focus on the SEC filings. If you’re an analyst or a portfolio manager at a hedge fund, check them out at https://fintool.com?utm_source=substack&utm_campaign=yavb&utm_content=podcast280See our legal disclaimer here: https://www.yetanothervalueblog.com/p/legal-and-disclaimer

Today's podcast is sponsored by FinTool. Still control effing through SEC filings? There's a better way. Meet FinTool, the chat GPT for SEC filings and earnings transcripts. Just ask your question, and FinTool pulls the best answer from relevant documents, whether it's a snippet from an earnings call or a key point from a filing. The best part, every answer comes with sources, so you know exactly where the information came from.

That's their read, but let me give you a little bit of my experience. I've been really interested in gaming stocks recently. So I was looking at Caesars, and Caesars has been built up through a slew of mergers over the year. I wanted to see their thoughts on how the synergies from these mergers had evolved to kind of judge them on these mergers. Previously, that would have been a half-day project that need to pull up every earnings call for the past five years. I need to go through it, control that for synergies, probably read through it because sometimes they don't say synergies directly. They say like merger benefits or something. And I'd have to go through all of them and see how the thoughts on synergies had evolved.

But with FinTool, it only takes two seconds. I just say, what did Caesar say about their El Dorado deal synergies? And FinTool gives me a quote and a link to the source. In this case, it points me to Caesar's 2023 proxy. That's noted they'd exceeded their stated synergy target. And then if I want to do more, I can ask them some different questions and find out more about how the synergies have evolved over time. Good luck finding that with Control F in just two seconds.

So look, if you're an analyst or a portfolio manager, try FinTool today at finTool.com and take your research to the next level. That's fintool.com. Byrne Hobart from the Diff's Monthly Book Club. I'm Andrew Walker, one of the co-hosts of this. With me today is from the Diff, one of my favorite newsletters. One of the few things I read, 85% of the days, I would say. My friend, Byrne Hobart, Bernd, how's it going?

Hey, great to be here. I always wonder what the right reader frequency stat to track is, incidentally, because I feel like what you want is an audience that will not actually, an audience that is too busy doing cool things to actually read 100% of what any one person writes. And then within that audience, you want to get the highest reader percentage you possibly can. It's a great point.

You know, if somebody's reading 100% either they're not busy enough or you're not publishing enough, then possibly it's both. But look, I love it. I also listen to not sponsor competitive podcast, but I listen to the Rift pretty religiously too, which I can just hear your deeper thoughts on it. But any man who can bring down a multi-billion dollar domestic bank

purely through the power of his newsletter, you have to pay attention to what you're saying. Anyway, now they're here nor there, book we're going to talk about advanced portfolio management. How do you say the author's name? Do you know? I think he has alluded to being called Gappy a lot, and so that tells me that I should just call him Gapy. Great. We'll call him that. Link in the show notes, if anyone hasn't read it yet, and wants to buy the book. But let's start. First Thoughts, Overall Impressions.

What did you think of the book? I thought it was great. I thought it was a just really clean explanation for, why the factor-neutral model works as well as it does and the context where it does. And there's a lot to unpack there, but it is, I thought it was just a really good explanation of how stock picking can work in a context where we recognize that if you pick one, if you buy one stock, you do get some kind of excess return from the fact that you picked Nvidia and someone else picked Apple or vice versa. But also that a lot of what drives those stock prices is something that is not

specific to that company and um you know sometimes people just get lucky and if you uh like i forget who originally made the joke that um the best financial decision you could ever make in your life was just get a job in finance probably in bonds in the mid 1970s like anyone who did that anyone who first got their first job out of college working in something that was bond facing or stock facing between like 1975 and 1985 they were just set for life um and then then there's the, you know, once that's true, investors start to ask themselves, why are we paying someone for

mostly being lucky? So if your returns, if someone has an exceptional career over multiple decades and they earn 13% a year buying, you know, owning stocks and the market did 10% a year over the same period, they're getting paid on that 13%

in many cases, but they actually deserve credit for about three points of that. And so this kind of explains some of the rise of indexing. If we're paying these fees for people who don't even necessarily get the market or often don't beat the market, why don't we just buy the market as a whole? But I think the other way to take it seriously is say, okay, if you are skilled, you want to minimize luck because you want to get paid on skill. And if you can measure your skill well and you can deliver a turn stream that is

as pure skill as possible, then you actually deserve a much larger cut of that than before. And these trends are all, you can't not be aware of them if you just regularly read the Wall Street Journal Financial Times in Bloomberg. You will see stories about larger funds, you'll read about them being factor neutral, and you'll have some awareness of what that is. You'll know that indexing is getting more popular, et cetera. So, like, it's talking

about a lot of trends that exist, and what it actually does is walk through the math of why you would do it that way. And it explains a lot of the idiosyncrasies of how these firms are structured. And it's also just, there's this sort of, there's this vibe that a lot of people

have where just, they've in a field for a while. They've done well in that field. And they've basically seen enough things that they have recurring jokes and, you know, observations and just things that give you the vibe that they've, uh, they have a lot of experience and kind of know what they're doing. And I felt like that was, that was another perk of this book was that you, when you're reading this book, you are reading the, the output of thousands of conversations with risk takers and with other risk managers and probably with LPs and at least, you know, people who allocate capital. So you're getting a lot of condensed wisdom from this book. No, look, I would just

and everything. You said, I was really skeptical when you chose this. I was like, look, what does that have to do you? But then when you read it, like a lot of it, a lot of it is applicable, at least my understanding of the reading. And I think he kind of knew that because he's got so many different things where he kind of talks about, hey, even if you're not like in the risk management shop at a pod chap, like here's why you need to be thinking about and here's how you can be improving it. And I'll tell you just like so many, I knew this book was going to hit with me because there was one specific line in the first chapter. He said, hey, if you collect all the trophies when you're playing a video game, here's exactly the parts of the books where you're

going to read. And I was like, oh, man, this guy knew me. He knew I was coming to read this book and he knew exactly what he wanted to. But let me start with a broad question. I think my listenership, your reader are individual investors and some, most of viewers, to individual investors in some way, shape, or form, or maybe, like, running a concentrated fundamental fund. So I just want to ask you, when you read this book, like, which pieces do you think of, actually one specific question was, which piece do you and Byrne use to implement into your investing that you kind of picked up from this book?

So I'll just toss it over to you. Yeah. So one of the things that this book really clarified for me is that I should be, even if I'm not trying to neutralize my exposure to every factor, I should be thinking about it. about what factors I have exposure to and also what factors I will just tend to have exposure to if I have a particular process. And the process is not, okay, I'm going to sort everything by momentum, but part of the process is, okay, I know that I have a tendency to double down on things when I'm obviously wrong. And most of my big losses started out as small losses. And I felt, you know,

really, you know, I felt like I would be totally vindicated if I just double the size of the position after it was down 10%. And then I have a larger position that goes down another, you know, It was down 15% after that. So for me, it was, part of what I realized was, okay, so I've kind of dealt with that by just trying to be more aggressive with stop losses. But if you do that, what do you end up with? You end up with a portfolio that is always by default long momentum. It always has momentum exposure because you're stopping out of everything. Tell us.

Can I pause you there? The stop losses piece was one of the most interesting piece of the book. And I'd just be curious. You said it sounds like you're implementing stop losses in some way or form. And I'm thinking about it. how are you implementing stop losses? Yeah, so I do the kind of lazy thing where I just have a mental stop loss. And it's like the rigorous way to do it would be to write out exactly what my thesis is and to write out what would prove it, what would disprove it,

in what case is the absence of evidence going to disprove it. And I think this actually gets to part of why stop losses work and where they work, which is that the more that you are making a bet on very near-term. changes in sentiment, the more what the stock does immediately after that tells you whether or not you were right. Whereas if you are a deep value investor and, you know, you are the person who is somehow digging up the last annual report that this company ever issued 15 years ago and figuring out what real estate they own and you're looking at the real estate records and finding out did they sell the building or not. And then, you know, you're buying, you're buying these random pink sheets companies at, you know, one-tenth of what the real estate is. There's also no momentum, right?

like your stop. Yeah, there's no momentum and there's like there's no real information from the stock price moving other than probably, you know, I get the impression that respect for MNPI is a lot looser in that corner of the market. So yeah, if the stock doubles, it is probably going to keep going up. But other than that, like you have very little new information there. But if you have some view on Nvidia and you think that this is the cadence of new model releases and this is how fast people will be building out data centers.

And you also think, okay, here's what investors are getting wrong about this. You get information every day on whether investors are converging on your viewpoint or whether they are developing stronger conviction that something else is true. And if you did map out all the right variables, if you did the pod shop approach of figuring out everything that moves the stock and then getting a variant view on a handful of those things, and if you did all of that and the stock is not moving in the direction you expect, then either you don't have a good model for how changes in fundamentals affect sentiment,

or you don't have a good read on those fundamentals. No, it's a great point. I'll use a hybrid of your two to do the ones like I struggle with the most. So something like DISH Networks, it's now Ecosar, Sats is the ticker if anybody wants it. I do not have a position. But, you know, that is a play on Charlie Ergen figuring out a way to monetize the spectrum. Now, whether he monetize it in a sell-up for $100 billion and everybody's happy, or sell it for $20 billion, but dividend it out through the bondholders in the way, you know,

maybe the market's a little soft or maybe just something there. It's not like in video where, as you mentioned, In video, a lot of it is people are playing games with, hey, a new $500 billion data centers coming in or like there are very short-term checks, whereas something like DISH, I don't know, like you could go 25 to 20 in a heartbeat, and then I'm kind of like, hey, am I stopping out? Or do, am I stopped out if you're kind of applying about a 20% or should I be doubling down because there's been no new news, but obviously you go tell people that a 20% move wasn't a big deal. I promise you 20% is about where it really starts to start. You know, those are the types

of things where I'm like, how can you as a fundamental investor implement kind of stop losses when you might just be selling out to some of your top ideas on random noise? And I'm not going to say an earnings puked down 20% like most earnings pukes down 20% is because there's something going on. But, you know, it feels like as a fundamental investor, you're always getting punched out. But then the counter is, I don't know many stocks that I've written down 50% that I've ever ended up making money on, you know? So it feels like there's somewhere in between. I just struggle with that. Yeah. Like, I think, I think it does get down to the question of how much are you trying to understand a business and buy a piece of a business that you'd

like to own for a very long or indefinite period, in which case you probably actually want to bet more on mean reversion, especially if part of why you own the stock is you think that this management team is going to be very diligent about buying back shares when the stock is cheap and then when the stock is not cheap, that you think they have opportunities to reinvest in the business. In a case like that, your expected return goes up when the stock price goes down. But this is also a classic way for a lot of value investors to blow up is that they keep telling themselves, you know, I thought this was a dollar bill trading for 70 cents and now it's a dollar trading for 50 cents. So I need to take it from 10% of my portfolio to a quarter of my

portfolio. And if you're right, that's great. But a lot of the time when you feel your, when you start to come up with that kind of justification, it's just, I guess like the way I think about it is if you are losing money on a trade, that is information that you have that you don't quite know what drives the stock. Because even if you are long biased or long only, and, you know, you're right about the company. If you understood kind of how the news flow is shaping it up and you deeply understood what parts of the macro economy this company is exposed to, et cetera, you probably would have just waited to buy it until it went down.

And, you know, there are a lot of companies. I actually, I had this watch list that's very, my most annoying watch list because I put together a couple months ago, I said, you know, I've looked at a lot of high quality companies where my thought was, this is really interesting and I should just buy it later. Like, I should buy it when they have a bad quarter because I think they will probably recover from that. And so late last year, I started putting together that list and then I forgot about it for a couple months. And I looked back at the list. And it was like two companies, like one had gone up 50 percent. And the other had gone up 20 percent. And it was immensely

frustrating to me. But, you know, when you're, I guess when you, if you are losing money on something and your first thought is, I'm going to make even more money on this, it's very likely that you're not thinking about is what, what could I have done to avoid this loss and where, you know, what is the actual distribution of outcomes that I'm looking for? Like my last, my last big loss, well, my last big loss was actually a company where they just had a, an uncertain event that just resolved in an unpleasant way and the stock dropped by half in a day. You may know this one. But, uh, wait, no, I wonder what it is now. Target hospitality. Oh, yeah. Yeah, there was, uh, no position anymore, but yeah, I know that one, uh, extremely well.

Yeah, yeah. So there was, but then the one before that was that I lost a lot of money on very quickly was a short position in a particularly scammy category of companies. And I just wasn't covering fast enough. And it went, went pretty vertical. The stock went from 7 to 50 in about a week and a half. Look, that's what short positions especially, like, that's what really breaks me. Because if you suck your thumb a little bit when a long position moves 20% against you, like, it's got. and smaller, right? So if you're not adding on the way down, like, at least you can suck. You have to be so glued to it. It's one of the tough things about shorts. Yeah. And this is actually a case where I feel like the podshops are in, they have the right vibe for handling a portfolio that has, like, probably in terms of a number of positions, it's mostly short positions. And so a lot of what they're doing day to day is figuring out the next thing to short. And if they are just constantly adjusting their portfolio and constantly

trying to get closer to being on target in terms of their exposure to different factors, then they are pretty much automatically covering the short positions that are moving against them and pressing the short positions that are working. And that is sort of what you're supposed to do. And it also does depend on whether this is a short position in the sense of stocks at 20, and I think that demand is getting softer in this industry and the stock will probably be at 15 and six months versus stocks at 20, it's a fraud, CEO should be in prison. I The only question is, how much do I pay and borrow costs before it goes to zero? And there are companies like that where I just, I look at them as entertainment because the borrow is too high.

But, yeah, they're, so in cases like that, like it is, it is really annoying to recalibrate because you do feel like you are, you're giving somebody else money when you cover part of this position because it's gone up and your conviction should be higher. But also, you don't really, nobody gets credit for finding the stock at 10 and really, realizing it's going to zero, but first it's going to go to 200. You don't really get credit for having been earlier recognizing it as a short if it just blew up beforehand. My favorite, my favorite one of all time is the, I think it was about GameStop. Somebody was like, hey, GameStop's at 50. You sell the $100 calls next to $2. And it might have been six weeks. And six weeks later, you came closing the leap on this, closing this out for a profitable trade, and GameStop had gone from 50 to 600 in the meantime.

It's like, hey, so you sold the $100 calls for two. So at peak, you were down for a little, but, you know, you died on the way there. Yeah, yeah. And that is, I guess that's another part of the Podshot model is that they are looking at what is the path that gets you to that long-term return. And they care about that a lot, one, because it is, it just does make more sense to compare a bunch of different strategies by looking at their volatility and looking at how correlated those volatilities are. But two, there just is a meaningful difference, especially when you lever up, between 15% plus or minus 5 and 15% plus or minus 40. And I think that if you look back at financial media coverage of good investors in, say, the 90s or the early 2000s, in a lot of cases, these people just really like beta. They were owning the more volatile slice of the market, and they did really well and the market did well and did badly when the market did badly.

And they actually had a nice narrative there, which was that they do get good returns over time and that the investors who keep the faith with them and don't redeem when the market's down by a third and they're down by two thirds, the ones who don't redeem do make money. And then I would suspect that for a lot of them, if you ran a regression on just what would you get from investing with them versus just using S&P futures to get 200% long exposure to the market, you'd probably find that, okay, they were like as good as being 2x along the market, but a little bit more volatile. No, no, I completely agree. A reminder that today's podcast is sponsored by fintool.com, the chat GPT for SEC bylings and earnings transcripts. Just ask your question, and FinTool will pull the best answer from the relevant documents, whether it's a snippet from an earnings call or a key point from a filing. If you're an analyst or portfolio manager, try FinTool today at fintool.com and take your research to the next level. That's fintool.com.

Let me go. The thing I thought was the most interesting question here. I love this line. He used to, if I remember correctly, twice in the book. His line was basically, it's not enough to have great ideas. He says it in a few different ways in the books. But that's almost a direct quote twice. Another one, having good ideas is useless without the knowledge of how it's turned them into the money.

That's actually saying something different. But he's basically saying it's not enough to have ideas and you have to have portfolio and risk management on your side. And I did think that was interesting because like Charlie Munger is famous for the it only takes one, right? You have one great idea, you plow it all into your one great idea, and that's all it takes. And both, I think, have a lot of elements of correctness. But when you mash the two together, like, they couldn't be more divergent. And this does come back a little bit to the pod shop versus running as, you know, six sock concentrated book divergence.

But I thought and I was just trying to marry two divergent thoughts in my head. Yeah, I think it is, that's actually probably just a realistic look at how the world works in general, that you can look back at someone's career and you can say, well, if I had made the one big decision they had made, that I would be very, very successful. If I had realized that electric cars were feasible in 2008, or if I had realized that the PC is going to be a really big deal in 1975, if I realized Bitcoin, it didn't matter that you couldn't use it to transact. All that mattered was it was digital gold,

and there'd be these little memes of doges. Like, who knew? Yeah, but then you can go back and find that actually a lot of people had that same key insight, and the people who executed well on those insights were often just making a lot of iterated good decisions along the way. I forget if I use this in our last chat, but it's something I've been obsessed

with. There was this thought experiment that was going around rationalist Twitter many years ago, which was something about how if you have a time machine, you can send a message to yourself 10 years in the past, but you can only send a certain number of bits of information. What do you send to make as much money as possible?

And I thought about that in light of the Winglewost twins, because you could imagine them actually getting this message from their future selves, the time machine saying, okay, the two best conceivable investments you can make in the early 2000s are, one, by Bitcoin, and they got that one right, and two, by as much equity as you can in whatever social network Mark Zuckerberg is working on. So they nailed it. Like they were, they were investing in, they were trying to max out their equity position in the Mark Zuckerberg social network

before anyone else realized that there was a way to make a fortune. And yet, they did not perfectly execute on that and ended up not making as much money as they probably could have. And, you know, it's obviously a very, very carefully chosen example. But I think it is, it is revealing. And it's especially revealing because there is this narrative fallacy around investors where you look at their one best idea. But you also have to ask, how were they in a position to actually monetize that idea? Like, if you had the same idea, but, But, you know, the same idea at the same time, but also at that time, your net worth was $1,000 in your checking account and you were otherwise pretty much broke. Like, yeah, you could have multiplied that money a lot by putting some of it into Netflix.

But being in a position where you can actually have a material stake in a company means having made a series of pretty good, but not, you know, not narrative bias creating decisions beforehand. And also, I think that there's natural resistance to that idea that coming up with good trading ideas is nice. It is an important ingredient, but it's actually not the key thing. The key thing is coming up with a portfolio construction that gets you the most value out of those. Because I think people who get into finance and in particular get into the capital market side and certainly to stock picking, they don't dream of coming up with the right way to diversify. a portfolio and they don't dream of coming up with the right cutoff where you figure out whether your 20th long idea is going to be a net contributor to expected to your to sharp ratio

or if it doesn't quite hit that threshold. What people dream of is being the person who figured out that Netflix was a good buy in 2011 or being the person who bought Google right at the IPO or Visa or MasterCard at the IPO, et cetera, or, you know, someone who figured out that GameStop was going to rock it up. That's what gets people excited, is those, individual discrete decisions. But if you're running a fund and you're selling this stream of returns, what you're actually selling is the output of a process where there shouldn't really be one big winning decision because those don't come out every year. And so you can't deliver consistent year-to-year returns with one big win. You know, it's interesting you say that.

So Ackman has been in the market with this Howard Hughes bidded recently. And I've owned it recently in the past. I do not currently own it. But I do think it's interesting. You say that because I said he was going to offer to take them private. And then it's morphed into this thing where he said, hey, your stock was at 70 unaffected. I'll put in almost a billion dollars at 90. But then you give me a management contract and I charged 1.5% of the equity market cap. And as I'm saying it, I was worried Byrne didn't notice it. But as I'm saying, I remember, Burn had one of the funniest lines on it where, well, I'll let you tell the line if you want.

But I'm sure by it because Ackman, if you think about it, he's made some killer decisions over the past 10 years. He's held Chipotle all the way up, several others. But the real reason his returns are even passable over the past 10 years is because he nailed an inflation trade in 2022, and he made maybe the best investment of all time with the COVID puts and the COVID CDS that he did. And it's just interesting because he wants, that's what he wants to monetize when he's going to Howard Hughes and say, I'll do the management contract, right? A, I'll get deals with B.

When it's time, I'm going to have some really interesting trade. I'm going to put you in. and we're going to make multiples and doubt pay for everything. And I'm just struck with that because that's what he's trying to manage, but a comment's going to hit the earth, and I'm going to be able to monetize that in some way, shape, or form. Yeah. So I guess, you know, in one sense it makes sense to backtrack slightly and say that in

some domains, you do want to make a handful of really good calls over your career. Like if you're doing early stage investing, it is pretty much, did you invest in, Stripe, did you invest in Databricks? Like, if you got one of those deals, then you were pretty much set. On the other hand, there's a big difference between I was able to put 5K into the seed round for this company, or I was able to put 500K into a very early round in the same company at a similar valuation. You do get very different outcomes from that. I forget who I was reading about.

It was one of the Huffington Post people where he'd been co-founder of Huffington Post and had also done a very early investment in Uber. and actually made more money from his 5K check into Uber than he made from co-founding and playing an important role in running the Huffington Post. So in cases like that, yeah, you have these nonlinear returns. And in macro, you have this combination of one, there are these one-off opportunities where we just, we didn't have, like, the only data we had on what do you do during a pandemic is you buy anything with China exposure because it's all going to bounce back. And that was not, that was not the COVID trade.

it was a COVID trade at various times. So it was not the main one. So in cases like that, like it does come down to making the right call in this one special circumstance. But I still think that thinking about being in a position to make that call,

like if your destiny is to predict some out-of-left-field recession and just make a ridiculous amount of money on CDS and deeply out-of-the-money index puts and whatever, the biggest impact that that decision will have on your net worth comes down to how much capital do you have to deploy in that trade and how much capital do you have access to. So even if you do, even if you are the kind of person who can just swing for fences like

that, the way to maximize the value of that is to have a good track record with these smaller scale bets that maybe use a similar kind of thought process. But then there's also this paradox of alpha where in the book talks a little bit about this, there are some strategies that used to be alpha. You used to have to pay someone to implement these strategies, and now the strategies are just wrapped in an ETF that costs tens of basis points if it's expensive to execute that strategy. So things like just buying value stocks that used to be a fairly difficult thing.

You had to get the Moody's manual. You had to page through it. I see this all the time. Somebody comes to me and they're like, hey, this stock is really cheap. It trades for eight times per starting. I'm like, hey, in the 1970s, that was. a grade analysis. Now there's a computer that's done that. So unless you've got, you know, a little bit more than that, maybe on the whole you're right, maybe you're not, but you've

provided no edge there, you know, like just saying that is not an edge. Yeah. I've almost felt like the right way to do stock screens at this point is screen for everything you don't like and then look at a random selection of companies and figure out, okay, which of these companies, you know, which of these low margin companies could actually inflect to being high margin or which of these no growth companies could actually start growing. I actually had a question slash insight on that that I wanted to ask you. So I'll just switch to that. He mentions when he's the MSCI, Standard & Poor's, there's four levels,

and then there's a bunch of different subsections beyond it, right? My question is, I will, like, he's talking about that in terms of using it to, you know, you've got Apple, Microsoft. If you buy Apple, you might need to shorts of Microsoft to get some of the, some of that factor out. My question to you is, I had this to be. a lot with people where they'll be looking at a company and they'll be like, hey, 60% of its revenue is from what's the lowest multiple thing you can do? Producing coal. And 40% of its

revenue is from AI generation. And one of the issues, they don't get labeled with the AI objects and then like all the pod shop money will be able to rush in or something. So I guess what I'm asking you here is you were talking about, hey, one of the ways you might make money is screen for the things you hate is do you think there is anything to buy things that are classified wrong in order to generate alpha? Or is that, again, it's not like the Jigs classification. It's not like this is unknown. Do you think people are diluting themselves and thinking, I've got a little edgy way to do this thing? I think what you, probably the synthesis of this is just analyze a lot of companies and have some process where either you have some source of randomness

in what you look at or you are just constantly traversing the graph of companies where, let's say you spend a lot of time on AI, and your big uncertain question is how fast does this get implemented on the enterprise side? And then you start looking at a bunch of enterprise AI users, and maybe you find that one of them is actually not just using chat GPT in smarter ways than everybody else, but has actually been doing a bunch of other smart things. And so your AI thesis turns into a, you know, I'm going to buy this well-managed industrial conglomerate over the other ones kind of thesis. I think you do want some of that randomness. And I think a lot of the, a lot of the strategies, you know, like I was, I was being a little bit cute and I said, okay, yeah,

you just screen for everything you hate and then find companies. But it is, it is the case that everything you can screen for is in someone is trying to price that in. And they might be mispricing it in. And there can be cases where when you screen for high margin, what you look for within that is some qualitative view that margins are not just high, but actually going keep going up. And then then you have a very pod shoppy kind of thesis, because not only do you have a fundamental view, and not only is that a variant view that you can sort of underwrite through whatever data you're able to gather, but you actually have a view for how that flows through into changes in the price. So instead of just saying, I like the stock and here's my DCF,

you say, okay, people are valuating, people are valuing this on EBITDA and here's what they think EBIT will be at the end of this year. Here's why I think it will be higher. And so here's my event path for getting to a higher number. And by the way, once they hit that higher number, they will probably be at a higher multiple. So we get to win on both sides of that. I think that kind of thought process where you're basically, you're taking it as a given,

you're taking the things you screen for as a given, and then you're saying, okay, but how are these things changing and can I have an edge in predicting those changes? I think that is pretty valuable. Let me switch. But back to the point I was getting to with one-off trades and the nature of alpha.

There are a lot of trades where they were repeatable but required a lot of manual work. They have since become more commoditized. And so those, I think, used to be alpha. They're now just a different flavor of beta. And as we go through the list of strategies you could turn into purely systematic strategies, what we're left was the stuff where you can't even really describe it. Like there are things like, say, the mortgage short or the magnetar. trade and mortgages. Like, you can't really describe that as a repeatable thing, as a kind of thing

that people do. All you can describe it as is you want to be smart, pay attention what's going on in the world, find things that don't make sense, find the best levered way to express that they don't make sense, and either find, either get the timing right or find a way to make this trade with little negative carry or ideally positive carry, which is part of what Magnetard did incredibly well is that they actually, they figured out this stuff is unsustainable, but they also figured out a way to earn positive carry while betting on it. And that meant that they did not have to actually call the top in housing. And they also didn't have to figure out exactly how the bubble falls apart. Let me, let me switch into a slightly different topic. Over the past eight weeks, I think you were one of the first people to kind of really note this in the market. But you saw a lot, there was the deep seek news, right? And a lot of the AI stocks fell on that. But interestingly, what fell more than the NVIDIA or the people who were directly exposed for it was the power plays, you know, the IPPs that had nukes that were going to have this unlimited demand from data. And you were one of the first people noted it and you're like, hey, if you're a pod shop

and you were running a fund, there was a limit on how much NVIDIA you could buy and chip exposure. And a way to increase your AI exposure even more was to go buy these utilities that, you know, you weren't getting penalized on the risk factor. In fact, you were probably getting like, Invidia might be five units of risk and a utility might be one unit of risk, so you might have been able to lever it up five times more or something. How much just in your experience do you think these factor in pod shop models, if you're on the inside of them, how much are they gamable by that type of,

by that type of thing? Because one thing I do always kind of get the sense of when I talk to my pod shop friends is, hey, I like this company over this company. And one of the reasons is like, it gets me a backdoor play that gets me a little bit more beta than I'm going to get charged for or something. So I always, I think it's a good idea to be really cautious about that because you read a book like this, everything makes perfect sense, and you're like, okay, I have a model in my head for how these companies operate. But of course, by the time the book got written, it was slightly obsolete. And the specific way that it gets obsolete is that people

are constantly trying to figure out how to the game is because it is just fundamentally true that if you give someone a cut of P&L and it's not their money that's invested, you've given them a call option. And you can try to construct whatever constraints and other incentives and things you want, but you've still given them a call option on the performance of their portfolio. So they do have an incentive to seek volatility. And also, you've hired people, you know, you've had this very rigorous hiring process and the people you've hired are really, really high performers. They basically haven't had any big career mistakes or if they have, they're actually really impressive, really original mistakes. And so you're selecting for people

who also have really high ego. And so you're basically hiring people. people who are all going to think to themselves, these risk rules are meant for people dumber than me. And if I find some edge case where not only do I have a good trade, but I have a trade wrap outsmarted the risk system, of course I'm going to take it. But then you hire all your risk people to stop them from doing that. And from the perspective of the actual manager of that fund, someone who's doing a trade like that is basically stealing.

They're stealing office supplies. In this case, the office supply is like market, you know, is beta, you know, beta, you know, back to exposure, whatever. So, um, always good to be cautious on that. So like, I don't, I don't know for sure who would have the, who would have been positioned that way exactly, but it is absolutely true that if you are, you're looking at these stocks from the perspective of their, they're, they're pretty low, you know, pretty low volatility. We know what their exposures are. It's a kind of quiet corner of the market that, yeah, if you, if you have some variant view on how much their profits will grow, um, like how much demand will grow,

then you have a really, really clean trade. But also, if, if you have some, if, If you were thinking that way, and everyone else who's analyzing these companies is thinking in a totally different way, and you and your peers who are long these particular utilities for the AI trade push them up 20 or 30% in a couple months, there isn't a utilities-focused buyer who is waiting for a 5% or 10% pullback. There's just someone mystified that you are so wildly bullish on this, and they're not even going to pay attention until it's backed down to where it was six months before. So I think there was like an air gap there where there were just not very many people who were somewhat optimistic about the AI utilities. Because to bet on the utilities trade, you are betting on the situational awareness paper model of the world. We're going to have this massive deployment, the scaling laws will hold, and there will be so much value created that we will increase electricity consumption in the U.S. by a third over a fairly short period. No, I completely agree with you. I guess the reason I asked was I do wonder how much being a successful pod shop manager,

there's a lot of different ways, but I wonder how much over the past two years was, hey, I have this bullish view on AI, and I'm going to express it in every way I can. And the easiest way is I buy Nvidia, I buy Microsoft, whatever, right? But the system limits me. The other, whatever it is in my book, I find ways to get exposure to the AI play. even if it's a little clunky or something. And then when I am right, I get rewarded. But as you said, you have a hidden AI beta because AI,

it's probably a factor now, but at the time you're putting it on, I wonder how much is like kind of filing. It sounds trend following, but, you know, kind of I've got one view. I put it all on and I do it in a way that the system doesn't realize that's what I'm taking the exposure to. So I guess the steelman argument against my, yeah, AI is a factor. People clearly weren't hedging it, which I think is broadly true. But I think the population of people who weren't hedging it might not be the pot shops. If there were some set of investors who were actually coming up with, who were the first ones to put a label on new factors and not just call them momentum, it's probably at one of the top or all of the top pod shops.

But even setting that aside, you could also say that part of the job of a pod shop manager is identify emerging factors, bet on those factors when they are being under-exploited. And then you capture some of the upside from people realizing that AI is a theme that we can allocate to, just like oil, you know, like midstream energy is a theme we can allocate to. And if you were early to that, the other thing is you're in a really good position in terms of information gathering. Like you, you have been following invidia slightly longer than many of your peers, and you have been thinking of it as an AI play slightly longer than many other people. and you, the experts who are being talked to now or the people you talked to three months ago and at conferences, you know, when you listen to the questions, you can kind of figure out, okay, how many quarters behind me are people in terms of understanding this thesis. And then when, so what that means is that you actually do have to have this very, very zen attitude,

which is basically if you have a portfolio where you're tilted towards a bet on one theme, and you understand that people are waking up to that theme and that's a big deal, the time to actually exit that trade is when you go to a conference and you learn a lot and you hear really great questions you hadn't thought of, because that means that other people are actually thinking a little bit past where you are, and their positioning is based on, their positioning is the position you'd have if that were the next question you were asking. So if you start to hear really, really good questions on fundamental feces that you're involved in,

then you know that at that point, your alpha has completed most of this evolution to beta. That is a really fascinating way to think about it. And one, look, I'm not saying I'm the smartest person in the world, but there's been one or two companies where I've really known it. And, you know, somebody would be like, oh, I just made this big position. I'm talking about things I'm an expert, but it's interesting to say, hey, when all of your conversations with peers are kind of asking the same questions, that is when it's played out. It's when people are asking what to you are the basic facts. That's when you have the most potential edge. That's a really fascinating way of framing it.

Let me ask one last question here. There's one line that jumped out to me that people, this book was one line jumps out to me. It is, this is a quote for our listeners. The ability to combine these alpha forecasts in non-trivial ways from a variety of sources and to process a large number of unstructured data is a competitive advantage of fundamental investing and one that will not soon go away. That's a really interesting line to me, because if I'm remembering correctly, and this is written in the pre-COVID period, it's before AI. We have AI today, and it is getting markedly, markedly better. And you know, you're the one who I think

put it the best. A good macro person reads an article today about the Japanese yen. He matches it to an article he read a year ago about the BOV interest rate policy and he realized, oh my God, the yen's about to break out one way or the other. It's not. That's an advantage for fundamental human investors. I wonder, like, how much do you think matching uncorrelated data is going to be taken over by AI versus kind of fundamental investors going forward? Yeah, I think we will actually just develop a more elaborate taxonomy for what that looks like.

Like, there are already things where I used to do them with a lot of manual processes, and now I can do them a lot more and automate a lot more. I'm just looking at, no. There are just parts of my parts of my idea sourcing process where it used to involve just getting a long list of things to read and skimming a bunch of them

and now it is feeding a long list of things to read into one of the Open AI APIs and just getting a summary that is a lot quicker to go through. So for stuff like that where you have this sort of fixed process and you are looking like you know roughly what the frequency of needles is and you know the size of your haystack, in that case, you can't automate it pretty straightforwardly. I don't know exactly where the human in the loop thing is going to stick around.

But I think that part of what will happen is just it will be as if you had an extra 10 hours a day or an extra 50 hours a day to read. And you'll basically have the knowledge base that you'd have if you had almost infinite time to read and we're looking for particular things. Where you, what I do think gets tough is just getting the right level of serendipity because whenever you are trying to automate some process,

you're implicitly saying that you know roughly what that process is. And in some cases, you don't really know what the process is until you've done it manually. And I think this will actually be an interesting barrier or like an interesting trait to look for among new investors who started doing fundamental analysis after LLMs became available and who are used to the idea that there are some things that you read yourself, there are some things that Claude reads for you and, you know, there's a whole mix of different kinds of content is that it will be, I think, harder for them to just do the boring work. and just like, you know, when they're learning an industry, just read a bunch of 10Ks of different companies in the industry and try to understand it because they know that they can get a summary, but they also know that if you just ask an LLM to summarize a 10K, it's probably going to give you a pretty high level bullet pointy description of what the company does and may tell you a little bit how they're growing, et cetera, but you won't know what the weird distinctions are between companies until you've read, you know, your 11th 10K and you realize, wait, this is the first. time that I've ever heard a company mentioned this thing. And I wonder if it's unique to this company or if nobody else talked about it or whatever. So I think that kind of thing, you still,

you basically want, you want a lot of tokens in your own personal context window, even though you have these other context windows. And so what it actually means is that a lot more people, and this was, I think, my first, my first piece on just, hey, AI is going to change a lot of the way that we work, which, let me see when I wrote that. it was called Working with a Co-Pilot. Working with a co-pilot. Okay, April, 2023. My basic argument was you have just been promoted to management.

Your direct reports are all electronic, but they are, you know, like anyone who works at a successful growing company, your direct reports are much less wise than you. They need a lot of specific instructions, but they're also a lot smarter than you, and they have just a higher energy level. like the company's standards have kept going up since you joined and so everyone who reports to you is just objectively better than you in every sense except they have less experience and therefore

they have worse judgment and so your job is like impart a lot of judgment but try to outsource as much of the actual cognition to them as possible because they will be better at it but then make sure you know enough about how they're thinking to be able to spot flaws in it so just yeah it changes the balance of a lot of things that that people do but I would go back to the the point about low multiple stocks and how it used to be alpha that you could just calculate this, hey, I actually calculated earnings and calculated the price and it looks like the PE is only eight, and that's no longer a source of alpha the way it once was. That didn't, that certainly did not reduce the amount of total time that people spend analyzing stocks. It just changed what they

spend that time on. And I think, yeah, also the world is just going to get more complicated as we have more, more, like, the AI tools make it faster to analyze the world as it is, but also make the world a lot more complicated. So there are a lot of things that are just going to be harder to measure. There will be this fuzzier, like think of things like the network effects in a social network. That used to be a pretty straightforward concept to think about. It's a lot fuzzier now if people are increasingly interacting with LLMs and using LLMs to produce comments because at one level it means there's more user-to-user interaction because the

LOM will just suggest something for you to say in response to your friend's status update on Facebook. Like they will just have these pre-filled options for you, which I always like, I always used to have this sensation where if, especially if someone posted about just some really tragic life event, and I would read the comments and I'd want to say something to comfort them. And every single time I'd see a comment, I'd think to myself, wow, so sorry for your loss. Like, that is a really good one.

I wish I had come up with that one first. And then it's like, you know, the next one is like, oh, man, you're in our thoughts and prayers. I'm like, oh, man, what a good thing to say. I wish I had said that. And now the LLMs actually solved that for you. Like, you can do that sort of token indication. But that also means that when you're receiving that, you recognize that, you know, Lama 3.7 feels really, really bad that your dog died and that maybe your friends do too. But you're actually more confident that the, you know, you're actually experiencing this mediated through the AI with your friends kind of giving their stamp of approval on comments written by something else.

That's just a fuzzier concept of the network effect and of the connectivity that we get from social networks, which means we just have to think about what the social network model actually is in an AI context. I think a bunch of other industries will also just have to rethink just what do they do, what are the economic drivers, what do they actually charge for, et cetera. So even though the AI does speed up that process, it also, the existence of AI also means there is more stuff to learn and more stuff to figure out. and it's trickier to think about because the barrier between what's a deterministic computer process and what's this somewhat random, unpredictable human process, it's now a continuum. A reminder that today's podcast is sponsored by fintool.com, the chat GPT for SEC filings

and earnings transcripts. Just ask your question, and FinTool will pull the best answer from the relevant documents, whether it's a snippet from an earnings call or a key point from a filing. If you're an analyst or portfolio manager, try FinTool today at fintool.com and take your research to the next level. That's fintool.com. As you say this, it brings me back to our conversation on Jix or GCS or whatever. I wonder if in the future, just as you're saying, more investors are trained. And I think you're right. Like the investors who are just a few years younger than me, their first thing is everything goes into chat, GPT.

And I'm trying to train myself to do that, but it takes a while to digest, maybe a little more bespoke. I wonder if that is increased, do they demand a higher risk premium so the companies trade cheaper? Or do they are there, is there more alpha in there and kind of like finding something that doesn't code to? I guess I'm underestimated an AI's ability to rationalize because it's really gosh darn good at at this point. But I'm kind of thinking back to I know companies I've certainly been accused, hey, this doesn't screen well on Bloomberg and that's one of the reasons nobody gets it.

I could see a company that's, hey, their financials are formatted different than every other company. So when AI glances through its financials, it doesn't quite pick it off the same way. I'm not sure where I'm going with it. but I do wonder if there's like the beat the bots. Sorry, you cut out for a second. I was just saying rambling about beat the bots. Yeah.

Okay, cool. Yeah. So the tough thing there is that I don't actually know, I don't have a good model for what I'd be good at that an LLM is probably not better at. Like, one of the ways that LLM has really helped me is things like, hey, this company was cheap because they only produced financials and their annual reports in Japanese and it's all Japanese accounting. But now you just save the PDF and dump it into chat TPT, and you get a nice summary.

And so something like that, you know, that's fine. I think maybe SIN stocks, maybe the AIs are less willing to talk about that. Oh, that's an interesting one. Yeah. That's a really interesting one. Yeah. I'm sure. I'm sure chat TV is not going to say, hey, alcohol is dangerous.

I'm not going to give you a cocktail recipe. No. So it's, it will probably be able to analyze. around Foreman, just fine. Maybe for a porn company or something, it wouldn't do that, but I don't think there are any pure play foreign companies left. Like, Reddit is probably the closest thing to that, but that's not really the main driver of the business.

What else could, you know, cannabis, I'm sure Shad TPT is fine talking about the cannabis business. So, like, a lot of those, I think the LLMs are probably going to be perfectly okay with just telling you about the financials. And for everything else, like everything. it's almost like you want to invert your instincts, like everything that used to be inconvenient because it's not in digitized, easily searchable form, but you can do the schlep and find out about it. Now those are the things where actually the AI can do the

slot faster than you. So if the problem with, you know, let's say there's an interesting conglomerate that's one of those, you know, mini Berkshire, like great capital allocator type companies, but one of the problems with analyzing them is the CEO only, you know, he'll do podcast interviews, but the annual letter doesn't tell you much, but the podcast interviews a lot of depth and a lot of information on the thought process, you might think to yourself, I'm not going to listen to like 20 hours of podcast interviews to get five nuggets on capital allocation. But if I can just convert those into text, dump the text into the LLM choice, and say, hey, I'm an investor looking for information on this company's capital allocation approach,

stick through this transcript and tell me what they said about it, then you probably do get your answer. So suddenly those things become. efficiently priced really fast. And maybe the closest you get is you ask yourself, what would a big sophisticated fund that has an enormous tech budget, many full-time, many employees whose full-time job is make investment researchers more productive and allow them to ingest more data, like increase the number of tickers that one analyst,

one pot-shop analyst can cover is basically their mandate. You ask yourself, okay, what would the first couple things they build be? And then you build a kind of janky, hacky version of that, and you only use it for a company to the market cap of $300 million or less. And you're just pretty confident that no, 0.72 is probably not going to have one of their very, very expensive analysts looking at this nearly bankrupt clothing retailer or, you know, this like Bulgarian energy company or whatever. I'm laughing because I know a few funds who have argued, hey, what is our edge? And they'll be like, look, we apply credit card data and they'd say more than this, but it basically has to be. We apply credit from 500 to 500 million to 2 billion, and that's our edge.

Like, we're much more sophisticated and the big guys will not play in these small ponds. And so I'm laughing. And then you're kind of like, well, why would the big, why will the big guys not play there? Why is it enough liquidity for you and not them? But that is the pitch. Byrne, let's drive it up there because, A, our internet connection keeps getting a little sticky, so I don't want us to miss any big things. And B, we're almost an hour anyway.

This was awesome. You've given me a lot to think of this book. This book, I will tell you again, like there were so many pieces. For somebody who runs an ideas podcast to hear the line, it is not sufficient to have great ideas. Hit me a little bit in the chest, but at the same time I was like, no, like it was just a really thoughtful thing. And again, I was skeptical as a fundamental coming into it,

but I was really pleased with it. Burn Hobart, writes the diff. I read 85% of them. He tells me maybe I should get a little more busy and read a few less. But it's been awesome. Oh, if you've got suggestions, book three, we're going to do another one in March. So Lobbin, Book 3,

Byrne and I will have to read it and discuss it next week. We'll chat soon. All righty. Talk soon. A quick disclaimer. Nothing on this podcast should be considered an investment advice. Guests or the hosts may have positions in any of the stocks mentioned during this podcast. Please do your own work and consult a financial advisor. Thanks.