The “Card Counter” Formula That Beat Wall Street For 20 Years| SIH
Summary
Modeling and Markets: The guest emphasizes that mathematical models are powerful but limited tools, urging investors to understand assumptions and where models can fail.
Tail Risk: Drawing on Mandelbrot, he explains fat tails and the higher likelihood of extreme events than normal distributions imply, with implications for pricing and risk.
Options Pricing: A deep dive into Bachelier, Thorp, and Black-Scholes leads to how implied volatility is inferred and why assumptions matter in derivatives markets.
Volatility Smile: Post-1987, traders price in fatter tails, producing a volatility smile/surface that both relies on and contradicts Black-Scholes, offering insight into market expectations.
Risk Management: Survival and position sizing via Thorp’s Kelly criterion are highlighted as essential to avoid ruin, especially amid rare but severe drawdowns.
Passive Investing: The rise of passive ETFs may impair price discovery and increase concentration-driven risks, potentially causing sharp market dislocations.
Quant Strategies and Scale: Many quant approaches work at small AUM but degrade at large scale; similar scalability concerns apply to the growth of private markets and shrinking public market breadth.
AI Impact: AI is transforming coding, research workflows, and potentially labor markets, offering utility but raising displacement and educational integrity concerns.
Transcript
Get out your pens and pencils, folks. Today's guest is a really smart guy, not a finance guy. He's a physics guy who wrote a book called The Physics of Wall Street, James Owen Weatherall. It's a fun book. It's not a hard crazy technical book, and we're going to talk about a lot of the ideas in it, which are very important for investors. It's important to know historically where all this stuff is coming from, and all the crazy math that influences, I guarantee you, influences places [music] that you are putting your money in your 401k. So, pay attention, and and have fun with this, because this stuff is pure fun to learn about. So, let's [music] do it. Let's talk with our guest, James Owen Weatherall. Let's do it right now. James, welcome to the show. Thanks for being here, man. Oh, thanks so much for having me, Dan. We have to start with one with one thing that I need to tell you. I love your book, um, The Physics of Wall Street. Thanks. >> Now, I say that not having read the whole thing, but I've used this thing. I was trying to sit here and count the number of times, at least three times I've used it as a reference, because you do a fantastic job of writing about some of the charac- the great characters, um, in in math history and the history of sort of math and finance together. Um, Edward Thorp, Mandelbrot, Bachelier. I think you were like one of my primary sources when I wrote my story about Bachelier. There I think I had three sources, and you were one of the best ones. So, thank you for that. And one day >> Of you, yeah. I'm I'm glad I could be of help. Yes, one day I I hope to read the whole thing, just so you know. I'm sure [snorts] there's even more good stuff in there. I'll probably go back to it for some, but, you know, Pascal or whoever else you got in there. Um, so, I want to start there if I may. Like where did the book come from? Are you a Wall Street guy or are you a physics guy or both? I know I'm I'm 100% a physics guy. So, I was I was a grad student in physics at Stevens Institute of Technology which is in Hoboken, New Jersey. Yeah. Um right across the river from from Wall Street Yep. >> uh in um 2007 and 2008. Got you. >> Uh I had a lot I was I was like right out of college. I had a lot of friends who had been physics majors who had gone into finance. I mean some of them are still still working in the industry now. I'm sorry James. I'm sorry to interrupt you but I have to wonder like physics seems like this cool thing. What what are you doing the right now in the year 2026? Like what what cool stuff are you doing? You must be doing cool stuff. >> Yes. I So, I'm actually I mean I'm I'm not a physicist now. I'm a I'm a philosopher. I Yeah. So, so I I am a professor at in logic and philosophy of science. It's a niche department at University of California, Irvine in Southern California for almost 20 years now. Wow. >> And yeah. So, I mean I I work in philosophy of physics and sort of mathematical physics but also you know abstract science and stuff like that. >> That's cool. Very cool. Um so that makes a lot more the then now the book makes more sense to me. Right? Because what what do we what else does a mathematician or a physicist would a philosophize about except for this grand machine of Wall Street. And if you know Bachelier then you got the connection from the word go. Uh and away you go. Well, what what do you think of it? Like what When I read your book, like I said I just read pieces of it. So, I didn't get the full sweep of it. What does James Owen Weatherall think of of this combination of really from my perspective super advanced mathematics plus Wall Street where you really can't predict what the hell's going to happen next. What do you make of that? Yeah, I mean, so I am interested in it from so many different perspectives, right? So, I'm interested in the math. I mean, I just I like math. I like math applied to stuff. I'm I'm super interested in the history of applied mathematics. Mhm. I'm also just interested in in the history of big ideas. Uh and, you know, what could be more important in the history of the 20th century and now the 21st century than this this movement from, you know, these these big technical ideas from mathematical physics and and um related areas uh into financial markets. Um but also, you know, I'm interested in the philosophy, right? I mean, we're building these mathematical models to try to do something in the world. How are we supposed to think about that? Why does that work? Why does it sometimes not work? Um I mean, that that to me is just core philosophy of science and uh you know, applied in in a case that really matters. I'm glad you brought models up because, like, we just we lived through this thing, you know, whatever, 20 years ago almost, where people lost their homes. And that was at the end of a chain of events that started out with with risk models that didn't work. Right? >> Yeah. So, it's real. It's impact All these applications are impacting us. Yeah, absolutely. Um I mean, so much happened in that that uh couple of years. You know, we had a long history of applying a different kind of of uh model um mostly in derivatives trading. Uh what went wrong, I think, in 2007 and 2008, though, was um it it involved involved derivatives trading, but it also just involved uh how to to rate risk on these um uh CDOs. Mhm. And uh I think it was a combination of uh overconfidence in mathematical models, but also just an ecosystem of, you know, confluences of interest between ratings agencies and big banks and um very, very difficult to sort of cut through the practice and understand how the models were working. Um and I think, you know, that's I think that's characteristic of the the sort of widespread use of these models, which, you know, I think are ultimately sound when used correctly, but it's very, very difficult once they get integrated into the markets for ordinary traders or even people who are pretty sophisticated to see where they matter and where they're going to break and how that's going to affect things. It seems like the models are great as far as they go, but people just some people thought they went a lot farther than they really do. You know, you They're good as far as they they they work, right? You know, some are good some are useful and some are not. I forget I I forget what the quote was. Who was that? George Box or somebody? Um you know, all models are are wrong, some are useful. I think I think it was the original. Yeah. So, um but you but you're okay with it. Like you're not an anti-model guy in in >> No, no, no, not at all. Okay. All right. But but I I am a uh uh uh a pro, you know, examine your assumptions and and make sure that um you know, we have to think about these things as a kind of technology and we have to understand what the limitations of the technology are, right? Like the kinds of the kinds of things that you know, the the you know, the technology that that runs my car isn't going to take us to the moon. And so, we need to understand when we're trying to adapt something beyond the scope where it it's going to to work. Let's do this. Let's I I started out talking about how I had used your book. Let's do the listener service because you I can tell you've got this stuff like online in your head and you you I hear your love for history coming out um with every other word you say. So, let's start with Bachelier and tell our listener in your words why he's important to it. Why what's the connection between this 19th century mathematician and investors in the year 2026? Why do we care about him? So, you know, Bachelier is this really fascinating, way ahead of his time figure uh who right around the turn of the 20th century writes this dissertation um in in mathematics in which he applies um ideas that we would now think of as stochastic calculus. I mean, the kinds of things that would go into like areas of physics and thermodynamics. Mhm. Um to uh basically options pricing. He's the first person in a sort of long history. I mean, it goes from 1900 to 1973 basically of people reinventing this idea of using probability theory to understand the relationship between some underlying asset and an option written on that asset. Um and so he sort of comes out of nowhere um applying these statistical ideas to markets and then immediately jumping to you know, a a kind of derivative Mhm. trading. Um people sort of sleep on him for for about 50 years and then his ideas start getting rediscovered and and reapplied and then you have a number of sort of key figures that I talk about in the book leading up to uh Black and Scholes in in the early 1970s and you know, the Black-Scholes model is is famous. That won the Nobel Prize. But there are all these people who are sort of discovering pieces of those ideas um in the early part of the 20th century. So, all these zero DTE option traders who are speculating wildly every day is all Louis Bachelier's fault. >> [laughter] >> Well, you know, Ed Ed Thorp and and and Fisher Black had something to do with it. >> [laughter] >> All right. So, so this guy basically applies this probability theory and everybody goes, "What? No." and forgets about him for 50 years. Um I guess the other guy I know anything about and if there's somebody else we should talk about first. The other guy I know anything about in this in in your book um is uh is Thorp, the the beat the dealer guy. I think it to me he's the blackjack guy cuz he's the you know, the mathematician who went to Las Vegas and beat the beat the odds and made money um and managed money as well. How does he figure into this? Why do we care about him today? And we do. Uh just so the listeners clear, we do care about Ed Thorp. Yeah, I mean Ed Thorp is also he's just an incredible incredible person. Um in fact, you know, he uh was a uh early faculty member here at UC Irvine and so he's he's around Southern California. Um yeah, so so he wrote this book as you say, you know, beat the dealer. Um sort of gave a mathematical proof that uh card counting could work. Um you know, I think lots of people in the industry are are like that kind of thing. Um but then, you know, a few years later he is trying to apply the same kind of ideas, the same kind of like, you know, statistical modeling of of strategic scenarios basically. Uh and comes up with the idea that you can model um the underlying like I mean, you can treat the underlying asset of some derivative of of like an option as um you know, governed by some statistical process. It's basically the same idea that that Bachelier has except he takes it a step further and comes up with a formula for the fair value of a mean, they were detachable warrants at that point. You didn't have options markets yet. Um but uh of of something like an option. Um and he he wrote a book with a guy named Sheen Kasouf who's an economist here in Irvine called uh beat the market. Mhm. And um this book comes out and no one reads it. I mean, it's you know, it literally has a uh a cheat code for options trading written into it. It tells you how to price them, how to trade them, how to manage your money. And it doesn't get any um any attraction. And so he decides he's just going to do it himself. He teams up with a broker-dealer and they start Princeton Newport Partners, which was you know, I think the first modern quantitative hedge fund. Um you know, not exactly the fee structure of a modern hedge fund, but the the sort of trading strategy. You know, James, I have to interrupt you. I have to interrupt you for just 1 second because it's so funny, isn't it? For you to immediately say not the fee structure because that's what it's become. That's what a hedge fund is. It's a fee structure. >> Yeah, that's what a hedge fund is now. Yeah, no, but back then there was a time where they were really good at hedging. Yeah. >> [laughter] >> It used to be a hedged fund. Now it's two and 20 or whatever. That's that's a funny thing. All right, I'm sorry. Just you know, finance geek jokes here. But but he started this first modern quant hedge fund. Yeah, exactly. And I mean, they were just they were unbelievably successful. I think they had two down quarters the entire time the fund ran um went a little over 20 years and uh Wow, two down quarters in 20 years. Yeah. Yeah, no, they they were incredible. And then Yeah, you know, and and he's I think he's been been very very influential uh um in in the field since then. Sort of the the godfather of of quants. Very cool. And the other person who I mentioned earlier, who I actually Don't take it personally, James. I didn't read your whole book. I read his whole book like more than once. Um was Mandelbrot. Uh he He wrote a book called Misbehavior of Markets, which is just brilliant. He's got the 10 heresies of finance in there, which I love. Um By your estimation, where does he figure in all of this? Yeah, so, you know, he's a another really interesting figure. I think someone who um >> [clears throat] >> I mean, very, very famous mathematician, very influential, but uh conceived of himself as and always sort of was a little bit of an outsider, a kind of a gadfly type of figure criticizing the mainstream. Even though I think his ideas now are are very, very widely accepted. Um I mean, the point that he was making was a little bit technical. It has to do with what kinds of distributions we use to think about the the stochastic processes that we use to model um uh market returns. And basically, what he observed was that when you look at actual market data, the kinds of assumptions that people like Bachelier and Thorp were making and, you know, Black and Scholes later on were um uh mis- mispricing uh tail risk. Basically, the the probability of extreme events according to the way that Bachelier and and and Thorp were doing it is actually higher. I was uh Mandelbrot was saying that the probability is actually higher than what Bachelier and and Thorp were implicitly assuming. Right. And so he developed, you know, important probabilistic tools for thinking about probability distributions that have these fat tails, that have more uh probability for extreme events, a little bit less for for sort of central behavior. Um and, you know, there's a lot that you can do. And sort of once you recognize that, there's a lot you can do to sort of revise your um uh trading assumptions, try to revise your modeling assumptions and and improve on your your pricing models. Um but, you know, he noticed this um right around the same time that that people were developing the models in the first place. But, it took until the 1987 crash for people to really take this seriously. And it's really, I think, only in the early 1990s that you start seeing options models that fully um you know, account for Mandelbrot's observations. Right. My favorite thing um for no apparent reason, just be just the quirkiness and weirdness of it about him is that is the connection between his um you know, he started out with cotton prices, I believe it was. >> Yeah, that's right. And ends up at at some point um like listening to submarine sonar, you know, noise or something and and looking at it, you know, looking at the data that it that it the submarine sonar driving through Puget Sound generated. I just thought it was the weirdest connection in the world. And and that's in his book and it connects all to finance. Mhm. And the the connection as I like I'm not a mathematician, so I'm basically I have you on the show to like check me out and make sure I'm right about all this stuff. But but I thought the point there seemed like um you know, the cotton prices behaved wilder than people said, you know, the prices didn't weren't all linear movements. They were, you know, herky-jerky but you know, non-linear movements. They didn't just you know, glide, they leaped. That was one of his 10 heresies. Prices don't just glide, they leap. And >> [clears throat] >> then when you get to the sonar data, that was about turbulence. And and what you know, what we would call volatility in markets behaves like turbulence. I'm still not like crystal clear. Like if I had to re-explain it, I'd have to almost go back and like look at what I wrote because I'm like on the edge of understanding what the turbulence data told us. Can you help me out with that at all? >> [laughter] >> Like at all. Yeah, I mean so so you know his his big idea that kind of pervades everything that he did um was the idea of of self-similarity and fractals. Right. >> Right. And so this this idea that um realistic processes in nature tend to have this um scaling property where they they look similar on different scales. Um turbulence is a an example of a physical process that seems to have that property. It's kind of chaotic, but it's chaotic on on many many scales. It's you know you see turbulence on very small scales, but then it sort of comes together, something that looks like turbulence on larger scales and and so on. Mhm. Um physical processes that have that character tend to be governed by these power-law type scaling properties. Right. What does that mean? >> And well that that basically is is just the kind of function that's going to give you these fat tails. Right? So the sorts of distributions that he was saying governed or or characterized um uh cotton prices also equity prices um are are sort of power-law distributions. Um they're called Levy stable distributions, but they're they're uh not normal distributions. They've got this other kind of property and they have the sort of self-similar scaling behavior. Okay, so the other thing that I need to check with a honest-to-goodness physicist a real mathematician um it's my impression that these like what we call fat-tail events like a you know, October 1987 event or something. Like they're still it's still correct to think of them as unlikely. Oh, yeah. >> They're just they're just far more likely than the traditional sort of bell curvy you know statistical distribution that people used, you know, um before Mandelbrot, let's just say um would would tell you, right? They're still unlikely. Yeah, that No, that's exactly Exactly they're still unlikely. Um you know >> want our readers to be and listeners to be too scared of this stuff. That's right. I mean so so they're they're still unlikely. Um it's true that they're more unlikely and it's also true that, you know, with a normal distribution it's not just that um something that's uh extreme is unlikely. It's that you know, the more extreme it is, the more unlikely it is. And that's suppressed very, very fast. The probabilities go down and down and down very fast. Where with these as you sort of get into these extreme events, that that territory, you start seeing um uh more and more probability relative to a normal distribution for more and more extreme events. And so something like, you know, Black Monday 1987 is is uh uh unlikely, um but you know, it's also a much more extreme event than than um you would have expected from from other sort of market moves um in the decades leading up to it. >> I'm you know, I'm an investor and I'm trying to think of all the how how to think of all this stuff. But when I listen to when I read all the stuff that I've read over the years, you know, and all these physicists and guys like you like it suggests to me that I ought to do something about it. Now, I know you're not a professional investor, so I'm not going to ask you what to do about it. But [clears throat] let me ask you like do you know, you maybe you you you've been with a university 20 years. You got a 401k. Like, do you ever make the connection go, "Hmm, let's see. I've got this 401k. And I know this thing is more likely than most people, and I better than anyone, you know, almost, right? You're among the group of people who know better than most that this these extreme events are more likely. Do you actually do anything with your 401k? Well, I mean, I'm not like day trading in my 401k. No, you're not. That's that's part of the point here. >> some, you know, look, you know, the thing about these these sorts of extreme events is that um you know, if you're uh highly leveraged, if you're doing a lot of derivatives trading, Mhm. Um or if you're doing really short-term stuff, these extreme events can can kill you. You know, if you're if you're buying and holding for retirement, Mhm. we've had them, and they, you know, the markets bounced back. Maybe it maybe it takes 2 minutes like with the flash crash. Maybe it takes 10 years, but um you know, we have the this long-term this long-term uh pattern of of equities growth that we've seen over the last, you know, century plus is robust against these these extreme events. >> The optimists have triumphed. Yeah. One might say. Yeah. Well, you know, >> [clears throat] >> I I can I one of the things that I'm absolutely fascinated by are the places where current practices build in assumptions that uh are in conflict with with other sort of fundamental assumptions that go into to markets. Um you know, where are the conditions today where something like the 1987 crash could happen again. Um and you know, I still think that the optimists will will triumph in the long run, but I think that there there can be some some unexpected dislocations in the meantime. >> Okay, here's what I want our listeners to know. This fascinates me. You admittedly not a professional investor, not a guy like me writing about finance, like philosophy logic professor, and you got us to the same place where I often get us and where our our finance professional guests often get us, which is to this idea of ruin. Right? [laughter] You can't set yourself up It's true. You can't set yourself up for ruin. That's why I made fun of the zero DTE traders a moment ago, because they're setting themselves up for ruin every time they do this, you know? Um and if they continue to do it, they're, you know, repeatedly setting themselves up, virtually guaranteeing it, in my opinion. Even even though probability may not work that way. I mean they keep doing this and they keep, you know, yoloing um lots of money into something super risky, they're going to blow up. And and that's that seems to be like the lesson that Nassim Taleb, one of the many lessons that he taught, is about survival. And the really great traders, it's all about survival. Yeah. Uh and that's in the math that you that you've written about in your book. >> I I think of this actually as um Ed Thorp's big contribution. I mean so the options trading stuff is is one big contribution, but the other thing that he brought into trading was uh a money management strategy based on this this work by um a mathematician at Bell Labs called uh John Kelly. I mean introduced this idea of the the Kelly betting criterion, which basically is, you know, uh you assess what your edge is, right? So you have some idea that you you understand what's going to happen better than the market does. You you try to assess what your edge is, uh and then um you you use that plus your current um Well, I mean you basically you that to to to manage your your bankroll, right? There's a a theorem that says what the optimal way of of doing that is. This is something that Thorp came up with in his blackjack days for figuring out how to survive at the blackjack table, right? So, he could count cards. Mhm. Counting cards gives you a a big advantage when you get to the end of the deck. Early on in the deck, you don't have that big advantage. And so, you need to figure out how to manage your bets, manage your money so you're still there when the advantage comes. But, the fact that you have an edge also doesn't mean you win every time. It just means that you have a higher chance of winning. And so, again, it's all about survival. It's all about just managing your money at the table until you get to the place where you do have the edge and then sort of optimizing. Right. That's where all the traders take us. It It It doesn't matter. It matters how much you you know, how much you win on your winners and how how little your losses are on your losers. You know, that's the That's the key thing that you're watching all the time. You know, people think that they're going to get certainty. Like, it's a typical novice investor thinks they're going to figure out a way to just snap their fingers and double their money. Uh and then they find out that that is not the way it is. And then they learn all this stuff we're talking about about managing money, managing position size, etc., etc. It's really cool. Let me ask you something. The book came out in what, 2014? Physics of >> I think it was January 2013. 2013. Okay. So, 13 years on, um w- like how would you characterize Like, if you could char- you know, sum up the book like the 2013 version that that exists, and and then say, you know, what I would do Is there anything you'd say or do differently now, do you think? Yeah. Well, I mean, look, that that that book was written uh at a particular time where there were particular debates happening, right? So, it was really a response to the response to the 2007-2008 crisis. Mhm. Um and, you know, I I think as a response to how people were talking about that, I I I wouldn't change I wouldn't change anything. Uh I I I think I stand by what I I wrote there. I mean, you know, maybe maybe if I went and read it again, I wouldn't I wouldn't I wouldn't feel that way. But, um but I I do think that a lot has changed in market since then. Um Okay. And just so for our listener, James, like have we covered the main conclusion which is like, you know, models screwed us up, but models are are good if they're used responsibly? What more is there to it than that? For you? Uh yeah, so so no, I I I think of that as the the main the main message, right? So, Okay, great. let's understand where these models came from, what people were trying to do. Okay. Let's look at the way in which the people who introduced these models talked about them. I mean, we've talked about um uh Mandelbrot and Thorp and and Bachelier. Um you know, Fisher Black is a another person who I think is really fascinating. Mhm. Uh he wrote a a an article um [clears throat] in the 1980s um called The Holes in Black-Scholes where I mean, it it it could have been written by Mandelbrot. He just goes through and and gives all of these assumptions that are, you know, unrealistic that go into the Black-Scholes equation. And you just list them. And he's like, "Look, you know, these are the ways in which this thing doesn't track reality. Here are the places where it's going to fail. Here are the problems with it. Um and here's why we think you you can use it anyway." Um so Wow. Uh I think when you look at how the people who were originally developing these ideas thought about them, you see a uh like a kind of humility about what the models can do and a kind of sensitivity to figuring out when they're going to fail. And I think that that's that's just a lesson that very easily gets lost, right? So I think most of the people who work with this kind of modeling stuff nowadays, many of them are are coming out of financial engineering backgrounds, they're coming out of sort of specialized training in applying these models. I think it's hard to get the full perspective that the originators had. But I think that that's something that's important to to to focus on. And so that's that's what I see as the big the big lesson. Hey, you don't happen to know a guy named Emanuel Derman, do you? Yeah, yeah, sure. You You know him personally or you know of him? I mean I'm trying to remember if I No, no, I I met him personally. Yeah, I mean we've corresponded over the years, but I went and and visited him at at Columbia when I was writing the book. I like his book Models Behaving Badly. Yeah. >> I think it's I think it's wonderful and I think He's got an awesome book on the volatility smile. I mean don't get me started on that, but but he has a he has a another book that that he wrote on the volatility smile. I'm happy to get you started. I'm happy to pull the rip cord on the volatility smile. I don't even know what that term means. Okay, so the volatility smile So this is this is like a little you know history and philosophy of of Black-Scholes here. When the Black-Scholes model was first introduced, I mean this is the same as when when Ed Thorp was doing his stuff. Options markets were very inefficient. And so you could use the model plus your you know knowledge about the the sort of historical statistics of the underlying asset to come up with a fair market price and you can be pretty confident that when the option expired its value was going to be very close to to what you predicted the option price like the the fair price was going to be. And so you could look for mispriced assets in options markets. Um by the early '80s you couldn't do this anymore cuz everybody was using Black-Scholes. Okay, so um at that point Black-Scholes sort of has a different kind of role. Um it's more like just the the thing that we use to uh set prices amongst ourselves. Um but you can start using it in a kind of inverse way as well where you can infer backwards from the market price of something to what the implied volatility of the underlying asset is. Like what is what does the market think these statistical properties of this thing are going to be? Okay, now fast forward to 1987 after the the um the uh market crash. I mean that whole crash is driven by um failures of Black-Scholes and the the uh portfolio insurance based on it. People start thinking like what's going on with this this equation. They start trusting it less and trading differently and all of the sudden now if you do this inverse calculation this you know by like 1989 you start seeing this very clearly um you start doing this inverse calculation to what the implied volatility is is no longer a number. It's a curve. And it's a curve that depends on how far out of the money you are. Basically how unlikely it is that an option is going to be valuable. Mhm. And one way of interpreting this is that traders kind of got wise to Mandelbrot's ideas. They figured out that actually these extreme events are more likely than Black-Scholes was assuming. Okay, I was going to ask you to make it a little more concrete for me but that's it right there. That's it. Um and so so nowadays the volatility smile is the main thing people use the Black-Scholes equation for. It's to figure out uh sort of what the the whole I mean it it'll become like a volatility surface or a multi-dimensional analysis where um you can see how other traders in the market are thinking about the likelihood of extreme events. Um And so yeah, so that's that's like a common quant technique now and uh uh Derman has a whole book just on that. That's cool. Yeah, I'm looking at I was I was clicking around while you were talking. That's why I was looking at at my computer here. Um at at the book and it's um the blurb sounds great if you're into this stuff and if you know, you know, Derman's writing which I don't know if the writing in this is good as the other one. Um but I really I enjoyed the other book so who knows? Maybe I'll pick this one up, too. Um is there like Let me ask you, do you know the book well? Like is am I going to go in and find a million equations I'm not going to understand? Uh no, there are equations in the book. I mean it's written it's written for a um a sophisticated professional audience. It's not like a Okay. It's not like a um a textbook. Uh-huh. Um but it is a a a book for uh quants. And you know, he's you know, he's he's really interested in in the history and he's really interested in how to think about the volatility smile because it's something that um I mean, the the way I think about it is uh the volatility volatility smile is a kind of contradiction within the Black-Scholes model. The Black-Scholes model is correct, you don't get a volatility smile. We get a volatility smile so the Black-Scholes model's incorrect. But you only get a volatility smile, like you only can can see what it is by applying the Black-Scholes model. And so it's like as a philosopher, it's like a a great example of a kind of circular paradoxical reasoning where it works and it seems to provide a lot of information, but you're contradicting yourself somewhere in there. >> Right. And that smile demonstrates the contradiction for you. Yeah. All right. Interesting. All right. Well, um I'm going to I'm going to add that one to my cart. We'll see. I'm not I'm [laughter] not pulling it. Sometimes I'll just click right as you're talking on the show, you know, cuz I like to I like to read and I I just like good recommendations, but this one seems a little heady even for me. So, um let's let's move on and so and and I just want to actually conclude this. So, you wouldn't change anything that you said about models. You wouldn't change the point of the book from 2013, right? We said that. Yeah. Um but if you were writing it today, it sounds like, you know, with that event farther in the past, you know, maybe you'd have something else to say. I don't know. Well, I I think I would say more about um algorithmic trading. I think I would say more about AI. There are sort of other changes that have happened in in markets. Okay. Um AI, of course, is the one that everybody listening to this just heard you say that. They said, "Oh, tell me about that." Yeah. I mean, the the the other thing that I I think is is really important is the rise of of um passive trading. I mean, so uh you know, ETFs are huge. Passive ETFs are huge. Yes. >> Um and you know, I think I think this is an example of a place where uh look, here's how markets are supposed to work. There's price discovery. You've got people with different amounts of information who are trading on that information and and the market price reflects some kind of, you know, um clearing of of all of this information. Uh um and uh if everyone's trading passively, that doesn't happen, right? Everything is sort of in lockstep by capitalization, and most of the money that's flowing in isn't information doesn't have information at all. Uh and so, I I at some point um markets break under that kind of scenario. Right. And there's um I guess we've had on the show before Mike Green and another guest we've had on the show Hari Krishnan who recently put out some work suggesting, you know, uh the basic suggestion as I understand it is that passive trading could become so large a part of the market it could generate so much and that would generate so much more volatility that there's a non-zero chance the market could actually go to zero. And that we're probably getting to that level of concentration within 5 years. And and Mike uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh >> [laughter] >> Uh but others have suggested, you know, there there's something there there must be something wrong with there because they don't, you know, see the the market going to zero guar- you know, um w- they don't see that high of a probability of the market going to zero. Uh yeah, so I I agree that that market going to zero is a pretty pretty extreme outcome. >> Well, it's extreme, yeah. Uh but you know, I I I I I I do think that that we can see really significant really significant dislocations. Um you know, Right. um And you know, and this is connected to all sorts of other related trends like the underperformance of value over the last two decades compared to, you know, the five decades before that and um wild price-to-earning ratios. I mean, this is just normal, you know, um this isn't like you don't need physics to to No. talk about this stuff. No, but but understanding extreme events is is kind of key here, though, right? Yeah, that's right. And I I also think that this this general theme of saying, "Look, you know, our once once models, including portfolio theory, get built into products and into standard practices, we stop seeing the assumptions. Um right, and so you know, once portfolio insurance was a product that you would sell people as an like you conceived of it as a kind of insurance in the '80s, all of a sudden the assumptions that go into Black-Scholes are invisible to you. Um similarly, when you talk about options trading in terms of the Greeks, the Greeks are all based on a Black-Scholes model. And if if the Black-Scholes model's not right, then those parameters aren't the right ones to look at in all scenarios. And I think that the ETF the passive the passive um uh trading is another example where this idea from portfolio theory that you should just diversify and buy the market is a great idea when everyone else is trying to trade on information, but it doesn't make as much sense once that's what everyone's doing. And um yeah, I mean it's a a place where an assumption about how markets work um it kind of goes out of um touch with with what markets are actually doing. >> And and this and this idea that you know, everybody doing what seems like the right thing and then and more and more and more people do it, it's that old thing, you know, the the the fool does it in the end what the wise man does in the beginning. And and it and even more to the point for investors, um you know, there's you you could pick the greatest strategy anyone's ever thought of and if too many people are doing it, it's just going to work less and less well and create more and more risk is really the point. Um >> Yeah. It's the increase in risk, yeah. That's right. Markets like you can't sleep on markets. There it's always changing. Well, there's also there's you know, there's just a general scalability problem, right? So and you I think you can see that in a few different places. One is, you know, the rise of of of private markets. Mhm. Um I I won't go into I won't go in at length about all my views on on private markets, but um it's it's a strategy that worked great 20 years ago. Mhm. Uh or okay, 10 years ago. Mhm. Um it's not clear that it it's sustainable under arbitrary growth. But of course, the more the more equities are traded privately, the more private debt you have, all of that, um the the less diverse the public markets are. Mhm. And um the harder it is to get out of, you know, the S&P 500 blue chip type you know, um uh stocks. Um and then, you know, uh another place is is, you know, quant strategies. I I think they're really interesting examples of strategies that worked really really well when you were um you know, in the the hundreds of millions, a few billion. When you try to scale them up to hundreds of billions or trillions, they just the opportunities aren't there. Um and so there I think there are some examples of very very successful firms that have tried to offer products that they hoped could scale differently. And um you know, just haven't it hasn't worked. It's interesting. Like um everything changes with size. Size is so important. Just um The guy named Vaclav Smil, who you probably heard of, has a whole book called size. And it's just about scale in nature and in life and so on. It's it's it's fascinating to me because you know, we investors see it all the time. Any investor who's been to the to the Berkshire Hathaway annual meetings hears it has heard it every year for decades. Well, well, we're bigger than we used to be, so we're not going to make as much money. And Buffett has said this like every single year for 20 some years now. Um at [clears throat] least, you know, that I've that I've been going or that I you know, I've gone a couple times the past 20 years. So, it's super duper important, the issue of size. What size company are you buying? What size pool of capital are you are your assets in? And frankly, I have to say it's not it's one of those things you discover it's nothing you ever would have thought would have been important, but it really becomes very important. Yeah, I completely agree. All right. Um A minute ago, >> [laughter] >> I don't know if you would have go here, but a minute ago you said I won't tell you all my views on uh on um uh private private markets. It sounded like there was some stuff there. I don't know if you want to get into it. >> [laughter] >> Uh yeah, I I think I better not get into my views on private markets. >> Okay. All right. Maybe it gets a little political or something. That's And that's cool. I respect that. And we And we thank you. Like we try to keep it, you know, real and and uh and not get into areas that like that. So, I appreciate it. >> [clears throat] >> Um But maybe we we could talk a bit more we we I mentioned AI and you and and you know, you you had something that you want to talk about, but I was just going to ask you, do you want to you had you had said at one point that you wanted to talk about von Neumann. Should we talk about von Neumann? Um yes. If only because wow, what a genius, right? I mean, unbelievable character there. Yeah. Yeah, so so, you know, um I've been writing a biography of von Neumann for the last almost a decade now. Um Cool. Very cool. And uh yeah, so so um incredible guy. I mean I think part of what I find so compelling about him is how he thought about applied mathematics, right? That these same sorts of themes that I think have become so important to how we think about financial modeling are themes that really originate with him. I mean applied math wasn't the kind of thing that it is today before really before World War II. And so much of what we now think of as sort of places where you can apply math. I mean game theory, mathematical economics, but also computer science, you know, all of this is stuff that that von Neumann played a central role in in developing including AI. You know, his his last book which was published right after he died was called the computer and the brain where he sort of sketches the idea of an artificial neural network and and and revisit some assumptions that he had made earlier on in his career about the relationship between electronic computers and and brains and sort of began to lay the foundations for for modern AI. Yeah, people who who saw this coming fascinate me cuz I sure as hell didn't. You know, it was all it was all like sci-fi until it practically until it arrived. And then 2022 came along and chat GPT came out and I was as you know, blown away as the average person. You know, I've never been a technical guy. But I let me ask you this. I use Claude every single day. I use Claude every and not only that, I use it on the weekends. Like it helped me clear out my Amazon save for later cart which had over 100 books in it, you know. And you see behind me like this is like not half of it. So, you know, I got to I got to chill out with the books. And I use it for for lots of stuff. Taxes, all kinds of things. And I use it to help me do research, financial research. What do you do with it? Do you use AI? Yeah, so I mean, I I use it for for a couple of things. Um I I use it in in my teaching in in some ways. Um I try to you know, I I try to incorporate into my teaching um ideas about uh what how AI works, but also what its limitations are, what it's good for, where it can cause problems. Um I've experimented with using it in in research and uh I have not yet had great successes. I have colleagues who swear by it. Um and and including like mathematicians who >> I was going to say like hard sciences swear by it. Yeah, and so so, you know, some people claim that that it can help them do new math. Um the kinds [clears throat] of of examples that I've seen that I'm most convinced by is really effective as a um a deep literature review tool. Finding things that are already out there that you wouldn't know how to to look for. That that characterizes my use actually rather well. Um I use something called Notebook LM that you can just dump a bunch of 10Ks and other SEC filings and you know, investor presentations from a company website. You can just put it all in one place and basically tell Notebook LM only look at these sources. Mhm. And it's a fantastic way to run through a whole, you know, say you have 10 or 15 10Ks and you want to know historical stuff. Um fan- fantastic tool for doing that. But the thing that really gets me is um how how Claude has like a voice. I mean, it's like it's you know, the latest and I I think I forget what that latest model is it called Opus or something? Um and it's um it makes connections or helps me make connections that I might not have made otherwise. Um partially I realized just by the sheer volume of of ideas that it can put in front of your face. And and it'll connect something and it'll say, "Oh, this connects to your earlier idea." And I'll say, "Whoa!" You know, just because I'm a human and I can't make it can make 30 connections and you know, in a split second or something and I can make 30 in like 30 years. You know what I'm saying? It's just uh it's a little bit faster than Dan at the age of 64 is all I'm saying, I guess. >> [snorts] >> But it's useful. Yeah. Of course, that's the problem with it, isn't it? That it's so human and so useful. And I I I'm trying to convince my stepdaughter, you know, it's a terrible therapist. Don't tell it your personal problems. [laughter] You know, and she keeps telling it her personal problems. Like, "Chat GPT said I should do this." So so you know, there's genuine concern out there and are you are you Where do you come down on the scale of, you know, people saying it's going to kill us all and you know, become Skynet and you know, kill us all in the future or put us all out of work? Do you have those fears? Um I mean, put us all out of work. Yeah, I work at a university. Uh Right. >> There are a lot of ways in which it can put us out of work. Mhm. Um you know, one way is that we can't figure out how to to continue to make you know, education a meaningful experience for students, right? So if if it becomes a situation where students are using AI to do their homework and professors are using AI to grade the homework, what's the point? I mean, why why are any of us doing this? And uh once that goes away, you know, >> the time, I might add, James, when the university it's not the sort of hallowed institution that, you know, before like we didn't question it. Everybody should go to college. Now, that is not a that is not as popular an idea at all, is it? Yeah, that's right. Well, and you know, I'll I'll tell you, it's the computer science majors who engineered themselves out of jobs. Right. Right. And so, we had this I think very very strange, you know, labor market situation where AI is booming. Everyone says, "I want to get involved in the AI boom." And so, I'm going to to learn what I need to learn to go and get these jobs. But then AI got good enough even before they had graduated that you just don't need programmers anymore. I mean, this is the place in my my research where I use AI the most. Um, you know, it's just integrated into to sort of Generating code? Generating code? Is that it? >> Generating code. Exactly. Yeah, and it's it's incredible at that. I mean, like it can read your mind. Like, you know, you type the name of a function and then it produces the function. It's like, how where in my in my code? I mean, I I believe I'm the first one writing this program. Like, where in my code did you figure this out from? But, yeah, it it it can do it. Wow. Okay, so maybe um you know, maybe we'll I don't know, maybe one day we'll get you back and we'll just talk about AI. But, it's it's it's been great talking with you. It's time for our final question, which is the same for every guest, no matter what the topic, even a non, you know, financial professional guest like yourself. Um same identical question, no matter what. But, it's simply this. It's for our listeners' sake. If you could give them one takeaway, one thought today, what would you like to leave them with? It can be anything. Yeah, I mean, the the my one thought is think meta. Not That's not a stock tip, okay? Right, right. Think meta about the ways in which um the financial products and the financial services that we're participating in are built on now these mathematical models that assume stuff about what the world is like. And it's very very hard to see those assumptions, but everything that we do can break when those assumptions break. The more integrated they are into our practices, the the harder they are to see and the more dangerous they become. All right. Wise words. Thanks for being here, James, um and I look forward to having you back on the show sometime. Yeah, thanks a lot, Dan. This was fun. That was incredible. For me, at least, I hope it was for you, too. It was very highly educational. I recommend James' book, The Physics of Wall Street. Um it's it's actually very well written, and it's not full of technical mumbo jumbo and equations. It's full of really good stories about important things that people mostly don't know a whole lot about, which is fun, right? It's fun to learn about these historical figures and why they're important and what they said that nobody was saying at the time. Um and it it'll teach you something about kind of how markets work in a way that you're just probably not going to read a lot of other places. I mentioned the book by um Mandelbrot, which is uh Misbehavior of Markets. We mentioned Edward Thorp's book, Beat the Market. Um we talked about John von Neumann and some other people. So, there there's other books by, you know, the the innovators [music] themselves, the thinkers themselves. Um but you usually you don't often get a smart guy like James who wants to make it so that everybody can understand it so that he can explain it to you. So, you know, that's why I like the Physics of Wall Street. That's why we wanted to have James on the show. I was like, a guy who can tell stories about math and finance so that a guy like me can understand it and really enjoy it. That's got some value. All right, I think I hope it has value for you, too. So, I recommend the book and I recommend you learn about all these characters, Bachelier and Thorp and and Mandelbrot and all the rest of them. So, that was another great interview and another episode of the Stansberry Investor Hour. I hope you enjoyed it as much as we did and remember, hit like, hit subscribe, and by all means sign up for our free daily email. Opinions expressed on this program are solely those of the contributor and do not necessarily reflect the opinions of Stansberry Research, its parent company, or affiliates.
The “Card Counter” Formula That Beat Wall Street For 20 Years| SIH
Summary
Transcript
Get out your pens and pencils, folks. Today's guest is a really smart guy, not a finance guy. He's a physics guy who wrote a book called The Physics of Wall Street, James Owen Weatherall. It's a fun book. It's not a hard crazy technical book, and we're going to talk about a lot of the ideas in it, which are very important for investors. It's important to know historically where all this stuff is coming from, and all the crazy math that influences, I guarantee you, influences places [music] that you are putting your money in your 401k. So, pay attention, and and have fun with this, because this stuff is pure fun to learn about. So, let's [music] do it. Let's talk with our guest, James Owen Weatherall. Let's do it right now. James, welcome to the show. Thanks for being here, man. Oh, thanks so much for having me, Dan. We have to start with one with one thing that I need to tell you. I love your book, um, The Physics of Wall Street. Thanks. >> Now, I say that not having read the whole thing, but I've used this thing. I was trying to sit here and count the number of times, at least three times I've used it as a reference, because you do a fantastic job of writing about some of the charac- the great characters, um, in in math history and the history of sort of math and finance together. Um, Edward Thorp, Mandelbrot, Bachelier. I think you were like one of my primary sources when I wrote my story about Bachelier. There I think I had three sources, and you were one of the best ones. So, thank you for that. And one day >> Of you, yeah. I'm I'm glad I could be of help. Yes, one day I I hope to read the whole thing, just so you know. I'm sure [snorts] there's even more good stuff in there. I'll probably go back to it for some, but, you know, Pascal or whoever else you got in there. Um, so, I want to start there if I may. Like where did the book come from? Are you a Wall Street guy or are you a physics guy or both? I know I'm I'm 100% a physics guy. So, I was I was a grad student in physics at Stevens Institute of Technology which is in Hoboken, New Jersey. Yeah. Um right across the river from from Wall Street Yep. >> uh in um 2007 and 2008. Got you. >> Uh I had a lot I was I was like right out of college. I had a lot of friends who had been physics majors who had gone into finance. I mean some of them are still still working in the industry now. I'm sorry James. I'm sorry to interrupt you but I have to wonder like physics seems like this cool thing. What what are you doing the right now in the year 2026? Like what what cool stuff are you doing? You must be doing cool stuff. >> Yes. I So, I'm actually I mean I'm I'm not a physicist now. I'm a I'm a philosopher. I Yeah. So, so I I am a professor at in logic and philosophy of science. It's a niche department at University of California, Irvine in Southern California for almost 20 years now. Wow. >> And yeah. So, I mean I I work in philosophy of physics and sort of mathematical physics but also you know abstract science and stuff like that. >> That's cool. Very cool. Um so that makes a lot more the then now the book makes more sense to me. Right? Because what what do we what else does a mathematician or a physicist would a philosophize about except for this grand machine of Wall Street. And if you know Bachelier then you got the connection from the word go. Uh and away you go. Well, what what do you think of it? Like what When I read your book, like I said I just read pieces of it. So, I didn't get the full sweep of it. What does James Owen Weatherall think of of this combination of really from my perspective super advanced mathematics plus Wall Street where you really can't predict what the hell's going to happen next. What do you make of that? Yeah, I mean, so I am interested in it from so many different perspectives, right? So, I'm interested in the math. I mean, I just I like math. I like math applied to stuff. I'm I'm super interested in the history of applied mathematics. Mhm. I'm also just interested in in the history of big ideas. Uh and, you know, what could be more important in the history of the 20th century and now the 21st century than this this movement from, you know, these these big technical ideas from mathematical physics and and um related areas uh into financial markets. Um but also, you know, I'm interested in the philosophy, right? I mean, we're building these mathematical models to try to do something in the world. How are we supposed to think about that? Why does that work? Why does it sometimes not work? Um I mean, that that to me is just core philosophy of science and uh you know, applied in in a case that really matters. I'm glad you brought models up because, like, we just we lived through this thing, you know, whatever, 20 years ago almost, where people lost their homes. And that was at the end of a chain of events that started out with with risk models that didn't work. Right? >> Yeah. So, it's real. It's impact All these applications are impacting us. Yeah, absolutely. Um I mean, so much happened in that that uh couple of years. You know, we had a long history of applying a different kind of of uh model um mostly in derivatives trading. Uh what went wrong, I think, in 2007 and 2008, though, was um it it involved involved derivatives trading, but it also just involved uh how to to rate risk on these um uh CDOs. Mhm. And uh I think it was a combination of uh overconfidence in mathematical models, but also just an ecosystem of, you know, confluences of interest between ratings agencies and big banks and um very, very difficult to sort of cut through the practice and understand how the models were working. Um and I think, you know, that's I think that's characteristic of the the sort of widespread use of these models, which, you know, I think are ultimately sound when used correctly, but it's very, very difficult once they get integrated into the markets for ordinary traders or even people who are pretty sophisticated to see where they matter and where they're going to break and how that's going to affect things. It seems like the models are great as far as they go, but people just some people thought they went a lot farther than they really do. You know, you They're good as far as they they they work, right? You know, some are good some are useful and some are not. I forget I I forget what the quote was. Who was that? George Box or somebody? Um you know, all models are are wrong, some are useful. I think I think it was the original. Yeah. So, um but you but you're okay with it. Like you're not an anti-model guy in in >> No, no, no, not at all. Okay. All right. But but I I am a uh uh uh a pro, you know, examine your assumptions and and make sure that um you know, we have to think about these things as a kind of technology and we have to understand what the limitations of the technology are, right? Like the kinds of the kinds of things that you know, the the you know, the technology that that runs my car isn't going to take us to the moon. And so, we need to understand when we're trying to adapt something beyond the scope where it it's going to to work. Let's do this. Let's I I started out talking about how I had used your book. Let's do the listener service because you I can tell you've got this stuff like online in your head and you you I hear your love for history coming out um with every other word you say. So, let's start with Bachelier and tell our listener in your words why he's important to it. Why what's the connection between this 19th century mathematician and investors in the year 2026? Why do we care about him? So, you know, Bachelier is this really fascinating, way ahead of his time figure uh who right around the turn of the 20th century writes this dissertation um in in mathematics in which he applies um ideas that we would now think of as stochastic calculus. I mean, the kinds of things that would go into like areas of physics and thermodynamics. Mhm. Um to uh basically options pricing. He's the first person in a sort of long history. I mean, it goes from 1900 to 1973 basically of people reinventing this idea of using probability theory to understand the relationship between some underlying asset and an option written on that asset. Um and so he sort of comes out of nowhere um applying these statistical ideas to markets and then immediately jumping to you know, a a kind of derivative Mhm. trading. Um people sort of sleep on him for for about 50 years and then his ideas start getting rediscovered and and reapplied and then you have a number of sort of key figures that I talk about in the book leading up to uh Black and Scholes in in the early 1970s and you know, the Black-Scholes model is is famous. That won the Nobel Prize. But there are all these people who are sort of discovering pieces of those ideas um in the early part of the 20th century. So, all these zero DTE option traders who are speculating wildly every day is all Louis Bachelier's fault. >> [laughter] >> Well, you know, Ed Ed Thorp and and and Fisher Black had something to do with it. >> [laughter] >> All right. So, so this guy basically applies this probability theory and everybody goes, "What? No." and forgets about him for 50 years. Um I guess the other guy I know anything about and if there's somebody else we should talk about first. The other guy I know anything about in this in in your book um is uh is Thorp, the the beat the dealer guy. I think it to me he's the blackjack guy cuz he's the you know, the mathematician who went to Las Vegas and beat the beat the odds and made money um and managed money as well. How does he figure into this? Why do we care about him today? And we do. Uh just so the listeners clear, we do care about Ed Thorp. Yeah, I mean Ed Thorp is also he's just an incredible incredible person. Um in fact, you know, he uh was a uh early faculty member here at UC Irvine and so he's he's around Southern California. Um yeah, so so he wrote this book as you say, you know, beat the dealer. Um sort of gave a mathematical proof that uh card counting could work. Um you know, I think lots of people in the industry are are like that kind of thing. Um but then, you know, a few years later he is trying to apply the same kind of ideas, the same kind of like, you know, statistical modeling of of strategic scenarios basically. Uh and comes up with the idea that you can model um the underlying like I mean, you can treat the underlying asset of some derivative of of like an option as um you know, governed by some statistical process. It's basically the same idea that that Bachelier has except he takes it a step further and comes up with a formula for the fair value of a mean, they were detachable warrants at that point. You didn't have options markets yet. Um but uh of of something like an option. Um and he he wrote a book with a guy named Sheen Kasouf who's an economist here in Irvine called uh beat the market. Mhm. And um this book comes out and no one reads it. I mean, it's you know, it literally has a uh a cheat code for options trading written into it. It tells you how to price them, how to trade them, how to manage your money. And it doesn't get any um any attraction. And so he decides he's just going to do it himself. He teams up with a broker-dealer and they start Princeton Newport Partners, which was you know, I think the first modern quantitative hedge fund. Um you know, not exactly the fee structure of a modern hedge fund, but the the sort of trading strategy. You know, James, I have to interrupt you. I have to interrupt you for just 1 second because it's so funny, isn't it? For you to immediately say not the fee structure because that's what it's become. That's what a hedge fund is. It's a fee structure. >> Yeah, that's what a hedge fund is now. Yeah, no, but back then there was a time where they were really good at hedging. Yeah. >> [laughter] >> It used to be a hedged fund. Now it's two and 20 or whatever. That's that's a funny thing. All right, I'm sorry. Just you know, finance geek jokes here. But but he started this first modern quant hedge fund. Yeah, exactly. And I mean, they were just they were unbelievably successful. I think they had two down quarters the entire time the fund ran um went a little over 20 years and uh Wow, two down quarters in 20 years. Yeah. Yeah, no, they they were incredible. And then Yeah, you know, and and he's I think he's been been very very influential uh um in in the field since then. Sort of the the godfather of of quants. Very cool. And the other person who I mentioned earlier, who I actually Don't take it personally, James. I didn't read your whole book. I read his whole book like more than once. Um was Mandelbrot. Uh he He wrote a book called Misbehavior of Markets, which is just brilliant. He's got the 10 heresies of finance in there, which I love. Um By your estimation, where does he figure in all of this? Yeah, so, you know, he's a another really interesting figure. I think someone who um >> [clears throat] >> I mean, very, very famous mathematician, very influential, but uh conceived of himself as and always sort of was a little bit of an outsider, a kind of a gadfly type of figure criticizing the mainstream. Even though I think his ideas now are are very, very widely accepted. Um I mean, the point that he was making was a little bit technical. It has to do with what kinds of distributions we use to think about the the stochastic processes that we use to model um uh market returns. And basically, what he observed was that when you look at actual market data, the kinds of assumptions that people like Bachelier and Thorp were making and, you know, Black and Scholes later on were um uh mis- mispricing uh tail risk. Basically, the the probability of extreme events according to the way that Bachelier and and and Thorp were doing it is actually higher. I was uh Mandelbrot was saying that the probability is actually higher than what Bachelier and and Thorp were implicitly assuming. Right. And so he developed, you know, important probabilistic tools for thinking about probability distributions that have these fat tails, that have more uh probability for extreme events, a little bit less for for sort of central behavior. Um and, you know, there's a lot that you can do. And sort of once you recognize that, there's a lot you can do to sort of revise your um uh trading assumptions, try to revise your modeling assumptions and and improve on your your pricing models. Um but, you know, he noticed this um right around the same time that that people were developing the models in the first place. But, it took until the 1987 crash for people to really take this seriously. And it's really, I think, only in the early 1990s that you start seeing options models that fully um you know, account for Mandelbrot's observations. Right. My favorite thing um for no apparent reason, just be just the quirkiness and weirdness of it about him is that is the connection between his um you know, he started out with cotton prices, I believe it was. >> Yeah, that's right. And ends up at at some point um like listening to submarine sonar, you know, noise or something and and looking at it, you know, looking at the data that it that it the submarine sonar driving through Puget Sound generated. I just thought it was the weirdest connection in the world. And and that's in his book and it connects all to finance. Mhm. And the the connection as I like I'm not a mathematician, so I'm basically I have you on the show to like check me out and make sure I'm right about all this stuff. But but I thought the point there seemed like um you know, the cotton prices behaved wilder than people said, you know, the prices didn't weren't all linear movements. They were, you know, herky-jerky but you know, non-linear movements. They didn't just you know, glide, they leaped. That was one of his 10 heresies. Prices don't just glide, they leap. And >> [clears throat] >> then when you get to the sonar data, that was about turbulence. And and what you know, what we would call volatility in markets behaves like turbulence. I'm still not like crystal clear. Like if I had to re-explain it, I'd have to almost go back and like look at what I wrote because I'm like on the edge of understanding what the turbulence data told us. Can you help me out with that at all? >> [laughter] >> Like at all. Yeah, I mean so so you know his his big idea that kind of pervades everything that he did um was the idea of of self-similarity and fractals. Right. >> Right. And so this this idea that um realistic processes in nature tend to have this um scaling property where they they look similar on different scales. Um turbulence is a an example of a physical process that seems to have that property. It's kind of chaotic, but it's chaotic on on many many scales. It's you know you see turbulence on very small scales, but then it sort of comes together, something that looks like turbulence on larger scales and and so on. Mhm. Um physical processes that have that character tend to be governed by these power-law type scaling properties. Right. What does that mean? >> And well that that basically is is just the kind of function that's going to give you these fat tails. Right? So the sorts of distributions that he was saying governed or or characterized um uh cotton prices also equity prices um are are sort of power-law distributions. Um they're called Levy stable distributions, but they're they're uh not normal distributions. They've got this other kind of property and they have the sort of self-similar scaling behavior. Okay, so the other thing that I need to check with a honest-to-goodness physicist a real mathematician um it's my impression that these like what we call fat-tail events like a you know, October 1987 event or something. Like they're still it's still correct to think of them as unlikely. Oh, yeah. >> They're just they're just far more likely than the traditional sort of bell curvy you know statistical distribution that people used, you know, um before Mandelbrot, let's just say um would would tell you, right? They're still unlikely. Yeah, that No, that's exactly Exactly they're still unlikely. Um you know >> want our readers to be and listeners to be too scared of this stuff. That's right. I mean so so they're they're still unlikely. Um it's true that they're more unlikely and it's also true that, you know, with a normal distribution it's not just that um something that's uh extreme is unlikely. It's that you know, the more extreme it is, the more unlikely it is. And that's suppressed very, very fast. The probabilities go down and down and down very fast. Where with these as you sort of get into these extreme events, that that territory, you start seeing um uh more and more probability relative to a normal distribution for more and more extreme events. And so something like, you know, Black Monday 1987 is is uh uh unlikely, um but you know, it's also a much more extreme event than than um you would have expected from from other sort of market moves um in the decades leading up to it. >> I'm you know, I'm an investor and I'm trying to think of all the how how to think of all this stuff. But when I listen to when I read all the stuff that I've read over the years, you know, and all these physicists and guys like you like it suggests to me that I ought to do something about it. Now, I know you're not a professional investor, so I'm not going to ask you what to do about it. But [clears throat] let me ask you like do you know, you maybe you you you've been with a university 20 years. You got a 401k. Like, do you ever make the connection go, "Hmm, let's see. I've got this 401k. And I know this thing is more likely than most people, and I better than anyone, you know, almost, right? You're among the group of people who know better than most that this these extreme events are more likely. Do you actually do anything with your 401k? Well, I mean, I'm not like day trading in my 401k. No, you're not. That's that's part of the point here. >> some, you know, look, you know, the thing about these these sorts of extreme events is that um you know, if you're uh highly leveraged, if you're doing a lot of derivatives trading, Mhm. Um or if you're doing really short-term stuff, these extreme events can can kill you. You know, if you're if you're buying and holding for retirement, Mhm. we've had them, and they, you know, the markets bounced back. Maybe it maybe it takes 2 minutes like with the flash crash. Maybe it takes 10 years, but um you know, we have the this long-term this long-term uh pattern of of equities growth that we've seen over the last, you know, century plus is robust against these these extreme events. >> The optimists have triumphed. Yeah. One might say. Yeah. Well, you know, >> [clears throat] >> I I can I one of the things that I'm absolutely fascinated by are the places where current practices build in assumptions that uh are in conflict with with other sort of fundamental assumptions that go into to markets. Um you know, where are the conditions today where something like the 1987 crash could happen again. Um and you know, I still think that the optimists will will triumph in the long run, but I think that there there can be some some unexpected dislocations in the meantime. >> Okay, here's what I want our listeners to know. This fascinates me. You admittedly not a professional investor, not a guy like me writing about finance, like philosophy logic professor, and you got us to the same place where I often get us and where our our finance professional guests often get us, which is to this idea of ruin. Right? [laughter] You can't set yourself up It's true. You can't set yourself up for ruin. That's why I made fun of the zero DTE traders a moment ago, because they're setting themselves up for ruin every time they do this, you know? Um and if they continue to do it, they're, you know, repeatedly setting themselves up, virtually guaranteeing it, in my opinion. Even even though probability may not work that way. I mean they keep doing this and they keep, you know, yoloing um lots of money into something super risky, they're going to blow up. And and that's that seems to be like the lesson that Nassim Taleb, one of the many lessons that he taught, is about survival. And the really great traders, it's all about survival. Yeah. Uh and that's in the math that you that you've written about in your book. >> I I think of this actually as um Ed Thorp's big contribution. I mean so the options trading stuff is is one big contribution, but the other thing that he brought into trading was uh a money management strategy based on this this work by um a mathematician at Bell Labs called uh John Kelly. I mean introduced this idea of the the Kelly betting criterion, which basically is, you know, uh you assess what your edge is, right? So you have some idea that you you understand what's going to happen better than the market does. You you try to assess what your edge is, uh and then um you you use that plus your current um Well, I mean you basically you that to to to manage your your bankroll, right? There's a a theorem that says what the optimal way of of doing that is. This is something that Thorp came up with in his blackjack days for figuring out how to survive at the blackjack table, right? So, he could count cards. Mhm. Counting cards gives you a a big advantage when you get to the end of the deck. Early on in the deck, you don't have that big advantage. And so, you need to figure out how to manage your bets, manage your money so you're still there when the advantage comes. But, the fact that you have an edge also doesn't mean you win every time. It just means that you have a higher chance of winning. And so, again, it's all about survival. It's all about just managing your money at the table until you get to the place where you do have the edge and then sort of optimizing. Right. That's where all the traders take us. It It It doesn't matter. It matters how much you you know, how much you win on your winners and how how little your losses are on your losers. You know, that's the That's the key thing that you're watching all the time. You know, people think that they're going to get certainty. Like, it's a typical novice investor thinks they're going to figure out a way to just snap their fingers and double their money. Uh and then they find out that that is not the way it is. And then they learn all this stuff we're talking about about managing money, managing position size, etc., etc. It's really cool. Let me ask you something. The book came out in what, 2014? Physics of >> I think it was January 2013. 2013. Okay. So, 13 years on, um w- like how would you characterize Like, if you could char- you know, sum up the book like the 2013 version that that exists, and and then say, you know, what I would do Is there anything you'd say or do differently now, do you think? Yeah. Well, I mean, look, that that that book was written uh at a particular time where there were particular debates happening, right? So, it was really a response to the response to the 2007-2008 crisis. Mhm. Um and, you know, I I think as a response to how people were talking about that, I I I wouldn't change I wouldn't change anything. Uh I I I think I stand by what I I wrote there. I mean, you know, maybe maybe if I went and read it again, I wouldn't I wouldn't I wouldn't feel that way. But, um but I I do think that a lot has changed in market since then. Um Okay. And just so for our listener, James, like have we covered the main conclusion which is like, you know, models screwed us up, but models are are good if they're used responsibly? What more is there to it than that? For you? Uh yeah, so so no, I I I think of that as the the main the main message, right? So, Okay, great. let's understand where these models came from, what people were trying to do. Okay. Let's look at the way in which the people who introduced these models talked about them. I mean, we've talked about um uh Mandelbrot and Thorp and and Bachelier. Um you know, Fisher Black is a another person who I think is really fascinating. Mhm. Uh he wrote a a an article um [clears throat] in the 1980s um called The Holes in Black-Scholes where I mean, it it it could have been written by Mandelbrot. He just goes through and and gives all of these assumptions that are, you know, unrealistic that go into the Black-Scholes equation. And you just list them. And he's like, "Look, you know, these are the ways in which this thing doesn't track reality. Here are the places where it's going to fail. Here are the problems with it. Um and here's why we think you you can use it anyway." Um so Wow. Uh I think when you look at how the people who were originally developing these ideas thought about them, you see a uh like a kind of humility about what the models can do and a kind of sensitivity to figuring out when they're going to fail. And I think that that's that's just a lesson that very easily gets lost, right? So I think most of the people who work with this kind of modeling stuff nowadays, many of them are are coming out of financial engineering backgrounds, they're coming out of sort of specialized training in applying these models. I think it's hard to get the full perspective that the originators had. But I think that that's something that's important to to to focus on. And so that's that's what I see as the big the big lesson. Hey, you don't happen to know a guy named Emanuel Derman, do you? Yeah, yeah, sure. You You know him personally or you know of him? I mean I'm trying to remember if I No, no, I I met him personally. Yeah, I mean we've corresponded over the years, but I went and and visited him at at Columbia when I was writing the book. I like his book Models Behaving Badly. Yeah. >> I think it's I think it's wonderful and I think He's got an awesome book on the volatility smile. I mean don't get me started on that, but but he has a he has a another book that that he wrote on the volatility smile. I'm happy to get you started. I'm happy to pull the rip cord on the volatility smile. I don't even know what that term means. Okay, so the volatility smile So this is this is like a little you know history and philosophy of of Black-Scholes here. When the Black-Scholes model was first introduced, I mean this is the same as when when Ed Thorp was doing his stuff. Options markets were very inefficient. And so you could use the model plus your you know knowledge about the the sort of historical statistics of the underlying asset to come up with a fair market price and you can be pretty confident that when the option expired its value was going to be very close to to what you predicted the option price like the the fair price was going to be. And so you could look for mispriced assets in options markets. Um by the early '80s you couldn't do this anymore cuz everybody was using Black-Scholes. Okay, so um at that point Black-Scholes sort of has a different kind of role. Um it's more like just the the thing that we use to uh set prices amongst ourselves. Um but you can start using it in a kind of inverse way as well where you can infer backwards from the market price of something to what the implied volatility of the underlying asset is. Like what is what does the market think these statistical properties of this thing are going to be? Okay, now fast forward to 1987 after the the um the uh market crash. I mean that whole crash is driven by um failures of Black-Scholes and the the uh portfolio insurance based on it. People start thinking like what's going on with this this equation. They start trusting it less and trading differently and all of the sudden now if you do this inverse calculation this you know by like 1989 you start seeing this very clearly um you start doing this inverse calculation to what the implied volatility is is no longer a number. It's a curve. And it's a curve that depends on how far out of the money you are. Basically how unlikely it is that an option is going to be valuable. Mhm. And one way of interpreting this is that traders kind of got wise to Mandelbrot's ideas. They figured out that actually these extreme events are more likely than Black-Scholes was assuming. Okay, I was going to ask you to make it a little more concrete for me but that's it right there. That's it. Um and so so nowadays the volatility smile is the main thing people use the Black-Scholes equation for. It's to figure out uh sort of what the the whole I mean it it'll become like a volatility surface or a multi-dimensional analysis where um you can see how other traders in the market are thinking about the likelihood of extreme events. Um And so yeah, so that's that's like a common quant technique now and uh uh Derman has a whole book just on that. That's cool. Yeah, I'm looking at I was I was clicking around while you were talking. That's why I was looking at at my computer here. Um at at the book and it's um the blurb sounds great if you're into this stuff and if you know, you know, Derman's writing which I don't know if the writing in this is good as the other one. Um but I really I enjoyed the other book so who knows? Maybe I'll pick this one up, too. Um is there like Let me ask you, do you know the book well? Like is am I going to go in and find a million equations I'm not going to understand? Uh no, there are equations in the book. I mean it's written it's written for a um a sophisticated professional audience. It's not like a Okay. It's not like a um a textbook. Uh-huh. Um but it is a a a book for uh quants. And you know, he's you know, he's he's really interested in in the history and he's really interested in how to think about the volatility smile because it's something that um I mean, the the way I think about it is uh the volatility volatility smile is a kind of contradiction within the Black-Scholes model. The Black-Scholes model is correct, you don't get a volatility smile. We get a volatility smile so the Black-Scholes model's incorrect. But you only get a volatility smile, like you only can can see what it is by applying the Black-Scholes model. And so it's like as a philosopher, it's like a a great example of a kind of circular paradoxical reasoning where it works and it seems to provide a lot of information, but you're contradicting yourself somewhere in there. >> Right. And that smile demonstrates the contradiction for you. Yeah. All right. Interesting. All right. Well, um I'm going to I'm going to add that one to my cart. We'll see. I'm not I'm [laughter] not pulling it. Sometimes I'll just click right as you're talking on the show, you know, cuz I like to I like to read and I I just like good recommendations, but this one seems a little heady even for me. So, um let's let's move on and so and and I just want to actually conclude this. So, you wouldn't change anything that you said about models. You wouldn't change the point of the book from 2013, right? We said that. Yeah. Um but if you were writing it today, it sounds like, you know, with that event farther in the past, you know, maybe you'd have something else to say. I don't know. Well, I I think I would say more about um algorithmic trading. I think I would say more about AI. There are sort of other changes that have happened in in markets. Okay. Um AI, of course, is the one that everybody listening to this just heard you say that. They said, "Oh, tell me about that." Yeah. I mean, the the the other thing that I I think is is really important is the rise of of um passive trading. I mean, so uh you know, ETFs are huge. Passive ETFs are huge. Yes. >> Um and you know, I think I think this is an example of a place where uh look, here's how markets are supposed to work. There's price discovery. You've got people with different amounts of information who are trading on that information and and the market price reflects some kind of, you know, um clearing of of all of this information. Uh um and uh if everyone's trading passively, that doesn't happen, right? Everything is sort of in lockstep by capitalization, and most of the money that's flowing in isn't information doesn't have information at all. Uh and so, I I at some point um markets break under that kind of scenario. Right. And there's um I guess we've had on the show before Mike Green and another guest we've had on the show Hari Krishnan who recently put out some work suggesting, you know, uh the basic suggestion as I understand it is that passive trading could become so large a part of the market it could generate so much and that would generate so much more volatility that there's a non-zero chance the market could actually go to zero. And that we're probably getting to that level of concentration within 5 years. And and Mike uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh uh >> [laughter] >> Uh but others have suggested, you know, there there's something there there must be something wrong with there because they don't, you know, see the the market going to zero guar- you know, um w- they don't see that high of a probability of the market going to zero. Uh yeah, so I I agree that that market going to zero is a pretty pretty extreme outcome. >> Well, it's extreme, yeah. Uh but you know, I I I I I I do think that that we can see really significant really significant dislocations. Um you know, Right. um And you know, and this is connected to all sorts of other related trends like the underperformance of value over the last two decades compared to, you know, the five decades before that and um wild price-to-earning ratios. I mean, this is just normal, you know, um this isn't like you don't need physics to to No. talk about this stuff. No, but but understanding extreme events is is kind of key here, though, right? Yeah, that's right. And I I also think that this this general theme of saying, "Look, you know, our once once models, including portfolio theory, get built into products and into standard practices, we stop seeing the assumptions. Um right, and so you know, once portfolio insurance was a product that you would sell people as an like you conceived of it as a kind of insurance in the '80s, all of a sudden the assumptions that go into Black-Scholes are invisible to you. Um similarly, when you talk about options trading in terms of the Greeks, the Greeks are all based on a Black-Scholes model. And if if the Black-Scholes model's not right, then those parameters aren't the right ones to look at in all scenarios. And I think that the ETF the passive the passive um uh trading is another example where this idea from portfolio theory that you should just diversify and buy the market is a great idea when everyone else is trying to trade on information, but it doesn't make as much sense once that's what everyone's doing. And um yeah, I mean it's a a place where an assumption about how markets work um it kind of goes out of um touch with with what markets are actually doing. >> And and this and this idea that you know, everybody doing what seems like the right thing and then and more and more and more people do it, it's that old thing, you know, the the the fool does it in the end what the wise man does in the beginning. And and it and even more to the point for investors, um you know, there's you you could pick the greatest strategy anyone's ever thought of and if too many people are doing it, it's just going to work less and less well and create more and more risk is really the point. Um >> Yeah. It's the increase in risk, yeah. That's right. Markets like you can't sleep on markets. There it's always changing. Well, there's also there's you know, there's just a general scalability problem, right? So and you I think you can see that in a few different places. One is, you know, the rise of of of private markets. Mhm. Um I I won't go into I won't go in at length about all my views on on private markets, but um it's it's a strategy that worked great 20 years ago. Mhm. Uh or okay, 10 years ago. Mhm. Um it's not clear that it it's sustainable under arbitrary growth. But of course, the more the more equities are traded privately, the more private debt you have, all of that, um the the less diverse the public markets are. Mhm. And um the harder it is to get out of, you know, the S&P 500 blue chip type you know, um uh stocks. Um and then, you know, uh another place is is, you know, quant strategies. I I think they're really interesting examples of strategies that worked really really well when you were um you know, in the the hundreds of millions, a few billion. When you try to scale them up to hundreds of billions or trillions, they just the opportunities aren't there. Um and so there I think there are some examples of very very successful firms that have tried to offer products that they hoped could scale differently. And um you know, just haven't it hasn't worked. It's interesting. Like um everything changes with size. Size is so important. Just um The guy named Vaclav Smil, who you probably heard of, has a whole book called size. And it's just about scale in nature and in life and so on. It's it's it's fascinating to me because you know, we investors see it all the time. Any investor who's been to the to the Berkshire Hathaway annual meetings hears it has heard it every year for decades. Well, well, we're bigger than we used to be, so we're not going to make as much money. And Buffett has said this like every single year for 20 some years now. Um at [clears throat] least, you know, that I've that I've been going or that I you know, I've gone a couple times the past 20 years. So, it's super duper important, the issue of size. What size company are you buying? What size pool of capital are you are your assets in? And frankly, I have to say it's not it's one of those things you discover it's nothing you ever would have thought would have been important, but it really becomes very important. Yeah, I completely agree. All right. Um A minute ago, >> [laughter] >> I don't know if you would have go here, but a minute ago you said I won't tell you all my views on uh on um uh private private markets. It sounded like there was some stuff there. I don't know if you want to get into it. >> [laughter] >> Uh yeah, I I think I better not get into my views on private markets. >> Okay. All right. Maybe it gets a little political or something. That's And that's cool. I respect that. And we And we thank you. Like we try to keep it, you know, real and and uh and not get into areas that like that. So, I appreciate it. >> [clears throat] >> Um But maybe we we could talk a bit more we we I mentioned AI and you and and you know, you you had something that you want to talk about, but I was just going to ask you, do you want to you had you had said at one point that you wanted to talk about von Neumann. Should we talk about von Neumann? Um yes. If only because wow, what a genius, right? I mean, unbelievable character there. Yeah. Yeah, so so, you know, um I've been writing a biography of von Neumann for the last almost a decade now. Um Cool. Very cool. And uh yeah, so so um incredible guy. I mean I think part of what I find so compelling about him is how he thought about applied mathematics, right? That these same sorts of themes that I think have become so important to how we think about financial modeling are themes that really originate with him. I mean applied math wasn't the kind of thing that it is today before really before World War II. And so much of what we now think of as sort of places where you can apply math. I mean game theory, mathematical economics, but also computer science, you know, all of this is stuff that that von Neumann played a central role in in developing including AI. You know, his his last book which was published right after he died was called the computer and the brain where he sort of sketches the idea of an artificial neural network and and and revisit some assumptions that he had made earlier on in his career about the relationship between electronic computers and and brains and sort of began to lay the foundations for for modern AI. Yeah, people who who saw this coming fascinate me cuz I sure as hell didn't. You know, it was all it was all like sci-fi until it practically until it arrived. And then 2022 came along and chat GPT came out and I was as you know, blown away as the average person. You know, I've never been a technical guy. But I let me ask you this. I use Claude every single day. I use Claude every and not only that, I use it on the weekends. Like it helped me clear out my Amazon save for later cart which had over 100 books in it, you know. And you see behind me like this is like not half of it. So, you know, I got to I got to chill out with the books. And I use it for for lots of stuff. Taxes, all kinds of things. And I use it to help me do research, financial research. What do you do with it? Do you use AI? Yeah, so I mean, I I use it for for a couple of things. Um I I use it in in my teaching in in some ways. Um I try to you know, I I try to incorporate into my teaching um ideas about uh what how AI works, but also what its limitations are, what it's good for, where it can cause problems. Um I've experimented with using it in in research and uh I have not yet had great successes. I have colleagues who swear by it. Um and and including like mathematicians who >> I was going to say like hard sciences swear by it. Yeah, and so so, you know, some people claim that that it can help them do new math. Um the kinds [clears throat] of of examples that I've seen that I'm most convinced by is really effective as a um a deep literature review tool. Finding things that are already out there that you wouldn't know how to to look for. That that characterizes my use actually rather well. Um I use something called Notebook LM that you can just dump a bunch of 10Ks and other SEC filings and you know, investor presentations from a company website. You can just put it all in one place and basically tell Notebook LM only look at these sources. Mhm. And it's a fantastic way to run through a whole, you know, say you have 10 or 15 10Ks and you want to know historical stuff. Um fan- fantastic tool for doing that. But the thing that really gets me is um how how Claude has like a voice. I mean, it's like it's you know, the latest and I I think I forget what that latest model is it called Opus or something? Um and it's um it makes connections or helps me make connections that I might not have made otherwise. Um partially I realized just by the sheer volume of of ideas that it can put in front of your face. And and it'll connect something and it'll say, "Oh, this connects to your earlier idea." And I'll say, "Whoa!" You know, just because I'm a human and I can't make it can make 30 connections and you know, in a split second or something and I can make 30 in like 30 years. You know what I'm saying? It's just uh it's a little bit faster than Dan at the age of 64 is all I'm saying, I guess. >> [snorts] >> But it's useful. Yeah. Of course, that's the problem with it, isn't it? That it's so human and so useful. And I I I'm trying to convince my stepdaughter, you know, it's a terrible therapist. Don't tell it your personal problems. [laughter] You know, and she keeps telling it her personal problems. Like, "Chat GPT said I should do this." So so you know, there's genuine concern out there and are you are you Where do you come down on the scale of, you know, people saying it's going to kill us all and you know, become Skynet and you know, kill us all in the future or put us all out of work? Do you have those fears? Um I mean, put us all out of work. Yeah, I work at a university. Uh Right. >> There are a lot of ways in which it can put us out of work. Mhm. Um you know, one way is that we can't figure out how to to continue to make you know, education a meaningful experience for students, right? So if if it becomes a situation where students are using AI to do their homework and professors are using AI to grade the homework, what's the point? I mean, why why are any of us doing this? And uh once that goes away, you know, >> the time, I might add, James, when the university it's not the sort of hallowed institution that, you know, before like we didn't question it. Everybody should go to college. Now, that is not a that is not as popular an idea at all, is it? Yeah, that's right. Well, and you know, I'll I'll tell you, it's the computer science majors who engineered themselves out of jobs. Right. Right. And so, we had this I think very very strange, you know, labor market situation where AI is booming. Everyone says, "I want to get involved in the AI boom." And so, I'm going to to learn what I need to learn to go and get these jobs. But then AI got good enough even before they had graduated that you just don't need programmers anymore. I mean, this is the place in my my research where I use AI the most. Um, you know, it's just integrated into to sort of Generating code? Generating code? Is that it? >> Generating code. Exactly. Yeah, and it's it's incredible at that. I mean, like it can read your mind. Like, you know, you type the name of a function and then it produces the function. It's like, how where in my in my code? I mean, I I believe I'm the first one writing this program. Like, where in my code did you figure this out from? But, yeah, it it it can do it. Wow. Okay, so maybe um you know, maybe we'll I don't know, maybe one day we'll get you back and we'll just talk about AI. But, it's it's it's been great talking with you. It's time for our final question, which is the same for every guest, no matter what the topic, even a non, you know, financial professional guest like yourself. Um same identical question, no matter what. But, it's simply this. It's for our listeners' sake. If you could give them one takeaway, one thought today, what would you like to leave them with? It can be anything. Yeah, I mean, the the my one thought is think meta. Not That's not a stock tip, okay? Right, right. Think meta about the ways in which um the financial products and the financial services that we're participating in are built on now these mathematical models that assume stuff about what the world is like. And it's very very hard to see those assumptions, but everything that we do can break when those assumptions break. The more integrated they are into our practices, the the harder they are to see and the more dangerous they become. All right. Wise words. Thanks for being here, James, um and I look forward to having you back on the show sometime. Yeah, thanks a lot, Dan. This was fun. That was incredible. For me, at least, I hope it was for you, too. It was very highly educational. I recommend James' book, The Physics of Wall Street. Um it's it's actually very well written, and it's not full of technical mumbo jumbo and equations. It's full of really good stories about important things that people mostly don't know a whole lot about, which is fun, right? It's fun to learn about these historical figures and why they're important and what they said that nobody was saying at the time. Um and it it'll teach you something about kind of how markets work in a way that you're just probably not going to read a lot of other places. I mentioned the book by um Mandelbrot, which is uh Misbehavior of Markets. We mentioned Edward Thorp's book, Beat the Market. Um we talked about John von Neumann and some other people. So, there there's other books by, you know, the the innovators [music] themselves, the thinkers themselves. Um but you usually you don't often get a smart guy like James who wants to make it so that everybody can understand it so that he can explain it to you. So, you know, that's why I like the Physics of Wall Street. That's why we wanted to have James on the show. I was like, a guy who can tell stories about math and finance so that a guy like me can understand it and really enjoy it. That's got some value. All right, I think I hope it has value for you, too. So, I recommend the book and I recommend you learn about all these characters, Bachelier and Thorp and and Mandelbrot and all the rest of them. So, that was another great interview and another episode of the Stansberry Investor Hour. I hope you enjoyed it as much as we did and remember, hit like, hit subscribe, and by all means sign up for our free daily email. Opinions expressed on this program are solely those of the contributor and do not necessarily reflect the opinions of Stansberry Research, its parent company, or affiliates.