The Surprisal — The Rationality of Gambling With Rob Carver and Burak Yenigun 00:00 Opening Rob · 00:00 So you're a bricklayer, you're sixty-four years old. This is your state. The only way you'll get to a better state is by doing the risky investment. And the question is, are you willing to pay fifty-three basis points of your income to have that possibility, no matter how remote? And to me, I don't think that's an irrational decision in the sense that I think that's a fairly reasonable utility function to have. 01:33 Should billionaires and bricklayers have the same investments? Burak · 01:33 Rob, I've been waiting to do this for a long time. I think I've planned this for couple of years now, but everything had to fall in place. The right podcast music had to be in place and stuff like that. Rob · 01:44 The important things. Burak · 01:45 Yeah, this is— Rob · 01:46 The music is most important. Burak · 01:47 Exactly. So finally we are here, and I wanted to record this first episode with you because your blog post, "Should the Billionaires and Bricklayers Have the Same Investments," that was probably the single biggest delivery of surprisal that I read. Since 2008, I guess, I've been in finance full time. And so after all that time and all that reading over the years, that single post delivered the biggest change in my understanding and the best explanation for a puzzle that had been bothering me. And the puzzle is basically, why does it make sense to gamble for some people, effectively? The rationality of gambling. And this was an open question for me since I was in university, because I remember a friend's father, he would basically buy a lottery ticket every week, and he would say, "You know what? If I win the lottery, I'm not gonna go to work on Monday." So he would do this every week. And then obviously, you take a bunch of finance courses or you study finance and investing, and everybody tells you that that's irrational. But then I couldn't quite shake off the feeling that there is something fair about what the guy is trying to do. And we will dig into that further. But yeah, so I guess maybe a good place to start is my favorite line from your blog post, where you say, "People who don't/can't use leverage and need or want higher returns will pay for risky investments, even if they have a negative expectation." So why is that the case? 03:24 Why people pay for risk Rob · 03:24 Yeah. So this is actually not an original idea. Almost nothing I say is original. A lot of it is kind of stuff that's sort of reprocessed from, in many cases, quite old ideas. And in classical economics, this idea goes back at least fifty years, probably longer, maybe even a hundred years: you wouldn't necessarily have what an economist would call constant appetite for— or constant risk aversion, right? So the idea is that whenever anyone's evaluating any kind of uncertain outcomes, they have to weigh up two different things, right? The potential gains, but also the potential variation in those gains, the potential risk involved. And economists love setting these stylized puzzles, and sometimes actually they do it in laboratories. So they'll get a bunch of students and they'll say, "Okay, do you wanna play a game where you can earn ten dollars for sure or have a fifty-fifty choice between earning six dollars and losing four dollars?" All this kind of stuff, right? And these games are, to be honest, a bit pointless because of the budget of these projects isn't that high. They can't really offer you a million dollars, which is what you can win in a lottery. And because the stakes are so low, even for a student, ten dollars is not a lot of money, probably. It's like a couple of coffees at best. I don't think it necessarily reveals that much about how people think about risk and return in a sort of an informed way. So there's that sort of strand of the literature. So there's for a long time been this idea in economics that essentially the richer you get, the more risk-averse you should get, which doesn't seem that illogical, because a lot of the really rich people in the world who have got what I would call generational wealth, many of them mostly concerned with keeping that money, right? And they're happy to accept quite low returns, and they're happy to pay high fees to private banks and investment managers and all this kind of stuff. But then you have some exceptions, right? So you've got someone who, as you know, Burak, I love and admire so much. He's one of my favorite people, and that's Elon Musk, who… Burak · 05:30 I know how much you love him, yes. Rob · 05:31 Oh, how much I love him. I actually own six Teslas. That's how much I love him. He's the richest guy in the world. But also a massive risk taker. His whole career has been a series of risks. So it's an interesting question as to whether he's an anomaly, but you're a big fan of founders, right, Burak? One of the things I know you're quite interested in is the psychology and the motivation and the sort of possibilities of people founding businesses. And those are mostly risk takers, right? They're people who naturally take more risks. So that's an interesting kind of counterexample. And then the other thing is that there's sort of well-known research in financial economics where the idea that essentially the market should pay you a certain amount for the risk you're earning. Basically, this idea from the nineteen fifties about the beta of a stock explaining all your returns. There's this well-known effect that high beta riskier stocks actually earn much less on a risk-adjusted basis than low beta stocks. Burak · 06:31 Yeah. So people pay for risk. People are willing to lose better risk-adjusted returns. Rob · 06:38 There is almost no risky investment has a positive expected return, right? If you analyze the odds in, say, a horse race, so you compare the odds you're being offered to the actual outcome or some other statistical model, you'll find that on the favorites, you have a much better kind of net positive expected return in the long run. But on the outsiders, you have a negative expected return. So people are willing to pay effectively over the long run for the possibility of winning on, say, a hundred to one horse, rather than accepting the relatively low returns on a two to one favorite, a one to two favorite. Burak · 07:18 Can I add something here? Rob · 07:19 Sure, of course. 07:20 Leverage & the Kelly criterion Burak · 07:20 So, you say that this idea has been around in economics or finance for some time. But I don't remember ever seeing anything about why it is understandable for somebody to gamble. Of course, it's fantastic if you can accept a positive expected return. But I think in your post, basically what you say is, look, as you said, with the high beta stuff, as you go further along the risk curve, you get hit with expected returns. So investing with positive expected return eventually, as you take more and more risk, becomes gambling with negative expected return. There are a few mechanisms for this, I suppose. One is that people overpay for riskier stocks, the high beta assets, so they're actually more expensive and therefore their expected return goes down. And the other mechanism is even if you have a fantastic investment, as you say in your post, if you have a one Sharpe ratio expected return investment, but if it has a volatility of twenty percent per annum on a unlevered basis, if you are somebody like my friend's father, what you need is a ten X return in the next one, two, five years. So to achieve your desired returns, you have to lever up that investment. And when you lever it up, because of the Kelly criterion, the maximum that you can run that with is like one hundred percent volatility target. And the moment you get to two hundred percent volatility target, it basically becomes zero expected return. So you don't really have a good menu of options. Rob · 08:49 No, you haven't. I mean, this is a thing, right? So it's all very well for economists to say, "Oh, you should leverage up as appropriately." But that assumes leverage is available, and it's not necessarily that available. But there's a few different things going on here. So effectively, by adding leverage to a, quote, unquote, "good investment," ultimately you effectively turn it into a lottery ticket because you turn it into something that has a negative expected return, but a small probability of a very high payout. So you can actually manufacture negative returned assets with massive possible outcomes just by taking leverage or by trading options or by doing various other things. But if you go beyond Kelly criteria, essentially you're then putting yourself into a situation where you've now got that negative expected return. So in a sense, the lottery ticket and the stock are not different. It's just that one already has the leverage built into it, in effect. You look at the payouts of them, you can manufacture a lottery-like payout out of a Sharpe ratio one investment just by adding enough leverage, essentially. And it may be actually after you've done that, you may get to the point where the payoff profile of the Sharpe ratio one asset is potentially even worse than the lottery ticket. Who knows? Because it's an insane level of leverage. So it's less about Sharpe ratio one assets potentially with leverage, which you may not be able to access versus lottery tickets. It's more why, as you say, why are people willing to effectively pay money for the possibility, a small possibility of winning a lot of money? 10:29 Gambling & utility functions Rob · 10:29 If I think nearly all the work I've done, a lot of it comes down to this same core idea, which is the idea that people don't have the utility function that classical economics says that they've got, and different people have different utility functions. This is from a book called Fortune's Formula, which is about the Kelly criteria, which I highly recommend to you. So this is not an original example, but think about the utility function of, say, someone who's relatively young, like yourself, Burak, you're much younger than me, who's got a reasonable income, can save money, has got a long time before they retire. Think about their utility function or think about the utility function of, as I said, someone who's like third generation wealth, who's got a few hundred million dollars of family money in the bank. Think about their utility function. And now let's contrast that with someone whose utility function is absolutely extreme. Okay? This is not my original example. And this is the version of your relative. Was it your uncle, did you say? Burak · 11:30 No, it was a friend's father. Rob · 11:31 Not a relative. Your friend's friend. Burak · 11:32 Yeah. Rob · 11:32 But this is like an extreme version of that guy, right? This is the guy kind of pushed to the absolute maximum of the continuum. So I want you to imagine a rather unpleasant scenario. I want you to imagine that your wife and your child has been kidnapped, and the kidnappers say to you, "We need to have, say, ten million dollars in an hour's time or they're both going to be executed." Okay? And let's also suppose, just for the sake of argument, I'm not saying this is true or not, suppose that you have in cash liquid available to you at the moment about, say, three hundred and fifty thousand dollars of cash. So suppose, okay? The rational thing for you to do is not to ring your broker and find a Sharpe ratio one investment, okay? The rational thing for you to do is to find a lottery-like investment that will pay off ten million dollars. It's about a thirty-to-one option, so that's roughly what you get placing one number on a roulette wheel, right? The rational thing for you to do is just to take that three hundred and odd thousand dollars, go to the casino, put all the money on any arbitrary number, and cross your fingers and hope like hell. Even though you know that's a negative expected bet, placing roulettes, there's no skill in roulette. You can't count cards. The people have tried to predict it with computers and that's another story. But you have to get that negative expected return because your utility function says that you have basically a very binary thing where you need to have ten million dollars in an hour or else. That's it. Burak · 12:58 Here is the problem I have. I completely agree with, obviously, this clear example. I feel like there is a lot of people who are perhaps quite not as extreme as— Rob · 13:08 Well, I hope, I hope there's not many people in that situation, Burak. Burak · 13:11 Right. Rob · 13:11 I really hope. Burak · 13:12 Before we had this conversation, I asked ChatGPT, I said, "Look, I'm looking for any papers that basically talks about this, the rationality of gambling." Or maybe rationality may not be the correct word here, but the sensibility of it. And it didn't bring up anything at all. But apart from that extreme example that you gave in your blog post, there is not a lot of text out there that says, "Well, here is why the negative expected return bets in the form of a lottery can make sense for a lot of people." So I think that, for example, again, to go back to your example from your blog, you say here, "Imagine you're a sixty-four-year-old bricklayer who will be retiring next week. You only have a state pension and no other investments except ten thousand dollars in cash. Economically, you own an annuity, the pension, worth perhaps hundred and eighty thousand pounds plus the cash, which is roughly five percent of your net worth. Is the best use of ten thousand dollars to invest it in a Sharpe ratio of one opportunity, which will return ten percent, or to buy the lottery tickets? The latter is more likely and also makes more sense. A thousand pounds isn't going to make any difference at all, I guess a thousand pounds per year. But in the two million to one or so chance of a lottery jackpot and winning ten million pounds, the bricklayer could be much better off." Now, this is another example, right? This is not an extreme example. Rob · 14:42 But this is actually… We're on a continuum here, Burak. We've just moved a little bit away from casinos. And we're probably now a bit closer to your relative, although not quite there because I'm guessing they're not sixty-four, so they've still got a bit of time. Yeah. Burak · 14:55 Yeah. And I feel like there's a lot of normal, regular people who are in similar situations where it makes sense to gamble, and it's not talked about. Another example. Imagine a young person, perhaps they have a reasonable income, and they expect to earn more because maybe they expect to do better in their career in the future. They can think of ways to increase their income and increase their savings and then just build wealth slowly. But maybe there's a large group of people who don't feel like they have those options. I like to think of it almost like a intangible balance sheet, right, like intangible assets of a young person, right? What is that? They have the time, they have the future earnings potential, et cetera. The larger that balance sheet, the less you have to take those risks. Or maybe you feel the ten thousand dollars that you have in bank right now isn't that much money compared to what you think you will earn in the future. So you feel comfortable effectively gambling with it. That's why you do things like, I don't know, one to ten leverage crypto bets or stuff like that. Rob · 16:00 Yeah. But I don't think these examples are that different, actually, because in the case of the bricklayer, the kind of existing capital they have, you can see it. It's a value of an annuity. The state pension, you can actually annuitize that and say, "What value does that have?" In the case of the young person, it's perhaps a bit more intangible. But in both cases, you're sort of saying, "Well, I have a certain amount of expected income in the future." Okay? And that could be guaranteed by the government, by an annuity in the case of the bricklayer, or it could be kind of future expectation of the future. And then you're saying, "Right, given I have some amount of sort of excess capital," if you want to call it cash, spare cash, whatever, "ultimately I have two options. I can either invest that in a careful way, and it might… it'll add basis points to my future income." Now, we can get a bit of an argument about that because I think a lot of people misunderstand, for example, the power of compounding and the fact that a thousand dollars invested at the age of twenty-one can actually be worth quite a lot of money by the time you get to my age or sixty or seventy. Whereas for the bricklayer, actually, they're gonna die before the young person is expected to, right? So the power of compounding is gonna work less in their favor. So you can be a bit more precise about… In fact, you can even almost put a price on what would happen to that marginal bit of money if they put it into their pension as an additional amount of money. For them, it probably will be basis points. I think I put fifty-three basis points in my post. So the question is, we're trying to consider these two options. Basically, we have a small probability of being very wealthy in the future, much wealthier, and having a much better lifestyle. Or the alternative of having a more certain increase in our wealth or income that will be much smaller. Rob · 17:48 So it's basically weighing up those two probabilities. Now, maybe there aren't any papers specifically on that topic, but there is, of course, and I think it's in my blog, the whole field of prospect theory, which is the core part of behavioral finance that potentially explains things like momentum effects, which you and I are both very fond of. And underlying the prospect theory is the psychological idea that basically people are rubbish with small probabilities. Okay? People always overestimate what the likelihood of a small probability thing happening. And you can use this to explain vast numbers of cognitive biases. So for example, people are really scared of being attacked by terrorists on planes, even if you include 9/11, I mean, you're more likely to die crossing the road in London by a factor of ten thousand. But it also explains, of course, these, to us, potentially irrational investment decisions. Because although we know that the odds of winning the lottery are millions, perhaps tens of millions to one, a small part of us is like, well, I mean, actually, the National Lottery had this brilliant advertising slogan a few years ago, which was the phrase, "It could be you." And that kind of sums it up, right? So people focus on this tiny probability. And I think part of the thing that drives that is another cognitive bias, which is the fact that people overrate anecdotes versus statistics. So when they think about "Will I win the lottery?" they don't think about the thirty-million-to-one odds. They think about the stories, which lottery people, of course, publish, of all the people who've won and so on and so forth. So that's the kind of rational behavioral economics theory about it. Now, I also have a theory, if you will forgive me, and this potentially is an original idea, but I also have a theory as to why I think it's got worse in the last… 19:28 Land theory of gambling culture Burak · 19:40 Oh, I have my own theory, so I— Rob · 19:42 You have your own theory? Burak · 19:43 Yeah. Rob · 19:43 Eighteen years. I'll say eighteen years. Well, let's hear your theory first, and we'll see if we're thinking on the same page. Let's do that. Burak · 19:51 I'm pretty sure that we're gonna have different theories, but I blame all— Rob · 19:53 Oh, right. Okay. Burak · 19:54 I blame all of this on land, basically. Rob · 19:57 Okay. Whoa. No, I did not see that coming. Burak · 20:01 As you know, I guess I'm always worried about putting these labels, but I guess I'm gonna call myself a Georgist, I suppose. But I think Georgism has had incredible predictive power. I mean, the guy wrote a bunch of books in, like, 1880s and predicted that San Francisco would be rubbish, or more rubbish than New York in terms of wealth inequality or just the extreme desolation or poverty if the land prices go up. And that's exactly what happened. But also, more importantly, the whole thing checks out. Taxing the land rent more heavily seems to be a rare case of economists agreeing on the same thing. But basically, I think, and again, this goes back to targeting high returns and high risk with a positive expected return problem, right? You always hear this story in the US, and I think in the UK as well. There was that TV show that followed young people, well, from six years old to, like, sixty or something. Rob · 21:01 Yeah. Yeah, yeah, yeah. Burak · 21:02 And they're just… Rob · 21:02 Yeah, it's really interesting, actually. Burak · 21:05 Yeah, that's a very, very cool one. And I think it was the taxi driver gentleman who bought his home in a… I think it was a London borough and on a single-person salary, right? So whereas obviously that's no longer possible in London, that's no longer possible in San Francisco or New York. It's just the land has gotten very expensive. And so the common theme there is it was possible to buy cheap property and lever it up with a mortgage, right? What is the volatility of a real estate… Well, let's just call it fifteen percent, right? Obviously, it's hidden. You don't get to see—really see a price ticker, but the hidden volatility, let's call it fifteen, twenty percent. But you had that investment, and what is the typical mortgage? Maybe you did twenty percent down payment, so you had a five X leveraged asset, fifteen percent vol, like, seventy-five percent vol target, let's call it. Of course, this is an imperfect analogy, but— Rob · 21:54 No, no, I don't think your numbers are far off, and I kind of agree with you, but go ahead. Burak · 21:58 Yeah. And then obviously, the land has gotten more expensive as the productivity went up over the decades. And Henry George's argument has been that a lot of the gains of productivity and better economy, it just accrues to landowners in the form of higher rents, right? We've all experienced this in London, probably, or UK more generally. So this was the people's path to wealth-building in the past, among other things, of course. But in the US, certainly, in the UK, and even in Turkey, land getting more expensive has been a tremendous wealth generation tool. Young people today don't have that option. First of all, you can't even save the down payment for the house like that. Rob · 22:40 No. Burak · 22:41 And there's policy after policy that tries to fix this, but it might make the problem worse by subsidizing demand. So I feel like there's a lot of people who feel that they don't have as many options on that front, just having a regular income plus levering up the house isn't there. That's my theory. 22:58 Envy theory of gambling culture Rob · 22:59 I one hundred percent agree with you. So, I think it's a good theory. I think it's actually, though, only part of the what I would call the grand overarching theory I'm gonna introduce. So— Burak · 23:10 All right. Well, if you do— Rob · 23:11 Well, okay. So it's definitely true that the opportunities and the perceived opportunities for younger people are not as good as they were, and there's lots of reasons for that. I think you're absolutely right. I think house prices have a big part to do with that, definitely. I think that the two thousand and eight financial crisis has a lot to do with it. I think the sort of huge effect of the baby boomers. Just to say, anyone who's listening to this who assumes I'm a boomer, I'm not a boomer, okay? I'm Gen X. I'm actually kind of right in the bang in the middle of Gen X, so I'm not a boomer. But I think of the boomers as a sort of python eating a small deer. And they're sort of going through the stomach of the python. There's a big bulge. So the demographic bulge of the boomers had a big effect, and now they're pretty much all retiring. That, again, is gonna have a big effect, and there's lots of potential implications of that. But those guys have all the money, and they're not necessarily willing to share that. The sort of two thousand and eight financial crisis, I think, had a almost permanent scarring effect on the opportunities available to people who— Burak · 24:22 How so, though? How do you think that two thousand and eight had that effect in terms of opportunity? What is the mechanism? Rob · 24:29 Well, obviously it caused a recession, and it was probably one of the most significant recession for quite a long time. Anyone who kind of graduates into a recession, it kind of causes a sort of scarring effect on the labor market. Because obviously what caused the two thousand and eight financial crisis goes back to land, doesn't it? Right? Mortgages, land and so on, it's all tied together. So I think that that kind of made this sort of mechanism of cheap money coming to people through house price appreciation, being able to borrow very easily. It all kind of got completely out of hand. Two thousand and eight happened, and since then, there's been much larger restrictions on things like, for example, borrowing one hundred percent of the purchase price of a house. And don't get me wrong, I think those restrictions are potentially good because that wasn't necessarily a good thing, but it has scarred that. So anyway, let's put that to one side, because that's the bit I think we both agree on. Rob · 25:23 Although I think there's slightly more to it than house prices, for sure, that's a big part of it. The other thing I think that's happening is what I would call just a generalized increase in envy. And you could argue this goes back to when magazines first came out, because if you go back five hundred years, the peasants knew that the lord had a much better lifestyle than them, but there was no way they would aspire to it. There was no possibility of their children ever getting into that. The absolutely best-case scenario was your son, and it would be a son, sadly, would learn how to read and get a job, quote-unquote, as a monk or something, writing scriptures. Or perhaps, they were smart and they could become a kind of small independent businessman or a cobbler or something like that, right? You weren't ever gonna become a lord. There's no possibility of that. But then gradually, as magazines were introduced and then TV and then films and then the sort of rise of celebrity culture over the last forty years, and now it's all come to a head with social media and the influencer. Young people in particular are surrounded by pictures and images saying, "You should be like this. This is what you should have. This is what you deserve." Right? And so it's focusing people's minds on the possibility, even though that possibility is almost certainly incredibly unlikely, and it's really reinforcing the cognitive bias and saying, "Well, there's only a small chance of this happening, but it could be you. You deserve, you deserve it for it to be you. You deserve to be the person having all of these nice things and all these wonderful things." And it, of course, doesn't help that simultaneously a lot of these influencers are peddling methods which will supposedly make it very easy for you to, whether it be crypto or multi-level marketing or what have you, to get into that space. So I think that's also potentially a big issue, because it's not just that people are being shown the possibility of something that is frankly unobtainable to most people. It's also the fact that it's making them feel bad if they're not at least trying to get into that situation. What do you think? 27:32 Markov chains & diminishing marginal utility of wealth Burak · 27:34 Yeah, I think… I like my theory better. Rob · 27:39 The nice thing about, Burak, is that both of our theories can be simultaneously true. They're not mutually exclusive. But anyway— Burak · 27:45 I will say, though, I think the version of your theory that I believe in is perhaps this one. I feel like envy or kind of the desire for more was always there, perhaps. Rob · 27:55 Yes. Burak · 27:55 But I think there were two things that happened, I suspect. One was declining marginal utility of wealth. Somebody introduced me to this concept. I saw this on Twitter, I think back in, it must be like six, seven years ago. And I think she had a very strong point in that let's imagine with some arbitrary numbers, right? Like, if you get from one hundred thousand dollars of net worth to two hundred thousand dollars net worth, that is not a massive difference in today's living standards, right? In terms of what more you can do with it. And I think it was Nassim Taleb talking about this. I think he was talking about with these Markov chains, right? So in the Markov chain, and I have a hazy memory of this from back in university, but some Markov chains have a absorbing state, and wealth is an absorbing state. Arguably, it shouldn't be. I guess this is the part where Nassim Taleb was talking about it, like wealth is an absorbing state right now, and once you get wealthy, you cannot really lose it. Well, unless you're Elon Musk or something and taking crazy risks, and maybe that shouldn't be the case. So the problem is you have various states and you're trying to get from one of them to the other one. So, for example, you may not like your job, just like my friend's father. You may want to be in a situation where you no longer have to work, right? Or maybe you no longer have to work, but instead of a house in London, maybe you want a house in London and, I don't know, like a summer place in Spain or something. I don't know. But then… Actually, this is a bad example because the difference between these two is very, very low. I don't know, maybe you want a super yacht or something. You know what I mean? There are maybe five, three, four, five distinct categories of wealth. And this is a technology-driven thing in part, I think, because we all use the same tech tools. We can all get on a plane and go somewhere. So you don't really need that much money for a lot of things. To make an actual difference in your life, you need a lot of money. And that's why I think that small ramp-up in wealth is no longer satisfactory to you. You push for bigger jumps, perhaps. I feel like this is another effect overlaid on top of what you're talking about. Young people just don't want to be going to work and slowly building wealth. They want to get to, I don't know, financial independence or maybe they want a yacht, to actually feel like there's a difference in their life. And of course, and this is the optimistic view, and the pessimistic view, I guess, is that they legitimately don't really have a lot of good financial options in front of them. Obviously, all of these happens at the same time, but I think part of the kind of Instagram effect that you're describing is, in a way, attached to this effect. Rob · 30:24 I think there's definitely something in that. I think actually the other person who talked about wealth as an absorbing state that way is Thomas Piketty in his massive book, "Capital in the Twenty-First Century." Yeah, no, I think there is something in that. You're right. 'Cause the thing about wealth is in terms of as an income generation, let's say you can earn four percent, say. So an extra hundred thousand dollars of wealth sounds great, but it's only four thousand dollars a year of extra income. And if you're already earning, say, a hundred thousand dollars, well, that's four percent extra on your income. As you say, that's not gonna take you from kind of comfortable middle-class life in London to billionaire status, right? To get to the next step, whatever that might be, next step is, I don't know, kids at private school, house in the country, second home in France. You're gonna need a lot more than that. Let's track this back to utility functions. Essentially, many, many people have sort of binary utility functions or utility functions a step up where they don't really value any sort of increase in wealth until it gets to the point where they can get a meaningful new, let's say, let's be really harsh and call it a new toy, right? So until you can get a meaningful new toy, the extra step up in wealth isn't really gonna help you. And of course, once you're at the sort of tens of billions of dollars stage, actually, I guess the only meaningful new toy you can get is Mars. So maybe at that point you— Burak · 31:58 A villa on Mars is gonna be the new flex after we get all this AI-driven abundance. Rob · 32:02 So maybe that's why Elon is still taking risks, 'cause he sees that next step up in wealth. He's not happy with the five hundred billion dollars. He needs the extra step up in wealth so he can afford to buy Mars or get to Mars or control Mars or whatever. Yeah, I think that's an interesting theory. And as you say, if you're a relatively young person stuck at the kind of almost the first state and what you're seeing on Instagram is almost three, four states ahead of that. So, you're almost not gonna value anything between where you are and getting to that future point. Burak · 32:34 Exactly. And you need ten X more. And how are you gonna get— Rob · 32:36 At least. Burak · 32:36 Ten X more in five years? Well, okay— Rob · 32:38 At least, yeah. Burak · 32:39 If we had to run the numbers, like ten X in… Is that like fifty percent per year over ten, five years or something? Rob · 32:44 Well, the rule of thumb is the divide by seven rule, isn't it, really? Yeah, you've probably got access to a computer. Burak · 32:49 Oh, okay. Okay, so actually this is not bad at all. If you do fifty percent per year over ten years, you get like sixty times your money, which is—we'll take it. Rob · 32:59 Easy. Burak · 33:00 Easy, easy. Rob · 33:00 Easy. Burak · 33:00 Exactly. Here's the problem, though. Young people don't want to wait for ten years. Rob · 33:05 No. Burak · 33:05 Maybe they're gonna wait for five years, in which case what you get is just eight times your money, like seven point five times, which is good. This is the ballpark, I suppose. Rob · 33:14 Yeah, but it's still fifty percent consistently. Burak · 33:17 That's the problem. So you have to have a one Sharpe— Rob · 33:20 If you're running at half Kelly… Burak · 33:22 Yeah, fifty percent vol target or something like that, right? Rob · 33:24 And also, fifty percent vol is terrifying. Most people don't understand how risky it is. Burak · 33:28 But it's not terrifying to people who are levering up crypto by ten. Rob · 33:31 Well, no. Quite, indeed. Yeah. 33:34 Gambling: why is it rational for some? Burak · 33:34 Yes. Actually, to tie this back to the original point, I read your blog post, and I think you're laying out a framework, a way of thinking. And at the end of that thinking, you reach the point where you say gambling is rational. Would you agree with that, or am I taking it too far? Rob · 33:51 So, okay, rationality is a very difficult concept. Basically, the way I see it is people have utility functions and then they maximize those utility functions, right? If you have a certain utility function, then gambling makes sense to you, right? It optimizes your utility. Now, then we take a step back and say, okay, the utility function you have—is it rational? Okay, so we've discussed a couple of cases where it seems eminently reasonable for someone to have a utility function that means they should gamble, the sixty-four-year-old bricklayer, and the more extreme version is you in the casino with your wife sort of in the house with a gun to her head. Okay. But for the vast majority of people who are gambling, and by gambling I also include any investment that has a negative expected value but a high risk, okay, including crypto. Done wrong, of course. For most of those people, I would argue their utility function is not rational. We can justify it, and we have justified it, and we've theorized as to why it might be the case, and I think our theories are interesting and may well be true. But they do not necessarily justify the behavior that people have, because ultimately I think what people are doing is focusing on an outcome that's less likely. In terms of my framework for how I think about judging probabilities and expectations, and this is something that I keep coming back to again and again and again in my work. So my most recent book, I discuss it, that's coming out in December. In my next book, which will be on backtesting, again, I'm going to discuss it. Basically, this is the idea that whenever you're making any judgment about an uncertain outcome, you need to think about which point of the distribution you're optimizing for, okay? And if you are like what an economist would call a risk-neutral person, a completely rational person, maybe a complete rational person, but someone who is not concerned about things going wrong or the probability of downside, then you should focus on the median fifty percent of the distribution. If you've got a choice of different distributions, different lottery games, investments you can play, you should look at the median of those distributions. Now, the median payout of a lottery ticket is zero. Okay? The median payout of a Sharpe ratio one investment with twenty percent vol is like nine percent, roughly. So in that situation, you should choose one over the other. And then I would further argue that most people should actually be optimizing for a point lower than the median. So if you're an ultra, ultra cautious person, you maybe focus on the one percent or fifth percentile. And that means, for example, you would do things like if you couldn't use leverage, you'd buy a portfolio mostly consisting of bonds. So actually, this also kind of is a way of expressing the idea of different portfolio risk preferences in a way I think is more natural than the way that economics traditionally does it, but that's another point. Now, what these people who are buying these lottery tickets and so on and so forth, is they're focusing on a much smaller percentage of the distribution above the median. And that, I believe, is not rational unless, as I said, there is some compelling reason as to why you should do that. And I think— Burak · 37:06 So just trying to get to an absorbing state. Rob · 37:09 Exactly, yeah. So you're a bricklayer, you're sixty-four years old. This is your state. The only way you'll get to a better state is by doing the risky investment. And the question is, are you willing to pay fifty-three basis points of your income to have that possibility, no matter how remote? And to me, I don't think that's an irrational decision in the sense that I think that's a fairly reasonable utility function to have. But on the other hand, if the price of that jump was fifty percent of your income, I'd say, "Well, no, that's completely insane. That makes no sense at all." Because you're actually gonna go down a state, right? You're gonna be in a state where you're barely surviving. To do that for a million-to-one chance of jumping to a much higher state, no, that doesn't make any sense. Rob · 37:59 So I do think it does come back down to, again, another economic concept, which is the price of risk. But it's more like, what's the price I'm willing… So I had a very smart guy I work with called Darren, and we were discussing lottery tickets. And we discovered that despite being quants, and I think he had a PhD in math, and being very familiar with all of this stuff, we both occasionally bought lottery tickets. And I still do buy lottery tickets. 38:05 Why Rob buys lottery tickets Burak · 38:24 A big reveal. Rob · 38:25 I should have said that up front at the start of the episode. But actually, my rule of thumb is I only buy lottery tickets when they've got positive expectation, which on the EuroMillions jackpot is, broadly speaking, when the jackpot has rolled over to be more than about one hundred million pounds. So I actually don't buy that many lottery tickets. But we both bought lottery tickets occasionally. And he said, "Well, the way I think about it is this. Me buying a lottery ticket is me buying insurance against being lucky." You know what? If I put that in my blog post, then maybe we could have saved all of this, because actually that's quite a nice way of thinking about it, right? But the question is, how much are you willing to pay for that insurance? Well, I'm probably barely spending like three basis points of my income on lottery tickets a year, maybe, I don't know. A tiny amount. Burak · 39:10 So basically, you're not risking your current state while trying to… No matter how faint that probability is, and as long as you're not risking your current state, that nonzero probability is better than a zero probability. Rob · 39:24 Exactly. So then it comes down to how much you're willing to pay for it. And as I said, fifty basis points a year, yeah. A hundred maybe. When you get to ten percent of your income on lottery tickets, I think that that's just silly. And there are people who gamble and who spend very large percentages of their income. Burak · 39:44 There's only one scenario where I feel like that could be rational, or understandable at least. If you are really, really low income, and let's say, you don't see any potential path to having a more middle-class life, maybe that ten percent of your annual income is very small. Rob · 40:04 Well— Burak · 40:04 And then you're trying to get from that desolation state to a higher state. Rob · 40:07 I'm gonna disagree with you, Burak, because the thing is, even let's say I spent ten percent of my income on lottery tickets. It would have no real perceptible, substantial effect on my lifestyle, right? Because, without giving away any secrets, I have a pretty good lifestyle, right? And it's not really gonna make much difference to me. I think once you get down to fairly low levels of income, then ten percent could be the difference between eating and not eating. Burak · 40:37 So we're talking about state change again. Rob · 40:38 Yeah, I think there is a point at which you get sufficiently poor where even a tiny percentage of your income being lost could actually substantially affect your… So I'm not sure I agree with that. So maybe the most rational person to gamble is the person who's sort of in the middle and for whom either they're a young person who, as you say, has got a lot of potential future capital, and actually this is quite a small amount of that. Burak · 41:06 Yeah, they just got a tiny amount of cash on the side that is not gonna— Rob · 41:09 Yeah, exactly. But I don't think it's necessarily rational for very poor people. And I don't think it's necessarily rational for rich people apart from Elon Musk, for the reasons we've already described, because… 41:19 The “permanent underclass” meme & why Rob quit his hedge fund job Burak · 41:21 By the way, something that I've been trying to sneak in for a while. You know how we talk about absorbing states and going from one state to the other? There is this other meme going around on Twitter these days with the impact of AI, this permanent underclass meme. Rob · 41:35 Yeah, I've seen it, yeah. Burak · 41:35 And there are probably like top one percent, maybe top one in a thousand earners on Twitter arguing that you should do such and such to escape the permanent underclass. Well, I mean, it's a wrong premise for a lot of reasons, I think, but that's a different conversation. But it's funny, even when you get to that top one percent of income or top zero point one percent of income, there can be these sources of anxiety that you're trying to kind of go from one state to the further state. Rob · 42:02 Yeah, I mean, this state thing is quite interesting and actually, I mean, it comes into other decisions as well. I quit my job in finance, like, thirteen years ago. And I actually left about six months later. But, arguably, if I'd stayed there for another five years or another ten years, I'd have gone up a state. Broadly speaking. Because I worked in a hedge fund for seven years, and that basically moved me up a few states. And if I'd stayed for another five, ten years, I probably would have gone up another couple of states. But to me personally, the sort of utility of those—of that move was not worth it. Because I just not that interested in— Burak · 42:48 And this is more common than people realize. There are people, I think… I forgot his name. Patrick McKenzie. Yes, I remember that name. Patrick McKenzie talks about this. He talks about how the revealed preferences suggest that software engineers, instead of working, let's say, at a big company, big tech company, they do something else. It could be joining a startup, it could be joining a small company. But basically, he was comparing these two things, and the revealed preferences would suggest people value working on things that they like with people that they like at hundreds of millions of dollars potentially. Which doesn't sound that rational, but it feels very rational to me. Let me put it that way. And I think it feels rational to a lot of people, a lot more people than we realize, because I see this all the time. More recently, I guess, there's Meta AI handing out billion-dollar pay packages to researchers at other companies who were probably earning a fraction of that, and yet they refuse it. So this is more common than… And then it goes back to utility functions. Rob · 44:00 Can I just say, if Meta AI is listening, I would throw all of my principles in the bin to work for you guys for a billion dollars, but only for one year. I think that's probably all I could manage. Burak · 44:12 Yes, exactly, exactly. Yeah, there are deals to be made in that framework. Rob · 44:17 I don't mind. I mean, that's a state jump that seems like a worthy sacrifice to jump a few states just for a year of prostituting myself to the great gods of AI. Burak · 44:28 Maybe it was like a five-year deal. That was the deal. Rob · 44:30 No, no, I'm sorry. I'm sorry. We need to negotiate. Burak · 44:33 I agree. Rob · 44:34 Five years is too long. Burak · 44:35 Too long. Well, that's fantastic, Rob. I have more questions, but that will take it to hours and hours. So maybe we'll come back and revisit this sometime. Rob · 44:45 Yeah, have a few more guests on, and then I'm happy to come back on, definitely. Burak · 44:49 Yes, exactly. Fantastic, Rob. Thanks a lot for this. Rob · 44:52 Cheers. Burak · 44:52 Cheers.