1 00:00:03,240 --> 00:00:07,560 Speaker 1: This is Masters in Business with Barry Ridholds on Bloomberg Radio. 2 00:00:08,480 --> 00:00:11,320 Speaker 1: This week on Masters in Business on Bloomberg Radio, we 3 00:00:11,400 --> 00:00:14,520 Speaker 1: have a special guest. If you are a fan of 4 00:00:14,800 --> 00:00:21,799 Speaker 1: quantitative finance, modeling, any application of mathematics to the world 5 00:00:21,880 --> 00:00:26,119 Speaker 1: of investing. UH. This is really a master class in 6 00:00:27,120 --> 00:00:30,200 Speaker 1: what the world used to be like in finance when 7 00:00:30,240 --> 00:00:34,400 Speaker 1: people really you know, math was optional as opposed to 8 00:00:34,400 --> 00:00:38,000 Speaker 1: how things have developed today. Professor Emmanuel Derman is the 9 00:00:38,159 --> 00:00:42,400 Speaker 1: head of Financial Engineering UH coursework in the master's program 10 00:00:42,400 --> 00:00:47,280 Speaker 1: at Columbia University. His background is really quite amazing, and 11 00:00:47,320 --> 00:00:51,000 Speaker 1: I go into a lot of detail UH in the program. 12 00:00:51,400 --> 00:00:57,120 Speaker 1: Suffice it to say at Goldman Sachs, where he UH 13 00:00:57,520 --> 00:01:04,000 Speaker 1: eventually became head of the renowned quantitative Strategies group. Gives 14 00:01:04,000 --> 00:01:08,480 Speaker 1: you an insight of of what sort of UH mathematical 15 00:01:08,520 --> 00:01:14,640 Speaker 1: and programming background he has helped Goldman Sacks make ungodly 16 00:01:14,760 --> 00:01:19,800 Speaker 1: gobs of money by the intelligent application of modeling and 17 00:01:19,920 --> 00:01:23,240 Speaker 1: risk management. A lot of people don't think about the 18 00:01:23,360 --> 00:01:27,679 Speaker 1: blocking and tackling and the basic approach that you have 19 00:01:27,880 --> 00:01:31,840 Speaker 1: to engage in when you're dealing with things like stock 20 00:01:31,840 --> 00:01:36,520 Speaker 1: spoons and options of a liquid assets that that are 21 00:01:36,560 --> 00:01:39,000 Speaker 1: hard to come up with a price because they don't 22 00:01:39,040 --> 00:01:43,880 Speaker 1: necessarily trade all the time anyway. Really a fascinating conversation. 23 00:01:44,360 --> 00:01:47,960 Speaker 1: If you're at all interested in in quantitative finance and modeling, 24 00:01:48,560 --> 00:01:52,640 Speaker 1: um it's and you're not a student at Colombia, you 25 00:01:52,680 --> 00:01:55,720 Speaker 1: don't often get access to someone like Professor Norman, and 26 00:01:55,800 --> 00:01:59,960 Speaker 1: I think you'll find this to be a really interesting conversation. 27 00:02:00,040 --> 00:02:05,240 Speaker 1: And so, without any further ado, my interview with Professor 28 00:02:05,240 --> 00:02:13,000 Speaker 1: Emmanuel Derman. This is Masters in Business with Barry Ridholts 29 00:02:13,200 --> 00:02:17,040 Speaker 1: on Bloomberg Radio. My special guest this week is a 30 00:02:17,200 --> 00:02:22,360 Speaker 1: Manuel Derman. He is a particle physicist and better known 31 00:02:22,400 --> 00:02:25,760 Speaker 1: in the world of finance as a quant His background 32 00:02:26,080 --> 00:02:30,480 Speaker 1: is quite fascinating. Born in South Africa, came to the 33 00:02:30,560 --> 00:02:36,240 Speaker 1: United States later in life, spent seventeen years on Wall Street, 34 00:02:36,360 --> 00:02:43,040 Speaker 1: eventually becoming head of the renowned Quantitative Strategies group at 35 00:02:43,080 --> 00:02:47,560 Speaker 1: Goldman Sachs, where he co developed the Black Derman Toy 36 00:02:47,680 --> 00:02:51,959 Speaker 1: interest Rate model and the Derman Connie local Volatility model, 37 00:02:52,200 --> 00:02:56,519 Speaker 1: both of which have become industry standards. He won the 38 00:02:56,720 --> 00:03:00,960 Speaker 1: I a f E. Sun Guard Financial Engineer of the 39 00:03:01,040 --> 00:03:05,000 Speaker 1: Year Award in two thousand. He's the director of the 40 00:03:05,040 --> 00:03:09,760 Speaker 1: Master's Programming in Financial Engineering at Columbia and author of 41 00:03:10,080 --> 00:03:13,959 Speaker 1: two books, the first, My Life as a quant Reflections 42 00:03:14,000 --> 00:03:19,440 Speaker 1: on Physics and Finance, and more recently, Models Behaving Badly. 43 00:03:19,960 --> 00:03:23,360 Speaker 1: Emmanuel Derman, Welcome to Bloomberg. Thanks, I'm very glad to 44 00:03:23,400 --> 00:03:26,919 Speaker 1: be here. So you're someone whose career I have followed 45 00:03:26,960 --> 00:03:29,639 Speaker 1: for a long time. Uh, and I would imagine a 46 00:03:29,720 --> 00:03:33,079 Speaker 1: lot of our listeners are are probably not familiar with 47 00:03:33,960 --> 00:03:37,840 Speaker 1: either yourself personally or what you do. So let's start 48 00:03:37,880 --> 00:03:43,280 Speaker 1: with a really basic question. What is financial engineering? Okay, 49 00:03:43,280 --> 00:03:46,480 Speaker 1: financial engineering, it's sort of a polyglock field. It's not 50 00:03:46,560 --> 00:03:49,160 Speaker 1: really one simple thing. It developed over the last twenty 51 00:03:49,200 --> 00:03:53,000 Speaker 1: five or thirty years. Um. It's a mix of building 52 00:03:53,040 --> 00:04:00,560 Speaker 1: models for describing businesses and more particularly securities, involving mathematics, artistics, 53 00:04:00,680 --> 00:04:05,080 Speaker 1: the use of computer science programming, all inspired somewhat by 54 00:04:05,680 --> 00:04:08,320 Speaker 1: physics or scientific type models, and using the kind of 55 00:04:08,320 --> 00:04:12,920 Speaker 1: math that's traditionally used in describing the inanimate world of 56 00:04:12,960 --> 00:04:15,960 Speaker 1: the material world, but now applied, for better or for worse, 57 00:04:16,120 --> 00:04:19,680 Speaker 1: to the world of stocks, bonds, securities, options, etcetera. So 58 00:04:19,680 --> 00:04:22,320 Speaker 1: so that naturally leads to the next question, how does 59 00:04:22,400 --> 00:04:27,880 Speaker 1: someone transition from being a specialist in particle physics to 60 00:04:27,960 --> 00:04:31,760 Speaker 1: a specialist in quantitative finance. Yeah, it's like how do 61 00:04:31,760 --> 00:04:35,640 Speaker 1: you get to Carnegie whole practice practice practice. So then, 62 00:04:36,160 --> 00:04:40,800 Speaker 1: so there are obviously some similarities math. Yes, so a 63 00:04:40,800 --> 00:04:43,160 Speaker 1: lot of a lot of them. Modeling in particularly in 64 00:04:43,240 --> 00:04:46,960 Speaker 1: options and in describing fixed income instruments actually has its 65 00:04:47,040 --> 00:04:52,080 Speaker 1: origin in in economists who were trained in physics and 66 00:04:52,120 --> 00:04:54,799 Speaker 1: started applying that kind of math to physics or stochastic 67 00:04:54,839 --> 00:04:59,839 Speaker 1: calculus series discunning already are physics, mathematics and pargular physics 68 00:04:59,880 --> 00:05:02,559 Speaker 1: in spired models, and so it's kind of a fairly 69 00:05:02,680 --> 00:05:06,200 Speaker 1: natural transision for people to make. The trouble is a 70 00:05:06,200 --> 00:05:10,040 Speaker 1: lot of physicists don't have the economics of the finance background. 71 00:05:10,080 --> 00:05:12,240 Speaker 1: And thirty years ago that wasn't a problem. When I 72 00:05:12,320 --> 00:05:14,720 Speaker 1: did it, um, they didn't expect you to know anything. 73 00:05:14,920 --> 00:05:17,479 Speaker 1: The whole quantitative finance field was sort of amateur heaven. 74 00:05:17,520 --> 00:05:20,400 Speaker 1: You just came along and people told you to pick up. 75 00:05:20,480 --> 00:05:22,200 Speaker 1: It happened to me when I came to Golden they said, 76 00:05:22,200 --> 00:05:24,920 Speaker 1: read the Cox Ross rubens Steen model for pricing options 77 00:05:25,360 --> 00:05:28,120 Speaker 1: and start working. And now so it was fairly easy 78 00:05:28,120 --> 00:05:30,159 Speaker 1: in though, is that they hired you if you had potential. 79 00:05:30,240 --> 00:05:34,560 Speaker 1: Now it's a much tougher transcision. It's become a professional field. 80 00:05:34,600 --> 00:05:37,479 Speaker 1: I'm in charge of a program at Columbia where people 81 00:05:37,520 --> 00:05:40,839 Speaker 1: get degrees in the field, professional degrees. And you can't. 82 00:05:41,080 --> 00:05:44,159 Speaker 1: You can't just from an economic point of view. There's 83 00:05:44,160 --> 00:05:46,159 Speaker 1: a glut of people not so easy to get in now. 84 00:05:46,279 --> 00:05:49,680 Speaker 1: So when we look at physics, you're dealing with inanimate 85 00:05:49,760 --> 00:05:54,920 Speaker 1: objects that aren't pushing back. They're not getting excited about rumors, 86 00:05:54,960 --> 00:05:58,320 Speaker 1: they're not getting panicky. But in finance you have the 87 00:05:58,360 --> 00:06:03,800 Speaker 1: individual players who are all suffer from cognitive foibles and 88 00:06:04,279 --> 00:06:08,400 Speaker 1: emotional extremes. How do you adapt to that shift from 89 00:06:08,680 --> 00:06:13,480 Speaker 1: particles which are pretty clean and interesting to humans which 90 00:06:13,480 --> 00:06:17,840 Speaker 1: are messy and volatile. Yeah, that's um. You put your 91 00:06:17,839 --> 00:06:20,080 Speaker 1: finger on the key problem in the whole field, which 92 00:06:20,120 --> 00:06:24,240 Speaker 1: is some people. Yeah, people exactly. And and that's why 93 00:06:24,279 --> 00:06:26,120 Speaker 1: the models don't really work as well. When you make 94 00:06:26,160 --> 00:06:29,880 Speaker 1: a model, even Newtonian mechanics for describing how planets go 95 00:06:29,960 --> 00:06:31,800 Speaker 1: around the Sun, the planets don't really care what you 96 00:06:31,800 --> 00:06:33,960 Speaker 1: say about them, and if you publish an article about 97 00:06:34,000 --> 00:06:36,279 Speaker 1: them that don't change their position even if you have 98 00:06:36,320 --> 00:06:38,080 Speaker 1: the wrong theory about them. But really, if you look 99 00:06:38,080 --> 00:06:41,000 Speaker 1: at what happens in financial markets, they interact with people 100 00:06:41,040 --> 00:06:44,000 Speaker 1: in financial markets all about opinions, and opinions affect the 101 00:06:44,040 --> 00:06:47,160 Speaker 1: future and affect the presence. So um, it's it's a 102 00:06:47,279 --> 00:06:51,400 Speaker 1: much iffier field. And models don't work as well. They 103 00:06:51,400 --> 00:06:54,360 Speaker 1: don't describe the system is accurately, but they're not useless. 104 00:06:55,200 --> 00:06:58,120 Speaker 1: That's the famous George Box quote. All models are wrong, 105 00:06:58,200 --> 00:07:01,520 Speaker 1: but some are useful. Obviously, if you have a useful 106 00:07:01,560 --> 00:07:04,760 Speaker 1: model in finance, it can make a lot of money 107 00:07:04,839 --> 00:07:09,080 Speaker 1: for for its owners. Um, how does somebody like George 108 00:07:09,080 --> 00:07:14,200 Speaker 1: Soros and his theory of reflexivity, how market prices subsequently 109 00:07:14,240 --> 00:07:17,680 Speaker 1: affect market prices? How does that impact someone who's trying 110 00:07:17,720 --> 00:07:20,880 Speaker 1: to create a financial model. Well, I think it impacts 111 00:07:21,000 --> 00:07:23,360 Speaker 1: I think most financial models don't really take a kind 112 00:07:23,360 --> 00:07:26,400 Speaker 1: of reflexivity. There's some some models. Maybe you're starting to 113 00:07:26,400 --> 00:07:29,920 Speaker 1: do that, but I wouldn't say they trading models. Um. 114 00:07:30,160 --> 00:07:31,880 Speaker 1: If I can summarize in one sentence, I would say 115 00:07:31,880 --> 00:07:34,400 Speaker 1: that anytime you build a model, it's a financial model 116 00:07:34,480 --> 00:07:36,600 Speaker 1: is mathematical. You have to keep in the back of 117 00:07:36,640 --> 00:07:39,520 Speaker 1: your head that you're actually short volatility, meaning if the 118 00:07:39,520 --> 00:07:41,560 Speaker 1: world changes, your model is going to be wrong. It 119 00:07:41,600 --> 00:07:43,480 Speaker 1: may make money for you, may lose money for you, 120 00:07:43,520 --> 00:07:45,920 Speaker 1: but it's certainly going to be wrong. Most models only 121 00:07:45,960 --> 00:07:48,240 Speaker 1: work in a very narrow regime where things are more 122 00:07:48,320 --> 00:07:50,840 Speaker 1: or less like the world you're in currently. And you 123 00:07:50,880 --> 00:07:53,120 Speaker 1: see when you get to negative interest rate, for example, 124 00:07:53,280 --> 00:07:56,160 Speaker 1: or when volatility blows up, or when you get you know, 125 00:07:56,240 --> 00:07:59,720 Speaker 1: the Great Financial Crisis, all of these models stopped working. 126 00:08:00,640 --> 00:08:03,480 Speaker 1: Since you you hinted at this in terms of a 127 00:08:03,560 --> 00:08:06,640 Speaker 1: changing regime, I'm thinking about the models that were put 128 00:08:06,680 --> 00:08:10,360 Speaker 1: together by a firm like Long Term Capital Management in 129 00:08:10,360 --> 00:08:14,880 Speaker 1: the late nineties, in the last minute we have. Is 130 00:08:14,920 --> 00:08:17,920 Speaker 1: that why they blew up so spectacularly The world changed 131 00:08:18,000 --> 00:08:22,040 Speaker 1: and their model failed to adapt plus hundred one leverage. Yeah. 132 00:08:22,080 --> 00:08:24,560 Speaker 1: I think my impression of what happened to Long Term 133 00:08:24,560 --> 00:08:28,560 Speaker 1: Capital was that they were basically looking for very small 134 00:08:28,720 --> 00:08:32,120 Speaker 1: deviations most things that earn your money in market, So 135 00:08:32,240 --> 00:08:34,600 Speaker 1: when you buy something liquid, because the liquid things tend 136 00:08:34,640 --> 00:08:36,120 Speaker 1: to be cheap, and so they were buying off the 137 00:08:36,200 --> 00:08:39,320 Speaker 1: run treasuries out of the money options. I don't know 138 00:08:39,360 --> 00:08:41,600 Speaker 1: the details anymore, but essentially things that were cheap, but 139 00:08:41,640 --> 00:08:43,760 Speaker 1: in the long run would revert to the mean and 140 00:08:43,800 --> 00:08:45,640 Speaker 1: give you the full amount. But they leveraged up like 141 00:08:45,760 --> 00:08:50,680 Speaker 1: crazy to accentuate small pennies differences. And then when Russia defaulted, 142 00:08:51,040 --> 00:08:53,160 Speaker 1: what happened was everybody in the world got scared. There's 143 00:08:53,160 --> 00:08:55,880 Speaker 1: a flight to quality. Nobody wants to hold the liquid stuff, 144 00:08:55,920 --> 00:08:58,280 Speaker 1: and they were so overleveraged that they were they were 145 00:08:58,280 --> 00:09:01,160 Speaker 1: put out of business. I'm very ridults. You're listening to 146 00:09:01,320 --> 00:09:05,080 Speaker 1: Masters and Business on Bloomberg Radio. My special guest today 147 00:09:05,120 --> 00:09:08,400 Speaker 1: is a Manuel Derman. He is a professor of financial 148 00:09:08,440 --> 00:09:12,000 Speaker 1: engineering at Columbia University. He used to work at Goldman 149 00:09:12,040 --> 00:09:15,560 Speaker 1: Sachs where he was head of the Quantitative Strategies group. 150 00:09:16,120 --> 00:09:17,840 Speaker 1: And I want to talk a little bit about a 151 00:09:17,880 --> 00:09:21,000 Speaker 1: book you wrote a few years ago after the financial 152 00:09:21,040 --> 00:09:26,200 Speaker 1: crisis called Models Behaving Badly. And one of the first 153 00:09:26,280 --> 00:09:29,600 Speaker 1: things that stood out to me from that book was 154 00:09:30,840 --> 00:09:35,000 Speaker 1: sentence you had written, models are metaphors that explain the 155 00:09:35,080 --> 00:09:40,080 Speaker 1: world we don't understand in terms that we do understand. 156 00:09:40,480 --> 00:09:45,280 Speaker 1: Expand on that. Okay, Um, yeah, that's that's some that's 157 00:09:45,320 --> 00:09:48,160 Speaker 1: my feeling. I can give an example. So what I 158 00:09:48,240 --> 00:09:50,760 Speaker 1: tried to do in that book was distinguished between models 159 00:09:50,760 --> 00:09:54,080 Speaker 1: and theories, as as different means of approaching trying to 160 00:09:54,120 --> 00:09:59,640 Speaker 1: explain the world animate or inanimate, and models seemed to 161 00:09:59,640 --> 00:10:01,720 Speaker 1: me to the analogies in the sense, for example, you 162 00:10:01,760 --> 00:10:04,280 Speaker 1: say the brain is like a computer, or the computer 163 00:10:04,400 --> 00:10:06,040 Speaker 1: is like a brain. People used to say now that 164 00:10:06,240 --> 00:10:08,560 Speaker 1: now they go the other way. Or they say the 165 00:10:08,600 --> 00:10:10,920 Speaker 1: heart is like a water pump. Or there's a great 166 00:10:10,960 --> 00:10:13,960 Speaker 1: quote I like about fixed income from Schopenhauer where he says, 167 00:10:14,800 --> 00:10:16,800 Speaker 1: sleep is the interest that we have to pay on 168 00:10:16,880 --> 00:10:19,280 Speaker 1: the capital which is called in a death. And the 169 00:10:19,360 --> 00:10:21,679 Speaker 1: more regularly the interest is paid, the further the date 170 00:10:21,720 --> 00:10:25,319 Speaker 1: of redemption is postponed. So in other words, it's important 171 00:10:25,320 --> 00:10:27,520 Speaker 1: to get eight hours nights. Yes, and if you can 172 00:10:27,559 --> 00:10:29,439 Speaker 1: get eight hours and I'd sleep. But if you think 173 00:10:29,440 --> 00:10:31,800 Speaker 1: about what he's doing, he's sort of there's a small 174 00:10:31,840 --> 00:10:35,240 Speaker 1: overlap between sleep and between paying interest on a bond, 175 00:10:35,280 --> 00:10:37,160 Speaker 1: which is that you sleep regular and you pay interest 176 00:10:37,200 --> 00:10:39,240 Speaker 1: on a bond. Rey that's really only analogy. And then 177 00:10:39,559 --> 00:10:41,400 Speaker 1: but on a bond, you've borrowed money and you have 178 00:10:41,440 --> 00:10:42,920 Speaker 1: to pay it back at the end. And now he's saying, 179 00:10:42,960 --> 00:10:44,719 Speaker 1: you've borrowed your life from the darkness, and at the 180 00:10:44,800 --> 00:10:46,240 Speaker 1: end you have to pay it back again, and you're 181 00:10:46,240 --> 00:10:48,880 Speaker 1: paying back darkness all the way. So I think that's 182 00:10:48,880 --> 00:10:53,720 Speaker 1: a good example. And most of the models in finance 183 00:10:53,880 --> 00:10:57,480 Speaker 1: or analogies where you say, for example, um in the 184 00:10:57,559 --> 00:11:01,319 Speaker 1: Kape model or in or in in modern finance, you 185 00:11:01,360 --> 00:11:05,120 Speaker 1: say stock prices behave like smoke diffusing from from a 186 00:11:05,120 --> 00:11:07,720 Speaker 1: cigarette end, you know, doing Branny in motion. That's not 187 00:11:07,840 --> 00:11:11,080 Speaker 1: literally true. It could be true, but it isn't. So 188 00:11:11,120 --> 00:11:14,000 Speaker 1: in other words, stock prices don't randomly fill a room 189 00:11:14,040 --> 00:11:17,600 Speaker 1: evenly distributed like the behavior of a gas right, But 190 00:11:17,760 --> 00:11:21,040 Speaker 1: you're creating a metaphor that, hey, it's somewhat random, it's 191 00:11:21,080 --> 00:11:24,720 Speaker 1: a little predictable, but not very predictable, especially over the 192 00:11:24,760 --> 00:11:28,480 Speaker 1: long term exactly. And and for example, in physics, that's 193 00:11:28,480 --> 00:11:31,680 Speaker 1: called Brannyan motion, this motion of diffusion in physics, that 194 00:11:31,880 --> 00:11:34,680 Speaker 1: really is a theory. It's an accurate description of the 195 00:11:34,679 --> 00:11:37,880 Speaker 1: way smoke behaves. But stock prices don't behave that way. 196 00:11:37,960 --> 00:11:41,719 Speaker 1: Volatility jumps, stock prices crash, apple rises are dumps buck 197 00:11:41,760 --> 00:11:44,520 Speaker 1: twenty or thirty bucks in a day more actually son 198 00:11:44,559 --> 00:11:46,760 Speaker 1: of ten percent in a day, So it's not it's 199 00:11:46,760 --> 00:11:50,679 Speaker 1: not a diffusion, it's something that's that's violent. Um, So 200 00:11:50,760 --> 00:11:53,040 Speaker 1: why do we why are we so enamored of all 201 00:11:53,040 --> 00:11:58,120 Speaker 1: these metaphors because we can't do better? And the difference 202 00:11:58,120 --> 00:12:00,720 Speaker 1: is I think economists don't understand. I'll to not be 203 00:12:00,720 --> 00:12:03,280 Speaker 1: too rude about economists, but they don't understand the difference 204 00:12:03,320 --> 00:12:05,840 Speaker 1: between a model and a theory, or a metaphor and 205 00:12:05,840 --> 00:12:09,600 Speaker 1: an accurate description. And physicists, which is my original background, 206 00:12:09,600 --> 00:12:12,480 Speaker 1: actually understand that very well. So for example, if you 207 00:12:12,559 --> 00:12:16,200 Speaker 1: say Newton's laws, which destrive the planets going around the Sun, 208 00:12:16,360 --> 00:12:20,120 Speaker 1: they say four sequals mass times acceleration. That's really accurate. 209 00:12:20,559 --> 00:12:23,240 Speaker 1: It may not be perfect, but it's very accurate. On 210 00:12:23,280 --> 00:12:25,960 Speaker 1: the other hand, they might say a nucleus at the 211 00:12:26,000 --> 00:12:28,959 Speaker 1: center of an atom is like a liquid drop, and 212 00:12:29,320 --> 00:12:31,360 Speaker 1: they understand that when they say a nucleus is like 213 00:12:31,400 --> 00:12:34,000 Speaker 1: a liquid drop, it isn't really a liquid drop. It's 214 00:12:34,040 --> 00:12:35,760 Speaker 1: just a lot like it. But at some point that 215 00:12:35,760 --> 00:12:38,320 Speaker 1: analogy is going to break down, and people get Nobel 216 00:12:38,360 --> 00:12:40,800 Speaker 1: prizes for saying a nucleus is like a liquid drop, 217 00:12:41,280 --> 00:12:43,480 Speaker 1: and they get Nobel prizes for doing what Feynman did, 218 00:12:43,520 --> 00:12:46,679 Speaker 1: which is an accurate description of electrons. But but physicists 219 00:12:46,720 --> 00:12:49,440 Speaker 1: understand that one of them is truth or close to truth, 220 00:12:49,440 --> 00:12:51,920 Speaker 1: and the other one's just a model. So let me 221 00:12:51,960 --> 00:12:55,560 Speaker 1: go off script and say, um, how accurate is it 222 00:12:55,640 --> 00:13:00,240 Speaker 1: to say that economists suffer from physics envy? Uh? I 223 00:13:00,280 --> 00:13:02,440 Speaker 1: don't know who originally coined that. I think maybe Andy Low, 224 00:13:02,440 --> 00:13:06,240 Speaker 1: who's a professor at M I T. I think that's accurate. 225 00:13:06,280 --> 00:13:08,280 Speaker 1: I think maybe it's fading to some extent now The 226 00:13:08,280 --> 00:13:12,439 Speaker 1: behavioral finance people clearly don't. But but yeah they do. 227 00:13:12,920 --> 00:13:16,040 Speaker 1: And you know Bachelier, the guy who started a lot 228 00:13:16,080 --> 00:13:18,040 Speaker 1: of the stuff, and Paul Samuelson, they were all ahead 229 00:13:18,040 --> 00:13:21,120 Speaker 1: of physics smith background somewhere in there. And I noticed 230 00:13:21,160 --> 00:13:24,920 Speaker 1: that you drop a lot of philosophical references in discussion. 231 00:13:25,080 --> 00:13:28,200 Speaker 1: Is that from the financial engineering side or is that 232 00:13:28,240 --> 00:13:31,480 Speaker 1: from the physics side. Um, I've got a late life 233 00:13:31,520 --> 00:13:35,400 Speaker 1: interest in philosophy. Well, it's throughout your books. I notice 234 00:13:35,440 --> 00:13:37,920 Speaker 1: you're you're always referring to that. We'll come back to 235 00:13:38,040 --> 00:13:41,400 Speaker 1: that a little later. Um, let me give you a 236 00:13:41,480 --> 00:13:46,199 Speaker 1: quote of yours that I found intriguing and wildly overlooked 237 00:13:46,200 --> 00:13:48,800 Speaker 1: by a lot of investors. If you want to take 238 00:13:48,840 --> 00:13:52,520 Speaker 1: a chance on the upside, you have also to take 239 00:13:52,559 --> 00:13:55,679 Speaker 1: a chance on the down side. Explain what you mean 240 00:13:55,760 --> 00:13:58,920 Speaker 1: by that, what I mean by that. There was a 241 00:13:58,960 --> 00:14:01,199 Speaker 1: response to I think you wrote a book about this, 242 00:14:01,320 --> 00:14:04,400 Speaker 1: two to the bailouts of two thousand seven eight nine, 243 00:14:05,200 --> 00:14:08,680 Speaker 1: when I felt that there were a lot of companies around, 244 00:14:08,800 --> 00:14:12,280 Speaker 1: banks in particular, but financial companies, that we're taking risk 245 00:14:12,360 --> 00:14:14,800 Speaker 1: and saying that's the essence of capitalism and sort of 246 00:14:14,880 --> 00:14:16,520 Speaker 1: leave me alone and let me, let me do what 247 00:14:16,559 --> 00:14:20,200 Speaker 1: I'm good at. And then suddenly, when everything collapsed, they 248 00:14:20,280 --> 00:14:24,600 Speaker 1: wanted to be saved from death. And I found that incredibly, 249 00:14:25,280 --> 00:14:29,440 Speaker 1: I don't know whether ethically or morally, sort of yeah, 250 00:14:29,520 --> 00:14:31,280 Speaker 1: and saying, oh, the whole system will die if you 251 00:14:31,320 --> 00:14:33,760 Speaker 1: don't save us. And and that's what I meant. If 252 00:14:33,760 --> 00:14:36,000 Speaker 1: you want to be somebody that benefits from taking risk, 253 00:14:36,120 --> 00:14:39,120 Speaker 1: you also have to benefit when taking risk kills you. 254 00:14:39,600 --> 00:14:41,960 Speaker 1: So in other words, you can't have privatized gains and 255 00:14:42,040 --> 00:14:44,840 Speaker 1: socialized laws. Yes, and I like to think that what 256 00:14:44,920 --> 00:14:46,640 Speaker 1: a lot of the banks should have done if they 257 00:14:46,640 --> 00:14:51,320 Speaker 1: were saved by the saved by the government, was they 258 00:14:51,360 --> 00:14:52,960 Speaker 1: were essentially given a put if you look at it 259 00:14:52,960 --> 00:14:54,840 Speaker 1: from an options point of view, they were safe from death, 260 00:14:55,280 --> 00:14:57,520 Speaker 1: and if they were given a put they should have 261 00:14:57,520 --> 00:15:01,040 Speaker 1: given away a call to the when they were When 262 00:15:01,080 --> 00:15:04,760 Speaker 1: they survived, companies had a little a I G. Gave 263 00:15:04,800 --> 00:15:07,880 Speaker 1: the government some uh, I mean it was forced upon them, 264 00:15:07,880 --> 00:15:10,280 Speaker 1: and then the G S, C S, Fannie and Freddie 265 00:15:10,400 --> 00:15:15,240 Speaker 1: they essentially became Yeah, yeah, so that's fair. You can't 266 00:15:15,240 --> 00:15:19,360 Speaker 1: get ten million dollar bonuses the rfter you was saved from. 267 00:15:19,960 --> 00:15:22,960 Speaker 1: Makes sense that should we have nationalized those companies, cleaned 268 00:15:23,000 --> 00:15:25,840 Speaker 1: them up, and then spun them out as free standing 269 00:15:25,880 --> 00:15:28,360 Speaker 1: companies with the benefits of those I p o s 270 00:15:28,400 --> 00:15:31,400 Speaker 1: going to the taxpayer? Is that a better ethical way 271 00:15:31,400 --> 00:15:33,280 Speaker 1: to do that. I think it's a better ethical way 272 00:15:33,280 --> 00:15:34,960 Speaker 1: to do it. I have friends who argue with this 273 00:15:35,000 --> 00:15:37,160 Speaker 1: about him. They say the whole system would have collapsed 274 00:15:37,200 --> 00:15:39,840 Speaker 1: if we hadn't bailed them out. Um, people are not 275 00:15:39,920 --> 00:15:42,480 Speaker 1: what the FED who say everything was credit was on 276 00:15:42,520 --> 00:15:46,200 Speaker 1: the verge of sort of freezing up. But I think 277 00:15:46,200 --> 00:15:49,520 Speaker 1: it's left a permanent um, a permanent bad taste in 278 00:15:49,560 --> 00:15:54,600 Speaker 1: everybody's mouth. Certainly created moral hazard that Hey, you know, look, 279 00:15:54,800 --> 00:15:58,880 Speaker 1: it's arguable to say we rescued bear Sterns and the 280 00:15:58,960 --> 00:16:02,120 Speaker 1: FED backed up JP Morgan's purchase, and that might have 281 00:16:02,280 --> 00:16:06,240 Speaker 1: given the impetus to Lehman Brothers to say someone will 282 00:16:06,280 --> 00:16:09,360 Speaker 1: come along and rescue us. Right? Is that fair? Fair 283 00:16:09,400 --> 00:16:11,560 Speaker 1: assessment of that? There was a funny cartoon. This was 284 00:16:11,560 --> 00:16:14,240 Speaker 1: a little maybe not right, but there was a funny 285 00:16:14,240 --> 00:16:17,400 Speaker 1: cartoon by Barry Blitzer or Berry Blitzer, forget his name 286 00:16:17,480 --> 00:16:20,440 Speaker 1: is in New York and New Yorker cartoonists who had 287 00:16:20,440 --> 00:16:23,160 Speaker 1: a picture somewhere in two thousand and eight of Obama 288 00:16:23,320 --> 00:16:26,360 Speaker 1: dressed as a as a New York City policeman with 289 00:16:26,400 --> 00:16:29,960 Speaker 1: a hat and a navy navy uniform and walking down 290 00:16:30,040 --> 00:16:32,920 Speaker 1: the street tooling of a ton with these eyes cast 291 00:16:33,000 --> 00:16:34,520 Speaker 1: up in heaven as though he couldn't see what was 292 00:16:34,520 --> 00:16:36,480 Speaker 1: around him, and meanwhile people behind him or running with 293 00:16:36,520 --> 00:16:39,880 Speaker 1: bags of money into buildings. And that's sort of typified, 294 00:16:40,760 --> 00:16:44,120 Speaker 1: typified it for me. But it was certainly an interesting period. 295 00:16:44,520 --> 00:16:47,320 Speaker 1: I'm Barry Ridults. You're listening to Masters in Business on 296 00:16:47,360 --> 00:16:50,760 Speaker 1: Bloomberg Radio. My special guest today is a manual German. 297 00:16:51,120 --> 00:16:56,080 Speaker 1: He spent seventeen years at Goldman Sachs, eventually becoming head 298 00:16:56,200 --> 00:17:00,480 Speaker 1: of the renowned quantitative strategies group. He all So is 299 00:17:00,600 --> 00:17:06,280 Speaker 1: currently uh, the director of financial Engineering at Columbia University. 300 00:17:06,280 --> 00:17:09,600 Speaker 1: It's a master's program, I believe is that the master's program. 301 00:17:09,600 --> 00:17:12,520 Speaker 1: It's a sort of travel month eighteen month professional degree. 302 00:17:12,720 --> 00:17:16,440 Speaker 1: He also is the author of Models Behaving Badly and 303 00:17:17,200 --> 00:17:21,359 Speaker 1: My Life as a Quant Reflections on Physics and Finance. 304 00:17:21,560 --> 00:17:24,719 Speaker 1: Let's talk a little bit about quants on Wall Street. 305 00:17:25,040 --> 00:17:28,360 Speaker 1: You mentioned when you began twenty five years ago or so, 306 00:17:28,760 --> 00:17:31,439 Speaker 1: it was a wide open field. There was no playbook. 307 00:17:32,359 --> 00:17:35,800 Speaker 1: But here it is. We have high frequency trading, we 308 00:17:35,840 --> 00:17:40,040 Speaker 1: have al go driven strategies, we we have all manners 309 00:17:40,200 --> 00:17:48,880 Speaker 1: of mathematical um investing, mathematically based investing. Is this the 310 00:17:48,960 --> 00:17:52,920 Speaker 1: age of the quant? Yeah? I think it is. Actually, 311 00:17:53,000 --> 00:17:55,800 Speaker 1: I mean when I started out, very few traders were 312 00:17:55,920 --> 00:17:57,959 Speaker 1: very numerous. They didn't know a lot of math, they 313 00:17:58,000 --> 00:18:00,840 Speaker 1: hadn't studied. A lot of them came from law backgrounds 314 00:18:00,840 --> 00:18:03,320 Speaker 1: and stuff like that. And now, in fact, quant was 315 00:18:03,359 --> 00:18:07,200 Speaker 1: a derogatory word. I've got a dictionary from the late 316 00:18:07,320 --> 00:18:10,200 Speaker 1: nineties where somebody it's it's by Mark Chritsman, called a 317 00:18:10,280 --> 00:18:14,280 Speaker 1: dictionary financial terms, and he says quanton. He says quantity 318 00:18:14,320 --> 00:18:16,800 Speaker 1: of analysts often used pejoratively and when I wrote that 319 00:18:16,840 --> 00:18:19,200 Speaker 1: book called My Life as a Quant. What was actually 320 00:18:19,200 --> 00:18:20,520 Speaker 1: in the back of my head was I took my 321 00:18:20,600 --> 00:18:23,640 Speaker 1: kids twenty five years ago to see us a Swedish 322 00:18:23,640 --> 00:18:26,760 Speaker 1: movie called My Life as a Dog. You see, there 323 00:18:26,800 --> 00:18:29,280 Speaker 1: was a great movie as for kids and parents to 324 00:18:29,320 --> 00:18:32,040 Speaker 1: see simultaneously, and I was seeing on My Life as 325 00:18:32,040 --> 00:18:33,480 Speaker 1: a Dog. That was really the back in the back 326 00:18:33,480 --> 00:18:35,359 Speaker 1: of my mind. When I came to Goldman, quant was 327 00:18:35,400 --> 00:18:37,040 Speaker 1: a sort of like being a geek or being a 328 00:18:38,320 --> 00:18:40,720 Speaker 1: and people laughed at you, although they kind of treated 329 00:18:40,760 --> 00:18:42,520 Speaker 1: you with respect. I liked it, but it was a 330 00:18:42,600 --> 00:18:45,240 Speaker 1: kind of mocking kind of respect. So since you brought 331 00:18:45,280 --> 00:18:47,639 Speaker 1: that up, let me go back to a question I missed. 332 00:18:48,000 --> 00:18:50,840 Speaker 1: So you were at Goldman Sachs in the ninety nineties, 333 00:18:50,880 --> 00:18:53,480 Speaker 1: that had to be one of the questions I was 334 00:18:53,480 --> 00:18:56,320 Speaker 1: going to ask you, was how were the quants treated? 335 00:18:56,480 --> 00:18:59,120 Speaker 1: So you you said there was some degree of respect, 336 00:18:59,560 --> 00:19:03,240 Speaker 1: but was it was there? Were you treated as one 337 00:19:03,320 --> 00:19:07,159 Speaker 1: of their own? Or were you guys kind of on 338 00:19:07,240 --> 00:19:10,480 Speaker 1: the outskirts and people, you know, slipped pizza under the 339 00:19:10,520 --> 00:19:14,359 Speaker 1: door and kind of left you alone. Um, that's funny. 340 00:19:14,400 --> 00:19:18,280 Speaker 1: It varied, it varied um for a while for a 341 00:19:18,280 --> 00:19:20,760 Speaker 1: while I worked for Fisher Black in his group, and 342 00:19:20,800 --> 00:19:23,600 Speaker 1: then he had everybody sit on the desk and get 343 00:19:23,680 --> 00:19:25,879 Speaker 1: lunch when the traders got lunch, and you were kind 344 00:19:25,880 --> 00:19:28,440 Speaker 1: of more equal. But generally I would say, that didn't 345 00:19:28,520 --> 00:19:32,639 Speaker 1: last long, and people people treated you like like geeks, 346 00:19:32,680 --> 00:19:35,399 Speaker 1: but like useful but useful people that they spoke to you. 347 00:19:35,440 --> 00:19:37,920 Speaker 1: What's the expression useful idiots? Is that how they looked 348 00:19:37,920 --> 00:19:42,119 Speaker 1: at did was that the action when I got in 349 00:19:42,320 --> 00:19:44,240 Speaker 1: When I got there in eighty five, which was earlier, 350 00:19:44,240 --> 00:19:46,119 Speaker 1: people would get in the elevator and sort of make 351 00:19:46,200 --> 00:19:47,920 Speaker 1: jokes and say, oh, all of you allowed to travel 352 00:19:47,960 --> 00:19:51,080 Speaker 1: in the same elevator at the same time. Or I 353 00:19:51,080 --> 00:19:52,520 Speaker 1: would be with I wrote about it in my life 354 00:19:52,520 --> 00:19:53,879 Speaker 1: as a quant. I would be with some guy and 355 00:19:53,920 --> 00:19:55,720 Speaker 1: I would I was very excited about being on Wall 356 00:19:55,760 --> 00:19:57,439 Speaker 1: Street and it was interesting. I would say something to 357 00:19:57,520 --> 00:20:00,840 Speaker 1: him about duration or convexity, and he would get embarrassed 358 00:20:00,880 --> 00:20:02,720 Speaker 1: and shrink away and say, what do you think of 359 00:20:02,760 --> 00:20:04,640 Speaker 1: the Yankees last night? You know? So, in other words, 360 00:20:04,640 --> 00:20:06,800 Speaker 1: he didn't know what convects it your duration? Actually now 361 00:20:06,800 --> 00:20:08,199 Speaker 1: he knew what it was. He just didn't want to 362 00:20:08,200 --> 00:20:11,119 Speaker 1: be outed in the fation in other words, he was 363 00:20:11,160 --> 00:20:14,359 Speaker 1: a geek, but he didn't he didn't like me talking 364 00:20:14,359 --> 00:20:17,040 Speaker 1: about it in public. But then, but then, even in 365 00:20:17,040 --> 00:20:20,840 Speaker 1: the early nineties, you know, I remember when not just Quantz, 366 00:20:20,880 --> 00:20:24,080 Speaker 1: but everybody. If you had a PhD, you didn't put 367 00:20:24,080 --> 00:20:25,600 Speaker 1: it on your business card, and if you had an 368 00:20:25,640 --> 00:20:27,600 Speaker 1: email address, you didn't put it on your business card 369 00:20:27,640 --> 00:20:29,479 Speaker 1: because that was like the brand of Caine. You know, 370 00:20:30,440 --> 00:20:35,160 Speaker 1: that's amazing. I think that that's early nineties. People didn't 371 00:20:35,160 --> 00:20:38,840 Speaker 1: want to put email addresses or or or PhD degrees 372 00:20:38,880 --> 00:20:41,040 Speaker 1: on their business card. Now, of course they'd love to 373 00:20:41,080 --> 00:20:45,600 Speaker 1: do it. That's wow. Become things have changed. So you 374 00:20:45,600 --> 00:20:49,399 Speaker 1: were Goldman when they went public, right, what what was 375 00:20:49,480 --> 00:20:53,600 Speaker 1: that like from the experience of someone with your background 376 00:20:54,119 --> 00:20:56,560 Speaker 1: on Wall Street? You know, I think things were steadily 377 00:20:56,600 --> 00:20:58,879 Speaker 1: improving for quans Alo at the same time Goldman was 378 00:20:58,880 --> 00:21:01,800 Speaker 1: getting much more bureaucrats. But by that time, you've been 379 00:21:01,880 --> 00:21:05,800 Speaker 1: through UM long Term Capital, You've been through d Shore, 380 00:21:06,000 --> 00:21:10,280 Speaker 1: so LTCM, so all sorts of hedge funds starting to 381 00:21:10,280 --> 00:21:13,119 Speaker 1: make a lot of money out of supposedly quantitative strategies, 382 00:21:13,160 --> 00:21:17,080 Speaker 1: and and particularly the whole Internet Internet and not Bubble, 383 00:21:17,160 --> 00:21:21,600 Speaker 1: but the whole Internet excitement where suddenly um being being 384 00:21:21,640 --> 00:21:24,760 Speaker 1: technologically competent became a way to make money. And so 385 00:21:24,800 --> 00:21:29,000 Speaker 1: that even destroying destroying the whole system, like LTC M 386 00:21:29,840 --> 00:21:32,000 Speaker 1: gave you, gave you respect. So things were getting better 387 00:21:32,000 --> 00:21:34,800 Speaker 1: for quants. But by the late nineties, I have to 388 00:21:34,840 --> 00:21:39,399 Speaker 1: think renaissance technologies had been putting up absurd returns for 389 00:21:39,440 --> 00:21:43,600 Speaker 1: a long time, right, And if memory serves didn't Jim 390 00:21:43,640 --> 00:21:48,720 Speaker 1: Simon's return outside investors money sometime around then is maybe 391 00:21:48,720 --> 00:21:50,439 Speaker 1: a little bit later, but yeah, they only have their 392 00:21:50,480 --> 00:21:53,840 Speaker 1: own money now, which is which is tells you, hey, 393 00:21:53,880 --> 00:21:57,439 Speaker 1: these guys are really putting up huge returns without a 394 00:21:57,440 --> 00:22:01,439 Speaker 1: whole lot of capacity to do it. In the tens 395 00:22:01,440 --> 00:22:05,160 Speaker 1: of billions your d shore as well returns outside money. 396 00:22:05,480 --> 00:22:07,840 Speaker 1: And no, I didn't return outside money, but we're very 397 00:22:07,920 --> 00:22:11,800 Speaker 1: visible and and at that point don't people look around 398 00:22:11,840 --> 00:22:14,360 Speaker 1: and say, hey, these guys are minting money. Let's get 399 00:22:14,400 --> 00:22:16,600 Speaker 1: us some quants, and maybe we need to pay these 400 00:22:16,640 --> 00:22:19,840 Speaker 1: people proper respect. Well, they gave them respect, and they 401 00:22:19,840 --> 00:22:22,320 Speaker 1: paid them kind of decently, but there was still, at 402 00:22:22,400 --> 00:22:24,160 Speaker 1: least in the area I worked in at that point, 403 00:22:24,160 --> 00:22:26,360 Speaker 1: there was still a very fine oute gap between being 404 00:22:26,400 --> 00:22:28,200 Speaker 1: a trader and being a quant Now, I think that's 405 00:22:28,240 --> 00:22:31,720 Speaker 1: vanished a lot. But then you were a support person, 406 00:22:31,800 --> 00:22:35,280 Speaker 1: you went a position taker. I'm Barry Ridholts. You're listening 407 00:22:35,320 --> 00:22:38,520 Speaker 1: to Masters in Business on Bloomberg Radio. My special guest 408 00:22:38,600 --> 00:22:44,000 Speaker 1: today is a Manual German. He is a renowned quantitative thinker. 409 00:22:44,320 --> 00:22:48,840 Speaker 1: UH teachers financial engineering at Columbia. You surround the quant 410 00:22:48,880 --> 00:22:52,639 Speaker 1: group at Goldman Sachs. I think that's a pretty good 411 00:22:52,760 --> 00:22:56,560 Speaker 1: CVUH to chat with. Little O me. Let's let's talk 412 00:22:56,600 --> 00:22:59,840 Speaker 1: a little bit about models and what they can and 413 00:23:00,080 --> 00:23:03,520 Speaker 1: can't do. Because you've every answer you've given me has 414 00:23:03,640 --> 00:23:06,000 Speaker 1: has set off in my mind the number of digressions. 415 00:23:06,040 --> 00:23:10,000 Speaker 1: But I wanted to stick with some of the questions. 416 00:23:10,119 --> 00:23:13,040 Speaker 1: So you look at models, and you look at what 417 00:23:13,119 --> 00:23:17,040 Speaker 1: they can do. What is it that people can do 418 00:23:17,200 --> 00:23:21,399 Speaker 1: that models can't Well, I think the right way to 419 00:23:21,480 --> 00:23:24,439 Speaker 1: use models is to sort of quantify your intuition. Like 420 00:23:24,560 --> 00:23:26,960 Speaker 1: people can have intuition, but it's hard to translate that 421 00:23:27,040 --> 00:23:30,240 Speaker 1: intuition into a number. And so for example, if somebody 422 00:23:30,280 --> 00:23:33,560 Speaker 1: said to you, um, what should I pay for an option? 423 00:23:33,720 --> 00:23:35,920 Speaker 1: Another price of an option? At the money? What should 424 00:23:35,960 --> 00:23:38,000 Speaker 1: I pay for an out of the money option. You 425 00:23:38,040 --> 00:23:39,920 Speaker 1: know it should be less, but it's hard to actually 426 00:23:40,000 --> 00:23:43,280 Speaker 1: quantify how much less and if you have something. But 427 00:23:43,400 --> 00:23:46,240 Speaker 1: when you look at black shoals or options pricing, they 428 00:23:46,240 --> 00:23:49,560 Speaker 1: invent the whole notion of volatility and measuring volatility. And 429 00:23:49,640 --> 00:23:51,800 Speaker 1: just like you can have intuition about interest rates, a 430 00:23:51,880 --> 00:23:54,439 Speaker 1: human being can have intuition about will volatility go up, 431 00:23:54,480 --> 00:23:56,119 Speaker 1: how much will it go up by? How much would 432 00:23:56,119 --> 00:23:58,400 Speaker 1: go down by? And what the model does is let 433 00:23:58,440 --> 00:24:02,560 Speaker 1: you take something you can think about in your head, 434 00:24:02,640 --> 00:24:07,119 Speaker 1: like interest rates, abstractly and converted into a dollar price. 435 00:24:07,520 --> 00:24:10,040 Speaker 1: So now let me ask the opposite of that question. 436 00:24:10,119 --> 00:24:13,920 Speaker 1: Let's let's say you've developed a model and it's working well. 437 00:24:13,960 --> 00:24:18,680 Speaker 1: At what point should a human intervene and say, hey, 438 00:24:18,680 --> 00:24:21,360 Speaker 1: this model is no longer producing the output we expect. 439 00:24:21,720 --> 00:24:24,640 Speaker 1: We need to make modifications. How does one even begin 440 00:24:24,680 --> 00:24:27,840 Speaker 1: to think about that? Um. When I worked at Golden 441 00:24:27,880 --> 00:24:29,680 Speaker 1: we kind of try to write to the last year. 442 00:24:29,920 --> 00:24:31,600 Speaker 1: The last years I was there. I was in a 443 00:24:31,640 --> 00:24:34,320 Speaker 1: group called firm wide Risk not in Quantitative Strategies, where 444 00:24:34,320 --> 00:24:37,320 Speaker 1: we were trying to look at derivatives risk throughout the firm, 445 00:24:37,359 --> 00:24:39,720 Speaker 1: and we sort of had a recommendation that every time 446 00:24:39,760 --> 00:24:43,440 Speaker 1: somebody write a model, they'd be forced to specify all 447 00:24:43,480 --> 00:24:46,320 Speaker 1: the assumptions and conditions they were they were making so 448 00:24:46,359 --> 00:24:49,639 Speaker 1: that they could specify when these things wouldn't hold. So, 449 00:24:49,720 --> 00:24:53,440 Speaker 1: for example, if you build a model to price options 450 00:24:53,480 --> 00:24:57,080 Speaker 1: on Apple, you're pretending interest rates will be pretty much stable. 451 00:24:57,160 --> 00:24:59,000 Speaker 1: You don't worry about interest rates. But if you suddenly 452 00:24:59,040 --> 00:25:01,320 Speaker 1: go to some emerging my icket county or interest rates 453 00:25:01,520 --> 00:25:04,280 Speaker 1: can rock at somebody should understand that that's not the 454 00:25:04,359 --> 00:25:09,320 Speaker 1: right model to use. That's interesting. Um. On a related note, 455 00:25:09,960 --> 00:25:13,199 Speaker 1: how could you tell the difference between a model working 456 00:25:13,200 --> 00:25:15,199 Speaker 1: its way through a rough patch and a model that 457 00:25:15,320 --> 00:25:19,280 Speaker 1: just no longer works. That's a really tough question. That's 458 00:25:19,440 --> 00:25:23,040 Speaker 1: especially applicable to people who do a statistical arbitrage or 459 00:25:23,560 --> 00:25:27,840 Speaker 1: people you mentioned like Renaissance or UM. I think that's 460 00:25:27,920 --> 00:25:30,360 Speaker 1: very difficult. You have to two ways. The first way 461 00:25:30,400 --> 00:25:32,639 Speaker 1: is statistical. You sort of have to have some idea 462 00:25:32,680 --> 00:25:36,840 Speaker 1: of when you're getting results that are statistically unlikely. So 463 00:25:36,880 --> 00:25:39,080 Speaker 1: if your model is correct, maybe you'll get, you know, 464 00:25:39,760 --> 00:25:41,760 Speaker 1: some fraction of the time something that's one or two 465 00:25:41,760 --> 00:25:44,720 Speaker 1: standard deviations away. If that persists, you start to sam 466 00:25:44,720 --> 00:25:46,800 Speaker 1: in a different regime or the model is broken down. 467 00:25:47,400 --> 00:25:49,679 Speaker 1: And then the second way, which which I also like, 468 00:25:49,840 --> 00:25:52,840 Speaker 1: is you really ought to have some model. Isn't just mathematics. 469 00:25:52,840 --> 00:25:55,639 Speaker 1: A model has some idea beneath its structural about the 470 00:25:55,640 --> 00:25:58,760 Speaker 1: way things behave, how people respond, how markets behave, and 471 00:25:58,800 --> 00:26:00,040 Speaker 1: you ought to be able to make some sort of 472 00:26:00,119 --> 00:26:02,520 Speaker 1: judgment as to whether the world is still behaving according 473 00:26:02,520 --> 00:26:04,240 Speaker 1: to the assumptions that you make. So it has to 474 00:26:04,280 --> 00:26:07,399 Speaker 1: be disprovable if X and y and z happen, therefore 475 00:26:07,400 --> 00:26:10,040 Speaker 1: the underlying thesis and the model no longer applies. Yes, 476 00:26:10,760 --> 00:26:13,760 Speaker 1: that's quite that's quite interesting. I like the idea that 477 00:26:13,760 --> 00:26:16,240 Speaker 1: that I was going to say this earlier. I sort 478 00:26:16,240 --> 00:26:19,000 Speaker 1: of think just using mathematics blindly is kind of stupid, 479 00:26:19,359 --> 00:26:22,840 Speaker 1: And most models get their inspiration out of some economic 480 00:26:23,000 --> 00:26:25,960 Speaker 1: or or or financial idea first, and the mathematics is 481 00:26:26,000 --> 00:26:28,200 Speaker 1: just the implementation. Where you get into trouble is when 482 00:26:28,240 --> 00:26:30,800 Speaker 1: you think that the mathematics is the thing in itself 483 00:26:30,960 --> 00:26:33,960 Speaker 1: rather than the idea that's behind it. That that goes 484 00:26:34,000 --> 00:26:37,760 Speaker 1: back to the George Box quote. They're wrong, but some 485 00:26:37,840 --> 00:26:41,480 Speaker 1: are useful. Let's so let's talk a little bit about 486 00:26:42,440 --> 00:26:46,720 Speaker 1: some changes that that quants have forced on on both 487 00:26:46,800 --> 00:26:49,679 Speaker 1: markets and investing. There was recently a column in The 488 00:26:49,680 --> 00:26:53,520 Speaker 1: Financial Times that talked about the secret source of hedge 489 00:26:53,520 --> 00:26:59,399 Speaker 1: funds and how quants are essentially reducing what some people 490 00:26:59,440 --> 00:27:03,680 Speaker 1: have previously called as alpha and identifying it as a 491 00:27:03,720 --> 00:27:08,119 Speaker 1: factor that can be reduced to mathematics. And others have 492 00:27:08,280 --> 00:27:12,080 Speaker 1: suggested that the quants are part of the reason why 493 00:27:12,200 --> 00:27:15,919 Speaker 1: hedge fund performance has been so mediocre over the past 494 00:27:15,960 --> 00:27:19,480 Speaker 1: decade or so. As the quants have risen in prominence 495 00:27:19,520 --> 00:27:23,359 Speaker 1: and stature and influence, the ability of a person working 496 00:27:23,359 --> 00:27:27,280 Speaker 1: in a hedge fund to create out performance, to develop 497 00:27:27,320 --> 00:27:32,920 Speaker 1: alpha is going away because of of what quants are doing. 498 00:27:33,119 --> 00:27:35,960 Speaker 1: What What are your thoughts on that? I've sort of 499 00:27:36,000 --> 00:27:38,320 Speaker 1: got a rush of thoughts coming to my head. Say 500 00:27:38,320 --> 00:27:40,520 Speaker 1: that because I think all of the things you said 501 00:27:40,520 --> 00:27:42,760 Speaker 1: are partially true. I think the first one, it's true 502 00:27:42,760 --> 00:27:46,280 Speaker 1: that the whole hedge fund and st allocation or asset 503 00:27:46,320 --> 00:27:49,360 Speaker 1: management world has become much more quantitative. When I started out, 504 00:27:49,720 --> 00:27:52,000 Speaker 1: nobody in those areas knew a lot of math or 505 00:27:52,080 --> 00:27:56,119 Speaker 1: use a lot of math. Now they all do um. 506 00:27:56,160 --> 00:27:58,080 Speaker 1: As a result, I think they're all in competition with 507 00:27:58,119 --> 00:28:00,320 Speaker 1: each other, And I think if somebody is a good 508 00:28:00,320 --> 00:28:02,560 Speaker 1: idea some way people move around a lot, and these 509 00:28:02,840 --> 00:28:05,480 Speaker 1: I've literally seen examples of somebody from one firm going 510 00:28:05,520 --> 00:28:08,200 Speaker 1: somewhere else bringing an idea there, they use it, they 511 00:28:08,240 --> 00:28:11,600 Speaker 1: get irritated because somebody else leaves and takes it somewhere else. 512 00:28:11,600 --> 00:28:14,360 Speaker 1: So I think these models propagate a lot and and 513 00:28:14,480 --> 00:28:17,439 Speaker 1: become become widely used, and that does cut into the 514 00:28:17,480 --> 00:28:21,480 Speaker 1: so called alpha of of everybody. Um. We saw a 515 00:28:21,560 --> 00:28:25,000 Speaker 1: little bit of that with LTCM when people had left 516 00:28:25,000 --> 00:28:28,800 Speaker 1: at or the people covering them at different brokerage firms 517 00:28:28,920 --> 00:28:31,239 Speaker 1: ended up moving around learning a little bit about what 518 00:28:31,280 --> 00:28:33,400 Speaker 1: they were doing, and so a lot of people were 519 00:28:33,400 --> 00:28:37,800 Speaker 1: piggybacking those trades as bad as the leverage was. When 520 00:28:37,840 --> 00:28:41,080 Speaker 1: every desk on the streets imitating it, it's really a 521 00:28:41,080 --> 00:28:43,719 Speaker 1: crowded trade. Yes, And now I think people actually have 522 00:28:43,800 --> 00:28:47,640 Speaker 1: sort of statistical mathematical models that they fit to the 523 00:28:47,680 --> 00:28:50,800 Speaker 1: whole the whole surface of of stock prices and decide 524 00:28:50,800 --> 00:28:53,000 Speaker 1: which ones are cheap and which ones are rich. And 525 00:28:53,120 --> 00:28:56,200 Speaker 1: somebody takes that model somewhere else or or or maybe 526 00:28:56,240 --> 00:28:59,800 Speaker 1: finds the inspiration for that model in some finance paper 527 00:28:59,840 --> 00:29:02,840 Speaker 1: that buried somewhere, and people people actually mind finance papers 528 00:29:02,840 --> 00:29:05,800 Speaker 1: for these sort of anomalies or behavioral anomalies and sought 529 00:29:05,880 --> 00:29:08,880 Speaker 1: to implement them everywhere. So I think these methods are 530 00:29:08,880 --> 00:29:12,120 Speaker 1: finding alpha getting a short run, shorter lifetime. So in 531 00:29:12,160 --> 00:29:16,080 Speaker 1: other words, it's either arbitraged away or just imitated and 532 00:29:16,080 --> 00:29:18,920 Speaker 1: and it loses its ability. That that's my impression. Yeah, 533 00:29:19,000 --> 00:29:21,440 Speaker 1: But at the same time, so there is a lot 534 00:29:21,520 --> 00:29:23,320 Speaker 1: more of this quantitative stuff. But at the same time, 535 00:29:23,360 --> 00:29:25,080 Speaker 1: to be a little on the cynical side, all of 536 00:29:25,120 --> 00:29:27,959 Speaker 1: the hitge funds and asset managers are in a competition 537 00:29:28,000 --> 00:29:31,480 Speaker 1: for assets under management, and they all like to pretend 538 00:29:31,520 --> 00:29:35,080 Speaker 1: that they have more secret source than maybe they actually do. Right, 539 00:29:35,560 --> 00:29:38,680 Speaker 1: So what does that tell us about the future of 540 00:29:38,680 --> 00:29:42,840 Speaker 1: of asset management. We've watched lots of money flow to Vanguard, 541 00:29:43,200 --> 00:29:46,800 Speaker 1: which is primarily a huge indexing shop, but we've also 542 00:29:46,920 --> 00:29:50,280 Speaker 1: seen a lot of money flow to hedge funds despite 543 00:29:50,320 --> 00:29:55,680 Speaker 1: a pretty bad run of underperformance. What explains that is it? 544 00:29:55,720 --> 00:29:58,240 Speaker 1: Is it just marketing wizardry or is there something more 545 00:29:58,280 --> 00:30:00,320 Speaker 1: than that? No, I think there are people and hedge 546 00:30:00,360 --> 00:30:04,960 Speaker 1: funds that really do have um some skill that that 547 00:30:04,960 --> 00:30:07,200 Speaker 1: that people who are just doing indexing, which doesn't take 548 00:30:07,200 --> 00:30:09,560 Speaker 1: any skill at all. Really, I think they do have 549 00:30:09,560 --> 00:30:12,960 Speaker 1: a skill um. I've seen studies that show that, especially 550 00:30:12,960 --> 00:30:16,560 Speaker 1: in in very liquid complex introments like mortgages UM as 551 00:30:16,560 --> 00:30:19,600 Speaker 1: opposed to equity long short, which is fairly simple, that 552 00:30:19,720 --> 00:30:22,320 Speaker 1: hedge funds that tend to do better than average one year, 553 00:30:22,360 --> 00:30:24,560 Speaker 1: there's some persistence they tend to do better than average 554 00:30:24,560 --> 00:30:26,440 Speaker 1: the next year. So it's not just a random district. 555 00:30:26,480 --> 00:30:28,280 Speaker 1: I don't think it's just random. I think if you 556 00:30:28,320 --> 00:30:31,160 Speaker 1: have skill, that tends to persist. But on the other hand, 557 00:30:31,880 --> 00:30:36,120 Speaker 1: the whole world is becoming so swept by people trading mechanically. 558 00:30:36,160 --> 00:30:38,880 Speaker 1: For example, I think if you're a value investor now, 559 00:30:38,960 --> 00:30:40,920 Speaker 1: then you kind of buy things when they go down 560 00:30:41,440 --> 00:30:42,880 Speaker 1: and you think they're cheap. On the other hand, the 561 00:30:43,000 --> 00:30:46,280 Speaker 1: momentum investors who think momentum is a factor and they 562 00:30:46,320 --> 00:30:48,160 Speaker 1: sell things that are going down and buy things that 563 00:30:48,200 --> 00:30:51,160 Speaker 1: are going up. So I think the large number of 564 00:30:51,160 --> 00:30:53,240 Speaker 1: people that are now acting in this mechanical way with 565 00:30:53,320 --> 00:30:57,000 Speaker 1: some model like following momentum tends to mess up the 566 00:30:57,040 --> 00:30:59,200 Speaker 1: people who are looking for value, and it's much harder 567 00:30:59,240 --> 00:31:02,160 Speaker 1: for hedge funds to compete in this world. There's a 568 00:31:02,280 --> 00:31:06,520 Speaker 1: quant name West Gray who writes Alpha Architect, and they 569 00:31:06,520 --> 00:31:09,560 Speaker 1: have a model that is both momentum and value. It's 570 00:31:09,600 --> 00:31:12,680 Speaker 1: actually two sleeves, and he said one sleeve is always 571 00:31:12,680 --> 00:31:14,440 Speaker 1: doing great while the other one is doing terrible, and 572 00:31:14,560 --> 00:31:17,240 Speaker 1: vice versa. But over the long haul, both of them 573 00:31:17,280 --> 00:31:20,239 Speaker 1: actually work out pretty well. That just gives you an 574 00:31:20,240 --> 00:31:23,720 Speaker 1: idea of the complexity of what we've seen come along. 575 00:31:24,360 --> 00:31:27,040 Speaker 1: That just there was nothing like that ten or twenty 576 00:31:27,160 --> 00:31:29,080 Speaker 1: years ago. You know, you look at the risk parity 577 00:31:29,120 --> 00:31:31,760 Speaker 1: people that say you should have instead of being sixty 578 00:31:32,120 --> 00:31:36,240 Speaker 1: equities bonds along the same philosophy you bedroo of having 579 00:31:36,400 --> 00:31:38,880 Speaker 1: one third of your risk in commodities, one third of 580 00:31:38,880 --> 00:31:40,840 Speaker 1: your risk and equities and one third of your risk 581 00:31:40,880 --> 00:31:43,360 Speaker 1: and bonds, which means livering leveraging a bonds a lot 582 00:31:43,720 --> 00:31:46,960 Speaker 1: because they have such low low volatility, and they also 583 00:31:47,320 --> 00:31:49,240 Speaker 1: believe that in the long run that will work best. 584 00:31:49,280 --> 00:31:51,320 Speaker 1: But in fact they've done badly for the last few months. 585 00:31:51,320 --> 00:31:53,680 Speaker 1: So well, you know, we saw a lot of money 586 00:31:53,720 --> 00:31:57,520 Speaker 1: floated risk parity, as as bonds looked like as rates 587 00:31:57,560 --> 00:32:00,360 Speaker 1: looked like they were bottoming, and as lot of the 588 00:32:00,400 --> 00:32:04,240 Speaker 1: commodity complex seemed to be topping out. About five years ago. 589 00:32:04,600 --> 00:32:06,800 Speaker 1: So if you're a third commodities and gold is down 590 00:32:06,800 --> 00:32:09,440 Speaker 1: thirty percent and oil is cut in half, that's gonna 591 00:32:09,440 --> 00:32:12,480 Speaker 1: have a pretty big impact on on your returns. Yeah. 592 00:32:12,480 --> 00:32:14,480 Speaker 1: I like to be a little philosophical and say that 593 00:32:15,080 --> 00:32:17,280 Speaker 1: if you take a model too seriously, it's a kind 594 00:32:17,280 --> 00:32:20,880 Speaker 1: of idolatry in the sense you're assuming that something somebody. 595 00:32:21,200 --> 00:32:23,280 Speaker 1: You're assuming that you can write down a formula that's 596 00:32:23,320 --> 00:32:26,200 Speaker 1: gonna mimic the way people behave. But people are too 597 00:32:26,200 --> 00:32:29,280 Speaker 1: complicated and and if you really believe that you can 598 00:32:29,520 --> 00:32:32,360 Speaker 1: capture people in a formula equation, you're you're looking for 599 00:32:32,400 --> 00:32:34,880 Speaker 1: trouble in the long run. I know you have a website, 600 00:32:34,920 --> 00:32:37,400 Speaker 1: a Manual Derman dot com. If people want to find 601 00:32:37,440 --> 00:32:40,200 Speaker 1: more of your your writings, it's my life is a 602 00:32:40,280 --> 00:32:44,360 Speaker 1: quant and models behaving badly, And you post regular papers 603 00:32:44,360 --> 00:32:47,400 Speaker 1: at a manual Derman dot com not that often anymore. 604 00:32:47,440 --> 00:32:51,120 Speaker 1: I do go on Twitter regularly. I'm a fan. Um 605 00:32:51,120 --> 00:32:54,160 Speaker 1: what is your Twitter handle? ETI manual dermant in one 606 00:32:54,200 --> 00:32:58,080 Speaker 1: word at a Manual Derman. If you've enjoyed this conversation, 607 00:32:58,320 --> 00:33:01,320 Speaker 1: be sure and check out our podcast extras, where the 608 00:33:01,320 --> 00:33:05,080 Speaker 1: tape keeps rolling, and we continue the conversation about physics 609 00:33:05,200 --> 00:33:09,240 Speaker 1: and quantitative trading and philosophy. Be sure and check out 610 00:33:09,320 --> 00:33:12,920 Speaker 1: my daily column on Bloomberg View dot com. Follow me 611 00:33:13,000 --> 00:33:16,520 Speaker 1: on Twitter at rid Halts. I'm Barry Ridholts. You're listening 612 00:33:16,520 --> 00:33:20,400 Speaker 1: to Masters in Business on Bloomberg Radio. Welcome back to 613 00:33:20,480 --> 00:33:24,120 Speaker 1: the podcast. Thank you so much, Professor Derman for doing this. 614 00:33:24,120 --> 00:33:27,120 Speaker 1: This is really fascinating stuff. And I don't think people 615 00:33:27,160 --> 00:33:30,760 Speaker 1: get to hear you often enough other than a handful 616 00:33:30,800 --> 00:33:34,240 Speaker 1: of Columbia students. Um. And I've been following your career 617 00:33:34,280 --> 00:33:36,840 Speaker 1: long enough that I really wanted to get you in 618 00:33:36,920 --> 00:33:41,160 Speaker 1: here and and put you under the microscope. So so 619 00:33:41,240 --> 00:33:43,880 Speaker 1: let's go over a couple of questions that we missed 620 00:33:43,960 --> 00:33:46,840 Speaker 1: earlier on And I know I only have you for 621 00:33:46,880 --> 00:33:49,840 Speaker 1: a finite amount of time, but we're we got plenty 622 00:33:49,840 --> 00:33:53,320 Speaker 1: of time to go. Um, So first let's go back 623 00:33:53,360 --> 00:33:59,440 Speaker 1: to this was way ahead of the curve in you 624 00:33:59,480 --> 00:34:04,640 Speaker 1: wrote an article titled model risk, pointing out the dangers 625 00:34:04,720 --> 00:34:08,759 Speaker 1: that inevitably accompany the use of models. How did that 626 00:34:08,800 --> 00:34:12,360 Speaker 1: play out? I think it played out pretty accurately. Actually, 627 00:34:12,400 --> 00:34:15,960 Speaker 1: although I was considered I was sort of looking I 628 00:34:15,960 --> 00:34:17,560 Speaker 1: can't remember too clearly. But I was looking at two 629 00:34:17,640 --> 00:34:20,080 Speaker 1: kinds of risk. One was implementation risk, where you have 630 00:34:20,120 --> 00:34:22,320 Speaker 1: the idea right, but you've got all sorts of computer 631 00:34:22,400 --> 00:34:25,120 Speaker 1: problems or efficiency problems. Then I was looking at what 632 00:34:25,160 --> 00:34:28,160 Speaker 1: happens when you actually have the idea wrong. And yeah, 633 00:34:28,160 --> 00:34:29,920 Speaker 1: there was the first paper I think ever written on 634 00:34:30,040 --> 00:34:33,000 Speaker 1: model risk, and I think it played out kind of accurately. 635 00:34:33,040 --> 00:34:35,520 Speaker 1: That's what sort of what sort of happened. So so 636 00:34:35,560 --> 00:34:37,719 Speaker 1: the next question, let's see if I can fix this 637 00:34:37,840 --> 00:34:42,839 Speaker 1: little thing here that is there you go. So the 638 00:34:42,920 --> 00:34:47,120 Speaker 1: crisis in two thousand seven two thou eight, how much 639 00:34:47,160 --> 00:34:52,200 Speaker 1: of that was a function of UM models not working well? 640 00:34:54,200 --> 00:34:56,640 Speaker 1: You know, as part of it definitely was, But I 641 00:34:56,640 --> 00:34:58,600 Speaker 1: don't think it was the fundamental cause there are a 642 00:34:58,640 --> 00:35:03,120 Speaker 1: lot of things happening. My sort of slightly biased view 643 00:35:03,239 --> 00:35:07,120 Speaker 1: is that UM really interest rates were very low, and 644 00:35:07,640 --> 00:35:09,879 Speaker 1: everybody was trying to stretch for yield and do anything 645 00:35:09,920 --> 00:35:12,640 Speaker 1: that would get more yield. Absolutely, and that's partly the 646 00:35:12,640 --> 00:35:17,280 Speaker 1: federal reserves fault, but whatever. And I think models played 647 00:35:17,280 --> 00:35:19,799 Speaker 1: a secondary role in that they were used to construct, 648 00:35:19,800 --> 00:35:24,000 Speaker 1: perhaps somewhat deceptively, through the rating agencies instruments that that 649 00:35:24,160 --> 00:35:26,560 Speaker 1: purported to give you a high yield with a low risk, 650 00:35:27,040 --> 00:35:30,680 Speaker 1: and some models were, um, we're a tool in trying 651 00:35:30,719 --> 00:35:33,360 Speaker 1: to cater to high risk, but I don't cater to 652 00:35:33,360 --> 00:35:36,760 Speaker 1: to disaster. But I don't think they were the fundamental cause. 653 00:35:37,640 --> 00:35:40,560 Speaker 1: It's certainly when you look at the rating agencies. I'll 654 00:35:40,560 --> 00:35:44,640 Speaker 1: never forget being in a conference room when a bunch 655 00:35:44,680 --> 00:35:47,840 Speaker 1: of salespeople came in and we're pitching us this new 656 00:35:48,200 --> 00:35:53,040 Speaker 1: fangled subprime product, and the phrase that resonated me with 657 00:35:53,080 --> 00:35:56,680 Speaker 1: me was this is just as safe as U S treasuries, 658 00:35:56,719 --> 00:36:01,840 Speaker 1: but it pays two fifty basis points more. And to 659 00:36:01,920 --> 00:36:04,160 Speaker 1: me it was, well, which is it? It can't pay 660 00:36:04,320 --> 00:36:06,520 Speaker 1: that much more? If it pays a little more, and 661 00:36:06,600 --> 00:36:10,920 Speaker 1: get arbitraged away, but how could it pay almost three more? 662 00:36:11,080 --> 00:36:13,880 Speaker 1: That either you guys are gonna win the Nobel Prize 663 00:36:13,960 --> 00:36:16,800 Speaker 1: or you're gonna be wearing aren't jumpsuits picking up trash 664 00:36:16,840 --> 00:36:18,879 Speaker 1: on the side of the road. It turned out neither 665 00:36:18,880 --> 00:36:22,640 Speaker 1: were true. They most of the people who participate in that, well, 666 00:36:22,680 --> 00:36:24,600 Speaker 1: they just moved on to the next thing and there 667 00:36:24,640 --> 00:36:29,440 Speaker 1: was no subsequent accountability. But the rating agencies and the 668 00:36:29,640 --> 00:36:34,279 Speaker 1: idea that hey, we've created taken draws and turned it 669 00:36:34,280 --> 00:36:39,439 Speaker 1: into gold. People wildly believed that that was possible, wasn't it. Yeah, 670 00:36:39,480 --> 00:36:41,520 Speaker 1: you know, I always like to make the analogy that 671 00:36:41,560 --> 00:36:45,040 Speaker 1: finances a lot like nutrition, in that people take a 672 00:36:45,120 --> 00:36:48,200 Speaker 1: small amount of information and extrapolated like crazy. You know, 673 00:36:48,320 --> 00:36:51,400 Speaker 1: so they will tell you women should or shouldn't get estrogenally, 674 00:36:51,400 --> 00:36:53,319 Speaker 1: you should or shouldn't eat eggs, and then they changed 675 00:36:53,320 --> 00:36:55,840 Speaker 1: in mind dramatically a few months a few years later, 676 00:36:56,320 --> 00:37:00,759 Speaker 1: and it's hard to tell who's a crank and who's correct, because, um, 677 00:37:00,800 --> 00:37:02,680 Speaker 1: basically they're all aimed at I don't know, they're all 678 00:37:02,719 --> 00:37:05,800 Speaker 1: aimed at marketing. In some sense, I'm fond of saying 679 00:37:05,880 --> 00:37:09,799 Speaker 1: predictions and forecasts and marketing because that's all they are. 680 00:37:09,960 --> 00:37:16,480 Speaker 1: But it's amazing how common sense, and granted this is 681 00:37:16,520 --> 00:37:19,720 Speaker 1: all after the fact, but a lot of common sense 682 00:37:19,719 --> 00:37:22,319 Speaker 1: could have saved people a lot of money if wait, 683 00:37:22,800 --> 00:37:25,400 Speaker 1: how could it be as safest treasuring yet pay so 684 00:37:25,560 --> 00:37:28,280 Speaker 1: much more? But nobody really stopped to ask those questions. 685 00:37:28,280 --> 00:37:32,600 Speaker 1: Then it it's amazing that nineties paper is really a 686 00:37:32,719 --> 00:37:35,720 Speaker 1: decade ahead of its time and it turned out, um 687 00:37:35,760 --> 00:37:38,120 Speaker 1: really to be significant, which leads to my next question. 688 00:37:38,560 --> 00:37:43,319 Speaker 1: So models behaving badly. How did that impact capitalism and 689 00:37:43,400 --> 00:37:48,960 Speaker 1: the Great Financial Crisis? Yeah, I mentioned this earlier, but 690 00:37:49,080 --> 00:37:51,600 Speaker 1: I've found the whole thing really disappointing. I'm sort of 691 00:37:51,640 --> 00:37:54,520 Speaker 1: disillusioned in that. That's what I said earlier. Some I 692 00:37:54,560 --> 00:37:58,640 Speaker 1: expected that. Um, I sort of like capitalism, but as 693 00:37:58,680 --> 00:38:01,120 Speaker 1: I said before, I think if you to benefit from 694 00:38:01,120 --> 00:38:04,520 Speaker 1: from taking risk, you've got to suffer the consequences. And 695 00:38:04,600 --> 00:38:08,520 Speaker 1: that really hasn't happened. And I'm with you on that also. 696 00:38:08,640 --> 00:38:14,600 Speaker 1: It's well from my perspective, it's like everyone's a capital nation. 697 00:38:14,719 --> 00:38:18,160 Speaker 1: Everybody's a capitalist until you know they're in trouble, and 698 00:38:18,200 --> 00:38:22,880 Speaker 1: then suddenly they they become temporary socialists until they get there, right. 699 00:38:23,760 --> 00:38:26,359 Speaker 1: I don't know, it was it was astonishing to see 700 00:38:26,360 --> 00:38:30,560 Speaker 1: how quickly people flipped from that. Um, let me go 701 00:38:30,600 --> 00:38:34,120 Speaker 1: through some more of my questions. We went over, Oh, 702 00:38:34,200 --> 00:38:36,239 Speaker 1: I have something that I've been dying to ask you. 703 00:38:36,600 --> 00:38:39,840 Speaker 1: I love this quote of yours. The efficient market model 704 00:38:40,480 --> 00:38:44,560 Speaker 1: a model and an analogy, but not a valid theory. 705 00:38:45,520 --> 00:38:48,480 Speaker 1: How is that not a valid theory. Let's hold aside 706 00:38:49,200 --> 00:38:52,840 Speaker 1: whether or not the efficient market hypothesis is accurate, but 707 00:38:53,040 --> 00:38:57,400 Speaker 1: why isn't it a theory, it's not a theory because 708 00:38:57,440 --> 00:39:00,880 Speaker 1: it doesn't describe the way things really behave. If I 709 00:39:00,880 --> 00:39:03,319 Speaker 1: can give you an example of I had a sort 710 00:39:03,360 --> 00:39:05,479 Speaker 1: of anecdot to tell you what I meant by theory 711 00:39:05,520 --> 00:39:07,560 Speaker 1: in a corny sort of way. When my son was 712 00:39:07,640 --> 00:39:09,359 Speaker 1: very smaller, used to put him my knee and play 713 00:39:09,440 --> 00:39:11,640 Speaker 1: with him and bounce him up and down. And my 714 00:39:11,680 --> 00:39:13,520 Speaker 1: sister used to do it to me. Say half a 715 00:39:13,520 --> 00:39:15,879 Speaker 1: pound of tupanny rice, half a pound of treacle, mix 716 00:39:15,960 --> 00:39:17,840 Speaker 1: them up and make them last. Pop goes the weasel. 717 00:39:17,880 --> 00:39:19,839 Speaker 1: It's an old English thing. And then you drop your 718 00:39:19,880 --> 00:39:22,160 Speaker 1: knees and the kid drops down to the floor and 719 00:39:22,320 --> 00:39:25,120 Speaker 1: he chortles with glee. And then he said again, and 720 00:39:25,160 --> 00:39:26,719 Speaker 1: I did it again, and then he said again, and 721 00:39:26,760 --> 00:39:28,640 Speaker 1: I did again. When I did it the fifteenth time, 722 00:39:29,239 --> 00:39:31,080 Speaker 1: when I dropped him down, he said to me, why 723 00:39:31,080 --> 00:39:36,920 Speaker 1: it's not funny anymore. And so there's no explanation of 724 00:39:36,920 --> 00:39:38,680 Speaker 1: why it's not funny anymore. The fact is, if you 725 00:39:38,719 --> 00:39:41,400 Speaker 1: tell the drug fifteen times, it's not funny anymore. And 726 00:39:41,480 --> 00:39:44,440 Speaker 1: that's both the fact and the theory. It's true, and 727 00:39:45,080 --> 00:39:47,400 Speaker 1: there's no explanation for it. That's just so I was 728 00:39:47,440 --> 00:39:49,680 Speaker 1: trying to say earlier about GOODA. When you describe something 729 00:39:49,719 --> 00:39:52,759 Speaker 1: that's true, you can't say why is it true? You 730 00:39:52,840 --> 00:39:55,520 Speaker 1: just say it's true, and the words are the theory 731 00:39:55,560 --> 00:39:57,600 Speaker 1: and the fact is that that's the way it behaves. 732 00:39:57,760 --> 00:40:01,400 Speaker 1: The efficient market model is not like that, Um, markets 733 00:40:01,400 --> 00:40:05,360 Speaker 1: are inefficient. Um. It's a model. It's not a description 734 00:40:05,400 --> 00:40:07,080 Speaker 1: of the way the world actually is. The world could 735 00:40:07,120 --> 00:40:10,040 Speaker 1: be that way, but it isn't. So so I'm fond 736 00:40:10,080 --> 00:40:14,280 Speaker 1: of saying that markets are kind of sort of eventually 737 00:40:15,200 --> 00:40:18,040 Speaker 1: somewhat efficient, yes, in the longer, in the long run, 738 00:40:18,160 --> 00:40:20,200 Speaker 1: more or less, I liked I used to like to 739 00:40:20,239 --> 00:40:23,560 Speaker 1: say that the efficient market hypothesis and model was a 740 00:40:23,640 --> 00:40:27,160 Speaker 1: very clever jiu jitsu trick by economists who couldn't predict 741 00:40:27,200 --> 00:40:29,440 Speaker 1: what was going to happen, and instead of giving up 742 00:40:29,560 --> 00:40:31,880 Speaker 1: society to make that a principle, so turning sort of 743 00:40:31,880 --> 00:40:35,759 Speaker 1: weakness into strength. That's that's interesting, and and it is true. 744 00:40:35,800 --> 00:40:39,640 Speaker 1: Everybody who believes in the efficient market hypothesis also believe 745 00:40:39,719 --> 00:40:42,239 Speaker 1: that in the random walk, we don't know what's going 746 00:40:42,280 --> 00:40:46,040 Speaker 1: to happen. So therefore by low cost indexes and rebalance 747 00:40:46,080 --> 00:40:48,480 Speaker 1: on a regular basis, And that's good. As long as 748 00:40:48,480 --> 00:40:51,000 Speaker 1: not everybody does that, you still need somebody out there 749 00:40:51,040 --> 00:40:55,680 Speaker 1: looking for value otherwise. Otherwise, Yeah, as much as money 750 00:40:55,760 --> 00:40:59,560 Speaker 1: has been flowing to places like Black Rock, I Shares 751 00:40:59,640 --> 00:41:03,000 Speaker 1: and and Guard, it's still a tiny percentage of the 752 00:41:03,080 --> 00:41:06,800 Speaker 1: overall investment. What is the three trillion dollars in indicries 753 00:41:06,840 --> 00:41:09,880 Speaker 1: now three trillion dollars in ETFs area that today not 754 00:41:10,640 --> 00:41:13,480 Speaker 1: a whole lot of money relative to what's the total 755 00:41:13,600 --> 00:41:19,319 Speaker 1: investable universe sixty eight or seventy five trillion dollars worldwide? Yeah, 756 00:41:19,360 --> 00:41:23,719 Speaker 1: I think so, yeah, that maybe trillion. I think the 757 00:41:23,920 --> 00:41:28,759 Speaker 1: US equity markets are are are about that, the bond 758 00:41:28,840 --> 00:41:30,480 Speaker 1: markets are more, and then you have the rest of 759 00:41:30,520 --> 00:41:32,920 Speaker 1: the world. At one point in time, US was almost 760 00:41:33,000 --> 00:41:35,800 Speaker 1: half of the investible assets and in the globe that 761 00:41:35,920 --> 00:41:39,520 Speaker 1: has since since changed. So so I think a lot 762 00:41:39,560 --> 00:41:41,680 Speaker 1: of models work well. Related to what you were saying, 763 00:41:42,200 --> 00:41:44,280 Speaker 1: they work well as long as you're just a rupple 764 00:41:44,360 --> 00:41:46,959 Speaker 1: on the sea. But when you become the whole, see, 765 00:41:47,120 --> 00:41:52,319 Speaker 1: then the model right, it's it's you. You no longer 766 00:41:52,520 --> 00:41:54,839 Speaker 1: trading in the market, you are the market, and that's 767 00:41:54,840 --> 00:41:59,800 Speaker 1: a very different um. I love this line. Explain for 768 00:42:00,040 --> 00:42:04,480 Speaker 1: us what is the unbearable futility of modeling. Oh is 769 00:42:04,520 --> 00:42:08,480 Speaker 1: that a line of mine? Yes? Okay, Um, well, I'm 770 00:42:08,520 --> 00:42:11,399 Speaker 1: not sure where that comes from, but I believe it's 771 00:42:11,440 --> 00:42:18,719 Speaker 1: a chapter or sub chapter heading in towards the ending. Um. 772 00:42:20,080 --> 00:42:21,640 Speaker 1: I'm not sure what I meant by that, to be honest, 773 00:42:21,680 --> 00:42:23,680 Speaker 1: I think I probably just meant what I said before 774 00:42:23,760 --> 00:42:26,239 Speaker 1: that in the end, you will always be wrong, and 775 00:42:27,239 --> 00:42:29,279 Speaker 1: you will you will never you will never be you know, 776 00:42:29,400 --> 00:42:31,080 Speaker 1: if if you it's what you were saying earlier. If 777 00:42:31,120 --> 00:42:33,360 Speaker 1: you find some law of nature, it kind of holds forever, 778 00:42:33,520 --> 00:42:37,239 Speaker 1: and if you find a model, it's inevitably going to 779 00:42:37,520 --> 00:42:42,400 Speaker 1: going to require revamping, changing alteration as people start to 780 00:42:42,840 --> 00:42:46,000 Speaker 1: change your behavior. Do you miss physics at all, given 781 00:42:46,080 --> 00:42:50,440 Speaker 1: that the certitude that you get in physics versus always 782 00:42:50,480 --> 00:42:54,560 Speaker 1: dealing with uncertainty about what's happening and about how long 783 00:42:54,640 --> 00:42:58,040 Speaker 1: a model is going to have a Literally, No, I don't. 784 00:42:58,080 --> 00:43:00,040 Speaker 1: I'm very glad I got a background in physics, but 785 00:43:00,120 --> 00:43:01,600 Speaker 1: I'm a most at this when glad I got out 786 00:43:01,640 --> 00:43:03,280 Speaker 1: of it. It's like a rough life being in physics 787 00:43:03,400 --> 00:43:06,640 Speaker 1: because there's so many incredibly smart people that you actually 788 00:43:06,719 --> 00:43:09,839 Speaker 1: run into people and you say to yourself. I can, 789 00:43:09,920 --> 00:43:11,800 Speaker 1: I can sort of understand what they're doing, but I 790 00:43:11,880 --> 00:43:14,160 Speaker 1: could never have done that myself. So it's actually a 791 00:43:14,239 --> 00:43:18,319 Speaker 1: bit of a relish. So you come to finance, where 792 00:43:18,600 --> 00:43:22,080 Speaker 1: essentially it's easy to be the smart I'm not, I'm 793 00:43:22,280 --> 00:43:26,200 Speaker 1: reasonably smart, but but finance also Actually, physics and finance 794 00:43:26,239 --> 00:43:28,200 Speaker 1: theres lots of room for taste, I think, and where 795 00:43:28,239 --> 00:43:30,040 Speaker 1: people go wrong is when they don't have taste and 796 00:43:30,080 --> 00:43:33,279 Speaker 1: they start slavishly sort of following the mathematics too much. 797 00:43:33,320 --> 00:43:37,000 Speaker 1: Physics is really driven by ideas and then mathematics to 798 00:43:37,160 --> 00:43:39,120 Speaker 1: implement them, and I think it should be that way 799 00:43:39,160 --> 00:43:41,320 Speaker 1: in finance too. I'm gonna push back at you know 800 00:43:41,360 --> 00:43:43,920 Speaker 1: a little bit, because I think you were a grad 801 00:43:44,080 --> 00:43:47,800 Speaker 1: student and you wrote a paper on it wasn't Higgs, Boston, 802 00:43:47,880 --> 00:43:50,080 Speaker 1: but maybe it was one of the Higgs particles that 803 00:43:50,200 --> 00:43:55,680 Speaker 1: became widely um, you know, widely regarded, And to do 804 00:43:55,880 --> 00:43:59,640 Speaker 1: that in physics as a grad student is no easy thing. No. Yeah, 805 00:43:59,640 --> 00:44:02,160 Speaker 1: I wrote a thesis on writing tests for how to 806 00:44:02,320 --> 00:44:05,759 Speaker 1: find a w boson that was cause don't ask me 807 00:44:05,840 --> 00:44:09,399 Speaker 1: where I pull that from than Wikipedia, but that's true. 808 00:44:09,440 --> 00:44:11,759 Speaker 1: And then they eventually discovered that although I was. Yeah, 809 00:44:12,520 --> 00:44:15,480 Speaker 1: now I was. I was a reasonably good physicist. But nevertheless, 810 00:44:15,600 --> 00:44:17,719 Speaker 1: and maybe in finance toopid. Let's say, in physics you 811 00:44:17,800 --> 00:44:20,279 Speaker 1: run into people I heard Fineman talk or people like that, 812 00:44:20,400 --> 00:44:26,040 Speaker 1: or Richard Fineman over in California was at cal Tech. Yeah, 813 00:44:26,239 --> 00:44:30,440 Speaker 1: and so did that Was that intimidating or impressive? It 814 00:44:30,960 --> 00:44:34,680 Speaker 1: was kind of intimidating or even I sort of wrote 815 00:44:34,719 --> 00:44:36,640 Speaker 1: in My Life as a Quantuent, my memoir, I wrote, 816 00:44:36,719 --> 00:44:39,600 Speaker 1: sort of, you see your ambitions slowly getting degraded. When 817 00:44:39,640 --> 00:44:42,000 Speaker 1: you're like eighteen, you want to be like Einstein, and 818 00:44:42,080 --> 00:44:43,840 Speaker 1: when you're like twenty five, you want to be like 819 00:44:44,000 --> 00:44:46,160 Speaker 1: the professor that you know in your department. And then 820 00:44:46,160 --> 00:44:48,399 Speaker 1: when you're like thirty two, you say, gee, the guy 821 00:44:48,440 --> 00:44:50,480 Speaker 1: at the desk next to me is getting more invitations 822 00:44:50,520 --> 00:44:53,719 Speaker 1: to give seminars than me. And then you suddenly think, well, 823 00:44:54,719 --> 00:44:57,719 Speaker 1: look where I've got to. So it's that competitive, and 824 00:44:57,840 --> 00:45:00,920 Speaker 1: it's that it is competitive. I got a job as 825 00:45:01,000 --> 00:45:04,520 Speaker 1: a particle physicist at Boulder in nine was my last 826 00:45:04,600 --> 00:45:06,800 Speaker 1: job in physics, and they were like a hundred and 827 00:45:06,880 --> 00:45:10,000 Speaker 1: forty people applying for one job. It was really rough. 828 00:45:10,040 --> 00:45:11,719 Speaker 1: I mean it was after the Vietnam War, there was 829 00:45:11,760 --> 00:45:14,920 Speaker 1: no money for research anymore. All the physics jobs that 830 00:45:15,000 --> 00:45:16,799 Speaker 1: filled up when there was still a lot of money 831 00:45:16,840 --> 00:45:19,640 Speaker 1: before the head Field govern Amendment which stopped the money 832 00:45:19,719 --> 00:45:23,080 Speaker 1: going to research. And I was maybe number four on 833 00:45:23,160 --> 00:45:25,520 Speaker 1: the list or something in the first three either had 834 00:45:26,080 --> 00:45:29,400 Speaker 1: had dual career problems or got better jobs and I 835 00:45:29,480 --> 00:45:32,160 Speaker 1: got the job. But but um, yeah, it was kind 836 00:45:32,200 --> 00:45:36,360 Speaker 1: of it was kind of discouraging. So what about today 837 00:45:36,480 --> 00:45:39,440 Speaker 1: for financial engineers? You mentioned there's lots of quants around. 838 00:45:40,000 --> 00:45:42,839 Speaker 1: Are there are there still opportunities for people who want 839 00:45:42,880 --> 00:45:47,839 Speaker 1: to go into this field and and apply mathematics to finance. Yeah, 840 00:45:47,840 --> 00:45:49,759 Speaker 1: I think there are. There's a much bigger market for 841 00:45:49,880 --> 00:45:52,120 Speaker 1: risk management now. Some of it is PR. People just 842 00:45:52,160 --> 00:45:54,160 Speaker 1: want to say they have risk management, but sometimes they 843 00:45:54,160 --> 00:45:56,919 Speaker 1: have real risk management. And ye know what I liked 844 00:45:56,960 --> 00:45:59,000 Speaker 1: about if I talk pass what I liked about getting 845 00:45:59,000 --> 00:46:01,240 Speaker 1: to Goldman as opposed to in physics. I like physics, 846 00:46:01,640 --> 00:46:03,880 Speaker 1: but in physics someone had the feeling like you had 847 00:46:03,920 --> 00:46:06,160 Speaker 1: to be really superb or otherwise you were wasting your 848 00:46:06,239 --> 00:46:08,200 Speaker 1: life because you spent all your time. I was doing 849 00:46:08,200 --> 00:46:11,120 Speaker 1: theoretical physics singing in an office and banging your head 850 00:46:11,120 --> 00:46:13,560 Speaker 1: against the wall trying to solve something difficult, and you 851 00:46:13,600 --> 00:46:16,120 Speaker 1: could spend half your time sort of depressed. It was 852 00:46:16,200 --> 00:46:19,160 Speaker 1: nice about being a Goldman. What I actually liked about 853 00:46:19,200 --> 00:46:21,520 Speaker 1: finance was it was a more it was a more 854 00:46:21,640 --> 00:46:23,879 Speaker 1: multi sided world where you spend part of your time 855 00:46:23,920 --> 00:46:26,520 Speaker 1: doing theory, but part of your time writing programs, and 856 00:46:26,600 --> 00:46:28,880 Speaker 1: part of your time interacting with clients and with traders, 857 00:46:29,360 --> 00:46:31,759 Speaker 1: and so there were many more sources of satisfaction. You 858 00:46:31,840 --> 00:46:34,200 Speaker 1: had long term projects and short term projects, and I 859 00:46:34,239 --> 00:46:36,400 Speaker 1: found that very satisfying. I got quite excited when I 860 00:46:36,440 --> 00:46:38,800 Speaker 1: went to it goes. So, after almost twenty years on 861 00:46:38,840 --> 00:46:40,920 Speaker 1: Wall Street, you transition to being a professor. How do 862 00:46:40,960 --> 00:46:42,320 Speaker 1: you How do you like teaching? How do you like 863 00:46:42,440 --> 00:46:47,120 Speaker 1: working with with young people? I like it, um I 864 00:46:47,120 --> 00:46:49,200 Speaker 1: would say it's not as exciting as as being on 865 00:46:49,320 --> 00:46:52,840 Speaker 1: Wall Street. People in universities are sort of siloed, and 866 00:46:52,880 --> 00:46:54,520 Speaker 1: that everybody is there because they want to get on 867 00:46:54,600 --> 00:46:57,600 Speaker 1: with their research and have students, and so I would 868 00:46:57,640 --> 00:47:00,960 Speaker 1: say in a sense, being at Goldman was more collegial 869 00:47:01,040 --> 00:47:03,480 Speaker 1: in being in a university for me, despite the university 870 00:47:03,560 --> 00:47:06,840 Speaker 1: being a college, because actually everybody wants to sort of 871 00:47:06,960 --> 00:47:08,680 Speaker 1: I'm exaggerating a little bit, but leave me alone. I 872 00:47:08,719 --> 00:47:10,759 Speaker 1: have work to do. Was it, Goldmen? You could go 873 00:47:10,840 --> 00:47:13,200 Speaker 1: to somebody and say, I'm thinking about this problem. Can 874 00:47:13,280 --> 00:47:16,440 Speaker 1: you help me? And because you're all paid to pull 875 00:47:16,480 --> 00:47:18,880 Speaker 1: in the same direction, they sort of helped you more willingly. 876 00:47:18,920 --> 00:47:20,640 Speaker 1: I would say, so a little more of a team 877 00:47:20,719 --> 00:47:25,080 Speaker 1: effort on Wall Street than in a university where everybody 878 00:47:25,160 --> 00:47:28,480 Speaker 1: has a list of things to do and students to 879 00:47:28,560 --> 00:47:31,560 Speaker 1: deal with and mid terms and papers exactly and their tenures. 880 00:47:31,640 --> 00:47:36,560 Speaker 1: So go away, I'm busy right that that that's fascinating. Um, 881 00:47:36,880 --> 00:47:41,320 Speaker 1: how has quantitative investing evolved since you began in the 882 00:47:41,400 --> 00:47:46,920 Speaker 1: industry all those years ago? Um? Well, first of all, 883 00:47:47,080 --> 00:47:50,319 Speaker 1: everybody on trading disks is, as I said, simple times before, 884 00:47:50,440 --> 00:47:52,919 Speaker 1: much more numerous. Now a lot of them can write 885 00:47:52,920 --> 00:47:55,240 Speaker 1: their own models. A lot of them studied math, especially 886 00:47:55,280 --> 00:47:57,640 Speaker 1: in derivatives. You get a lot of French or or 887 00:47:58,080 --> 00:48:00,719 Speaker 1: even American students who have all gotten advant degrees or 888 00:48:00,880 --> 00:48:04,080 Speaker 1: PhDs in finance, So that's become much more complicated. The 889 00:48:04,160 --> 00:48:07,759 Speaker 1: second thing that's changed a lot is is electronic price 890 00:48:07,880 --> 00:48:11,719 Speaker 1: feeds and electronic settlements. So suddenly computers have become much 891 00:48:11,800 --> 00:48:16,000 Speaker 1: more important. Jeff Gunlock tells the story about his one 892 00:48:16,040 --> 00:48:19,120 Speaker 1: of his first jobs on a bond desk. He shows 893 00:48:19,200 --> 00:48:21,680 Speaker 1: up with a math background and is able to do 894 00:48:21,880 --> 00:48:26,040 Speaker 1: some really basic calculations in his head, and everybody there 895 00:48:26,120 --> 00:48:30,279 Speaker 1: thinks he's a wizard, because they're not math guys yet 896 00:48:30,440 --> 00:48:33,320 Speaker 1: they're on a bond desk. Thirty years ago. Yeah, I 897 00:48:33,400 --> 00:48:35,680 Speaker 1: started out on a bond desk, and actually along the 898 00:48:35,760 --> 00:48:38,200 Speaker 1: lines of what you're saying, I worked on this model 899 00:48:38,480 --> 00:48:40,880 Speaker 1: which you mentioned, the Black Derman toy model. But I 900 00:48:40,960 --> 00:48:42,880 Speaker 1: think one of the things that almost as much impact 901 00:48:43,000 --> 00:48:45,000 Speaker 1: was that I could program well in a in a 902 00:48:45,160 --> 00:48:47,040 Speaker 1: in a world where people couldn't, and I built a 903 00:48:47,160 --> 00:48:51,360 Speaker 1: user interface that let the traders enter trade, save it, 904 00:48:51,480 --> 00:48:53,399 Speaker 1: think about it, come back the next day, and talk 905 00:48:53,560 --> 00:48:57,200 Speaker 1: the salespeople talk to a client again. And user interface 906 00:48:57,360 --> 00:48:59,480 Speaker 1: makes a lot of difference. So in those days, if 907 00:48:59,520 --> 00:49:02,040 Speaker 1: you could act really in those days, you couldn't get 908 00:49:02,040 --> 00:49:03,279 Speaker 1: something to do it for you. If you could do 909 00:49:03,320 --> 00:49:05,520 Speaker 1: it yourself, you could be that much more effective if 910 00:49:05,560 --> 00:49:08,839 Speaker 1: you knew the theory and the programming. That's a that's 911 00:49:08,920 --> 00:49:11,920 Speaker 1: quite interesting. So that was leads to what was the 912 00:49:12,000 --> 00:49:15,040 Speaker 1: first model you built at Goldman it was the some 913 00:49:15,960 --> 00:49:19,239 Speaker 1: dormantory model for interest rates. It was a model for 914 00:49:19,320 --> 00:49:22,640 Speaker 1: options on treasury bonds, options on interest rates. Goldman was 915 00:49:22,719 --> 00:49:24,840 Speaker 1: doing a lot of business. People were beginning to extend 916 00:49:25,320 --> 00:49:28,840 Speaker 1: black shoals, which is for options on stock, options on bonds. 917 00:49:29,440 --> 00:49:32,239 Speaker 1: And it's actually quite a complicated problem to do consistently. 918 00:49:32,320 --> 00:49:35,600 Speaker 1: And I can imagine, yeah, because bonds bonds are all 919 00:49:35,760 --> 00:49:38,680 Speaker 1: you can pretend apple and Apple and Walmart have nothing 920 00:49:38,719 --> 00:49:40,680 Speaker 1: to do with each other and write an option model 921 00:49:40,719 --> 00:49:42,919 Speaker 1: for each. But you can't pretend to five year bond 922 00:49:43,000 --> 00:49:44,360 Speaker 1: and a three year bond have nothing to do with 923 00:49:44,440 --> 00:49:46,400 Speaker 1: each other because the five year bond will become a 924 00:49:46,440 --> 00:49:49,439 Speaker 1: three year bond. Yeah, And so you have to really 925 00:49:49,719 --> 00:49:53,520 Speaker 1: many more constraints on building the model. So what is 926 00:49:53,600 --> 00:49:57,840 Speaker 1: it that people misunderstand about quants and model in finance? 927 00:49:57,960 --> 00:50:01,520 Speaker 1: The lay person has an idea year about a person 928 00:50:01,600 --> 00:50:04,400 Speaker 1: sitting in front of a computer, But what do the 929 00:50:04,520 --> 00:50:08,279 Speaker 1: average what does the average person not know about what's 930 00:50:08,320 --> 00:50:13,719 Speaker 1: happening behind the scenes that is telling us to as 931 00:50:13,800 --> 00:50:19,080 Speaker 1: to just the perspective of plants in finance. You know, 932 00:50:19,600 --> 00:50:22,759 Speaker 1: I think people who academics or other people haven't worked 933 00:50:22,800 --> 00:50:25,319 Speaker 1: with trading desks. You imagine that you're making predictions all 934 00:50:25,360 --> 00:50:28,320 Speaker 1: the time, and I think the truth is, most of 935 00:50:28,320 --> 00:50:31,760 Speaker 1: the time, certainly derivatives, you're not making predictions about the future. 936 00:50:32,719 --> 00:50:35,320 Speaker 1: You're trying to figure out in the present what's costs 937 00:50:35,360 --> 00:50:37,640 Speaker 1: too much and what what's too rich and what's too dear. 938 00:50:38,080 --> 00:50:40,440 Speaker 1: And your models are much more good to saying this 939 00:50:40,680 --> 00:50:42,480 Speaker 1: is more expensive than it should be, and this is 940 00:50:42,760 --> 00:50:46,800 Speaker 1: richer than it should be, And then the prediction is, Okay, 941 00:50:46,840 --> 00:50:48,640 Speaker 1: the rich things will become cheaper and the cheap things 942 00:50:48,680 --> 00:50:51,399 Speaker 1: will become richer. But most of your time is spent 943 00:50:51,520 --> 00:50:53,520 Speaker 1: trying to at least most of my time, we've spent 944 00:50:53,600 --> 00:50:57,120 Speaker 1: trying to figure out how to tell the price of 945 00:50:57,239 --> 00:51:01,000 Speaker 1: something liquid from a lot of liquid things. So, for example, 946 00:51:01,040 --> 00:51:03,680 Speaker 1: if you want to buy an option, how you figure 947 00:51:03,680 --> 00:51:05,799 Speaker 1: out the option price from the stock on the bond price, 948 00:51:05,880 --> 00:51:08,840 Speaker 1: which are both liquid and so the example I like 949 00:51:08,960 --> 00:51:11,080 Speaker 1: to give in a way it's also a metaphor, is 950 00:51:11,160 --> 00:51:13,640 Speaker 1: if a lot of the problems you faced with is 951 00:51:13,719 --> 00:51:15,640 Speaker 1: not what will happen to the price of apples in 952 00:51:15,680 --> 00:51:18,560 Speaker 1: the future, if you're dealing with with with fruit, but 953 00:51:19,040 --> 00:51:21,080 Speaker 1: what should a pay for fruit salad? Given the price 954 00:51:21,160 --> 00:51:25,279 Speaker 1: of apples, pears and peaches, and and so you say, 955 00:51:25,400 --> 00:51:27,359 Speaker 1: what's the cost of canning? What's the cost of buying 956 00:51:27,360 --> 00:51:29,359 Speaker 1: the apples, pears and peaches, unless should be the price 957 00:51:29,440 --> 00:51:31,920 Speaker 1: of fruit salad. And then sometimes you do the reverse, 958 00:51:32,000 --> 00:51:34,000 Speaker 1: you say, I know the price of fruit salad, I 959 00:51:34,120 --> 00:51:36,400 Speaker 1: know the price of apples and oranges, so what's the 960 00:51:36,520 --> 00:51:39,479 Speaker 1: right price for pears. But it's always trying to figure 961 00:51:39,480 --> 00:51:42,319 Speaker 1: out in the present how to get the liquid thing, 962 00:51:42,560 --> 00:51:44,400 Speaker 1: given the price of the liquid things, how to get 963 00:51:44,440 --> 00:51:47,759 Speaker 1: the value of the liquid things. Am I making sense? Yeah? No, 964 00:51:47,840 --> 00:51:51,080 Speaker 1: that's that's quite fascinating. You're absolutely less prediction than trying 965 00:51:51,120 --> 00:51:53,800 Speaker 1: to figure out what the current price of something should be. 966 00:51:54,000 --> 00:51:57,200 Speaker 1: And very often there's incomplete information. If if something is 967 00:51:57,239 --> 00:51:59,160 Speaker 1: liquid and trading, well, when then we know what the 968 00:51:59,239 --> 00:52:02,839 Speaker 1: price is. But if something is illiquid doesn't trade a lot, 969 00:52:03,280 --> 00:52:06,279 Speaker 1: you really don't have a market based price, so you 970 00:52:06,360 --> 00:52:09,680 Speaker 1: have to come up with a different way to figure 971 00:52:09,719 --> 00:52:11,680 Speaker 1: out what you should or shouldn't pay for this. Is 972 00:52:11,960 --> 00:52:15,160 Speaker 1: that a fair yes? And the prediction comes less in 973 00:52:15,280 --> 00:52:18,120 Speaker 1: predicting what will happen, then in saying this is cheap. 974 00:52:18,200 --> 00:52:20,719 Speaker 1: And eventually I think my model says this is cheap, 975 00:52:20,760 --> 00:52:22,839 Speaker 1: and so eventually it will come to what I think 976 00:52:22,960 --> 00:52:25,920 Speaker 1: is fair of value. Everything comes back down to reversion 977 00:52:25,960 --> 00:52:33,680 Speaker 1: to the means that's fair, I think. Um, So what 978 00:52:33,880 --> 00:52:37,319 Speaker 1: other quants do you admire? What? What quants have moved 979 00:52:37,400 --> 00:52:41,440 Speaker 1: the industry forward? I think Fisher Black most of all. 980 00:52:41,480 --> 00:52:43,040 Speaker 1: I wrote a lot about him in the book that 981 00:52:43,080 --> 00:52:45,799 Speaker 1: I wrote in my memoir My Life is a quant 982 00:52:45,920 --> 00:52:48,399 Speaker 1: because I think he was kind of an exceptional guy, 983 00:52:48,560 --> 00:52:51,200 Speaker 1: both from a character difficult guy, but exceptional from a 984 00:52:51,280 --> 00:52:53,760 Speaker 1: character point of view, and that he liked to tackle 985 00:52:53,840 --> 00:52:57,160 Speaker 1: everything ab initio. You know, he would from scratch, from 986 00:52:57,280 --> 00:52:59,719 Speaker 1: right from start, right from the start, somebody who was 987 00:52:59,760 --> 00:53:01,520 Speaker 1: totally unknown could come to him and send him an 988 00:53:01,560 --> 00:53:03,560 Speaker 1: email about something, and he would think about it, you know, 989 00:53:03,640 --> 00:53:06,560 Speaker 1: without any prejudice. And he thought about everything in his 990 00:53:06,640 --> 00:53:09,440 Speaker 1: own way. So I kind of admire him. Um. I 991 00:53:09,560 --> 00:53:13,080 Speaker 1: kind of admired Steve Ross and Mark Rubinstein. They were 992 00:53:13,120 --> 00:53:16,960 Speaker 1: more PhD academics, but for for their setting out the 993 00:53:17,000 --> 00:53:19,680 Speaker 1: whole basis of them of option pricing, which I spent 994 00:53:19,800 --> 00:53:22,280 Speaker 1: most of my life on. I kind of admired Paul Wilmot, 995 00:53:22,360 --> 00:53:25,319 Speaker 1: who I once wrote a paper with. We actually wrote 996 00:53:25,520 --> 00:53:29,120 Speaker 1: based on the based on the on the financial crisis, 997 00:53:29,200 --> 00:53:31,520 Speaker 1: we wrote a financial model as manifesto. It was a 998 00:53:31,560 --> 00:53:33,960 Speaker 1: bit of a joke, I recall. I recall that now 999 00:53:34,080 --> 00:53:37,759 Speaker 1: for people who may not know. Wilmot puts out a 1000 00:53:37,920 --> 00:53:45,400 Speaker 1: Quantitative quarterly Quantitative magazine magazine on really developments Paul Wilmot, right, 1001 00:53:45,480 --> 00:53:49,400 Speaker 1: developments in in the world of quantitative trading and analysis, 1002 00:53:50,040 --> 00:53:53,960 Speaker 1: and it's really high level stuff when when you sift 1003 00:53:54,080 --> 00:53:58,560 Speaker 1: through it um as I on occasion have done, most 1004 00:53:58,640 --> 00:54:00,520 Speaker 1: of it's going to be over the head of the 1005 00:54:00,600 --> 00:54:04,439 Speaker 1: average trader, the average investor, the average person. But within 1006 00:54:04,560 --> 00:54:08,880 Speaker 1: the industry, I have to think that that's become practically 1007 00:54:09,040 --> 00:54:12,000 Speaker 1: the go to bible. Is that. Am I overstating it? 1008 00:54:12,200 --> 00:54:13,840 Speaker 1: Maybe a little bit, but that's one of them. But 1009 00:54:13,920 --> 00:54:16,239 Speaker 1: I sort of admiring because he's written some I mean 1010 00:54:16,400 --> 00:54:18,440 Speaker 1: for that too, but he's written some good textbooks, and 1011 00:54:19,040 --> 00:54:21,279 Speaker 1: I like to think, like me, he tries to make 1012 00:54:21,360 --> 00:54:24,200 Speaker 1: things simpler rather than complexify them. There's been a tendency 1013 00:54:24,680 --> 00:54:27,439 Speaker 1: for finance academics to do everything in a very formal, 1014 00:54:27,520 --> 00:54:30,480 Speaker 1: axiomatic way as of the teaching math, and Paul and 1015 00:54:30,600 --> 00:54:33,520 Speaker 1: myself both like to use the least amount of math 1016 00:54:33,640 --> 00:54:38,160 Speaker 1: possible that that's quite interesting. So let's keep plowing through 1017 00:54:38,239 --> 00:54:41,719 Speaker 1: some of some of these questions here. Um, we talked 1018 00:54:41,760 --> 00:54:45,960 Speaker 1: about the ft, we talked about alpha. What's been the 1019 00:54:46,040 --> 00:54:50,000 Speaker 1: impact of of high frequency trading and al go driven 1020 00:54:50,239 --> 00:54:53,680 Speaker 1: trading strategies on the sort of work that that you do? 1021 00:54:55,080 --> 00:54:57,719 Speaker 1: Very broad if if I talk from a sociological point 1022 00:54:57,760 --> 00:54:59,640 Speaker 1: of view, ill see from the students at Columbia until 1023 00:54:59,680 --> 00:55:02,600 Speaker 1: a few years ago, and they all wanted to take derivatives, 1024 00:55:02,640 --> 00:55:05,600 Speaker 1: and the last year or two everybody's being tempted probably 1025 00:55:05,600 --> 00:55:07,800 Speaker 1: about the job market and by excitement to work on 1026 00:55:08,200 --> 00:55:11,000 Speaker 1: high frequency trading, algorithmic trading. And now for the last 1027 00:55:11,080 --> 00:55:15,360 Speaker 1: year or two machine learning. You know, you're programming machines 1028 00:55:15,480 --> 00:55:18,359 Speaker 1: to to to just grow through all the data that's 1029 00:55:18,360 --> 00:55:20,960 Speaker 1: out there and figure out their own rules. Yes, or 1030 00:55:21,080 --> 00:55:23,800 Speaker 1: even even things like can you scour the internet to 1031 00:55:23,880 --> 00:55:26,279 Speaker 1: find out whether there were a lot of cars in 1032 00:55:26,360 --> 00:55:28,960 Speaker 1: Walmart parking lots the last year, you know, to try 1033 00:55:29,000 --> 00:55:32,640 Speaker 1: to get them get source advice? There? There have been 1034 00:55:32,719 --> 00:55:35,720 Speaker 1: people doing that with Twitter. Can we identify sentiment shifts 1035 00:55:35,760 --> 00:55:41,120 Speaker 1: on Twitter and then generate a tradeable algorithm from that? Yeah, 1036 00:55:41,280 --> 00:55:42,799 Speaker 1: I've noticed that, So a lot of that, A lot 1037 00:55:42,840 --> 00:55:45,880 Speaker 1: of that stuff, some sentiment based, some really objective, like 1038 00:55:46,040 --> 00:55:48,360 Speaker 1: can you figure out in some way how much sales 1039 00:55:48,400 --> 00:55:51,280 Speaker 1: are happening in various places from data you can collect 1040 00:55:51,360 --> 00:55:54,480 Speaker 1: on the internet. Um, what was your question again, us, 1041 00:55:54,560 --> 00:55:58,319 Speaker 1: I have no idea, so I have no recollection. Um, 1042 00:55:59,360 --> 00:56:03,360 Speaker 1: what's the pact of high frequency trading on model development 1043 00:56:04,080 --> 00:56:08,280 Speaker 1: and and model construction and then actually having these models 1044 00:56:09,000 --> 00:56:14,520 Speaker 1: execute in the market if there are theoretically predatory algorithms 1045 00:56:14,800 --> 00:56:17,680 Speaker 1: looking for whatever it is they're trying to do. Yeah, 1046 00:56:17,719 --> 00:56:19,840 Speaker 1: I think if I get a little bit general that 1047 00:56:19,920 --> 00:56:22,319 Speaker 1: there've been sort of two classes of models that people 1048 00:56:22,400 --> 00:56:26,279 Speaker 1: use in finance. One or structural models, where like for derivatives, 1049 00:56:26,320 --> 00:56:28,680 Speaker 1: where you say an option is really a hybrid of 1050 00:56:28,760 --> 00:56:30,600 Speaker 1: a stock in a bond, and I've got to figure 1051 00:56:30,600 --> 00:56:32,920 Speaker 1: out exactly how it works, And that's what black shouls does. 1052 00:56:33,320 --> 00:56:35,560 Speaker 1: So it's really saying like an option is like a 1053 00:56:35,640 --> 00:56:37,839 Speaker 1: molecule made out of atoms and there's a structure there. 1054 00:56:38,320 --> 00:56:41,080 Speaker 1: And those were always what quantum mostly did. But what's happened, 1055 00:56:41,080 --> 00:56:42,800 Speaker 1: as you point out, in the last few years is 1056 00:56:43,080 --> 00:56:46,279 Speaker 1: econometrical statistical models have become much more the rage and 1057 00:56:46,360 --> 00:56:49,160 Speaker 1: students are gaining that direction where you just say, cannot 1058 00:56:49,200 --> 00:56:53,200 Speaker 1: find a regression between um, you know, various factors that 1059 00:56:53,440 --> 00:56:56,239 Speaker 1: that seem to predict the market. I don't care. I 1060 00:56:56,320 --> 00:56:58,120 Speaker 1: care a little bit why, but I care less. Why 1061 00:56:58,200 --> 00:57:01,800 Speaker 1: then finding the actual pattern? And that's become so statistics 1062 00:57:01,840 --> 00:57:05,040 Speaker 1: an econometrics have become much more fashionable as a result 1063 00:57:05,080 --> 00:57:09,680 Speaker 1: of high frequency trading and algorithmic trading. That's interesting why econometrics? 1064 00:57:09,760 --> 00:57:12,640 Speaker 1: How how did that work its way into the sort 1065 00:57:12,680 --> 00:57:17,120 Speaker 1: of modeling. I'm looking for relations between time series, between 1066 00:57:18,000 --> 00:57:22,440 Speaker 1: FED behavior, between interest rates, between UM the behavior effectors 1067 00:57:22,520 --> 00:57:26,120 Speaker 1: and and particular stocks. So let's let's caring about why 1068 00:57:26,160 --> 00:57:28,640 Speaker 1: it's happening, the detecting a pattern. So I see a 1069 00:57:28,720 --> 00:57:32,440 Speaker 1: lot of these sort of correlations, and they always make 1070 00:57:32,600 --> 00:57:34,760 Speaker 1: me and I see people writing about them, and I 1071 00:57:34,800 --> 00:57:37,440 Speaker 1: read articles and I see charts. But the back of 1072 00:57:37,480 --> 00:57:41,120 Speaker 1: my head, I'm always saying, well, yes, but is one 1073 00:57:41,200 --> 00:57:45,160 Speaker 1: thing causing another? Is this a temporary you know? Sometimes 1074 00:57:45,200 --> 00:57:47,880 Speaker 1: they're in sync, and so I can't tell. I can't 1075 00:57:47,960 --> 00:57:50,680 Speaker 1: count how many people have been insisting, well, look the 1076 00:57:50,760 --> 00:57:52,720 Speaker 1: FED cut rates and here's what the market did, and 1077 00:57:52,800 --> 00:57:55,080 Speaker 1: now we have quantitative easing, and here's what the market did. 1078 00:57:55,360 --> 00:57:57,280 Speaker 1: And as soon as this unwinds, the opposite is going 1079 00:57:57,320 --> 00:58:00,320 Speaker 1: to happen. And they run in parallel for long periods 1080 00:58:00,320 --> 00:58:03,120 Speaker 1: of time and then suddenly they just go their own way. 1081 00:58:03,520 --> 00:58:06,520 Speaker 1: How much of a risk is that on the econometric side, 1082 00:58:06,640 --> 00:58:11,120 Speaker 1: that we're gonna be fooled by randomness. We're gonna see 1083 00:58:11,160 --> 00:58:14,400 Speaker 1: a correlation and it works until it stops working. I 1084 00:58:14,720 --> 00:58:17,000 Speaker 1: agree with you totally. I'm I mean, I understand why 1085 00:58:17,080 --> 00:58:18,920 Speaker 1: people do this stuff, but I'm a bit of a 1086 00:58:18,960 --> 00:58:22,439 Speaker 1: skeptic about it myself. I think, UM, just what you said, 1087 00:58:22,760 --> 00:58:25,240 Speaker 1: there's a standard phrase which you are sort of paralleling, 1088 00:58:25,360 --> 00:58:28,800 Speaker 1: saying correlation is not causation, and I think that's true. 1089 00:58:28,920 --> 00:58:31,640 Speaker 1: But but people are so keen to make money for 1090 00:58:31,800 --> 00:58:34,640 Speaker 1: legitimate reasons that that's the easiest thing to do, and 1091 00:58:34,760 --> 00:58:37,680 Speaker 1: sometimes it works. All right, I have to ask you 1092 00:58:37,800 --> 00:58:41,600 Speaker 1: before I get to some of my favorite UM questions. 1093 00:58:41,640 --> 00:58:45,120 Speaker 1: I have to ask you about Spinosa's theory of emotions, 1094 00:58:45,680 --> 00:58:50,480 Speaker 1: which you mentioned in UM models behavior behaving badly, moral 1095 00:58:50,560 --> 00:58:53,800 Speaker 1: concepts of good and evil, virtue, and perspective have a 1096 00:58:53,960 --> 00:58:58,000 Speaker 1: basis in human psychology, which naturally leads to the question, 1097 00:58:58,560 --> 00:59:02,680 Speaker 1: what does this have to do with trading models. Well, 1098 00:59:03,040 --> 00:59:05,320 Speaker 1: Spinoza was like three years ahead of his time. I 1099 00:59:05,400 --> 00:59:07,479 Speaker 1: tried to read The Ethics once when I was writing 1100 00:59:07,560 --> 00:59:10,160 Speaker 1: my book because I was trying to find an example 1101 00:59:10,240 --> 00:59:12,240 Speaker 1: of something that I thought was the theory rather than 1102 00:59:12,320 --> 00:59:15,840 Speaker 1: a model. And what Spinoza did in the Ethics, so 1103 00:59:15,880 --> 00:59:20,160 Speaker 1: that it was only published posthumously, was trying to figure 1104 00:59:20,200 --> 00:59:23,960 Speaker 1: out actually very a lot like behavioral economists, but three 1105 00:59:24,360 --> 00:59:26,840 Speaker 1: years earlier, trying to figure out what drives people to 1106 00:59:26,920 --> 00:59:30,560 Speaker 1: behave the way they do. And his argument was that, um, 1107 00:59:31,160 --> 00:59:34,080 Speaker 1: if you can understand the passions of the emotions, as 1108 00:59:34,120 --> 00:59:36,600 Speaker 1: he calls them, then he'll be able to understand why 1109 00:59:36,640 --> 00:59:39,040 Speaker 1: people are unhappy and then be able to figure out 1110 00:59:39,040 --> 00:59:40,960 Speaker 1: how they should live their lives. And so what he 1111 00:59:41,000 --> 00:59:46,080 Speaker 1: starts out doing is analyzing all the emotions or as 1112 00:59:46,120 --> 00:59:47,720 Speaker 1: he called with the passions that they are. And he 1113 00:59:47,760 --> 00:59:51,000 Speaker 1: calls him passions because they affect you as a passive 1114 00:59:51,080 --> 00:59:53,320 Speaker 1: person rather than an active person. They sweep over you 1115 00:59:53,480 --> 00:59:55,400 Speaker 1: rather than you deciding you would you would want to 1116 00:59:56,040 --> 00:59:59,440 Speaker 1: want to behave that way. And it's actually so it's 1117 00:59:59,520 --> 01:00:02,920 Speaker 1: very interesting and be it's really closely related to derivatives. 1118 01:00:02,960 --> 01:00:05,320 Speaker 1: It's sort of astonishing because what he does is he says, 1119 01:00:06,000 --> 01:00:08,400 Speaker 1: at the bottom of the chain, there are only three things, 1120 01:00:08,480 --> 01:00:13,120 Speaker 1: which is pleasure, pain, and desire. And everybody knows what pleasure, 1121 01:00:13,160 --> 01:00:16,080 Speaker 1: pain and desire, probably actually defines it in some philosophical way. 1122 01:00:16,600 --> 01:00:20,480 Speaker 1: And then he defines every other emotion, and there are 1123 01:00:20,800 --> 01:00:24,000 Speaker 1: hundreds of them verbally in terms of how they reduced 1124 01:00:24,040 --> 01:00:27,919 Speaker 1: to pleasure, pain, and desire. So, for example, um, love 1125 01:00:28,080 --> 01:00:31,200 Speaker 1: is expectation of pleasure from some other person, and pain 1126 01:00:31,360 --> 01:00:34,040 Speaker 1: is hatred is the opposite expectation of pain. That's a 1127 01:00:34,040 --> 01:00:38,360 Speaker 1: little bit corny. But then for example, he says, um um, 1128 01:00:39,200 --> 01:00:41,480 Speaker 1: so envy, Yeah, envy is the one I wanted. I 1129 01:00:41,520 --> 01:00:45,400 Speaker 1: couldn't think, so he says, envy, Um, I have to 1130 01:00:45,440 --> 01:00:48,200 Speaker 1: try to remember the pain of absence of pleasure. Yeah, 1131 01:00:48,400 --> 01:00:51,040 Speaker 1: And and envy is um wow, it's suddenly slipping me. 1132 01:00:51,480 --> 01:00:56,080 Speaker 1: I remember cruelty better. Cruelty. Cruelty is what you call 1133 01:00:56,520 --> 01:00:59,360 Speaker 1: the desire of somebody else to inflict pain on someone 1134 01:00:59,440 --> 01:01:03,720 Speaker 1: that you love. So cruelty eventually goes down to pleasure, pain, 1135 01:01:03,800 --> 01:01:06,400 Speaker 1: and desire. And I like to say it's like a 1136 01:01:08,000 --> 01:01:10,480 Speaker 1: it's um it's like a convertible bond that has both 1137 01:01:11,080 --> 01:01:14,640 Speaker 1: equity exposure, credit exposure, and fixed income exposure. It goes 1138 01:01:14,680 --> 01:01:16,560 Speaker 1: down to all three of the derivatives. He's really got 1139 01:01:16,680 --> 01:01:19,000 Speaker 1: to actually do a chart, which is on a giant 1140 01:01:19,080 --> 01:01:25,840 Speaker 1: then diagram pleasure, pain and eventually goes down through some 1141 01:01:26,040 --> 01:01:29,080 Speaker 1: chain to these three things at the bottom. I wonder 1142 01:01:29,120 --> 01:01:31,000 Speaker 1: if there's a graphic of that somewhere. I've got a 1143 01:01:31,040 --> 01:01:33,280 Speaker 1: graphic of it. Yeah, it's on my website, and I 1144 01:01:33,960 --> 01:01:35,320 Speaker 1: was only going to make a post out of it 1145 01:01:35,360 --> 01:01:37,680 Speaker 1: because it's really in fact, I once I once submitted 1146 01:01:37,720 --> 01:01:41,560 Speaker 1: that somebody, somebody at the Serpentine Gallery in London had 1147 01:01:41,600 --> 01:01:44,160 Speaker 1: a map competition that somebody connected me to and I 1148 01:01:44,240 --> 01:01:46,600 Speaker 1: submitted that as a map, and it's in a book 1149 01:01:46,680 --> 01:01:52,840 Speaker 1: somewhere maps of all kinds of things. That's fascinating. Um, 1150 01:01:52,960 --> 01:01:56,480 Speaker 1: So now let's move over to some of my favorite questions. 1151 01:01:56,920 --> 01:02:01,720 Speaker 1: These are the standard questions we ask everybody. So we 1152 01:02:01,840 --> 01:02:04,840 Speaker 1: went over how you how you got into the finance. Well, 1153 01:02:04,880 --> 01:02:08,040 Speaker 1: actually we really didn't get that question. So you started 1154 01:02:08,760 --> 01:02:13,080 Speaker 1: as a physicist, how do you go from bould of 1155 01:02:13,200 --> 01:02:18,080 Speaker 1: Colorado studying particle physics? Too? I'm gonna do quantitative work 1156 01:02:18,320 --> 01:02:21,320 Speaker 1: on Wall Street. How did that transition actually? You come 1157 01:02:23,320 --> 01:02:27,160 Speaker 1: kind of um, not planned. I I got a PhD 1158 01:02:27,280 --> 01:02:30,040 Speaker 1: in nineteen seventy three from Colombia. And then, as I said, 1159 01:02:30,120 --> 01:02:32,040 Speaker 1: jobs were hard to find, and I had a post 1160 01:02:32,120 --> 01:02:34,760 Speaker 1: stock at University of Pennsylvania, and then I had a 1161 01:02:34,840 --> 01:02:37,040 Speaker 1: job at Oxford in England. And then I had a 1162 01:02:37,120 --> 01:02:39,320 Speaker 1: job at Rockefeller University of New York. And my wife 1163 01:02:39,440 --> 01:02:41,360 Speaker 1: was in biology, and we were moving all over the 1164 01:02:41,400 --> 01:02:44,000 Speaker 1: world out of sync with each other actually, and then 1165 01:02:44,040 --> 01:02:45,920 Speaker 1: when we were in England together we had a kid, 1166 01:02:46,560 --> 01:02:48,680 Speaker 1: and and then we both had jobs in New York. 1167 01:02:48,920 --> 01:02:51,160 Speaker 1: And then it was very hard to get permanent jobs. 1168 01:02:51,200 --> 01:02:52,920 Speaker 1: I got a job in Boulder. She couldn't get a 1169 01:02:53,000 --> 01:02:55,560 Speaker 1: job there. I lived there for a while, um, and 1170 01:02:55,640 --> 01:02:57,280 Speaker 1: eventually I sort of threw in the towel. It was 1171 01:02:57,320 --> 01:03:01,439 Speaker 1: getting too complicated. And then this is nineteen eighty where 1172 01:03:01,520 --> 01:03:04,600 Speaker 1: people went then was there was the oil crisis and 1173 01:03:04,920 --> 01:03:07,440 Speaker 1: telecom was building up, and you have the people who 1174 01:03:07,520 --> 01:03:10,800 Speaker 1: were physicists either went to work for Exxon or Mobile, 1175 01:03:11,400 --> 01:03:14,120 Speaker 1: or they went to Bell Labs. And I took a 1176 01:03:14,200 --> 01:03:16,680 Speaker 1: job at Bell Labs later became loosen. You were a 1177 01:03:16,720 --> 01:03:18,600 Speaker 1: good couple of years. I was there for five years 1178 01:03:18,960 --> 01:03:22,440 Speaker 1: and and so then I stopped doing physics and I 1179 01:03:22,560 --> 01:03:25,680 Speaker 1: worked in a business analysis center that sort of was 1180 01:03:25,720 --> 01:03:29,240 Speaker 1: applying physics techniques or mostly programming to A T and 1181 01:03:29,320 --> 01:03:32,320 Speaker 1: T business problems. That's funny because when I saw your 1182 01:03:32,360 --> 01:03:35,560 Speaker 1: background and you were at at Bell Labs at Loosen, 1183 01:03:36,080 --> 01:03:39,920 Speaker 1: I assumed you were doing some fine physics work as 1184 01:03:39,960 --> 01:03:43,400 Speaker 1: an telecom I never for a second said, oh, well, 1185 01:03:43,480 --> 01:03:45,960 Speaker 1: he must be working in the in the finances there were. 1186 01:03:46,320 --> 01:03:48,120 Speaker 1: I was a sort of retread when I went there, 1187 01:03:48,280 --> 01:03:52,360 Speaker 1: and I'm aladd of that lead to Goldman Sachs. Well, 1188 01:03:52,880 --> 01:03:55,560 Speaker 1: that was when I went there and I started working 1189 01:03:55,640 --> 01:03:57,920 Speaker 1: on actually much more computer program I learned to be 1190 01:03:58,000 --> 01:04:00,720 Speaker 1: a good program at the time, and then we were 1191 01:04:00,760 --> 01:04:03,560 Speaker 1: tackling sort of problems that A T and T had, 1192 01:04:04,520 --> 01:04:09,320 Speaker 1: where you could use sort of computer modeling and financial modeling. UM. 1193 01:04:09,640 --> 01:04:11,240 Speaker 1: I didn't like it very much. I learned a lot 1194 01:04:11,280 --> 01:04:13,280 Speaker 1: of useful stuff, but I always really hanchred to go 1195 01:04:13,400 --> 01:04:15,880 Speaker 1: back to a more academic job. But that was two 1196 01:04:15,920 --> 01:04:18,760 Speaker 1: difficult and then all the head and just started knocking 1197 01:04:18,800 --> 01:04:21,080 Speaker 1: on the doors of people because interest rates were high, 1198 01:04:21,160 --> 01:04:23,800 Speaker 1: and Solomon and Goldman in these places were hiring people. 1199 01:04:24,640 --> 01:04:26,480 Speaker 1: And it took me a few years to adjust the idea, 1200 01:04:26,560 --> 01:04:29,960 Speaker 1: but I decided I would I would take a leap 1201 01:04:30,080 --> 01:04:33,160 Speaker 1: out of out of the whole sphere and into the 1202 01:04:33,200 --> 01:04:34,800 Speaker 1: business world. It was sort of a shock for me. 1203 01:04:34,840 --> 01:04:37,040 Speaker 1: I never expected to ever do that, and it worked 1204 01:04:37,080 --> 01:04:39,360 Speaker 1: out pretty well. Yeah, I got very excited, as I say, 1205 01:04:39,400 --> 01:04:40,960 Speaker 1: when I went to Golden and eighty five. It was 1206 01:04:41,040 --> 01:04:43,320 Speaker 1: like a shot in the arm for me. And so 1207 01:04:43,480 --> 01:04:48,440 Speaker 1: and you were there through a brief period. I was 1208 01:04:48,480 --> 01:04:50,920 Speaker 1: at Solomon for one year a little bit more, and 1209 01:04:51,120 --> 01:04:53,440 Speaker 1: then in mortgages, and then I went back to Goldman. 1210 01:04:53,480 --> 01:04:56,240 Speaker 1: I started in fixing income. I went to Solomon mortgages, 1211 01:04:56,320 --> 01:04:58,040 Speaker 1: and then I came back and I was in equity 1212 01:04:58,040 --> 01:05:00,360 Speaker 1: to Routers for like ten years, which was really my favorite. 1213 01:05:00,640 --> 01:05:04,400 Speaker 1: And it sounds like you thrived and did really well there. 1214 01:05:04,480 --> 01:05:08,120 Speaker 1: So let's talk about mentors. Who are some of your 1215 01:05:08,240 --> 01:05:13,120 Speaker 1: early mentors? Oh, in physics are in well both? Yeah, 1216 01:05:13,800 --> 01:05:16,120 Speaker 1: in physics, there were a couple of professors in South 1217 01:05:16,160 --> 01:05:18,280 Speaker 1: Africa that I that I worked for. There was a 1218 01:05:18,280 --> 01:05:21,400 Speaker 1: guy called Professor Whiteman who sort of tutored me. It's 1219 01:05:21,480 --> 01:05:25,680 Speaker 1: long dead, probably, um And how about in in the 1220 01:05:25,760 --> 01:05:28,560 Speaker 1: finance and finance I would say Fisher black the most 1221 01:05:28,640 --> 01:05:31,680 Speaker 1: in that I got there and there was still not 1222 01:05:31,720 --> 01:05:35,840 Speaker 1: a lot of concert Goldman and and I got involved 1223 01:05:35,880 --> 01:05:38,560 Speaker 1: immediately in working for the bond options trading desk, and 1224 01:05:38,600 --> 01:05:40,800 Speaker 1: they connected me with Fisher because he was the expert 1225 01:05:41,440 --> 01:05:43,720 Speaker 1: and m he really had a very big impact on me. 1226 01:05:43,840 --> 01:05:45,320 Speaker 1: So I would say I don't think he I don't 1227 01:05:45,320 --> 01:05:46,920 Speaker 1: think he set out to mentor me. He was kind 1228 01:05:46,920 --> 01:05:49,560 Speaker 1: of a bit of a cold fish in an ice way. 1229 01:05:50,720 --> 01:05:53,080 Speaker 1: But I really learned a lot a lot from him, 1230 01:05:53,520 --> 01:05:57,480 Speaker 1: both about perseverance and about not taking models too seriously. 1231 01:05:58,120 --> 01:06:03,320 Speaker 1: Quite quite interesting. Um, what investors influenced the way you 1232 01:06:03,480 --> 01:06:09,120 Speaker 1: think about modeling or investment? Um, that's a good question. 1233 01:06:09,320 --> 01:06:11,600 Speaker 1: I to be honest, I don't invest that much. I'm 1234 01:06:11,680 --> 01:06:14,280 Speaker 1: a I'm a E T F mutual fund guy. In 1235 01:06:14,320 --> 01:06:17,240 Speaker 1: the old days E d F Now UM. For a 1236 01:06:17,280 --> 01:06:19,320 Speaker 1: long time when I worked at Goldman and I recently 1237 01:06:19,400 --> 01:06:21,280 Speaker 1: worked at a fund of funds, I wasn't allowed to 1238 01:06:21,320 --> 01:06:25,560 Speaker 1: buy individual stocks. Um. So so I'm so you're an 1239 01:06:25,600 --> 01:06:28,560 Speaker 1: index sir. I'm an index a pretty much. And by 1240 01:06:28,560 --> 01:06:32,480 Speaker 1: the way, that is not uncommon amongst academics who say 1241 01:06:32,840 --> 01:06:34,840 Speaker 1: I don't have the interest, the time, the effort. I 1242 01:06:34,840 --> 01:06:37,960 Speaker 1: don't want to babysit a bunch of stocks on the 1243 01:06:38,040 --> 01:06:40,840 Speaker 1: possibility about performing. Let me just go. I'm kind of 1244 01:06:41,040 --> 01:06:42,920 Speaker 1: I'm kind of like that. I like it intellectually, but 1245 01:06:43,040 --> 01:06:45,480 Speaker 1: I don't and I like following the markets. But um, 1246 01:06:45,600 --> 01:06:47,680 Speaker 1: I have at times board options years ago when I 1247 01:06:47,760 --> 01:06:51,680 Speaker 1: was allowed to, but I don't really do that anymore. Uh, 1248 01:06:51,920 --> 01:06:56,080 Speaker 1: let's talk about books. Always like Peter Lynch, interesting guy, right, 1249 01:06:56,240 --> 01:07:01,160 Speaker 1: fascinating and I was fascinated. But yeah, a quick digression. 1250 01:07:01,440 --> 01:07:05,680 Speaker 1: I always thought Peter Lynch and the idea that when 1251 01:07:05,720 --> 01:07:08,440 Speaker 1: you're out looking at things on your own and you 1252 01:07:08,720 --> 01:07:13,120 Speaker 1: discover stuff. I don't know if that still exists anymore. Yeah, 1253 01:07:13,120 --> 01:07:15,120 Speaker 1: I kind of like that a few times. I used 1254 01:07:15,120 --> 01:07:18,680 Speaker 1: to do that. I remember noticing with my kids years 1255 01:07:18,720 --> 01:07:21,120 Speaker 1: ago that all of the all of all of my 1256 01:07:21,240 --> 01:07:23,880 Speaker 1: kids friends mothers were wearing rebox shoes before I've ever 1257 01:07:23,920 --> 01:07:26,320 Speaker 1: heard of Reebok when they were doing aerobics, and they 1258 01:07:26,360 --> 01:07:29,360 Speaker 1: were all drinking clearly Canadian at some point. He's all 1259 01:07:29,520 --> 01:07:32,600 Speaker 1: like twenty five years ago. And I remember going into 1260 01:07:32,640 --> 01:07:35,120 Speaker 1: Lulu Lemon a few years ago and being amazed by 1261 01:07:35,160 --> 01:07:38,400 Speaker 1: the good running gear they had and seeing people drink 1262 01:07:38,480 --> 01:07:41,560 Speaker 1: them drink what do you call it? Curry Green mountains. 1263 01:07:41,600 --> 01:07:43,680 Speaker 1: So I still like the idea that you that you 1264 01:07:43,760 --> 01:07:45,640 Speaker 1: spot something that you like, or I was like that 1265 01:07:45,720 --> 01:07:47,840 Speaker 1: about Apple actually, where you just love the product and 1266 01:07:47,920 --> 01:07:50,120 Speaker 1: you say I'm going to go with it without reading 1267 01:07:50,160 --> 01:07:53,000 Speaker 1: about EBIT, orbit, d A or anything like that. The 1268 01:07:53,680 --> 01:08:00,360 Speaker 1: question is given, how sophisticated technology and biotech and and 1269 01:08:01,160 --> 01:08:05,560 Speaker 1: all sorts of things that require an expertise everything you've 1270 01:08:05,640 --> 01:08:11,280 Speaker 1: described as either a consumer product in fact they're so 1271 01:08:11,480 --> 01:08:15,040 Speaker 1: the does the Peter Lynch approach is it's still valid 1272 01:08:15,200 --> 01:08:19,400 Speaker 1: given that so much on the market is not related 1273 01:08:19,439 --> 01:08:22,920 Speaker 1: to retail or consumer spending. It's like, stop and thinking 1274 01:08:22,920 --> 01:08:25,760 Speaker 1: about you're not doing biotech, you're not doing pharma, you're 1275 01:08:25,760 --> 01:08:31,560 Speaker 1: not doing almost all these different software, hardware, networking technology companies. 1276 01:08:32,520 --> 01:08:36,000 Speaker 1: The I wonder if Peter Lynch is of an era, 1277 01:08:36,160 --> 01:08:39,280 Speaker 1: and maybe that era no longer exists. I'm inclined to 1278 01:08:39,320 --> 01:08:42,800 Speaker 1: agree with you. I have friends who went into money 1279 01:08:42,880 --> 01:08:45,599 Speaker 1: management after they left Golden into their own money management firms, 1280 01:08:45,640 --> 01:08:48,680 Speaker 1: and occasionally one of them says to me, well, this 1281 01:08:48,760 --> 01:08:50,559 Speaker 1: has been a really hard time to be a money 1282 01:08:50,560 --> 01:08:53,519 Speaker 1: manager for the last fifteen years, and then I think, well, 1283 01:08:53,960 --> 01:08:56,640 Speaker 1: kind of, what are you getting paid for? Right? But 1284 01:08:57,080 --> 01:08:59,719 Speaker 1: how do you It's hard to do better than other people, 1285 01:08:59,800 --> 01:09:02,680 Speaker 1: and you know, he sort of imagines maybe that the 1286 01:09:02,720 --> 01:09:05,200 Speaker 1: world should keep prices should keep going up, and he 1287 01:09:05,200 --> 01:09:09,880 Speaker 1: should get paid. Well, that's that's you know, if you're 1288 01:09:09,920 --> 01:09:11,560 Speaker 1: just buying a trend, you might as well just do 1289 01:09:11,680 --> 01:09:15,040 Speaker 1: the t f my Peter Lynch experiences. I moved out 1290 01:09:15,080 --> 01:09:18,439 Speaker 1: of the city to the suburbs. It's like fifteen years ago, 1291 01:09:18,600 --> 01:09:21,160 Speaker 1: maybe even a little. It was before nine eleven, so 1292 01:09:21,320 --> 01:09:24,040 Speaker 1: it was just before two thousand and one, and I 1293 01:09:24,240 --> 01:09:27,040 Speaker 1: discovered this new company that nobody had ever heard of 1294 01:09:27,200 --> 01:09:30,880 Speaker 1: called home Depot, and I'm like, wow, this place is amazing. 1295 01:09:30,960 --> 01:09:35,840 Speaker 1: They have everything. A yeah, we're fixing up the house. 1296 01:09:35,960 --> 01:09:38,160 Speaker 1: Is the first thing I do is like I have 1297 01:09:38,400 --> 01:09:41,280 Speaker 1: to buy some of this company. And I punched it 1298 01:09:41,400 --> 01:09:44,479 Speaker 1: up in a on the on the Bloomberg and it's 1299 01:09:44,520 --> 01:09:47,120 Speaker 1: just done nothing but go up for like fifteen years. 1300 01:09:47,520 --> 01:09:50,440 Speaker 1: I'm like the last person on the planet who discovered 1301 01:09:50,560 --> 01:09:53,240 Speaker 1: home Depot. So that was my like, oh, I guess 1302 01:09:53,280 --> 01:09:55,920 Speaker 1: I'm a little late to the Peter Lynch party with 1303 01:09:56,640 --> 01:10:00,519 Speaker 1: with Home with Home Depot. I know you like books, 1304 01:10:00,560 --> 01:10:02,960 Speaker 1: and I know you like philosophers, so I have to ask, 1305 01:10:03,400 --> 01:10:07,120 Speaker 1: what are some of your favorite books? Okay, fiction, nonfiction, 1306 01:10:07,280 --> 01:10:11,320 Speaker 1: finance related, non it doesn't matter. Fiction, I like sort 1307 01:10:11,360 --> 01:10:13,640 Speaker 1: of Um, I like good romantic books. So I like 1308 01:10:13,760 --> 01:10:17,080 Speaker 1: Anna Karenina and I like Madame Bovary. Those are two 1309 01:10:17,120 --> 01:10:18,639 Speaker 1: of my favorite and low leader I have to say, 1310 01:10:18,720 --> 01:10:24,840 Speaker 1: so that's similar theme people people obsessively, obsessively in love. 1311 01:10:25,080 --> 01:10:30,840 Speaker 1: It's always a good story. Um. Nonfiction, Yeah, I like history, 1312 01:10:30,880 --> 01:10:32,800 Speaker 1: although I kind of like I have a hard time 1313 01:10:32,840 --> 01:10:34,679 Speaker 1: suddenly thinking of things, but I sort of like good 1314 01:10:34,800 --> 01:10:37,320 Speaker 1: I read a lot of modern I read mostly a 1315 01:10:37,360 --> 01:10:41,040 Speaker 1: lot of modern fiction. But um, but nonfiction I like. 1316 01:10:41,479 --> 01:10:44,240 Speaker 1: I like some philosophy. I like Schopenhauer. I like him. 1317 01:10:44,760 --> 01:10:46,400 Speaker 1: I like Chopin out the most because he's kind of 1318 01:10:46,439 --> 01:10:50,280 Speaker 1: cynical in a very real several times. So what what's 1319 01:10:50,280 --> 01:10:52,840 Speaker 1: your favorite work of Chopin. There's a collection that I 1320 01:10:52,960 --> 01:10:56,680 Speaker 1: have called him um Essays and Aphorisms. It's a thin 1321 01:10:56,760 --> 01:10:59,280 Speaker 1: Penguin book with a lot of essays about everything from 1322 01:10:59,479 --> 01:11:02,720 Speaker 1: getting old to wisdom to how you should not read 1323 01:11:02,840 --> 01:11:07,000 Speaker 1: until you've until you've exhausted thinking. And I'm very good 1324 01:11:07,080 --> 01:11:09,439 Speaker 1: and very very beautifully written, a little bit like Freud, 1325 01:11:09,600 --> 01:11:11,640 Speaker 1: like you could get a Nobel Prize just for just 1326 01:11:11,800 --> 01:11:15,160 Speaker 1: for his writing style. Really that there's a thin penguin. 1327 01:11:15,160 --> 01:11:16,760 Speaker 1: I've got it for years. But it's called essays and 1328 01:11:16,800 --> 01:11:20,680 Speaker 1: effort and really funny. Oh really yeah, I mean in 1329 01:11:20,760 --> 01:11:22,800 Speaker 1: a in a in a in a slightly bitter sort 1330 01:11:22,840 --> 01:11:25,000 Speaker 1: of way, but very sharp. So you strike me as 1331 01:11:25,000 --> 01:11:28,120 Speaker 1: someone who would read Fineman, who is brilliant and a 1332 01:11:28,240 --> 01:11:32,280 Speaker 1: serbic and funny all at the same time. Yes, I 1333 01:11:32,479 --> 01:11:35,040 Speaker 1: I studied the Fineman lectures and I actually actually shook 1334 01:11:35,080 --> 01:11:37,559 Speaker 1: his hand once in a men's room at a conference 1335 01:11:37,600 --> 01:11:40,080 Speaker 1: in not shook his hand. I spoke to him in 1336 01:11:40,200 --> 01:11:44,920 Speaker 1: in nineteen. But he was a phenomenal guy. I heard 1337 01:11:45,000 --> 01:11:47,960 Speaker 1: him speak several times. The essays is supposed to be 1338 01:11:48,920 --> 01:11:51,600 Speaker 1: the lectures are supposed to be phenomenal. I got a 1339 01:11:51,680 --> 01:11:55,280 Speaker 1: birthday gift of the read books. No, I got someone gave. 1340 01:11:55,400 --> 01:11:57,240 Speaker 1: Actually it was a gift to somebody else that they 1341 01:11:57,320 --> 01:11:59,840 Speaker 1: re gifted to me. It was all the c D 1342 01:12:00,000 --> 01:12:03,439 Speaker 1: ease of the and and they were all none of 1343 01:12:03,520 --> 01:12:06,040 Speaker 1: them worked. It was I don't know what the heck 1344 01:12:06,120 --> 01:12:09,520 Speaker 1: to do. I have like twelve CDs. They're essentially paperweights, 1345 01:12:09,840 --> 01:12:11,760 Speaker 1: but it's like the whole box set and I don't 1346 01:12:11,760 --> 01:12:15,000 Speaker 1: know what was DVD or it was because it was 1347 01:12:15,040 --> 01:12:18,280 Speaker 1: done in the sixties and it's um. I'm sure I could. 1348 01:12:18,360 --> 01:12:20,400 Speaker 1: I sure I could pick up another set somewhere, but 1349 01:12:20,520 --> 01:12:23,760 Speaker 1: it's sitting somewhere in my basement in a box. The 1350 01:12:23,880 --> 01:12:27,120 Speaker 1: Fineman Lectures all cd s, they're all completely but don't 1351 01:12:27,160 --> 01:12:30,120 Speaker 1: read much physics anymore. I have them. I like to 1352 01:12:30,240 --> 01:12:32,080 Speaker 1: keep up in a popular way, but I don't. I 1353 01:12:32,120 --> 01:12:35,639 Speaker 1: don't read anything. It's a handful of of astrophysics blogs. 1354 01:12:35,720 --> 01:12:39,639 Speaker 1: I still track. Phil Plate is a guy whose Twitter 1355 01:12:39,760 --> 01:12:44,280 Speaker 1: handle is bad Astronomer, and he writes some really interesting stuff. 1356 01:12:44,400 --> 01:12:49,960 Speaker 1: He can bring, um, some really complex things to a 1357 01:12:50,200 --> 01:12:53,080 Speaker 1: to an understandable At this point, I'm a lay person, 1358 01:12:53,280 --> 01:12:56,000 Speaker 1: not a you know, not able to keep up with 1359 01:12:56,040 --> 01:13:01,519 Speaker 1: any of the high mass. But the ongoing advancements in 1360 01:13:01,640 --> 01:13:06,320 Speaker 1: physics are just phenomenal, most recently the gravitational waves. And 1361 01:13:06,479 --> 01:13:09,960 Speaker 1: it's just of all the sciences that are eventually going 1362 01:13:10,000 --> 01:13:12,760 Speaker 1: to crack the secrets to the universe, it looks like 1363 01:13:12,880 --> 01:13:15,960 Speaker 1: physics is way out ahead of everybody. Yeah, it's astonishing 1364 01:13:16,040 --> 01:13:18,000 Speaker 1: if you if you sat out in physics, you kind 1365 01:13:18,040 --> 01:13:20,400 Speaker 1: of get spoiled for everything else. That's that's a good 1366 01:13:20,479 --> 01:13:23,400 Speaker 1: way to and it's it's nice to sit back and say, well, 1367 01:13:23,439 --> 01:13:26,400 Speaker 1: I'm glad I don't have double labs anymore. But it's 1368 01:13:26,479 --> 01:13:30,200 Speaker 1: interesting to watch this and and it almost seems that 1369 01:13:30,360 --> 01:13:36,200 Speaker 1: the pace of new knowledge is accelerating. Look, we just 1370 01:13:36,360 --> 01:13:40,439 Speaker 1: landed on that comet last year. These things were hard 1371 01:13:40,479 --> 01:13:44,280 Speaker 1: to even imagine outside of science fiction a decade ago. 1372 01:13:44,439 --> 01:13:46,800 Speaker 1: And it's just it's really to me that stuff is 1373 01:13:47,560 --> 01:13:50,639 Speaker 1: is utterly fascinating. Yeah, I almost have a slightly religious 1374 01:13:50,680 --> 01:13:53,000 Speaker 1: feeling about it. I'm not religious, but you think about 1375 01:13:53,040 --> 01:13:55,720 Speaker 1: this gravitational way thing, and you say, okay, it took 1376 01:13:56,160 --> 01:13:58,920 Speaker 1: a hundred years for a thousand people to verify the 1377 01:13:59,000 --> 01:14:04,800 Speaker 1: prediction of one guy years ago, and the fact that 1378 01:14:04,920 --> 01:14:07,120 Speaker 1: somebody could figure out the way the universe works just 1379 01:14:07,320 --> 01:14:09,720 Speaker 1: by pure thought. It's sort of an intuition which he 1380 01:14:09,760 --> 01:14:12,800 Speaker 1: then elaborates into a model. It's just sort of yeah, 1381 01:14:12,800 --> 01:14:16,519 Speaker 1: it's enough to make your believerable. There's there's a fascinating book. 1382 01:14:16,520 --> 01:14:21,439 Speaker 1: As long as we're talking about this, So um, there's 1383 01:14:21,520 --> 01:14:23,639 Speaker 1: a and now I'm drawing a blank on the name. 1384 01:14:24,320 --> 01:14:30,080 Speaker 1: There's a very famous physicist who asked about why we've 1385 01:14:30,160 --> 01:14:34,880 Speaker 1: never come across life elsewhere where? Is a firm's paradox. 1386 01:14:35,640 --> 01:14:39,080 Speaker 1: Where is everybody? So if if you have you have 1387 01:14:39,920 --> 01:14:44,160 Speaker 1: two billion stars in in a galaxy, there are billions 1388 01:14:44,200 --> 01:14:47,479 Speaker 1: and billions of galaxies, how is it possible that we've 1389 01:14:47,560 --> 01:14:51,600 Speaker 1: never come across any other intelligent life anywhere else? And 1390 01:14:52,280 --> 01:14:56,519 Speaker 1: some interesting biologists and physicists put together what they call 1391 01:14:57,240 --> 01:15:01,240 Speaker 1: the rare Earth thesis, which is essentially, life is common 1392 01:15:01,360 --> 01:15:06,120 Speaker 1: on the planet butte in the universe, but intelligent life 1393 01:15:06,240 --> 01:15:09,720 Speaker 1: is relatively rare because the universe turns out to be 1394 01:15:09,840 --> 01:15:14,560 Speaker 1: a very hostile place. So so that gravitational wave that 1395 01:15:14,680 --> 01:15:18,480 Speaker 1: hit us was relatively modest. But there are there are magnetars, 1396 01:15:18,560 --> 01:15:22,639 Speaker 1: and there are all sorts of pulsars, and if you're 1397 01:15:23,080 --> 01:15:27,240 Speaker 1: even like a few hundred light years away, the gamma 1398 01:15:27,280 --> 01:15:31,640 Speaker 1: waves and the radiations that wash over essentially sterilize the 1399 01:15:31,680 --> 01:15:35,040 Speaker 1: planet of all life. It eventually comes back, but intelligent 1400 01:15:35,160 --> 01:15:37,880 Speaker 1: life has needs a lot of stability for a long time. 1401 01:15:38,400 --> 01:15:41,640 Speaker 1: I find that sort of stuff endlessly fascinated to and 1402 01:15:42,040 --> 01:15:45,800 Speaker 1: the parallels to finance are very much there if you 1403 01:15:45,960 --> 01:15:48,320 Speaker 1: sit back and think about it a little bit. That's 1404 01:15:48,320 --> 01:15:50,000 Speaker 1: really I think I've seen a vision of this that's 1405 01:15:50,040 --> 01:15:52,600 Speaker 1: really interesting, and I think I've only read about descriptively, 1406 01:15:52,680 --> 01:15:55,160 Speaker 1: but where they say you might get intelligent life, but 1407 01:15:55,240 --> 01:15:57,680 Speaker 1: soon or later they destroy themselves and they unable to 1408 01:15:57,800 --> 01:16:00,200 Speaker 1: communicate with the rest of the universe by reverting to 1409 01:16:00,600 --> 01:16:04,080 Speaker 1: any primitive state that either either universe. So first of all, 1410 01:16:04,200 --> 01:16:06,080 Speaker 1: you have to be in the exact right sweet spot 1411 01:16:06,360 --> 01:16:09,120 Speaker 1: distance from from the Sun. You have to hope a 1412 01:16:09,200 --> 01:16:12,840 Speaker 1: meteor doesn't come along or any of the early Solar 1413 01:16:12,960 --> 01:16:18,040 Speaker 1: System formations that hits you. There's another thesis that says, 1414 01:16:18,920 --> 01:16:22,640 Speaker 1: so the Earth has this giant, oversized moon relative to 1415 01:16:22,720 --> 01:16:26,040 Speaker 1: all the other planets, and but for that, you may 1416 01:16:26,160 --> 01:16:31,080 Speaker 1: not have had tides, which really lead to accelerating a 1417 01:16:31,160 --> 01:16:35,200 Speaker 1: lot of of just taking protein strings and leading to 1418 01:16:36,600 --> 01:16:40,000 Speaker 1: minerals and and taking tides all the waing inland and 1419 01:16:40,040 --> 01:16:42,479 Speaker 1: then having the tides leave. You get that with an 1420 01:16:42,520 --> 01:16:46,240 Speaker 1: oversized moon without that, So you know, once, once you 1421 01:16:46,280 --> 01:16:49,120 Speaker 1: study physics, you can never really let it go. It's 1422 01:16:49,120 --> 01:16:51,439 Speaker 1: always there. I find that stuff I don't know how 1423 01:16:51,479 --> 01:16:57,480 Speaker 1: we had this with this digression from other than Schopenhauer's 1424 01:16:57,600 --> 01:17:01,080 Speaker 1: essays and aphorisms, I guess and and finement stuff, which 1425 01:17:01,200 --> 01:17:05,040 Speaker 1: is uh, which is fascinating. So so since you joined 1426 01:17:05,160 --> 01:17:08,439 Speaker 1: the industry, boy, that was like a I don't know 1427 01:17:08,560 --> 01:17:12,240 Speaker 1: where that digression came from, but it's in the back 1428 01:17:12,320 --> 01:17:14,080 Speaker 1: of my head. And it's uh, I never heard that 1429 01:17:14,120 --> 01:17:19,599 Speaker 1: about the moon, the moon making making them cheer. Well, 1430 01:17:19,880 --> 01:17:21,880 Speaker 1: it's not so much that it made it possible, but 1431 01:17:23,080 --> 01:17:26,479 Speaker 1: you end up with these really large tidal Now keep 1432 01:17:26,479 --> 01:17:28,400 Speaker 1: in mind the moon is slowly moving away from the 1433 01:17:28,479 --> 01:17:33,360 Speaker 1: Earth um and and a few millennia ago, a few 1434 01:17:33,479 --> 01:17:36,479 Speaker 1: billion years ago, it was much closer, which, by the way, 1435 01:17:36,560 --> 01:17:39,960 Speaker 1: raises a whole another question. How did this moon end 1436 01:17:40,040 --> 01:17:43,599 Speaker 1: up around this planet? Was it captured? Was there were 1437 01:17:43,640 --> 01:17:47,080 Speaker 1: there two moons um one of which got absorbed into 1438 01:17:47,160 --> 01:17:49,439 Speaker 1: the other? I mean, there's all sorts of thesis is 1439 01:17:49,479 --> 01:17:52,479 Speaker 1: as to how you end up with a really big 1440 01:17:52,600 --> 01:17:56,200 Speaker 1: moon relative to a mid sized planet. When we look 1441 01:17:56,240 --> 01:17:58,920 Speaker 1: at Jupiter of saner and they have what is it 1442 01:17:58,960 --> 01:18:04,639 Speaker 1: a hundred dozens and dozens of moons, all of which yeah, 1443 01:18:04,720 --> 01:18:07,320 Speaker 1: so it's it's I'll dig up the name of that 1444 01:18:07,840 --> 01:18:10,120 Speaker 1: in a while ago. I think it's called Rare Earth. 1445 01:18:10,680 --> 01:18:14,439 Speaker 1: But if you like the occasional physics nonfiction, you might 1446 01:18:14,520 --> 01:18:18,400 Speaker 1: find this um and it is a little uh, not 1447 01:18:18,600 --> 01:18:24,600 Speaker 1: religious or spiritual, but anything that changes your perspective of 1448 01:18:24,880 --> 01:18:28,440 Speaker 1: our place in the universe is really kind of fascinating 1449 01:18:28,479 --> 01:18:31,639 Speaker 1: and spiritual and curious. Although there are people who insist 1450 01:18:31,800 --> 01:18:35,920 Speaker 1: firm's right and will eventually find people, but so far. 1451 01:18:37,680 --> 01:18:40,120 Speaker 1: And then there's another thesis that says, why are you 1452 01:18:40,280 --> 01:18:43,320 Speaker 1: looking for them? If they can communicate with you, they 1453 01:18:43,360 --> 01:18:47,240 Speaker 1: could come here and basically, oh, a nice planet to Uh. 1454 01:18:47,760 --> 01:18:49,800 Speaker 1: There's a third one that says, don't let anybody see you, 1455 01:18:49,920 --> 01:18:53,200 Speaker 1: because people in the Amazon or to the bush exactly 1456 01:18:53,280 --> 01:18:57,639 Speaker 1: that we we are, the American Indians, and anybody who's 1457 01:18:57,680 --> 01:19:02,840 Speaker 1: coming superior technology is superior here ability to do whatever 1458 01:19:02,920 --> 01:19:06,080 Speaker 1: they want. And let's hope that doesn't happen anytime soon. 1459 01:19:06,280 --> 01:19:10,639 Speaker 1: Um So, but we digress. So so let's go back 1460 01:19:10,760 --> 01:19:16,320 Speaker 1: to uh, quantitative finance. So since you started in that industry, 1461 01:19:16,479 --> 01:19:22,360 Speaker 1: what do you think are the biggest changes that have 1462 01:19:22,640 --> 01:19:29,960 Speaker 1: affected finance? I think electronic trading and electronic markets and 1463 01:19:30,520 --> 01:19:33,080 Speaker 1: the effect that said filtering down on everything so kind 1464 01:19:33,080 --> 01:19:36,400 Speaker 1: of computers in a sense no more open outcry um, 1465 01:19:36,960 --> 01:19:40,839 Speaker 1: you know, everything matched by computer, and that certainly affected 1466 01:19:40,880 --> 01:19:45,040 Speaker 1: the careers of people and and um did that did 1467 01:19:45,240 --> 01:19:46,920 Speaker 1: that enable a lot of what we see on the 1468 01:19:47,000 --> 01:19:50,400 Speaker 1: quantitative side to progress. Without that, you kind of at 1469 01:19:50,439 --> 01:19:52,760 Speaker 1: an impasse, aren't you. Yeah, they kind of go in 1470 01:19:52,880 --> 01:19:56,519 Speaker 1: lock step. You know, you can now get good price data, 1471 01:19:56,600 --> 01:19:59,280 Speaker 1: good volatility that I guess the bond market isn't totally 1472 01:19:59,439 --> 01:20:02,400 Speaker 1: isn't really electronic yet, but hitting in that direction. And 1473 01:20:02,520 --> 01:20:05,280 Speaker 1: currencies are still over the counter. But I think that's 1474 01:20:05,320 --> 01:20:07,920 Speaker 1: the biggest trend that people can now trade by algorithms 1475 01:20:07,960 --> 01:20:11,200 Speaker 1: and trade by computer, and um, those skills are more 1476 01:20:11,240 --> 01:20:14,040 Speaker 1: and more more and more more and more in demand. 1477 01:20:14,280 --> 01:20:16,360 Speaker 1: So when you when you look at models, so you 1478 01:20:16,439 --> 01:20:20,080 Speaker 1: could create models for equities and derivatives, and you can 1479 01:20:20,120 --> 01:20:25,479 Speaker 1: create models for fixed income. Can you can we not 1480 01:20:25,720 --> 01:20:29,080 Speaker 1: create models for currencies because of the way they're traded. 1481 01:20:30,280 --> 01:20:32,280 Speaker 1: You can still make models, but they're kind of harder 1482 01:20:32,320 --> 01:20:34,400 Speaker 1: to implement because you I don't know how it is 1483 01:20:34,439 --> 01:20:35,720 Speaker 1: now a little out of date, but you had to 1484 01:20:35,760 --> 01:20:40,719 Speaker 1: call somebody and you can't really do statistical arbitrage easily 1485 01:20:40,760 --> 01:20:43,240 Speaker 1: where you have an algorithm that just sends out orders 1486 01:20:43,320 --> 01:20:45,000 Speaker 1: and buys when it needs to in cells when it 1487 01:20:45,040 --> 01:20:46,560 Speaker 1: needs to, because you still have to call somebody on 1488 01:20:46,640 --> 01:20:50,400 Speaker 1: the telephone. Had to do that in equities to sort 1489 01:20:50,439 --> 01:20:53,080 Speaker 1: of you know, twenty five years ago, whereas now now 1490 01:20:53,680 --> 01:20:58,040 Speaker 1: nobody makes schools. I wonder why currencies aren't as automated 1491 01:20:58,160 --> 01:21:01,559 Speaker 1: or as electronically driven as as some of these other 1492 01:21:01,640 --> 01:21:03,920 Speaker 1: asset classes. Yeah, because actually they should be in the 1493 01:21:04,040 --> 01:21:05,880 Speaker 1: sense that there you know, it's hard to do with 1494 01:21:05,960 --> 01:21:09,040 Speaker 1: bonds because bunds are very idiosyncratic and there aren't one 1495 01:21:09,160 --> 01:21:12,720 Speaker 1: stock millions of different bonds and yeah, but currencies are 1496 01:21:12,800 --> 01:21:14,240 Speaker 1: and I think it will go that way. I think 1497 01:21:14,280 --> 01:21:16,840 Speaker 1: partly because in a cynical sort of way, there's a 1498 01:21:16,880 --> 01:21:19,720 Speaker 1: lot of money to be made by I think part 1499 01:21:19,720 --> 01:21:21,640 Speaker 1: of the people who trade currencies, my guests, would be 1500 01:21:21,680 --> 01:21:23,920 Speaker 1: that they're reluctant to go electronic because there's a lot 1501 01:21:24,000 --> 01:21:26,639 Speaker 1: of them. You look at what Chase charging you three 1502 01:21:27,160 --> 01:21:29,840 Speaker 1: for for every time you use your credit card in 1503 01:21:29,880 --> 01:21:32,920 Speaker 1: a foreign country. There's very big, very big margins over there. 1504 01:21:33,200 --> 01:21:36,040 Speaker 1: So it hasn't gone electronic because there's a big incentive. 1505 01:21:36,080 --> 01:21:38,120 Speaker 1: I think that's incentive to not to Did you see 1506 01:21:38,120 --> 01:21:40,720 Speaker 1: somebody got into trouble some bank a few a few 1507 01:21:40,760 --> 01:21:45,639 Speaker 1: months ago for giving giving clients UM the worst price, 1508 01:21:46,040 --> 01:21:49,720 Speaker 1: the worst currency price of the day, consistently consistently giving 1509 01:21:49,760 --> 01:21:51,840 Speaker 1: them the low if they were if they were selling 1510 01:21:51,880 --> 01:21:55,479 Speaker 1: in the hive, they were buying. That's amazing. Why why 1511 01:21:55,520 --> 01:21:59,479 Speaker 1: are we not surprised by that? UM? So we've we've 1512 01:21:59,520 --> 01:22:02,080 Speaker 1: talked about what's changed in the past. What are the 1513 01:22:02,240 --> 01:22:05,919 Speaker 1: upcoming shifts that you see, UM that's going to impact 1514 01:22:06,080 --> 01:22:12,439 Speaker 1: quants or impact the concept of quantitative UH modeling. UM. 1515 01:22:15,280 --> 01:22:17,320 Speaker 1: Try to think a little bit about this as you're talking. 1516 01:22:19,880 --> 01:22:22,360 Speaker 1: Just the extensions of what's happening. I think the vanishing 1517 01:22:22,400 --> 01:22:25,200 Speaker 1: of small investors, which has been happening and maybe not 1518 01:22:25,320 --> 01:22:27,439 Speaker 1: so much in Japan, but it's certainly happened here. Nothing 1519 01:22:27,560 --> 01:22:30,080 Speaker 1: is now starting to happen in Japan. What about China, 1520 01:22:30,160 --> 01:22:32,400 Speaker 1: which I was going to say, China seems to be 1521 01:22:32,520 --> 01:22:36,720 Speaker 1: all mom and pop investors. When does that become I 1522 01:22:36,840 --> 01:22:39,960 Speaker 1: think they're more mature and institutional. I think that's that's 1523 01:22:40,040 --> 01:22:43,040 Speaker 1: doomed in the long run. I mean, I think mom 1524 01:22:43,120 --> 01:22:45,120 Speaker 1: and poppy investors are doomed in the long run, and 1525 01:22:45,160 --> 01:22:48,240 Speaker 1: in China to China is still manipulating their markets a lot, 1526 01:22:48,280 --> 01:22:51,160 Speaker 1: as far as I can tell, um, So how does 1527 01:22:51,200 --> 01:22:55,720 Speaker 1: that play out? I suppose, I suppose badly. At some 1528 01:22:55,840 --> 01:22:59,240 Speaker 1: point Chinese markets will collapse and I'm worse than what 1529 01:22:59,360 --> 01:23:02,360 Speaker 1: we've just four correction we've seen, I would guess. So 1530 01:23:02,520 --> 01:23:04,960 Speaker 1: I'm not really good at predicting the future about the stuff, 1531 01:23:05,000 --> 01:23:06,720 Speaker 1: but I think as China is going to have to 1532 01:23:07,520 --> 01:23:10,040 Speaker 1: democratize them now you see a lot of people are 1533 01:23:10,040 --> 01:23:11,639 Speaker 1: trying to get money out of the country. They're gonna 1534 01:23:11,640 --> 01:23:15,080 Speaker 1: have to put in capital controls. Um. It'll be interesting 1535 01:23:15,160 --> 01:23:17,400 Speaker 1: to see what happens. Certainly, in the program I teach, 1536 01:23:17,439 --> 01:23:20,479 Speaker 1: it's sort of astonishing of the of the students are 1537 01:23:20,520 --> 01:23:25,240 Speaker 1: basically from mainland China. Really that's a wild number, and 1538 01:23:25,360 --> 01:23:27,839 Speaker 1: that's true in most of the sciences in this country. 1539 01:23:28,320 --> 01:23:30,479 Speaker 1: Are they going back home or is this a way 1540 01:23:30,560 --> 01:23:32,720 Speaker 1: to get out of the country? A mix, a mix, 1541 01:23:33,000 --> 01:23:35,120 Speaker 1: but a lot of them. You know, it's an expensive program. 1542 01:23:35,320 --> 01:23:38,320 Speaker 1: People pay you know, I don't know, sixty or eighty 1543 01:23:38,360 --> 01:23:42,000 Speaker 1: thousand dollars a year to study plus living costs. Um. 1544 01:23:42,920 --> 01:23:46,120 Speaker 1: Chinese Chinese have money, they get they can get money 1545 01:23:46,160 --> 01:23:49,559 Speaker 1: out of the country until now to pay for tuition 1546 01:23:49,720 --> 01:23:53,320 Speaker 1: or is this just a way to know this is 1547 01:23:53,400 --> 01:23:56,719 Speaker 1: for tuition? And and so yeah, they're they're, they're millions 1548 01:23:56,760 --> 01:24:00,000 Speaker 1: of It's good. The millions of Chinese students in this country. 1549 01:24:00,200 --> 01:24:02,320 Speaker 1: Some stay, some go back. But really the graduate schools 1550 01:24:02,360 --> 01:24:05,519 Speaker 1: in the sciences and certainly in the finance engineering programs, 1551 01:24:05,560 --> 01:24:08,200 Speaker 1: but I think in physics general will run on all 1552 01:24:08,280 --> 01:24:10,920 Speaker 1: run on on foreign students in China. Why is that? 1553 01:24:11,120 --> 01:24:15,040 Speaker 1: That's fascinating. It's a good And when I worked at 1554 01:24:15,080 --> 01:24:17,640 Speaker 1: Goldman I ran this con group. You know, the majority 1555 01:24:17,720 --> 01:24:21,439 Speaker 1: of foreign, not necessarily Chinese, but the majority of I 1556 01:24:21,479 --> 01:24:23,439 Speaker 1: don't know, Rember. What I see in a lot of 1557 01:24:23,600 --> 01:24:28,479 Speaker 1: math is so so it's it's Chinese, it's Korean, it's Indian. 1558 01:24:29,600 --> 01:24:32,040 Speaker 1: Used to be Jews sort of fifty years ago. And 1559 01:24:32,600 --> 01:24:36,759 Speaker 1: so what is it Each subsequent generation of immigrants takes 1560 01:24:36,880 --> 01:24:41,400 Speaker 1: the hardest working area until the next. General, How does 1561 01:24:41,479 --> 01:24:43,320 Speaker 1: that work? Yeah, I think I think that's what kind 1562 01:24:43,320 --> 01:24:47,320 Speaker 1: of happens. They immigrants, um immigrants come in and and 1563 01:24:48,720 --> 01:24:51,360 Speaker 1: do the stuff that's that's hard. Not so much hard 1564 01:24:51,400 --> 01:24:53,879 Speaker 1: in a difficult sense, but hard in that very concrete 1565 01:24:54,400 --> 01:24:56,479 Speaker 1: that you know, what successes and you know what, you 1566 01:24:56,560 --> 01:24:58,960 Speaker 1: know what what being good at it means. And then 1567 01:24:59,120 --> 01:25:01,920 Speaker 1: their kids want to be businessmen. They don't want to 1568 01:25:01,920 --> 01:25:05,000 Speaker 1: work that hard. So then that opens up the flow 1569 01:25:05,040 --> 01:25:07,679 Speaker 1: as first who was the Jews, and then it's Asians, 1570 01:25:07,760 --> 01:25:11,000 Speaker 1: and then it's another word, it's each subsequent generation of immigrants. 1571 01:25:11,520 --> 01:25:13,240 Speaker 1: So you know, a friend of mine is fond of 1572 01:25:13,320 --> 01:25:16,439 Speaker 1: pointing out that when you mentioned a lot of the 1573 01:25:16,520 --> 01:25:20,120 Speaker 1: Jews had come over in a way of immigration, studying 1574 01:25:20,200 --> 01:25:23,519 Speaker 1: the sciences at one point in time, the predecessor to 1575 01:25:23,640 --> 01:25:30,120 Speaker 1: the NBA, which back then was heavily represented with Jewish 1576 01:25:30,200 --> 01:25:33,439 Speaker 1: basketball players. And I was astonished when I first heard it, 1577 01:25:33,840 --> 01:25:35,840 Speaker 1: But neither did I the first time I heard that, 1578 01:25:35,880 --> 01:25:38,320 Speaker 1: I'm like, what come on, you're pulling my leg. But 1579 01:25:38,439 --> 01:25:41,960 Speaker 1: it's the same thesis of a wave of immigrants comes. 1580 01:25:42,320 --> 01:25:45,560 Speaker 1: They're willing to do stuff that everybody who's here and 1581 01:25:45,680 --> 01:25:48,360 Speaker 1: somewhat wealthy and a little spoiled perhaps don't want to do. 1582 01:25:49,120 --> 01:25:51,400 Speaker 1: And then after they go through that and achieve some 1583 01:25:51,479 --> 01:25:54,080 Speaker 1: degree of success, their kids don't want to do that, 1584 01:25:54,320 --> 01:25:57,559 Speaker 1: and it creates yet another opening, and then the next 1585 01:25:57,640 --> 01:26:02,720 Speaker 1: wave of immigrants come. That was, uh, I wonder how 1586 01:26:02,800 --> 01:26:05,839 Speaker 1: far this continues, who who's going to follow the Chinese 1587 01:26:05,920 --> 01:26:11,240 Speaker 1: after that? But eight of your students as Chinese in 1588 01:26:11,360 --> 01:26:13,720 Speaker 1: this master's program, about like a hundred students, So I 1589 01:26:13,760 --> 01:26:16,240 Speaker 1: would say I'm guessing a little bit. I would say 1590 01:26:17,240 --> 01:26:21,320 Speaker 1: comes straight from mainland China. Another fort come from America 1591 01:26:21,360 --> 01:26:24,000 Speaker 1: or Europe, but the Chinese citizens and went their undergrad 1592 01:26:24,920 --> 01:26:26,600 Speaker 1: In fact, this is sort of a little funny. But 1593 01:26:26,760 --> 01:26:29,639 Speaker 1: my son who lives in Hong Kong, he teaches history 1594 01:26:29,720 --> 01:26:32,599 Speaker 1: at Hong Kong University of Science and Technology. And when 1595 01:26:32,600 --> 01:26:34,320 Speaker 1: I've been there, of setting on his class and his 1596 01:26:34,439 --> 01:26:38,719 Speaker 1: classroom looks more cosmopolitan than mine. Really, yeah, I wonder 1597 01:26:38,800 --> 01:26:42,680 Speaker 1: what that's about. That's quite fashionating. He's got some fair 1598 01:26:42,720 --> 01:26:46,960 Speaker 1: amount of and expacts and people or Australians who come there, 1599 01:26:47,000 --> 01:26:49,880 Speaker 1: whereas our classes are pretty much nothing wrong with it. 1600 01:26:50,000 --> 01:26:52,599 Speaker 1: But that just shows you where it's going. So your students, 1601 01:26:52,640 --> 01:26:56,920 Speaker 1: who who are who graduate the master's degree in financial engineering? 1602 01:26:57,240 --> 01:27:01,080 Speaker 1: Where they end up working after that? Some in China, 1603 01:27:01,560 --> 01:27:05,479 Speaker 1: some in money management, some in the risk management, a 1604 01:27:05,600 --> 01:27:10,120 Speaker 1: few in trading UM, some in non purely financial firms. 1605 01:27:10,240 --> 01:27:14,080 Speaker 1: Now there's a big trend towards machine learning and big data. UM. 1606 01:27:14,560 --> 01:27:17,720 Speaker 1: How many how many what percentage of your students stay 1607 01:27:17,760 --> 01:27:20,160 Speaker 1: in the United States and work in that field? Um, 1608 01:27:20,920 --> 01:27:23,600 Speaker 1: I would guess fifty or sixty, but I'm actually on 1609 01:27:23,640 --> 01:27:26,519 Speaker 1: shaky ground now. I don't know. For the ballpark about 1610 01:27:26,560 --> 01:27:31,920 Speaker 1: half that wouldn't surprise you. Huh, that's quite fascinating. Um. So, 1611 01:27:32,360 --> 01:27:35,160 Speaker 1: speaking of students, what what sort of advice would you 1612 01:27:35,280 --> 01:27:39,479 Speaker 1: give to a millennial or someone who's just graduating from 1613 01:27:39,520 --> 01:27:43,040 Speaker 1: school and is interested in in a career in modeling 1614 01:27:43,080 --> 01:27:47,840 Speaker 1: and quantitative finance. Okay, I'm somewhat cynical stuff. I don't know. 1615 01:27:47,840 --> 01:27:50,360 Speaker 1: I would say, first of all, yeah, I started out 1616 01:27:50,400 --> 01:27:53,000 Speaker 1: thinking I was going to have one career and you're 1617 01:27:53,080 --> 01:27:54,560 Speaker 1: the same from what you were telling And then you 1618 01:27:54,640 --> 01:27:57,040 Speaker 1: discover that you're going to change and life has defeated 1619 01:27:57,080 --> 01:27:59,080 Speaker 1: to day. Yeah too, So I've kind of had three 1620 01:27:59,120 --> 01:28:02,000 Speaker 1: in a sense. I started in physics, then I went 1621 01:28:02,080 --> 01:28:03,920 Speaker 1: to Bell Labs for a while, and then I ended 1622 01:28:04,000 --> 01:28:06,240 Speaker 1: up in finance. So I would say, expect to have 1623 01:28:06,320 --> 01:28:09,360 Speaker 1: more than one career. Don't do don't if it you're 1624 01:28:09,360 --> 01:28:10,799 Speaker 1: going to do one thing for the same time, it's 1625 01:28:10,800 --> 01:28:14,320 Speaker 1: actually quite invigorating to change. Yeah, it's a whole different 1626 01:28:14,360 --> 01:28:17,560 Speaker 1: set of muscles, and after twenty years, it's kind of 1627 01:28:17,640 --> 01:28:20,120 Speaker 1: nice to have a difference. Um, I would say, get 1628 01:28:20,160 --> 01:28:23,000 Speaker 1: good at programming, at least that's my experience in almost anything. 1629 01:28:23,040 --> 01:28:24,600 Speaker 1: If you can do your own Yeah, I'd like to 1630 01:28:24,680 --> 01:28:27,200 Speaker 1: tell people to be willing to get your hands dirty 1631 01:28:27,240 --> 01:28:28,920 Speaker 1: and do your own dirty work. Don't just be a 1632 01:28:29,000 --> 01:28:31,479 Speaker 1: manager maybe one day. But that's the thing I was 1633 01:28:31,520 --> 01:28:33,160 Speaker 1: going to say earlier. It summer seems to me a 1634 01:28:33,200 --> 01:28:35,920 Speaker 1: lot of maybe not now after the internet craze, but 1635 01:28:36,439 --> 01:28:38,559 Speaker 1: in the nineties, most Americans and when I worked at 1636 01:28:38,920 --> 01:28:42,400 Speaker 1: most Americans wanted to be managers and most foreigners wanted 1637 01:28:42,439 --> 01:28:45,599 Speaker 1: to work with their hands or their heads sort of spread. 1638 01:28:46,200 --> 01:28:48,639 Speaker 1: And now it's changed a little bit. I'm getting off topic. 1639 01:28:48,720 --> 01:28:52,040 Speaker 1: It's changed a little bit because Americans are suddenly discovering 1640 01:28:52,080 --> 01:28:54,280 Speaker 1: that you can get rich by by being a good 1641 01:28:54,320 --> 01:28:57,400 Speaker 1: program and like like like Mark Zuckerberg or somebody like that, 1642 01:28:58,240 --> 01:28:59,760 Speaker 1: and so it's changed a little bit. But for a 1643 01:28:59,840 --> 01:29:01,240 Speaker 1: long time it seemed to me where you can just 1644 01:29:01,280 --> 01:29:03,639 Speaker 1: wanted to be managers, so why go do financial engineering, 1645 01:29:03,680 --> 01:29:06,640 Speaker 1: whereas foreigners sort of had no choice. I wonder how 1646 01:29:06,760 --> 01:29:09,000 Speaker 1: much of that was a post War War two phenomena, 1647 01:29:09,439 --> 01:29:12,040 Speaker 1: because if you think about the era that fouled the 1648 01:29:12,520 --> 01:29:16,440 Speaker 1: Second World War, you had a huge rise of corporations 1649 01:29:16,720 --> 01:29:22,400 Speaker 1: and what we almost derisively described as middle management today, 1650 01:29:22,880 --> 01:29:27,280 Speaker 1: it was the path to a reasonably safe, certain comfortable job. 1651 01:29:27,720 --> 01:29:30,080 Speaker 1: But that's all gone away a long time ago. I wonder, 1652 01:29:30,120 --> 01:29:33,240 Speaker 1: I wonder how much of that is demographics and how 1653 01:29:33,320 --> 01:29:38,240 Speaker 1: that's changed. That's interesting, Probably probably a large part. Yeah, 1654 01:29:38,240 --> 01:29:40,080 Speaker 1: I was trying to think. So I was saying, be 1655 01:29:40,120 --> 01:29:42,320 Speaker 1: wanting to do your own dirty work, get your hands dirty, 1656 01:29:42,400 --> 01:29:46,880 Speaker 1: and learned to program. You learned to program, don't ignore 1657 01:29:46,960 --> 01:29:49,360 Speaker 1: I don't know I was gonna say him. One thing 1658 01:29:49,400 --> 01:29:51,200 Speaker 1: I kind of learned is when I left physics. When 1659 01:29:51,240 --> 01:29:53,960 Speaker 1: I left particle physics, I am I was sort of 1660 01:29:53,960 --> 01:29:56,280 Speaker 1: a little bit disgusted with myself that I thought I 1661 01:29:56,320 --> 01:29:57,800 Speaker 1: was going to be a physicist and I wasn't gonna 1662 01:29:57,840 --> 01:30:00,519 Speaker 1: be one. And they were jobs in apply physics as 1663 01:30:00,560 --> 01:30:03,880 Speaker 1: opposed to pure research like an energy or or heat, 1664 01:30:04,200 --> 01:30:06,720 Speaker 1: you know, heat stuff, and I didn't want to do that. 1665 01:30:06,760 --> 01:30:08,280 Speaker 1: I thought, if I'm going to get out of physics, 1666 01:30:08,360 --> 01:30:09,760 Speaker 1: I'm going to go sort of all the way. And 1667 01:30:09,800 --> 01:30:11,280 Speaker 1: it works for me. But at the same time, I 1668 01:30:11,400 --> 01:30:14,320 Speaker 1: realized over the years that almost everything is interesting, and 1669 01:30:14,400 --> 01:30:16,000 Speaker 1: that maybe I was wrong. I could have gotten just 1670 01:30:16,160 --> 01:30:19,720 Speaker 1: as interested in doing something else. So I think the 1671 01:30:19,800 --> 01:30:22,519 Speaker 1: more applied as opposed to theoretical Yes, and maybe it 1672 01:30:22,600 --> 01:30:24,960 Speaker 1: might have still been interesting. And but but I sort 1673 01:30:24,960 --> 01:30:27,040 Speaker 1: of scandered at the time in a snobby sort of way. 1674 01:30:27,560 --> 01:30:30,560 Speaker 1: And um, my experience over the years with lots of 1675 01:30:30,640 --> 01:30:34,839 Speaker 1: things is that when you get involved in them hardcore 1676 01:30:34,920 --> 01:30:37,320 Speaker 1: and deeply, you find all sorts of interesting things that 1677 01:30:37,400 --> 01:30:41,240 Speaker 1: you didn't expect. And so I think it's that's the 1678 01:30:41,280 --> 01:30:44,679 Speaker 1: advice I would give people is m is um plunge 1679 01:30:44,760 --> 01:30:46,880 Speaker 1: in and if you have to do something different, and 1680 01:30:47,520 --> 01:30:49,320 Speaker 1: and when you get deep inside you find a whole 1681 01:30:49,320 --> 01:30:52,880 Speaker 1: world opening up. And and then our final question, what 1682 01:30:53,080 --> 01:30:56,800 Speaker 1: is it that you know about modeling, about investing, about 1683 01:30:56,880 --> 01:31:00,479 Speaker 1: quantitative finance today you wish you knew thirty years ago 1684 01:31:00,560 --> 01:31:03,960 Speaker 1: when you you were first stepping into the field. Don't 1685 01:31:04,000 --> 01:31:07,280 Speaker 1: get out when things look bad. Don't get out when 1686 01:31:07,360 --> 01:31:10,360 Speaker 1: things look bad, because in the long run, don't sell 1687 01:31:10,400 --> 01:31:13,519 Speaker 1: at the bottom. So, in other words, mean regression, that 1688 01:31:13,680 --> 01:31:17,360 Speaker 1: reversion to one of these people that that when everything 1689 01:31:17,520 --> 01:31:19,280 Speaker 1: was about to collapse, I thought, oh, my God, like 1690 01:31:19,360 --> 01:31:21,240 Speaker 1: now I've been a soul before it goes to zero 1691 01:31:22,040 --> 01:31:25,479 Speaker 1: and I'm but it doesn't go so far. It doesn't 1692 01:31:25,479 --> 01:31:27,840 Speaker 1: go to zero. So I would say take a long 1693 01:31:28,000 --> 01:31:31,559 Speaker 1: term VI you and ignore the fluctuations. But let's talk 1694 01:31:31,680 --> 01:31:34,320 Speaker 1: to do Professor Dermott. I have to thank you for 1695 01:31:34,400 --> 01:31:37,000 Speaker 1: being so generous with you with your time. This has 1696 01:31:37,080 --> 01:31:41,960 Speaker 1: really been a fascinating conversation. I could sit here for 1697 01:31:42,200 --> 01:31:45,040 Speaker 1: hours longer, but I know you have places um to 1698 01:31:45,160 --> 01:31:48,080 Speaker 1: go and and things to do to review if people 1699 01:31:48,160 --> 01:31:50,479 Speaker 1: want to find your work. My life is a quant 1700 01:31:50,960 --> 01:31:54,320 Speaker 1: is on Amazon. Models Behaving Badly can be found just 1701 01:31:54,479 --> 01:31:59,519 Speaker 1: about anywhere. A Manual Derman dot com at a Manual 1702 01:31:59,600 --> 01:32:03,840 Speaker 1: Derman on Twitter, and your homepage is on at Columbia. 1703 01:32:04,160 --> 01:32:05,680 Speaker 1: I don't actually have one in Columbia. I'll just have 1704 01:32:05,800 --> 01:32:08,080 Speaker 1: Emmanuel Deman dot com. And I once wrote a book 1705 01:32:08,120 --> 01:32:10,920 Speaker 1: of columns and short stories which you can get on 1706 01:32:11,080 --> 01:32:13,320 Speaker 1: Amazon as an e book that I put them myself. 1707 01:32:13,800 --> 01:32:17,560 Speaker 1: And what's the name of that's called bad Behavior? Bad Behavior. 1708 01:32:18,439 --> 01:32:20,519 Speaker 1: That's not a big best seller by any means. But 1709 01:32:20,560 --> 01:32:23,479 Speaker 1: I once wrote a bunch of columns for a German newspaper. 1710 01:32:23,520 --> 01:32:26,479 Speaker 1: There was an editor there. He actually died shortly episode 1711 01:32:26,520 --> 01:32:29,280 Speaker 1: the Frank foot or argument at Siton who liked models 1712 01:32:29,280 --> 01:32:31,320 Speaker 1: behaving badly, And for about a year I wrote a 1713 01:32:31,400 --> 01:32:34,080 Speaker 1: column for them every two weeks, and I took most 1714 01:32:34,160 --> 01:32:36,720 Speaker 1: of it. They put it into German, but I had 1715 01:32:36,760 --> 01:32:38,280 Speaker 1: it in English, and I put most of it into 1716 01:32:38,360 --> 01:32:41,280 Speaker 1: this some little book. I'll put a link up to 1717 01:32:41,360 --> 01:32:44,160 Speaker 1: this when this goes up. Thank you UM so much 1718 01:32:44,200 --> 01:32:47,320 Speaker 1: for your time. For those of you who enjoy this conversation, 1719 01:32:47,600 --> 01:32:51,439 Speaker 1: look upward down an inch on iTunes and you could 1720 01:32:51,520 --> 01:32:56,719 Speaker 1: see all of our prior conversations. I would be remiss 1721 01:32:56,760 --> 01:33:00,160 Speaker 1: if I did not thank UH my research direct jer 1722 01:33:00,240 --> 01:33:03,519 Speaker 1: Michael Batnick, who helped do the deep dive UH into 1723 01:33:03,720 --> 01:33:08,280 Speaker 1: Professor Derman's background. UH special thanks to Taylor Riggs for 1724 01:33:08,840 --> 01:33:12,120 Speaker 1: handling all the booking, and Charlie Vohmer for being our producer. 1725 01:33:12,600 --> 01:33:16,000 Speaker 1: You're listening to Masters in Business on Bloomberg Radio.