Acquired - Renaissance

00:00:00,000 --> 00:00:05,860 I always used to misspell renaissance as I was typing it out at R-E-N and then I would sort of
00:00:05,860 --> 00:00:10,400 like not really know what came from there but I learned a mnemonic to make sure I get it right
00:00:10,400 --> 00:00:14,840 oh I thought you're gonna say you've typed it so many times now over the past month
00:00:14,840 --> 00:00:20,280 well there's that too but you ready for this you can't spell renaissance without A-I
00:00:20,280 --> 00:00:21,120 oh
00:00:21,120 --> 00:00:25,380 touche touche
00:00:25,380 --> 00:00:28,480 all right let's do it
00:00:28,480 --> 00:00:35,040 who got the truth is it you is it you is it you who got the truth now
00:00:35,040 --> 00:00:43,720 is it you is it you is it you sit me down say it straight another story on the way
00:00:43,720 --> 00:00:51,040 welcome to season 14 episode 3 of acquired the podcast about great companies and the stories
00:00:51,040 --> 00:00:56,020 and playbooks behind them i'm ben gilbert i'm david rosenthal and we are your hosts
00:00:56,360 --> 00:00:58,300 they say david that as an
00:00:58,300 --> 00:00:58,460 investor
00:00:58,460 --> 00:01:05,120 you can't beat the market or time the market that you're better off indexing and dollar
00:01:05,120 --> 00:01:10,720 cost averaging rather than trying to be an active stock picker they say there's no persistence of
00:01:10,720 --> 00:01:16,000 returns for hedge funds that this year's big winner can be next year's big loser and that
00:01:16,000 --> 00:01:21,800 nobody gets huge outperformance without taking huge risk when i was in college i actually took
00:01:21,800 --> 00:01:26,800 an economics class with burton malkiel who of course you know was involved in starting vanguard
00:01:26,800 --> 00:01:28,280 and is a big proponent of all that and i was in college and i was in college and i was in college
00:01:28,280 --> 00:01:30,340 in all the part and i was the beginning of all that and that is what i learned ben
00:01:30,940 --> 00:01:33,020 well david it turns out they were wrong
00:01:33,020 --> 00:01:35,820 today listeners we tell the story
00:01:35,820 --> 00:01:39,000 of the best performing investment firm in history
00:01:39,000 --> 00:01:42,080 renaissance technologies or rentek
00:01:42,080 --> 00:01:47,860 their 30-year track record managing billions of dollars has better returns than anyone you have
00:01:47,860 --> 00:01:53,660 ever heard of including berkshire hathaway bridgewater george soros peter lynch or anyone
00:01:53,660 --> 00:01:58,260 else so why haven't you heard of them or if you have why don't you know much about them
00:01:58,260 --> 00:02:03,300 about them? Well, their eye-popping performance is matched only by their extreme secrecy,
00:02:03,300 --> 00:02:09,480 and they are unusual in almost every way. Their founder, Jim Simons, worked for the U.S. government
00:02:09,480 --> 00:02:15,000 in the Cold War as a codebreaker before starting Renaissance. None of the founders or early
00:02:15,000 --> 00:02:20,560 employees had any investing background, and they built the entire thing by hiring PhD physicists,
00:02:20,900 --> 00:02:26,040 astronomers, and speech recognition researchers. They're located in the middle of nowhere in a
00:02:26,040 --> 00:02:31,240 tiny town on Long Island. They don't pay attention to revenues, profits, or even who the CEOs are of
00:02:31,240 --> 00:02:35,880 the companies that they invest in. And at any given time, they probably couldn't even tell you
00:02:35,880 --> 00:02:41,060 what actual stocks they own. Now, you may be thinking, okay, great, I just learned about this
00:02:41,060 --> 00:02:47,380 insane fund with unbelievable performance. And to be specific, listeners, that's 66% annual returns
00:02:47,380 --> 00:02:54,200 before fees. And, you know, well, I want to invest. Well, you can't. To add to everything else that I
00:02:54,200 --> 00:02:55,940 just said, Rentex flagship
00:02:55,940 --> 00:03:00,620 medallion fund doesn't take any outside investors. The partners of the firm have
00:03:00,620 --> 00:03:05,780 become so wealthy from the billions that the fund has generated that the only investors they allow
00:03:05,780 --> 00:03:13,080 in are themselves. Oh, we are going to talk a lot about that towards the end of the episode,
00:03:13,080 --> 00:03:19,200 because I think it's kind of the key to the whole thing. Ooh, cliffhanger, David. I'm excited. So
00:03:19,200 --> 00:03:25,120 what exactly does Renaissance do? Why does it work? And how did it evolve to be the way it is today?
00:03:25,280 --> 00:03:25,920 And while there's a lot of talk about it, there's a lot of talk about it. And I think it's a lot of
00:03:25,940 --> 00:03:30,900 resources are out there are scarce because for one, employees sign a lifetime non-disclosure
00:03:30,900 --> 00:03:35,760 agreement. David and I are going to take you through everything we've learned about the firm
00:03:35,760 --> 00:03:40,420 from our research dating all the way back before Jim Simons started as a math professor
00:03:40,420 --> 00:03:46,160 to understand it all. This episode was selected by our acquired limited partners. And to be honest,
00:03:46,300 --> 00:03:51,140 I didn't think enough people knew what Rentex was to pick it. But when we put it out for a vote,
00:03:51,260 --> 00:03:55,880 the people have spoken. So if you want to become a limited partner and pick one episode each season,
00:03:55,940 --> 00:04:02,500 enjoy the quarterly zoom calls with us, you can join at acquired.fm slash LP. If you want to know
00:04:02,500 --> 00:04:07,380 every time a new episode drops, sign up at acquired.fm slash email. These emails also
00:04:07,380 --> 00:04:12,500 contain hints at what the next episode will be and follow up facts from previous episodes. For
00:04:12,500 --> 00:04:18,460 example, we had a listener Nicholas Cullen email us this time who found the actual document with
00:04:18,460 --> 00:04:24,880 the bylaws of Hermes's controlling family shareholder H51, which we linked to in this
00:04:24,880 --> 00:04:25,860 most recent email.
00:04:26,500 --> 00:04:30,840 Come talk about this episode with us after listening at acquired.fm slash slack. If you
00:04:30,840 --> 00:04:36,240 want more from David and I check out ACQ2. Our most recent episode was with Lata Bjerg-Nudsen,
00:04:36,480 --> 00:04:42,300 who led the team that created the first GLP-1s at Novo Nordisk. So awesome follow up to the
00:04:42,300 --> 00:04:46,680 Novo episode if you liked that one. Before we dive in, we want to briefly share our presenting
00:04:46,680 --> 00:04:50,980 sponsor this season is JP Morgan, specifically their incredible payments business.
00:04:51,220 --> 00:04:55,680 Yeah, just like how we say every company has a story, every company's story is powered by
00:04:55,680 --> 00:05:00,480 payments. And JP Morgan Payments is a part of so many companies that we talk about on Acquired.
00:05:00,840 --> 00:05:06,100 It's not just the Fortune 500, too. They're also helping companies grow from seed to IPO and beyond.
00:05:06,760 --> 00:05:10,880 Yep. So with that, the show is not investment advice. David and I may have investments in
00:05:10,880 --> 00:05:14,780 the companies we discuss or perhaps wish we did. And this show is for informational
00:05:14,780 --> 00:05:19,440 and entertainment purposes only. David, where do we start our story today?
00:05:19,700 --> 00:05:24,940 Ah, well, we start in 1938 in Newton, Massachusetts.
00:05:25,680 --> 00:05:31,940 Which is a fairly wealthy suburb just outside of Boston, where one James Simons is born.
00:05:32,560 --> 00:05:39,000 Both of Jim's parents were very, very smart, especially his mother, Marsha. His dad was a
00:05:39,000 --> 00:05:44,460 salesman for 20th Century Fox, the movie company. His job was he went around to theaters in the
00:05:44,460 --> 00:05:49,980 Northeast and sold packages of movies to them. Super cool. By the way, we knew all this because
00:05:49,980 --> 00:05:55,180 we have to thank Greg Zuckerman, author of The Man Who Solved the Market, which is the
00:05:55,680 --> 00:06:00,680 book out there that is solely dedicated to Rentech and Jim Simons. And we actually got to talk to
00:06:00,680 --> 00:06:03,000 Greg in our research. He helped us out a bunch. Thank you, Greg.
00:06:03,540 --> 00:06:07,000 And help fact check a few of our assumptions of what happened after the book came out.
00:06:07,780 --> 00:06:14,580 So that was Jim's parents. But really, a major influence on him growing up was his grandfather,
00:06:15,260 --> 00:06:20,440 Marsha's dad. There's already kind of echoes of the Bezos story here with the grandfather,
00:06:20,960 --> 00:06:25,660 the mother's father, and spending a bunch of time with him and rubbing off on young Jeff or
00:06:25,680 --> 00:06:31,400 young Jim in this case. And Bezos, of course, would get his start in his career at D.E. Shaw.
00:06:32,160 --> 00:06:34,960 A quant fund coming up at the same time as Rentech.
00:06:35,360 --> 00:06:43,320 But back to Jim here in the 1940s, his grandfather, Peter, owned a shoe factory that made
00:06:43,320 --> 00:06:49,740 women's dress shoes. Jim spends a ton of time there growing up at the factory. So Jim's grandfather,
00:06:49,740 --> 00:06:55,540 Peter, was quite the character. He was a Russian immigrant, and he's,
00:06:55,540 --> 00:06:58,620 kind of like, still more Russia than Boston at this point in time.
00:06:59,260 --> 00:07:04,480 As Greg puts it in the book, Peter reveled in telling Jim and his cousins stories of the
00:07:04,480 --> 00:07:10,720 motherland involving wolves, women, caviar, and vodka. And he teaches young Jim when he's a child
00:07:10,720 --> 00:07:16,540 here in the factory to say Russian phrases like, give me a cigarette and kiss my ass.
00:07:17,240 --> 00:07:20,340 Which I think he probably would say that thousands of times the rest of his life.
00:07:20,340 --> 00:07:25,460 I think so. If you watch interviews with Jim, his hands are always twitching,
00:07:25,540 --> 00:07:30,840 because he has chain-smoked his entire life, probably going back to, like, age 10 in the
00:07:30,840 --> 00:07:36,520 factory. Three packs of merits a day. Unbelievable. Although I think he quit later in life, but he
00:07:36,520 --> 00:07:40,000 definitely chain-smoked the better part of the first, call it, 75 years or something.
00:07:40,180 --> 00:07:45,020 I mean, these famous stories of the conference rooms at Rentech and the war rooms when the
00:07:45,020 --> 00:07:49,300 market is going through, like, a crazy gyration and it's just filled with cigarette smoke and
00:07:49,300 --> 00:07:55,520 it's all Jim. Different time. Different time. So back to Jim's childhood, though, here in the,
00:07:55,540 --> 00:08:02,480 like, Boston suburbs. He grows up certainly not uber-wealthy or uber-rich, but very, very solidly
00:08:02,480 --> 00:08:07,240 upper-middle class. And especially, he's an only child. He has all the resources of his parents,
00:08:07,340 --> 00:08:12,420 his family, his grandfather's this sort of well-to-do entrepreneur. And Jim, you know,
00:08:12,440 --> 00:08:18,840 he gets to rub shoulders in the Boston area with people who are really rich. And he says later,
00:08:19,000 --> 00:08:24,340 I observed that it's very nice to be rich. I had no interest in business, which is not to say I
00:08:24,340 --> 00:08:25,500 had no interest in money.
00:08:26,080 --> 00:08:29,000 Yes. Important to tease out the difference between those two things.
00:08:29,220 --> 00:08:33,900 Yes. Very, very important. And what he means when he says he has no interest in business,
00:08:34,100 --> 00:08:41,280 it's because from a pretty young age, he gets really into math. So the legend has it when Jim
00:08:41,280 --> 00:08:46,800 is four years old, he stumbles into one of Zeno's famous paradoxes from ancient Greek times.
00:08:47,380 --> 00:08:53,000 Yep. This is great. The basic gist of Zeno's paradox is if you are always taking a quantity
00:08:53,000 --> 00:08:55,520 and dividing it by two, you will never get the same amount of money. And that's why he's so
00:08:56,060 --> 00:09:01,240 zero. You will asymptotically approach zero, but you will never actually touch zero. You need to do
00:09:01,240 --> 00:09:06,420 addition or subtraction to do that. Division won't cut it. And so Jim, as a four-year-old,
00:09:06,780 --> 00:09:11,840 when he observes they need to go to the gas station to fill up the tank, he throws out the idea,
00:09:12,080 --> 00:09:18,180 well, let's just use only half the gas in the tank because then we'll still be able to after
00:09:18,180 --> 00:09:21,940 that only use half the gas in the tank. And, you know, the funny thing that doesn't occur to a
00:09:21,940 --> 00:09:24,780 four-year-old is, well, then we're just not going to get very far.
00:09:24,780 --> 00:09:31,520 So Jim's dream is to go to MIT down the street in Cambridge and study math. He graduates high
00:09:31,520 --> 00:09:36,580 school in three years. And during the second semester of Jim's freshman year there, he
00:09:36,580 --> 00:09:42,360 enrolls in a graduate math seminar on abstract algebra. So pretty, you know, heady stuff.
00:09:42,960 --> 00:09:48,580 Yeah. And Jim would go on to finish his undergrad at MIT in three years and get a master's in one
00:09:48,580 --> 00:09:54,640 year. Yeah. Pretty, pretty smart. But it turns out that that freshman year grad seminar,
00:09:54,640 --> 00:10:01,020 he took, actually has a big impact on him because he doesn't do well in the class. He can't keep up.
00:10:01,600 --> 00:10:08,220 And Jim's pretty self-aware here. There are other people at MIT who never run into problems.
00:10:08,520 --> 00:10:14,840 They never hit a limit. They never struggle understanding any concept. And he realizes that,
00:10:15,200 --> 00:10:19,480 oh, I'm smart. I'm very, very smart. I'm smarter than most other people here.
00:10:19,760 --> 00:10:24,000 But I'm not one of those people. Right. Which is, you know,
00:10:24,000 --> 00:10:24,620 what do you do with that?
00:10:24,640 --> 00:10:29,780 You realize you have to add a few of your skills together to become the best at something. You have
00:10:29,780 --> 00:10:34,480 to be smart and something else. Yes. So Jim's own words on this are,
00:10:34,620 --> 00:10:38,580 I was a good mathematician. I wasn't the greatest in the world, but I was pretty good.
00:10:38,920 --> 00:10:41,820 But he recognizes, like you said, Ben, that he has a different advantage
00:10:41,820 --> 00:10:47,580 that most of the super geniuses lacked. And that's that, as he put it, he had good taste.
00:10:47,920 --> 00:10:53,760 So these are his words. Taste in science is very important. To distinguish what's a good problem
00:10:53,760 --> 00:10:54,200 and what's a problem that's not, that's not true. And that's the best advantage that Jim has.
00:10:54,200 --> 00:10:59,140 what's a problem that no one's going to care about the answer to anyway. That's taste. And I think I
00:10:59,140 --> 00:11:06,100 have good taste. By the way, this is exactly the same thing as Jeff Bezos in college realizing he
00:11:06,100 --> 00:11:12,120 wanted to be a theoretical physicist. He met some of the extreme brainpower people that would go on
00:11:12,120 --> 00:11:16,240 to become the best theoretical physicist in the world. And he said, I'm smart, but I'm not that
00:11:16,240 --> 00:11:23,760 smart. And so switch to computer science. I think the analogy here is like sports. There are all
00:11:23,760 --> 00:11:30,700 star players. There are Hall of Famers. And then there's LeBron and MJ. And Jim ends up being a
00:11:30,700 --> 00:11:35,560 Hall of Famer mathematician, but he's not Tom Brady. I mean, he's got a pretty important theorem
00:11:35,560 --> 00:11:39,880 named after him. That goes on to become a foundation of string theory and physics, which
00:11:39,880 --> 00:11:46,520 isn't even Jim's field. Crazy. So this realization that Jim has about himself, though, both that he's
00:11:46,520 --> 00:11:52,680 not the smartest person in the room at a place like MIT, but he can hang with them and that he
00:11:52,680 --> 00:11:53,400 has this.
00:11:53,760 --> 00:12:00,760 Taste concept, I think, becomes one of the most important keys to the secret sauce that ends up
00:12:00,760 --> 00:12:07,460 getting built at Rentech, which is that he can relate to everybody. He understands what's going
00:12:07,460 --> 00:12:11,820 on. Any person off the street probably couldn't even really have a conversation with these folks,
00:12:12,340 --> 00:12:18,340 but he can. And yet he also has the perspective. Maybe some of this is from his grandfather of
00:12:18,340 --> 00:12:23,740 what is important out there in the real world. And as a result, all of his friends at MIT,
00:12:23,760 --> 00:12:30,020 and these super smart people, they look up to him because you aren't like the kid in the corner at
00:12:30,020 --> 00:12:34,020 the high school dance. You're cool. He's the extroverted theoretical mathematician.
00:12:34,740 --> 00:12:40,660 Yes. So he was elected class president in high school. You know, he smokes cigarettes. He's
00:12:40,660 --> 00:12:46,060 popular with the ladies. He kind of looks like Humphrey Bogart. He's a popular dude,
00:12:46,360 --> 00:12:50,180 especially at this point in time. We're now in the late 50s when Jim's at MIT.
00:12:50,180 --> 00:12:53,740 You know, this is kind of James Dean rebel without a cause.
00:12:53,760 --> 00:13:02,100 Yep. So after graduation, Jim leads his buddies on a road trip with motor scooters.
00:13:02,240 --> 00:13:07,600 You can't make this stuff up from Boston down to Bogota, where one of his classmates is from.
00:13:07,820 --> 00:13:11,980 The idea is that they're going to do something so epic that the newspapers are going to have
00:13:11,980 --> 00:13:17,460 to write about it. So they all load up on scooters and drive down to Bogota. They get
00:13:17,460 --> 00:13:21,920 into all sorts of adventures. There's knives and guns and they get thrown in jail.
00:13:22,100 --> 00:13:23,540 It's honestly crazy that,
00:13:23,760 --> 00:13:25,620 this group of people took this type of risk.
00:13:26,240 --> 00:13:30,420 Totally crazy. So after he's done at MIT and after the road trip,
00:13:31,100 --> 00:13:36,460 Jim heads out to Berkeley in California so that he could do his PhD with the professor
00:13:36,460 --> 00:13:43,280 Xingxian Chern. And much later in life, Jim would collaborate with Chern for the Chern-Simons
00:13:43,280 --> 00:13:46,740 theory that we talked about earlier that becomes one of the foundational parts of
00:13:46,740 --> 00:13:52,700 string theory in physics. But before Jim leaves for the West Coast, he meets a girl in Boston
00:13:52,700 --> 00:13:53,740 and they do a little bit of a talk. And she's like, I'm going to do a little bit of a talk.
00:13:53,740 --> 00:14:01,540 So they decide to get engaged in four days. I mean, this is, this is him back then. These
00:14:01,540 --> 00:14:09,600 were the times. And when they get to California and they get married, Jim takes the $5,000 wedding
00:14:09,600 --> 00:14:15,520 gift that I believe they got from her parents. And he decides, I want to multiply this. So he
00:14:15,520 --> 00:14:20,780 starts driving from Berkeley into San Francisco every morning to go hang out at the Merrill Lynch
00:14:20,780 --> 00:14:23,560 brokerage office and just be a rat.
00:14:23,560 --> 00:14:23,720 And he's like, I'm going to do a little bit of a talk. And he's like, I'm going to do a little bit of a talk.
00:14:23,720 --> 00:14:27,180 So they start driving around the brokerage and find ways to trade and turn this money into
00:14:27,180 --> 00:14:30,900 something more. Which is so interesting to think about because at that point in time,
00:14:31,040 --> 00:14:35,560 there was such an advantage to just being there. This wasn't even the trading floor,
00:14:35,860 --> 00:14:40,280 but information is all so manual and also relationship driven in the markets
00:14:40,280 --> 00:14:44,720 that there was basically no way to be in on the action unless you were physically there
00:14:44,720 --> 00:14:50,020 in on the action. Exactly. Yeah. You couldn't just log into Yahoo finance or something or
00:14:50,020 --> 00:14:53,540 open the stocks app on your iPhone, which even the information they were
00:14:53,540 --> 00:14:58,200 getting was God knows how long delayed from New York or from Chicago for the futures and
00:14:58,200 --> 00:15:02,820 commodities that are being traded that Jim gets into. He's as close to the action as he can
00:15:02,820 --> 00:15:09,120 possibly be, but he's a long, long way from the action. Yep. Nonetheless, when he starts out
00:15:09,120 --> 00:15:15,800 doing this, Jim hits a hot streak and he goes up 50% in a few days. Trading is easy.
00:15:16,200 --> 00:15:22,660 Trading is easy. He says, I was hooked. It was kind of a rush. I bet. Except he ends up losing
00:15:22,660 --> 00:15:27,340 all of his profits just as quickly. Yeah. Important to learn that lesson early.
00:15:27,860 --> 00:15:33,020 Yes. And also right around this time, Barbara, his wife, gets pregnant with their first child
00:15:33,020 --> 00:15:38,180 and is like, you can't be driving into San Francisco every morning at gambling our future
00:15:38,180 --> 00:15:44,500 like this. Right. Effectively playing the ponies. Yeah, exactly. So Jim's like, okay, okay. I'll
00:15:44,500 --> 00:15:50,780 stop. I'll focus on academia for now. So he finishes his PhD in two years. They come back
00:15:50,780 --> 00:15:52,500 to Boston and he joins MIT.
00:15:52,660 --> 00:15:59,300 As a junior professor at age 23. So they stay one year in Boston, but Jim, even though he's got a
00:15:59,300 --> 00:16:05,240 family, even though he's super successful as a young academic here, he's got kids, he's restless.
00:16:05,700 --> 00:16:11,000 So one of his buddies from the scooter trip to Bogota is from Bogota and lives there. His family's
00:16:11,000 --> 00:16:17,160 there. He has an idea to start a flooring tile manufacturing company because he's like, you know,
00:16:17,160 --> 00:16:21,800 the flooring at MIT and in Boston, it's so much nicer than a Bogota. We should start a company and
00:16:22,000 --> 00:16:22,640 make this.
00:16:22,660 --> 00:16:23,960 Same kind of flooring here.
00:16:24,320 --> 00:16:30,280 When I read this, I couldn't believe that this was Jim Simon's first business venture. Like it's so
00:16:30,280 --> 00:16:35,920 random, but it really is emblematic of just how much he was thrill-seeking and just looking for
00:16:35,920 --> 00:16:41,080 anything that was unexpected, different, exciting. He just gets bored fast.
00:16:41,760 --> 00:16:45,920 Totally. Not just is this the start of his entrepreneurial career,
00:16:46,500 --> 00:16:50,420 the seeds of this financially are what go on to start Rentech.
00:16:50,420 --> 00:16:51,080 It's wild.
00:16:51,660 --> 00:16:55,560 Totally wild. So Jim takes a year off and goes down to Bogota.
00:16:56,100 --> 00:17:03,960 This is a guy with an MIT undergrad and master's and a Berkeley PhD in theoretical math.
00:17:04,060 --> 00:17:06,020 Who's now a professor at MIT.
00:17:06,500 --> 00:17:09,600 Who is taking a year off to go work on a flooring company in Bogota.
00:17:10,160 --> 00:17:14,480 Yes. Accurate. So he does that for a year. They get it set up. He gets bored again. He's like,
00:17:14,500 --> 00:17:17,980 all right, I don't want to just run this company. I've helped set it up. I have an ownership stake
00:17:17,980 --> 00:17:20,320 in it now. He bounces back.
00:17:21,080 --> 00:17:25,360 To Boston, this time to Harvard as a professor there for a year.
00:17:25,560 --> 00:17:26,600 He's really racking them up.
00:17:27,100 --> 00:17:32,180 But he spends a year there and he's like, ah, got the itch again. And you know,
00:17:32,180 --> 00:17:37,360 the junior professor's salary isn't that much. And like we said about him back from his childhood
00:17:37,360 --> 00:17:41,780 days, he sees the appeal in being rich. And he's like, this is not a path to being rich.
00:17:43,760 --> 00:17:49,840 So he's like, I'm going to go put my skills out on the open market. He gets a job in Princeton,
00:17:50,000 --> 00:17:51,000 New Jersey, not at Princeton.
00:17:51,000 --> 00:17:51,060 Princeton, New Jersey, not at Princeton, New Jersey, not at Princeton, New Jersey, not at Princeton,
00:17:51,060 --> 00:17:59,880 University, but at the Institute for Defense Analyses, which is a nonprofit organization
00:17:59,880 --> 00:18:07,740 that consults exclusively for the U.S. government, specifically the Defense Department,
00:18:08,100 --> 00:18:14,260 and specifically the NSA. These are the civilian codebreakers.
00:18:14,680 --> 00:18:20,160 Yes. It was basically formed with this idea that one, across various branches of our government,
00:18:20,260 --> 00:18:20,980 we need to be able to do this. And so we're going to do this. And we're going to do this. And we're
00:18:20,980 --> 00:18:27,100 going to need better collaboration and crossfunding of the same initiatives. And two, there are going
00:18:27,100 --> 00:18:31,120 to be a lot of people who don't work for the government that we're going to want to hire to do
00:18:31,120 --> 00:18:36,640 some pretty secret work. Yep. So the IDA there in Princeton
00:18:36,640 --> 00:18:42,600 kind of functioned like the Institute for Advanced Study, which is also in Princeton.
00:18:42,740 --> 00:18:48,400 That's where Einstein went when he came to America, kind of an independent think tank research group,
00:18:48,400 --> 00:18:50,840 except it's solely focused on...
00:18:50,980 --> 00:18:55,200 Codebreaking and signal intelligence with the Russians during the Cold War.
00:18:55,660 --> 00:19:01,220 Yeah. It's a pretty wild charter, and especially how special of an organization it was. The way
00:19:01,220 --> 00:19:06,880 these people would spend their time is part codebreaking, but part kind of goofing around
00:19:06,880 --> 00:19:11,440 because the creativity of mathematicians working together on passion projects
00:19:11,440 --> 00:19:19,440 is important to discovering clever new algorithms. Yes. This is so, so key. And this culture
00:19:19,440 --> 00:19:20,960 ends up getting translated into a lot of other things. And I think it's a really important
00:19:20,980 --> 00:19:21,140 thing.
00:19:21,140 --> 00:19:21,160 Yeah.
00:19:28,380 --> 00:19:29,000 And I think it's a really important thing.
00:19:29,000 --> 00:19:34,960 They recruited top mathematicians and academics to come be codebreakers there. They would double
00:19:34,960 --> 00:19:35,700 their salaries.
00:19:36,260 --> 00:19:39,280 And importantly, it couldn't have been a government division if they were going to
00:19:39,280 --> 00:19:44,100 be doing that because there's very specific congressionally approved budgets for payroll.
00:19:44,460 --> 00:19:49,940 Exactly. They figured out that they needed to attract the smartest people in the world who
00:19:49,940 --> 00:19:50,860 weren't going to come just go...
00:19:50,860 --> 00:19:56,640 go work for the Department of Defense. This was the way to do it. So like you said, Ben,
00:19:57,160 --> 00:20:02,420 the charter of the group was that employees had to spend 50% of their time doing code breaking.
00:20:02,680 --> 00:20:08,480 But the other 50% of the time, they were free to do whatever they wanted, like research,
00:20:08,900 --> 00:20:14,860 pursue whatever they were doing in academia, publish papers. Kind of the appeal of going
00:20:14,860 --> 00:20:20,420 there was, hey, it's the same thing as being a professor at MIT or Princeton or Harvard or
00:20:20,420 --> 00:20:26,160 whatever, except you're doing code breaking instead of teaching. And there's no bureaucracy
00:20:26,160 --> 00:20:31,480 to worry about. There's no politics. It's just like, hey, you do your code breaking work and
00:20:31,480 --> 00:20:36,940 then you publish it. You can collaborate with your colleagues there. Now, this is pretty crazy.
00:20:37,520 --> 00:20:44,000 Very quickly after Jim arrives at IDA, remember, he's in moneymaking mode at this point in time.
00:20:45,340 --> 00:20:50,400 He recruits a bunch of his very brilliant colleagues to come work with him in
00:20:50,400 --> 00:20:58,440 their 50% free time on an idea to apply the same work and technologies that they're using in
00:20:58,440 --> 00:21:06,060 code breaking and signal intelligence to trading in the stock market. So they come together and
00:21:06,060 --> 00:21:12,260 they publish a paper called Probabilistic Models for and Prediction of Stock Market Behavior.
00:21:13,460 --> 00:21:19,840 And everything that they suggest in this paper really is rent tech. Just,
00:21:20,400 --> 00:21:24,740 20 years before rent tech. It's crazy. 1964, this was published?
00:21:25,260 --> 00:21:32,680 Yes. Now, at this point in time, fundamental analysis was then, as in most of the world
00:21:32,680 --> 00:21:38,220 today, still is the primary way of investing in things of, hey, I know this company. I'm
00:21:38,220 --> 00:21:43,060 going to analyze their revenues, their price multiple, or I'm going to think about what's
00:21:43,060 --> 00:21:48,520 happening in the currency markets or in the commodity markets and why copper is moving here
00:21:48,520 --> 00:21:50,360 or the British pound is moving there.
00:21:50,400 --> 00:21:52,260 I'm going to invest on those insights.
00:21:52,720 --> 00:21:56,300 You're effectively looking at the intrinsic value of an asset,
00:21:56,440 --> 00:21:59,620 trying to assign it a value and make investments based on that.
00:21:59,940 --> 00:22:08,840 Yes. Fundamental investing. There also existed in the 60s, technical investing, which kind of is
00:22:08,840 --> 00:22:16,880 voodoo. This is like, I'm looking at a stock chart and I've got a feeling that it's going to go up.
00:22:16,880 --> 00:22:20,120 I'm tracing this pattern and it's going up, baby. Or,
00:22:20,180 --> 00:22:20,380 no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no, no.
00:22:20,400 --> 00:22:22,020 This pattern is going down.
00:22:22,500 --> 00:22:24,740 Yeah. Using the phrase technical might be a little generous,
00:22:24,740 --> 00:22:31,640 but what they're looking for basically trying to mine trading behavior for signal about the way
00:22:31,640 --> 00:22:36,380 that it will trade in the future, rather than mining the intrinsic information about an asset
00:22:36,380 --> 00:22:40,740 for what you think it will do in the future. Right. And what Jim and his colleagues here
00:22:40,740 --> 00:22:50,180 suggesting is that, but just not really done by humans. It's that with a lot more data and a lot
00:22:50,180 --> 00:22:55,660 more sophisticated signal processing. And importantly, you might say, why is it this
00:22:55,660 --> 00:23:02,640 group of people that came to that conclusion of applying computational signal analysis to
00:23:02,640 --> 00:23:08,420 investing? Well, it's effectively the same thing as code breaking. You are looking for signal in
00:23:08,420 --> 00:23:14,060 the noise and trying to use computers and algorithms to mine signal from something that
00:23:14,060 --> 00:23:19,800 otherwise kind of looks random. Totally. When Jim started working on code breaking, I think he just
00:23:19,800 --> 00:23:24,500 looked right back to his experience trading in the markets and was like, whoa, this is the same thing.
00:23:24,740 --> 00:23:30,000 Which is not an insight other people had. That was the amazing thing about his background,
00:23:30,220 --> 00:23:35,180 priming him to realize that. Yes. There's all this noise in this data,
00:23:35,520 --> 00:23:38,800 and it is impossible for a human to sit here and look at this data and say,
00:23:39,060 --> 00:23:43,900 oh, I know what the Soviets are saying. No, no, you have to use mathematical models and
00:23:43,900 --> 00:23:49,500 statistical analysis to extract the patterns. So mathematical models, statistical analysis,
00:23:49,800 --> 00:23:56,260 we actually hear a lot of that in the world today because machine learning is a thing.
00:23:57,020 --> 00:24:04,760 Yes. What they are really doing here at IDA and then soon in Rentech is early machine learning.
00:24:05,480 --> 00:24:11,600 And Jim just had this incredibly brilliant insight that you can use these techniques and
00:24:11,600 --> 00:24:16,880 this technology for making investments, which makes this the perfect time to talk about our
00:24:16,880 --> 00:24:19,620 presenting sponsor for this season, JP Morgan Payments.
00:24:19,800 --> 00:24:25,340 Yes. The finance industry has a rich history of innovating, dating all the way back to the
00:24:25,340 --> 00:24:31,180 literal renaissance, where double-entry bookkeeping and letters of credit revolutionized global trade
00:24:31,180 --> 00:24:36,320 and economic development. And JP Morgan Payments really continues that tradition in their technology
00:24:36,320 --> 00:24:42,660 investments today. They move $10 trillion a day securely. That is a quarter of all U.S.
00:24:42,740 --> 00:24:48,780 dollar flows globally. Just think about the sheer volume of data at 5,000 transactions per second
00:24:48,780 --> 00:24:49,080 and how important it is to us. And that's what we're doing here at IDA.
00:24:49,080 --> 00:24:49,100 And that's what we're doing here at IDA. And that's what we're doing here at IDA.
00:24:49,100 --> 00:24:49,200 And that's what we're doing here at IDA. And that's what we're doing here at IDA.
00:24:49,200 --> 00:24:49,260 And that's what we're doing here at IDA. And that's what we're doing here at IDA.
00:24:49,260 --> 00:24:51,260 And how important that is to the global economy.
00:24:52,000 --> 00:24:56,020 Unsurprisingly, JP Morgan Payments has been in the AI game for years now.
00:24:56,420 --> 00:25:00,400 Similar to Rentech, they were also early to recognize the value of AI to gather,
00:25:00,860 --> 00:25:05,400 process, and analyze those massive troves of data to provide solutions for their customers
00:25:05,400 --> 00:25:09,800 and mitigate risk, like when they incorporated AI into their cash flow forecasting tool,
00:25:10,140 --> 00:25:14,500 which helps businesses manage liquidity. And that proved especially valuable during the pandemic.
00:25:14,620 --> 00:25:19,200 Yep. So also, unsurprisingly, JP Morgan was ranked number one in a recent global
00:25:19,200 --> 00:25:22,860 banking index of AI capabilities, with Fortune saying they were, quote,
00:25:22,860 --> 00:25:27,920 head and shoulders above the others. Their customers get AI-powered payment solutions
00:25:27,920 --> 00:25:33,220 for fraud prevention, customer insights, and treasury insights, all of which grows the bottom
00:25:33,220 --> 00:25:38,840 line. They can even analyze transaction data to predict and mitigate fraud patterns in real time
00:25:38,840 --> 00:25:43,860 with their validation services, helping stop millions of dollars for customers in attempted
00:25:43,860 --> 00:25:48,780 fraud. Yep. We were doing some research to prep for this segment, and we came across something
00:25:48,780 --> 00:25:55,740 pretty wild. The United States Treasury Department has started using AI to detect suspected check fraud
00:25:55,740 --> 00:26:03,200 and recovered over $375 million in 2023 utilizing the new tools. The U.S. Treasury Department
00:26:03,200 --> 00:26:08,380 disperses trillions of dollars annually, so if they continue to employ new technologies like this,
00:26:08,640 --> 00:26:13,120 it could really add up to the tune of billions. So how does this fit in? Well,
00:26:13,120 --> 00:26:18,680 the Treasury Department recently selected JP Morgan to provide account validation services
00:26:18,780 --> 00:26:24,240 for federal agencies. Obviously, payment integrity and this issue of improper payments is top of mind
00:26:24,240 --> 00:26:30,140 for them, and at enormous scale. So whether you are one of the largest institutions in the world
00:26:30,140 --> 00:26:35,680 or a small business like us here at Acquired, JP Morgan offers you peace of mind and protection.
00:26:36,500 --> 00:26:40,480 Yeah. One more playbook theme in common between Rentech and JP Morgan Payments,
00:26:40,680 --> 00:26:45,480 they both analyze data to uncover patterns and insights you may never think to look for.
00:26:45,480 --> 00:26:47,700 One of their clients, a furniture store,
00:26:48,780 --> 00:26:53,940 has a correlation with customers who also shop at pet stores where shoppers spent 76%
00:26:53,940 --> 00:26:59,120 more than the average customer when this was the case. So the furniture store launched a line of
00:26:59,120 --> 00:27:03,320 pet-friendly furnishings for that audience. These are the sorts of insights that drive
00:27:03,320 --> 00:27:07,760 growth with JP Morgan Payments as your partner. When it comes to payments technology, businesses
00:27:07,760 --> 00:27:12,800 of all sizes can benefit from having end-to-end AI-powered solutions that are constantly learning.
00:27:12,800 --> 00:27:17,600 And JP Morgan's API-first infrastructure across all aspects of treasury and payments
00:27:17,600 --> 00:27:18,760 is a one-stop shop.
00:27:18,780 --> 00:27:23,140 To learn more, check out jpmorgan.com slash acquired. And fun fact, listeners,
00:27:23,340 --> 00:27:29,200 it is Fraud Prevention Month. So listeners can learn even more by following JP Morgan on LinkedIn.
00:27:30,040 --> 00:27:36,600 Okay, David, so this paper is published. They're going to trade and make a whole bunch of money
00:27:36,600 --> 00:27:42,880 in the stock market by applying this code-breaking signal processing data analysis approach to
00:27:42,880 --> 00:27:43,600 investing.
00:27:43,860 --> 00:27:48,240 Yep. So then the natural question is, okay, what is the model here?
00:27:48,240 --> 00:27:52,880 How are they going to do this? And it turns out that one of the employees of IDA at this time,
00:27:52,960 --> 00:27:57,120 and one of the members of this sort of rebel group, shall we say, within the organization
00:27:57,120 --> 00:28:05,540 is a guy named Lenny Baum. And Lenny just happens to be the world expert in a mathematical concept
00:28:05,540 --> 00:28:12,640 called a Markov model, specifically a version of the Markov model called a hidden Markov model.
00:28:12,800 --> 00:28:18,140 Now, a Markov model is a statistical concept that's used,
00:28:18,240 --> 00:28:25,820 to model pseudo-random or chaotic situations. Basically, it says, let's abandon any attempt
00:28:25,820 --> 00:28:33,020 to actually understand what is going on in all of this data that we have, and instead just focus on
00:28:33,020 --> 00:28:39,940 what are the observable states that we can see of the situation? Can we identify different states
00:28:39,940 --> 00:28:47,560 that the situation is in? And if we just do that, can we predict future states based on what we've
00:28:47,560 --> 00:28:48,220 observed about the situation? And if we just do that, can we predict future states based on what we've observed about the situation?
00:28:48,240 --> 00:28:54,600 patterns of past states? And the answer to that is usually yes, even if you don't know anything
00:28:54,600 --> 00:28:59,520 about fundamentally how the system operates. So the great example that Greg Zuckerman gives
00:28:59,520 --> 00:29:06,660 in the book is, yes, a baseball game. There's three balls and two strikes. That state has a
00:29:06,660 --> 00:29:11,840 narrow set of states after it. It's going to be a strikeout. They're going to get on base. It's
00:29:11,840 --> 00:29:16,360 going to be a walk, or maybe they foul it off and it keeps going. There's only really a narrow set
00:29:16,360 --> 00:29:18,220 of things that could happen after that. Whereas, if you're going to do a strikeout, you're going to
00:29:18,240 --> 00:29:22,940 when it's zero balls and zero strikes, there's a lot that could happen. They could just keep
00:29:22,940 --> 00:29:26,020 pitching. And if you don't know the rules, you're like, why do they just keep pitching?
00:29:26,460 --> 00:29:30,100 And so it's this sort of great way to explain this idea of the black box that
00:29:30,100 --> 00:29:36,180 if nobody tells you the rules to the game by observing the outputs enough and observing,
00:29:36,480 --> 00:29:44,340 okay, in this state, these outputs are possible, you actually can kind of get pretty good at at
00:29:44,340 --> 00:29:47,960 least, if not predicting, understanding the probability distribution,
00:29:48,240 --> 00:29:51,240 of the outcomes for any given state in the game.
00:29:51,900 --> 00:29:59,380 So we brought up machine learning and AI a minute ago. This is a foundational concept to modern day
00:29:59,380 --> 00:30:04,500 AI. If you think about large language models and predicting what comes next, it's not like these
00:30:04,500 --> 00:30:10,760 large language models necessarily understand English. They're just really, really good at
00:30:10,760 --> 00:30:16,580 predicting states and the next state, i.e. characters and the next character or pixels and
00:30:16,580 --> 00:30:18,220 the next set of pixels or frame.
00:30:18,240 --> 00:30:23,880 And obviously, they're much fancier than that. But that is kind of the underpinning of it all. I
00:30:23,880 --> 00:30:28,440 mean, I remember in my sophomore year of college computer science class, I had a Markov chain
00:30:28,440 --> 00:30:33,660 assignment. And it was basically write a Java program to ingest this public domain book. And
00:30:33,660 --> 00:30:38,140 then I would give it a seed word, you know, the first word of each sentence and press return,
00:30:38,260 --> 00:30:42,780 return, return, return, return. And it would scan through the probability tree and give me the most
00:30:42,780 --> 00:30:47,900 probable word based on the corpus of the book that it just read to create some sentence. And it feels
00:30:47,900 --> 00:30:51,940 like magic. And of course, in these early rudimentary Markov chain things like the one I did in
00:30:51,940 --> 00:30:58,220 college, it kind of spits out nonsense. But that would evolve to be the LLMs that we know of today.
00:30:59,000 --> 00:31:04,180 Yes, totally. And that is what they were using at IDA to do code breaking. And that's what they
00:31:04,180 --> 00:31:08,200 propose in this paper that they could use in the stock market too.
00:31:08,900 --> 00:31:15,700 Exactly. And the way that this applies to investing is just like you might not know the rules of
00:31:15,700 --> 00:31:17,880 baseball, but if you've watched enough of it, you know, you're going to know the rules of baseball.
00:31:17,900 --> 00:31:23,920 You can kind of guess at what the probabilities of the next thing to happen are based on the state.
00:31:24,520 --> 00:31:28,020 Investing is kind of the same thing, or at least the stock market movements are where
00:31:28,020 --> 00:31:31,340 you don't know the future. You don't know what's going to happen. You don't know if
00:31:31,340 --> 00:31:36,900 stock X affects stock Y in some way, because you don't know in what way those companies do
00:31:36,900 --> 00:31:42,020 business together or who holds both stocks. Are they overlapping investors? Like you don't know
00:31:42,020 --> 00:31:47,560 the relationship between those companies. So you can't forecast with 100% certainty what is going
00:31:47,560 --> 00:31:47,880 to happen. And so you can't forecast with 100% certainty what is going to happen. And so you can
00:31:47,900 --> 00:31:53,440 However, if you suck in enough data about what has happened in the past and the probability
00:31:53,440 --> 00:31:58,340 distribution from every given state in the past, you probably could make some educated guesses
00:31:58,340 --> 00:32:04,500 or at least understand the probability of any individual outcome based on a state today of
00:32:04,500 --> 00:32:12,400 what could happen next. Yes, exactly. So Jim and Lenny and this whole little crew,
00:32:12,640 --> 00:32:17,180 they're pretty fired up. They're like, oh, great. Let's go.
00:32:17,180 --> 00:32:23,120 Let's go raise a fund and invest in the markets using this strategy.
00:32:23,760 --> 00:32:25,840 Certainly, we're going to be successful at raising that fund. And certainly,
00:32:25,940 --> 00:32:28,160 we're going to be very profitable because we've got this great idea.
00:32:28,720 --> 00:32:35,740 Totally. What could go wrong? Well, in the mid-60s, the idea that some wonky academics
00:32:35,740 --> 00:32:44,200 at some random secretive agency in Princeton, New Jersey could go raise money was non-viable.
00:32:44,200 --> 00:32:47,160 I mean, it was hard enough for Warren Buffett to,
00:32:47,180 --> 00:32:53,140 at this point in time, for his fund. And he was Benjamin Graham's anointed appointed disciple.
00:32:53,740 --> 00:32:59,540 And here are these academics who are working at some random unknown non-profit saying,
00:33:00,000 --> 00:33:03,720 give us money. We don't know anything about these companies that we're going to invest in.
00:33:03,900 --> 00:33:07,420 We don't know anything about fundamentals, but we've got a really good algorithm.
00:33:07,900 --> 00:33:11,960 People are probably like, what is an algorithm? So they just have no access to capital.
00:33:11,960 --> 00:33:17,000 Right. This was decades before it became high pedigree to come from a
00:33:17,000 --> 00:33:20,060 technical computer science background in the world of investing.
00:33:20,760 --> 00:33:27,160 Yes. So a bunch of kind of Keystone Cops style fundraising happens here. They're going around
00:33:27,160 --> 00:33:31,100 in secret. They're trying to keep the IDA bosses from knowing what they're doing.
00:33:31,920 --> 00:33:37,560 One of the group ends up leaving a copy of the investment prospectus on the copy machine
00:33:37,560 --> 00:33:43,500 at work one night, and the boss discovers it and calls them all into his office and is like,
00:33:43,500 --> 00:33:46,820 guys, what are you doing here? Right. It's a little bit of a clown show on the
00:33:46,820 --> 00:33:53,000 operational side, even if the idea is good. Yes. So they end up abandoning the effort,
00:33:53,160 --> 00:33:57,440 both because they can't raise money and because IDA has found out about this and they're not too
00:33:57,440 --> 00:34:03,680 pleased. Shortly after all of this, though, Jim ends up moving on anyway, because the Vietnam
00:34:03,680 --> 00:34:09,400 War starts. And he, as you can imagine from his background, is not a supporter of the Vietnam
00:34:09,400 --> 00:34:16,500 War at this point in time. Jim writes an op-ed in the New York Times denouncing the Vietnam
00:34:16,500 --> 00:34:20,540 War and saying like, yeah, he's, you know, sort of part of the Defense Department, but like not
00:34:20,540 --> 00:34:25,540 everybody in the Defense Department is for the war. Which is so naive, thinking you can write
00:34:25,540 --> 00:34:31,420 an op-ed in the New York freaking Times and that's not going to create issues for you in your job.
00:34:31,620 --> 00:34:37,340 Even more than that, amazingly, nobody really paid attention to it except a reporter at Newsweek
00:34:37,340 --> 00:34:43,280 who then comes to interview Jim and ask him some more questions. And he just doubles down on this.
00:34:43,280 --> 00:34:45,480 And when the Newsweek piece comes out,
00:34:45,800 --> 00:34:46,340 that's when the
00:34:46,340 --> 00:34:48,940 Department of Defense is like, all right, you got to fire this guy.
00:34:50,700 --> 00:34:58,400 Yeah. So Jim gets fired in 1967. Even though he's a star codebreaker, he made supposedly huge
00:34:58,400 --> 00:35:03,700 contributions to the group, which are still classified. But at age 30, with a wife and
00:35:03,700 --> 00:35:09,960 three kids, he's out on the street. And even though he's super smart, his colleagues love him
00:35:09,960 --> 00:35:15,880 clearly. He's now bounced out of MIT. He's bounced out of Harvard. He's gone to
00:35:15,880 --> 00:35:16,320 the University of New York. He's gone to the University of New York. He's gone to the
00:35:16,340 --> 00:35:16,500 University of New York. He's gone to the University of New York. He's gone to the
00:35:16,500 --> 00:35:21,780 This seemingly final home for him, great place at IDA. He gets bounced out of there, too.
00:35:22,740 --> 00:35:31,300 His job prospects are not great. Yeah. So he takes pretty much the only halfway decent paying job
00:35:31,300 --> 00:35:37,560 that he could get, which is to be the chair of the newly established or maybe reestablished
00:35:37,560 --> 00:35:45,540 math department at the State University of New York Stony Brook, which is the Long Island campus
00:35:45,540 --> 00:35:45,700 of the State University of New York. And he's got a job there. He's got a job there. He's got a job there.
00:35:45,700 --> 00:35:46,320 He's got a job there. He's got a job there. He's got a job there. He's got a job there. He's got a job there.
00:35:46,340 --> 00:35:55,880 University of New York. This is not Harvard. This is not MIT. No, it is not. But it did have one
00:35:55,880 --> 00:36:01,000 very important thing going for it, which is why Jim ended up there. And that is that Nelson
00:36:01,000 --> 00:36:06,920 Rockefeller, who was then the governor of New York, had launched a campaign, a hundred million
00:36:06,920 --> 00:36:15,080 dollar campaign to try and turn this Long Island campus of the State University of New York into a
00:36:15,080 --> 00:36:21,960 mathematical powerhouse, to become the Berkeley of the East. I sort of thought MIT was the Berkeley
00:36:21,960 --> 00:36:28,740 of the East already, but Rockefeller is waging a campaign that he wants Stony Brook to become
00:36:28,740 --> 00:36:36,820 a math and sciences powerhouse. And Jim is the key. He wouldn't be able to recruit somebody like Jim
00:36:36,820 --> 00:36:43,520 otherwise, but because he's now kind of tarnished his career, here's a very talented mathematician
00:36:43,520 --> 00:36:45,060 that they can convince to come back. And he's got a job there. And he's got a job there. And he's
00:36:45,080 --> 00:36:51,320 going to be chair of the department. Yep. So they basically give Jim an unlimited budget and leeway
00:36:51,320 --> 00:36:57,920 to go try and poach math professors from departments all over the country in the world and bring them
00:36:57,920 --> 00:37:03,600 there to Long Island. And part of how Jim goes and recruits folks is money, like the old, hey,
00:37:03,620 --> 00:37:11,480 I'll double your salary line. But the other part of it too, is he's given such leeway and Stony
00:37:11,480 --> 00:37:15,060 Brook is so different from the politics of an MIT or a university. And he's given such leeway to
00:37:15,080 --> 00:37:20,060 Harvard or Princeton. He says, hey, come here, I'll pay you more. But even more importantly,
00:37:20,680 --> 00:37:25,540 you can just focus on your research. You're not going to have to deal with committees. You're
00:37:25,540 --> 00:37:29,500 not going to have to do all this stuff. There is none of this stuff here. You might have to teach
00:37:29,500 --> 00:37:33,800 a little bit, but that's not even the point. Rockefeller doesn't want this necessarily become
00:37:33,800 --> 00:37:39,180 a great teaching institution. He just wants to assemble talent there. Yep. And amazingly,
00:37:39,640 --> 00:37:44,540 it works. Jim starts getting a bunch of great talent, including James Axe, who is a superstar
00:37:45,080 --> 00:37:51,640 at number theory from Cornell. And he ends up at Stony Brook recruiting and building
00:37:51,640 --> 00:37:55,180 one of the best math departments in the world. Amazing.
00:37:55,960 --> 00:38:02,040 Totally amazing. But in true Jim fashion, after a couple of years of this and also his marriage
00:38:02,040 --> 00:38:07,580 with Barbara falling apart, he starts getting restless again. He decides that he wants to go
00:38:07,660 --> 00:38:12,560 on a sabbatical and go back to Berkeley and reunite with his old advisor there and go
00:38:12,560 --> 00:38:14,360 spend some time out on the coast in California.
00:38:14,360 --> 00:38:19,160 Yeah. And this is where Chern and Simons end up collaborating and developing the Chern-Simons
00:38:19,160 --> 00:38:23,920 theory that ends up winning the highest award in geometry from the American Mathematical Society
00:38:23,920 --> 00:38:32,760 and really kind of is Jim's personal mark on mathematics. Yep. Now also, right around the
00:38:32,760 --> 00:38:40,900 same time, remember the Columbian Flooring Company? It gets acquired and Jim and his buddies who are
00:38:40,900 --> 00:38:47,540 partners in it come into a good amount of money. And Jim is newly divorced. He's restless in
00:38:47,540 --> 00:38:55,900 academia. He has these ideas back from when he was at IDA about what you could do in the markets
00:38:55,900 --> 00:39:03,260 if you had capital. He starts trading again and he gets more and more into it. Meanwhile, like we
00:39:03,260 --> 00:39:09,260 said, he's becoming disillusioned again and restless at academia. And in 1978, he leaves
00:39:09,260 --> 00:39:10,880 to focus full-time on mathematics. And he's like, I don't know, I don't know, I don't know, I don't
00:39:10,880 --> 00:39:15,360 know. I'm on trading, which is a huge shock to the academic community. Remember, he's assembled
00:39:15,360 --> 00:39:20,140 this superstar team there at Stony Brook. There's a quote in Greg's book from another mathematician
00:39:20,140 --> 00:39:25,420 at Cornell. We looked down on him when he did this, like he had been corrupted and had sold
00:39:25,420 --> 00:39:30,980 his soul to the devil. Yeah. I mean, it was really viewed in the math community as anyone who's going
00:39:30,980 --> 00:39:36,000 to do investing is throwing away their talent. And it wasn't even that it was common the way
00:39:36,000 --> 00:39:40,500 that it sort of is today. Right. Jim was the first one. But the idea that you would leave to do
00:39:40,500 --> 00:39:40,860 anything.
00:39:40,880 --> 00:39:48,380 Commercial. You're doing a disservice to humanity. Yes, exactly. And leaving to do anything. Sure. But
00:39:48,380 --> 00:39:53,280 leaving to do investing was almost just seen as dirty. Like it's this rich person's game that
00:39:53,280 --> 00:39:59,200 provides no value to society. Right. Yeah. I don't think it was that the rest of the math world was
00:39:59,200 --> 00:40:03,220 skeptical that it could work. They probably were like, oh, yeah, this could work. But they were
00:40:03,220 --> 00:40:09,660 like, ew. Academics tend to be much more motivated by prestige than money. So I could totally see
00:40:09,660 --> 00:40:13,320 this other people being like, oh, I could do that if I wanted. But I have this higher calling and
00:40:13,320 --> 00:40:17,840 everyone respects me for this higher calling. And my currency is the papers I publish and the awards
00:40:17,840 --> 00:40:22,680 that I win. And that's what I want. Yep. Now, Stony Brook, we should say, too, like it's a very nice
00:40:22,680 --> 00:40:28,460 place. Yes. But it's in the middle of Long Island on the North Shore. This is not the Hamptons.
00:40:28,820 --> 00:40:34,540 It's like the Long Island suburbs. Yep. The wooded Long Island suburbs. Yes. The wooded
00:40:34,540 --> 00:40:38,940 Long Island suburbs. Here's Jim in a strip mall next to a pizza joint,
00:40:38,940 --> 00:40:44,380 setting up his trading operation that he decides very cleverly to call monometrics,
00:40:44,880 --> 00:40:53,820 a combination of money and metrics or econometrics. And he recruits his old IDA buddy,
00:40:54,540 --> 00:41:03,660 original partner in crime on the trading idea, Lenny Baum, to come and join him. And this time,
00:41:03,740 --> 00:41:08,540 though, they have some capital from the sale of the flooring company. And how much did he make on
00:41:08,940 --> 00:41:16,600 sale? I think together with Jim, his partners and whatever money Lenny put in, they had a little
00:41:16,600 --> 00:41:24,220 less than four million dollars in this initial capital in 1978. Yep. Now, Jim also has another
00:41:24,220 --> 00:41:29,980 advantage at this point in time, which is he's right down the street from Stony Brook and he's
00:41:29,980 --> 00:41:36,660 just recruited all of these superstar mathematicians. The table has been set. Yes. And those
00:41:36,840 --> 00:41:38,880 folks are more loyal to Jim than they are.
00:41:38,880 --> 00:41:45,260 But they're more loyal right now to academia than they are to finance. This is not a paved
00:41:45,260 --> 00:41:50,540 pathway until Jim paves this pathway. Yes, in general, but some of them, and in particular,
00:41:51,520 --> 00:41:58,060 the superstar James Axe, Jim convinces to come join him in his trading operations.
00:41:58,720 --> 00:42:06,340 So having Baum and Axe and Simons, it's like suddenly this extremely credible team in the
00:42:06,340 --> 00:42:07,900 world. Yes.
00:42:08,880 --> 00:42:13,140 Uncredible. Right. All the theorems that a lot of mathematicians are using every day
00:42:13,140 --> 00:42:16,580 are all named after these three guys who are now at the same firm trading.
00:42:17,120 --> 00:42:23,540 Yes. And it's led by Jim, who's somebody that they respect as an academic, but even more important
00:42:23,540 --> 00:42:28,800 is somebody they want to work for and they look up to and they think is cool. And he's out there
00:42:28,800 --> 00:42:35,520 being like, hey, I think we can make money. Right. Now, at this point, they're primarily
00:42:35,520 --> 00:42:38,440 trading currencies, not stocks.
00:42:38,880 --> 00:42:45,500 And currencies are obviously large markets, but they aren't impacted by as many signals and as
00:42:45,500 --> 00:42:51,560 many factors as stocks are, or really even slightly more complex commodities like, I don't
00:42:51,560 --> 00:42:56,080 know, soybeans or whatever. And it seemed to me like a lot of the trading of currencies they were
00:42:56,080 --> 00:43:02,540 doing was basically based on feelings that they had around how a central bank was acting. Like
00:43:02,540 --> 00:43:06,380 if the head of state of a certain country was going to do something or not, it's basically like
00:43:06,380 --> 00:43:08,860 betting on how one single currency is going to do something or not. And it's basically like
00:43:08,880 --> 00:43:15,680 who was in control of currencies at governments would act. So to your point about very few signals
00:43:15,680 --> 00:43:19,480 impacting price, it's knowing what one person is going to do.
00:43:20,120 --> 00:43:25,020 Yes. And this is super important. At the end of the day, they build some models there. They're
00:43:25,020 --> 00:43:32,820 getting the early versions and infrastructure and scaffolding of this quantitative approach set up.
00:43:33,360 --> 00:43:37,820 But in terms of the actual trades they're putting on, they're still doing all of it by hand.
00:43:37,820 --> 00:43:38,320 And they're still doing it by hand.
00:43:38,320 --> 00:43:38,860 And they're still doing it by hand.
00:43:38,880 --> 00:43:45,220 They're still all really going on a fundamental type analysis. They'll take some signals from
00:43:45,220 --> 00:43:48,940 the model. They'll see it's interesting what they spit out, but they're not going to act on
00:43:48,940 --> 00:43:55,000 anything unless they can be like, oh yeah, I see what is going on here. I have a hypothesis.
00:43:55,660 --> 00:43:58,980 Right. The computers are by no means running loose at this point.
00:43:59,320 --> 00:44:04,880 By no means at all. Yeah. They're just suggesting patterns and ideas. And Jim and Lenny and James,
00:44:04,880 --> 00:44:08,440 they have to then decide, hey, are we going to do this or not? Or are we going to do something
00:44:08,440 --> 00:44:11,000 just totally different that we think is what's going to happen?
00:44:11,340 --> 00:44:11,480 Yep.
00:44:12,100 --> 00:44:20,180 And this actually does make sense, really for two reasons. One, computers and computing power just
00:44:20,180 --> 00:44:29,220 wasn't sophisticated enough yet to really build AI in a way that's powerful enough that it could
00:44:29,220 --> 00:44:35,300 work well enough you could really trust it. That's one part. The other part is these folks
00:44:35,300 --> 00:44:38,260 are mathematicians. They're not computer scientists.
00:44:38,440 --> 00:44:39,040 Right.
00:44:39,460 --> 00:44:47,020 And they're really, really good at building models, decoding signals, obviously, but they're
00:44:47,020 --> 00:44:52,640 much more from this realm of theory. And I actually spoke with Howard Morgan, who's going to come up
00:44:52,640 --> 00:44:57,140 here in a second, and he made this point to me. He's like, in math, there's this concept of
00:44:57,140 --> 00:45:02,620 traceability that's a really, really important cultural tenet. It's like proving a proof or
00:45:02,620 --> 00:45:08,420 proving a theorem or something like that. You really need to understand what's going on. And
00:45:08,420 --> 00:45:13,220 why to get ahead in the field. It's not like you can just say, oh, hey, the data suggests this. It's
00:45:13,220 --> 00:45:17,260 like, no, no, no, you need proof. And that's the world that these guys are coming from. They're
00:45:17,260 --> 00:45:22,880 like, oh, we can use data to sort of help us here. But ultimately, we want to have a rock solid
00:45:22,880 --> 00:45:28,320 theory of what is fundamentally happening here. Fascinating, which is very different than we'll
00:45:28,320 --> 00:45:32,060 cram a huge amount of data in and then whatever the data suggests, we know it's true because the
00:45:32,060 --> 00:45:37,160 data suggests it, which is sort of where they would end up many years later once they had both
00:45:37,160 --> 00:45:38,400 the hardware you're referring to.
00:45:38,400 --> 00:45:43,160 Sophisticated computers, the clean data that would be required to make all of those
00:45:43,160 --> 00:45:49,560 incredibly numerous and fast calculations, and also the real computer engineering architecture
00:45:49,560 --> 00:45:55,880 to build these scale systems to actually act on large amounts of signals and understand them all
00:45:55,880 --> 00:46:00,880 to come up with results. They just didn't have any of that at the time. So it was hunches and
00:46:00,880 --> 00:46:08,160 chalkboards. Yes. And so much so that even Jim is ringleader here. He's far from convinced that he
00:46:08,160 --> 00:46:11,880 should put all of his wealth into this thing. He's like, oh, yeah, this is interesting. We're
00:46:11,880 --> 00:46:17,260 building. We're experimenting like great. But I also want to put my money somewhere else, too,
00:46:17,320 --> 00:46:23,020 for some diversification. So this is where Howard Morgan comes in. You know, we used to talk about
00:46:23,020 --> 00:46:27,840 this on old acquired episodes that in the early days of Silicon Valley, there were only 10 people
00:46:27,840 --> 00:46:31,280 out here and they all knew each other and they were all doing the same thing.
00:46:31,820 --> 00:46:37,920 This was also the case in East Coast finance and technology and early VC in these days.
00:46:38,160 --> 00:46:41,800 Howard Morgan would go on to be one of the co-founders of First Round Capital.
00:46:42,520 --> 00:46:47,060 Which was essentially spun out of Renaissance. Like it was kind of the venture capital
00:46:47,060 --> 00:46:51,220 work that they were doing at Renaissance that didn't fit with the rest of Renaissance.
00:46:51,940 --> 00:46:56,260 Yes. So here's how it all went down. And this is so poorly understood out there.
00:46:56,400 --> 00:47:02,920 Yes. Howard was a computer science and business school professor at the University of Pennsylvania.
00:47:02,920 --> 00:47:07,040 So he taught CS at Penn and business at Wharton.
00:47:08,160 --> 00:47:14,180 And he had been involved in bringing ARPANET to Penn and was kind of like early, early
00:47:14,180 --> 00:47:22,500 internet pioneer. And so as a result, he was super plugged into tech and early startups
00:47:22,500 --> 00:47:29,000 and really early, early proto internet stuff. And Jim gets excited about investing together
00:47:29,000 --> 00:47:35,320 with Howard. So they say like, hey, maybe we should partner together. And in 1982,
00:47:36,120 --> 00:47:38,140 Jim actually winds down Monomath. He's like, I don't know. I don't know. I don't know. I don't
00:47:38,140 --> 00:47:45,460 And he and Howard co-found a new firm together that's going to reflect both of their backgrounds
00:47:45,460 --> 00:47:51,740 and be a great diversification. Jim and his group are going to bring in the quantitative trading
00:47:51,740 --> 00:47:58,180 thing. And again, trading on currencies and commodities at this point. And Howard's going
00:47:58,180 --> 00:48:06,080 to bring in private company technology investing. And they pick a name for a firm that is going to
00:48:06,080 --> 00:48:06,980 reflect this.
00:48:08,140 --> 00:48:10,120 It's crazy.
00:48:10,120 --> 00:48:12,760 And that is why Rentech is called Rentech.
00:48:13,260 --> 00:48:17,660 I could not, when we figured this out in the research, I could not believe that this is not a
00:48:17,660 --> 00:48:23,600 more widely understood story, that this is the origins of what is today a fantastic venture
00:48:23,600 --> 00:48:29,280 capital firm, first round capital. But you could not name two more different strategies
00:48:29,280 --> 00:48:36,140 in investing. I mean, a long-term illiquid thing like venture capital, highly speculative
00:48:36,140 --> 00:48:38,120 versus, you know, we're going to do this. We're going to do this. We're going to do this. We're
00:48:38,140 --> 00:48:43,160 going to trade whether we think the French franc is going to go up or down tomorrow based on the
00:48:43,160 --> 00:48:47,220 whim of some government leader. It's unbelievable these were under the same roof.
00:48:48,080 --> 00:48:51,160 Totally. But when you know the whole background in history, it kind of makes sense because
00:48:51,160 --> 00:48:57,540 this is their personal money. This is Jim and his buddies and Lenny and James and Howard.
00:48:58,080 --> 00:49:02,120 There's not institutional capital here. They're not out pitching LPs of like,
00:49:02,200 --> 00:49:07,440 oh, you should invest in my diversified strategy of currency trading and private technology startups.
00:49:07,440 --> 00:49:07,640 Yeah.
00:49:08,140 --> 00:49:10,640 Multi-strategy. This is really multi-strategy.
00:49:10,920 --> 00:49:16,780 Yeah. We'll get into what multi-strategy today means later. But in these early days of rent tech,
00:49:17,180 --> 00:49:23,840 50% of the portfolio was venture capital and 50% was currency trading. And in fact,
00:49:24,380 --> 00:49:30,160 a couple of years after they get started, the currency trading side of the firm almost blows
00:49:30,160 --> 00:49:36,800 up when Lenny goes super long on government bonds and the market goes against him and the whole
00:49:36,800 --> 00:49:38,080 portfolio drops 40%.
00:49:38,140 --> 00:49:46,900 Which is wild. That ends up triggering a clause in Lenny's agreement with Jim and they sell off
00:49:46,900 --> 00:49:51,860 Lenny's entire portfolio and he leaves the firm. This is crazy. I mean,
00:49:52,280 --> 00:49:56,560 blow up risk is always an issue in the markets, but this happened to rent tech.
00:49:57,200 --> 00:50:00,060 And because we quickly got to this point in the story, it would be easy to say,
00:50:00,140 --> 00:50:04,760 well, that's a clause that has a lot of teeth. There were many sort of rumbles of something
00:50:04,760 --> 00:50:08,060 like this potentially happening. Simon's going to Lenny.
00:50:08,140 --> 00:50:13,660 Hey, maybe we should cut some of our losses and it's okay to trade out of these positions. And
00:50:13,660 --> 00:50:18,460 Lenny was just very dug in on I'm a true believer. And that's how you can get into a situation where
00:50:18,460 --> 00:50:24,360 you trigger a covenant like this. Totally. And again, also shows they weren't doing model-based
00:50:24,360 --> 00:50:30,900 quantitative trading really at this point in time. No, so much gut. So as a result of that,
00:50:31,180 --> 00:50:37,980 for a while, rent tech is truly almost entirely a venture capital firm. At one point on,
00:50:38,140 --> 00:50:42,360 the venture side, just one investment, Franklin Dictionaries. Do you remember Ben?
00:50:42,540 --> 00:50:42,800 Yes.
00:50:43,120 --> 00:50:45,960 Franklin Electronic Dictionaries. Yeah. That was one of their biggest investments.
00:50:46,520 --> 00:50:50,000 That one investment is half of Jim's net worth.
00:50:50,360 --> 00:50:50,800 What?
00:50:51,180 --> 00:50:53,360 At this low point for the trading side. Yes.
00:50:54,000 --> 00:50:55,560 I had no idea. That's crazy.
00:50:56,280 --> 00:51:02,260 Yeah. So in the book, Greg talks about, oh, Jim was focused on venture capital and that's kind
00:51:02,260 --> 00:51:05,500 of the story out there. It's like, well, he was focused on venture capital because that was the
00:51:05,500 --> 00:51:07,120 only thing working and making money.
00:51:07,260 --> 00:51:07,740 Yeah.
00:51:08,140 --> 00:51:12,080 The only thing where they actually had an edge from Howard's access to deal flow,
00:51:12,200 --> 00:51:14,860 because they certainly didn't have an edge in the global currency markets.
00:51:15,720 --> 00:51:21,900 So I think perhaps in part because of the trading losses, James Axe starts to get a
00:51:21,900 --> 00:51:27,400 little disillusioned too. And he tells Jim that he wants to move out to California with Sandor
00:51:27,400 --> 00:51:31,860 Strauss, who started working with them at this point. Sandor was another Stony Brook alum that
00:51:31,860 --> 00:51:36,880 joined them. And the two of them want to move out to California and do trading out there.
00:51:37,580 --> 00:51:38,100 Yeah.
00:51:38,100 --> 00:51:38,120 Yeah.
00:51:38,120 --> 00:51:43,800 And Jim says, sir, fine. I'm here with Howard. I'm doing venture capital stuff. Why don't you go
00:51:43,800 --> 00:51:50,580 move out to California? You can start your own firm, which they do. It's called Axe Com, A-X-C-O-M.
00:51:50,980 --> 00:51:58,240 And we'll contract with Axe Com to run what's left of the trading operations here for Rentech.
00:51:59,060 --> 00:52:04,700 So it's this interesting arm's length thing where Jim strikes a deal where he's going to own a part
00:52:04,700 --> 00:52:07,980 of Axe Com in exchange for this very favorable deal.
00:52:07,980 --> 00:52:11,580 Favorable contractual relationship where they're going to hire them to be the manager
00:52:11,580 --> 00:52:16,920 for this pot of money that Renaissance has raised. But, you know, it's technically not
00:52:16,920 --> 00:52:18,260 Renaissance. It's Axe Com.
00:52:18,700 --> 00:52:22,140 Right. It's another company that is now doing the quantitative trading.
00:52:22,480 --> 00:52:25,000 Yep. And I think Jim owned a quarter of it. Is that right?
00:52:25,500 --> 00:52:26,140 Yes, that's right.
00:52:26,640 --> 00:52:31,380 And importantly, I don't think anyone had any idea what Axe Com would become
00:52:31,380 --> 00:52:35,260 or how unbelievably profitable it would be.
00:52:35,960 --> 00:52:37,040 Uh, no.
00:52:37,040 --> 00:52:41,540 Nobody would have done what they did had they known what was coming.
00:52:42,140 --> 00:52:43,280 Yes. Wouldn't have spun it out.
00:52:43,720 --> 00:52:52,640 No. So once Axe and Strauss get out to California, Strauss, he's kind of on the computing data
00:52:52,640 --> 00:52:56,840 infrastructure side. That's what he was doing at Stony Brook. And that's what he came into
00:52:56,840 --> 00:53:02,940 Renaissance to build. He starts getting really into data and he starts collecting
00:53:03,520 --> 00:53:06,080 intraday pricing movements on securities.
00:53:06,080 --> 00:53:13,920 At this point in time, I think really the best data you could get from providers out there was
00:53:13,920 --> 00:53:21,340 maybe open and close data on securities pricing. Strauss finds a way to get tick data,
00:53:21,700 --> 00:53:26,780 like every 20 minute data on these securities throughout the day.
00:53:27,540 --> 00:53:32,580 Not only that, he's getting historical data that predates what your traditional data providers
00:53:32,580 --> 00:53:35,880 would give you and then ingesting it into computers.
00:53:36,080 --> 00:53:40,860 And cleaning the data to get it into the same format as the tick data. So he's getting early
00:53:40,860 --> 00:53:47,580 1900s, even 1800s stuff to try to just say, at some point, hopefully we'll be able to make use
00:53:47,580 --> 00:53:52,020 of this. And I want to have this just really, really clean data set about the way that these
00:53:52,020 --> 00:53:57,520 markets interact. Yeah. I mean, he's doing ETL on the data. Yes. I think before anybody knew
00:53:57,520 --> 00:54:02,400 what ETL was. Again, no one told him to do that. That was just a self-motivated, almost like
00:54:02,400 --> 00:54:05,980 obsession of like, well, if we're going to have data, it should be well formatted. And,
00:54:05,980 --> 00:54:11,100 well understood and labeled and all that. So that's one thing that happens. The other thing
00:54:11,100 --> 00:54:16,440 is Jim says, oh, you're going out to California. Let me hook you up with my buddy. Who's a Berkeley
00:54:16,440 --> 00:54:25,060 professor out there, Elwin Burlakamp. And Burlakamp had studied with folks like John Nash
00:54:25,060 --> 00:54:29,920 and Claude Shannon at MIT. I love that Claude Shannon is coming in again.
00:54:30,080 --> 00:54:34,560 I know. We talked about it a lot on the Qualcomm episode, father of information theory,
00:54:35,000 --> 00:54:35,960 really this.
00:54:35,980 --> 00:54:41,120 Center of gravity for attracting tons of talent to MIT and kind of paving the way for what would
00:54:41,120 --> 00:54:46,500 become phone technology and telecommunications broadly in the future. But the fact that
00:54:46,500 --> 00:54:51,180 Burlakamp is crossing paths at MIT with Claude Shannon, so cool.
00:54:51,800 --> 00:54:58,080 So cool. And most importantly for this specific use case, Burlakamp had worked with John Kelly,
00:54:58,280 --> 00:55:04,580 who developed the Kelly criterion on bet sizing, which poker players will likely be well familiar
00:55:04,580 --> 00:55:05,020 with.
00:55:05,360 --> 00:55:05,560 Yep.
00:55:05,980 --> 00:55:10,540 So with this combination now of much, much, much better and deeper data from Strauss
00:55:10,540 --> 00:55:14,620 and Burlakamp coming in and working with Axe on the models and saying,
00:55:14,840 --> 00:55:21,100 hey, we should be smart about the bet sizing that we're doing in the trades that are coming out of
00:55:21,100 --> 00:55:25,140 these models versus I don't know what they were doing before. Maybe it was naive of like every
00:55:25,140 --> 00:55:30,860 trade was the same or just like we should actually be systematic about this. The models start
00:55:30,860 --> 00:55:32,200 really working.
00:55:32,780 --> 00:55:34,700 Yep. This is the turning point.
00:55:35,140 --> 00:55:35,620 Yeah.
00:55:35,980 --> 00:55:42,720 In these kind of mid-80s years, Axe Com is generating IRRs of like 20 plus percent on the
00:55:42,720 --> 00:55:48,400 trading side. You know, not necessarily going to beat venture capital IRRs, but liquid.
00:55:48,640 --> 00:55:48,960 Yes.
00:55:49,200 --> 00:55:49,640 Reliable.
00:55:50,200 --> 00:55:54,220 Well, that's the thing. They don't know how reliable yet. They know they've done it kind
00:55:54,220 --> 00:55:59,160 of a few years in a row here. But the question is how uncorrelated to the stock market over a
00:55:59,160 --> 00:56:04,980 long period of time and how predictable are these returns or is it just super high variance?
00:56:05,140 --> 00:56:05,740 Yes.
00:56:05,980 --> 00:56:13,560 But the early results are really good. And Jim and Burlakamp especially are very encouraged by
00:56:13,560 --> 00:56:21,140 this. So in 1988, Jim and Howard Morgan decide to spin out the venture investments and Howard goes
00:56:21,140 --> 00:56:29,000 to manage those with basically their own money. Fun coda on this. When Howard starts first round
00:56:29,000 --> 00:56:33,580 a number of years later with Josh Koppelman, Jim, of course, is a large LP.
00:56:33,820 --> 00:56:33,960 Ah.
00:56:34,860 --> 00:56:35,840 And Howard, of course, is a large LP.
00:56:35,840 --> 00:56:44,700 Howard, of course, remains an investor in Rentech. The first institutional fund that
00:56:44,700 --> 00:56:53,080 first round ended up raising was a 50X on $125 million fund. It had Roblox, Uber, and Square.
00:56:53,900 --> 00:57:00,280 So I believe this is right. I think Jim made as much money from his investments in first round
00:57:00,280 --> 00:57:04,100 as Howard did from his LP stake in Rentech.
00:57:04,720 --> 00:57:05,680 That's wild.
00:57:05,840 --> 00:57:07,000 Isn't that amazing?
00:57:07,700 --> 00:57:13,720 Wow. That is a untold story about Jim Simons. I think I read basically every primary source
00:57:13,720 --> 00:57:18,700 thing on Jim or Renaissance on the whole internet, but I assume you got that from Howard.
00:57:19,100 --> 00:57:23,120 Yeah. It was super fun talking to Howard about this and just the history of how first round
00:57:23,120 --> 00:57:27,360 started and early super angel investing and everything that became.
00:57:27,680 --> 00:57:33,220 I also didn't realize that first rounds fund one was a 50X on $125 million fund.
00:57:34,220 --> 00:57:35,660 First institutional fund?
00:57:35,840 --> 00:57:37,680 Which I believe they called fund two.
00:57:38,080 --> 00:57:40,900 I mean, wild, wild stuff.
00:57:41,420 --> 00:57:50,340 Totally wild. So when Howard spins out the venture activities, Jim then decides to set up a new
00:57:50,340 --> 00:57:57,660 fund as a joint venture between Rentech and Axcom. And they decide to name it after all of the
00:57:57,660 --> 00:58:04,720 collective mathematical awards that Jim and James and Burlakamp and all these prestigious mathematicians
00:58:04,720 --> 00:58:05,700 have won.
00:58:05,700 --> 00:58:09,840 in their careers. They name it the Medallion Fund.
00:58:10,460 --> 00:58:11,020 Ba-da-da.
00:58:11,600 --> 00:58:12,160 Ba-da-da.
00:58:12,580 --> 00:58:16,920 And listeners, we've arrived. This is the part of the story that matters. The Medallion Fund
00:58:16,920 --> 00:58:23,320 is the crown jewel, or you might even say actually the only interesting thing about Renaissance. And
00:58:23,320 --> 00:58:28,740 it is born out of this observation that, oh my God, what they're doing over there at Axcom is
00:58:28,740 --> 00:58:34,020 really interesting. Maybe they shouldn't be doing it all the way over there. Maybe that should be a
00:58:34,020 --> 00:58:35,680 deeper part of the fold here at Rentech.
00:58:35,680 --> 00:58:41,520 And we shouldn't have let that get away or frankly given up on the quantitative trading strategies too early.
00:58:41,520 --> 00:59:04,440 And again, still just currencies, still just commodities futures, not playing the stock market at all, but the seeds and the ideas, the huge amount of clean data, the robust engineering infrastructure to process all that data, the mining of signals from data to figure out what trading strategies to execute, that is really starting to form here in this moment.
00:59:04,440 --> 00:59:05,120 in this moment.
00:59:05,120 --> 00:59:05,620 in this moment.
00:59:05,620 --> 00:59:07,860 This new joint venture, this Medallion Fund.
00:59:08,020 --> 00:59:17,240 Those ideas had all existed before. This is the first time that it's all brought together and actually working and operationalized.
00:59:17,760 --> 00:59:21,760 And frankly, that computers got good enough to actually do it too. That's another big piece of this.
00:59:22,260 --> 00:59:29,300 Yeah, I don't know that Strauss could have done his data engineering too much earlier in time.
00:59:29,620 --> 00:59:29,840 Yeah.
00:59:30,240 --> 00:59:35,380 But before we get into the just absolutely insane run that this
00:59:35,620 --> 00:59:43,620 Medallion Fund is about to go on, that continues right through to this day, now is the perfect time for another story about ServiceNow.
00:59:44,160 --> 00:59:48,680 ServiceNow is one of our big partners here in Season 14 and is just an incredible company.
00:59:48,680 --> 00:59:49,240 Yep.
00:59:49,580 --> 00:59:58,780 ServiceNow digitally transforms your enterprise, helping automate processes, improve service delivery, and increase operational efficiency all in one intelligent platform.
00:59:59,140 --> 01:00:05,600 Over 85% of the Fortune 500 runs on them, and they have quickly joined the Microsofts and the NVIDIAs as one of the most powerful companies in the world.
01:00:05,600 --> 01:00:18,600 So we talked on our Novo Nordisk episode about how ServiceNow founder Fred Luddy discovered this core insight that software can transform and eliminate manual tasks.
01:00:19,120 --> 01:00:29,100 And on Hermes, we told the story of how current CEO Bill McDermott came in and turbocharged that into an absolute monster $150 billion market cap global behemoth.
01:00:29,380 --> 01:00:35,560 The key thread that connects those two eras is that from day one, Fred knew the ServiceNow platform.
01:00:35,600 --> 01:00:37,780 could be used across the whole enterprise.
01:00:38,200 --> 01:00:47,280 But at the same time, he also knew from his decades of prior software experience that launching a broad, horizontal offering right out of the gate as a startup was a recipe for failure.
01:00:47,800 --> 01:00:50,200 You need to start with a specific vertical use case.
01:00:50,560 --> 01:00:53,140 And in this case, he chose IT service management.
01:00:53,700 --> 01:00:53,780 Yep.
01:00:53,860 --> 01:00:55,680 And that's been true for us here on Acquired, too.
01:00:55,800 --> 01:01:04,460 David, if we didn't name it Acquired and cover technology acquisitions that actually went well, we never could have broadened and become the podcast that tells the stories of great companies.
01:01:04,460 --> 01:01:05,500 You can't just start.
01:01:05,600 --> 01:01:05,940 You can't just pass that.
01:01:06,300 --> 01:01:06,660 Totally.
01:01:07,160 --> 01:01:10,480 Well, this is what's so cool and where I think the playbook lesson really is for listeners.
01:01:10,800 --> 01:01:14,540 Because you can't just pick any use case, you have to be strategic about it.
01:01:14,540 --> 01:01:21,120 And IT was the perfect vertical because every other department has to interface with them from the CEO on down.
01:01:21,240 --> 01:01:24,840 So they're going to notice when IT service management rapidly improves.
01:01:25,060 --> 01:01:29,940 All of those support tickets that used to take forever are now just magically resolved.
01:01:30,100 --> 01:01:35,580 And that greases the wheels for the other departments to say, hey, maybe we should adopt ServiceNow to turbocharge.
01:01:35,600 --> 01:01:38,220 And digitally transform our service levels, too.
01:01:38,800 --> 01:01:38,900 Yep.
01:01:39,160 --> 01:01:46,100 Once those other departments do pull the trigger on joining the ServiceNow platform, who is in charge of rolling it out for them?
01:01:46,400 --> 01:01:51,180 Of course, it's IT, who are already true ServiceNow believers.
01:01:51,640 --> 01:01:55,980 I'm honestly not sure that there's a better enterprise software playbook in history than ServiceNow's.
01:01:56,560 --> 01:02:02,680 So once they established the beachhead in IT, they then took the same platform to HR with employee experience.
01:02:02,680 --> 01:02:05,500 They took it to CSM with customer service requests.
01:02:05,900 --> 01:02:09,800 They took it to finance with regulatory reporting, audit, and expense approvals.
01:02:10,200 --> 01:02:13,100 And now they're adding AI, which will take everything to the next level.
01:02:13,640 --> 01:02:13,740 Yep.
01:02:14,180 --> 01:02:19,780 So if you want to learn more about the ServiceNow platform and playbook and hear how it can transform your business,
01:02:20,140 --> 01:02:22,360 head on over to servicenow.com slash acquired.
01:02:22,440 --> 01:02:25,420 And when you get in touch, just tell them that Ben and David sent you.
01:02:26,300 --> 01:02:30,840 So they've got this grand new plan and vision with the Medallion Fund.
01:02:32,360 --> 01:02:34,940 Unfortunately, right out of the gate.
01:02:35,600 --> 01:02:37,340 The fund stumbles a bit.
01:02:38,280 --> 01:02:40,880 And Axe ends up getting burned out.
01:02:41,780 --> 01:02:43,560 Burlakamp, though, is like, no, no, no, no.
01:02:43,780 --> 01:02:44,760 This is an anomaly.
01:02:44,980 --> 01:02:46,160 Like, we're going to fix this.
01:02:46,420 --> 01:02:51,760 I really, really believe that what we're doing with these models is going to be extremely profitable.
01:02:52,280 --> 01:02:58,000 So he buys out most of Axe's stake in the summer of 1989.
01:02:58,920 --> 01:03:01,660 And he moves the offices up to Berkeley.
01:03:01,940 --> 01:03:05,580 And there he comes up with the idea that,
01:03:05,600 --> 01:03:11,180 hey, we should trade more frequently, a lot more frequently.
01:03:11,620 --> 01:03:15,720 Because if what we're trying to do is understand the state of the market from the data we have
01:03:15,720 --> 01:03:17,540 and then predict the future state of the market,
01:03:17,540 --> 01:03:21,760 and then combine that with figuring out the right bet sizing to make,
01:03:22,220 --> 01:03:26,460 we actually want to make a lot more trades to get a lot more data points
01:03:26,460 --> 01:03:31,080 and learn a lot more about the bets we're making so that we can then size them up or size them down.
01:03:31,300 --> 01:03:33,660 It's that and it's two other things.
01:03:33,780 --> 01:03:34,660 One is the further...
01:03:35,600 --> 01:03:39,100 The further into the future you look, the less certain you can be about it.
01:03:39,340 --> 01:03:42,080 If you know something is worth $10 right now,
01:03:42,460 --> 01:03:45,380 what you know five minutes from now is it's probably going to be worth about $10.
01:03:46,120 --> 01:03:49,440 The most likely situation is it's within 5% of that.
01:03:49,720 --> 01:03:53,420 If you ask me three years from now, I have almost no intuition about that.
01:03:53,740 --> 01:03:55,740 And a state machine is the same way.
01:03:55,900 --> 01:03:57,920 If you flash forward a whole bunch of states,
01:03:57,980 --> 01:04:01,800 you sort of lose predictability as you sort of continue down that chain.
01:04:01,980 --> 01:04:05,140 The second thing is, if your models are showing that,
01:04:05,140 --> 01:04:09,560 you're going to be right, call it something like 50.25% of the time,
01:04:10,100 --> 01:04:17,880 then the amount of money you can make is gated by the number of bets you can make at a quarter percent edge.
01:04:18,000 --> 01:04:23,080 If I walk up to the casino and I think I'm right about this particular roulette wheel,
01:04:23,180 --> 01:04:25,580 which of course you're not, 50.25% of the time,
01:04:25,600 --> 01:04:28,680 and I decide to play once or play twice or play five times,
01:04:29,180 --> 01:04:30,820 there's a chance I could lose all my money.
01:04:30,900 --> 01:04:34,620 Or if I have tiny little bet sizes, then I'm just not going to make that much money.
01:04:34,620 --> 01:04:34,680 Right?
01:04:34,680 --> 01:04:34,700 Right?
01:04:34,700 --> 01:04:34,720 Right?
01:04:34,720 --> 01:04:34,740 Right?
01:04:34,740 --> 01:04:34,760 Right?
01:04:34,760 --> 01:04:34,780 Right?
01:04:34,780 --> 01:04:34,840 Right?
01:04:34,840 --> 01:04:34,860 Right?
01:04:34,860 --> 01:04:34,880 Right?
01:04:34,880 --> 01:04:34,900 Right?
01:04:34,900 --> 01:04:34,920 Right?
01:04:34,920 --> 01:04:34,980 Right?
01:04:34,980 --> 01:04:35,120 Right?
01:04:35,140 --> 01:04:40,640 If I walk up to said game with a little bit of edge and I use small bet sizes and I play 10,000 times,
01:04:40,760 --> 01:04:42,380 I'm going to walk out with a lot of money.
01:04:42,860 --> 01:04:46,400 There is a great Bob Mercer quote about this later.
01:04:46,620 --> 01:04:50,660 He says, we're right 50.75% of the time.
01:04:50,900 --> 01:04:52,440 And I do think he's making up that number.
01:04:52,520 --> 01:04:53,200 I think it's illustrative.
01:04:53,620 --> 01:04:53,940 Right.
01:04:54,500 --> 01:05:00,480 But we're 100% right 50.75% of the time.
01:05:00,580 --> 01:05:02,660 You can make billions that way.
01:05:03,160 --> 01:05:03,860 It's so true.
01:05:03,860 --> 01:05:05,060 When you have that little,
01:05:05,060 --> 01:05:14,760 it's about making sure that you're not betting so much that a few bets that don't break your way can take you down to zero and to make sure you can just play the game a lot.
01:05:15,220 --> 01:05:15,500 A lot.
01:05:15,720 --> 01:05:15,960 Yes.
01:05:16,480 --> 01:05:21,380 And then back to the Kelly criterion, adjust your bet sizes over time as you're making those bets.
01:05:21,940 --> 01:05:22,060 Yeah.
01:05:22,580 --> 01:05:24,500 Now, of course, this is all great in the abstract.
01:05:24,500 --> 01:05:31,980 If it's that you're literally sitting at a casino and you're somehow perfectly making these bets and you're just sitting right there at the table and then you can walk over to the cashier.
01:05:32,180 --> 01:05:33,700 It gets a little bit different in the market.
01:05:33,700 --> 01:05:34,660 For example,
01:05:35,060 --> 01:05:44,540 there are real transaction costs, especially at this point in history before some of these more innovative trading business models with pay for order flow and zero transaction fees and all this stuff.
01:05:44,760 --> 01:05:47,340 There's real transaction costs to putting on these trades.
01:05:47,720 --> 01:05:50,520 And of course, you're going to move the market when you put on these trades.
01:05:51,020 --> 01:05:52,540 Yes, this is slippage.
01:05:52,800 --> 01:05:55,220 There's all sorts of practical consideration.
01:05:55,780 --> 01:05:57,500 You could get front run by other people.
01:05:57,960 --> 01:06:00,200 It's not just a computer program that gets executed.
01:06:00,200 --> 01:06:04,380 You actually have to meet the constraints of the real world when you're deciding.
01:06:04,560 --> 01:06:04,940 Instead of,
01:06:04,940 --> 01:06:07,520 we're going to have a few big bets, we're going to have 100,000 tiny bets.
01:06:08,020 --> 01:06:08,520 Yes.
01:06:08,520 --> 01:06:20,160 And as time goes on and the whole quant industry emerges and becomes much more sophisticated, I think it's particularly the slippage there that becomes the governor on how high velocity you can actually be on this.
01:06:20,480 --> 01:06:25,660 And the slippage is that once you are at a certain scale, you are going to move the market with your trades.
01:06:26,040 --> 01:06:34,320 So the deeper you get into the order book, like let's say you want to buy $5 million of something, maybe your first $100,000, you're pretty sure you can get the quoted price.
01:06:34,320 --> 01:06:40,320 But by your last $100,000 of that $5 million buy, the price might have gotten pretty different already.
01:06:40,980 --> 01:06:41,000 Yeah.
01:06:41,440 --> 01:06:42,980 We're going to come back to this in just a minute.
01:06:43,160 --> 01:06:51,600 But this certainly for early rent tech and then even now still for all of quantitative finance is a really, really, really important thing.
01:06:52,340 --> 01:06:52,520 Yep.
01:06:52,520 --> 01:06:56,840 And David, in a very crude way, calls back to last episode on Hermes.
01:06:57,240 --> 01:07:04,300 The idea that the price would be highest for the family member that is willing to sell now and sort of goes down over time.
01:07:04,300 --> 01:07:11,440 If the family was going to sell to Bernard Arnault, it would behoove you to be first in the order book, not last in the order book.
01:07:11,840 --> 01:07:11,860 Yes.
01:07:12,620 --> 01:07:19,320 I feel like there's this metal lesson that I've been learning through Acquired and my own personal investing over the past couple of years.
01:07:19,920 --> 01:07:22,480 Every market is dependent on supply and demand.
01:07:23,200 --> 01:07:30,040 You can see quoted valuations and quoted price streams, but oftentimes that's like the mistake of just looking at averages.
01:07:30,680 --> 01:07:31,200 Exactly.
01:07:31,200 --> 01:07:31,560 Yes.
01:07:31,700 --> 01:07:33,700 Looking at the quoted price of an asset.
01:07:33,700 --> 01:07:34,880 That is wrong.
01:07:35,300 --> 01:07:40,960 You actually should be looking at what is the volume that is willing to buy and what is the volume that is willing to sell.
01:07:41,080 --> 01:07:45,540 And for all of those buyers and all of those sellers, what are the price at which they are willing to transact?
01:07:46,060 --> 01:07:51,060 And the way that tends to manifest on a stock chart is here's the price of a share right now.
01:07:51,140 --> 01:07:54,100 But that's not actually what's going on under the surface.
01:07:54,300 --> 01:08:00,420 It's a whole bunch of buyers and sellers who have different willingness to pay and have different amounts that they're trying to buy or sell.
01:08:00,720 --> 01:08:01,080 Yes.
01:08:01,720 --> 01:08:03,420 Now, at this point in time, when the Medallion Fund...
01:08:03,700 --> 01:08:11,780 When the Medallion Fund is first starting to work in, say, late 1989, early 1990, it's small enough that this isn't a big consideration yet.
01:08:11,940 --> 01:08:12,160 Yeah.
01:08:12,580 --> 01:08:12,840 Right.
01:08:13,120 --> 01:08:18,400 Medallion was about $27 million under management when Burleigh Camp bought out Axe.
01:08:18,940 --> 01:08:32,180 In 1990, the first full year after that, the fund gains 77.8% gross, which after fees and carry was 55% net.
01:08:32,620 --> 01:08:33,680 Now, what were the fees?
01:08:33,680 --> 01:08:38,540 I mean, either one of those numbers is shooting the freaking lights out.
01:08:38,740 --> 01:08:46,400 Assuming that this is not a crazy high-risk strategy that they executed and it'll completely fall apart under different market conditions.
01:08:46,400 --> 01:08:52,460 Like, if this is an actual repeatable strategy that produces the numbers you just said, unbelievable.
01:08:52,840 --> 01:08:53,820 World-changing.
01:08:54,580 --> 01:08:55,000 Hell yeah.
01:08:55,400 --> 01:08:56,300 Let's go.
01:08:56,640 --> 01:08:56,780 Yes.
01:08:57,340 --> 01:09:00,980 And indeed, it was a hell yeah, let's go situation.
01:09:01,320 --> 01:09:03,640 So, the numbers you quoted me, the gross and the net.
01:09:03,680 --> 01:09:04,520 It sounded quite different.
01:09:04,700 --> 01:09:05,920 Talk to me about the fees and carry.
01:09:06,480 --> 01:09:11,480 So, carry, I've seen different sources of whether it was 20% or 25% in the early days.
01:09:11,700 --> 01:09:15,940 But the management fee on the fund was 5%, which is crazy.
01:09:15,940 --> 01:09:20,620 The top venture capital firms in the world charge a 3% management fee.
01:09:20,740 --> 01:09:24,120 And even that is like, everybody holds their nose and is like, this is ridiculous.
01:09:24,340 --> 01:09:31,960 How on earth were these nobodies charging a 5% management fee out the gate to their investors?
01:09:32,580 --> 01:09:33,100 Well.
01:09:33,680 --> 01:09:34,940 A couple things.
01:09:35,560 --> 01:09:37,120 One, their investors were not sophisticated.
01:09:37,500 --> 01:09:39,920 It was mostly their own money and their buddy's money.
01:09:40,320 --> 01:09:41,300 So, they set that precedent.
01:09:41,780 --> 01:09:42,500 They set that precedent.
01:09:42,920 --> 01:09:46,540 But two, though, they actually needed the money.
01:09:46,640 --> 01:09:46,900 Yes.
01:09:47,060 --> 01:09:52,100 Because Strauss's infrastructure costs were about $800,000 a year.
01:09:52,320 --> 01:09:58,260 So, they just backed into the management fee based on like, hey, we need $800,000 a year to run the infrastructure.
01:09:58,260 --> 01:10:01,200 Plus, we need some money to pay folks and whatnot.
01:10:01,200 --> 01:10:02,120 Like, great.
01:10:02,400 --> 01:10:03,520 5% management fee.
01:10:03,680 --> 01:10:12,660 And so, the pitch they're making to the investor base is like, if you believe that we should be able to massively outperform the market doing quantitative trading, well, we're going to need a lot of fees to do that.
01:10:12,760 --> 01:10:16,440 And so, the investors basically took the deal, if they thought about it enough.
01:10:17,080 --> 01:10:17,280 Okay.
01:10:17,320 --> 01:10:18,100 So, that's the fees.
01:10:18,360 --> 01:10:25,160 On the performance, that 20% or 25%, it's just not actually that far above market, if it's above market at all.
01:10:25,500 --> 01:10:30,040 What you're seeing is a high fee, normal-ish performance fee fund at this point in time.
01:10:30,460 --> 01:10:30,760 Yes.
01:10:30,760 --> 01:10:33,640 High management fee, normal-ish carrier performance.
01:10:33,640 --> 01:10:33,960 That's the performance element.
01:10:34,260 --> 01:10:34,460 Yep.
01:10:34,900 --> 01:10:46,460 So, at the end of 1990, Simons is so jazzed about what's going on that he tells Burlick Camp, hey, you should move here to Long Island.
01:10:46,620 --> 01:10:49,040 Let's re-centralize everything here.
01:10:49,420 --> 01:10:51,480 I want to go all in on this.
01:10:51,600 --> 01:10:55,680 I think with some tweaks, we can be up 80% after fees next year.
01:10:56,740 --> 01:10:59,680 Burlick Camp is a little more circumspect.
01:11:00,260 --> 01:11:01,620 A, he wants to stay in Berkeley.
01:11:01,720 --> 01:11:03,620 He doesn't have any desire to move to Long Island.
01:11:03,640 --> 01:11:13,580 And B, I couldn't tell how much of this is just he's a little more conservative than Jim, or how much of this actually might be his, hey, whole poker bet-sizing thing.
01:11:14,140 --> 01:11:19,600 He turns to Jim and he says, well, if you're so optimistic, why don't you buy me out?
01:11:20,320 --> 01:11:28,280 So, Jim does at 6X the basis that Burlick Camp had paid Axe a year earlier.
01:11:29,000 --> 01:11:31,680 On the one hand, making a 6X in one year sounds great.
01:11:31,680 --> 01:11:32,680 On the other hand…
01:11:33,640 --> 01:11:46,620 This is the equivalent of when Don Valentine sold Sequoia's Apple Steak before the IPO to lock in a great gain, but miss out on all the upside to come.
01:11:47,120 --> 01:11:50,640 David, I think we should throw this out so people understand the volume of this.
01:11:51,040 --> 01:12:00,260 They've generated on the order of $60 billion of performance fees for the owners of the fund over their entire lifetime.
01:12:00,520 --> 01:12:03,600 So, on the one hand, 6X in a year ain't…
01:12:03,640 --> 01:12:03,880 Not bad.
01:12:03,980 --> 01:12:11,340 On the other hand, you owned a giant part of something that has dividended $60 billion in cash out to its owners.
01:12:12,020 --> 01:12:12,300 Oof.
01:12:12,600 --> 01:12:12,760 Yeah.
01:12:13,300 --> 01:12:14,660 That's just on the carry side.
01:12:14,820 --> 01:12:17,520 I mean, the owners are the principals.
01:12:18,020 --> 01:12:21,220 So, just like dollars out of the firm, it's probably twice that.
01:12:21,640 --> 01:12:28,720 I would estimate probably $150, $200 billion that have come out of Medallion over the last 35 years.
01:12:29,280 --> 01:12:33,280 So, Jim buys out Burlick Camp.
01:12:33,640 --> 01:12:41,120 He rolls everything in the Medallion fund back into Rentech itself, moves everything back to Stony Brook.
01:12:41,540 --> 01:12:43,100 Strauss moves to Stony Brook.
01:12:43,720 --> 01:12:50,500 So, it's now the Jim Simons show in New York with Strauss building the engineering systems and Axe, I think, still had a small stake.
01:12:51,120 --> 01:12:52,820 Yes, that's right. And Strauss had a stake as well.
01:12:53,220 --> 01:13:02,820 So, once Jim takes control and moves everything back, he basically decides that he's going to turn…
01:13:03,640 --> 01:13:14,960 into an even better, even more idealized version of IDA and the math department at Stony Brook.
01:13:14,960 --> 01:13:32,260 He's going to make this an academics paradise, where if you are one of the absolute smartest mathematicians or systems engineers in the world, this is where you want to be.
01:13:32,520 --> 01:13:32,960 So…
01:13:33,640 --> 01:13:37,720 Of course, he starts raiding the Stony Brook department itself again.
01:13:38,080 --> 01:13:41,840 And this is when Henry Laufer joins full-time.
01:13:42,600 --> 01:13:51,140 Laufer had been consulting with Medallion in the early days and working with Burlick Camp as they're doing bet sizing, as they're making more frequent trades.
01:13:51,500 --> 01:13:56,260 But now, once the whole operation has moved back to Long Island, Laufer's like,
01:13:56,340 --> 01:13:58,860 Oh, okay, great. I'll come full-time. I'm here at Stony Brook anyway.
01:13:59,220 --> 01:14:01,000 This is way more fun than teaching.
01:14:01,480 --> 01:14:03,520 And listeners, I imagine this is probably the point where you're starting…
01:14:03,640 --> 01:14:07,040 to get confused and saying, there are so many people in this story.
01:14:07,100 --> 01:14:09,800 I think we're on eight or nine. We just keep introducing more people.
01:14:10,380 --> 01:14:13,360 And that is the story of renaissance.
01:14:13,360 --> 01:14:16,840 It is not this singular, clean narrative.
01:14:17,040 --> 01:14:25,400 It is a very complex reality of a whole bunch of different people that came in and out at different eras,
01:14:25,520 --> 01:14:32,840 where the firm was trying different things and eventually became phenomenally successful with a very particular approach.
01:14:32,840 --> 01:14:36,480 But while they were figuring it out along the way, it took a lot of people.
01:14:37,020 --> 01:14:39,720 A lot of people and just a lot of time, too.
01:14:39,980 --> 01:14:52,780 This is 25 years. This is a quarter century from the time that Baum and Simons write the paper at IDA until Medallion really starts to work.
01:14:52,960 --> 01:14:54,760 It takes a long time.
01:14:54,940 --> 01:15:01,900 And we haven't even introduced the two people who would become the co-CEOs of this company for 20 years.
01:15:02,140 --> 01:15:02,620 Yes.
01:15:02,840 --> 01:15:05,080 Well, let's get to that.
01:15:05,720 --> 01:15:12,480 So, Jim moves everything back to Long Island, sets it up as this academic paradise, is recruiting the smartest people in the world.
01:15:13,160 --> 01:15:24,500 In 1991, the next year, the firm does 54.3% gross returns and 39.4% net returns after fees.
01:15:24,620 --> 01:15:28,520 So, not Jim's bogey of 80%, but still pretty freaking great.
01:15:28,520 --> 01:15:32,520 And we should say the years of modest performance…
01:15:32,840 --> 01:15:36,060 From every single year forward, they shoot the lights out.
01:15:36,420 --> 01:15:40,460 From 1990 onward, they never lose money.
01:15:41,080 --> 01:15:45,180 And on a gross basis, they never even do less than 30%.
01:15:45,180 --> 01:15:47,540 It's working. It's going.
01:15:47,740 --> 01:15:54,220 The whole rest of the story is about hold on, keep the machine working, and we're on the train.
01:15:54,940 --> 01:15:57,900 The historic run has begun, let's just say.
01:15:58,140 --> 01:15:58,360 Yep.
01:15:58,360 --> 01:16:02,360 So, 1992, gross returns are 47%.
01:16:03,460 --> 01:16:06,320 93, they're 54%.
01:16:06,320 --> 01:16:13,960 At the end of 1993, Simons decides to close the fund and not allow new LPs in.
01:16:14,240 --> 01:16:18,720 So, if you're an existing LP, you can stay in, but they're no longer open for new inflows.
01:16:19,000 --> 01:16:28,500 He has so much confidence in what they're doing that he thinks they're all going to make more money without accepting new capital by just keeping it to the existing investor base.
01:16:28,940 --> 01:16:32,500 1994, gross returns are 93%.
01:16:32,840 --> 01:16:34,660 90 freaking percent.
01:16:35,460 --> 01:16:39,440 Medallion, at this point, is stacking up cash.
01:16:39,900 --> 01:16:42,460 It is a meaningful fund.
01:16:42,460 --> 01:16:56,180 It's about $250 million total at this point in time, which is small, but we're talking about 1994 with a bunch of outsiders and academics that have managed to amass a quarter billion dollars here.
01:16:56,520 --> 01:16:57,860 People start to pay attention.
01:16:58,680 --> 01:17:02,460 And the performance fees on this are $7 million, $13 million.
01:17:03,460 --> 01:17:04,500 $52 million.
01:17:05,120 --> 01:17:09,780 The free cash flow flowing to partners here is certainly becoming real, too.
01:17:10,400 --> 01:17:10,660 Yes.
01:17:11,160 --> 01:17:24,100 But as they get into that, call it on the order of magnitude of a billion-dollar scale, they start bumping into the moving markets problem and the slippage that we were talking about earlier.
01:17:24,340 --> 01:17:24,460 Yep.
01:17:24,520 --> 01:17:25,820 And that's sort of in the mid-90s?
01:17:26,260 --> 01:17:26,500 Yep.
01:17:26,560 --> 01:17:29,660 As they're hitting this $250 million, half a billion-dollar scale.
01:17:30,040 --> 01:17:30,260 Right.
01:17:30,360 --> 01:17:32,820 The computer model spits out, we should go buy.
01:17:32,820 --> 01:17:35,460 We should go buy this huge amount of something at this price.
01:17:35,560 --> 01:17:36,200 They go to do it.
01:17:36,400 --> 01:17:40,200 They can only buy 10%, 20%, 30% of the amount they want at that price.
01:17:40,280 --> 01:17:41,840 And then suddenly, the price is very different.
01:17:42,480 --> 01:17:42,580 Yep.
01:17:43,000 --> 01:17:52,420 Up to this point, the vast majority of what Medallion is doing is trading currencies and commodities, not equities.
01:17:53,000 --> 01:17:56,700 Because you might be thinking, okay, yeah, I hear you.
01:17:56,820 --> 01:17:58,260 The 90s was a different era.
01:17:58,480 --> 01:18:01,400 But half a billion-dollar fund doesn't sound that big.
01:18:01,400 --> 01:18:02,640 How are they moving markets?
01:18:02,820 --> 01:18:03,980 With half a billion dollars.
01:18:04,400 --> 01:18:05,440 It's not the equity markets.
01:18:05,620 --> 01:18:07,660 It's because they're in these thinner markets.
01:18:07,660 --> 01:18:11,100 It's not that commodities and futures are small markets.
01:18:11,260 --> 01:18:13,760 They're large, but they're thin compared to equities.
01:18:14,060 --> 01:18:19,500 There's just not that much volume, and you just can't trade that much without slippage becoming a huge issue.
01:18:20,000 --> 01:18:21,780 And Medallion is now hitting that limit.
01:18:22,780 --> 01:18:31,140 So Simons decides the only thing we can do here to expand, which I'm such a believer in what we're doing, we need to expand,
01:18:31,140 --> 01:18:33,480 is we need to move into equities.
01:18:34,160 --> 01:18:35,880 Equities are the holy grail.
01:18:35,880 --> 01:18:42,960 If we can make this work there, the depth in those markets will let us scale way, way, way bigger than we are now.
01:18:43,720 --> 01:18:50,160 And there's so much more data about equities pricing that we can feed into our models,
01:18:50,320 --> 01:18:55,440 and the signal processing that we can do and the signals that we can find are going to be even better.
01:18:56,020 --> 01:18:56,040 Right.
01:18:56,140 --> 01:19:00,940 There's so many buyers and sellers every day showing up to trade so many different companies.
01:19:00,940 --> 01:19:06,600 At such high velocity, it's almost this honeypot for Renaissance's systems.
01:19:07,120 --> 01:19:08,080 This is sort of their moment.
01:19:08,180 --> 01:19:13,040 This is what they were built for, and it's kind of funny that they've just been in kid-glove land the whole time
01:19:13,040 --> 01:19:15,700 with these thinly traded markets with minimal data.
01:19:16,400 --> 01:19:16,420 Yes.
01:19:16,880 --> 01:19:20,060 And this brings us to Peter Brown and Bob Mercer.
01:19:20,680 --> 01:19:28,800 And in 1993, one of the mathematicians that Jim had recruited to Rentech, a guy named Nick Patterson,
01:19:29,060 --> 01:19:29,800 gets especially...
01:19:30,940 --> 01:19:34,800 Really passionate about going out and recruiting new talent along with Jim.
01:19:35,200 --> 01:19:37,460 And this is, I think, one of the keys to Rentech and the culture there.
01:19:38,120 --> 01:19:41,660 People want other smart people to come be there, too.
01:19:42,120 --> 01:19:43,620 Nick's sitting there like, this is a joy.
01:19:43,840 --> 01:19:46,900 I want to go find other best people in the world to hang out with.
01:19:47,660 --> 01:19:54,820 And he had read in the newspaper that IBM was going through cost-cutting and was about to do layoffs.
01:19:55,240 --> 01:20:00,820 And he also knew that the speech recognition group at IBM had some absolutely...
01:20:00,940 --> 01:20:04,680 Absolutely fantastic mathematical talent.
01:20:04,680 --> 01:20:12,860 And really, what they were doing was, again, another vector in the early AI machine learning research.
01:20:13,440 --> 01:20:20,240 Specifically, IBM's deep blue chess project of the time had come out of this group.
01:20:20,900 --> 01:20:24,500 And Peter Brown there was the one that actually spearheaded the project.
01:20:25,260 --> 01:20:25,340 Yep.
01:20:25,760 --> 01:20:29,680 And it's interesting that you talk about speech recognition as...
01:20:30,940 --> 01:20:33,540 The perfect fit for what they were doing.
01:20:33,580 --> 01:20:34,660 And you might say, why is that?
01:20:34,660 --> 01:20:39,660 Well, the actual work that goes into speech recognition, natural language processing,
01:20:39,920 --> 01:20:44,800 is kind of the same signal processing that Renaissance is doing to analyze the market.
01:20:45,200 --> 01:20:45,940 It's not just kind of.
01:20:46,000 --> 01:20:47,760 It's exactly the same signal processing.
01:20:47,980 --> 01:20:48,300 Right.
01:20:48,500 --> 01:20:55,580 Speech recognition is a hidden Markov process where the computer that's listening to the sounds
01:20:55,580 --> 01:20:59,760 to try to turn it into language doesn't actually know English, right?
01:21:00,000 --> 01:21:00,440 Obviously.
01:21:00,440 --> 01:21:05,620 But what it does know is, when I hear this set of frequencies and tonalities and sounds,
01:21:06,000 --> 01:21:09,340 there's a limited set of likely things that could come after it.
01:21:09,640 --> 01:21:12,820 And in Greg's book, he greatly points out this perfect example.
01:21:13,160 --> 01:21:15,320 When I say apple, you might say pie.
01:21:15,780 --> 01:21:20,560 The probability that pie is going to be the next word following apple is significantly higher.
01:21:20,980 --> 01:21:24,800 And so these people who have spent their careers not only doing the math
01:21:24,800 --> 01:21:29,500 and the theoretical computer science behind speech recognition to help figure out and
01:21:29,500 --> 01:21:33,800 predict the next words, that you have a narrow set of likely words to choose from.
01:21:33,900 --> 01:21:38,140 So when you're listening to those frequencies, you can say, it's probably going to be one
01:21:38,140 --> 01:21:41,820 of these three, rather than search the entire dictionary for any word that it could be to
01:21:41,820 --> 01:21:43,000 narrow the processing power.
01:21:43,580 --> 01:21:48,700 It's not only the theoretical side, but it's also people who have built those systems at
01:21:48,700 --> 01:21:51,680 IBM, like a real operational computer company.
01:21:52,300 --> 01:21:54,560 Yes, at operational scale.
01:21:54,560 --> 01:21:58,280 And this is what's so important and why the two of them become,
01:21:58,380 --> 01:21:59,480 probably.
01:21:59,500 --> 01:22:05,260 The most critical hires in Rentec's history, even including all the great academics that
01:22:05,260 --> 01:22:10,820 came before them, because they're good on the math side, but they have this large systems
01:22:10,820 --> 01:22:11,480 experience.
01:22:12,280 --> 01:22:18,000 And Jim and Nick know that if they're going to move into equities because of the volume of data
01:22:18,000 --> 01:22:23,580 and because of how much more complex that market is, they need more complex systems.
01:22:24,020 --> 01:22:28,100 And the current talent at Rentec coming from academia has just never experienced that or
01:22:28,100 --> 01:22:29,060 built anything like it.
01:22:29,060 --> 01:22:29,340 Yeah.
01:22:29,340 --> 01:22:29,480 Yeah.
01:22:29,500 --> 01:22:34,620 And the world that they're entering is just exploding in complexity and dimensionality.
01:22:35,200 --> 01:22:36,520 And when I say that, here's what I mean.
01:22:37,100 --> 01:22:42,480 The data that they are mining, that they're looking for, is this intraday tick data between
01:22:42,480 --> 01:22:44,100 every stock trading.
01:22:44,800 --> 01:22:49,760 So they're in this sort of trying to map the relationship between one stock and every other
01:22:49,760 --> 01:22:53,220 stock, not just at that moment in time, but every time before it and every time after it.
01:22:53,560 --> 01:22:59,220 They're also, once they do identify patterns, which this is key, the algorithms identify,
01:22:59,500 --> 01:23:00,080 the patterns.
01:23:00,080 --> 01:23:05,780 It's not a human with a hunch saying, I think when oil prices go up, the airline prices are
01:23:05,780 --> 01:23:06,280 going to get hit.
01:23:06,580 --> 01:23:10,260 It's computers doing machine learning to discover the patterns in the data.
01:23:10,720 --> 01:23:16,840 Then there's the second piece of, well, what trades do you actually put on to be profitable
01:23:16,840 --> 01:23:20,340 from the probabilities that you just discovered?
01:23:20,440 --> 01:23:23,520 All these weights of relationships between all of these different companies.
01:23:23,680 --> 01:23:25,180 You're not just putting on one trade.
01:23:25,180 --> 01:23:29,180 You're putting on 10, 100, thousands of simultaneous trades.
01:23:29,180 --> 01:23:33,400 And both to hedge, to be able to isolate some particular variable that you're looking for,
01:23:33,520 --> 01:23:35,540 again, not you, but a computer is looking for.
01:23:35,920 --> 01:23:42,600 And you also need to do it in such specific bite sizes so that you don't move the market.
01:23:42,960 --> 01:23:49,380 So you're looking for a super multivariate, multidimensional problem, both on the data
01:23:49,380 --> 01:23:52,980 ingestion side and on the how do I actually react to it side.
01:23:52,980 --> 01:23:58,820 And all of this computation can't take a long time because you must act, you know, not in
01:23:58,820 --> 01:23:59,600 milliseconds.
01:23:59,960 --> 01:24:03,020 It's not a high frequency trading that's front running the market.
01:24:03,180 --> 01:24:04,320 That's not actually what they do.
01:24:04,420 --> 01:24:06,500 A lot of people think it is, but we'll get to that later.
01:24:06,920 --> 01:24:10,820 But they do need to act with reasonable quickness, probably on the order of minutes.
01:24:11,260 --> 01:24:14,280 So these need to be really efficient computer systems too.
01:24:15,540 --> 01:24:22,080 And the universe of equities is so much more multidimensional and interrelated.
01:24:22,080 --> 01:24:26,480 There are only so many currencies in the world, and there are especially only so many currencies
01:24:26,480 --> 01:24:28,660 that are large enough trading markets.
01:24:28,660 --> 01:24:28,800 There are so many currencies that are large enough trading markets.
01:24:28,800 --> 01:24:34,920 And there's not infinite, but thousands and thousands of equities in the world that are
01:24:34,920 --> 01:24:36,800 deep enough markets that you can operate in.
01:24:37,240 --> 01:24:40,160 And to some degree, they're all correlated with one another.
01:24:41,040 --> 01:24:43,680 And just keep adding layers of complexity here.
01:24:43,780 --> 01:24:46,180 Keep adding new things to multiply by.
01:24:46,560 --> 01:24:48,600 Many of these are traded on multiple exchanges.
01:24:48,600 --> 01:24:54,700 So you might also be looking for pricing disparities on the same equity on different markets at
01:24:54,700 --> 01:24:55,580 different points in time.
01:24:55,580 --> 01:24:58,680 So there's just dimensions upon dimensions of things.
01:24:58,800 --> 01:25:00,540 And that's something that we need to analyze, correlate, and act upon.
01:25:01,540 --> 01:25:04,860 So Patterson and Simons go raid IBM.
01:25:05,060 --> 01:25:07,660 They're like Steve Jobs raiding Xerox PARC.
01:25:07,880 --> 01:25:14,980 They bring Peter and Bob and one of their programming colleagues, David Magerman, over
01:25:14,980 --> 01:25:16,920 from IBM into Rentech.
01:25:17,480 --> 01:25:20,320 And they get started on building the equities model.
01:25:20,820 --> 01:25:25,160 But it turns out, A, they're obviously very successful at that.
01:25:25,160 --> 01:25:28,720 But the impact that they have and what they build.
01:25:28,800 --> 01:25:38,120 It's even bigger because Bob and Peter realize that, hey, actually, we should just have
01:25:38,120 --> 01:25:44,480 one model for everything here, for currencies, for commodities, for equities.
01:25:45,000 --> 01:25:46,520 Everything is correlated.
01:25:46,960 --> 01:25:48,180 Everything is a signal.
01:25:48,460 --> 01:25:53,840 It's not like the equities market is wholly independent and separate from what's happening
01:25:53,840 --> 01:25:55,760 in currencies or what's happening in commodities.
01:25:56,260 --> 01:25:58,580 There are relationships everywhere.
01:25:59,120 --> 01:26:01,340 We really want just one model.
01:26:01,960 --> 01:26:06,940 This is like a fantastical undertaking, especially in the early to mid-90s.
01:26:07,600 --> 01:26:07,700 Right.
01:26:07,700 --> 01:26:12,940 But if you can nail it, it means that you can do interesting things like, hey, we don't
01:26:12,940 --> 01:26:19,500 have a lot of data on this particular market, but it looks a lot like something we do have
01:26:19,500 --> 01:26:20,160 data on.
01:26:20,320 --> 01:26:24,320 So if it's all part of the same model, we can kind of just apply all the learnings from
01:26:24,320 --> 01:26:28,420 this other thing onto this brand new thing that we're looking at with little data for
01:26:28,420 --> 01:26:28,780 the first time.
01:26:28,780 --> 01:26:34,480 And because we're putting it all in one model and no one else in the world is, we can discover
01:26:34,480 --> 01:26:35,880 patterns that no one else knows about.
01:26:36,540 --> 01:26:42,080 It turns out that this was actually the second most important innovation that Bob and Peter
01:26:42,080 --> 01:26:46,400 bring to Rentech, the actual product and performance of having one model.
01:26:46,780 --> 01:26:54,680 The most important thing is that if you have only one model, one infrastructure, everybody
01:26:54,680 --> 01:26:58,320 in the firm is working on that same model.
01:26:58,780 --> 01:27:05,120 You can all collaborate all together, which is especially important when you have the
01:27:05,120 --> 01:27:08,360 smartest people in the entire world, all in one building.
01:27:08,960 --> 01:27:12,120 Before this, there were separate models within Rentech.
01:27:12,260 --> 01:27:18,960 So insights and innovations and work that one team was doing on one model wouldn't get
01:27:18,960 --> 01:27:23,820 applied or translate over to work that was happening by another team on another model.
01:27:24,140 --> 01:27:27,600 They did have the cultural element where it was encouraged that you share your learnings,
01:27:27,660 --> 01:27:28,740 but someone would have to...
01:27:28,780 --> 01:27:32,860 Take the time during their lunch break and go learn from you about those and then implement
01:27:32,860 --> 01:27:33,740 it in their version.
01:27:33,980 --> 01:27:36,200 There's a lag and it may actually not get implemented.
01:27:36,200 --> 01:27:41,540 Yeah, this is wholly unique and revolutionary.
01:27:42,360 --> 01:27:50,880 No other at-scale investment firm, period, and especially quant firm, operates this way
01:27:50,880 --> 01:27:52,400 today with just one model.
01:27:52,920 --> 01:27:58,280 There are portfolio managers and teams and multi-strategy people are culturally competitive
01:27:58,280 --> 01:27:58,760 with one another.
01:27:58,780 --> 01:28:03,040 But even if they're not, the work that you're doing on this side of Citadel is not impacting
01:28:03,040 --> 01:28:04,900 the work that you're doing on that side of Citadel.
01:28:06,000 --> 01:28:09,700 What Bob and Peter do is they unify everything at Rentech.
01:28:09,780 --> 01:28:12,340 So all the wood is going behind one arrow.
01:28:13,080 --> 01:28:13,340 Yes.
01:28:14,040 --> 01:28:20,100 And before we talk about the impact of that, we want to thank our longtime friend of the
01:28:20,100 --> 01:28:23,540 show, Vanta, the leading trust management platform.
01:28:24,180 --> 01:28:27,840 Vanta, of course, automates your security reviews and compliance efforts.
01:28:27,840 --> 01:28:28,640 So frameworks like...
01:28:28,780 --> 01:28:35,080 Like SOC 2, ISO 27001, GDPR, and HIPAA compliance and monitoring, which is quite topical if
01:28:35,080 --> 01:28:39,620 you are in the heavily regulated finance industry and you need a lot of security and compliance.
01:28:40,080 --> 01:28:43,980 Vanta takes care of these otherwise incredibly time and resource draining efforts for your
01:28:43,980 --> 01:28:46,180 organization and makes them fast and simple.
01:28:46,420 --> 01:28:50,540 Yeah, Vanta is the perfect example of the quote that we talk about all the time here
01:28:50,540 --> 01:28:51,000 on Acquired.
01:28:51,380 --> 01:28:56,060 Jeff Bezos, his idea that a company should only focus on what actually makes your beer
01:28:56,060 --> 01:28:57,020 taste better.
01:28:57,020 --> 01:28:57,760 I.e.
01:28:57,760 --> 01:29:01,560 spend your time and resources only on what's actually going to move the needle for your
01:29:01,560 --> 01:29:05,220 product and your customers and outsource everything else that doesn't.
01:29:05,220 --> 01:29:07,880 In Rentech's case, this would be the model.
01:29:08,540 --> 01:29:11,320 Every company needs compliance and trust with their vendors and customers.
01:29:11,320 --> 01:29:15,480 It plays a major role in enabling revenue because customers and partners demand it,
01:29:15,840 --> 01:29:18,600 but yet it adds zero flavor to your actual product.
01:29:19,140 --> 01:29:20,340 Vanta takes care of all of it for you.
01:29:20,480 --> 01:29:24,240 No more spreadsheets, no fragmented tools, no manual reviews to cobble together your
01:29:24,240 --> 01:29:25,840 security and compliance requirements.
01:29:26,200 --> 01:29:27,600 It is one single.
01:29:27,600 --> 01:29:32,040 Software pane of glass, just like one model that connects to all of your services via
01:29:32,040 --> 01:29:35,800 APIs and eliminates countless hours of work for your organization.
01:29:36,280 --> 01:29:40,280 There are now AI capabilities to make this even more powerful, and they even integrate
01:29:40,280 --> 01:29:42,460 with over 300 external tools.
01:29:42,700 --> 01:29:45,980 Plus, they let customers build private integrations with their internal systems.
01:29:46,500 --> 01:29:50,540 And perhaps most importantly, your security reviews are now real time instead of static,
01:29:50,540 --> 01:29:54,720 so you can monitor and share with your customers and partners to give them added confidence.
01:29:54,720 --> 01:29:56,880 So whether you're a startup or a large enterprise,
01:29:56,880 --> 01:30:00,480 and your company is ready to automate compliance and streamline security reviews
01:30:00,480 --> 01:30:05,100 like Vanta's 7,000 customers around the globe, and go back to making your beer taste better,
01:30:05,340 --> 01:30:09,940 head on over to vanta.com slash acquired and just tell them that Ben and David sent you.
01:30:10,360 --> 01:30:13,380 And thanks to friend of the show, Christina, Vanta's CEO,
01:30:13,700 --> 01:30:17,160 all acquired listeners get $1,000 of free credit.
01:30:17,560 --> 01:30:19,320 Vanta.com slash acquired.
01:30:20,140 --> 01:30:22,700 So David, the equities machine.
01:30:23,440 --> 01:30:26,120 Yes, and indeed a machine it is.
01:30:26,360 --> 01:30:26,860 So David, the equities machine.
01:30:26,860 --> 01:30:32,460 So Peter and Bob come in in 1993, and 1994, 1995, they're building this,
01:30:32,820 --> 01:30:34,180 Rentech is getting into equities.
01:30:34,680 --> 01:30:39,700 And yeah, just imagine the computers that you were using during 1994 and 1995.
01:30:40,200 --> 01:30:45,880 It is astonishing the level of computational complexity and coordination and results
01:30:45,880 --> 01:30:49,960 that they are pulling off, again, in real time, analyzing these markets
01:30:49,960 --> 01:30:52,660 with the technology that was available during those years.
01:30:52,660 --> 01:30:55,320 Yes, and here's what's amazing.
01:30:55,320 --> 01:31:02,780 Returns go down maybe slightly, certainly a bit from the blowout year that 1994 was,
01:31:03,100 --> 01:31:08,980 but they're still above 30% every single year, most years above 40%.
01:31:08,980 --> 01:31:12,580 This is unbelievable that they're maintaining this performance
01:31:12,580 --> 01:31:16,300 as they're going into this hugely more complex market,
01:31:16,300 --> 01:31:18,840 and they're scaling assets under management.
01:31:19,120 --> 01:31:24,480 So by the end of the 1990s, Medallion has almost $2 billion in assets under management.
01:31:24,480 --> 01:31:30,100 While maintaining roughly the same performance by getting into equities.
01:31:30,500 --> 01:31:32,220 This is huge.
01:31:32,620 --> 01:31:33,140 Yep.
01:31:33,600 --> 01:31:35,960 And David, if you just kind of look at this and do the math.
01:31:36,140 --> 01:31:43,660 Okay, so 94, their AUM was $276 million, and they grew 93%.
01:31:43,660 --> 01:31:48,800 And then their AUM the next year was $462 million, and then they grew 52%.
01:31:48,800 --> 01:31:51,680 And their AUM the next year was $637 million.
01:31:51,680 --> 01:31:53,960 You kind of quickly get where I'm going here, which is,
01:31:53,960 --> 01:31:57,540 oh, they're scaling AUM not by bringing in new investors.
01:31:58,060 --> 01:31:59,160 Right, it's closed to new investors.
01:31:59,380 --> 01:32:00,780 It's all just compounding.
01:32:01,120 --> 01:32:04,420 This is the same capital that they had in 1993
01:32:04,420 --> 01:32:11,640 that has gone from $122 million at the beginning of that year to 1999 being $1.5 billion.
01:32:12,320 --> 01:32:12,640 Yes.
01:32:13,040 --> 01:32:18,520 And then in the year 2000, they just totally blow the doors off.
01:32:18,520 --> 01:32:23,740 A 128% gross returns.
01:32:23,960 --> 01:32:30,600 Net returns after fees of 98.5%.
01:32:30,600 --> 01:32:32,500 This is bananas.
01:32:32,880 --> 01:32:38,680 They grow the fund from $1.9 billion to $3.8 billion of asset center management.
01:32:38,800 --> 01:32:42,840 Again, purely by investing gains, not by getting any new investors.
01:32:43,280 --> 01:32:45,420 The year the tech bubble burst.
01:32:45,960 --> 01:32:46,240 Yes.
01:32:46,720 --> 01:32:50,420 While the whole rest of the market is down big time,
01:32:51,100 --> 01:32:53,560 Medallion is up 128%.
01:32:53,960 --> 01:32:55,520 Gross on the air.
01:32:55,580 --> 01:32:56,440 And this becomes a theme.
01:32:56,440 --> 01:32:59,960 High volatility is when Medallion really shines.
01:33:00,480 --> 01:33:02,540 And here you go, uncorrelated.
01:33:02,700 --> 01:33:06,440 They have their final stamp of approval right here of,
01:33:07,000 --> 01:33:08,920 not only are we a money printing machine,
01:33:09,260 --> 01:33:12,160 we are a money printing machine in all environments,
01:33:12,340 --> 01:33:14,160 regardless of the state of the broad market.
01:33:14,580 --> 01:33:15,600 And David, as you said,
01:33:15,940 --> 01:33:18,960 volatility actually makes their algorithms work even better.
01:33:19,040 --> 01:33:19,860 Because what are they doing?
01:33:20,160 --> 01:33:23,780 They're looking for scenarios where the market's going to
01:33:23,780 --> 01:33:23,940 actually be able to do something.
01:33:23,960 --> 01:33:28,460 And they can take advantage of people making decisions that they shouldn't.
01:33:28,780 --> 01:33:30,820 And any time any investors are under pressure,
01:33:31,280 --> 01:33:33,920 there's a little bit of edge that's going to accrue to a Medallion
01:33:33,920 --> 01:33:36,760 that's saying, oh, okay, you're fear selling right now.
01:33:37,020 --> 01:33:40,240 Well, I can determine if you should be fear selling or not.
01:33:40,340 --> 01:33:43,460 And if I determine that you shouldn't be dumping that asset, I'm buying it from you.
01:33:44,120 --> 01:33:47,660 So there's a really fun story around this that really illustrates
01:33:47,660 --> 01:33:52,360 Jim's genius in managing the firm and the people.
01:33:53,080 --> 01:33:53,940 And, you know, I don't know if you've heard of it, but I've heard of it.
01:33:53,940 --> 01:33:57,320 How this year was when they really figured this out.
01:33:57,480 --> 01:34:02,040 So the first couple days of the tech bubble bursting,
01:34:02,800 --> 01:34:06,040 Medallion actually takes a bunch of large losses.
01:34:06,540 --> 01:34:10,160 And part of it might be that the model wasn't tuned right yet
01:34:10,160 --> 01:34:13,440 because nobody at Rentech had seen this type of behavior in the market before.
01:34:14,020 --> 01:34:17,920 Part of it might also be, too, that it didn't perform well for those couple days.
01:34:18,200 --> 01:34:21,020 It's a really stressful time for everybody.
01:34:21,020 --> 01:34:23,600 You know, everybody's in Jim's office.
01:34:23,940 --> 01:34:25,100 Jim's smoking his cigarettes.
01:34:25,300 --> 01:34:26,200 It's a cloud of smoke.
01:34:26,600 --> 01:34:28,020 And they're debating what to do.
01:34:28,100 --> 01:34:30,600 And Jim makes the call to take some risk off.
01:34:30,920 --> 01:34:32,280 He's worried about blowing up.
01:34:32,440 --> 01:34:35,580 We're not very far removed at this point from long-term capital management.
01:34:36,160 --> 01:34:40,640 The model may be saying we should stay long here, but let's not blow up the firm.
01:34:41,140 --> 01:34:41,240 Yep.
01:34:41,980 --> 01:34:47,280 After this goes down, Peter Brown comes to Jim and offers to resign,
01:34:47,820 --> 01:34:50,840 given the losses that they incurred over these couple days.
01:34:51,040 --> 01:34:53,420 And Jim says, what are you talking about?
01:34:53,420 --> 01:34:55,000 Of course you shouldn't resign.
01:34:55,220 --> 01:34:59,460 You are way more valuable to the firm now that you've lived through this,
01:34:59,560 --> 01:35:04,160 and you now know not to 100% trust the model in all situations.
01:35:04,440 --> 01:35:05,200 It's fascinating.
01:35:05,520 --> 01:35:06,740 It's such a good insight.
01:35:06,940 --> 01:35:09,240 That illustrates Jim as a leader right there.
01:35:09,500 --> 01:35:10,520 It totally does.
01:35:10,700 --> 01:35:16,400 There's a parallel story when Jim ultimately does retire in 2009
01:35:16,400 --> 01:35:19,300 and Peter and Bob take over as co-CEOs,
01:35:19,560 --> 01:35:23,240 where a year or so before the quote-unquote quant quake,
01:35:23,420 --> 01:35:26,080 had happened, where similar to the tech bubble bursting,
01:35:26,580 --> 01:35:30,000 there was all of a sudden very large drawdowns
01:35:30,000 --> 01:35:33,340 among all quantitative firms in the market and rent tech gets hit.
01:35:34,120 --> 01:35:38,520 And during that period, Peter argued very strenuously
01:35:38,520 --> 01:35:41,860 that we should trust the model, stay risk on,
01:35:41,960 --> 01:35:44,340 this is going to be an incredibly profitable time for us.
01:35:44,820 --> 01:35:48,940 And Jim pumped the brakes and stepped in, intervened and took risk off.
01:35:49,720 --> 01:35:53,360 And Peter goes to Jim again around the CEO,
01:35:53,420 --> 01:35:54,320 transition and says,
01:35:54,380 --> 01:35:58,040 hey, Jim, aren't you worried that with me running the place now,
01:35:58,100 --> 01:36:00,520 I'm going to be too aggressive and blow it up one of these days?
01:36:00,820 --> 01:36:03,500 And Jim says, no, I'm not worried at all.
01:36:03,900 --> 01:36:07,040 I know you were only so aggressive in that moment
01:36:07,040 --> 01:36:09,140 because I was there pushing back on you.
01:36:09,220 --> 01:36:12,120 And when you're in the seat, you're going to be less aggressive.
01:36:12,260 --> 01:36:15,360 He's just such a master at insight into human behavior.
01:36:15,680 --> 01:36:16,580 It is so true, though.
01:36:16,680 --> 01:36:20,360 I even find this about myself that I will naturally take the position
01:36:20,360 --> 01:36:22,460 of the foil to the person across from me.
01:36:22,460 --> 01:36:23,360 So if somebody,
01:36:23,420 --> 01:36:24,980 if somebody's being pushy in some way,
01:36:25,040 --> 01:36:28,240 I'll find myself taking a position where if I pause and reflect,
01:36:28,300 --> 01:36:30,500 I'm like, I don't think I expected to take this position
01:36:30,500 --> 01:36:31,560 coming into this conversation.
01:36:31,740 --> 01:36:35,820 But, you know, you naturally want to sort of play the other side
01:36:35,820 --> 01:36:38,220 to balance out the person sitting across from you.
01:36:38,880 --> 01:36:38,900 Yeah.
01:36:39,540 --> 01:36:42,960 So back to the year 2000 and this incredible performance.
01:36:43,240 --> 01:36:46,300 Ben, to what you were saying earlier about uncorrelated returns,
01:36:46,660 --> 01:36:48,900 not only did they shoot the lights out that year,
01:36:49,020 --> 01:36:50,840 they're doing it when the market is down.
01:36:51,460 --> 01:36:52,940 We got to introduce this concept
01:36:52,940 --> 01:36:53,400 of a sharp return.
01:36:53,420 --> 01:36:56,820 Which for all of you listeners that are in the finance world,
01:36:56,940 --> 01:36:57,480 you'll know this.
01:36:57,740 --> 01:37:00,200 But for everybody else, this is a really important concept.
01:37:00,800 --> 01:37:02,420 And I think people grasp it intuitively.
01:37:02,920 --> 01:37:06,120 We've mentioned this concept a couple times this episode where,
01:37:06,760 --> 01:37:07,380 OK, great.
01:37:07,660 --> 01:37:10,960 It's amazing to have a fund that 25Xs
01:37:10,960 --> 01:37:14,200 or a year where you have 100% investment return
01:37:14,200 --> 01:37:17,520 or I bought Bitcoin yesterday and it doubled overnight.
01:37:17,980 --> 01:37:20,280 Does that make you one of the best investors in the world?
01:37:20,360 --> 01:37:21,540 We all intuitively know.
01:37:21,540 --> 01:37:22,520 No, it doesn't.
01:37:22,520 --> 01:37:24,960 Because maybe that was a fluke.
01:37:25,400 --> 01:37:28,000 Maybe you're taking on an extreme amount of risk.
01:37:28,400 --> 01:37:31,560 And then the question is always adjusting for the risk that you're taking.
01:37:31,920 --> 01:37:35,840 Can you produce a superior return taking the risk into that account?
01:37:36,180 --> 01:37:40,720 And so you basically can provide value to investors as a fund manager in two ways.
01:37:41,020 --> 01:37:45,360 You can outperform the market or you can be entirely uncorrelated with the market
01:37:45,360 --> 01:37:46,520 and get market returns.
01:37:46,760 --> 01:37:49,640 Or what you can do as rent tech is both.
01:37:49,640 --> 01:37:51,640 You can be uncorrelated and,
01:37:52,520 --> 01:37:53,120 effectively, outperform,
01:37:53,320 --> 01:37:55,940 which is effectively the holy grail of money management.
01:37:56,580 --> 01:37:56,660 Yes.
01:37:56,860 --> 01:38:00,920 And so the Sharpe ratio is a measurement combining these two concepts.
01:38:01,080 --> 01:38:01,460 Exactly.
01:38:01,900 --> 01:38:04,700 So it's named after the economist William F. Sharpe.
01:38:04,780 --> 01:38:06,240 It was pioneered in 1966.
01:38:07,000 --> 01:38:12,020 It is effectively the measure of a fund's performance relative to the risk-free rate.
01:38:12,340 --> 01:38:17,200 So if you performed at 15% that year and the risk-free rate was 3%,
01:38:17,720 --> 01:38:20,880 then your numerator is going to be 12%.
01:38:20,880 --> 01:38:22,200 And it is compared, again,
01:38:22,520 --> 01:38:26,280 against the volatility or the standard deviation is technically what it is.
01:38:26,360 --> 01:38:30,800 But effectively, how volatile have you been the last X years?
01:38:30,920 --> 01:38:34,740 And typically, it's looked at as a 3-year Sharpe or a 5-year Sharpe or a 10-year Sharpe.
01:38:35,140 --> 01:38:38,560 The Sharpe ratio represents the additional amount of return
01:38:38,560 --> 01:38:43,320 that an investor receives per unit of an increase in risk.
01:38:43,700 --> 01:38:45,360 And so, David, you're starting to throw out numbers.
01:38:45,820 --> 01:38:47,280 Low Sharpe ratios are bad.
01:38:47,560 --> 01:38:50,620 Negative Sharpe ratios are worse because that means you're underperforming
01:38:50,620 --> 01:38:51,580 the risk-free rate.
01:38:51,580 --> 01:38:51,860 High Sharpe ratios are bad.
01:38:51,860 --> 01:38:54,980 High Sharpe ratios are good because it means that you're producing lots of returns
01:38:54,980 --> 01:38:59,020 and your variance or your standard deviation or your sort of risk is low.
01:38:59,500 --> 01:39:06,400 So in 1990, they had a Sharpe of 2.0, which was twice that of the S&P 500 benchmark.
01:39:06,620 --> 01:39:06,840 Awesome.
01:39:07,220 --> 01:39:07,360 Yep.
01:39:07,640 --> 01:39:07,840 Good.
01:39:08,200 --> 01:39:11,540 1995 to 2000, Sharpe ratio of 2.5.
01:39:11,920 --> 01:39:12,960 Really starting to hum.
01:39:13,140 --> 01:39:13,800 Pretty unbelievable.
01:39:14,400 --> 01:39:14,620 Good.
01:39:15,080 --> 01:39:16,400 Where do I sign up to invest?
01:39:16,400 --> 01:39:21,760 At some point, they added foreign markets and achieved a Sharpe ratio of 6.3.
01:39:21,860 --> 01:39:24,440 Which is double the best quant firms.
01:39:25,040 --> 01:39:29,620 This is a firm that has almost no chance of losing money, at least historically,
01:39:29,860 --> 01:39:34,000 and massively outperforms the market on an uncorrelated basis.
01:39:35,020 --> 01:39:39,560 And I believe, if I have my research right, in 2004,
01:39:40,300 --> 01:39:43,720 they actually achieved a Sharpe ratio of 7.5.
01:39:44,120 --> 01:39:44,720 Astonishing.
01:39:45,300 --> 01:39:47,440 You know, again, back to our sports analogy here.
01:39:47,500 --> 01:39:48,600 These aren't Hall of Fame numbers.
01:39:48,600 --> 01:39:50,720 These are like, I don't know,
01:39:50,820 --> 01:39:51,700 make Tom Brady look like a star.
01:39:51,700 --> 01:39:51,840 These are like, I don't know, make Tom Brady look like a star.
01:39:51,860 --> 01:39:52,700 Like a third stringer.
01:39:52,860 --> 01:39:53,680 Yes, exactly.
01:39:54,660 --> 01:39:57,760 So, on the back of 2000 and this rise,
01:39:58,300 --> 01:39:59,900 the next year in 2001,
01:40:00,380 --> 01:40:05,040 they raise the carried interest on the fund to 36%,
01:40:05,040 --> 01:40:08,260 up from either 20 or 25%, whatever it was before.
01:40:08,800 --> 01:40:12,260 Now, remember, they've already closed the fund to new investors.
01:40:12,420 --> 01:40:14,480 So, there's still outside investors in the fund,
01:40:15,040 --> 01:40:16,840 but no new investors are coming in.
01:40:17,460 --> 01:40:19,900 And then, the next year, in 2002,
01:40:20,440 --> 01:40:21,840 they raise the carried interest on the fund to 36%.
01:40:21,840 --> 01:40:30,400 to 44%. I mean, great work if you can get it. But for context, the Sequoias, the benchmarks out
01:40:30,400 --> 01:40:37,020 there, they have obscene carry of 30%. 44% is unprecedented. There's two interesting ways to
01:40:37,020 --> 01:40:42,220 look at this. One, they're just trying to jack it up so high that they just purge their existing
01:40:42,220 --> 01:40:45,600 investors out, where they're saying, we're not going to kick anyone out yet, but we've been
01:40:45,600 --> 01:40:49,540 closed to new business for a long time now. You should see yourself out at some point.
01:40:49,540 --> 01:40:53,360 The other way to look at this, which I think is probably the right way to look at it, is
01:40:53,360 --> 01:41:00,240 investors are arbitragers. They see a mispricing, they come into the market,
01:41:00,520 --> 01:41:07,100 they fix that mispricing. So anytime that there's an opportunity to bring the way that a currency
01:41:07,100 --> 01:41:11,680 is trading on two different exchanges closer together, investors are serving their purpose
01:41:11,680 --> 01:41:16,480 of coming in, arbitraging that difference, taking a little bit of profit as a thank you,
01:41:16,480 --> 01:41:19,360 and then sort of fixing the market to make the market a true
01:41:19,360 --> 01:41:19,520 market.
01:41:19,540 --> 01:41:22,880 Weighing machine, not a voting machine, but making it so that all prices reflect
01:41:22,880 --> 01:41:28,620 the value of what something is actually worth. And in some ways, that's what Renaissance is
01:41:28,620 --> 01:41:32,460 doing here to themselves or to their investors. They're coming in and saying, look, this is
01:41:32,460 --> 01:41:38,440 obscene. We so clearly outperformed the market. You're still going to take this deal, even if we
01:41:38,440 --> 01:41:43,720 take more of this, because there's just a mispricing here. This product should not be priced at 20,
01:41:43,800 --> 01:41:48,380 25% carry. This product should be priced at a much higher carried interest, and you're still
01:41:48,380 --> 01:41:48,920 going to love it.
01:41:48,920 --> 01:41:55,840 You should pay 20% carry for a firm that delivers you 15% annual returns. We're delivering you
01:41:55,840 --> 01:41:58,480 50% annual returns.
01:41:59,020 --> 01:42:02,620 Totally. So I have to imagine it didn't go over well with the existing investors,
01:42:02,780 --> 01:42:05,520 but they just have so much leverage that what's going to happen?
01:42:06,120 --> 01:42:10,460 Okay. Once again, I'm sorry, audience. I have to say, hold on one more minute
01:42:10,460 --> 01:42:16,640 for another perspective that I have to offer on the carry element, but I want to finish the story
01:42:16,640 --> 01:42:18,900 first. Okay. So 2001, they raised a carry. They raised a carry. They raised a carry. They raised a carry.
01:42:18,900 --> 01:42:26,080 They raised a carry to 36%. 2002, they raised it to 44%. And then in 2003, they actually say,
01:42:26,300 --> 01:42:30,300 hey, we can't incentivize you out of the fund, outside investors. We are going to kick you out.
01:42:30,660 --> 01:42:35,100 So starting in 2003, everybody who's an outside investor, who's not part of the rent tech
01:42:35,100 --> 01:42:40,480 family, current employee or alumni of the firm, gets kicked out.
01:42:41,040 --> 01:42:44,700 And not all alumni get to stay. There's select alumni that get grandfathered in.
01:42:44,980 --> 01:42:48,720 Yes. Now, why did we do this? I'm going to talk about one reason,
01:42:48,740 --> 01:42:48,880 and I'm going to talk about another reason, and I'm going to talk about another reason,
01:42:48,900 --> 01:42:54,840 but one reason is super obvious. The Medallion Fund is now at $5 billion in assets under
01:42:54,840 --> 01:42:59,920 management that they're trading. Even in the equities market, they are now hitting up against
01:42:59,920 --> 01:43:07,320 slippage. And so if they want to maintain this crazy, crazy performance, they just can't get
01:43:07,320 --> 01:43:11,880 that much bigger. This is the problem that Warren Buffett talks about all the time and why he has
01:43:11,880 --> 01:43:16,520 to basically just increase his position in Apple rather than going and buying the next great family
01:43:16,520 --> 01:43:18,740 owned business. The things that move the needle for the next great family owned business are the
01:43:18,740 --> 01:43:24,600 are so big that that's really all they can do. And when you are big, you're going to move any
01:43:24,600 --> 01:43:30,380 market that you enter into. And the strategy that rent tech is employing right now, they're just
01:43:30,380 --> 01:43:36,820 deeming doesn't work at north of $5 billion. So in 2003, they start kicking all the outside
01:43:36,820 --> 01:43:42,880 investors out of Medallion. But clearly, there's still lots of institutional demand to invest
01:43:42,880 --> 01:43:48,280 with Renaissance. So what do they do? Well, time to start another fund.
01:43:48,280 --> 01:43:55,060 So they start the Renaissance Institutional Equities Fund. And there's a couple of things
01:43:55,060 --> 01:43:59,340 to add a little bit of context to really why they decide to do this. Well, the first one is
01:43:59,340 --> 01:44:04,720 sometimes there's just more profitable strategies than they had the capital to take advantage of
01:44:04,720 --> 01:44:09,900 in Medallion, but they weren't sure it would be on a durable basis. If they were sure that
01:44:09,900 --> 01:44:16,040 they could manage 10, 15, 20, 25 billion in Medallion all the time, then they would grow
01:44:16,040 --> 01:44:18,260 to that. But if just sometimes there's these strategies that are more profitable, then they
01:44:18,280 --> 01:44:22,780 appear, well, we don't want to commit to a much higher fund size and then not always have those
01:44:22,780 --> 01:44:28,760 strategies available. The other thing is that a lot of the times those strategies aren't really
01:44:28,760 --> 01:44:34,500 what Medallion is set up to do. They require longer hold times. And so there's a little bit
01:44:34,500 --> 01:44:39,640 of downside to that because these new strategies, the predictive abilities are less because they
01:44:39,640 --> 01:44:44,040 have to predict further into the future to understand what the exit prices will be on
01:44:44,040 --> 01:44:48,220 these longer term holds. But they still figure, hey, even though it's not quite our
01:44:48,280 --> 01:44:52,340 bread and butter with the short term stuff, we should be able to make some money doing it.
01:44:52,960 --> 01:44:59,040 Yeah, there's a fun story around this that Peter Brown tells of Jim came into his office one day
01:44:59,040 --> 01:45:05,920 and said, Peter, I got a thought exercise for you. If you married a Rockefeller, would you advise
01:45:05,920 --> 01:45:12,300 the family that they should invest a large portion of their wealth in the S&P 500? And Peter says,
01:45:12,380 --> 01:45:17,400 no, of course not. That's not a great risk adjusted return. And these guys are very used to
01:45:17,400 --> 01:45:18,240 sharp ratios.
01:45:18,280 --> 01:45:19,760 That are far better than the S&P.
01:45:20,280 --> 01:45:26,160 Right. And so Jim says, yes, exactly. Now get to work on designing the product that they should
01:45:26,160 --> 01:45:26,760 invest in.
01:45:27,260 --> 01:45:31,760 Right. And so that's basically what they come up with is, can we create something that's
01:45:31,760 --> 01:45:37,320 like an S&P 500 with a higher sharp ratio? Can we beat the market by a few percentage points or
01:45:37,320 --> 01:45:41,360 frankly, even match the market each year with lower volatility than if they were buying an
01:45:41,360 --> 01:45:46,280 index fund? And you can see who this would be very attractive to pensions, large institutions,
01:45:47,120 --> 01:45:48,260 firms that want to come in and buy an index fund. And so they come up with a very attractive
01:45:48,280 --> 01:45:54,800 compound at market or slightly above market rate, but don't want to risk these massive drawdowns or
01:45:54,800 --> 01:45:59,480 frankly, just big volatility in general, should they need to pull the capital earlier. And the
01:45:59,480 --> 01:46:03,840 nice thing about being invested in a hedge fund versus a venture fund is you can do redemptions.
01:46:03,920 --> 01:46:08,940 Like if you look at the 13 Fs, the SEC documents that the Renaissance Institutional Equities Fund
01:46:08,940 --> 01:46:12,800 files over time, it changes every quarter because there's new people putting money in,
01:46:12,860 --> 01:46:17,600 there's people doing redemption. So it's a pretty good product, or at least the theory behind it is
01:46:17,600 --> 01:46:18,260 a pretty good product. And so if you're investing in a hedge fund versus a venture fund, you can do
01:46:18,280 --> 01:46:23,640 of a lower risk, similar return thing to the S&P 500.
01:46:24,460 --> 01:46:29,540 And the marketing is built in. It's not like there's any lack of demand of outside capital
01:46:29,540 --> 01:46:33,360 that wants to invest with rent tech. Right. It's really funny. There's
01:46:33,360 --> 01:46:36,460 all these stories about how the marketing documents literally say, this is not the
01:46:36,460 --> 01:46:40,700 Medallion Fund. We don't promise returns like the Medallion Fund. In fact, we're not charging for it
01:46:40,700 --> 01:46:45,540 like the Medallion Fund. David, you said that the fees and carry on Medallion went up to what,
01:46:45,540 --> 01:46:48,100 5 and 44. Well, on the institutional,
01:46:48,280 --> 01:46:53,720 fund, the fees are 1 in 10. You're only taking 1% annual fee and 10% of the performance.
01:46:54,040 --> 01:46:58,420 Clearly, this is a very different product. But people did not perceive that. People were very
01:46:58,420 --> 01:47:01,980 excited. It's a Renaissance product. It's the same analysts. They're using all their fancy
01:47:01,980 --> 01:47:05,880 computers. I'm sure we're going to get this crazy outperformance. And at the end of the day,
01:47:05,940 --> 01:47:11,740 it is an extremely different vehicle. Yeah. That has not performed anywhere
01:47:11,740 --> 01:47:18,180 near how Medallion has performed. Correct. Has it served its purpose? Yeah. But is it Medallion? No.
01:47:18,280 --> 01:47:24,460 It's not special in the way that Medallion is special. Yes. A couple other funny things on the
01:47:24,460 --> 01:47:30,300 institutional fund. So I spent a bunch of time scrolling through 13Fs over the last decade from
01:47:30,300 --> 01:47:34,120 the Medallion filings. And they're all from, I think they have two institutional funds.
01:47:34,700 --> 01:47:37,800 Yeah. There's Institutional Equities and Diversified Alpha.
01:47:38,440 --> 01:47:43,400 So the funniest thing is they file these 13Fs. And David and I are very used to looking at the 13Fs of
01:47:43,400 --> 01:47:47,800 friends of the show who run hedge funds, who we've had on as guests, or perhaps,
01:47:47,800 --> 01:47:51,240 really just any investor where you want to see like, or what are they buying and selling this
01:47:51,240 --> 01:47:57,740 quarter? And usually you see 15, 25, maybe 50 different names on there. Well, the 13F
01:47:57,740 --> 01:48:04,280 for Renaissance has 4,300 stocks in these tiny little chunks. And there's a little bit of
01:48:04,280 --> 01:48:08,440 persistence quarter to quarter. For example, weirdly, Novo Nordisk has been one of their
01:48:08,440 --> 01:48:12,960 biggest holdings, biggest, I say at like one to 2%. That's their biggest position
01:48:12,960 --> 01:48:16,220 for several quarters in a row. Hey, they've been listening to a choir.
01:48:16,220 --> 01:48:16,900 That's right.
01:48:17,800 --> 01:48:18,540 Signals in the model.
01:48:19,880 --> 01:48:27,180 You kind of get the sense from looking at these filings that these things were flying all over
01:48:27,180 --> 01:48:31,020 the place. And this was just the moment in time where they decided to take a snapshot
01:48:31,020 --> 01:48:35,220 and put it on a piece of paper. And even though this is the end of quarter filing of what their
01:48:35,220 --> 01:48:40,260 ownership was, if you had taken it a day or a week earlier, it could look completely different.
01:48:41,100 --> 01:48:46,820 Yes. The way that some folks we talked to described the difference between the institutional
01:48:46,820 --> 01:48:47,780 funds and medallions is that they're not going to be able to do that. They're going to have to
01:48:47,780 --> 01:48:48,160 hold their position for a whole month. And they're going to have to hold their position for a whole month.
01:48:48,160 --> 01:48:49,300 So that's the way that medallions are doing it. And that's what I call the medallion to us is that
01:48:49,300 --> 01:48:56,980 medallions average hold time for their trades and positions is call it like a day, maybe a day and a
01:48:56,980 --> 01:49:03,120 half. Whereas the average hold time for the institutional funds positions is like a couple
01:49:03,120 --> 01:49:10,560 months. So across 4,300 stocks in the portfolio, there's a lot of trading activity that happens on
01:49:10,560 --> 01:49:17,460 any given day. But it's a lot slower in any given name than medallion would be.
01:49:17,780 --> 01:49:22,680 Which makes sense. Again, it gets back to this slippage concept. If you have a bigger fund and
01:49:22,680 --> 01:49:27,460 you're investing larger amounts, which the institutional funds are, you can't be trading
01:49:27,460 --> 01:49:32,940 as frequently or all of your gains are going to slip away. Yep. And frankly, it just looks a lot
01:49:32,940 --> 01:49:38,440 like the S&P 500. Like when you look at as of November 23, so 11 of the 12 months of the year
01:49:38,440 --> 01:49:45,020 had happened, they were up 8.6%. Okay. That sounds like an index type return. You look at the first
01:49:45,020 --> 01:49:50,920 four months of 2020, right after the crazy dip from the pandemic, they were down 10.4%. Less
01:49:50,920 --> 01:49:55,520 than the broader market, but they still were sort of a mirror of the broader market. So I think
01:49:55,520 --> 01:50:03,180 the RIEF, their institutional fund, yes, it works as expected. No, it's not medallion. And if it
01:50:03,180 --> 01:50:07,020 were standing on its own, there's zero chance that we would be covering the organization behind it
01:50:07,020 --> 01:50:12,520 unacquired. Zero percent chance. Speaking of the fund, that is the reason why we are covering
01:50:12,520 --> 01:50:14,400 this company on this show.
01:50:15,020 --> 01:50:20,540 We set up during the tech bubble crash that volatility is when medallion really shines.
01:50:20,960 --> 01:50:31,160 Well, there's no more volatile periods than 2007 and 2008. Yep. 2007, medallion does 136%
01:50:31,160 --> 01:50:43,320 gross. 2008, medallion does 152% gross. Like get out of here. Crazy. This is 2008 while the rest
01:50:43,320 --> 01:50:45,000 of the financial world is melting.
01:50:45,020 --> 01:50:49,300 And so this really does illustrate where do they make their money from? Who is on the other side
01:50:49,300 --> 01:50:54,080 of these trades? It's people acting emotionally. They have effectively these really robust models
01:50:54,080 --> 01:51:00,380 that are highly unemotional, that are making these super intricate multi-security bets.
01:51:00,620 --> 01:51:05,440 And they are putting on exactly the right set of trades to achieve the risk and exposure that the
01:51:05,440 --> 01:51:11,220 system wants them to have. And who is on the other side of those trades? It's panic sellers.
01:51:11,460 --> 01:51:15,000 It's dentists. It's hedge funds who don't trust their computer.
01:51:15,020 --> 01:51:19,540 They're like, ah, crap. We got to just take risk off, even though it's a negative expected value
01:51:19,540 --> 01:51:24,700 move for us. They're basically trading against human nature. And importantly, in this business
01:51:24,700 --> 01:51:28,860 versus every other business that we cover here on Acquired or most other businesses, this is truly
01:51:28,860 --> 01:51:34,380 zero sum. It's not like they're here in an industry that's a growth industry and lots of competitors
01:51:34,380 --> 01:51:39,700 can take different approaches, but the whole pie is growing so much that I don't care if, no,
01:51:39,700 --> 01:51:44,740 you're fighting over a fixed pie here. I'm trading against someone else. I win, they lose.
01:51:45,020 --> 01:51:50,880 Yes. Well, there's one slight nuance to that, but I don't know how much it holds water.
01:51:51,280 --> 01:51:58,200 And the apologist nuance would be, well, Warren Buffett could be on the other side of the trade
01:51:58,200 --> 01:52:05,000 and Medallion could make money on that trade with Warren over its time horizon of a day and a half.
01:52:05,080 --> 01:52:09,660 And Warren could make money over his time horizon of, you know, 50 years.
01:52:10,040 --> 01:52:10,320 Super fair.
01:52:10,320 --> 01:52:14,400 So I think the argument against that, though,
01:52:15,020 --> 01:52:21,520 is that Medallion sold after a day and a half to somebody else who bought at that lower price.
01:52:22,080 --> 01:52:27,120 And so somewhere along the chain, that loss is getting offloaded to somebody.
01:52:27,340 --> 01:52:33,820 The direct counterparty of Medallion and the quant industry writ large might not take the loss,
01:52:33,920 --> 01:52:38,500 but somebody is going to take the loss along the way. It is, as you say, a zero sum game.
01:52:39,000 --> 01:52:43,980 Yeah. But I think the important thing is, can you and your adversary both benefit? And I think
01:52:43,980 --> 01:52:45,000 in this case, you and your adversary both benefit. And I think in this case, you and your adversary
01:52:45,000 --> 01:52:48,360 and you and your counterparty, the person you're trading against, yes, you have two different
01:52:48,360 --> 01:52:53,740 objective outcomes. Like, can I get a penny over on Warren Buffett by managing to take him on this
01:52:53,740 --> 01:52:57,000 one trade? Sure. But his strategy is such that that is irrelevant.
01:52:58,060 --> 01:53:04,100 So after the historic performance during the financial crisis, as I alluded to earlier,
01:53:04,660 --> 01:53:11,900 Jim retires at the end of 2009 and Peter and Bob become co-CEOs, co-heads of the firm in 2010.
01:53:13,000 --> 01:53:14,900 They take the portfolio.
01:53:15,000 --> 01:53:19,640 They take the portfolio size up to $10 billion when they take over. It had been at five for the
01:53:19,640 --> 01:53:27,580 last few years of Jim's tenure. They take it up to 10 and really with no impact, which I assume
01:53:27,580 --> 01:53:31,920 means that rent tech was getting better and the models were getting better because otherwise they
01:53:31,920 --> 01:53:39,320 would have gone to 10 before. Right. They gained confidence that they had enough profitable trades
01:53:39,320 --> 01:53:44,440 they could make that they could raise the capacity without dampening returns. Yes. And perhaps they
01:53:44,440 --> 01:53:47,800 could have done it earlier and they just didn't have the confidence that it would work at larger
01:53:47,800 --> 01:53:53,820 size. But I bet they're very good at knowing how large can our strategy work up to before it starts
01:53:53,820 --> 01:54:01,560 having diminishing returns. Yeah. And importantly, during periods of peak volatility, like say 2020,
01:54:02,380 --> 01:54:08,480 Medallion continues to shoot the lights out. So from at least the data that we were able to find
01:54:08,480 --> 01:54:14,160 on Medallion's performance over the past few years, 2020, they were up 149%.
01:54:14,440 --> 01:54:25,640 Gross and 76% net. So the magic is still there. And one way to look at it, which may not be the be-all
01:54:25,640 --> 01:54:32,600 and end-all, but I think is a good way to compare Jim's era at Medallion versus Peter and Bob's era
01:54:32,600 --> 01:54:41,760 during Jim's tenure, Medallion's total aggregate IRR from 1988 when the fund was formed to 2009
01:54:41,760 --> 01:54:42,560 when he retired.
01:54:44,440 --> 01:54:54,640 0.5% gross annual returns and 40.1% net annual returns, which of course did include many periods
01:54:54,640 --> 01:55:02,920 of lower carry, 20% versus the 44%. During the post-Jim era, the Peter and Bob era from 2010
01:55:02,920 --> 01:55:14,160 to 2022 was when we were able to get the latest data. IRRs are 77.3% gross and 40.3%
01:55:14,440 --> 01:55:22,680 net. So better on both fronts, even with much higher average fees. So yeah, I think Medallion is
01:55:22,680 --> 01:55:23,280 doing fine.
01:55:23,900 --> 01:55:28,260 That's amazing. And we weren't able to tell, there's some sources that report that they've
01:55:28,260 --> 01:55:33,940 grown from $10 billion in the last few years to being comfortable at a $15 billion fund size.
01:55:34,400 --> 01:55:39,480 And if so, that just means that they continue to find more profitable strategies within Medallion
01:55:39,480 --> 01:55:42,840 to keep those same unbelievable returns at larger sizes.
01:55:43,280 --> 01:55:43,760 Yeah.
01:55:44,440 --> 01:55:50,800 Ben, at the end of the day, this is all just insane. So as far as we can tell, Ben, you alluded
01:55:50,800 --> 01:55:56,340 to this a bit at the beginning of the episode. And as far as anybody else can tell, Medallion
01:55:56,340 --> 01:56:03,480 has by far the best investing track record of any single investment vehicle in history.
01:56:03,740 --> 01:56:05,180 So give me those net numbers.
01:56:05,180 --> 01:56:14,180 So during the entire lifetime so far of Medallion from 1988 to 2022, that's 34 years,
01:56:14,440 --> 01:56:26,340 the total net annual return number is 40%, 4-0, over 34 years after fees. It's 68%
01:56:26,340 --> 01:56:33,520 before fees, which equates to total lifetime carry dollars for the whole firm
01:56:33,520 --> 01:56:37,140 of $60 billion, just in carry by our calculations.
01:56:37,920 --> 01:56:38,220 Astonishing.
01:56:38,420 --> 01:56:41,120 That is a lot of money.
01:56:41,320 --> 01:56:43,760 Also, David Rosenthal, good spreadsheet work on this.
01:56:43,960 --> 01:56:44,320 You have not done...
01:56:44,440 --> 01:56:48,060 I haven't done a spreadsheet for an episode in a while, so I admire your work on this one.
01:56:48,380 --> 01:56:57,320 Yeah. I still know how to use Excel. Barely. It's going to be a dying art now with Copilot and
01:56:57,320 --> 01:56:58,160 GPTs.
01:56:58,520 --> 01:57:01,820 That's right. Okay, so $60 billion in total carry.
01:57:02,140 --> 01:57:09,240 So $60 billion in total carry is a lot of money. And, well, speaking of a lot of money,
01:57:10,040 --> 01:57:13,280 we do need to mention before we finish the story here,
01:57:13,280 --> 01:57:20,900 that that Rentech money has bought a lot of influence in society. So Bob Mercer,
01:57:21,260 --> 01:57:27,760 that name may have sounded familiar to many of you along the way, Bob was the primary funder
01:57:27,760 --> 01:57:34,320 of Breitbart and Cambridge Analytica, and one of the major financial backers of both the 2016 Trump
01:57:34,320 --> 01:57:41,940 campaign and the Brexit campaign in Great Britain. Now, lest you think that Rentech dollars are
01:57:41,940 --> 01:57:43,260 solely being funneled into the market, I'm not going to go into that. I'm going to go into the
01:57:43,260 --> 01:57:50,100 one side of the political spectrum. Jim Simons is a major Democratic donor, as are many other folks
01:57:50,100 --> 01:57:50,940 at Rentech.
01:57:51,280 --> 01:57:57,220 Yeah. Henry Laufer and other folks are also huge donors, approximately to the same tune as what
01:57:57,220 --> 01:57:58,380 Bob Mercer is on the right.
01:57:58,900 --> 01:58:04,200 Yeah. Tens of millions of dollars, many tens of millions of dollars on all sides and through many
01:58:04,200 --> 01:58:11,640 campaign cycles here from Rentech employees and alumni. This did become a flashpoint for the firm
01:58:11,640 --> 01:58:13,220 in the wake of the 2016 election.
01:58:13,260 --> 01:58:20,540 Mercer obviously became a controversial figure, both externally and internally within the firm.
01:58:21,360 --> 01:58:24,940 Especially once people realized he was the through line through Breitbart,
01:58:25,100 --> 01:58:27,820 Cambridge Analytica, the Trump election, and Brexit.
01:58:28,420 --> 01:58:36,520 Yes. Ultimately, Jim asked Bob to step down as co-CEO in 2017, which he did, but he did remain
01:58:36,520 --> 01:58:41,660 a scientist at the firm and a contributor to the models, even though he wasn't leading the
01:58:41,660 --> 01:58:43,240 organization with Peter from the beginning.
01:58:43,260 --> 01:58:51,680 Ultimately, the thing that surprised me the most is how these people all still work together, despite
01:58:51,680 --> 01:58:56,020 having about the most opposite political beliefs you could possibly have.
01:58:56,280 --> 01:58:58,240 Yeah. Understatement of the century.
01:58:58,240 --> 01:59:05,720 And all being extremely influential and active in those political systems. Yes, Bob Mercer is no
01:59:05,720 --> 01:59:11,820 longer the CEO of Renaissance Technologies or the co-CEO. He still works there. He's still associated.
01:59:12,280 --> 01:59:13,240 They all still speak high-profile.
01:59:13,260 --> 01:59:14,640 They all speak highly of each other. It's unexpected.
01:59:15,440 --> 01:59:18,460 Yeah. I think unexpected is the best way to put it.
01:59:19,040 --> 01:59:22,780 Like everything with Renaissance, it works a little bit different than the rest of the world.
01:59:23,540 --> 01:59:32,480 Yes. Okay. Speaking of, let's transition to analysis. And I have a fun little monologue
01:59:32,480 --> 01:59:38,440 I want to go on, if you will bear with me, Ben. I think this qualifies as the Rentech playbook,
01:59:38,440 --> 01:59:42,420 but I really kind of think of it as the Rentech tapestry.
01:59:42,620 --> 01:59:43,120 Ooh.
01:59:43,120 --> 01:59:46,380 I was inspired by Costco here because we were talking to folks in the research and
01:59:46,380 --> 01:59:52,620 everybody said, you know, Rentech, it just has these puzzle pieces that fit together.
01:59:53,000 --> 02:00:01,020 On the surface, Rentech does the same things that Citadel, D. Shaw, Two Sigma,
02:00:01,880 --> 02:00:06,260 Jane Street, others, et cetera, do. They hire the smartest people in the world
02:00:06,260 --> 02:00:12,800 and they give them the best data and infrastructure in the world to work on. And they say,
02:00:13,120 --> 02:00:20,020 go to town and make profitable trades. Those are very expensive commodities, those two things,
02:00:20,080 --> 02:00:24,140 the smartest people in the world and the best data and infrastructure, but they are commodities.
02:00:24,140 --> 02:00:29,680 Like Citadel can say the exact same things, just the same as like Walmart and Amazon can say,
02:00:29,880 --> 02:00:33,980 we too have large-scale supplier relationships that we leverage to provide low prices to
02:00:33,980 --> 02:00:39,360 customers, just like Costco. But it's underneath that where I think the magic lies. There are three
02:00:39,360 --> 02:00:42,920 very interrelated things that make Rentech,
02:00:43,120 --> 02:00:50,680 unique. So number one, they get the smartest people in the world to collaborate and not
02:00:50,680 --> 02:00:59,480 compete. Pretty much every other financial firm out there, employees and teams within the firm
02:00:59,480 --> 02:01:06,000 quasi-compete with one another. Yeah. I mean, typically in kind of a friendly way, but yeah.
02:01:06,680 --> 02:01:11,800 Let's take like in a venture firm, you've got your lead partner on a deal or a deal team.
02:01:11,960 --> 02:01:13,080 They're working that.
02:01:13,120 --> 02:01:18,700 Deal. And maybe some of the other partners help a little bit, but mostly they're off prosecuting
02:01:18,700 --> 02:01:23,640 their own deals. Yep. And I think that's the most collegial way that this happens in finance.
02:01:23,640 --> 02:01:28,200 Then you've got multi-strategy hedge funds out there where literally firms are being pitted
02:01:28,200 --> 02:01:33,140 against one another to be weighted in the ultimate trading model for the firm. Yep.
02:01:33,620 --> 02:01:41,160 At Rentech though, because of the one model architecture, everyone works together on the same
02:01:41,160 --> 02:01:42,640 investment strategy.
02:01:43,120 --> 02:01:48,780 And the same investment infrastructure. That means everyone sees everybody else's work.
02:01:49,140 --> 02:01:52,360 Everybody who works at Rentech on the research team, on the infrastructure team,
02:01:52,520 --> 02:01:57,420 they have access to the whole model. That's not true anywhere else.
02:01:58,000 --> 02:02:00,840 Yeah, that's a good point. The whole code base is completely visible.
02:02:01,480 --> 02:02:09,180 And that also means because it's just one model, just one strategy, when somebody else improves
02:02:09,180 --> 02:02:13,100 that model's performance, that directly impacts you.
02:02:13,120 --> 02:02:19,000 As much as it impacts them. This is really different than any other hedge fund out there.
02:02:19,560 --> 02:02:23,320 So why is that different than if I roll some of my compensation into a multi-strategy hedge fund
02:02:23,320 --> 02:02:27,260 that I work at? Don't I love other teams creating high performance also?
02:02:27,880 --> 02:02:32,880 Sure, but you don't love it as much as your team because either compensation or career-wise,
02:02:33,240 --> 02:02:38,400 you are much more dependent on your performance than you are other people's performance.
02:02:38,640 --> 02:02:42,820 Oh, yes. This is a big thing. You intend to have a job,
02:02:43,120 --> 02:02:46,860 after that job, at most places, most of the time. So you care about credit,
02:02:47,200 --> 02:02:51,160 and you care about smashing the pinata and then going elsewhere, or building reputation
02:02:51,160 --> 02:02:54,540 and then going elsewhere. Most of the people at Rentech are not going to have another job.
02:02:55,260 --> 02:03:00,680 What did you find on LinkedIn? At least the median tenure of employees is like 16 years.
02:03:01,060 --> 02:03:05,100 Yeah, I just got LinkedIn Premium, and you can see median tenure. And it's crazy. There's only
02:03:05,100 --> 02:03:10,320 like 300, 400 employees at Renaissance, and the median tenure, at least as reported by LinkedIn,
02:03:10,320 --> 02:03:12,820 is like 14 years. Yes.
02:03:13,120 --> 02:03:19,540 Okay, this brings me to point number two, which you said, this is an absurdly small team.
02:03:19,720 --> 02:03:26,580 There are less than 400 employees that work at Rentech, only half of which work in research and
02:03:26,580 --> 02:03:31,680 engineering, and the other half are either back office or institutional sales for the open funds.
02:03:31,820 --> 02:03:37,860 So let's call it, I don't know, 150, 200 people max who are like hands on the wheel here for
02:03:37,860 --> 02:03:39,140 Medallion. Yep.
02:03:39,800 --> 02:03:42,900 Every other peer firm of Rentech,
02:03:43,120 --> 02:03:49,980 Citadel, D, Shaw, Two Sigma, et cetera, all of them, you lump Jane Street, jump the high
02:03:49,980 --> 02:03:54,500 frequency guys in here. Minimum, 2,000 to 5,000 people work at those places.
02:03:55,040 --> 02:03:56,560 Wow, I didn't realize it was that big.
02:03:56,940 --> 02:04:02,960 It is an order of magnitude more people who are working at the other firms versus who are working
02:04:02,960 --> 02:04:08,740 at Rentech. And lest you think that it's like a capital-based thing, no, the institutional funds
02:04:08,740 --> 02:04:13,000 have gotten big. They peaked at over a hundred billion, but they're currently between 60 and 70
02:04:13,000 --> 02:04:13,100 billion. And that's a lot of people. And that's a lot of people. And that's a lot of people. And that's a lot of people.
02:04:13,120 --> 02:04:16,460 They peaked at over a hundred billion that they manage on top of the 10 or 15 that's in the
02:04:16,460 --> 02:04:23,780 Medallion fund. Yeah. So AUM is like the same as these big funds. This has all sorts of benefits.
02:04:24,360 --> 02:04:29,600 Number one, there's like the Hermes Atelier workshop benefit. Everyone knows each other by
02:04:29,600 --> 02:04:33,120 name. You know, your colleagues' kids, you know, your colleagues' families.
02:04:33,580 --> 02:04:38,040 Yep. They put right on their website, there are 90 PhDs in mathematics, physics, computer science,
02:04:38,040 --> 02:04:42,680 and related fields. The about page has these 10 kind of random bullet points, and that's one of
02:04:42,680 --> 02:04:42,820 them.
02:04:43,120 --> 02:04:48,700 Yes. Then there's the related aspect to all this. The firm is in the middle of nowhere on Long
02:04:48,700 --> 02:04:54,300 Island. You actually know your colleagues' families and kids because you're not going out and getting
02:04:54,300 --> 02:04:59,520 drinks with someone from Two Sigma in New York City. You're not comparing notes or measuring
02:04:59,520 --> 02:05:03,000 parts of your anatomy with someone else. You're like hanging out at the swimming pool.
02:05:03,740 --> 02:05:07,200 Totally. And since Renaissance doesn't recruit from finance jobs,
02:05:07,940 --> 02:05:13,080 it's kind of unlikely that you know someone else in finance. You came out of a science-related
02:05:13,080 --> 02:05:18,760 field. You now work in East Setauket, Long Island, which has, it's like 10,000 people or something
02:05:18,760 --> 02:05:23,240 or less that live there. So you're in this little town. You're not actually going into the city that
02:05:23,240 --> 02:05:28,860 often. And if you are, it's, again, not to grab drinks with other finance people. So even if you
02:05:28,860 --> 02:05:37,740 didn't have a many-page non-compete and a lifetime NDA, you're very unlikely to be in the social
02:05:37,740 --> 02:05:42,720 circles. You're just not getting exposed. Exactly. And Rentex Hiring,
02:05:43,080 --> 02:05:47,680 established scientists and PhDs, they're not hiring kids out of undergrad like
02:05:47,680 --> 02:05:53,020 Jane Street or Bridgewater is. My sense is that the place is like a college campus without any
02:05:53,020 --> 02:05:57,700 students. Have you seen the pictures online? Yeah. If you look up Renaissance Technologies
02:05:57,700 --> 02:06:02,820 at Google and you go and look at the photos on campus, it's a little courtyard and winding
02:06:02,820 --> 02:06:10,440 walking path and woods all around it, tennis courts. Yep. So then there's the last piece of
02:06:10,440 --> 02:06:13,060 the small team element, which is just the magnitude.
02:06:13,080 --> 02:06:18,860 of the financial impact, which I don't think is true. But let's say that there were another
02:06:18,860 --> 02:06:25,780 quant fund that made the same number of dollars of performance returns that Rentex does. At Rentex,
02:06:25,820 --> 02:06:30,280 you're splitting that a couple hundred ways. At Citadel, you're splitting that 5,000 ways.
02:06:30,920 --> 02:06:35,880 It just doesn't make sense to go anywhere else. We were chatting with someone to prep for this
02:06:35,880 --> 02:06:39,740 episode and they told us, you can't ever compete with them, but they'll pay you enough that you
02:06:39,740 --> 02:06:42,240 won't want to. Yes. Okay.
02:06:43,080 --> 02:06:46,320 This brings me to what I've been kind of teasing and I'm super excited about.
02:06:46,320 --> 02:06:53,600 I think the third puzzle piece of what makes Rentex so unique and defensible
02:06:53,600 --> 02:07:06,740 is Medallion's structure itself. That it is a LPGP fund with 5% management fee and 44% carry.
02:07:07,440 --> 02:07:12,920 So it's not like a prop shop or like proprietary, it's just one pot of money. It's literally a
02:07:12,920 --> 02:07:16,360 GPLP, even though the GPs and the LPs are the same people.
02:07:16,960 --> 02:07:21,140 So here's my thinking on this. Now, I don't know how it is actually structured, but
02:07:21,140 --> 02:07:27,740 there was something about this whole crazy 44% carry that just wasn't sitting with me right
02:07:27,740 --> 02:07:31,160 throughout the research because I kept asking myself, why?
02:07:31,620 --> 02:07:35,860 Right. They've already kicked out most of the LPs, if not all. So why are they raising the carry?
02:07:36,260 --> 02:07:42,900 Right. It's all themselves. It's all insiders. Why do they charge themselves 44% carry and 5%
02:07:42,920 --> 02:07:47,180 management fees? I think Jim talks about this though. Oh, I pay the fees just like everybody
02:07:47,180 --> 02:07:50,500 else. Yes. It's always a funny argument. It's like, who are you paying the fees to?
02:07:50,940 --> 02:07:58,640 Right. So I was like, what is happening here? So, okay, here's my hypothesis. This is not about
02:07:58,640 --> 02:08:03,740 having crazy performance fees. This is not about having the highest carry in the industry.
02:08:05,120 --> 02:08:11,980 This is a value transfer mechanism within the firm from the tenure base,
02:08:12,260 --> 02:08:12,900 to the company, to the company, to the company, to the company.
02:08:12,920 --> 02:08:17,780 People who are working on Medallion in any given year. So here's how I think it works.
02:08:18,440 --> 02:08:25,140 When people come into Rentech, they obviously have way less wealth than the people who've been there
02:08:25,140 --> 02:08:30,460 for a long time, both from the direct returns that you're getting every year from working there
02:08:30,460 --> 02:08:37,360 and just your investment percentage of the Medallion fund, which by the way, I think they
02:08:37,360 --> 02:08:42,900 took, it was either the state of New York or the federal government to court to be able to
02:08:42,920 --> 02:08:49,880 have the 401k plan at Rentech be the Medallion fund. No way. Yeah. So like if you work there,
02:08:49,980 --> 02:08:54,740 your 401k is the Medallion fund. That's crazy. So it really doesn't take more than a few years
02:08:54,740 --> 02:08:58,780 before you're set for life. Totally. I mean, depending on your definition of set for life,
02:08:58,840 --> 02:09:05,260 I think it happens very, very quickly. Yep. Okay. So given that though, how do you avoid
02:09:05,260 --> 02:09:10,680 the incentive for a group of talented younger folks to split off and go start their own
02:09:10,680 --> 02:09:12,620 Medallion fund? Right.
02:09:12,920 --> 02:09:17,660 Especially when they all have access to the whole code base. The whole thing is meant to
02:09:17,660 --> 02:09:22,920 function like a university math department where everyone's constantly knowledge sharing because
02:09:22,920 --> 02:09:26,520 we're going to create better peer-reviewed research when we all share all the knowledge
02:09:26,520 --> 02:09:31,120 all the time. You would think that's a super risky thing to give everyone all the keys.
02:09:31,700 --> 02:09:37,820 Right. So I think it's the 44% carry structure that does it. Because basically what you're
02:09:37,820 --> 02:09:42,560 saying is every year, 5% management fee.
02:09:42,920 --> 02:09:48,980 So 5% off the top and then 44% of performance. So let's say Medallion is on the order of,
02:09:49,120 --> 02:09:52,580 call it doubling every year. Let's round that up and just add them and say
02:09:52,580 --> 02:10:02,400 49% of the economic returns in any given year go to the current team and 51% of the economic returns
02:10:02,400 --> 02:10:07,380 go to the tenure base. I was like, what is the equivalent here? I think it's kind of
02:10:07,380 --> 02:10:11,720 like a academic tenure kind of thing. The longer tenure you are at the firm,
02:10:11,880 --> 02:10:12,900 the more you're better at it. So it's like, what is the equivalent here?
02:10:12,900 --> 02:10:18,700 The balance shifts to the LP side of things. And the younger you are at the firm,
02:10:18,700 --> 02:10:24,620 the more your balance is on the GP side of things. But at the end of the day, it's 51-49.
02:10:25,140 --> 02:10:30,860 So there's this very natural value transfer mechanism to keep the people that are working
02:10:30,860 --> 02:10:38,400 in any given year super incentivized. And as you stay there longer, you are paying
02:10:38,400 --> 02:10:41,100 your younger colleagues to work for you.
02:10:41,420 --> 02:10:42,060 Right.
02:10:42,900 --> 02:10:46,720 It's funny. I think it's a good insight that it's structured like a university department tenure.
02:10:47,160 --> 02:10:53,440 Well, I just kept asking myself, why? Why? Why do they have this if there's no outside LPs? And this
02:10:53,440 --> 02:10:58,740 was the best thing I could come up with. And I actually think it's kind of genius.
02:10:59,020 --> 02:11:03,120 Yeah. It's more elegant than it's all one person's money and they're deciding to bonus
02:11:03,120 --> 02:11:07,120 out the current team every year and just give them enough money to make sure you retain them.
02:11:07,480 --> 02:11:11,480 Right. Which is how I think most prop shops work. Like Jane Street is mostly a prop shop.
02:11:11,480 --> 02:11:12,600 I think it is mostly...
02:11:12,900 --> 02:11:17,060 The principal's money. But that's a static situation. It's not like,
02:11:17,540 --> 02:11:20,880 you know, if that were true, then Jim would just own this thing forever.
02:11:21,720 --> 02:11:23,900 And I don't think that's true here at Rentech.
02:11:24,220 --> 02:11:30,240 Yeah. So essentially, David, the real magic is they've got one fund. It's evergreen. And
02:11:30,240 --> 02:11:35,720 when you start at the firm, you're only getting sort of paid the carry amount. But over time,
02:11:36,020 --> 02:11:42,000 you become a meaningful investor in the firm and you sort of shift to that 51%. You're kind of the
02:11:42,000 --> 02:11:42,280 LP.
02:11:42,900 --> 02:11:46,720 And then over time, you eventually graduate out entirely and you're only an LP. And so you're
02:11:46,720 --> 02:11:52,760 right. It's a value transfer mechanism from the old guard to the new guard in a way that is clear,
02:11:53,000 --> 02:11:58,160 well understood, probably tax advantaged versus just doing, I'm the owner and I'm giving everyone
02:11:58,160 --> 02:11:58,980 arbitrary bonuses.
02:11:59,580 --> 02:12:05,900 Yeah. And at the end of the day, I think these three pieces, to me, are the core of this sort
02:12:05,900 --> 02:12:12,880 of tapestry of Rentech. One model that everybody collaborates on together, a super small team,
02:12:12,900 --> 02:12:18,860 where we all know each other and the financial impact that any of us make to that one model is
02:12:18,860 --> 02:12:27,800 great to all of us. And three, this LPGP model with very high carry performance fees that creates
02:12:27,800 --> 02:12:32,160 the right set of incentives, both for new talent on the way in and old talent on the way out.
02:12:32,820 --> 02:12:37,280 Yep. I think that's right. Okay. There's a few other parts of the story that we skipped along
02:12:37,280 --> 02:12:42,480 the way because there was no real good place to put them in. But these are objectively fascinating
02:12:42,480 --> 02:12:42,880 historical pieces. And I think that's a really good way to put them in. And I think that's a really
02:12:42,880 --> 02:12:47,620 historical events that are totally worth knowing about. And the first one is called basket options.
02:12:48,180 --> 02:12:56,640 So the year is 2002. Rentech has 13 years of knowing that they basically have a machine that
02:12:56,640 --> 02:13:02,520 prints money. So what should you do when you have a machine that prints money? Leverage. Now, there
02:13:02,520 --> 02:13:07,060 are all sorts of restrictions around firms like this and how much leverage they can take on. You
02:13:07,060 --> 02:13:12,360 can't just go and say, I'm going to borrow $100 for every dollar of equity capital that I have in
02:13:12,360 --> 02:13:12,560 here.
02:13:12,880 --> 02:13:18,600 So you need to sort of get clever to borrow a whole bunch of money from banks or from any lender
02:13:18,600 --> 02:13:24,140 to basically juice your returns. If, again, you have a money printing machine that's reliable.
02:13:24,660 --> 02:13:28,500 Most people don't. Most people probably shouldn't take leverage because they're just as likely to
02:13:28,500 --> 02:13:33,760 blow the whole thing up as they are to be successful. So basket options. I am going to
02:13:33,760 --> 02:13:37,580 read directly from the man who solved the market because Greg Zuckerman just put it perfectly.
02:13:38,340 --> 02:13:42,460 Basket options are financial instruments whose values are pegged to the performance of a specific
02:13:42,460 --> 02:13:48,180 basket of stocks. While most options are based on an individual stock or a financial instrument,
02:13:48,600 --> 02:13:53,440 basket options are linked to a group of shares. If these underlying stocks rise,
02:13:53,520 --> 02:13:58,080 the value of the option goes up. It's like owning the shares without actually doing so.
02:13:58,540 --> 02:14:02,820 Indeed, the banks who, of course, loaned the money, who put the money in the basket option,
02:14:03,100 --> 02:14:06,740 were legal owners of the shares in the basket. But for all intents and purposes,
02:14:07,120 --> 02:14:11,060 they were Medallion's property. So this is very clever. Medallion's saying, well,
02:14:11,060 --> 02:14:12,240 the way we're going to lever up is,
02:14:12,460 --> 02:14:17,600 there's a basket. We have an option to purchase that basket. Most of the capital in that basket
02:14:17,600 --> 02:14:23,080 is actually the bank's capital. But the bank has hired us to trade the options in the basket. And
02:14:23,080 --> 02:14:30,000 then after a year, when long-term capital gains tax kicks in, we have the option to buy that
02:14:30,000 --> 02:14:34,420 basket. So anyway, all day, Medallion's computer sent automated instructions to the banks,
02:14:34,560 --> 02:14:40,040 sometimes in order a minute or even a second. The options gave Medallion the ability to borrow
02:14:40,040 --> 02:14:42,360 significantly more than it otherwise would be allowed.
02:14:42,460 --> 02:14:47,140 Competitors generally had about $7 of financial instruments for every dollar of cash.
02:14:47,500 --> 02:14:52,620 By contrast, Medallion's option strategy allowed it to have $12.50 worth of financial instruments
02:14:52,620 --> 02:14:56,920 for every dollar of cash, making it easier to trounce rivals, assuming they could keep
02:14:56,920 --> 02:15:00,780 finding profitable trades. When Medallion spied an especially juicy opportunity,
02:15:01,080 --> 02:15:06,840 it could boost leverage, holding close to $20 of asset for every dollar of cash. In 2002,
02:15:07,000 --> 02:15:11,580 Medallion managed over $5 billion, but it controlled over $60 billion of investment positions.
02:15:12,300 --> 02:15:12,440 David, thank you so much for joining us.
02:15:12,460 --> 02:15:16,120 This exposes something we haven't shared yet on the episode, which is it's not just that they
02:15:16,120 --> 02:15:22,020 could find $5 billion worth of profitable trades. It's that they wanted to lever the crap out of
02:15:22,020 --> 02:15:28,060 $5 billion and find $60 billion of profitable trades to make. And basket options gave them
02:15:28,060 --> 02:15:33,000 a legal way to have an incredible amount of leverage in a way that they felt safe about.
02:15:33,540 --> 02:15:40,460 Yeah, the unlevered returns if you were running this strategy would be much lower.
02:15:40,580 --> 02:15:42,400 Yeah. So a big piece of this playbook,
02:15:42,400 --> 02:15:45,680 we didn't talk about is leverage, but every quant fund does leverage. And so
02:15:45,680 --> 02:15:47,920 Renaissance was just more clever than everyone else.
02:15:48,460 --> 02:15:54,160 Yeah. It's an important point, though. Nine out of every 10 companies that we cover on Acquired,
02:15:54,780 --> 02:15:56,680 leverage is zero part of the story.
02:15:56,900 --> 02:15:57,020 Right.
02:15:57,340 --> 02:16:00,560 And for us coming from the world we come from in tech and venture capital,
02:16:01,080 --> 02:16:03,280 leverage is like a dirty word. Like, I'm scared of it.
02:16:03,760 --> 02:16:08,100 Right. I mean, you could imagine, let's say it wasn't, they were right 50.25% of the time,
02:16:08,100 --> 02:16:11,940 but they were right 50.0001% of the time. They would need to do
02:16:11,940 --> 02:16:18,660 a ton of trades in order to generate enough profits. So that's why you need $60 billion of
02:16:18,660 --> 02:16:24,900 cash to actually execute the strategy to produce the returns that they were looking for on $5
02:16:24,900 --> 02:16:29,580 billion of equity. Anyway, there's a second chapter to this, which is it's all well and
02:16:29,580 --> 02:16:34,200 good that this is how they get a bunch of leverage. That's one piece of it. The other piece is they
02:16:34,200 --> 02:16:38,460 thought this was a remarkably tax efficient vehicle. The way that they were filing their
02:16:38,460 --> 02:16:41,760 taxes said, oh, sure, there's stuff in that basket.
02:16:41,940 --> 02:16:46,760 But the thing that we actually own is an option to buy that basket or sell that basket. And we
02:16:46,760 --> 02:16:51,600 only exercise that once every 13 months or so. I don't know the exact number, but something like
02:16:51,600 --> 02:16:56,060 that over a year. And so therefore, we're buying something, we're holding it for a year, we're
02:16:56,060 --> 02:17:00,480 selling it. Oh, of course, there's millions and millions of trades going on inside the basket,
02:17:00,600 --> 02:17:05,040 but we don't own that basket. The banks do. We're just advising them. You can kind of see the logic
02:17:05,040 --> 02:17:11,880 here. Over time, eventually in 2021, the IRS said, no, you made all those trades. That was,
02:17:11,940 --> 02:17:19,200 not a completely separate entity. And so you guys owed $6.8 billion in taxes that you didn't pay.
02:17:19,680 --> 02:17:24,440 You're going to need to pay that with interest, with penalties. And by the way, Jim Simons,
02:17:24,600 --> 02:17:28,500 we're going to want you and the other few partners to really bear the load of that.
02:17:28,500 --> 02:17:35,300 And they did. So for Simons alone, he paid $670 million to the IRS in back taxes for this basket
02:17:35,300 --> 02:17:40,360 option strategy that turned out not to be a long-term capital gain. All right. So numbers
02:17:40,360 --> 02:17:41,920 on the business today. And then we'll be back in just a moment.
02:17:41,920 --> 02:17:46,560 We will dive into power and playbook. So today we've talked about Medallion,
02:17:46,800 --> 02:17:50,880 10 or 15 billion, depending on who you ask. Historically, it was more like 5 or 10 billion.
02:17:51,200 --> 02:17:56,940 The institutional fund is about 60 to 70 billion. And at one point was 100 billion.
02:17:57,400 --> 02:18:02,820 The total carry generated, David, you said is $60 billion. Forbes estimates that Jim Simons alone
02:18:02,820 --> 02:18:07,340 is worth about $30 billion today, which kind of pencils with a bunch of other
02:18:07,640 --> 02:18:10,880 stats over the years that he owned about half of Renaissance.
02:18:10,880 --> 02:18:17,640 The returns, obviously, the Medallion fund generated approximately 66% annualized from
02:18:17,640 --> 02:18:26,300 1988 to 2020 after those fees was about 39% wild. So an interesting thing to understand,
02:18:26,720 --> 02:18:32,420 I ran a hypothetical scenario of how much money do you think Renaissance the business makes a year
02:18:32,420 --> 02:18:39,260 in revenue? And so the institutional fund, let's call it 10% on 60 billion of assets. So that's
02:18:39,260 --> 02:18:40,860 $600 million from fees and $600 million from returns. So that's about $60 billion of assets.
02:18:40,880 --> 02:18:48,460 From performance. So 1.2 billion a year in revenue to the firm from the institutional side of the
02:18:48,460 --> 02:18:52,680 business. Because I always ask myself the question, does that actually matter? They did all this work
02:18:52,680 --> 02:18:58,480 to stand up the institutional side. Who cares? Well, let's say Medallion does their average 66%
02:18:58,480 --> 02:19:09,660 gross on 15 billion. That is 750 million in fees and 4.3 billion on performance. So a total of 5
02:19:09,660 --> 02:19:10,760 billion from Medallion.
02:19:10,880 --> 02:19:16,460 And 1.2 billion from the institutional side of the business. Now, of course, the employees are
02:19:16,460 --> 02:19:20,280 the investors in Medallion. So you could just argue it's actually silly to cut them up. But
02:19:20,280 --> 02:19:23,740 I don't know. It's a $7, $8, $9 billion revenue business.
02:19:24,520 --> 02:19:27,720 Right. Because that's not including the LP return on Medallion.
02:19:27,800 --> 02:19:29,120 A hundred percent. It's not.
02:19:29,260 --> 02:19:32,220 Which again, as we spent a long time talking about, it's all the same thing.
02:19:32,440 --> 02:19:36,980 Yes. But it's kind of interesting just to compare it against other companies to have
02:19:36,980 --> 02:19:40,560 this in the back of your head. This is a $7, $8 billion a year revenue business.
02:19:40,880 --> 02:19:44,920 Now, I think there are a lot of expenses on the infrastructure side.
02:19:45,360 --> 02:19:48,080 Totally. That was another thing I wanted to talk about. The fact that they do,
02:19:48,220 --> 02:19:53,880 let's say Medallion alone. So they have $750 million in fees. I don't think they come close
02:19:53,880 --> 02:19:59,560 to $750 million a year in expenses, but they are running who knows what infrastructure,
02:19:59,740 --> 02:20:04,620 some kind of supercomputing cluster. What does it cost to run one Amazon data center? I mean,
02:20:04,660 --> 02:20:06,420 it's, I think, much smaller scale.
02:20:06,740 --> 02:20:10,080 I don't know. I mean, you're talking about a lot of data here.
02:20:10,360 --> 02:20:10,680 Yeah.
02:20:10,880 --> 02:20:17,120 It says right on their website, they have 50,000 computer cores with 150 gigabits per second of
02:20:17,120 --> 02:20:22,380 global connectivity and a research database that grows by more than 40 terabytes a day.
02:20:22,860 --> 02:20:24,000 That's a lot of data.
02:20:24,000 --> 02:20:28,720 Right. Is that $750 million a year? I don't know, but it's not zero.
02:20:29,120 --> 02:20:33,320 I don't think so. They're certainly not losing money on the fees,
02:20:33,560 --> 02:20:36,440 but there are actual hard costs to this business.
02:20:36,780 --> 02:20:40,400 Right. I wonder too, if the fee element,
02:20:40,400 --> 02:20:45,860 of Medallion, basically pays the base salaries for the current team.
02:20:46,720 --> 02:20:52,560 That feels like it's right. If you're someone who has done a data center build out before,
02:20:52,780 --> 02:20:58,620 or has any way to sort of back into what the costs of Medallion's operating expenses are on the
02:20:58,620 --> 02:21:03,400 compute and data and network side, we would love to hear from you. Hello at Acquired.fm.
02:21:04,200 --> 02:21:05,100 Okay. Power?
02:21:05,820 --> 02:21:09,040 Power. This is a fun one.
02:21:09,040 --> 02:21:10,340 Yeah. So listeners who are new,
02:21:10,400 --> 02:21:15,820 this is Hamilton Helmer's framework from the book Seven Powers. What is it that enables a business
02:21:15,820 --> 02:21:20,980 to achieve persistent differential returns to be more profitable than their closest competitor on
02:21:20,980 --> 02:21:26,440 a sustainable basis? And the seven are counter-positioning, scale economies, switching
02:21:26,440 --> 02:21:33,140 costs, network economies, process power, branding, and cornered resource. And David,
02:21:33,760 --> 02:21:39,960 my question to you to open this section is specifically about Rentech's lifelong non-competes.
02:21:40,400 --> 02:21:44,800 That feels like a big reason that they maintain their competitive advantage.
02:21:44,980 --> 02:21:48,320 And I'm curious, if you agree with that, what would you put that under?
02:21:48,320 --> 02:21:54,340 Well, I think it's lifelong NDAs and non-competes as long as the state of New York
02:21:54,340 --> 02:21:58,860 legally allows for, but that is not lifetime. I've heard various figures,
02:21:59,280 --> 02:22:04,960 six years, five years, something like that. I mean, at the end of the day, non-competes are more like,
02:22:05,140 --> 02:22:09,820 what is one side willing to go to court over? But the reality is,
02:22:10,400 --> 02:22:14,840 people don't leave. People don't leave, period. And people especially don't leave and start their
02:22:14,840 --> 02:22:21,240 own firms. I was thinking about this in the middle of the night, and I think there's three
02:22:21,240 --> 02:22:30,920 layers to the effective non-compete that happens with Rentech. There's the legal layer, the base
02:22:30,920 --> 02:22:34,520 layer that you're talking about. It's like the agreements you sign. Then there's the economic
02:22:34,520 --> 02:22:40,040 layer of what we spent a long time talking about in Tapestry of, it would just be dumb
02:22:40,400 --> 02:22:45,220 leave. You are better off staying there as part of that team with a smaller number of people
02:22:45,220 --> 02:22:50,640 than going to Sigma with a lot more people. I think that's the next level of, and then I think
02:22:50,640 --> 02:22:55,000 the highest level is just probably the social layer. You're there with the smartest people
02:22:55,000 --> 02:22:59,680 in the world in a collegial atmosphere where you're all working hard on something that has
02:22:59,680 --> 02:23:04,320 direct impact on you. Right. It's your community. It's your community. Totally. You're not in New
02:23:04,320 --> 02:23:10,240 York City. You're not in the Hamptons. You're not in Silicon Valley. You are selecting into
02:23:10,240 --> 02:23:17,140 that. I think if that's what you want, what better place in the world? All right. Classify it.
02:23:17,140 --> 02:23:22,520 What power does that fall under? Well, I think the people specifically you would
02:23:22,520 --> 02:23:27,220 put into cornered resource, but I'm not actually sure that fully captures it here.
02:23:27,680 --> 02:23:32,960 I was thinking more process power because I think it is the combination of the people
02:23:32,960 --> 02:23:40,220 and the model and the incentive structures. Yep. I think that's right. I also had my,
02:23:40,240 --> 02:23:45,540 biggest one being process power. You actually can develop intricate knowledge of how a system
02:23:45,540 --> 02:23:50,540 works and then build processes around that that are hard to replicate elsewhere. I think these
02:23:50,540 --> 02:23:55,100 systems have been layered over time also, where anyone who's come into the firm in the last five
02:23:55,100 --> 02:24:02,300 years doesn't know how it works start to finish. I didn't ask anyone to verify that, but it's over
02:24:02,300 --> 02:24:09,600 10 million lines of code. And the level of complexity of the system of when it's putting
02:24:09,600 --> 02:24:14,920 on trades, what trades is putting on, why, the speed at which they need to happen. I actually
02:24:14,920 --> 02:24:20,480 don't think anyone holds the whole model in their head. And so I think there's process power just
02:24:20,480 --> 02:24:27,100 because it's 30 plus years of complexity that's been built up. Yep. I totally agree with that,
02:24:27,360 --> 02:24:32,720 particularly in the model itself. I mean, maybe you could argue the model is a cornered resource.
02:24:33,240 --> 02:24:38,420 I am going to argue that the data is a cornered resource. I don't know for sure about the model.
02:24:38,420 --> 02:24:39,320 Maybe. I mean,
02:24:39,320 --> 02:24:43,580 I guess that's the same thing as saying the knowledge of what the 10 million lines of code
02:24:43,580 --> 02:24:48,860 does. That's the model. But I actually think the fact that they have clean data and they've been
02:24:48,860 --> 02:24:54,840 creating systems, like they have the best PhDs in the world thinking about data cleaning. That's
02:24:54,840 --> 02:25:02,820 not a sexy job. And yet they have probably the treasure trove of historical market data in the
02:25:02,820 --> 02:25:08,960 best format that nobody else has. That's an actual cornered resource. I have a couple nuances on this.
02:25:09,320 --> 02:25:13,940 One, I think it probably is true that they have better data than any other firm,
02:25:14,140 --> 02:25:19,560 thanks to Sandor Strauss and the work that he started doing in the 80s before anybody else
02:25:19,560 --> 02:25:27,540 was really doing this. So they have that and other firms don't. That said, certainly all the
02:25:27,540 --> 02:25:33,820 other quant firms are throwing untold resources at all this too. Right. They want to do this
02:25:33,820 --> 02:25:38,620 and money is not the issue. So in chatting with a few folks
02:25:38,620 --> 02:25:47,080 about this episode, I had more than one person say to me, there's two ways that rent tech could
02:25:47,080 --> 02:25:55,320 work. And one version of how it works is they discovered something 20 plus years ago that is
02:25:55,320 --> 02:26:00,140 a timeless secret and they've been trading on that for 20 plus years. Right. There's one particular
02:26:00,140 --> 02:26:04,380 relationship between types of equities that they've just been exploiting and no one can figure out
02:26:04,380 --> 02:26:08,540 except them. Right. And that may entirely be possible. Isn't that crazy?
02:26:08,620 --> 02:26:12,880 Right. Now rent tech will say, they will all say that is a hundred percent, not the way that it
02:26:12,880 --> 02:26:17,260 works. It's not that at all. If that were the way that it works, they would of course still say that
02:26:17,260 --> 02:26:21,200 because they don't want anybody to know. Right. Don't look at the relationship between soybean
02:26:21,200 --> 02:26:28,200 futures and GM. Just don't do it. Right. So let's accept that there is a possibility that that might
02:26:28,200 --> 02:26:36,040 be true. More likely though, is that what rent tech does say is true, which is no, there is no
02:26:36,040 --> 02:26:42,140 holy grail. What we do here is we completely reinvent the whole system continuously on a
02:26:42,140 --> 02:26:49,440 two-year cycle. Two years is kind of what I heard that the model is fully restructured every two
02:26:49,440 --> 02:26:54,200 years. It's not like on a date every two years, it's being restructured every day, but collectively
02:26:54,200 --> 02:26:58,400 it's about a two-year cycle. So that would be an argument then that the people actually could,
02:26:58,720 --> 02:27:03,160 with five people left, they probably could go recreate it and all they would need is the data.
02:27:03,460 --> 02:27:05,840 It's also an argument that there is no actual,
02:27:06,040 --> 02:27:09,900 cornered resource here in terms of either the model itself and maybe not the data either.
02:27:10,660 --> 02:27:14,280 I bet the data is though. Let's say you've been working there for 10 years.
02:27:14,620 --> 02:27:21,360 You don't know how the 1955 soybean futures data ended up in the database. Even if you're used to
02:27:21,360 --> 02:27:25,600 using that data and you're able to go recreate the model elsewhere, you don't know how it
02:27:25,600 --> 02:27:30,060 originally found its way in. I think that's fair. I think there
02:27:30,060 --> 02:27:36,020 might also be some argument to the data that that older data is helpful, but its value decays,
02:27:36,040 --> 02:27:42,040 as markets evolve. The broader point I want to make here is just that every other major
02:27:42,040 --> 02:27:45,960 quant firm out there is also spending hundreds of millions, if not billions on this stuff too.
02:27:46,560 --> 02:27:50,600 And people are looking for alt data everywhere. The Bridgewaters of the world are paying gobs of
02:27:50,600 --> 02:27:55,640 money for things that you would never dream could possibly have an effect on the stock market. And
02:27:55,640 --> 02:27:59,040 yet they're paying millions or tens of millions or hundreds of millions of dollars for it.
02:27:59,440 --> 02:28:04,500 Yep. So I think we can rule out scale economies for sure. If anything,
02:28:04,500 --> 02:28:06,000 they're anti-scale economies.
02:28:06,040 --> 02:28:12,160 Oh, yes. Totally. There's dis-economies of scale. Your strategies stop working when you get too much
02:28:12,160 --> 02:28:12,580 AUM.
02:28:13,040 --> 02:28:18,020 Yep. You get slippage. I don't think there's any network economies here. I mean, they literally
02:28:18,020 --> 02:28:19,140 don't talk to anybody.
02:28:21,000 --> 02:28:28,160 Although, well, they do have some very well-established relationships with electronic
02:28:28,160 --> 02:28:32,940 brokerages and different players in the trade execution chain. I think they have very good
02:28:32,940 --> 02:28:35,940 trade execution and very fast marketability.
02:28:36,040 --> 02:28:36,540 Yeah. I mean, I think they have very good trade execution and very fast marketability.
02:28:36,540 --> 02:28:39,620 Their ability to pull data out of the market is very high quality.
02:28:39,960 --> 02:28:42,180 Do you think it's actually better than their competitors though?
02:28:42,340 --> 02:28:44,140 I don't know. That's probably not the secret sauce.
02:28:44,340 --> 02:28:45,320 Yeah. I don't think so.
02:28:45,700 --> 02:28:46,660 It's the table stakes.
02:28:47,020 --> 02:28:51,860 Switching costs I don't think apply. Branding maybe applies in their ability to raise money
02:28:51,860 --> 02:28:54,640 for the institutional funds, but that's not a big part of the business.
02:28:55,000 --> 02:28:58,600 The fee stream on the institutional fund may entirely belong to branding.
02:28:59,100 --> 02:28:59,420 Yes.
02:28:59,840 --> 02:29:03,800 But I think there's a lot of public equity firms and a lot of hedge funds that have a lot of
02:29:03,800 --> 02:29:06,020 branding power that have, on average, a lot of money. And I think that's a big part of the
02:29:06,040 --> 02:29:11,640 market returns with decent sharp ratios and are able to raise because they've built a brand.
02:29:12,080 --> 02:29:12,200 Yep.
02:29:12,600 --> 02:29:13,780 Venture firms the same way.
02:29:14,120 --> 02:29:18,460 Totally. So for me, this kind of leaves counter-positioning. I actually think there's
02:29:18,460 --> 02:29:21,880 some counter-positioning here, and I think we're going to have two episodes in a row
02:29:21,880 --> 02:29:23,680 of counter-positioning at scale.
02:29:24,100 --> 02:29:27,580 Tell me about your counter-positioning. Who is being counter-positioned in what way?
02:29:28,340 --> 02:29:32,440 They're direct competitors in the market, the other quant firms. And when I say direct
02:29:32,440 --> 02:29:35,380 competitors, I obviously don't mean for LP dollars. I mean for like
02:29:35,380 --> 02:29:35,980 the same amount of money that's being put into the market.
02:29:35,980 --> 02:29:36,020 So they're not going to be able to raise money.
02:29:36,020 --> 02:29:37,760 They're not going to be able to raise money for the same type of trading activity.
02:29:38,220 --> 02:29:39,620 Like they're counterparties in trades.
02:29:40,120 --> 02:29:44,480 I don't think they are counterparties. I think they are all seeking to exploit
02:29:44,480 --> 02:29:47,960 similar types of trades. I think the counterparties are the people there,
02:29:48,220 --> 02:29:49,960 the dentists that they're taking advantage of.
02:29:50,120 --> 02:29:52,620 Well, but quant funds are often counterparties to each other.
02:29:52,940 --> 02:29:58,680 That's true. But I think, yes, adversaries in finding the similar types of trades. And I think
02:29:58,680 --> 02:30:05,780 the counter-positioning for Rentech, or for Medallion specifically, is one,
02:30:05,780 --> 02:30:10,960 and I do think the single model approach versus the multi-model, multi-strategy approach that
02:30:10,960 --> 02:30:14,660 most others have, does have benefits like I was talking about in the tapestries.
02:30:14,660 --> 02:30:22,400 But I think also, and maybe bigger, is every incentive at Rentech is fully aligned
02:30:22,400 --> 02:30:29,660 to optimize fund size for performance in a way that is not true just about everywhere else.
02:30:30,380 --> 02:30:35,720 I think they have the most incentive of anybody to truly maximize,
02:30:35,720 --> 02:30:37,820 maximize performance we're able to achieve.
02:30:38,400 --> 02:30:43,240 Right. Even though the dollars would continue to rise because they get fee dollars from more
02:30:43,240 --> 02:30:49,420 money in the door, they are incentivized in a unique way that makes it so they're not willing
02:30:49,420 --> 02:30:53,940 to trade the dampener on performance to get those dollars.
02:30:54,560 --> 02:31:00,460 Yes. Particularly because it's all the same people on the GP and LP side.
02:31:00,560 --> 02:31:04,980 Oh, we keep going round and round that axle. I loosely buy the counter-positioning thing.
02:31:05,200 --> 02:31:05,700 I just don't.
02:31:05,700 --> 02:31:09,300 I just think the answer is disgustingly simple and kind of annoying here, which is
02:31:09,300 --> 02:31:14,320 they're just better than everyone else at this particular type of math and machine learning,
02:31:14,320 --> 02:31:17,120 and they've been doing it for longer, so they're just going to keep beating you.
02:31:17,720 --> 02:31:23,140 Oh, that's another argument I heard from people, in that Rentech basically is a math department
02:31:23,140 --> 02:31:26,800 in a way that none of these other firms are.
02:31:26,980 --> 02:31:27,720 It could be culture.
02:31:28,080 --> 02:31:28,800 Yeah, it could be culture.
02:31:29,200 --> 02:31:31,680 I mean, honest to God, it could just be that the culture is set up
02:31:31,680 --> 02:31:35,180 in a way that continues to attract the right people and incentivize them
02:31:35,180 --> 02:31:37,640 in a sort of fake altruistic way.
02:31:38,100 --> 02:31:42,340 Like, this is just a fun place to do my work, and yeah, the outcome is getting really rich,
02:31:42,520 --> 02:31:44,440 but I wouldn't go work at Citadel.
02:31:45,060 --> 02:31:46,640 Yeah, I think that could be.
02:31:46,820 --> 02:31:48,380 So maybe that feeds into process power.
02:31:48,800 --> 02:31:49,000 Yeah.
02:31:49,580 --> 02:31:53,380 Okay, for me, it is some combination of process power and counter-positioning,
02:31:53,440 --> 02:31:55,280 and I don't think it's any of the other powers.
02:31:55,660 --> 02:31:58,200 For me, it is process power and cornered resource.
02:31:58,740 --> 02:31:59,980 Yeah. Okay, I buy that.
02:32:00,140 --> 02:32:05,020 And a thing that's not captured in seven powers is tactical, like execution.
02:32:05,180 --> 02:32:09,400 The whole point of seven powers is strategy is different than tactics.
02:32:09,520 --> 02:32:10,920 And I think legitimately,
02:32:11,200 --> 02:32:15,820 Rentech may just have persistently been able to out-execute their competitors.
02:32:16,220 --> 02:32:18,460 There's part of it that's just like, they're smarter than you.
02:32:19,100 --> 02:32:19,120 Yeah.
02:32:19,860 --> 02:32:25,860 Well, if you buy the whole thing gets reinvented continuously every two years, then yes.
02:32:26,400 --> 02:32:27,680 And there's remnant knowledge.
02:32:27,760 --> 02:32:34,320 Like, if you started building a machine learning system in 19-whatever it was,
02:32:35,180 --> 02:32:39,140 64, you're going to be really good at machine learning today.
02:32:39,220 --> 02:32:42,740 And the people that you've been spending time with for the last 15 years,
02:32:42,800 --> 02:32:46,640 learning all of your historical knowledge and working in your systems are also going
02:32:46,640 --> 02:32:50,460 to be better at machine learning than probably the other people who are out in the world,
02:32:50,460 --> 02:32:55,560 learning it from people that just got inspired to start learning machine learning
02:32:55,560 --> 02:32:57,500 based on the new hotness.
02:32:57,960 --> 02:32:59,620 So learnings compound is my answer.
02:33:00,120 --> 02:33:00,400 Great.
02:33:01,060 --> 02:33:01,400 Okay.
02:33:02,220 --> 02:33:02,900 Playbook.
02:33:03,160 --> 02:33:05,120 So in addition to the three-part David.
02:33:05,120 --> 02:33:07,540 David Rosenthal tapestry that you have woven.
02:33:07,800 --> 02:33:08,720 I have nothing more to add.
02:33:09,120 --> 02:33:12,520 There are a handful of things that I think are worth hitting.
02:33:12,700 --> 02:33:17,580 So the first one is signal processing is signal processing is signal processing.
02:33:18,020 --> 02:33:24,880 They, by not caring about the underlying assets, they literally don't trade on fundamentals,
02:33:25,040 --> 02:33:27,600 except in the institutional fund when they trade on fundamentals a little bit.
02:33:27,600 --> 02:33:31,240 They use price-to-earnings ratios and stuff like that in the institutional fund,
02:33:31,360 --> 02:33:34,400 which is kind of funny because that's a completely different skill set.
02:33:34,400 --> 02:33:40,400 But if you just look at medallion, it's all just abstract numbers.
02:33:41,240 --> 02:33:45,360 You don't actually have to care about what underlies those numbers.
02:33:45,540 --> 02:33:50,500 You just have to look for whether it's linear regression or any of the fancier stuff that
02:33:50,500 --> 02:33:52,920 they do, just relationships between data.
02:33:53,320 --> 02:34:00,660 And once you reduce it to that, it is so brilliant that they can just recruit from any field.
02:34:00,660 --> 02:34:03,660 It's not relevant how someone has done sophisticated.
02:34:04,400 --> 02:34:08,620 Signal processing in the past, whether it's being an astronomer and trying to denoise
02:34:08,620 --> 02:34:12,480 a quote-unquote photo of a star super far away,
02:34:12,580 --> 02:34:16,540 or whether they've tried to do like natural language processing, it's just signal.
02:34:16,840 --> 02:34:23,020 There's this really funny line that Jim and Peter and others will say when asked about why
02:34:23,020 --> 02:34:26,340 they only hire academics and not from Wall Street and whatnot. And they're like,
02:34:26,340 --> 02:34:32,900 well, we found it's easier to teach smart people the investing business than teach investing people
02:34:32,900 --> 02:34:33,720 how to be smart.
02:34:34,400 --> 02:34:37,380 Right. That's ridiculous. They don't teach anybody anything about investing.
02:34:37,700 --> 02:34:42,600 They're just doing signal processing. I bet at least half the people at Rentech
02:34:42,600 --> 02:34:45,000 on the research side could not read a balance sheet.
02:34:45,220 --> 02:34:48,220 It's so funny. It's a whole bunch of people who are in the investment business,
02:34:48,300 --> 02:34:49,180 none of which are investors.
02:34:49,640 --> 02:34:49,880 Yes.
02:34:50,260 --> 02:34:55,840 Another one that you can decide if this fits or not. I was thinking a lot about complex
02:34:55,840 --> 02:35:00,460 adaptive systems. It's always been on my mind since we had the NZS Capital guys on a few years
02:35:00,460 --> 02:35:03,560 ago and read their work and the Santa Fe Institute's work on this.
02:35:04,040 --> 02:35:04,220 Right.
02:35:04,220 --> 02:35:04,280 Right.
02:35:04,280 --> 02:35:04,300 Right.
02:35:04,300 --> 02:35:04,360 Right.
02:35:04,360 --> 02:35:04,380 Right.
02:35:04,380 --> 02:35:04,400 Right.
02:35:04,400 --> 02:35:08,640 In a complex adaptive system, it's really difficult to actually understand how one thing
02:35:08,640 --> 02:35:14,060 affects everything else. Because the idea is the relationships are so combinatorially complex that
02:35:14,060 --> 02:35:19,240 you can't deterministically nail down this one thing as the cause of that other thing. It's
02:35:19,240 --> 02:35:25,060 the butterfly flapping its wings. But there are relationships between entities that you can't
02:35:25,060 --> 02:35:30,400 understand or see on the surface. Do you remember way back when we did our second NVIDIA episode,
02:35:30,400 --> 02:35:34,360 I opened with the idea that when I was a kid, I always used to look at fire,
02:35:34,360 --> 02:36:04,140 and think like, if you actually knew the composition of the atoms in the wood, and you actually knew the way the wind was blowing, and you actually knew that, like all the, could you actually model the fire? And when I was a kid, and you always just assumed no. But actually, the answer is yes. This is a known thing of what will happen when you light this log on fire for the next three hours. And can you see exactly the flames? I think Rentech has basically, they haven't figured that out for the market. They can't predict the future.
02:36:04,140 --> 02:36:34,120 But if they have a 50.01% chance of being correct, then they can sort of take a complex adaptive system and say, we don't really care that it's a complex adaptive system. Our models understand enough about the relationships between all these entities, that we're just going to run the simulation a bunch of times, and we're going to be profitable enough from all the little pennies that we're collecting on all the little coin flips, where we have a slight edge over and over and over and over again, that they're sort of the closest in the world to being able to
02:36:34,140 --> 02:36:44,020 actually predict how the complex adaptive system of the market will work. Now, I don't think they can back out to it. No person could explain it, but I think their computers can.
02:36:44,180 --> 02:37:04,020 Yes. And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:37:04,140 --> 02:37:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:37:34,140 --> 02:38:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:38:04,140 --> 02:38:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:38:34,140 --> 02:39:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:39:04,140 --> 02:39:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:39:34,140 --> 02:40:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:40:04,140 --> 02:40:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:40:34,140 --> 02:41:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:41:04,140 --> 02:41:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:41:34,140 --> 02:42:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:42:04,140 --> 02:42:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:42:34,140 --> 02:43:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:43:04,140 --> 02:43:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:43:34,140 --> 02:44:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:44:04,140 --> 02:44:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:44:34,140 --> 02:45:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:45:04,140 --> 02:45:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:45:34,140 --> 02:46:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:46:04,140 --> 02:46:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:46:34,140 --> 02:47:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:47:04,140 --> 02:47:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:47:34,140 --> 02:48:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:48:04,140 --> 02:48:34,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:48:34,140 --> 02:49:04,120 And I think when I've heard people from Rentech talk about this, they will all say, the model does not actually understand the market, but it can predict, and we can be so confident in its predictions about what the market will do.
02:49:04,140 --> 02:49:11,840 But it's literally the same business model as a casino. You have a slight edge, and you let a whole bunch of patrons come in and lose money to you in your slight edge.
02:49:11,940 --> 02:49:22,660 Well, where I was going with the service provider, I think casinos are service providers. They are providing entertainment to their customers. Everybody knows that the games are stacked in the casino's favor.
02:49:23,360 --> 02:49:34,120 Similarly, I think you could make an argument, and I think this is probably quite accurate, that Rentech and all other quant firms like them are providing a service to the casino.
02:49:34,140 --> 02:49:57,340 Absolutely. That is the undeniable, yes, quant funds create value in the world thing, which I think it's very easy to say quant funds provide no value, because it's like it's zero-sum, they're not actually providing the capital to businesses to do something with.
02:49:57,340 --> 02:50:02,440 They're purely looking to do an arbitrage or any of the strategies we've talked about this episode.
02:50:03,060 --> 02:50:04,120 But you're totally right.
02:50:04,140 --> 02:50:23,420 You're totally right that there is a value to market liquidity. Creating more depth to a market makes it so that if we go back to the era that Renaissance was started, there's no chance that retail is able to function like it does today with zero transaction fees and people able to invest in all these different companies at near real time.
02:50:23,420 --> 02:50:33,280 And any single one of us can go buy a security in just about any market at just about any time of day.
02:50:33,520 --> 02:50:34,120 Pretty.
02:50:34,140 --> 02:50:39,360 Pretty much instantaneously and get a very, very, very granular price on it.
02:50:39,760 --> 02:50:39,960 Yep.
02:50:40,180 --> 02:50:41,520 None of which used to be true.
02:50:42,000 --> 02:50:42,220 Nope.
02:50:42,760 --> 02:50:51,860 The fact that there is a whole bunch of quant funds, hedge funds out there that are ready to be willing counterparties to anyone who wants to trade, that is a service.
02:50:52,120 --> 02:50:52,500 You're right.
02:50:53,120 --> 02:50:55,380 They're also not all medallion.
02:50:55,700 --> 02:50:58,680 They actually don't all have an edge, even though they might purport to.
02:50:59,080 --> 02:51:00,980 Lots of them are going to lose money to you.
02:51:01,220 --> 02:51:01,660 Right.
02:51:01,920 --> 02:51:02,900 Lots of them lose money.
02:51:03,360 --> 02:51:04,120 You too, listeners.
02:51:04,300 --> 02:51:04,920 Could beat the market.
02:51:05,220 --> 02:51:06,040 Not investment advice.
02:51:06,240 --> 02:51:06,880 Please don't try.
02:51:07,380 --> 02:51:07,600 Right.
02:51:07,600 --> 02:51:09,800 On average, medallion will not lose money to you.
02:51:09,940 --> 02:51:17,020 But, you know, there are plenty of other hedge funds out there and high frequency shops and counterparties for you where you could take them.
02:51:17,240 --> 02:51:18,500 It's just not Jim Simons.
02:51:22,840 --> 02:51:26,300 There's this great, great vignette at the end of Greg's book.
02:51:26,640 --> 02:51:33,560 It was during one of the like sell-offs in the mid-20-teens in the market where Jim calls the head of his family office.
02:51:33,560 --> 02:51:40,760 He's, you know, long retired from rent tech at this point, calls the head of his family office and says, what should we do with all the sell-off in the market?
02:51:40,820 --> 02:51:42,620 And it's like, you're Jim Simons.
02:51:42,740 --> 02:51:42,980 Right.
02:51:43,620 --> 02:51:44,380 You're Jim Simons.
02:51:44,460 --> 02:51:45,140 What should we do?
02:51:45,280 --> 02:51:46,140 What should we do?
02:51:46,460 --> 02:51:46,660 Yeah.
02:51:48,760 --> 02:51:49,160 Yeah.
02:51:49,240 --> 02:51:50,240 All humans are fallible.
02:51:50,820 --> 02:51:51,140 Totally.
02:51:51,840 --> 02:51:53,940 A couple of other are squintable.
02:51:54,160 --> 02:51:55,280 The value creation exists.
02:51:55,820 --> 02:52:01,920 It's easy to knock that all these smart people are going into finance and you wish they were doing something more productive for the world.
02:52:01,920 --> 02:52:06,600 At the end of the day, humans are going to do what they are incented to do.
02:52:07,180 --> 02:52:12,460 And so absent a larger global concern that is incredibly motivating to people.
02:52:12,560 --> 02:52:20,360 I mean, you look at World War II, people's level of patriotism and wanting to go save the world from evil was a huge, unbelievable motivating factor to move mountains.
02:52:20,980 --> 02:52:29,780 When that is absent or when people feel that there's some existential thing that is absent, they're going to go do what's best for them and their family.
02:52:29,780 --> 02:52:31,780 And if they're an empire builder, go build empires.
02:52:32,480 --> 02:52:34,800 And if they're a fierce capitalist, go make a bunch of money.
02:52:35,260 --> 02:52:38,000 And so the system is set up the way that it is.
02:52:38,040 --> 02:52:39,360 So, like, you can be mad about that.
02:52:40,120 --> 02:52:45,780 Given that, okay, people are going to go engage in quantitative finance as a lucrative profession.
02:52:46,520 --> 02:52:49,860 Fortunately, there's a bunch of valuable stuff that comes out of that.
02:52:49,860 --> 02:53:01,900 And I think that is often missed is that these really lucrative professions and businesses can often produce R&D that becomes valuable elsewhere.
02:53:02,460 --> 02:53:04,780 For example, we just did this big NVIDIA series.
02:53:05,360 --> 02:53:09,060 What do you think Mellanox was used for before large language models?
02:53:09,600 --> 02:53:10,380 Oh, yes.
02:53:10,960 --> 02:53:16,920 This is such a really mind-blowing point here in value creation, value capture.
02:53:17,480 --> 02:53:17,980 Go for it.
02:53:18,060 --> 02:53:18,440 Take it away.
02:53:18,440 --> 02:53:24,960 Well, there's not much to it other than a huge amount of InfiniBand was used by high-frequency trading firms.
02:53:25,360 --> 02:53:29,680 And I don't know for sure, but I kind of think Mellanox built their business on quant finance.
02:53:30,320 --> 02:53:30,500 Yes.
02:53:31,040 --> 02:53:31,440 That's one of...
02:53:31,920 --> 02:53:32,680 Of many examples.
02:53:33,000 --> 02:53:35,000 But now, you know, that has limits.
02:53:35,480 --> 02:53:39,800 But I think it goes overlooked that there's a lot of technology innovation here.
02:53:40,240 --> 02:53:40,460 Yep.
02:53:40,860 --> 02:53:42,360 These are all great points.
02:53:42,540 --> 02:53:43,840 They all came up in the research.
02:53:44,320 --> 02:53:46,740 I totally agree with all of them.
02:53:46,920 --> 02:53:56,400 It is, in my opinion, false to say that quantitative finance does not create value for the world.
02:53:56,640 --> 02:53:58,280 It definitely does, in my opinion.
02:53:58,760 --> 02:54:01,100 But does it create anywhere near as much as it captures?
02:54:01,940 --> 02:54:06,080 That said, they're really, really good at value capture.
02:54:06,620 --> 02:54:07,020 Yes.
02:54:07,460 --> 02:54:09,040 This is not Wikipedia here.
02:54:09,400 --> 02:54:12,300 This is about as far away on the spectrum as you can get.
02:54:12,420 --> 02:54:19,960 There's a great Always Sunny in Philadelphia where Frank, Danny DeVito, sort of goes back to his whatever business he founded in the 80s.
02:54:20,000 --> 02:54:22,260 And he's, like, dressing in his pinstripes and stuff again.
02:54:22,320 --> 02:54:23,140 And he's taken back over.
02:54:23,200 --> 02:54:24,140 He brings Charlie with him.
02:54:24,540 --> 02:54:26,920 And Charlie, you know, he's like, so, Frank, what is the business?
02:54:27,220 --> 02:54:28,140 What do we do here?
02:54:28,620 --> 02:54:29,680 What does the business make?
02:54:30,440 --> 02:54:30,940 And Danny...
02:54:30,940 --> 02:54:32,800 Danny DeVito looks at him and he goes, what do you mean?
02:54:33,040 --> 02:54:33,640 We make money.
02:54:34,160 --> 02:54:34,900 He's like, no, no.
02:54:34,940 --> 02:54:35,840 Like, what do you build?
02:54:36,060 --> 02:54:37,340 He goes, we build wealth.
02:54:37,680 --> 02:54:41,320 I think that's a pretty good meme for kind of what's going on here.
02:54:42,040 --> 02:54:42,060 Yeah.
02:54:42,560 --> 02:54:42,960 Totally.
02:54:43,380 --> 02:54:45,020 Very, very good at value capture, too.
02:54:45,320 --> 02:54:45,560 Yes.
02:54:46,400 --> 02:54:46,780 Okay.
02:54:47,140 --> 02:54:48,020 Bear, bull.
02:54:48,220 --> 02:54:52,160 So, this was a section that we had for a long time that we did not put in the last episode.
02:54:52,260 --> 02:54:53,420 And boy, did we hear about it.
02:54:53,500 --> 02:54:56,540 So, listeners, thank you so much for expressing your concern.
02:54:57,100 --> 02:54:59,920 Bear versus bull is unkilled and it is back.
02:55:00,400 --> 02:55:00,820 Resurrected.
02:55:00,940 --> 02:55:01,500 Like a phoenix.
02:55:02,180 --> 02:55:02,580 Resurrected.
02:55:03,060 --> 02:55:05,940 However, this is about the lamest episode to resurrect it on.
02:55:06,200 --> 02:55:07,740 What's the bull case for Rentech?
02:55:07,960 --> 02:55:10,780 Past performance is an indicator of future success.
02:55:10,920 --> 02:55:11,100 Right.
02:55:11,160 --> 02:55:13,360 Like, they're going to keep attracting all the smartest people in the world.
02:55:13,360 --> 02:55:16,980 They're going to have the ability to keep their incredibly unique culture.
02:55:17,320 --> 02:55:23,020 They're not going to get tempted to let the business of institutional funds become the
02:55:23,020 --> 02:55:23,860 dominant business.
02:55:24,640 --> 02:55:27,240 You know, keep on keeping on is basically the bull case.
02:55:27,440 --> 02:55:29,700 Maybe that they're actually still ahead.
02:55:30,180 --> 02:55:30,720 The bull case.
02:55:30,720 --> 02:55:37,860 It's for the GP and LP stakeholders in Medallion, which is, I don't know, 500 people in the
02:55:37,860 --> 02:55:41,660 world and none of the rest of us can get any exposure to it.
02:55:42,140 --> 02:55:42,280 Yeah.
02:55:42,540 --> 02:55:44,580 The bear case is things are changing.
02:55:44,920 --> 02:55:49,740 And I think things are changing basically on any axis is the bear case for them.
02:55:49,860 --> 02:55:52,160 So, things are changing where competitors are catching up.
02:55:52,640 --> 02:55:53,000 Maybe.
02:55:53,440 --> 02:55:58,860 Maybe the fact that the tech industry has figured out these large language models, maybe
02:55:58,860 --> 02:56:00,340 that trickles into.
02:56:00,720 --> 02:56:02,500 Making it easier to compete with Rentech.
02:56:02,620 --> 02:56:04,860 It's a blurry line, but it is plausible.
02:56:05,040 --> 02:56:07,680 Like, maybe Rentech actually was here a decade before everyone else.
02:56:07,760 --> 02:56:09,720 And now everyone else has arrived to the party.
02:56:10,880 --> 02:56:13,540 There's things that are changing maybe about their culture.
02:56:13,840 --> 02:56:15,920 Like, Jim Symes has been gone for a long time.
02:56:16,400 --> 02:56:18,880 Bob Mercer is no longer a co-CEO.
02:56:19,220 --> 02:56:20,660 Peter Brown is a co-CEO.
02:56:20,720 --> 02:56:24,900 And they just announced that they're making the guy who was in charge of the institutional
02:56:24,900 --> 02:56:29,320 funds, David Lippe, he is becoming a co-CEO as well.
02:56:29,740 --> 02:56:30,700 So, maybe there's a.
02:56:30,720 --> 02:56:34,840 There's a bear case around that, that someone from the institutional side of the house is
02:56:34,840 --> 02:56:40,540 becoming the current co-CEO and maybe eventually CEO if you believe the medallion is the special
02:56:40,540 --> 02:56:44,080 thing and the institutional funds are sort of a blemish on the business.
02:56:44,700 --> 02:56:48,820 You know, they're the Hermes Apple Watch strap in David's parlance.
02:56:49,620 --> 02:56:51,020 Maybe that's a bear case.
02:56:51,320 --> 02:56:56,320 Maybe there's a bear case that their talent is becoming kind of the same as everyone else's
02:56:56,320 --> 02:56:56,620 talent.
02:56:56,620 --> 02:57:00,700 When you look on LinkedIn, I recognize a lot of the companies that people worked at.
02:57:00,700 --> 02:57:02,980 Who are more junior at Rentech.
02:57:03,200 --> 02:57:07,200 And in the past, I think it would have been all people just out of university research
02:57:07,200 --> 02:57:07,800 shops.
02:57:08,000 --> 02:57:14,360 So, I think if it's true that they're starting to see the same talent flow as everyone else,
02:57:14,380 --> 02:57:15,240 that would be concerning.
02:57:15,640 --> 02:57:19,180 These things are all sort of narratives you can concoct and really no way to know if they're
02:57:19,180 --> 02:57:19,680 true or not.
02:57:20,020 --> 02:57:20,100 Right.
02:57:20,400 --> 02:57:23,960 There's no way for us to know any of this because there's no way to know any of this.
02:57:24,280 --> 02:57:24,680 Right.
02:57:24,820 --> 02:57:26,040 It's all the secret.
02:57:26,660 --> 02:57:26,860 Yep.
02:57:27,580 --> 02:57:27,900 Okay.
02:57:27,900 --> 02:57:29,880 Our new ending section.
02:57:29,880 --> 02:57:32,220 The splinter in our minds.
02:57:32,380 --> 02:57:32,860 The takeaway.
02:57:33,520 --> 02:57:35,440 The one thing you can't stop thinking about.
02:57:35,660 --> 02:57:42,920 What is the one thing for each of us, personally, from doing this work over the past month on
02:57:42,920 --> 02:57:45,080 Rentech that sticks with us?
02:57:45,780 --> 02:57:50,520 For me, perhaps this is obvious from my little diatribe on the tapestry.
02:57:51,060 --> 02:57:58,220 I just think this is such a powerful example of the power of incentives and getting them
02:57:58,220 --> 02:57:58,660 right.
02:57:58,660 --> 02:57:58,720 Right.
02:57:58,720 --> 02:57:58,960 Right.
02:57:58,960 --> 02:57:58,980 Right.
02:57:58,980 --> 02:57:59,000 Right.
02:57:59,000 --> 02:57:59,020 Right.
02:57:59,020 --> 02:57:59,060 Right.
02:57:59,060 --> 02:57:59,080 Right.
02:57:59,080 --> 02:57:59,100 Right.
02:57:59,100 --> 02:57:59,120 Right.
02:57:59,120 --> 02:57:59,140 Right.
02:57:59,140 --> 02:57:59,180 Right.
02:57:59,180 --> 02:57:59,200 Right.
02:57:59,200 --> 02:57:59,220 Right.
02:57:59,220 --> 02:57:59,240 Right.
02:57:59,240 --> 02:57:59,260 Right.
02:57:59,260 --> 02:57:59,540 Right.
02:57:59,540 --> 02:57:59,560 Right.
02:57:59,560 --> 02:57:59,580 Right.
02:57:59,580 --> 02:57:59,640 Right.
02:57:59,640 --> 02:57:59,740 Right.
02:57:59,880 --> 02:58:00,140 Right.
02:58:00,320 --> 02:58:00,440 Right.
02:58:00,440 --> 02:58:01,120 And culture, too.
02:58:01,480 --> 02:58:02,580 I don't want to shortchange that.
02:58:02,700 --> 02:58:10,400 I think the culture of managing an academic environment in a fashion like a lab, but
02:58:10,400 --> 02:58:16,140 without letting it spin into the frivolity of a lab that Jim Simons set up.
02:58:16,420 --> 02:58:16,440 Right.
02:58:16,580 --> 02:58:17,700 In other words, early Google.
02:58:18,200 --> 02:58:18,360 Yeah.
02:58:18,420 --> 02:58:19,500 This is like early Google.
02:58:19,640 --> 02:58:20,100 Exactly.
02:58:21,000 --> 02:58:26,220 There historically has not, from our research, and as best as we can tell currently, is
02:58:26,220 --> 02:58:29,400 not anything going on at Rentech.
02:58:29,400 --> 02:58:31,960 There is no one thing that is frivolous.
02:58:31,960 --> 02:58:36,880 They are all very focused, which again, to me, then speaks back to the power of incentives.
02:58:36,880 --> 02:58:42,420 When you're there with less than 400 people, and on the research and engineering side,
02:58:42,420 --> 02:58:50,800 less than 200 people, and those colleagues who you work with are the sole purveyors,
02:58:50,800 --> 02:58:55,920 supervisors and beneficiaries of all of this that you're doing, like that is so powerful.
02:58:55,920 --> 02:58:56,920 Yeah.
02:58:56,920 --> 02:58:57,960 I can't think of anywhere else like that in the world.
02:58:57,960 --> 02:58:58,960 I mean, maybe not in the United States, but maybe in the United States.
02:58:58,960 --> 02:58:59,960 I can't think of anywhere else in the world.
02:58:59,960 --> 02:59:04,860 I mean, maybe some venture funds or other investment firms, but not on a day-to-day,
02:59:04,860 --> 02:59:07,940 fully liquid, with returns like this.
02:59:07,940 --> 02:59:08,940 There's nothing like it.
02:59:08,940 --> 02:59:09,940 Nope.
02:59:09,940 --> 02:59:11,940 Pure gasoline right into the veins.
02:59:11,940 --> 02:59:13,120 Yeah.
02:59:13,120 --> 02:59:15,140 Which is not to say I would necessarily want to work there.
02:59:15,140 --> 02:59:16,140 I think I would not.
02:59:16,140 --> 02:59:17,140 Totally.
02:59:17,140 --> 02:59:18,140 Yeah.
02:59:18,140 --> 02:59:19,140 But it is truly unique.
02:59:19,140 --> 02:59:20,140 Yep.
02:59:20,140 --> 02:59:24,480 The one thing I can't stop thinking about is the idea of the complex adaptive system
02:59:24,480 --> 02:59:26,120 that I was talking about earlier.
02:59:26,120 --> 02:59:28,520 I think, from everything we can tell from the outside, run a lab, run a lab, run a lab,
02:59:28,520 --> 02:59:35,400 but renaissance actually has built a large-scale computer system that discovers relationships
02:59:35,400 --> 02:59:42,440 between different entities in the world, stocks, commodities, bond prices.
02:59:42,440 --> 02:59:47,460 And whether it can explain them or not it is correct most of the time.
02:59:47,460 --> 02:59:52,420 And it might be a small most, but all you need is most, and then you can operate a casino
02:59:52,420 --> 02:59:53,520 business.
02:59:53,520 --> 02:59:57,540 That is my takeaway, is that they are the house and they have an edge, and that edge
02:59:57,540 --> 03:00:04,400 is predicated on a graph of all the relationships between these entities that we think are just
03:00:04,400 --> 03:00:09,940 noise and they know the signal. It does make you wonder, to what you were talking about with the
03:00:09,940 --> 03:00:18,620 tech industry catching up, quote unquote, in recent years, how hard is it to build this now,
03:00:18,760 --> 03:00:23,080 given the technology, open source and otherwise, that's available for sale out there?
03:00:23,080 --> 03:00:24,940 That's the bear case. I don't know.
03:00:25,300 --> 03:00:29,740 Yeah. And then what's going to happen? By nature, given that it's a complex adaptive system,
03:00:30,340 --> 03:00:34,300 if you can now buy and build this, well, the returns will get arbitraged down.
03:00:34,860 --> 03:00:37,120 Yep. All right. Should we have some fun? Carve-outs?
03:00:37,600 --> 03:00:38,360 Let's have some fun.
03:00:38,780 --> 03:00:42,860 Sweet. All right, listeners, I have three. People have been expressing that they're
03:00:42,860 --> 03:00:45,880 loving the carve-out section, so I decided to load them up a little bit more.
03:00:46,160 --> 03:00:47,580 That's right. Let's indulge.
03:00:47,800 --> 03:00:49,560 We'll spin off a whole new podcast called Carve-outs.
03:00:49,840 --> 03:00:51,560 44% carry. Let's go.
03:00:51,900 --> 03:00:52,740 So I have one announcement.
03:00:53,080 --> 03:00:58,200 One TV show and one other fun thing for listeners. So first, the announcement,
03:00:58,920 --> 03:01:04,260 David and I are going to be emceeing Modern Treasury Transfer again this year. And so if
03:01:04,260 --> 03:01:09,360 payments are your thing, you should come join us. It was awesome last year. It'll be May 15th,
03:01:09,360 --> 03:01:14,280 2024 in San Francisco. And we'll put a link in the show notes to register. We would love to see
03:01:14,280 --> 03:01:14,820 you there.
03:01:15,300 --> 03:01:15,680 Can't wait.
03:01:16,120 --> 03:01:22,160 My second one is a TV show, and it is actually Acquired related. It is called
03:01:22,160 --> 03:01:22,880 The New...
03:01:23,080 --> 03:01:25,260 Look on Apple TV+.
03:01:25,260 --> 03:01:30,140 Oh, yes. But Christiane, it is such a new look.
03:01:31,620 --> 03:01:35,380 Exactly. So for anyone who listened to the LVMH episode, remember we were talking about
03:01:35,380 --> 03:01:40,500 the groundbreaking thing that Christian Dior did was his collection, The New Look,
03:01:40,640 --> 03:01:44,280 that was a post-World War II explosion onto the scene.
03:01:44,740 --> 03:01:45,720 Celebration of life.
03:01:45,720 --> 03:01:51,840 Yes. Gone are the days of the militaristic, boxy clothing, and now we're in with these
03:01:51,840 --> 03:01:52,920 seductive...
03:01:53,080 --> 03:01:55,140 Conductive and, dare I say...
03:01:55,140 --> 03:01:58,300 Sumptuous materials. War rationing is over.
03:01:58,300 --> 03:02:07,060 Exactly. Yes. Provocative dresses. The Apple TV show is this incredible drama of kind of
03:02:07,060 --> 03:02:14,520 flashbacks to the wartime experiences, harrowing wartime experiences of Christian Dior, of
03:02:14,520 --> 03:02:19,840 Balenciaga, of Coco Chanel, and everything they went through and how all their paths crossed.
03:02:20,340 --> 03:02:21,720 Oh, Coco's in it.
03:02:21,800 --> 03:02:22,160 Yes.
03:02:22,160 --> 03:02:23,040 Oh, wow.
03:02:23,100 --> 03:02:24,000 How do they treat that?
03:02:24,540 --> 03:02:29,740 It will be very interesting if a lot of people watch this show to see if that affects product
03:02:29,740 --> 03:02:34,000 sales of Chanel. I'm also very curious, for people who are watching, feel free to put a
03:02:34,000 --> 03:02:38,400 thing in the Slack in carve-outs. Do you think she's a sympathetic figure? Do you think she's
03:02:38,400 --> 03:02:42,620 a villainous figure? I'm curious how you think of her portrayal versus reality.
03:02:42,920 --> 03:02:48,480 Well, there's the whole crazy thing with Chanel where the company ends up getting bought by
03:02:48,480 --> 03:02:52,460 Chanel the Perfume Division, which is the two Jewish brothers in New York.
03:02:52,460 --> 03:02:53,060 The Worth.
03:02:53,080 --> 03:02:58,220 The Worth, indeed. Oh, God, we got to do a Chanel episode at some point. But the new look on Apple
03:02:58,220 --> 03:03:03,760 TV+, I promise you, whether or not fashion luxury is your thing, it's a beautiful and harrowing story.
03:03:04,260 --> 03:03:08,800 Oh, as you and listeners know, I'm not a TV guy, but this is so up my alley.
03:03:09,020 --> 03:03:11,460 The whole thing, it takes place in wartime Paris.
03:03:11,960 --> 03:03:13,580 Oh, all right. I got to watch it.
03:03:13,620 --> 03:03:18,800 You got to watch it. Okay. My third one is a fun thing for listeners. So after our Nike episode,
03:03:18,800 --> 03:03:22,860 the president of Cole Haan, which if you listen to the episode, you now know,
03:03:22,860 --> 03:03:27,140 is its own company, spun out from Nike years ago. The president of Cole Haan reached out,
03:03:27,260 --> 03:03:31,840 and it turns out he, like all of you, is also an acquired listener. And so we were chatting,
03:03:32,040 --> 03:03:36,380 and he brought up the idea that they'd be happy to create a specific Cole Haan deal
03:03:36,380 --> 03:03:40,800 for acquired listeners. And I told him, frankly, if it's good enough, then I'll share it on air.
03:03:41,180 --> 03:03:45,820 To be clear, this is not a sponsorship. This is just like, he's a fan and reached out to us.
03:03:46,140 --> 03:03:49,000 And I've owned a bunch of Cole Haan products over the years and I've really liked them. So
03:03:49,000 --> 03:03:52,420 for 35% off anything,
03:03:52,860 --> 03:03:59,880 you can go to colehaan.com slash acquired or use the code acquired35 at checkout. And thank you to
03:03:59,880 --> 03:04:04,700 Dave for providing this to us. This is only live, I think for a couple of weeks. So if you're
03:04:04,700 --> 03:04:09,420 listening to this episode soon after it drops, go check it out. I think they intentionally want
03:04:09,420 --> 03:04:13,660 to cut it off at some point so it doesn't get shared around all the coupon sites, but fun thing
03:04:13,660 --> 03:04:17,760 for acquired listeners. Super cool. He likes acquired and wanted to share the love back.
03:04:17,920 --> 03:04:20,680 Love it. I love it. All right, David, your carve outs.
03:04:21,300 --> 03:04:22,300 My carve out is
03:04:22,300 --> 03:04:31,180 related to the new look in a very different way, but both video consumption and fashion and luxury
03:04:31,180 --> 03:04:41,860 and style. It is the class of Palm Beach, Instagram and TikTok account. This is so great.
03:04:42,100 --> 03:04:46,240 David, you and I go to Palm Beach for two days and you get hooked on.
03:04:46,500 --> 03:04:51,080 This is amazing. So Ben and I went to Palm Beach for a couple of days for a speaking event recently,
03:04:51,080 --> 03:04:52,280 which was amazing.
03:04:52,300 --> 03:04:53,800 I'd never been to Palm Beach before.
03:04:54,260 --> 03:04:54,820 Ooh, it is nice.
03:04:55,140 --> 03:05:00,020 So great. We didn't knowingly spot any Rentech people there, but we may have.
03:05:00,280 --> 03:05:02,260 We did knowingly spot some Birkin bags though.
03:05:02,780 --> 03:05:10,860 Yes. The style in Palm Beach. We had just recorded the Hermes episode and oh man,
03:05:11,120 --> 03:05:17,300 I was so pleased to be there. And then I got home and Jenny, my wife was like,
03:05:17,780 --> 03:05:19,920 do you not know the class of Palm Beach TikTok account?
03:05:19,920 --> 03:05:22,280 And David's like, I'm a thousand. I have no idea what you're talking.
03:05:22,300 --> 03:05:25,580 Yeah, right, right, right. I live under a rock. I'm a dad.
03:05:26,780 --> 03:05:31,480 And she showed it to me. This is a woman who lives in Palm Beach and she goes around,
03:05:31,540 --> 03:05:35,540 she posts on Instagram and on TikTok and she just interviews people on the street about
03:05:35,540 --> 03:05:40,880 what they're wearing, what brands they're wearing, their style. It is magnificent.
03:05:41,340 --> 03:05:44,400 My favorite is, we'll see if we can find it and link to it in the show notes.
03:05:44,660 --> 03:05:51,220 There's a video of one woman who's being interviewed who has a mini Kelly inside her Birkin.
03:05:51,220 --> 03:05:52,220 Oh, excellent.
03:05:52,300 --> 03:05:53,860 Excess. Truly excess.
03:05:54,220 --> 03:05:58,160 And that's when I was hooked. I was just like, this is the greatest thing I have ever watched.
03:06:01,000 --> 03:06:01,780 I'm obsessed.
03:06:02,540 --> 03:06:05,280 All right. If I used TikTok, I would subscribe.
03:06:05,960 --> 03:06:07,240 No, you can get it on Instagram too.
03:06:07,320 --> 03:06:08,020 Oh, all right. Good.
03:06:08,180 --> 03:06:12,980 I actually subscribed the acquired account on Instagram to class of Palm Beach. I don't know
03:06:12,980 --> 03:06:16,040 how many people we're following. It's not many, but we are following class of Palm Beach.
03:06:16,220 --> 03:06:19,680 Look at David opening up our Instagram account. You're so youthful.
03:06:20,080 --> 03:06:20,520 Oh, no.
03:06:20,820 --> 03:06:21,880 All right, listeners.
03:06:22,300 --> 03:06:28,420 Well, a huge, huge thank you to JP Morgan Payments, ServiceNow, and Vanta. You can click
03:06:28,420 --> 03:06:34,060 the link in the show notes to learn more. David, I know you've got some thank yous from folks you
03:06:34,060 --> 03:06:35,860 talked with and a few of them we did together.
03:06:36,780 --> 03:06:41,940 Yes. For sources for this episode who were so generous with their time and thoughts. First,
03:06:42,120 --> 03:06:45,540 huge thank you to Greg Zuckerman, author of The Man Who Solved the Market,
03:06:46,100 --> 03:06:50,440 the canonical book out there about Rentech and Jim Simons. Greg was super generous,
03:06:50,640 --> 03:06:52,200 spending time talking to us.
03:06:52,300 --> 03:06:57,920 Emailing with us, making sure we're getting things right. He also, he and the book is
03:06:57,920 --> 03:07:05,820 the canonical source of Medallion's investment returns. And I know he worked so hard to get
03:07:05,820 --> 03:07:09,260 that table together that is now all over the internet as it should be.
03:07:09,780 --> 03:07:14,980 It is crazy. Everywhere you hear that 66% number quoted, and that is from Greg's analysis.
03:07:15,600 --> 03:07:20,660 Yes. Truly a service to us and to corporate historians and financial historians everywhere
03:07:20,660 --> 03:07:22,140 that he did that research and got those returns.
03:07:22,140 --> 03:07:22,220 Thank you.
03:07:22,220 --> 03:07:22,260 Thank you.
03:07:22,260 --> 03:07:22,280 Thank you.
03:07:22,300 --> 03:07:27,520 And there's a few other primary sources. There's really not much. So we can actually list
03:07:27,520 --> 03:07:32,920 all of them here. There's a congressional testimony of Peter Brown about the basket
03:07:32,920 --> 03:07:37,360 options thing. There's Peter Brown doing an interview at GS Exchanges, which again,
03:07:37,480 --> 03:07:41,260 many of the questions were straight out of Greg's book and the stories told.
03:07:42,080 --> 03:07:45,640 Yeah. It's a funny moment where Peter's like, where are you getting these questions? How do
03:07:45,640 --> 03:07:48,140 you know all this stuff? And I'm like, come on. They read the book.
03:07:48,140 --> 03:07:52,260 Clearly. There's a great book called The Quants, which is,
03:07:52,260 --> 03:07:57,480 it's a little bit earlier. I think it's 2011. So it's not as updated as The Man Who Solved the
03:07:57,480 --> 03:08:01,600 Market. And there's only sort of a couple chapters about Rentech, but some good stuff in there. And
03:08:01,600 --> 03:08:06,680 then there's a good Bloomberg piece from 2016 that we'll link to that. I think between that
03:08:06,680 --> 03:08:10,840 and The Quants, it was sort of the first time there was really anything at all that was published
03:08:10,840 --> 03:08:15,420 about Rentech. So all those will be in the show notes. Other people to thank, David.
03:08:16,060 --> 03:08:18,720 Other people to thank, Howard Morgan, who we spoke to, which was
03:08:18,720 --> 03:08:22,100 so fun to get a bunch of the first round history from him. And then of course,
03:08:22,100 --> 03:08:27,260 the founding of Rentech and partnering with Jim and investing in each other's funds and all that.
03:08:27,520 --> 03:08:34,200 So fun. Brett Harrison, who you mentioned, Ben. Brett is now building Architect, which I love
03:08:34,200 --> 03:08:38,720 this. This is so needed in the world. It's the interactive brokers for the 21st century.
03:08:39,840 --> 03:08:45,540 Well, anybody who uses interactive brokers knows exactly the opportunity there. So thank you,
03:08:45,540 --> 03:08:51,820 Brett. And then Matthew Grenade, who I spoke with. Matt is the co-founder of Domino
03:08:51,820 --> 03:08:56,920 Data Lab, which is a great enterprise AI ops platform backed by Sequoia and many others.
03:08:57,100 --> 03:09:02,240 It allows model-driven businesses and products to accelerate research, increase collaboration,
03:09:02,880 --> 03:09:07,400 rapidly deliver new machine learning models, all of the sorts of things that we were talking about
03:09:07,400 --> 03:09:15,120 here with Rentech. Matt, before starting Domino Data, came out of the quant world. He was at 0.72
03:09:15,120 --> 03:09:20,040 and Bridgewater, which isn't really quant, sort of its own thing, but he was a longtime senior
03:09:20,040 --> 03:09:21,800 employee at both of those firms. And Matt, who is the co-founder of Domino Data Lab,
03:09:21,820 --> 03:09:27,960 he gave us great, great perspective on the landscape of everybody out there and where
03:09:27,960 --> 03:09:32,640 Rentech fits in. Awesome. Well, if you liked this episode, you should check out our Berkshire
03:09:32,640 --> 03:09:38,680 Hathaway episodes from a few years ago for a very different style to investing. You can sign up for
03:09:38,680 --> 03:09:44,080 new episode emails at acquired.fm slash email. We'll be including little tidbits that we learn
03:09:44,080 --> 03:09:50,120 after releasing each episode, including listener corrections. You can listen to ACQ2, search and
03:09:50,120 --> 03:09:51,800 subscribe in any podcast player. And if you're interested in learning more about Rentech, you can
03:09:51,800 --> 03:09:58,000 and listen for our most recent episode with the, well, really creator or a person who led the team
03:09:58,000 --> 03:10:02,260 that created Lyra Glutide, which went on to become Semiclutide, which of course is Ozempic, Wegovi,
03:10:02,620 --> 03:10:10,100 et cetera. Yep. All modern GLP1s. Lata Bier-Nudson from Novo Nordisk was awesome to have her on the
03:10:10,100 --> 03:10:14,080 show. And after you finish this episode, come talk about it with other smart members of the
03:10:14,080 --> 03:10:20,480 Acquired community at acquired.fm slash slack. If you want some merch, we've got some. Acquired.fm
03:10:20,480 --> 03:10:21,780 slash store. And we'll see you next time. Bye.
03:10:21,800 --> 03:10:25,060 Fat listeners. We'll see you next time. We'll see you next time.