The Exchanges

Q Cliff, before I let you go, I want to ask you a couple fun closing questions. What's your favorite hobby or activity outside of work and family?

A You know, this is embarrassing because it's so on the nose, but I collect comic books and comic book art. It's on the nose for a quant geek to be a radical comic book fan. Possibly outside of my children, the greatest development in my life was when movies, special effects finally caught up. I was watching the Lou Ferrigno was a tough guy, but he didn't look like the Hulk. I still collect those. Thankfully, I have a money manager's income with a child's taste. I'm not competing with Stephen A. Cohn for impressionist paintings. I'm trying to buy the cover of Avengers number 147 in pencil, but I have a lot of fun with that, and I love that stuff.

AI assessment note: “I collect comic books and comic book art.”

Q And your thoughts on the cognitive dissonance of performance chasing?

A I think if you asked most allocators, investment committees, individuals, is it a good idea just to buy what's up in the last three to five years and sell what's down? They say probably not. No, that's too superficial. There could be good reasons. Maybe it's cheap. And then as an industry, we all collectively do it anyway. That one doesn't have a great subtlety. I think it's just pure cognitive dissonance. We say one thing. And we do another. Again, the eternal paradox is there wouldn't be a whole lot of stuff for active managers to do if people didn't make some crazy errors like this. Be careful what you wish for. Not that I'm going to single-handedly do this, but if myself and a thousand other ever convince the world to be perfectly rational, we're also out of a job. The perfect world is I get to complain about this and whine about this and convince some subset of people I'll probably write and not convince most of the world, which I think is the path I'm on.

AI assessment note: “I think it's just pure cognitive dissonance. We say one thing. And we do another.”

Q What was the first spark for you going from studying under Gene Fama to practicing that what you were learning at Chicago Efficient Markets wasn't quite right?

A I think you're being a little unfair to Gene Fama when you say that. There's a moment after getting the preliminary ideas out there, the first two or three weeks, Gene looks at the class and said, markets are almost assuredly not perfectly efficient. And you get a gasp. And only at the University of Chicago and only in Gene Fama's class do you probably get a gasp at that statement. It was really simple. To be fair, he probably thinks they're more efficient than I do these days. Perfect efficiency is an extreme hypothesis. It's perfection. Grossman and Stiglitz wrote a paper introducing a paradox about a perfectly efficient market a long time ago that people need to spend a lot of money making something perfectly efficient on research, on time, and why are they going to do that if it's perfectly efficient? They're not going to make it perfectly efficient. They're going to spend an optimal amount of money to get it close. Fahman French later wrote a paper about taste and agreement, saying if any part of the market has an irrational love for a set of stocks, Unless somebody else has a perfectly balanced, irrational hate for those set of stocks, those stocks are going to be overpriced, because the perfect arbitrage is not possible. If you, Ted, think a bunch of stocks, you're crazy, you're a bad investor, I'm choosing you as an example, and you think a bunch of stocks are way overv…

AI assessment note: “Gene looks at the class and said, markets are almost assuredly not perfectly efficient.”

Q You gotta bring me back. What were you like as a kid?

A Often people don't believe me, but I was certainly an underachiever. Sounds obnoxious, but I was considered pretty smart on standardized tests and whatnot, but I got B's and even some C's in high school. Nowadays, I don't think kids could get away with this as much. I went to a fairly decent But public school on Long Island. And the teacher would come around occasionally and go, you didn't turn in the homework again. And then I'd get either a zero or the famous check minus. I played ball. I was never very good, but I had fun in high school. My wife gets mad when I still say senior year of high school is probably the most fun year because I didn't know her back then. But I was already into college because I got an early decision. I had a girlfriend. I had a car. I had a job. My parents always make me work. But I was the last year I was What you might call type B happy-go-lucky. I had told everyone, my parents, teachers, whenever they yelled at me that I didn't try and had potential, I'd be like, I'll try when I get to college, it'll count. I will say, when I went to college, I probably had a very tense freshman year where it was, all right, I've been telling people I'll pick it up and I'll be great once I get here. And no matter who you are, you're not that confident that you're right, that you'll be able to do that. And I was a wise-ass. Which has never changed, but I didn't ev…

AI assessment note: “Often people don't believe me, but I was certainly an underachiever.”

Q In that very beginning. What happens in that early part of the tenure of a business where performance is just moving against you?

A When you go through a bad period, my partners and I, the ones who were there, then only back in the early days, would discuss what was more painful, a really horrible period from inception, the first year and a half, or really horrible two and a half year periods when you're running hundreds of billions of dollars and have many more clients. And in one sense, it's clearly the first one because you face existential risk. At no point in the depths of COVID when value was particularly Doing its crescendo of losses, which turned out to be totally irrational. People thought all you needed was Peloton and Tesla forever, but at no point did we face going out of business risk. It was just excruciating. We were losing money for clients. We hate, quite obviously, I hope any money manager hates that. Did a lot of business pressure, a lot of have you guys lost it kind of questions, and you got to defend yourself, but not an existential risk. The only existential risk would be if we gave up, and that was not going to happen. But as you survive in this industry, you will find Duration of pain is as important as intensity of pain to how damaging it is to both your future life expectancy and to your business. A crash that quickly reverses itself. Almost by definition, people are tense. Why'd that crash? You have to talk about it. You have to figure out what happened. I would argue a three-year…

AI assessment note: “it's clearly the first one because you face existential risk.”

Q In the early years, when you're writing your thesis, you're then going to practice this. The world doesn't know about factors. You could drive a truck through the differences. So how did you, over time, calibrate regimes changing or the markets catching up with what historically looked very good?

A Most things get somewhat attenuated over time. They get somewhat arbitraged away. They don't have to go all the way away. This is a key. It's the same concept I was talking about before. If there's an error people make on average, and people catch on to it, maybe they invest enough dollars to arbitrage half that error away. But then it gets to be a pretty low-risk-adjusted return to do the last half. Literally, since our Goldman Sachs days, this is more than 25 years ago, we've generally used half a backtest out of sample as a bogey. And life has worked out fairly close to that. Now, we've built more factors, improved our backtest, so it's not necessarily half of what it used to be, but that's not been far off, and we would consider that a victory. If you have a decent backtest that you don't think is over data mine, that you think is reasonable, that you understand the economic spirit, that you've tested robustly in a whole bunch of places, all of that, still assume half going forward. And half is obviously a rule of thumb. You may sometimes think two-thirds, one-third makes more sense, but It's good to have a simple thing you don't deviate from. Some things, I think, are much more susceptible to being arbitraged away. Strategies that are about acting quickly, maybe they don't go to zero, but as you can imagine, someone's always quicker, and maybe someone can still make money …

AI assessment note: “we've generally used half a backtest out of sample as a bogey”

Q So that freshman year stress period, as you mentioned, could be a little bit of foreshadowing of sticking to some conviction. What was that first experience like of trying and not being sure you're going to get there?

A Everyone feels anxiety at some point. It's not like I went through high school even as type B is with no anxiety, but it was by far the most I've ever experienced. I'm immediately revealing more to you than I think I've ever revealed publicly, but the only time in my life up to that point I ever talked to a counselor or something. This was shocking for me because I never tried that hard or cared that much, but at some point mid-freshman semester, first semester, I went to the university's counselor and just said, I'm really tense. I think I ended up talking about my dad the whole time, which was odd. It was just cathartic. It was not like I was in deep trouble or needed great lessons, but it's a statement on how serious it was for me. And I just fought through it. It was a good ultimate experience because I realized stress sinks at the time, but it has a purpose. I did my freshman year, which quelled many of the fears. Basically, the fear was I wouldn't do well. The fear was that I told everyone, I'm really smart. I just haven't tried yet. And then I'd get there and discover that actually even trying, I wasn't that good. But I overcame that fear freshman year, and I had a good time freshman year, but it was definitely my first experience, not my last, as you alluded to, but my first experience with type A, anxiety, is this going to work out, permeating your thoughts at times, t…

AI assessment note: “Basically, the fear was I wouldn't do well... I just haven't tried yet.”

Q I want to play around a little bit with your thoughts on private equity and volatility laundering. How do you respond to the thought that maybe the private markets are right? If the public markets are some combination of behavioral implication on returns and fundamentals, and the private markets are a little more tied to just fundamentals, that you would expect there not to be those movements?

A If you say it this way, let me say it back to you, that markets are not perfectly efficient, and there are occasional bubbles. Therefore, the private equity prices are closer to true. That doesn't necessarily have to be the case. Maybe halfway is true, and they don't move them, but it's pretty plausible that they could be closer to accurate and metaphysical, what is this really worth, sense. My question then does become one of competitive jealousy. For both private equity and what we do, you can always value things at least one of two ways. You can say, what would the market bear if I sold this tomorrow? Or you can go, what do I actually think it's worth? Private equity managers are better at valuing a company than I can ever dream of. They don't have to tell you, so they don't. But if the market crashed 30% tomorrow, and they had to come up with, what do we think we could sell our portfolio for today? They could absolutely tell you. No one would be happy with it, but they could absolutely tell you. When we are suffering in the public markets, in March of 2000, when NASDAQ was at 5000, and we were down 30 some odd percent Since inception, I could tell you what I think if prices were fair, what the portfolio is actually worth. And in fact, I did those exercises. I did it more in the sense of when it comes back, what I think we'll make, but it's the same calculation. What you'll …

AI assessment note: “it's pretty plausible that they could be closer to accurate and metaphysical”

Q Which two people have had the biggest impact on your professional life?

A I got to go with three, but one of them comes in a pair. My dad, my dad gave me that advice, what program to go to. Something else I didn't tell you. When I was an undergrad, I always assumed I'd go to law school, even though I was studying computer science and business. My whole family were lawyers. Tiny exaggeration, but one of my cousins was actually John Gotti's lawyer in the eighties. That's a great story that we can do at another time. I promise you, I've never had any interaction with the mafia. I don't want to get tarred with that brush. But up through junior year, I was signed up to take the LSATs, and my dad, again, and my dad was not, again, a quant, not a math guy, but he just said, you're really good at math. Shouldn't you do something that actually uses that? And at the same time, I was doing some research for some professors, which I mentioned earlier, that I really liked. I was an impressionable kid. It took me about eight minutes to go, yeah, I'll switch to the GMATs, and then I decided to get a PhD. So my dad, he had some very specific Influence on me. And then I got to go with my two dissertation chairs, Fama and French. We don't agree on everything these days. Again, I probably think markets are more behavioral than Gene does. Ken might be in between Gene and I. When I told Gene I want to write a dissertation on price momentum, he said if it's in the data, w…

AI assessment note: “I got to go with three... My dad... my two dissertation chairs, Fama and French.”

Q So that messaging based on valuation that you tried, And some people stuck with you through the pain and others didn't. At the same time, for those that are coming out, you have your organization that's contracting. How did you go about either keeping morale or keeping people motivated to the mission when some people who presumably believed in it couldn't stay?

A That's the human cost when you have to shrink, and the financial industry is too cyclical. I preach against it, but financial firms shrink in assets and people when they go through really tough times. The math doesn't work. Some of that, this will sound really cold, but we had a pretty good decade prior to that, and we got a little fat and happy. So the initial shrinking's not that hard. It can be hard on a human level. I dread giving these messages. If I have to let someone go who's a good person who's done good for us, but we're just in a bad Situation. I'll spend three days very tense about it. I'll never say it's, this is harder on me than you. That's a lie. But it is hard. It's not a fun thing. But the early ones are hard to do, but are easy decisions. You've grown too much. As a drawdown goes on, the decisions are more painful. But there is a pretty big self-selection bias that works for you. People who are not true believers. Here's another irrationality. Someone who's been with you through another drawdown, seen you come back. It's far more likely the second time to say, I've seen this movie before, and we're going to prosper. Someone who joined an hour and a half before the current drawdown began, and has only seen you lose money for two and a half years, even if they are just as smart and intellectually get it as the other person, doesn't feel it the same way. So even…

AI assessment note: “you have a selection bias where a fair amount of people who don't believe leave”

Q When you move from equities, which is a lot of what we're talking about, to other asset classes, you could make the case that the behavioral aspects of investing are the same. You don't hear as much about this type of activity in the kind of scale in credit, in rates, and some of the other assets. What have you found as you've explored those different areas?

A I'm going to go back to 1995. So this is a twenty-nine-year-old deep cut when we did our first version of a long, short, aggressive fund at Goldman. I think there are three of us left from that experience. Three of seven strategies were macro strategies. Early on, it was, all right, if value and momentum work for individual stocks, if we added up the valuation of the DAX and the valuation of the FTSE, And said the DAX was selling for price to sales double the FTSE and had the similar momentum. If you're a behavioral finance person, people are over optimistic about Germany and under optimistic about the UK and not super high sharp, but clearly positive. It held up over time. So forever, we've done both. And mainly for the old school factors, we think of this as a beautiful out of sample test. If we went to equity markets, we wrote our first paper on this in On showing that then it was only value and momentum. The world has expanded tremendously past that. And size, though we never traded size. Showing that the same things hold up for country selection as they do for individual stocks. If we had tested that and it had failed, maybe we wouldn't throw out the individual stock results, but if you're honest, it would give you a little pause. Because you'd go, why did it fail? Why would the same story, if my story's real? And you don't want to be so flexible that you're always coming …

AI assessment note: “Showing that the same things hold up for country selection as they do for individual stocks.”

Q today, and you started to come back, it's easier to say, we believe this all along, it was clear in the numbers. When you're going through it, as you mentioned, there's always some doubt. How do you go about thinking about sticking to the models that have worked alongside of trying to figure out whether it's through innovation or new research that something isn't working and may not come back?

A I see these as two very parallel things that That interact, maybe a little less than you might think. We've done very little research. It's a really tough period. We just need better stuff. We got to go find better stuff. We do a lot of research at that time, as I said, to say, is our stuff broken? Is there a reason why this is going on that's not going to reverse, or even go back to working normally? It doesn't always have to reverse. So I'm not saying we don't apply that, but that's only because it's a fairly empty statement, because We always have a big research effort going on. Some of my favorite things we've done in 25 years, we did in the last five to 10 years, and they were not in response to the drawdown, but I think consistent, things are often oversold. The internet in 1999, 2000 was oversold, but it does appear to be lasting. Maybe the biggest improvement I've seen at AQR, and I'm going to get in trouble because I love all my AQR children, you're not allowed to have favorite children. I have four kids, and my wife would get very mad if I had a favorite. Katie, by the way. I'm joking, Grace, Charlie, and Leo. That is just a joke. What we've done in trend following, which 10 years ago was the simplest of all strategies, it was one to 12 month price trend on the major markets. Of course, not everyone did the same one. A different trend follower will use a moving averag…

AI assessment note: “I see these as two very parallel things that That interact, maybe a little less”

Q I want to turn to asking you about being on the other side of the table on investment committees. You mentioned earlier you don't know how Podchops, for example, might select managers. How have you thought about the manager evaluation process?

A I have no magic bullets, but I've been on plenty of investment committees. Whenever I get involved with a new organization, a charity, a school, they almost always stick me on the investment committee. They often regret it for various reasons, but largely because if they have a philosophy that's worked for them, I try not to be too much of a jerk. I don't come in and go, I make this all quant. No, I'm not. I'll make my observations at the edges, but having participated in many, and I'll brag and say some of these have been absolutely stellar committees with luminaries. Of investing, and I'm not being sarcastic, people I really respect, including myself, we can collectively be an idiot, because there's a dynamic to a committee. Individual members of the committee get far more punished for disasters than they get for good portfolio returns. That's true in a lot of walks of investing. I think it's particularly true of committees where they don't own anything, they don't get the upside, they just get the opprobrium. There are other weird things. My partner, John Liu, who ran a major investment committee for a long time for his school, he pointed out to me once an interesting effect. Let me be crass here. Those on the committee are often some of the biggest donors over the last 10 years. Let's just be honest. The biggest donors are the most successful parts of the financial world ov…

AI assessment note: “there's a dynamic to a committee... dominated by what has worked the best”

Q How do you think about the risk of that game of Go when the machine does pick out the move that you just were never going to see?

A I don't think we are even trying to use it at that kind of level. Our techniques train on Go from scratch. They literally had it learn the game on its own, and as many permutations there are of Go, it's still, I think, probably a more tractable problem Than every stock and other tradable asset over a hundred years. This is a matter of computing power, and maybe there's a way to constrain the problem where there will be some insight. Those insights would probably take the form of weird non-linearities. If, yeah, value's good, momentum's good, but not if you have these other three good things at the same time. We can use our techniques to look for those. You just got to put a big penalty function on so you don't overfit. It's an interesting question. I haven't gotten it before. I will check with the experts afterwards.

AI assessment note: “I don't think we are even trying to use it at that kind of level.”

Q decided advantage. That's why customers call it miraculous, game-changing, and an awakening. If that's not how you would describe your investment management tech, request a demo at ridgeline.ai. And now, back to the show. On the other side, we think a lot of the pod shops as the sharks of the sharks. How has what you've seen of the activity and the growth of the pods impacted any of this?

A First, let me start out with an important caveat. I don't really know. I keep asking Izzy and Ken to send me exactly what their trades are and what all their strategies are, and they keep politely refusing. Actually, they don't even answer the emails. Everyone knows that's a joke, right? I don't actually ask them that, though they would not tell me. I will tell you this, something I would have gotten dramatically wrong. If you had told me 20 years ago that here's my strategy, I'm gonna go find people I think are really good at this, Pay them a lot. Charge that to the underlying investor. So the fees are, when you add it all up, are gigantic. And then I'm going to fire them rather quickly if they have losses. I would have said, and obviously I would have been wrong, because the results speak for themselves. That's a terrible strategy. There's not that much alpha out there, and the end investor is paying a ton. And even good strategies, this theme we were talking about before, go through bad periods, and I'll be firing a lot of good managers. That leads me to believe that a lot of it is how good some of these shops are at identifying that skill ex-ante. Something I've never tried to do, and sadly we'll be sharing no special insight on, but I'm narrowing it down to what it has to be. If you just told me that's our process, and a lot of people have tried to set up these shops, and …

AI assessment note: “First, let me start out with an important caveat. I don't really know.”