• Home
  • /
  • Blog
  • /
  • Bogleheads on Investing with Dr. Cliff Asness – Episode 27

Bogleheads on Investing with Dr. Cliff Asness – Episode 27

Post on: November 5, 2020 by Rick Ferri

Dr. Cliff Asness is a Founder, Managing Principal, and Chief Investment Officer at AQR Capital Management, a quantitative money manager overseeing $186 billion in assets as of December 2019. Prior to co-founding AQR Capital Management, he was a Managing Director and Director of Quantitative Research for the Asset Management Division of Goldman, Sachs & Co.  He is an award-winning researcher on quantitative investment strategies and has authored articles for many publications, including The Journal of Portfolio Management, Financial Analysts Journal, The Journal of Finance, and The Journal of Financial Economics

Cliff received a B.S. in economics from the Wharton School and a B.S. in engineering from the Moore School of Electrical Engineering at the University of Pennsylvania, graduating summa cum laude in both. He received an M.B.A. with high honors and a Ph.D. in finance from the University of Chicago, where he was Eugene Fama’s student and teaching assistant for two years.

You can discuss this podcast in the Bogleheads forum here.

Listen On


Rick Ferri: Welcome everyone to Bogleheads on Investing, podcast number 27. Today our special guest is Cliff Asness, founder, managing principal, and chief investment officer of AQR Capital Management. Cliff has a PhD from the University of Chicago. Jack Bogle called him a brilliant academic.


My name is Rick Ferri, and this is Bogleheads on Investing, podcast number 27. This podcast, as with all podcasts, are brought to you by the John C Bogle Center for Financial Literacy, a 501(c)3 nonprofit organization at Bogelcenter.net.Today we have a special guest, Cliff Asness. Cliff is the founder, managing principal, and chief investment officer of AQR Capital Management. At the end of 2019 the firm was managing over 186 billion dollars of assets, in part on strategies that Cliff developed at the University of Chicago as a PhD student under Nobel Laureate Gene Fama. Cliff has received many awards, including the Bernstein-Fabozzi-Jacobs-Levy award from the Journal of Portfolio Management, the Graham and Dodd award from the Financial Analyst Journal, the James Verton award from the CFA Institute, and probably his biggest award, Jack Bogle referred to Cliff Asness as a brilliant academic.  I will warn you that some of this podcast is going to be a little geeky, so put on your math hat and here we go.  With no further ado let me introduce Cliff Asness. Welcome Dr Asness.

Cliff Asness: Thank you Rick, but please call me Cliff. Doctor makes me think you want me to help with an appendectomy.

Rick Ferri: Well thank you Cliff. It’s a real pleasure to have you on Bogleheads on Investing. You have a lot of followers on the Bogleheads. All of your research and your background is really quite fascinating, and that’s what I want to start out with. If we could start as far back as you are comfortable going, How did you get to be where you are today?

Cliff Asness: Oh, I was, I was a very unmotivated high school student. My parents thought I was an underachiever. I did very well on standardized tests and very poorly relative to those tests in classes. Partially back then you can get away with actually not doing your homework and I took advantage of that. When applying to college my dad found this program at the University of Pennsylvania that was a dual degree program between the Wharton School and the engineering school. I had not really given any of this much thought at all and it was my father who suggested, “Hey why don’t you study two things that both could help with a career because you don’t know what the heck you want to do and you have no direction.” So I really didn’t start out to go into school and doing that because I had a great love of finance or really wanted to, you know, be a computer scientist. It was more of a default. I ended up loving both those things. A little luck is always a good thing. You know you’ve got to take advantage of good luck when you get it. You have to minimize bad luck when you get it, but if things work out well and you don’t acknowledge the role of luck you’re just not telling the truth.

So I went to this program. At one point I started doing research assistant work for a few different professors, one of which is quite well known now, Andrew Lo of MIT was at Wharton at the time. He was a very young man. I ended up enjoying even the research assistant work, and I think I said something to him, kind of like I might want to do what you do, and he was like, that’s cool, why don’t you think about where you want to go see where you get in. I applied to a bunch of PhD programs, got in to Chicago, Stanford and Wharton. Did not get into MIT, for which I  am still bitter about. But you know, eventually I’m on year 32 of being bitter about that. I think at some point I gotta let it go, but I’m not not quite there yet. They said something along the lines of, you know, we don’t take people directly from undergrad but I think that was kind of like when someone breaks up with you and says it’s not you it’s me.

Rick Ferri: It probably wasn’t because of your grades because according to the bio that I read you were summa cum laude in both engineering and economics.

Cliff Asness: Yeah there was a day where I was pretty smart. Hey you know that day has faded and I couldn’t do any of those things again, but sometime in the distant past the brain used to work. Let’s not dwell on the one school I did not get in, though I will one day extract my revenge upon the entire institution. I ended up, you know Wharton’s a great place but it didn’t stack up against Chicago or Stanford when it came to PhD programs. I’m not talking of course about MBA programs or whatnot, but PhD programs always have their own kind of unique rankings and I basically went around to 10 professors at Wharton, told them that I’m looking to get a PhD  in this. I got in these three schools and almost to a man they said, shut the door, largely because they were going to tell me not to go to Wharton and again no knock on Wharton, I love my alma mater, but that was how PhD programs ranked those days, and nine out of ten told me to go to Chicago and many of them had gone to Chicago themselves.

One, a famous professor named Robert Litzenberger, who was the next Stanford professor, he suggested Stanford, but at that point in history Chicago is where you want to be. As an aside Chicago offered to fly me out and Stanford didn’t. And I asked and they said, “no it’s just not in the budget” and I had no money at the time, so I took Chicago up on their offer, visited on one of those gorgeous spring days Chicago has ever produced. So for years I am fond of saying that I am the world’s only person to choose Chicago over Stanford on the weather, which turned out to be a bit of a bait and switch. But the school worked out well.

Rick Ferri: And not only did it work out well but you got to study under some future Nobel laureates and one of them was Gene Fama.

Cliff Asness: Yes. Yeah and you know what’s amazing is back then we pretty much knew he was a future Nobel laureate. Frankly I thought they gave him the prize, you know, 15 years after they should have. But you knew it at the time, he was an amazing man. You know, again it was a lucky break for me to be at Chicago in the late 1980s when Fama and French were doing a lot of the early work on what we now call systematic value investing or factor investing. We didn’t even use those terms back then. We were just a bunch of geeks running regressions and sorting stocks on different variables looking at returns. The terminology came later but I was lucky enough for Gene to ask me to be his teaching assistant at the end of the year and I did that for two years, so it basically meant I kind of took his class three times. So I sat in on it pretty much every day for three years. I got to teach review sessions and got to know Gene. In PhD  programs, getting good advisors, it’s just a big part of it– getting advisers who are interested in you, who think well of you. I think that work helped me get him and Ken French to co-chair my dissertation committee, which was quite wonderful, more luck.

Rick Ferri: Professor Fama was a big efficient market guy but your thesis really had nothing to do with efficient market or maybe almost to disprove efficient markets in some way.

Cliff Asness: Oh, first of all we never prove or disprove things. We’re statisticians so we’re cowards. I joke about the cowardly statisticians that we are, we don’t disprove we just say there’s only a one out of 100 chance we’re wrong.

Rick Ferri: Yeah that sounds like a Wall Street Analyst.

Cliff Asness: But it is– you know it’s funny, it sounds kind of wimpy and cowardly but it’s also just intellectually true. Data can never prove something, it can just increase the probability of your hypothesis or against your hypothesis. If you see a 99 out of 100 event it may be in fact that it’s true or maybe you got a 99 out of 100 random event. But yeah, a big part of my dissertation, not all of it, but a big part was one of the very early pieces of work on price momentum strategy, you know pride of place is first, has to go to professors Jegadeesh and Titman who wrote the first big published paper on the price momentum strategy. I was doing fairly concurrent with them, and I am proud that my definition — partly because Fama and French adopted it for their, for Ken French’s kind of data website, and used in their papers — it’s kind of the most common one people use today, UMD, “up minus down” for basically one year price momentum. 

I do remember and very, very distinctly being nervous about precisely what you just said, that here I am, I’m lucky to have this relationship with the premier advocate for the efficient market hypothesis in the world and I have all these preliminary results saying that a fairly straightforward momentum strategy was effective. And I distinctly remember telling him “Gene” — well I say Gene now, Professor Fama, certainly it took me 15 years after the program until I would call him Gene even though he kept telling me to — “Professor Fama, I want to write a dissertation on price momentum,” and then I mumbled the second part out of fear: “and it works really well.” And he’s like what? I’m like, “and it works really well.” And he thought for a bit and he said something that has always stayed with me, it was actually in a weird geeky way almost kind of moving. He said, “If it’s in the data write the paper.”  Which to me said, “look this might not be something I love. I might not love this result. I might consider it the exception that proves the rule” but data is kind of holy to him and that kind of stuck with me and it also allowed me to get on the path to graduation.

Which was nice though even then I was not crazy enough to say, “Gee I found something better than your value factor.”  First of all, that would have been bad economics, they work better — if you believe both of them — they work better together, systematic value can make money over the long term, systematic momentum can, and they’re fairly negatively correlated which is a wonderful thing to have if both strategies make money. Would have been bad graduation strategy to tell Fama “we’re going to do this instead of value” saying “it’s a compliment to your one wonderful value strategy” was some sucking up I was absolutely willing to do.

Rick Ferri: Well let me ask just a couple of questions on what you just said. Did Fama ever call what he did value investing?

Cliff Asness: You know, it’s funny you ask that because I keep meaning to go back and check the original papers. I don’t believe so. It’s certainly possible when the world started calling it that that they eventually referred to it that way. I don’t even know that, but I don’t remember them calling it a value factor or whatnot. I think that it was index providers and maybe other papers that did it. I think they were just looking at multiples and whatnot to sort stocks. I could be dead wrong about this but from memory I don’t remember being in Chicago and talking about value investing.

But certainly that label caught on and I actually don’t even think it’s the best label. It creates tremendous confusion because your kind of Graham-and-Dodd old-fashioned concentrated portfolio active management — value managers, people call themselves value managers get mad because they look at the quantitative factor and they say that’s not value. Value is the price against fundamentals in consideration of things like which are better companies, safer companies, faster growing companies, more profitable companies. The answer is quants don’t think price divided by anything — book, sales, cash flow, earnings — is the only thing you should judge on. We agree with those managers and we just call them different factors, things like the profitability factor, earnings momentum, low risk investing. Those are all in the direction of the things the Graham and Dodd people want us to consider. But they just put the whole thing together and call it  a cheap company, one that’s cheap considering these other factors. We break it up and call them other factors, and you know a thousand fights have been called caused by that when they’re really through very different methods, concentration versus diversification, judgment versus systematic, but through very different methods, I think they’re often looking for the same kind of things in the stocks.

Rick Ferri: I always thought that beta can be agreed upon if you take all the stocks on the market, weight them by either the shares outstanding in the public float or total shares outstanding, something, but we all can come to an agreement on what beta, the beta of the market should be, and that size, meaning large cap stocks, mid cap stocks, small cap stocks, even micro cap stocks, in general all of the index providers can sort of agree what that is. I mean they are all pretty close. But when it comes to value I always say value is in the eyes of the beholder. Everybody has a different equation.

Cliff Asness: Sure, first I think that’s fair, but let’s talk about the value factor. What I would prefer to call the price factor, not the holistic considering other aspects, just price divided by some fundamental. You could choose price to cash flow; I could choose price to book; someone else could choose a weighted average of how it looks on many different reasonable evaluation factors. You could do what we do. We wrote a paper on this in 1994 and try to use this value factor to bet within industries, meaning we don’t think it’s very effective across industries. It’s pretty hard to compare valuations. The standard work in almost all of academia and a fair amount of the indices allows very large industry bets. In fact it always strikes me that the way we invest in value, we try very hard not to take an industry bet, and the way many speak about value is only in terms of industry bets, you know tech versus textiles. All of these things make your value possibly different than my value, but if we both use a handful of reasonable measures and have a portfolio construction technique — I also didn’t mention how you weight them can be different — Fama and French is a cap weighted. You could do a signal weighted, where the degree of the valuation signal tells you your position. You could do equal weight. Then of course those will have somewhat different results and have somewhat different degrees of implementability–is that a word, I think it’s a word–implementability.

Rick Ferri: Yeah we’ll call it a word.

Cliff Asness: We’ll call it, yeah it’s our podcast.

Rick Ferri: Write a paper on that when you get off this podcast.Rick Ferri: Write a paper on that when you get off this podcast.

Cliff Asness: With that said, I do think if you create your favorite value strategy, and it’s reasonable, deflating price by four or five of the major things out there. I might do it within industries, you might allow an industry bet. We’re going to be, I’m making this up, don’t hold me to it, we’re going to be 0.8 correlated. What’s amazing is, over a couple years, how different sometimes “point A” correlated things can be.  That often surprises people. So these choices they may not be provable, which one’s better, but over the the short even the medium term can matter but “point A” correlated says to me, it’s you are right, you know there is no God-given way to measure value, but I think it matters a little less than people worry about because price is the main thing driving the ship.

Price compared to anything reasonable leads to at least fairly decently correlated strategies. I’d also say a cap-weighted market, even that, has a little degree of ambiguity. How deep do you go in the capital structure, you know, Wilshire 5000 versus S&P 500. Size is actually even a little more complicated. I would agree still with your statement that size in the market is more common. If you and I both come up with a definition, we’re probably going to be more correlated across those. But I still think  value, there’s a concept that any decent value factor follows:  deflating price by reasonable things, and going long the cheap and short the expensive on that measure, and again yours or mine, we might fight to the death over whose is better. We’re kind of co-religionists who differ slightly in our dogma at that point — and as we’ve seen in history those are some of the most vicious fights — but I don’t think it’s really and I would include myself I think, of course after years of doing this, but our way’s the best humanly possible. And I don’t mean that arrogantly, I mean that in the sense of we would do it differently if we didn’t think it was the best way to do it, and I’m sure our competitors would but again I think it’s a great question. I think you’re right directionally. But I think it is less you don’t even say it was a problem but I don’t think it’s a problem. I think most reasonable ways to come up with value will be more similar than you think.

Rick Ferri: Some people believe that a multi-factor approach to value might solve the problem, at least for retail investors. In other words, instead of trying to pick which price to something is value, you just have a formula that takes a little bit of everything.

Cliff Asness: I like that approach. We’ve written on that approach. You know everything is going a little too far, price divided by, you know, you can imagine crazy things. I think I even used this phrase a few minutes ago, price divided by anything reasonable. Make up your favorite four or five price divided by book, price divided by trailing earnings, price divided by forecasted earnings, price divided by free cash flow, price divided by sales. All reasonable and I can tell you advantages and disadvantages to each one. I would prefer a stock that is the most attractive on a weighted average of those five more than I would prefer a stock on any one. I’m not smart enough and I don’t know if anyone else is, to pick out which of those five is actually, I’m saying five, it can be many more than that of course, but I don’t  think we’re really smart enough to figure out collectively. I don’t think value is that, as you pointed out earlier, that well-defined a concept. I would rather take the average. 

Rick Ferri: So let’s get to another debate. Is the value premium, at least historically a result of risk or is it a result of a behavior?

Cliff Asness: Oh you’re going to get me in trouble.

Rick Ferri: Oh I am sorry.

Cliff Asness: I like to put it this way. First they both can have some influence. They both can contribute to a positive value premium. I know you’ve seen some of my work before. I’m not always a peacemaker. Sometimes I like to throw gasoline on a fire. In this particular case I often throw water on the fire because I’ve been in academic seminars where I point out you guys are arguing about this but you know the real world’s complicated. It could have a risk premium component and a behavioral component. And just to really mess with everyone, those can vary through time. Right now I think we’re in a fairly crazy period and I’ll go out on a limb and say I  think there are more mispricings than normal. So maybe behavioralism matters more right now, and maybe there’s sometimes it matters less and it’s more of a risk premium. I will tell you — and this is why I said you’re going to get me in trouble — even though I wrote my dissertation on momentum as a Fama student. I was probably two-thirds, one-third towards the risk story.

Doing this live for now 26 years, going back to ’94. I’ve certainly drifted more towards the behavioral story. I think living through about six cataclysmic events and at least three for value and being frustrated at those times, to be honest, that a world I thought was going a little crazy has certainly pushed me. But again, I’m not dismissive of the risk. I’m probably two-thirds/one-third the other way now. I’m not someone who feels, you know, intellectually confident saying we know the answer here. But at this point in my career if I had to bet on one and couldn’t split my vote I would lean behavioral. Though I warn you I say that on the coast. If I’m ever in the midwest and feel like Fama’s within a thousand miles I change my story.

Rick Ferri: It’s kind of odd though, that you would say that and here’s why. It’s because there’s been so much work done on behavioral finance and biases and everything, that you would think that we would have resolved some of our internal human issues and that we would notice that. But what you’re really saying is people are acting worse than they used to be even though the research by various top Nobel laureate behavioral economists are putting this information out. People are reading the books, right .

Cliff Asness: No, I think that’s a fair observation and I have a few thoughts on it. I think it comes down to this. You know value has been on a terrible streak, it’s kind of a 10 year drawdown. I think the first seven years of that drawdown were very different than the last three. I think the last three have been far less rational than the first seven. Remember my distinction that behavioral can matter but it’s not always behavioral. But even a few years ago, not ten years ago, even two, three years ago, maybe the most common question I got was along the lines of what you’re saying. Not just value, but other factors too, but we’ll talk about value here. If this is true, why doesn’t it get arbitraged away?

It could get arbitraged away by too much capital, or people could learn and stop making these biases. The learning part, which is more where you’re coming from, again I don’t see a whole lot of evidence that human biases are going away, or even ameliorating around the world. You know, we live in a little bit of a bubble where we’re very well versed in all these behavioral issues and we read all the literature, but I think your — not to pick on them it’s become an easy punch line — but you go to your average Robinhood trader and I don’t think we’ve gotten through on the arbitrage. Why aren’t these arbitraged away?

Remember strategy can go away for two reasons. It can go away for either demand or supply. Demand for strategy is too much capital chasing it, which even if there is a certain amount of bias in the world, you can’t extract an infinite amount of money from that, you can only take the other side and close the gap on valuation. So if the first trillion dollars closes most of the gap, the second trillion dollars doesn’t make any money. It’s closed, no one makes any money. Or it can go away because supply. And this is a little bit cynical, but this is behavioral. This is the supply of investor error. People have to continue to make errors and a limited amount of capital has to pursue them.

You know, again two and a half, three years ago the question was why do these things persist. After two and a half, three painful years to these strategies the question I get now is more how can anyone stick with these even if they’re great long term. How can anyone stick with these. And I love to point out that these are in a real way the opposite question. One is saying this is so obvious everyone should do it and it should go away. The other is saying no one can do it and I think there’s an element of truth to both. The real life strategies that can be done in large capacity, I mean a strategy like value that can be done at fairly large scale, is not a super high Sharpe Ratio. I think it’s real. And so a real positive Sharpe Ratio you can add to a diversified portfolio, that’s a wonderful thing, but it goes through as we’ve seen even before the last 10 years, we’ve seen horrible periods for value.

And the very fact that it can be very hard to stick which drives a fair amount of people out, makes it hard to seize the whole round trip. So you do have this kind of yin yang on the one hand, it should go away, on the other hand it’s really hard to stick with, and perversely I think low Sharpe Ratio, high capacity strategies that are a long-term investors good friend if you really can stick with them, they can improve a portfolio and I think they don’t go away precisely because they’re at that kind of risk-adjusted return that’s on the cusp. You can recognize it’s good some people can stick with them but a lot of people think they can but can’t. And so if something is real but you know ninety percent of the people throw in the towel when it’s had a bad three years, it’s going to stay an opportunity and it’s going to occasionally really hurt.

Rick Ferri: I think that the overwhelming interest in value investing that occurred in the early 2000s, after tech tumbled and value just had an incredible run, small cap value in particular, created this wave of fundamental weighted ETFs and a whole range of other factor based ETFs just came out of the woodwork and people were inundated with this. Because of all this interest and because of all this new knowledge that people were getting, it seems to me made this value factor not work. But also not work for a longer period of time than normally would have occurred perhaps.

Cliff Asness: It’s certainly possible. We certainly live in a cyclical business where people do, in fact, overreact to the prior big event, and it’s possible we all collectively did that at one point. I will say ever since 1999 when we started our firm, right before the key to the tech bubble really took off, we invented and have been tracking since then something that has come to be called the value spread.

This was — all right, Fama and French and many others in academia sort stocks on some valuation ratio and they look at the spread between the cheap and expensive in returns, but what was really missing, we couldn’t find it anywhere. It’s certainly possible someone privately had done it, but we didn’t see it in the literature — was anyone who looked at and thought the information was interesting in the magnitude: how cheap is cheap. Imagine all of value is price to cash flow and again, like we talked about before, I would much rather use many. But just to make it simple you’re going long low price to cash flow and short high price to cash flow. The expensive ones will always be more expensive on your own metric, but how much more expensive, and how much cheaper, what the spread is between them will vary through time.

And we did this in the teeth of a very terrible value market and not surprisingly discovered that after it being terrible, that spread had blown out to record levels. That didn’t have to happen. You can lose in a value strategy or any strategy, and have it not get cheaper. Simplest example. Imagine you bought a stock because it had a low PE and the price fell fifty percent. A naive simple approach to say, “Oh God, it got cheaper.” But if the earnings fell 75, your favorite metric PE doubled. It didn’t get cheaper, it got more expensive.

So you can actually lose in a strategy if you lose because the fundamentals move against you. It doesn’t necessarily get cheaper, so it was worth checking. And we did find it had gotten massively cheaper. Which means most of what was going on in the tech bubble, actually I think more than all, because I don’t think these were high quality companies even, was from price action, it was not fundamentals. We’ve been monitoring this spread in many different ways, in many different places, ever since. Halfway, of course it’s symmetric. Halfway to worry about periods like now where that spread is back out to approximately record levels.

If you’re beaten up in value is that money gone? Or meaning you lost on fundamentals, which you don’t give up. You may continue your strategy but you never certainly don’t expect to make that back. Or did you lose on price where you may have an expectation that it’s more attractive than normal now the premium is larger than normal. But we didn’t track this only for times like today. Maybe even more so we tracked it to see what if that spread ever compressed well beyond history. The only one I have memorized is price to book and I think if you look at the standard Fama/French construction the price to book of the expensive 30 percent against the cheap, call it 30 percent, it averages maybe five or six times. So the expensive are five or six times more expensive than the cheap. The low, it is maybe the high threes, maybe four was the low, and you’ve hit that a few times and the high was about 12. In the tech bubble 12 to 14 depending upon the precise measurement.

And we are back to similar levels to that, but one reason we’ve monitored this forever is what if that spread smashes down to one and a half, to a compressed level where the differences in valuation are just not very large, and in fact record small compared to history and then stayed there for a few years. Your hypothesis that maybe the world has figured this out, maybe they’re not making that error, or maybe too much arbitrage capital has come in, would be hard to argue with. You know if a value strategy exists because there’s a spread in valuations, if that spread in valuations is much smaller than it used to be, maybe the world has figured it out. And we’ve always felt monitoring that, in case it was time not to do it was the case. And we’ve seen lower and we’ve seen higher, and we certainly see extremely high today, but we’ve never seen it smashed down to levels that say value’s never going to work again because the world has figured it out. It could certainly still happen in the future.

Rick Ferri: Let’s pivot back to momentum.

Cliff Asness: Sure.

Rick Ferri: In a very short and eloquent, simple way if you can explain the formula for momentum, that’s number one, and then number two, you said that the momentum factor and the value factor are uncorrelated, or at least lower–

Cliff Asness: No, negatively correlated.

Rick Ferri: They’re negatively correlated, okay. So if they’re negatively correlated and you had them both together in the same portfolio. I mean how does that reconcile, or do you not have it in the same portfolio?

Cliff Asness: The simplest measure of momentum I know is still the one I used in my dissertation. It is simply one year total return, leaving off some of the very recent returns. I didn’t even use daily data, so I left off the last month in my dissertation. You don’t really need to leave a whole month off, you need to leave off a few days. That could be because we see some true reversal at the very short term, liquidity provision — you know if someone tries to trade a lot they push a price and it’s temporary price pressure. Or they’re just some microstructure issues if the last trade is at the bid, the price looks low and the momentum looks worse. But if the next trade is half at the bid and half of the offer, half the time on average you get a bounce, so it creates a little bit of an artificial contrarian strategy. So I often shorten it to one year price momentum and leave off that all the geeky stuff about having to be sure that you don’t accidentally get that contrarian part that may or may not be real at the very short term. But call it simply one year price momentum.

I actually called this the “fool strategy” when I presented it to Fama, and I think I was just currying favor there. I knew he wasn’t going to like the momentum strategy so I might as well be self-deprecating about it. But on the face of it you can’t get simpler, right. Buy a diversified portfolio, again you can do this, you can weight it in different ways, but buy a diversified portfolio of stocks with strong one-year returns and sell a diversified portfolio of stocks with weak one-year returns — on a relative basis, it doesn’t matter which direction the market has gone.

Now you’re, of course I won’t get into this, but just like value you can do this in more complex perhaps better, perhaps over engineered, we can always debate that, ways you can take out the industry bet. From this you can look not at total return, but total return net of say, a beta or multiple betas to different factors, what’s the excess return. There are a lot of different things you could do, and people do pursue these. There’s a lot of papers on tweaking the momentum factor just like there is with value. Same problem and/or opportunity depending on your perspective that there’s no one way.

But I think the base way in academia is still this one year price momentum. Now negatively correlated strategies, the the way I think about this is imagine your value strategy long cheap short expensive has a five percent average return and a five percent long-term volatility. And imagine so does your momentum strategy. If they are uncorrelated, you take their volatility and divide by the square root of two. So you do get, if you put half your money in each, you get a diversification benefit. If they are negatively correlated the volatility comes in even lower. An implausible negative correlation, if you thought they were negative one correlated, they both can’t have a positive mean, that’s a perpetual motion machine. Right, to say I have two strategies that are perfect hedges, perfectly opposite but both have a positive average that would be no volatility ever and making money every day.

Rick Ferri: It’s like going long the S&P 500 and short the S&P 500. That is absolute negative correlation, but you don’t make any money.

Cliff Asness: Exactly. But in real life value and momentum at times it’s higher, times it’s lower, but call it a negative 0.5 correlation and that is actually pretty reasonable. If you say it in English what you’re trying to buy, if you give, and I don’t say you have to give half your weight to each, and there are other things besides value and momentum, but if you gave half your weight to value and half your weight to momentum, the actual stocks you’re trying to buy — forget the quant stuff, just think about describing the stocks — are cheap stocks that are not disastrous on momentum, or even mildly cheap stocks that have good momentum, that are getting better. I used to laugh and be kind of obnoxious because when more traditional managers would all describe their process as “we’re looking for value plus a catalyst” and I used to roll my eyes and say, “You all say the same thing.” And then a friend who was an active stock picker said, “Don’t you do value and momentum.” I said, “Yeah.” and he said, ” Don’t you think you can describe that as value plus a catalyst.” And I said ,”Damn, I think I see where you’re going with that.” 

So just the mechanical math is if you have two strategies that are not unrealistically negatively correlated but are decent, they’re not perfect, hedges they just tend to move in the opposite direction and both make money. I think that is both possible at a minus point five correlation and you create a better risk return. Any time you create a better risk return you can take that benefit through getting basically the same average return at lower risk or you can be more aggressive and try to translate that into higher returns for similar risk. But I think the idea that there are two strategies that can both make money and partially hedge each other not fully because that can’t work, but partially hedge each other, I just think is what’s going on with value and momentum. And I think it’s quite intuitive. All else equal the price goes up, the momentum gets better and the value gets worse.

Rick Ferri: So let me ask a question. Dimensional Funds does this. They first screen for their value stocks, the stocks they want to buy based on their value screens, and then they buy the ones that have good momentum. I mean, so in other words value first and then momentum. How does that work?

Cliff Asness: Sure. I mean I have great respect, of course, for the guys at Dimensional. Fama

and French still do a ton with them, and there’s nobody in the industry I respect more. We do differ on this, like I said people of basically the same religion with small differences. You get your 30 Years War, you get your Shia/Sunni kind of thing where essentially we agree on almost everything but we disagree about this small point — let’s fight about that. So that definitely does happen. I am fond of saying momentum is very hard to reconcile with efficient markets. I think Ken and Gene would agree. By DFA using it as a screen, in my view, is very close to just using less of it than we do. You’re giving it a say over when you trade, but that’s it. You’re not giving it any kind of equal status. So I think it’s just a point on the spectrum.

I think their philosophy started out more value-based. I think they recognize that momentum has been a real phenomena and they don’t want to trade against it. We want it to actively contribute, not to dominate value but to actively contribute, and people who agree on a lot, who even like each other’s processes, and think each are good, can differ with where on the spectrum you would be. If you and I both designed a process with similar data and similar empirical results, we wouldn’t necessarily come out with the same weights. A screen is not technically the same thing, but I think it’s pretty close to thinking of it as them believing in momentum but less than we do. And I will say if they ever say something negative about momentum, I do make the “a little bit pregnant joke.” You’ve already shown you believe in it, you just believe in it less than we do. So I think we’re more similar than we are different on this, but yes, clearly we would give momentum closer, not necessarily precisely equal– we don’t have to weight these 50/50–  but we would give momentum more weight  than they would.

Rick Ferri: Well we sort of got off track a little bit because we’re actually-

Cliff Asness: That’s my specialty.

Rick Ferri: No, no, no. This is fine. This is perfect but we haven’t even gotten out of college yet. I mean you didn’t, you obviously got your PhD. so they accepted you.

Cliff Asness: Well I’m more comfortable with the geek stuff than with the autobiography.

Rick Ferri: But then you went to Goldman Sachs.

Cliff Asness: I went to Chicago. I started a dissertation on momentum, when I got an offer from friends who were at Goldman Sachs Asset Management, starting up, a taking over, it was really, it was close to a startup, but taking over a fixed income area. Initially I went for a summer which was great. They then said, why don’t you come for a year and see if you like this as compared to being a professor. And I still thought I wanted to be a professor at that point, but I did recognize that free options are good things. They teach us that in finance. So I said, let me try this.

 I ended up being there about a year and a half. Was working on my dissertation at night. That was an odd experience, to be a portfolio manager, a trader and a modeler in fixed income during the day, and go home and work on quant equity at night. I was getting closer, to very close to being done. Fama and French were great about doing my dissertation from afar. I went back to Chicago, I proposed it, I went back and defended it. They wanted me to go on the academic job market. I got again very lucky. PIMCO, the west coast asset manager, wanted to start a quant research group. Quant was getting more popular. It was still a relatively new thing, believe it or not, back then. PIMCO ended up offering me that job. 

As an aside, the guy who took me out to lunch asked if sushi was okay and I had never had sushi, and hardly ever had fish at that point in my life, and being a cowardly 24 year old or whatever I was something 24 – 25, I said, “Yeah sure, sushi would be great.” And one point, when he wasn’t looking, I did spit my food into a napkin, gotta come clean about that.

But I went to Goldman and said I’m thinking of doing this.This seems perfect. I’m trying to decide between academia and being a practitioner, and this seems to be the best of both worlds. And to be brutally honest and I feel bad, it was kind of PIMCO’s idea and we ran with it. But Goldman swore to God they were looking to start the same group, and they were thinking of talking to me. Again, I made the odd life decision not to live, first not to live in Palo Alto, and then not to live in Newport Beach. I have managed to avoid good weather rather consistently in my career. But Goldman says start this group and I said, “Yeah my family’s from here.” I like Goldman, they were treating me well. I’m an east coast guy, okay let’s start this.

And I hired three people.That gave me a budget for a four-person group. Two were fellow PhD students at the University of Chicago.  One is still my partner today at AQR. We ended up both building direct quantitative processes to run client money over the course of kind of the next three years. We ended up running probably just about seven billion dollars, about six billion in traditional kind of beat the benchmark, no leverage, no shorting just regular stock picking, though in a quantitative manner. And about a billion dollars in long/short, more hedge fund kind of work. To be honest, the hedge fund work being less constrained, we think is a more pure version of what we do, but it’s not that different. When our factors beat the benchmark we tend to get positive returns in a similar based hedge fund product. The hedge fund is kind of the alpha part.

Life went well. This group though had much broader responsibilities than running money. We were also in charge of pretty much anything quantitative. We were like door to door quants. Especially before we were running money we were desperate for someone to give us a task. So we were doing asset liability analysis, running efficient frontiers for their marketers, and while there’s absolutely nothing wrong with that–it was good stuff– it was really awkward to be running a group that was running billions of dollars and growing and doing well, and be in charge of the graphics for the marketing people, for other people’s products.

Sometime, I think it was ’96, I had one of my other partners, still to today, a guy named David Cabiller, started working on me, particularly on the hedge fund side saying, “You know, we really could do this on our own. You guys are at Goldman Sachs but you buy all outside data. You don’t trade off Goldman Sachs’ strategists, it’s all self-contained.” And there have been various hedge funds booms and busts but that was a boom time and he thought we could do it, and I spent the better part of ’96 doing my Hamlet imitation, you know, should I do this, should I not, indecisive. Mid ’96, I went to Goldman and actually said, “I’m thinking about doing this.” It was very straightforward, “I’m thinking of going off on my own.” 

They said, “What would it take to keep you and the whole team, not just me.” I said, “Well certainly money is part of it.” And I won’t pretend it wasn’t. We didn’t even propose a deal where we’d make anything like if we ran our own successful hedge fund, but we proposed something in the middle, some kind of sharing arrangement. What was really odd was they certainly didn’t  jump at our first offer, but they met us a decent part of the way on the more crass issue just of money. But they were very insistent that we kept doing the support work. I ended up saying yes. Goldman Sachs does many things well but retaining employees they want to retain, it’s something they are excellent at. Suddenly you are sitting with Jon Corzine, then president of Goldman Sachs, multiple times being told how much they want to keep you. And it turns your 20 something-year old head, and I ended up along with the rest of the group deciding to stay. And Rick, you ever have one of these giant decisions, that as soon as you finally make the decision after agonizing, you get a wave of relief, partly just because it’s over, but partly it often seems obvious what was not obvious when you were agonizing seems obvious after the fact. You’re like, well yeah, I don’t know what I was agonizing about.

I never got my waiver relief. You know, about a month later I was still going “we should be doing this on our own.” And David, again he was the Mephistopheles, the tempter, in this story. About a month into it he and I were talking and I said, “You know, is that still an option?” and he–you don’t say that to David– that was, I gave away the whole ball game at that point. So we agonized for a few more months, but at the end of the year we did decide to go and do it on our own.

But the first time you try to leave, when there’s someone you want, when you’re someone they want to keep, you get the full charm offensive. The second time you do not get the charm offensive you get quite the opposite, largely because when you come back the second time they know you’re going. But the partner I reported to was not happy with me. Was still one of the scariest things in my life was telling him I was resigning. I remember that I was a total child. I played like psych up music on my Sony cassette walkman in my office before I went down to his. I was, I was not calm, but it worked out.

Rick Ferri: Well, it sure did. I mean you ended up creating an incredible powerhouse, AQR. By the end of last year you had 186 billion dollars in assets under management. That’s a phenomenal number. So clearly you made the right decision and, of course, to be congratulated So all that your father would be proud of you.

Cliff Asness: I hope so.

Rick Ferri: So I have just a couple more questions from the Bogleheads, and I’ve asked you some already, but I just have a few more. Now this has to do with things that you’ve wrote about over the years, and one of the things you wrote about was the 60/40 portfolio, and is it a good solution going forward. And just realize that a lot of investors kind of invest along the 60/40. What’s wrong with 60/40?  We hear a lot about this recently. I mean, what do you think?

Cliff Asness: Yeah. Well first of all, you know pretty much ever since we and others have been saying that 60/40 just chugs along. Some wild volatility at times, you know mid-March this year it didn’t seem like it was chugging along. But on net it’s continued to do well. My issues with 60/40: there are two, a long-term one, and a conditional one on where we are today. The long-term one is there’s nothing wrong with it. I think it’s great to have that as a core of your portfolio. There’s no magic to the particular weights of 60/40. Of course like everything else we pick, kind of a benchmark and then it gets holy over time.. And I don’t think  65/35 or 55/45 is going to be any great sin. But the general idea of being diversified across stocks and bonds as a core is great.

We do think you could do better, both in a traditional sense, you know all the classic things we talk about: being global, at least to some extent. Yeah you can still have a home bias, but you know 60/40 is often thought of as the domestic US portfolio. There are other ways to make it better but if you believe that there is a value premium, there is a momentum premium, then you can just view that as another investment that’s not very correlated to 60/40.  So we do think long term you can at least somewhat improve the risk adjusted return or the total return if you choose to take enough risk of 60/40. 

Today it’s a little more serious. These are very long-term phenomena. We always, thank God, we caution don’t use this to time the market, don’t use this to make a one-year forecast. But stocks you know, pick your favorite measure. I remember when the Shiller CAPE was a brand new baby in 1996, but it’s become kind of the lingua franca and talking about how expensive the stock market is. Just like any valuation measure, it’s not the only one by any means, there are a whole set of them. The Shiller CAPE, I think last time I looked just a few days ago was about 32. It hit the low 40s at the end in ’99 –  2000. So if you graph it back 100 years, the only way it looks reasonable is if you squint and compare it to the tech bubble. If you compare it to the whole series it’s somewhere in the, again don’t hold me to the precise number, the 92nd percentile expensive versus history. And statistically it’s actually hard to say strong things about long-term returns because you don’t get to observe enough, let’s call it 10-year or longer periods, but what we do get to observe seems to fit with intuition. When you buy stocks more expensive, the average return is lower going forward, and vice versa. So therefore we’d expect positive returns out of stocks. We’re not people who are forecasting giant mean reversion in a crash, but we do forecast the lower risk premium for long-term investors from here.

And this is even easier because it fits intuition even more so, the same for bonds. When bonds are selling for near zero real yield against economists’ forecasts of long-term inflation, not many of us need great convincing that they’re probably from here in the next call at 10 years going to make less on their bonds than history, in a real sense against inflation. If stocks are less than normal and bonds are less than normal, then you could do some very complicated math, this is stochastic calculus. You take sixty percent of your estimate for stock returns and you add it to forty percent of your estimate for bond returns and you get your estimate for 60/40.

Rick Ferri: Four percent is really what you’re looking at, maybe.

Cliff Asness: Nominal. We think it’s about one and a half percent real.

Rick Ferri: Nominal.

Cliff Asness: No you’re dead on. And I prefer real because it just adjusts for the environment. But you know that comes out to about one and a half percent real, right. Now history is more like five percent, meaning if you could buy a diversified portfolio at a low transaction cost and a low fee, that’s a big if because there’s a long part of history where you really couldn’t do that, but if you could you got five percent real just for showing up. We think you get one and a half, two percent real just for showing up today.

And that’s not negative, you can make it negative if you want to forecast big mean reversion, but that’s market timing. That’s saying we know these prices are going to drop. I don’t know that this could be the new normal. Yeah if the prices drop it’ll be even worse but I’m negative enough at one and a half, two percent If you do have alternatives that can make a portfolio better on average I think they’re more important when the core portfolio is offering you less.

Rick Ferri: Let me ask a question about international stocks because I look at the international market and I look at the industry group weightings of ex-US or international stocks versus US stocks and I see very different industry groups and I see–

Cliff Asness: And the US is certainly more tech.

Rick Ferri: And more healthcare and now more communications so it’s a different makeup. It’s like 50 years ago. We did everything in this country. Now we import all our desks and computers and just about everything I’m looking at right here in front of me, but my point is that if you look at the valuations of international stocks — just use the Vanguard Total International Index Fund — and I didn’t know I was looking at the Vanguard Total International Index Fund, I was looking at the price to earnings ratio, right, price to book ratio, return on equity growth rate. I thought I’d be looking at a mid cap value index and then, lo and behold, you know you pull the cover off and it’s the international market.

Can’t people who are doing say a 60 /40 portfolio and to your point that a lot of people are thinking about just US stocks, can’t they diversify, or maybe get some of the value premium that you’re talking about as simply as adding instead of having all US, having 40 percent US and 20 percent international. Would that work? 

Cliff Asness: Absolutely. I would say you really don’t have to choose one or the other, the international versus the US, even when international is cheap, is not highly correlated at all to when value works, as we do it within industries, for picking individual stocks. You know the basic math of portfolios says you should do everything you believe in that’s diversifying. You should just have a high standard for what you believe in so you don’t end up doing a ridiculous amount. But I don’t think these are necessarily competitive solutions. I think they’re complementary solutions. You can do half of what you would have done over each and and have less risk of disaster if you’re wrong on either one.

But absolutely I do think even if you adjust for the industry –a long, long time ago we wrote a paper in the mid 90s showing that cheap countries with good momentum tend to outperform expensive countries with bad momentum– if you just pretend the index is a single stock and add up the numbers for the underlying components. But years later we did notice exactly what you are talking about, that of course you can get some very different industry components, particularly in some of the smaller countries, where you don’t have a very large cross-section. You go to Belgium and you get a ton of financials. You go to Australia and you get a lot of resource stocks.

So this gets really geeky and this is your fault Rick that I’m doing this, you brought it up. But years ago we modified what we do in comparing countries. We still compare them straight out against each other as you were doing, but we also do something where we look at a country’s industry composition. We look at the global — including the US — total cap weighted global industry, get its valuation forecast based only on its industry composition, what we think the valuation ratio of that country would be if everything was selling at an average for the industry and then call it cheaper/expensive only if it’s lower or higher than that forecasted price. So it’s another, it’s kind of like a similar thing. Remember earlier I said we take our value bet without taking a whole lot of industry exposure, at least I think I said that earlier. I say that most of the time.

Rick Ferri: Yeah that you did. You said that.

Cliff Asness: But we found that makes the country betting a little higher of a risk adjusted return, but we still do both, we give weight to both and both we’ll say, even adjusting for industries, we think the US is expensive against the world. Like everything else value alone, particularly for big decisions like US versus global, you could be wrong for a long, long time. It’s just a question of is it a smart bet. And we do think it’s a smart bet to have some outside the US partly because we always think that, because we like diversification, but also conditionally as you’re saying because it has more of a value tilt than it normally does, right now.

Rick Ferri: Well since you agreed with me on the last point and made me look very smart, I think I’m going to end the podcast here.

Cliff Asness: You’re afraid one more question, I’ll disagree and it’ll all collapse. That’s smart. Remember Seinfeld, that’s George Costanza has learned that if he ever says something smart in a meeting he excuses himself from the meeting immediately and says thank you, and just, you know, claims victory.

Rick Ferri: And that’s what I’m going to do. Cliff, thank you very much for being our guest, and we really appreciate your insights.  It’s been a real pleasure having you on the show.

Cliff Asness: Thank you Rick. I enjoyed it very much.

Rick Ferri:This concludes Bogleheads on Investing, podcast number 27. I’m your host Rick Ferri. Join us each month to hear a new special guest. In the meantime visit Bogleheads.org and the Bogleheads wiki. Participate in the forum, and help others find the forum.Thanks for listening.



About the author 

Rick Ferri

Investment adviser, analyst, author and industry consultant



You may also like