Are financial markets becoming less efficient? Famous investor Cliff Asness certainly thinks so. In his paper published last year, “The Less-Efficient Market Hypothesis,” Asness argues that social media and low interest rates, among other factors, have distorted market information so that stocks have become disconnected from their true values. This distortion has directed funds toward undeserving assets and firms and staved off necessary market corrections. Asness is the founder, managing principal, and chief investment officer at AQR Capital Management. He is an active researcher on various financial and investment topics and received an MBA and PhD in finance from the University of Chicago Booth School of Business. From her early days as a journalist reporting on Wall Street, Bethany recounts Asness as an outspoken, successful quant investor: one who invests based primarily on the fundamentals of the market rather than those of the firm. She also remembers him being “colloquial” and willing to be “experimental” with ideas. Asness’s recent paper continues that experimental style as he challenges the legacy of the efficient market hypothesis on which his PhD advisor, Nobel Prize laureate Eugene Fama, made his name, and which argues that asset prices reflect all available information, making it impossible to “beat” or outperform the market. Asness joins Bethany and Luigi to discuss how the market has fundamentally changed due to new technologies and macroeconomic trends and how investment strategies must adapt, what these changes mean for long-term productivity and growth, how researchers and investors should think about emerging market factors like tariffs and artificial intelligence, and why he's not investing in TrumpCoin anytime soon. Disclosure: In October 2024, Chicago Booth received a $60 million gift from Cliff Asness and John Liew to name its Master in Finance program. Bonus: Revisit our recent episode with Eugene Fama, Why This Nobel Economist Thinks Bitcoin is Going to Zero
Are financial markets becoming less efficient? Famous investor Cliff Asness certainly thinks so. In his paper published last year, “The Less-Efficient Market Hypothesis,” Asness argues that social media and low interest rates, among other factors, have distorted market information so that stocks have become disconnected from their true values. This distortion has directed funds toward undeserving assets and firms and staved off necessary market corrections.
Asness is the founder, managing principal, and chief investment officer at AQR Capital Management. He is an active researcher on various financial and investment topics and received an MBA and PhD in finance from the University of Chicago Booth School of Business. From her early days as a journalist reporting on Wall Street, Bethany recounts Asness as an outspoken, successful quant investor: one who invests based primarily on the fundamentals of the market rather than those of the firm. She also remembers him being “colloquial” and willing to be “experimental” with ideas. Asness’s recent paper continues that experimental style as he challenges the legacy of the efficient market hypothesis on which his PhD advisor, Nobel Prize laureate Eugene Fama, made his name, and which argues that asset prices reflect all available information, making it impossible to “beat” or outperform the market.
Asness joins Bethany and Luigi to discuss how the market has fundamentally changed due to new technologies and macroeconomic trends and how investment strategies must adapt, what these changes mean for long-term productivity and growth, how researchers and investors should think about emerging market factors like tariffs and artificial intelligence, and why he's not investing in TrumpCoin anytime soon.
Disclosure: In October 2024, Chicago Booth received a $60 million gift from Cliff Asness and John Liew to name its Master in Finance program.
Bonus: Revisit our recent episode with Eugene Fama, Why This Nobel Economist Thinks Bitcoin is Going to Zero
Cliff Asness: Our goal is to make our clients money, not to make markets more efficient. That is a lovely secondary thing that I believe we help with, I hope we help with, but we don’t wake up every day saying, “Our job today is just to make markets better.” I probably should not have admitted that on a major podcast.
Bethany: I’m Bethany McLean.
Phil Donahue: Did you ever have a moment of doubt about capitalism and whether greed’s a good idea?
Luigi: And I’m Luigi Zingales.
Bernie Sanders: We have socialism for the very rich, rugged individualism for the poor.
Bethany: And this is Capitalisn’t, a podcast about what is working in capitalism.
Milton Friedman: First of all, tell me, is there some society you know that doesn’t run on greed?
Luigi: And, most importantly, what isn’t.
Warren Buffett: We ought to do better by the people that get left behind. I don’t think we should kill the capitalist system in the process.
Bethany: From my very early days as a journalist, I remember the famous investor Cliff Asness, not just because he was a hugely successful quant—meaning someone who invests not so much on the fundamentals of a company but, rather, due to quantitative factors in the market—but because he was willing to say things other people weren’t and in such a memorably biting way.
In 2000, he published a piece called “Bubble Logic,” which exposed the fallacies being used to justify crazy stock prices like that of Cisco. Then, in 2004, he wrote a piece entitled “Stock Options and the Lying Liars Who Don’t Want to Expense Them.” Cliff now manages $128 billion across various strategies, and those who follow him on X know that he’s as outspoken on lots of issues as ever before.
When fellow billionaire Bill Ackman described Trump’s tariff policy change as brilliantly executed, Asness replied: “One of the main benefits of making some money is not having to wear a gimp suit for anybody. To each his own.”
Cliff later apologized for the language, but he still said Ackman was being illogical, and he said: “It may or may not have been good negotiating, but the man clearly feels any trade deficit with any country is stealing from us and has believed this for 40 years. So, there’s clearly some idiocy about the actual topic at hand to go with potential luck or brilliance.” Wow, certainly outspoken.
Luigi: Indeed. He also wrote a fascinating paper arguing that the market is less efficient than ever before, in part due to the rise of social media, about which he says, “Has there ever been a better vehicle for turning a wise, independent crowd into a coordinated, clueless, even dangerous mob than social media?”
Bethany: He’s also willing to be colloquial and experimental with ideas in a way that I don’t think are typical of University of Chicago finance types. Sorry, Luigi.
Another factor he thinks explains the growing inefficiency of markets is these several decades of super-low interest rates, and he wrote this: “Well, perhaps super-low interest rates for a long time make investors go cray-cray. Yes, I know cray-cray is not covered in the standard CFA exam and rarely the result of formal analysis, but it seems at least possible to me.”
Luigi: The reason we wanted to have Cliff on the show is precisely because he’s so opinionated, and it’s so refreshing to see him attacking left and right. He’s not somebody who is in a position trying to push something. He’s really a free spirit. He once described himself as a part-time Republican, a full-time libertarian, and he has been outspoken in his defense of capitalism as leading to a better living standard for everyone: “If you go by living standard, you cannot escape that things are wildly better. I will attribute most of that to capitalism.”
Bethany: Cliff has no shortage of things to say for himself, so we’ll stop waxing on and start talking to him. He is the founder, managing principal, and chief investment officer at AQR Capital Management. And a quick disclosure we need to make, which is that Chicago Booth a year and a half ago received a $60 million gift from Cliff Asness and John Liew to name its Master in Finance program.
One of the things you’ve been talking about a lot recently is your less-efficient market theory, and I think, the market is reasonably close to efficient, but there are lots of little inefficiencies. Anyway, you’ve argued that for a bunch of reasons, the markets have gotten less efficient.
You wrote this in one piece, “Hence, if those prices don’t reflect reality, there are real consequences to long-term productivity and growth.” How much do you worry about that? How important is it that the market actually does reflect back real-time information about prices?
Cliff Asness: Well, you started out with a hard one. First, let’s back up. It’s a total straw man that markets are perfect. Literally, no one believes that.
I was Gene Fama’s TA for two years. The first two weeks of class he teaches us what the efficient market hypothesis is, and the third week of class he says something like, “Markets are assuredly not perfectly efficient,” and you get a gasp, which you get nowhere else in the world. The rest of the world either has no idea what you’re talking about, but if they do, they’re like: “Of course it’s not perfect. Perfect is silly.”
Gene, I’m pretty sure he thinks they’re more perfect than I do—and I might think they’re more perfect than many others—but once you acknowledge they’re at all imperfect, the questions of how imperfect, and has that changed over time, become still very hard to answer, but very legitimate questions to ask.
We are mostly, in the piece I wrote . . . I start out at the very beginning saying this is a highly opinionated piece. There’s not a lot of hard data. It’s life experience and a few vignettes of some huge things that I would call bubbles. That’s not a word Gene Fama likes, but it’s my contention that markets have gotten materially . . . In the piece I think I kept saying less efficient. After writing it, I decided it would be more accurate to say “prone to bouts of extreme inefficiency.”
I have two periods I really point to. What we’ve lived through always carries more weight with us than what we can look at in the CRSP data going back to 1926. But the dot-com bubble at the end of the ’90s—I reveal my opinion by even calling it the dot-com bubble. Gene Fama would call it “the time the dot-coms were priced as very low-risk assets to a very low expected return.” It rhymed more than it didn’t but also was not perfectly the same.
But 2019 and 2020 culminating in COVID—COVID was not the sole thing, but even before COVID, we saw models for what we considered spreads between cheap and expensive stocks getting to at least tech-bubble-like levels. They were getting there before COVID, but then COVID kicked them up to, as they say in Spinal Tap, “They kicked it up to 11.”
You remember COVID, when all you were supposed to own was Tesla and Peloton? We didn’t see that stuff, at least in the data, if you measure things similarly for the prior, call it 50 years. The dot-com bubble was the largest event in this disparity we look at, ever in the data. You can argue with how good the data is as you go back in time. You can literally go back to the ’20s. I’d probably go back to the ’50s, where I feel like I’m at all confident in it.
But it was the biggest by far we’d ever seen, and you could say that was a once-in-a-lifetime event, and then, almost exactly 20 years later, it became a twice-in-a-lifetime event. That sent me down a road of thinking, “I probably can’t solve this, I can’t put three decimal points on it, but what might have changed to cause this?”
Like everything else in our field, it can be completely random, and we could all be trying to explain randomness, but I’m of the belief that we have seen at least two . . . You could say the GFC was one also, though the GFC was probably more of an economic event that became a market event, as opposed to just a pure euphoric mispricing. But call it two-and-a-half-times in my career that we’ve seen craziness that I didn’t expect to see more than once every 50 years.
Luigi: Bob Schiller, in his book Irrational Exuberance, has a very interesting observation that bubbles—and he does believe in bubbles—got started with the beginning of media. The first bubble was in the Netherlands, the tulip bubble, and this was at the same time when newspapers got started, because the only way you can have a bubble is that contemporaneous irrational exuberance. You need to be coordinated by something, and the media plays a role. You have mentioned social media as being part of this craziness that made markets less efficient.
But I want to go to your roles, like Lloyd Blankfein of Goldman Sachs said that he was doing God’s work. So, to what extent has your work —
Cliff Asness: I happen to love Lloyd, but I think he’d take that one back if he could.
Luigi: Now you have your chance. I want to know, to what extent is your work making the market more efficient? And to what extent is the work of a lot of other people in finance making the market more efficient or not, and can you tell them apart?
Cliff Asness: Sometimes they’re not easy to tell apart. If you are doing an old-school, Graham and Dodd-type strategy, like low multiples, high profitability, low volatility, I think it’s pretty unambiguous that that manager is putting their money behind things that are actually cheap, selling, or at least avoiding, things that are actually expensive, and they are a force for making the market more efficient.
Take a strategy like the price momentum strategy. Luigi, you want to do something scary? In 1989, you tell Gene Fama you want to write a dissertation saying that if you buy the stocks that have been going up lately and sell the ones that have been going down lately, you beat the market. He was very kind about it, to be honest. He said, “If it’s in the data, write the paper.” But he’s never liked the fact that it seems to work.
Ultimately, let me give you a truth that will occur with many of my answers. They will end up unsatisfying, because I won’t know the final answer. But to give you an example where it gets more ambiguous, say you believe price momentum works on average. Is that making the market more efficient or less efficient?
You might instantly think: “Well, momentum traders are crazy. They make the market less efficient.” But there are at least two major hypotheses for why momentum or trend-following strategies have worked long term, and when I list the two, you will see that they’re somewhat embarrassing to the field that these are the two main competitors.
One is markets tend to underreact to new information. The other is markets tend to overreact to new information. When you’ve narrowed something down to those two, you do occasionally have to step back and go, “What have I done with my life?”
But I don’t want to get existential about this. I am trying to be funny about it, and it is, I think, amusing, but both can be part of it. Markets can underreact to new information. The classic example is there’s an earnings announcement and there’s post-announcement drift, where the whole impact of the earnings doesn’t get into the price immediately.
Dick Thaler types will tell you behavioral psychology has a concept called anchoring and adjustment. When you get new information, you move towards it, but you don’t fully incorporate it. So, markets might underreact.
On the other hand, pure feedback trading, where people are just chasing prior returns, can have a self-induced, almost tautological effect that caused that strategy to work for a while, and maybe it leads to big busts occasionally for that strategy, but price momentum can feed on itself.
I think it’s pretty clear, if it was all underreaction, the momentum trader is moving the markets towards equilibrium. They are pro-equilibrium. Because, by definition, the market has underreacted in that case. It should have moved more. Anyone trying to buy on that is going to make a little bit of money, but they’re going to help push the price towards where it should have gotten immediately.
If it’s trading on just positive feedback, I think more often than not—not always, sometimes the feedback randomly will be in the direction of equilibrium—but I think it’s a much stronger case that those people would be making the markets move away from equilibrium and be less stable and cause larger and crazier events.
Of course, in real life, everybody—they do it in politics, they do it in economics—wants to come up with the solution. My theory, underreaction, overreaction, is it. You’ve read more than I have, but we’ve both read a million academic papers that say: “Here are these 12 theories. I can rule out 11, and this is the real one.”
Real life is probably the case where half the theories are totally nonsense, and the other half have some degree of truth to them. Overreaction and underreaction can exist at the same time. There can be positive-feedback traders, and there can be an underreaction to true information.
Just to make it even worse, for those of us who’d like the world to be tractable, they can vary through time in which one is driving momentum returns. In call it normal times, momentum may be mostly an underreaction strategy. It’s a little boring. It says, when information comes out, prices don’t move all the way. So, you bet it continues on.
But occasionally—and maybe this is plus or minus two years around major, I’ll use the B word again, bubbles—maybe feedback-loop strategies are more important at those times.
By and large, I think quants, factor traders, even active managers who are following a disciplined valuation-based strategy . . . I know, from our experience, we have suffered in what I consider major bubbles and more than made it back in the aftermath of major bubbles. I don’t know what it would have looked like without people like us, but I think we were fighting to keep the markets more towards equilibrium and to bring them back to equilibrium. On net, there’s that.
But I absolutely admit, on occasion, if we’re following price momentum, and it’s currently irrational price momentum, just based on chasing trends, we could be adding to it. I think, on net, we probably make markets more efficient, but I also admit our goal is to make our clients money, not to make markets more efficient. That is a lovely secondary thing that I believe we help with, I hope we help with, but we don’t wake up every day saying, “Our job today is just to make markets better.” I probably should not have admitted that on a major podcast.
Bethany: I was thinking, as I read through your interviews, about that comment you made about Fama telling you to go where the data is, but one of the things that seems to stand out about your brain is this oscillation between the data and the idea. You don’t simply follow the data. You have an idea, and then you look to see if the data will justify it. But has your relationship to data and how you think about data changed over the years, and what comes first for you? The data pointing you in a direction, or having the direction in the idea and then seeing if the data can back it up?
Cliff Asness: Oh, you’re going to regret asking me that because I might just talk the rest of the podcast. The bête noire of quant trading or academic work is over data mining, overfitting the data. For many, many years—forever, actually—if you go back to the beginnings of AQR or the nucleus of our group that started at Goldman Sachs, we would have told you roughly—there’s no way to quantify this precisely—"We think we give about half influence to each.”
We don’t trade something with a great theory—that has never ever worked in the history of mankind—and there are examples like this. The capital asset pricing model is an example where there’s pretty much no evidence it’s ever worked anywhere, ever, even though it is a beautiful theory; it’s a gorgeous theory.
On the other hand, pure data with utterly no theory . . . In the US, timing the stock market with the Super Bowl effect: buy when the NFC wins, sell when the AFC wins. In fact, it’s not just that—I’m going to get it a little wrong, I don’t remember the specifics—but there’s a tweak to the rule where if there’s a team in the AFC that moved over from the old NFL, you don’t buy, or you do buy in the opposite direction, and I think it was because the Pittsburgh Steelers screwed up the rule. Your nose for data mining should be on fire at that point, if you not only have a Super Bowl rule that has nothing to do with markets, but you’ve got to have a little codicil to it.
I did once present that by accident to a board where one of the owners of the Pittsburgh Steelers was on the board, and everyone starts laughing, and I’m like, “I try to make a lot of jokes, but I don’t think I made one.” And they’re like, “He owns them.” And I’m like, “OK, that’s kind of funny.” That, we would not trade.
That’s art, not science, saying what that mix is. But let me give you one twist. Probably in the last five years we’ve moved in the direction of data mattering more relative to story than in the past. We have not thrown out story. This is not a binary, zero or one, but if it used to be 50/50, is it 65/35 now?
I hate being so cliché—a quant talking about machine learning is very cliché—but we think we have made progress on using some of those tools. Almost by definition, if the machine is learning, if it’s ridiculously, obviously intuitive, you didn’t need the machine. The machine is doing something subtler than you’re able to pick up on your own, and we do believe that in many cases—not every one—we can add some value with that.
One of my colleagues, a Yale professor, Bryan Kelly, he’s an AQR partner and a Yale professor. Some of you guys in academia get a great deal, Luigi, you get two full-time jobs. It’s beautiful. He has a paper called “The Virtue of Complexity,” which is exactly this counterintuitive idea that sometimes having more parameters than data can actually be less of a sin than we used to think it is.
But there are still problems, huge problems, in finance. What actually is the equity risk premium? How much should an investor expect to make on stocks versus bonds, or the value premium, or the small-cap premium? I don’t think ML is going to get us a drop closer to that. The problem there is we don’t have nearly enough data, and I don’t even know how I’d go about applying ML to how much stocks should beat bonds by and what they have done over the past.
Luigi: You know, Cliff, that Robert Novy-Marx has a paper now that automatically identifies some factors and then puts the factors into ChatGPT and has ChatGPT come up with a story. ChatGPT is pretty good at coming up with a story, so the ex-post rationalization is alive and well.
But since you raised the issue of machine learning and general AI application to finance, how is this going to change your field? Is the market going to become more efficient, or is it going to become less efficient with the use of AI?
Cliff Asness: Oh, you guys, do you have any easy questions on your list? I think net more efficient, but most of the ways it’ll become more efficient are the small pockets of inefficiency.
When I talked about the less-efficient market hypothesis, I’m talking about big, obvious things. I doubt ML is what’s going on there. I think it’s much more mob psychology, and the hard part there is not coming up with the right model. The hard part there is not throwing out your model at the bottom voluntarily or not having your client or whoever’s funding you throw it out on you involuntarily by taking back the money.
There are times at these extremes when the challenge isn’t figuring out what’s going on. The challenge is holding on. I don’t see ML making that task much easier unless I can go, “ChatGPT, how can I convince a recalcitrant client to stick with something that I know is going to work in the next five years that has worked horribly over the last two and a half years?” And it just gives me a magic paragraph. That’s the only way I could see ML changing.
Having said that, one thing I’m fascinated by is the very fact that I think ML is finding things that are harder to put into words that we think are profitable to trade means, if that’s true . . . And I step back, and I go, “I always say we could be wrong.” But if that’s true, that means there’s a much subtler structure to markets going on than we realize. There are nonlinearities going on. There are things that work differently within different industries. There are things where one factor has to be crossed within another factor. Markets might change in what they reward more through times. There might be more seasonalities that ML is pretty good at incorporating.
I think this stuff is great. It makes the market more efficient, but in a more narrow, short, “This stock is slightly off, and we’re going to fix that,” sense. Project far enough in the future and none of us has a job due to ML, so maybe one day it’ll do it all perfectly and it’ll also be doing the podcast for us, but I don’t see ML making us more bubble resistant in the near term.
Bethany: To use that unspeakable word, bubble, we actually had Gene Fama on the podcast, and he said—a view I think you might share—that Bitcoin actually is a bubble. He said that if it doesn’t go to zero, he’s going to have to rethink all of monetary theory. Do you share his point of view on that, or would you caveat it in some way?
Cliff Asness: I would caveat it, but very slightly. I’m mostly in the Gene direction. Here’s a bold statement: I’m not as cynical about Bitcoin as I am about Trumpcoin or Fartcoin. I blame these people that I’m on a sophisticated podcast and I have to say the word Fartcoin. It’s their fault, not my fault.
Bitcoin, I am still net cynical. I am still net in Gene’s direction. What would disprove it to me one day is if there’s a massive use case. I’ve yet to see one outside of smuggling. When your best use case is criminality, it feels a little weird that you’re going to base a whole currency on the fact that you can ransom people with it. But it’s not impossible.
One of the things I’ve admitted to myself multiple times in my career, and then I forget, is I don’t understand macroeconomics and particularly monetary economics very well. About four times, over five-year spaces, I’ve read or reread one of the major macro textbooks. I’ve read Paul Krugman’s, I’ve read Robert Barrow’s, I’ve read Tyler Cowen’s, I’ve read Greg Mankiw’s. These are all great authors, but I will tell you, at the end I come out saying, well, the bad news is, I still don’t think I really understand macroeconomics and particularly monetary economics. The less bad news, at least for me, is I don’t think they do, either.
Hard money, we all understand. I’m not a hard-money person. I’m not saying we should be on a gold standard, but at least we understand what it is. Fiat money has a little magic to it. It always has a little magic to it. There’s always a little faith-based nature.
So, I am not so confident that I can look at Bitcoin and say I’m 100 percent confident that the world won’t suddenly decide that we have more faith in this than anything else and use it. I think it’s a very unlikely scenario. I think everyone into Bitcoin claims that they don’t care about the US dollar. One Bitcoin is one Bitcoin; it’s all about the Bitcoin. I’m pretty sure they’re very excited about their gains in US dollars.
They make some arguments I can’t stand, like, the supply is ultimately capped. That’s better than uncapped, but they act like that in itself has value. That doesn’t create value. A use case plus limited supply creates value. My poetry—there’s only so much poetry I can produce. I promise you it has zero value.
Someone in AQR compliance would like me to disclose that we do trade some crypto. We only trade it in pure price trending models, because that’s the only thing we understand about it.
Luigi: I think you are a little bit too negative about your poetry. I promise you that if you gave one piece of poetry, we could raise good money at a Booth charity auction. A lot of people would bid for it.
Cliff Asness: Maybe I could do a limerick. I can tell you, to the extent I’m cynical about crypto, I’m cynical squared about a sovereign wealth fund of crypto. Considering I’m cynical about sovereign wealth funds in general, and I’m cynical about crypto in general and ex Bitcoin to the extreme, you put the two together, you’ve got a witch’s brew of horrible there.
Luigi: What I want to discuss is, I read that many of your clients are interested in ESG. You wrote an interesting piece about how to consider shorting for ESG. Can you tell our listeners how the short dimension should enter into ESG strategy?
Cliff Asness: I want to back up to an earlier piece I wrote on ESG. First, I was dealing with one kind of ESG. There are a lot of different ways to do it. You can be activist and get on boards and try to change policy, but the classic long-term way of ESG is a client tells you: “Don’t own these stocks or these industries. These companies do things we don’t like, and we don’t want to put our money behind them.”
I’ve invested that way for clients since 1996. I had my first account that didn’t let us do a whole bunch of what we then called sin stocks. I won’t say the name, but some people could probably figure it out. It was a major religion that included caffeinated beverages on the list of things we’re not allowed to own, so we couldn’t own Coca-Cola.
I wrote a piece that if you give us an ESG restriction that you can’t own the following subset, you should expect us to do worse for you. People hated it, because everyone wants to be told life’s a free lunch. You could be a good person and change the world, and you’ll make more money.
Well, not to be a cynic, you know I’m a big capitalist, but if you’re doing something thinking you’re making more money, your motives are mixed. How do you actually change the world by not owning things?
Again, there are other forms of ESG. I don’t mean to be reductionist and say that’s the only way, but this is still one of the main ones, restricted lists. How do you actually change the world if you say, “I won’t own energy companies”? Well, there’s really one way. You can do, it by the way, even if you don’t think you change the world. You could do it on first principles, “I will not benefit from something I consider harmful to the world.” More power to you. That’s your ethics. You’re totally allowed to do that.
But if you’re trying to say, “How do I change the world?” The main way I can think of is changing the cost of capital to companies doing things you consider bad things. I’m going to again be reductionist and just call them the bad guys. A lot of people might disagree with that. I might disagree with that, but just call them the bad guys.
If enough of the market won’t own the bad guys, the subset of the market that is amoral, that only cares about money, has to own more than they want to own, more than they would have wanted to own. Explaining these things to Luigi makes me very paranoid, because it’s coals to Newcastle. He knows more about the stuff than I do.
But to get a market to clear, the price has to be lower. If one set of the market has to own more than it wants to because another set dumped it, to induce those people to own more than they wanted, the price has to be lower, and a lower price, all else equal, makes expected returns higher.
That makes one of the horrible things for fans of ESG to have to live with is that the amoral people should have a higher expected return, and the ESG people should have a lower expected return. It’s also the way you change the world, because one person’s expected return is another person’s cost of capital.
A lot of the people listening to your podcast will have been subjected to an MBA class, probably taught by Luigi at some point, where they plug expected cash flows and discount rates in, they do a net present value calculation, and they decide if the project should be done.
All else equal, for most projects . . . I’m saying this before Luigi does. If you have really weird time and flipping of signs of cash flows, you can have odd results. But for the standard project, where you pay up today and you get benefits in the future, a higher discount rate means you do less. Fewer projects meet your cutoff. And that’s what you want. If you think energy companies are terrible, and you raise the cost of capital on them, they do less because the cost of capital is higher.
We were nervous about me writing this, because we manage ESG money for clients. We have moral grounds we wouldn’t do, but if a client doesn’t want ESG or does, within any realm of reason, we’ll do what our client wants. We just try to be honest about the trade-offs. “We expect to make less” is not the most popular thing to tell a client, but it would be really weird, otherwise.
That went well, but then we came to shorting. It’s a pretty simple result, actually. Imagine you don’t just not own the bad guys, but you’re allowed to short the bad guys. Well, if you short something, that means the rest of the market, the amoral people, have to own even more of it. It’s the exact same thing. There’s no magic about the zero point. Not owning something is just a stopover point on the way to shorting something.
If you do the math, I actually think it’s pretty hard. I told this whole story about raising the cost of capital by not owning things. You need a lot of the market to do that at once to really raise the cost of capital. Shorting puts it on steroids. You need less of the market to do it. Also, it’s our belief that if you are an active manager, it allows some much more interesting trade-offs.
If there were two equally bad companies on ESG, but one your model thinks is going up and the other thing your model thinks is going down, you can be short $10 of the one going down and long $5 of the one going up. You’re still, on net, moving the world to a better place on ESG, but the trade-off is a lot less binding on your client. You’re giving up less in this world.
We think if you believe in this kind of thing, and many of our clients do, relaxing the shorting constraint is actually just a logical extension of ESG to begin with.
Bethany: You said on the Reason podcast recently that the way in which Trump has moved the Overton window is staggering. What are the opportunities, and what are the risks in that movement of the Overton window, both from an investing standpoint and for our world?
Cliff Asness: Well, a lot of this will depend on your politics, but in the investing world, a year ago we probably would have thought 10 percent tariffs across the board was a big, big change in the US.
We now have a market that’s net up since he announced this stuff on the relief that it might—and we still don’t know—only be 10 percent. Not only does he move the Overton window, he moves it with great rapidity. Suddenly, something that might have been beyond the pale outside that window is dead set in the middle, and we’re all going, “Ooh, thank God it’s only that.”
In my less-efficient markets paper, I talked a lot about how politics and markets are not that dissimilar. They’re both voting mechanisms. People tend to think of markets . . . I am taking your question a slightly different way, and yes, that is partially to avoid having one side or the other wanting to kill me for my political answer. But people mistakenly often think markets are a pure arbitrage mechanism. Most things are not arbitrage. Arbitrage is riskless profit. I’ve not seen many in my years. If a bunch of people are making a mistake in the market, the people taking the other side of that mistake will not take that price all the way back to fair. They’ll take it towards fair.
The intuition here is, let’s say the world is making a very big mistake, and the three of us realize it. The first hundred dollars we put in this, a small part of our portfolio is fairly low risk and fairly high expected return, because the market is wildly mispriced.
As we close that gap, we’ve moved more of our portfolio into it, meaning it’s higher risk to us with each marginal amount, and the return is less, because we’ve helped close some of that gap. As long as there’s a net error in the market, it will be in prices. This is part of why perfect efficiency is a straw man. It’s never going to happen.
I think most people would agree with the statement, “Social media . . .” And, yes, I’m back to that. I’m a near-60-year-old man who’s harping on social media. I know how cliché that is. I’m sitting here saying, “It’s the damn kids.” I feel like it’s like the end of a Scooby-Doo cartoon.
But for a market or for our politics to work, it depends a lot on this famous idea that we’ve all heard a million times, the wisdom of crowds. But the wisdom of crowds itself has a sneaky little thing it depends on. It depends on relatively independent opinions. That’s how you get a wise crowd. We might all be on average clueless, but if we all have just a little clue, if we’re all only 90 percent clueless, but that 90 percent is random, that goes away. And the 10 percent part where we have a clue doesn’t go away.
All three of us are old enough to remember Regis Philbin’s Who Wants to Be a Millionaire? All of you who aren’t ancient enough to remember this TV show, it’s a game show with multiple-choice questions where, when you hit a question you couldn’t answer, you had several cheats you could use, but they were of limited quantity.
One was to call your smartest friend. This was nearly useless. Most people seemed to pick friends in their bubble who knew the same stuff. The other one that was also very useful, obviously, was eliminate two of the answers.
I didn’t do a full empirical study, but every time I watched it, by far the best was to poll the audience. The audience had little electronic keypads. The reason it worked is obvious when you say it. Imagine only 20 percent of the audience knows the answer, but the key is the other ones are randomly distributed. Well, the 80 percent will be 20, 20, 20, 20 on the four. The 20 who know the answer will all be on the right one. So, what you’ll see with some noise is 40, 20, 20 20. Right answer. Got it.
Markets are supposed to work that way. Once we all can talk to each other, our ability to coordinate by accident . . . We’re not trying to eliminate the wisdom of crowds, but we are. I do think the most political thing I say is Donald Trump has been a master and a major beneficiary of this phenomenon.
Luigi: You criticize social media, but you are a big user of Twitter, now X. In fact, if people want to enjoy your poetry, they can see your tweets, because they are really funny and very insightful. One recent tweet that you wrote was: “We now have two economically far left and economically ignorant parties. They just differ in their preferred pronouns.”
Cliff Asness: It’s even funny when you say it.
Luigi: Even me. Believe it or not, even me.
Cliff Asness: The accent makes it funnier.
Luigi: But tell us about this idea.
Cliff Asness: OK. Well, first of all, I didn’t invent it. There’s this concept called horseshoe theory in political economics that when you go far enough left and far enough right . . . and I mean people I would consider the fringes in both parties. I don’t necessarily mean even the center of both parties, though increasingly, the center is feeling fringy to me.
You go far enough left or far enough right, they end up agreeing on a lot, because far enough left and far enough right are basically populist. They might blame different people. The left will blame the rich. The right will blame people who don’t look like the right, to put it as kindly as I possibly can, and occasionally my tribe, if I want to get really ugly, but they end up recommending a lot of the same things.
Donald Trump is a big spender who’s vowed not to cut entitlements. I’m not arguing this point. I’m not saying we should cut them, by any means. I’m just saying, he’s floated taxing the wealthy more. I’m not against that. We have a huge deficit. If you’ve got to tax someone, it probably should be the wealthy, but it was not a very traditional Republican point of view.
Tariffs, for most of my life, were a province of the left. And the New Left, that Bill Clinton left, eschewed them. The right always at least said they eschewed them. Again, you don’t need me to tell you they are no longer the province of the left. Trump raised them the first time. Biden had his chance to lower them. He did not. And then Trump, as usual, again, Spinal Tap, Trump’s amps go to 11. Even rhetoric against big pharma, big finance.
Don’t get me wrong, I don’t think big pharma or big finance is guiltless in all cases. That’s not a hill I’ll die on. But the parties are starting to sound a lot more like each other.
What happens from here? I have no idea. Twitter, pithy Twitter observations. I will tell you, Luigi, my use of Twitter and my belief that it and things like it are ruining the planet, I don’t think that’s hypocritical. It’s just like voting. My individual actions, I don’t think I can single-handedly ruin the planet. And I’m still too libertarian to say we should have rules against it.
Where I get even a little optimistic, and I don’t get optimistic too often . . . My wife, one of her favorite quips about me is: “Your glass is not half full or half empty. Your glass is bone dry.” But where I’m a little optimistic is big technological change, like, I think, this mob psychology we can get from social media, I think if it doesn’t kill us, eventually the world endogenizes it and starts to get a little more rational about it.
Maybe these things do fix themselves through the natural pull and tug. But right now, I think we’re at a pretty scary point where both markets and politics are much more susceptible to this kind of mob psychology.
Luigi: You’re very outspoken in every direction, and that’s one of the many things I like about you. But I heard you say recently that now you are starting to fear because you’re so outspoken. Can you elaborate on that?
Cliff Asness: I can repeat it. Look, this will be self-fulfilling. I’ll insult the current administration by explaining this and I’ll make it happen. It’s like not saying Voldemort’s name.
My first thing outside of finance I wrote was really a ranting screed about Obama. It was during the GFC when he said that bondholders should not exercise their rights. They should be easy on companies, because it’s good for those companies and good for the country. And I’m like: “Most of us either act like or legally are fiduciaries. We’re not allowed to do that.”
It may be good for the country. Giving to the American Cancer Society is probably a good idea, but if Luigi gives me money to manage, he has to tell me if he wants me to give it to the American Cancer Society, I don’t get to just tell him: “Hey, this is a great idea. Shut up, I gave some of your money away.”
I’ve been very bipartisan. I’m also a complainer and a whiner. Whoever’s in power, I’m going to write about. I might have joked about it. I think that Obama piece, I joked about being afraid. I really wasn’t.
I’m still not very afraid. I still believe in this country. I still think we have the rule of law. I still think I lead such a boring life that exactly what way someone would take out retribution on me is hard to figure out. By the way, I’m not challenging someone to get creative with that.
It may be just a feeling. But do I feel more nervous saying things in the current environment? And, yes, with the current administration? Yeah, I do. Yet I’m not stopping, which is really an odd choice. My partners and my wife would prefer I do less of this.
I did have someone recently say to me: “God, you share everything you’re thinking. That’s wonderful.” And I’m like: “You have no idea what I’m really thinking. It might be much worse than what I’m sharing.”
Bethany: Thank you so much. This was really fun.
Cliff Asness: I really enjoyed it. This was really one of my favorite conversations I’ve had in a long time. I can’t wait to see the final product.
Luigi: What really emerged from our conversation is that he actually seems much more a student of Richard Thaler than a student of Gene Fama, because he’s fairly behavioral in his analysis of the data. That’s perfectly fine, but I think that it’s an interesting contrast, because he did not overlap with Richard Thaler at Chicago. It seems to suggest that minds were going in that direction, anyway.
Bethany: I was really interested in this question. We talked to him about the insight versus the data and which comes first. I thought his answer on that was really, really fascinating. They are moving more toward data analysis and a little bit away from the theory.
I always think that’s a really interesting question about life. Are you guided by your insights, or are you guided by the data? And how do you not let one control the other?
It’s important for life as well as for how we think about things, because if you’re too guided by the data, you never have an insight. If you’re too guided to believe that your insights are brilliant, then you never look to make sure that it can be corroborated by the data.
Luigi: I would say, his answer is very much in line with this quantitative approach, because data and data analysis have become much better. It would be strange if they didn’t weight more on something that is better than in the past.
I think that theory made a lot of progress in the ’70s and ’80s, but I would say that we didn’t have dramatic changes in the theory of finance in recent years, while data haves become much better. I understand completely why he answered that way.
Bethany: How widely accepted is it in finance that these relationships between things are actually not linear, that there’s almost this quantum aspect to the way in which things relate, and that is making machine learning so much more relevant or applicable to the markets than he had ever thought that it might have been? Is most economic or financial theory still fairly linear in its way of understanding the world? Is what he mentioned revolutionary, or is it something that’s well accepted?
Luigi: There is a lot of linearity in finance for a very simple reason. You can’t easily make money. In fact, you should be unable to make any money by simply combining and decomposing portfolios, and that’s a linear operation.
When you sum up returns or you break up returns, that’s a linear operation. And so, you would expect that everything that matters in finance should be somehow linear, because if it weren’t, it would be easy to make money just by combining and breaking down portfolios.
That said, I think that this is an influence of machine learning, because the technique to capture these nonlinearities have become much, much better. And so, people now pay more attention. In the past, the analysis of nonlinear data was more complicated, but most importantly, there wasn’t a structure to analyze all the possible permutations of nonlinear combinations. Now machine learning is basically providing a systematic way to do so.
Bethany: Is it changing the way people do papers or the way people think about unproven aspects or things that they want to write about in finance?
Luigi: I think it has started to. I saw a beautiful presentation at the American Economic Association by Sendhil Mullainathan. Remember, we had him on our podcast. He was trying to explain how you should take machine learning and, in general, AI as a way to gain more insights into what we’re doing.
The simple, and if you want, more mechanical way is how you can do better than a human on certain tasks. Take, for example, recognizing X-rays. Now, machines do better than humans in recognizing problems in X-rays. But the interesting insight is why and how machines are doing better than humans, so that humans can learn how to capture that aspect that was basically unnoticed before.
In a sense, that’s exactly what Cliff was saying about asset pricing. They are trying to see some irregularities that they didn’t see before, but the machine is better able at capturing.
Bethany: I was actually so happy to hear him talk about the unknowability of some aspects of finance or the unprovability of them. I remember one of the most depressing things or one of the most unmooring things to me was when I started at Goldman Sachs in 1992 after majoring in math.
When you were inputting the discount rate into your model, you would just put in whatever number made the outcome come out the way you wanted it to be. Of course, I worked in an M&A department, so when we were running these models, usually somebody, the senior banker, had an answer in mind.
But I remember thinking: “Isn’t there some fundamentally correct way to come up with what this number should be? Isn’t there some way that will be actually scientifically and mathematically completely sound?” I remember being very upset that if there was, I couldn’t figure it out and couldn’t understand it. So, I feel like one of the central questions of my youth has now been addressed.
Luigi: It’s one of the central questions in finance, and it’s still not fully addressed. So, you were aiming high.
Bethany: Well, I think I was always told in those years that it was just knowable. It was just one of those things that you were supposed to understand. I feel, all these years later, some measure of justification or something that I was not able to understand it.
Luigi, I really liked your question invoking Lloyd Blankfein’s idea of, “I’m doing God’s work,” and this parallel you made between the market and finding efficient market prices as somehow being akin to God’s work. You know, in a way, it sort of is, for all that we on this podcast, and for all that the current vogue right now is to be critical of the market, or critical of how . . . Being bottom-line driven and believing in the market are not the same thing.
But still, I thought, “Wow, what would it be like if the market is no longer a mechanism of pricing?” If you can’t look to the market as a gauge of pricing, if it just becomes this animal that is completely detached from any kind of efficient market theory, if the inefficiencies become the rule rather than the efficiencies, we might all long for that old market that we love to despise again, if that actually is what starts to happen.
Luigi: That’s for sure. But also, think about it, what is the alternative? If you don’t go for some decision with market prices, you go with political decisions. The alternative looks worse every day. I think that it is a relative statement.
Here, I think that Cliff was very honest. We are not looking at perfection because nothing is perfect in this world. We are looking at less imperfection. I have to say, with all the inefficiency of the market, the market looks much less imperfect these days than the alternative.
Bethany: Has there ever been a third mechanism? Forgive my ignorance of economic theory or economic history, but has there ever been another method for setting prices? Has it always been either the market or government control, or has there ever been a third way or an exploration of any other means that there might be in a society to set prices?
Luigi: You can make decisions on expertise if you don’t like the political system, but at the end of the day, you need to empower somebody to make the decision. The politician might delegate that to an expert, but he or she picks the expert. At the end of the day, it is a political decision that can be reverted if you don’t like the expert.
Bethany: It’s funny. It definitely made me think that I may be perhaps too cynical of our market-driven environment, but if I really had to think about a world where the market actually isn’t any kind of gauge either in the short term or in the long term of prices, and we’re just reliant on authority figures, the experts, to tell us how things should be priced, I’ll take the market any day.
Luigi: OK, so you prefer the market over me? Of course.
Bethany: Yep.