He Wrote the Papers Finance Didn't Want to Hear | Cliff Asness Breaks Down His Greatest Hits

You have to be a crazy person to say you are certain of something. I just think the idea of a permanent loss of capital, yeah, if you're certain it's going to come back, it's not risky. Risk control to me is how bad is it if you are wrong or how bad is it if the market doesn't recognize you're right for a long time. There is no cash on the sidelines. When someone says there's a lot of cash on the sidelines, I always ask, 'All right, well, when that cash gets spent on stocks, they have to buy them from someone, right?' A lot of these have like the S&P 500 sitting out at 17% annual vol, privates 5%. And that would just send steam coming out, and when you have no hair on your head, the steam is really obvious. We all know the prices are moving. You're just not observing them.

I'm good. You're already getting us in trouble because most artists want to do the new stuff. And then they get the audience all mad. But I will acquiesce and do the greatest hits. I saw the Stones at Jazzfest two or three years ago and I think they were over 80 at the time. And they were still jumping around the stage. I don't know if they put them in some hyperbaric chamber when they're not performing, but it was incredible. Like if I'm doing that when I'm 80, I win.

No, no, it was really good.

Oh, well then you're definitely going to play the hits.

How to learn to stop worrying and love the bull. Great movie reference. We can't have enough Dr. Strangelove references. Just inject them straight into all.

I'll take it. I have to admit, Matt, I didn't get it. So, you're doing better than I am, Jack. There's no fighting in the war room.

Well, again, this is ancient history for a lot of your listeners. For me, our firm, we launched one month before the great dot-com bubble of '99-2000 really hit its explosive stage. It was an unpleasant year and a half for us. Value is less of what we do these days but still an important part, but it was more of the process back then. And to say probably what's obvious, it was not a very fun period to be a value investor. Momentum certainly helped but not nearly as much as value hurt. So we were hurting and in this book draft I started out like first line is, 'Okay, I'm a money manager currently losing.' By the way, the story worked out quite well for round trip. I just want to take the suspense away. But so I admitted that you always should say, you know, almost every paper is some degree of self-serving to some degree. So I like to get that stuff out there. But it was first, I should never have made it a book. I was trying to write a book. And the first draft was like the end of '99, right near the peak. And I kept revising it. And the last draft I did was June of 2000 when it was still crazy, but it had started to come down. And then I'm like, well, it's coming down and down and down and we're making a lot of money as a firm now. So, do I really need to write this book? So, it's permanently a book draft. The goal was both to make the argument that both the level of the overall market and I stressed we're not any better at market timing than everyone else. We're talking the 5 to 10 year perspective returns looked very low, not the next month's perspective returns. That is when we famously hit, depending on your version, a Shiller CAPE of like 45. But even more so intra the stock market. I don't like calling things value and growth. I'd prefer overpriced and underpriced or something like that. Because the opposite of cheap is not growing. But that's the vernacular, certainly was at the time. Growth was just destroying value. We've seen other periods like that since then but this was it probably, you know, COVID gave it a run for its money. Those were the two periods where I've seen the most extreme version of this but this was the most extreme deviation in those two that we had ever seen in 50 years of data. So part of it was kind of a more mundane, can we justify this? You know, if you try a million ways to justify it and fail, you haven't proven you're right. It's hard to disprove a negative, but it's the best you can do, right? Eventually, you take a bet. The bubble logic concept came from me like trying to get in the heads of what people were thinking. How would they argue for their, I thought patently in some cases insane positions. So I'll just give you a couple examples. The whole thing is chock full of them. One was I made fun of the book which only the old on this podcast will remember, Dow 36,000. This was a perfect foil for me because I'm talking about bubble logic. This is a book saying the Dow should be many times higher than it is today. When it was already super high compared to the fundamentals. By the way, if anyone out there is saying, well, the Dow hit 36,000 eventually, that's not the freaking point. If it hits it 25 years later and you made no money between now and then, it's not. So there's always some smart alec who says that. But here was the logic that was kind of fun to dissect of this book. The logic was this: over long-term horizons, call them at least 10, but probably more like 20 years, stocks have never lost to bonds. That's actually mostly true. You could probably, are they real returns? Whatever. But it's mostly true in the US. And it's not as true outside the US. But let's just take it as a given. Stocks have beaten bonds all the time. Therefore, for a long-term investor, which we all should really be, they're not riskier. If you never lose, how are you riskier? Therefore, they shouldn't have to offer an expected return premium. You shouldn't need more in a rational market. The only reason you should, and we could talk about irrational markets later, but in a rational market, the reason you get paid a higher expected return is you're taking on some risk. But if I just prove these things were riskless to the relevant investors, why don't stocks and bonds offer the same return? Now, this is idiocy. And I am mincing words, believe it or not. Because it's extremely circular logic. Imagine you got to a world where they were priced to the same expected long-term return. So you didn't expect to make any more in your stocks. Stocks are still three to four times more volatile at pretty much any horizon. Even at 10 and 20 year horizons, they have much wider dispersions of outcome. The reason they've won over time is they've had a higher average return. The dispersions around that average are still much higher at any horizon. So if you cut them down to no advantage in return, suddenly you have something with the same end point but much more dispersion in where it may end up. That's called risk. And suddenly you go, you know, I need a risk premium to own these things. And you're right back where you started from. This bestseller at the time. My other example of somewhat circular logic, not exactly circular but similar was I think the statute of limitations on this is out. I wanted, I was watching the message boards. The message boards was like proto social media. It was text-based. It was not as cool looking as today. Frankly I thought the interface was better than Facebook but that's an entirely different story.

I was Yahoo. I was going on the NDX 100 and the Cisco message board a lot, but I was just a lurker because we have all these rules against posting. I can't post like, 'You should short this stock that we're short.' I asked compliance, can I just pose this question? And they said yes because it's a very innocuous question. It was on the Cisco message board. Cisco was the internet pick-and-shovel company that was selling, that was one of the biggest companies in the world selling for 100 PE on very high earnings. So this is not a venture capital low E. You know, some you could have a super high PE if the E is non-existent. That's just a bet on the future. But for huge E, the E is for earnings in case anyone's not following along. So I posted a question. Everybody on this message board is bullish, right? They're all rah-rah. I think it's going to go to this. Oh, you're a wuss. I think it's going to go to this. I posted a question that I don't even know if I intended to do this, but I framed as a value investor. I said, 'All right, everyone loves Cisco on this board. I get it. At what price would you be a seller?' And I'm looking for the concept that no matter how much you like it, but you may like it a hundred times more than me. You might disagree with me. I think it's wildly overvalued right now. You might disagree. At what point do you agree? If it's a thousand times more is it? And I got no answers in the direction that I wanted of a higher price. But I had multiple people saying, 'Oh, if it was down 25%, I'd be out.' So I cracked up. It was bittersweet because we were losing money at the time, but I was cracking up that I posed what I thought was a really cool value question to get people to get their minds reframed. You know, once you accept that there is a price you wouldn't want to own something for the long term. And I think I did say for the long term, so it's not a trade. Then you start thinking what should that price be? And you need a framework. And I thought it would be helpful. Didn't help at all. It was only, so there were tons of things like this. Wall Street analysts did not bring honor upon themselves at the time. I was very sarcastic about Wall Street. I said like cars, they sell new and used stocks. And would you go to General Motors to get their strategist's opinion of which car you should buy? I don't think so. So it was both a tough time for us, but ultimately it worked out very well and I had a lot of fun. This was also, I think that bubble kind of broke something in me where it was the first thing I wrote in something that I've done again and again where I tried to inject some humor and sarcasm and fun into it. Everything till then I was trying to be a Fama clone. I was a pretty good clone actually. But it was dry and right down the middle. That was the one where I'm like, all right this is insane. I'm going to have some fun with this. So not everything, I still write plenty of things that are dry. But I've gone bubble logic on a bunch of things since then.

I was very clear with compliance. They were like, 'Post it.' No responses. None of you know my handle, so you'll never be able to find it. Actually, now that I gave you the, it's probably, everything, nothing dies on the internet, right? You probably could find it, but I was good. I followed the rules.

Yeah, that one I'm going to be disappointing. I'm not going to give you a great answer on this one because let me tell you the truth behind Rubble Logic. You may notice it rhymes. I was very bitter that I never published Bubble Logic. So I took all the things I did in Bubble Logic and didn't pretend to be predicting anything because by 2005 when I wrote this the bubble had collapsed. But I said, have people learned these lessons? And I just went through the same stuff of Bubble Logic and I think I posted some cynical stuff like I don't think they fully learned these lessons with some examples. But there was hardly a shred of originality to Rubble Logic. It was a thinly veiled attempt to get some credit for a piece I loved and never wrote. With some success, I might add.

Yeah, the Russian. Then I would have had a write in 1998 in August of '98 when Russia defaulted. That would have been a really good title.

You know, one day they'll default again and I'll get to use that title. So...

Sure.

You know, a lot of people make the analogy, everyone makes is to the same dot-com bubble where we're talking about, and the whole history doesn't rhyme or does rhyme but it doesn't repeat is kind of true but how much it rhymes is important. I don't think this is anything like the GFC. I think the GFC was, there was one hidden terrible assumption in everyone's model that real estate can't go down. You know, when you have a lower bound of zero, you get some odd results. And then there was a derivative structure that was just unwise. I can let the politicians and the professors debate that. But that's even, people make some private credit analogies now. I don't think it's close to a GFC-like problem. But in the dot-com bubble we saw Shiller CAPEs of about 45. We see it close to 40 now. Which would be the second most in history. The spread between cheap and expensive, something called the value spread, which we first wrote about in 1999, was higher than the dot-com bubble at one point. It actually, I never thought I'd see that in my career. And we did. And it took COVID to get you there. It was getting there before COVID and then COVID sent it into the stratosphere. Expensive stuff went to the roof. And you remember the six months after COVID you were only supposed to own Peloton. That was your diversified portfolio. Maybe Peloton and Tesla if you wanted to be a diversified portfolio. You can drive or you can bike in place. But the dot-com bubble, it was a tech bubble and at least that's what people are worried about now. I think it certainly does rhyme but the value spreads now are about 75th percentile. And I should be careful. The way we do it like a lot of quants do is we try very hard not to take an industry bet. So I don't track that so well. So I am not making a strong statement that AI versus everything else is fair. But on our versions of value it's good. It's better than normal and a lot of things in the world are less attractive than normal. But I wouldn't make that as a super strong analogy. I would defend the analogy in one way. All the time you hear people say, 'Yeah, but back then they were all dot-coms and worthless companies who made no money.' That's just not true. I saw a post by I think it was Andy Constant.

The actual dot-coms were insane. They were stupid. There was no chance that most of them weren't going to zero. But they were a sideshow. They didn't make up most of the index. If you look at the S&P 500, the NASDAQ 100, they were more like firms like Cisco, which made a ton of money and was still selling for a ridiculous price. So what I can't tell you, and you should never come to a quant for one grand thesis, is AI worth it? That's not a quant thing. Quants want to make a thousand longs and a thousand shorts and play the odds. So it could still turn out to be worth it, but it does look more similar just on the numbers. Just because I reject some of these arguments that the numbers don't apply. So the whole market looks almost as expensive. The spread between cheap and expensive, I imagine if you did it not like us, but allowed industry bets, you'd see something a little more comparable. So in a lot of ways I think the market is similar but something like will AI or back then will the internet pay off, that's a single bet and again a quant doesn't do that. And just it may be 19 out of 20 times when you see crazy prices you haven't seen before you want to bet against that but 19 out of 20, this will be the entire math we'll have on this podcast is not 20 out of 20. So I am not making any heroic bets. We won't do this at AQR, but shorting the market, shorting AI stocks, but call me in the middle. I think the market rhymes with back then and the outcome will, if I had to guess, it'll be at least directionally similar. But it doesn't have to be. This is not an arbitrage.

It's always different. We have this graph and I, we've moved so far beyond this. This is Fama and French book-to-market from 1990. And I love Fama and French. And by the way, I'd still bet this will make money in the next hundred years. But I don't want anyone to think this is our current model. I get very paranoid that it, about that when you say, can we imagine this doing well? No one can imagine this doing well at this point. When meanwhile famously it's lost for like 30 years or at least been flat. It's been a great disappointment since the original papers came out. Like 80% of that is from the fact that growth got more expensive against value. It was a capital loss, not a fundamental loss. Adjusted for that, which you can't do in real life, by the way. If you bet on this thing, you actually did lose the money. But when it comes to estimating what the true long-term return is, I don't think you ever want to put in a valuation change because if you put that in you're baking in, it goes on forever. And we've learned valuation changes can go on for a long time, but forever is a really long time. So when you redo that, I wrote a piece on our website called 'The Long Run Is Lying to You' that the value premium has actually been about two-thirds as good in the out-of-sample period as the in-sample period. So these things are hard. That's a long time, that's 30 years. That's about since the tech bubble broke. I'm not saying that I'm any better than anyone else at getting someone to stick with a losing strategy for 30 years. I find about two, two and a half years is as long as an investor, most of your investors can take. I have never experienced longer than that thankfully but I have experienced that and it's very hard, it's very hard even when you've made a lot of money for many years and have a great story for why things are messed up and will come back. So I can't imagine 30 years in. Can you imagine the review, we've lost for 29 and a half years, but we really feel... No, can't happen. I totally forgot your original question. I talked too long.

Thank you. I had fun with it.

Well, I don't know if you guys know me well enough to know that no peeves roll off. They simply accumulate. I am open to being proven wrong, and I won't share any of those with you on this podcast, but believe it or not, in my career, I have been wrong about things. But I did glance through them and some of them are less relevant because some of them were more topical about stuff going on at the time. But there are none I would like, there's one where I'm talking about the GFC and who's to blame and I kind of divide it up and I think people tell a political story where the left wants it to be the banks and the right wants it to be the government subsidizing housing. And the truth is a little of both. The government subsidizing housing is partly what got real estate going. I'm not sure you can even say real estate was in a bubble at the time. It's way higher now, but clearly it was screaming up for a while. And that led to some silliness and then Wall Street with the help of some deregulation that was in the '90s created this. So that was one that I think people just bored talking about. It's a long time ago. One was defending high-frequency trading. I think that ship has sailed. It's just part of the firmament now. But some other ones, a perennial one was I was defending volatility as a risk measure. And all the time you hear people kind of diss volatility as some silly quant short-term thing. Volatility is not perfect. A, you have to estimate it and estimates can always be wrong. B, if you're trying to estimate either an asymmetric security, like an option-like security, symmetric volatility doesn't make a lot of sense. And like any risk system looking backwards, it could always be an event that's just never happened in history that happens. And markets are not that well behaved. But the question is whether it's useful. And the critique of it though particularly annoys me because every single person who critiques this, including some super smart people, says the risk isn't volatility. The risk is the permanent loss of capital. To which I always respond, do you delete expletives on this?

I always respond, no. You swear to me that this loss of capital is not permanent. I agree. It's not risk. Can you swear that to me? I've lost money and easily made it back over time, but it's been painful. And it's been painful mostly because you have to deal with the world thinking you're an idiot. But also, as confident as I've been at these times, only a maniac is 100% confident. I don't care what I write in a bubble logic or if I write something around COVID or about the value spreads. I damn well think I'm right and I'm betting a lot of my own money that I'm right. But a crazy person, you have to be a crazy person to say you are certain of something. So I just think the idea of a permanent loss of capital, yeah, if you're certain it's going to come back, it's not risky. Otherwise, is volatility useful? Well, for most securities, again, options might be, for most securities, volatility, if you're short-term measured with daily or monthly numbers are very high. And then you look at five-year or 10-year outcomes. Some narrow a bit because of mean reversions. Some don't, but they're a more volatile security on monthly vol also will generally have a much wider dispersion of five and 10-year returns. That is a useful thing for me to know in building a portfolio. I make other, I think I included this in the thing, permanent loss of capital. I, did I get to say this will come up again if we talk about privates, but when we were getting killed in the dot-com bubble, did I get to tell my investors, 'Yeah, we're down a ton for the risk we're taking, but it's not permanent and we're going to be right, so please leave me alone.' You know, jump cut. They did not leave me alone. So permanent loss of capital is to me correct and a fairly vacuous statement. Another way I don't like it, I'm sorry I'm going on and on about this is it also says the method you do risk control, the way you do risk control is not being wrong. Because that's what avoiding a permanent loss of capital. It says you have to really not be long-term wrong. That is nice work if you can get it and we all try to get it. We all try not to be wrong, but risk control to me is how bad is it if you are wrong or how bad is it if the market doesn't recognize you're right for a long time. Both those are two versions of risk control. A risk control is a separate part of the firm even if we all work together. There's what do we think's going to happen and then how risky is it should not be the same question. So I had fun defending that. I led with that. I think that was number one. I convince no one. By the way, every one of these peeves, it's not like you don't see them anymore because I spoke. I always have this fantasy I will write something so convincing that suddenly it'll all change and people like, well, we don't say that anymore. It's rare. It's rare.

There is no cash on the sidelines. When someone says there's a lot of cash on the sidelines, I always ask, 'All right, well, when that cash gets spent on stocks, they have to buy them from someone, right?' Like, yeah, that person now has less stock and more cash. We have preserved the amount of cash in the system. There is no such thing. You can try to create sentiment indicators like if I think there's a lot of money in a particular kind of short-term instrument that tends to be a staging area for stocks. I'm making this up. It's not something we trade, tends to be a staging area for stocks. Maybe you're making a behavioral that when money's here, but the notion that there's cash on the sidelines, it's only true in the long term. Issuance and buybacks can change the total amount of stock. And if you change the total amount of stock, you also change what's on the sidelines. If you buy back shares, there's less stock and there's more cash out there. But that is a much slower phenomenon. That is not what people mean when they say this. I'm giving them too much slack. I always try to make the best argument for the other side. Best argument is that the long-term there are sidelines, but that's not how people use it. They use it as a short-term market timing tool. And I just love to just go, yeah, it's a little bit of a, from The Matrix, there is no spoon. Again, I'm dating myself. There are no sidelines. I like to be very Morpheus about that. It's a good thing to be very Morpheus about.

Oh, okay. Well, somebody's going to tell us in the comments.