US Videos

Asness: Splitting the Middle on Market Efficiency

Scott Burns

Scott Burns: Are markets efficient? Hi there, I'm Scott Burns, Morningstar's global director of manager research, coming to you from the Morningstar Investment Conference. Joining me today is AQR's Cliff Asness. Cliff, thanks for joining us.

Cliff Asness: Thank you.

Burns: Now, recently the Nobel Committee handed out their prize for economics to Bob Shiller, Eugene Fama and a third person who you are going to bring up for me.

Asness: Lars Hansen.

Burns: Lars Hansen. Why does everybody forget Lars?

Asness: Because he is not fighting with the other two.

Burns: That's right. That's right. With the prize, you couldn't have two different camps. Bob is saying the markets aren't efficient; Eugene, of course, is on the efficient market [hypothesis]. You want to split the middle. How do we resolve this?

Asness: You've known me for a long time, and as I joked at the Conference, when I see a fire, I'm more often a gasoline thrower than a water thrower. I like to be in the extremes, if warranted. Here, actually I think the middle is the right place to be.

I think if you have one extreme that markets are very close to efficient, even Gene Fama will tell you they are not going to be perfectly efficient. He likes to shock the class by telling them that.

Burns: Does he know that you are telling people this? You know we're on camera.

Asness: He's said it in class for like 30 years, and he's written it. Gene, I'm revealing your secrets. Perfect efficiency as Gene would say is a very extreme hypothesis. But clearly, I think it's fair to say he thinks they are a lot closer than Bob does. Gene will not use the word bubble; Bob will use the word bubble a lot. There really are some big differences of where on the spectrum of "how efficient?" I think the Nobel Committee got it right because I think the efficient market hypothesis as created by Gene Fama was pretty much the most important contribution to modern financial economics. It is modern financial economics. We had no starting point. We had no, to be a geek for second, null hypothesis before Gene gave us one. We casually assumed markets were wildly inefficient. Index funds were un-American. Treasurers could issue stock or split stock or do anything. Obviously, they had great skill at doing that. Gene said, hey, wait a second, and he gave us his null hypothesis. 

I think if you go through time, my judgment of the evidence is both things academics love to study. They often call they anomalies; things that are hard to explain. Why cheap beats expensive; why good momentum beats bad momentum; why low risk seems to outperform high risk, especially when you take account for that risk, and other things. You get into both camps on these. There is the Fama efficient markets camp. I'm personifying with these two men maybe a bit too much, but let me call the Fama camp that will say, "These things work because it's risk." And the Shiller camp would say, "No, the market gets it wrong. These things work because the market makes errors."

When you step back, one thing complicated about the world and makes for very complicated model, is both can have a great degree of truth and which one is a more important factor can change through time. The real world can be much more complicated and kind of binary. One person is right, and the other is right. For instance, value investing, cheap beating expensive. I come out like I do on a lot of this, somewhere in the middle. I think both the risk stories and the inefficient-market-era stories have good logic to them, and the effect is really ridiculously strong when you add up all the evidence that we have of looking everywhere for all periods of time. It doesn't mean it's the same all the time. During, for instance, the period I'm kind of obsessed, the tech bubble, I think much of the difference in valuation was driven by irrationality. I think that was exception that maybe proves the rule, but I will use that world, bubble, about the tech bubble. I won't use it very often, putting me firmly in between.

Read Full Transcript

Burns: I thought it was great when during your session, you said, "An efficient market person would never call it a 'tech bubble.'" What would they have called that?

Asness: I'm going to have trouble getting it right myself. The tech stocks again, I mentioned in the conference, it's more than just tech, but for shorthand let's call it tech. Tech stocks are very expensive and offering a low expected return because they are currently less risky than other stocks, and therefore, I rationally choose to buy them to accept this lower expected return.

I happen to think that might often be the case. I don't think that was the case during the tech bubble. I don't think people were buying dot-coms and Cisco. And again I mentioned this in the conference, but expensive was beating cheap within every industry, not just technology. I don't think people were doing this because they thought it was less risky. I think they were doing it because they thought they were going to triple their money out.

Burns: I really like the way you presented that. You've got the Fama side, the efficient market hypothesis is the null set, and then folks like Shiller come out. Is it the rejection of the null, or is it the proof that the null works most of the time?

Asness: You could phrase it either way. It depends. It's a half full/half empty kind of way to view it. If you believe the real world is such that markets work very well, I certainly think they work better than whatever alternatives mankind has ever come up with to markets. But they work real well, but are not perfect. As Gene will start out saying it then becomes an empirical and logical question. You can't just look at data. But it becomes a question of how inefficient are they?

Once you accept that they are somewhat inefficient, that inefficiency can vary through time. Is it ever inefficient enough where the word bubble is warranted? I don't think something that works a little bit more than average and can generate some return over time would warrant the word bubble. I don't think a market that's a little bit more expensive than normal would warrant the word bubble. I do think--I assume and I call it the tech bubble--I think the tech bubble was a bubble. But I do think if you look at industry, if you look even at academics who are very much on that side, they use the word "bubble" to mean, "We kind of don't like that. It's not as attractive as, as normal." And you devalue the word to me. So you used it--the exception that proves the rule. The tech bubble and maybe a couple other episodes, I will use that word. I will use it less than--I don't know about Bob Shiller personally--but less than the inefficient market camp, far less. Far less than a typical stock-picker who thinks anything he doesn't like is in a bubble. Gene won't use the word. He'll allow it to be used in his presence; he won't believe they occur. And I do think they occur seldom, so I am between the two men.

Burns: This kind of down-the-middle-of-the-road path, I think one of the interesting things that that you bring up about that is that a lot of other things then start to matter that in either of those camps aren't necessarily that important. You mentioned regulation, can you talk about that a little bit?

Asness: I like to focus on accounting as my favorite example. If you are a strong believer in efficient markets, in some sense it doesn't matter. Our accounting standards, and my favorite example--again its tech bubble obsessed--is the fight after the tech bubble about expensing stock options. I was involved in that fight; I think it was very clear that they should be expensed, that when you hand someone something worth something, it's an expense. But there was a big fight at the time. I expected a lot of criticism from people who have interest in the other side, people who disagree.

I got some flak from some friends who are big believers in efficient markets, who were kind of, of the view, "Yeah, you're probably right, but who cares. The market will see through it."

If you accept that we're in the middle of Bob and Gene's world, that markets are darn good and they price things well over the long-term, but they're not perfect and can occasionally become very imperfect, every impediment you put up, you don't know what matters. You don't know what the tipping point to an irrational market is. You don't know what causes an error in the price. I like to say it's a geeky comment that only accounting and finance people would like: "Why make the market look in the footnotes? If we know the right answer, put the right answer up top and put the wrong answer in the footnotes." So I think everything matters once you accept markets aren't perfect.

Burns: Right, and then that's just so clear. For a market to be efficient, there has to be a lot of transparency. You have to be able to see these things. I mean, that's almost like definitionally required.

Asness: That's what I believe. Gene would accept this, too. There are limits to arbitrage. He and Ken [French] wrote a great paper on tastes in agreement showing that if people make errors, unless people happen to make the exact opposite errors, it will influence prices. So whatever you can do to reduce those errors, you want to do them. It matters.

Burns: Your take, I'm going to borrow from your terms, was the Nobel Committee wimps, did they wimp out, or how do you see it?

Asness: I think in this case they were more Solomonlike than wimps, unless of course you find Solomon to be one of the wimpier biblical figures.

I think [the Nobel Committee] did the right thing. I think the efficient market hypothesis as created by Gene is the single most important part of finance. We'd be nowhere without it. Bob is one of the most prominent critics of it and I think truth does lie in between them. So giving it to both of them and to Lars I think that was a pretty reasonable act.

Burns: Well, Cliff, thank you very much for joining us.

Asness: Thank you. It's been great.

Burns: Thanks for watching. I'm Scott Burns with Morningstar.