Eugene Fama — For Whom is the Market Efficient? (#161)

Eugene Fama is a 2013 Nobel laureate in economic sciences, and is widely recognised as the "father of modern finance." He is currently the Robert R. McCormick Distinguished Service Professor of Finance at the University of Chicago.

As of 2024, the Research Papers in Economics project ranks him as the 10th-most influential economist of all time.

Sponsors

This episode is sponsored by GiveWell. GiveWell is an independent nonprofit focused on helping people do as much good as possible with their donations.

GiveWell conducts in-depth research to determine how much good a given charity program accomplishes (in terms of lives saved, lives improved, etc.) per dollar spent, and publishes the full details of their analysis, for free, to help donors decide where to give.

If you’ve never used GiveWell to donate, you can have your donation matched up to one hundred dollars before the end of the year or as long as matching funds last.

To claim your match, go to GiveWell's website, pick "Podcast" and enter "The Joe Walker Podcast" at checkout. To donate or find out more, go here.

Transcript

JOSEPH WALKER: Eugene Fama, welcome to the podcast. 

EUGENE FAMA: Thank you.

WALKER: Gene, I was talking with a few friends who work in high finance in preparation for this conversation. And one of my impressions is that a lot of people think of you as holding this extreme position that markets are perfectly rational. But I know that you don't believe that, and I've also heard people who've taken your classes at Chicago say that you repeat ad nauseam that models aren't real and the question is really: how efficient are markets?

FAMA: There’s a different way of putting it, actually: who is it efficient for? That's another way to put it.

WALKER: Can you elaborate on that?

FAMA: Well, for almost everybody, the market is efficient in the sense that they don't have information that's not already built into prices. People who have special information, the market's not efficient for them. So let's say insiders, for example, typically have special information. So as far as they're concerned, this stock is not priced totally efficiently because they have information they know will change the price. But for everybody else, assuming it's efficient, it may be a really good approximation.

WALKER: So does that mean you think that the semi-strong version of the efficient markets hypothesis is the most plausible version?

FAMA: No, I don't think those words really make any sense anymore. I invented those words 100 years ago. That was describing the nature of the tests that people were doing, not the reality of what the market is like. It's just what kind of tests are likely to expose whatever shortcomings the efficient market hypothesis has.

And there are some that are not strong enough to expose. So that would be what I used to call weak-form tests. And semi-strong would be a little bit better. And then a strong form would be the way I described it: when you find people who can beat the market.

WALKER: Let me ask you then, if you had to quantify in some very crude way where you sit on the continuum between a market that's perfectly efficient for everyone and a market that's perfectly inefficient for everyone, how close are you to the perfectly-efficient-for-everyone end of the spectrum? 95% of the way there? 80%?

FAMA: I can't put a number on it because I'm a data person. So that would require data that's not sufficient to answer that question. If you give me a group of people, I'll have a guess. 

So if you say, tell me about professional investors, I'll say a very small fraction of them show evidence of having information that isn't already built into the price. So that's going to the top of the food chain, saying even among the professionals, there are very few that have information that isn't already in the price. 

If I go out to the public, alright, the market's efficient for everybody out there.

WALKER: So rather than asking you to define the efficient markets hypothesis for probably the umpteenth time in your life, perhaps I could give a four-point summary, not just of it, but its implications. And then you can grade me on it if you like. 

FAMA: Okay.

WALKER: So, number one, the efficient markets hypothesis is simply the claim that prices fully reflect available information. 

FAMA: Correct.

WALKER: Number two, that also means that prices should roughly follow a random walk?

FAMA: False. 

So the problem is, this is what I call the joint hypothesis problem. You can't tell me that prices reflect all available information unless you take a stance on what the price should be. So you have to have some model that tells me, for example, what is risk and what's the relation between risk and expected return. And then we can look at deviations from that and see if the market is efficient. So there is what I call this joint hypothesis problem. You need a model that tells you how prices get formed. So in the jargon that's called a model of market equilibrium. You need to join that with efficiency, then I can test it in the context of whatever model you tell me is determining prices.

WALKER: I feel like that's perhaps the deepest insight of your work, the joint hypothesis problem. 

FAMA: Right.

WALKER: I do want to come back to it because I want to talk about asset pricing models. But that was actually going to be my third and fourth points in my summary. So the two key implications of the efficient markets hypothesis are, to borrow Dick Thaler's summary, that the price is right and that there's no free lunch. And then I guess relatedly and finally...

FAMA: You have to tell me what that means. That's the problem, right?

WALKER: Yeah, exactly. That's where the joint hypothesis problem comes in. Okay, anything else you would add or change about that summary?

FAMA: No.

WALKER: So, I'd like to talk about some specific anecdotes and then just get your interpretation of them. And of course, these are just anecdotes, but it still might be fun to hear how you think about them. 

FAMA: Alright.

WALKER: So first is GameStop.

FAMA: I didn't follow that very carefully, so tell me what happened.

WALKER: Yeah, just in brief, at one point the stock price was about $3. About a week later it was $100. One of those prices was probably massively wrong. And I think it's almost certainly the high one.

FAMA: Where is it now?

WALKER: Good question. Maybe I should check. But I think it's come back down.

FAMA: Okay, well, there's no doubt that a price can get out of line, that the market for that stock can become inefficient if enough people pile in on it. So finance is just a branch of economics. Every branch of economics says supply and demand determine prices. So if demand gets high, the price is going to go up and maybe that will last, maybe it won't. But now in this case, if people are just doing it because other people are doing it, then it's going to bail out eventually.

So those things happen. But they're exceptions. They're not the rule. It's not everybody's investing that you're talking about there. It's not big things either. You're dealing with a little company in that case. So you can distort prices with enough demand, or you can distort them the other way by not having enough demand.

WALKER: I'd like to talk about some potentially bigger things a little bit later, such as the US housing market, but we can come to that.

FAMA: Yeah, I'm working on that right now, actually.

WALKER: Oh, fascinating. Okay. Alright, well, I'd definitely like to hear about that. But a few more anecdotes first. So how do you interpret the success of outlier investors or firms? So people with incredible track records, like for example, Renaissance and Jim Simons. Are they examples of survivorship bias or are they exceptions to the rule of the efficient markets hypothesis?

FAMA: Suppose we take 10,000 tosses of a coin—and I'm telling you we're going to do it 10,000 times and the probability of a head is 0.5, and I'm going to say, how many heads can you get in a row? Well, if you do it 10,000 times, you're going to get a lot of runs of heads and tails. So you're going to get a lot of big winners and a lot of big losers. So my message is, be careful; after the fact, what looks like good performance could just be luck.

So when people have studied this, what they've found is if you look at the winners of the past and then you follow them in the future, they don't look like winners anymore. Most of them just look random after you anoint them. And that's pretty typical. But of course, even some fraction of those will continue to do well solely by chance.

So the problem is 2020 hindsight: it doesn't work. You have to identify before the fact. Or you can do as Ken French and I did in that paper. We said, well, given the game that's played, what fraction of people would you expect to win by chance? And if you look at it (I always chuckle when there's bad news coming),  if you look at the cast of mutual fund managers, what you find is before fees and expenses, in other words, not returns to investors, but just returns on their portfolio before you take out the expenses that they usually take back themselves, well, then what you see is there's a very small fraction that you can't explain by chance. 

So there are some people out there that do have special information. If you take away fees and expenses, though... it's a terrible game for investors. They lose. 

Now even in that game, there are winners. There are people that do better than their fees and expenses. But you expect lots of those by chance, and there is a much smaller number than you get if you look before fees and expenses. 

But chance alone will produce such results. That's the problem. So when we anoint people, most of the time we're anointing them based on chance. They were just lucky.

WALKER: Is it less likely that they were just lucky if they also have a compelling causal theory as to how they were successful? So say someone like George Soros comes in with reflexivity, or Michael Burry has his theory of what's going to happen to the US housing market and all of that before it actually happens—if there's a compelling causal theory, does that increase the likelihood that it's actually due to skill rather than luck?

FAMA: Well, you have to do the test. If you do the test after the fact, there's always a causal theory that somebody will come up with. But you have to do it looking forward. You have to tell me your causal theory and then we'll follow it, and we'll see how it does. And if it works on a better than chance basis, fine. I'm not one that says these things can't happen. They can happen, but that's the way you have to test them. You have to test them going forward. You can't test them looking backward.

WALKER: Yeah. Otherwise the causal theory is just like an adornment.

FAMA: Otherwise it's just rationalization.

WALKER: Post hoc.

FAMA: Right, post hoc.

WALKER: I'm curious, what do you make of George Soros's theory of reflexivity, just intuitively?

FAMA: I have no idea what it is.

WALKER: The idea that investors react to price increases and it's like a self-fulfilling...

FAMA: So he thinks there's momentum basically.

WALKER: Exactly.

FAMA: Well, in stocks there is a little bit of momentum. Very short lived. And I've not seen evidence of professional managers that can effectively gain from it. So everybody knows it's in the data. That's well documented. But it's very short term, so not clear you can capitalize on it. It's not something that I look now and then two weeks later I can capitalize on it. It's very short term.

But on a statistical basis that is one of the embarrassments of market efficiency: the existence of this momentum that doesn't seem to be tied to risk in any sense. Because momentum changes so much over short periods of time, it moves across stocks so much in short periods of time that you can't attribute it to risk. It's too short term. So that... No, I have no problem. I mean it was one of my PhD students that discovered momentum and when he came to me and thought I was going to be mad at him, I said “No, it's in the data. It's in the data. That's it.”

WALKER: That was Cliff Asness?

FAMA: Right.

WALKER: So this next question is high variance. It could either be very dumb or very interesting, but I'll give it a try. So I guess I'm trying to gesture at how solid the assumptions like information theory are underpinning the efficient markets hypothesis. So in a million years, if human civilization and stock markets still exist, do you predict that...

FAMA: Will markets be more or less efficient? So usually the way this is asked is we've gotten so much better at collecting information, has the market become more efficient? Well, the problem is that's so hard to test in the data. You don't really know what the answer is. 

So the way I answer it is the market has always looked pretty efficient. When I did my thesis in 1963, we didn't have all of the high-speed stuff that we have now for getting information. But it still looked very efficient at that point. I'm not sure there's any evidence that it's more or less efficient now.

WALKER: Got it. We discussed the joint hypothesis problem. Perhaps you could just elaborate on that a little further?

FAMA: Okay, so you cannot test market efficiency without a story about risk and return, which is a market equilibrium issue. The reverse is also true. You can't test models of market equilibrium without market efficiency. So these two things are like joined at the hip. They can't be separated.

People who do market efficiency, they almost don't exist anymore. Everybody takes it for granted in the academic sphere. It's considered uninteresting to test. But everybody that does market efficiency understands the joint hypothesis problem. But it's not that widely recognized among the people who do asset risk and return models. It's implicitly assumed, but they never make it explicit.

WALKER: I see, so they're not so interested in the efficiency questions.

FAMA: No, they take it for granted.

WALKER: So there are a few asset pricing models, obviously the Capital Asset Pricing Model or CAPM. Then there was the three-factor model that you and Ken French created to extend the CAPM. And then after that, more recently, the five-factor model. There are also models that incorporate momentum as a factor. I have some questions about the CAPM. Perhaps we could just begin if you could just very briefly outline what the CAPM actually is.

FAMA: Well, the CAPM was a brilliant insight of Bill Sharpe. I don't think it was his PhD thesis, but it was the next paper that he wrote. It was published in 1963 I think, and it was the first asset pricing model. So it was the first formal story about what is risk and what's the relation between risk and expected return.

And the model is really simple. So it basically says in a simple world everybody would hold a combination of risk-free security and the market portfolio, and you would vary risk by how much you put in the risk-free security and how much you put in the market portfolio. And then by varying the proportions in the two in that model. And that model, if everybody followed it and they all had the same information, in other words, market was efficient, it would say that the measure of risk is your sensitivity to the market portfolio. The sensitivity of your security's return to the market return. That was all you would have to know in order to describe the expected return on the security.

So that was a powerful idea and it was tested up and down. It looked very good for… That’s the way these models go. They look very good for 10 years. And then so-called anomalies come along that say, well, they can't explain this, can't explain that, can't explain the other thing. That's what happened to that one. That's what happens to every asset pricing model basically.

WALKER: Enough anomalies accrue that it starts to look pretty bad.

FAMA: Right. People look beyond it. So that’s why we came up with the three-factor model and then the five-factor model.

WALKER: Yeah. I'm not sure whether you saw this new paper by Nicolas Hommel and a couple of other authors where they compared different discounting approaches on their ability to predict actual market prices. And they found that discounting based on expected returns, such as variance on the CAPM or multi-factor model actually performs very poorly. And I know that you and Ken French have questioned the CAPM for a few decades now. I guess my question is: is there anything left of it? Is it just utterly useless now at this point?

FAMA: It's not utterly useless. I would never use it. So people who don't understand asset pricing use that model in their classes, for example. So the corporate finance people, when they teach capital budgeting, will tell you to use the CAPM to calculate your cost of capital. The problem is that's terrible in practice. Basically what the evidence says is you can't use this market sensitivity as a measure of differential risk. You do that, it looks like everything has the same expected return. It just doesn't work.

So its applicability is basically gone. But the insight of the model was incredible. So I'll never take that away from Bill Sharpe: as the models go, this one really opened the field up. There's no way around that. The field of asset pricing basically starts with that model. 

And it evolved. Bob Merton made a huge contribution to it.

Three-factor and five-factor model... Well, those are kind of ad hoc. Those are trying to pick up things that we observed in the data. It's not something we came up without looking at the data. So it's a bit suspicious from that perspective. But people have kind of lost interest in asset pricing because stuff we had that looked good doesn't look that good. It's easy to find stuff that doesn't work for us. 

So asset pricing is kind of at a slow point right now. The young people in finance are doing things that really don't look like asset pricing. They're kind of branching off into other areas. Otherwise they’re worried they won't get tenure, I guess.

WALKER: What are the current fashionable areas? Behavioral finance, or?

FAMA: Behavioral finance, well, that had its time, but everything is behavioral after all. All of economics is behavioral. I would say that what is called behavioral finance or behavioral economics is really irrational finance, irrational economics. What's the irrational behavior of people? What's the effect of that on prices, not just of stocks, but of everything? That's what that stuff is about.

Now the problem is that behavioral finance, behavioral economics, doesn't have any models of their own. It's just a criticism of other models. 

So I've always chided Dick Thaler and told him, “Hey, it's easy to criticize my models if that's what you guys do. Give me a model of yours that I can criticize.” 

Never happens. 

I really get under his skin when I say, well, there's no real behavioral economics. It's just a branch of efficient markets. You don't have a model of your own. You just have a criticism of efficient markets. So they're really just my cousin.

WALKER: I heard a debate between you and Thaler where you said that you were the most important person in behavioral finance.

FAMA: That's what I said. That's another one of my lines. Without efficient markets they'd have nothing to criticize.

WALKER: Yeah. Do you think eventually the anomalies will coalesce into a theory?

FAMA: That is the hope, but hasn't happened so far.

WALKER: Are there any good efforts that you've noticed? Stuff maybe by Andrei Shleifer or…?

FAMA: No, Andre is trying to develop behavioral models. So he's trying to give content to what I would call plus-content to the behavioral aspect. But I haven't seen anything from that school yet.

WALKER: In your opinion, what's the current best asset pricing model? Is it the five-factor model?

FAMA: I don't know. I wouldn't claim that. It does well on the things it was designed to explain, both nationally and internationally. But there are contradictions of it. So it's like every other model. There are things that it can't explain. So I would say it explains the things it was designed to explain, and they're really important. A lot of money is managed based on those things. But is it the best model? I hope not.

I would like to see... I don't want more factors. I want less. I want simpler models that work, not more complicated models. So I'm still hoping that it'll last—I will last—to the point where something good comes along that says, “I don't need five; here are two that'll do the trick.”

WALKER: Yeah. More parsimonious.

FAMA: Yeah, right, exactly.

WALKER: Because the models with more factors feel like you're just overfitting to the data.

FAMA: Right. You're just data dredging.

WALKER: Data dredging, yeah. Have you developed any theories behind any of the factors that you added to the CAPM?

FAMA: Well, so the three-factor model basically added a size factor, small stocks versus big stocks, and a value-growth factor, value versus growth being the second factor. And there was a little bit of intuition in those, in the sense that everybody would think that small stocks are more risky than big stocks. Everybody would kind of agree that value stocks tend to be poorly performing companies. Maybe the market requires higher expected returns for those.

But multifactor asset pricing requires something in people's tastes that make them have negative attitudes that will persist. So if you tell me that after this discovery of these things, value factor, small stock factor, people pile into them because they're really not concerned that the stocks are small or that they're poorly performing companies, they only care about the expected return. Well, then I get a problem because I think that will erase it. I think that'll nullify the model on its own.

The problem is you won't know if that happened or not. So those models have not done as well in the last 15 to 20 years of data. But that's a drop in the bucket as far as model testing goes. That's the reality of it. You basically need a lifetime of data to test an asset pricing model.

WALKER: A whole lifetime is like the minimum? 

FAMA: Right.

WALKER: So obviously these pricing models are about relating risk and return, as you said. How do you think about black swan events in the context of pricing models?

FAMA: Well, I wrote my thesis on those. So it is the case that outliers are much more frequent than would be expected if returns were normally distributed. Yet they're far from normally distributed. They have fat tails in both directions.

Now, one of my first papers was a version of the capital asset pricing model that took account of these outliers. But the problem was if the distributions are symmetric, the CAPM works. And that was my basic point. You really didn't have to do much to accommodate these fat tails.

So there's no specific model that addresses that.

WALKER: Yeah. Are there specific tools or approaches that help deal with the existence of black swan events?

FAMA: Yeah. Don't invest in stocks. I mean, that's what it comes down to, right? If you don't like the fact that there's been several days in history when prices have gone down by more than 15%, you can't live with that, you shouldn't be there.

WALKER: Let's talk about housing. So there's good empirical evidence that housing markets are relatively less efficient than stock markets.

FAMA: What do you mean by that, though?

WALKER: So, for example, I think there's a paper by Case and Shiller where they find enormous inertia, momentum, in house prices. 

FAMA: Right, right.

WALKER: So I guess firstly, do you agree with that claim that housing markets are relatively less efficient?

FAMA: It's very difficult to tell because the data are not that good. I don't think you can really test... I'd love to do it. I don't think you can really test efficiency in the housing market. So they constructed these indices which... They're very good. They're the best housing indices available. But they basically are moving averages. So you're not going to test market efficiency with moving averages. You're building lags into the data.

So I think that's a really difficult question.

WALKER: Okay, so it's hard to test.

FAMA: It's hard to test in that market. The housing market is very difficult to do those kinds of tests.

WALKER: Yeah, but aren't there good reasons to think a priori that housing markets would be less efficient? So, for example: very high transactions costs; they're less liquid; you can't short sell houses; they're simultaneously investment and consumption goods, so you have a lot of amateur investors, you have homeowners.

FAMA: I don't know. Some of those things are common to lots of markets, and they don't seem to destroy market efficiency. I don't know why they would in this one. Common stocks are very expensive to short. So I'm not sure that... It could make it more difficult, but I'm not sure that it should destroy efficiency in that market.

So the issue is, doesn't everybody that buys a house want to get the best possible price? The buyer and the seller both want the best possible price, so they have all kinds of incentives to investigate whether they're getting a good price or the right price. If that doesn't work, I don't know. I don't know how to test it unless you gave me really good data on all the transactions that take place.

And even then you get a quality problem. So every house is different. So it's not like they're all comparable. There are price series on General Motors and whatever. That's the beauty of the Case-Shiller thing is they [take] repeat sales of the same house.

But they still have the quality problem because they're looking across houses and constructing indices.

WALKER: Right. And people can modify their homes over time.

FAMA: Sure. Right. It's a difficult issue. But I wouldn't say a priori that there are problems, something about that market that makes it less efficient automatically, because other markets have problems too.

WALKER: But I guess it's a difference of degree.

FAMA: Maybe, maybe. But I don't think we have the data that allow us to tell how bad it really is—if it is bad.

WALKER: Should we have futures markets for house prices?

FAMA: I think Bob Shiller tried to do that. And I think the... I forget whether it was the Merc or whatever... One of the Chicago exchanges tried to develop an index that people would trade on, and there wasn't enough interest in it, so they gave it up.

WALKER: Yeah, that was my next question. Why haven't such markets taken off in the U.S.?

FAMA: Well, that's a good question. The stock answer would be there just isn't enough volatility. So futures markets exist on volatility and there's just not enough volatility in house prices to keep them going, I guess.

WALKER: And should we be encouraging those markets?

FAMA: We should be encouraging markets. Shiller, who doesn't really believe in market efficiency but who has done really good work, thinks you still want to develop markets and they'll make things better. So I would say the same thing, really. I mean, you do want to develop futures markets in these things if people want to trade in them, but if they don't, nobody's going to have an incentive to keep them going.

WALKER: Yeah. So you mentioned that you're doing some work on housing at the moment. Can you share a little bit about that?

FAMA: Okay, well, we're trying to not test efficiency, because I think that's impossible, but one of the papers is trying to extract the information from house prices about expected future rents. That's very difficult because what you find is what you also find in stock prices—that is, house prices vary much more than rents.

When we think of that in terms of stocks, what we think is that the discount rate for expected future earnings and dividends is varying through time, and that's creating variation in addition to variation associated with dividends earnings. Well, we get the same problem with houses. There's a discount rate for expected rents that seems to vary a lot through time. 

It's highly correlated across areas. West coast, east coast, central areas. 

It's much more extreme in the coast than it is in the central areas. It’s really interesting.

WALKER: Why do you think that is? 

FAMA: I don’t know.

WALKER: Any hunches?

FAMA: A big ingredient in volatility is: how restricted is the housing market? So, is land very expensive because the rules about building houses are very restrictive, like right here [in Los Angeles], they're very restrictive. So the value of land here, it's incredible.

So there's those kinds of things. Maybe they are more extreme in the coast than they are in the central areas. But there are people who work on that. And that's... I think that's one of the conclusions they come to. 

So we're looking at things that are very, not high level, but an aggregate view. So we're looking at basically 11 metro areas and looking at prices and rents in those areas and seeing if we can extract information and prices about expected rents. And I think we've done it to the extent it can be done.

WALKER: Okay, so you famously canceled your subscription to The Economist because they were throwing the B word... Am I allowed to say it?

FAMA: Yeah, yeah.

WALKER: Bubble around too lightly. Can you explain your problem with the concept of bubbles?

FAMA: The concept bubble, the way they're using the word, is something you identify with hindsight. Not something you identify going forward. 

So I think that in the actual empirical literature, there's no evidence of bubbles, because there are huge price swings, but they're basically unpredictable. If they're not predictable, that kind of violates the definition of a bubble. That was my problem with their sloppy use of the term.

WALKER: So when you're talking about huge price swings, it's kind of reminiscent of a weak-form efficient markets test. But what about a semi-strong form kind of test?

FAMA: No, that's fine. If you can find things that you can use to predict prices, good luck to you. I mean, fine. That's a higher level test than just looking at statistical behavior of prices. That's fine. I mean, that was a popular area. I, along with two of my students, wrote the first paper testing that, and after that, hundreds of papers were written looking at different events and seeing how well prices adjusted to those events. Usually it was very good. So those event studies were among the best evidence on how efficient the market is.

WALKER: Excluding you and David Booth, does Dimensional use any behavioral finance insights in its strategies?

FAMA: Not that I know of.

WALKER: You've expunged them?

FAMA: No. I mean, basically, they're an efficient market shop, so they're buying and creating products based on basically the three-factor model and the five-factor model, mostly the three-factor model. So they're assuming that that is a model for expected returns, and they'll give people mechanisms, the products that will allow them to put their money in, the ones that seem to have higher expected returns—and the ones that seem to have lower expected returns. If you want to do that.

WALKER: Your experience with Dimensional, how easy did you find it translating academic theory into practice?

FAMA: Well, they were my former students, so David's one of the best students I've ever had. He had no problem grasping these things. And the people on our line there, initially they had one Caltech guy who learned finance in about 12 minutes, and then we got another Caltech guy who learned finance in about 10 minutes. So the stuff we're saying was not a problem for them to grasp, and they implemented it without a problem. They were never in the stock picking game. They were pure scientists.

WALKER: I'm just curious generally how easy or difficult it is to translate academic research into practice.

FAMA: Oh. Well, it used to be incredibly difficult. 

WALKER: Why’s that?

FAMA: Well, go back to 1963 when I wrote my thesis. I mean, there was nobody doing passive investing at that point. And it took a long time before it caught on. And it took a long time before there was a substantial fraction of total investing done passively. And it's only recently, I think, that it's gotten around 50% in the stock market, and we're talking 70 years here—’65. So it takes a long time for the stuff to penetrate. And there are always people who don't believe it, which is fine.

WALKER: Is there a way in which efficient market hypothesists and behavioral finance people converge on the same prescription for how average people should manage their money?

FAMA: Absolutely.

WALKER: So efficient market purists would say that the market is efficient, so just, you know, invest in an index fund. Behavioral finance people would say people are irrational and dumb and it's nihilistic, so just invest in an index fund.

FAMA: Right. So they come to the same conclusion, different reasons.

WALKER: Yeah, but I guess those reasons could mutually coexist.

FAMA: They could. But how many people do you know that will volunteer that they're dumb? But according to the behaviorists, everybody's dumb. It's not just the people you think are dumb, but experts are also dumb. So they make mistakes. They make consistent mistakes that should be easily avoidable. You know, doctors and… it's not just finance people. 

That's their problem, not mine.

WALKER: Yeah, all the experts are dumb except for the behavioral finance people!

FAMA: Well, that's a good point actually. So why aren't they among the dumb? Maybe they are the dumb and the other people are the smarts.

WALKER: Have you heard of this idea of bias bias?

FAMA: No.

WALKER: Gerd Gigerenzer has a paper, I think with that in the title. But it's about the tendency to see bias everywhere.

FAMA: Do the two negatives make a positive?

WALKER: Well, it's more about the tendency to see bias in places where it doesn't exist or can be explained by something else. (But I guess you could apply that to the behavioral finance people.)

I'm curious about the link between libertarianism and the efficient markets hypothesis. Correct me if I'm wrong, but you would describe yourself as a hardcore libertarian. Were you a hardcore libertarian before you started getting interested in the efficient markets hypothesis? Or was it the other way around?

FAMA: In college, I guess I would say that the professors were... They weren't libertarians, that's for sure, but they were more liberals than libertarians back in the ‘50s, so I would have been influenced by them. And then when I went to the University of Chicago, took Friedman's course and listened to the goings on in various workshops and started thinking about stuff, I became a libertarian. That was before I wrote my thesis. 

So I think I turned my political coat before actually dealing with efficient markets. 

Milton never believed markets were efficient. He didn't. He thought he could beat the market. Never saw any evidence to that effect.

WALKER: Milton Friedman?

FAMA: Yeah.

WALKER: Did he beat the market?

FAMA: No, I don't think so. But he thought he could.

WALKER: Did he accept the logic of the efficient market's hypothesis or...

FAMA: He accepted the logic, but I don't know...

WALKER: He just thought he was in that fraction.

FAMA: Yeah, right. Everybody does actually.

WALKER: But maybe that's a bias. Overconfidence.

FAMA: Well, it is a bias. I'd like to see his portfolio.

WALKER: You're free to reject that premise that I guess was implied in my question that there's perhaps some kind of connection between a libertarian worldview and belief in the efficient markets hypothesis. I don't necessarily like that word belief, but do you understand my point?

FAMA: Yeah, I do understand it. I'm thinking... So, I don't know Thaler's politics, but you know, it used to be easy. People's politics and their economics used to be easy to figure out. It isn't anymore. So Chicago's a free market school still. But you know, one of my colleagues just became... was Barack Obama's chief economics person and is now going to the Fed.

WALKER: Who was that?

FAMA: Austan Goolsbee. 

So my guess is Austan's a libertarian when you come down to it, but he's not... He's a Democrat. But his ideas about economics are quite libertarian.I don't think he likes a lot of government interference. I don't know anybody that does. Among economists that's pretty general. They're suspicious of what government is likely to do given a free hand. 

So the way I classify libertarians is: we don't trust Republicans or Democrats and we don't favor one over the other. We think they both are self-seeking.

WALKER: I'm interested in talking about some applications of the efficient markets hypothesis generally. Have there been any situations outside of your direct field of research where you've noticed that it was analogous to the efficient markets hypothesis? Is it generalizable in any interesting ways?

FAMA: Well, basically the presumption in all of economics until recently was that behavior is rational; the prices in all contexts, not just financial markets, are rational. So in that sense the idea of efficient markets is ancient. It was the basis for all of economics. Now it's being questioned all over the place, not just in finance, but now all kinds of markets.

So remains to be seen how true it is in individual markets outside of finance and how many profit opportunities are out there that people have not yet exploited. But that's the issue. It's the same across all areas basically. And some of them you got to make a bigger bet and a more concentrated bet than you do buying a diversified portfolio in the market. But it's basically the same problem.

WALKER: Let me give another concrete example. Moving to labor markets, I look at the US labor market in particular and view it as quite inefficient (inefficient in obviously a broader sense). There's a lot of credentialism. Employers obsess over pedigree more than skills. There are a lot of jobs that require four-year degrees for just inexplicable reasons. Why has someone not arbitraged that away? Could you use efficient markets thinking to explain why the...

FAMA: You could use empirical work for sure. To see if the presumptions that you just stated are actually true. Now if I went and looked, would I find that people with four-year degrees are actually in the end more productive than people that don't have them in the same jobs? So I think there's a presumption that there are a lot of jobs where that isn't true. But what's the evidence? Now that's always the bottom line question. What's the evidence? So I don't think these questions have ex ante answers to them. They require tests. 

WALKER: Couple of questions to finish on. So as I mentioned, I was interviewing Danny Kahneman in New York. I'm curious what you make of... Obviously his research with Amos was the research that kind of kicked off the behavioral economics program, including in many ways behavioral finance. What do you make of their research generally, Danny and Amos?

FAMA: It's had an impact, obviously, in the sense that people have become much more aware that there are pretty systematic biases that people have, that you can avoid with simple cures. Dick Thaler's a genius at that, coming up with how do you arrange the choices in your retirement plan to make people do things that you know they think they should do, but they don't get around to doing it. So how do you set up the choice? How do you set up the decision problem to make them do the things they want to do? But what an economist would say is seems to them too costly to do it. He wants to lower those costs. 

But so take the book Thinking Fast, Thinking Slow. Okay. That's Kahneman's big seller. So I threw this one at Thaler and he didn't have an answer to it. I said, "Dick, that's not a scientific theory. What can't I explain with thinking fast and thinking slow? It's a tautology." 

And he thought about it and I think he agreed that was not... And that's an incredibly popular book that people think is full of insights. But the basic presumption is a tautology.

WALKER: Being dual process theory.

FAMA: Yeah, right. If you tell me, okay, I'm going to explain what you did because you were thinking too fast, and I'm going to explain what you did because you were thinking too slow, what can't I explain then?

WALKER: But shouldn't we think of that as the underlying conceptual framework. It's like evolution—evolution by natural selection is a tautology as well. In a way, isn't the efficient markets hypothesis a tautology?

FAMA: No, because I can contradict it. I can get evidence that contradicts it. Evolution too. I mean, that could have gone a different way, right? I mean, you couldn't perfectly predict what's going to evolve from selection.

WALKER: But the concept of: you've got variation and then the things that get selected for the things that survive and reproduce the best.

FAMA: Yeah, there're probably exceptions to that too, right? 

Still, I think the behavioral stuff in systematizing the mistakes that people make in different circumstances and how you can explain them—I mean, that doesn't have anything to do with Kahneman's book—I think that's very useful. I mean, there's just no way around it.

What I object to among the behaviorists is that they go into a problem looking for those things and they're not willing to go the other way. That, to me, is not scientific. 

So I would go in and say: you've got a problem, it could be because markets work, or it could be because you've discovered something that's inconsistent with rational behavior in markets. But you don't go in with a presumption about the answer to that. And I think most of them do. I don't go in with a presumption, or at least I try not to.

And I'm willing to be contradicted by evidence that tells me, like Cliff's momentum, for example—that's a big embarrassment to market efficiency. But it's there. So I'm not going to argue with it. I mean, statistically it's unassailable. And I think that's the way everybody should approach these issues. You shouldn't go in looking for something or a particular slant and then move on to something else if you don't find it.

WALKER: It's just data dredging.

FAMA: Right.

WALKER: How much does momentum undermine the efficient markets hypothesis? How much less of an efficient market hypothesist did you become after Cliff's discovery?

FAMA: It's something that's so short term that it doesn't seem like anybody can make any money off of it. So it's basically a curiosity item. But it is a violation of market efficiency. So no way around that. But if it's just in the realm of a curiosity item, I don't think it affects anything in the end.

There's also evidence of reversals. They're not reversals. It's just negative autocorrelation in the long term. But you expect that if expected returns are varying through time, so that's not so embarrassing. There are explanations for that. But short-term momentum is kind of a killer, at least logically.

WALKER: Final question, apart from the housing stuff, what are you working on at the moment, and what excites you? Or is it just mainly the housing stuff?

FAMA: I'm chuckling because that's kind of the way I am. So I’ll usually start with a small thing, write a paper on it, and then that suggests some extension that you want to do. And eventually when you put it all together, all the papers that you write in the same thing, it looks like something much bigger.

So that's the way I've always worked. I'm not the person that can jump ahead with my mind and say, here's where I want to be in 10 years. I can't do that. I make little steps and try to build on them. So I don't know where this real estate stuff's going to go. 

And I'm doing this with Ken French. So I'm hoping we can find data that allow us to investigate more questions. But it remains to be seen.

WALKER: Yeah, I guess that's the nature of your very empirical approach: just follow things where they lead.

FAMA: Right.

WALKER: Well, it's been great chatting with you, Gene. Thank you so much for having me here today, and it's been an honor.

FAMA: My pleasure.