Russ Roberts

Fama on Finance

EconTalk Episode with Eugene Fama
Hosted by Russ Roberts
PRINT
David Rose on the Moral Founda... William Black on Financial Fra...

Eugene Fama of the University of Chicago talks with EconTalk host Russ Roberts about the evolution of finance, the efficient market hypothesis, the current crisis, the economics of stimulus, and the role of empirical work in finance and economics.

Size: 28.3 MB
Right-click or Option-click, and select "Save Link/Target As MP3.

Readings and Links related to this podcast

Podcast Readings
HIDE READINGS
About this week's guest: About ideas and people mentioned in this podcast:

Highlights

Time
Podcast Highlights
HIDE HIGHLIGHTS
0:36Intro. [Recording date: January 17, 2012.] Russ: Your impact on the field of finance has been immense--in a whole bunch of areas, but one that stands out is the efficient markets hypothesis (EMH). I'd like you to sketch out the evolution of that idea in the field, how it was understood initially, and how it has changed over time. Guest: How much time do we have? Russ: Well, four or five hours, but let's try to keep it to under 10 minutes for this first question, if you can. Guest: Okay. I'll go back to the beginning. The way Harry Roberts tells it, Holbrook Working in the 1930s started to become interested in whether speculative prices moved randomly. He was mostly an agricultural economist, looking at agricultural commodities, and he took a series of random numbers, simulated them, and brought them to his faculty at Stanford, faculty lounge, I guess; showed them to them and they agreed they were an agricultural series. So he thought from that that maybe a random walk kind of model would work pretty well for agricultural prices, prices of other commodities. But then there was a big gap from there to like the end of the 1950s. And what opened things up was the coming of computers, which made computations much easier. And the most readily available data was stock price data. So, basically, statisticians, econometricians took the data and started doing calculations on it, calculating autocorrelations with their estimates of how predictable returns are based on past returns. And then they stopped. Economists got into the mix and said: Okay, how would we expect prices to behave if they were set based on all available information? Which is basically the EMH, but it wasn't stated in those terms at that time. So, they said: I think they should be a random walk, an hypothesis pulled out of the air. Russ: When you say it's a random walk, explain what that means. Guest: That means that expected changes are successive changes are independent of one another. It also means they have identical distributions, but that part is not important. It's basically the independence part that's important. It basically means that you can't predict future returns based on past returns. Russ: And yesterday doesn't tell you anything about tomorrow. Guest: Right. Returns from day to day are basically independent of past returns. Now that was a very extreme hypothesis. Let me give you an example. You wouldn't say that about tomatoes, for example. Tomatoes are going to be cheaper in August than in January, for the most part, because they are seasonal. It has to do with supply and demand--mostly supply of tomatoes. There's a similar thing operating in prices of stocks, bonds, whatever. Basically, there's an expected return component, what people would require in order to hold those securities; and there's no reason that that has to be independent through time. There's no reason why that's not predictable or why it doesn't go--and there's lots of evidence that it is--higher on stocks during recessions and lower during good times. So there can be predictability in returns that is consistency with efficient markets. What people didn't understand in the beginning was that propositions about how prices should behave had to be joined to a statement about how you think they ought to behave. In other words, what you need is some statement about what we call a market equilibrium. What is the risk-return model you have in mind underlying the behavior of the prices in returns? So, for example, stocks are very risky; they require a higher expected return than bonds; and you have to take that into account in the tests. So there is this, what I call the joint hypothesis problem, which is basically what I added to the mix, but it's kind of an important part of it. It says whenever you are testing market efficiency you are jointly testing efficiency with some story about risk and return. And the two are joined at the hip. You can't separate them. So, people infer from that, it means market efficiency is not testable on its own. And that's true. But the reverse is also true. A risk-return model is untestable without market efficiency. Most risk-return models assume that markets are efficient. With very few exceptions.
6:18Russ: And so when we say markets are efficient, what do you mean by that? Guest: What you mean is that prices at any point in time reflect all available information. Russ: Now that idea--what's the distinction between the weak form and the strong form that people talk about? Guest: Two words that I used in 1970 that I came to regret. Because I was trying to categorize various tests that were done. So, I called weak form tests, tests that only used past prices and returns to predict future prices and returns. And I called semi-strong form tests, tests that used other kinds of public information to predict returns, like an earnings announcement or something like that. And then I called strong form tests, tests that look at all available information; and those are basically tests of if you look at groups of investment managers and you look at returns that they generate, you are basically looking at all the information they had to generate to [?] securities, and what's the evidence that the information they had wasn't in prices. Russ: And empirically, where do we stand today, do you believe and what has been established about those various hypotheses? Guest: Well, believe it or not, the weak form one has been the one that has been subject to the most, what people call anomalies, in finance. Things that are inconsistent with either market efficiency or some model of risk and return. The big one at the moment is what people call momentum--prices seem to move in the same direction for short periods of time. So, the winners of last year tend to be winners for a few more months, and the losers tend to be losers for a few more months. In the strong form tests, Ken French and I just published a paper called "Luck Versus Skill in Mutual Fund Performance," and basically looked at performance of the whole mutual fund industry--in the aggregate, together, and fund by fund, and try to distinguish to what extent returns are due to luck versus skill. And the evidence basically says the tests it's skill in the extreme. But you've got skill in both extremes. That's something people have trouble accepting. But it comes down to a simple proposition, which is that active management in trying to pick stocks has to be a zero sum game, because the winners have to win at the expense of losers. And that's kind of a difficult concept. But it shows up when you look at the cross section of mutual fund returns, in other words the returns for all funds over very long periods of time. What you find is, if you give them back all their costs, there are people in the left tail that look too extreme and there are people in the right tail that look too extreme, and the right tail and left tail basically offset each other. If you look at the industry as a whole; the industry basically holds a market portfolio. That's all before costs. If you look at returns to investors then there is no evidence that anybody surely has information sufficient to cover their costs.
10:11Russ: Which says that for any individual investing, certainly someone like me, that is, who doesn't spend any time or very much time at all looking--in my case no time, but let's suppose even a little time--trying to look at what would be a good investment. The implication is to go with index mutual funds because actively managed funds can't outperform. Guest: Well, no, it's more subtle than that. What's more subtle about it is, even if you spent time, you are unlikely to be able to pick the funds that will be successful because so much of what happens is due to chance. Russ: So, for me the lesson is: buy index mutual funds because the transaction costs of those are the smallest, and since very few actively managed funds can generate returns with any expectation other than chance to overcome those higher costs, I can make more money with an index fund. Guest: Right. Now, it's very counterintuitive, because we look at the whole history of every fund's returns, and sort them, and really the ones in the right tail are really extreme. Russ: Some great ones. Guest: They beat their benchmarks by 3-6% a year. Nevertheless, only 3% of them do about as well as you would expect by chance. Now what's subtle there is that by chance, with 3000-plus funds, you expect lots of them to do extremely well over their whole lifetime. So, these are the people that books get written about. Russ: Because they look smart. Guest: What this basically says is that there is a pretty good chance they are just lucky. And they had sustained periods of luck--which you expect in a big sample of funds. Russ: Of course, they don't see it that way. Guest: No, of course not. Russ: A friend of mine who is a hedge fund manager--before I made this call I asked him what he would ask you, and he said, well, his assessment is that efficient markets explain some tiny proportion of volatility of stock prices but there's still plenty of opportunity for a person to make money before markets adjust. And of course in doing so, make that adjustment actually happen and bring markets to equilibrium. Somebody has to provide the information or act on the information that is at least public and maybe only semi-public. What's your reaction to that comment? Guest: That's the standard comment from an active manager. It's not true. Merton Miller always liked to emphasize that you could have full adjustment to information without trading. If all the information were available at very low cost, prices could adjust without any trading taking place. Just bid-ask prices. So, it's not true that somebody has to do it. But the issue is--this goes back to a famous paper by Grossman and Stiglitz--the issue really is what is the cost of the information? And I have a very simple model in mind. In my mind, information is available, available at very low cost, then the cost function gets very steep. Basically goes off to infinity very quickly. Russ: And therefore? Guest: And therefore prices are very efficient because the information that's available is costless. Russ: But what's the implication of that steep incline? That information is not very-- Guest: It doesn't pay to try to take advantage of additional information. Russ: It's not very valuable. Guest: No, it's very valuable. If you were able to perfectly predict the future, of course that would be very valuable. But you can't. It becomes infinitely costly to do that. Russ: So, your assessment, that you just gave me of the state of our knowledge of this area, I would say remains what it's been for some time--that at the individual certainly there is no return to--prices reflect all publicly available information for practical purposes for an individual investor. Guest: For an individual investor? Even for an institutional investor. Russ: Correct. So, what proportion of the economics and finance areas do you think agree with that? Guest: Finance has developed quite a lot in the last 50 years that I've been in it. I would say the people who do asset pricing--portfolio theory, risk and return--those people think markets are pretty efficient. If you go to people in other areas who are not so familiar with the evidence in asset pricing, well, then there is more skepticism. I attribute that to the fact that finance, like other areas of economics, have become more specialized. And people just can't know all the stuff that's available. Russ: Sure. Guest: There's an incredible demand for market inefficiency. The whole investment management business is based on the idea that the market is not efficient. I say to my students when they take my course: If you really believe what I say and go out and recruit and tell people you think markets are efficient, you'll never get a job. Russ: Yes, it's true. And so there's a certain bias, you are saying, to how people assess the evidence. Guest: There's a bias. The bias is based, among professional money managers, the bias comes from the fact that they make more money from portraying themselves as active managers. Russ: That's true in macroeconomics as well. We'll get to that a little later in the conversation.
16:50Russ: I was going to ask you about the current crisis. Guest: I have some unusual views on that, too. Russ: I'd say that the mainstream view--and I recently saw a survey that said--it was an esteemed panel of economists; you weren't on it but it was still esteemed, both in finance and out of finance. And they asked them whether prices reflected information and there was near unanimity. Some strongly agreed; some just agreed. But there was also near unanimity that the housing market had been a bubble. Guest: The nasty b-word. Russ: Yes; and was showing some form of what we might call irrationality. Guest: Okay, so they had strong feelings about that, getting mad about the word bubble. Russ: Why? Guest: Because I think people see bubbles with 20-20 hindsight. The term has lost its meaning. It used to mean something that had a more or less predictable ending. Now people use it to mean a big swing in prices, that after the fact is wrong. But all prices changes after the fact are wrong. Because new information comes out that makes what people thought two minutes ago wrong two minutes later. Housing bubble--if you think there was a housing bubble, there might have been; if you had predicted it, that would be fine; but the reality is, all markets did the same thing at the same time. So you have to really face that fact that if you think it was a housing bubble, it was a stock price bubble, it was a corporate bond bubble, it was a commodities bubble. Are economists really willing to live with a world where there are bubbles in everything at the same time? Russ: And your explanation then of that phenomenon? Guest: My explanation is you had a big recession. I think you can explain almost everything just by saying you had a big recession. A really big recession. Russ: And why do you think we had a really big recession? Guest: I've heard some of your podcasts; I'm with you. I don't think macroeconomists have ever been good at knowing why we have recessions. We still don't understand the Great Depression. Russ: True. Although Ben Bernanke would argue, and Milton Friedman would argue and he did before he passed away, that monetary policy is a huge part of it. Guest: Let me reflect. I had this discussion with Milton, actually; and what I pointed out was from your own data, they show that there were massive free reserves throughout the Great Depression. And my point is: we can't force people to move demand deposits. Or to make love to anyone. Russ: Well, you can but it's not very productive. Guest: It's not very productive. M1 and M2--those things are basically endogenous. Russ: I have the same feeling. Guest: The only thing that's sort of exogenous is the monetary base. Russ: What did Milton say to that? Guest: All I gathered from Milton was: Interesting. Even when you won you thought you lost. Russ: Yes, I know. I had plenty of those. So, are you saying that that's analogous to our current situation? Guest: Oh, no. What I'm saying is that for example people want to blame the recession on the housing sector crashing and subprime mortgages. But if you are an economist and you are thinking about that, you have to be saying that there was some misallocation across markets, that margins weren't being equated across markets. That's pretty hard to accept because people are acting in all markets, working in all markets. That's a pretty tough one to follow. Russ: Well, a lot of people swallow it. Here's their version. They say things like there are these things called animal spirits that you can't measure, but that doesn't mean they are not real; that people get all excited about a particular asset class--in this case it was housing. And as those prices start to rise it becomes rational to speculate that it will continue to rise. And as that happens--as you would admit, people are making money along the way--and then they don't. They stop making money; the prices collapse. And this happens from time to time because of irrational exuberance; and that's just an aspect of capitalism. That's the standard counterpoint. Guest: Okay, but it wasn't just housing. That was my point when we started. The same thing was going on in all asset markets. Russ: Well, the timing isn't quite identical for all asset markets, right? The stock market--the housing market starts to collapse I think around early-mid-2006. Guest: It stops rising, right. Russ: And then begins a steady decline. Guest: That decline was nothing compared to the stock market decline. Russ: But when did that happen? Guest: I don't know the exact timing. Russ: It's not around then. It's later. Guest: The onset of the recession started with the collapse of the stock market. The recession and the collapse of the stock market, the corporate bond market, all of that basically coincides. But that also coincides with the collapse of the securitized bond market. Russ: Mortgage-backed securities. Guest: The subprime mortgages and all of that. Russ: Well, yes; that happens through 2007, 2008. I guess there is some parallel. So, you are going to reverse the causation. Guest: I'm not saying I know. What I'm saying is I can tell the whole story just based on the recession. And I don't think you can come up with evidence that contradicts that. But I'm not saying I know I'm right. I don't know. I'm just saying people read the evidence through a narrow lens. Russ: Yes, they do. Confirmation bias. Guest: And the rhetoric acquires a life of its own; so there are books written that basically all say the same thing about the crisis. Russ: And you are arguing that they have essentially cherry-picked the data. Guest: Well, they just look at pieces of the data and the fact that the housing market collapsed is taken to be the cause; but the housing market could collapse for other reasons. People don't just decide that prices aren't high any more. They have to look at supply and demand somewhere in the background. Russ: We did have people holding second and third homes who didn't have the income and capability of repaying the first one. Guest: Sure. Standards were relaxed. But then you have to look on the supply side, the lending side. The people who were lending to these people had the information. Russ: Yes, they knew it. I don't think that they were fooled. They were not overly optimistic about the value of those loans. They were willing to do that because they could sell them. Guest: The puzzle is why they were able to sell them.
24:17Russ: Correct. Now my claim is the people who bought them did it with largely borrowed money. Guest: No, that's not true. These were bought by people all over the world. Russ: Correct. Guest: No one borrowed money. Remember now: savings has to equal lending. For everyone that's short bonds, somebody is on the other side. The net amount of leverage in the world is always zero. Russ: That's true. Guest: So you can't tell a story based on leverage. Russ: So what's your story? I have to think that through. It's undeniably true, and I'm not going to argue with that point. So, what's your explanation of why people bought these things? Guest: Well, I have no explanation. Again, I'd say the market crashes because of the big recession. Even a minor depression if you like. Remember that all the people buying these subprime mortgages all over the world, they are the ones making the loans in the end, they were sophisticated investors. Institutions, big banks all over the world. They thought these things were appropriately priced. They might have been at that time, but they weren't ex post. Russ: So you are not going to allow me to make the claim that the incentives they faced to worry about how appropriately priced were distorted. Guest: The incentives to make money are always there. The question is whether the market lets you make money. So, these people that wanted to securitize all these mortgages, they could have failed at any time in the process; and they would have failed big time because in order to do these things, you have to initially finance them yourself. So when the investment bankers were bringing out the securitized mortgages and other kinds of securitized assets, they initially held them. And they held them afterwards, too. Russ: They held many of them. Guest: Well, initially they held them all, because they are bundling them together; they have to come up with the capital and then they can sell them. So, they could have failed right at that point because the market says: Forget it. We're not paying you par value for these things. Russ: But when they did fail, which they fundamentally did because, at least for them, even though the world wasn't leveraged, they were leveraged, they should have gone out of business. Guest: Right, exactly. Russ: But they did not. Guest: That's awful. That's the worst consequence of this whole episode. Russ: So, my narrative is the anticipation of that distorted their decision-making. Guest: Sure, but that doesn't satisfy what address what goes on on the demand side. Russ: Why? Guest: Because people on the demand side have to buy these things. Russ: Well, the people who were buying them, and selling them, were fundamentally the same people, right? Guest: Okay, so if greed causes me to put out securities that I know are no good, why would I hold them? Russ: Because I can hold them at a very low cost. I have uncertainty; I don't know what's going to happen. There's an upside; there's a downside. Guest: It's really a low cost if you know you are going to get bailed out. Russ: Right. My argument is it dulls your senses. Guest: It does; I agree with you there. Any probability that you are going to be bailed out is going to distort your decision. Russ: So, is your argument then that that was relatively unimportant? Guest: No, no. My argument is it can't explain why people who weren't generating these things and weren't going to be bailed out by us, investors in Norway, whatever--why were they buying? Russ: Well, I'm happy to admit that some people just made a mistake. After the fact. Ex ante they certainly didn't think they were throwing away their money. And a lot of those people making those investments around the world, we bailed them out, too. The European banks got some of the benefits. Guest: Yes, because they were mixed into the same piles that involved our own investment banks. And so they got bailed out in the process. If they were holding credit default swaps (CDSs) that were sold by AIG, they got bailed out. Russ: Although I think Goldman was the number 2 holder of those. The first was--I can't remember; it was a foreign bank, either French or German.
29:19Russ: So, you have publicly said that that was a mistake, those bailouts; we should have let them fail. Guest: It's irrelevant because there is no political regime that will let that happen. Russ: Correct. But let's suppose, let's live in a fantasy world for forty seconds. Suppose on March of 2008, Ben Bernanke and Hank Paulson and the others who got together to talk about the impending bankruptcy of Bear Stearns had just let them go. They would have opened for business Monday morning without enough cash to cover their positions; they would have had to tell their creditors: Sorry; I can't honor the promise I made to you the other day or the other money; and you won't be getting the payment you anticipated. The justification for the intervention was that if we had let that happen there would have been an enormous crisis: credit markets would have frozen up and we would have had a worldwide depression. Guest: I don't know about that last part. That's what we'll never know. The issue is: How long would it take to straighten things out? And I think it's really overrated that it would have taken a large amount of time. So, banks fail all the time, and the FDIC goes in and draws a line in the sand about who is going to get paid and who isn't; stuff is put up for sale and everything goes on. I don't know how long it would take to solve a multiple failure problem. We'll never know. Russ: Well, the Lehman Brother's bankruptcy is still in process. Which is now three years old. This was the argument made at the time--like you, I'm skeptical about it but it has some legitimacy--it's that bankruptcy is complicated enough as it is; when it's a large investment bank with international creditors like Bear, Lehman, it would take a long time. In the meanwhile everybody would be thrown into turmoil. Blah, blah, blah. Do you think there's anything to that? Guest: It's possible. What happened in the Lehman case is it's held up by multiple jurisdictions. So, you have to settle with the British shareholders. Russ: The Japanese, Korean. Guest: Who all have their own set of laws about what happens in a bankruptcy. And that's what I think they've been fighting over for three years. It's pretty clear what assets [?]. Russ: But isn't that an argument for justifying what Bernanke and Paulson did? Guest: I don't know. Because who knows what would have been done if all of them went down. The problem really is that the investment banks weren't subject to the same disposition rules that would face an ordinary commercial bank. They are not subject to the FDIC. And the FDIC can come in and arbitrarily do it. That's what you buy into when you sign up for it. Whereas for the investment banks, they are not really banks; and they are not subject to those rules. The ongoing problem is that you haven't killed their incentive to finance things the way they always have. Russ: Well, I guess my claim is that part of the problem is that we gave a regulatory advantage to triple-A rated stuff, which allowed very large and different amounts of leverage compared to other stuff. That gave an incentive to these folks to find more triple-A. The amount of triple-A is essentially, until recently, there's just not enough of it to go around, if that's the most profitable thing you can do, because that's the thing you can leverage; so they found a way to invent more of it. And that included not just the things we are talking about, but European sovereign debt. Hey, that's safe; let's leverage that, too. Guest: Right. Russ: So, once we said: this is the stuff that you can make scads of money on because you can leverage it and use other people's money. Guest: You are slipping back again, though. Russ: Because? Guest: You are saying that people will buy this stuff even though it isn't triple-A. Russ: Correct. Guest: Why? Russ: Well, that's the puzzle. Is it because they were stupid, ex ante? Guest: We are talking about the world's most sophisticated people who invest. Russ: So is the alternative argument that people just made a mistake? Guest: After the fact, definitely. Whether it was a mistake before the fact, that involves estimating the probabilities of extreme tail events, which, as you know, are very difficult. Russ: So, where does that leave us? Story-telling, of course. Guest: Which is very entertaining but it's not convincing. I don't find it convincing.
34:45Russ: Before I forget, I was going to ask you--I don't want to miss this chance to ask you this: Does your research inform your own personal portfolio decisions? And has it over time? Guest: Oh, sure, always. Russ: Has it changed over time? Guest: Well, I'm not as young as I used to be. Russ: That's part of the theory, too. Guest: Right. So, my portfolio has become somewhat more conservative. I'm also a stockholder in an investment management company, so that part of it is very unconservative. Russ: That's true. Recently--a related question to what we were just talking about before that--the government published the transcripts of the Federal Reserve deliberations in 2006. I don't know if you've looked at that. Guest: No. Russ: Well, one of the most obvious things you learn from reading those transcripts--well, first of all, this is 15 really smart people, very savvy. Their job is to try to figure out what could happen next that could be dangerous. And in 2006, we were on the edge of a collapse in the housing market. And as you argue, maybe just a general problem coming that would be unforeseeable. But what was interesting was that they made the same mistake that I made at the time; and I heard lots of other people much smarter than I am made the same mistake. They said: Well, it's true that there could be a housing price fall; it's been going up for a long time, but the subprimes are essentially only a small part of the whole housing market; housing is only a small part of the overall investment market. So, if this does occur, there's not going to be much of a consequence and we don't have to worry about it. Now, one of the things I think was mistaken, certainly for me as someone not very well versed in finance, and I think most economists are not very well versed in finance, is that we did not understand the role that leverage would play if asset prices fell by a relatively small amount. Do you think that has been a lesson that some people have learned from this crisis? And should we learn that lesson? Guest: Well, leverage will put some people out of business. Russ: Correct. Guest: So, what's the problem? Russ: Well, the problem is that if lots of people go out of business at the same time it allegedly has a multiplier effect--I hate to use that phrase--but that there is some credit market contagion, systemic risk, etc. Guest: That's a word I don't think existed 20 years ago. Russ: Which one? Guest: Systemic. Russ: But let's go back to our mutual friend, Milton. Certainly Milton would argue that the contraction of the money supply at the onset of the Great Depression precipitated by bank failures was something that the Federal Reserve should have paid attention to. Guest: What could they do? Russ: They should have injected liquidity into the system. Guest: Well, but if you have massive free reserves, what is that going to do? Russ: Well, that's a problem. Again, I wish Milton were here. I'm mystified by monetary policy generally, as anyone who has listened to these podcasts knows. Guest: Well, I am too. In the podcasts of this program that I've listened to, I've heard everybody talk about the Fed controlling the interest rates. That's always escaped me how they can do that. Russ: Yes, I'm mystified by it myself. Guest: But I'm in finance, so you've got an excuse. Russ: When I interviewed Milton in 2006 and I asked him why there had been a change in public discussion at least of what the Fed does from changing the money supply to instead manipulating interest rates, his answer was: Well, that's what they say but that's not what they do. They like to say they manipulate interest rates because it makes them feel powerful. All they really do is change the monetary base. And in fact he said, if you look at M2, that's the thing to look at. Guest: That's the thing to look at if you want to know what's happening to business activity. But it's not something you can do anything about.
39:28Russ: I'm with you there. While we're on that subject, do you have any thoughts on why the Fed is paying interest on reserves? Guest: Oh, absolutely. Because they know that if there is an opportunity cost from these massive reserves they've injected into the system, we are going to have a hyperinflation. Russ: So what's the point of injecting the reserves if you are going to keep them in the system? Guest: Exactly. Russ: So what's the answer? Guest: The answer is: this is just posturing. What's actually happened? That debt is now almost fully interest-bearing, all the liquidity that they've injected. So, they've actually made the problem of controlling inflation more difficult. Controlling inflation when they didn't pay any interest focused on the base: cash plus reserves. But now the reserves are interest-bearing, so they play no role in inflation. It all comes to cash, to currency. How do you know? Currency and reserves were completely interchangeable; that's what the Federal Reserve is all about. So I think they've lost it. Now what happened, they went and bought bonds, long-maturing bonds, and issued short-maturing bonds. It's nothing. They didn't do anything. Russ: But they are smart people. Guest: Right. Russ: Ben Bernanke is not a fool. If you could get him alone in a quiet place with nobody else listening and say: Ben, what were you thinking? What do you think he'd say? Guest: I don't know, but I wouldn't believe it. In the sense that at most he could have thought he could twist the yield curve. Lower the long-term bond rate. Now I'm looking at the long-term bond market--it's wide open. Even though they are doing big things, they are not that big relative to the size of the market. Russ: Yes, I am mystified by that as well. I don't have an explanation. Guest: Let me put it differently. So, if I look at the evolution of interest rates, is it credible that in the early 1980s the Fed wanted the short term interest rate to be 13-14%? Russ: No. You are making the argument that it's endogenous; that they can't control it. Guest: Maybe they can tweak it a bit; they can do a lot with inflationary expectations. That will affect interest rates. Turn it around--all international banks think they can control interest rates; and at the same time they agree that international bond markets are open. Inconsistent. Russ: Correct. It reminds of this CNN reporter, credible insight into economic policy. He said: Macroeconomics generally--and fiscal policy, but he could equally as well be talking about Central Bank policy--he said: Politicians who think they can control the economy are like a little kid who is playing a video game; he hasn't put the money in yet and he is watching the arcade game do all its bangs and bells and whistles and noises. Which is an advertisement for the game. And he's pushing the buttons, and he's attributing all the successes on the screen to himself even though he hasn't put the money in yet because he misunderstands the underlying process that generates what he is seeing on the screen. There is some truth to that. Guest: There's a lot of truth to it.
42:51Let's turn to fiscal policy, which you've written some interesting things on lately. You have been very skeptical, as have a few others. And by the way, I should add, before we get into this I should just mention: your view that it's an open question about whether the crisis was averted by these rather remarkable open interventions by the Fed and the Treasury Department in the last few years--it's not a mainstream view. Certainly most economists believe--and I'm with you--but most economists believe that the Fed and the Treasury and the policy makers did a good thing. Guest: That's not taking into account the long term costs. Russ: For sure. And that would be true of most of these interventions. I always find it remarkable that the auto bailout was a success, quote, "because very few people lost their jobs." As if that's the only effect we would ever want to look at. Guest: The long term effects of that are horrendous. Russ: And it's not clear that they saved very many jobs, either. Clearly they changed the incentives. Guest: Not just changed the incentives--they changed the ordering of precedence in contracts. That's something that's really dangerous. Russ: Yes, they abrogated the rule of law. It's very depressing. But on this issue of fiscal stimulus, most economists believe it's a good thing, it works. We are in the minority who suggest that maybe it isn't effective. And recently you wrote a piece suggesting, I would argue, that it's never effective--unless it's well-spent. And I would contrast it with the Keynesian view, which I heard come out of Joe Stiglitz's mouth personally--people can't be what they actually believe--I heard him actually say: It doesn't matter what you spend the money on; it's all stimulus. You are very much on the other side. So, explain why. Guest: When he says it doesn't matter what you spend the money on, I think he thinks there are multiple choices that would all be good. He doesn't think that if you just wash it down the sink, that's good. Russ: Oh, no; he said, when pressed and he was asked: If you ask people to dig ditches and fill them back in, would that stimulate the economy? And he said: Yes; but it's not as good as doing something productive. I can't explain it. It's a mystery to me. Guest: It's a mystery to me, too. Russ: But he's not on the show right now; I wish he were; I'll try to get him down the road. But in your view, talk about what you think the effect of stimulus is and why you are skeptical. Guest: This is a case where you can't be sure. If you look at the empirical evidence, it basically allows you to say anything you want, because the estimates of the effects of stimulus are subject to so much uncertainty. So, I think, though, if I interpreted Christina Romer's stuff properly, or she and her husband's stuff, what it says is that the only thing that clearly gets a pretty good statistical support is permanent [?]intervention [?]. And the other stuff is just [?]. I think that's probably--I'm an empiricist in the end, so that's probably, I don't know. I have my position that I think it's a waste of money, because it will all be wasted. Eventually, you have to finance it. You have to finance it now, which means eventually you have to pay back, future generations have to pay back, for things that are then mostly useless maybe. But the evidence doesn't, like you say. So it's possible for Stiglitz to say one thing; it's possible for you and I to say something entirely different. And neither one can point to the evidence. Russ: I don't view it as a very scientific enterprise. I view it as essentially ideology being wrapped up in scientism, scientific looking, statistical estimation. It seems to me there is too much noise. Guest: I don't agree with what you said when you started; I don't think most economists do think it works. Maybe I'm in the wrong cocoon. Russ: Yes, you need to get out more, Gene, I think. Although I'm in a different cocoon over here on the East Coast; I'm in the only cocoon, I'm at George Mason University and occasionally I'm at Stanford; so we just happen to talk about the three places where there is an overwhelming majority that is skeptical; but outside of those three, I think it's pretty much the other way. Guest: Well, Bob Barro.Russ: Lonely voice, in that enclave. Guest: I think with Barro, famous macroeconomist at Harvard, there's a younger guy. Russ: Alesina. Guest: Council of Economic Advisers. Russ: Oh, Mankiw. Guest: He's skeptical, but what he says is: Once you get into politics, you become a Keynesian. The political pressures are enormous. I think that's right. Russ: It's a terrible view of our intellectual opponents, though. It's not very nice. We don't like it when they attribute our views to being friends of business, which I find repugnant. So, it seems embarrassing to suggest that they hold their views because they like being powerful. I think there's some truth to it, but it's not very nice. You want to hold that view? Guest: Hold which view? I don't know. I don't think economists are different from other people. They all like, have their views, excepted [accepted?] by everybody else, no matter what their views are. Russ: We're prone to incentives; there's no doubt about that. Guest: I've had a tough time for a long time because I believe in efficient markets. Russ: Get a lot of flack.
49:13Russ: Let's go back to finance for a minute. I will put a link up to your recent article on stimulus where you make a theoretical argument against stimulus. Guest: There's no data, right. Russ: And I think basically--it's interesting how the Chicago school has been pushing this--you are using what I would call accounting identities. The money has got to come from somewhere. I expressed it as the resources have to come from somewhere. Guest: That's the right way to say it, actually. Russ: And so I don't understand where the free lunch comes from. Guest: There is no free lunch. Russ: But the counterpoint is that there is a free lunch because there are all these resources laying around. And then it's a question--Milton said this also--how much of the stimulus goes towards the unused, so-called-- Guest: But that's the problem of implementation, which is horrendous. The same problem in regulation: implementation, which is always the killer. Russ: But let's go back to finance. There's been a big trend in recent years towards what's called behavioral finance. What's your assessment of that? Guest: I think the behavioral people are very good at describing microeconomic behavior--the behavior of individuals--that doesn't seem quite rational. I think they are very good at that. The jump from there to markets is much more shaky. Russ: Explain. Guest: There are two types of behavioral economists. There are guys like my friend and colleague Richard Thaler, who are solidly based in psychology, reasoned economics but he's become a psychologist, basically, and he is coming from the research in psychology. Now there are other finance people who are basically what I call anomaly chasers. What they are doing is scouring the data for things that look like market inefficiency, and they classify that as behavioral finance. But to me it's just data judging [?]. Russ: They don't tell you about the times they can't find the anomaly. Guest: Exactly. In all economics research, there is a multiple comparisons problem that never gets stated. Russ: A multiple what? Guest: The fact that the data have been used by so many other people and the people using it now use it in so many different ways that they don't report, that you have no real statistical basis to evaluate and come to a conclusion. Russ: My view is you should video your keyboard so we can see your keystrokes and then we can see what didn't come out. The dishes that didn't come out of the kitchen because you didn't like the way they tasted. Guest: Right. I've had people say to me that the people who do this anomaly stuff, when they come and give a paper and I'll say, when you do this, that, or the other thing, and they'll say Yes. And I'll say, why don't you report it? And they'll say it wasn't interesting. Russ: Not publishable, either. Guest: Well, that's the problem, that there's a counting process [?] and a publication process as well. You do this, that, and the other thing and I'll say, yes, why don't you report it? And they say it wasn't interesting. Russ: It wasn't interesting. Not publishable, either. Guest: Well, that's the problem, there's a publishing process and a culling process as well. This stuff makes it through.
52:37Russ: So, we started off this conversation talking about efficient markets, and we haven't talked about a zillion other things that you've studied that are important in the field of finance. One question I'd like to hear you talk about is the issue of a non-specialist. Let's say I'm just a smart, everyday person and I want to be educated out in the world. What are the lessons for me that finance has learned that are important? There are obviously of findings that have stood up, findings that have had to be modified over the last 50 years that has become more empirical that an educated person should be able to understand and use? Guest: I'm obviously going to be biased. I think all of our stuff on efficient markets would qualify. I think there is a lot of stuff in the corporate area, corporate governance and all of that, a huge field--that has penetrated to the practical level. The Black-Scholes option pricing paper in view is the most important economics paper of the century. Russ: Why? Guest: Because every academic, every economist whether he went into finance or not, read that paper. And it created an industry. In the applied financial domain. What else can claim that? So, I think we've learned a lot about risk and return. Some of it is intuitive. But there is a lot of stuff on which stocks are more or less risky. A lot of stuff on international markets. Now, what should an ordinary, intelligent person know? That's an interesting question. Let me turn it over. What should an ordinary, intelligent person know about pricing? Russ: Well, I have some thoughts on that. You probably do, too. Guest: Yes. It's difficult. I know lots of very intelligent business people who need some knowledge of what's going on in finance. But not an awful lot. Because that's not a big part of their business. Russ: Correct. I would worry about what people think they know that isn't so and the things they should know. Guest: That happens to me all the time. I'll be playing golf with somebody and they ask me what I do, and I tell them I teach finance and they say, oh my goodness, they don't know anything about finance. And then they give me a lesson in finance. Russ: What do they usually say? Guest: They tell me all about their smart investments. Russ: Do you smile and just take another strike at the ball? Guest: I do. Russ: Do you say anything back? Guest: No, I don't; but I avoid them in the future. Russ: One of the fascinating things about our profession--it must be true for other professions as well--but everybody's an expert. So, your 50 years of empirical, in-the-trenches work is meaningless because that guy had a good month. Guest: Right. Russ: He doesn't think he has much to learn from you. Guest: Indeed. He's making a lot of money. Russ: He is, sure. He's doing great. He knows. I was really thinking more of practitioners than experts, not of lay-people, so I assume there are some things we know in finance that may not turn out to be so. Guest: Oh, absolutely. Russ: I guess I'm thinking about macro, which I know a little bit more about when thinking about the Great Moderation and the comfort that people had that we had mastered the business cycle; and that turned out not to be true. So, I assume there are some aspects of finance that may turn out not to be true. Guest: Oh, absolutely. What I say to my students is: I'm showing you the stuff that people have done in the last 30 years, but in 20 years, it may all be irrelevant; so the best I can do is to train you about how to think about these things, so you can absorb stuff that comes along in the future that may overturn what's there now. That's what makes this profession fun, I think--the fact that stuff can get overturned. Russ: Of course, if we only have the illusion of understanding, or what Hayek called the pretense of knowledge, we could be doing some dangerous and stupid things under the guise of thinking we are making progress. Guest: Right. Russ: So, you do have to be careful. Where do you think in the near future finance is going? Guest: Oh, gee, I don't know. That's part of the fun of it. You just don't know. I wouldn't have been able to predict 30 years ago the stuff that evolved during those intervening 30 years. No way. Russ: It's kind of a random walk. Guest: I don't think it pays to think about it very much. There's so much serendipity in what happens in research. My best stuff has always been--I didn't start thinking about writing a great paper. I started thinking about a little problem; it just kept working in circles into a bigger problem. Or had offshoots that were related. I've beaten many topics to death, with the consequence I've got a lot of recognition; what started as a little thing developed into something much bigger. That's not a predictable process. Lots of little things end up as nothing. Russ: And? Guest: A student comes to me, a Ph.D. student, and says: I want to write a great paper. You can't start out to do that. You have to pick a problem and hope it works out into something that will get you a job, and hopefully a good one. But if you start saying: I want to come up with a great topic, you won't come up with anything. Russ: You recently wrote a very nice essay, "My Life in Finance," that gives an overview of some of your contributions and some of your thinking along the way and all those little problems. You started out by talking about your thesis topic, where you had five ideas and Merton Miller said four of them weren't very good. Guest: Right. Russ: Did you ever go back to any of those four? Guest: No, actually. Merton was incredible. He had a great eye for stuff that would work and wouldn't work. I went to Belgium for two years to teach, and I came back and showed him the stuff I'd been working on, and I think he discarded like 8 out of 10 things. He was right on all of them. Russ: Such is life. Guest: It taught me that nobody can work in a vacuum. You really need colleagues around you to enrich your work. You get credit for it in the end, but there are a lot of inputs from other people that go into it in the meantime.

Comments and Sharing



TWITTER: Follow Russ Roberts @EconTalker

COMMENTS (44 to date)
rhhardin writes:

Roger Pinkham, in a signal processing course in the mid 60s, remarked on methods of stock price cycle filtering, that you could determine the filter they used from the cycles they spotted.

Which is to say that the input they're finding their cycles in is white noise, and the filter itself generates the cycles.

My own favorite book on the subject was Burton Malkeil's A Random Walk Down Wall Street, in the 70s, which apparently is still around. I don't know what "the latest investment advice" could be that's touted on the front cover, though.

Possibly it was put there by a publishing drone.

Abe writes:

I was surprised by how much I enjoyed this podcast. Fama was earnest. He's critical of Macroeconomists for storytelling and - to his credit - really refuses to participate in the ex post analysis.

Sometimes I wonder what Fama's life must be like. The efficient markets hypothesis is incredibly important... and yet the whole world (and especially the asset allocation business) shuns him.

rhhardin writes:

Controlling interest rates vs controlling money supply:

Money is wealth to an individual but isn't wealth to a country. It's a ticket in line to say what the economy does next, presumably something for you. The Fed regulates the number of tickets so that they don't bid against each other for more than the economy can do, nor are there so few that the economy goes idle.

That keeps the economy busy and keeps prices stable.

The Fed in normal times looks at leading indicators of inflation and decides whether there's too many tickets or too few. If too many, is soaks some up by selling debt; if too few, it creates some by buying back debt.

The Fed doesn't care what the number of tickets is, just that it's too many or too few. It changes the amount by changing the trigger for buying or selling debt, which is the short interest rate. It changes it a tiny bit in the direction that leading indicators of inflation suggest, and then waits until the next meeting to see what to do next.

Two remarks: the interest rate is an output from the economy, not an input from the Fed. The Fed uses that output to say whether to buy or sell debt so as to regulate the money supply as they've decided, either a little more or a little less. This method distributes the giving or taking of money over the economy in a minimally invasive way.

Second, the leading indicators of inflation used have to be orthogonal to interest rate, so that the Fed does not blind itself with its own interventions. This is not hard to arrange mathematically.

In unusual times, namely today, the Fed is pretty completely blind as to what it's doing. None if its indicators are clear of its own interventions.

PS: they pay interest on reserves because the money is there as a confidence builder for creditors of the respective banks, and they want the money to stay there.

David writes:

My theory as to why these savy/educated investors bought these junk financial instruments is that by doing so they were to extract a significant enough windfall through immediate compensation that they frankly didn't care that eventually everything would collapse. Simple incentives, nothing more complex.

Steve writes:

Russ, please expand your ditch digger joke.

You tell it as something self evident, yet it assumes a larger context that I think is unjustified. Why isn't ditch digging "productive" or "good?" I also agree it is not, but I think there is a lesson that is not being drawn, or I misunderstand the point. I am not sure what story you would tell, but I think you would say something about how the demand for ditches is already filled. Any extra production would be waste. Of course there are people who dig and fill ditches, so it is valuable to some degree. So the question comes down to the fundamental issue you two explicitly skirted during the interview, what causes demand? Can you imagine any scenario for which it would be good and productive for us to stimulate ditch digging and ditch filling?

I know you have a problem with external/government subsidizing of ANY industry. But there must be more than that or else the issue of which industry wouldn't be relevant.

I have a funny imagined case, what if ditches were our currency? It would be a bad currency in that the transaction cost of its use would be very high. That is a fuzzy concept though. What is 'too high?' Disadvantaged relative to another competing economy? Obviously if a large part of our economy was devoted to filling and digging ditches the transaction cost would become progressively lower. And perhaps ditches would eventually become miniaturized or abstracted to electronic 'ditches.' That ditch industry would be then exactly as beneficial as our the analogous part of our actual financial system, (minus the actually useful irrigation ditches, ect.)

We could also just have a society that loved making or owning ditches, like we have one that loves hub caps and green lawns. In any case, I may be wrong, but I think you would have to appeal to an understanding of 'human nature' and/or 'market nature' to explain the absurdity of an actual ditch digging work program. Neither human nature nor market nature seems self evident or uncontroversial. To the degree you could unpack the assumptions behind the example, you might be able to apply it in a more constructive way. Although I am not positive how successful unpacking the joke would actually be...

"Russ: Oh, no; he said, when pressed and he was asked: If you ask people to dig ditches and fill them back in, would that stimulate the economy? And he said: Yes; but it's not as good as doing something productive. I can't explain it. It's a mystery to me. Guest: It's a mystery to me, too. "

Greg G writes:

The Efficient Markets Hypothesis is the ultimate example of how no bad idea ever dies in economics. No doubt it is true that market prices are based on all available information.....and all available emotion and irrationality as well.

The fact that you cannot predict future market prices does not vindicate the efficiency of today's prices. Nor does it prove that "successive changes are independent of one another". Market prices are one of many phenomena that are unpredictable due to sensitive dependence on initial conditions.

The Efficient Markets Hypothesis is one big exercise in assuming its' own conclusions and defining away all counter arguments.

Matt S. writes:

Good interview. However it would've been interesting to hear Fama's response to a challenge of the EMH on Hayekian grounds. I'm thinking of Hayek's "Use of Knowledge in Society", in which he argues that information is decentralized. I think Hayek would have hated the idea that information could be perfectly assimilated without trades taking place, as Fama indicated when he said the bid/ask spread could just change smoothly in response to new information. I mean, if that's the case then why couldn't a central planner just utilize information in the exact same way?

Also, please do get Stiglitz if you can Russ. I actually think he's probably right about the ditch digging, though of course the amount of stimulus would be very small. Regardless it'd be very interesting to hear a discussion along these lines.

AngryKrugman writes:

@David

Yes, to me, this seems like a simple case of principal-agent problems at work.

Schepp writes:

I was thinking that investment managers do make a difference. Maybe small at the margins but with no one watching the market no information gets shared to set index prices. The Free riding of investors, quite similar to me that simply let their employer recommended mutual funds ride, provide opportunity to skilled investment managers. Of course the average return is the index rate, but how many market entrances and exits are controlled by the equilibrium how many indexed sheep are in the market.

William B writes:

I feel like Fama's reasoning is completely illogical. He argues in circles. Also, why does he say that savings equals lending? That doesn't seem to be true in a fractional reserve system.

Lawrence writes:

Great podcast. However, I would like to suggest that you missed an opportunity to compare EMH with BF and the irrational behavior associated with bubbles. I understand Fama to say that the existence of irrational behavior does not disprove EMH, so long as it is random irrational behavior. But is it? If irrational behavior is systematic, does it provide a profit opportunity? If opportunities exist, are they based on publically available information or are they open to manipulation? Or are they simply unpredictable as Keynes argued in his well-known quote. I would really like to hear a discussion on the apparent conflict between these ideas. (BTW, I graduated from U of C 30 years ago and took the class based on Fama's "Foundations in Finance". A masterpiece, to be sure, but even Newton's laws had to be revised.)

Greg writes:

Anyone else a bit surprised that Fama seemed more Austrian than expected? I was...

Overall, I really enjoyed this podcast. Fama's work is overwhelmingly technical so I wasn't sure he would translate to podcast well. I thought he was great. I found him consistent, reasoned, and very insightful.

They've all been very good of late - but this is one of my favorites from the last six months or so.

Mike G. writes:

@William B (this piggybacks on what you said)

I have a question about the identity that Fama states: total savings must equal total investment.

Is that identity consistent with the assertion that banks "create money" by making loans?

It would seem that a given amount of savings could support a larger amount of investment because the loans made for investment purposes are backed only by reserves and the banks' promises to redeem savings deposits, not by the full amount of the deposits themselves.

Therefore, it would seem that the two propositions mentioned above are inconsistent.

Chambana writes:

Problem with Fama’s view of the world is that he must defend EHF hypothesis in all-or-nothing fashion. And that is fine, as long as one does not disqualifies behavioral folks in ad hominem fashion. E.g., Thaler is just a psychologist working in finance or, behavioral researchers are just looking for anomalies, and therefore, such research is illegitimate!?

Yet, Fama fails to tell us, what exactly is wrong with ‘fishing’ for anomalies (as a starting point of one’s research)? And secondly, if you disqualify one’s empirical research stream based on the argument that they are fishing for statistical significance on favored coefficients, could not we also disqualify Fama’s empirical research by the same logic? In my mind, any research must stand the scrutiny of replicability. To the best of my knowledge, I am still waiting to discover a research paper(s) that disqualifies entire field that shows systematic failures in human judgment. I guess I am just looking for a more convincing explanation on why behavioral finance is an utter failure.

Lastly, I was expecting an unbiased approach to this interview from Russ. I recently listened to an earlier interview with Thaler, which was somewhat hostile. Although, in his own words, Russ is not well versed in finance, he still felt comfortable challenging Thaler on all issues, while Fama’s words were taken at their face value.

Even if Russ agrees with Fama (who is undoubtedly a towering figure), I expected that he would be a bit more of a devil’s advocate.

P.S. Here is another puzzle: If all financial management (hedge funds as such) is just a dumb luck and fund managers make money by charging high fees until their luck runs out, could not we consider naive investments in high fees funds a form of a behavioral anomaly?

A. Zarkov writes:

How does Fama explain the apparent success of Renaissance Technologies Medallion Fund? The financial press says the fund has had an average return of 35% per year since 1989. James Simon, a mathematician, started Renaissance Technologies in 1982. He hires physicists, mathematicians, statisticians ... , but few or no economists, finance people or MBAs (a policy I think wise). Simon does what Fama says can't be done. Simon evidently finds information from past data to make successful predictions. From Wikipedia,

For the 11 years ending in December 1999, Medallion’s cumulative returns were 2,478.6 percent. Among all offshore funds over that same period, according to the database run by hedge fund observer Antoine Bernheim, the next-best performer was George Soros’ Quantum Fund, with a 1,710.1 percent return.[11] A measurement of the risk (e.g., beta, volatility, or leverage figures) which accompanied Medallion's high annual returns is not publicly available. In 2009 the Medallion fund topped the list of the most profitable hedge funds with profits of over $1 billion.

Renaissance Technologies seems to provide a counter example to Fama's assertions about efficient markets. It has outperformed the market for so long, it can't be chance, or a survivor effect. Perhaps Renaissance is actually functioning as a market maker or exploiting very short time correlations.

Then we have the work of Edward O. Thorp who has found inefficiencies and exploited them for profit for an extended period. Again a counter example to Fama. I think Fama and other advocates of EMH deal with these counter examples by ignoring them, or coming up with not very convincing explanations.

Russ Roberts writes:

A. Zarkov,

Do you know the story of Bill Miller and Legg Mason? Miller was also a genius who beat the market year in and year out. Until he didn't. One data point (or even two or three) in a very large ocean doesn't tell you very much. When the ocean is very large, there will always be unusual winners and losers.

Howard writes:

2 thoughts/questions:
(1) Why no discussion of Limits to Arbitrage... isn't that at the heart of any EMH critique?

(2) If we take the Fama/EMH position as true, what is the implication for optimal tax policy on investment managers? If all returns in that space are due to "luck/statistical anomaly", is there any real impact if tax policy changes incentives?

A. Zarkov writes:

Russ,

Legg Mason Capital Management Value Trust (LMVTX) "beat" the S&P 500 stock index every year from 1991 to 2005. "Beating" means exceeding, and that's a lot different from achieving a consistently very high absolute return from 1989 to 2011. Moreover, I don't think LMVTX beat every index.

There are about 10,000 mutual funds (more if you count the ones that went out of business), while that's big pool, it's not an ocean. An ocean would be millions or tens of millions of mutual funds. Not only that, strictly speaking, the Medallion Fund is a hedge fund, not a mutual fund like LMVTX. The hedge fund universe was much smaller (until recently). In 1984 Tremont Partners could only identify 68 hedge funds. After the 1990s, the hedge fund industry took off. In 1998, according to data from BarclayHedge, hedge funds had $143 billion in assets under management, and by 2010 they had $1.694 trillion under management. So until recently the hedge fund universe was more like a shallow pond than an ocean.

I'm well aware of the fact that by chance alone a small number of players in a large universe will beat the average. But I think the probability of achieving yearly returns of about 30%, year after year for more than 20 years is just too small not to refute the EMH. That probability is tricky to calculate, but I think if we do it, you will see that Renaissance Technologies, Soros, D.E. Shaw and a few other super performers can't be explained by pure luck.

The common element among these super performers is the use of mathematics, and perhaps advanced hardware (for high frequency trading). I also doubt they make their money on pure stock picking. I think some of them amount to a market maker. However, I don't think Soros uses mathematics, I think he has inside information.

Years ago I did accept the EMH, but as I learned more about finance and investing, I came to reject it. To be sure markets are very efficient and it's hard to make high returns without getting lucky, but not impossible. There are just too many counter examples to the EMH. The there is Thorpe, which is another big story.

DougT writes:

Zarkov:

Without risk information, returns are meaningless. If someone is taking outsized risk, they can claim anything about their returns.

Renaissance is indeed legendary. So let's call it skill. His 5/50 fee structure fits with Fama's hypothesis, though: that the managers capture a lot of the excess returns via fees. Perhaps he's so far to the right in the distribution that there are some crumbs left over for investors, but consider the risk-adjusted returns they enjoy versus Simon.

Consistent outperformance is rare, but it does happen. Usually the excess return is garnered via fees, but not always. The behavioral anomaly of rational investors using active funds is like people buying lottery tickets: hey, they might get lucky.

Charlie writes:

A. Zarkov,

Hedge funds complicate the matter with leverage. Leverage can be used to increase the expected average return of the fund by adding kurtosis (a higher chance the fund will suffer a large loss). For instance, Long Term Capital Management had several impressive years and then "blew up." Thus, looking at average returns can be misleading. What seems like a streak of impossibly high returns, could just be reflected in an unrealized tail.

Unfortunately, hedge fund data is sparse. For instance, if I try to find the sources the wikipedia article used, it leads back to a puff piece on Simmons with no explanation of where the 35% figure came from. When good data has been found, it has mostly not found large returns to management. The data also finds evidence of skewness (a tendency to blow up).

It's certainly possible that some people have very good, very secret strategies that create abnormally high risk adjusted average returns. If someone could show that empirically in an academic paper, there would be an extremely high reward. A puff piece in the WSJ does not meet that burden though.

Charlie writes:

Abe

"Sometimes I wonder what Fama's life must be like. The efficient markets hypothesis is incredibly important... and yet the whole world (and especially the asset allocation business) shuns him."

I wouldn't worry too much about Fama. He enjoys extremely high social status as an academic. He is probably the most revered financial economist. Second, there is a whole mountain of academic research on asset allocation built around the EMH. The simple premise is that in a world with EMH an investor should think about how his individual risk profile differs from the risk profile of the representative investor. In practice, it would mean a good investment adviser would spend a lot of time finding out about their client. What's their risk tolerance? Does there wage fluctuate heavily with the business cycle (a top surgeon's probably fluctuates a lot less than a successful small business owner)? Are they old or young? That, in practice, probably isn't the focus as much as it should be, but my sense is that there is much more focus on those sorts of things than 20 years ago.

RGV writes:

Fama comes out well in the podcast. But this wasn't Russ' finest attempt at getting to the depth of the issues. No mention of counter examples (Rennaisance/Buffet etc.) or Taleb's critique of finance. What happened Russ? You barely challenged the guy.

Sri writes:

So is Fama implying the fund managers who provide above average return are either lucky or acting on insider information quickly before it is publicly available?

I think when Fama meant loans equals to savings, thereby net leverage is zero in fractional reserve system- he means the drop in exchange rate takes care of large sum of money created by the fractional banking system of the lender country.

I disagree with Fama that buyers of CDOs had all the information at hand. People who created the CDO had all the information, but still kept making them because the CDS against them where backed by AIG and like, whom they knew will bailed out by the government.

I also disagree with him on his assertion that how did the bond market, housing market & commodities all fail roughly at the same time. All of them were feeding the same beast which was consumption created by leveraging increasing house prices.

Lawrence writes:

A. Zarkov,

A minor correction in your math. The odds of a mutual fund beating the average is 2^(-14) or one in 16,000 approximately. I don't have exact figures, but I think that the number of mutual funds that have existed during those 14 years is well in excess of 16,000 (including those that went bust, as you must do). In other words, it is expected that at least one mutual fund will beat the average every year for 14 years by chance alone. Just sayin.

John Berg writes:

A delightful podcast: made me laugh out load several times.

As a former editor of a peer reviewed journal, was pleased to learn that getting published is as important as ever and may ever determine the course of research trends.

Will audit again in order to make more substantive comments.

John Berg

A. Zarkov writes:

Lawrence,

The math is more complicated. Miller's LMVTX fund started in 1982. The Wikipedia entry for Bill Miller says the fund beat the S&P 500 index for 15 consecutive years from 1991 to 2005. Let's assume a constant universe of mutual funds, meaning there is a fixed number and they all stay in business. Let's suppose all the funds have a 50% chance of beating the S&P 500 in any one year. So we have 24 years in operation (as of 2005) and he had a run of successes of length 15. What's the probability of 1 or more runs of length 15 in 24 independent trials? This is a little complicated, but the solution is given here. A "run" is an unbroken series of successes. The answer is 0.000167847, which is over five times 1/2^15. Thus the probability of at least one fund out of 10,000 having a run of length 15 or more is 1-(1-0.000167847)^10,000= 0.81. If you like 16,000 better, then we get .93.

The real world is much more complicated. Funds with a run of bad luck are often closed. The fund manager will then open a new fund with a different name, and with much the same investment strategy. This "survivor bias" effect is pretty strong. Nevertheless we can see from this toy example that with a big universe of funds a few managers will get lucky and have an impressive looking run like Bill Miller. Today LMVTX has been in business for 29 years. Now the probability of one or more runs of length 15 about doubles to 0.000244141, and that 81% becomes 91%.

The assumption of independence is also not quite right. I suspect from the name of the fund LMVTX started out as a value fund. Today Morningstar lists LMVTX as a "blend fund." Sometimes market conditions favor value funds and they do well. They are also more risky so you would expect then to do better than the S&P 500. I suspect Miller got lucky by choosing a value style of investing at the right time and enjoyed a string of successes.

Russ provided us with a good example, but not one that really supports the EMH.

A. Zarkov writes:

Charlie,

Increasing the kurtosis should not affect the mean of a symmetrical distribution. With increased leverage a fund will have a higher chance of a big loss as well as a big gain. I doubt Renaissance Technologies success can be explained by high leverage. Eventually it wipes you out as LTCM found out as well as the banking industry in 2008-2009. Now if the government steps and rescues the firms the firms in the lower tail, that changes everything. I don't think Renaissance Technologies has ever been bailed out.

A. Zarkov writes:

DougT,

Again if Renaissance Technologies is simply running a high risk strategy they should have gotten wiped out by now. According to Wikipedia, that 35% average annual return since 1989 is after fees, so if that's correct, there is more than "crumbs" left over for the investors in the fund. What we don't know is their internal rate of return. They might have had high returns with a small amount of assets under management, and low returns with a large amount of assets under management. In which case, the IRR could be way below 35%. It gets harder and harder to get high returns when the fund gets bigger and bigger. Moreover, they do more than simply pick stocks with the Medallion fund; they run a trading desk. They might be earning large fees as a market maker of some kind, which juices up their profits.

Since the hedge fund universe was small when they started in the early 1980s, and didn't get big until after 2000, their survival and possible high returns might serve as a counter example to the EMH. But I do agree that we need more information.

A. Zarkov writes:

Slight correction on my reply to Lawrence. I wrote, "What's the probability of 1 or more runs of length 15 in 24 independent trials?" I should have written, What's the probability of a run of length 15 or more in 24 independent trials?

muirgeo writes:

Ah... more of the same. When reality is destroying your life's work go for a stalemate and hide behind incomprehensible complexity claiming no one could no the truth. Problem is I've been following this debate since 2005 and people are on record all over the Internet predicting the bubble and crash.

We have a major global financial crisis that basically has the world economy set up as a casino and all Prof Fama can due is provide more cover for the elites gathered at Davos who love the complexity argument.

The economics profession has failed modern society more than any other.

Mike G. writes:

Sometimes the best conversations aren't debates. It can be useful to go along and give someone the benefit of the doubt and just get that person's views out there.

Silent Troubadour writes:

Fama is defensive at his best, most of the time being just dishonest.

On one hand he states that prices incorporate all the market information for making decisions(supplementary information being costly) , yet on the other hand when faced by Russ with the bubble argument, he quickly accuses supply and demand for price dysfunctions and deviates to erroneous facts about the timing of various market crashes.


There is a documentary available on Youtube and Google videos called 'Quants: The Alchemists of Wall Street (Marije Meerman, VPRO Backlight 2010)', in which one of the actively involved quants in the risk analysis software implementation during the boom period, expresses his doubts on whether the buyers of CDSs really knew what they were buying.

The question is: could they have really known, since all the securities were AAA rated and nicely packaged?

In other words, EMH is an interesting piece of Chicago-based corporation-sponsored pseudo-science, which clearly doesn't find its place in the REAL world.

Hudson Cashdan writes:

Thanks Russ- excellent podcast as usual. Four paths I wished you went down that I would have liked to hear Fama counter:

1-Half of me understands and believes in Fama’s EMH but the other half of me wonders what would happen if everyone believed in it. Put simply: isn’t the accurate pricing of assets the direct result of so many people laboring to discover mis-priced assets and arbitrage them back to fair market value? What would the result be if everyone assumed asset prices were right and passively invested in an index? What if everyone owned an index and there was just two pieces of relevant news- one good, one bad- that cancelled out; e.g., if news meant RIMM is worth 50% less and AAPL 50% more, how would those prices adjust if everyone assumed efficient markets?

2-OPM. A large portion of the financial industry is managing other people’s money. Within this there are different types of institutions managing money for different types of clients operating under different incentive structures. What was the incentive of the pension fund manager getting $200K/yr to manage a defined benefit program for the teachers of Springfield? Perhaps his motivation was to keep his job by performing about in-line with the other pension funds. Or maybe the plan he managed was severely under-funded and he had to reach for yield; and maybe he was only allowed to invest in AAA assets (rated by one of the three rating agencies, obviously). For this pension fund manager, managing career risk was more important than managing to maximize his client’s economic interest. Economists respect prices because they convey a message about the scarcity/utility of an item; but don’t you think a warped incentive structure can influence prices in a manner that creates false price signals? And, if so, wouldn’t one get very wealth spotting the mis-pricings and correcting them via the financial markets? And for that they would earn above average returns…

3-There is also regulatory arbitrage to consider. I already touched on the ratings cartel (Moodys, S&P, Fitch) that is enabled by government laws and regulations. Additionally, European banks were holding large amounts of Italian sovereign bonds with leverage because Basel III said they are to be treated as a risk-free asset and thus no capital must be held against them. This despite the fact that the bond traders down the hall from the risk managers were trading the same Italian bonds- which were trading 5%+ above Germans- as if they were as risky as Junk. The whole concept of a risk-free asset underpins our financial system and is, quite frankly, preposterous. Nothing is risk-free. Nothing. Not even US Treasuries. Incidentally, if/when the risk-free status of Treasuries comes into question by the market the move in yields will be like a damn breaking….and everyone will claim to have predicted it but only a few will get rich by having spotted the inefficiency.

4-Fama is backwards w/respect to leverage- banks do not need savings to lend; the lending comes first and the savings follows. That lending becomes the bank’s asset and another party’s liability. As far as that lending is not backed by an asset it is indeed a problem. Furthermore, if the assets and liabilities are not properly accounted for- as seems to be the case with a lot banks holding liabilities off balance sheet- then it is also a problem. The aggregation of these problems may lead to mis-priced assets that people smarter and more connected than I tend to discover.

Sigh writes:

[Comment removed pending confirmation of email address and for rudeness. Email the webmaster@econlib.org to request restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Jennifer Yoon writes:

Thank you, Professor Roberts, for a vey entertaining and informative podcast, as always. And please join LinkedIn.org Friends of EconTalk user group.

To other commenters,
I think Roberts was not confrontational because he agreed with many of Fama's views, and because most of the time Fama was saying he did not know the correct solution. It's tough to be confrontational when your guest keeps saying he doesn't know.

To me this "I don't know" position is not unusual for Professor Fama. When he presents his empirical papers, he presents many of the experiments he ran that did not pan out, although he presents only the major ones (else the paper will never end).

The EMH as applied to mutual fund performance says we cannot know whether a fund's top performance was due to managers' skill or luck. It may have been skill, or it may have been luck. Our tools in statistics does not allow us to determine the cause.

Also all mutual fund companies are notorious for merging in good performing funds to boost performance and closing bad performing funds. And if the manger changed during a funds life, the entire period cannot be used. Otherwise, we are assuming that a twin took over the fund and that there has not been a material change. I don't remember where I read this, but I recall that after adjusting for fund mergers and manager turnover, only Buffet's Birkshire Hathaways had some indication at a statistically significant level that the Fund's return may be due to managers' skill. Think about it. How many funds did not change its managers or merger with a sister funds and had a stellar return for close to 60 years?

Fama co-founded and is a director at the Dimensional Fund. It does not believe in inefficient market information, so it uses market microstructures that may lead to certain segments of the market to trade at a disadvantaged price for a while, perhaps years, until the market segment is revised (sometimes due to regulation, sometime due to mechanics of trade execution). I think the Dimensional Fund has done well over a long period (15+ years?) but I don't think Fama is that actively involved in the daily investment decisions. You can get better information directly from http://www.google.com/search?q=Dimension%20Fund .

I probably have many errors here since I didn't check any of this before posting, so please take with a grain of salt. ;-)

George writes:

Great podcast. Always a pleasure listening to Eugene Fama. A couple of things:

1) A link to "The Arithmetic of Active Management" by William Sharpe might be relevant and helpful.

2) For a future podcast on some of these topics (and a bit contra-Fama), a great guest could be Emanuel Derman, who recently wrote Models.Behaving.Badly. Derman was a physicist in academia who became a leading quant at GS. In his book, Derman covers philosophy, economics (Hayek), finance, and physics in examining how economic/financial models should -- or should not -- be applied in real life.

Confused writes:

Can someone explain Fama's statement that savings have to equal loans. I fail to see why- given fractional reserve banking- this is true.

Blades writes:

You guys probably know a lot more about probability than do. However, I think that, Prof. Fama engaged in a rookie mistake when he described a recent study he did on returns to fund managers. Essentially, he analyzed the data and found a bell curve. Then he looked at the outliers and said that the number of outliers were no greater than would be expected by chance. In other words, consistent good performers were there, but not at levels greater than you would expect from chance. Same with consistent bad performance.
This seems like a misinterpretation of probability to me. A bell curve simply describes results. It says nothing about why the outliers are where they are. I think it's perfectly reasonable to assume that financial returns would be in a normal distribution. Most things are. However, I believe that the reason that some funds consistently perform well is not purely because of chance, but because of a better strategy or better information, which is found either through greater resources or hard work or access.
It sounds like if professor Fama saw a distribution of height, he would say that people are tall by chance. Well, no, some of the reason that people are tall is because of their genetic inheritance.
Perhaps I am being too simplistic, but I don't think that Prof. Fama’s argument should convince anyone that success in finance is purely a matter of luck.

LowcountryJoe writes:

I thought Friedman was only an advocate for substantially increasing the money supply during bank runs. And that his rationale was to make it painful for people to be sitting on cash while they panic; providing a disincentive to hoard cash and an incentive to at least earn a real return by keeping loanable fund on deposit.

RWood writes:

I enjoyed the podcast with Dr. Fama. But I felt that I understood a smaller percentage of it than any other Econtalk discussion.

I'd like to see you have him back just for a discussion of the efficient markets hypothesis.

Frank Howland writes:

Fama is obviously very smart and says intelligent things about empirical research. However, I was mystified by his claims that high leverage can't be an important part of the story and that the recession caused asset prices to decline (as opposed to the story about asset price declines leading to collapse of prices of mortgage backed securities, causing a financial crisis, which in turn led to a recession). A recent NBER Working Paper by Gorton and Metrick which is scheduled to appear as a review article in the Journal of Economic Literature includes this quote: "Reinhart and Rogoff (2011) and Schularick and Taylor (2012), are the products of Herculean data collection efforts on long historical time series about government and private debt. Both of these papers demonstrate the strong association between accelerations in economy‐wide leverage and subsequent banking crises. That finding deserves emphasis as the main empirical fact about historical predicates to financial crises." It seems that these authors are using a different definition of leverage from Fama's.

See:
Reinhart, Carmen and Kenneth Rogoff (2011), “From Financial Crash to Debt Crisis,” American Economic Review 101, 1676‐1706.
Schularick, Moritz and Alan M. Taylor (2012), “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles and Financial Crises, 1870‐2008,” American Economic Review, forthcoming

Jason writes:

BERK AND GREEN, read it

http://finance.martinsewell.com/fund-performance/BerkGreen2004.pdf

One of the best recent papers in Finance. Lack of persistence in mutual funds does not imply efficiency in stocks, but just efficiency of the mutual fund industry itself. The gains from trade of skill goes to the people that possess the skill through management fees.

weird professor writes:

Interestingly, Fama's interpretation of the EMH appears to have given up the claim that prices are "right", and fallen back on the claim that - on average - people cannot beat the market.

Given that - on average - poeple ARE the market, the usefullness of the EMH seems rather limited.

Dr. Duru writes:

I am relieved to see there are several people here mystified by Fama's claim that lending must equal savings. I sensed Professor Roberts had to hesitate on that. I would LOVE to know whether he has done further thinking on this.

I am confused because it seems in a world where the Fed prints money that is then used to buy bonds, lending does not have to equal savings at all. Unless our understanding of "savings" is too limited?

Comments for this podcast episode have been closed
Return to top