On December 3, 1996, Robert J. Shiller, the Stanley P. Resor Professor of Economics at Yale University, made a sobering presentation to the Board of Governors of the Federal Reserve System. The stock market, he warned, was priced beyond reason; valuation ratios had reached historical extremes. If history was a reliable guide, the market would be a poor investment over the next 10 years. Evidently, Shiller's testimony made a deep impression on Alan Greenspan. Two days later, the Fed chairman delivered the speech in which he coined the famous phrase, "irrational exuberance."
That was then, when the Dow Jones Industrial Average hovered around 6,400. This is now, and the Dow has skyrocketed, exceeding 11,200 in early September. Today, Greenspan isn't talking much aboutinvestor irrationality (in public, at least). Instead, he's cautiously optimistic about the productivity-boosting effects of information technology and the Internet, and he seems to lend support to the widespread notion that a new era of the economy is at hand. Professor Shiller, on the other hand, is more worried than ever, and earlier this year, he published a book on the state of the stock market. Its title? Irrational Exuberance (Princeton University Press).
Written for a nonacademic reader, the book examines the stock market boom from the point of view of a behavioral economist--someone who believes that describing how investors actually think and behave, and how various forces shape that thinking and behavior, can yield far more insight into the "messier aspects of market reality" than elegant mathematical models can. No devotee of efficient-markets theory, Shiller draws on the disciplines of psychology, sociology, demography, and history to explain why people invest the way they do--and how market values can soar to irrational levels.
Of course, not everyone thinks that the stock market is grossly overvalued. Indeed, many observers believe that the U.S. economy is poised on the cusp of an even longer boom, and that the greatest bull market in history will continue unabated for years to come. Is this rational optimism, or more irrational exuberance? Recently, CFO articles editor Edward Teach discussed this question with the 54-year-old Shiller in his office on the Yale campus in New Haven, Connecticut.
You write that today's stock market is "a speculative bubble of historic proportions." What's the evidence for this belief?
Well, I think the evidence is many-faceted. Calling something a speculative bubble is a judgment call.
The common view is that the market is the outcome of a vote among many people who are trying to guess at value, and therefore, since so many people have voted, it must be an accurate indicator of value. So, in order to understand whether that is a plausible argument, you have to look at the whole picture and what people are doing and what kind of information the typical person is using. And looking at the whole picture, it seemed obvious to me that it is not a vote of people carefully trying to evaluate the value of the market. Instead, there are social, historical, and institutional forces that shape their decisions.
Let's compare historical numbers. I'm struck by a couple of charts in your book. One displays the monthly price/earnings ratio of the S&P Composite Stock Price Index from 1881 to January 2000. The second-highest peak is September 1929, when the P/E reached 33.6. The highest peak, by far, is January 2000-- 44.3.
The other chart, a scatter diagram, suggests that the price/earnings ratio predicts 10-year returns. That is, the higher the P/E of the S&P Composite Index in January of a given year, the lower the ensuing 10-year real return of the index. The lower the P/E, the higher the 10-year returns.
Yes, this is a very important part of the argument because there is so much attention paid to the random-law theory. The random-law theory would imply that there should be nothing in that scatter, no sense of slope.
According to the scatter diagram, which slopes downward from left to right, the annualized 10-year real return following January 1929-- the farthest point along the x- axis--is practically zero. What does that suggest about the 10-year annualized return of the S&P in 2010?
That's a good question. I've fitted lines through this--I just did it with your business card--it comes out to about minus 10 or 15 percent a year. But I never felt comfortable about extrapolating that line out of samples so far. This is a fundamental dilemma; we're out of the historical range. And some people would say, "Well, that proves we're in a new era. History is irrelevant."
Maybe we are in a new era, thanks to the Internet. Investors seem to think so.
If you go through any decade in U.S. history--excluding maybe the war decades, World War I and World War II--there was always some thing that was a really important transformation. We have to remember that whenever a new technology appears, it fuels growth for a time, while it's being adopted, and then it becomes mature, and then it's done.
An example that I've been thinking about lately is the interstate highway system. In 1919, the U.S. Army decided to send a truck and military car convoy coast-to- coast as an experiment. It took them 62 days. This made a big impression on the young Eisenhower, and later, after he saw the Autobahn, he decided to build this highway system, which created 43,000 miles of highway. It changed the whole structure of the economy.
But, that's done. That was a very important boost to our economy, but the 1950s didn't have the same kind of boom. People could sense that we were in a new era, but it didn't have the same impact.
Well, the 1950s were a time when the market, especially earlier, was very low. The New York Stock Exchange did some surveys of individual investors to see why they were prejudiced against the market--that was how they thought, that the public was prejudiced against the stock market. One thing they found was great ignorance: people didn't even know what the stock market was. They couldn't really define what a stock was.
Now, if you read John Kenneth Galbraith's 1929 crash book, or Frederick Lewis Allen's Only Yesterday, they talk about a culture preoccupied by the market. And you might try to imagine what that culture is, but we've got a better example now. We're more preoccupied.
Still, the Internet revolution isn't done yet. Alan Greenspan, for one, has given a lot of credit for recent productivity growth to new information technology.
We don't have particularly high productivity growth. If you look at a plot of the productivity figures that the U.S. Department of Labor publishes, it's quite variable, and it was higher in the '50s or the '60s than it is now. It's really only in the last few years that we've seen a lot of productivity growth. And to say that we understand that and understand its connection to the Internet or something is ridiculous.
Lots of advances in technology haven't had this extreme effect on the market in the past, except maybe in the 1920s. They've had effects on productivity, they've had effects on earnings, but no consistent effects on the market. It seems to me that to understand these unusual stock market reactions, you have to look at culture.
In the 1920s, technological progress was exceptionally visible. One thing that was happening was radio. That really was a change, because you were living in your little house with no contact with the outside world, and by the late '20s, you're tuning in every night to comedy shows and celebrities. Also, that's the decade when most people got their first car. So now we're driving around, we've got a radio, we've got a refrigerator, and the electrification of the economy is essentially completed. The working TV was demonstrated. For many people, it looked like a new era.
I thought it was an article of faith among economists that markets are efficient. That, as you define it, "all financial prices accurately reflect all public information at all times." But you don't really believe this.
You believe there's a huge psychological, herd-thinking element involved in setting stock prices.
This has been a theme of mine for the past 20 years, and I think there's a lot of evidence that market psychology is very important. Remember, the idea that market psychology plays an important role is commonplace. It was the people who proposed the efficient-markets hypothesis who were proposing something dramatic. The burden of proof is on them, I think.
Some efficient-markets thinkers believe that market bubbles are rational, too.
Well, there is this rational-bubble literature, which I can't understand. We don't need rational bubbles--we've got irrational ones already! But the idea that markets are efficient is both attractive and ridiculous at the same time. It's a neat, clean theory, and if you get creative with it, you can explain anything.
Speaking of efficient- markets proponents, Burton Malkiel, the author of A Random Walk Down Wall Street, reviewed your book. He said that while you're to be commended for sounding the alarm about overvalued stocks, it isn't the entire market that's overvalued, just technology and Internet stocks. He claims many Old Economy stocks are undervalued. How do you respond to that?
My book isn't about which stocks to buy. The other side of it, though, is that the conventional wisdom that's been pushed by people-- even by Malkiel somewhat--is that the smart thing to do is diversify and hold stocks, a lot of stocks. They would tell you that there's no picking stocks.
So what seems to be the conclusion of Malkiel's book--I haven't seen the latest edition--is that you should hold a broad portfolio of stocks and shouldn't try to pick stocks. And then he comes at me by saying, "Well, but there are some stocks that are underpriced." Well, then should one try to pick stocks? I actually believe that one should probably try to pick stocks, though it's not going to make you rich overnight.
When this interview is published in October, it will be 13 years to the month that the New York Stock Exchange had its biggest one-day loss in history. By 1988, the market had recovered. Today it's over 11,000, even after a one-day drop of 618 points in April. Why shouldn't investors continue to think that if we are in a speculative bubble, there will just be another correction like 1987, the Fed will step in, and sooner or later, the market will go back up?
First of all, you're correct in describing how people interpret the '87 crash. People feel their confidence boosted by that event; it's the experience of hearing a doomsday scenario and finding out it wasn't so bad. There's a tendency to think the market's just going to keep on going up.
But it is a judgmental error to reach that conclusion, because the more this goes on, the more overpriced the market is. Now of course, the other side of it is, I don't know that the market isn't going to go up for a while. We've seen the irrational exuberance building.
Some critics, though, accuse you of trying to predict the market. They say that had investors sold their stocks in 1996, when you gave a pessimistic presentation to the Fed, they would have missed out on tremendous gains. The authors of Dow 36,000, which I think you know--
Oh yes, I know that book very well. [Laughs]
The authors contend that you're relying on broken valuation models--that we need new models to explain what's going on. They think the market is ready to skyrocket.
They're going way more out on a limb than I am. See, they're talking about a few years from now-- they said by 2003 it'll probably be [at 36,000], but they're a little vague. This is wild. That's a tripling. I'm not saying anything so dramatic. I'm saying that the market looks very high, so I would think that returns would be poor over the long haul.
The theme of their book is that the market is less risky than bonds, that there shouldn't be any risk premium, or virtually no risk premium, and that people will have finished learning this in three to five years. How do they know that people are going to finish learning this in three to five years?
Jeremy Siegel, in Stocks for the Long Run, has shown that over 30-year periods since the 1870s, stocks do outperform bonds.
They have. It's funny how people put that in the present tense- -and you just did!
It sounds to me like a risk worth taking. Why not invest in stocks for the long run?
Well, that's what I've done. In fact, for many years, [my investments were] 100 percent in stocks. But you have to remember that stock market returns were puny for the 20- year periods that followed the peaks of the price/earnings ratio in 1901, 1929, and 1966--under 2 percent.
Earlier, you said that maybe we're entering a new era. But now you're saying, "Well, it's done this for the last 100 years, so it has to continue doing that." Well, that's the fundamental tension. We don't know whether we're entering into something different or continuing the same. And what is the same? What kind of thing are we thinking that's continuing?
One way of putting it is that the 20th century was a great century for this country. If you go through history, different countries tend to dominate in different centuries. Now, we're embarking on a whole new century with lots of other countries aspiring to do what we did in the last century. It doesn't seem very sensible to say that we're sure to do exactly as well in [this century] as we did in [the last] one.
In sum, you believe that investors have lost a healthy respect for risk.
That's right. A lot of people are not saving very much, because they have this idea that, "I don't even have to save now; I can just wait a while, put it in the market, and the power of compound interest will make me a millionaire." It's a mistake.
You're often called a behavioral economist, along with academics like Andrei Shleifer and Richard Thaler. How do you define behavioral finance?
It's a discipline that approaches finance with the assumption that human psychology is important. And not just human psychology, but that also we have to look more at people and institutions to understand--not assume--what their objectives are.
Do your colleagues regard your work as out of the mainstream?
In some sense what I'm doing is mainstream economics, because economists like to think that history will repeat itself. They look at the historical trend. So when I say that real earnings growth has only been 2 percent a year for the last century up to 1990, and that that's a reasonable forecast for the next century, that sits quite well with academic economists. It's this new-era thinking that is more foreign to them.
Last question: What do you think the Dow will be on January 1, 2001?
This is a question I don't like to answer. I often repeat a statement I made last year. I said, "Let me talk about 20 years out, in 2020. That's an extremely hard thing to forecast. But I would say that as good a prediction as any is 10,000." That statement got quite a reaction, because people think, "That's ridiculous! How could the Dow be at 10,000 20 years from now? How could that possibly be?" But if you look at history, there have been many periods when the stock market has languished for 20 years. So, I made that statement just as a reasonable guess.
More reasonable than 36,000?
It could very well be at 36,000 in 20 years --but I doubt that it'd be there in 2003.