Why AMZN hit 95 today:
Why Do Investors Overpay? It's All in the Game
By Steven Pearlstein
Washington Post Staff Writer
Thursday, April 2, 1998; Page C01
With the stock market again hovering near record levels, everyone is wondering whether prices reflect a reasonable expectation about future corporate profits or merely a speculative bubble that could burst at any time.
The Trendlines staff has a policy of never giving market advice. But some interesting experiments done over the years show that investors appear to have strong tendencies to get caught up in a speculative fervor and pay much more for stocks than they are really worth.
A recent example is a game devised by Sheryl Ball of Virginia Polytechnic Institute and Charles Holt of the University of Virginia for some undergraduate classes.
In the game's little market, each trader starts with a few shares of stock in an imaginary company, plus a small amount of cash -- say, $20. The traders buy and sell shares over 10 rounds of trading, and a $1 dividend is distributed to the holder of each share at the end of each round. But there is a hitch: After the dividends are distributed, there is a one-in-six chance that any share will suddenly become worthless, as determined by the roll of a die. At the end of the game, players get $6 for each active share they hold.
When Ball and Holt run their experiment, as shown in the chart at the right, the actual trading price almost always rises well above the underlying value of the shares -- which should be $6 in any round. (Pencils ready? That's $1 for the dividend, plus a five-out-of-six chance of cashing in the $6 at the end.) Only in the final rounds does reality begin to sink in. Suddenly, the price plunges back to near its fundamental level. As a result, most players end the game with a portfolio worth less than when they started.
Such experiments cast some doubt on the widely held assumption among economists that people generally act rationally and markets do a good job of pricing things. Here is case of a simple stock market in which the usual uncertainties about risks and profits have been removed and all traders have access to the same information -- and even then, speculative bubbles almost always develop.
By the way, it's not just undergraduates who behave this way. At the University of Arizona, professor Vernon Smith has been running variations on the same game for a decade. He says he got similar results when he used graduate students, business executives and even a group of commodity traders flown in from Chicago.
According to these three experimental economists, one reason the traders are willing to overpay for the stock is that they don't take the time to figure out the fundamental value or they simply can't do the calculation correctly. But even when participants are invited back to play the game a second time and have the value calculations done for them, a speculative bubble still develops, albeit a smaller one. Even forewarned, it seems, traders can't help but get caught up in the excitement.
At Arizona, for example, Smith found that it is only with the third try at the game that most traders finally adopt the safe strategy of buying whenever the price is under $6 and selling whenever it is over. When that happens, the price settles in permanently at $6 and trading effectively stops.
It was most famously John Maynard Keynes who commented on the tendency of stock markets to develop speculative bubbles. Back in 1936, Keynes observed that many traders are concerned not with what a stock is really worth but only with whether there will be another trader, caught up in the market's herd psychology, who is willing to pay more. Among Keynes's insights was the role such behavior plays in the booms and busts that classical economics could never quite explain.
Fifty years later, there is a small cottage industry armed with powerful computers and doctorates in mathematics trying to figure out whether there is a predictable pattern to the irrational behavior of traders. After all, if you could come up with a formula that could predict at what point the bubble usually bursts, then the "correct" trading strategy would be to continue trading in the market even after prices rise above the fundamental value.
But the problem of such exercises is that once other people figure out the same formula -- or at least figure out that you have one and begin to copy your behavior -- then the formula quickly loses its magic. If everyone knows the precise point at which to become a seller, then who will be a buyer?
In the end, it is this dynamic quality to markets -- the fact that I can adjust my behavior in anticipation of your behavior, but you could then adjust yours in anticipation of mine -- that makes it impossible for anyone to say when it is time to sell out of a bull market. And that's true even when you feel certain that stocks are trading at speculative levels.
The broader lesson from these experiments seems to be that traders aren't as perfectly logical as economic models would predict. But neither are they stupid. They learn from past experience and are constantly adjusting their strategies to incorporate that new knowledge.
c Copyright 1998 The Washington Post Company
|
