Crowds Can Outsmart Pros' Market Forecasts
By SPENCER JAKAB
A DOW JONES NEWSWIRES COLUMN

NEW YORK -- Investors talk a lot about the "madness of crowds" when explaining irrational episodes like the dot-com bubble, but they give far too little credit to the wisdom of crowds.

That's starting to change as Wall Street comes to the realization that, for many important questions, groups of less-informed novices are able to make better predictions than highly paid experts or even the consensus of several experts. This doesn't mean that economists and strategists are becoming obsolete, but it may change the way investors view their advice. One of those forecasters, Legg Mason strategist Michael Mauboussin, posed the question in a note to clients last month of whether these concepts can help an investor beat the market.

"Our answer is an unmitigated yes, for lots of reasons," he wrote. "To begin, the wisdom of crowds appears to be a viable and robust way to explain market behavior."

Crowds Are Mad But Wise Too

Business texts are full of cautionary tales about the disastrous consequences when previously sober executives ignored conventional wisdom in order to buy into the collective wisdom, such as when Time Warner's (TWX) bosses rushed to exchange stock in their valuable media empire for high-flying AOL six years ago. With such fiascoes in mind, it's easy to dismiss a recent poll by the Los Angeles Times showing that 60% of Americans expect a recession this year. Considering that not a single economist out of a sample of 30 expects even a single quarter of negative growth in 2007, it requires a leap of faith to lend much credence to a survey from a population in which only a small minority understand the basics of macroeconomics.

Despite the reluctance to lend much or any credence to the predictions of the unwashed masses on something as tough to predict as economic growth, Wall Street would be wise to pay attention. This is the sort of question that crowds are well-equipped to answer, even if the 60% probability seems high.

"I think it would be stupid to take literally, but there's information in this survey," said George Mason University economics professor Robin Hanson. "In general, asking ordinary people to forecast things isn't a good idea, but it is important because these people are part of the economy. It's also information about their mood, which is independently influential."

Markets Beat Surveys

Surveys tend to be less accurate than so-called "decision markets," Hanson's area of expertise. These involve wagering and setting odds, much the same way that the point spread in a football game or the odds on a certain horse are a function of many bettors' views. Someone placing a $100 bet will influence the odds more than a less-confident gambler wagering $10, whereas both would have equal weight in a random survey. Betting actual money also weeds out less-savvy participants over time.

"In their heart of hearts, they know how much they know," said Hanson.

An investor placing zero odds on a recession in 2007 because 100% of 30 experts said so might be making a mistake. A better approximation comes from a source like Irish betting site Intrade, which puts the odds at 17.6% based on cash trades, down from as high as 28% in mid-March.

Whether it is in small-scale experiments like students in Mauboussin's classes at Columbia Business School guessing the winners of the Oscars or large-scale wagers like the Iowa Electronic Marketplace used to predict presidential election winners, the track record of large crowds is better than almost any individual for certain types of problems.

A number of financial institutions have begun futures auctions both as profit centers and to predict periodic indicators. For example, Goldman Sachs (GS) has organized auctions before data releases like nonfarm payrolls and futures broker ICAP has had great success with a weekly auction to determine natural gas inventory changes, usually beating consensus surveys. Some companies even use them internally for predicting important information like future revenues or prices of components.

Diversity Is Good, Homogeneity Bad

If the odds of a recession are almost one in five and were recently seen as close to one in three, why didn't even one of the 30 economists go out on a limb to make such a prediction? For that matter, why do stocks with unanimous buy ratings from Wall Street analysts fail to outperform those on which they are more pessimistic? Hanson explains that, when someone is in the forecasting business, sticking with the crowd is the best career move.

"A known bias is that everyone has the same forecast and you go with it and, if you're wrong, you suffer less than if you stuck your neck out and were wrong," he said.

Indeed, the last time the Los Angeles Times survey showed 60% odds of a recession seven years ago one ensued later in the year, but Wall Street economists had failed to predict it.

Mauboussin cites an example from Scott Page's book "The Difference" that explains why even a relatively uninformed group can beat individual experts. In the TV show "Who Wants to be a Millionaire," the audience tends to be able to answer a multiple-choice question correctly even when only a small percentage know the exact answer and another small group knows enough to eliminate one or two of the wrong answers. Random errors from the majority who don't know cancel out, leaving the right answer as the one chosen by the highest percentage. This is the same principle used by herds of animals or schools of fish in nature in choosing the right course most of the time.

Turning to predictions about the economy or the stock market, Mauboussin cites the importance of "cognitive diversity." Collective predictions by individuals with a wide variation in views tend to do much better than when there is a broad consensus that a stock or the market will go a certain way, such as during speculative manias. He also notes that, even though behavioral finance has shown that investors often make dumb decisions on their own, a group of people with these human emotions of excessive fear or greed should do well overall as long as they are not in the middle of a panic or a euphoric period.

Mauboussin's conclusion is that markets are mostly efficient, making it unlikely that even savvy investors can outsmart the market during normal times. The exception is when there is too much consensus such as in a major bull or bear market, allowing the crowd to be very wrong.

"The logic of diversity shows that, in solving hard problems, the crowd is almost always going to be better than most people if diversity, aggregation and incentives are operative," he wrote. "Investors aware of this reality will remain humble, while seeking occasions when the crowd's wisdom gives way to whim, and hence opportunity."

(Spencer Jakab, a columnist who provides insightful and unique takes on the stock market, previously wrote about the energy industry).

--- By Spencer Jakab, Dow Jones Newswires, 201-938-2429; spencer.jakab@dowjones.com

 
URL for this article:
http://www.djtopstories.com/article/0,,SB117734296931161607,00.html