0

Beyond forecasting

A trader takes a break from the floor outside the New York Stock Exchange May 17, 2006 in New York City. Stocks fell sharply in afternoon trading following a stronger-than-expected rise in consumer inflation.

TEXT OF STORY

MARK AUSTIN THOMAS: You know we told you earlier about that revised GDP report that's coming out today. We're always talking about what analysts predicted. Do you ever wonder who those people are? How they make their predictions? What happens when they're wrong? Marketplace's Amy Scott looked into it.

AMY SCOTT: When the latest inflation report came out last week, the markets tanked. Analysts had under-estimated how high prices had risen in the month of April. But they were off by just one-tenth of a percent. Economist Nariman Behravesh says it was enough to catch investors way off guard. The Dow dropped almost two percent that day.

NARIMAN BEHRAVESH: The consensus up until maybe a couple weeks ago was that inflation and interest rates wouldn't rise that much. In fact some in the market had convinced themselves that the Fed was done hiking rates, and then they were very unpleasantly surprised.

This sort of thing happens all the time. Just before the government releases new unemployment figures, or IBM reports its quarterly earnings, an army of analysts and economists guess what those numbers will be. Seems harmless enough. But when the actual numbers beat or fall short of those predictions, it can set off a whole chain reaction in the markets. And it seems forecasters are rarely spot on. Ernie Ankrim with Russell Investment Group says he's happy if he's right even 60 percent of the time.

ERNIE ANKRIM: But you know they build really big casinos in Las Vegas out of games that only pay out 54 percent of the house. So, what we hope to do is make only those bets where we think the odds are substantially in our favor and still recognize that we're probably be wrong a third of the time at least.

It could be worse. UC Berkeley psychology professor Phil Tetlock wrote a book about political forecasting. He found that the average "expert" was no better than the proverbial dart-throwing chimpanzee at guessing the fall of a government, or the end of apartheid. Economists tend to fare slightly better. They have hard data and fancy computer models at their disposal. And Tetlock says large corporations and governments can't afford not to listen.

PHIL TELOCK: Even a small improvement on chance can have multi-billion dollar consequences. It's also considered part of due diligence, for senior executives in business and government to consult with experts. And if they fail to consult with experts and things go wrong, they're in a lot of trouble.

Tetlock has some advice for how to judge the economic forecasts you hear. He says the more famous a person is, the less likely he or she is to be right. That's because the kind of people who get on television or appear in the paper tend to make simpler, more passionate arguments. Thus the saying in the business, "often wrong, never in doubt." Tetlock says the more sure someone is about a prediction, the more skeptical you and I should be.

In New York, I'm Amy Scott for Marketplace.

About the author

Amy Scott is Marketplace’s education correspondent covering the K-12 and higher education beats, as well as general business and economic stories.
Log in to post0 Comments
With Generous Support From...