Newspaper article THE JOURNAL RECORD

Why Are Forecasts by Top Stock Analysts Faring So Poorly?

Newspaper article THE JOURNAL RECORD

Why Are Forecasts by Top Stock Analysts Faring So Poorly?

Article excerpt

By Susan Antilla

N.Y. Times News Service

NEW YORK _ Wall Street analysts can earn as much as $1 million a year, and superstars among them can demand _ and get _ their own private session to interrogate a company's top management about business strategy.

So why aren't these powerful analysts better at predicting earnings of the companies they follow?

From the looks of the secondarter results of the nation's publicly held companies, even the heavy hitters are falling short at anticipating the fates of the companies they cover. And that can wreak havoc with stock prices and cause more than a little discomfort to investors who make decisions based on those predictions.

Earnings surprises "are the single biggest driver of stocks," said Daniel Benton, the "Allerica" analyst at Goldman, Sachs Co. who covers Compaq Computer.

Shares of a number of wellown companies soared or plunged in the recent reporting period as those companies announced earnings that took analysts by surprise.

Cray Research's shares, for example, dropped $4.625, to $23.875, after the company said it earned 5 cents a share for the quarter, less than half the 13-cent average of analysts' predictions. Kodak's shares gained $1.125, to $43.875, after that company said its earnings were better than analysts had thought.

And on Friday, shares of Hewlettckard Co. plunged $12.625 to $58.875, responding to the company's estimate Thursday evening of earnings for the quarter that ended July 31. The company said earnings would come in at 76 cents to 87 cents a share; analysts had been expecting $1.12.

It is a tricky business divining the dollars and cents of a corporation's fate on a quarterly basis. But this job, in addition to writing research reports and making buy and sell recommendations on stocks, is a chief reason why analysts get paid so well.

Of 1,950 companies that had reported earnings through Aug. 3, only 12 percent came in at exactly the consensus average of analysts' predictions, said Ben Zachs of Zachs Investment Research. Fifty percent came in better than expected, he added, and 38 percent came in worse than the average prediction of the analysts covering them.

"There's no question that even on a quarterly basis, the forecasts are missing actual results more than in the past," said David Dreman, a money manager at Dreman Value Management.

Dreman conducted a study last year of earnings estimates of 1,221 stocks followed by at least six analysts, and found a steady erosion of analysts' accuracy from 1973 to 1990.

By 1990, the average earnings estimate was either 65 percent too high or 65 percent too low. In 1973, the first year of his study, the average estimate was off by 31 percent from actual earnings. The trend to greater inaccuracy continues, Dreman said.

Why can't Wall Street pros equipped with a wealth of research tools and M.B.A. degrees from the best training grounds in the country hit the mark?

For one thing, while they are called research analysts, they have surprisingly little time to do research. This year in particular, securities analysts spent an inordinate amount of time working with their firms' corporate finance departments.

When the initial public offering market is hot, as it was earlier this year, analysts help smooth the way for colleagues in corporate finance to meet key executives at companies in the industries they follow _ a helpful ingredient in snaring a client. Analysts also make themselves available as a brain trust for corporate finance executives on the lookout for new prospects.

But even without the unusual demands of initial public offering work, analysts already had a full plate, with decreasing room for research. The "morning call" that is broadcast over firms' squawkx systems regularly requires analysts to spout their latest views on companies; after that, the sales force wants a report in writing to show to customers. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.