CNET también está disponible en español.

Ir a español

Don't show this again

Tech Industry

The Starting Line: More to companies than earnings estimates

Stocks often rise or fall based on whether companies miss or beat analyst forecasts. Investors should take a broader view.

    Don't base everything on analysts' earnings estimates.

    Earnings tracking firm First Call might be the most powerful company on Wall Street when it comes to short-term shifts in stock prices. Your favorite company can double earnings, triple sales and predict a quadrupling of profits and revenue in the coming year, and yet the stock might be crushed if the its latest profits fall short of First Call's consensus estimates. Your favorite short-selling target might call for skyrocketing losses during the next six months, but the stock could soar if revenue races past the firm's forecasts.

    For instance, Nokia plummeted 19 percent Tuesday when it warned of results in the near-term that would be lower than what First Call analysts expected.

    First Call numbers are the closest thing to a benchmark for corporate performance. But the numbers aren't always right and don't indicate as much as one might think.

    "You see those situations where the sell-off is overdone, but that's the nature of things," said Chuck Hill, First Call's director of research. "There a lot of people who say 'Well, I don't want to take the risk that this could turn out to be a long-term problem.' It's a judgment call, and people are going to make mistakes on occasion."

    A study of stocks in the late 1990s and early 2000 indicated that during the three-day period from the session before an earnings announcement to the session after, a company that tops the First Call forecast will see its stock outpace the Standard & Poor's 500 index by 0.6 percent for every penny of earnings beyond the consensus prediction, according to Mark Bagnoli, an associate professor at Purdue University. Companies that report lower-than-expected earnings underperform the S&P 500 by the same ratio, Bagnoli said.

    Bagnoli and Susan Watts, also a professor at Purdue, have been studying earnings estimates and their correlation to stock prices in the past few years. Their current project involves First Call's revenue estimates. The project isn't finished yet, but they said it is obvious that the stock prices of companies that report losses (rather than earnings) are affected not by the posted loss, but by reported revenue vs. analyst expectations, Bagnoli said. If a money-losing company generates more revenue than First Call predicted, the stock generally will rise.

    The fact isn't surprising, given that certain traditional measures of a stock price, such as price-to-earnings ratio, are impossible without earnings. But revenue growth and ratios such as price-to-sales offer some gauge.

    Incomplete picture
    Unfortunately, consensus estimates don't give a complete picture of a company, author Michael Thomsett said.

    "A lot of times, calling the estimates is sort of like predicting the weather," said Thomsett, whose books offer advice on how to filter information from Wall Street. "Trying to invest on the basis of how earnings estimates come out is very short-term thinking in some respect. I think it's sort of a game that a lot of investors play, but I wonder what that really tells you about a company. It says nothing about the fundamentals."

    Companies often manipulate earnings estimates, Thomsett said. Accounting rules can let companies shift revenue between quarters to some extent, and all companies try to "guide" analysts to publish estimates that are easy to surpass.

    "There's so much a company can do to control that," Thomsett said.

    Any Wall Street research firm that advises fund managers employs a bevy of analysts who predict quarterly sales and earnings of publicly traded companies of their choice, though the market forces some of the stocks on them.

    Practically every semiconductor analyst on Wall Street, for instance, tries to forecast earnings for Intel, the world's biggest source of PC processors. Nearly all communications equipment analysts track profits for Cisco Systems, the largest vendor of routers and switches for corporate networks.

    First Call polls analysts for each public stock and averages their earnings and revenue projections to produce consensus forecasts. Technically, it's not the consensus estimate; other organizations, such as Zack's Investment Research and IBES International, compile their own consensus estimates. Thomson Financial Securities Data recently bought IBES and plans to merge it with First Call, although the companies have separate estimates for now.

    Yet First Call's product for years has been Wall Street's established standard for analyst forecasts. First Call's Hill will tell you his company gained an early lead because it was the first company to provide something close to real-time estimates. "We had a time advantage back in the beginning," he said. "That's since diminished. Now it's minimal."

    Investors beware
    Hill believes First Call's real strength lies in its ability to get analysts to use the same standard for each company. Publicly traded companies in recent years have made a habit out of excluding certain items--such as restructuring charges and goodwill write-downs--from "cash earnings" reports. The trick is getting all the analysts to agree on what items can be left out of earnings estimates.

    First Call simply learns what standard the majority of analysts are using and compels the rest to go along. The company screens the forecasts to make sure all analysts are on the same page, but that doesn't always stop mistakes from coming through. Last week alone saw least two errors.

    When Loudcloud released quarterly results Tuesday, some news reports declared that the company missed estimates. It turned out that one analyst left out the effects of Loudcloud's deferred compensation, something that was included by the other research firms following the stock. Loudcloud actually beat estimates.

    Adobe Systemd told analysts to look for third-quarter revenue growth to remain the same year over year, which would appear to be better than what analysts expected; First Call, at the time, listed a third-quarter revenue forecast of about $290 million, which would be down 11.5 percent from the year-ago period. However, that "consensus" number included a low estimate of $34.9 million--likely the result of a misplaced decimal point, analysts said. Analysts were actually looking for a slight increase in Adobe's third-quarter revenue, prior to the company's new projection.

    First Call didn't know whether the error was its own or the fault of an analyst. "In either case, we should have caught it," Hill said.

    The point is moot, since Adobe analysts sent in new revenue estimates Friday, after the company's conference call. But regardless of whether estimates are accurately recorded, investors would be better served to ignore them, Thomsett said.

    "I don't know that analysts have been right enough of the time," Thomsett said. "You have to wonder what they're basing their estimates on."

    Loudcloud and Adobe are two examples of stocks that moved against the trend outlined in the Purdue study. Loudcloud topped the consensus prediction, but shares fell the next day as analysts issued a slew of downgrades. Adobe warned of lower-than-expected results, yet share rose, because the company's market position remains solid.

    Thomsett looks at investing from a long-term point of view. Purdue's study indicates a correlation between a company's performance against estimates and its near-term stock performance. But even Bagnoli wouldn't rely solely on analyst forecasts. "Revenue forecasts have gotten more accurate and significantly less biased," the Purdue associate professor said. "But they're by no means the whole story."