IS the glass half full or half empty? For several thousand analysts who make a living assessing the value of publicly traded stocks, the answer depends on which week it happens to be.

That’s what a study of the Standard Poor’s 500 earnings cycle by Thomson Reuters/I/B/E/S shows.

The distant future sometimes looks better than the mundane day-to-day of the moment. For stock analysts who crunch numbers to come up with earnings estimates for individual companies, the far horizon is often just one year down the road. And the numbers show that when analysts estimate quarterly earnings a year in advance, they tend to be unrealistically optimistic about the prospects of companies they cover, according to Greg Harrison, the senior research analyst at Thomson Reuters who did the study.

Mr. Harrison’s day job involves compiling consensus earnings estimates for the overall stock market, figures he derives from the collective appraisals published by thousands of individual analysts. Does the market expect earnings for the S. P. 500-stock index to rise or fall, and by how much? Have companies met expectations for the quarter, or will they disappoint the market? Some of the answers come from the data he gathers each week.

But while doing his work, he noticed a consistent pattern in the numbers, which he describes in a fascinating study of earnings since 2008, titled “Estimates Too High, Low? Check the Calendar.”

Except for several quarters in the Great Recession, he found, early earnings estimates are generally rosy, and become predictably and progressively gloomier as time goes on. As analysts revise estimates downward, it becomes easier for companies to beat the market consensus, creating what Wall Street calls a “positive earnings surprise” roughly two-thirds of the time.

Positive surprises, of course, are good for share prices. Negative surprises are not. And by being optimistic about the long-term future, and relatively pessimistic about immediate results, the quarterly cycle of stock market earnings estimates has the effect of bolstering the market.

Analysts, of course, are encouraged in this practice by corporate executives who routinely issue warnings — “guidance,” in Wall Street parlance — that their companies won’t really meet the analysts’ lofty targets. The analysts respond by lowering their targets.

Typically, Mr. Harrison finds, analysts are most accurate — neither too optimistic nor too pessimistic — about seven weeks before companies actually release earnings.

“That’s when analysts’ estimates and the eventual, real numbers meet,” Mr. Harrison said. But the analysts don’t leave well enough alone. Instead, they keep cutting their forecasts and end up being gloomier than reality warrants. “By the time earnings season actually ends,” he says, “it turns out that the analysts have been too pessimistic — and we end up with a lot of ‘earnings surprises.’ ”

We’re now near the end of the earnings season for the second quarter. Most big companies have already issued their final numbers for the period, and the current pattern fits the overall picture fairly well, Mr. Harrison says.

On July 2, 2012, for example, when he compiled the first market consensus for the second quarter of this year, analysts as a group were projecting great things for stocks one year ahead. They said earnings would grow at the blistering pace of 14.4 percent in the second quarter of 2013.

Reality hasn’t come close to matching that early optimism — but because analysts repeatedly ratcheted their projections downward, earnings reports have been surpassing the relatively pessimistic estimates of recent weeks.

On Friday, with 462 members of the S. P. 500 reporting, Mr. Harrison found that the actual growth rate so far has been only 4.9 percent. Yet 67 percent of those companies beat the analysts’ estimates, producing positive surprises. How was that possible? Analysts collectively dropped their estimates for the quarter to only 2.9 percent on July 1, when earnings season began. The actual results were much better than that.

The rough pattern held for many major companies. Consider General Motors. On July 2, 2012, analysts covering G.M. estimated that it would have earnings per share of $1.25 in the second quarter of 2013. That July, the estimate of Ryan Brinkman, an analyst at J.P. Morgan, was $1.12. G.M. has “best-in-class leverage to global growth markets, ongoing operational turnaround, and improving product cadence,” he wrote.

G.M. has had problems, however. The company acknowledged that it was doing poorly in Europe. By mid-April, after a report that industrywide sales there had plummeted to a 30-year low, and that G.M.’s Opel brand was lagging, analysts’ estimates fell to 73 cents a share. Mr. Brinkman’s was 72 cents. For all analysts, they stood at 74 cents at the beginning of July and edged up to 75 cents the week of July 12.

But that was still way off the mark. G.M.’s actual earnings, released on July 25, were 84 cents a share. Although earnings declined compared with a year earlier, news coverage generally treated the announcement as a positive surprise.

Mark Bradshaw, an accounting professor at Boston College, says what we are seeing is probably overconfidence by analysts and deft maneuvering by corporate executives, who have leeway in adjusting accounting to improve reported profits and in choosing what information to reveal. “For companies, issuing ‘guidance’ has become an art form,” he said. “The analysts seem to try to do what they can, but they’re often at a loss.”

Aswath Damodaran, a finance professor at New York University, called the earnings season “a Kabuki dance” in which “analysts are trying to forecast; companies are trying on the other side, with accounting choices, to affect those earnings and to lower the forecasts; the companies watch the analysts; the analysts watch the earnings; and it’s all a big game. And it’s a game that the companies generally win.”

Frequent traders scrutinize these rituals, he said, seeking nuance. For them, he said, “it’s not enough now just to beat the earnings forecast. That’s too common. Now, you’ve got to beat the forecast enough — by a big-enough number that it really is a surprise — if you want to stir up the market.”

In his view, most of us would be better off ignoring short-term earnings reports. “None of this matters much to long-term investors,” he said. “It’s the long-term picture that’s important, and that is revealed eventually, even if it isn’t clear now.”

Is the glass half full or half empty? Don’t even try to figure that out during earnings season.

Article source: http://www.nytimes.com/2013/08/18/your-money/rosy-earnings-forecasts-at-least-at-first.html?partner=rss&emc=rss