Interesting article in The Economist "How companies massage their profits to beat market forecasts", some snippets:
Executives have every incentive to match or beat forecasts as the market punishes those that fail to do so. That, in turn, hurts the value of the share options which are the best hope of making the executives rich.
A recent academic paper looks in detail at this process (The Valuation Premium for a String of Positive Earnings Surprises: The Role of Earnings Manipulation by Jenny Chu, Patricia Dechow, Kai Wai Hui and Annika Yu Wang). As the authors point out, it is hard to know whether the ability of the corporate sector to beat forecasts is due to good management, a growing economy or outright manipulation. So they focus on companies that the SEC has identified as indulging in manipulation.
Sure enough they find that 53% of such firms have a record of four straight quarters of beating forecasts, compared with just 43% of all firms. Secondly, they find that firms tend to indulge in earnings manipulation when they already have a high stock market multiple; they are trying to prop up their share price, not inflate it. The average price-earnings ratio during the manipulation period is 35.
Third, they find that 42% of manipulating firms beat profits for eight quarters, compared with 32% of all firms. After that, the difference is not statistically significant. So two years seems to be the limit for cooking the books. And fourth, they find that executives focus more on beating forecasts than on beating last year's numbers.
The link to the research paper can be found here.