Still Missing the Mark

Brokerage house strategists make a big deal out of announcing their annual projections for earnings of the S&P500. Sometimes firms trumpet the fact that the estimates are "bottoms up," that is put together by sectors, relying on the expertise of analysts following the companies in a sector. Based on these magical numbers, and a wave of the hand to calculate "normalized P/Es" comes the pronouncement that the market is undervalued and the recommendation to mortgage the farm to purchase the firm's Buy List. It was always hard enough to analyze and value individual companies, but put it together into projecting an index and a Ouija board would have been a better tool.

Recent work by the McKinsey Corporate Finance Practice shows that in every year from 1985-2008, the actual S&P 500 earnings number was significantly lower than the earliest estimate for the year, and the estimates migrated only slowly towards the actual number.

Over the past 25 years, annual EPS growth estimates have ranged from 10-12% per year versus actual growth of 6%. This excludes a period of extreme estimate giddiness from 1998-2001!

The McKinsey researchers note that the November 2009 implied market P/E of 14 projected an earnings growth rate of 5%, which is in line with the growth of nominal GDP. Over the long run, earnings can't deviate too far from this growth rate, excluding post-recession snapbacks of earnings.

Executives at firms whose earnings projections are allowed to wander off into fantasy land should hold themselves accountable for not reining in estimates. They should also focus on clearly articulating their business models and on shaping their disclosures to allow capital markets actors to do better at consistently estimating earnings.