As Wall Street's top strategists publish their 2013 stock market forecasts, Gerard Minack is mulling over Wall Street's embarassingly poor track record of predicting earnings, a key driver of stocks.
Minack, Morgan Stanley's Head of Global Developed Markets, notes that "consensus forecasts almost always imply that 2-3 years ahead economic and earnings growth will be at trend."
More from Minack:
As they say, forecasting is difficult – particularly the future. Consensus earnings forecasts show just how difficult: Exhibit 1 illustrates consensus EPS forecasts for the S&P 500 from 1985, with the initial forecast for each year indexed to 100.
Here's that chart. The farther the colored line strays from 100, the more inaccurate the initial forecast:
Here's Minack's take on earnings forecasts as seen in this chart:
First, they are almost always wrong – almost always because they are too high. The initial forecast has been too low only twice in the history of this series; otherwise, it was too high. On average, the forecasts have been 61⁄2% too high one year ahead (that is, actual EPS were on average 61⁄2% below the EPS forecast a year earlier). The average error two years ahead has been 111⁄4%.
Second, forecasts for growth rates are also typically too high, with a larger error than for EPS levels. On average, forecasts have implied 151⁄2% EPS growth over the next 12 months; on average, EPS have risen 71⁄2%.
Third, revisions to two-year-ahead EPS largely reflect the error to current year forecasts. Exhibit 2 shows revisions to the two-year-ahead EPS forecast and the error in the current year forecast. To illustrate: EPS over the year to September 2012 were 5.5% lower than the forecast made a year ahead (in September 2011). As that error appeared, the two-year-ahead EPS forecast (earnings for the year to September 2013) were cut by 7.7% between September 2011 and September 2012.
Here's that second chart:
Please follow Money Game on Twitter and Facebook.
Join the conversation about this story »
