Standard & Poor’s Capital IQ research unit said it has developed a method of forecasting negative and positive earnings surprises of a magnitude of 5% or larger.

These, it said, will be forecast using an “intelligent estimate” on existing earnings estimates for stocks, which will in turn be weighted by:

  • Age of estimate, since newer estimates contain more information and represent the most complete forecast of what a company will report.
  • Broker size, since larger brokers tend to attract and retain analysts that provide more accurate forecasts.
  • Forecast horizon, because forecasts made closer to the company report date presumably will embed more information into the forecast.
  • Analyst tenure, since individuals covering stocks for a long time tend to provide more accurate forecasts.

"Intelligent Estimates provide an edge for investment professionals, particularly during the busy, high workload earnings season," said Carson Boneck, managing director, S&P Capital IQ. "By comparing the intelligent estimate to the traditional consensus estimate, investment professionals will be provided with a quick snap shot of where earnings surprises are likely to occur."
S&P Capital IQ said it expects the intelligent estimates to reduce forecast errors about 14 percent in the United States and 10 percent internationally.

In addition, it expects to predict 62 percent of earnings surprises greater than 5 percent in either direction in the United States and 58 percent abroad.

The S&P Capital IQ Estimates database covers more than 18,600 companies in more than 100 countries. History goes back to 1999 in the United States and 1996 internationally.

Here is the S&P Capital IQ Methodology for Forecasting Earnings Surprises

Tom Steinert-Threlkeld writes for Securities Technology Monitor.