Template-Type: ReDIF-Paper 1.0
Series: Tinbergen Institute Discussion Papers
Creation-Date: 2015-03-06
Number: 15-032/III
Author-Name: Bert de Bruijn
Author-Workplace-Name: Erasmus University Rotterdam, the Netherlands
Author-Name: Philip Hans Franses
Author-Workplace-Name: Erasmus University Rotterdam, the Netherlands
Title: How Informative are the Unpredictable Components of Earnings Forecasts?
Abstract: An analysis of about 300000 earnings forecasts, created by 18000 individual forecasters for earnings of over 300 S&P listed firms, shows that these forecasts are predictable to a large extent using a statistical model that includes publicly available information. When we focus on the unpredictable components, which may be viewed as the personal expertise of the earnings forecasters, we see that small adjustments to the model forecasts lead to more forecast accuracy. Based on past track records, it is possible to predict the future track record of individual forecasters.
Classification-JEL: G17, G24, M41
Keywords: Earnings Forecasts, Earnings Announcements, Financial Markets, Financial Analysts
File-Url: https://papers.tinbergen.nl/15032.pdf
File-Format: application/pdf
File-Size: 336945 bytes
Handle: RePEc:tin:wpaper:20150032