Choose an application
Choose an application
Choose an application
Current research aims at uncovering empirically the link between the diversity of analysts' forecasts and the accuracy of the consensus forecast given by the average of the individual forecasts. James Surowiecki (2004) asserts that diversity in analysts' forecasts increases the consensus forecast accuracy. We test his assertion against the I/B/E/S database, which provides analysts' consensus forecasts. We find the exact opposite: The higher diversity, the larger the forecast error observed ex post. Having established that fact, we argue that diversity is an ex ante measure of estimation risk. We find that increasing the number of forecasts contained in the consensus further than 10 does not significantly improves the consensus forecast accuracy.
Choose an application
Choose an application
Choose an application
Choose an application
Choose an application
Choose an application
Choose an application