Comparison of Cross-Sectional Earnings Forecasting Models for the European Market

Abstract

This paper aims to compare the cross-sectional models of earnings forecasts to the analysts’ earnings forecasts in the European markets. We introduce the models of Fama and French (2006), Novy-Marx (2013), and Hou et al. (2012). The target of the forecasting methods is to minimize the bias and to have a high forecasting accuracy in comparison to real earnings. Our results show that for almost all regions, the cross-sectional models of earnings forecast have a lower bias, but they are not as accurate as the analyst’s earnings forecast. In addition, the outcomes also show that cross-sectional models of earnings forecasts tend to be more accurate for long-term earnings forecasts, whereas for short-term forecasts analysts’ earnings forecasts are more reliable. The best cross-sectional earnings forecasts are calculated with the model of Novy-Marx (2013) in all the observed regions. Finally, the bias results show that even the cross-sectional forecasts are over-optimistic.