Consensus analysts' earnings forecasts and security returns

This study examines the information content and informational efficiency of consensus analysts' forecasts in an Australian setting. Consistent with Lys and Sohn (1990) and Abarbanelll (1991), analysts' forecasts were found to possess information content but did not incorporate all publicly available information. The empirical analysis in this paper suggests that negative security returns are associated with higher forecast errors. This contrasts with the results of Lys and Sohn (1990) and is consistent with a positive bias in forecasts. However, this did not preclude analysts' forecasts from being viable proxies for the market's expectations of future earnings.