CONSENSUS EARNINGS FORECASTS FOR LOSS-MAKING FIRMS

Abstract

The purpose of this research was to examine the accuracy of consensus earnings forecasts for loss-making firms and non-loss-making firms listed on the JSE Securities Exchange South Africa (JSE) and to test whether systematic differences existed between the two groups. Other key areas studied were whether analysts:  incorporated time series information in their forecasts;  could differentiate between the temporary and permanent components of earnings shocks and assess the level of earnings persistence. Systematic differences were found in the forecast error between loss-making firms and non-loss-making firms. Evidence was found that analysts may not be using past forecast errors and past stock returns when setting their forecasts. Other evidence found was that analysts appeared to be unable to differentiate between the temporary and permanent components of earnings shocks or to assess the level of earnings persistence in firms.