On the Performance of Naïve, Analyst and Composite Earnings Forecasts: Evidence from Hong Kong

Abstract

AbstractIn this paper, we compare the information content and performance of naïve, analyst and composite forecasts in Hong Kong. Empirical evidence shows that superior performance can be obtained by a composite measure combining both analyst and naïve forecasts. In addition, analyst forecasts become more conditionally efficient over the naïve model as the actual announcement approaches. The superiority and timing advantage of analyst forecasts suggest that more emphasis should be placed on the services of analysts for predicting future earnings figures, particularly when the announcement is approaching.