Analysts' forecast error: a robust prediction model and its short‐term trading profitability

This paper contributes to the empirical evidence on the investment horizon salient to trading based on predicting the error in analysts' earnings forecasts. An econometric framework is proposed that accommodates the stylized fact of extreme values in the forecast error series. We find that between 1998 and 2010, the strategy of taking a long (short) position in stocks with the most pessimistic (optimistic) I/B/E/S forecast has an annual risk‐adjusted return of 16.56 per cent before transaction costs. The robust method used to predict this pessimism (optimism) and the one‐week investment horizon are the key drivers of the strategy's profitability.