TY - JOUR T1 - Equity Return Performance from a Prediction Model of Meeting or Missing Analysts’ Forecasts JF - The Journal of Investing SP - 52 LP - 66 DO - 10.3905/JOI.2010.19.1.052 VL - 19 IS - 1 AU - Lynn Rees Y1 - 2010/02/28 UR - https://pm-research.com/content/19/1/52.abstract N2 - This article develops a trading strategy based on a prediction model as to whether firms will meet their analyst forecasts for the next two quarters. The author shows that a predictionmodel applied to quarterly data in a recent time period is profitable. In fact, hedge returns are positive for every calendar quarter from the third quarter of 2003 to the last quarter of 2007. Strikingly, the average return magnitude for the entire sample is about 17-23%, depending on the return metric used and the investment horizon (three or six months). This magnitude is similar to a return strategy with perfect foresight as to whether the firm will meet future forecasts, although the model is not perfect in its predictions. Further analysis reveals that this result is due to firms realizing significantly greater returns in the higher-probability deciles relative to the lower deciles, regardless of whether the firm ultimately met its future forecasts. Thus, it appears that the model provides information as to not only the likelihood of the firm meeting its earnings targets, but also how the market is likely to interpret this event at the time it occurs.TOPICS: Exchanges/markets/clearinghouses, performance measurement, statistical methods ER -