TY - JOUR T1 - Trading Strategies Based on Analyst Estimate Revision Clusters and Associated Corporate Information Events JF - The Journal of Investing SP - 32 LP - 42 DO - 10.3905/joi.2006.616842 VL - 15 IS - 1 AU - Mark Bagnoli AU - Stanley Levine AU - Susan G. Watts Y1 - 2006/02/28 UR - https://pm-research.com/content/15/1/32.abstract N2 - This paper examines trading strategies based on clusters of analysts' earnings estimate revisions and their triggering corporate information events (CIEs). We find that the profitability of long and short positions depends on the nature of triggering event as well as the sign of price movements at the beginning of the cluster. Long positions based on revision clusters associated with earnings and stand-alone guidance CIEs result in meaningful average excess returns. Short positions based on revision clusters associated with earnings result in small but statistically significant excess returns. Short positions in revision clusters associated with stand-alone guidance and positions in clusters associated with other types of corporate information events (e.g., corporate presentations, IPO filings, M&A, and other joint venture or strategic alliance announcements) generally do not yield positive excess returns. This indicates that there is value to matching revision clusters and triggering events when developing trading strategies based on revision clusters.TOPICS: In markets, big data/machine learning ER -