Contextual Fundamental Analysis in the Prediction of Extreme Returns
2001, Review of Accounting Studies
Messod Daniel Beneish, Charles M. C. Lee, Robin L. Tarpley
This study examines the usefulness of contextual fundamental analysis for the prediction of extreme stock returns. Specifically, we use a two-stage approach to predict firms that are about to experience an extreme (up or down) price movement in the next quarter. In the first stage, we define the context for analysis by identifying extreme performers; in the second stage we develop a context-specific forecasting model to separate winners from losers. We show that extreme performers share many common market-related attributes, and that the incremental forecasting power of accounting variables with respect to future returns increases after controlling for these attributes. Collectively, these results illustrate the usefulness of conducting fundamental analysis in context.
Beneish, Messod D., Charles M.C. Lee, and Robin L.Tarpley (2001), "Contextual Fundamental Analysis in the Prediction of Extreme Returns," The Review of Accounting Studies, Vol. 2/3, pp. 165-191.