Performance of forward-looking value drivers in stock screening: making automated recommendations based on future forecasts

Abstract

PURPOSE OF THE STUDY The purpose of this study is to analyze different value drivers and their capability of explaining future stock returns. The value drivers are evaluated based on their performance using backward- and forward-looking data. The forward-looking value drivers are tested both with perfect foresight, i.e. actual future fundamentals, and with analyst forecasts. The main objective is to find the best value drivers that can be used to create automatic stock recommendations based on future forecasts. Finally, the performance of these auto-recommendations is compared against the performance of analysts’ own consensus stock recommendations.

DATA In this study, the US stock market data is used as the primary source of data. The data is gathered from Compustat and CRSP databases and it covers years 1975-2007. In total, the sample consists of 98,688 company-year combinations. The average number of stocks in a single year is 2,990. This data is complemented with equity analyst forecasts and recommendations gathered from I/B/E/S database.

RESULTS The main finding of this study is that the P/E ratio and to some extent the PEG ratio are good forward-looking value drivers that provide significant stock screening potential when applied to earnings forecasts of good quality. However, when applying analyst consensus forecasts these value drivers do no result in significantly better performance than using only backward-looking value drivers with historical financial figures. Despite the relatively low performance in absolute terms the auto-recommendations made with the value driver approach outperform significantly the consensus recommendations of the analysts. As a conclusion, investors and analysts would be better off making recommendations based on systematic value drivers rather than the current ad-hoc methods.