E/P Ratios v. Changes in Earnings in Forecasting Future Price Changes

Abstract

THE PURPOSE of this paper is to describe the results of an extensive empirical investigation of the relationship of earnings/price and earnings change ratios to subsequent stock price performance. The results allow a comparison of the usefulness of two basically different variables for predicting future price changes: a fundamental variable where earnings are considered relative to price, and a trend variable where earnings movements are considered without regard to price. Both deseasonalized and non-deseasonalized (unadjusted) earnings per share are employed. The original purpose of the paper was to see if an existed for quarterly data similar to that reported on earlier by the authors for annual data.' The annual data studies had shown that future period earnings (for example, relative to current price) were more valuable as predictors of future price changes than current period earnings which were in turn more valuable than previous period earnings. In fact, the latter variable-the only one known to the investor when he had to make his decision-was found to be of little value in selecting stocks for above average price appreciation. It was hypothesized that this same information effect would be found by employing leads and lags in the quarterly data, although more favorable results from known (previous quarter) earnings information were hoped for. The results were more striking than were anticipated.