[Excerpt] Security valuation techniques have become increasingly sophisticated, as evidenced by advances in option pricing theory and Arbitrage Pricing Theory (APT). Despite the availability of these paradigms, the price/earnings (P/E) approach to security valuation has maintained its popularity among practicing security analysts. Much of this lasting popularity can be attributed to the apparent simplicity of the P/E approach and the difficulties inherent in implementing complex valuation models. To implement the P/E approach, analysts must estimate the appropriate P/E multiple, which typically requires the development of a model. Unfortunately, most of the P/E models that have been developed use ad hoc empirical tests and historical data. One notable exception is a study by Cragg and Malkiel, which tested a theoretical P/E model using analysts' forecasts for a nonrandom sample of 175 large firms. This note extends the work of Cragg and Malkiel in two ways. First, it is not restricted to large firms; all firms with complete data are included in the empirical tests. Second, the model includes some factors not considered by other researchers.