T HE three possible hypotheses with respect to what an investor pays for when he acquires a share of common stock are that he is buying (i) both the dividends and the earnings, (2) the dividends, and (3) the earnings. It may be argued that most commonly he is buying the price at some future date, but if the future price will be related to the expected dividends and/or earnings on that date, we need not go beyond the three hypotheses stated. This paper will critically evaluate the hypotheses by deriving the relation among the variables that follows from each hypothesis and then testing the theories with cross-section sample data. That is, price, dividend, and earnings data for a sample of corporations as of a point in time will be used to test the relation among the variables predicted by each hypothesis. variation in price among common stocks is of considerable interest for the discovery of profitable investment opportunities, for the guidance of corporate financial policy, and for the understanding of the psychology of investment behavior.' Although one would expect that this interest would find expression in cross-section statistical studies, a search of the literature is unrewarding. Cross-section studies of a sort are used extensively by security analysts to arrive at buy and sell recommendations. values of certain attributes such as the dividend yield, growth in sales, and management ability are obtained and compared for two or more stocks. Then, by some weighting process, a conclusion is reached from this information that a stock is or is not an attractive buy at its current price.2 Graham and Dodd go so far as to state that stock prices should bear a specified relation to earnings and dividends, but they neither present nor cite data to support the generalization.3 distinguished theoretical book on investment value by J. B. Williams contains several chapters devoted to the application of the theory, but his empirical work is in the tradition of the investment analyst's approach.4 only study along the lines suggested here that is known to the writer is a recent one on bank stocks by David Durand.5 In contrast with the dearth of published studies the writer has encountered a number of unpublished cross-section regressions of stock prices on dividends, earnings, and sometimes other variables. In these the correlations were high, but the values of the regression coefficients and their variation among samples (different industries or different years) made the economic significance of the results so questionable that the investigators were persuaded to abandon their studies. There is reason to believe that the unsatisfactory nature of the findings is due in large measure to the inadequacy of the theory employed in interpreting the model, and it is hoped that this paper will contribute to a more effective use of cross-section stock price studies by presenting what might be called the elementary theory of the variation in stock prices with dividends and earnings. Before proceeding, it may be noted that there have been some time series studies of the variation in stock prices with dividends and other variables. focus of these studies has been the relation between the stock market and the business cycle6 and the discovery of profitable * research for this paper was supported by the Sloan Research Fund of the School of Industrial Management at Massachusetts Institute of Technology. author has benefited from the advice of Professors Edwin Kuh, Eli Shapiro, and Gregory Chow. computations were done in part at the M.I.T. Computation Center. 'Assume that the hypothesis stock price, P f (xi, X2,...), is stated so that it can be tested, and it is found to do a good job of explaining the variation in price among stocks. model and its coefficients thereby shed light on what investors consider and the weight they give these variables in buying common stocks. This information is valuable to corporations insofar as the prices of their stocks influence their financial plans. It is also true that a stock selling at a price above or below that predicted by the model deserves special consideration by investors. 2 Illustrations of this method of analysis may be found in texts on investment analysis such as: Graham and Dodd, Security Analysis, 3rd ed. (New York, i95i); and Dowrie and Fuller, Investments (New York, I94I). ' Graham and Dodd, op. cit., 454 ff. 'The Theory of Investment Value (Cambridge, I938). 5Bank Stock Prices and the Bank Capital Problem, Occasional Paper 54, National Bureau of Economic Research (New York, I957). 'J. Tinbergen, The Dynamics of Share-Price Formation, this REVIEW, XXi (November I939), 153-60; and Paul G. Darling, A Surrogative Measure of Business Con-

This article presents a response to comments on the proposed a model of transfer pricing in a socialist economy setting. Over the years an enormous body of literature has accumulated on the limitations of market determined prices as an instrument for the administration of an economic system. There are problems when perfect competition and/or government intervention are present in addition to the problems that arise when these conditions do not exist. Considering the brevity of Kenneth Most's comment, it is remarkable for its coverage of the literature on the subject. Furthermore, it may be useful to be reminded of the fact that these problems do not disappear when a socialist government employs transfer prices to facilitate the administration of its economy. How such a system would deal with these problems and whether these solutions would be more or less effective than the solutions available to a regulated capitalist economy is a subject for further research. A. Rashad Abdel-khalik argues that under the conditions specified, it appears that the proposed transfer price system will augment the employees' bonus for any period as compared to the preceding period by a function of the unfavorable deviation in fixed costs in period. If this statement were correct, the transfer price system would be less than useless, since a firm's management could maximize its income by maximizing costs. Abdel-khalik's argument is incorrect, but his concern with the control of organization costs is well placed.

20, 3 leaves : ; Includes bibliographical references. ; Cover title.;"January, 1978."

In the early sixties a number of papers developed stock value models in which the cost of equity capital is a function of the firm's equity investment rate.The conclusions reached, however, have for the most part been rejected in the subsequent literature.Instead, it has become widely accepted both in the theoretical work and in the textbooks on finance that the cost of equity capital is equal to the yield at which a firm's stock is selling, and the latter is independent of the firm's investment rate.In fact, one frequently finds the two terms, share yield and cost of equity capital used interchangeably in the literature.However, contributions to the literature during the last few years provide additional support for and clarification of the alternative theory, and a broad reconsideration of the subject is in order.Part I below critically reviews the literature of the early sixties on both sides of the subject and certain subsequent papers which contributed to the conclusion that share yield and cost of equity capital are equal and independent of investment policy.Parts II and III examine the recent contributions which have materially strengthened the theoretical basis for the contrary conclusion, that the cost of equity for a firm is an increasing function of its investment rate.Part IV reviews the empirical evidence on the subject, and Part V both summarizes the conclusions reached and discusses the areas in which the model requires further development.