A Computer Program for Common Stock Valuation

Abstract

FINANCIAL LITERATURE recent years has contained an increasing number of articles expounding the characteristics of one or another mathematical model of common stock valuation.' Yet most indications are that the majority of practicing security analysts still estimate reasonableness of price for particular stocks by a process which is part comparison with other available at the same time (regardless of whether the .'other issues are reasonably priced) and part sheer seat-of-the-pants Seldom does a potential investor hear from a security analyst a firm statement to the effect that, that analyst's professional opinion, a particular stock offers a probable return, including appreciation, of such-and-such percent. If he does manage to obtain advice to the effect that a particular stock is a good buy at a particular price, he still has no clear idea as to the logic or assumptions underlying the recommendation. In fact, he probably doesn't even know what good buy (or its slightly more precise pseudonym, selected for above average appreciation,) really means terms of expected annual gain. Why the difficulty blending current theoretical knowledge, which generally favors measuring investment gains by internal rates of return, with current practice? In part, the answer to this question lies the difficulty of reducing mathematical models to an operational level. Consider problems confronting an eager young MBA, newly hired as a junior financial analyst and aglow with a scholarly understanding of the latest analytical tools for scientific valuation of common stocks: 1. He may, if he wishes, calculate for each stock assigned to him a proper value, using any of several similar mathematical models based upon obtaining the present value of expected future cash receipts. Unfortunately, the calculation time involved is formidable, and the young man will spend most of his time doing math rather than analysis. 2. Somewhat wiser, the young man may turn to a number of prepared tables published for the express purpose of permitting him to plug in his expectancy for a number of variables and read out the values. Normally these tables are expressed terms of an initial one dollar of dividends or of earnings, so only a simple multiplication is necessary to arrive at a scientific value for the stock. The best and most comprehensive of these tables are those published by Soldofsky and Murphy.2 In their book, a large number of tables are available for various combinations of dividend growth from 3% to 12%, rates of return from 2% to 25%, and several combinations of growth patterns. In a recent Financial Analysts Journal article, Molodovsky, May, and Chottiner presented another collection of tables. In order to value earnings streams at 6%, 7%, 8%, and 9%, some 80 tables were presented for varying combinations of growth patterns. The table user must compromise his solution by limiting his expectancies to the number of variables provided for by the tables or by obtaining a fantastically large number of tables. Neither solution is extremely satisfactory. 3. A third alternative is to adopt a visual approach to scientific stock valuation. A nomograph developed by Ferguson, permits the analyst to combine expected growth rate with a standard price-earnings ratio and expected growth period.4 Given two of the variables, the implications for the third may be read from the nomograph. An entirely separate approach to the graphic expression of the variables frequently found stock value tables was developed a few years ago by the present author.5 Although somewhat faster than tables use, and having the added advantage of portraying visually the interrelationships among a greater number of variables, the graphs are nevertheless also limited treating the interaction of all the combinations of variables which might occur.