marked advantage in the depression years from mid-1930 to mid-1933. During this time it earned and paid dividends while Continental Steel was reporting moderate losses. It is curious to observe that in the more recent recession the tables were exactly turned, and Continental Steel did very well while Granite City fared badly. Obviously the 1937–1938 results would command more attention than those in the longer past. Nevertheless, the thorough analyst would endeavor to learn as much as possible about the basic reasons underlying the change in the relative performance of the two companies.
Study of Qualitative Factors Also Necessary. Our last observation leads to the more general remark that conclusions suggested by comparative tabulations of this sort should not be accepted until careful thought has been given to the qualitative factors. When one issue seems to be selling much too low on the basis of the exhibit in relation to that of another in the same field, there may be adequate reasons for this disparity that the statistics do not disclose. Among such valid reasons may be a definitely poorer outlook or a questionable management. A lower dividend return for a common stock should not ordinarily be considered as a strong offsetting factor, since the dividend is usually adjusted to the earning power within a reasonable time.
Although overconservative dividend policies are sometimes followed for a considerable period (a subject referred to in Chap. 29), there is a well-defined tendency even in these cases for the market price to reflect the earning power sooner or later.
Relative popularity and relative market activity are two elements not connected with intrinsic value that nevertheless exert a powerful and often a continuing effect upon the market quotation. The analyst must give these factors respectful heed, but his work would be stultified if he always favored the more active and the more popular issue.
The recommendation of an exchange of one security for another seems to involve a greater personal accountability on the part of the analyst than the selection of an issue for original purchase. The reason is that holders of securities for investment are loath to make changes, and thus they are particularly irritated if the subsequent market action makes the move appear to have been unwise. Speculative holders will naturally gage all advice by the test of market results—usually immediate results. Bearing these human-nature factors in mind, the analyst must avoid suggesting common-stock exchanges to speculators (except possibly if accompanied by an emphatic disclaimer of responsibility for subsequent market action), and he must hesitate to suggest such exchanges to holders for investment unless the statistical superiority of the issue recommended is quite impressive. As an arbitrary rule, we might say that there should be good reason to believe that by making the exchange the investor would be getting at least 50% more for his money.
Variations in Homogeneity Affect the Values of Comparative Analysis. The dependability of industrial comparisons will vary with
C OMPARISON OF C ONTINENTAL S TEEL AND G RANITE C ITY S TEEL
(000 OMITTED, EXCEPT THOSE PER SHARE )
the nature of the industry considered. The basic question, of course, is whether future developments are likely to affect all the companies in the group similarly or dissimilarly. If similarly, then substantial weight may be accorded to the relative performance in the past, as shown by the statistical exhibit. An industrial group of this type may be called “homogeneous.” But, if the individual companies in the field are likely to respond quite variously to new conditions, then the relative showing must be regarded as a much less reliable guide. A group of this kind may be termed “heterogeneous.”
With certain exceptions for traffic and geographical variations,
e.g.
, in particular, the Pocohantas soft-coal carriers, the railroads must be considered a