Tabell’s Market Letter – July 06, 1979

TABELL'S MARKET LETTER , J 909 STATE ROA.D, PRINCETON. NEW JERSEY 08540 DIVISION OF MEMBER NEW YORI( STOCK EXCHANGE, INC MEMBER A.MERICAN STOCK EXCHANGe July 6, 1979 and Merck in; Chrysler arid 'Esmark out; from readers. It was asked, for example, why the split in du Pont caused the weight of that stock in the Dow to be reduced substantially. The answer lies in the method of computation chosen by Dow-Jones in 1897, when the Average was started, and maintained ever since. That method was the eminently simple and sensible one of adding the prices of the individual component stocks and dividing by the number of those components. However, this technique, exactly the same one we were all taught to use when com- puting an average in grade school, posed a question of how to adjust for stock splits. One method that could have been applied would have been, if a stock split, say, 2 for I, forever after to multiply the price of that stock by two. Dow-Jones rejected this method. They decided instead that, when a stock split occurred, they would adjust the divisor, so that the reduction in the price of the component stock due to the split would not affect the average. The same technique of divisor adjustment would be used when individual stocks were added or removed. Needless to say, there have been a good many such changes since 1897, and these, in most cases, resulted in a reduction of the divisor, which is now 1.465. This relatively low divisor accounts for the fact that the Dow is now well above 800, although none of its individual components sells above 80. This difference between the level of the average and the level of the average stock in the average is one of the commonly-voiced objections to the Dow as an indicator. Another is the arbitrary weighting of each stock by the price at which it happens to sell, a weighting that is radically changed by splits. Thus du Pont around 120 constituted more than 10 percent of the average. At around 40 it constitutes about 3 percent. Procter & Gamble and IBM, because of their high price, are about four times as important to the Dow as Goodyear and Sears Roebuck for the same reason. Why, it may be asked, did Dow-Jones not choose the first alternative referred to above, multiplying the split prices by the amount of the split ,or did it not choose to weight equal dollar investment newspaper , and it was, a time, market indicator. One of the components, however, was IBM, and, as most of us are aware, the price history of IBM in the 1940's and 1950's was somewhat spectacular. Under these conditions the weight of IBM in the index in question became so heavy that, as an average, it was entirely useless. It simply tended to reflect what its major component was doing. The average has since been discontinued. What other methods for computing averages exist Standard & Poors has chosen to weight its indices by capitalizations, in effect taking the price of each issue in its 500-stock universe and multiplying by the number of shares outstanding. The resultant total market value is then divided by a figure calcu- lated on a given base period, which happens to be 1941-1943. This method has the advantage of obviat- ing any adjustment for splits. Moreover, the inclusion of a large number of stocks prevents any single issue from ever obtaining a disproportional weight. Thus IBM, although the largest component, accounts for only 6 percent of the index. – The difficulty is, in our view, that it is fairly difficult to justify capitalization as a criterion for relative weighting. This is especially true when an index is used as a standard against which to measure portfolio performance, as the S & P often is. The small investor, at least, has an equal opportunity to purchase any of the components of the average. Why, therefore, should Kroehler Manufacturing, the smallest component, carry only 1/300Oth the weight of the largest It is possible, of course, to construct an average which weights all stocks equally. This is precisely what our Cumulative Index does for all issues on the New York Stock Exchange, essentially taking the percentage change for all these issues, averaging it, and applying that percentage change to the index each day. An indicator of this type is called a geometric average. We think this index useful, which is why we compute and publish it, but it also has a limitation, in that to duplicate its performance in an actual portfolio, one would have to keep the individual components of that portfolio equally weighted on a daily basis. This is, of course, impossible. The moral of this whole discussion is that the construction of a market average is hardly the simple task it appears on the surface, and there exists no single perfect method of computation. Different market indicators are appropriately used for different purposes, and it is the responsibility of the analyst to justify the use of a market average in any particular context. Dow-Jones Industrials (1200 PM) S & P Composite (1200 PM) Cumulative Index (7/6/79) 837.54 102.66 759.83 ANTHONY W. TAB ELL DELAFIELD, HARVEY, TABELL AWTsla No statement or expression of OpInion or any other matter herein contained IS, or IS 10 be deemed to be, directly or lOdHectly, on oHer or the Ollcltatlon of on offer to buy or sell any secuflly referred 10 or mel'lfloned The matter IS presented merely for Ihe convellenCE of the subscriber While oNe believe the sources of our Information to be reliable, we In no way represent or guarantee Ihe accuracy thereof nor of Ihe statements mode herein Any !lellon to be token by ihe subscriber should be b!lsed on hiS own Invesllg!ltlon !lnd 'nfarmai'on Janney Montgomery SCali, tnc, !l a corporotlon, ond lIs officers or employees, may now have, or may laler toke, positions or irades In respect to any securities mentioned In Ihls or any future Issue, and such position may be different from ony View! now or hereafter eprened In thiS or ony other IHue Janney Montgomery Scott, Inc, whICh IS registered With the SEC as on Invesiment odvlsor, may give adVICe to tts ,vestment odvlsory and othel customers Independently of any stotements mode In thiS or In ony other Issue Further Informalron on any securrly mentioned herein IS available on request