Lapas attēli
PDF
ePub

The

to account for any considerable proportion of the total holdings. corporations with assets of $100,000,000 and over, on the other hand, accounted for 70 percent of the total shareholdings and about 67 percent of their aggregate value. The 17 issues of corporations with assets of $200,000,000 and over accounted alone for almost 60 percent of all shareholdings in the closely-held category and about 51 percent. of the total estimated value.

As among the common stocks, there was greater concentration of shareholdings in the lower price brackets for closely-held than widelyheld preferred stocks. Almost half of all shareholdings in the closelyheld group was accounted for by issues selling at under $10 a share (table 82). Another 16 percent was in issues ranging between $10 and $30 in price, about 18 percent was in issues priced between $30 and $60 and almost 17 percent was in issues selling at $60 or more per share. The latter group of shareholdings accounted, however, for over 60 percent of the total value for the group.

Of the 86 closely-held issues, 63 had less than 1,000 shareholdings each. In 22 issues the number of shareholdings ranged from 1,000 to 10,000 each, and there was only one issue in which the number of shareholdings exceeded 10,000 (table 83). The largest of the closelyheld issues had a little over 23,000 shareholdings-about 24 percent of the total number of holdings in all 86 issues-but accounted for only around 1 percent of the total estimated value for the group.

Thirty-five of the eighty-six closely-held issues were listed on the New York Stock Exchange, 28 were fully listed on some other exchange, and 23 were either not listed or admitted to unlisted trading privileges only (table 85). The fully listed issues accounted for almost 90 percent of the total shareholdings of which about 61 percent was in issues listed on the New York Stock Exchange alone; corresponding proportions for the widely-held issues were 85 percent and 58 percent, respectively. In the closely-held preferred stocks, as in the closely-held common stocks, the distribution by estimated value of individual holding was more heavily tipped toward the lower end of the scale than in the widely-held preferred stock issues. Of the 96,000 total shareholdings in the 86 closely-held issues, about 67 percent had a value of $500 or less and only about 3 percent was valued in excess of $10,000 (table 88); in the much more numerous widely-held group (table 94), 44 percent of the total shareholdings was valued at $500 or less and almost 5 percent had a value in excess of $10,000 each. Of the remaining 30 percent of all shareholdings in the closely-held preferred stock issues, a little over 11 percent had a value of $501 to $1,000, almost 16 percent ranged in value from $1,001 to $5,000, and about 3 percent was valued from $5,001 to $10,000; corresponding proportions for the widely-held issues were 19 percent, 26 percent, and 6 percent, respectively, or a combined total of 51 percent distributed among these 3 intermediate value classes.

CHAPTER VI

CONCENTRATION OF STOCK OWNERSHIP IN THE
1,710 CORPORATIONS

INTRODUCTION

The preceding chapters have dealt with certain totals of shareholdings in the 1,710 corporations. They have given a concrete idea of the number of shareholdings of different size and their relative importance among the 1,710 corporations, and have indicated differences in size distribution of shareholdings by type of stock, industry, and asset size of the issuer, price of issue, number of shareholdings, and average value of shareholding per issue. These chapters have also touched briefly upon the concentration of ownership prevailing among the 1,710 corporations by indicating the relatively small number of large shareholdings and the relatively large number of shares included in these not too numerous holdings. This chapter, however, is specifically devoted to a discussion of the degree of concentration of ownership in the 1,710 corporations and of differences in concentration between different types of stocks and issuers.

1. THE MEASUREMENT OF CONCENTRATION1

As in the preceding chapters, the basic material consists of the data on the size distribution of shareholdings in each of the 2,381 common and preferred stock issues of the 1,710 corporations. Two distributions are available. In the first distribution the number of shareholdings and shares outstanding are arranged in seven groups on the basis of the number of shares in each individual holding. The second distribution, derived from the first as described in chapter III, is based, on the other hand, on the estimated market value at the end of 1937 of each individual shareholding; it shows the number of shareholdings falling within five value classes but not the number or the aggregate market value of the shares included in the holdings in each value class.

It is important to remember that both distributions are based on record shareholdings as they appear on the books of the 1,710 corporations with the result that shares owned by numerous individual stockholders, generally in relatively small blocks, frequently appear as a smaller number of larger shareholdings registered in the names of nominees, mainly brokers and banks. The available figures thus tend to exaggerate somewhat the degree of concentration existing among the beneficial owners of the stock of the 1,710 corporations. This tendency has already been discussed in chapter I (sec. 4) and the

1 This section is similar in approach to the treatment of this subject in the report on the "Distribution of Ownership in the 200 Largest Non-financial Corporations," ch. III. sec. 6.

? On the other hand, there are a number of instances in which several record shareholdings in the same stock are owned beneficially by the same person through nominees. These are considered, however, to have only a relatively small effect on the results.

272188-41-No. 30- 5

47

conclusion was drawn therein that the distribution of ownership is probably only slightly less concentrated on the basis of beneficial ownership than on the basis of record ownership, although the difference may be quite considerable in individual corporations and undoubtedly is in a number of instances. Fairly sizable differences in this respect may even characterize whole groups of corporations but it has been attempted to make allowance for this factor in interpreting the data.

Ownership of an issue of stock may be regarded as equally distributed if every shareholding is equally large or, in other words, if every stockholder owns the same proportion of stock outstanding. The more the actual distribution deviates from this perfectly equal distribution, the more concentrated the ownership. This concept of concentration of ownership has been utilized to construct graphs, generally known as Lorenz curves, which indicate visually the degree of concentration of record ownership existing in any stock issue. These Lorenz curves are constructed by connecting a number of points derived from the distribution data, each of which indicates the percentage of the total issue outstanding included in a certain percentage of the shareholdings, cumulated from the largest shareholding downward. By such linking of points a broken line is obtained which will ordinarily approximate a smooth curve more and more closely as the number of points increases. Since the limited number of points available for this study does not permit drawing a smooth curve, all the charts show the broken line obtained by linking the actual points, as derived from the data for each issue (see appendix, tables 18 to 41). The size of the area between the broken line and the line of equal distribution indicates the degree of concentration; the larger this area the greater the concentration.5

In this section, the term "concentration of ownership" in an individual corporation refers to the extent of the inequality of the distribution of ownership among the stockholders of that corporation. More specifi cally, the concentration of ownership in one corporation is said to be greater than the concentration of ownership in another corporation when it takes, on the average, a smaller proportion of the shareholdings in the first corporation to account for a designated proportion of the shares. The particular measure of concentration used in this section is the area between the Lorenz curve and the line of equal distribution. There are, however, other aspects of concentration of ownership of a corporation which are not covered in this concept. Possibly the most important limitation of the concept used here is the fact that it relates to the distribution of ownership of some corporation or issue among a group of stockholders without regard to their number. Thus a corporation might be closely-held and yet not at all concentrated in its ownership according to this concept-if each of the few stockholders owned the same amount of stock-even though its ownership is unquestionably concentrated from the point of view of all stockholders. For some purposes, therefore, concentration might be measured by a second and entirely independent figure, the reciprocal of the number of shareholdings, which may be used in conjunction with the measure derived from the Lorenz curve. This second measure is not used in this chapter, but comparison of the degree of concentration among individual issues or groups is limited to issues or groups with a considerable number of shareholdings.

4 There are only 8 points available on the basis of size of shareholdings (namely, the point representing the percentage of all shareholdings constituted by those of more than 5,000 shares and the proportion of all shares outstanding included in these holdings, and so on downward), and only 6 points where the distribution by value groups of holdings is utilized. In the latter case the proportion of the total value of the issue represented by shareholdings in a certain value group must be estimated, generally by multiplication of the number of shareholdings by an estimated average value.

The area between the broken line and the line of equal distribution will always be smaller than the area between the line of equal distribution and the curve which would be obtained if all points were available. Consequently, the estimated degree of concentration will always be smaller than the actual degree of concentration. The size of this error is not constant, being larger for issues with a relatively low degree of concentration of ownership than for issues with high concentration. This factor, however, has been taken into consideration in comparing various groups of issues with respect to significant differences in the distribution of their ownership.

The concept and measure of concentration of ownership, as described above, are readily applicable to an individual corporation or, rather, to an individual issue of stock. Certain difficulties arise, however, when it is attempted to characterize a group of corporations or stock issues in a similar manner, i. e., to measure the average degree of concentration of ownership prevailing in the group. An obvious solution to this problem is to use the median area under the Lorenz curve, together with some measure of its representativeness. Such a measure is based on an entire issue as a unit and each shareholding receives a weight based on its size relative to all shareholdings of the same issue only. Another measure which can be utilized is the area under the Lorenz curve obtained by combining all the shareholdings of the issues covered by the report or of some smaller group of issues. This aggregate measure is based on the shareholding as a unit and each shareholding receives a weight based on its size relative to all the shareholdings of the issues included in the group. In such an approach, the shareholdings in a number of corporations are grouped together and treated as if they all formed part of one large issue. Since the data for the aggregate Lorenz curves have already been obtained as a basis for the discussion of the value and share distributions of individual shareholdings (chs. III and IV), and since it was not feasible to follow the median approach in view of the large number of issues involved, the aggregate Lorenz curves are used instead of the median Lorenz curves in the graphic presentation of concentration of ownership of the various groups of corporations. However, the conclusions arrived at on the basis of the aggregate measures of concentration were checked against median measures of concentration, obtained from a 10 percent sample of issues, and any important differences between the two measures of concentration will be pointed out in the textual discussion."

Two measures of aggregate concentration of ownership in a group of corporations have actually been used, one based on the value distribution of shareholdings in all corporations in the group, the other based on the share distribution of holdings. For an individual corporation, or rather an individual issue, both measures of concentration are identical. However, for a group of corporations this is no longer true because of the different weighting inherent in the two measures. Thus in the aggregate value distribution of shareholdings the same weight is given to shareholdings of equal value regardless of the number of shares in each holding. Conversely, in the share distribution the same weight is given to shareholdings comprising the same number of shares regardless of their value. The share distribution has the advantage of being derived directly from the original data while in the value distribution it was necessary to resort to interpolation with the attendant possibilities of error. The value distribution, on the other hand, has the advantage of putting issues of various prices on a comparable basis; furthermore, the information it provides-viz, the number and relative importance of shareholdings of a certain valueis more interesting than that given by the share distribution. The

• For almost all subgroups of companies considered there is a rather marked clustering of measures of concentration of ownership in individual corporations (i. e., the areas under the Lorenz curves for individual corporations) about the median for the group (i. e., the median area under the Lorenz curve). In other words, there is a rather high degree of similarity among the patterns of distribution of ownership in different corporations in the group.

relatively small differences between the concentration of ownership indicated by the value and share distributions are probably due mainly to the different weights given to the same issue by the two approaches. However, they also reflect to some extent errors in the interpolation used to derive the value distribution of the shareholdings in individual corporations from the share distributions.

The procedure described above makes it possible to depict by a single curve the distribution of ownership of all the 1,710 corporations

CHART VIII

CONCENTRATION OF OWNERSHIP OF THE STOCK OF 1710 CORPORATIONS
WITH SECURITIES LISTED ON A NATIONAL SECURITIES EXCHANGE

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][subsumed][ocr errors][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small]

or large segments thereof. The composite nature of such averages or aggregates of concentration must, however, be borne in mind in their interpretation.

2. CONCENTRATION IN COMMON VERSUS PREFERRED STOCK ISSUES

Chart VIII shows the Lorenz curves for all stock issues of the 1,710 corporations, based on the estimated distribution by the endof-1937 value of all 14,000,000 individual shareholdings. One curve

« iepriekšējāTurpināt »