The Ownership Structure Created by Share Issue Privatizations

Boutchkova and Megginson (2000) examine stock ownership patterns in privatized firms by comparing the numbers of stockholders in the privatized firms in the

1999 Business Week Global 1000 list (discussed above) to capitalization-matched private sector firms from the same national markets. They also compare the number and fractional ownership of institutional investors in these two lists. Next, BM perform a similar comparison of privatized and private-sector companies in the 1999 Top 200 Emerging Market list, but do not examine institutional shareholdings owing to lack of data. For each privatized firm, BM select as a match that private-sector company with the closest total market value in the Business Week lists, and they then collect the most recent data on the total number of shareholders for both sets of firms from the June 1999 Worldscope Disclosure CD-ROM database. While this data item is far from universally available, BM are able to collect values for 97 of the 153 privatized companies, and for 99 of the matching privately owned firms. In the majority of those cases where data are available for both the privatized and the matching firm, the privatized companies have a much larger number of shareholders, in spite of the fact that governments usually retain sizable stakes in these firms. This reduces the effective float of privatized firms, since these stakes remain unsold to private investors.

BM use the Wilcoxon signed-rank test to show that the mean number of shareholders of the privatized firms is significantly higher than that of the non-privatized matching firms. The frequency distribution of the number of shareholders in the Global Company Database on WorldScope is strongly skewed to the left. Roughly 91 percent of the 6,410 companies with data on the number of shareholders have fewer than 50,000 individual shareholders, 7.2 percent have between 50,000 and 250,000, and 1.8 percent have more than 250,000. The frequency distribution of the capitalization of their sample of privatized and matching nonprivatized firms is also markedly skewed to the left. However, BM focus on the companies with the highest market capitalizations, which also tend to have the largest number of shareholders, implying a higher proportion of companies with more than 250,000 shareholders. Owing to the limited availability of information on the number of shareholders — especially for the large, traditionally widely held companies — BM are able to construct a sample with complete information on both privatized and nonprivatized companies for only 86 pairs. Using these pairs, they conclude that the number of shareholders of the privatized companies is significantly higher, at the 0.01 level, than the number of shareholders in the matching private-sector (nonprivatized) sample companies.

There are three peculiar cases among the nonprivatized companies that have very large numbers of shareholders: Britain's Abbey National and Woolwich, with 2,028,141 and 1,216,932 stockholders, respectively, and Brazil's Banco Bradesco, with 2,414,603 stockholders. All three of these companies are financial institutions, and the two British firms were very large "de-mutualizations" that by their very nature created a great many new shareholders out of existing depositors. BM do not exclude these companies, and their testing procedure takes into account the magnitude of the differences between the number of shareholders of every pair. Even including these three firms, however, they still find that SIPs have (highly) significantly more shareholders than do the matching firms. The subsample of firms with complete data constructed from table 7.10 shows that the matching private companies have a total market capitalization of $1.2 trillion and 14 million table 7.10. Law and Finance—English Common Law Systems Promote Capital Market Growth

External

GDP Growth

Domestic

Rule of

Antidirector

Creditor

Cap/GDP

Debt/GDP

Rate

Firms/Pop

Law

Rights

Rights

Country (1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

Australia

0.49

0.76

3.06%

63.55

10.00

4

1

Canada

0.39

0.72

3.36%

40.86

10.00

4

1

Israel

0.25

0.66

4.39%

127.60

4.82

3

1

United Kingdom

1.00

1.13

2.27%

35.68

8.57

4

4

United States

0.58

0.81

2.74%

30.11

10.00

5

1

English origin

0.60

0.68

4.30%

35.45

6.46

3.39

3.11

average

Belgium

0.17

0.38

2.46%

15.59

10.00

0

2

France

0.23

0.96

2.54%

8.05

8.98

2

0

Greece

0.07

0.23

2.46%

21.60

6.18

1

1

Italy

0.06

0.55

2.82%

3.91

8.33

0

2

Spain

0.17

0.75

3.27%

9.71

7.80

2

2

French origin

0.21

0.45

3.18%

10.00

6.05

1.76

1.58

average

Austria

0.06

0.79

2.74%

13.87

10.00

2

3

Germany

0.13

1.12

2.60%

5.14

9.23

1

3

Japan

0.62

1.22

4.13%

17.78

8.98

3

2

Korea

0.44

0.74

9.52%

15.88

5.35

2

3

Switzerland

0.62

% 8

33.85

10.00

1

1

German origin

0.46

0.97

5.29%

16.79

8.68

2.00

2.33

average

Denmark

0.21

0.34

2.09%

50.40

10.00

3

3

Finland

0.25

0.75

2.40%

13.00

10.00

2

1

Norway

0.22

0.64

3.43%

33.00

10.00

3

2

Sweden

0.51

0.55

1.79%

12.66

10.00

2

2

Scandinavian

0.30

0.57

2.42%

27.26

10.00

2.50

2.00

origin average

Sample average

0.44

0.59

3.79%

21.59

6.85

2.44

Source: LaPorta, Lopez-de-Silanes, Shleifer, and Vishny (1997).

(44 countries)

Source: LaPorta, Lopez-de-Silanes, Shleifer, and Vishny (1997).

Note: This table details the relationship between the type of legal system upon which a country's commercial code is based and the size of that nation's capital markets for selected countries in 1994. Column 2 is the ratio of the Stock market capitalization held by minority (noncontrolling) shareholders to GDP, and column 3 provides a similar measure for private-sector debt (bank loans and bonds). Column 4 presents the country's average annual GDP growth rate during 1970-1993 and column 5 is the ratio of the number of domestic firms in a country to its population, in millions. Columns 6-8 present summary measures of the law and order traditions in a country (column 6) and of how well its legal code protects the rights of shareholders (column 7) and creditors (column 8).

shareholders, whereas the total market capitalization of the privatized firms ($1.6 trillion) is held by more than twice as many shareholders (37.6 million).

BM also compare institutional shareholdings in developed-country privatized firms to those of the matching private-sector companies. The mean (293 versus 281) and median (242 versus 231) number of institutional investors in the privatized and matching firms is surprisingly close. The same is true for mean (15.46 versus 15.78 percent) and median (12.81 versus 12.79 percent) percent shareholdings by these institutional investors. Using the Wilcoxon test of paired differences,

BM cannot reject the null hypothesis that the mean and median values are equal at conventional significance levels. The fact that governments retain sizable stakes in privatized firms — making the fraction of shares available for trading substantially smaller than for matched firms — suggests that institutional investors are at least as interested in investing in privatized companies as they are in private-sector firms of similar size. Additionally, while individual investors are offered the opportunity to profit by initial underpricing by government issuers, institutions are generally not offered similar discounts in the initial offering. Therefore, institutions must reap superior returns by providing monitoring and/or other valuable services.3

Was this article helpful?

0 0

Post a comment