Corporate Governance, Market Valuation and Dividend Policy in Brazil

André Carvalhal-da-Silva
Assistant Professor of Finance
COPPEAD Graduate Business School
Federal University of Rio de Janeiro
Tel: (55 21) 3235-3444
Fax: (55 21) 2256-6985
Email: andres@[Link]

Ricardo Leal
Professor of Finance
COPPEAD Graduate Business School
Federal University of Rio de Janeiro
Tel: (55 21) 2598-9802
Fax: (55 21) 2598-9817
Email: rleal@[Link]

JEL Classification: G30, G32

Keywords: Ownership structure; corporate control; agency costs; Brazil

Abstract

This study investigates the effects of the corporate governance structure on market valuation and dividend payout of

Brazilian companies. The empirical results indicate a high degree of ownership and control concentration. We can also note

a significant difference between the voting and total capital owned by the largest shareholders, mainly through the existence

of non-voting shares, pyramidal structures, and shareholding agreements. These mechanisms seem to be used by controlling

shareholders to keep the firm’s control without having to own 50% of the total capital. The evidence also reveals that there

is a relationship between governance structure, market valuation, and dividend policy in Brazil.

I - Introduction

The concept of corporate governance is not a particularly recent subject, but it was only in the 80´s that corporate

governance models were studied to evaluate their effects in the valuation and dividend policy of companies in emerging

markets. In Brazil, the debate about corporate governance structures was intensified in the last decade, when factors such as

privatizations, the opening process of the Brazilian economy, the entrance of new investors – especially foreign and

institutional ones, have stimulated new efforts towards better corporate governance practices.

The corporate governance concept itself is very broad, but the analysis can be centered in the ownership (cash flow

rights) and control (voting rights) structure of companies, since this dimension is intimately related with all the others.

Shleifer and Vishny (1997) found that the ownership structure, along with the country legal protection, is one of the most

important determinants of corporate governance. The main purpose of this paper is to analyze the ownership and control

structure of Brazilian companies and its effect on market valuation and dividend policy.

The understanding of ownership structure is very important since it influences directly the efficiency of the market

for corporate control. First it shows the degree of risk diversification of shareholders. When ownership is concentrated,

there is less risk diversification by shareholders. A second important point is that it shows a potential agency problem in the

management of the firm. There may be an agency problem between managers and shareholders because managers may not

be maximizing shareholder’s value. When there is a stockholder that can influence the control of a company, a new agency

problem can arise between controlling and minority shareholders. We will examine these points by looking into direct and

indirect ownership concentration of Brazilian companies.

Most of the literature that first studied the problem of the separation between ownership and control has done it in

an environment where ownership was diffuse, i.e., there were a lot of small shareholders, each of whom with a very little

2
portion of the capital. Berle and Means (1932) studied the ownership structure of large firms in the United States and

observed that most of them had its capital diluted among many small shareholders. This idea was extensively accepted as

the corporation model in modern economies. However, recent studies concluded that very few countries are actually

characterized by diffuse ownership firms. Many developed countries, such as France, Italy, and Germany, and emerging

markets are characterized by a high concentration of ownership and control.

This paper follows an extensive literature on the effects of corporate ownership structures on valuation. Jensen and

Meckling (1976) and Morck et al (1988) have provided important contributions to the research on ownership structures and

corporate valuation. Jensen and Meckling concluded that concentrated ownership is beneficial for corporate valuation,

because large investors are better at monitoring managers. Morck et al distinguish between the negative control effects and

the positive incentive effects of higher shares of ownership. They suggest that the absence of separation between ownership

and control reduces conflicts of interest and thus increases shareholder value.

Recent research suggests that higher cash flow rights are associated with higher valuation. In contrast, the

concentration of control rights and the separation of voting from cash flow rights have a negative effect on firm value.

Shleifer and Vishny (1997), La Porta et al (1998,1999), Morck et al (1988) and Claesens et al (2000a, 2000b) studied the

conflicts of interest between large and small shareholders. When large investors control a corporation, their policies may

result in the expropriation of minority shareholders. Such companies are unnattractive to small shareholders and their shares

have lower valuation.

The Brazilian literature on corporate governance is composed mainly by research made since the 90´s. Valadares

and Leal (2000) and Leal, Carvalhal-da-Silva and Valadares (2002) analyze the direct and indirect ownership structures and

find a high degree of ownership concentration in Brazilian firms. The concentration occurs mainly through the violation of

the one share –one vote rule.

In this paper, we test 6 hypotheses in order to analyze the effects of cash flow and voting rights on market valuation

and dividend policy of Brazilian companies. Recent research (Shleifer and Vishny (1997), La Porta et al (1998, 1999,

2000a, 2002), Morck et al (1988) and Claessens et al (2000a, 2000b)) suggests that the concentration of voting rights by the

controlling shareholders is associated with more expropriation of minority shareholders since large owners may prefer to

generate private benefits of control that are not shared by minority shareholders. So the first hypothesis to be tested is:

H1: Higher concentration of voting rights by the controlling shareholder is associated with lower corporate valuation.

Burkart, Gromb and Panunzi (1998) argue that, in general, expropriation is costly and therefore higher cash flow

ownership should lead to lower expropriation, other things equal. La Porta et al (1999) argue that the power of the

3
controlling shareholders to expropriate outside investors is moderated by their financial incentives not to do so. An

important source of such inventives is equity or cash flow ownership by the controlling shareholder. The second and third

hypotheses are derived from the above statements with respect to market valuation and the potential expropriation of

minority shareholders.

H2: Higher cash flow ownership by the controlling shareholder is associated with higher corporate valuation.

H3: Higher separation of voting from cash flow rights by the controlling shareholder is associated with lower corporate

valuation.

Recent literature suggests that corporate governance structures are related not only to market valuation, but also to

dividend policy. La Porta et al (2000b) state that firms located in countries with a higher legal protection to minority

shareholders (based on common law systems) pay higher dividends, compared to countries where legal protection is weak

(civil law systems). Johnson and Shleifer (2001) consider the payment of a higher dividend as a means to establish a

reputation to adequately treat minority shareholders.

Since control concentration tends to decrease firm valuation, due to the possibility of expropriation of minority

shareholders, we can expect that firms with high concentration of control have a low payout, since the controlling

shareholder will tend to distribute a low dividend, expropriating minority shareholders. Therefore, the fourth hypothesis is:

H4: Higher concentration of voting rights is associated with lower dividend payout

On the other hand, an increase in the concentration of cash flow rights is associated with a higher firm valuation,

due to the alignment of interests between the controlling and minority shareholders. Therefore, we can expect that firms

with a high concentration of cash flow rights distribute high dividends, which formulates our fifth and sixth hypotheses.

H5: Higher concentration of cash flow rights is associated with higher dividend payout.

H6: Higher separation of voting from cash flow rights is associated with lower dividend payout.

The paper is structured as follows. This section presented the theoretical framework, with some of the main studies

on the topic, along with the hypotheses to be tested. Next section describes the data set and the methodology used in the

tests. Section III presents the results of the ownership and control structures and their relationship with market valuation and

dividend payout of Brazilian firms. Section IV concludes.

4
II – Data and Methodology

Our sample consists of firms listed in Sao Paulo Stock Exchange (Bovespa) in the year of 2000. We collected

information on the shareholding structure from the Infoinvest Database (Browne Global Solutions). Our sample does not

include financial institutions, companies with incomplete or unavailable information, and firms whose shares were not

traded in Bovespa during December 2000. The final sample consists of 225 firms, which represent about 45% of the number

of firms, and approximately 70% of total market capitalization of Bovespa.

The study of the ownership structure should not be limited to the direct ownership, but should also focus on who is

the final owner of the firms. Therefore, we analyze two forms of shareholding composition: direct and indirect. Direct

shareholders are those who own shares in the company itself. We consider all shareholders with 5% or more of the voting

capital, because 5% is the threshold for mandatory identification of shareholders in Brazil. Indirect composition represents

stockholders who ultimately own the company. For example, if a shareholder owns, directly, 60% of firm A, which owns

51% of firm B, we can say that this shareholder owns, indirectly, 30.60% (60% times 51%, i.e, multiplying the levels of

shareholding interest) and controls 51% (the minimum between 51% and 60%, i.e, the weakest link in the shareholding

channel) of firm B.

This analysis was possible since the Annual Reports show the shareholding composition of parent companies when

they exist. Thus, we analyzed the shareholding composition backwards until we were able to classify the true owners into

one of the following groups: (i) individuals or fa milies; (ii) foreign investors (individuals or institutions); (iii) government;

(iv) institutional investors (banks, insurance firms, pension funds or investment funds).

We analyzed the direct and indirect structure of both control (voting capital) and ownership (total capital, i.e, the

sum of voting and non-voting capital) of Brazilian firms. Cash flow and voting rights can be rather different due to the use

of non-voting shares and indirect structures (“pyramids”). A pyramid is a structure where a shareho lder controls a firm,

which controls other firms and so on. Shareholding agreements are another way to increase control, because shareholders

can form groups through written contracts in order to exercise jointly their voting rights.

The 225 companies on ht e sample were divided into two main groups: firms with and without a majority

shareholder. A company with a majority shareholder is one where a single shareholder has directly more than 50% of the

voting capital. For the direct and indirect shareholding composition we computed the voting and total capital owned by the

largest, the three largest and the five largest shareholders of the firm.

Since corporate governance is related to the control of firms, it is fundamental to identify the control and the

ownership structure. The control and ownership categories differ with respect to the concentration of ownership and the

origin of capital. Therefore, the ownership has two main dimensions: (i) the identity of the largest owner; and (b) the

5
concentration of his ownership. In this context, another classification was made, based on the type of the owners, and the

firms with a majority shareholder were further divided according to the origin of capital: foreigners, government, familiy

and institutional investors.

The purpose of the classification of firms according to the type of the majority shareholder and to his concentration

of ownership is to verify the existence of a significative difference among firms, regarding their market valuation and

dividend policy. In order to measure the market value, we used Tobin’s Q, which represents the market value of assets,

divided by their replacement cost. Many authors (Morck et al (1988), McConnel and Servaes (1990), La Porta et al (2000a,

2002)) have used Tobin’s Q to measure the relative market value. The numerator of Tobin’s Q is the market value of the

firm´s assets, computed as the book value of assets minus the book value of common equity and deferred taxes plus the

market value of common equity. The denominator is the replacement value of assets, and we used the book value of assets

as a proxy for this variable. The dividend payout is measured as the dividend/net profit ratio.

The first analysis is a parametric test that compares averages, in order to evaluate possible differences between

market valuation and dividend payout of the firms, classified according to the origin of capital and to the ownership and

control concentration. Then, we conduct a more formal analysis using multiple linear regression procedures. Using this

technique, we are able to study how independent variables, specially the direct and indirect control and ownership structure,

affects the market valuation and dividend payout of Brazilian firms. Therefore, we assume a causality relation among

variables, such that the dependent variable is explained by the independent variables.

We estimated this model for each of the two dependent variables (Tobin’s Q and dividend payout). The

independent variables include measures of control and ownership structure (voting capital, total capital, voting/total capital

ratio), and variables that might influence the dependent variables, previously identified and selected from the literature, such

as leverage (debt/asset ratio), size (ln (assets)), ROA (EBITDA/Asset ratio), risk (stock volatility), current asset/total asset

ratio. Specifications that included the squared variables (voting capital)2 , (total capital)2 , (voting/total capital ratio)2 , and

dummy variables that indicated the type of the shareholder were also tes ted. The inclusion of squared variables is consistent

with a curvilinear relationship between firm valuation and ownership structure (Morck, Shleifer and Vishny (1988), and

McConnel and Servaes (1990)). Equations 1 and 2 show the variables included in each model.

(1) Tobin’s Q = Constant + θ 1 Voting Capital + θ 2 Total Capital + θ 3 Voting/Total Capital Ratio + θ4 (Voting Capital)2 +
θ 5 (Total Capital)2 + θ 6 (Voting/Total Capital)2 + θ 7 Dummy Shareholder Type + θ8 Leverage + θ 9 Size + θ10 ROA

(2) Payout = Constant + θ1 Voting Capital + θ 2 Total Capital + θ 3 Voting/Total Capital Ratio + θ 4 (Voting Capital) 2 +
θ 5 (Total Capital)2 + θ 6 (Voting/Total Capital)2 + θ 7 Dummy Shareholder Type + θ8 Leverage + θ 9 Size + θ10 ROA +
θ 11 Risk + θ12 Current/total asset ratio

6
 In all specifications, we also included industry dummy variables, to control inherent characteristics of specific

sectors of the economy. The idea behind this adjustment is that each industry may be in a different stage of maturity, growth

and present some peculiarities that determine the firm valuation and dividend policy.

III – Empirical Results

Table I shows the direct structure of ownership and control of Brazilian companies in 2000. Out of 225 firms, 203

(90%) have one shareholder that owns more than 50% of the voting capital. This shareholder owns on average 76% of the

voting capital. Among the firms where the control is not held by one shareholder (22), the largest shareholder owns on

average 37% of the voting capital. This demonstrates that, even when one single shareholder does not have the majority of

votes, the largest shareholder holds a considerable portion of them. Considering the sample as a whole, the largest, the 3

largest and the 5 largest shareholders have, respectively, 72%, 85% and 87% of the voting capital.

Our results show a high degree of concentration of the voting capital. Even when there is no majority shareholder,

the largest one owns a significant portion of the voting capital, and the company is, on average, controlled by its 3 largest

shareholders. Besides this, 87% of the voting capital of companies are in the hands of the 5 largest shareholders.

Table I
Direct Shareholding Composition of Brazilian Companies in 2000
Direct shareholding composition of 225 Brazilian companies. A company with a majority shareholder is one where a single
shareholder has more than 50% of the voting capital. Data collected from Annual Reports, referring to year-end 2000.
Companies with a majority Companies without a Total Sample
shareholder (203) majority shareholder (22) (225)
Shareholder Voting Total Voting Total Voting Total
Capital Capital Capital Capital Capital Capital
Largest 76% 54% 37% 23% 72% 51%
3 Largest 88% 65% 62% 41% 85% 62%
5 Largest 89% 65% 66% 44% 87% 63%

We also can note a reasonable difference between the percentage of voting and total capital held by large

shareholders. In Brazil, the issuance of non-voting shares appears to be used by large shareholders to maintain control of the

firm without having to hold 50% of the total capital. In companies with a single shareholder, this investor has on average

76% of the votes but only 54% of the total capital. Considering the entire sample, the five largest shareholders have 87% of

the voting capital but only 63% of the total capital.

In Brazil companies are allowed to issue shares without voting rights in an amount up to two-thirds of the total

capital (Law 6404 - Law of Corporations). In 2001, the New Law of Corporations (Law 10303) changed the maximum

amount of non-voting shares from 2/3 to 50% of total capital, but this rule is obligatory only to non-public firms that decide

7
to go public after October 2001 and for new companies. This mechanism allows companies to issue shares without

relinquishing control and is therefore a way of separating ownership from control. Control of a company can be guaranteed

with only one-sixth of its total capital. Thus, this group may represent companies where a pyramidal structure is used to

separate ownership and control, or to maintain distance from the one share-one vote rule.

Table II shows the indirect structure of control and ownership of Brazilian companies in 2000. In the case of

companies where the major shareholder holds more than 50% of the voting capital the indirect ownership is more dilluted.

In direct form, the average majority shareholder owns 76% of the voting capital and 54% of the total capital, while

indirectly the figures are 69% and 40% respectively. Nevertheless, this reduced participation of the major shareholder does

not occur in the case of companies where there is no single majority shareholder. On the contrary, the data actually show a

small increase in the invested capital. Directly, the largest shareholder has on average 37% of the voting capital and 23% of

the total capital, while indirectly the figures are, respectively, 40% and 24%. This fact may indicate the use of pyramidal

structures to maintain control with reduced investment in the firm.

Table II
Indirect Shareholding Composition of Brazilian Companies in 2000
Indirect shareholding composition of 225 Brazilian companies. The indirect composition shows the indirect interest of
shareholders. Such participation is analyzed backwards until the effective shareholder is revealed to be from one of the
following groups: (i) individuals or families; (ii) foreign investors (individuals or institutions); (iii) government; (iv)
institutional investors (banks, insurance firms, pension funds or investment funds). A company with a majority shareholder
is one where a single shareholder has more than 50% of the voting capital directly. Data collected from Annual Reports,
referring to year-end 2000.
Companies with a majority Companies without a Total Sample
shareholder (203) majority shareholder (22) (225)
Shareholder Voting Total Voting Total Voting Total
Capital Capital Capital Capital Capital Capital
Largest 69% 40% 40% 24% 66% 38%
3 Largest 83% 51% 61% 39% 81% 50%
5 Largest 85% 54% 64% 41% 83% 52%

Then, the 203 firms that had a controlling shareholder were classified according to the origin of capital. Table III

shows the direct and indirect structure of control and ownership of firms according to the identity of the largest shareholder

(foreigners, government, family and institutional investors). Among these 203 firms , 108 are controlled by families, 60 by

foreign investors, 19 by institutional investors and 16 by the government. On average, institutional investors own directly

80% of the voting capital, while foreign investors, government and families own respectively 79%, 75% and 73%.

Indirectly, the voting capital of institutional investors drops to 64%, while foreign investors, government and families own,

respectively, 74%, 77% and 66% of the voting capital.

8
 Table III
Shareholder Composition of Controlling Groups in 2000
Firms that have a controlling shareholder were classified according to the origin of capital (foreigners, government, family
and institutional investors) and the shareholder composition of each of those groups was analyzed.
Direct Structure Indirect Structure
No Firms % Firms Voting Total Voting Total
Capital Capital Capital Capital
Total Sample 225 100% 72% 51% 66% 38%
Family 108 48% 73% 46% 66% 31%
Firms with a Government 16 7% 75% 57% 77% 51%
controlling Foreigners 60 27% 79% 62% 74% 56%
shareholder Institutional 19 8% 80% 66% 64% 33%
Total 203 90% 76% 54% 69% 40%
Firms without a controlling shareholder 22 10% 37% 23% 40% 24%

Table IV shows the existence of shareholding agreements, pyramidal structures and the percentage of voting capital

on total capital of Brazilian firms. These 3 mechanisms are closely related to the control and ownership structure and to the

possibility of expropriation of minority shareholders, since they can increase the separation between voting and cash flow

rights. Shareholding agreements exist in 27% of the family-owned firms, as opposed to only 6% of government-owned

firms that have those agreements. Generally, 23% of Brazilian firms have shareholding agreements, influencing the

corporate governance, in the sense that shareholding agreements on voting rights are a way of increasing firm control.

Most of the firms (86%) have pyramidal structures, mainly family-owned companies (91%) and lessly government-

owned firms (63%). The issue of non-voting shares is a common practice in Brazil, and voting shares represent on average

53% of the total capital. The percentage of non-voting shares on the total capital is smaller in firms controlled by families

(49%) and institutional investors (51%). On the other hand, government-owned firms have the highest percentage of voting

shares on total capital (64%).

Table IV
Mechanisms of Separation Between Control and Ownership in Brazil
Firms with a controlling shareholder were classified according to the origin of capital (foreigners, government, family and
institutional investors), and mechanisms of separation between control and ownership were analyzed: shareholding
agreements, pyramids and the percentage of voting shares on total capital.
% Firms with % Firms with % Voting Capital/Total
Shareholding Agreements Pyramid Structure Capital
Total Sample 23% 86% 53%
Family 27% 91% 49%
Firms with a Government 6% 63% 64%
controlling Foreigners 20% 87% 56%
shareholder Institutional 21% 79% 51%
Total 23% 86% 53%
Firms without a controlling shareholder 27% 82% 59%

9
 Table V shows the market valuation (Tobin’s Q), and the payout of Brazilian firms, according to the identity of the

controlling shareholder (foreigners, government, family and institutional investors). Firms without a controlling shareholder

have a higher Tobin’s Q (1.19) than those with a controlling shareholder (1.07). Since this analysis is focused on voting

rights (control), this fact seems to confirm hypothesis 1, i.e., that a higher concentration of voting rights is associated with a

lower firm valuation. Therefore, although the differences are not statistically significant, there is some evidence of the

negative relationship between control concentration and market valuation. Moreover, there is statis tical evidence that firms

controled by the government tend to be undervalued (Q=0.76) when compared to firms controlled by families (Q=1.09),

foreigners (Q=1.10) and institutional investors (Q=1.16).

There is no statistically significant difference between average payout of firms without (35%) or with (31%) a

controlling shareholder. Although there is no statistical significance, companies controlled by the government tend to have a

higher payout (36%) than the ones controlled by families (28%), institutional investors (34%) and foreigners (35%).

Table V
Market Valuation and Dividend Payout of Brazilian Firms in 2000
Comparative analysis of market valuation and dividend payout of Brazilian companies according to
the type of controlling shareholders (foreigner, government, family and institutional investors)
Tobin’s Q Payout
Total Sample 1.08 32%
Family 1.09 28%
Firms with Government 0.76* 36%
a controlling Foreigners 1.10 35%
shareholder Institutional 1.16 34%
Total 1.07 31%
Firms without a controlling shareholder 1.19 35%
* indicates Tobin´s Q differences significant at the 1% level.

Then, we performed a more formal analysis using multiple linear regressions. Table VI shows the results of the 6

model specifications for the study of the market valuation of Brazilian companies, considering the direct structure of

ownership and control. Although only some variables of the ownership and control structure were statistically significant,

the coefficient signs confirm what is predicted from the theory. The negative coefficient of the voting capital confirms

hypothesis 1, i.e., that a higher concentration of voting rights is associated with a lower firm valuation. The positive

coefficient of the total capital confirms hypothesis 2, i.e., the higher the concentration of cash flow rights, the higher is the

firm valuation. And the negative coefficient of the voting/total capital ratio confirms hypothesis 3, i.e., that the higher the

voting/total capital ratio, the lower is the firm valuation.

10
 Table VI
The Effects of the Direct Structure of Ownership and Control on Market Valuation of Brazilian Firms
Multiple linear regression analysis to evaluate the effects of the direct structure of ownership and control on the market
valuation of 225 Brazilian companies in 2000. The dependent variable is the Tobin’s Q and the independent variables are:
voting capital; total capital; voting/total capital ratio; leverage; firm size (log assets); ROA; (voting capital)2; (total capital)2 ;
(voting/total capital)2 ; and dummy indicating the type of controlling shareholders. Six model specifications were tested: 2
for the largest, 2 for the three largest, and 2 for the five largest shareholders. In all specifications, we included industry
dummies (coefficients are not reported here). The p-values of the t -tests are shown in parenthesis.
Largest 3 Largest 5 Largest
Shareholder Shareholders Shareholders
(1) (2) (3) (4) (5) (6)
Constant 0.22 -0.11 0.30 -0.30 0.27 -0.25
(0.50) (0.74) (0.42) (0.51) (0.48) (0.65)
Leverage 1.06* 1.06* 1.06* 1.06* 1.06* 1.06*
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Size 0.01 0.01 0.01 0.00 0.01 0.01
(0.47) (0.50) (0.51) (0.81) (0.47) (0.65)
ROA 1.10* 0.82** 1.11* 0.85* 1.09* 0.89*
(0.00) (0.02) (0.00) (0.01) (0.00) (0.01)
Voting Capital -0.21 -0.55 -0.35 -1.31 -0.19 -0.86
(0.47) (0.41) (0.25) (0.29) (0.54) (0.51)
Total Capital 0.10 0.50 0.14 0.40 0.03 0.58
(0.79) (0.46) (0.71) (0.71) (0.95) (0.68)
Voting/Total Capital Ratio -0.02 -0.01 -0.00 -0.04 -0.03 -0.05
(0.83) (0.44) (0.99) (0.62) (0.81) (0.72)
(Voting Capital)2 -0.69 -1.17 -0.86
(0.16) (0.14) (0.28)
(Total Capital)2 0.33 0.23 0.41
(0.55) (0.77) (0.65)
(Voting/Total Capital Ratio)2 -0.00 -0.00 -0.00
(0.43) (0.72) (0.70)
Dummy Foreigners 0.12*** 0.14*** 0.14***
(0.10) (0.07) (0.06)
Dummy Government -0.05 -0.02 -0.03
(0.65) (0.85) (0.78)
Dummy Institutional 0.12 0.13 0.15
(0.23) (0.19) (0.12)
Adjusted R2 0.80 0.80 0.80 0.80 0.80 0.80
*, ** and *** indicate statistical significance at the 1%, 5% and 10% levels, respectively.

Although with little statistical significance, the firm market valuation is different, as a function of the origin of

capital. Companies with a foreign or institutional controlling shareholder tend to present a higher valuation, when compared

to family-owned companies, while government-owned firms tend to have the lowest valuation. A possible explanation is

that in government-owned firms, governance is more complex due to the presence of one extra agent: politicians. In this

context, the control rights (in the government’s hands) are totally dissociated from the cash flow rights, since ownership is

spread among the citizens, who are the ultimate owners of these firms. Squared variables (voting capital, total capital, and

11
voting/total capital ratio) generally do not present statistically significant coefficients. Leverage, ROA and size have a

positive relationship with the firm valuation, as predicted by the theory.

Table VII shows the results of the 6 model specifications for the study of the market valuation of Brazilian

companies, considering the indirect structure of ownership and control. The results are mainly the same as the direct

structure, but in the indirect structure the p-values of variables related to corporate governance tend to be lower than in the

direct structure, which means that those variables have a higher statistical power in the indirect structure. For example,

when we analyze only the largest shareholder, the negative relationship between voting capital concentration and market

valuation becomes statistically significant at the 10% level.

Table VII
The Effects of the Indirect Structure of Ownership and Control on Market Valuation of Brazilian Firms
Multiple linear regression analysis to evaluate the effect of the indirect structure of ownership and control on the market
valuation of 225 Brazilian companies in 2000. The dependent variable is the Tobin’s Q and the independent variables are:
voting capital; total capital; voting/total capital ratio; leverage; firm size (log assets); ROA; (voting capital)2; (total capital)2 ;
(voting/total capital)2 ; and dummy indicating the type of controlling shareholders. Six model specifications were tested: 2
for the largest, 2 for the three largest, and 2 for the five largest shareholders. In all specifications, we included industry
dummies (coefficients are not reported here). The p-values of the t -tests are shown in parenthesis.
Largest 3 Largest 5 Largest
Shareholder Shareholders Shareholders
(1) (2) (3) (4) (5) (6)
Constant 0.21 -0.15 0.27 -0.86 0.30 -1.16
(0.38) (0.89) (0.32) (0.58) (0.28) (0.47)
Leverage 1.06* 1.06* 1.06* 1.06* 1.06* 1.06*
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Size 0.01 0.01 0.01 0.01 0.01 0.01
(0.48) (0.58) (0.61) (0.58) (0.48) (0.56)
ROA 1.03* 0.85* 1.08* 0.87* 1.11* 0.84*
(0.00) (0.01) (0.00) (0.01) (0.00) (0.01)
Voting Capital -0.25*** -2.82 -0.30 -4.22 -0.26 -2.51
(0.10) (0.13) (0.11) (0.12) (0.20) (0.36)
Total Capital 0.11 1.61 0.15 2.10 0.03 0.59
(0.44) (0.54) (0.35) (0.58) (0.87) (0.88)
Voting/Total Capital Ratio -0.00 -0.09 -0.01 -0.33 -0.01 -0.72
(0.89) (0.91) (0.49) (0.78) (0.72) (0.54)
(Voting Capital)2 -1.91** -2.83** -1.88
(0.03) (0.02) (0.12)
(Total Capital)2 1.00 1.46 0.65**
(0.46) (0.42) (0.72)
(Voting/Total Capital Ratio)2 -0.02 -0.13 -0.19
(0.87) (0.48) (0.31)
Dummy Foreigners 0.12*** 0.11 0.11
(0.08) (0.13) (0.13)
Dummy Government -0.07 -0.06 -0.07
(0.52) (0.60) (0.56)
Dummy Institutional 0.16 0.14 0.15
(0.11) (0.15) (0.13)
Adjusted R2 0.80 0.81 0.80 0.81 0.80 0.80
*, ** and *** indicate statistical significance at the 1%, 5% and 10% levels, respectively.
12
 Table VIII shows the results of the 6 model specifications for the study of the payout of Brazilian firms,

considering the direct structure of ownership and control. Some variables present statistically significant coefficients, with

signs that confim what is predicted by the theory. The negative coefficient of the voting capital, statistically significant at

the 1% level, depending on the specification, confirms hypothesis 4, i.e., firms with a high concentration of voting rights

have a low payout. The positive coefficient of total capital, which is statistically significant at the 5% or 10% level,

depending on the specification, confirms hypothesis 5, i.e., companies with a high concentration of cash flow rights have a

high payout. Finally, the negative coefficient of the voting/total capital ratio confirms hypothesis 6, i.e., firms with a high

separation between voting and cash flow rights have a low payout. The quadratic variables (voting capital, total capital, and

voting/total capital ratio) did not present statistically significant coefficients.

There is no statistically significant difference between the payout of firms, as a function of the type of controlling

shareholders. Family-owned fims tend to present a lower payout when compared to the rest, and government-owned firms

tend to present the highest payout. According to what is predicted by the theory, the current asset/total asset ratio has a

positive coefficient, although it is not statistically significant. The size and the ROA present positive coefficients,

statistically significant at the 1%, 5% or 10% levels, depending on the specification. The risk and leverage have negative

coefficients, and the first is statistically significant at the 1% level.

Table IX shows the results of the 6 model specifications for the study of the payout of Brazilian firms, considering

the indirect structure of ownership and control. The results are mainly the same as the direct structure, but in the indirect

structure the p-values of variables related to corporate governance tend to be lower than in the direct structure, which means

that those variables have a higher statistical power in the indirect structure. In this way, the relationship between payout,

voting capital, total capital, and the separation between voting and total capital becomes statistically significant at the 1%,

5% or 10% levels, depending on the specification.

13
 Table VIII
The Effects of the Direct Structure of Ownership and Control on Dividend Payout of Brazilian Firms
Multiple linear regression analysis to evaluate the effect of the direct structure of ownership and control on the payout of
225 Brazilian companies in 2000. The dependent variable is the payout and the independent variables are: voting capital;
total capital; voting/total capital ratio; leverage; current asset/total asset ratio, firm size (log assets); ROA; risk (stock
volatility); (voting capital)2 ; (total capital) 2 ; (voting/total capital)2 ; and dummy variables indicating the type of controlling
shareholders. Six model specifications were tested: 2 for the largest, 2 for the three largest, and 2 for the five largest
shareholders. In all specifications, we included industry dummies (coefficients are not reported here). The p-values of the t-
tests are shown in parenthesis.
Largest 3 Largest 5 Largest
Shareholder Shareholders Shareholders
(1) (2) (3) (4) (5) (6)
Constant 0.02 -0.04 0.17 -0.35 0.17 -0.31
(0.90) (0.86) (0.39) (0.30) (0.40) (0.44)
Leverage -0.03 -0.04 -0.03 -0.03 -0.03 -0.03
(0.21) (0.19) (0.21) (0.22) (0.20) (0.20)
Current Asset/Total Asset Ratio 0.11 0.13 0.12 0.14 0.11 0.13
(0.54) (0.46) (0.48) (0.44) (0.52) (0.46)
Size 0.03* 0.04* 0.03* 0.03** 0.03* 0.03*
(0.00) (0.01) (0.01) (0.02) (0.01) (0.01)
ROA 0.35 0.36 0.39*** 0.39 0.42** 0.43***
(0.13) (0.14) (0.09) (0.11) (0.07) (0.08)
Risk -0.11* -0.11* -0.11* -0.11* -0.11* -0.11*
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Voting Capital -0.26* -0.21 -0.42* -0.57 -0.46* -0.46
(0.01) (0.67) (0.00) (0.51) (0.00) (0.62)
Total Capital 0.18*** 0.27 0.22** 0.55 0.26** 0.49
(0.10) (0.57) (0.05) (0.48) (0.05) (0.63)
Voting/Total Capital Ratio -0.00 -0.00 -0.01 -0.07 -0.02 -0.07
(0.62) (0.94) (0.46) (0.15) (0.43) (0.47)
(Voting Capital)2 -0.35 -0.78 -0.69
(0.33) (0.16) (0.22)
(Total Capital)2 0.46 0.07 0.05
(0.25) (0.90) (0.93)
(Voting/Total Capital Ratio)2 -0.00 -0.00 -0.00
(0.97) (0.13) (0.49)
Dummy Foreigners 0.02 0.02 0.02
(0.78) (0.73) (0.70)
Dummy Government 0.03 0.02 0.02
(0.72) (0.82) (0.77)
Dummy Institutional 0.01 0.02 0.00
(0.88) (0.75) (1.00)
Adjusted R2 0.19 0.17 0.20 0.20 0.20 0.19
*, ** and *** indicate statistical significance at the 1%, 5% and 10% levels, respectively.

14
 Table IX
The Effects of the Indirect Structure of Ownership and Control on Dividend Payout of Brazilian Firms
Multiple linear regression analysis to evaluate the effect of the indirect structure of ownership and control on the payout of
225 Brazilian companies in 2000. The dependent variable is the payout and the independent variables are: voting capital;
total capital; voting/total capital ratio; leverage; current asset/total asset ratio, firm size (log assets); ROA; risk (stock
volatility); (voting capital)2 ; (total capital) 2 ; (voting/total capital)2 ; and dummy variables indicating the type of controlling
shareholders. Six model specifications were tested: 2 for the largest, 2 for the three largest, and 2 for the five largest
shareholders. In all specifications, we included industry dummies (coefficients are not reported here). The p-values of the t-
tests are shown in parenthesis.
Largest 3 Largest 5 Largest
Shareholder Shareholders Shareholders
(1) (2) (3) (4) (5) (6)
Constant -0.29 -0.11 0.10 -0.70 0.20 -0.50
(0.21) (0.89) (0.71) (0.54) (0.45) (0.66)
Leverage -0.04 -0.04 -0.04 -0.04 -0.04 -0.04
(0.18) (0.19) (0.14) (0.15) (0.14) (0.15)
Current Asset/Total Asset Ratio 0.06 0.09 0.09 0.11 0.09 0.12
(0.71) (0.63) (0.60) (0.53) (0.59) (0.51)
Size 0.04* 0.04* 0.03* 0.03* 0.03* 0.03*
(0.00) (0.00) (0.01) (0.01) (0.01) (0.01)
ROA 0.44*** 0.44*** 0.44*** 0.44*** 0.45** 0.45***
(0.06) (0.08) (0.06) (0.07) (0.05) (0.07)
Risk -0.12* -0.12* -0.11* -0.10* -0.11* -0.10*
(0.00) (0.00) (0.00) (0.01) (0.00) (0.01)
Voting Capital -0.56* -0.15 -0.62* 1.62 -0.63* -0.74*
(0.01) (0.91) (0.00) (0.41) (0.00) (0.70)
Total Capital 0.72* 0.16 0.43 3.01 0.36 2.24
(0.01) (0.93) (0.12) (0.27) (0.21) (0.41)
Voting/Total Capital Ratio -0.14** -0.08 -0.07 -0.38 -0.05 -0.15
(0.04) (0.89) (0.42) (0.65) (0.60) (0.86)
(Voting Capital)2 -0.33 -0.23 -0.24
(0.60) (0.80) (0.78)
(Total Capital)2 0.47 1.39 1.07
(0.63) (0.29) (0.41)
(Voting/Total Capital Ratio)2 -0.03 -0.00 -0.03
(0.74) (0.97) (0.85)
Dummy Foreigners 0.02 0.01 0.00
(0.68) (0.92) (0.93)
Dummy Government 0.06 0.03 0.03
(0.46) (0.72) (0.67)
Dummy Institutional 0.03 0.01 0.01
(0.71) (0.91) (0.88)
Adjusted R2 0.20 0.18 0.20 0.19 0.21 0.20
*, ** and *** indicate statistically significant at 1%, 5% and 10%, respectively.

15
IV – Conclusion

The debate on corporate governance in Brazil gained importance in the last decade, when the relationships between

controlling and minority shareholders changed due to the privatizations and the entry of new investors in the economy,

specially foreign and institutional investors. The purpose of this paper is to analyze the effects of the ownership and control

structure on market valuation and dividend policy of Brazilian firms.

Results show a high degree of voting capital concentration. Even when there is no controlling shareholder, the

largest shareholder owns a significative portion of the voting capital. The firm is controlled, on average, by its 3 largest

shareholders. We can also note a significant difference between the voting and total capital owned by the largest

shareholders. This mechanism seems to be used by majority shareholders to keep the firm’s control without having to own

50% of the total capital.

Most firms are controlled by family groups, followed by foreign investors and, lessly, by institutional investors and

the government. On average, shareholding agreements are present in 23% of the Brazilian firms. Most of the companies

have a pyramidal structure, and it tends to be less used in government-owned firms and more frequent in family and foreign-

owned firms. The issue of non-voting shares is common in Brazil, and voting shares represent, on average, 53% of the total

capital. The percentage of voting shares on total capital is lower in firms controlled by families and institutional investors,

while government-owned companies have the highest percentage of voting shares on total capital.

The results of the tests show that there is a relationship, which is statistically significant in many cases, between

governance structure, market valuation, and dividend policy of Brazilian firms. The results are basically the same when we

use the direct and indirect structures, but indirect structure variables tend to have a higher statistical power. A possible

explanation is that indirect structure variables really measure who is the actual owner of the firm. Therefore, the study of

ownership and control should focus not only on the direct structure, but also on who is the ultimate owner of the companies.

16
V – References

BERLE, A.; MEANS, G. The modern corporation and private ownership. NY: MacMillan, 1932.

BURKART, M., GROMB, D., PANUNZI, F. Why High Takeover Premia Protect Minority Shareholders, Journal of

Political Economy, 106, 1: pp. 172-204, 1998.

CLAESSENS, S.; DJANKOV, S.; LANG, L. The Separation of ownership and control in East Asian corporations. Journal

of Financial Economics, v. 58, p. 81-112, 2000a.

CLAESSENS, S.; DJANKOV, S.; FAN, J.; LANG, L. Expropriation of minority shareholders: evidence from East Asia.

World Bank Policy Research Paper , n. 2088, Washington, 2000b.

JENSEN, M.; MECKLING, W. Theory of the firm: managerial behavior, agency costs, and ownership structure. Journal of

Financial Economics, v. 11, p. 5-50, 1976.

JOHNSON, S.; SHLEIFER, A. Privatization and corporate governance. In: 12 th Annual East Asian Seminar on

Economics, 2001.

LA PORTA, R.; LOPEZ-DE-SILANES, F.; SHLEIFER, A.; VISHNY, R. Law and finance. Journal of Political Economy,

v. 106, p. 1113-1155, 1998.

________. Corporate ownership around the world, Journal of Finance, v. 54, p. 471 -518, 1999.

________. Investor protection and corporate governance. Journal of Financial Economics , v. 58, p. 3-28, 2000a.

________. Agency problems and dividend policies around the world. Journal of Finance, v. 55, p. 1-33, 2000b.

________. Investor protection and corporate valuation. Journal of Finance, v. 53, n. 3, 2002.

LEAL, R.; CARVALHAL-DA-SILVA, A.; VALADARES, S. Estrutura de controle das companhias brasileiras de capital

aberto. Revista de Administração Contemporânea, v. 6, p. 07-18, 2002.

McCONNEL, J.; SERVAES, H. Additional evidence on equity ownership and corporate value, Journal of Financial

Economics, v. 27, p. 595-612, 1990.

MORCK, R.; SHLEIFER, A.; VISHNY, R. Management ownership and market valuation: an empirical analysis. Journal

of Financial Economics, v. 20, p. 293-315, 1988.

SHLEIFER, A.; VISHNY, R. A survey of corporate governance. Journal of Finance , v. 52, p. 737-783, 1997.

VALADARES, S.; LEAL, R. Ownership and control structure of Brazilian companies. Abante, v. 3, 2000.

17