why do managerial misconducts persist? the role of controlling … · 2017-03-29 · 2 i....

1

Why do managerial misconducts persist? The role of controlling shareholders in corporate governance

Kyung Suh Park*, Eunjung Lee**, Hasung Jang*

Summary

This paper investigates the role of controlling shareholders in the corporate governance of listed firms in Korea. It focuses on the conflicting incentives of insiders and outsiders of a firm at the stage of determining the corporate governance of their firm. Empirical tests show that the controlling shareholder of a firm in general resists the introduction of governance mechanisms that strengthen the monitoring power of outside investors, who, on the other hand, prefer to have more of them. On the other hand, controlling shareholders with large enough ownerships are shown to welcome improved governance, possibly due to their larger economic interests and confidence in their control.

Such a result implies that corporate governance as a monitoring mechanism may not properly be in place due to resisting insiders, and their exploitation of minority shareholders may persist. It suggests a necessity for regulatory intervention in setting corporate governance, and also for an active market for control to complement internal governance.

2

I. Introduction

It is a well established hypothesis that managers pursue their own interests at the expense of shareholders (Jensen and Meckling (1976)). Controlling shareholders are not an exception either, in that they also seek for their own private interests despite their large shareholdings. Johnson, La Porta, Lopez de Silanes, and Shleifer (2002) show that the controlling shareholders of European firms misappropriate corporate assets for their own benefit. Bae, Kang and Kim (2002) also show that the controlling shareholders of Korean firms also pursue their own profit through ‘tunneling’ based on the complicated ownership structure of their conglomerates.

An interesting question is why this kind of misconducts on the part of controlling shareholders persists despite the existence of diverse monitoring mechanisms in place. Existing literatures have mainly focused on the issues of moral hazard and pursuit of private benefit by controlling shareholders, while they assume that monitoring by outsiders cannot be perfect due to information asymmetry and related monitoring costs.

In this paper, we investigate whether controlling shareholders actively intervene in the early stage of setting the corporate governance of their firms and succeed to manipulate it to their advantage. We conjecture that controlling shareholders would find it much easier to manipulate the governance structure of their firms since they usually control the board of directors, which has the ultimate power to decide the overall structure of corporate governance. In this sense, controlling shareholders with concentrated ownership structure are harder to be monitored and checked for their behavior than professional managers.

The paper also provides an answer to the observed positive relationship between corporate governance and firm value. As many papers such as La Porta, Lopez de Silanes, Shleifer and Vishny (1999), Mitton (2002), Durnev and Kim (2004), and Black, Jang, and Kim (2003) confirm, corporate governance matters and affect firm values. Then, it is a question why firms do not optimize their governance so that shareholder values are maximized. Again, our answer is that the private incentives of insiders lead to a suboptimal corporate governance level, which is another form of moral hazard on the part of managers and controlling shareholders.

For the convenience of analysis, we select three items of governance mechanism that can be easily identified and quantified. They are the proportion of outside directors on the board of directors, the number of committees on the boards, and cumulative voting scheme. These are relatively new governance devices that have been introduced into Korean firms mainly after the Asian economic crisis devastated the Korean economy. The board of directors and its structure is believed to be the key factor that effectuates the monitoring function on management. Cumulative voting scheme, a fairly new concept in Korea, allows a shareholder to cast all of his votes, which is the number of shares owned multiplied by the number of directors to be elected, for one candidate. It is supposed to strengthen shareholder right by increasing the possibility of electing a director supported by minority shareholders, thereby can be a direct threat to existing management.

We also use the corporate governance scores collected by the Korea Stock Exchange and the Korea Corporate Governance Services over the past 3 years, and analyze their determining

3

factors. The results of empirical tests provide a different perspective on the incentives of controlling shareholders.

As conjectured, the empirical analyses show that insiders reveal strong resistance to the introduction of new governance schemes that will improve the monitoring on their behavior, despite the efforts of outside shareholders and regulatory authorities to strengthen the monitoring system.

2. Existing Literatures Recently, existing literatures on corporate governance have been focused on the

relationship between corporate governance and firm value. Morck, Shleifer and Vishny (1988), La Porta, Lopez de Silanes, Shleifer and Vishny (1999, LLSV hereafter), Mitton (2002), Claessens, Djankov, Fan and Lang (2002) are among the many on this line and Joh (2003) and Baek, Kang and Park (2004) cover Korean firms. Surprisingly, they mostly confirm the positive relationship between corporate governance and firm value, regardless of how they measure them.

On the other hand, a couple of papers have dealt with the determinants of corporate governance. Durnev and Kim (2003) provide a theoretical model and empirically test the relationship between the financial characteristics of a firm and corporate governance. They show that firms with good investment opportunity, higher sales growth rates and higher dependency on external financing would maintain a better corporate governance not to lose those good investment opportunities. Concentration of ownership to certain level also led to a better governance. However, their focus is not on the incentives of controlling shareholders, and they use governance scores, as evaluated by outside institutions, as a proxy for corporate governance, while we use specific governance schemes in addition.

Agrawal and Knoeber (1996) use simultaneous empirical models that analyze the relationship between ownership, corporate governance and firm value, and find that firms with founding managers tend to have lower proportion of outside directors. Weisbach (1988) and Klein (2002) also look into the incentives of insiders and show that there exists a negative correlation between the ownership of managers and the proportion of outside directors on the boards of directors, or on audit committees. Shivdasani and Yermack (1999) claim that CEO exercises major influence on the selection of new directors when the ownership distribution of his firm is dispersed, while it is the controlling shareholder under concentrated ownership structures. Yeh and Woidtke (2003) also show that, in Taiwanese firms, controlling shareholders with official position in his firm or with larger ownership tend to appoint more of his men on their boards.

In this paper, we focus on the incentives of controlling shareholders and outside investors as well in the context of affiliated business structure of Korean conglomerates. This paper tests two competing hypotheses regarding the relationship between ownership and corporate governance. First, it would be a natural choice for a firm to optimize on the use of governance mechanism since it is costly, and there would be a substitution effect between governance and

4

the concentration of ownership. For example, institutional investors can be a good monitor on the management and as such it can substitute for other governance mechanism such as outside directors. Second, block shareholders who have major ownerships in their firms might prefer a stronger monitoring system to protect their stakes, in which case we would observe a positive correlation between ownership and governance, which we term as a complement hypothesis. The paper tests whether specific types of investors tend to substitute for governance mechanisms or reinforce them.

The Korean data also provides an advantage in the analysis of the relationship between ownership and corporate governance. In this line of research, the issue of endogeneity among these two groups of variables has been an obstacle to overcome. As we have implicitly assumed in the introduction, and as many existing literatures have confirmed, ownership structure generally affects governance structure. On the other hand, governance structure might also affect ownership structure in a long run. For example, institutional investors would prefer to invest in companies with good corporate governance, which naturally increases the ownership of outside investors while lowering that of insiders.

In dealing with the endogeneity issue, researchers generally use two or three stage simultaneous models, but harder part of such model is to find an instrument variable that is closely related with one dependent variable, but not so with other dependent variables. However, in Korean case, the corporate governance mechanisms have been introduced only recently and the data allows us to resort to one-direction empirical analysis.

The Korean economy is also an interesting subject of analysis since it is dominated by Chaebols. Even though they are not necessarily classified as conglomerates or Chaebols, most of the listed companies in Korea have affiliates, and independent firms in its pure sense are rather exceptions in the Korean economy.5

Also, most of the listed companies on the Korea Stock Exchange have controlling shareholders, a feature that differentiates them from Keiretsu of Japan, eg.6 They exercise full control on the management of their firms, and are naturally termed ‘owner-managers’. In a separate paper by Jang, Kang and Park(20030, we could identify controlling shareholders for 364 firms out of 425 manufacturing firms sampled. The controlling shareholders of Korean Chaebols maintain their control with the help of affiliated ownership as well as their family ownership. As of the end of 2001, the average family ownership for those firms that belong to the 30 largest Chaebols in Korea was 8.27%, while their average affiliated ownership is 20.1%. On the other hand, for those firms that do not belong to the largest 30 Chaebols, the numbers are 21.51% and 10.79% respectively, showing a reversed pattern of ownership structure between Chaebol firms and non-Chaebol firms.7 In either case anyway, no outsider could

5 Jang, Kang and Park (2004) report that the proportion of listed companies without affiliates is less 5% of the manufacturing companies listed on the KSE in 2003 6 A few exceptions are commercial banks, and privatized public corporations such as KT and POSCO. 7 The dependence on affiliated ownership tends to enlarge with the size of the Chaebols. As of the end of 2003, the controlling family’s average ownership for the ten largest chaebol groups is only 4.63% of outstanding shares, while their affiliated ownership is 44.6%

5

possibly challenge their control, mainly due to the interlocking ownership structures among affiliates, even though their capability and integrity as managers were in doubt.8

This is also the first paper that analyzes the determining factors of the cumulative voting system. We expect that the analysis will provide an interesting perspective on the incentives of controlling shareholders or managers as the new voting scheme is directly related with the ownership structure of a firm, and challenges controlling shareholders on the control of their companies.

II. Empirical Hypotheses and Data 1. Hypotheses and Variables This section provides empirical hypotheses that relate corporate governance with firm

characteristics, based on existing theories and empirical results, and identifies variables that will be used to test the hypotheses.

Ownership and Corporate Governance As confirmed in existing researches, ownership structure is a part of corporate governance

in its broad sense, and it also affects other elements of corporate governance. Controlling shareholders, e.g., have a strong incentive to monitor the management of firms and can be the most important part of corporate governance.

Existing theories and empirical analyses generally approach the ownership structure of a company by identifying the ownership of block shareholders such as controlling shareholders, corporate shareholders, institutional investors and financial institutions. In this paper, block shareholders affect corporate governance of a firm in two ways, which presents competing hypotheses. The first one, which we term ‘substitute hypothesis’, assumes that higher ownerships of block shareholders would substitute for other governance mechanisms as the latter incurs costs on companies. Firms thereby adjust the level of corporate governance given the monitoring role of block shareholders. This would be more of the case if block shareholders actively monitor the management of their firm.

On the other hand, as Durnev and Kim (2003) have claimed, higher ownership may induce block shareholders to further improve the corporate governance of their firm as they will have larger economic stake to protect. This is what we call ‘complement hypothesis’.

It is our conjecture that one of these hypotheses would more likely hold depending on who the block shareholders are. In case of a controlling shareholder, who usually participates in the management of his firm, he may not find it palatable to have a governance structure which monitors the management too tightly if he derives private benefit of control as well as

8 The challenge by the Sovereign Fund on SK Corporation in 2003 was one of the first in kind.

6

economic profit from his ownership in the firm.9 This, however, would not be true for institutional investors who have no such control benefits and only seek for higher firm value.

Therefore, we may observe less strict monitoring mechanism with increasing ownership by controlling shareholders, which we may term ‘control hypothesis’ to further differentiate it from the substitute hypothesis as their purpose is not to save monitoring costs, but to secure more control. Of course, it is not easy to differentiate between these two hypotheses empirically since we would observe the same direction of signs of the coefficients for the controlling ownership variable in both cases. We test alternative empirical models and use diverse proxy variables to have a better understanding of the incentives of controlling shareholders.

We also analyze the role of affiliated ownership, which provides interesting information about the incentives of controlling shareholders. As affiliated firms are under the control of controlling shareholders and usually do not intervene in the management of other affiliates, their existence would not substitute for internal monitoring function. Therefore, if we observe a negative effect of affiliated ownership on governance scheme, it is a strong indication that controlling shareholders exploit the affiliated ownership only to fortify their control by rejecting outside monitoring.

LLSV (1999), Mitton (2002), and Baek, Kang and Park (2004) have interpreted affiliated ownerships as representing the discrepancy between the cash rights and the control rights of controlling shareholders, which tends to lower firm value. We interpret the affiliated ownership as a device to resist the introduction of the new monitoring mechanism, thus eventually leading to lower firm values.

It would also make some difference if a block shareholder assumes a position on the management and officially participate in the management of his firm. A dummy variable, which takes the value of 1 if CEO has more than 5% of ownership and 0 otherwise, would be used.10 We conjecture that its coefficient would be negative as the owner-manager would have stronger incentive to resist outside monitoring since he is now more of a manager than a shareholder. On the other hand, controlling shareholders in Korean firms practically have a full control on the management even without any official positions. In this case, the CEO dummy may not have any effect on the governance of a firm.

One technical issue needs to be resolved regarding the use of ownership variables in the empirical model. We have considered only the effect of ownership on corporate governance in the discussion. But, the truth is that governance can also affect ownership structure. A good example would be an investment strategy based on corporate governance, employed by some institutional investors in their portfolio management. In that case, firms with good corporate governance would have higher outside ownership and naturally lower inside ownership, and we would observe a positive correlation between institutional ownership and corporate governance, and a negative relationship between controlling ownership and corporate governance, thus 9 In this paper, we define controlling shareholder as an individual whose ownership including that of his relatives exceeds 5% of the total shares issued. 10 It would have been better if we had a dummy variable denoting whether the controlling shareholder has a position in his company or not.

7

falsely accepting the complementary hypothesis for institutional investors and control hypothesis for controlling shareholders.

Existing literature such as Mak and Li (2001) and Agrawal and Knoeber (1996), minding the endogeneity issue, used simultaneous empirical models. One problem with using a simultaneous model is that we need an instrumental variable which is correlated with one dependent variable but not with others. Existing papers are very loose in this aspect mainly because identifying such a variable is not an easy task. In our case, we need to find a variable that is correlated with ownership, but not with governance structure, which we guess would be a difficult job.

In this regard, our Korean samples offer a good solution for the endogeneity issue since the corporate governance mechanisms we are going to analyze had been introduced mainly after the economic crisis and not much time has passed for them to affect the ownership structure of Korean firms. Even Mak and Li (2001) and Agrawal and Knoeber (1996) showed that it is ownership that affects corporate governance, but not the other way round under their simultaneous models. We also use lagged variables for ownership and other firm specific variables to further minimize the endogeneity problem.

Business Structure and Corporate Governance Another major factor that can affect the governance structure of a firm is business structure,

among which conglomerates have been a focus of interest since they offer a very nice environment for controlling shareholders to pursue their own benefits through transactions among affiliated firms. Tunneling, as termed by existing literature, has been widely reported in European conglomerates by Johnson, La Porta, Lopez de Silanes, and Shleifer (2002), and also in Korean conglomerates by Bae, Kang and Kim (2002). Conglomerate business structure also allows controlling shareholders to maintain their control through affiliated ownerships.

In this paper, we use a dummy variable which takes the value of 1 if a firm belongs to one of the 30 largest Chaebols as defined by the Fair Trade Commission for their regulatory purpose. We conjecture those firms that belong to a Chaebol suffer from agency problem more than independent firms do, and therefore may have more stringent monitoring mechanism as demanded by outsider investors. But, the dominance of controlling shareholders through affiliated ownership may also weaken it on the contrary, which is to be confirmed by empirical analysis.

Firm Size and Corporate Governance Since governance mechanisms consume corporate resources, we expect that larger firms

would have better corporate governance and we include asset size as an explanatory variable. Most of the monitoring system such as the board of directors, subcommittees including audit committee, internal control system, financial reporting and disclosure system incur financial costs, most of which is of fixed component and can be borne more efficiently by larger firms.

8

More complicated business structure of large firms may also require better corporate governance.

We also use a dummy to accommodate the effect of regulatory requirements on corporate governance based on asset size. 11 A dummy variable, which takes the value of 1 if the total asset size of a firm exceeds 2 trillion won, and 0 otherwise, is included. We also separate the sample firms depending on whether their asset sizes exceed 2 trillion or not, and run the regressions for the two groups since variables other than asset size may have different effects on governance decision in these two groups.

Other Financial Characteristics and Corporate Governance We also expect some financial characteristics would affect the governance decision and

need to be controlled. We include control variables that represent profitability, liquidity, financial structure, and growth rates of firms.

The effects of profitability on corporate governance may be two way. High profitability implies good capability of management and monitoring on them may not be necessary. On the other hand, high profitability affords the company better governance system. Outside investors may also demand for better governance as they have more economic stakes to lose.

Higher liquidity as measured by the amount of free cash flow would lead to better governance mechanism since it can be appropriated by the management for their private benefit. It also allows firms to maintain costly monitoring system. The growth potential would be also related with better governance since those firms with high growth rates have more to lose from lack of investment capital, and would try to satisfy outside investors with better governance (Durnev and Kim (2003)).

We also include debt ratio and bank loan ratio. Higher debt ratio implies larger amount of interests and principals to be paid periodically, and management would be under pressure to make enough cash flows to cover the debt payment, which can be done through more efficient management (Grossman and Hart (1982)). We expect debt ratio is negatively correlated with corporate governance mechanism, thus substitutes for them. Among the diverse types of debt, bank loan is of interest since banks, as larger creditors with long run relationship with firms, are supposed have an incentive and informational advantage to monitor their client firms (Park and Yoon (2001)).

2. Samples and Data We analyze the corporate governance of Korean firms listed on the Korea Stock Exchange

(KSE) as of the end of 2001. For the financial data, we use the one from the Korea Listed

11 The listing law requires one quarter of the boards of listed firms to be filled with outside directors with its minimum number to be one. As for the firms with asset sizes over 2 trillion won, the minimum proportion of outside directors is one half, with the minimum number of three.

9

Company Association. Ownership data have been collected using the Electronic Disclosure System of the KSE, and the governance data was provided by the Korea Corporate Governance Service, which is a public governance scoring agency in Korea. We exclude financial companies from our samples, and are left with 438 manufacturing companies listed on the KSE.

<Table 1> shows the summary statistics of major variables. The average ratio of outside directors on the boards of the sample firms is 0.3095, which is relatively low compared with the case of the U.S. firms, which on average have 80% of their boards filled with outside directors.12 The average number of committees for Korean listed firms is 0.283, implying many of the sample firms do not have any committees. This would be mainly due to the small number of directors to form any committee at all. On the other hand, the largest number of board committees is six, suggesting there exists huge difference in the corporate governance of Korean firms.

Only 5% of the sample firms have introduced the cumulative voting system. This is rather surprising since listed firms are required by the law to explicitly specify on their corporate charter that they are not introducing the new scheme. Considering that controlling shareholders of the sample firms secure their control with the average ownership of 29.85%, also that it is not easy to aggregate dispersed votes owned by outside shareholders partly due to some institutional aspects, and that the new voting scheme is effective only when two or more of outside directors are to be elected, such a low introduction rate is rather surprising. Finally, 22.6% of the CEOs of the sample firms have more than 5% of the shares including the ownership by their families.

<Table 2> shows the correlation coefficients of the variables. Controlling ownership is negatively correlated with the ratio of outside directors, number of committees, and the introduction of cumulative voting scheme, while institutional and foreign ownerships are positively correlated with them as expected. Surprisingly, affiliated ownership is positively correlated with ratio of outside directors. Asset size and Chaebol dummy also show strong correlations with the ratio of outside directors and the number of committees.

[<Table 1> and <Table 2> here]

III. Empirical Models and Results

In this section, we set up empirical models and test our hypotheses. Dependent variables

are the ratio of outside directors, the number of committees on a board, and a dummy for the introduction of cumulative voting scheme. Our focus is on the role and the incentives of controlling shareholders in deciding the governance mechanisms. Cross-sectional regressions are employed testing the hypotheses reviewed in the previous section.

12 Korn/Ferry international 29th annual board of directors study 2002.

10

1. Ownership and The Composition of the Board One of the key factors in evaluating the corporate governance system of a firm is the

number of outside directors on its board, therefore its proportion. The boards of directors are known to be the most important institution that can directly and effectively monitor and check the management of a company. The ratio of outside directors, which we use in the following models, takes its importance as the major managerial decisions delegated by shareholders to the boards are decided by the boards. If more than half of the directors are executive directors, the management would be able to force whatever agendas they want to see implemented.13 Even though the listing regulation requires the minimum number and the proportion of outside directors, we find that many listed firms are electing outside directors beyond the required level, which suggests that some internal factors are affecting the decision.

The explanatory variables are block ownerships such as family ownership, affiliated ownership, institutional ownership, and foreign ownership. CEO dummy, which takes the value of 1 if the manager has more than 5% of shares issued, and 0 otherwise, and Chaebol dummy, which takes the value of 1 if the firm belongs to one of the 30 largest Chaebols are also included.

As for financial characteristics of firms, we use cash flow from operation to control for the agency problem and also for profitability. In addition to asset size, asset size dummy which takes the value of 1 if asset size exceeds 2 trillion won and 0 otherwise is also included to reflect the effect of the regulatory requirement.14

Debt ratio and bank loan ratio are also included to reflect their monitoring role, and sales growth rates measured over the prior 3 year periods control for the opportunity costs of the shareholders. Finally, the total number of directors is also included as it shows a negative correlation with the ratio of outside directors in Mak and Li (2001). Following is the basic empirical model and the variables.15

εββββββββββ

βββββ

+∗+∗+∗+∗+∗+∗+∗+∗+∗+∗+

∗+∗+∗+∗+=

mmyIndustryDuLoanRatioLeverageBoardSizeSizeDummySizeGrowthCFnInstitutioForeign

myChaebolDumCEODummyAffiliatesFamilyOBOARD

14131211

1098765

43210

OBOARD : Proportion of outside directors on the board Family : Family ownership of controlling shareholders Affiliates : Affiliated ownership CEO Dummy: Dummy variable, which takes the value of 1 if the CEO owns more than 5%

of shares.

13 We also tried a model using the number of outside directors exceeding the minimum requirement as the dependent variable, but the explanatory power of the model was much low than the current model. From the management’s point of view, the proportion of outsider directors matters because the boards generally use majority rules in their decision.

14 We also divided and analyzed the sample firms into two groups on their asset sizes exceeding 2 trillion or not, but the results were the same as the current models. 15 We controlled for multicollinearity problem using the state factor test and VIF test

11

Chaebol Dummy: Dummy variable, which takes the value of 1 if the firm belongs to the 30 largest Chaebols

Foreign : Foreign ownership Institution : Institutional ownership CF : Cash flow from operations/Total assets Growth : Average growth rates of sales over the last 5 years Size : log(Total asset) Size Dummy : Dummy variable, which takes the value of 1 if the asset size of the firm is

over 2 trillion wons Board Size : Total number of directors on the board Leverage : Total debt/Total assets Loan ratio: Total bank loans/Total debt Industry Dummy : Industry dummy <Table 3> shows the results of the empirical tests. In regression (1) of <Table 3-a>, the

family ownership has the coefficient of -0.1876, which is significant at the 1% level, and suggests that higher ownership tend to lower the proportion of outside directors as we conjectured. It is still significant at the 5% level even after controlling for other financial characteristics in the regression (2). The result supports the ‘substitute hypothesis’ which argues that existence of a controlling shareholder would substitute for other monitoring mechanism as he can efficiently monitor the managements, saving the costs of other mechanism. However, it also supports the ‘control hypothesis’ where controlling shareholders do not like outside directors since they would be the subjects of monitoring.

The negative linear relationship between family ownership and the proportion of outside directors is possibly due to the increasing controlling power of controlling shareholders with their ownership, which allows them to control the board of directors and therefore affects the decision on the issues of setting their corporate governance.16 We come back to this issue later on with further empirical evidence.

Of the control variables, asset size and asset size dummy both show a very strong influence on the proportion of outside directors. Larger asset size allows firms to bear monitoring costs more easily. The coefficient of size dummy implies that the proportion of outside directors increases by 17.31% if a firm has an asset size over 2 trillion won even after controlling for asset size. Considering the average board size of six directors for the sample firms, the regulation requiring half the board to be filled with outside directors has the effect of adding about one more outside director on their boards of directors. Consistent with existing literature, the size of the board itself has a negative impact on the ratio of outside directors, and significant at the 10% level.

Sales growth rate, cash flow and CEO dummy do not show any significance. Sales growth rate and cash flow represent profitability and therefore the opportunity cost of controlling

16 We checked and found no non-linearity problem in case of the family ownership variable.

12

shareholders as they have more to lose if they cannot obtain external financing due to bad corporate governance (Durnev and Kim (2003)). However, the empirical results show that Korean owner-managers do not have such a long-term incentive to maintain good governance. The insignificance of the CEO dummy also suggests that controlling shareholders of Korean firms are de facto managers, and therefore it does not make any difference whether they are officially on the management or not.

Regression (3) shows the effect of affiliated ownership on the board composition. The coefficient is positive with significance at the 10% level. However, with the control variables added in regression (4), it loses the significance and even turns negative in regression (5) with control variables. The family ownership is still negative and significant at the 5% level.

<Table 3-b> shows the influence of outside shareholders. In regression (1), higher foreign ownership leads to higher proportion of outside directors with significance at the 1% level, which supports our ‘complement hypothesis’ as outside investors require stronger monitoring mechanism to protect their interests. Note again that we don’t have to worry about the reversed causality since ownership precedes the new governance mechanism.

However, in regression (2) where an interaction variable between foreign ownership and Chaebol dummy is added, the coefficient of the foreign ownership variable loses its significance and only the coefficient of the interaction variable shows a positive effect on board composition, which is significant at the 1% level. It would be only in those Chaebol firms that foreign investors ask for higher proportion of outside directors, while they are not very much concerned about the composition of the boards of non-Chaebol firms. This is possibly due to their worry on higher possibility of tunneling among affiliated firms of Chaebols.

The coefficient of institutional ownership in regression (3) also shows positive significance at the 1% level, while its effect is absorbed again by the interaction variable between the ownership and the Chaebol dummy in regression (4). Both the results support the ‘complementary hypothesis’, and confirm our conjecture that outside investors would prefer a stronger monitoring system.

In regression (5), we have both the inside and outside ownerships with control variables added. However, the results are not consistent with the previous models as the coefficient of the interaction variable between the institutional ownership and the Chaebol dummy is significant and negative at the 10% level, while other ownership variables lose their significance. We guess this is partly due to the multicollinearity between the variables. First of all, the family ownership has a very strong negative correlation with other ownership variables in <Table 2>. Also, we observe strong positive correlations between foreign ownership and some of the control variables such as asset size and cash flow. Institutional ownership also shows positive correlation with asset size as well. We controlled for the multicollinearity problem for all the empirical models, but we still suspect that strong correlations among the variables distort the results of the regressions.

In summary, the empirical results in <Table 3> show that the insiders of Korean firms prefer not to have outside directors. This is mainly because they are managers as well as shareholders, and enjoy private benefit from control. On the other hand, outside investors prefer

13

to have stronger monitoring system, especially for those firms belonging to Chaebols. The results imply that firms with controlling shareholders would set their corporate governance below the level asked for by outside investors or by regulatory authorities despite its negative impact on firm value. The shirking behavior of owner-managers is another form of moral hazard and is aimed at setting the governance of their firms to their advantage.

Existing theory claims that increasing ownership by the management would align its incentive with that of shareholders, thus reducing moral hazard. However, our empirical results show that higher controlling ownership may lead to lax monitoring system and more moral hazard. In this sense, it provides another explanation why we observe declining Tobin’s Q after certain level of controlling ownership in existing literature (Morck, Shleifer and Vishny (1988), and McConnell and Servaes (1990)). They ascribe it to decreasing possibility of hostile takeover and managerial entrenchment, thereby. We claim that it is also due to the lack of internal monitoring and increasing conflicts of interest between insiders and outsiders as the former find it easier to manipulate internal governance structure to their taste.

[<Table 3> around here]

2. Ownership and The Number of Committees of the Boards The committees of the board of directors have their importance as practical channels for

implementing managerial functions such as nomination, evaluation and compensation of directors including CEO above all. Of course, those functions can be also taken care of by the board itself, but meetings where all the board members participate can be inefficient, and expertise of participants and conflict of interests on some agendas can be of an issue. The fact that there exist committees that are specialized on their major functions also signals the determination of the management to maintain an efficient and effective board.

A priori, we can expect the number of committees would be closely related with the number of directors since partially different sets of directors should be on the committees. Otherwise, there is no use for additional committees. The correlation between these two variables is 0.361 and significant at the 1% level. We set up similar empirical models as in the case of the board composition, while the dependent variable is the number of committees.

<Table 4> shows the results of the empirical tests. In regression (1) of <Table 4-a>, we regress the number of committees on family ownership, whose coefficient is -0.7852 and significant at the 1% level. However it loses its significance when the control variables are added in regression (2). On the other hand, the coefficient of affiliated ownership is significant and negative at the 1% level even after controlling for other variables in regression (4), supporting either the substitute or the control hypothesis. As affiliated firms are not generally known for their monitoring on other affiliates, the result cannot support the substitute hypothesis. Their ownerships are rather exploited by controlling shareholders for the purpose of neutralizing their boards. In regression (5) with control variables, we again confirm the negative and significant coefficient of the affiliated ownership significant.

14

Existing literature interprets affiliated ownership as increasing the discrepancy between the cash rights and the control rights of controlling shareholders, and found negative relationship between the wedge and firm values. LLSV (1999), for example, reasoned that the negative relationship was mainly due to unchallenged monopolistic control by controlling shareholders who pursued their own interests at the expense of outside investors. This paper confirms that the increasing discrepancy through higher affiliated ownership may lead to malfunctioning of the boards, and eventually help controlling shareholders misappropriate the property of their firms.

Of the control variables, asset size and asset size dummy both have positive influence on the number of committees as expected with significance at the 10% and 1% levels respectively. On the other hand, the coefficient of bank loan ratio is negative and significant at the 5% level, suggesting their monitoring function replaces the role of the committees, and supporting the substitute hypothesis. Sales growth rate, debt ratio, Chaebol dummy, and CEO dummy do not show any significance again as in the board composition models.

<Table 4-b> analyzes the effect of outside ownership on the number of committees. In regression (1), the coefficient of foreign ownership is positive and significant at the 1% level, again confirming the complement hypothesis. It maintains its significance in regression (2) where the interaction variable between the foreign ownership and Chaebol dummy, whose coefficient is also significant and positive, is added. It shows again that the monitoring effect of foreign investors is more intensified for those firms that belong to Chaebols.

The institutional ownership also shows positive and significant effect on the board committees in regression (3), supporting the complement hypothesis. But, in regression (4), it loses the significance and the interaction variable between institutional ownership and Chaebol dummy shows significantly positive coefficient at the 1% level. However, in regression (5) where all the outside ownerships and control variables are included, only the foreign ownership survives and maintains the positive significance. Unlike in the board structure model, asset size is not significant, while asset size dummy is significant and positive at the 1% level reflecting the regulatory effect.17 Finally, the bank loan ratio is negative and significant at the 5% level, again supporting the substitute hypothesis.

<Table 4-c> show the results of analysis of the models that include both the inside and outside ownerships, and other control variables. In regressions (1), (2), (3) and (4), the control effect of inside ownership and the complement effect of outside ownership are confirmed again with varying degree of significance.

In summary, inside ownerships tend to reduce the number of board committees while outside ownerships tend to increase it. However, the overall impact of the former is shown to be larger than the impact of the latter as we measure it from the sizes of the coefficients. It again confirms our previous results on board composition, and shows that controlling shareholders do not like a well functioning board and that they actually succeed in partially neutralizing it.

17 Listed firms with asset size over 2 trillion wons are required to have a nomination committee and an audit committee, at least two thirds of which should be outside directors.

15


3. Ownership and Cumulative Voting Scheme Cumulative voting system is a relatively new concept in the field of corporate governance

and has been introduced to Korean firms for the purpose of electing directors favored by minority shareholders. 18However, given the existence of a controlling shareholder with a reasonable level of friendly ownership, minority shareholders with dispersed ownerships would need to mount a severe effort to elect a director of their choice.19 As such, before the introduction of the new scheme, we expected many firms would introduce it. However, the initial reaction by Korean firms beat our expectation. Out of the 438 sample firms, only 23 firms introduced the cumulative voting system as of the end of 2001 and has not since changed much, reflecting the strong reservation on the part of management or controlling shareholders. The number is more surprising since those listed firms need to explicitly reject the introduction of the new scheme on their corporate charter.

In this section, we analyze the types of firms that introduced the new voting scheme. As the dependent variable is a dummy variable that takes the value of one or zero depending on whether firms introduce the scheme or not, we use a Logit model for empirical tests. Since it is the total inside ownership that matters in the voting for an election of a new director, we use the sum of family ownership and affiliated ownership as an explanatory variable.

<Table 5> shows the results of the analysis, which reveal new facts related with the incentives of controlling shareholders. The inside ownership is not any more in linear relationship with corporate governance. As the regression (1) shows, the linear term of inside ownership shows a negative coefficient, which is significant at the 1% level, while the squared term shows a positive coefficient, also significant at the 1% level. The result suggests that controlling shareholders with relatively smaller ownership are more reluctant to introduce the cumulative voting scheme, while those with larger ownership are willing to introduce it.

To have a further understanding on the different incentives of controlling shareholders of different ownership levels, we use stepwise ownership variables as in the case of Morck, Shleifer and Vishny (1988). The first ownership dummy takes the value of the inside ownership if it is less than or equal to 20%, and takes the value of 20% otherwise. The second ownership dummy takes the value of 0 if the inside ownership is less than or equal to 20%, takes the value of the ownership minus 20% if the inside ownership is between 20% and 50%, and 30%

18 Korean Commercial Law, Article 388, Clause 2-3 requires Korean firms to introduce the cumulative voting scheme, but allows them to reject it through their articles of incorporation. 19 One simple way to estimate the number of shares needed to elect directors of your choice is (ND * TS)/(TD+1) + 1, where ND is the number of directors a shareholder wants to elect, TS is the total number of shares issued, and TD is the total number of directors to be elected (See Solomon and Palmiter (1994)). For a typical firm listed on the KSE, minority shareholders would need about 25% of collective ownership to elect an outside director of their choice when there are 3 slots of directors to be elected.

16

otherwise. The third ownership dummy takes the value of 0 if the inside ownership is less than 50%, and takes the value of the ownership minus 50% otherwise.

In regression (3), the coefficient of the inside ownership dummy 1 is negative and significant at the 6% level, suggesting that controlling shareholders with less than 20% of controlled ownership are strongly against the introduction of the new voting scheme. On the other hand, the coefficient of ownership dummy 2 is also negative but insignificant, which implies that the managements with controlled ownership between 20% and 50% are less worried about a potential threat to their control. And, the inside ownership dummy 3 shows a positive coefficient, which is significant at the 6% level. We guess the result is due to their confidence about their control as well as their larger economic stakes in their firms. The statistical significances of those variables are still larger as we add control variables in regression (4). We interpret the increasing willingness of controlling shareholders with larger ownership as the sign of confidence in their control over the election of new directors.

Both foreign and institutional ownership do not show any significance in regression (5) and (6). It looks like that it is mainly the management and controlling shareholders who are more concerned about the new voting scheme, while outside investors do not pay much attention to it.

As for financial control variables, they do not show significant effects on the introduction of the new voting scheme except the case of asset size, which is contrasted with the cases of board structures. We conjecture that this is mainly because the voting scheme does not require much financial resources, unlike the new board systems which can incur substantial and direct costs on firms.

In summary, we interpret the results as another piece of compelling evidence that supports the control hypothesis of inside ownerships. The new voting system affects who would be on the board, and it directly hurts controlling shareholders or management since outside shareholders would resort to the scheme only when they want to challenge the insiders. It has been actually a major criticism on the outside director system in Korea and also in other countries that outside directors are rather disoriented and are not active enough to represent minority shareholders. But, in case of the new voting system, outside directors who would monitor and bother the management would be elected. This potential threat is exactly reflected in the low introduction ratio of the voting scheme in the listed firms of Korea.


4. Ownership and Corporate Governance Scores A stylized fact in corporate governance area is that there exists a positive correlation

between corporate governance and firm value. As LLSV (1999), Mitton (2002), Durnev and Kim (2003), and Black, Jang, and Kim (2003) have confirmed, corporate governance matters and affects firm values.

Then, it is a question why firms do not improve their corporate governance so that their shareholder values are further increased. One possible answer is that the current state of

17

corporate governance is optimal already. That is, it is too costly for a firm to improve its corporate governance. However, Park and Lee (2004) show that the difference in the average Tobin’s Qs of those firms in the highest quartile of corporate governance scores and those in the lowest quartile is about 0.39. Considering that the average market value of those Korean firms is 0.8 billion U.S. dollars, potential increase in shareholder value due to improved corporate governance would amount to 0.32 billion U.S. dollars on average, which would well exceed any costs related with upgrading the corporate governance of those firms. Park and Lee (2004) even show that the individual category of governance mechanism such as board composition or disclosure policy, which can be rather easily upgraded, has a positive effect on firm value.20

In the following, we conjecture again that the private interests of controlling shareholders would deter firms to attain optimal corporate governance. For the empirical analysis, we use the governance scores of Korean firms collected over the last 3 years from 2001 through 2003.21 The annual surveys contain over 100 questions on corporate governance of Korean firms, and evaluate shareholder rights, structure and operation of the boards, disclosure and managerial transparency, and internal control system among others (See <Table 8> for a summary of the questionnaire). According to the surveys, the corporate governance of Korean firms shows wide difference among themselves (see <Figure 1> for the distribution of scores for the year 2003). The highest score was 219 while the lowest was 117. Again, as in existing literature, we can confirm that there exists a positive correlation between corporate governance score and firm value as measured by Tobin’s Q (see <Figure 2>).

The advantage of using the scores instead of individual governance mechanism is that the governance scores are more comprehensive in evaluating the overall corporate governance of a firm than a specific governance mechanism, and also that it allows us a larger number of samples to increase the power of the models.

<Table 6> shows the results of a panel data analysis that covers 3 years of corporate governance scoring. As we can find from the table, the explanatory power of the regressions have increased as the higher R-squares of the regressions and the increased significance of the coefficients of the explanatory variables show.

In regression (1) of <Table 6-a>, the coefficient of family ownership is negative and significant at the 1% level, again confirming our conjecture that controlling shareholders do not like a good corporate governance. The significance is maintained even if we add control variables in regression (2). The coefficient of affiliated ownership in regression (3) is not significant, but with control variables added, its significance increases to the 1% level in regression (5), where both the inside ownerships show negative and significant influence on the

20 In LLSV (1999) or Mitton (2002), they use the ‘wedge’ between cash rights and control rights of controlling shareholders as a proxy for the conflicts of interest between controlling shareholders and minority shareholders. As the ownerships of controlling shareholders or affiliated firms cannot be adjusted easily, we can expect the negative correlation between the wedge and firm value would persist in their model. 21 The KSE initially, and then, the Korea Corporate Governance Service (KCGS), a subsidiary of the KSE, has been in charge of evaluating the corporate governance of listed companies in Korea

18

governance score of Korean firms. Other financial variables also show expected signs and significance. We did not use the asset size dummy in the model since the evaluation process already reflects the size effect in their scoring.

<Table 6-b> confirms again that outside shareholders tend to strengthen the internal governance of their firms. In regression (2), the interaction variable between foreign ownership and Chaebol dummy is again positive and significant at 1% level, confirming previous results, but loses its significance in regression (5). .

5. Ownership and Change in Corporate Governance Scores In this section, we further analyze the incentives of controlling shareholders by delving into

their effect on ‘the change’ in corporate governance over time. Since we have conjectured that controlling shareholders would prefer weak corporate governance, their preference will be also reflected over time such that those firms with higher controlling ownership tend to show slower improvement in their corporate governance scores; they will be slow to increase outside directors, slow to improve the transparency of management, slow to enhance minority shareholder rights, and also slow to tighten their internal control system.

<Table 7> shows the results of the analysis. The dependent variable is the change in the corporate governance scores of sample firms between 2002 and 2003.22 Here, again we observe that the inside ownership, which is defined as the sum of family and affiliated ownership, is not in linear relationship with corporate governance. As the regression (1) shows, the linear term of inside ownership shows a negative coefficient, which is significant at the 1% level, and the squared term shows a positive coefficient, also significant at the 1% level. The result suggests that controlling shareholders with relatively smaller ownership are more reluctant to improve the corporate governance of their firms, while those with larger ownership are more willing to improve it.

In regression (3) and (4), we use stepwise ownership variables as used in the previous section, where the coefficient of the inside ownership dummy 1 is negative and significant at the 5% level, suggesting that controlling shareholders with less than 20% of controlled ownership are less willing to improve the corporate governance of their firm. This implies that their private benefits of control under weak monitoring by outside investors exceed any incremental value they can obtain from higher market value of their shares under better corporate governance.

On the other hand, the coefficient of ownership dummy 2 is positive, even though it is not significant, which implies that the managements with controlled ownership between 20% and 50% are rather neutral to the level of corporate governance. Finally, the inside ownership dummy 3 shows a positive coefficient, which is significant at the 10% level. We interpret that the increasing confidence about their control as well as their larger economic stakes in their

22 We also checked the difference in governance scores of the year 2001 and 2002, and came up with similar results.

19

firms renders controlling shareholders to improve the corporate governance of their firms. The statistical significance of those variables is maintained as we add control variables in regression (4).

One possible reason why the effect of controlling ownership is linear in the analysis of ‘the level’ of corporate governance while it is not so in the case of the analysis of ‘the change’ in corporate governance would be that those controlling shareholders with higher ownership did not have enough time to improve the corporate governance of their firms in Korea. If the pattern repeats over time so that the corporate governance of those firms with relatively high controlling ownerships continues to improve, we also expect to observe a non-linear relationship between controlling ownership and the level of corporate governance itself.

Finally, <Table 7b> shows the effect of outside ownerships on the change in the corporate governance of sample firms. We do not observe any significant variables in this case. Possibly, it is the governance that affects the ownership of outside investors, not vice versa.

IV. Summary

The paper analyzes the determinants of the corporate governance of Korean firms, focusing

on the incentives of controlling shareholders. It uses the data on corporate governance scores of Korean firms over the period of 2001 through 2003 as well as their individual governance mechanisms for an empirical analysis.

The paper shows that controlling shareholders of Korean firms tend to resist good corporate governance that will monitor and check their decisions. The result is possibly due to the fact that they tend to assume managerial role as well as monitoring role as block shareholders. Naturally, they would enjoy the private benefit of control and would not like being monitored or challenged by outside investors in the matter of management.

This paper contributes to existing literature on corporate governance by showing why the conflicts between management and outside investors are not easily resolved by internal corporate governance mechanism. It shows that controlling shareholders as managers take measures to avoid monitoring. Since they have controlling ownership, they are in a better position to deter the introduction of any monitoring mechanisms. However, in a stepwise analysis, the paper shows that those controlling shareholders with large enough ownership are shown to welcome improved governance, possibly due to their larger economic interests and confidence in their control.

In summary, the paper shows that there exists a limit to internal governance mechanisms as they can be neutralized by insiders, and that appropriation of corporate wealth by the insiders may persist. It suggests a need for regulatory intervention by authorities in the issue of setting the corporate governance of public firms, and also a need for an active market for control as a complement to the internal governance mechanism.

20

<Table 1> Summary Statistics The sample includes 410 manufacturing firms listed on the Korea Stock Exchange during the period of 2000 and 2001. Governance variables are measured as of the end of 2001, while the ownership and financial variables are measured as of the end of 2000.

Average Minimum Maximum Proportion of outside directors 0.3095 0 0.75

Total number of directors 6.32 3 16 Number of committees 0.2831 0 6

Proportion of firms with cumulative voting scheme

0.0525 0 1

Family ownership 0.1849 0 0.7673 Affiliated ownership 0.1136 0 0.7362 Foreign ownership 0.0620 0 0.8294

Institutional ownership 0.0584 0 0.9107 CEO dummy 0.2259 0 1

Chaebol dummy 0.2162 0 1 Cash flow from operation 0.0547 -0.7248 0.5362

Sales growth rates 0.1015 -0.4359 4.6825 Asset size(billion won) 964 28 50,900

Debt to Asset 0.5779 0.0437 3.2192 Bank loan to Debt 0.8108 0 1

21

<Table 2> Correlation Coefficients The sample includes 410 non-financial firms listed on the Korea Stock Exchange between 2000 and 2001. The governance related statistics are as of the end of 2001, while the financial statistics are as of the end of 2000.

Proportion of outside directors

Number of committees

Proportion of firms with cumulative

voting scheme

Family ownership

Affiliated ownership

Foreign ownership

Institutional ownership

CEO dummy

Chaebol dummy Cash flow Sales

growth rates Asset size

Number of committees 0.361***

Proportion of firms with cumulative

voting scheme

0.058 -0.007

Family ownership -0.244*** -0.157*** -0.095**

Affiliated ownership 0.101** -0.015 -0.016 -0.301***

Foreign ownership 0.166*** 0.318*** -0.038 -0.131*** 0.055

Institutional ownership 0.231*** 0.036 0.164*** -0.269*** -0.060 -0.009

CEO dummy -0.122** -0.068 0.016 0.397*** -0.262*** -0.074* -0.149*** Chaebol dummy 0.299*** 0.215*** -0.011 -0.296*** 0.266*** 0.246*** 0.037 -0.239***

Cash flow 0.130*** 0.034 -0.027 0.021 0.080* 0.217*** 0.065 -0.049 0.114*** Sales growth

rates -0.036 0.019 -0.041 -0.011 0.039 0.058 -0.062 -0.017 0.029 0.020

Asset size 0.475*** 0.406*** 0.072 -0.254*** 0.196*** 0.432*** 0.220*** -0.280*** 0.475*** 0.238*** 0.044 Debt to asset 0.095** 0.041 0.072 -0.281*** -0.011 -0.169*** 0.069 -0.103** 0.056 -0.212*** -0.059 0.009

22

<Table 3> Ownership and Board Composition The sample includes 410 non-financial firms listed on the Korea Stock Exchange. The dependent variable is the proportion of outside directors on a board, measured as of the end of 2001, while the explanatory variables are measured as of the end of 2000. The numbers in parentheses are t-values, and ***, **, * denote the significance at the 1%, 5%, and 10% significance levels respectively.

a. Inside ownership and board composition (1) (2) (3) (4) (5)

Constants 0.3746***

(17.70) 0.0801 (0.99)

0.3258***(15.42)

0.0441 (0.55)

0.0769 (0.91)

Family ownership -0.1876***

(-6.21) -0.0539**

(-2.10)

-0.0674**(-2.31)

Affiliated owner ship 0.0514* (1.73)

-0.0032 (-0.14)

-0.0257 (-1.03)

CEO dummy -0.0001 (-0.01)

Chaebol dummy -0.0029 (-0.28)

Cash flow -0.0293 (-0.67)

-0.0284 (-0.64)

-0.0211 (-0.47)

Sales growth rates -0.0129 (-0.98)

-0.0104 (-0.79)

-0.0131 (-0.98)

Asset size 0.0127***

(3.12)

0.0138*** (3.37)

0.0132***(3.09)

Asset size dummy 0.1731***

(10.09)

0.1747*** (10.12)

0.1724***(9.85)

Total number of directors -0.0031* (-1.69)

-0.0030* (-1.67)

-0.0032* (-1.71)

Debt ratio -0.0026 (-0.14)

0.0059 (0.32)

-0.0027 (-0.16)

Bank loan ratio -0.0016 (-0.12)

-0.0065 (-0.48)

-0.0022 (-0.16)

Industry dummy Yes Yes Yes Yes Yes Adjusted R-sq 0.0800 0.4655 0.0039 0.4593 0.4630

F Value 3.07 15.66 0.91 15.30 13.90

23

b. Outside ownership and board composition

(1) (2) (3) (4) (5)

Constants 0.2934***

(31.09) 0.2981***

(31.75) 0.2978***

(30.88) 0.2988***

(31.23) 0.0079 (0.09)

Foreign ownership 0.2019***

(4.75) 0.0272 (0.42)

-0.0457 (-0.85)

Institutional ownership 0.1141***

(2.71) 0.0527 (1.11)

0.0573 (1.55)

Foreign ownership*Chaebol dummy

0.2741***

(3.57)

0.0384 (0.59)

Institutional ownership*Chaebol dummy

0.2102***

(2.77) -0.1181* (-1.88)

Cash flow -0.0328 (-0.72)



(3.72)


(9.59)

Total number of directors -0.0030 (-1.61)

Debt ratio 0.0019 (0.10)


Industrial dummy Yes Yes Yes Yes Yes Adjusted R-sq 0.0615 0.0890 0.0253 0.0417 0.4576

F Value 2.56 3.20 1.62 1.98 14.14

24

c. Ownership and board composition (1) (2) (3) (4)

Constants 0.3361***

(23.80) 0.0883 (1.07)

0.3375*** (24.06)

0.0759 (0.89)


(-4.76) -0.0690**

(-2.40) -0.1528***

(-4.45) -0.0699**

(-2.37)

Affiliated ownership -0.0349 (-1.12)

-0.0262 (-1.06)

-0.0403 (-1.30)

-0.0248 (-0.99)


(4.10) -0.0131 (-0.34)

0.0577 (0.99)

-0.0104 (-0.19)

Institutional ownership 0.0464 (1.14)

0.0109 (0.33)

0.0274 (0.60)

0.0391 (1.06)

Foreign ownership*Chaebol dummy 0.1808**

(2.16) -0.0054 (-0.08)


0.0774 (1.02)

-0.1074* (-1.75)

CEO dummy 0.0005 (0.05)

Cash flow -0.0212 (-0.47)

-0.0162 (-0.36)


-0.0142 (-1.09)


(3.04)

0.0138*** (3.12)


(9.89)

0.1741*** (9.70)

Total number of directors -0.0029 (-1.62)

-0.0031* (-1.70)

Debt ratio -0.0041 (-0.21)

-0.0031 (-0.16)


-0.0053 (-0.38)

Industrial Dummy Yes Yes Yes Yes Adjusted R-sq 0.1304 0.4657 0.1432 0.4662

F Value 4.02 14.51 4.06 13.13

25

<Table 4> Ownership and the Number of Committees The sample includes 410 non-financial firms listed on the Korea Stock Exchange in 2000 and 2001.The dependent variable is the number of committees on a board, and measured as of the end of 2001, while the explanatory variables are measured as of the end of 2000. The numbers in parentheses are t-values, and ***, **, * denote the significance at the 1%, 5%, and 10% significance levels respectively.

a. Inside ownership and the number of committees (1) (2) (3) (4) (5)

Constants 0.4502***

(4.73) -0.5577 (-0.73)

0.3085***(3.69)

-0.7103 (-0.95)

-0.6281 (-0.80)


(-3.06) -0.0052 (-0.02)

-0.3513 (-1.24)

Affiliated ownership -0.2389 (-0.96)

-0.5743*** (-2.58)

-0.6573***(-2.66)

CEO dummy 0.0149 (0.17)


Cash flows 0.0417 (0.10)

0.1835 (0.43)

0.2358 (0.54)

Sales growth rates 0.0366 (0.28)

0.0418 (0.33)

0.0337 (0.26)

Asset size 0.0564 (1.50)

0.0676* (1.82)

0.0688* (1.77)


(5.85)

0.9627*** (5.73)

0.9609***(5.61)

Debt ratio -0.2490 (-1.35)

-0.2141 (-1.20)

-0.2472 (-1.35)

Bank loan ratio -0.2724**

(-2.05)

-0.2929** (-2.24)

-0.2864**(-2.14)

Industrial dummy Yes Yes Yes Yes Yes Adjusted R-sq 0.0400 0.2432 0.0195 0.2560 0.2536

F Value 2.01 6.73 1.48 7.13 6.36

26

b. Outside ownership and the number of committees (1) (2) (3) (4) (5)

Constants 0.1662**

(2.19) 0.4504***

(3.44) 0.2389***

(3.01) 0.2493***

(3.17) 0.0099 (0.01)


(5.88) 1.0032**

(2.26)

0.8656* (1.91)

Institutional ownership 0.5788* (1.66)

0.0032 (0.01)

0.1570 (0.44)


1.5624***

(2.70)

-0.0728 (-0.12)


1.9937***

(3.19) 0.0403 (0.07)

Cash flow -0.1961 (-0.44)


Asset size 0.0246 (0.62)


(5.88)

Debt ratio -0.2242 (-1.23)


(-2.14) Industrial dummy Yes Yes Yes Yes Yes

Adjusted R-sq 0.0965 0.1104 0.0239 0.0462 0.2492 F Value 3.58 4.24 1.59 2.10 6.46

27

c. Ownership and the number of the committees (1) (2) (3) (4)

Constants 0.4038***

(3.40) -0.0719 (-0.09)

0.4131*** (3.51)

-0.0829 (-0.10)

Family ownership -0.7267**

(-2.52) -0.1979 (-0.71)

-0.6326** (-2.21)

-0.2095 (-0.72)

Affiliated ownership -0.6041**

(-2.31) -0.6151**

(-2.51) -0.6663***

(-2.56) -0.6106**

(-2.46)


(5.28) 0.7724**

(2.19) 1.0521**

(2.24) 0.7889* (1.69)


0.1251 (0.39)

-0.1041 (-0.28)

0.0929 (0.26)


1.1574* (1.89)

-0.0318 (-0.05)


1.1832* (1.86)

0.1362 (0.22)

CEO dummy 0.0163 (0.18)

Cash flow -0.0378 (-0.09)

-0.0418 (-0.09)


0.0285 (0.22)

Asset size 0.0348 (0.87)

0.0351 (0.85)


(5.64)

0.9496*** (5.37)

Debt ratio -0.2128 (-1.15)

-0.2121 (-1.14)


(-2.16)

-0.2840** (-2.09)

Industrial dummy Yes Yes Yes Yes Adjusted R-sq 0.1131 0.2619 0.1311 0.2562

F Value 3.61 6.58 3.80 5.86

28

<Table 5> Ownership and Cumulative Voting Scheme The sample includes 410 non-financial firms listed on the Korea Stock Exchange. We use a Logit model, where the dependent variable is a dummy variable which takes the value of 1 if a sample firm has introduced the cumulative voting scheme, or 0 otherwise. The dependent variable is measured as of the end of 2001, while the explanatory variables are measured as of the end of 2000. The inside ownership dummy 1 takes the value of the inside ownership if it is less than 20%, and takes the value of 20% otherwise. The inside ownership dummy 2 takes the value of 0 if the inside ownership is less than 20%, takes the value of the ownership minus 20% if the inside ownership is between 20% and 50%, and 30% otherwise. The inside ownership dummy 3 takes the value of 0 if the inside ownership is less than 50%, and takes the value of the ownership minus 50% otherwise. The numbers in parentheses are p-values, and ***, **, * denote the significance at the 1%, 5%, and 10% significance levels respectively.

(1) (2) (3) (4) (5) (6)

Constants -1.8329***

(0.0001)-1.9471(0.5014)

-1.9061***(0.0001)

-2.2096 (0.4457)

-3.1875*** (0.0001)

-9.0298***(0.0042)

Inside ownership -9.1106***

(0.0010)-12.665***

(0.0001)

(Inside ownership)2 11.2177***

(0.0016)14.2745***

(0.0006)

Inside ownership dummy 1 -6.8388*(0.0515)

-8.4997**(0.0222)

Inside ownership dummy 2 -1.8676(0.4925)

-4.8643 (0.1183)

Inside ownership dummy 3 6.5933*(0.0516)

8.7869**(0.0325)

Foreign ownership 0.7265

(0.6996) -1.3838(0.5854)

Institutional ownership 0.0029

(0.9987) -0.6229(0.7709)

Cash flow 2.7710

(0.2855)

3.1052 (0.2423)

1.4758 (0.6380)

Sales growth rate -0.0059(0.9882)

-0.0116 (0.9769)

-0.5399(0.7339)

Asset size 0.0718

(0.6034)

0.0733 (0.5949)

0.3411**(0.0428)

Debt ratio -1.4273(0.1897)

-1.3966 (0.2019)

-1.2329(0.3497)

Max-rescaled R-sq 0.0569 0.1236 0.0534 0.1211 0.0010 0.0393

29

<Table 6> Ownership and Corporate Governance (3 year panel data analysis) The sample includes 251 non-financial firms listed on the Korea Stock Exchange during the period. The dependent variable is the corporate governance scores of Korean firms over the 3 year periods between 2001 and 2003, and we use random effect model for the control of firm specific effects. The numbers in parentheses are t-values, and ***, **, * denote the significance at the 1%, 5%, and 10% significance levels respectively. a. Inside Ownership and Corporate Governance

(1) (2) (3) (4) (5)

Constants 4.6631***

(57.05) -2.1245***

(-3.72) 4.2147***

(64.43) -2.9127***

(-5.73) -2.0917***

(-3.48)


(-6.19) -0.6274**

(-2.39)

-0.9173***(-3.26)

Affiliated ownership 0.4074 (1.22)

-0.3966 (-1.54)

-0.7281***(-2.66)


Cash flows 0.7585* (1.65)

0.8051* (1.74)

0.8554* (1.86)

Sales growth rate 0.6164***

(2.78)

0.6049*** (2.72)

0.6006***(2.73)


(11.66)

0.3649*** (13.32)

0.3394***(11.16)

Debt ratio -0.1284 (-0.63)

-0.0026 (-0.01)

-0.1718 (-0.84)

R-Square 0.0557 0.2616 0.0023 0.2560 0.2755

30

b. Outside Ownership and Corporate Governance (1) (2) (3) (4) (5)

Constants 4.0918***

(71.47) 4.0905***

(73.52) 4.1535***

(69.09) 4.1586***

(68.75) -2.1802***

(-3.38)


(5.98) 1.3039***

(3.01)

0.2569 (0.64)

Institutional ownership 1.4249***

(3.52) 1.1401**

(2.13) 0.8848* (1.89)


2.5188***

(3.91)

0.7194 (1.16)


0.5691 (0.81)

-0.8637 (-1.37)

Cash flow 0.5042 (1.05)

Sales growth rates 0.5942***

(2.57)


(9.10)

Debt ratio 0.0815 (0.39)

R-Square 0.0523 0.0782 0.0188 0.0197 0.2473

31

<Table 7> Ownership and Changes in Corporate Governance The sample includes 251 non-financial firms listed on the Korea Stock Exchange as of the end of 2002 and 2003. The dependent variable is the difference in the corporate governance scores of Korean firms for 2002 and 2003. The independent variables are measured as of the end of 2002. The inside ownership dummy 1 takes the value of the inside ownership if it is less than 20%, and takes the value of 20% otherwise. The inside ownership dummy 2 takes the value of 0 if the inside ownership is less than 20%, takes the value of the ownership minus 20% if the inside ownership is between 20% and 50%, and 30% otherwise. The inside ownership dummy 3 takes the value of 0 if the inside ownership is less than 50%, and takes the value of the ownership minus 50% otherwise. The numbers in parentheses are t-values, and ***, **, * denote the significance at the 1%, 5%, and 10% significance levels respectively. a. Inside ownership and corporate governance

(1) (2) (3) (4)

Constants 3..0972***

(2.54) 2.6256 (0.44)

3.1380** (2.40)

2.0833 (0.35)

Inside ownership -20.9678***

(-3.36) -19.6902***

(-3.07)

(Inside ownership)2 29.4181***

(3.45) 28.5377***

(3.27)

Inside ownership dummy 1 -18.3853**

(-2.42) -18.2651**

(-2.39)

Inside ownership dummy 2 0.8790 (0.19)

2.5108 (0.53)

Inside ownership dummy 3 14.6660*

(1.80) 13.3602*

(1.69)

Chaebol dummy 0.6813 (0.66)

0.6622 (0.64)

Cash flow -1.5249 (-0.32)

-1.3899 (-0.29)

Sales growth rate 2.1553 (0.65)

2.1810 (0.66)

Asset size -0.0625 (-0.21)

-0.0311 (-0.10)

Debt ratio 2.8524 (1.41)

2.9288 (1.42)

Industry dummy Yes Yes Yes Yes Adjusted R-sq 0.0398 0.0328 0.0321 0.0254

F Value 1.70 1.43 1.52 1.31

32

b. Outside ownership and corporate governance (1) (2) (3) (4) (5)

Constants 0.3672 (0.49)

0.4309 (0.57)

-0.2196 (-0.29)

-0.2419 (-0.31)

0.3713 (0.05)

Foreign ownership -1.0403 (-0.37)

-2.3424 (-0.68)

-1.8413 (-0.48)


5.4465 (1.32)

6.2326 (1.45)


3.4196 (0.67)

3.6312 (0.61)


-0.6774 (-0.14)

-2.5644 (-0.46)

Cash flow -1.1410 (-0.23)

Sales growth rate 1.2715 (0.37)

Asset size -0.1144 (-0.31)

Debt ratio 3.8702* (1.85)

Industry dummy Yes Yes Yes Yes Yes Adjusted R-sq 0.0059 0.0083 0.0041 0.0001 0.0074

F value 0.89 0.86 1.07 1.00 0.91

33

<Figure 1> Distribution of Corporate Governance Scores of Korean Firms (2003)

37

223

106

2313 5

0

50

100

150

200

250

below 90 90~120 120~150 150~180 180~210 over 210

<Figure 2> Normalized Corporate Governance Score and Firm Value (2003)

0

0.5

1

1.5

2

2.5

3

3.5

0 20 40 60 80 100

CG Scores

Tobin's Q

34

<Table 8> Summary of Survey Questionnaire by the Korea Corporate Governance Services (2003)

1. Protection of Shareholder Rights - Introduction of corporate governance charters and managerial code of ethics - Ownership structure - Amount of related party transactions including sales, loans and debt guarantees - Investments in affiliated firms - Adoption of cumulative voting scheme - Observation of one-share-one-vote rule - Restriction on the issuance of equity related securities - Adoption of anti-takeover measures - Shareholder participation in annual shareholders’ meeting 2. Composition and Operation of the Board of Directors - Proportion of outside directors - Independence of outside directors - Activities of outside directors - Availability of corporate information by outside directors - Frequency of board meetings - Board attendance rates - Board minutes - Managerial compensation schemes - Committees of the board 3. Managerial Transparency and Disclosure - Frequency of IR activities - Frequency and contents of disclosures 4. Audit and Internal Control System - Composition of audit committee - Functions of audit committee - Organization of internal control system 5. Distribution of Profit - Dividend policy - Stock repurchase activities

35

References

Park, R. and S. Yoon, “Corporate monitoring roles of inside shareholders, outside shareholders, and Banks”, The Korean Journal of Financial Intermediations Vol. 6, No. 1, 2001, pp.29-61.

Agrawal, Anup and C. Knoeber, "Firm performance and mechanisms to control agency problems between managers and shareholders", Journal of Financial and Quantitative Analysis, Vol. 31 No. 3, 1996. pp 377-397

Bae, K., J. Kang, and Chan W. Lim, "The value of durable bank relationships : Evidence from Korean banking shocks", Journal of Financial Economics, 64, 2002, pp.181-214

Bae, K., J. Kang, and J. Kim, "Tunneling or Value Added: Evidence form Mergers by Korean Business Groups", Journal of Finance, 62, 2002, pp.2695-2740

Baek, J., J. Kang and Kyung S. Park, "Corporate Governance and Firm Value: Evidence from the Korean Financial Crisis", Journal of Financial Economics, Vol. 71, 2004, pp.265-313

Black, B., H. Jang, and W. Kim, "Does corporate governance matter" evidence from the Korean market," working paper, The 3rd Asia Corporate Governance Conference, 2003

Claessens, S., S. Djankov, J. Fan, and H. Lang, "Disentangling the Incentive and Entrenchment Effects of Large Shareholdings." Policy Research Working Paper 2088, The World Bank, 2002.

Durnev, A. and E. Han Kim, "To steal or not to steal: Firm attributes, legal environment, and valuation," working paper, Michigan University, USA, 2003.

Erickson, J., Park, Y.W., Reising, J. and Shin, H.H., "Board of directors and endogenously determined institution and firm value: The Canadian evidence," working paper, The 3rd Asia Corporate Governance Conference, 2002.

Gilson, S., "Bankruptcy, Boards, Banks, and Blockholders: Evidence on Change in ownership and control when firms default," Journal of Financial Economics 27, 1990, pp.355-388

Grossman, S. and O. Hart, "Corporate Financial Structure and Managerial Incentives" in the Economics of Information and Uncertainty, edited by J. McCall, 1982.

Hermalin, B. and Weisbach, M., "The determinants of board composition," The RAND Journal of Economics 19, 1988, pp.589-606

Jang, H.S, H. Kang and Kyung S. Park, “Determinants of ownership structure of family firms: Evidence from family controlled Korean firms”, AICG Working Paper, 2004, Korea University

Jensen, M., "Agency costs of free cash flow, corporate finance and takeovers", The American Economic Review 76, 1986, pp.323-329

Jensen, Michael C. and William H. Meckling, "Theory of the firm: Managerial behavior, agency costs and ownership structure," Journal of Financial Economics 3, 1976, pp.305-360

36

Joh, Sung W., "Corporate Governance and Firm Profitability: Evidence form Korea Before the Economic Crisis," Journal of Financial Economics 68, 2003 pp. 287-322

Johnson, S., La Porta, R., Lopez-de-Silanes, F., Shleifer, A., "Tunneling," The American Economic Review 90, 2000, pp.22-27

Kaplan S. N., and B.A. Minton, "Appointments of outsiders to Japanese Boards: Determinants and Implications for Managers," Journal of Financial Economics 36, 1994, pp.225-258

Klein, A., "Economic Determinants of audit committee independence," The Accounting Review 77, 2002, pp.435-452

La Porta, R., F. Lopez-de-Silanes, and A. Shleifer, "Corporate Ownership around the world," Journal of Finance 54, 1999, pp.471-517

Mak, Y. T. and Y. Li, “Determinants of corporate ownership and board structure: evidence from Singapore,” Journal of Corporate Finance 7, 2001, pp.235-256

McConnell, J.J. and H. Servaes, 1990, "Additional evidence on corporate takeovers and management turnover", Journal of Financial Economics 46, pp. 451-477.

Mitton, T., "A cross-firm analysis of the impact of corporate governance on the East Asian financial crisis," Journal of Financial Economics 64, 2002, pp.215-242

Morck, R., A. Shleifer, and R. Vishny, "Management ownership and market valuation: An empirical analysis," Journal of Financial Economics 20, 1988, pp.293-315

Park, Kyung S. and E. Lee, “Does regulatory monitoring replace market discipline? Corporate governance of financial companies in Korea”, Working Paper, Asian Institute of Corporate Governance, Korea University, 2004.

Shivdasani, A., and D., Yermack, "CEO involvement in the selection of New Board Members: An Empirical Analysis," Journal of Finance 54, 1999, pp.1829-1853

Weisbach, M., "Outside directors and CEO turnover", Journal of Financial Economics 20, 1988, pp.431-460.

Yeh, Y.H., T.S. Lee and T. Woidtke, "Family Control and Corporate Governance: Evidence for Taiwan," International Review of Finance, forthcoming

Yeh, Y.H., and T. Woidtke, "Commitment or Entrenchment?: Controlling shareholders and Board composition," working paper, The 3rd Asia Corporate Governance Conference, 2003.

Korn/Ferry international 29th annual board of directors study 2002 Solomon/Palmiter, Corporation, 2nd ed, Little, Brown and Company, 1994, p183.

why do managerial misconducts persist? the role of controlling … · 2017-03-29 · 2 i....

Documents