Table - Ross School of Business



Does Corporate Governance Matter?

Evidence from Korea

BERNARD S. BLACK*

Stanford Law School

HASUNG JANG**

Korea University Business School

WOOCHAN KIM***

KDI School of Public Policy and Management

First Draft: May 8, 2002

This Draft: September 16, 2002

ABSTRACT

In this paper, we show that corporate governance is an important factor explaining firm value and firm return in the Korean market. Such analysis is made possible by making use of an extensive survey data compiled by the Korea Stock Exchange in the first half of 2001, from which we construct a corporate governance index (0~100) for a cross-section of 540 companies. Evidence shows that a moderate improvement in corporate governance, say an increase of index by 10 units, on average, results in an increase of market capitalization by 16 percent of the company’s book asset value or 35 percent of the company’s book value of common equity. 2SLS is used to control for the endogeneity of corporate governance. Also, evidence shows that an increase in corporate governance index by 10 units, on average, increases the level of buy-and-hold return by 4~6 percent over a year period.

Key words: corporate governance, firm valuation, and stock return

JEL classification: G300

Introduction

Since the aftermath of the financial crisis of 1997/98, Korean government has been aggressive in transplanting legal institutions that were deemed necessary to raise the corporate governance standard in Korea. Outside directors became legally mandatory. Audit and nomination committees were introduced. Chaebol firms were required to disclose consolidated statements. Number of shares required to file derivative suits and inspect financial books have been reduced dramatically. The list goes on. At least during the few years immediately after the crisis, even the business community acknowledged that corporate governance reform is an imperative to overcome the crisis.

With the economic recovery, however, complacency started to creep in. Business managers started to focus on the immediate costs of adapting to the new institutions, but began to ignore their long-term benefits. They even started to lobby the government so that it reverse the hard-won reform measures. Regulations imposed for the sake of better governance are now portrayed as regulations choking freedom and creativity. Skeptics started to question the link between corporate governance and firm performance.

Against this backdrop, in this paper, we document that corporate governance is indeed an important factor explaining the variation of firm value and firm return in Korea. We make use of an extensive survey data set complied by the Korea Stock Exchange (KSE) in the first half of 2001, from which we construct a corporate governance index (0~100) for a cross-section of 540 listed companies. Evidence shows that a moderate improvement in corporate governance, say an increase of the index by 10 units, increases market capitalization by 16 percent of the company’s book asset value or 35 percent of the company’s book value of common equity. 2SLS is used to control for the endogenous nature of corporate governance. Also, evidence shows that an increase in corporate governance index by 10 units, on average, increases the level of buy-and-hold return by 4~6 percent over a year period.

A number of features differentiate this paper from the existing literature. First, this paper studies the linkage between corporate governance and firm performance in an emerging market setting. Most of the existing work investigating the linkage use firm-level U.S. data alone or country-level cross-country data.[1] While many studies found a strong linkage between corporate governance and performance in an international setting, not many papers found a robust relationship between the two using firm-level data within a country.[2] Empirical findings of this paper suggest that corporate governance do explain the variation of firm value or firm return within a country, and that its explanatory power is greater in emerging markets than in mature ones.

Second, this paper tries to construct an overall corporate governance index in investigating the linkage between corporate governance and firm performance.[3] Many existing papers concentrate on a particular aspect of governance such as board, shareholders’ activism, compensation, anti-takeover provisions, investor protection, and so on.[4] On the other hand, the index adopted in this paper includes 42 variables covering four major areas of governance: investor protection, board of directors in general, outside directors, and transparency.

Third, the corporate governance index used in this paper is relatively less subjective than those compiled by the financial industry.[5] Out of the original 123 variables, questions asking for management view or plan are all dropped when constructing the index. The index is constructed using the remaining 42 variables asking for verifiable facts. This is in contrast to those indices that make use of subjective views expressed by fund managers or equity analysts.

Fourth, this paper attempts to control for the endogeneity of corporate governance when explaining firm value. In doing so, we take advantage of the legal system that shapes the firm-level governance structure in Korea. That is, we make use of the fact that law mandates firms with book asset size over KRW 2 trillion have higher governance standards. Thus, asset size dummy, taking a value of 1 if book asset value is greater than KRW 2 trillion, and 0 otherwise, became our natural choice of an instrumental variable.

This paper is organized as follows. We first describe the data in Section II and explain how we constructed our corporate governance index in Section III. Then, we report our results in Section IV and Section V. In Section VI, we conclude.

Data

1 Data on Corporate Governance

This paper makes use of a survey data provided by the Korea Stock Exchange (KSE). In March 2001, KSE sent out an extensive survey of corporate governance practices and attitudes to the disclosure officers of all listed companies.[6] The survey was completed between March and July 2001. Given the nature of the institution that conducted the survey, the response rate was very high.[7]

Among the 699 listed companies in the KSE during the survey period, 139 companies on the watch list for possible de-listing were not surveyed. Also, 20 companies that were surveyed, but did not respond to any of the survey questions, were also dropped from our sample. This left us with 540 companies that can be used for analyses.

The survey questions were grouped into five categories: (i) shareholders’ right, (ii) board of directors in general, (iii) outside directors, (iv) auditing, and (v) other stakeholders. Some questions were objective, asking about the company’s practices. Others were more subjective, asking about the company’s attitudes or plans.

The survey contains 113 questions, some with several parts. From the survey questions, we extracted 123 variables. In constructing the corporate governance index, however, we made use of only 42 variables. See Section III for the details.

2 Financial and Stock Market Data

Balance sheet and income statement data used in this paper is from a database called TS2000, which is provided by the Korea Listed Companies Association (KLCA). Stock market data, on the other hand, is obtained from a database complied by the Korea Stock Exchange (KSE). The lists of top-30 Chaebol companies are from various press releases from the Korean Fair Trade Commission (KFTC).

Construction of Corporate Governance Index

Construction of the Index

We extract 123 variables from the survey questions. Of these, we exclude 81 for various reasons. First, we exclude all those variables asking for management view. Among the variables that ask for facts, we exclude those that are not relevant to corporate governance, ambiguous as to which answer indicates better governance, had minimal variation between firms, or had very few responses. Such exclusion rule left us with 42 variables.

The 42 variables are categorized into four groups: (i) shareholder rights, (ii) board of directors in general, (iii) outside directors, and (iv) disclosure and transparency.[8] From each group, sub-indices are constructed, each having an equal weight and standardized to have a value between 0 and 25. Thus, the overall corporate governance index is constructed to have a value between 0 and 100. The index is constructed so that better governed firms have higher index.

Each variable in the sub-index is designed to have a value between 0 and 1. To obtain a sub-index, first, we compute a simple sum over the variables, and second, divide by the number of ‘non-missing’ variables.[9] We then multiply this ratio by 25, so that the resulting figure takes a value between 0 and 25. The overall indices and sub-indices are constructed before running any of the statistical tests described in the following sections. Table 1 describes the variables used in constructing the indices.

Description of the Index

Figure 1 shows the histogram of the overall corporate governance index. A normal distribution curve is superimposed. By comparing the histogram and the normal distribution curve, one can easily see that the distribution of corporate governance index is slightly skewed to the right (long tails to the right).[10] That is, many companies concentrated below the mean and few companies located at very high scores. Table 2 shows that the mean is 29.3, the minimum is 0, and the maximum is 77.5.[11]

Corporate Governance and Firm Value

We first test if companies with higher corporate governance index have greater firm value. Here, we use two measures of firm value: Tobin’s Q and market-to-book ratio. In our analyses, Tobin’s Q is computed by book value of debt plus market value of equity over book value of asset.[12] That is, [book value of debt + market value of equity] / [book value of asset]. Market-to-book ratio is computed by market value of common equity over book value of common equity.

In principle, all book values are measured at the end of year 2000. If fiscal year does not end in December, they are measured during a period between July 2000 and June 2001. If multiple figures exist during this period, the most recent figure is chosen. Market value of equity, on the other hand, is measured as of June 29, 2001, which is closer to the survey period (March ~ July, 2001). Table 2 describes the details of each variable.

Difference-in-Mean Test

540 companies in our sample are categorized into five groups according to their corporate governance scores and, for each quintile, averages of Tobin’s Q and market-to-book ratio are computed.[13] Before computing the averages, outliers are identified by computing studentized residuals and dropped from the sample. That is, as for Tobin’s Q, a regression of Tobin’s Q on corporate governance index is estimated and studentized residuals are obtained from the estimated regression equation. Observations are considered as outliers if such studentized residuals are greater than 1.96 or smaller than -1.96. This resulted in 24 outliers.[14] Using the same method, we found 16 outliers as for market-to-book ratio.[15]

Table 4 shows the results. The third column shows that Tobin’s Qs are not increasing monotonically with corporate governance scores. However, when the first (lowest corporate governance score) and the fifth (highest corporate governance score) quintiles are compared, the difference is 0.1017 and statistically significant at 1% level. That is, when corporate governance improves from the level of bottom quintile companies to the level of top quintile companies, Tobin’s Q increases by 0.1017%. More intuitively put, market value of equity will increase by 10.17% of the company’s book value of asset.

As for market-to-book ratio, the difference between the first and the fifth quintile is not statistically significant.[16] This result, however, is overturned in the follow sub-section, where we run regressions. Figure 2 and 3 show scatter plots between corporate governance index and measures of firm value. The fitted lines are clearly upward sloping.

Regression Analyses

In Table 5, Tobin’s Q and market-to-book ratio are regressed on corporate governance index. Equations (1), (2), and (3) regress Tobin’s Q, while equations (4), (5), and (6) regress market-to-book ratio. Following Shin and Stulz (2000) and Gompers, Ishii, and Metrick (2001), we use the log of book value of assets, the log of years of listing, and industry dummy variables, as the basic controlling variables. We use the 2-digit KSIC to capture industry effects.[17] Debt-to-asset ratio is additionally included to capture any effect from capital structure. Sales growth is included in the equation to capture growth opportunities. Detailed description of each variable can be found in Table 2.

The result shows that the coefficients on corporate governance are statistically significant and economically meaningful. In the Tobin’s Q equations, the coefficients are around 0.002 and statistically significant at 1% level. The coefficient of 0.002 implies that an increase in corporate governance index by 10 units (e.g. improvement of corporate governance index by 10 units) results in an increase of market capitalization by 2% of the company’s book asset value. Notice that corporate governance index ranges from 0 to 100. All the control variables turned out to be significant at 1% level.

When market-to-book ratio is regressed, the coefficient is 0.0065 and statistically significant at 1% level as can be seen in equation (6), where we included all the control variables. The coefficient of 0.0065 implies that an increase in corporate governance index by 10 units (e.g. improvement of corporate governance index by 10 units) results in an increase of market capitalization by 6.5% of the company’s book value of common equity. All the control variables with the exception of debt-to-asset ratio turned out to be significant.

Regression on Different Sub-samples

The governance structure of Korean firms is in large part determined by law and because of the way the laws are written, the governance structure is heavily influenced by the following three factors: book asset value, affiliation to banking industry, and affiliation to chaebol. For example, the Securities & Exchange Act sets the minimum ratio and number of outside directors, requires the establishment of audit and nomination committees, and sets the minimum ratio of outside directors in the audit committee.

Such minimum ratios and requirements, however, vary according to the company’s book asset value. That is, listed companies with book value of asset greater than KRW 2 trillion must have at least three outside directors and the ratio of outside directors must be at least one half. Those below the KRW 2 trillion threshold need only to have one quarter of their directors from outside. Also, listed companies with book value of asset greater than KRW 2 trillion must establish audit and nomination committees, while those below the KRW 2 trillion threshold do not have to. When the audit committee is required, two third of the members must be filled with outside directors and the chairperson must come from outside. All these suggest that the book value of asset is a very important determinant of corporate governance in Korea.[18]

In Korea, the Banking Act shapes the governance structure of banks differently from others. That is, most of the requirements in the Securities & Exchange Act that is applied to companies with book asset value greater than KRW 2 trillion are applied to commercial banks and merchant banks, regardless of their size. Thus, one can easily see that affiliation to banking industry is another important factor determining corporate governance in Korea.[19]

In Korea, the Monopoly Regulation and Fair Trade Act shapes the governance structure of chaebol-affiliated firms differently from others. That is, for companies affiliated to the 30 largest chaebols, the Act requires board approval for related-party transactions if transaction size is greater than 10% of equity capital or KRW 10 billion. Such requirement was not in effect for non-chaebol firms at the time the survey was conducted.[20] Thus, affiliation to chaebol can be considered as another factor influencing the level of corporate governance in Korea.[21]

If such factors are not controlled for, one may have a biased result. That is, if corporate governance index and firm value measure are both positively influenced by asset size, one may see a positive relationship between corporate governance index and firm value even when there is no direct link between the two.

As a way of controlling for such factors, in this sub-section, we run the same set of regression equations used in Table 5 on the following six sub-samples: banks (commercial and merchant), non-banks, firms affiliated to chaebol, firms not affiliated to chaebols, firms with book asset value greater than KRW 2 trillion, and firms with book asset value below KRW 2 trillion.[22]

Panel A and B of Table 6 show the results. Panel A reports the results for Tobin’s Q and Panel B reports the results for market-to-book ratio. Equation (1) and (5) in each of the Panels show that corporate governance does not have much power explaining the variations within banks or within firms with book asset value greater than KRW 2 trillion. Other equations, however, show that corporate governance is still an important factor explaining the variation of firm value.

In fact, equations (3) and (4) in both of the Panels show that the explanatory power of corporate governance is particularly high when applied to the chaebol sub-sample. When explaining Tobin’s Q, the coefficient on corporate governance index is 0.0032 when applied to the chaebol sub-sample, while it is only 0.002 when applied to the non-chaebol sub-sample. When explaining market-to-book ratio, the coefficient on corporate governance index is 0.0138 when applied to the chaebol sub-sample, while it is only 0.0044 when applied to the non-chaebol sub-sample.

Endogeneity

Analysis up to this point assumes that the level corporate governance is exogenously given. Corporate governance, however, can be endogenously determined. As demonstrated by Durnev and Kim (2002), firms with profitable investment opportunities and more reliance on external financing may have higher-quality governance. In such cases, the corporate governance index variable and the error term will be correlated and thus lead to a biased coefficient.

To test for the endogeneity of corporate governance, we follow the method suggested by Wooldridge (2000).[23] The test results show that whenever firm value is the dependent variable, corporate governance is negatively correlated with the error term. This suggests that the coefficients estimated by OLS are downward biased. The test results, however, show that corporate governance is not correlated with the error term when firm return is the dependent variable.

To control for the endogeneity, 2SLS is use to estimate the coefficients. Asset size dummy is chosen to be the exogenous variable that is highly correlated with corporate governance, but that does not appear in the firm value equation.[24] There are three reasons underlying this choice of an IV. First, as explained in the previous sub-section, the Securities & Exchange Act made the asset size dummy an important determinant of corporate governance in Korea. Firms with book asset size greater than KRW 2 trillion are required to have an audit committee, an outside director nomination committee, and a greater portion of outside directors. Second, since the asset size dummy represents the legally mandated portion of corporate governance, it can be safely said to be exogenous. That is, firms have no choice but to meet the legal requirement.[25] Third, when asset size is already controlled for, it is hard to imagine that asset size dummy will have any additional explanatory power over firm value other than through strengthened corporate governance. Bank and chaebol dummies are not used as instrumental variables since they might be correlated with firm value without changing the quality of governance.

Table 7 shows that the coefficient on corporate governance increases significantly when estimated by 2SLS.[26] When Tobin’s Q is the dependent variable, the coefficient on corporate governance increases almost eightfold. When market-to-book ratio is the dependent variable, the coefficient increases almost fivefold. More specifically, the magnitude of the coefficients imply that a 10-unit increase in the corporate governance index, on average, leads to an increase of market capitalization by 16 percent of the company’s book asset value or 35 percent of the company’s book value of common equity.

Corporate Governance and Return

Difference-in-Test

If the quality of corporate governance does not get fully reflected into price, but gets incorporated only gradually over time, one may find a positive relationship between the level of corporate governance and return. To verify this in the Korean market, we again categorized firms into five groups according to their corporate governance index scores and for each quintile, computed equal-weighted average returns. To this end, buy-and-hold returns are computed for each company over a year period from December 26, 2000 to December 28, 2001. Stock-splits and stock-mergers are adjusted.

Table 7 shows the details. Column (1) shows gross returns and columns (2), (3), (4), and (5) show different measures of excess returns. In column (2), excess return is defined as gross return minus market return (e.g. KOSPI return). In column (3), excess return is defined as gross return minus required rate of return from the market model.[27] In column (4), excess return is defined as gross return minus required rate of return from CAPM.[28] Lastly, in column (5), excess return is defined as gross return minus average return of firms in the same size category.

Before computing the averages, however, 34 observations are dropped from the sample. Among the 34 observations, 23 are dropped as outliers. Observations are considered as outliers if a studentized residual obtained from a regression of gross return on corporate governance index is greater than 1.96 or smaller than –1.96. Such method identified the 23 highest figures as outliers. In addition, 11 firms that did not exist over the entire sample period - year 2001 - are also dropped from the sample.

Table 7 shows the results. The overall picture is that average returns do not vary much within the bottom four quintiles, but increase significantly when it comes to the top quintile (e.g. firms with high corporate governance index scores). When the top quintile and the rest are compared, the difference ranges from 13.8% to 28.8%. The difference is also statistically significant at the 1% level regardless of the way excess returns are measured. Figure 4 shows the scatter plot between corporate governance index and firm return. The fitted line is clearly upward sloping.

Regression Analyses

Excess returns reported in Table 7 are regressed on the corporate governance index and on other control variables. Following Brennan, Chordia, and Subrahmanyam (1998) and Gompers, Ishii, and Metrick (2001), we control for the log of size, market-to-book ratio, past return, the log of per share price, the log of trading volume, dividend yield, past sales growth, and industry dummies. We use the 2-digit KSIC to capture industry effects. Detailed description of each variable can be found in Table 2.

In Table 8, we report two sets of analyses. In Panel A, regression equations are estimated without market-to-book ratio in the right-hand side of the equation. In Panel B, we include market-to-book ratio. The results show that the coefficient on corporate governance index is materially large and statistically significant at either a 5 or a 10% level. In Panel A (without market-to-book ratio), the coefficients are between 0.5166 and 0.5799 and statistically significant at 5% level regardless of the way excess returns are measured. Intuitively put, it means that whenever corporate governance index score increases by 10 unites (e.g. corporate governance improves by 10 unites), excess return increases by 5.166~5.799% over a year period.

Results in Panel B (without market-to-book ratio) basically tells the same story, except that the coefficient magnitudes are slightly smaller. The coefficients are between 0.4016 and 0.5420 and statistically significant either at a 5 or a 10% level regardless of the way excess return is measured. More intuitively put, it means that whenever corporate governance index score increases by 10 unites (e.g. corporate governance improves by 10 unites), excess return increases by 4.016~5.420% over a year period.

Regression on Different Sub-samples

As stated in Section IV, asset size, affiliation to banking industry, and affiliation to chaebol, are important factors shaping the governance structure of Korean firms. As such, in this section, we control for such factors by running regressions on the following six sub-samples: banks (commercial and merchant), non-banks, firms affiliated to chaebol, firms not affiliated to chaebols, firms with book asset value greater than KRW 2 trillion, and firms with book asset value below KRW 2 trillion.[29]

Table 9 shows the results. Notice that we do not use all the control variables that appear in Table 8. The log of size, market-to-book ratio, the log of trading volume, dividend yield, and past sales growth are dropped from the equation. In Table 8, the coefficients on such variables are individually insignificant. Wald-test also shows that the variables are jointly insignificant, which justifies their omission. In Table 9, we only use one measure of excess return, which is defined by gross return minus KOSPI return.[30]

Equations (2), (4), and (6) in Table 9 show that corporate governance is still an important factor in explaining the variation of excess returns when the samples are restricted to non-banks, non-chaebols, or firms with book asset value below KRW 2 trillion. In case of the bank sub-sample, the coefficient increases dramatically. But, given the small sample size (N = 13), the coefficient does not deserve any series interpretation. In case of the chaebol sub-sample, the coefficient is nearly zero and statistically insignificant. This result is in contrast to the one reported in Table 6, which basically says that the explanatory power of corporate governance is greater among chaebol-affiliated firms than among non-chaebol-affiliated firms. The two results, however, can be compatible. That is, if corporate governance gets incorporated into price quickly for chaebol-affiliated firms, the level of corporate governance would not have any influence on ex-post buy-and-hold returns.

Conclusion

In this paper, we document that corporate governance is an important factor explaining firm value and firm return in the Korean market. Such analysis is made possible by making use of a survey data collected by the Korea Stock Exchange in the first half of 2001, from which we construct a corporate governance index (0~100) for a cross-section of 540 companies.

Evidence shows that firms with better governance have higher Tobin’s Q or higher market-to-book ratio. More specifically, an increase in corporate governance index by 10 units results into an increase of Tobin’s Q by 0.016 or market-to-book ratio by 0.035.[31] This implies that an improvement of corporate governance by 10 units increases market capitalization by 16 percent of the company’s book asset value or 35 percent of the company’s book value of common equity. Such figures are obtained after controlling for the endogeneity of corporate governance by follow the 2SLS procedure. Empirical analysis also shows that the explanatory power of corporate governance is particularly high when explaining the variation of firm value of chaebol-affiliated firms.

This paper also documents that firms with better governance experience higher excess return. More specifically, an increase in corporate governance index by 10 units increases the level of excess return by 4-6 percent over a year period. This is so regardless of the way excess returns are measured. Test for endogeneity shows that corporate governance is not correlated with the error term when excess return is the dependent variable.

This study has three important implications. First, our empirical finding helps to clear the skepticism on the linkage between corporate governance and firm performance in the Korean market. Thus, bolstering the rationale behind the reform measures that took place since the financial crisis in 1997/98. Second, our empirical finding justifies the commercial existence of corporate governance funds that reap capital gain by influencing the firms they invest to adopt better governance structure. Third, given the evidence that strong linkage cannot be found easily in mature markets, our empirical findings suggest that corporate governance is a much more important determinant of firm value or firm return in emerging markets.

REFERENCES

Bhagat, Sanjai and Bernard Black (1999), “Uncertain Relationship Between Board Composition and

Firm Performance, “ Business Lawyer Vol. 54 No.3

Bhagat, Sanjai, Dennis Carey, and Charles Elson (1999), “Director Ownership, Corporate

Performance, and Management Turnover,” Business Lawyer Vol.54 No.3

Black, Bernard (2001), “Does Corporate Governance Matter? A Crude Test Using Russian Data,”

University of Pennsylvania Law Review Vol.149, pp.2131-2150

Brennan, Michael, Tarun Chrodia, and Avanidhar Subrahmanyam (1998), “Alternative Factor

Specifications, Security Characteristics, and the Cross-Section of Expected Stock Returns,”

Journal of Financial Economics Vol.49, pp 345-375

Carleton, Willard, James Nelson, and Michaels Weisbach (1998), “The Influence of Institutions on

Corporate Governance through Private Negotiations: Evidence from TIAA-CREF,” The Journal

of Finance Vol.53 No.4 pp 1335-1362

Credit Lyonnais Securities Asia (2001), CG Watch: Corporate Governance in Emerging Markets, April

Durnev, Art and E. Han Kim (2002), To Steal or Not to Steal: Firm Attributes, Legal Environment,

and Valuation, paper presented at the 2nd Asian Corporate Governance Conference in Seoul,

Korea.

Gompers, Paul, Joy Ishii, and Andrew Metrick (2001), “Corporate Governance and Equity Prices,”

NBER Working Paper 8449

Karpoff, Jonathan, Paul Malatesta, and Ralph Walkling (1996), “Corporate Governance and

Shareholder Initiatives: Empirical Evidence,” Journal of Financial Economics, Vol. 42 pp 365-395

La Porta, Rafael, Florencio Lopez de-Silanes, Andrei Shleifer, and Robert Vishny (1997), “Legal

Determinants of external finance, Journal of Finance, Vol. 52, pp 1131-1150

_______________________________________________________________________(2002), “Investor

Protection and Corporate Valuation,” Journal of Finance, Vol. 57 pp 1147-1170

Millstein, Ira and Paul MacAvoy (1998), “Active Board of Directors and Performance of the Large

Publicly Traded Corporation,” Columbia Law Review, 98

Shin, Hyun-Han and Rene Stulz (2000), “Firm Value, Risk, and Growth Opportunities,” NBER

Working Paper 7808

Sundaramurthy, Chamu, James Mahoney, and Joseph Mahoney (1997), “Board Structure, Anti-

takeover Provisions, and Stockholder Wealth,” Strategic Management Journal Vol.18:3

Wooldridge, Jeffrey (2000), Introductory Econometrics: A Modern Approadh (South-Western College

Publishing)

Figure 1. Distribution of Corporate Governance Index

[pic]

Figure 2. Corporate Governance and Tobin’s Q

[pic]

Figure 3. Corporate Governance and Market-to-Book Ratio

[pic]

Figure 4. Corporate Governance and Return

[pic]

Table 1. Corporate Governance Index

|A. Shareholder Rights Sub-index | | | |

|Variable |Summary of the Variable |Sample Size |Number of “1” |Mean |

| | | |Responses | |

|A.1 |Adopts cumulative voting for electing directors |540 |34 |0.06 |

|A.2 |Permits voting by mail |540 |71 |0.13 |

|A.3 |Shareholder meeting date or location set to encourage attendance |508 |89 |0.18 |

|A.4 |Director candidates disclosed before shareholder meeting |540 |96 |0.18 |

|A.5 |Board approval required for related party transactions |540 |181 |0.34 |

|B. Board of Directors Sub-index | | | |

|Variable |Summary of the Variable |Sample Size |Number of “1” |Mean |

| | | |Responses | |

|B.1 |Directors’ average attendance rate is greater than 75% |482 |330 |0.68 |

|B.2 |Disapproved item exists at board meeting |498 |30 |0.06 |

|B.3 |Directors’ positions on board meeting agenda recorded in minutes |540 |223 |0.41 |

|B.4 |CEO and board chairman are different people |540 |25 |0.05 |

|B.5 |System for evaluating directors exists |540 |36 |0.07 |

|B.6 |By-law to govern board meetings exists |540 |381 |0.71 |

|B.7 |Number of regular board meetings no less than four times a year |356 |260 |0.73 |

|C. Outside Directors Sub-index | | | |

|Variable |Summary of the Variable |Sample Size |Number of “1” |Mean |

| | | |Responses | |

|C.1 |Ratio of outside directors |527 |N/A |0.33 |

|C.2 |Firm has foreign outside directors |540 |37 |0.07 |

|C.3 |Outside director nominating committee exists |540 |75 |0.14 |

|C.4 |Outside directors do not receive retirement pay |320 |281 |0.88 |

|C.5 |Outside directors can get advice from outside experts at the |320 |77 |0.24 |

| |company’s expense | | | |

|C.6 |System of evaluating outside directors exists or planning to have |509 |155 |0.30 |

| |one | | | |

|C.7 |Outside directors’ pay in aggregate determined separately at |482 |48 |0.10 |

| |shareholders’ meeting | | | |

|C.8 |Outside directors’ average attendance rate is greater than 75% |465 |197 |0.42 |

|C.9 |Bard meeting item opposed by outside director exists |540 |41 |0.08 |

|C.10 |Outside directors made suggestions on management issues |540 |54 |0.10 |

|C.11 |Provide information beyond board meeting items |495 |383 |0.77 |

|C.12 |One-week prior notice of agenda to outside directors |471 |235 |0.50 |

|C.13 |Do not hesitate to provide confidential information to outside |501 |251 |0.50 |

| |directors | | | |

|C.14 |Code of conduct for outside directors exits |540 |43 |0.08 |

|C.15 |Designate contact person to support outside directors |540 |274 |0.51 |

|C.16 |A meeting composed exclusively of outside directors exists |540 |24 |0.04 |

|D. Disclosure & Transparency Sub-index | | | |

|Variable |Summary of the Variable |Sample Size |Number of “1” |Mean |

| | | |Responses | |

|D.1 |Conducted IR activity in year 2000 |540 |21 |0.04 |

|D.2 |Resume of board members found in company website |540 |47 |0.09 |

|D.3 |Audit committee exists |518 |95 |0.18 |

|D.4 |Ratio of outside directors in the audit committee |103 |N/A |0.77 |

|D.5 |By-laws governing audit committee (or auditor) exits |489 |321 |0.66 |

|D.6 |Accounting experts in the audit committee |324 |163 |0.50 |

|D.7 |Audit committee (or auditor) recommends external auditor at AGM |495 |369 |0.75 |

|D.8 |Each audit committee member’s (or auditor’s) pay determined |474 |72 |0.15 |

| |separately at the AGM | | | |

|D.9 |Audit committee (or auditor) approves the appointment of internal |404 |192 |0.48 |

| |audit head | | | |

|D.10 |Minutes written for each audit committee (auditor) meeting |267 |164 |0.61 |

|D.11 |Report audit committee’s (or auditor’s) activities at the AGM |473 |422 |0.90 |

|D.12 |Rejected item exits at the audit committee meeting |101 |4 |0.04 |

|D.13 |Audit committee members’ average attendance rate is greater than |99 |93 |0.94 |

| |75% | | | |

|D.14 |English disclosure exists |498 |24 |0.05 |

Table 2. Other Variables

|Variables |Descriptions |

|Tobin’s Q |Book value of debt plus market value of equity over book value of asset. That is, [book value of |

| |debt + market value of equity]/book value of asset. Here market value of equity includes common |

| |stocks and preferred stocks. |

|Market-to-Book Ratio |Market value of common equity over book value common equity |

|Book Value of Debt |Total liability as of December 26, 2000 in billion KRW. If fiscal year does not end at the end of|

| |December, the liability available during a period between July 2000 and June 2001. If there are |

| |multiple figures during the period, the most recent figure is chosen [Source: Korea Listed |

| |Companies Association]. |

|Book Value of Asset |Asset value as of December 26, 2000 in billion KRW. If fiscal year does not end at the end of |

| |December, the asset value available during a period between July 2000 and June 2001. If there are|

| |multiple figures during the period, the most recent figure is chosen [Source: Korea Listed |

| |Companies Association]. |

|Book Value of Equity |Book value of asset subtracted by book value of debt and book value of preferred stocks. That is |

| |[book value of asset – book value of debt – book value of preferred stocks]. Book value of |

| |preferred stocks are in billion KRW and obtained as of December 26, 2000. |

|Market Value of Equity |Market capitalization of common equity as of June 29, 2001 in billion KRW. If a company is |

| |de-listed before June 29, 2001, the most recent figure is chosen [Source: Korea Stock Exchange]. |

|Years of Listing |Number of years since listing [Source: Korea Listed Companies Association]. |

|Sales Growth |Average growth rate of sales during the past five years in ratios. That is, between fiscal year |

| |1996 and fiscal year 2000. If sales figures are available over less than five years, the average |

| |growth rate during that sub-period [Source: Korea Listed Companies Association]. |

|Asset Size Dummy |1 if book value asset is greater than KRW 2 trillion; 0 otherwise |

|Bank Dummy |1 if the firm is either a commercial bank or a merchant bank; 0 otherwise |

|Chaebol Dummy |1 if affiliated to top-30 Chaebols during a period between April 2000 and April 2001; 0 otherwise |

| |[Source: Fair Trade Commission, press releases]. |

|Size |Market capitalization of common equity as of December 26, 2000 in Million KRW [Source: Korea Stock|

| |Exchange]. |

|Past Return |Percentage return on common shares between December 26, 2000 and December 28, 2001 adjusted for |

| |stock-splits and stock-mergers [Source: Korea Stock Exchange]. |

|Price |Closing common share price as of January 2, 2001 [Source: Korea Stock Exchange]. |

|Trading Volume |Trading volume in KRW as of January 2, 2001 [Source: Korea Stock Exchange]. |

|Dividend Yield |Dividend yield in percentage as of January 2, 2001 [Source: Korea Stock Exchange]. |

Table 3. Descriptive Statistics

A. Corporate Governance Index

| |No. of Obs |Mean |Stand. Dev. |Min |Max |

|Shareholders’ Right |540 |4.377 |4.614 |0.000 |20.000 |

|Board of Directors |540 |8.555 |4.819 |0.000 |21.429 |

|Outside Directors |540 |7.462 |3.658 |0.000 |20.000 |

|Disclosure & Transparency |540 |8.918 |4.629 |0.000 |20.313 |

|Overall Index |540 |29.312 |13.007 |0.000 |77.500 |

B. Other Variables

| |No. of Obs |Mean |Stand. Dev. |Min |Max |

|Tobin’s Q |539 |0.854 |0.285 |0.317 |3.040 |

|Market-to-Book Ratio |538 |0.788 |1.636 |-7.554 |21.611 |

|Debt-to-Asset Ratio |539 |0.556 |0.213 |0.093 |1.238 |

|Book Value of Asset |539 |1923 |7617 |10 |81522 |

|Book Value of Equity |539 |439 |1791 |-121 |31835 |

|Market Value of Equity |539 |357 |1861 |2 |29038 |

|Years of Listing |540 |15.8 |9.5 |0.0 |45.0 |

|Sales Growth |517 |0.144 |0.311 |-0.295 |5.852 |

|Asset Size Dummy |540 |0.133 |0.340 |0 |1 |

|Bank Dummy |540 |0.206 |0.404 |0 |1 |

|Chaebol Dummy |540 |0.031 |0.175 |0 |1 |

|Size |535 |304380 |1861266 |1560 |23900000 |

|Past Return |535 |-32.626 |45.557 |-94.795 |467.963 |

|Price |540 |13242 |26521 |445 |329500 |

|Trading Volume |535 |2.13E+09 |8.61E+09 |0 |1.30E+11 |

|Dividend Yield |540 |3.633 |5.753 |0.000 |59.520 |

B. Correlation Matrix of Selected Variables

*, **, and *** respectively indicates significant level at 10%, 5%, and 1%

| |CG Index |Tobin’s Q |M/B |D/A |Asset |Years of listing|Sales Growth (%)|Asset Size |Bank |Chaebol |

| | |(in ratio) |(in ratio) |(in ratio) |(in logs) | | |Dummy |Dummy |Dummy |

| | | | | | |(in logs) | | | | |

|CG Index |1.0000 | | | | | | | | | |

|Tobin’s Q |0.1489*** |1.0000 | | | | | | | | |

|(in ratio) | | | | | | | | | | |

|M/B |0.0240 |0.5208*** |1.0000 | | | | | | | |

|(in ratio) | | | | | | | | | | |

|D/A |0.3007*** |0.3009*** |0.1593*** |1.0000 | | | | | | |

|(in ratio) | | | | | | | | | | |

|Asset |0.5799*** |0.0142 |-0.0702 |0.4110*** |1.0000 | | | | | |

|(in logs) | | | | | | | | | | |

|Years of listing |0.0811* |-0.1317*** |-0.0514 |0.1521*** |0.2086*** |1.0000 | | | | |

|(in logs) | | | | | | | | | | |

|Sales Growth (%) |0.1064** |0.1020** |0.0705 |-0.0587 |0.0774* |-0.054 |1.0000 | | | |

|Asset Size |0.5501*** |0.1603*** |-0.0162 |0.3325*** |0.7439*** |0.1001** |0.0708 |1.0000 | | |

|Dummy | | | | | | | | | | |

|Bank |0.3881*** |0.0886** |0.0089 |0.3402*** |0.4399*** |0.1031** |0.0487 |0.3973*** |1.0000 | |

|Dummy | | | | | | | | | | |

|Chaebol |0.2354*** |0.0380 |-0.0392 |0.1399*** |0.4315*** |0.1376*** |0.1209*** |0.3667*** |-0.0392 |1.0000 |

|Dummy | | | | | | | | | | |

Table 4. Corporate Governance and Firm Value

(Difference-in-Mean Test)

Firms are categorized into five groups according to their corporate governance scores and for each quintile (equally-weighted) averages of Tobin’s Q and market-to-book ratio are computed. Tobin’s Q is computed by [(book value of debt + market value of equity) / book value of asset]. Market-to-book ratio is computed by [market value of common equity/(book value of asset – book value of debt – book value of preferred stocks)]. When computing average Tobin’s Q for each quintile, 24 outliers are identified and dropped from the sample. Observations are considered as outliers if a studentized residual obtained from a regression of Tobin’s Q on corporate governance index is greater than 1.96 or smaller than –1.96. Such method identified the 24 largest figures as outliers. When computing average market-to-book ratio for each quintile, 16 outliers are identified and dropped from the sample. Observations are identified as outliers if a studentized residual obtained from a regression of market-to-book ratio on corporate governance index is greater than 1.96 or smaller than –1.96. Such method identified the 6 smallest figures and 10 largest figures as outliers. The 6 smallest figures include all the negative market-to-book ratios. Numbers of observations are reported in the brackets. The last row reports the t-test results. Standard errors assuming unequal variances are reported in the parentheses. ***, **, and *, respectively indicate significance levels at 1%, 5%, and 10%.

|Groups according to Corporate |Corporate Governance Index |Tobin’s Q |Market-to-Book Ratio |

|Governance Scores | | | |

|(1) Low |0 ~ 18.80 |0.8054 [102] |0.6708 [106] |

|(2) |18.83 ~ 25.16 |0.7616 [104] |0.5008 [103] |

|(3) |25.27 ~ 31.04 |0.7810 [101] |0.6442 [104] |

|(4) |31.19 ~ 39.00 |0.8090 [105] |0.6105 [103] |

|(5) High |39.02 ~ 75.28 |0.9071 [104] |0.7708 [106] |

|(5) – (1) | |0.1017*** |0.0999 |

| | |(0.0265) |(0.0780) |

Table 5. Corporate Governance and Firm Value

(Regression Analyses)

Tobin’s Q and market-to-book ratio are regressed on corporate governance index. Equations (1), (2), and (3) regress Tobin’s Q, while equations (4), (5), and (6) regress market-to-book ratio. Tobin’s Q is computed by [(book value of debt + market value of equity) / book value of asset]. Market-to-book ratio is computed by [market value of common equity/(book value of asset – book value of debt – book value of preferred stocks)]. 24 outliers are identified and dropped from the sample when estimating equation (1), (2), and (3). Observations are identified as outliers if a studentized residual obtained from a regression of Tobin’s Q on corporate governance index is greater than 1.96 or smaller than –1.96. Such method identified the top 24 figures as outliers. 16 outliers are identified and dropped from the sample when estimating equation (4), (5), and (6). Observations are identified as outliers if a studentized residual obtained from a regression of market-to-book ratio on corporate governance index is greater than 1.96 or smaller than –1.96. Such method identified the bottom 6 figures and top 10 figures as outliers. The bottom 6 figures include all the negative market-to-book ratios. White’s heteroscedasticity-consistent robust standard errors are reported in the parentheses. ***, **, and *, respectively indicate significance levels at 1%, 5%, and 10%.

| |Tobin’s Q |Market-to-Book Ratio |

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

|Corporate Governance |0.0035*** |0.0025*** |0.0022*** |0.0049*** |0.0032* |0.0065*** |

| |(0.0006) |(0.0007) |(0.0006) |(0.0019) |(0.0019) |(0.0021) |

|Debt / Asset | | |0.5464*** | | |0.1643 |

| | | |(0.0451) | | |(0.1434) |

|Log (asset) | | |-0.0224*** | | |-0.0595*** |

| | | |(0.0062) | | |(0.0218) |

|Log (years of listing) | | |-0.0604*** | | |-0.1341*** |

| | | |(0.0113) | | |(0.0286) |

|Sales Growth (ratio) | | |0.0581*** | | |0.1916*** |

| | | |(0.0113) | | |(0.0837) |

|Intercept |0.7110*** |0.6966*** |0.6411*** |0.4962*** |0.3614*** |0.8258*** |

| |(0.208) |(0.0832) |(0.0592) |(0.0559) |(0.1363) |(0.1810) |

|Industry Dummies |No |Yes |Yes |No |Yes |Yes |

|Number of Observations |516 |516 |496 |522 |522 |501 |

|Adjusted R-Square |0.0524 |0.0872 |0.3643 |0.0163 |0.0702 |0.1516 |

Table 6. Corporate Governance and Firm Value

(Regression on Different Sub-Samples)

Regression of firm value on corporate governance index is conducted for a number of sub-samples. Samples are divided in three ways: bank versus non-banks; chaebol affiliated firms versus non-chaebol firms; and firms with book value of asset greater than KRW 2 trillion versus those less than KRW 2 trillion. Panel A shows the results for Tobin’s Q and Panel B shows the results for market-to-book ratio. White’s heteroscedasticity-consistent robust standard errors are reported in the parentheses. ***, **, and *, respectively indicate significance levels at 1%, 5%, and 10%.

A. Tobin’s Q

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

| |Banks |Non-Banks |Chaebol |Non-Chaebol |Asset > KRW 2 trillion |Asset < KRW 2 trillion |

|Corporate Governance |0.0001 |0.0023*** |0.0032** |0.0020*** |0.0018 |0.0014* |

| |(0.0003) |(0.0007) |(0.0015) |(0.0007) |(0.0013) |(0.0008) |

|Debt / Asset |0.2212 |0.5473*** |0.4978*** |0.5775*** |-0.0612 |0.5821*** |

| |(0.4735) |(0.0452) |(0.1020) |(0.0531) |(0.2101) |(0.0456) |

|Log (Asset) |0.0106* |-0.0228*** |0.0050 |-0.0305*** |-0.0193 |-0.0438*** |

| |(0.0056) |(0.0068) |(0.0119) |(0.0090) |(0.0192) |(0.0091) |

|Log (years of listing) |-0.0268*** |-0.0616*** |-0.0405 |-0.0612*** |-0.0551** |-0.0536*** |

| |(0.0070) |(0.0118) |(0.0224) |(0.0136) |(0.0212) |(0.0131) |

|Sales Growth (ratio) |-0.0033 |0.0574*** |0.0659* |0.0706 |0.0448 |0.0583*** |

| |(0.0208) |(0.0146) |(0.0119) |(0.0553) |(0.1343) |(0.0130) |

|Intercept |0.7559 |0.6432*** |0.5036*** |0.6695*** |0.9886*** |0.7314*** |

| |(0.4265) |(0.0621) |(0.1141) |(0.0667) |(0.2348) |(0.0689) |

|Industry Dummies |No |Yes |Yes |Yes |Yes |Yes |

|Number of Observations |17 |479 |104 |392 |64 |432 |

|Adjusted R-Square |0.6965 |0.3458 |0.4386 |0.3746 |0.4270 |0.3703 |

B. Market-to-Book Ratio

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

| |Banks |Non-Banks |Chaebol |Non-Chaebol |Asset > KRW 2 trillion |Asset < KRW 2 trillion |

|Corporate Governance |0.0041 |0.0066*** |0.0138** |0.0044** |0.0104 |0.0038* |

| |(0.0038) |(0.0022) |(0.0053) |(0.0022) |(0.0072) |(0.0023) |

|Debt / Asset |-4.2275 |0.1961 |-0.6829* |0.4037** |-1.6684** |0.3046** |

| |(5.3210) |(0.1452) |(0.3883) |(0.1582) |(0.6628) |(0.1453) |

|Log (asset) |0.0642 |-0.0629*** |0.0211 |-0.0826*** |0.0015 |-0.1205*** |

| |(0.0779) |(0.0228) |(0.0394) |(0.0284) |(0.0754) |(0.0290) |

|Log (years of listing) |-0.3724** |-0.1207*** |-0.1336* |-0.1323*** |-0.1096 |-0.1112*** |

| |(0.1245) |(0.0292) |(0.0712) |(0.0324) |(0.0736) |(0.0349) |

|Sales Growth (ratio) |0.7514 |0.1809** |0.0915* |0.5178*** |0.5270 |0.1763** |

| |(0.4662) |(0.0800) |(0.0493) |(0.1820) |(0.4030) |(0.0769) |

|Intercept |4.7733 |0.7917*** |0.7898** |0.8326*** |1.0561 |1.0615*** |

| |(4.8104) |(0.1859) |(0.3266) |(0.1497) |(0.8196) |(0.1943) |

|Industry Dummies |No |Yes |Yes |Yes |Yes |Yes |

|Number of Observations |17 |484 |107 |394 |64 |437 |

|Adjusted R-Square |0.7652 |0.1493 |0.3911 |0.1435 |0.4277 |0.1581 |

Table 7. Corporate Governance and Firm Value

(2 SLS Estimation)

To control for the endogeneity of corporate governance index variable, regression of firm value on corporate governance index is estimated using 2SLS. Asset size dummy is assumed to be an exogenous variable that is highly correlated with corporate governance, but that does not appear in the firm value equation. Equations (1) and (2) regress corporate governance on asset size dummy plus all other exogenous variables. Equations (2) and (4) are estimated by OLS. Equations (3) and (4) are estimated by 2SLS. White’s heteroscedasticity-consistent robust standard errors are reported in the parentheses. ***, **, and *, respectively indicate significance levels at 1%, 5%, and 10%.

| |CG (1st Stage) |Tobin’s Q |CG (1st Stage) |Market-to-Book Ratio |

| | |OLS |2SLS | |OLS |2SLS |

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

|Corporate Governance | |0.0022*** |0.0162*** | |0.0065*** |0.0353*** |

| | |(0.0006) |(0.0039) | |(0.0021) |(0.0111) |

|Debt / Asset |4.5932* |0.5464*** |0.4725*** |5.4379* |0.1643 |-0.0115 |

| |(2.6486) |(0.0451) |(0.0499) |(2.8000) |(0.1434) |(0.1596) |

|Log (asset) |2.8701*** |-0.0224*** |-0.0826*** |2.6037*** |-0.0595*** |-0.1791*** |

| |(0.4930) |(0.0062) |(0.0177) |(0.4868) |(0.0218) |(0.0499) |

|Log (years of listing) |-0.3615 |-0.0604*** |-0.0499*** |-0.2137 |-0.1341*** |-0.1188*** |

| |(0.7180) |(0.0113) |(0.0114) |(0.7108) |(0.0286) |(0.0305) |

|Sales Growth (ratio) |2.4996** |0.0581*** |0.0233 |1.9973** |0.1916*** |0.1310 |

| |(1.2524) |(0.0113) |(0.0164) |(1.0036) |(0.0837) |(0.0816) |

|Asset Size Dummy |8.7018*** | | |9.7425*** | | |

| |(2.34090 | | |(2.3822) | | |

|Intercept |3.9634 |0.6411*** |0.6812*** |4.5040 |0.8258*** |0.9186*** |

| |(3.8526) |(0.0592) |(0.0603) |(3.8662) |(0.1810) |(0.1733) |

|Industry Dummies |Yes |Yes |Yes |Yes |Yes |Yes |

|Number of Observations |496 |496 |496 |504 |501 |501 |

|Adjusted R-Square |0.3933 |0.3643 |0.3757 |0.4067 |0.1516 |0.1554 |

Table 8. Corporate Governance and Firm Return

(Difference-in-Mean Test)

Firms are categorized into five groups according to their corporate governance scores and for each quintile (equally-weighted) average returns are computed. 34 observations are dropped from the sample. Among the 34 observations, 23 are dropped as outliers. Observations are considered as outliers if a studentized residual obtained from a regression of gross return on corporate governance index is greater than 1.96 or smaller than –1.96. Such method identified the 23 highest figures as outliers. In addition, 11 firms that did not exist over the entire sample period - year 2001 - are also dropped from the sample. Column (1) reports buy-and-hold gross return for each group from December 26, 2000 to December 28, 2001. Stock-splits and stock-mergers are adjusted. Columns (2), (3), (4), and (5) report average excess returns. Column (2) reports return in excess of KOSPI return. Column (3) reports return in excess of the required rate of return from the market model. Column (4) reports return in excess of the required rate of return from CAPM. Beta values are obtained from FnGuide. Lastly, column (5) reports return in excess of the average return of firms in the same size category. Such matching portfolio does not include the stock being concerned. Size is computed by market capitalization on December 26, 2000. The last row reports the t-test results. Standard errors assuming unequal variances are reported in the parentheses. ***, **, and *, respectively indicate significance levels at 1%, 5%, and 10%.

|Groups by Corporate |Corporate Governance |Number of Observation|Gross Returns |Excess Returns |

|Governance Index |Index Score Range | | | |

|Scores | | | | |

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

| | | | |(1) – KOSPI Return |(1) – Required Return |(1) – Required Rate of |(1) – Return in the Same |

| | | | | |from Market Model |Return from CAPM |Size Category |

|(1) Low |0 ~ 18.80 |105 |33.7740 |-3.6959 |9.5824 |8.7054 |-7.0371 |

|(2) |18.83 ~ 25.16 |101 |30.5950 |-6.8748 |5.0934 |4.1472 |-10.6418 |

|(3) |25.27 ~ 31.04 |104 |33.5834 |-3.8864 |7.9371 |6.6445 |-7.2328 |

|(4) |31.19 ~ 39.00 |100 |32.7056 |-4.7642 |2.2931 |2.1666 |-9.9352 |

|(5) High |39.02 ~ 75.28 |96 |58.9681 |21.4983 |20.0920 |20.5739 |20.0817 |

|(5) – [(1) ~ (4)] | | |26.2861*** |26.2861*** |13.8107** |15.1090** |28.7633*** |

| | | |(6.8556) |(6.8556) |(6.6893) |(6.7421) |(6.9929) |

Table 9. Corporate Governance and Firm Return

(Regression Analyses)

Excess returns reported in Table 7 are regressed on the corporate governance index and on other control variables. Control variables include size (log of market capitalization), market-to-book ratio, past return (over year 2000), share price (in logs), trading volume (in logs), dividend yield, sales growth (over past five years), and industry dummies. Panel A shows results without B/M ratio as a control variable. Panel B shows results with B/M ratio as a control variable. White’s heteroscedasticity-consistent robust standard errors are reported in the parentheses. ***, **, and *, respectively indicate significance levels at 1%, 5%, and 10%.

A. Without Market-to-Book Raito

| |(1) |(3) |(5) |(6) |

| |Gross Return – KOSPI Return |Gross Return – Required Return from |Gross Return – Required Rate of |Gross Return – Return in the Same |

| | |Market Model |Return from CAPM |Size Category |

|Corporate Governance Index |0.5799** |0.5230** |0.5166** |0.5400** |

| |(0.2294) |(0.2386) |(0.2374) |(0.2283) |

|Log (Size) |-2.9620 |-3.8805 |-4.2420* | |

| |(2.4968) |(2.4861) |(2.4970) | |

|Past Return |-0.1284*** |-0.0001 |-0.0070 |-0.1109** |

| |(0.0490) |(0.0640) |(0.0628) |(0.0492) |

|Log (Price) |11.8471*** |12.5823*** |13.0378*** |11.5376*** |

| |(2.9125) |(2.9795) |(2.9923) |(2.6434) |

|Log (Trading Volume) |2.2531 |-0.7344 |0.1749 |1.2244 |

| |(1.7694) |(1.7274) |(1.7420) |(1.4368) |

|Dividend Yield (ratio) |0.4898 |0.4404 |0.4385 |0.3572 |

| |(0.4660) |(0.4548) |(0.4587) |(0.4640) |

|Sales Growth (ratio) |0.0197 |-10.1179 |-7.0063 |-3.4819 |

| |(7.0256) |(6.9033) |(7.0117) |(7.5516) |

|Intercept |-124.1687*** |-39.0214 |-59.1761* |-130.2550*** |

| |(37.0815) |(35.4489) |(35.9095) |(37.1313) |

|Industry Dummy |Yes |Yes |Yes |Yes |

|Number of Observation |480 |479 |479 |480 |

|Adjusted R Square |0.2432 |0.1915 |0.1967 |0.2456 |

B. With Market-to-Book Raito

| |(1) |(3) |(5) |(6) |

| |Gross Return – KOSPI Return |Gross Return – Required Return from |Gross Return – Required Rate of |Gross Return – Return in the Same |

| | |Market Model |Return from CAPM |Size Category |

|Corporate Governance Index |0.4700** |0.4030* |0.4016* |0.5420** |

| |(0.2237) |(0.2324) |(0.2308) |(0.2295) |

|Log (Size) |-3.6628 |-4.4684* |-4.8813* | |

| |(2.4910) |(2.4927) |(2.5013) | |

|Market-to-Book Ratio |-3.6834 |-3.4863 |-2.6161 |0.3338 |

| |(4.6958) |(5.0957) |(4.9870) |(1.4813) |

|Past Return |-0.1358*** |-0.0080 |-0.0160 |-0.1111** |

| |(0.0485) |(0.0647) |(0.0631) |(0.0492) |

|Log (Price) |12.3008*** |13.0252*** |13.3979*** |11.5301*** |

| |(2.9042) |(2.9623) |(2.9776) |(2.6475) |

|Log (Trading Volume) |2.7195 |-0.3267 |0.5343 |1.2077 |

| |(1.8155) |(1.7807) |(1.7935) |(1.4409) |

|Dividend Yield (ratio) |0.1167 |0.0712 |0.0875 |0.3707 |

| |(0.4614) |(0.4582) |(0.4584) |(0.4721) |

|Sales Growth (ratio) |0.3766 |-9.8976 |-6.9713 |-3.5591 |

| |(7.1000) |(7.0507) |(7.1838) |(7.5874) |

|Intercept |-124.9280*** |-39.6143 |-58.2528 |-130.1156*** |

| |(37.9103) |(36.0553) |(36.5552) |(37.1861) |

|Industry Dummy |Yes |Yes |Yes |Yes |

|Number of Observation |469 |468 |468 |480 |

|Adjusted R Square |0.2549 |0.2032 |0.2084 |0.2457 |

Table 10. Corporate Governance and Firm Return

(Regression on Different Sub-Samples)

Regression of firm’s excess return on corporate governance index is conducted for a number of sub-samples. Samples are divided in three ways: bank versus non-banks; chaebol affiliated firms versus non-chaebol firms; and firms with book value of asset greater than KRW 2 trillion versus those less than KRW 2 trillion. Excess return is defined by gross return – KOSPI return. White’s heteroscedasticity-consistent robust standard errors are reported in the parentheses. ***, **, and *, respectively indicate significance levels at 1%, 5%, and 10%.

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

| |Banks |Non-Banks |Chaebol |Non-Chaebol |Asset > KRW 2 trillion |Asset < KRW 2 trillion |

|Corporate Governance |2.7372** |0.3470* |-0.0224 |0.5568** |0.4712 |0.3973* |

| |(1.0520) |(0.1958) |(0.4967) |(0.2390) |(0.6398) |(0.2260) |

|Log (price) |44.3212* |11.6091*** |17.7592*** |10.0059*** |21.2165*** |10.5744*** |

| |(21.7784) |(2.5528) |(6.4754) |(2.7530) |(6.0784) |(2.8325) |

|Past return |0.1713 |-0.1444*** |-0.2234 |-0.1185*** |-0.1603 |-0.1481*** |

| |(0.8439) |(0.0426) |(0.1649) |(0.0461) |(0.4467) |(0.0438) |

|Intercept |-439.8873** |-104.0005*** |-116.1743** |-93.9368*** |-291.6720*** |-96.4283*** |

| |(176.9082) |(27.1785) |(57.3918) |(29.2991) |(71.2835) |(29.5866) |

|Industry Dummies |No |Yes |Yes |Yes |Yes |Yes |

|Number of Observations |13 |493 |102 |404 |58 |448 |

|Adjusted R-Square |0.6703 |0.2162 |0.4002 |0.1653 |0.6039 |0.1661 |

-----------------------

* Professor of Law, Stanford Law School, Stanford, California U.S.A. 84305. Tel: (+1) 650-725-9845, fax (+1) 650-725-0684, e-mail: bblack@stanford.edu.

** Professor of Finance, Korea University Business School, Anam-Dong, Sungbuk-Ku, Seoul, Korea 136-701. Tel: (+82-2) 3290-1929, fax: (+82-2) 929-3405, e-mail: jangya@

*** Professor of Finance, KDI School of Public Policy and Management, Chongyangri-Dong Dongdaemun-Ku, Seoul, Korea 130-868. Tel: (+82-2) 3299-1030, fax: (+82-2) 968-5072, e-mail: wc_kim@kdischool.ac.kr.

[1] The pioneering work in cross-country studies is La Porta et al. (1997).

[2] Exception to this would be Gompers, Ishii, and Metrick (2001) who find that anti-takeover provisions and shareholder rights are significantly related to firm value in the U.S.

[3] Recently, there are a number of attempts in this direction. Gompers, Ishii, and Metrick (2001) combine anti-takeover provisions and shareholder rights to construct a corporate governance index. Black (2001) constructs corporate governance risk of 16 Russians firms. Durnev and Kim (2002) make use of corporate governance and transparency rankings produced by Credit Lyonnais Securities Asia (CLSA) and Standard and Poor’s.

[4] To name a few, Millstein and MacAvoy (1998) and Bhagat and Black (1999) investigate the relationship between board characteristics and firm performance. Karpoff, Malatesta, and Walking (1996) and Carleton, Nelson, and Weisbach (1998) link firm performance with shareholders’ activism. Bhagat, Carey, and Elson (1999) look at the relationship between outside directors’ pay and firm performance. Sundaramurthy, Mahoney, and Mahoney (1997) links firm performance with anti-takeover provisions and LLSV (2002) analyses the relationship between investor protection and firm performance.

[5] Governance rankings produced by Credit Lyonnaise Securities Asia (CLSA) is one example.

[6] The original intension of this survey was to award companies that stand out in terms of their corporate governance.

[7] 540 companies responded out of the 560 companies that were surveyed.

[8] The KSE survey is composed of questions in five categories: (i) shareholders, (ii) directors & board, (iii) outside directors, (iv) auditing, and (v) other stakeholders. The questions on “auditing” cover both auditing as such, and disclosure practices in general. Thus, “disclosure and transparency” is a more descriptive term. The questions on “other stakeholders” are in fact a miscellaneous set of questions, some relating to stakeholders, but others relating to other survey categories. We reassigned the corporate governance related questions in this category to one of the other four categories, as appropriate. We also reassigned other selected questions, where this seemed appropriate.

[9] Since we are computing a ratio, we do not have to worry directly about missing values.

[10] Skewness is computed to be 0.5890. The median is 28.0202.

[11] Please not that KSE requested that we do not disclose the names of the companies.

[12] We do not have a measure for the market value of debt.

[13] We compute equal-weighted averages.

[14] This coincided with the 24 largest figures.

[15] This coincided with 6 smallest figures and 10 largest figures. The bottom 6 figures include all the negative market-to-book ratios.

[16] The p-value is 0.2.

[17] Industry dummies can also capture product market competition that can influence firm value.

[18] The correlation coefficient between corporate governance index and asset size dummy (=1 if greater than KRW 2 trillion) in our sample is 0.5501 and statistically different from zero at 1% significance level.

[19] The correlation coefficient between corporate governance index and banking dummy (=1 if a commercial bank or a merchant bank) in our sample is 0.3881 and statistically different from zero at 1% significance level.

[20] With the amendment of the Securities & Exchange Act in March 28, 2001, all listed companies with book asset value greater than KRW 2 trillion must obtain board approval for related party transactions if the size is greater than 1% of book asset value or 1% of total sales. Such amendment, however, was not in effect at the time when the survey was conducted.

[21] The correlation coefficient between corporate governance index and chaebol dummy (=1 if affiliated to a chaebol) in our sample is 0.2354 and statistically different from zero at 1% significance level.

[22] We also ran a regression with intercept dummy variables representing asset size, affiliation to banking industry, and affiliation to chaebol. The coefficient on corporate governance index turned out to be still significant with a similar magnitude.

[23] The method suggested by Wooldridge follows two steps. First, the corporate governance index variable is regressed on all exogenous variables and residuals are obtained from the estimation. Second, the residual is added in the original equation and OLS is estimated. If the coefficient on the residual is statistically different from zero, we conclude that corporate governance is endogenous.

[24] Asset size dummy takes a value of 1 if book value of asset is greater than KRW 2 trillion, and 0 otherwise.

[25] Of course some firms with asset size close to the KRW 2 trillion level might intentionally manage their balance sheet so as not to be subject to the legal requirement.

[26] This is consistent with the test result for endogeneity that there exists a negative correlation between the corporate governance index variable and the error term.

[27] Intercept and beta estimates are obtained from fnGuide, a commercial vender of financial data in Korea. According to fnGuide, coefficients are estimated at a monthly frequency over a 36-month period from April 1999 to April 2001.

[28] KOSPI return of 37.47% and risk-free rate of 5.88% is used to compute required rate of return. Average commercial bank deposit rate is used as a proxy for risk-free rate.

[29] We also ran a regression with intercept dummy variables representing asset size, affiliation to banking industry, and affiliation to chaebol. The coefficient on corporate governance index turned out to be still significant with a similar magnitude.

[30] We tried other measures of excess return, but no material difference was found.

[31] Corporate governance index ranges from 0 to 100.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download