Restructuring of Family Firms after the East Asian Financial Crisis: Shareholder Expropriation or Alignment? Piruna Polsiri * and Yupana Wiwattanakantang ** This version: February 2004 (Preliminary: Do not quote without permission) Abstract We investigate whether controlling shareholders expropriate corporate resources during a financial crisis using the data of Thai. Our results are consistent with the argument of Friedman, Johnson, and Mitton (2003) that the propensity to tunnel and prop is higher for business groups in particular if they are organized in pyramids. Specifically, we find that that belong to the top 30 business groups implement a number of restructuring activities such as expansion, executive turnover, and dividend cuts, more often than non group. Among the business group, we find that with higher ratio of cash flow rights to voting rights are less likely to implement the following restructuring measures: downsizing, expansion, and executive turnover. Interestingly, even in the business group that are not organized in pyramids, debt increases the probability of operational restructuring. This evidence is consistent with the view that debt increases the incentives to prop. JEL classification: G32; G33; G34; P12; K12 Keywords: Business Groups; Family Firms; Corporate Governance; East Asian Financial Crisis; Ownership Structure; Restructuring *Department of Finance, University of Melbourne, Parkville, VIC 3010, Australia, and Department of Finance, Dhurakijpundit University, Thailand. Tel.: +61-3-8344-0804; e-mail: piruna@yahoo.com. **Center for Economic Institutions, Institute of Economic Research, Hitotsubashi University, 2-1 Naka, Kunitachi, Tokyo 186-8602, Japan. Tel.: +81-42-580-8374; Fax: +81-42-580-8333; e-mail: yupana@ier.hit-u.ac.jp.
1. Introduction The conflicts of interests between a firm s controlling shareholders and minority shareholders have been crucial issues in the discussion of family at least since the analysis of La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1999), Claessens, Djankov, and Lang (2000), and Faccio and Lang (2002) who show that family ownership is universal around the world. La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1997, 1998, 2000), Johnson, La Porta, Lopez-de-Silanes, and Shleifer (2000), and Burkart, Panunzi, and Shleifer (2003) show that legal protection of minority shareholders varies across countries, and this variation determines the level of the ownership concentration, the existence of family worldwide, the patterns of separation between ownership and management, and the degree of expropriation by corporate insiders. In the countries with moderate legal protection of outside investors, the controlling shareholder may be beneficial to the firm as a monitor (Anderson and Reeb (2003), Burkart, Panunzi, and Shleifer (2003), and Morck and Yeung (2003)). However, in the emerging economies where the legal and regulatory systems are weak, the controlling shareholder is likely to expropriate the firm s resources. Under some circumstances, for example when are doing well, the controlling shareholder family is likely to tunnel resources out of the for their own benefits (Johnson, Boone, Breach, and Friedman (2000), Johnson, La Porta, Lopez-de-Silanes, and Shleifer (2000), Bertrand, Mehta, and Mullainathan (2002), and Morck and Yeung (2003)). Friedman, Johnson, and Mitton (2003), however, argue that tunneling does not occur all the time. During the time when the return on investment is temporarily low, the controlling shareholder may prop up or bailing out the by injecting their private funds or in other forms. The incentives of propping are to keep the alive so that the controlling shareholders can keep their option to expropriate in the future. In the extreme situation when the expected rate of return is extremely low, however, the controlling shareholders are likely to choose to abandon the instead of rescuing them (Johnson, Boone, Breach, and Friedman (2000)). The literature suggests that expropriation (tunneling and propping) is likely to occur in business groups in which the ownership and control structures are often organized as pyramids. The pyramidal structure creates the separation between cash flow and voting rights. So, while maintaining control over in the pyramid without bearing too much of the cash 2
flows, the controlling shareholder, therefore, has strong incentives to divert resources by tunneling (Wolfenzon (1999), Bebchuk, Kraakman, and Triantis (2000), and Shleifer and Wolfenzon (2000)). Friedman, Johnson, and Mitton (2003) contend that if tunneling and propping are a symmetric behavior, the chance of propping should be higher for that are affiliated with a family owned business group. In addition, given that business groups are often diversified and hence have a number of affiliations, they are likely to be less transparent. A higher level of information asymmetries might facilitate expropriation (Claessens, Djankov, Fan, and Lang (2003), Lins and Servaes (2002), Mitton (2002)). To investigate the expropriation effect empirically, the literature typically has concentrated on linking ownership and performance (Wiwattanakantang (2001), Khanna and Palepu (2000), Claessens, Djankov, Fan, and Lang (2002), Mitton (2002), Anderson and Reeb (2003), Attig, Fischer, and Gadhoum (2003), Joh (2003), Lemmon and Lins (2003), Lins (2003), Baek, Kang, and Park (2004)). Considerably less attention has been placed on measuring the extent of expropriation via corporate activities (Bae, Kang, and Kim (2002), Bertrand, Mehta, Mullainathan (2002), Volpin (2002), and Friedman, Johnson, and Mitton (2003)). This dimension is the focus of our study. Specifically, in this study, we investigate the propping and tunneling argument of Friedman, Johnson, and Mitton (2003). They contend that propping is likely to occur in a weak legal and regulatory environment, and when there is unexpected and moderate economic shock. Otherwise, looting instead of propping would occur if the shock is too strong. To study this expropriation effect, a number of studies choose to focus on the East Asian crisis (Johnson, Boone, Breach, and Friedman (2000), Mitton (2002), Friedman, Johnson, and Mitton (2003), Mitton (2003), Baek, Kang, and Park (2004)). In this study, we focus on a single country, Thailand which provides a natural setting to study this issue. An advantage of investigating one country is that we can control for the institutional effects (such as legal and regulatory effects) because all the are operating in the same environment. Also, it allows us to construct governance variables at a detailed level. Our methodology is to how Thai respond to the 1997 East Asian financial crisis. Our sample includes all non financial listed on the Stock Exchange of Thailand. We examine restructuring activities undertaken by the following the 1997 East Asian crisis. 3
More precisely, following the financial distress literature, we consider both operational actions (asset downsizing, expansion, and executive turnover) and financial actions (dividend cuts, debt restructuring, and capital raising) over the period 1997-2000. Following the expropriation hypothesis stated earlier, we use the country s top 30 business groups as a measure for the propensity to prop. In other words, we hypothesize that business groups are more likely to implement restructuring actions than non business groups because they would like to keep the option to expropriate corporate resources in the future. Our empirical results are consistent with the hypothesis. We find that that belong to families who own the top 30 business groups implement a number of restructuring activities such as expansion, executive turnover, and dividend cuts, more often than non group. Among the business group, we find that with a higher ratio of cash flow to voting rights are less likely to implement the following restructuring measures: downsizing, expansion, and executive turnover. The results are consistent with the argument that the propensity to tunnel and prop is higher for business groups, in particular if they are organized in pyramids (Claessens and Fan (2002), Friedman, Johnson, and Mitton (2003) and Wolfenzon (1999)). Interestingly, in the business group that use less pyramidal ownership structure, debt increases the probability of restructuring actions, in particular downsizing, expansion, and executive turnover. This evidence supports the argument of Friedman, Johnson, and Mitton (2003) that debt increases the incentives to prop. This study is organized as follow. Section 2 discusses the data and sample design. Section 3 presents the empirical results. Section 4 concludes with a summary of this research s main findings. 2. Data In this section, we discuss our sample selection criteria and data sources. We also define business groups and describe their characteristics. In addition, we classify restructuring actions. 4
2.1 The sample Our interest is to investigate the nature of firm responses to the East Asian financial crisis that hit Thailand in July 1997. We define 1997 as the base year since this is when experienced the economic shock and might start undertaking various restructuring actions in response. As may not have responded to the shock immediately, we think that it is more appropriate to investigate restructuring actions over a longer period. Specifically, our investigation covers the period from 1997 to 2000. In this aspect, we are different from Baek, Kang, and Park (2002) who investigate the immediate responses of in Korea focusing on the period between November 1997 and December 1998. Our samples include all non-financial companies listed on the Stock Exchange of Thailand during the period 1997-2000. The sample contains 1,328 observations. The main data source is the I-SIMS database produced by the Stock Exchange of Thailand. This database provides the information on the ownership, board of directors, and the financial data. 2.2 The ownership data We construct a comprehensive database of ownership using the I-SIMS database which provides detailed information on shareholders with shareholdings of at least 0.5%. We hand-collect additional information on the family relationships between major shareholders, the names of affiliated companies of the in our sample, the family relationships between the board members from the document FM 56-1. This document is a company annual report that is submitted to the Stock Exchange annually. It is available at the library of the Stock Exchange of Thailand and its website. In addition to the family relationship information provided in the document FM 56-1, we try to identify the family relationships via marriage by using various books namely Johnstone et al. (2001), Pornkulwat (1996), Sappaiboon (2000a, 2000b, and 2001). To trace the ultimate ownership of private companies who appear as corporate shareholders of our sample, we use the BOL database produced by the Business OnLine Ltd. Basically, the BOL databank includes all registered companies in Thailand. The Business OnLine Ltd. has the license from the Ministry of Commerce to reproduce this information. With all the above data sources, we believe that we are able to trace the ultimate owners of all privately owned 5
companies that are the shareholders of in our focus 1. We calculate the cash-flow and voting rights using the standard approach introduced by La Porta, Lopez-de-Silanes, Shleifer, and Vishny (1999), Claessens Djankov and Lang (2000b), Faccio and Lang (2002), and Lemmon and Lins (2003). We use a slightly different definition of controlling shareholders or ultimate owners, however. Instead of using the 20% cut-off which is commonly used in the literature, we use the 25% cut-off following the definition of the Stock Exchange of Thailand (Stock Exchange of Thailand, 1998). We justify this choice by referring to the Thai corporate law. According to the Public Limited Companies Act B.E. 2535, a shareholder needs at least 75% of a firm s voting rights to have veto rights. So, a shareholder who owns more than 25% of the voting rights can have sufficient voting rights to control a firm because in which case no other single shareholder would own enough voting rights to veto his decisions (see also Wiwattanakantang (2000 and 2001) and Khantavit, Polsiri, and Wiwattanakantang (2003) for further discussion). 2.3 The business groups Our focus is the top 30 largest business groups in Thailand. We define the business groups following Charumilind, Kali, and Wiwattanakantang (forthcoming). Specifically, they use the ranking of business groups in Thailand done by Suehiro (2000). Suehiro (2000) ranks the business groups based on sales of the top 1000 companies in 1994 that are ranked and published by Advance Research Group (1995). Note here that if all the affiliations of the business groups were not included in the top 1000 companies, we would encounter some biases. Nevertheless, as our focus is the wealth of the business groups and hence the ability to bail out the, we believe that this ranking should provide a reasonable measure of Thailand large groups. The names of these top 30 groups are shown in Table 2 in Charumilind, Kali, and Wiwattanakantang (forthcoming). A firm is classified as an affiliation of the top 30 groups if at least 25% of its shares are owned by one of the families that own the top 30 business groups. We think that this level 1 Khantavit, Polsiri, Wiwattanakantang (2003), and Charumilind, Kali, and Wiwattanakantang (forthcoming) provide further discussion of the ownership of Thai. 6
of shareholdings should be sufficiently high to provide the incentives for the owner to bail out the firm during its difficulties. Table 1 provides the information on the ownership, governance, and financial characteristics of business group in comparison with non group during 1996-2000. The information on the onset of the crisis (1996) is given for reference. Panel A shows the ownership and corporate governance structure. The ownership of both group and non group is very concentrated in the hands of the largest shareholder s family. In general, the voting rights held by the largest shareholder for business group are significantly higher than those of non group in all periods. For example, in 1997, the average voting rights held by the largest shareholder of group are 46.3% while those of non group are 40.6%. The average ratio of the cash flow rights to the voting rights is lower than one for both group and non group and for all periods. This evidence indicates that pyramids and cross-shareholdings are used by the sample. Compared to non business group, business group do not significantly have a lower ratio of cash flow rights to voting rights. Regarding the management and control structure, the median values of the board size for business groups are 13 for all the years which are significantly larger than the median values of the board size of 10 for non group. We find that there is at least one person from the largest shareholder s family serving as top executives in 38%, 40%, 42%, and 44% of the business group in 1997, 1998, 1999, and 2000, respectively. Non group also have a similar structure. Here, top executives include the following positions: Honorary chairman, chairman, vice chairman, president, vice president, chief executive officer, managing director, deputy managing director, and assistant managing director. We find that the family of the largest shareholder also sits in the board of directors. The median number of directors held by the largest shareholder s family is 3 for all the years for group, which is significantly higher than that of non group. Panel B shows the financial characteristics of our sample. On average business group are significantly much larger than non group in terms of total assets and sales in all periods. For example, the mean values of total assets of business group in 1997 are about three times of those of non group (Baht 16,788.4 million vs. Baht 7
5,986.2 million). In all periods, the average values of the debt to asset ratio (both measured by the book and market values) of business group are not significantly different from those of non business group. The debt to asset ratio of both business group and non group increases after the crisis. For business group, the average values of the ratio of debt to asset ratio increase from 39% in 1996 to 51% in 1997, and 66% in 1998. For non business group, the average values of the ratio of debt to asset ratio increase from 42% in 1996 to 53% in 1997, and slightly decline to 52% in 1998. In all periods, group are not significantly profitable compared to non group. On average, both group and non group in our sample were vulnerable to the crisis. For group, the average ratio of the EBIT to total assets declines from 12.2% in 1996 to 3.9% in 1997. For non group, the average ratio of the EBIT declines from 7.5% in 1996 to 1.58% in 1997. After taking account of interest expenses, by the end of 1997, the average values of the net profit ratio for group and non group are -0.49% and -3.59%, respectively. < Table 1 here > 2.4 The response to the crisis Following the literature, we consider both operational and financial restructuring actions as follow. 1. Operational actions: 1.1 Asset downsizing. This action includes selling assets (e.g., financial securities, land, properties, and stakes in other businesses or joint ventures), closing plants, reducing production capacities, discontinuing or suspending production operations, shutting down a division/office/branch/subsidiary. 1.2 Expansion. This action includes forming a joint venture or strategic alliance, fully or partially acquiring other businesses, diversifying into a new line of business, constructing a new facility, establishing a new division/office/ branch/subsidiary, expanding an existing production facility, and increasing investment in an existing subsidiary. 8
1.3 Top management turnover: A turnover occurs when at least one of the top incumbent management is replaced. Here, the top management includes the chairman of the board, president, vice president, chief executive officer, managing director, and deputy managing director. 2. Financial actions: 2.1 Dividend cut. This occurs when the firm reduces its dividend payout from the previous year or omits its dividend payout after paying dividend in the previous year. 2.2 Debt restructuring. A firm reduces required interest or principal payment on debt agreement, extends debt maturity, exchanges equity securities (common stocks or securities convertible to common stocks) for debt or offers creditors the firm s equity securities, and appoints a financial advisor to assist in debt restructuring process. 2.3 Capital raising. This action includes issuing new loans, debentures, common stocks, and hybrid securities (preferred stocks, warrants, and convertible debentures). Due to data unavailability, we are not able include lay offs and management compensation reduction which are common restructuring actions documented in the literature. Many do not report such information. Also, the number of employees reported appears to be unreliable. We hand-collect the restructuring actions from the company daily news database available at the Stock Exchange of Thailand. In principle, the Stock Exchange of Thailand requires listed companies to report various corporate actions. This information is then disclosed at the website of the Stock Exchange of Thailand for six months, and later on is kept in the company daily news database. We also cross check the data with the company annual reports available at the library of the Stock Exchange of Thailand and its website, and a local English newspaper, the Bangkok Post. Table 2 shows the number of in the sample that engage in restructurings to overcome the shock. Note that the actions are not mutually exclusive. Hence, when a firm implements more than one action, it will enter our count more than once. Overall, 1,806 restructuring actions are implemented by our sample during the period 1997-2000. About one-third of the engage in multiple actions. Restructuring occurs most often in 1997 which is the onset of the crisis. About 87.64% of the in 1997 undertake some 9
kinds of restructuring. In total, about 624 restructuring actions are implemented in this year. Dividend cuts represent the most frequently observed response occurring in about 216 out of 356 which accounts for 60.67% of the. Expansionary actions and capital raising are the second and third most often implemented actions. About 48.03% of the take expansionary actions, and about 35.67% of the raise additional capital. Asset downsizing is taken by 19.66% of the. Changes in top management are taken by 8.71% of the. Debt restructuring occurs in only 2.53% of the. The evidence on Thai is in line with previous studies in that dividend cuts are most common and debt restructuring appears to be the least common restructuring actions. Ofek (1993) finds that dividend cuts and debt restructuring occur in 47% and 11% of financial distressed in the U.S., respectively. However, compared to other countries, asset downsizing is taken less frequently in Thai. Denis and Kruse (2000) find that asset downsizing occurs in 44% of financial distressed. Ofek (1993) reports that 23% of his U.S. sample engage in some form of asset downsizing. Similarly, Baek, Kang, and Park (2002) document that about 42.4% of in the Korean financial crisis engage in asset downsizing actions. < Table 2 here > 3. Empirical Analysis The primary focus is to test whether business groups are more likely to implement restructuring actions. First, we test whether business groups are more vulnerable to the crisis when compared with non group. Second, we perform univariate analysis comparing the responses between group and non group. Third, we perform multivariate analysis including various control variables. We also investigate the effects of corporate governance variables on the likelihood of restructuring. 3.1 The effect of the crisis: Are business groups more vulnerable than non business groups? To examine the effect of the 1997 financial crisis on the performance of business groups and non business groups, we use the model specification of Mitton (2002), Joh (2003), 10
and Baek, Kang, and Park (2004). Following Joh (2003), we use the accounting based profitability, the ratio of net income to total assets. We control for the effects of firm size, debt ratio, and industries. All the variables are measured at the end of 1997. Consistent with the results based on the univariate tests in Table 1, the results in equations (1) of Table 3 show that the estimated coefficient on the business group dummy is positive but not statistically significant at the conventional level. This result implies that business group are not more vulnerable to the crisis compared with non group. Our results are similar to Baek, Kang, and Park (2004) who show that the average ratio of net income to assets of the top 30 cheabol is insignificantly different from non chaebol at the onset of the crisis. For robustness checks, we run another regression using the ratio of the EBIT to total assets as a measure of profitability. The results remain the same. In addition, we run similar regressions using the data at the end of 1998. The results shown in the column (3) and (4) indicate that business groups are not more vulnerable compared to non business group in 1998. < Table 3 here > 3.2 The univariate analysis: The responses to the crisis Table 4 reports the frequency of restructuring actions taken by business group and non group. Overall, on average 61.49% of group implement at least one operational action during the period 1997-2000. More precisely, in about 69.32%, 55.06%, 60.24%, and 61.33% of business group undertake at least one operational action in 1997, 1998, 1999, and 2000, respectively. Compared to business group, operational restructurings are taken significantly less often, occurring in about 46.18% of non group over the period 1997-2000. Like group, operation actions occur more often in 1997 when about 52.7% of non group undertake operation actions. Of the three operational restructuring actions, expansion occurs most often. Some sort of expansionary actions are undertaken in 58%, 43.8%, 44.9%, and 45.3% of group in 1997, 1998, 1999, and 2000, respectively. The occurrence rate of expansion for non group is 43.8%, 27.1%, 26.9%, and 29.3% in 1997, 1998, 1999, and 2000, respectively. Non group appear to take expansionary actions significantly less than group. 11
Downsizing actions were taken less often in 24.5% and 20.5% of group and non group during the period of 1997-2000, respectively. Among all operational actions, top executive turnover is observed least often. There are attempts to change control in 15.9%, 12.4%, 18.1%, and 18.1% in 1997, 1998, 1999, and 2000, respectively for group, and in 6.6%, 9.2%, 11.4%, and 12% in 1997, 1998, 1999, and 2000, respectively for non group. The frequency of changes in top management has been higher for group compared to non group for the period 1997-2000. However, the differences are statistically significant only for 1997. Dividend cuts are the actions that are carried out most often as immediate responses to the crisis. In 1997, dividend cuts are observed in about 70.45% and 56.25% of business groups and non business group, respectively. The dividend cut likelihood is significantly higher is business group than non business group. The proportion of taking this action declines substantially afterwards. Dividend cuts are carried out in only 8.99%, 10.84%, and 18.67% of business groups in 1998, 1999, and 2000, respectively. Similarly, dividend cuts are taken by about 13.55%, 9.39%, and 12.40% of non business group in 1998, 1999, and 2000, respectively. We also find that business group engage in capital raising significantly more often than non business group in 1997-1999. This action is taken by about 45.45% 31.46%, 44.58%, and 34.67% of business group in 1997, 1998, 1999, and 2000, respectively. There are about 33.20%, 20.72%, 26.94%, and 30.58% of non business group in 1997, 1998, 1999, and 2000, respectively. In contrast to other restructuring actions, debt restructuring is taken significantly less often in business group compared to non business group in all years. The differences, however, are statistically significant only in 1997 and 1999. The proportion of adopting this action increases over time. The action was taken in 0%, 5.62%, 6.02%, and 14.67%, in business group in 1997, 1998, 1999, and 2000, respectively. In non business group, about 3.52%, 9.56%, 13.06%, and 17.36% restructure their debt in 1997, 1998, 1999, and 2000, respectively. < Table 4 here > 12
3.3 Regression Results In this section, we use probit regressions to estimate the relation between business group affiliations and the likelihood of various restructuring actions. The dependent variables are binary variables taking a value of one if a certain restructuring action occurs during the period 1997-2000, and zero otherwise. The dependent variables are not mutually exclusive. If a firm undertakes more than one type of restructuring actions, it will enter the regression more than once. In fact, we find that conducting multiple types of restructuring actions is common. Following the literature, we control for a number of firm specific factors. 1. Capital structure. High leverage are likely to respond more rapidly to a shock to avoid defaulting (Jensen (1989), Wruck (1990), and Ofek (1993)). For example, high leverage are likely to raise cash by selling their assets. However, leverage may not trigger restructuring in the countries with poor corporate governance where debt is soft and provided based on connections as suggested by Laeven (2001), La Porta, Lopez-de-Silanes, and Shleifer (2003), and Charumilind, Kali, and Wiwattanakantang (forthcoming). Here, leverage is measured by the ratio of total debt to total assets. 2. Firm size. Larger have a larger asset base, greater diversification, and greater number of employees. Thus, they are more likely to undertake restructuring actions such as asset sales and staff layoffs relative to small (Ofek (1993), Kang and Shivdasani (1997), and Baek, Kang, and Park (2002)). In addition, large tend to have better access to external sources of funds compared with small because they are well established and have a large asset base which can be used as collateral. Large, hence, are more likely to undertake expansionary actions and capital raising. We measure firm size by the log of total assets. 3. Performance. Previous studies show that past performance has both positive and negative effects on restructuring. Ofek (1993) documents a marginal positive relationship between annual stock returns and the likelihood that poorly performing undertake asset restructuring and make dividend cuts. In contrast, Kang and Shivdasani (1997) find that returns on assets are negatively associated with the likelihood of downsizing in that suffer a substantial performance decline. Baek, Kang, and Park (2002) show that higher 13
holding period returns decrease the probability of downsizing and internal reorganization taken by Korean during the East Asian crisis. Following Denis and Kruse (2000), we measure firm performance as a change in the ratio of EBIT to total assets from the previous year to the year in which restructure. This measure allows us to control for the impact of specific-firm performance deterioration during the economy-wide crisis. 4. Industry performance. Existing literature suggests that industry performance affects the probability of restructuring positively and negatively. On one hand, Kang and Shivdasani (1997) find that when the industry is performing well, financial distressed in Japan tend to acquire more assets, while US are likely to take some form of downsizing actions. Similarly, Shleifer and Vishny (1992) document that are less inclined to sell assets if their industry condition is poor. In contrast, Morck, Shleifer and Vishny (1989) show that top executive turnover is lower in that outperform their industry standard. Following Kang and Shivdasani (1997), industry performance is measured by the median ratio of operating income to total assets for an industry. 5. Liquidity. Liquidity may reduce the incentives to implement restructuring actions since liquidity is related to a firm s cash flows. Ofek (1993) finds that highly liquid are less likely to replace top managers and to carry out asset restructuring. We calculate liquidity as the ratio of current assets to current liabilities. All explanatory variables, except firm and industry performances, are measured as of the year prior to the year in which restructuring is taken (Year -1). Firm performance is measured as a change in operating performance from Year -1 to the year in which restructure (year 0), while industry performance is measured as of Year 0. We also control for fix effects by including year dummies and 19 industry dummies. These 19 industries are following the Standard Industrial Classification (SIC) codes. To conform to the SIC codes, we re-categorize the which are grouped under the classification of the Stock Exchange of Thailand. It is likely that would make decisions on taking all kinds of restructuring actions simultaneously. Hence, each action may affect the likelihood that others would occur, and vice versa. To control for the problems, we use the multivariate probit model. Specifically, we estimate a system of six regressions representing each type of the restructuring actions. 14
Table 5 presents the results of the six probit regressions that characterize the relationship between the top 30 business groups and probabilities of the six types of responses. The estimated coefficients of the business group dummy are positive as expected in all models, except equation (5) where the dependent variable represents the probability of debt restructuring. However, the coefficients are statistically significant in equations (2), (3), and (4). The results indicate that the top 30 business groups are more likely to engage in expansion, top executive replacement, and dividend cuts than non business group. There is a significant and negative relation between debt restructuring and the business group dummy (equation (5)). An affiliation with the top 30 business group appears to reduce the probability of engaging in debt restructuring. A plausible reason for this evidence might be something to do with the maturity structure of their debt. As shown by Charumilind, Kali, and Wiwattanakantang (forthcoming), due to close connections with banks, these business group have more long term debt compared to non group in 1996. Therefore, the business group may not have the need to restructure their debt structure. Among the firm characteristics, firm size appears to have quite strong influence on the likelihood of restructuring. Firm size is positively related to the probability of all restructuring actions, and strongly significant at the 1% level in all models except model (4) in which the dependent variable is the probability of dividend cuts. Our results also indicate that leverage is negatively and significantly related to the likelihood of expansion and dividend cuts. Consistent with the US evidence shown in Ofek (1993), with a high level of debt are more likely to engage in debt restructuring. Firm performance has a significant and negative effect on the probability of dividend cuts and capital raising, and a significant and negative effect on the probability of expansion. The results imply that that experience a performance decline are likely to engage in dividend cuts and capital raising, but less likely to take expansionary actions. Consistent with Shleifer and Vishny (1992), our results indicate that are less inclined to sell assets and engage in debt restructuring if their industry condition is poor. However, the dividend cut likelihood is significantly high if their industry condition is poor. Finally, the results show that greater liquidity decreases the probability that debt restructuring and capital raising will be taken. 15
< Table 5 here > 3.4 Restructuring of business groups: Propping or shareholder alignment? This section evaluates incentives to restructure in response to the crisis for business group. We investigate whether restructuring actions are implemented to improve a firm s value, and the benefits will be for both controlling shareholders and minority shareholders. To test this issue, we follow Friedman, Johnson, and Mitton (2003) who argue that the propensity to tunnel and prop is likely to be higher for that are organized in pyramids. Accordingly, we expect that pyramid are more likely to restructure more often than other. Otherwise, the ownership structure should have no effects on the restructuring likelihood. We use the ratio of cash flow to voting rights (CFVR) to measure how far pyramidal structures are used. As suggested by Claessens, Djankov, Fan, and Lang (2002), and Lins (2003), this variable indicates the degree of minority shareholders expropriation. The low ratio represents a high divergence between ownership and control by the ultimate shareholder, and hence high incentives to prop and tunnel. We also include the voting rights (VR) held by the controlling shareholder to control for its effects on the likelihood of restructuring. However, we do not have a clear prediction for the relationship between these two variables. As suggested by the literature, the voting rights may be positively or negatively related to the agency problems. On one hand, higher voting rights may enable the controlling shareholder to become more entrenched since he cannot be ousted (Morck, Shelifer, and, Vishny (1988) and Stluz (1988)). On the other hand, the controlling shareholder s entrenchment is less acute if he holds the cash flow rights in the same proportion to the voting rights because he would also bear a large expropriation cost. We also include a dummy variable, CS-manager, set to one if the controlling shareholder and his family are involved in the top management. Large shareholders who are involved in managing the are prone to be entrenched from holding dominant influence over corporate policy and being able to take actions for their own interests which may not be aligned with those of minority shareholders (Mitton (2002)). Lai and Sudarsanam (1997) find 16
that manager-owner dominated in U.K. are more likely to implement operational restructuring and acquisitions. Volpin (2002) finds that in Italy top executive turnover of the where the controlling shareholders serve as top executives is less sensitive to performance. He contends that this is because the controlling shareholders are entrenched against the interest of other shareholders in order to preserve the opportunity to extract benefits. Accordingly, we predict that propping incentives and the presence of the controlling shareholder as top executives are positively related. Table 6 shows the results of multivariate probit regressions that characterize the effects of business group affiliations, voting rights, the level of ownership and control disparity (i.e., the use of pyramids), owner-manager dominance, and debt, on the probability of restructuring activities. Similar to the regressions in Table 5, we also control for other firm characteristics, the industry and year effects. The coefficients of the business group dummy are positive and significant in regressions (2) and (3) at the 10% level. The results indicate that compared to non business group, business group are more likely to conduct expansion and replace top managers. While the estimated coefficients on the ratio of cash flow rights to voting rights (CFVR) are not statistically significant in all models, the coefficients of the interaction terms between the business group dummy and the ratio of cash flow to voting rights (CFVR) are negative and significant in regressions (1), (2), and (3) at the 10%, 5%, and 10%, respectively. The results suggest that business group that are controlled via pyramids and hence have higher agency costs (or a lower ratio of cash flow to voting rights) are more likely to undertake operational actions namely, asset downsizing, expansion, and top management replacement than other. Friedman, Johnson, and Mitton (2003) argue that debt commits the controlling shareholder to bail out. They also contend that the propensity to prop may be higher in particular for interconnected group. So, these two arguments together suggest that business groups with higher debt might have more incentive to prop, and hence are more active in engaging in restructuring activities. To test these hypotheses, we create two variables. First, to measure the effects of debt of the business group on the restructuring likelihood, we interact the business group dummy and the debt ratio. Second, to capture the effects of 17
debt of the pyramidal group on the restructuring likelihood, we interact the business group dummy, the debt ratio, and the ratio of cash flow to voting rights. The results show that the estimated coefficients on the interaction term between the business group dummy and the debt ratio are significant at the 10% level only in regression (4) where the dependent variable represents the probability of cutting dividends. The striking result in regression (4) is that while the coefficients on the debt ratio is negative and significant at the 1% level, the coefficients on the interaction term between the business group dummy are positive. These results suggest that high leverage are less inclined to cut dividends. In contrast, business group with high debt are more likely to engage in dividend cuts. The debt influence on restructuring likelihood is more pronounced in pyramid belonging to the business groups. The coefficients of the interaction term between the business group dummy, the debt ratio, and the ratio of cash flow to voting rights are significant and positive in regressions (1), (2), and (3) where the dependent variables represent the probability of asset downsizing, expansion, and top management replacement, respectively. We have noted before that business groups with a higher ratio of cash flow to voting rights (lower agency costs) are less likely to engage in all three operational restructuring actions. However, debt appears to increase the probability of operational actions to be taken by business group that have use less pyramid (with high ratio of cash flow to voting rights). The controlling shareholder dominance in top management (CS-manager) has both positive and negative effects on the restructuring likelihood. The estimated coefficient on the owner-manager dummy is negative and highly significant at the 1% level in regression (3). The results indicate that controlling shareholder-manager dominated appear to be less likely to sack their top management, which are in line with the finding of Denis, Denis, and Sarin (1997), Lai and Sudarsanam (1997), and Volpin (2002). However, the estimated coefficient on CS-manager is positive and significant at the 5% level in regression (5). This result suggests that controlling shareholder-manager dominated are more likely to take debt restructuring actions. The coefficients of the interaction term between the business group dummy and on 18
CS-manager are positive and significant at the 10% level in regression (4), where the dependent variables represent the probability of dividend cuts. This result implies that controlling shareholder-manager dominated group are more likely to undertake dividend cuts. Overall, our results support the argument of Friedman, Johnson, and Mitton (2003) that higher debt increases the probability to prop by providing more commitment for the controlling shareholders to bail out business group with low agency costs. < Table 6 here > 4. Summary and Conclusion In this paper, we investigate the effects of the presence of controlling shareholders in family that are prevalent world wide. Our focus is in emerging economies in which legal and regulatory frameworks are weak. In this environment, many scholars argue that controlling shareholders are likely to expropriate corporate assets. We investigate how controlling shareholders respond to an economic crisis. Friedman, Johnson, and Mitton (2003) argue that when returns on investment are temporarily low, controlling shareholders are likely to prop up the so that they can tunnel corporate resources when the get back in good shape. This is, however, under the assumption that propping and tunneling are symmetric. We use the data from Thailand to study this issue. Thailand provides a natural research setting because it was hit by the East Asian financial crisis in July 1997. We classify our sample into two groups: affiliations of the top 30 business groups and non business group. Our results show that business group are as vulnerable to the crisis as non group. However, group are more likely to undertake some kinds of restructuring actions namely expansion and replace top executives to cope with the economic downturn. We also find that group that are organized via pyramids are more likely to take operational restructuring actions namely asset downsizing, expansion, and management turnover. Our results are also consistent with the argument of Friedman, Johnson, and Mitton (2003) that debt increases the controlling shareholders incentives to prop. 19
References Advance Research Group. 1995, Thailand Company Information, Bangkok: Business and Management Publications. Anderson, R., Reeb, D., 2003. Founding-family ownership and firm performance: Evidence from the S&P 500. Journal of Finance 58, 1301-1327. Attig, N., Fischer, K., Gadhoum, Y., 2003. On the Determinants of Pyramidal Ownership: Evidence on Dilution of Minority Interests. Unpublished working paper. Bae, K.H., Kang, J.K., Kim, J.M., 2002. Tunneling or value added? Evidence from Korean business groups. Journal of Finance 57, 2695-2740. Baek, J.S., Kang, J.K., Park, K.S., 2002. Economic shock, business group, and determinants of firm value and restructuring: Evidence from the Korean financial crisis. Mimeo. Baek, J.S., Kang, J.K., Park, K.S., 2004. Corporate governance and firm value: Evidence from the Korean financial crisis, Journal of Financial Economics 71, 265-313. Bebchuk, L., Kraakman, R., Triantis, G., 2000. Stock pyramids, cross-ownership, and dual class equity: The creation and agency costs of separating control from cash flow rights. In: Morck, R. (Ed.), Concentrated Corporate Ownership. Chicago: University of Chicago Press. Bertrand, M., Mehta, P., Mullainathan, S., 2002. Ferreting out tunneling: An application to Indian business groups. The Quarterly Journal of Economic 117, 121-148. Burkart, M., Panunzi, F., Shleifer, A., 2003. Family. Journal of Finance 58, 2167-2202. Charumilind, C., Kali, R., Wiwattanakantang, Y., 2003. Connected lending: Thailand before the financial crisis. Journal of Business, forthcoming. Claessens, S., Djankov, S., Fan, J.P.H., Lang, L., 2002. Disentangling the incentive and entrenchment effects of large shareholdings. Journal of Finance 57, 2741-2711. Claessens, S., Djankov, S., Fan, J.P.H., Lang, L., 2003. The benefits and costs of internal markets: Evidence from East Asia s financial crisis. In: Fan, J.P.H., Hanazaki, M., Teranishi, J. (Eds.), Designing Financial Systems in East Asia and Japan Toward a Twenty-First Century Paradigm. London: Routledge. Claessens, S., Djankov, S., Lang, L., 2000. The separation of ownership and control in East Asian corporations. Journal of Financial Economics 58, 81-112. Claessens, S., Fan, J.P.H., 2002. Corporate governance in Asia: A survey. International Review of Finance 3, 71-103. 20
Denis, D., Denis, D., Sarin, A., 1997a. Ownership structure and top executive turnover. Journal of Financial Economics 45, 193-221. Denis, D., Kruse, T., 2000. Managerial discipline and corporate restructuring following performance declines. Journal of Financial Economics 55, 391-424. Faccio, M., Lang, L., 2002. The ultimate ownership of Western European corporations. Journal of Financial Economics 65, 365-395. Friedman, E., Johnson, S., Mitton, T., 2003. Propping and tunneling. Unpublished working paper, NBER. Jensen, M., 1989. Eclipse of the public corporation. Harvard Business Review 5, 61-74. Joh, S.W., 2003. Corporate governance and firm profitability: Evidence from Korea before the economic crisis. Journal of Financial Economics 68, 287-322. Johnson, S., Boone, P., Breach, A., Friedman, E., 2000. Corporate governance in the East Asian financial crisis. Journal of Financial Economics 58, 141-86. Johnson, S., La Porta, R., Lopez-de-Silanes, L., Shleifer, A., 2000a. Tunneling. American Economic Review (Papers and Proceedings) 90, 2, 22-27 Johnstone, W., Neilsen, D. and Henderson, A. (2001). Thai business groups 2001: A unique guide to who owns what. Bangkok: The Brooker Group Public Company Limited. Kang, J.K., Shivdasani, A., 1997, Corporate restructuring during performance declines in Japan. Journal of Financial Economics 46, 29-65. Khanna, T., Palepu, K., 2000. Is group affiliation profitable in emerging markets? An analysis of diversified Indian business groups. Journal of Finance 55, 867-891. Khanthavit, A., Pholsiri, P., Wiwattanakantang, Y., 2003. Did families lose or gain control? Thai after the East Asian financial crisis. In: Fan, J.P.H., Hanazaki, M., Teranishi, J. (Eds.), Designing Financial Systems in East Asia and Japan Toward a Twenty-First Century Paradigm. London: Routledge. Laeven, L., 2001. Insider lending and bank ownership: The case of Russia. Journal of Comparative Economics 29, 207-229. Lai, J., Sudarsanam, S., 1997. Corporate restructuring in Response to performance decline: Impact of ownership, governance and lenders. European Finance Review 1, 197-223. 21
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1997. Legal determinants of external finance. Journal of Finance 52, 1131-1150. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1998. Law and Finance. Journal of Political Economy 106, 1113-1155. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 1999. Corporate ownership around the world. Journal of Finance 54, 471-517. La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2000. Investor protection and corporate governance. Journal of Financial Economics 58, 3-27. La Porta, R., Lopez-de-Silanes, F., Zamarripa, G., 2003. Related lending. Quarterly Journal of Economics 118, 231-268. Lemmon, M., Lins, K., 2003. Ownership structure, corporate governance, and firm value: Evidence from the East Asian financial crisis. Journal of Finance 58, 1445-1468. Lins, K., 2003. Equity ownership and firm value in emerging markets. Journal of Financial and Quantitative Analysis 38, 159-184. Lins, K., Servaes, H., 2002. Is corporate diversification beneficial in emerging markets?. Financial Management Summer, 5-31. Mitton, T., 2002. A cross-firm analysis of the impact of corporate governance on the East Asian financial crisis. Journal of Financial Economics 64, 215-241. Morck, R., Shleifer, A., Vishny, R., 1988. Management ownership and market valuation: An empirical analysis. Journal of Financial Economics 20, 293-315. Morck, R., Shleifer, A., Vishny, R., 1989. Alternative mechanisms for corporate control. American Economic Review 79, 842-852. Morck, R., Yeung, B., 2003. Agency Problems in Large Family Business Groups. Entrepreneurship Theory and Practice 27, 367-382. Ofek, E, 1993. Capital structure and firm response to poor performance. Journal of Financial Economics 34, 3-30. Pornkulwat, S., 1996. The development of business strategies of overseas Chinese: A case study of two families. Unpublished Master thesis, Thammasat University, Thailand (in Thai). Sappaiboon, T., 2000a. The Lamsam Family. Bangkok: Nation Multi Media Group (in Thai). 22