Corporate Governance and Diversification*

Transcription

1 Corporate Governance and Diversification* Kimberly C. Gleason Dept of Finance Florida Atlantic University Inho Kim Dept of Finance University of Cincinnati Yong H. Kim Dept of Finance University of Cincinnati Young Sang Kim** Haile/US Bank College of Business Northern Kentucky University This version: Nov 10, 2011 JEL classification: G34, G32 Keywords: Corporate Governance; Mergers and Acquisitions; Diversification Discount * We thank seminar participants at the Bowling Green State University, Florida Atlantic University, Northern Kentucky University, the 2006 Financial Management Association meetings, 2006 Global Finance Conference meetings, the Sixth International Conference on Asia-Pacific Financial Markets, 2011 for helpful comments to improve the paper. We gratefully acknowlege the best paper award in capital market at the 13th annual meeting of Global Finance Conference, Brazil, ** Correspondence Address: Young Sang Kim, Associate Professor of Finance, Haile/US Bank College of Business, Northern Kentucky University, Highland Heights, KY 41099, Tel: (859) , Fax: (859) , kimy1@nku.edu

2 Corporate Governance and Diversification Abstract In this paper, we investigate the relation between corporate governance and returns to bidders and targets, using 1,640 observations of completed acquisitions from 1996 to We find that the cumulative abnormal returns for acquirers are significantly negative upon announcement of acquisitions for the full sample and for the related and diversifying sub-samples. However, we find that diversifying acquisitions, when conducted by firms with a higher percentage of outsiders on the board, improve returns. Furthermore, we separately examine high-tech and non high-tech firms to test the relation between board characteristics and announcement returns in different information asymmetry environments. We also find that diversifying acquirers with independent boards perform better than those with insider dominated boards and the results are especially pronounced for high-tech firms. Taken together, the results suggest that firms with better incentive alignment will be more likely to be perceived by the market as stronger performers in acquisitions. In sum, we find that corporate governance plays an important role in determining wealth creation for our sample of acquiring firms. JEL classification: G34, G32 Keywords: Corporate Governance; Mergers and Acquisitions; Diversification Discount 2

3 Corporate Governance and Diversification I. Introduction Following corporate governance scandals involving numerous U.S. firms, regulators took action to reduce the scope for corporate misbehavior. The Sarbanes-Oxley Act of 2002 imposed new restrictions on the structure of the board of directors, in the hope that independent outside directors would be more impartial in assessing managerial decisions than inside directors, who may have conflicting interests and hence, loathe disciplining top level managers. The Sarbanes-Oxley Act requires that boards have audit committees that consist only of independent outside directors, and encourages firms to adopt smaller boards, which may provide better (or more careful) oversight than large boards. The role of board is to monitor and advise managers on important decisions, such as mergers and acquisitions. If the board is structured effectively so that sufficient monitoring takes place, proposals by managers that may lead to personal perquisite consumption or entrenchment, such as diversifying acquisitions, will be voted down. Alternatively, the proposal may not be initiated in the first place if it is clear that the board will provide disciplinary action when confronted with agency behavior by managers. To examine the effectiveness of the board as a monitor of the firm, many studies have investigated top management turnover, which is attributed to actions taken by the monitors of the firm, such as boards, and the market for corporate control (see, e.g., Mikkelson and Partch, 1997; Huson et al., 2004; Goyal and Park, 2002). Bhagat et al. (2006) indicate that share ownership by directors impacts the ability of the board to 3

4 discipline management. While effective corporate governance provides a higher ex-ante threat of dismissal, using top management turnover detects only those firms where the threat is ex-post exercised. Thus, it is necessary to investigate the monitoring mechanisms providing ex-ante impact on crucial investment decision. The examination of the relation between corporate governance and mergers and acquisitions provides important insights of ex-ante monitoring role. In addition, it is important to incorporate the impact of a broader set of governance quality indicators such as percentage of independent directors, board size, percentage of director shares, CEO/Chair duality, and institutional block holdings when we investigate the relation between mergers and acquisitions and corporate governance. In this paper, we empirically investigate whether board structure and outside blockholdings affect acquisition announcement returns, using 1,640 observations of completed acquisitions from 1996 to A large body of literature provides insights into wealth destruction and diversification 1, but there is no clear consensus regarding the relationship between governance and the performance of diversification. We document a link between corporate governance and diversification, while diversifying acquisitions are value-destructive for the sample as a whole, the negative impact is modified when firms have strong independent boards 2 and effective governance monitors such as outside blockholders, so that managerial incentives are aligned with shareholders interests. Thus, our results imply that high quality governance firms make appropriate diversification decisions, in support of the monitoring hypothesis, which indicates that managers of firms 1 Several studies find negative announcement returns for diversifying takeovers (see, e.g., Morck, Shleifer, and Vishny, 1990; Moeller, Schlingemann, and Stulz, 2005). 2 We define strong independent as a board comprised of at least 50% of outsiders. Also we use the average value of the percentage of independent board in our sample period as an alternative definition for strong board. 4

5 that have independent board and outside blockholdings are better monitored and hence, make value enhancing investment decisions. Our study contributes to the literature in four ways. First, we provide additional evidence that diversifying acquisitions are viewed by shareholders as value destructive, though a high percentage of independent directors on diversifying firms boards mitigate the negative abnormal returns. This suggests that for firms with strong boards, diversification proposals that are not in the shareholders interest may not be initiated by managers or may not be approved by the board. This result may shed light on the puzzle of the empirically observed negative relationship between board composition and firm value. The effective board can reduce potential agency conflicts by prohibiting value destructive proposals in an earlier stage (ex-ante). Our results imply that internal controls provided by boards add value when it comes to mergers and acquisitions. Secondly, by designing our empirical test using event study method with announcement returns as the proxy for firm performance, we reduce the potential endogenous problems associated with typical corporate governance studies that analyze firm value differences. Our results indicate that corporate governance plays an important role in the shareholders perception of the valuation effects of diversifying mergers and acquisitions. Consistent with earlier studies (e.g., Morck et al., 1990; Lang and Stulz, 1994; Berger and Ofek, 1995; Moeller et al., 2004; and Hoechle et al., 2011) we show that diversifying mergers and acquisitions reduce firm value as measured by cumulative abnormal returns upon announcement of the acquisition bid. However, in regression specifications with an interaction variable between diversification and internal controls, we show that diversifying firms with higher board independence experience less of a 5

6 wealth decline after controlling other factors: CEO age and duality, deal characteristics (i.e., multiple bids, method of payment, deal size, target organizational form), institutional ownership, and firm characteristics (i.e., size, leverage, and growth opportunities). This result provides evidence of a positive relationship between properly structured boards and firm value, and is consistent with other studies which indicate empirically that outside directors have reputation and human capital incentives to maximize shareholder wealth (Coles and Hoi, 2003; Brickley et al., 1999; Harford, 2003). As a result, outsider dominated boards lead to better performance (Hermalin and Weisbach, 1991). Third, we examine high-tech firms and non high-tech firms separately. Denis and Sarin (1999) and Lasfer (2004) segment their samples into high and low growth firms, and find significant differences in the relation between board structure and performance, and attribute these differences to a necessity of stronger monitoring of firms with high information asymmetry (such as technology firms). Accordingly, we observe that for firms in non high-technology industries, there is a weak relationship between the percentage of the board comprised of independent directors and announcement returns, while for firms in high-technology industries, there is a strong relationship between independent dominated boards and firm value. Raheja (2005) argues that for firms that are more difficult to verify, such as technology firms, private benefits are lower, and there is less of a need for outsiders. For firms that are easier to verify, private benefits increase, and more outsiders are optimal. Our results provide further empirical evidence on this issue. Consistent with Raheja (2005), we find that insiders play an important role for high-tech firms engaging in related acquisitions, and independent directors do not add value. For diversifying 6

7 acquisitions in high-technology industries, however, independent directors play an important role and add value. Fourth, to our knowledge, our paper is the most comprehensive study thus far to analyze the relation between various aspects of corporate governance and diversifying acquisitions. Earlier studies are limited in terms of sample size and scope in examining this issue. 3 Notably, a recent study by Masulis et al. (2006) examines the impact of firm s anti-takeover provisions and the shareholder wealth effects of its acquisitions. They find that firms with higher anti-takeover provisions are more likely to make value destructive diversifying acquisitions. This result complements our findings of a relation between corporate governance and firm value. The rest of the paper is organized as follows. The testable hypotheses regarding the relation between corporate governance and diversification are presented in Section II. Section III describes the sample selection. Section IV provides evidence concerning the characteristics of firms engaging in value enhancing versus value destructive corporate diversification. Section V provides empirical tests of our hypotheses. Section VI summarizes the conclusions. II. Corporate Governance and Performance Several studies point towards a paradoxical insignificant or negative relationship between corporate governance quality, as proxied by the percentage of outside directors on the board, and firm value. Morck et al. (1988), Hermalin and Weisbach (1991) and Bhagat and Black (2001) find no significant relationship between board independence and firm value (using Tobin s q as a proxy for firm value). Coles et al. (2006) examine 3 For example, Bradley, Desai, and Kim (1988), Byrd and Hickman (1992), Bhagat and Black (2001) and others examine with small sample size. 7

8 the relationship between board structure and firm value, and find that one board size or composition does not provide the same monitoring benefits for all firms. Furthermore, Hermalin and Weisbach (1991), Mehran (1995), Klein (2000), and Bhagat and Black (2001) find an insignificant relationship between board independence and accounting performance. Agrawal and Knoeber (1996) find that Tobin s q decreases with an increase in the proportion of outside directors. Thus, the evidence regarding the merits of independent boards is inconclusive. We attempt to shed more light on this issue by examining the relationship between corporate governance and acquisition returns to bidders and targets. II.1. Monitoring and Scope for Value Destruction The scope for value destruction of mergers and acquisitions is well documented in the finance literature. Roll (1986) asserts that hubris may play a role in value destructive acquisitions; managers may overestimate synergies generated by the merger and hence, overvalue takeover targets. Jensen (1986) argues that managers have an inherent conflict with shareholders due to the separation of ownership and control, and when they have large stocks of free cash flows, they tend to make value destructive managerial decisions. Thus, it can be anticipated that in the absence of proper internal controls, managers will make decisions that benefit themselves at the expense of shareholders. We demonstrate that internal controls via independent boards and outside blockholdings are alterative mechanisms for disciplining managerial decision regarding mergers and acquisitions. Our paper attempts to evaluate these two theoretical arguments of hubris and free cash flow, and to empirically demonstrate their interactions with various monitoring mechanisms of corporate governance. 8

9 II.2. The Diversification Discount and Wealth Losses from Diversifying Acquisitions Several studies have revealed that diversified firms suffer from a diversification discount (e.g., Berger and Ofek, 1995; Lang and Stulz, 1994; Servaes, 1996; and Denis et al., 1997) where a conglomerate firm is worth less than the sum of imputed value of its segment components, measured by multipliers of the comparable industry focused firms. Comment and Jarrell (1995) and Berger and Ofek (1999) find that firms that begin refocusing programs are able to improve value by reversing costly prior diversification. Lins and Servaes (1999) find evidence of a diversification discount in the U.K. and Japan, while German firms do not exhibit a diversification discount, arguably due to high inside ownership in Germany that aligns incentives. In essence, these studies show that diversifying acquisitions destroy shareholder wealth, and hence, are met with disapproval from the market. Specifically, Berger and Ofek (1999) show that abnormal returns to bidding firms diversifying into new product markets are significantly negative. However, there is some debate over whether diversification is value destructive. Bodnar et al. (1999) find that the losses due to product market diversification are offset somewhat by positive gains from geographic diversification, and argue that the losses to diversification have been overstated. Graham et al. (2002) find that a large percentage of excess value occurs because firms acquire already discounted business units, not because diversification is value destructive per se. Mansi and Reeb (2002) find that diversified firms are less risky than non-diversified firms, and that losses to diversified firm shareholders are offset by gains to diversified firm bondholders. Also, utilizing Heckman s (1979) two-stage estimator to control for the propensity to diversify, Villalonga (2004) reports no diversification discount in her sample. Whited (2001) argues 9

10 that the diversification discount documented in earlier studies is misleading, due to a miscalculation of Tobin s q. Finally, Campa and Kedia (2002) argue that firms selfselection explains the diversification discount. Thus, the debate over the relationship between diversification and firm value is ongoing. However, no studies to date have examined whether it may be the case that firms that diversify have poorly structured boards which drive this result. Anderson et al. (2000) examine the structure of corporate governance in diversified firms. However, they fail to find evidence that governance characteristics explain the value discount from diversification. 4 In this paper, we contribute to the literature by demonstrating that proper governance may reverse the losses observed in prior studies from diversification activities. II.3. Corporate Governance and Acquisitions The role of the board of directors is to ensure that managers make shareholder value maximizing decisions, and to discipline managers if they fail to do so. Agency theory and the free cash flow hypothesis indicate that managers have incentives to consume perquisites at the expense of shareholders (Jensen and Meckling, 1976). Mergers and acquisitions are important managerial decisions that require vetting from the shareholders and the board of directors, and the potential for value destructive decision making is great. Moeller et al. (2004) find that conglomerate firms destroyed on average $25.2 million of wealth per year upon announcement between 1980 and 2001 and a total loss of $240 billion from 1998 to In a recent paper by Hoechle et al. (2011) examine the relation between diversification discount and corporate governance. They find the discount disappears entirely after controlling for governance variables in dynamic panel GMM regression. This finding is also consistent with our results. 10

11 Several studies examine the link between corporate governance, decision making regarding acquisitions, and post-acquisition performance. Paul (2006) examines that firms with independent boards are more likely to withdraw acquisition bids, following a negative market response. She attributes this to independent boards being more willing to facilitate the correction of managerial mistakes. Masulis et al. (2006) find that firms with anti-takeover provisions in place are more likely to make diversifying acquisitions. They attribute this to the anti-takeover provision value destruction- hypothesis, which posits that as the disciplinary capabilities of the market for corporate control are removed, managers have more scope for engaging in wealth destruction, because they do not perceive their human capital to be at risk. Masulis et al. (2006) find negative abnormal returns upon announcement of diversifying acquisitions, and note that this effect is particularly strong for firms with weak shareholder rights. These results suggest that firms that are not properly governed may be more likely to make value destructive decisions. Byrd and Hickman (1992) provide evidence regarding the relationship between board of director characteristics and bidder returns. They find that bidders with independent boards (where independent is defined as a board comprised of at least 50% of outsiders) experience higher abnormal returns than those that do not have independent boards. 5 Li and Srinivasan (2011) compare board with founder-director firms and nonfounder firms on their CEO compensation, CEO retention policies, and M&A decisions. Board with founder-director firms is more independent, less agency problems and more pay-for-performance sensitive than board with nonfounder firms. They report 5 However, Byrd and Hickman (1992) find no evidence that board independence is relevant when it is defined as the percent of the total board comprised of outside directors, a measure traditionally used in the literature. 11

12 that founder-director firms represent higher three-day M&A announcement returns than nonfounder firms. In another empirical analysis, Malmendier and Tate (2005) find support for the Jensen (2003) view, that overconfident managers are more likely to undertake acquisitions, and that the negative market reaction is significantly stronger for overconfident managers as opposed to rational managers. II.4. Board Structure Prior research on the effectiveness of board indicates that certain board qualities enhance board performance. Yermack (1996) finds that as board size increases, the ability to monitor the firm declines, and argues that smaller boards are more effective at monitoring than larger boards. However, Coles et al. (2006) find that not all firms are best served by small boards. Their research indicates that for high-tech firms (i.e., R&D intensive firms), a larger fraction of insiders on the board leads to better performance. We investigate whether the board size and board structure affects the share price response to acquisitions differently for high-tech firms and non high-tech firms. High-tech firms, in general, have a substantial amount of information asymmetry, and thus the verification costs by boards are higher. Thus, we expect that the effect of board composition on announcement returns would differ for high-tech and non high-tech firms. In a similar vein, Cicero et al. (2011) show that firms tend to change the structure of board by changes in underlying firm characteristics. II.5. Blockholders Impact on Gains from Diversification Several studies assert that blockholders - outside entities that hold five percent or more of the firm s stock - provide a significant monitoring role, while others argue that 12

13 blockholders are able to extract private benefits of control. Shleifer and Vishny (1989) argue that large blockholders have an incentive to monitor management and the power to influence the board and managers. Other studies indicate that pay per performance sensitivity is affected by the presence of large blockholders (Mehran, 1995; Gillan et al., 2003). Further evidence on the disciplinary mechanism of blockholders is provided by Denis and Serano (1996), who find that outside blockholders are likely to take the initiative to fire poorly performing managers. Even if blockholders do not directly discipline management, they are able to vote with their feet and provide a credible threat to management by selling their shares (Parrino et al., 2003). Evidence from the acquisitions literature also suggests that stock transactions are value enhancing because they introduce a new blockholder group which will provide additional monitoring (Chang, 1998). Thus, we anticipate finding that the percent of outside blockholdings positively affects managerial decision making, and hence, diversifying acquisitions of firms with high external blockholdings should be value enhancing, as blockholders will be more likely to take an activist role in governing the firm (and overseeing acquisition decisions). However, other research (i.e., Barclay and Holderness, 1989) indicates that blockholders may have a better ability to obtain private benefits from managerial decisions. Further, the evidence suggests that some blockholders provide better monitoring than others; pressure-insensitive blockholders lead to improved operating performance (Saunders et al., 2003). We incorporate pressure-insensitive institutional investors using a variable that measures pension plan holdings. Our paper is the first to examine the relationship between outside blockholdings and the efficacy of diversification decisions, as well as the impact of pressure insensitive versus pressure sensitive investors. 13

14 II.6. Technology Consider a small entrepreneurial venture in the biotechnology industry. It is often the case that scientists who work for large companies (such as the pharmaceutical sector, in drug development) decide to leave and pursue independent research, and so they create start-up firms together. Having worked together closely for a long time on a certain, specialized kind of technology, they form a management team knowledgeable in the technology, so that they can pursue product development. However, the technology of the venture is not widely understood by the public; when the firm goes public and structures its board, it may be beneficial to put people on the board who also understand that specific technology, and these are often people who worked for the large pharmaceutical firm. In that case, where technology is specialized and few people have these specialized skills, it is often useful to have a higher percentage of directors who are insiders. II.7. Control variables The literature finds that abnormal returns are lower for acquisitions by firms with low leverage (Maloney et al., 1993), low Tobin s q (Lang et al., 1989; Servaes, 1996), low managerial ownership (Lewellen et al., 1985), and large capitalization (Moeller et al., 2004). In addition, the existence of competing bids (Bradley et al., 1988) and deal size relative to bidder size are relevant determinants of abnormal returns (Asquith et al., 1983). Recently, Bradley and Sundaram (2005) argue that while acquirer size does matter larger acquirers realize greater losses the effects of the medium of exchange, the organizational type of target, and the relative size of the target are also relevant. Finally, Chang (1998) finds that while returns to shareholders of acquisitions of publicly traded companies are negative, abnormal returns to bidders of privately held firms are 14

15 significantly positive when they are paid for with stock. We incorporate these control variables in our analysis. III. Data We investigate a sample of acquisitions constructed from the Securities Data Company's (SDC) U.S. Mergers and Acquisitions Database. We obtain the initial sample of 5,429 acquisitions from SDC from year 1996 to In addition, the sample meets the following criteria: (1) The announcement date is in the 1996 to 2003; (2) The acquirer controls less than 50% of the shares of the target at the announcement date and obtains 100% of the target shares; (3) The deal value is equal to or greater than $1 million; 6 (4) Data on acquirer stock prices and accounting variables are available from CRSP and Compustat Research Insight, respectively. To investigate the effectiveness of corporate governance and internal control issues of impacting managerial decision making, we obtain the following data from Investor Responsibility Research Center (IRRC): board size, board composition (percent of insiders versus outsiders), CEO duality, CEO age, and director share ownership. The IRRC database provides details on the structure and practices of the boards of directors at a large number of U.S. companies from 1996 to Furthermore, we obtain institutional blockholdings and pension fund holdings data from Cremers and Nair (2005). 7 Blockholdings are defined as the percentage shareholding by largest institutional blockholder, which owns at least 5% of the firm s outstanding shares. Pension fund holdings are defined as the percentage of shares held by 6 The deal value corresponds to 10%, 5%, and 1% of the market value of the assets of the acquirer (defined as the book value of assets minus the book value of equity plus the market value of equity). We report results for the 1% threshold but our conclusions hold for the more restrictive samples. 7 We thank Cremers and Nair for providing their institutional ownership data. 15

16 the 18 largest public pension funds, which are generally more distant from conflicts of interest and corporate pressure than other institutional shareholders (i.e., pressure insensitive blockholders). Accounting data is collected from Compustat Research Insight. After all of the above restrictions and requirements, we finally obtain 1,640 mergers and acquisitions from 1996 to IV. Sample Characteristics Table 1 provides descriptive characteristics of sample deals. Panel A shows the sample breakdown by year, indicating the deal size and method of payment for the sample transactions. Panel B parses the sample into related transactions and diversifying transactions, and provides cumulative abnormal returns, percent of independent directors, and board size by year. [Table 1 About Here] Table 1, Panel A indicates that the mean (median) deal size of the sample period was $1, ($185.54) million. The year with the largest deals was 1999, with a mean (median) size of $1, (260.52) million. The year with the smallest average transaction size was 1997, with a mean (median) value of $ (171.20) million. About 29.5% of the deals were financed with cash only; 47.01% were financed with stock only, with the remaining being financed with a combination of considerations. The most frequent year for stock deals was 1997, with 57.89% of transactions financed with stock. Stock transactions declined substantially following the collapse of the internet bubble in 2001, to only 18.42% of deals in Cash payment was most frequent in 2003, when 51.31% were financed with cash. 16

17 Table 1, Panel B parses the data by related and diversifying acquisitions transaction. We use the Fama-French (FF) 49 industrial categorization to identify the relatedness of the acquisition. 8 We define related acquisitions as those where the bidder and target are in the same FF49 industry, while diversifying acquisitions occur when a firm acquires a target in a different FF49 industry. The results indicate that for the 1,007 related transactions, the mean (median) deal size was $1, ($217.80) million. The 633 diversifying deals tended to be smaller on average, with a mean (median) of $ ($150.00) million. Table 1, Panel B also indicates that the average percentage of independent directors is higher for diversifying transactions than for related transactions, although the median (66.67%) is the same for both. Mean board size for related transactions (10.96) exceeds that of diversifying transactions (9.87), though the medians are the same (10.0). Finally, Panel B provides cumulative abnormal returns by year for diversifying and related transactions. We calculate the cumulative abnormal returns (CAR) over the three-day window (-1, +1) surrounding the mergers and acquisitions announcement date. We use standard event study methodology of Brown and Warner (1985). The parameters for the market model are estimated over 200 days before the event date [-210, -11]. 9 We use the CRSP equally weighted market index and the significance levels are computed using time series and cross-sectional variation of abnormal returns. Mean (median) CARs for the [-1,+1] window for related transactions are % (-0.671%). For diversifying transactions, mean (median) cumulative abnormal returns 8 We also examine 2 digit SIC code to identify the relatedness, and the results (not reported) are similar to our findings in this paper. We obtain FF49 industry from Kenneth French website, 9 We also use various estimation periods such as 120 days, and 150 days. The results are qualitatively similar to the results reported in this paper. 17

18 appear smaller, but less negative than for related transactions, a finding inconsistent with previous literature. It can be explained by the issues of corporate governance and market for corporate control becoming more important in the sample period or other confounding factors need to be control for better explanation. For related transactions, CARs were highest in 1996 (0.258%) and lowest in 2000 (-3.424%) and for diversifying transactions, the highest CARs were in 1997 (0.818%) and lowest for 2001 (-1.972%). Table 2 provides summary statistics on the characteristics of sample firms. The mean (median) percent of independent directors was 63.33% (66.66%). The mean (median) number of independent directors was 6.80 (6.00). Mean (median) board size was (10). This is comparable with Coles et al. (2006). They find mean (median) board size is 10.4 (10) from for Execucomp firms. The percent of shares held by directors was a mean (median) of 8.629% (3.047%). The mean (median) CEO age is (54.0). CEOs of 69.57% of firms were also chairmen of their boards. Institutional pension fund holding is 2.59%, and blockholdings overall comprise a mean (median) of 6.631% (6.621%) of shares outstanding. Mean (median) deal size for related acquisitions was $ million ($ million). Mean (median) relative deal size is 8.58% (2.56%) of bidder s market value. About 1.83% of the transactions had multiple bidders. Private targets constituted about 36.3% of the transactions. Total assets of the mean (median) firm was $17,684 ($4,398.0) million. Mean (median) Tobin s q was 2.62 (1.66). Leverage, as proxied by long term debt to total assets, was 21.35%. Operating cash flow to total assets had a mean (median) of 7.30% (6.43%). [Table 2 About Here] 18

19 Table 3 provides univariate results for governance characteristics, deal characteristics, and firm characteristics by diversifying versus related acquisition. [Table 3 About Here] Table 3 indicates that the percent of independent directors for firms engaging in related diversification is insignificantly different from those engaging in diversifying acquisitions. Furthermore, both mean and median board size is significantly larger for firms announcing related transactions than for firms engaging in diversifying transactions. The percentage of shares owned by directors is significantly higher for firms engaging in related transactions than diversifying transactions. However, CEO age and duality do not differ for firms doing related versus diversifying transactions. Nor was pension fund shareholding for firms doing related transactions significantly greater than for diversifying firms. Surprisingly, diversifying firms had significantly greater blockholder participation than firms engaging in related acquisitions. However, these blockholders may be primarily pressure-sensitive, due to relationships with the firm, given that pension fund shareholding was not significantly different across the two groups. Table 3 also provides univariate tests of differences in deal characteristics. Deal size was significantly larger for related than diversifying transactions, and the relative size of related deals exceeds those of diversifying transactions. Significantly more related transactions were paid for with stock; significantly fewer were paid for with cash. Private targets were more prevalent in diversifying acquisitions than related acquisitions. Finally, we examine characteristics of firms engaging in related versus diversifying acquisitions. Firms doing related deals were significantly larger in terms of total assets. Firms doing related deals had a significantly lower mean (median) Tobin s q 19

20 than those doing diversifying transactions. Leverage was insignificantly different for firms announcing related versus diversifying transactions. The ratio of operating cash flows (OCF) to total assets was significantly larger for diversifying firms than for those doing related acquisitions. V. Empirical Results V.1. Univariate Analysis Moeller et al. (2005) find that diversifying acquisitions are value destructive, especially those conducted between 1998 and However, we argue that independent boards and outside blockholdings may improve the returns to bidders. We present event study results that investigate this issue in Table 4 (for bidder returns) and Table 5 (for value-weighted combined acquirer-target returns). [Table 4 About Here] Further, Table 4 indicates that mean and median bidder cumulative abnormal returns are significantly negative upon announcement of acquisitions for both the related (-1.227%) and diversifying (-0.790%) subsamples. However, only the mean overall abnormal returns for each category are significantly negative; and the difference between related versus diversifying acquisitions is insignificant. In Table 4, Panel A shows that bidders with insider dominated boards have abnormal returns that are less negative (though insignificantly so) than for those with outsider dominated boards. Moeller et al. (2005) show the large losses of acquiring firms, which are mainly driven by a small number of extremely large losses. In our sample, deal size is significantly larger in related acquisitions than in diversifying acquisitions. Size effect may drive the smaller (i.e., more negative) CARs in related acquisitions. 20

21 Table 4 also indicates that for bidders with insider dominated boards, related acquisitions result in higher mean abnormal returns than for those engaging in diversifying acquisitions, but lower median abnormal returns (neither the mean nor median difference is significant). For bidders with outsider dominated boards, diversifying acquisitions result in less negative abnormal returns than those that engage in related acquisitions, and the mean and median differences are significant at the 5% level. In Table 4, Panel B, we segment the sample by blockholding of greater than 5% and less than 5%. CARs (-1, +1) for bidders of firms with high blockholdings are higher than for those with low blockholdings (i.e., % versus %, respectively), and the difference is statistically significant. To summarize, bidders for diversifying acquisitions with independent boards experience higher abnormal returns than those for related acquisitions. Furthermore, for both high and low blockholding firms, bidders engaging in related acquisitions experience larger negative cumulative abnormal returns than do firms announcing diversifying acquisitions, though the difference is insignificant. However, for high blockholding firms, abnormal returns are significantly higher than for low blockholding firms in both related and diversifying acquisitions. Interestingly, bidder returns are highest in diversifying acquisition with high blockholdings. This result suggests that diversifying acquisitions may be less value destructive if appropriate monitoring mechanisms are put into place. This is consistent with the monitoring hypothesis, and suggests that governance quality affects the market s perception of the potential for wealth creation through mergers and acquisitions. 21

22 In order to examine the wealth generated by diversifying and related acquisition activity for both the bidders and targets combined, we present the combined CARs in Table 5. [Table 5 About Here] Table 5, Panel A shows the combined value weighted acquirer-target abnormal returns by level of board independence. Mean (median) overall combined abnormal returns for all categories of bidder-target combinations are 0.618% (0.525%). This combined abnormal return is smaller than that of Moeller et al. (2005). It is possible that our data include higher frequency of larger firms because of corporate governance data requirements. However, we also observe increase in wealth to the shareholders of both firms when we take into account the value created for the target as well as for the bidder. Given the mean abnormal return to the bidder-target combination, the average of total dollar value of wealth generated in each acquisition is approximately $15.51 million since the average of market value of equity in combined firm based on the fiscal year end value before the announcements is $2,509 million. Overall, the returns to firms engaging in diversification are positive; for related acquisitions, abnormal returns are insignificant. Furthermore, the abnormal returns for diversifying firms are higher than those for related firms, but the difference is not significant. For diversifying acquisitions, firms with independent boards experience greater abnormal returns than those of insider dominated boards. The results for the diversifying acquirers are consistent with the monitoring hypothesis, namely, that outsider boards can ensure that managers do not engage in activities that will destroy shareholder wealth. 22

23 Table 5, Panel B shows the comparisons of CARs by institutional blockholdings. Consistent with the monitoring hypothesis, firms with greater blockholdings exhibit a higher (and statistically significant) combined abnormal return than those with low blockholdings. This suggests that governance quality affects the market s perception of the potential for wealth creation through mergers and acquisitions. Firms with high levels of blockholding experience positive and significant combined CARs, regardless of whether the transaction was related or diversifying. For firms with both low and high levels of external monitoring (as proxied by blockholdings), diversifying acquisitions result in higher combined abnormal returns than those for related acquisitions. Overall, our results suggest that when both bidder and target abnormal returns are taken into consideration, acquisitions involving greater blockholdings and independent boards are wealth enhancing, and significantly more so than firms that have less stringent corporate governance controls in place. However, caution with the interpretation of the univariate results should be taken, since other factors are not controlled for. V.2. Multivariate Analysis We next investigate the relationship between board characteristics, blockholding and cumulative abnormal returns in a multivariate framework. 10 The results are shown in Table 6. [Table 6 About Here] For the full sample of merger and acquisition transactions, Model 1 indicates that the coefficient of the diversification variable is negative, but statistically insignificant. 10 We report the result of OLS regression model to test the hypothesis. Also we try to use Heckman two stage treatment models to test our hypothesis. The result of Heckman selection model provides the similar result on the main interaction variable in this paper and coefficient of inverse mills ratio is not statistically significant. 23

24 Hence, diversification per se does not lead to the market s perception of value destruction. Furthermore, the results indicate that the percent of the board comprised of independent directors is insignificant in determining abnormal returns, as is the log of board size. Multiple bidders do not appear to affect cumulative abnormal returns. However, cash deals are significantly and positively related to CARs. This reflects higher liquidity, and the result is consistent with Asquith et al. (1983). Size of the firm (log of total assets) and relative deal size are negatively and significantly related to abnormal returns, and Tobin s q is significantly and positively related to CARs, indicating that firms with higher growth options are expected to make more value enhancing acquisitions than low Tobin s q firms. Leverage is not significantly related to abnormal returns, nor is duality, the ratio of operating cash flow to total assets, and director ownership. Model 2 incorporates an interaction term between diversification and board independence. As with Model 1, the results indicate that diversifying acquisitions are significantly and negatively related to abnormal returns. This is consistent with prior evidence of wealth destruction and the diversification discount. However, and more importantly, the interaction term between the percent of independent directors and diversification is positive and significant. 11 This means that when boards are structured so that they are independent, diversifying acquisitions contribute positively to abnormal returns; outside directors do a better job of ensuring that managers do not engage in wealth destruction, consistent with our monitoring hypothesis (i.e., diversification may not necessarily be value destructive if managers are monitored appropriately). Other 11 We also use a dummy variable equal to 1 if independent director is higher than 50%, instead of percentage of independent director and interaction variable, as per the Byrd and Hickman (1992) analysis. The results are very similar to our results. However, an interaction variable with board size is not statistically significant, implying that board composition is a more important determinant of announcement returns than board size. 24

25 control variables retain the signs from in Model 1. Model 3 incorporates the public status of the target as a control, and the result indicates that announcements of acquisitions of private targets yield significantly higher abnormal returns than public targets. Further, Model 4 indicates that the interaction between private target and stock deal is significantly and positively related to abnormal returns. This result is consistent with Chang (1998), and supports the internal control hypothesis whereby when the private entity becomes a shareholder of the acquirer, a new monitoring group is introduced. The result of Model 5 is consistent with previous results. All other variables retain their signs and significance, with the addition of blockholding in Model 5, which is significant and positively related to CARs. This result suggests that blockholders, as well as independent board of directors, provide an effective monitoring role that ensures that diversification decisions will be made wisely and in accordance with shareholder wealth creation. In Model 5, the presence of a pension plan is negatively related to CARs. This result is consistent with Wahal (1996), Gillian, Hartzell, and Starks (2003), and Karpoff et al. (1996), who find that monitoring by public pension funds does not increase shareholder wealth. In all five models, diversification has a negative and significant coefficient; but the interaction term between diversification and the percentage of outside directors is significant and positive. Hence, our results show that diversifying mergers are, on average, value destructive. However, independent boards facilitate a lesser wealth destruction. Thus, our results suggest that it is board quality that determines whether a diversifying acquisition will be wealth enhancing or wealth destructive. V.3. High- Tech vs. Non high-tech Firms 25

26 As previously discussed, high-tech firms are characterized by substantial information asymmetry, as their lines of business are often difficult to understand. Because they have proprietary technology, these firms are better served by having boards with inside directors (who understand the technology well). Because the characteristics and optimal board structure of high-tech firms may differ from those of non high-tech firms, we next run regressions on cumulative abnormal returns for high-tech and non high-tech separately, as motivated by (Raheja, 2005). These results are shown in Table 7. [Table 7 About Here] We present three models for both non high-tech and high-tech firms. For the non high-tech firms, Model 1 indicates that the diversification variable is negative but insignificant. Thus, for non high-tech firms, diversifying acquisitions do not appear to significantly destroy wealth. The percentage of independent directors is also insignificant. This implies that the structure of the board whether insider or outsider dominated does not impact abnormal returns. The log of board size, multiple bid deals, and stock deals are insignificant determinants of abnormal returns. However, cash transactions are significantly and positively related to abnormal returns, indicating that the market views more liquid firms favorably when they engage in acquisitions. Relative deal size is significantly and negatively related to CARs, indicating that larger deals destroyed more wealth than smaller deals, consistent with Moeller et al. (2004). Firm size, as measured by the log of total assets, is significantly and negatively related to abnormal returns. CEO age is a negative and significant determinant of abnormal returns for non high-tech firms. None of the other control variables, including Tobin s q, leverage, market value, percent of director ownership, or duality are significant. 26

27 Model 2 introduces the interaction variable between diversification and board independence, which is insignificant in determining abnormal returns. The interaction variable is positive but insignificant. All other variables retain their sign and significance from Model 1. Model 3, again for non-high tech firms, is consistent with Models 1 and 2, with the exception of the stock transaction variable, indicating that stock financed transactions are viewed by the market as wealth destructive. The institutional holding of pension plans and blockholdings do not contribute significantly to abnormal returns. Regression results are next shown for the subsample of high tech firms. Model 4 indicates that diversification is positive but insignificantly related to abnormal returns. Percent of independent directors is insignificant in determining abnormal returns, as are board size, multiple bid deals, and stock payment. Cash transactions have significantly higher abnormal returns than stock transactions, consistent with the results for non hightech firms. Furthermore, larger transactions reduce wealth significantly. For high-tech firms, unlike non high-tech firms, Tobin s q is a significant positive determinant of abnormal returns, most likely because these are firms in nascent industries with high growth options. The other control variables (leverage, operating cash flow to total assets, director shareholdings, duality, and CEO age) are all insignificant determinants of abnormal returns. In Model 5, we examine the relationship between board structure and abnormal returns. We find a negative and statistically significant coefficient for independent directors, implying that the percentage of inside directors are more effective for governing high tech firms in related acquisitions, consistent with Boone et al. (2006) and Raheja (2005). The evidence indicates that the advisory role of insider directors is more 27

28 important for high tech firms than for non high-tech firms, because insiders understand complex technology well. Further, Model 5 introduces the interaction variable of diversification and percentage of the board consisting of independent directors. Unlike for non high-tech firms, the interaction term is positive and significant in determining abnormal returns, though the coefficient of diversification is negative and significant. This result suggests that for high-tech firms, diversifying acquisitions are less value destructive so long as the board is structured in a way that facilitates proper review of managerial decisions. Thus, the independent dominated board plays an important role in determining wealth creation for high-tech firms in diversifying mergers and acquisitions. As with non high-tech firms, relative transaction size is a significant and negative determinant of abnormal returns, with the same sign as in Model 4. Furthermore, leverage is significant and positive; indicating that high-tech firms that are more closely monitored by bondholders and have less free cash flow will have higher announcement returns. Model 6 is consistent with Model 5, and introduces pension plan and blockholding variables. The results suggest that while pension plans are insignificant in determining abnormal returns, blockholding is significant, a deviation from the results found for non-high tech firms. Thus, monitoring plays a stronger role in determining whether firms will make value enhancing or value destructive decisions for high tech firms. This is consistent with Denis et al. (1997) and Shleifer and Vishny (1986), who demonstrate that outside blockholders may have an influence on managers, preventing them from making value destructive diversification decisions. Since insider-dominated boards experience positive abnormal returns, perhaps blockholders and creditors act as additional monitors, and counteract any misbehavior on behalf of the insider dominated 28