Seasoned Equity Issues and Ownership Concentration in a Closely Held Market An Alignment Effect Test Free from the Unobserved Firm Heterogeneity Problem.docx

《Seasoned Equity Issues and Ownership Concentration in a Closely Held Market An Alignment Effect Test Free from the Unobserved Firm Heterogeneity Problem.docx》由会员分享，可在线阅读，更多相关《Seasoned Equity Issues and Ownership Concentration in a Closely Held Market An Alignment Effect Test Free from the Unobserved Firm Heterogeneity Problem.docx（37页珍藏版）》请在得力文库 - 分享文档赚钱的网站上搜索。

1、 Seasoned Equity Issues and Ownership Concentration in a Closely Held Market: An Alignment Effect Test Free from the Unobserved Firm Heterogeneity Problem Xueping Wu* Department of Economics and Finance, City University of Hong Kong Tat Chee Avenue 83, Kowloon, Hong Kong Tel: (852) 2788 7577; Fax: (

2、852) 2788 8806: email: efxpwucityu.edu.hk and Zheng Wang CITIC Fund Management 8/F, Tower A, CISTCC, 12 Yumin Road, Chaoyang District, Beijing, PR China Tel: (86) 10-82028888 (extension 316) email: efzwangcityu.edu.hk This version: March 2005 * Corresponding author. We wish to thank James Bergin, St

3、ephen Ching, Espen Eckbo, Lawrence Khoo, Jay Ritter, Piet Sercu, Qian Sun, John Wei and Jun Yao for helpful discussions and comments. Wu gratefully acknowledge financial support from the Research Grants Council of the Hong Kong SAR (Project No. CityU1181/03H). 1 Seasoned Equity Issues and Ownership

4、Concentration in a Closely Held Market: An Alignment Effect Test Free from the Unobserved Firm Heterogeneity Problem Abstract Opportunities for controlling shareholders to expropriate minority shareholders come largely from insider information. These opportunities and hence private benefits of contr

5、ol, largely under asymmetric information, can vary substantially across firms even within the same legal environment. As a result, this unobserved firm heterogeneity confounds the alignment effect of ownership concentration in reducing the private benefits. To control for this firm heterogeneity, we

6、 use a conditional test based on the valuation shocks of corporate financial decisions that have implications for private benefits. We find that in Hong Kong, rights offers signal large private benefits and control-diluting new issues signal small private benefits. This is consistent with a separati

7、ng equilibrium suggested by recent research. We also find that controlling ownership cannot predict the flotation method choice but the SEO announcement returns are conditionally and significantly related to the level of controlling ownership (with a negative slope for rights issuers and a positive

8、slope for new issuers). The conditional valuation shock patterns suggest a meaningful tradeoff between the incentives and the opportunities of controlling shareholders to both benefit and expropriate the minority shareholders. Key Words: Insider Ownership, Private Benefits, Unobserved Firm Heterogen

9、eity, SEO Flotation Method JEL Classification Code: G14, G32, G34 2 1. Introduction Recent literature has suggested that countries with concentrated ownership structures are associated with poor protection of minority shareholders (La Porta, Lopez-de-silanes and Shleifer, 1999), and that in contrast

10、 to managerial agency problems in widely held firms, the main concern by the market about ownership concentration is the extent to which controlling shareholders expropriate the minority shareholders (Shleifer and Vishney, 1986, 1997). Yet, given the legal environment, the expropriation from the min

11、ority shareholders is endogenously constrained because large controlling equity holdings should produce a positive alignment effect such that controlling shareholders are concerned about a potential loss in firm value as a result of bad corporate governance. Thus, it is important to understand how i

12、ncentive alignment works, as La Porta, et al. (1999, p. 474) put it: the theory of corporate finance relevant for most countries should focus on the incentives and opportunities of controlling shareholders to both benefit and expropriate the minority shareholders. Unobserved heterogeneity in the con

13、tracting environment across firms, however, gives rise to the difficulty for empirical research to detect a meaningful relationship between firm value and insider ownership (Himmelberg, Hubbard, and Palia, 1999).1 In concentrated ownership structures, unobserved 1 For U.S. firms, Morck, Shleifer and

14、 Vishney (1989) and McConnell and Servaes (1990) documented a non-linear relationship between firm value and insider ownership, consistent with a tradeoff between managerial incentive and entrenchment through an increase in insider ownership. But insider ownership is largely endogenous; besides obse

15、rved firm characteristics (Demsetz and Lehn, 1985), Himmelberg et al. (1999) suggest that the unobserved firm heterogeneity also determines insider ownership, making the observed relation between firm value and insider ownership spurious. While studies on concentrated ownership generally supported a

16、 strong positive relation between firm value and controlling ownership (see the survey paper by Denis and McConnell, 2002), these tests are equally prone to the endogeneity problem. However, the cure may be worse than the disease; controlling for the unobserved firm heterogeneity is difficult in emp

17、irical tests, as Zhou (2002) found that the use of firm fixed effects 3 firm heterogeneity is more likely to arise from information asymmetries about the private benefits. For example, controlling shareholders with the same level of controlling ownership may have different opportunities to expropria

18、te the minority shareholders where these opportunities are largely insider information. If an increase in controlling ownership tends to reduce the extent to which the controlling shareholders expropriate the minority shareholders, information asymmetries about private benefits will confound this in

19、centive alignment effect. This paper examines one important channel through which controlling ownership determines firm valuenamely, an increase in controlling ownership reduces private benefits of control, ceteris paribus. Wu and Wang (2003) suggest that under asymmetric information about private b

20、enefits, the choice of flotation method of seasoned equity offerings (SEOs) can significantly reveal the inside information about the issuers private benefits at the SEO announcement. Thus, corporate financial decisions that have strong implications for private benefits may significantly reduce the

21、unobserved firm heterogeneity problem and facilitate a clean alignment test. In SEOs worldwide, there are usually two major flotation methods: rights offerings and new issues (see the survey papery by Eckbo and Masulis, 1995). Rights offerings are new equity sales to the existing shareholders on a p

22、ro-rata basis, and new issues (or simply non-private placements in many countries) are firm-commitment underwritten offers to the public. Wu and Wang (2003) suggest that the SEO flotation method choice is mainly driven by issuers concern with a possible loss in some private benefits. Unlike rights o

23、fferings, control-diluting new issues weaken controlling shareholders control over their firms. While a hostile takeover is rare in countries with concentrated ownership structures, a in Himmelberg, et al. (1999) may simply throw away the genuine incentive alignment effect of insider ownership in th

24、e cross-sectional data. 4 reluctant sharing in the private benefits with other newly emergent large shareholders is likely to occur when the incumbents control is weakened (see the argument for private benefits sharing in Zwiebel, 1995, and Gomes and Novaes, 2001). If private benefits are large, the

25、 incumbent controlling shareholders may not afford to use control-diluting new issues and hence have to choose rights offers to safeguard their private benefits. The model of Wu and Wang (2003) predicts that with sufficient variations in private benefits under asymmetric information, a separating eq

26、uilibrium can emerge where issuers with large private benefits (or low value issuers) choose rights offers and those with small private benefits (or high value issuers) choose new issues. Slovin, Sushka and Lai (2000) observe a similar separating equilibrium in the U.K., although they rely on an und

27、erwriter certification argument. The revisions in the markets estimates of private benefits of control at the announcement of the SEO flotation method choice are rational but comprehensive. It is well known that new issues in the U.S. cause a negative market reaction, as explained by the Myers-Majlu

28、f (1984) adverse-selection effect. In rights offerings, if managers act in the interest of the existing shareholders, as assumed in Myers and Majluf (1984), they are unlikely to sell overvalued new equity to these shareholders. As a result, many have believed that rights offers protect the interest

29、of all existing shareholders. Empirical evidence from the U.S. appears largely consistent with the insight from Myers and Majluf (1984). But conflicts between large shareholders (rather than managers) and dispersed minority shareholders are at the center of corporate governance problems in concentra

30、ted ownership structures (Shleifer and Vishny, 1997). Such conflicts stem from the existence of private benefits of control, which arise from the deviation of the cash flow rights from the control rights.2 If self-interested controlling 2 In reality, cash flow rights never fully coincide with contro

31、l rights. In a dual-class ownership structure with superior voting rights for one class of shares, or in a pyramid structure, the deviation is obvious. In the one-share- one-vote structure advocated by Grossman and Hart (1988) and Harris and Raviv (1988), as is common in most 5 shareholders maximize

32、 their own wealth, which includes both the cash flow benefits of their equity ownership and the control benefits, the predictions of the Myers and Majluf (1984) model have to be modified. For example, a new issue is more likely to produce a positive announcement effect when the asymmetric informatio

33、n arises from growth more than from assets in placeit is the latter that mainly drives the adverse selection effect (Cooney and Kalay, 1993; Wu and Wang, 2004). On the other hand, rights offers may not protect the interest of minority shareholders if the private benefits under asymmetric information

34、 are large (Wu and Wang, 2003). In this paper, we show evidence consistent with these predictions. Using Hong Kong data from 1989 to 1997, we find that the rights offers and new issues have opposite announcement effects. More precisely, for a two-day announcement window, a sample of 180 rights offer

35、ings has, on average, a significantly negative cumulative average abnormal return (CAR) of 3.4 percent. Using a three-day announcement window, the CAR is 7.6 percent. In sharp contrast, a sample of 306 new issues or non- private placements produces, on average, a significantly positive CAR to the tu

36、ne of 1.9 with a two-day and 3.1 percent with a three-day window. 3 This phenomenon, especially regarding the negative announcement effect of rights offers, indicates that controlling shareholders are willing to choose value- listed firms, there is also a de facto deviation. Unlike the cash flow rig

37、hts, the value of voting rights from the same shares is asymmetric between large shareholders (or managers in a coalition with other blockholders) and dispersed shareholders. 3 These findings are the reverse of the rights issue puzzle observed in the U.S., originally described by Smith (1977) and la

38、ter reviewed in Eckbo and Masulis (1995). The Smith (1977) rights issue puzzle is perplexing because, despite rights issues having lower direct and indirect flotation costs, increasing numbers of firms have favored the more expensive placement flotation method (in the U.S.). 6 destroying rights offe

39、rs even though new issues are apparently a better alternative.4 All this indicates that despite strong common law enforcement, Hong Kong firms show substantial variations in private benefits of control. How do we know that the announcement effects documented here, in particular, the panic drops of r

40、ights issuers stock prices, are mainly related to the upward revisions in the markets estimates of private benefits? Further findings of this paper show that the slope estimates for market-to-book (MV/BV) and return on equity (ROE), proxies for corporate governance quality, are significantly negativ

41、e in our Logit model of the choice of rights offerings versus new issues. This makes the connection of rights issuers with poor corporate governance. The choice by rights issuers implies that control-diluting new issues would have otherwise put those less governance-conscious controlling shareholder

42、s in a situation that could significantly weaken their control. Despite some observed firm characteristics being able to predict the flotation method choice to some extent, the information asymmetries about private benefits in Hong Kong are so severe that the choice announcement still has significan

43、t information content. For one thing, our Logit regression results show that the level of controlling ownership cannot predict the flotation method choice. This suggests substantial uncertainty over the valuation effects of controlling ownership as a result of severe information asymmetries about pr

44、ivate benefits before the SEO announcement. Given that the 4 Significantly negative announcement effects of rights offerings are also documented by Gajewski and Ginglinger (2002) in France, Kabir and Roosenboom (2003) in the Netherlands, and Slovin, Sushka and Lai (2000) in the U.K. To the extent th

45、at rights offerings are much more popular in the other countries than in the U.S. (Eckbo and Masulis, 1995), the negative announcement effects of rights offerings in other countries have deepened the Smith (1977) rights issue puzzle (see the discussion in Wu and Wang, 2003). 7 announcement can signi

46、ficantly reveal the issuers private benefits, the conditional valuation effects of controlling ownership should become pronounced subsequently. We find that the cross-sectional announcement returns of SEOs are conditionally and significantly related to controlling ownership with the slope signs depe

47、nding on the flotation method choice. More precisely, while valuation shocks are negative for rights issuers and positive for new issuers, we see big market responses when the level of controlling ownership is high and small shocks when controlling ownership is relatively low. The patterns for the m

48、arket to rationally respond to the announcement by issuers with different levels of controlling ownership are consistent with our test hypothesis for the alignment effect of controlling ownership plus legal challenges able to limit excessive private benefits of control (see Section 4 for details). F

49、or example, the market sharply revises down the stock prices of issuers that have a high level of ownership concentration but choose rights offers (which signals large private benefits), because the high level of controlling ownership has indicated smaller private benefits unconditionally before the announcement. In contrast, the market makes smaller revisions for rights issuers with a low level of controlling ownership, because while the low level of controlling ownership has indicated large private benefits u