Does Regulation Substitute or Complement Governance?

David A. Bechera,b, Melissa B. Fryec

Department of Finance, Drexel University, Philadelphia, PA 19014 Wharton Financial Institutions Center, University of Pennsylvania, Philadelphia, PA 19014 c Department of Finance, University of Central Florida, Orlando, FL 32816
a

Forthcoming, Journal of Banking and Finance ________________________________________________________________________

Abstract We examine whether firms utilize governance systems and increased monitoring mechanisms when information asymmetry and managerial discretion are limited. Given that such monitoring is costly, we expect regulated firms to use less monitoring if regulation substitutes for governance. Using data from initial public offerings, we document that regulated firms have greater proportions of monitoring directors and larger boards as well as use similar amounts of equity-based compensation as non-regulated firms. Further, regulated and unregulated firms are analogous in terms of observed trade-offs between traditional monitoring mechanisms and insider ownership. Finally, regulated firms appear to decrease monitoring following a period of deregulation. These findings support the hypothesis that regulation and governance are complements and are consistent with the notion that regulators pressure firms to adopt effective monitoring structures. Keywords: Corporate Governance; Regulation JEL Classification: G21; G22, G28; G34

________________________________________________________________________
We

thank an anonymous referee, Tom Bates, Fabrizio Ferri,Laura Field, Michelle Lowry, Harold Mulherin, David Reeb, Dan Rogers, Chad Zutter, as well as seminar participants at the Financial Intermediation Research Society conference, the Financial Management Association meetings, American University, Lehigh University, University of Adelaide, and University of Central Florida for helpful comments and suggestions. We also thank Kinjal Desai for his research assistance. Corresponding author. Tel: +1-215-895-2274; e-mail: becher@drexel.edu.

Electronic copy available at: http://ssrn.com/abstract=1108309

1. Introduction Governance mechanisms are costly to implement (Shleifer and Vishny, 1997; Baker and Gompers, 2003). When there is a separation of ownership and control, however, the benefits of monitoring may outweigh the costs. In fact, Jensen and Meckling (1976) suggest that monitoring can alleviate agency problems when insider ownership is low. Firms adopt governance mechanisms to align manager and shareholder interests, thereby assuring suppliers of finance a return on their investment (Shleifer and Vishny 1997). In an environment where executive decision-making may be more transparent and opportunity sets may be limited, however, the benefits of monitoring may be reduced (Joskow, Rose, and Shepard, 1993). Governance mechanisms may be less important in regulated industries. Much of the literature to date has taken a literal interpretation of this relationship, arguing that regulation should substitute for governance. However, empirical evidence does not fully support this notion (e.g., Hadlock, Lee, and Parrino, 2002; Houston and James, 1995), raising the question as to why regulated firms adopt governance structures with greater levels of monitoring given the cost. In this paper, we provide an alternative explanation for the relation between regulation and governance that may shed light on why costly governance mechanisms are utilized by firms with restricted opportunity sets. Regulators do not have the same interests as shareholders. Their focus is on safety and soundness rather than wealth maximization (Joskow et al. 1993). While regulators do not control specific governance practices, the presence of regulators may pressure firms to adopt effective corporate governance structures that promote safety and soundness. In effect, regulators must ensure that firms comply with all procedures, but it is not feasible for them to monitor all activities. Stigler and Friedland (1962) note that it is very costly for regulators to monitor a firms actions as they cannot control daily operations. This suggests
1

Electronic copy available at: http://ssrn.com/abstract=1108309

regulators may rely on traditional governance systems to promote their goals. Booth, Cornett, and Tehranian (2002) and Joskow et al. (1993) note the threat of corrective actions by regulators and increased scrutiny on regulated firms pressures these firms to adopt effective monitoring systems. Joskow, Rose, and Wolfram (1996) contend that governance differences are the results of regulatory pressure rather than inherent productivity differences. Regulatory pressure may encourage greater monitoring (e.g., best practices approach). Essentially, regulation and governance may work together to ensure an effective governance structure. Our analysis is organized around a corporate event (initial public offering or IPO) rather than in calendar time. Baker and Gompers (2003) note that monitoring mechanisms are more likely optimally chosen at the IPO since existing shareholders bear the cost of suboptimal governance. Brown, Dittmar, and Servaes (2005) demonstrate that IPO firms with the proper initial governance structure have better operating and stock performance. Wang, Winton, and Yu (2009) contend monitoring mechanisms are less ambiguous at the IPO. Higgins and Gulati (2006) show young firms influence investor decisions by giving information that signals organizational legitimacy through effective governance structures at the IPO. Engel, Gordon, and Hayes (2002) find incentives to monitor newly public firms are stronger than at well established firms, suggesting governance structures are more important. Finally, Hartzell, Kallberg, and Liu (2008) focus on the IPOs of a regulated industry (REITs) to analyze the impact of corporate governance. The authors contend this approach mitigates the endogeneity problem present in studies of the impact of governance on seasoned firms valuation. By contrast, in calendar time, governance structures may be as much a consequence of past performance as a measure of the quality of governance. Analyzing structures at the IPO enables us to examine governance while decreasing the impact of prior performance. However,

firms do not arbitrarily decide to go public. It is a pre-determined choice that is planned for well in advance and requires the firm to establish a governance structure that will enable the firm to go public (Burton, Helliar, and Power, 2004). Any IPO effect will affect regulated and unregulated firms alike, thus minimizing any impact on our study. Our paper provides strong support that regulatory pressure, rather than substitution, influences governance. We document that regulated firms have governance structures with greater monitoring than unregulated firms at the IPO. If governance mechanisms are unnecessary (regulation substitutes for governance), regulated firms should have less monitoring. This heightened monitoring is not related to firm characteristics typically associated with regulated firms, such as leverage age, or size. We also examine trade-offs between monitoring mechanisms and ownership, controlling for firm characteristics. If regulation substitutes for governance, the degree of interdependency between traditional monitoring mechanisms and ownership should be lower at regulated firms (Booth et al., 2002). Regulatory pressure, however, suggests regulation and governance both ensure a system of governance where monitoring mechanisms are interchanged (Adams and Mehran, 2003). We document monitoring mechanisms serve as alternates for ownership at both regulated and unregulated firms at the IPO as well as two and four years post-IPO (using levels and changes). These results do not indicate that regulation substitutes for governance, rather regulation serves as a means of pressuring firms to adopt effective governance systems. We further examine the impact of regulation on governance by analyzing the role of deregulation. Deregulation increases the importance of the managerial role within a firm and the need for monitoring (Kole and Lehn 1997). Removing regulation introduces additional downside risk, increases managerial discretion, and may impact the sensitivity of firm value to the quality

of managerial decisions. If regulation substitutes for governance, relaxing regulation should lead regulated firms to increase their monitoring (Kole and Lehn, 1999). However, governance mechanisms of regulated IPOs post-deregulation do not increase. In fact, consistent with the removal of regulatory pressure, some monitoring levels actually decrease post-deregulation. While this paper focuses on regulated versus unregulated firms, our findings have implications for research on corporate governance more broadly. A debate exists as to whether corporate governance affects market values (Gompers, Ishii, and Metrick, 2003). Corporate governance can reduce agency problems and lead to more effective monitoring of managers. However, adopting these mechanisms is costly. In an environment where information asymmetry and managerial discretion are limited, monitoring systems would be redundant (regulation would substitute for governance). Our results, however, are not consistent with this substitution assumption, and provide additional support on the importance of corporate governance in protecting shareholders. The remainder of this paper is organized as follows. Section 2 details the motivation and hypotheses. In section 3, we describe our samples and summary statistics. Section 4 presents empirical results and differences between regulated and unregulated firms at the IPO. In sections 5 and 6, we provide additional specifications and robustness tests, while section 7 concludes. 2. Motivation 2.1. Regulation as a substitute or a complement for governance Adams and Ferreira (2008) note that the issue of whether regulation substitutes for governance remains an open question. Empirical research provides some evidence that regulation substitutes for traditional monitoring mechanisms. Joskow, et al. (1993) find lower pay for regulated CEOs; Kole and Lehn (1999) note that the governance structures at regulated firms move toward structures of unregulated firms post-deregulation. Crawford, Ezzell and Miles

(1995) document a stronger link between compensation and performance post-deregulation. Becher, Campbell and Frye (2005) find bank directors receive less incentive compensation than non-bank directors, while Booth et al. (2002) show internal monitoring of managers at regulated firms is less important. Caprio, Laeven, and Levine (2007) state bank regulations may be sufficiently pervasive that they render shareholder protection laws superfluous. Thus, the presence of regulators may substitute for traditional shareholder monitoring mechanisms by reducing the effect of managerial decisions on shareholder wealth. However, others are inconsistent with the substitution argument. Hadlock et al. (2002) find regulated CEOs are held at least as accountable for performance as non-regulated CEOs. Houston and James (1995) indicate CEO stock holdings and option-based compensation are lower in banking due to differences in investment opportunities and other firm characteristics rather than regulation. Adams and Mehran (2003), Booth et al. (2002) show that boards of regulated firms have greater independence than non-regulated firms. If regulated firms require less monitoring, boards should have a lower proportion of independent outside directors relative to non-regulated firms. Roengpitya (2007) finds that intrastate bank deregulation leads to a lower proportion of outsiders on the board. Pathan and Skully (2010) show that banks structure their boards in a way that is consistent with shareholder wealth maximization. These findings cast doubt on the substitution hypothesis. Furthermore, Joskow et al. (1993) note that regulators do not have the same financial interests as shareholders and focus on safety and soundness rather than wealth maximization. Similarly, Caprio et al. (2007) indicate that regulation does not impact firm value (positively or negatively), consistent with regulation serving a different role. Regulators do not set specific

monitoring levels1, board size, or board independence. However, regulatory presence may pressure firms to adopt effective corporate governance structures to ensure that rules and requirements are met. Given regulators are unable to effectively control daily operations, it is costly to monitor regulated firms actions (Stigler and Friedland, 1962). Masulis and Thomas (2008) argue financial firms need more monitoring because of their heightened risk exposure. To explain governance differences, Booth et al. (2002) point to the threat of corrective actions by regulators. Joskow et al. (1993) note regulation increases the visibility of corporate governance through enhanced public scrutiny and provides a set of instruments (price and allowable cost decisions) to penalize firms with poor governance structures. Joskow et al. (1993) and Joskow et al. (1996) note that compensation structures are most impacted with the degree of regulatory intensity, with electric utilities impacted the most. This impact comes from regulatory pressure rather than inherent productivity differences. It is important that banks are perceived by regulators as well managed. If their management rating becomes less than satisfactory, their financial holding company status could be jeopardized. Adams and Ferreira (2008) contend regulators view board oversight as an important complement to supervision rather than a substitute. They point to studies by the Comptroller of the Currency and General Accounting Office, which link bank failures and inadequate board monitoring. Regulatory pressure may be strongest with respect to board structures. While regulators do not implicitly influence governance structures, there may be a more direct link with compensation. Houston and James (1995) note the FDIC Improvement Act (FDICIA) provides regulators with oversight of senior management, requiring undercapitalized firms to receive approval to pay bonuses or increase compensation. John, Saunders, and Senbet
Savings institutions face limitations on stock ownership: no person, firm, or group can acquire over 10% of voting stock without approval (Friesen and Swift, 2009). However, this appears non-binding. To illustrate, ownership levels exceed 10% for all savings banks in our sample. 6
1

(2000) argue for a more prominent role for management compensation structures in bank regulation. Regulators have been reluctant to adopt strict guidelines limiting compensation for healthy banks. Joskow et al. (1993) attribute differences in executive compensation at regulated firms to political pressures. They cite the weaker link between pay and performance at regulated firms as an effort to reflect interests of both consumers and the firm. 2.2. Role of regulation While the type of regulation will vary by industry, our aim is to examine the impact of the overall regulatory environment versus testing the effect of specific regulatory changes, as in Booth et al. (2002). Prior research suggests deregulating an industry causes regulated firms to alter governance structures and adapt structures similar to unregulated firms. Even if regulation does not directly impact governance, relaxing restrictions on regulated firms has been shown to increase competition, change opportunity sets, require greater managerial effort and create the need for a different governance structure to deal with the increased opportunities and firm complexity (Houston and James, 1995; Kole and Lehn, 1999; Becher et al., 2005). Thus, governance is affected by the presence of regulators even if they do not directly dictate monitoring levels. Focusing on the overall regulatory environment is more appropriate to evaluate the role of regulation in governance, contrasting environments where information asymmetry and managerial discretion are limited to those which are not limited by regulation. Numerous studies have focused on the impact of specific regulatory changes on corporate governance. For example, FDICIA required a majority of independent directors on bank audit committees and gave regulators oversight of senior managers. The Riegle-Neal Act of 1994 removed restrictions on bank branch locations and managing operations across state lines; setting off a merger wave creating firms much more difficult to monitor firms. At the same time, federal

and state regulators relaxed limitations on banking activities (e.g., investment banking), leading to the Graham-Leach Bliley Act of 1999. This rapid geographic and activity diversification increased these firms complexity and scope; requiring greater expertise/effort by boards and managers. In addition, Becher et al. (2005) document these regulatory changes dramatically altered the structure of bank boards, equity compensation, and overall systems of governance. For utility firms, the Energy Policy Act of 1992 (EPACT) deregulated the industry by shifting from cost-plus to marginal pricing as well as relaxing constraints on ownership. These changes greatly increased the scope and complexity of utilities and impacted their governance structures. Rennie (2006) notes that utilities governance structures were greatly enhanced postderegulation (changes in ownership, compensation, and board structure). These results show that deregulation of a regulated industry (irrespective of the type of regulatory change) directly impacts the governance, compensation, and monitoring of these firms. For other industries, Lehn (2002) suggests deregulation of the telecommunications industry significantly altered its governance structures. Friedlaender et al. (1993) examine the governance structure of rail firms after the Staggers Act of 1980 (removal of rate restrictions) and find governance improves post-deregulation. In addition, Feng, Ghosh, and Sirmans (2007) note even in a regulated environment such as REITs monitoring efforts are tied to governance structures and compensation; regulatory pressures exists to improve governance. While these other industries have been through deregulation, they still face regulatory scrutiny that could impact governance structures. Securities firms are regulated by the NASD, which registers members, governs their behavior, examines for compliance, and disciplines those that fail to comply. REITs face regulations with respect to where they can derive income, assets they can own, and payout of taxable income. Other real estate firms also face regulations,

licensing requirements, and examinations by state agencies. Finally, the SEC has sought to impose broader governance requirements on the mutual fund industry (Mulherin, 2007). Overall, there is a wealth of evidence to suggest deregulation (in any form) can impact regulated firms corporate governance and cause them to alter their governance systems. 2.3. Hypotheses The extant literature has focused on the idea regulation substitutes for governance, but fails to address empirical inconsistencies. In addition to examining the substitution hypothesis in a new setting, we propose an alternative hypothesis: regulatory pressure. Regulators focus on safety and soundness rather than shareholder interests, which weakens the substitution argument. Essentially, the substitution hypothesis implies these goals are interchangeable or paired. Rather than substitute, regulators may pressure firms to adopt effective governance structures; regulation and governance may work together and serve as complements rather than substitutes. To explore if regulation substitutes or complements governance, we test three hypotheses. H1A Substitution. If regulation substitutes for governance, monitoring levels should be lower at regulated firms. H1B Pressure. If regulation pressures firms to adopt effective monitoring structures, monitoring levels should be higher at regulated firms. The substitution hypothesis predicts that regulated firms should have lower levels of monitoring directors, smaller boards, less monitoring shareholdings, and implement less equitybased compensation. In contrast, greater use of these mechanisms would be consistent with the regulatory pressure hypothesis, where there is pressure to adopt effective governance structures. Further, regulatory pressure may help ensure that firms are at their optimal governance structure. However, governance mechanisms are costly to implement and firms tend to establish

monitoring systems where one measure may be increased when another decreases, suggesting the firms overall system of governance is also important to examine. Note that we examine monitoring levels at the IPO, when firms experience a significant dilution of inside ownership.2 However, regulation would have been present prior to the IPO. If regulation substitutes for traditional monitoring, this would be true before and after the IPO, suggesting lower monitoring levels before and after. In contrast, if regulation serves as a complement, monitoring levels at the IPO would be higher or similar to unregulated firms. H2A Substitution. If regulation substitutes for governance, trade-offs between monitoring mechanisms and inside ownership will not exist. H2B Pressure. If regulation pressures firms to adopt effective monitoring structures, trade-offs between monitoring mechanisms and inside ownership will exist. As agency problems arise from the separation of ownership and control, effective monitoring should include alternate monitors for low inside ownership. However, the substitution hypothesis contends that regulators replace a traditional system of governance. Examining well-established firms, Booth et al. (2002) find the need for traditional monitoring is less critical to regulated firms, since regulators are an alternative monitor. Regulated firms may be less likely to establish systems of governance where traditional governance mechanisms are exchanged for one another. In contrast, if regulators pressure firms to establish controls through the board, monitoring mechanisms should still be alternatives to inside ownership. The IPO represents a time of considerable dilution in insider holdings. If regulation substitutes for governance, this dilution will not cause significant changes in governance. In other words, firms will not need to trade-off monitoring from other mechanisms for the decline

In the prospectus, only ownership information is provided immediately prior and after the IPO. Other governance variables (board size and structure, executive option valuations, etc.), however, can only be captured after the IPO. 10

in inside ownership. Lower insider holdings will not be associated with an increase in other monitoring mechanisms. In contrast, if regulation serves as a complement, the dilution will lead firms to alter monitoring to essentially take up the slack from this decrease. Lower insider holdings will be associated with an increase in other monitoring mechanisms. We examine levels and changes in inside ownership and other monitoring mechanisms to examine whether trade-offs exist. While our first two hypotheses rely on governance characteristics, our third considers governance dynamics by focusing on how these structures change in a deregulatory period. H3A Substitution. If regulation substitutes for governance, monitoring of regulated firms should increase following deregulation. H3B Pressure. If regulation pressures firms to adopt effective monitoring structures, monitoring utilized by regulated firms post-deregulation should decrease. When regulation is at least partially removed, substitution predicts traditional monitoring mechanisms will increase as regulators provide less monitoring (Kole and Lehn, 1999). Deregulation causes firms to adapt their governance to handle the increased opportunity set and managerial discretion. Without regulatory pressure, however, some firms may decide to reduce monitoring if it were too high under regulation or if agency problems lead firms to choose less monitoring post-deregulation. Essentially, the regulatory pressure hypothesis does not support an increase in monitoring following deregulation.3 3. Data and summary statistics

An alternative interpretation suggests no changes in governance post-deregulation if monitoring mechanisms are in equilibrium. If governance is in equilibrium, it is not clear monitoring levels would decrease. However, substitution clearly predicts an increase in monitoring; thus, no change in monitoring would not be consistent with substitution. 11

We collect data for IPOs in 1993, 1996, and 1998.4 Analyzing a sample during the 1990s allows us to examine how deregulation affected governance and test the pressure and substitution hypotheses. Also, Gompers et al. (2003) note legislation in the 1980s resulted in wide variation in governance structures for established firms while these structures were more stable in the 1990s and may be more representative of current governance structures rather than reactions to legal provisions. Further, using a more recent year would substantially decrease our sample of heavily regulated firms (4, 5, 4, 7, 13, and 5 potential IPOs from 2001 - 2006, respectively). For the 1993 sample, we start with 453 IPOs on the IPO Prospectus database developed by R. R. Donnelley Financial and IPO Crossroads. For 1996, we use the SEC's EDGAR database to get prospectuses for IPOs after May (not required to file electronically before this). For 1998, we again use EDGAR to obtain prospectuses. For banking firms and other financials we collect prospectus from SNL Interactives document archive. Compustat and CRSP data are utilized for financial data. Our final samples consist of 436 IPOs in 1993, 444 in 1996, and 281 in 1998. 3.1. Governance data We examine seven commonly used measures of the firm's governance structure, which include the proportion of independent outside directors, proportion of venture capitalist directors, size of the board, shareholdings of outside blockholders, shareholdings of venture capitalists, percentage of equity-based compensation, and officer and director shareholdings (insider holdings). We also examine monitoring directors, which we define to be the proportion of independent outside directors and venture capitalist directors combined, as well as monitoring holdings, which is the combination of outside blockholdings and venture capitalist holdings.

These years were chosen for the following reasons: data availability, internet bubble of 1999, changes in executive compensation disclosure in 1993, and to be consistent with Booth et al. (2002). While it is possible these years are anomalous, our sample is comparable to Chen and Ritter (2000) and others. 12

Adams and Mehran (2003) note that typical external governance mechanisms (e.g., hostile takeovers) are absent in regulated firms; it is more appropriate to focus on internal governance structures and block ownership than potential ineffectual external mechanisms. Roengpitya (2007) documents that external governance mechanisms do not replace internal mechanisms in banks. Using data on non-IPO firms, Booth et al. (2002) show other monitoring mechanisms are used in lieu of inside ownership at unregulated firms, but the interdependencies are weaker for regulated firms. We follow this approach by examining whether trade-offs exist between monitoring mechanisms and inside ownership for regulated and unregulated IPO firms. Our focus on inside ownership also relates to Jensen and Meckling (1976) who associate a lack of insider control with the need for shareholders to protect their interests. One means to monitor managers is through the board. John and Senbet (1998) contend how effective a board is in its monitoring function is determined by composition and size. Increasing board independence may be beneficial since these directors are more likely to monitor executives. Likewise, the boards monitoring potential may increase as more directors are added, especially for startup companies with relatively small boards.5 IPOs also may receive monitoring from venture capitalists on their board (Baker and Gompers, 2003; Suchard, 2009). Large blockholders may also monitor to protect their interests. Burkart and Panunzi (2006) argue large shareholders can be effective monitors. Barry, Muscarella, Peavy, and Vetsuypens (1990) show venture capitalists own economically significant equity positions and participate in the governance of portfolio firms.

For well-established firms with larger boards, the costs of poor communication and decision-making with a large group may outweigh benefits of additional monitoring. Some evidence on established firms suggests large boards are less efficient (Yermack, 1996). Adams and Mehran (2009) find the opposite for banking firms. Our focus is not on board efficiency but monitoring potential. Further, Baker and Gompers (2003) note IPO boards are substantially smaller than those of large, public companies, suggesting the large group issues may be less of a concern. 13

Equity-based compensation may also be traded off for inside ownership. One way to align manager/shareholder interests is by ex-ante contracting, where agency costs are mitigated by incentive compensation (Jensen and Murphy, 1990). Mehran (1995) finds firms with high inside ownership rely on less equity compensation for top executives. Data for these governance measures are collected from the prospectus and proxy statements. Independent outside directors exclude insiders and gray or quasi-outside directors. To illustrate, former executives, executive spouses, and lawyers or consultants with a working relation are not outsiders. Gray directors are not considered independent directors since they may have conflicting goals. Consistent with prior studies (e.g., Adams and Mehran, 2003), directors with lending relationships with a bank or savings institutions are not eliminated from being labeled independent. Board size is the total number of directors. Outside blockholders are institutions or companies that own at least 5% of shares outstanding. Inside ownership is the percentage of shares owned by officers and directors. The percentage of incentive compensation is measured as average percentage of compensation that is equity-based for top-executives, whose compensation is reported in the prospectus/proxy.6 Equity-based components include stock options, restricted stock, and performance shares from long-term incentive plans.7 3.2. Degree of regulation The literature is not always consistent as to which types of firms should be regard as regulated firms. We consider a spectrum of regulation from all regulated firms to heavily regulated firms to banking firms.

Results are robust to using the percentage of equity-based compensation for the CEO only. To calculate option grant values, we apply a variant of Black-Scholes (Noreen and Wolfson, 1981). In the year preoffering, we estimate stock return variance two ways: 20-day after market standard deviation (Beatty and Zajac, 1994) and industry median annual standard deviation of monthly returns for the year pre-IPO (Baker and Gompers, 1999). Industries are at the 4-digit SIC level (or 3-digit if insufficient data are available). The methods produce near identical results; correlation between these two is 0.99. Reported results use the Beatty and Zajac (1994) approach.
7

14

If regulation substitutes for traditional monitoring, regulation should provide a more complete substitute at heavily regulated firms (and similarly if regulation acts as a pressure). Heavily regulated firms include banks, savings institutions, and gas and electric utility companies. Prior studies suggest these industries face an enhanced level of regulatory influence and Booth et al. (2002) find the same interdependencies between governance mechanisms for banks and utilities. Both depository institutions and public utilities have experienced deregulation over recent years; however, remain substantially regulated. Note that we also separate out banking firms. If regulators play a different role in the banking and utility industries, costs and benefits of monitoring may vary. For example, bank regulators may be primarily concerned about safety and soundness whereas utility regulators may be more concerned with reliability. In short, banks and utilities may not face the same regulatory pressure.8 All regulated firms comprise these heavily regulated firms plus partially regulated firms (which includes transportation, telecommunications, and other financial, non-depository firms). Including all regulated firms allows us to measure the broad impact of any regulation and its effect on governance. The literature, however, is inconsistent on how to treat partially regulated firms. Booth et al. (2002) consider telecommunications firms as utilities, but transportation and other financials as unregulated. Others (e.g., Baker and Gompers, 2003) classify all financials as regulated. In terms of the regulatory environment, Federal rules prevent transportation firms from operating as freely as those in non-regulated industries. The telecommunications industry also has been partially deregulated. Other financial firms include insurance, securities brokers, mortgages, and real estate. Such companies act as financial intermediaries, but are less regulated and subject to greater market discipline than depository institutions. Nonetheless, these firms still face regulations and restrictions not faced by unregulated firms.
8

Results are qualitatively the same if we include utility firms in the all other regulated group or delete them. 15

To determine if a firm was a heavily regulated utility, we start with all IPOs with an SIC code of 4900-4939 (electric and gas), 1300 (oil and gas extraction); and 6710-6719 (holding companies). For banking firms, we start with all firms with an SIC code of 6020-6039 as well as 6710-6719 (holding companies). To ensure we only include regulated utilities and banking firms, primary and secondary operations were examined. In case of ambiguity, its history, productions and operations, industry, and top competitors from Hoovers, Moodys, Yahoo Finance, and a firms financial statements and annual reports were analyzed to determine its operations. For partially regulated firms, we start with SIC codes of 4000-4700 (transportation), 4800 (telecommunications), 4950-4959 (sanitary services), and all 6000s (financial companies) excluding banking firms. Of our 1,161 IPO firms, 928 are unregulated, 175 are partially regulated, 58 are heavily regulated (18 in 1993, 24 in 1996, and 16 in 1998) and 37 banks. Table 1 details summary statistics for all regulated, heavily regulated, banking, and unregulated firms at the IPO. As expected, banking and heavily regulated firms have the greatest amount of leverage. Regulated firms are significantly larger and older than unregulated firms but generally do not differ in terms of tangible assets or ROA. Banks and heavily regulated firms are less likely to have a founder involved, which may be consistent with these firms being older at the IPO. Heavily regulated firms and banks also have the lowest adjusted q values. 4. Complements or substitutes 4.1. Monitoring intensity To explore whether regulation substitutes for or complements governance, we examine monitoring intensity for regulated, heavily regulated, bank, and unregulated firms (Hypothesis 1). Summary statistics in Table 2 support the regulatory pressure rather than the substitution hypothesis. If regulation substitutes for monitoring, we would expect regulated firms to receive less monitoring from traditional governance measures. Specifically, the substitution hypothesis

16

predicts regulated firms should have fewer monitoring directors, smaller boards, less monitoring holdings, and less equity-based compensation than their unregulated counterparts. Contrary to this hypothesis, in Panel A of Table 2 heavily regulated firms and banks have greater proportions of monitoring directors and larger boards. Heavily regulated and banking firms have significantly lower monitoring holdings; however, all regulated firms and unregulated firms have similar amounts. The use of equity-based compensation is similar for all firms. Focusing on the components of our monitoring mechanisms in Panel B, banks, heavily, and all regulated firms all have significantly more independent directors than unregulated firms. Examining outside blockholdings highlights that lower monitoring holdings for banks and heavily regulated firms is largely attributed to the lack of venture capitalist holdings. Overall, our results are more consistent with the regulatory pressure hypothesis, where regulated firms feel pressure to adopt governance structures that provide heightened monitoring. The differences above may not be related to regulated firms versus unregulated firms per se; rather regulated firms may have specific characteristics that require extra monitoring. Prior studies detail that regulated firms are more highly levered, larger, and older than unregulated firms. In Table 3 we test whether differences in monitoring are driven by these characteristics. Following Houston and James (1995), we use one minus the ratio of the book value of equity to total assets to proxy for leverage and the natural log of total assets at the IPO to proxy for size.9 The number of years since a firms first date of incorporation until the IPO captures firm age. In Panels AC of Table 3, our sample is split into three groups (high, medium, low) based on leverage, size, and age. These groups do not consider if the firm was regulated. If firm characteristics drive observed differences in monitoring, these differences should follow a

While levels are different, results are qualitatively similar if we use total debt divided by total assets for leverage (Booth et al., 2002). Further, results remain unchanged if we remove potential outliers for leverage. 17

similar pattern for heavy regulated (banks), regulated, and unregulated firms. In other words, Table 3 patterns should mirror Table 2 patterns. To illustrate, firms with high (low) age, leverage, and size should use similar monitoring as heavily regulated (unregulated) firms. Unlike heavily regulated/banking firms, highly levered firms have significantly smaller boards and use significantly more equity-based compensation (from Table 2 heavily regulated firms have the largest boards and no difference in equity compensation). High and low levered firms also do not differ significantly in terms of monitoring holdings, while heavily regulated/banking firms have less monitoring holdings. Grouping by size, larger firms have significantly larger boards (like heavily regulated and banking firms), but do not have significantly more monitoring directors. Also, larger firms have more monitoring holdings, while heavily regulated firms and banks do not. Except for board size, the results with size are not characteristics of heavily regulated/banking firms. Segmenting by age also presents differences. Older firms have significantly less monitoring directors, smaller boards, and less equity-based compensation, which are not characteristics of heavily regulated or banking firms. Older firms do not differ from younger ones in terms of monitoring holdings. Nonetheless, we control for leverage, size, and age in multivariate analyses and utilize a matched sample. Overall, Table 3 suggests regulation is not simply a proxy for leverage age, or firm size. 4.2. Multivariate analyses Next, we examine the trade-offs between monitoring mechanisms and inside ownership (Hypothesis 2).10 In Table 4 we regress each measure of governance (monitoring directors and holdings, board size, and incentive compensation) on insider holdings, firm characteristics, and other controls. We include binary variables equal to one for all, heavily regulated, or banking

10

For all multivariate analyses, the impact of multicollinearity is assessed by examining variance inflation factors (VIFs). We find no evidence multicollinearity problems exist in any specification. 18

firms and an interaction term between these variables and insider holdings. We expect other monitoring mechanisms to serve as alternatives to insider holdings, suggesting a negative coefficient on insider holdings. With regulatory pressure, the interaction term should be insignificant, consistent with regulated and unregulated firms trading off mechanisms in a similar manner. In other words, trade-offs between mechanisms will be the same at regulated firms if regulatory pressure forces firms to have an optimal system. The notion of an optimal system of governance is similar to Agrawal and Knoeber (1996) where governance is insignificant in firm performance regressions. In contrast, the substitution hypothesis would suggest a positive and significant coefficient on the interaction term. In this case, regulated firms would not be establishing monitoring systems where mechanisms are traded off for one another. We control for additional factors that may affect the need for monitoring mechanisms. Gompers (1995) argues the need for monitoring increases as asset tangibility declines (ratio of tangible to total assets). Intangible assets are associated with higher agency costs since their liquidation values are lower. We also include a binary variable equal to one if a founding family member is an officer or director at the IPO, since governance may differ if a founder is involved. We control for shares outstanding minus inside holdings to control for the public float (Bartov, Mohanram, and Seethamraju 2002). We control for profitability using return on assets (ROA) or return on equity (unreported). Baker and Gompers (2003) show venture capitalist involvement shapes the board; we include a binary variable equal to one if the IPO firm was backed by a venture capitalist. We also include the average initial returns for IPOs in that year to capture market conditions around IPOs (Lowry and Schwert, 2002) to capture hot IPO markets.11

11

We also use the number of IPOs and year dummies. Since Ljungqvist and Wilhelm (2003) express concerns about VC and insider ownership during the bubble period, we add an interaction term between insider holdings and year dummy variables as well as the venture capital backing dummy and year dummy variables. Finally, we run analyses separately for each year. Results are robust, suggesting changes in market conditions are not driving our results. 19

Numerous studies argue a link exists between firm performance and governance. For firm performance and growth opportunities, we use Chung and Pruitts (1994) approximation of Tobins q, the sum of market value of common stock, long- and short-term debt, and preferred stock all divided by total assets. Research has shown that Tobin's q may also proxy for industry characteristics, making it important to adequately control for industry effects. We subtract median Tobin's q ratio from each firms performance measure for firms in the same four-digit SIC code. In banking, numerous studies implement a market-to-book ratio to control for investment opportunity set or market power (Houston and James 1995). Thus, our measure also controls for differing investment opportunity sets. Finally, we include proxies for leverage, age, and size. In Table 4, we show that trade-offs exist for insider holdings. The coefficient on insider holdings is negative and significant when monitoring directors and holdings, as well as equitybased compensation are dependent variables. All IPO firms replace insider holdings with these other monitoring mechanisms. However, firms do not appear to turn to monitoring from larger boards to compensate for lower inside ownership, which may be related to inefficiencies associated with larger boards. Yermack (1996) suggests costs of poor communication and decision-making with a large group outweigh the benefits of increased monitoring potential. Results in Table 4 support the pressure hypothesis; regulated and unregulated firms do not differ in the interdependences of mechanisms (Hypothesis 2B). The interaction term between the regulated dummy (heavily regulated and banks) and insider holdings is insignificant in all models. If regulation substitutes for governance, trade-offs should not exist for regulated firms (positive interaction term). However, we find no differences between regulated and unregulated firms. Using all regulated firms, the interaction term remains insignificant except with incentive

20

compensation.12 Overall, our results support the notion regulation is a complement to traditional monitoring at the IPO and contrasts Booth et al.s (2002) results on mature firms.13 Table 4 also provides support for Hypothesis 1B. With the regulatory pressure hypothesis, monitoring levels should be higher or the same for regulated firms. The coefficients on the heavily regulated dummy variable are positive and significant in the monitoring directors and board size equations and insignificant in the monitoring holdings and equity compensation equations. These results contradict the substitution hypothesis which predicts significantly lower levels of monitoring. If we change our definition to include all regulated or banking firms, we see a negative and marginally significant coefficient only in one equation (monitoring holdings). With all other equations, regulated firms have similar or higher levels of monitoring. 4.3. Deregulation Banks and utilities both experienced significant deregulation in the 1990s (Energy Policy Act of 1992, FIDICIA, Riegle-Neal Act of 1994, additional rules by the SEC and FERC, etc.).14 Kole and Lehn (1997) find that deregulation increases the importance of the managerial role within a firm and thus the need for monitoring. They note that removing regulation increases managerial discretion and thus firm value becomes more sensitive to the quality of managerial decisions. However, Hermalin (1992) proposes that the effects of competition on executive behavior are ambiguous, noting that knowing what an executives preferences are for agency goods (e.g., slacking, perquisites, empire building) is an empirical question.15

While this could suggest substitution for equity-based compensation, further results do not support this claim. Results from separate regressions for all, heavily regulated, banking, and unregulated firms are qualitatively similar. We find similar trade-offs between monitoring mechanisms for both regulated and unregulated firms. 14 See Roengpitya (2007) and Rennie (2006) for detailed discussions of specific deregulatory events and their effects on governance in banking and utility sectors. 15 Cuat and Guadalupe (2009) provide the most direct test of Hermalin (1992); increase in pay sensitivities and the substitution of fixed for variable pay after deregulation are results of increasing competition. These compensation changes would likely be consistent with shareholders providing motivation for executives to consume smaller amounts of agency goods. In addition, a considerable number of empirical studies designed to test the impact of
13

12

21

Cuat and Guadalupe (2009), Rennie (2006) and Becher et al. (2005) document that the deregulation of regulated industries in the 1990s led firms to alter governance structures. Thus, firms going public during a post-deregulatory period may adopt different monitoring structures than firms going public pre-deregulation. In Table 5, we examine regulated firms across the 1990s. The 1993 and 1996 samples (early period) proxy for a time when regulation was stricter, while by 1998 (late period) substantial deregulation occurred for our regulated firms.16 In Hypothesis 3, if regulation substitutes for governance, as regulated firms deregulate monitoring should increase. If regulation complements governance, deregulation should afford regulated firms to reduce monitoring. Results from Table 5 (Panels A and B) indicate monitoring of regulated firms does not significantly increase post-deregulation. Monitoring and outside holdings significantly decline from the early to late period. In addition, outside directors and equity-based compensation (all regulated), as well as VC directors, monitoring directors, and VC holdings (heavy and all regulated) all exhibit insignificant declines. Results for banks are similar to heavily regulated, but significance is lost in a few cases due to sample sizes. Board size is the only mechanism that significantly increases post-deregulation; however, firms do not trade-off board size for insider holdings (Table 4). Our results contradict the substitution hypothesis where deregulation leads to an increased need for monitoring. Table 6 details multivariate analyses of deregulation. For regulated firms, we regress each monitoring mechanism on variables identified previously to control for the need for monitoring plus a binary variable for 1998. Panel A segments heavily regulated firms from all regulated firms. If deregulation results in firms increasing monitoring, the 1998 variable should
competition on governance and performance find that increased competition leads to higher firm growth or productivity (e.g., Januszewski, Koke, and Winter, 2002). 16 Deregulatory results are qualitatively similar if we: (1) exclude 1996 and compare 1993 (early period) versus 1998 (late period), (2) run all analyses on heavily and all regulated firms separately, and (3) include an interaction term between our regulation dummy and post-deregulation (heavy*1998 dummy). 22

be positive and significant (higher monitoring levels post-deregulation). We again do not find support for the substitution hypothesis. In Panels A, the 1998 binary variable is significantly negative in two models (monitoring holdings and directors), suggesting monitoring is lower postderegulation. In Panel B, we include all firms and then utilize a bank dummy and a non-bank regulated dummy. This allows us to consider whether banks are different from other regulated firms and if all firms change over time. The interaction between banking firms and the 1998 binary variable is never significant. Likewise, the interaction using non-bank regulated firms is insignificant in all models except equity-based compensation, where it is negative. The negative sign suggests compensation decreased, which does not support the substitution hypothesis. Overall, Tables 5 and 6 provide evidence regulation is a complement for governance. Relaxing regulation does not appear to cause regulated firms to significantly increase monitoring. 5. Additional specifications 5.1. Post-IPO results Baker and Gompers (2003) suggest governance structures are more likely to be chosen optimally at the IPO. However, a firms governance structure may evolve following the IPO. If regulation substitutes for governance, the dilution in inside ownership around the IPO will not cause a substantial change in monitoring. Only if regulation complements governance will firms respond significantly to this dilution and a trade-off among monitoring mechanisms will occur. As a result, we examine governance structures two years post-IPO (Tables 7 and 8). Data are from proxy statements. The trade-off among monitoring mechanisms is well-established for non-regulated firms (post-IPO) in the literature (Berry et al., 2006), thus we focus on regulated firms.17 To this point, we have focused on governance levels at the IPO year; however, examining trade-offs among changes in monitoring mechanisms may better capture the role of
17

Unreported tests on non-regulated firms generally conform to these results. In addition, we focus on regulated firms instead of just banking firms due to sample sizes two years post-IPO. 23

regulation in governance. Specifically, we examine whether changes in monitoring directors, board size, monitoring holdings, and equity-based compensation are related to the change in inside ownership over this period. We utilize both changes in levels (Year +2 minus IPO Year) as well as percentage changes [(Year +2 IPO Year) / IPO Year]. Table 7 shows that insider holdings decline significantly following the IPO for all regulated and heavily regulated firms. Using changes in levels, Panel A of Table 7 provides support that all regulated firms as well as heavily regulated firms increase both the monitoring directors on their board and their use of equity-based compensation after the IPO. We also find some evidence that monitoring blockholders provide additional governance. In Panel B, we examine the percentage change in these monitoring mechanisms. We continue to document support for our regulatory pressure hypothesis. In particular, when insider holdings decline, firms turn toward more monitoring directors, larger boards, more monitoring blockholders, and increased equity-based compensation. Firms appear to be trading off

mechanisms, which is inconsistent with the substitution hypothesis. Table 8 tests for trade-offs among monitoring mechanisms and insider holdings using levels, change in levels, and percentage changes in a multivariate setting. Heavily regulated and all regulated firms rely more on monitoring directors when inside ownership levels are low and increase monitoring directors when insider holdings decline. For monitoring holdings, the coefficient on insider holdings is negative and significant for all regulated firms using levels, changes, and percentage change. With heavily regulated firms and monitoring holdings, the coefficient on insider ownership is not significant (p-value 0.19), but becomes significant if we remove the venture capitalist effect. For equity-based compensation, we find evidence all regulated firms trade off compensation for inside ownership using levels and percent changes.

24

Note that in the IPO year (Table 4), all regulated firms are less likely to use equity-based compensation when inside ownership is low. Two years post-IPO, however, we see evidence of a trade-off between compensation and insider holdings. Overall, Table 8 shows evidence of tradeoffs in monitoring mechanisms, which is consistent with the regulatory pressure hypothesis To more precisely follow prior literature (e.g., Berry, Fields, and Wilkins, 2006), in unreported results we use panel model regressions for all regulated firms, heavily regulated firms, and banking firms combining the IPO year and Year +2 data. The models implemented are the same models as those in Table 4 with the addition of time controls as well as firm random effects. Since we have time invariant independent variables, a fixed effect model cannot be implemented. The results from these stacked regressions are qualitatively similar to those results reported in Table 4: regulated firms trade-off other monitoring mechanisms for inside ownership. Thus, our evidence remains most consistent with regulation being a complementary force (regulatory pressure) rather than a substitute for governance. As a further robustness, we collect data on our sample firms four years after the IPO (when available). We repeat the above analyses in unreported results. We find results that mirror those in Table 4 (IPO year) and Table 8 (Year +2); trade-offs between inside ownership and monitoring directors and monitoring holdings as well as equity compensation. 5.2. Monitoring systems We have focused on trade offs for inside ownership. Given an IPO is an event designed to significantly alter ownership structures, it is likely that other monitoring mechanisms would be increased at the time of the IPO as a trade off to the reduction in monitoring. However, other interdependences between monitoring mechanisms exist (Berry et al. 2006; Burkart and Panunzi 2006), suggesting a system of equations. Since our previous results show that board size is not an alternative monitoring mechanism for insider holdings, we omit this mechanism to reduce the
25

number of equations and focus on insider holdings, monitoring directors and holdings, and equity-based compensation. Omitting board size is also consistent with Chhaochharia and Laeven (2009), who find board size is not a significant determinant of a firm's overall governance system. We use two methods to address this concern of interdependency. First, we follow Berry et al. (2006), where essentially each monitoring mechanism is treated as a dependent variable. The independent variables include all other monitoring mechanisms, interaction terms with the regulated dummy variable (all and heavily) and each monitoring mechanism, as well as control variables. We again find strong support for our pressure hypothesis. The interaction terms in these model specifications are almost all statistically insignificant. In the few cases where the interaction is significant, the sign supports the pressure not substitution hypothesis, except for one interaction term in one equation. Second, we utilize a simultaneous equations model. Defining an appropriate model is difficult with a large set of monitoring mechanisms. Given all mechanisms may be interrelated, identifying exogenous variables for such a large system is problematic (Himmelberg, Hubbard, and Palia, 1999). Nevertheless, we use first stage regressions to identify control variables (from the set in Table 4) that are statistically significant determinants of each monitoring mechanism. We then use two-stage least squares to estimate the system of equations where each monitoring mechanism has a separate equation and controls are selected based on significance in the first stage. Interaction terms between our regulated dummy variables (all and heavily separately) and each monitoring mechanism are included. The following shows the model specifications:
Monitoring Directors i i 1Monitoring Holdings i 2 Equity Compensati on i 3Inside Holdings i 4 Monitoring Holdings i * Regulated Dummy i 5Equity Compensati on i * Regulated Dummy i 6 Inside Holdings i * Regulated Dummy i 7 Tangible Assets i 8Founder Dummy i 9 VC Dummy i i

26

Monitoring Holdingsi i 1Monitoring Directorsi 2 Equity Compensation i 3Inside Holdingsi 4 Monitoring Directorsi * Regulated Dummyi 5 Equity Compensation i * Regulated Dummyi 6 Inside Holdingsi * Regulated Dummyi 7Sizei 8Leveragei 90 ROA i 10 Founderi 11Adjusted q i 12 VCDummyi 13IPOReturnsi i Equity Compensation i i 1Monitoring Directorsi 2 Monitoring Holdingsi 3Inside Holdingsi 4 Monitoring Directorsi * Regulated Dummyi 5Monitoring Holdingsi * Regulated Dummyi 6 Inside Holdingsi * Regulated Dummyi 7 Leveragei 8Tangible Assetsi 90 Adjusted q i 10 VCDummyi i

Consistent with pressure, interaction terms are insignificant in all specifications. While other monitoring mechanisms may be related, controlling for this does not alter our conclusions. 5.3. Matched sample Our results could be driven by the unbalanced nature of our sample, since the number of unregulated firms is larger than that of heavily regulated firms. To explore this, we construct a matched sample. We match every heavily regulated firm to an unregulated firm based on size (for consistency with past studies that suggest size may be explain differences in monitoring for regulated and unregulated firms). In Table 9, we report summary statistics using heavily regulated firms and their matches, Panel A focuses on our main monitoring measures while Panel B details the components of these measures. The patterns reported mirror those reported in Table 2 and support the regulatory pressure hypothesis. Regulated firms have greater proportions of monitoring directors, larger boards, but less monitoring holdings. Again we find no significant differences in percentages of equity-based compensation. Regression results for the matched sample (Table 10) are directly comparable to those reported in Table 4. The interaction term for heavily regulated firms and insider holdings is again insignificant in all specifications, which supports the pressure hypothesis. We continue to find evidence firms trade-off monitoring directors and monitoring holdings for insider holdings. For

27

board size, we find a positive relation, which may again reflect the inefficiencies of increasing board sizes. In other words, if inside ownership is low, firms may prefer smaller boards. We also run separate models for heavily regulated firms and their matches and obtain qualitatively similar results.18 Our results using the matched sample support the regulatory pressure hypothesis, where firms establish monitoring systems. We find alternative monitoring mechanisms are exchanged for inside ownership at regulated firms and unregulated firms alike. 6. Robustness Additional characteristics may drive differences between regulated and non-regulated firms and failure to control for these characteristics may lead to spurious conclusions. First, regulated firms may face differing political and legal environments, which may shape their boards. To illustrate, regulated firms may be more involved in politics because of regulation, making it optimal for these firms to add outsiders to the board to help manage the political landscape. Agrawal and Knoeber (2001) consider the role of politics and board structure. One measure implemented is the number of directors on the board with political and legal ties. The authors contend that if political pressures are high, a firm will appoint more politically connected directors, where politically connected directors include lawyers as well as directors with government work experience. Such a hypothesis provides a plausible explanation for results documented in the existing literature that are contrary to the substitution hypothesis, such as regulated firms having larger and more independent boards. To explore whether political pressures affect our results, we follow Agrawal and Knoeber (2001) and use director background information for nearly 3,000 individual directors for our 1998 sample. Analyses are limited to 1998 because reporting of past director experiences

18

To be consistent with Table 4, we repeat all these tests including IPO returns or year dummies with qualitatively similar results. We omit these in Table 8 because of our relatively small sample sizes. 28

dramatically improved in the 1990s as well as Agrawal and Knoeber (2001) find the number of political and legal directors increases over the 1990s for electric utility companies. We document a limited role for political directors at regulated and unregulated firms; 2% of executives and directors have political backgrounds (lawyer, regulatory employee, political officer at any observed point in their career).19 Limiting to board members only, we have 4% political directors (it is negligible if we require directors to currently be political). Further, no significant difference exists between regulated and unregulated firms (p-value of difference 0.33). For regulated firms, the mean (median) is 5% (0%) compared to 3.4% (0%) for unregulated firms. It does not appear differing political environments drive our results or that politics, in general, are likely a significant determinant of board members for IPO firms. While our analyses focus on inside ownership, venture capitalists often serve on the board and disentangling their ownership is difficult. We implement several methods to control for this. First, in multivariate analyses, we include a dummy variable for venture-backed IPOs. Second, for the 1998 sample we separate venture capitalists from insiders for our 3,000 directors. The correlation between inside ownership with and without venture capitalists is extremely high. However, we encounter difficulties splitting ownership. In many cases, total director and officer ownership does not equal the sum of ownership listed by these individuals. To reconcile ownership, we must create plug values for 25% of our firms. The average plug size is nontrivial: 184,605 shares (3.71%) pre-IPO and 186,396 shares (4.75%) post-IPO (maximum 1,309,238 shares or 44.78%). Given these uncertainties, we do not focus on these numbers. Third, we repeat analyses for non venture-backed IPOs only. Results are qualitatively similar, suggesting our findings are not biased by the inclusion of venture capitalists on the board.
19

It is still possible executives and directors were politically involved at some point in their careers, but it was not reported in the proxy. Focusing on a sample on the late 1990s likely diminishes this concern as reporting was much more detailed and all results hold if we base definitions on current profession. 29

Our measure of director and officer ownership may double count independent outsiders ownership if they own a significant stake in the firm (unaffiliated blockholder). We implement several measures to control for this potential bias. First, outside blockholdings appear similar at both regulated and unregulated firms; suggesting both groups would experience similar biases (if one exists). Second, for the 1998 sample we collect data on the 3,000 directors and document when independent outside directors are associated with outside blockholders. Only 50 outside directors at 28 firms (total includes 606 independent outside directors at 281 firms) are affiliated with outside blockholders. For 30 of these directors, no double counted shares exist; individual shares are zero or the proxy indicates shares are independent from those of the blockholder. Double counting could only exist for 20 directors at 17 firms. Third, segmenting ownership is difficult; the correlation between inside ownership with and without double counted shares is 0.94. Finally, we repeat all analyses for IPOs with no outside blockholdings. While this is a severe restriction, results are qualitatively similar. Overall, our measure of inside ownership does not appear biased by independent outside directors having a large stake in the firm. Tables 9 and 10 compare the heavily regulated sample to a set of matched firms based on asset size. Given the skewness in size for regulated firms, we also match on market value of equity; all results are qualitatively similar. To examine if leverage has a nonlinear relationship with our dependent variables we include a squared term in all tests. The squared term, however, is never significant and has no effect on results. For our unregulated firms, we include Fama French industry dummies (17 industries). The results remain unchanged as we continue to find trade-offs between monitoring mechanisms. Additionally, analyses in Tables 1, 2, 3, 5, and 9 are repeated testing for differences in medians rather than means. Results are qualitatively similar.

30

7. Conclusion Monitoring mechanisms are costly to adopt. In regulated industries, executive decisionmaking is more transparent and opportunity sets are limited, suggesting governance mechanisms may be less important. The literature to date has focused largely on the notion that regulation substitutes for governance. However, the empirical evidence is mixed. In this paper, we propose an alternative explanation for the relation between regulation and governance. Specifically, we contend that the presence of regulators may pressure firms to adopt effective corporate governance structures that promote safety and soundness. We examine governance structures at regulated and unregulated firms at the time of their IPO, when governance is likely to be optimal (Baker and Gompers 2003). Contrary to the substitution hypothesis, we find that regulated firms do not have significantly lower monitoring. These firms have greater proportions of monitoring directors, similar levels of equity-based compensation, and larger boards. These results support Adams and Ferreira (2008) contention that board oversight appears to be a complement to regulation. Further, we do not find evidence to suggest that firm characteristics such as leverage, size, or age proxy for regulation. In a multivariate setting, we again find support for regulatory pressure. Specifically, at both regulated and unregulated firms, trade-offs exist between traditional monitoring mechanisms and inside ownership, which is consistent with regulation serving to pressure firms to develop a system of governance. If regulation substitutes for governance, the degree of interdependencies would be less or not present at regulated firms. Finally, we examine if deregulation impacts regulated governance structures (Kole and Lehn 1997). Deregulation increases competition, opportunity sets, and sensitivity to managerial decisions. With substitution, deregulation would reduce monitoring, forcing firms to strengthen

31

their governance structures to more closely resemble those of unregulated firms. In contrast, the regulatory pressure hypothesis predicts a likely decrease in monitoring following deregulation due to the removal of pressure. Again, our results support the regulatory pressure hypothesis. Post-deregulation, governance structures of regulated firms significantly decrease in several cases. Monitoring does not increase as the substitution hypothesis would predict. Our results suggest that regulation and governance are complements where regulators may pressure firms to adopt effective monitoring structures. The pressure hypothesis provides an explanation for some puzzling empirical findings in the literature. This paper also has implications for the governance literature in general. Essentially, we examine whether firms utilize governance systems and high monitoring mechanisms when information asymmetry and managerial discretion are limited. Given that such monitoring is costly, we would expect firms to use less (none) if such monitoring were not important. However, our results are not consistent with substitution, implying governance systems are important to shareholders. Regulation does not replace traditional monitoring. Finally, our results provide support for the findings in Wang, Winton, and Yu (2009); regulators play an important role even in the presence of monitoring.

32

References Adams, R., Ferreira, D, 2008. Regulatory pressure and bank directors incentives to attend board meetings. Working paper, European Corporate Governance Institute. Adams, R., Mehran, H., 2003. Is corporate governance different for bank holding companies? FRBNY Economic Policy Review, 123-142. Adams, R., Mehran, H., 2009. Corporate performance, board structure and its determinants in the banking industry. Journal of Financial Intermediation, forthcoming. Agrawal, A., Knoeber, C.R., 1996. Firm performance and mechanisms to control agency problems between managers and shareholders. Journal of Financial and Quantitative Analysis 31, 377-397. Agrawal, A., Knoeber, C.R., 2001. Do some outside directors play a political role? Journal of Law and Economics 44, 179-198. Baker, M., Gompers, P.A., 2003. The determinants of board structure at the initial public offering. Journal of Law and Economics 46, 569-598. Barry, C., Muscarella, C., Peavy, J., Vetsuypens, M.R., 1990. The role of venture capital in the creation of public companies. Journal of Financial Economics 27, 447-471. Bartov, E., Mohanram, P., Seethamraju, C., 2002. Valuation of internet stocks: perspective. Journal of Accounting Research 40, 321-346. An IPO

Beatty, R.P., Zajac, E.J., 1994. Managerial incentives, monitoring, and risk bearing: A study of executive compensation, ownership and board structure in initial public offerings. Administrative Science Quarterly 39, 313-335. Becher, D.A., Campbell, T.L., Frye, M.B., 2005. Incentive compensation for bank directors: The impact of deregulation. Journal of Business 78, 1753-1777. Berry, T.L., Fields, P., Wilkins, M.S., 2006. The interaction among multiple governance mechanisms in young newly public firms. Journal of Corporate Finance 12, 449-466. Booth, J.R., Cornett, M.M., Tehranian, H., 2002. Boards of directors, ownership, and regulation. Journal of Banking and Finance 26, 1973-1996. Brown, K.C., Dittmar, A., Servaes, H., 2005. Corporate governance, incentives, and industry consolidation. Review of Financial Studies 18, 241-270. Burkart, M., Panunzi,F., 2006. Agency conflicts, ownership concentration, and legal shareholder protection. Journal of Financial Intermediation 15, 1-31. Burton, B., Helliar, C.V., Power, D.M., 2004. The benefits and costs of deeply-discounted rights issues - practitioners' viewpoints. Applied Economics Letters 11, 369-372.
33

Caprio, G., Laeven, L., Levine, R., 2007. Governance and bank valuation. Journal of Financial Intermediation 16, 584-617. Chen, H., Ritter, J., 2000. The seven percent solution. Journal of Finance 55, 1105 - 1131. Chhaochharia, V., Laeven, L., 2009. Corporate governance norms and practices. Journal of Financial Intermediation 18, 405-431. Chung, K.H., Pruitt, S.W., 1994. A simple approximation of Tobins q. Financial Management 23, 70-74. Cuat, V., Guadalupe, M., 2009. Executive compensation and competition in the banking and financial sectors. Journal of Banking and Finance 33, 495-504. Crawford, A., Ezzell, J.R., Miles, J.A., 1995. Bank CEO pay-performance relations and the effects of deregulation. Journal of Business 68, 231-256. Engel, E., Gordon, E.A., Hayes, R.M., 2002. The roles of performance measures and monitoring in annual governance decisions in entrepreneurial firms. Journal of Accounting Research 40, 285527. Feng, Z., Ghosh, C., Sirmans, C.F., 2007. CEO involvement in director selection: Implications for REIT dividend policy. Journal of Real Estate Finance and Economics 35, 385-410. Friedlaender, A.F., Berndt, E.R., Chiang, J., Showalter, M., Vellturo, C.A., 1993. Rail costs and capital adjustments in a quasi-regulated environment. Journal of Transport Economics and Policy 27, 131-152. Friesen, G.C. Swift, C., 2009. Overreaction in the thrift IPO aftermarket. Journal of Banking and Finance 33, 1285-1298. Gompers, P.A., 1995. Optimal investment, monitoring, and the staging of venture capital. Journal of Finance 50, 1461-1489. Gompers, P.A., Ishii, J., Metrick, A., 2003. Corporate governance and equity prices. The Quarterly Journal of Economics 118, 107-155. Hadlock, C., Lee, S., Parrino, R., 2002. CEO careers in regulated environments: Evidence from electric and gas utilities. Journal of Law and Economics 45, 535-563. Hartzell, J., Kallberg, J., Liu, C., 2008. The role of corporate governance in initial public offerings: Evidence from real estate investment trusts. Journal of Law and Economics 51, 539562. Hermalin, B.E., 1992. The effects of competition on executive behavior. Rand Journal of Economics 23, 350365.
34

Higgins, M.C., Gulati, R., 2006. Stacking the deck: The effects of top management backgrounds on investor decisions. Strategic Management Journal 27, 1-25. Himmelberg, C.P., Hubbard, R.G., Palia, D., 1999. Understanding the determinants of managerial ownership and the link between ownership and performance. Journal of Financial Economics 53, 353-384. Houston, J.F., James, C., 1995. CEO compensation and bank risk: Is compensation in banking structured to promote risk taking? Journal of Monetary Economics 36, 405-431. Januszewski, S.I., Koke, J., Winter, J.K., 2002. Product market competition, corporate governance and firm performance: An empirical analysis for Germany. Research in Economics. 56, 299-332. Jensen, M.C., Meckling, W.H., 1976. Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305-360. Jensen, M.C., Murphy, K.J., 1990. Performance pay and top-management incentives. Journal of Political Economy 98, 225-264. John, K., Saunders, A., Senbet, L.W., 2000. A theory of bank regulation and management compensation. Review of Financial Studies 13, 95125. John, K., Senbet, L.W., 1998. Corporate governance and board effectiveness. Journal of Banking and Finance 22, 371-403. Joskow, P.L., Rose, N.L., Shepard, A., 1993. Regulatory constraints on CEO compensation. Brookings Papers on Economic Activity, Microeconomics, 1-72. Joskow, P.L., Rose, N.L., Wolfram, C.D., 1996. Political constraints on executive compensation: Evidence from the electric utility industry. RAND Journal of Economics, 27, 165-182. Kole, S., Lehn, K., 1997. Deregulation, the evolution of corporate governance structure, and survival. American Economic Review 87, 421-425. Kole, S., Lehn, K., 1999. Deregulation and the adaptation of governance structure: The case of the U.S. airline industry. Journal of Financial Economics 52, 79-117. Lehn, K., 2002. Corporate governance in the deregulated telecommunications industry: Lessons from the airline industry. Telecommunications Policy 26, 225242. Ljungqvist, A., Wilhelm. W.J., 2003. IPO pricing in the dot-com bubble. Journal of Finance 58, 723-752. Lowry, M.B., Schwert, G.W., 2002. IPO market cycles: Bubbles or sequential learning?, Journal of Finance, 57, 1171-1200.
35

Masulis, R., Thomas, R., 2008. Does private equity create wealth? The effects of private equity and derivatives on corporate governance. University of Chicago Law Review 76, 219-259. Mehran, H., 1995. Executive compensation structure, ownership, and firm performance. Journal of Financial Economics 38, 163-184. Mulherin, J.H., 2007. Measuring the costs and benefits of regulation: Conceptual issues in securities markets. Journal of Corporate Finance 13, 421-437. Noreen, Eric and Mark Wolfson, 1981, Equilibrium warrant pricing models and accounting for executive stock options, Journal of Accounting Research 19, 384-398. Pathan, S., Skully, M., 2010. Endogenously structured boards of directors in banks. Journal of Banking and Finance 34, 1590-1606. Rennie, C.G., 2006. Governance structure changes and product market competition: Evidence from U.S. electric utility deregulation. Journal of Business 79, 1989-2017. Roengpitya, R., 2007. The effects of financial deregulation on bank governance: The panel data evidence of the 1990s, Working paper, University of Chicago. Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. Journal of Finance 52, 737783. Stigler, G.J., Friedland, C., 1962. What can regulators regulate? The case of electricity. Journal of Law and Economics 5, 1-16. Suchard, J., 2009. The impact of venture capital backing on the corporate governance of Australian initial public offerings. Journal of Banking and Finance 33, 765774. Wang, T.Y., Winton, A., Yu, X., 2009. Corporate fraud and business conditions: Evidence from IPOs. Journal of Finance, forthcoming. Yermack, D., 1996. Higher market valuation of companies with a small board of directors. Journal of Financial Economics 40, 185-211.

36

Table 1 Firm characteristics. This table reports means (medians) for firm characteristics for all regulated, heavily regulated, banks, and unregulated firms. The p-value reports the significance of the difference between the sample means. Leverage is defined as total debt divided by total assets. Age is the number of years from the first date of incorporation until the IPO. Assets are total firm assets reported in dollar millions. Tangible assets are the ratio of tangible to total assets. Founder is a binary variable equal to one if a founding family member is present at the IPO. Float is the number of shares outstanding minus inside holdings. Adjusted q is Chung and Pruitts (1994) approximation of Tobins q, adjusted for the industry median. The sample contains 928 unregulated, 233 regulated, 58 heavily regulated, and 37 banking firms. All Regulated includes partially and heavily regulated firms. The p-value reports the significance of the difference between the sample means. Float Adjusted Tangible Assets ROA Founder (millions) Age Leverage q Assets (millions) All Regulated (1) 36.50% (28.57%) Heavily Regulated (2) Banks (3) Unregulated (4) 63.23% (80.80%) 85.22% (87.41%) 18.26% (7.69%) 0.00 0.00 0.00 1,122.25 (201.34) 533.00 (183.68) 707.58 (292.56) 159.45 (51.96) 0.00 0.00 0.00 19.02 (8.00) 22.05 (11.00) 25.70 (14.00) 14.49 (8.00) 0.00 0.01 0.00 73.39% (98.69%) 70.87% (99.84%) 59.44% (98.53%) 74.51% (99.24) 0.69 0.49 0.02 48.67% (0.00) 32.14% (0.00) 21.62% (0.00) 54.18% (1.00) 0.13 0.00 0.00 21.62 (6.02) 9.62 (3.78) 4.05 (2.32) 8.88 (4.94) 0.00 0.70 0.04 1.64 (0.14) 0.17 (0.01) 0.01 (-0.01) 1.80 (0.62) 0.80 0.02 0.04 6.44% (6.37%) 5.25% (2.60%) 2.20% (2.31%) 3.56% (10.98%) 0.17 0.67 0.76

p-value (1 vs.4) p-value (2 vs. 4) p-value (3 vs. 4)

37

Table 2 Monitoring intensity. This table reports means of monitoring mechanisms for all regulated, heavily regulated, banks, and unregulated firms. Panel A details our main monitoring measures (monitoring directors and holdings, board size, and equity-base compensation) while Panel B provides the component monitoring measures. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity-based compensation is the average percentage of compensation that is equity-based for the top executives. Insider holdings is director and officer total holdings. The sample contains 928 unregulated, 233 regulated, 58 heavily regulated, and 37 banking firms. All Regulated includes partially and heavily regulated firms. The p-value reports the significance of the difference between the sample means. Panel A: Main Monitoring Measures Monitoring Directors All Regulated (1) Heavily Regulated (2) Banks (3) Unregulated (4) p-value (1 vs.4) p-value (2 vs. 4) p-value (3 vs. 4) 52.07% 57.32% 57.66% 50.30% 0.31 0.03 0.07 Board Size 7.54 9.27 10.49 6.28 0.00 0.00 0.00 Monitoring Holdings 23.11% 15.50% 4.61% 24.14% 0.61 0.01 0.00 Equity Comp 16.62% 13.27% 9.91% 14.44% 0.19 0.70 0.21

Panel B: Component Monitoring Measures Outside Directors All Regulated (1) Heavily Regulated (2) Banks (3) Unregulated (4) p-value (1 vs.4) p-value (2 vs. 4) p-value (3 vs. 4) 44.00% 54.11% 57.66% 33.24% 0.00 0.00 0.00 VC Directors 8.08% 3.21% 0.00% 17.07% 0.00 0.00 0.00 Outside Blockholdings 14.95% 11.51% 3.00% 9.80% 0.00 0.53 0.03 VC Holdings 8.16% 3.98% 1.61% 14.34% 0.00 0.00 0.00 Insider Holdings 33.01% 33.55% 36.33% 41.85% 0.00 0.01 0.17

38

Table 3 Firm characteristics and monitoring intensity. This table reports means for monitoring mechanisms based on whether various firm characteristics were high, medium, or low. In Panel A, the sample is divided by leverage (total debt divided by total assets). In Panel B, the sample is divided by size (natural log of total assets). In Panel C, the sample is divided by age (number of years from incorporation until IPO). The p-value reports the significance of the difference between sample means. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity-based compensation is the average percentage of incentive compensation for the top executives. The sample consists of 928 are unregulated firms, 233 regulated, 58 heavily regulated, and 37 banking firms. The p-value reports the significance of the difference between the sample means. Monitoring Monitoring Equity Directors Holdings Board Size Comp Panel A: Leverage High (1) Medium (2) Low (3) p-value (1 vs. 2) p-value (2 vs. 3) p-value (1 vs. 3) Panel B: Size High (1) Medium (2) Low (3) p-value (1 vs. 2) p-value (2 vs. 3) p-value (1 vs. 3) Panel C: Age High (1) Medium (2) Low (3) p-value (1 vs. 2) p-value (2 vs. 3) p-value (1 vs. 3) 47.19% 52.91% 51.62% 0.00 0.45 0.01 6.35 6.49 6.76 0.43 0.20 0.06 22.50% 24.58% 24.59% 0.26 0.99 0.30 12.06% 16.07% 16.56% 0.01 0.77 0.01 51.26% 50.26% 50.35% 0.55 0.96 0.60 7.43 6.28 5.82 0.00 0.00 0.00 29.60% 22.49% 19.32% 0.00 0.06 0.00 15.91% 15.18% 13.50% 0.66 0.29 0.14 53.75% 49.84% 48.28% 0.02 0.39 0.00 6.55 6.08 6.93 0.01 0.00 0.09 24.28% 23.67% 23.39% 0.74 0.89 0.64 17.56% 14.28% 12.81% 0.05 0.36 0.00

39

Table 4 Substitution versus pressure tests: Interaction terms. This table reports results from regression analysis using robust standard errors with the monitoring mechanisms as the dependent variables. All Regulated Dummy equals one if the firm is heavily or partially regulated. Heavily Regulated dummy equals one if the firm is heavily regulated. Regulation Dummy * Insider Holdings are interaction terms between the regulation dummy variables and insider holdings. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity-based compensation is the average percentage of incentive compensation for the top executives. Robust standard errors are in parentheses. Regulated = Regulated = Regulated = Regulated = Regulated = Regulated = All Heavy Bank All Heavy Bank Monitoring Directors Constant Insider Holdings Size Leverage ROA Ln(age) Tangible Assets Founder dummy Float Adjusted q VC Dummy IPO Returns Regulated Dummy Regulated Dummy * Insider Holdings Non-bank Regulated Dummy Non-bank Regulated * Insider Holdings Observations p-value R2 0.443*** (0.076) -0.195*** (0.030) 0.007 (0.006) -0.047 (0.030) -0.016 (0.028) -0.010 (0.007) 0.083*** (0.026) -0.036*** (0.014) -0.006* (0.003) 0.001 (0.001) 0.193*** (0.013) 0.087 (0.298) 0.029 (0.021) 0.032 (0.021) 0.384*** (0.086) -0.197*** (0.033) 0.013** (0.006) -0.079** (0.031) -0.023 (0.028) -0.004 (0.008) 0.099*** (0.029) -0.036** (0.015) -0.002 (0.003) -0.000 (0.001) 0.201*** (0.014) 0.158 (0.334) 0.128** (0.051) 0.018 (0.118) 0.463*** (0.072) -0.200*** (0.031) 0.007 (0.007) -0.086*** (0.027) -0.013 (0.026) -0.011 (0.007) 0.083*** (0.026) -0.031** (0.014) -0.003 (0.003) 0.001 (0.001) 0.196*** (0.013) 0.006 (0.288) 0.196** (0.082) 0.030 (0.186) -0.015 (0.031) 0.089 (0.067) 1,072 0.00 25.31% 5.289*** (0.847) 0.215 (0.340) 0.496*** (0.083) 0.318 (0.375) -1.372*** (0.269) -0.137 (0.088) -0.783** (0.322) -0.425*** (0.160) 0.004 (0.031) 0.018*** (0.006) 0.955*** (0.152) 9.326*** (3.128) 0.648** (0.269) -0.056 (0.231) Board Size 3.536*** (0.839) 0.207 (0.353) 0.428*** (0.073) 0.156 (0.333) -1.227*** (0.250) -0.184** (0.084) -0.764** (0.329) -0.401** (0.161) 0.042 (0.031) 0.031** (0.012) 0.892*** (0.151) 9.825*** (3.144) 1.477** (0.700) 2.775 (2.275) 3.597*** (0.842) 0.314 (0.368) 0.487*** (0.066) -0.345 (0.321) -1.253*** (0.301) -0.157* (0.081) -0.793*** (0.299) -0.315** (0.159) 0.033 (0.040) 0.018** (0.009) 1.012*** (0.156) 7.719** (3.374) 2.600*** (0.960) 2.934 (2.183) 0.531 (0.358) -1.198 (0.786) 1,072 0.00 22.00%

1,072 0.00 24.92%

927 0.00 27.74%

1,072 0.00 19.27%

927 0.00 23.47%

40

Table 4 (continued). Regulated = All Constant Insider Holdings Size Leverage ROA Ln(age) Tangible Assets Founder dummy Float Adjusted q VC Dummy IPO Returns Regulated Dummy 0.209*** (0.076) -0.266*** (0.035) 0.043*** (0.006) -0.046* (0.027) -0.042* (0.022) -0.010 (0.008) -0.025 (0.026) -0.079*** (0.013) 0.006** (0.003) 0.003*** (0.001) 0.197*** (0.014) -0.488* (0.296) -0.043* (0.025) 0.008 (0.029) Regulated = Heavy Monitoring Holdings 0.138* (0.079) -0.238*** (0.039) 0.046*** (0.007) -0.047 (0.029) -0.049** (0.023) -0.009 (0.009) 0.037 (0.026) -0.076*** (0.014) 0.007** (0.003) 0.002 (0.001) 0.202*** (0.015) -0.298 (0.309) -0.038 (0.092) -0.110 (0.212) 0.189*** (0.032) -0.243*** (0.006) 0.042*** (0.006) -0.010 (0.028) -0.044* (0.026) -0.010 (0.007) 0.025 (0.026) -0.083*** (0.014) 0.004 (0.003) 0.003*** (0.001) 0.196*** (0.013) -0.423 (0.289) -0.146* (0.082) -0.088 (0.187) 0.013 (0.031) -0.087 (0.067) 1,072 0.00 36.59% 0.181** (0.077) -0.139*** (0.029) 0.011* (0.006) -0.072*** (0.025) -0.028 (0.026) -0.012* (0.008) -0.063** (0.028) 0.010 (0.014) 0.005* (0.003) 0.003*** (0.001) 0.042*** (0.014) 0.182 (0.292) -0.010 (0.023) 0.068*** (0.019) Regulated = Bank Regulated = All Regulated = Heavy Equity-based Compensation 0.177** (0.082) -0.169*** (0.033) 0.011 (0.007) -0.090*** (0.027) -0.025 (0.027) -0.007 (0.008) -0.076** (0.030) 0.014 (0.015) 0.003 (0.003) 0.005*** (0.001) 0.037** (0.015) 0.293 (0.302) -0.039 (0.061) 0.209 (0.142) 0.186** (0.075) -0.148*** (0.032) 0.011* (0.006) -0.068** (0.028) -0.029 (0.026) -0.013* (0.007) -0.065** (0.026) 0.010 (0.014) 0.005 (0.003) 0.003*** (0.001) 0.043*** (0.014) 0.179 (0.298) -0.058 (0.086) 0.163 (0.200) -0.019 (0.032) 0.095 (0.070) 1,062 0.00 7.50% Regulated = Bank

Regulated Dummy * Insider Holdings Non-bank Regulated Dummy Non-bank Regulated * Insider Holdings Observations 1,072 p-value 0.00 R2 36.54% * Statistical significance at the 10% level. ** Statistical significance at the 5% level. *** Statistical significance at the 1% level.

927 0.00 37.13%

1,062 0.00 9.01%

919 0.00 10.20%

41

Table 5 The effect of deregulation. This table reports mean levels of monitoring mechanisms for all regulated, heavily regulated and banking firms. The first row for each regulated firm type reports data for the early-period (1993 and 1996) IPOs and the second row reports for the late period (1998) sample. Panel A details our main monitoring measures (monitoring directors and holdings, board size, and equity-base compensation) while Panel B provides the component monitoring measures. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity-based compensation is the average percentage of incentive compensation for the top executives. In the early (late) period, there are 173 (60) regulated and 42 (16) heavily regulated firms. P-values report significance of difference between sample means. Panel A: Main Monitoring Measures Monitoring Directors All regulated Early Late p-value Heavily regulated Early Late p-value Banks Early Late p-value 53.26% 48.21% 0.16 57.69% 56.32% 0.82 58.57% 56.32% 0.72 9.14 12.47 0.05 8.10 12.47 0.00 7.66% 0.14% 0.18 Board Size 7.07 9.06 0.00 21.12% 0.14% 0.01 6.78% 14.29% 0.24 Monitoring Holdings 26.76% 11.20% 0.00 12.90% 14.29% 0.83 Equity Comp 16.90% 15.76% 0.76

Panel B: Component Monitoring Measures Outside Directors All regulated Early Late p-value Heavily regulated Early Late p-value Banks Early Late p-value 44.91% 41.01% 0.29 53.31% 56.32% 0.62 58.57% 56.32% 0.72 0.00% 0.00% n/a 4.38% 0.00% 0.20 4.95% 0.14% 0.28 VC Directors 8.34% 7.20% 0.66 15.68% 0.14% 0.03 2.71% 0.00% 0.42 Outside Blockholdings 18.31% 3.97% 0.00 5.44% 0.00% 0.15 36.61% 35.94% 0.92 VC Holdings 8.45% 7.23% 0.68 32.67% 35.94% 0.62 Insider Holdings 31.93% 36.51% 0.26

42

Table 6 Multivariate analyses of deregulation. This table reports results from regression analysis using robust standard errors with four monitoring mechanisms as the dependent variable. Panel A segments heavily regulated from all regulated, while Panel B details banks versus all regulated. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity compensation is the average percentage of incentive compensation for top executives. Size is the natural log of total assets. Leverage is total debt divided by assets. Age is the number of years from date of incorporation until the IPO. Tangible assets are the ratio of tangible to total assets. Founder is a binary variable equal to one if a founding family member is present at the IPO. Float is the number of shares outstanding minus inside holdings. Adjusted q is Chung and Pruitts (1994) approximation of Tobins q, adjusted for the industry median. The sample includes 233 regulated firms (partially and heavily regulated firms). Robust standard errors are in parentheses. Panel A: Heavily Regulated versus All Regulated Firms
Monitoring Directors 0.695*** (0.091) 0.004 (0.012) -0.058 (0.058) -0.180* (0.099) -0.043*** (0.016) -0.096 (0.081) -0.044 (0.032) -0.011* (0.006) 0.003*** (0.001) 0.138*** (0.033) 0.120*** (0.039) -0.127* (0.079) 0.00 22.87% Board Size 4.902*** (1.718) 0.735*** (0.269) -0.370 (1.070) -1.318 (1.047) -0.193 (0.300) -1.669 (1.626) -0.570 (0.511) -0.189 (0.116) 0.015* (0.009) 1.440** (0.559) 2.253*** (0.597) 0.521 (1.560) 0.00 22.98% Monitoring Holdings 0.232 (0.147) 0.039** (0.017) -0.101 (0.074) -0.037 (0.101) -0.010 (0.020) -0.135 (0.124) -0.118*** (0.040) 0.017 (0.014) 0.003*** (0.001) 0.170*** (0.049) -0.027 (0.049) -0.262** (0.112) 0.00 30.16% Equity-based Compensation 0.199* (0.110) 0.007 (0.014) 0.012 (0.060) 0.063 (0.079) -0.032** (0.016) -0.032 (0.094) -0.022 (0.038) 0.011 (0.009) 0.001* (0.001) 0.088** (0.041) -0.004 (0.041) -0.056 (0.085) 0.11 8.64%

Constant Size Leverage ROA Ln(age) Tangible Assets Founder dummy Float Adjusted q VC Dummy Heavily Regulated Dummy 1998 Dummy

p-value R2

43

Table 6 (continued) Panel B: Banks versus All Firms

Monitoring Directors Constant Size Leverage ROA Ln(age) Tangible Assets Founder dummy Float Adjusted q VC Dummy 1998 Dummy Bank Dummy Non-bank Regulated Dummy Bank Dummy * 1998 Dummy Non-bank Regulated * 1998 Dummy p-value R2 0.494*** (0.051) 0.012* (0.006) -0.100*** (0.030) -0.026 (0.027) -0.015** (0.007) -0.033 (0.041) -0.052*** (0.013) 0.001 (0.003) 0.001 (0.001) 0.200*** (0.013) -0.138*** (0.037) 0.190*** (0.051) 0.015 (0.021) 0.086 (0.071) 0.059 (0.045) 0.00 20.82% Board Size 4.575*** (0.626) 0.491*** (0.084) -0.318 (0.347) -1.334*** (0.269) -0.156* (0.086) -0.483 (0.546) -0.322** (0.145) 0.043 (0.031) 0.021*** (0.007) 0.978*** (0.149) 0.762 (0.496) 2.971*** (0.910) 0.184 (0.239) 1.725 (1.775) -0.303 (0.818) 0.00 22.86% Monitoring Holdings 0.147** (0.061) 0.048*** (0.007) -0.030 (0.033) -0.060** (0.024) -0.013 (0.008) -0.105** (0.051) -0.113*** (0.014) 0.010*** (0.003) 0.003*** (0.001) 0.200*** (0.014) -0.161*** (0.048) -0.167*** (0.050) 0.009 (0.026) -0.010 (0.051) -0.069 (0.043) 0.00 21.24% Equity-based Compensation 0.173*** (0.058) 0.012* (0.006) -0.065** (0.027) -0.033 (0.026) -0.015** (0.008) -0.083* (0.047) -0.006 (0.014) 0.007*** (0.003) 0.003*** (0.001) 0.045*** (0.014) 0.001 (0.044) 0.007 (0.042) 0.045* (0.023) -0.005 (0.081) -0.108** (0.046) 0.00 7.48%

* Statistical significance at the 10% level. ** Statistical significance at the 5% level. *** Statistical significance at the 1% level.

44

Table 7 Changes in monitoring. This table reports mean and median monitoring mechanisms for all regulated and heavily regulated firms. Panel A details changes in the levels of our main monitoring measures (monitoring directors and holdings, board size, and equity-base compensation) while Panel B provides percentage change. Changes in levels are calculated as Year +2 values minus IPO Year values. Percentage changes are calculated as [(Year +2 IPO Year) / IPO Year]. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equitybased compensation is the average percentage of compensation that is equity-based for the top executives. Insider holdings is director and officer total holdings. The sample contains 233 regulated and 58 heavily regulated firms. All Regulated includes partially and heavily regulated firms. Panel A: Changes in Levels Change in Monitoring Directors All regulated Mean Median Heavily regulated Mean Median 5.28%*** 0.00%*** 5.69%** 1.52%** 0.17 0.00 Change in Board Size Change in Monitoring Holdings 1.70% 0.00%** -0.45% 0.00% Change in Equity Comp 14.83%*** 10.68%*** 13.25%*** 11.32%*** Change in Insider Holdings -7.48%*** -4.53%*** -5.91%** -4.20%**

0.12 0.00*

Panel B: Percent Changes % Change in Monitoring Directors All regulated Mean Median Heavily regulated Mean Median 14.27%*** 0.00%*** 15.73%*** 1.85%** 10.89%** 0.00% % Change in Board Size 15.32%*** 0.00%*** -10.80%** 0.00%** % Change in Monitoring Holdings 12.93% 0.00%* % Change in Equity Comp 113.13%* 0.00%* 230.89% 0.00% % Change in Insider Holdings 21.71% -13.54%*** -6.19% -12.05%**

* Statistical significance at the 10% level. ** Statistical significance at the 5% level. *** Statistical significance at the 1% level.

45

Table 8 Year + 2 for regulated firms. This table reports results from regression analysis using robust standard errors with four monitoring mechanisms as the
dependent variable. The sample includes levels for all regulated and heavily regulated firms as well as changes and percent change. Data are for two years after the IPO year. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity-based compensation is the average percentage of incentive compensation for top executives. Leverage is total debt divided by assets. Age is the number of years from date of incorporation until the IPO. Tangible assets are the ratio of tangible to total assets. Founder is a binary variable equal to one if a founding family member is present at the IPO. Adjusted q is Chung and Pruitts (1994) approximation of Tobins q, adjusted for the industry median. Robust standard errors are in parentheses. Monitoring Directors Board Size All Heavily All All All Heavily All All Regulated Regulated Regulated: Regulated: Regulated Regulated Regulated: Regulated: Changes % Change Changes % Change Constant 0.643*** 1.228*** -0.432*** -0.438* 0.948 9.440 0.390 0.260 (0.153) (0.452) (0.122) (0.260) (2.037) (9.197) (1.864) (0.484) Insider Holdings -0.135* -0.361** -0.147* -0.023* 0.604 0.554 -0.050 0.022 (0.073) (0.175) (0.077) (0.014) (1.096) (2.024) (1.514) (0.020) Size -0.002 -0.031 -0.028 0.142 0.679*** 0.668 0.485 0.101 (0.010) (0.029) (0.025) (0.277) (0.154) (0.469) (0.396) (0.223) Leverage -0.026 0.146 0.036 0.002 1.306 5.102** -0.257 0.002 (0.053) (0.134) (0.074) (0.006) (0.931) (2.141) (1.341) (0.003) ROA -0.024 -0.003 0.056** 0.001** -0.236 1.722 -1.429*** 0.001* (0.023) (0.051) (0.025) (0.000) (0.461) (1.199) (0.517) (0.001) Ln(age) -0.027 -0.020 0.062*** 0.022 -0.718*** -0.823 0.596* 0.055 (0.019) (0.035) (0.021) (0.112) (0.261) (0.739) (0.349) (0.058) Tangible Assets 0.047 -0.404* -0.093 -0.001 2.747** 0.437 1.715 -0.000 (0.090) (0.226) (0.060) (0.001) (1.077) (5.625) (1.167) (0.000) Founder dummy -0.018 0.034 0.009 -0.006 -0.368 -0.069 1.236* 0.052 (0.034) (0.062) (0.032) (0.131) (0.548) (1.371) (0.651) (0.129) Float -0.003 -0.012 0.014** 0.039* 0.076 0.333* 0.335*** 0.056*** (0.007) (0.014) (0.007) (0.021) (0.100) (0.166) (0.114) (0.021) Adjusted q -0.010 -0.056* 0.002*** 0.000 0.184** 0.197 0.003 0.000 (0.008) (0.032) (0.000) (0.000) (0.088) (0.451) (0.006) (0.000) VC Dummy 0.097*** 0.155 -0.057 -0.134 -0.243 0.906 -0.881 -0.038 (0.033) (0.097) (0.035) (0.117) (0.504) (2.520) (0.661) (0.192) IPO returns 0.219 -0.143 1.621*** 3.327 12.849* -38.979 -9.917 -2.481 (0.527) (1.354) (0.616) (2.327) (7.269) (26.585) (10.487) (2.206) Obs 168 44 146 115 168 44 146 115 p-value 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 R2 10.04% 31.58% 22.89% 8.03% 24.49% 40.70% 17.30% 7.79%

46

Table 8 (continued)
All Regulated Constant Insider Holdings Size Leverage ROA Ln(age) Tangible Assets Founder dummy Float Adjusted q VC Dummy IPO returns Obs p-value R2 0.329* (0.175) -0.235** (0.093) -0.001 (0.012) -0.093 (0.063) 0.050* (0.027) 0.044** (0.023) 0.004 (0.113) -0.082** (0.038) 0.018 (0.011) -0.024* (0.012) 0.069 (0.048) -1.044 (0.643) 168 0.00 29.10% Monitoring Holdings Heavily All Regulated Regulated: Changes 0.526 -0.194 (0.605) (0.167) -0.196 -0.297** (0.145) (0.137) -0.000 -0.008 (0.026) (0.021) -0.407*** -0.068 (0.146) (0.120) -0.024 0.022 (0.032) (0.024) -0.048 0.041* (0.031) (0.023) 0.379 0.015 (0.276) (0.069) -0.012 0.021 (0.065) (0.033) 0.002 -0.009 (0.014) (0.010) -0.071* 0.002*** (0.039) (0.000) -0.030 -0.058 (0.150) (0.048) -1.769 0.562 (1.733) (0.852) 44 146 0.00 0.00 58.53% 13.83% All Regulated: % Change 1.944 (1.424) -0.078** (0.031) -0.757 (1.134) -0.013** (0.007) 0.001 (0.001) -0.272 (0.210) -0.001 (0.001) 0.466** (0.233) 0.023 (0.034) 0.001*** (0.000) 0.327 (0.425) -7.970 (5.290) 115 0.00 12.21% All Regulated 0.124 (0.204) -0.205** (0.101) 0.046*** (0.015) -0.046 (0.075) -0.030 (0.031) -0.033 (0.023) -0.130 (0.116) 0.024 (0.043) 0.006 (0.016) 0.051*** (0.008) 0.083 (0.055) 1.044 (0.068) 153 0.00 29.53% Equity-based Compensation Heavily All Regulated Regulated: Changes -0.980 -0.309 (0.653) (0.197) 0.172 0.121 (0.181) (0.180) 0.115*** 0.107** (0.036) (0.046) -0.406** 0.004 (0.160) (0.216) -0.103 -0.020 (0.067) (0.040) -0.003 0.036 (0.045) (0.033) 0.149 -0.080 (0.388) (0.103) -0.037 -0.021 (0.077) (0.057) -0.027* 0.016 (0.015) (0.015) -0.011 0.001 (0.030) (0.001) 0.267 0.078 (0.179) (0.067) 4.141** 1.946* (0.180) (1.114) 44 131 0.00 0.00 39.08% 13.02% All Regulated: % Change 5.710 (5.733) -1.420** (0.602) 0.853 (0.910) 0.028 (0.032) 0.004 (0.007) -0.550 (1.374) 0.180*** (0.023) -2.664 (1.887) 0.146* (0.081) 0.000 (0.001) -0.474 (1.317) -18.553 (16.433) 109 0.00 70.83%

* Statistical significance at the 10% level. ** Statistical significance at the 5% level. *** Statistical significance at the 1% level.

47

Table 9 Monitoring intensity: Matched sample. This table reports means of monitoring mechanisms for heavily regulated firms and a matched sample of unregulated firms. Firms are matched based on total assets. Panel A details our main monitoring measures (monitoring directors and holdings, board size, and equitybase compensation) while Panel B provides the component monitoring measures. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity-based compensation is the average percentage of compensation that is equity-based for the top executives. The sample consists of 58 heavily regulated and 58 matched unregulated firms. The p-value reports the significance of the difference between the sample means. Heavily Regulated Matched p-value Panel A: Main Monitoring Measures Monitoring Directors Board Size Monitoring Holdings Equity-based compensation 57.32% 9.27 15.50% 13.27% 49.24% 5.85 28.98% 17.49% 0.04 0.00 0.00 0.29

Panel B: Component Monitoring Measures Outside Directors VC Directors Outside Blockholdings VC Holdings Insider Holdings 54.11% 3.21% 11.51% 3.98% 33.55% 31.06% 18.18% 10.63% 18.35% 37.98% 0.00 0.00 0.83 0.00 0.35

48

Table 10 Substitution versus pressure tests: Matched sample. This table reports results from
regression analysis using robust standard errors with four different monitoring mechanisms as the dependent variable. Monitoring directors is the proportion of independent outside directors and venture capitalist [VC] directors. Monitoring holdings is the sum of outside blockholdings and venture capitalist holdings. Equity-based compensation is the average percentage of incentive compensation for the top executives. Leverage is total debt divided by assets. Age is the number of years from date of incorporation until the IPO. Tangible assets are the ratio of tangible to total assets. Founder is a binary variable equal to one if a founding family member is present at the IPO. Adjusted q is Chung and Pruitts (1994) approximation of Tobins q, adjusted for the industry median. The sample includes 58 heavily regulated and 58 unregulated firms (selected by matching asset size to the heavily regulated firms). Robust standard errors are in parentheses. Monitoring Board Size Monitoring Equity-based Directors Holdings Compensation Constant 0.788*** 9.243*** 0.255 0.145 (0.267) (3.499) (0.319) (0.297) Insider Holdings Size Leverage ROA Ln(age) Tangible Assets Founder dummy Float Adjusted q VC Dummy IPO Returns
Heavily Reg Dummy Heavily Reg Dummy * Insider Holdings

-0.286*** (0.081) -0.040* (0.023) 0.110 (0.092) 0.328 (0.252) -0.001 (0.019) -0.031 (0.071) -0.029 (0.046) -0.003 (0.005) 0.030* (0.017) 0.093** (0.046) -0.330 (0.924) 0.060 (0.068) 0.081 (0.133) 104 0.00 26.65%

2.827* (1.674) -0.078 (0.304) 2.791** (1.247) -3.546 (2.781) -0.823* (0.358) -2.417* (1.302) -0.827 (0.725) 0.093 (0.083) -0.200 (0.165) 0.845 (0.693) -7.312 (12.586) 2.486** (0.957) -0.294 (2.825) 104 0.00 38.05%

-0.166 (0.138) 0.037 (0.026) -0.105 (0.080) 0.369 (0.300) 0.016 (0.022) -0.023 (0.073) -0.001 (0.043) 0.008 (0.009) -0.017 (0.013) 0.251*** (0.065) -1.568 (1.081) 0.095 (0.121) -0.243 (0.244) 104 0.00 46.45%

-0.144 (0.124) -0.005 (0.025) -0.021 (0.094) -0.010 (0.230) 0.023 (0.021) -0.066 (0.084) 0.038 (0.046) 0.006 (0.010) 0.006 (0.019) 0.088 (0.071) 0.202 (1.025) -0.041 (0.101) 0.158 (0.194) 102 0.92 8.59%

Obs p-value R2

* Statistical significance at the 10% level. ** Statistical significance at the 5% level. *** Statistical significance at the 1% level.

49