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

Does Regulation Substitute or Complement Governance?

David A. Bechera,b, Melissa B. Fryec

aDepartment of Finance, Drexel University, Philadelphia, PA 19014 bWharton Financial Institutions Center, University of Pennsylvania, Philadelphia, PA 19014

cDepartment of Finance, University of Central Florida, Orlando, FL 32816

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

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 firm’s actions as they cannot control daily operations. This suggests

2

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,

3

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

4

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

5

(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

6

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

1 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.

7

(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

8

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 post-

deregulation (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,

9

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 equity-

based 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

10

monitoring systems where one measure may be increased when another decreases, suggesting

the firm’s 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

2 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.

11

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

3 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.

12

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 Interactive’s 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.

4These 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.

13

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

board’s 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.

5 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.

14

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.

6 Results are robust to using the percentage of equity-based compensation for the CEO only. 7To calculate option grant values, we apply a variant of Black-Scholes (Noreen and Wolfson, 1981). In the year pre-offering, 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.

15

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.

16

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, Moody’s, Yahoo Finance, and a

firm’s 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

17

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 firm’s first date of incorporation until the IPO captures firm age.

In Panels A–C 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

9 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.

18

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.

19

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.

20

Numerous studies argue a link exists between firm performance and governance. For

firm performance and growth opportunities, we use Chung and Pruitt’s (1994) approximation of

Tobin’s 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 firm’s 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 equity-

based 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

21

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 executive’s preferences are for agency

goods (e.g., slacking, perquisites, empire building) is an empirical question.15

12 While this could suggest substitution for equity-based compensation, further results do not support this claim. 13Results 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 Cuñat 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

22

Cuñat 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).

23

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 post-

deregulation. 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 firm’s 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.

24

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.

25

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 trade-

offs 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

26

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:

i

i

iDummy VC9

iDummyFounder 8iAssets Tangible7iDummy Regulated*iHoldings Inside6

iDummy Regulated*ionCompensatiEquity 5iDummy Regulated*iHoldings Monitoring4

iHoldings Inside3ionCompensatiEquity 2iHoldings Monitoring1iDirectors Monitoring

27

i

i

i

i

i10

i90i8i7ii6

ii5ii4

i3i2i1i

i13i12i11i10

i90i8i7ii6

ii5ii4

i3i2i1i

DummyVC

q AdjustedAssets TangibleLeverageDummy Regulated*Holdings Inside

Dummy Regulated*Holdings MonitoringDummy Regulated*Directors Monitoring

Holdings InsideHoldings MonitoringDirectors MonitoringonCompensatiEquity

ReturnsIPODummyVCq AdjustedFounder

ROALeverageSizeDummy Regulated*Holdings Inside

Dummy Regulated*onCompensatiEquity Dummy Regulated*Directors Monitoring

Holdings InsideonCompensatiEquity Directors MonitoringHoldings Monitoring

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

28

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.

29

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 non-

trivial: 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.

30

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.

31

7. Conclusion

Monitoring mechanisms are costly to adopt. In regulated industries, executive decision-

making 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

32

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.

33

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: An IPO perspective. Journal of Accounting Research 40, 321-346. 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.

34

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 Tobin’s q. Financial Management 23, 70-74. Cuñat, 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, 285–527. 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, 539-562. Hermalin, B.E., 1992. The effects of competition on executive behavior. Rand Journal of Economics 23, 350–365.

35

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, 95–125. 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, 225–242. 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.

36

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, 737-783. 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, 765–774. 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.

37

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 Pruitt’s (1994) approximation of Tobin’s 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.

Leverage Assets

(millions)

Age

Tangible Assets

Founder

Float (millions)

Adjusted q

ROA

All Regulated (1) 36.50%

(28.57%)

1,122.25

(201.34)

19.02

(8.00)

73.39%

(98.69%)

48.67%

(0.00)

21.62

(6.02)

1.64

(0.14)

6.44%

(6.37%)

Heavily Regulated (2) 63.23%

(80.80%)

533.00

(183.68)

22.05

(11.00)

70.87%

(99.84%)

32.14%

(0.00)

9.62

(3.78)

0.17

(0.01)

5.25%

(2.60%)

Banks (3) 85.22%

(87.41%)

707.58

(292.56)

25.70

(14.00)

59.44%

(98.53%)

21.62%

(0.00)

4.05

(2.32)

0.01

(-0.01)

2.20%

(2.31%)

Unregulated (4) 18.26%

(7.69%)

159.45

(51.96)

14.49

(8.00)

74.51%

(99.24)

54.18%

(1.00)

8.88

(4.94)

1.80

(0.62)

3.56%

(10.98%)

p-value (1 vs.4) 0.00 0.00 0.00 0.69 0.13 0.00 0.80 0.17 p-value (2 vs. 4) 0.00 0.00 0.01 0.49 0.00 0.70 0.02 0.67 p-value (3 vs. 4) 0.00 0.00 0.00 0.02 0.00 0.04 0.04 0.76

38

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

Board Size Monitoring Holdings

Equity Comp

All Regulated (1) 52.07% 7.54 23.11% 16.62% Heavily Regulated (2) 57.32% 9.27 15.50% 13.27% Banks (3) 57.66% 10.49 4.61% 9.91% Unregulated (4) 50.30% 6.28 24.14% 14.44% p-value (1 vs.4) 0.31 0.00 0.61 0.19 p-value (2 vs. 4) 0.03 0.00 0.01 0.70 p-value (3 vs. 4) 0.07 0.00 0.00 0.21

Panel B: Component Monitoring Measures

Outside Directors

VC Directors

Outside Blockholdings

VC Holdings

Insider Holdings

All Regulated (1) 44.00% 8.08% 14.95% 8.16% 33.01% Heavily Regulated (2) 54.11% 3.21% 11.51% 3.98% 33.55% Banks (3) 57.66% 0.00% 3.00% 1.61% 36.33% Unregulated (4) 33.24% 17.07% 9.80% 14.34% 41.85% p-value (1 vs.4) 0.00 0.00 0.00 0.00 0.00 p-value (2 vs. 4) 0.00 0.00 0.53 0.00 0.01 p-value (3 vs. 4) 0.00 0.00 0.03 0.00 0.17

39

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

Directors

Board Size Monitoring

Holdings Equity Comp

Panel A: Leverage

High (1) 53.75% 6.55 24.28% 17.56% Medium (2) 49.84% 6.08 23.67% 14.28% Low (3) 48.28% 6.93 23.39% 12.81% p-value (1 vs. 2) 0.02 0.01 0.74 0.05 p-value (2 vs. 3) 0.39 0.00 0.89 0.36 p-value (1 vs. 3) 0.00 0.09 0.64 0.00

Panel B: Size

High (1) 51.26% 7.43 29.60% 15.91% Medium (2) 50.26% 6.28 22.49% 15.18% Low (3) 50.35% 5.82 19.32% 13.50% p-value (1 vs. 2) 0.55 0.00 0.00 0.66 p-value (2 vs. 3) 0.96 0.00 0.06 0.29 p-value (1 vs. 3) 0.60 0.00 0.00 0.14

Panel C: Age

High (1) 47.19% 6.35 22.50% 12.06% Medium (2) 52.91% 6.49 24.58% 16.07% Low (3) 51.62% 6.76 24.59% 16.56% p-value (1 vs. 2) 0.00 0.43 0.26 0.01 p-value (2 vs. 3) 0.45 0.20 0.99 0.77 p-value (1 vs. 3) 0.01 0.06 0.30 0.01

40

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 =

All Regulated =

Heavy Regulated =

Bank Regulated =

All Regulated =

Heavy Regulated =

Bank

Monitoring Directors Board Size

Constant 0.443*** (0.076)

0.384*** (0.086)

0.463*** (0.072)

5.289*** (0.847)

3.536*** (0.839)

3.597*** (0.842)

Insider Holdings -0.195*** (0.030)

-0.197*** (0.033)

-0.200*** (0.031)

0.215 (0.340)

0.207 (0.353)

0.314 (0.368)

Size 0.007 (0.006)

0.013** (0.006)

0.007 (0.007)

0.496*** (0.083)

0.428*** (0.073)

0.487*** (0.066)

Leverage -0.047 (0.030)

-0.079** (0.031)

-0.086*** (0.027)

0.318 (0.375)

0.156 (0.333)

-0.345 (0.321)

ROA -0.016 (0.028)

-0.023 (0.028)

-0.013 (0.026)

-1.372*** (0.269)

-1.227*** (0.250)

-1.253*** (0.301)

Ln(age) -0.010 (0.007)

-0.004 (0.008)

-0.011 (0.007)

-0.137 (0.088)

-0.184** (0.084)

-0.157* (0.081)

Tangible Assets 0.083*** (0.026)

0.099*** (0.029)

0.083*** (0.026)

-0.783** (0.322)

-0.764** (0.329)

-0.793*** (0.299)

Founder dummy -0.036*** (0.014)

-0.036** (0.015)

-0.031** (0.014)

-0.425*** (0.160)

-0.401** (0.161)

-0.315** (0.159)

Float -0.006* (0.003)

-0.002 (0.003)

-0.003 (0.003)

0.004 (0.031)

0.042 (0.031)

0.033 (0.040)

Adjusted q 0.001 (0.001)

-0.000 (0.001)

0.001 (0.001)

0.018*** (0.006)

0.031** (0.012)

0.018** (0.009)

VC Dummy 0.193*** (0.013)

0.201*** (0.014)

0.196*** (0.013)

0.955*** (0.152)

0.892*** (0.151)

1.012*** (0.156)

IPO Returns 0.087 (0.298)

0.158 (0.334)

0.006 (0.288)

9.326*** (3.128)

9.825*** (3.144)

7.719** (3.374)

Regulated Dummy 0.029 (0.021)

0.128** (0.051)

0.196** (0.082)

0.648** (0.269)

1.477** (0.700)

2.600*** (0.960)

Regulated Dummy * Insider Holdings

0.032 (0.021)

0.018 (0.118)

0.030 (0.186)

-0.056 (0.231)

2.775 (2.275)

2.934 (2.183)

Non-bank Regulated Dummy

-0.015 (0.031)

0.531 (0.358)

Non-bank Regulated * Insider Holdings

0.089 (0.067)

-1.198 (0.786)

Observations 1,072 927 1,072 1,072 927 1,072 p-value 0.00 0.00 0.00 0.00 0.00 0.00 R2 24.92% 27.74% 25.31% 19.27% 23.47% 22.00%

41

Table 4 (continued). Regulated =

All Regulated =

Heavy Regulated =

Bank Regulated =

All Regulated =

Heavy Regulated =

Bank

Monitoring Holdings Equity-based Compensation

Constant 0.209*** (0.076)

0.138* (0.079)

0.189*** (0.032)

0.181** (0.077)

0.177** (0.082)

0.186** (0.075)

Insider Holdings -0.266*** (0.035)

-0.238*** (0.039)

-0.243*** (0.006)

-0.139*** (0.029)

-0.169*** (0.033)

-0.148*** (0.032)

Size 0.043*** (0.006)

0.046*** (0.007)

0.042*** (0.006)

0.011* (0.006)

0.011 (0.007)

0.011* (0.006)

Leverage -0.046* (0.027)

-0.047 (0.029)

-0.010 (0.028)

-0.072*** (0.025)

-0.090*** (0.027)

-0.068** (0.028)

ROA -0.042* (0.022)

-0.049** (0.023)

-0.044* (0.026)

-0.028 (0.026)

-0.025 (0.027)

-0.029 (0.026)

Ln(age) -0.010 (0.008)

-0.009 (0.009)

-0.010 (0.007)

-0.012* (0.008)

-0.007 (0.008)

-0.013* (0.007)

Tangible Assets -0.025 (0.026)

0.037 (0.026)

0.025 (0.026)

-0.063** (0.028)

-0.076** (0.030)

-0.065** (0.026)

Founder dummy -0.079*** (0.013)

-0.076*** (0.014)

-0.083*** (0.014)

0.010 (0.014)

0.014 (0.015)

0.010 (0.014)

Float 0.006** (0.003)

0.007** (0.003)

0.004 (0.003)

0.005* (0.003)

0.003 (0.003)

0.005 (0.003)

Adjusted q 0.003*** (0.001)

0.002 (0.001)

0.003*** (0.001)

0.003*** (0.001)

0.005*** (0.001)

0.003*** (0.001)

VC Dummy 0.197*** (0.014)

0.202*** (0.015)

0.196*** (0.013)

0.042*** (0.014)

0.037** (0.015)

0.043*** (0.014)

IPO Returns -0.488* (0.296)

-0.298 (0.309)

-0.423 (0.289)

0.182 (0.292)

0.293 (0.302)

0.179 (0.298)

Regulated Dummy -0.043* (0.025)

-0.038 (0.092)

-0.146* (0.082)

-0.010 (0.023)

-0.039 (0.061)

-0.058 (0.086)

Regulated Dummy * Insider Holdings

0.008 (0.029)

-0.110 (0.212)

-0.088 (0.187)

0.068*** (0.019)

0.209 (0.142)

0.163 (0.200)

Non-bank Regulated Dummy

0.013 (0.031)

-0.019 (0.032)

Non-bank Regulated * Insider Holdings

-0.087 (0.067)

0.095 (0.070)

Observations 1,072 927 1,072 1,062 919 1,062 p-value 0.00 0.00 0.00 0.00 0.00 0.00 R2 36.54% 37.13% 36.59% 9.01% 10.20% 7.50% * Statistical significance at the 10% level. ** Statistical significance at the 5% level. *** Statistical significance at the 1% level.

42

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

Board Size

Monitoring Holdings

Equity Comp

All regulated Early 53.26% 7.07 26.76% 16.90% Late 48.21% 9.06 11.20% 15.76% p-value 0.16 0.00 0.00 0.76

Heavily regulated         Early 57.69% 8.10 21.12% 12.90% Late 56.32% 12.47 0.14% 14.29% p-value 0.82 0.00 0.01 0.83

Banks         Early 58.57% 9.14 7.66% 6.78% Late 56.32% 12.47 0.14% 14.29% p-value 0.72 0.05 0.18 0.24

Panel B: Component Monitoring Measures

Outside Directors

VC Directors

Outside Blockholdings

VC Holdings

Insider Holdings

All regulated Early 44.91% 8.34% 18.31% 8.45% 31.93% Late 41.01% 7.20% 3.97% 7.23% 36.51% p-value 0.29 0.66 0.00 0.68 0.26

Heavily regulated           Early 53.31% 4.38% 15.68% 5.44% 32.67% Late 56.32% 0.00% 0.14% 0.00% 35.94% p-value 0.62 0.20 0.03 0.15 0.62

Banks           Early 58.57% 0.00% 4.95% 2.71% 36.61% Late 56.32% 0.00% 0.14% 0.00% 35.94% p-value 0.72 n/a 0.28 0.42 0.92

43

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 Pruitt’s (1994) approximation of Tobin’s 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

Board Size

Monitoring Holdings

Equity-based Compensation

Constant 0.695*** (0.091)

4.902*** (1.718)

0.232 (0.147)

0.199* (0.110)

Size 0.004 (0.012)

0.735*** (0.269)

0.039** (0.017)

0.007 (0.014)

Leverage -0.058 (0.058)

-0.370 (1.070)

-0.101 (0.074)

0.012 (0.060)

ROA -0.180* (0.099)

-1.318 (1.047)

-0.037 (0.101)

0.063 (0.079)

Ln(age) -0.043*** (0.016)

-0.193 (0.300)

-0.010 (0.020)

-0.032** (0.016)

Tangible Assets -0.096 (0.081)

-1.669 (1.626)

-0.135 (0.124)

-0.032 (0.094)

Founder dummy -0.044 (0.032)

-0.570 (0.511)

-0.118*** (0.040)

-0.022 (0.038)

Float -0.011* (0.006)

-0.189 (0.116)

0.017 (0.014)

0.011 (0.009)

Adjusted q 0.003*** (0.001)

0.015* (0.009)

0.003*** (0.001)

0.001* (0.001)

VC Dummy 0.138*** (0.033)

1.440** (0.559)

0.170*** (0.049)

0.088** (0.041)

Heavily Regulated Dummy

0.120*** (0.039)

2.253*** (0.597)

-0.027 (0.049)

-0.004 (0.041)

1998 Dummy -0.127* (0.079)

0.521 (1.560)

-0.262** (0.112)

-0.056 (0.085)

p-value 0.00 0.00 0.00 0.11 R2 22.87% 22.98% 30.16% 8.64%

44

Table 6 (continued)

Panel B: Banks versus All Firms

Monitoring Directors

Board Size

Monitoring Holdings

Equity-based Compensation

Constant 0.494*** (0.051)

4.575*** (0.626)

0.147** (0.061)

0.173*** (0.058)

Size 0.012* (0.006)

0.491*** (0.084)

0.048*** (0.007)

0.012* (0.006)

Leverage -0.100*** (0.030)

-0.318 (0.347)

-0.030 (0.033)

-0.065** (0.027)

ROA -0.026 (0.027)

-1.334*** (0.269)

-0.060** (0.024)

-0.033 (0.026)

Ln(age) -0.015** (0.007)

-0.156* (0.086)

-0.013 (0.008)

-0.015** (0.008)

Tangible Assets -0.033 (0.041)

-0.483 (0.546)

-0.105** (0.051)

-0.083* (0.047)

Founder dummy -0.052*** (0.013)

-0.322** (0.145)

-0.113*** (0.014)

-0.006 (0.014)

Float 0.001 (0.003)

0.043 (0.031)

0.010*** (0.003)

0.007*** (0.003)

Adjusted q 0.001 (0.001)

0.021*** (0.007)

0.003*** (0.001)

0.003*** (0.001)

VC Dummy 0.200*** (0.013)

0.978*** (0.149)

0.200*** (0.014)

0.045*** (0.014)

1998 Dummy -0.138*** (0.037)

0.762 (0.496)

-0.161*** (0.048)

0.001 (0.044)

Bank Dummy 0.190*** (0.051)

2.971*** (0.910)

-0.167*** (0.050)

0.007 (0.042)

Non-bank Regulated Dummy

0.015 (0.021)

0.184 (0.239)

0.009 (0.026)

0.045* (0.023)

Bank Dummy * 1998 Dummy

0.086 (0.071)

1.725 (1.775)

-0.010 (0.051)

-0.005 (0.081)

Non-bank Regulated * 1998 Dummy

0.059 (0.045)

-0.303 (0.818)

-0.069 (0.043)

-0.108** (0.046)

p-value 0.00 0.00 0.00 0.00 R2 20.82% 22.86% 21.24% 7.48%

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

45

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. 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 233 regulated and 58 heavily regulated firms. All Regulated includes partially and heavily regulated firms.

Panel A: Changes in Levels

Change in Monitoring Directors

Change in Board Size

Change in Monitoring

Holdings

Change in Equity Comp

Change in Insider

Holdings

All regulated Mean 5.28%*** 0.12 1.70% 14.83%*** -7.48%*** Median 0.00%*** 0.00* 0.00%** 10.68%*** -4.53%***

Heavily regulated         Mean 5.69%** 0.17 -0.45% 13.25%*** -5.91%** Median 1.52%** 0.00 0.00% 11.32%*** -4.20%**

Panel B: Percent Changes

% Change in Monitoring Directors

% Change in Board

Size

% Change in Monitoring

Holdings

% Change in Equity

Comp

% Change in Insider Holdings

All regulated Mean 14.27%*** 15.32%*** 12.93% 113.13%* 21.71% Median 0.00%*** 0.00%*** 0.00%* 0.00%* -13.54%***

Heavily regulated           Mean 15.73%*** 10.89%** -10.80%** 230.89% -6.19% Median 1.85%** 0.00% 0.00%** 0.00% -12.05%**

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

46

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 Pruitt’s (1994) approximation of Tobin’s q, adjusted for the industry median. Robust standard errors are in parentheses. Monitoring Directors Board Size All

Regulated Heavily Regulated

All Regulated: Changes

All Regulated: % Change

All Regulated

Heavily Regulated

All Regulated: Changes

All Regulated: % Change

Constant 0.643*** (0.153)

1.228*** (0.452)

-0.432*** (0.122)

-0.438* (0.260)

0.948 (2.037)

9.440 (9.197)

0.390 (1.864)

0.260 (0.484)

Insider Holdings -0.135* (0.073)

-0.361** (0.175)

-0.147* (0.077)

-0.023* (0.014)

0.604 (1.096)

0.554 (2.024)

-0.050 (1.514)

0.022 (0.020)

Size -0.002 (0.010)

-0.031 (0.029)

-0.028 (0.025)

0.142 (0.277)

0.679*** (0.154)

0.668 (0.469)

0.485 (0.396)

0.101 (0.223)

Leverage -0.026 (0.053)

0.146 (0.134)

0.036 (0.074)

0.002 (0.006)

1.306 (0.931)

5.102** (2.141)

-0.257 (1.341)

0.002 (0.003)

ROA -0.024 (0.023)

-0.003 (0.051)

0.056** (0.025)

0.001** (0.000)

-0.236 (0.461)

1.722 (1.199)

-1.429*** (0.517)

0.001* (0.001)

Ln(age) -0.027 (0.019)

-0.020 (0.035)

0.062*** (0.021)

0.022 (0.112)

-0.718*** (0.261)

-0.823 (0.739)

0.596* (0.349)

0.055 (0.058)

Tangible Assets 0.047 (0.090)

-0.404* (0.226)

-0.093 (0.060)

-0.001 (0.001)

2.747** (1.077)

0.437 (5.625)

1.715 (1.167)

-0.000 (0.000)

Founder dummy -0.018 (0.034)

0.034 (0.062)

0.009 (0.032)

-0.006 (0.131)

-0.368 (0.548)

-0.069 (1.371)

1.236* (0.651)

0.052 (0.129)

Float -0.003 (0.007)

-0.012 (0.014)

0.014** (0.007)

0.039* (0.021)

0.076 (0.100)

0.333* (0.166)

0.335*** (0.114)

0.056*** (0.021)

Adjusted q -0.010 (0.008)

-0.056* (0.032)

0.002*** (0.000)

0.000 (0.000)

0.184** (0.088)

0.197 (0.451)

0.003 (0.006)

0.000 (0.000)

VC Dummy 0.097*** (0.033)

0.155 (0.097)

-0.057 (0.035)

-0.134 (0.117)

-0.243 (0.504)

0.906 (2.520)

-0.881 (0.661)

-0.038 (0.192)

IPO returns 0.219 (0.527)

-0.143 (1.354)

1.621*** (0.616)

3.327 (2.327)

12.849* (7.269)

-38.979 (26.585)

-9.917 (10.487)

-2.481 (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%

47

Table 8 (continued) Monitoring Holdings Equity-based Compensation All

Regulated Heavily Regulated

All Regulated: Changes

All Regulated: % Change

All Regulated

Heavily Regulated

All Regulated: Changes

All Regulated: % Change

Constant 0.329* (0.175)

0.526 (0.605)

-0.194 (0.167)

1.944 (1.424)

0.124 (0.204)

-0.980 (0.653)

-0.309 (0.197)

5.710 (5.733)

Insider Holdings -0.235** (0.093)

-0.196 (0.145)

-0.297** (0.137)

-0.078** (0.031)

-0.205** (0.101)

0.172 (0.181)

0.121 (0.180)

-1.420** (0.602)

Size -0.001 (0.012)

-0.000 (0.026)

-0.008 (0.021)

-0.757 (1.134)

0.046*** (0.015)

0.115*** (0.036)

0.107** (0.046)

0.853 (0.910)

Leverage -0.093 (0.063)

-0.407*** (0.146)

-0.068 (0.120)

-0.013** (0.007)

-0.046 (0.075)

-0.406** (0.160)

0.004 (0.216)

0.028 (0.032)

ROA 0.050* (0.027)

-0.024 (0.032)

0.022 (0.024)

0.001 (0.001)

-0.030 (0.031)

-0.103 (0.067)

-0.020 (0.040)

0.004 (0.007)

Ln(age) 0.044** (0.023)

-0.048 (0.031)

0.041* (0.023)

-0.272 (0.210)

-0.033 (0.023)

-0.003 (0.045)

0.036 (0.033)

-0.550 (1.374)

Tangible Assets 0.004 (0.113)

0.379 (0.276)

0.015 (0.069)

-0.001 (0.001)

-0.130 (0.116)

0.149 (0.388)

-0.080 (0.103)

0.180*** (0.023)

Founder dummy -0.082** (0.038)

-0.012 (0.065)

0.021 (0.033)

0.466** (0.233)

0.024 (0.043)

-0.037 (0.077)

-0.021 (0.057)

-2.664 (1.887)

Float 0.018 (0.011)

0.002 (0.014)

-0.009 (0.010)

0.023 (0.034)

0.006 (0.016)

-0.027* (0.015)

0.016 (0.015)

0.146* (0.081)

Adjusted q -0.024* (0.012)

-0.071* (0.039)

0.002*** (0.000)

0.001*** (0.000)

0.051*** (0.008)

-0.011 (0.030)

0.001 (0.001)

0.000 (0.001)

VC Dummy 0.069 (0.048)

-0.030 (0.150)

-0.058 (0.048)

0.327 (0.425)

0.083 (0.055)

0.267 (0.179)

0.078 (0.067)

-0.474 (1.317)

IPO returns -1.044 (0.643)

-1.769 (1.733)

0.562 (0.852)

-7.970 (5.290)

1.044 (0.068)

4.141** (0.180)

1.946* (1.114)

-18.553 (16.433)

Obs 168 44 146 115 153 44 131 109 p-value 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 R2 29.10% 58.53% 13.83% 12.21% 29.53% 39.08% 13.02% 70.83% * Statistical significance at the 10% level. ** Statistical significance at the 5% level. *** Statistical significance at the 1% level.

48

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 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. 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 57.32% 49.24% 0.04 Board Size 9.27 5.85 0.00 Monitoring Holdings 15.50% 28.98% 0.00 Equity-based compensation 13.27% 17.49% 0.29

Panel B: Component Monitoring Measures

Outside Directors 54.11% 31.06% 0.00 VC Directors 3.21% 18.18% 0.00 Outside Blockholdings 11.51% 10.63% 0.83 VC Holdings 3.98% 18.35% 0.00 Insider Holdings 33.55% 37.98% 0.35

49

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 Pruitt’s (1994) approximation of Tobin’s 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

Directors Board Size Monitoring

Holdings Equity-based Compensation

Constant 0.788*** (0.267)

9.243*** (3.499)

0.255 (0.319)

0.145 (0.297)

Insider Holdings -0.286*** (0.081)

2.827* (1.674)

-0.166 (0.138)

-0.144 (0.124)

Size -0.040* (0.023)

-0.078 (0.304)

0.037 (0.026)

-0.005 (0.025)

Leverage 0.110 (0.092)

2.791** (1.247)

-0.105 (0.080)

-0.021 (0.094)

ROA 0.328 (0.252)

-3.546 (2.781)

0.369 (0.300)

-0.010 (0.230)

Ln(age) -0.001 (0.019)

-0.823* (0.358)

0.016 (0.022)

0.023 (0.021)

Tangible Assets -0.031 (0.071)

-2.417* (1.302)

-0.023 (0.073)

-0.066 (0.084)

Founder dummy -0.029 (0.046)

-0.827 (0.725)

-0.001 (0.043)

0.038 (0.046)

Float -0.003 (0.005)

0.093 (0.083)

0.008 (0.009)

0.006 (0.010)

Adjusted q 0.030* (0.017)

-0.200 (0.165)

-0.017 (0.013)

0.006 (0.019)

VC Dummy 0.093** (0.046)

0.845 (0.693)

0.251*** (0.065)

0.088 (0.071)

IPO Returns -0.330 (0.924)

-7.312 (12.586)

-1.568 (1.081)

0.202 (1.025)

Heavily Reg Dummy 0.060 (0.068)

2.486** (0.957)

0.095 (0.121)

-0.041 (0.101)

Heavily Reg Dummy * Insider Holdings

0.081 (0.133)

-0.294 (2.825)

-0.243 (0.244)

0.158 (0.194)

Obs 104 104 104 102 p-value 0.00 0.00 0.00 0.92 R2 26.65% 38.05% 46.45% 8.59%

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