Do Private Equity Buyouts Create Value?

Examining the promise of operational engineering in buyout targets

Fernando R. Chaddad

1
To my wife, Jasmine Bush.

2
Contents

Preface 4
Acknowledgements 11
Abbreviations 14
Chapter 1: Introduction 15
Chapter 2: Background Theory on Corporate Governance 22
The Theory of Governance Engineering 22
The Promise of Operational Engineering 28
Chapter 3: The Buyouts Phenomenon 33
The 1980s‟ Leveraged Buyout (LBO) Wave 34
The Private Equity (PE) Buyout Cycle 41
A Definition of Buyouts: Scope of this Study 46
Chapter 4: An Empirical Study 50
Research Hypotheses 50
Research Methodology 58
Empirical Results 67
Chapter 5: Discussion and Implications 84
Contributions 85
Implications for theory and practice 90
Limitations and opportunities for future research 97
References 100
Appendix: Tables 114

3
Preface

This book is based on my doctoral work conducted at the Kenan-

Flagler Business School, University of North Carolina, in 2008 and 2009.

The choice of topic for a Ph.D. dissertation is never straightforward, given

the conflicting goals of finding a topic that is interesting, viable, and novel.

In a way, the doctoral candidate looking for a research topic is not unlike

the entrepreneur in search of a new business idea: the desire is to create

something innovative by recombining existing knowledge, and then to

mobilize scarce resources and deliver the plan against a ticking clock in a

competitive environment. This process is never linear, progress is usually

uneven, and the inevitable surprises along the way induce the researcher to

make decisions that ultimately lead to a destination quite removed from the

starting point.

Private equity buyouts were not my first topic of choice for a Ph.D.

dissertation. Instead, private equity buyouts chose me as I read an

intriguing press release in 2007. In those days, the private equity industry

in the US and around the world was on a roll, producing fabulously

4
wealthy businesspeople and attracting a lot of press attention in this

process. In an interesting press release following the announcement of yet

another record-breaking private equity deal, a buyout executive stated that

his firm – the private equity acquirer – was expected to significantly

improve operations in the newly acquired target, by bringing seasoned

executives with operational experience onto the board of directors of the

acquired target. This sounded strange to me.

I had been in management consulting for over ten years by then,

working around the world with clients mainly in the consumer goods

vertical, but also in mining, retailing, pharmaceuticals, and non-

governmental organizations. The consulting company I was aligned with

was a global organization with billions of dollars in revenues and a

workforce in the tens of thousands. I sold, ran or executed over 30

engagements in countries as different as Brazil, the United Kingdom, Peru,

the United States, Panama, the Netherlands, and Mexico. The type of

management consulting I was involved with was end-to-end, which means

we ate our own cooking: we were hired to strategize, and then to make it

happen. In some instances, we worked contingent on success fees, which

means we only got paid if we could deliver tangible and measurable

5
results. With regards to content, I was involved with large-scale

operational change in Fortune 500 clients. By large-scale operational

change I mean projects such as the planning and execution of mergers and

acquisitions (post-merger integration); planning and executing strategic

sourcing; or the design and implementation of new sales and distribution

strategies. Such work is transformational, as it usually mobilizes vast

amounts of people (both consultants and client personnel) numbering from

the tens to the hundreds; unfolds over a fairly long time frame (from 6

months to 2 years); and usually delivers bottom-line results measured in

the order of tens or even hundreds of millions of dollars.

With this experience under my belt, I formed a fairly clear opinion

on how to drive operational engineering. First, it takes hands-on effort:

large-scale operational change in large organizations is a full-time job that

requires scale in the form of sizeable teams. Second, it also requires a

broad set of skills and experiences. Such armies of operational engineers

are usually comprised of business strategists, process specialists, change

management professionals, information technology professionals, and

functional team members ranging from supply chain to finance to sales and

marketing. Third, it takes time: the larger the organization and the larger

6
the desired change, the longer it will take for operational engineering to

trickle down and reach the intended corners of the organization before it

starts appearing in the company‟s financial statements.

And then I read this press release about someone claiming to be able

to drive operational engineering in a large corporation via remote control

by placing a couple of new directors on the board. This flew in the face of

everything I had done and witnessed in the corporate world.

I shared these thoughts with Rich Bettis, who chaired the Strategy

Department at the Kenan-Flagler Business School, at the University of

North Carolina. A renowned scholar with decades of experience and a very

broad world view, Rich encouraged me to conduct thorough research

starting with the 1980s‟ leveraged buyouts (LBOs) and culminating with

the 2000s‟ private equity mania. The first picture that emerged looked

promising. First, the topic was interesting enough to the point of being

controversial. This controversy was captured in the 1988 movie Wall

Street, in which actor Michael Douglas played the buyout artist Gordon

Gekko with lines such as “greed is good” and “I create nothing”. Second,

there was a small literature stream in academia focusing on buyouts, with

strands in strategic management, economics and finance. This was quite

7
practical for the purposes of offering the curious researcher the possibility

of identifying potential research gaps, and then evaluating each one for

feasibility and impact. Third, and perhaps most importantly, it became very

clear to me that a key question remained open for both academics and

businesspeople despite the published research in the buyouts domain: do

buyouts create value? Relatedly, a sense of déjà-vu inevitably hits any

researcher reviewing buyouts from the 1980s through the 2000s. Why were

buyouts referred to as LBOs in 1980s, and then called private equity

buyouts in the 2000s? Has the phenomenon changed so much that it

justified a new nametag?

The question of buyouts‟ value creation may be approached from

different perspectives. From the vantage point of equity holders in buyout

firms, this is ultimately a question of shareholder returns. Steve Kaplan

from the University of Chicago has conducted extensive research on this

question, and concluded that the buyout industry is highly cyclical as

returns will depend on vintage years. The equity holder‟s perspective is not

the one I chose to examine in this book.

The perspective I examine here is the one of the acquired buyout

target, and specifically its top and bottom lines. How might a new owner

8
add value to the acquired company? Some options include bringing in

innovation that enables firm growth, or focusing on operational efficiency

to lower costs and improve profitability. A buyout firm might hire a team

of seasoned, former CEOs or other senior executives with deep operational

experience – as some buyout firms have done in the 2000s – in order to

work on the opportunity to drive operational engineering.

In theory, such former executives might create value of at least two

types. Ex ante (before the deal), the buyout executives with operational

experience may influence deal screening by helping identify the

underperforming targets in the direst need of a turnaround. Ex post, the

buyout executives with operational experience may create value after the

buyout deal is complete by getting their hands dirty and driving operational

engineering. In this book, I offer initial work on the question of whether

buyouts create value by designing and running an empirical study on the ex

ante type of value creation. Whereas it is not the ambition of this book to

exhaust the question on buyout value creation, it is a goal to redirect the

debate on buyouts towards value creation from the target‟s perspective.

This book may be of interest to distinct audiences. Businesspeople,

government professionals and the general reader may be most interested in

9
Chapter 1 (which sets up the context and discussion); Chapter 2 (which

summarizes a debate on corporate governance leading up to buyouts);

Chapter 3 (which offers a historical description of buyouts); and Chapter 5

(discussion and implications). In addition to these chapters, business

academics and scholars might be interested in Chapter 4, which describes

the empirical study including the developed methodology.

10
Acknowledgements

In developing the doctoral dissertation which led to this book, I was

fortunate to have access to a community of extraordinary individuals to

whom I am indebted. I first thank Rich Bettis, the Ellison Distinguished

Professor of Strategy and Entrepreneurship at the Kenan-Flagler Business

School, University of North Carolina. In addition to always offering me the

broad view and the best advice, Rich was overwhelmingly generous with

his time, which is a rather scarce commodity given Rich‟s leading role as

editor of the Strategic Management Journal.

I also thank Jeff Reuer, who is the Blake Family Endowed Chair in

Strategic Management and Governance at the Krannert School of

Management, Purdue University. Jeff is a very prolific researcher in the

domain of mergers and acquisitions and joint ventures, as well as a friend

who showed me how superior and careful research methodology very often

makes the difference.

Further, I thank Will Mitchell, the J. Rex Fuqua Professor of

International Management at the Fuqua School of Business, Duke

University. Will is an extraordinarily sharp mind who does not need more

11
than a few minutes in any given encounter – planned or unplanned – to

significantly contribute to a debate.

I am also indebted to committee members at the Kenan-Flagler

Business School, University of North Carolina: Hugh O‟Neill (the Edward

O'Herron, Jr. Distinguished Scholar and Professor of Strategy and

Entrepreneurship) and David Ravenscraft (the Fulton Global Business

Distinguished Professor of Finance) for input and encouragement.

This work benefitted immensely from informal conversations with

academic colleagues, and I therefore thank Nandini Lahiri, Matt Semadeni,

Tony Tong, Atul Nerkar, Isin Guler, Phil Rosenzweig, James Henderson,

Idie Kesner, Jeff Covin and Oliver Gottschalg for thoughts and ideas.

I also thank doctoral seminar participants at the Kelley School of

Business, Indiana University; the Queen‟s School of Business at Queen‟s

University in Canada; the IMD Business School in Switzerland; and the

ESSEC Business School in France.

Next, I am indebted to buyout market participants for sharing their

experiences with me. They are Dennis Schaecher (BOS LLC) and Clay

Hamner (Montrose Capital Corporation) on the buy side, as well as Mike

Bowers (CoMark) and an anonymous contributor on the sell side.

12
 Finally, I am also indebted to Amol Joshi, whom I first met during

our MBA years at the Tuck School of Business, Dartmouth College. Amol

is very successful technology entrepreneur, who taught me much about

innovation and value creation.

13
Abbreviations

CapEx Capital Expenditures

FCF Free Cash Flow

GP General Partner

LBO Leveraged Buyout

LP Limited Partner

MBO Management Buyout

PE Private Equity

POE Potential for Operational Engineering

14
CHAPTER 1
Introduction

Business or firm performance is the central focus of strategic

management research (Schendel and Hofer, 1979). The broad question of

why businesses underperform has been examined in strategic management

from several theoretical perspectives, including agency theory (Jensen and

Meckling, 1976). According to agency theorists, firms underperform due

to the separation of ownership and control and the resulting misalignment

of interests between shareholders (principals) and management (agents).

To resolve this problem, agency theorists prescribe the redesign of

managerial incentives in the form of pay-for-performance compensation.

This induces managers to focus on shareholder value as well as a more

concentrated ownership structure, since fewer active shareholders with

higher equity stakes are more apt to effectively monitor management and

curb opportunism (Jensen and Meckling, 1976; Eisenhardt, 1989).

Following redesigned managerial incentives and closer board monitoring,

the firm is expected to perform optimally. Kaplan (2007) refers to these

agency-related remedies as governance engineering.

15
 Governance engineering, though, may not always fully solve the

problem of underperformance for at least two reasons (e.g., Hendry, 2002).

The first reason is associated with the specification of objectives, known as

multitasking (Holmstrom and Milgrom, 1991). When a principal‟s goals

are complex and multidimensional and therefore difficult to capture in an

outcome-based contract, attempts to specify contract outcomes may be

dysfunctional, as agents will perform to the specific terms linked to

incentives, rather than in the more general interests of their principals. The

second reason is honest incompetence (Hendry, 2002) by the firm‟s

management team. In agency theory, managers are assumed to be

competent and always able to achieve desired outcomes provided

incentives are in place.

Yet, in empirical research, the competence of individuals is not

guaranteed given bounded rationality (Simon, 1957) and limitations of

rational understanding and communication arising from language, culture,

and cognition (e.g., Simon, 1991). In situations in which agents are called

upon to exercise judgment, or in which the achievement of goals depends

on cooperative efforts involving other people, outcomes are not guaranteed

regardless of the effort applied (Nilikant and Rao, 1994). The issues of

16
honest incompetence and multitasking may be reduced via advice in the

form of mentoring and guidance (Hendry, 2002). Principals may dedicate

effort not to monitor for opportunism, but to help agents develop their

technical competence via the transfer of skills or to improve the agents‟

understanding of goals beyond those specified contractually – including

circumstances, values, and the broader priorities of principals. Similarly,

the offering of advice to management has been often reported as one of the

key duties of boards of directors (e.g., Mace, 1971; Carter and Lorsch,

2004).

An underlying assumption of such an advisory role is that principals

must have industry-specific knowledge, operating expertise or expert

networks deep enough to be considered areas of valuable advice to agents.

Such valuable knowledge possessed by principals may manifest

themselves in the capacity for operational engineering (Kaplan, 2007).

Potential for operational engineering (POE) refers to the degree to which a

firm‟s operational inefficiency is higher than the competition‟s, and

therefore could be improved. The implication is that principals must know

how to identify and act upon potential for operational engineering in order

to be capable of providing valuable advice to agents. If this line of

17
reasoning is correct, then firms in which principals and agents implement

both governance engineering (via redesigned incentives and closer

monitoring) and operational engineering (enabled by the advisory role of

capable principals) at the same time should be least likely to underperform.

It is noteworthy that operational engineering and governance

engineering embody overlapping yet different concepts. They overlap as

their absence may lead to similar consequences (underperformance), yet

they differ because not all firms with high potential for operational

engineering suffer from high agency costs. As an example, there is no

separation of ownership and control in a founder-owned and managed

firm, yet this firm may underperform even in the presence of profit-

maximizing goals as a consequence of honest managerial incompetence.

The empirical setting chosen to examine governance engineering

and operational engineering is buyouts. Buyouts constitute an ideal setting

where governance engineering is always present, and yet the presence of

operational engineering has been disputed as will be explained in detail

below. Buyouts are defined as going-private transactions, in which a

financial acquirer (or a group of investors led by a financial acquirer) in the

form of a private equity partnership purchases the controlling equity stake

18
in a publicly-traded target, usually with debt financing. This transaction

results in the formation of a new privately-held firm with a typically high

debt-to-equity ratio, and whose equity is closely held by a small group of

large outside investors (including the lead financial acquirer) and a group

of managers with high-powered incentives. Thus, buyouts represent a

clear-cut example of governance engineering (Kaplan, 2007). Further, the

extent to which operational engineering opportunities may be found in

buyouts has been the subject of debate (e.g., Rappaport, 1990; Fox and

Marcus, 1992). Previously, buyouts were referred to as LBOs (leveraged

buyouts) and were considered one of the most controversial features of the

US takeover boom of the 1980s (Wiersema and Liebeskind, 1995). By the

2000s, buyouts were referred to as PE (private equity) buyouts.

Buyouts were the focus of a series of empirical studies on the 1980s‟

LBO wave in the US (e.g., Kaplan, 1989; Singh, 1990; Liebeskind,

Wiersema and Hansen, 1992; Long and Ravenscraft, 1993a, 1993b; Phan

and Hill, 1995; Wieserma and Liebeskind, 1995). Building on this research

stream, scholars have started to examine the more recent buyout wave that

gained momentum in the 2000s. Kaplan (2007) argues that one of the key

differences between 1980s‟ LBOs and recent PE buyouts is that LBOs

19
focus solely on governance engineering, whereas PE buyouts are

concerned with operational engineering as well as governance engineering.

An implication of this argument is that, whereas agency proxies were

found to be key drivers of buyout activity in the 1980s (Singh, 1990; Opler

and Titman, 1993), an additional predictor in the form of potential for

operational engineering must be considered in the examination of the

antecedents of recent buyouts. At the heart of this argument are the

dramatic changes in US corporate governance since the 1980s, as

explained in detail below.

In this book, I examine the interplay between governance

engineering and operational engineering in the context of buyouts. In doing

so, I aim to address two questions associated with the antecedents of

buyout activity. First, whereas the agency costs of free cash flows (Jensen,

1986) have been proposed as one of the key drivers of buyout activity,

prior work (e.g., Singh, 1990; Opler and Titman, 1993) has relied on

proxies rather than direct measures of agency costs which gauge a firm‟s

potential for governance engineering. In this book, I offer a more direct test

of this agency-related argument by using measures such as board

independence, CEO duality and insider equity ownership to predict buyout

20
activity. The second question I aim to address with this book centers on

Kaplan‟s (2007) proposition on operational engineering, which has yet to

be empirically tested. I thus examine the antecedents of PE buyouts, testing

Kaplan‟s (2007) proposition while controlling for governance-related

antecedents.

The remainder of this book is organized as follows. Chapter 2 starts

with a review of the background theory on corporate governance, agency

theory, the role of buyouts, and lastly the concept of potential for

operational engineering (POE). Chapter 3 offers a review of the scholarly

literature on the buyouts phenomenon from the 1980s‟ LBOs through the

2000s‟ private equity (PE) buyouts. Chapter 4 offers a set of research

hypotheses and a methods section including sample description, statistical

technique, variables and measurements, and results of the empirical

analysis. Chapter 5 concludes this book with a discussion including

contributions, implications for theory and practice, limitations and future

research directions.

21
CHAPTER 2

Background Theory on Corporate Governance

Given this book‟s goal to investigate the relationships among

corporate governance, potential for operational engineering and firm

underperformance in the empirical setting of buyouts, I start with a brief

review of the corporate governance literature (in which agency theory is

the dominant paradigm) to summarize the governance engineering

argument. I then extent this rich research stream to explain how the

argument for operational engineering may complement governance

engineering to explain firm underperformance in general, and in the

context of buyouts in specific.

The Theory of Governance Engineering

The argument for governance engineering can be traced back to

Berle and Means (1932), who identified problems arising from the

separation of ownership and control in large, listed (publicly-traded) firms

and argued that management ownership in such firms is too small to make

managers interested in profit maximization. This is the incentive-intensity

22
argument also explored by Baumol (1959), Marris (1964) and Williamson

(1964), according to whom managers have a natural incentive to increase

firm size rather than to focus on profitability. Jensen and Meckling (1976)

and Fama and Jensen (1983) developed this argument further into a

complete theory which became the subject of extensive empirical

examination. Agency theory has then become the predominant theoretical

paradigm in corporate governance research (e.g., Daily, Dalton and

Cannella, 2003).

According to agency theorists (e.g., Jensen and Meckling, 1976), the

separation of ownership and control in the modern corporation evidences

two distinct entities with different interests and risk profiles: management

(agents) and shareholders (principals). While shareholders may diversify

risks by investing in multiple firms, management is tied to a single

organization by virtue of their position (Fama, 1980). This difference in

risk profiles means that management and shareholders operate under

different sets of incentives. Jensen and Meckling define an agency

relationship as “a contract under which one or more persons (the

principal[s]) engage another person (the agent) to perform some service on

their behalf which involves delegating some decision-making authority to

23
the agent” (1976:308). However, it is difficult to specify ex ante contracts

that accommodate all possible future contingencies (e.g., Shleifer and

Vishny, 1997). Agency theorists make the explicit assumption of self-

interested individuals (Jensen and Meckling, 1976) prone to opportunism.

From a shareholder‟s perspective, this may lead to inefficient managerial

behavior, such as: making short-term, risk-averse investments (Lambert

and Larcker, 1985); empire-building (Amihud and Lev, 1981); shirking

(Jensen and Meckling, 1976); exploiting managerial perks (Williamson,

1985); and „the quiet life‟ (e.g., Stein, 2003).

To agency theorists, the corporation‟s board of directors is the

primary monitoring device aimed at protecting shareholder interests and

alleviating potential agency problems (Fama and Jensen, 1983; Jensen and

Meckling, 1976). Agency theorists posit that the primary responsibility of

the board of directors is to ensure that management actions are consistent

with shareholder interests (Alchian and Demsetz, 1972; Fama and Jensen,

1983). Thus, the board acts to separate decision management from decision

control, keeping for itself the roles of ratification and monitoring (Fama

and Jensen, 1983). Additionally, boards of directors also influence firm

performance by reducing agency costs arising from noncompliance by

24
management with established goals and procedures, by articulating

shareholder objectives, and by focusing the attention of management on

performance (Mizruchi, 1983).

However, severe limitations to the degree of discretion conferred to

boards of directors have been widely documented (e.g., Mace, 1971;

Lorsch and MacIver, 1989) even more recently (e.g., Carter and Lorsch,

2004). Finkelstein and Hambrick (1996) note that boards are not always

effective monitors of management, and conclude that the underlying reason

relates to the balance of power in the boardroom, which tends to shift

toward the dominant CEO (e.g., Kosnik, 1987). Further, Kerr and Bettis

(1987) show that boards often do not honor their fiduciary duties.

Corporate governance controls may be internal or external to the

firm. Walsh and Seward (1990) argue that boards of directors have two

classes of internal controls available: the adjustment of incentives, and

dismissal. In case of failure of these internal control mechanisms available

to boards, the market for corporate control is supposed to serve as an

external mechanism and the discipline of last resort (Manne, 1965; Fama,

1980). However, external controls may also fail given a host of external

25
entrenchment practices available to astute, opportunistic management

(Walsh and Seward, 1990).

Corporate governance scholars disagree on the effectiveness of the

existing mechanisms in the United States (Shleifer and Vishny, 1997).

Easterbrook and Fischel (1991) offer an optimistic assessment of the US

corporate governance system, whilst Jensen (1989; 1993) argues that US

listed corporations embody deeply flawed governance mechanisms.

According to Jensen (1986), one of the reasons why governance

mechanisms in the US are flawed is the agency costs of free cash flows.

Jensen defines free cash flow (FCF) as “cash flow in excess of that

required to fund all projects that have positive net present values when

discounted at the relevant cost of capital” (1986:323). Noting that conflicts

of interest between principals and agents over payout policies are

especially severe when the organization generates substantial free cash

flows, Jensen (1986) examines the problem of how to motivate managers

to disgorge cash in lieu of investing it in projects yielding returns lower

than the cost of capital, or wasting it on other organizational inefficiencies.

An underlying assumption is that free cash flows may allow corporate

management (agents) to finance low-return or even negative-return

26
projects, which would otherwise not be funded via external sources such as

the equity or bond markets. FCF theory implies management in firms with

unused borrowing power and large free cash flows are more likely to

undertake low-benefit or even value-destroying projects. In order to test

FCF propositions, Jensen (1986) examined the US oil industry, which had

earned substantial free cash flows in the 1970s and the early 1980s in the

aftermath of substantial increases in oil prices. The author found that,

consistent with the agency costs of free cash flow, managers in the US oil

industry did not pay out excess cash to shareholders. Instead, they

continued to spend heavily on activities such as exploration and

development in the 1980s as oil prices collapsed, even though average

returns were below the cost of capital.

As a potential solution available to firms with severe agency costs of

free cash flow, Jensen (1986) offered debt. Specifically, “levering the firm

so highly that it cannot continue to exist in its old form generates benefits.

It creates the crisis to motivate cuts in expansion programs and the sale of

those divisions which are more valuable outside the firm. The proceeds are

used to reduce debt to a more normal or permanent level. This process

results in a complete rethinking of the organization‟s strategy and its

27
structure. When successful, a much leaner and competitive organization

results” (1986:328-329).

The leveraged buyout (LBO) wave that swept the US in the 1980s

was in line with this reasoning. By taking the buyout target out of public

ownership and into private hands with high levels of debt, LBO firms

(acquirers) expected to reduce agency costs and deliver superior

performance in the buyout target. Following the LBO wave of the 1980s,

Jensen (1989) predicted the eclipse of the public corporation and its

replacement by superior governance structures such as LBOs, which were

associated with lower agency costs of free cash flow as a consequence of

more aggressive executive pay as well as a closer monitoring of

management actions by a small but powerful board of directors (Gupta and

Rosenthal, 1991).

The Promise of Operational Engineering

Kaplan (2007) argues that, whereas governance engineering is a key

antecedent of buyouts, an additional antecedent in the form of potential for

operational engineering must be considered. At the heart of this argument

28
are the profound changes in US corporate governance since the 1980s as

synthesized by Holmstrom and Kaplan (2001).

Internal control mechanisms in US corporations were mostly

inactive until the 1970s. Descriptive work on boards of directors tracking

back to Mace (1971), Vance (1983) and Whisler (1984) described boards

of directors as passive groups of individuals, therefore establishing a

considerable gap between what the laws stated boards of directors should

do, and what boards actually did.

Typical board meetings in the 1960s and 1970s were not regarded as

a proper forum for the discussion of issues raised by challenging questions.

Mace (1971) and Whisler (1984) pointed to the existence of a tacit code of

conduct in the boardroom, in which professional courtesy and corporate

good manners suggested that embarrassing questions should not be asked,

especially if some of the CEOs subordinates were inside directors.

Consequently, any doubts or concerns about policies, operations, or

management decisions were typically expressed to the CEO outside board

meetings, unless the outside director was prepared to resign. Abrasive

questions risked being interpreted as a vote of no-confidence in the CEO,

and thus were very rarely asked. This description of typical board behavior

29
in the 1960s and 1970s is consistent with the conclusions offered by

Holmstrom and Kaplan (2001), according to whom corporate

managements then tended to be loyal to the corporation rather than to

shareholders, focusing on growth rather than shareholder returns.

Relatedly, boards tended to be passive and the use of incentive pay by

corporate management was limited (Lorsch and McIver, 1989).

After the 1980s LBO wave, however, the use of internal control

mechanisms changed significantly in US corporations. First, CEO

compensation increased by a factor of six from the 1980s to the 2000s,

with a disproportionate increase in equity-based pay. This increase in

equity-based compensation led to a hike in CEO pay-to-performance

sensitivities by a factor of more than ten times from 1980 to 1999

(Holderness et al., 1999). Second, changes in boards of directors were

mandated in the US by new legislation in the form of the Sarbanes-Oxley

Act of 2002. A key board-related change was the increased power,

responsibility, and independence of the listed firms‟ audit committee.

Holmstrom and Kaplan (2003) argued that the Sarbanes-Oxley Act

strengthened internal corporate governance mechanisms of listed firms by

affecting board behavior, which became more exposed to pressure to more

30
aggressively monitor management. As a consequence, many US

corporations hired board consultants to help implement best practices after

the passage of Sarbanes Oxley (Carter and Lorsch, 2004).

External control mechanisms also became more robust since the

1980s (Holmstrom and Kaplan, 2003). Donaldson (1994) described the rise

of shareholder activism as institutional investors with high ownership

stakes publicly state their high expectations with regards to shareholder

returns. In addition, the revolution in information technology enabled a

faster dissemination of information in capital markets, facilitating the work

of institutional investors and also allowing smaller shareholders to cast

votes of no-confidence by selling shares when in disapproval of corporate

management. Kaplan (1997) noted that the resurgence of general takeover

activity in the 1990s was evidence that the market for corporate control

proposed by Fama (1980) was as active as ever.

Kaplan also argued that LBOs did not reappear in the 1990s because

“we are all [KKR‟s] Henry Kravis now” (1997:2), explaining that

shareholders, managers and boards by then applied the insights and

strengths of 1980s‟ LBOs, including closer board monitoring and more

aggressive managerial incentives. If this argument is valid, then buyouts

31
will not reappear in the US because corporations do not need to complete

buyouts in order to implement governance engineering. From the early

2000s onwards, however, buyout activity in the US increased sharply

despite the presence of significantly more robust mechanisms of corporate

governance in place. If this resurgence cannot be explained by governance

engineering alone, then the question remains: what else may explain this

phenomenon?

32
CHAPTER 3

The Buyouts Phenomenon

Buyouts are a longstanding feature of US financial markets as well

as corporate governance, and can be traced back to the first half of the 20 th

century, when aggressive financiers took control of target firms via

financial transactions then called bootstraps (Rickertsen, 2001). Among the

first recorded buyouts was McLean Industries‟ purchase of Pan-Atlantic

Steamship in 1955. Interestingly, many early buyouts considered as

milestone transactions were hostile in nature, such as Victor Posner‟s

takeover of Sharon Steel in 1969. In the 1980s, buyouts gained much

media attention in the US under the label of LBO (leveraged buyout).

Following the boom-and-bust cycle of the 1980s, buyouts remained

somehow subdued in the US throughout much of the 1990s. By the early

2000s, buyouts resurfaced and then attracted intense media attention again,

this time under the PE (private equity) label. Like the 1980s LBO wave,

the PE buyout boom in the US turned into a bust following the 2007 credit

crunch in US financial markets.

33
The 1980s’ Leveraged Buyout (LBO) Wave

As noted, buyouts predate the 1980s but did not attract the interest of

corporate governance scholars before the phenomenon grew in importance

as “the $3 million buyout of the mid 1970s [was] replaced by the $1 billion

buyout of the mid 1980s” (Lowenstein, 1985:735). 1980s‟ LBOs were

transactions in which the shareholders of a listed firm are bought out by a

new group of investors – usually including incumbent management, a

specialized buyout firm, commercial banks and public debt holders. With

aggressive compensation arrangements, management typically increased

their percentage stake in the buyout company, even though they actually

extracted a certain dollar amount of their previous stake in the pre-buyout

firm (Kaplan and Stein, 1993). Whereas the acquiring firm typically

bought enough equity to secure control in the buyout target, most of the

remaining financing was in the form of debt (usually provided by

commercial banks) and high-yield subordinated public debt, also known as

junk bonds in the 1980s. The use of significant leverage (debt)

differentiated LBOs from other types of buyouts.

The 1980s‟ LBOs became central to what was then the greatest

M&A wave in US history. Between 1981 and 1989, more than 2400 listed

34
corporations with a market value of around $300 billion unadjusted for

inflation underwent an LBO (Wiersema and Liebeskind, 1995). Typical

LBOs then featured debt-to-capital ratios exceeding 85%, and shareholder

premia exceeding 40% (Kaplan, 1991). By the late 1980s, some of the

largest companies in the US were being targeted by buyout firms. High-

profile 1980s‟ LBOs included RJR Nabisco, Southland, H.R. Macy and

Trans World Airlines – all of which were amongst the top 20 companies of

the 1990 Forbes Rank (Opler, 1992).

Total Transaction Value of 1980s LBOs

[US completed deals in 2007 US$ billions]

1981 $5.2

1982 $3.6

1983 $4.9

1984 $30.7

1985 $27.4

1986 $83.0

1987 $76.8

1988 $101.0

1989 $125.7

1990 $28.0

35
 The 1980s‟ LBO wave culminated with the RJR Nabisco buyout in

1989 for a then record $24.8 billion unadjusted for inflation. By the early

1990s, the LBO boom became a bust as buyout activity collapsed due

several factors including anti-takeover legislation and jurisprudence; overt

political pressure against leverage; the collapse of the junk bond market;

and a credit crunch (Jensen, 1991; Comment and Schwert, 1995).

In sum, the LBO wave of the 1980s was considered one of the most

controversial business phenomena of the 1980s (Wiersema and Liebeskind,

1995) given its highly asymmetrical impact on firm stakeholders

(Rappaport, 1990; Fox and Marcus, 1992; Shleifer and Summers, 1988)

and the controversy that followed. Bhagat, Shleifer, and Vishny (1990)

stated that the main purpose of LBOs in the 1980s was to serve as a

temporary financing tool for the implementation of drastic short-run

improvements such as divestitures. Firm stakeholders that were apt to be

affected negatively in the aftermath of LBOs included employees (who

risked being dismissed or witnessing pay or benefit cuts), the government

(which would experience a decrease in corporate tax revenues) and

bondholders (whose holdings would plummet in value given higher

leverage and lower ratings in the buyout target following the LBO).

36
 In addition to the controversy arising from asymmetric outcomes

from the perspective of stakeholders, another controversial aspect was that

LBOs usually constituted a temporary ownership and governance structure,

as the acquirer (buyout firm) typically sold the buyout target after a limited

period of time. Buyout firms typically count on one of four exit strategies

in order to reach liquidity (Rickertsen, 2001): take the buyout target public

again, in what is referred to as reverse buyouts; sell the buyout target to a

strategic buyer; sell the buyout target to another buyout firm in what is

known as a secondary deal; or issue new debt in exchange for the equity of

exiting investors. Kaplan (1991) examined the question of whether buyouts

are permanent organizations, or whether they return to public ownership

via IPO‟s (initial public offerings). The author found that the median firm

goes public again within five to six years after the buyout, suggesting that

LBOs embody a temporary ownership structure.

Antecedents of 1980s’ LBOs. LBOs were the focus of a series of

studies following the intense buyout activity in the 1980s (e.g., Kaplan,

1989; Lehn and Poulsen, 1989; Singh, 1990; Muscarella and Vetsuypens,

1990; Smith, 1990; Palepu, 1990; Lichtenberg and Siegel, 1991;

Liebeskind, Wiersema and Hansen, 1992; Opler, 1992; Kaplan and Stein,

37
1993; Long and Ravenscraft, 1993a, 1993b; Hoskisson and Hitt, 1994;

Phan and Hill, 1995; Wiersema and Liebeskind, 1995). Scholars offered

theoretical arguments and empirically tested at least four potential

explanations for 1980s‟ LBOs.

The first potential explanation for 1980s‟ LBOs was Jensen‟s (1986,

1989, 1991, 1993) free cash flow (FCF) argument. According to FCF

theory, LBOs occurred as a consequence of cash flows that were invested

in value-destroying projects rather than returned to shareholders in the

form of dividends. Kaplan (1997) noted that, if this argument were correct,

then firms should cut capital expenditures in the aftermath of LBOs. Yet,

the evidence for this assertion was mixed. Whereas Kaplan (1989) and

Kaplan and Stein (1993) found evidence in support of FCF theory, Servaes

(1994) found the opposite.

A second, potential explanation for 1980s‟ LBOs was the

shareholder disappointment with conglomerates, and the resulting shift

away from unsuccessful diversification efforts and the return to

specialization. Shleifer and Vishny (1990) argued that many corporations

had embarked into unrelated diversification in the 1960s and 1970s, and

that these corporations could create shareholder value by becoming less

38
diversified in the 1980s. This argument was also in line with Jensen (1986)

and his influential work on free cash flow theory, as Jensen (1986) argued

that diversification was more likely to destroy than to create value.

However, Kaplan (1997) noted that the empirical evidence in support of

this argument was mixed, arguing that, whereas US corporations became

less diversified in the 1980s, this decrease did not seem significantly large

(e.g., Montgomery, 1994; Liebeskind and Opler, 1994).

Several scholars (e.g. Donaldson, 1994; Singh, 1990; Fox and

Marcus, 1992) argued that a third potential explanation for 1980s‟ LBOs

was firm undervaluation. Donaldson (1994) explained that the rise of

institutional shareholders and the greater availability of information

enabled by information technology in capital markets increased the

pressure to maximize shareholder value. Donaldson (1994) referred to the

1980s as the decade of confrontation. Kaplan (1997) argued that the

undervaluation argument was the most convincing explanation for the

1980s‟ LBO wave.

A fourth potential explanation was offered and tested by Opler and

Titman (1993), who combined Jensen‟s (1986) FCF argument and the

undervaluation argument to explain LBO activity. In their empirical tests,

39
the authors found that the firms most likely to become LBO targets were

the ones that combined high cash flows with unfavorable investment

opportunities in the form of Tobin‟s Q. The authors reasoned that it was

the interaction of high potential to waste investments in value-destroying

projects (in the form of high free cash flows) and few investment

opportunities (in the form of Tobin‟s Q) that explained LBO activity.

Prior work in strategic management (Singh, 1990) and corporate

finance (Opler and Titman, 1993) empirically tested several potential

explanations for 1980s‟ LBOs. As will be explained in detail below, this

book builds on extant literature to examine the antecedents of present-day

buyout activity by applying more direct measures of agency costs

(governance engineering) and testing new theory of buyouts (operational

engineering following Kaplan, 2007) while offering a methodological

approach that accommodates controls for sample selection bias and

competing risks in the form of Cox regression models.

Consequences of the 1980s’ LBOs Wave. As consequences of the

1980s‟ LBO wave, scholars studied the impact of LBOs on the buyout

targets‟ operational performance (e.g., Kaplan, 1989; Long and

Ravenscraft, 1993a), the buyout targets‟ diversification policies (e.g.,

40
Wiersema and Liebeskind, 1995), and changes in governance structure in

the buyout target after the LBO (e.g., Singh, 1990). From a theoretical

standpoint, many of these LBO studies drew from the agency costs of FCF

to examine LBOs. Whereas many scholars (e.g., Kaplan, 1991; Jensen,

1993) argued that LBOs addressed agency problems, others (e.g.,

Rappaport, 1990; Fox and Marcus, 1992) questioned the viability of LBOs

as a durable governance form given its adverse impact on firm

stakeholders other than shareholders.

The Private Equity (PE) Buyout Cycle

As noted, buyout activity resurfaced again in the early 2000s under a

new label: private equity (PE) buyouts. Rickertsen (2001) argued that this

re-labeling was a reaction by buyout firms to the negative press that

followed polemical LBO transactions in the late 1980s, when up to a third

of all buyouts eventually led to distressed outcomes such as restructurings

or even bankruptcies (Kaplan, 1997).

According to figures from Thomson Financial‟s Securities Data

Company (SDC), the PE buyout boom of the 2000s came into existence as

the total transaction value of US buyouts increased sharply from $22

41
billion in 2001 to $353 billion in 2006. By mid 2006, more than 250 US

firms in the booming buyouts industry controlled some $800 billion in

capital, and observers noted that these buyout firms were preparing for

more. Writers at Buyouts magazine estimated that nearly $175 billion of

new money flowed into US buyout firms in 2005 alone (The Wall Street

Journal, 2006). Similar to what was observed in the 1980s, the average size

of a PE buyout transaction dramatically increased as the phenomenon

gathered momentum. Some high-profile PE buyouts in the US included

Chrysler, Toys-R-Us, Neiman Marcus, Hertz and La Quinta Inn. In 2007,

the Blackstone Group announced the buyout of Equity Office Properties

for $36 billion, breaking KKR‟s longstanding record set in 1989 with the

RJR Nabisco deal.

The impact of buyouts on US business was by then significant, as an

estimated 1/25th of the entire US economy (Hubbard, 2007) as well as 6

million workers representing around 2% of the entire US workforce

(Davis, 2007) were under the control of buyout firms.

The amount of capital involved was quite substantial. Buyout

opportunities were offered only to institutional investors and wealthy

individuals (accredited investors) as required by law. General partners

42
(GPs) typically charged a 2% management fee per annum and 20% of

annual profits in the form of carried interest (two-and-twenty

arrangement), which typically led to fabulously rich pay schemes. GP pay

became rather controversial by the mid 2000s as the average size and

visibility of buyouts increased. Some of the world‟s wealthiest individuals

in 2007 were buyout executives such as Blackstone‟s Steve Schwarzman

(with a net worth of $7.8 billion), KKR‟s Henry Kravis ($5.5 billion) and

Carlyle‟s David Rubinstein ($2.5 billion) according to Forbes magazine

(Douglas, 2007). Adding controversy to PE buyouts was the fact that these

partnerships were taxed in the form of carried interest at 15%, similar to

long-term capital gains taxes and quite unlike ordinary corporate tax rates

of 35%. By the late 1990s and early 2000s, the buyout phenomenon that

had previously been essentially US-centered became global as buyout

firms searched for targets around the world, developing particularly strong

presences in Europe.

Not unlike the previous boom-and-bust cycle of the 1980s, the PE

buyout boom peaked in the third quarter of 2007, as the value of buyout

transactions by PE firms fell 68% from the previous quarter as a liquidity

crisis reached the credit markets that had made such deals possible.

43
 Total Transaction Value of 1990s and 2000s Buyouts
[US completed deals in 2007 US$ billions]

1991 $10.1
1992 $11.3
1993 $13.4
1994 $12.4
1995 $17.0
1996 $25.9
1997 $18.4
1998 $26.2
1999 $29.5
2000 $45.9
2001 $18.0
2002 $16.1
2003 $35.1
2004 $55.1
2005 $100.9
2006 $176.1
2007 $408.3
2008 $89.4

Given the relative newness of the phenomenon, scholars have

examined recent PE buyouts less intensively than the 1980s‟ LBO wave. A

brief review of scholarly work on the recent PE buyout cycle follows.

Cumming, Siegel and Wright (2007) provide an overview of the literature,

distinguishing between financial returns and „real‟ (productivity-related)

returns to investors. Guo, Hotchkiss and Song (2007) examined the

question of how PE buyouts create value and found increases in operating

performance in buyout targets after the completion of the transaction. The

44
authors also found that the improvement of cash flows after the buyout

event was higher in targets where the CEO had been replaced. Cao and

Lerner (2007) examined the performance of reverse leveraged buyouts (R-

LBOs), which are buyout targets that subsequently go public again after a

limited time under PE ownership. Further, Kaplan and Schoar (2005)

examined the capital inflows and performance of PE partnerships while

Davis (2007) studied the impact of PE buyouts on employment in buyout

targets.

Commenting on the potential antecedents of recent buyout activity,

Kaplan (2007:11) argues that:

What makes [today’s buyouts] different from the ‘80s is

that most of the big [buyout] firms, though not all, are now
committed to operational engineering. That’s why most of
them now have a pool of former CEOs or operating
executives. They bring them in to advice on where there is fat
that can be taken out [...]. In the late 1980s and afterward,
[...] incentives and board monitoring at public companies
have also improved. The buyout firms have responded to such
changes by developing industry and operating expertise that
they can use to add value to their investments. Many of the
firms have differentiated themselves by acquiring the industry
knowledge to oversee the strategies and operations of their

45
 portfolio companies. And they have also created networks of
operating executives—in many cases, highly regarded former
CEOs – to ensure that their portfolio firms have the best
managers and advice.

As noted, Kaplan‟s (2007) proposition on the potential for

operational engineering has yet to be empirically tested, controlling for

direct measures of agency costs which gauge potential for governance

engineering in the context of buyouts.

A Definition of Buyouts: Scope of this Study

In this paper, buyouts are defined as going-private transactions in

which a financial acquirer (or a group of investors led by a financial

acquirer) in the form of a PE (private equity) partnership purchases the

controlling equity stake in a publicly-traded target, usually with debt

financing. This transaction results in the formation of a new privately-held

firm with a typically high debt-to-equity ratio, and whose equity is closely

held by a small group of large outside investors (including the lead

financial acquirer) and a group of managers with high-powered incentives.

Further, it is important to note that the buyout target may be a private firm,

46
a listed firm, or the division of a firm (either public or private). In this

paper, I focus solely on whole-firm buyouts of listed firms, which lie at the

heart of the debate around the agency costs of free cash flows.

PE firms are usually structured as limited partnerships with one or

more general partners (GPs) and one or more limited partners (LPs) in the

form of a PE partnership. GPs are in the same legal position as partners in

a conventional firm: they have management control, share firm profits in

predefined proportions, and have joint liability for debt. GPs also have

authority as agents of the firm to bind all the other partners in contracts

with third parties. Like shareholders in a corporation, LPs have limited

liability, are only liable on debts incurred by the firm to the extent of their

registered investment, and they have no management authority. GPs pay to

LPs the equivalent of a dividend on their investment, the nature and extent

of which is usually defined in the partnership agreement. Whereas the PE

structure described above was already in place during the 1980s‟ LBO

wave, the buyout cycle that started in the early 2000s became known as the

PE buyout phenomenon as noted above.

Some buyout firms may engage in other types of activities beyond

buyouts, including VC (venture capital), PIPE (private investment in

47
public equity), and special situations such as distressed debt. In this paper,

I exclude all of these cases and focus solely on whole-firm buyouts of

publicly-traded (listed) firms. Further, several types of buyouts are

considered here. As noted above, buyouts may be leveraged (LBOs) or not.

Whereas most buyouts involve leverage, some buyout firms may execute

transactions with little or no use of leverage. Unlevered buyouts are

defined as buyouts with a post-buyout leverage of 30% or less (following

Long and Ravenscraft, 1993a) and are also included in this study. Further,

MBOs (Management Buyouts) are buyouts in which the managers of a

company purchase the company‟s controlling interest from existing

shareholders, usually with the help of a buyout firm. In the present study,

MBOs are included. The inclusion of this special type of buyout also raises

potential methodological concerns associated with endogeneity, as will be

discussed in detail and addressed below.

Further, a practice that did not occur widely in 1980s‟ LBO wave

but became quite widespread with the 2000s‟ PE buyout cycle was the club

deal, in which a group of PE firms pools its assets together and executes

the buyout collectively, in a practice that allows PE firms to acquire larger

targets and to reduce risks by taking smaller individual investments. In this

48
paper, my definition of buyouts includes club deals. Finally,

recapitalizations involve the restructuring a company‟s debt and equity

mix, often with the aim of making a company‟s capital structure more

stable. Because this does not lead to a change in firm ownership, these

types of deals are not included in this paper. Now that this study‟s

definition of buyouts is complete, I next present the hypotheses to be

tested.

49
CHAPTER 4

An Empirical Study

Research Hypotheses

Agency theory is built on the premise that the separation of

ownership and control in the modern corporation potentially leads to self-

interested actions by entrenched agents and therefore firm

underperformance. The potential for this conflict of interest between agents

and principals leads to the need for monitoring mechanisms designed to

protect shareholders and avoid shareholder expropriation by self-interested

agents (e.g., Fama and Jensen, 1983; Jensen and Meckling, 1976). In this

study, I seek to draw from the most widely applied measures of agency

problems in strategic management (e.g., Dalton et al., 1998). Extant

literature reflects two common remedies that mitigate agency costs: board

monitoring and incentive alignment.

Board Monitoring. In theory, the firm‟s board of directors is the

primary monitoring device aimed at protecting shareholder interests and

alleviating potential agency problems. Agency theorists posit that the

primary responsibility of the board is to ensure that management actions

50
are consistent with shareholder interests (Alchian and Demsetz, 1972;

Fama and Jensen, 1983).

With regards to board composition, the consensus in the corporate

governance literature (e.g., Mizruchi, 1983; Dalton et al., 1998;

Holmstrom and Kaplan, 2001; Cannella, Finkelstein and Hambrick, 2008)

is that boards of directors comprised of predominantly outside directors

rather than insiders represent a more effective monitoring of managers.

This preference for outsider-dominated boards of directors is grounded on

agency theory. Outside directors are believed to provide stronger control as

a result of their independence from management as inside directors may

not be able or willing to monitor the CEO with equanimity (Lorsch and

MacIver, 1989). As noted by Dalton et al. (1998), however, outside

directors who maintain personal or professional relationships with the firm

or firm management (affiliated directors) are not independent and not

believed to be effective in fulfilling the board‟s control role. Affiliated

directors are those with significant business dealings with the firm, defined

by the SEC as involving $60,000 per year or more and can be identified

through proxy statements or information services such as the Investor

51
Responsibility Research Center – IRRC (Cannella, Finkelstein and

Hambrick, 2008).

In contrast, independent directors are expected to be more effective

monitors of management behavior because of their focus on financial

performance as a central component of monitoring (Fama and Jensen,

1983); the increased likelihood of CEO dismissal in case of poor

performance (e.g., Couglan and Schmidt, 1985); the incentive to protect

their personal reputations as directors (Fama and Jensen, 1983); and the

likelihood to exercise greater objectivity because they are not as beholden

to CEOs as are inside directors (e.g., Walsh and Seward, 1990). In the

context of buyouts, if the agency-theoretic logic proposed by Jensen (1986,

1989) is valid, then the firms with the least independent board of directors

are the ones expected to underperform and thus become the target of a

buyout. Therefore:

H1:The less board independence in a given firm, the higher the probability
that the firm will become the target of a buyout.

As with board composition, agency theorists posit that the CEO

should not serve simultaneously as chairperson of the board in an

52
arrangement called CEO duality. According to agency theory predictions,

CEO duality promotes CEO entrenchment by reducing board monitoring

effectiveness (Finkelstein and D‟Aveni, 1994). Powerful individuals with

dual CEO and board chair positions may reduce board monitoring

effectiveness via a host of activities. First, CEOs with dual positions may

influence director selection that further strengthens a CEO‟s power base

(Pfeffer, 1981), which is also seen as a sign of entrenchment (Fama and

Jensen, 1983). Second, a powerful CEO that takes the chair position on the

board gains control of both the agenda and the debate in board meetings

(Cannella and Holcomb, 2005; Finkelstein and D‟Aveni, 1994). Third,

duality may lead to further entrenchment because board chairs provide

outsiders with most of the information about the organization (Mallette and

Fowler, 1992).

In the context of buyouts, if the agency-based arguments by Jensen

(1986, 1989) hold, then the firms with CEO duality are more likely to

underperform due to the decreased capacity for board monitoring, thus

becoming the target of a buyout. Therefore:

H2: CEO duality in a given firm increases the probability that the firm will
become the target of a buyout.

53
 Incentive Alignment Agency theorists argue that one way to mitigate

agency costs is to increase managerial equity holdings or to grant stock

options, thus aligning the interests of managers and shareholders (Jensen

and Meckling, 1976). Agency theorists therefore suggest that when

managers own substantial equity stakes in the firms they run, they are more

likely to act in shareholders‟ interests given their shared financial interests

(e.g., Perry and Zenner, 2000). The reason for this alignment of interests is

that managerial stock ownership causes managers‟ wealth to vary directly

with firm performance (Jensen and Murphy, 1990). In the absence of

insider equity ownership, managers are more likely to act in a self-serving

way by pursuing projects that further their own interests at shareholders‟

expense.

The same logic has been extended to board members. Several

empirical studies rely on directors‟ and officers‟ (D&O) equity to capture

insider equity ownership (Jensen, 1993). Some directors or board members

also serve as officers (managers) in their firms and are thus inside

directors. Regardless of whether a given director is an officer or not, all

board members are subject to the same alignment incentives as officers. If

officers with high equity stakes have the incentive to act in the best

54
interests of shareholders, directors who are not officers but who have high

equity stakes in the firm also have the incentive to effectively monitor

management behavior. In line with this reasoning, Jensen argued that “the

idea that outside directors with little or no equity stake in the company

could effectively monitor and discipline the managers who selected them

has proven hollow at best” (1989: 64). Therefore:

H3:The lower the equity ownership stake held by insiders in a given firm,
the higher the probability that the firm will become the target of a
buyout.

Potential for Operational Engineering (POE). POE refers to the

degree to which a firm‟s operational inefficiency is higher than the

competition‟s, and therefore could be improved. According to Kaplan

(2007), the main difference between 1980s‟ LBOs and the more recent

wave of buyouts is the focus on operational engineering. Kaplan (2007:11)

argues that in recent-era buyouts, “financial and governance engineering

continue to be important. However, most buyout firms try to augment

financial and governance engineering with another discipline – what I call

operational engineering.” An implication of this statement is that, whereas

55
governance engineering in the form of agency-variables was a key driver

of buyout activity in the 1980s, an additional antecedent in the form of

potential for operational engineering must be considered in the

examination of 1990s and 2000s buyouts. If Kaplan‟s (2007) proposition

on the potential for operational engineering is valid, then buyout firms will

be especially interested in buyout targets in which acquirers believe to able

to affect operational engineering.

The personnel recruiting practices of buyout firms, which had

traditionally focused on dealmakers with prior experience in investment

banking, have more recently been extended to include former senior

executives with operational experience (Meerkatt et al., 2008; Kaplan,

2007). According to Meerkatt et al. (2008), former senior executives with

operational experience are apt to have access to social networks enabling

ex ante buyout opportunities, and also have hands-on experience in

working with incumbent managers in setting and implementing ex post

agendas for operational improvements. The recruitment of operating

executives thus enables the pursuit of operational engineering in buyout

targets via advice in the form of mentoring and guidance (Hendry, 2002),

in addition to the traditional role of monitoring. Buyout firms that employ

56
former operating executives will thus be better positioned to help

managements in buyout targets achieve goals by developing their technical

competence via the transfer of skills, or to improve their understanding of

goals beyond the ones stipulated contractually.

The pursuit of operational engineering in buyout targets is

essentially an efficiency debate (e.g., Fox and Marcus, 1992). More

generally, examples of efficiency measures in strategic management have

included turnover ratios for total assets, receivables, and fixed assets (see

Carton and Hofer, 2006, for a review). Specifically in the context of

buyouts, typical ways to measure buyout target inefficiency include a low

working capital turnover (e.g., Fox and Marcus, 1992; Kaplan, 1988),

which is a measure of asset inefficiency. Relatedly, the empirical evidence

shows that, at least in the short-term, cash management improves in buyout

targets after the buyout is complete (e.g., Kaplan, 1989; Long and

Ravenscraft, 1993a). One likely reason is offered by Singh (1990), who

argued that potential buyout targets with asset inefficiencies in the form of

higher levels of receivables and lower inventory turnovers are apt to have a

lower quality of operational controls and implementation.

57
 Potential buyout targets that underperform operationally in the form

of high asset inefficiency therefore will become especially appealing to

buyout firms, since the context of operational inefficiencies exacerbate the

effects of governance engineering (board independence, CEO duality and

insider equity ownership) on the probability of buyout. Therefore:

H4:The higher a firm’s asset inefficiency, the stronger the negative

relationship between board independence and the probability that the
firm will become the target of a buyout.

H5:The higher a firm’s asset inefficiency, the stronger the positive

relationship between CEO duality and the probability that the firm will
become the target of a buyout.

H6:The higher a firm’s asset inefficiency, the stronger the negative

relationship between insider equity ownership and the probability that
the firm will become the target of a buyout.

Research methodology

Qualitatively, the hypotheses laid out above were informed by

unstructured interviews with buyout market participants – both on the buy

58
side and on the sell side.1 In the subsections below, I describe the

quantitative features of this study, including the sampling, statistical

technique, variables and measures employed.

Sample. This study links three datasets. The first dataset is Thomson

Financial‟s SDC (Securities Data Company), which offers detailed data on

IPO and M&A activity amongst other types of information. Given the

comprehensiveness of its coverage, SDC has been widely used in prior

research in strategic management. Using SDC data, I start with a search for

buyouts as transactions coded as “complete deal”, “deal is a LBO” and

“acquirer is a LBO firm” from 1998 through 2007 inclusive. I exclude

buyouts of financial targets in order to keep consistency with prior work

(e.g., Opler and Titman (1993). This search yields 1121 buyouts, which

include divestitures or divisional buyouts (413 deals), whole-company

private-to-private buyouts (622 deals), and whole company public-to-

private buyouts (86 deals). Given this study‟s aim to build on agency-

theoretic predictions, I focus on the whole company public-to-private deals

only. Whereas divisional buyouts could also be of interest from an agency-

1
The interviewees were Dennis Schaecher (BOS LLC) and Clay Hamner (Montrose
Capital Corporation) on the buy side; Mike Bowers (CoMark) and an anonymous
contributor on the sell side.

59
theoretic perspective, data availability is a challenge since US corporations

are not required to report separate sets of financials for each one of their

subsidiaries or lines of businesses. Next, I then evaluate and confirm that

each one of these 86 whole-company buyouts is consistent with the

definition provided in this study via press releases and The Wall Street

Journal.

The second dataset I use in this study is the Investor Responsibility

Research Center (IRRC), which offers data on board structure and

executive equity holdings for US publicly-held corporations. Governance

data from IRRC is available for 65 of the 86 firms identified as undergoing

a whole-company buyout via the SDC search described above. The count

of 65 buyout events featured in this study is comparable with other studies

in this literature stream. The 65 whole-company US buyouts span 27

different industry segments at 2-digit SIC level, as shown on Table 1 in the

Appendix. Further, the rapid acceleration of overall buyout activity in the

US from 2005 through 2007 is reflected in my dataset as shown on Table 2

in the Appendix, which offers a sample description by year.

The third dataset I use in this study is S&P‟s Compustat, which

offers firm-level financial data for US listed firms. Compustat II (Business

60
Segments) offers limited data points from firms‟ income statements for

each of the firm‟s line of business, which is necessary for developing a

control measure of diversification as explained below. Unfortunately, the

balance sheet data that could be used to examine divisional buyouts is not

provided.

In order to construct a risk set of all US firms that could become a

buyout target from 1998 to 2007 inclusive, I merge the governance data

available from IRRC with firm financials from S&P‟s Compustat in the

form of a panel dataset. This data merger resulted in a master file with

1459 publicly-traded firms with complete data, totaling 8631 observations.

Of the 1459 firms, 65 became the target of a buyout, an additional 260

became the target of (non-buyout) M&A, 18 were delisted, and 1116 were

still operating independently at the end of the sample period. Given the

need to control for competing risks as explained below, I coded non-

buyout M&A transactions and delistings accordingly.

Statistical Technique. In general, event history analysis is used to

analyze the effects of predictor variables on the occurrence or non-

occurrence and the timing of specific events (Allison, 1984). In this study,

I use the Cox proportional hazards model (Cox, 1972). The Cox

61
proportional hazards model is semi-parametric and therefore more flexible

as well as robust in cases in which it is difficult to specify a particular

shape of the time dependence of the hazard rate (Blossfeld and Rohwer,

1995).This modeling approach is associated with several advantages as it

accounts for both discrete events and continuous timescale data,

accommodates left truncation and right censoring simultaneously, allows

time-dependent independent variables, accommodates competing risks, and

identifies both cross-sectional and longitudinal effects (Allison, 1995).

These properties are particularly useful to our data for investigating the

antecedents of buyout activity over time. The model can be specified as:

H(t) = h(t) exp [β X(t)]

where the hazard rate H(t) is the product of an unspecified baseline rate,

h(t), a second term X(t), specifying the values at time t of a vector of

independent variables, and β, representing a vector of parameters

embodying the effect of each independent variable. The model is

proportional in that the hazard is obtained by shifting the baseline hazard

as the independent variables change over time. The model assumes that,

62
whatever the shape of the baseline hazard, it is the same for all firms.

Therefore, the baseline hazard cancels out, and there are no intercepts.

Variables and Measurement. In the subsections below, I describe all

variables and measurement definitions, including the dependent variable,

independent variables and controls.

Dependent Variable. The dependent variable Buyout denotes the

probability of a buyout occurring at time t, given that the event has not

occurred prior to that instant. The dependent variable is dummy-coded to

indicate whether a buyout has occurred towards the end of each spell.

Firms that were still operating as independent entities without becoming

the target of a buyout at the end of the sample period (1998-2007) will be

treated as right-censored. As noted above, competing risks or outcomes

associated with non-buyout M&A transactions as well as delistings from

stock exchanges due to bankruptcy or outright liquidation were also coded

accordingly. Hazard models accommodate such competing outcomes by

treating them as distinct risks (Cox and Oakes, 1984), which is a statistical

approach adequate for buyouts research.

63
 Independent Variables. Consistent with the event history design, all

data were collected annually. All independent variables reported here are

time-varying, and all financial data were collected from S&P‟s Compustat.

The first independent variable is Board Independence, defined as the

percentage of board members that are independent and therefore not

affiliated directors according to the IRRC. Data were sourced from IRRC,

which defines an independent director as someone who is not a former or

current employee; a service provider, supplier, or customer (or employee

or director thereof); a recipient of charitable funds; or a family member of

a director or executive of the firm. Second, CEO Duality is a dummy

variable coded as one if the firm‟s CEO also serves as chairperson at the

same firm. Data were provided from IRRC. Third, Insider Equity

Ownership is defined as the percentage of equity controlled by the firms‟

directors and officers as provided by the IRRC. Fourth, Asset Inefficiency

combines several inefficiency items as introduced by Carton and Hofer

(2006) and applied by others (e.g., van Mourik, 2007). These include:

liquidity inefficiency, defined as cash and equivalents (Compustat #1)

scaled by net sales (#12); working capital inefficiency, defined as working

capital (#179) scaled by net sales (#12); fixed-asset inefficiency, defined as

64
gross property, plant and equipment (#7) scaled by net sales (#12); and

total asset inefficiency, defined as total assets (#6) scaled by net sales

(#12). These measures were highly correlated, so I used weights from

factor analysis to extract a unified construct. This analysis yielded a one-

factor weighted average of asset inefficiency with an eigenvalue of 3.38

that explained 84.5% of the total variance in the four items described

above.

Control Variables. I start by controlling for antecedents of buyout

activity that have been previously tested. The first control variable is Free

Cash Flow, defined as the firm‟s free cash flow divided by net sales. I

follow Lehn and Poulsen (1989) and define free cash flow as operating

income before depreciation (Compustat #13), minus total income taxes

(#16), minus the change in deferred taxes from the previous year to the

current year (change in #35), minus gross interest expenses on total debt

(#15), minus the total amount of preferred dividend requirement on

cumulative preferred stock and dividends paid on non-cumulative preferred

stock (#19), minus the total dollar amount of dividends declared on

common stock (#21).

65
 The second control variable is Unrelated Diversification. I follow

Wiersema and Liebeskind (1995) in their LBO study and use the

Jacquemin-Berry entropy measure of diversification at the 2-digit SIC

level. This measure is estimated as:

 P ln (1  P )
i 1
i i

where Pi is the share of total firm sales of the ith line of business.

Following Palepu (1985) and Wiersema and Liebeskind (1995), unrelated

diversification is measured using the 2-digit SIC lines of business of the

firm to estimate Pi.

Third, I follow Kaplan and Zingales (1997) and estimate Tobin’s Q

as market value divided by book value of assets, where market value is the

firm‟s market value of common stock (#24 * #25), plus the book value of

assets (#6), less the book value of common equity (#60), less deferred

taxes on the balance sheet (#74).

Fourth, I also control for Leverage defined as the firm‟s total debt

(#9 + #34) divided by total equity (#216) in order to control for the

66
alternative explanation according to which firms that are already highly

levered may become a less attractive target for buyout firms.

Fifth, I control for Taxes defined as the firm‟s income tax expenses

(#16) scaled by net sales, controlling for the potential explanation offered

by Lowenstein (1985).

Finally, I control for Year Fixed Effects to account for unobservable

macroeconomic effects. I also control for Industry Fixed Effects to ensure

comparability across potential buyout targets in the same industrial sector

at the 2-digit SIC level. For both year and industry fixed effects, I have

included k-1 dummy variables in all models. Table 3 in the Appendix

summarizes all variables employed in the analyses below.

Empirical Results

Descriptive Statistics. The table below presents descriptive statistics

and a correlation matrix of the variables. Some notable correlations as well

as comments to these follow.

67
Independent
Mean S.D. (1) (2) (3) (4) (5) (6) (7) (8)
Variables

1. Buyout .01 .09

2. Free cash
.05 1.94 .00
flow
3. Unrelated
.15 .24 -.01 .01
diversification
4. Tobin‟s Q 2.11 1.70 -.03* -.01 -.14***

5. Leverage .77 3.15 -.01 .00 .05*** -.01

6. Taxes .03 .08 .01 -.01 -.02† .17*** -.02

7. Board
independence 67.24 16.75 .00 -.01 .07*** -.04*** .00 .00

8. CEO duality .32 .46 .02 -.01 -.11*** .02* -.04** .00 -.11***
9. Insider
equity 10.73 14.34 .01 -.02† -.06*** .01 .02† -.02† -.41*** .04***
ownership
10. Asset
2.56 6.78 .00 -.79*** -.03** .04** .00 -.09*** .01 -.01
inefficiency

n = 8631. † p < .10, * p < .05, ** p < .01, *** p < .001

First, free cash flow is not significantly correlated with board

independence or CEO duality, and it is only weakly correlated with insider

equity ownership (p < .10). This finding highlights the importance of using

direct measures of agency variables (when available) rather than relying on

proxies. In light of prior scholarly work on 1980s LBOs employing free

cash flows as a proxy of agency cost, one way to interpret this finding is

that free cash flow as a proxy measure was adequate in the 1980s but

68
probably less so later on given the evolution of corporate governance

mechanisms in the US as described by Holmstrom and Kaplan (2003).

At the same time, it is noteworthy that the correlation between free

cash flow and asset inefficiency is very high at -.79 (p < .001). This strong,

negative correlation may suggest that prior work on buyouts (e.g., Singh,

1990; Opler and Titman, 1993) testing free cash flow theory with proxy

measures of agency costs may have rather offered tests of the POE

argument instead.

Next, the probability of buyout and Tobin‟s Q are negatively

correlated (p < .05), in line with one of the data panels analyzed by Opler

and Titman (1993). Further, board independence and CEO duality are

negatively correlated (p < .001), consistent with the notion that the

separation of the CEO and board chair positions is associated with board

independence.

Given some high correlations among the variables, I investigated

potential multicollinearity problems, yet found that the maximum variance

inflation factor (VIF) value for all models was 4.64, well below the rule-of-

thumb threshold of 10 (Neter, Wasserman and Kutner, 1985).

69
 The charts below present a trend analysis examining key agency-

theoretic variables (board independence, CEO duality and insider equity

ownership). This analysis shows that board independence increased from

1998 to 2007 and CEO duality decreased in the same period, consistent

with Holmstrom and Kaplan (2001). However, insider equity ownership

remained largely unchanged in this period.

Board Independence
75%

70%

65%

60%

55%
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Insider Equity Ownership

15.0%

12.5%

10.0%

7.5%

5.0%
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

70
 CEO Duality
80%

70%

60%

50%

40%
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Event History Analysis. The table below reports the main results of

the event history analysis examining how governance engineering interacts

with asset inefficiency to shape the probability of buyout. All models are

highly significant (p < 0.001). Model I is the baseline specification

comprising control variables only. Model II augments Model I by adding

governance engineering variables (board independence, CEO duality and

insider equity ownership) in addition to asset inefficiency. Models III

through V introduce the direct effects of asset inefficiency, as well as its

interaction with each one of the three governance engineering variables.

71
Independent Model I Model II Model III Model IV Model V
Variables (Controls) (H1-H3) (H4) (H5) (H6)
2.235 2.481 2.029 2.306 2.627
Free cash flow
(2.295) (2.355) (2.335) (2.315) (2.357)
Unrelated .076 .067 .069 .082 .064
diversification (.144) (.146) (.146) (.147) (.145)
-1.462*** -1.493*** -1.532*** -1.531*** -1.494***
Tobin’s Q
(.368) (.377) (.378) (.380) (.378)
-.029 -.031 -.026 -.033 -.029
Leverage
(.129) (.128) (.129) (.129) (.130)
.337 .370 .436† .407 .355
Taxes
(.227) (.251) (.253) (.251) (.249)
Year fixed effects 30.21*** 28.14*** 28.69*** 28.55*** 28.56***
Industry fixed effects 64.81*** 63.79*** 64.03*** 64.50*** 64.21***
.095 .089 .088 .096
Board independence ---
(.154) (.154) (.154) (.154)
.023 .027 .038 .026
CEO duality ---
(.127) (.127) (.128) (.127)
Insider equity .126 .140 .119 .092
---
ownership (.123) (.125) (.123) (.131)
-.059 -.226 -.135 -.068
Asset inefficiency ---
(.479) (.494) (.493) (.456)
Asset inefficiency * -.484*
--- --- --- ---
board independence (.244)
Asset inefficiency * .402
--- --- --- ---
CEO duality (.353)
Asset inefficiency *
-.358
insider equity --- --- --- ---
(.409)
ownership
Log Likelihood, L(β) -376.03 -375.48 -373.94 -374.87 -375.05
Model χ 2 131.95*** 133.03*** 136.11*** 134.26*** 133.89***
Number of
8631 8631 8631 8631 8631
observations
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65
†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

Hypotheses 1 through 3 test predictions from agency theory

according to which board independence, CEO duality and insider equity

72
ownership are associated with the probability of buyout. In the table above,

none of the coefficient estimates for these three variables on Model II are

significant, providing no support for H1, H2 or H3.

Hypotheses 4 through 6 state that asset inefficiency moderates the

effect of governance engineering variables on the probability of buyout.

Specifically, hypothesis 4 predicts that the higher a firm‟s asset

inefficiency, the stronger the negative relationship between board

independence and the probability of the firm becoming a buyout target. In

Model III above, the interaction term is significant and negative (p <.05).

Because Cox models are non-linear and semi-parametric, a plot is

helpful in interpreting this result. The figure below shows the baseline

survival function for different firms.2 The plot shows how the interaction

effect of asset inefficiency and board independence decreases the

probability of buyout, providing no support for H4.

2
The mean firm is a hypothetical firm in which the values for all variables of interest
are set equal to the sample‟s mean. Firm A is a hypothetical firm with high board
independence (75th- percentile) and high asset inefficiency (75th- percentile). Firm B is
a hypothetical firm with high board independence (75th- percentile) and low asset
inefficiency (25th- percentile). Firm C is a hypothetical firm with low board
independence (25th- percentile) and high asset inefficiency (75th- percentile). Firm D is
a hypothetical firm with low board independence (25th- percentile) and low asset
inefficiency (25th- percentile).

73
 Further, hypothesis 5 predicts that the higher a firm‟s asset

inefficiency, the stronger the positive relationship between CEO duality

and the probability that the firm will become the target of a buyout. In

Model IV above, the interaction term is not significant, providing no

support for H5. Finally, hypothesis 6 predicts that the higher a firm‟s asset

inefficiency, the stronger the negative relationship between insider equity

ownership and the probability that the firm will become the target of a

buyout. In Model V, the interaction term is not significant, providing no

support for H6.

74
 There are some noteworthy patterns of results for control variables.

First, Tobin‟s Q is strongly significant in all model specifications (p

<.001), lending support to the undervaluation argument of buyout activity

tested by Opler and Titman (1993). My results show that, all else equal,

firm undervaluation in the form of a low Tobin‟s Q in relation to industry

peers is a key predictor of buyout activity. In other words, the firms that

seem cheaper than competitors are the ones that tend to be bought out. Not

surprisingly, buyout firms thus seem to focus on the more inexpensive

targets. Second, taxes are marginally significant only in Models III (p <

.10), lending no support to the argument that firms with comparatively

high tax expenses are more apt to become buyout targets, as proposed by

Lowenstein (1985). Finally, year fixed effects and industry year effects are

jointly significant in all models.

Robustness Analyses. I also performed several changes in

specification in order to assess the robustness of the models and results

discussed above.

Potential Sample Selection. Further, there might be a concern that

potential sample selection bias might affect the interpretation of my results.

Specifically, I define the risk set of my analysis as the entire universe of

75
publicly-listed firms with available data as reported by S&P‟s Compustat

from 1998 to 2007 inclusive. Yet, IRRC offers governance variables for

only a subset of the firms that report financials to S&P‟s Compustat,

comprising 17% of publicly-traded firms. Firms covered by IRRC tend to

be larger, more established organizations. But this potential focus on larger

firms would work against the hypotheses laid out above, making the

statistical tests more conservative. Therefore, potential sample selection

may be a lesser concern in this study.

Despite this consideration, one adequate approach to empirically

account for potential sample selection bias is provided by Lee (1983) in the

generalization of the Heckman (1979) two-stage selection model. This

approach provides a correction for potential selection in the form of the

correction variable lambda (λ), which makes the estimates of predictor

variables more precise by mitigating the effects of omitted variable bias

(Greene, 2000). I follow this approach, which has been applied in strategic

management (e.g., Leiblein, Reuer and Dalsace, 2002; Turner, Mitchell

and Bettis, 2008), and employ a Cox regression model that predicts the

hazard of inclusion in the IRRC dataset for all 8422 firms featured in

76
Compustat in a first-stage model that generates lambda. The correction

variable lambda (λ) is defined as:

 [ ( F (t ))]
1


i

it
1  F (t )
i

where Fi (t) is the cumulative hazard function for firm i at time t, Ф is the

standard normal density function, and Ф-1 is the inverse of the standard

normal distribution function (Lee, 1983). However, as Table 4 in the

Appendix shows, the implementation of a two-stage model with correction

for potential selection bias did not change the results of the main

regressions reported on the table presented above. It is also noteworthy that

the selection correction variable (λ) is not significant in any of the second-

stage analyses as shown on Table 4.

Potential Endogeneity. I took steps to address potential concerns on

how endogeneity may affect the interpretation of the results presented

above. For example, it may be argued that self-interested managers may

purposefully drive lower employee and asset productivity, which would

manifest itself in the form of higher potential for operational engineering,

underperformance, and thus in a higher probability of buyout. In this

77
hypothetical case, self-interested management would be aiming for a

buyout that would enable them to stay on as managers of the buyout target

after the completion of the transaction, but with higher-powered incentives.

In order to evaluate this possibility, I sought to determine how many of the

65 buyouts in my sample are MBOs (management buyouts), since the self-

interested manager seeking a profitable buyout would only reap personal

rewards in the case of an MBO.

This approach has been applied by Long and Ravenscraft (1993a),

and represents a particularly controversial aspect of buyouts since

managers are insiders that have access to privileged information. However,

I found that only 7 of the 65 buyouts in my sample are MBOs, and

therefore conclude that endogeneity may be a lesser concern in the

interpretation of the results of this study.

Alternative Measures. First, the POE measure employed in this

study is based on asset inefficiency following Carton and Hofer (2006).

Yet, measures based on personnel inefficiency or productivity may also be

considered as potential antecedents of buyout activity (e.g., Fox and

Marcus, 1992). I have therefore developed two measures to reflect this

concept.

78
 The first one is Employee Inefficiency, defined as number of

employees per unit of net sales, gauging the level of overall employee

productivity in line and support functions of firms. Baker and Wruck

(1989) and Palepu (1990) argued that a post-buyout reorganization may

lead to stronger operational performance following personnel reductions in

the buyout target due to the removal of hierarchical layers, the elimination

of communication bottlenecks, a faster flow of information, and a quicker

pace of decision-making.

Second, a related albeit different personnel-centered measure is

Overheads Inefficiency in the form of SG&A (sales, general and

administrative expenses scaled by net sales), which gauges personnel

productivity in support functions only. As shown on Tables 5 and 6 in the

Appendix, however, no empirical support was found for either employee or

overheads inefficiency as a predictor of buyout activity.

Third, variables associated with the general market conditions

affecting overall buyout activity may also influence the regression results

presented above. Whereas the year fixed effects employed in my

regressions may capture some of these general market conditions, buyout

firm-level variables such as the availability of equity and debt capital or

79
recent experience with buyouts in specific industry factors may also

influence the results presented above. Therefore, I have included two

additional variables. Dry Powder gauges the overall availability of equity

capital in the buyout industry available from Thomson Financial. Credit

Spread, defined as the interest rate spread between US government bonds

and junk (speculative) bonds, measures the ease of access to debt financing

for buyouts. These data are available from the Federal Reserve Board.

However, as Table 7 shows, these additional variables did not improve

model fit above and beyond the year fixed effects, which had to be

excluded due to multicollinearity.

Next, Table 8 in the Appendix offers a robustness check without the

industry fixed effects reported on the table showing mains results above.

The results are qualitatively unchanged.

Further, I examine more closely on Tables 9 through 13 in the

Appendix each one of the four measures employed to build the one-factor

construct (asset inefficiency) which gauges POE as utilized on the main

regressions reported above. As the Pearson correlation coefficients on

Table 9 in the Appendix show, each one of the four measures (cash

inefficiency, working capital inefficiency, fixed-asset inefficiency and

80
total-asset inefficiency) are highly correlated with the one-factor construct

asset inefficiency (p<.001).

I therefore substitute the one-factor construct for each one of the

four measures on Tables 10, 11, 12 and 13 (also shown in the Appendix).

The results show that two of these measures (liquidity inefficiency and

working capital inefficiency) lower the probability of a buyout event as the

coefficients of their direct effects are negative and significant (p<.05),

whereas another measure (fixed-asset inefficiency) increases the

probability of a buyout event as its regression coefficient is positive and

significant (p<.05). These results highlight an opportunity for future

research, which I elaborate below.

On Tables 14 and 15 in the Appendix, I check for the robustness of

the results presented on the main regressions‟ table above by building on

the measure of free cash flows as applied by Lehn and Poulsen (1989).

Whereas this measure has been applied in buyout studies and in broader

corporate governance studies, it does not include what may be considered a

key input in the computation of free cash flows: capital expenditures. I

therefore create an adjusted measure of free cash flows by subtracting

capital expenditures (scaled by net sales) from the measure employed by

81
Lehn and Poulsen (1989). Pearson correlations with variables of interest

are shown on Table 14 in the Appendix. On Table 15 in the Appendix, I

provide Cox regression analyses with this adjusted measure of free cash

flows. The results are qualitatively similar to the ones reported with the

main regressions above.

Finally, I also provide robustness checks for the model specification

employed in this study. Specifically, prior work in the buyouts literature

stream has relied on Logit regressions to examine the antecedents of

buyout activity. As mentioned above, these models cannot accommodate

competing risks. Yet, the question remains as of whether the results

presented above are sensitive to model specification. On Table 16, I re-

estimate the main regressions reported above with Logit models, and find

that the results are qualitatively unchanged. Further, the extreme rareness

of buyouts in this sample (65 buyouts in a sample of 1459 firms and 8631

years) may bias the results of Logit regressions.

King and Zeng (1999) warn against the use of Logit models in rare-

event samples and propose the application of RE (Rare Events) Logit

models instead. On Table 17 in the Appendix, I therefore re-estimate the

82
main regressions reported above with RE-Logit models, and find that the

results remain qualitatively unchanged.

83
CHAPTER 5

Discussion and Implications

In the empirical study shown above, I examined the interplay

between governance engineering and operational engineering in the

context of buyouts. In doing so, I aimed to address two questions

associated with the antecedents of buyout activity. First, prior work (e.g.,

Singh, 1990; Opler and Titman, 1993) relied on proxies rather than direct

measures of agency costs which gauge a firm‟s potential for governance

engineering. In this empirical study, I offered a more direct test of the

agency-based argument by using measures such as board independence,

CEO duality and insider equity ownership to predict buyout activity.

The second question I aimed to address with this study centered on

Kaplan‟s (2007) proposition on operational engineering. I thus examined

potential antecedents of PE buyouts, testing Kaplan‟s (2007) proposition

while controlling for agency-related antecedents. In empirical tests where I

controlled for competing risks with a Cox proportional hazards model, I

did not find strong support for either proposition as an antecedent of

buyout activity.

84
 This research is associated with potential contributions to and

implications for corporate governance theory and practice. It is also

associated with several limitations, which future research may address. I

start by elaborating on contributions and implications as follows.

Contributions

This research offers at least four contributions to corporate

governance research in general and buyouts in specific.

Contributions to Theory. In this study, I tested a new theory

proposed by Kaplan (2007), according to whom POE (potential for

operational engineering) is a predictor of buyout activity in addition to

agency conditions first proposed by Jensen (1986). Taken as a whole, the

evidence I found failed to lend strong support for either proposition as an

antecedent of buyout activity in my sample of buyouts events from 1998

through 2007.

Yet, I found statistical significance in one specific instance. In a

context of high asset inefficiency (high POE) and high board

independence, the probability of buyout is decreased. This finding is

perhaps not counterintuitive. One potential explanation for this finding is

85
that buyout executives will avoid buying out targets that underperform

operationally and have independent boards at the same time because

buyout executives are more confident about their ability to address

governance problems (board monitoring) than to solve operational

problems, which – in addition to requiring expert knowledge and

experience – also require significantly more attention and effort by buyout

executives to address. Another potential explanation is that buyout

executives behave in a risk-averse fashion when considering buyout targets

that seem problematic operationally.

Nevertheless, taken as a whole, the lack of strong support for extant

theory (FCF-based and POE-based) as derived from my findings is more

consistent with the perspective offered by a manager who was interviewed

as part of this study. This interviewee is a successful founder-CEO who

decided to sell his manufacturing company to a prestigious buyout firm in

the early to mid 2000s. When asked whether the new owners were able to

add value in the form of advice in addition to the typical control

(monitoring) role, the interviewee reacted quite strongly:

You need to understand something. These buyouts

guys are some of the smartest people you will ever find, but

86
 I would never invite them to come see me on-site and give
me advice on how to run my business. They sure know a lot
about finance, but they don’t have a clue about the simplest
of principles on how to run an industrial operation and
handle messy issues with employees or clients. They would
destabilize my operation if they were allowed to get their
hands dirty. This is why I made sure that all our meetings
were in hotels or restaurants – before, during and after the
buyout.

Contributions to Measures. In this study, I employed more direct

measures of agency problems in the form of board independence, CEO

duality and insider equity ownership, rather than the agency proxy

employed in previous work in the form of free cash flows scaled by net

sales. Whereas high free cash flows may serve as a measure of agency

problems, it is also likely that this measure is associated with other

conditions which have little to do with agency problems.

This reasoning is consistent with descriptive statistics in my study,

in which free cash flow was not significantly correlated with board

independence or CEO duality, and was only weakly correlated with insider

equity ownership (p < .10). At the same time, the correlation between free

87
cash flow and asset inefficiency was highly significant (-.79, p < .001),

suggesting that prior work on buyouts based on proxy measures of agency

costs may have rather offered tests of the POE argument instead.

In light of prior work on 1980s‟ LBOs, one way to interpret my

findings is that free cash flow as an agency proxy was an adequate measure

in the 1980s, but probably in lesser degree from the 1990s onwards given

the evolution of US corporate governance as described by Holmstrom and

Kaplan (2003). An alternative interpretation of my finding is that free cash

flow was never an adequate measure of agency problems to begin with.

My study does not offer direct evidence that could answer this question,

which would require data from the 1980s onwards.

Contributions to Methods. This study departs methodologically from

prior work on the antecedents of buyout activity. Prior work has relied on

logit models to estimate the probability of buyout. Whereas adequate given

the nature of the binary dependent variable employed, logit models do not

accommodate situations in which competing risks must be taken into

account. In this study, I thus control for competing risks of M&A when

examining buyout activity with a Cox proportional hazards model.

Whereas the FCF- and POE-based arguments may lead to buyout activity,

88
they may also lead to M&A by other financial acquirers (e.g. banks) or

even strategic acquirers. If this is valid, then the proportional hazards

approach represents a contribution to the buyout research stream.

Contributions to the Phenomenon. Finally, I extend prior work on

LBOs in strategic management (Singh, 1990; Wiersema and Liebeskind,

1995) to examine the more recent PE buyout wave. Anecdotal evidence

has suggested that the more recent PE buyout wave is rather different from

1980s‟ LBOs. In the 1980s, the typical buyout transaction was usually

hostile, with heavy use of leverage, pursued by one buyout firm, and with

the goal to implement governance engineering. By the 2000s, the typical

buyout transaction was arguably friendly, with limited use of leverage,

pursued jointly by more than one buyout firm, and with the goal to

implement operational engineering. The question as of how different

1980s‟ LBOs are from 2000s‟ PE buyouts remains ultimately open. This

study offers an initial contribution to this question.

Implications

The empirical results of the present study also lead to implications

for theory and practice, as explained below.

89
 Implications for Theory. From a theoretical perspective, this study

leads to several implications. An immediate implication that follows from

the present study is related to the antecedents of PE buyouts. If the present

study leads to the conclusion that FCF theory does not predict which firms

become buyout targets, then strategic management researchers may look

for alternative explanations for the present-day PE phenomenon. As noted

above, it is possible that FCF was germane in explaining buyouts in the

1980s, but became less important from the 1990s onwards given the

profound changes in US corporate governance culminating with the

passage of Sarbanes Oxley in 2002. As explained by Holmstrom and

Kaplan (2003), internal and external governance mechanisms in US

corporations became progressively stronger from the 1990s onwards,

prompting Kaplan (1997) to produce a working paper entitled “we are all

Henry Kravis now”.

The second implication is linked with the consequences of private

equity ownership in situations where the new owners are more interested in

short-term financial gains than long-term firm competitiveness. Whilst

most research on buyouts evaluates consequences from the perspective of

key stakeholders of the firm such as equity investors, management,

90
employees, bondholders and tax authorities, the discussion has yet to focus

on the consequences to the firm as an on-going concern. This question is of

central importance to strategic management scholars. The present study

raises the question that buyouts may be associated with financial

improvements only rather than operational improvements. If private equity

buyouts lead to financial improvements only, then the economic logic of

these transactions must be questioned since they have no effect on firm

competitiveness.

A third implication concerns the relationship of risk and return, a

topic of interest in strategic management (e.g. Bowman, 1980; March and

Shapira, 1987; Andersen, Denrell and Bettis, 2005). A source of

controversy in PE buyouts is the significant use of leverage and the

resulting increase in risk in the buyout target. PE buyouts offer an

interesting context into which the debate on the relationship between risk

and return may be extended. Relatedly, a study commissioned by the Yale

Investments Office showed how buyouts, in the absence of operational

improvements, simply increase risk in the buyout target without offering

much in return. This study of 542 buyouts from 1987 to 1998 documented

a gross return to investors of 48% per year, equivalent to a 36% return net

91
of PE management fees and general partner compensation. This return

would have widely surpassed the return of comparably-sized investments

in the S&P 500 stock index, which delivered 17% per year in the same

period. However, this comparison does not adjust for the high levels of

leverage typically associated with PE buyouts. Comparably-timed,

comparably-sized and comparably-leveraged investments in the S&P 500

would have produced an impressive 86% annual return in the same period,

widely surpassing PE returns. Potential accusations of biased sampling in

this study are counterweighted by the fact that the data was gathered from

PE firms soliciting business from Yale. David Swensen, the longtime chief

investment officer of the Yale Endowment, concludes that (2005:135):

Pure financial engineering represents a commodity,

easily available to marketable securities investors through
margin accounts and futures markets. Buyout managers
deserve scant incremental compensation for adding debt
to corporate balance sheets.

However, this debate is far from over as Groh and Gottschalg (2006)

found opposite results in a study of 199 US buyouts from 1984 to 2004. I

echo Cumming, Siegel and Wright (2007) in stating that more research is

92
warranted on the intriguing relationship between risk and return in the

context of PE buyouts.

A fourth implication relates to an implicit assumption underlying the

FCF argument: the notion that managers are self-interested and will

rationally take courses of action that serve their interests, sometimes at the

expense of shareholder interests. Yet, managerial behavior has also been

examined from other perspectives. For example, bounded rationality (e.g.

Simon, 1997) offers an alternative explanation for dysfunctional

managerial behavior such as overinvestments in listed corporations.

Specifically, managers may overinvest in value-destroying opportunities

not because of the combination of self-interest and opportunism as

proposed by Jensen (1986) in his classic FCF argument, but due to human

decision-making processes based on heuristics and biases such as

overconfidence (e.g. Kahneman, Slovic and Tversky, 1982).

The analysis of overconfidence relates several branches of the

psychology literature (Malmendier and Tate, 2005), including the tendency

of individuals to consider themselves above average on positive

characteristics (e.g. Kruger, 1999); the tendency of individuals to be too

optimistic about their own future prospects (e.g., Weinstein, 1980); and the

93
tendency of top managers to be highly likely to face low-frequency and

noisy feedback – which are key predictors of biased decision making (e.g.,

Nisbett and Ross, 1980). Top-level executive decisions such as large-scale

investments, merger agreements, or capital restructuring are relatively rare

events in the life of one company, and each project has many distinct

features which make comparison to past experiences difficult (Malmendier

and Tate, 2005). From this perspective, the oil managers that ex post

seemed to have overinvested in capital expenditures in the mid 1980s may

have simply suffered from the adverse consequences of inaccurate oil price

forecasts in the context of long investment cycles that, once started, are

hard to be stopped or reversed.

Finally, moving beyond the specific context of buyouts, this study

may also lead to implications related to the broader question of why firms

underperform. If the role of agency-theoretic arguments (combined with

POE) in explaining firm underperformance are found to be limited in the

context of buyouts, then these relationships should also be re-examined in

a broader set of contexts in corporate governance, such as bankruptcies.

Implications for Practice. The present study also evidences

implications to practitioners in strategic management. In describing ways

94
in which academic researchers may successfully translate their work into

articles that are relevant for managers, McGahan (2007) highlighted the

value of showing that a widely used management practice violates

important principles. This study on buyouts leads to the implication that

PE buyouts may be difficult to reconcile with an important principle in

strategic management, according to which a key component of a firm‟s

strategy is the economic logic of actions undertaken by management (e.g.

Hambrick and Frederickson, 2005). Specifically, if little empirical support

is found for Kaplan‟s (2007) proposition on operational engineering, then

buyout-related benefits are more likely to be purely financial rather than

operational in line with Swensen (2000), potentially leading to the

questioning of the economic logic of buyouts as a governance structure.

Whereas operational engineering may lead to enhanced firm

competitiveness, financial engineering in the form of more debt may not.

A second implication from a practical perspective follows from an

intriguing question, namely whether the operational improvements

potentially associated with buyouts can be obtained without taking the

target private in the first place (Jensen, Kaplan, Ferenbach, Feldberg,

Moon, Hoesterey, Davis and Jones, 2006). At the heart of this debate is the

95
extent to which internal governance mechanisms may serve as substitutes

to their external counterparts (e.g. Walsh and Seward, 1990). This is an

intriguing question yet to be resolved from a practitioner‟s perspective.

Relatedly, a third practical implication concerns a closer

examination of PE firms, which grew in importance rapidly as the buyout

phenomenon disseminates throughout the US. Strategic management

scholars may be interested in examining in PE firms the very same agency

problems that some PE executives claim to solve in buyout targets. As

noted, the relationship between general partners (GPs) and limited partners

(LPs) is prone to severe agency problems. Poorly structured PE

arrangements often produce misaligned interests between GPs and LPs. As

an example, some GPs may receive carried interests of 20% of partnership

profits regardless of buyout performance, creating the incentive to increase

buyout volume without worrying about returns (e.g. Swensen, 2000). It is

thus not surprising to note that the relationship between GPs and LPs

remains a high-potential domain for scholars interested in corporate

governance.

96
Limitations and Opportunities for Future Research

This research is also associated with limitations that future research

may address. A first limitation as well as opportunity for future research

relates to my dataset. As noted, the data is time-varying and target-level.

Yet, it is possible that additional antecedents of buyout activity may also

involve dyad-level (buyout target and acquirer) elements. In these cases,

other research methods such as state-based sampling (e.g., Folta and

O‟Brien, 2004) are more adequate.

A second limitation as well as opportunity also relates to the fact

that my data source includes listed (publicly-traded) whole-firms only,

given the agency-theoretic backdrop of this study. Whereas I focus on

going-private (public to private) buyouts, buyout transactions are also a

common practice in private firms. This places limits to the potential for

generalization of this study. Indeed, for some types of private firms such as

family businesses, buyouts are sometimes the most viable path for a

succession plan leading to a change in ownership (Rickertsen, 2001).

Further, the characteristics of private-to-private buyouts are quite unique

(e.g., Malone, 1989; Zahra, 1995). Future research may therefore examine

97
the antecedents of buyouts of privately-held firms. This research

opportunity could requite primary data sources.

A third potential avenue for future research follows from a key

assumption that underlies the present work, namely that buyout firms

would actively seek candidate targets with high potential for operational

engineering. A related, equally intriguing question relates to what actually

happens to operations in the buyout target after the transaction is complete.

Whereas this question is beyond the scope of the present study which

focuses on the antecedents of buyouts, the question on the consequences of

buyouts is also of high importance to corporate governance scholars.

Specifically, an alternative way to test Kaplan‟s (2007) proposition would

be to focus on post-buyout operations. This research opportunity would

require data sources that span pre- and post-buyout observations, which

could be challenging given the nature of going-private transactions.

Finally, corporate governance researchers interested in buyouts

might also consider the development of new theory, if a combination of the

traditional agency-theoretic arguments and the more recent POE

proposition fails to be associated with strong empirical support. A common

theme that emerged anecdotally is how partners at buyout firms may

98
behave opportunistically when screening and chasing potential buyout

targets. If this is valid, then a social networks perspective might be useful

in further examining buyout activity. Whereas economic drivers have

traditionally been used to explain buyout activity in the extant literature,

little research has considered firms‟ embeddedness in social and economic

contexts that may enable buyouts. In a broader context, some researchers

(e.g., Haunschild, 1993; Haunschild and Beckman, 1998) have examined

the role of social context and organizational embeddedness in M&A

transactions. These theoretical considerations have yet to be extended to

the buyouts literature.

99
REFERENCES

Alchian A, Demsetz H. 1972. Production, information costs and economic

organization. American Economic Review, 62: 777-795.

Allison P. 1984. Event history analysis: regression for longitudinal data.

Sage: Thousand Oaks, CA.

Amihud Y, Lev B. 1981. Risk reduction as a managerial motive for

conglomerate mergers. Bell Journal of Economics, 12: 605-617.

Andersen T, Denrell J, Bettis R. 2007. Strategic responsiveness and

Bowman‟s paradox. Strategic Management Journal, 28:407-429.

Baker G, Wruck K. 1989. Organizational changes and value creation in

leveraged buyouts: the case of the O.M. Scott & Sons Company.
Journal of Financial Economics, 25: 163-190

Barney J. 1991. Firm resources and sustained competitive advantage.

Journal of Management, 17: 99-120.

Baumol W. 1959. Business behavior, value and growth. McMillan: New

York, NY.

Bhagat S, Shleifer A, Vishny R. 1990. Hostile takeovers in the 1980s: the

return to corporate specialization. Brookings Papers on Economic
Activity: Microeconomics, Special Issue, 1-72.

Berle A, Means G. [1932] 1968. The modern corporation and private

property. Harcourt, Brace & World: New York, NY.

Blossfeld H, Rower G. 1995. Techniques of event history modeling: new

approaches to causal analysis. Erlbaum: Mahwah, NJ.

Bowman E. 1980. A risk-return paradox for strategic management. Sloan

Management Review, 21:17-31.

100
Bowman E, Singh H. 1993. Corporate restructuring: reconfiguring the firm.
Strategic Management Journal, 14: 5-14.

Cannella A, Finkelstein S, Hambrick D. 2008. Strategic leadership: theory

and research on executives, top management teams, and boards.
Oxford University Press, in press.

Cannella A, Holcomb T. 2005. A multilevel analysis of the upper-echelons

model. In A. Dansereau, & F. J. Yammarino (Eds.), Research in
Multi-Level Issues, Vol. 4: 197-237. Elsevier: Oxford, UK.

Cao J, Lerner J. 2007. The performance of reverse leveraged buyouts.

Working paper.

Carter C, Lorsch J. 2004. Back to the drawing board. Harvard Business

School Publishing: Boston, MA.

Carton R, Hofer C. 2006. Measuring organizational performance. Edward

Elgar: Northampton, MA.

Chirinko R. 1987. Tobin‟s Q and financial policy. Journal of Monetary

Economics, 19: 69-87.

Comment R, Schwert G. 1995. Poison or placebo? Evidence on the

deterrence and wealth effects of modern antitakeover measures. Journal
of Financial Economics, 39: 67-88.

Couglan A, Schmidt R. 1985. Executive compensation, managerial

turnover, and firm performance: An empirical investigation. Journal
of Accounting and Economics, 7: 43-66.

Cox D. 1972. Regression models and life-tables. Journal of the Royal

Statistical Society (Series B), 34: 187–220.

Cox D, Oakes, D. 1984. Analysis of survival data. Chapman and Hall:

London.

101
Cressy R, Munari F, Malipiero A. 2007. Playing to their strengths?
Evidence that specialization in the private equity industry confers
competitive advantage. Journal of Corporate Finance, 13: 647–669.

Cumming D, Siegel D, Wright M. 2007. Private equity, leveraged buyouts

and governance. Working paper.

Daily C, Dalton D, Cannella A. 2003. Corporate governance: decades of

dialogues and data. Academy of Management Review, 28: 371-382.

Dalton D, Daily C, Ellstrand A, Johnson J. 1998. Meta-analytic reviews of

board composition, board leadership structure, and financial
performance. Strategic Management Journal, 19: 269-290.

Davis S. 2007. Private equity’s impact: productivity and labor market

effects. Working paper presented at the American Enterprise Institute
for Public Policy Research, Washington DC, November 27-28.

Donaldson G. 1994. Corporate restructuring. Harvard Business School

Press: Cambridge, MA.

Douglas E. 2007. The Forbes 400. Forbes Magazine, October 8.

Easterbrook F, Fischel D. 1991. The economic structure of corporate law.

Harvard University Press: Cambridge, MA.

Eisenhardt K. 1989. Agency theory: an assessment and review. Academy of

Management Review, 14: 195-223.

Fama E. 1980. Agency problems and the theory of the firm. Journal of
Political Economy, 88: 288-307.

Fama E, Jensen M. 1983. Separation of ownership and control. Journal of

Law and Economics, 26: 301-325.

102
Finkelstein S, D‟Aveni R. 1994. CEO duality as a double-edged sword: how
boards of directors balance entrenchment avoidance and unity of
command. The Academy of Management Journal, 37: 1079-1108.

Folta T, O‟Brien J. 2004. Entry in the presence of dueling options. Strategic

Management Journal, 25: 121-138.

Fox I, Marcus A. 1992. The causes and consequences of leveraged

management buyouts. Academy of Management Review, 17: 62-85.

Gonzalez P, Vasconcellos G, Kish R. 1998. Cross-border mergers and

acquisitions: the undervaluation hypothesis. Quarterly Review of
Economics and Finance, 38: 25-45.

Grant R. 1989. Contemporary strategy analysis. Basic Blackwell:

Cambridge, MA.

Greene W. 2000. Econometric analysis. Prentice-Hall: Upper Saddle River,

NJ.

Groh A, Gottschalg O. 2006. The risk-adjusted performance of US

buyouts. Working paper.
Guo S, Hotchkiss E, Song W. 2007. Do buyouts (still) create value?
Working paper.

Gupta A, Rosenthal L. 1991. Ownership structure, leverage, and firm

value: the case of leveraged recapitalizations. Financial Management,
20: 69-83.

Hall B. 2002. The financing of research and development. Oxford Review

of Economic Policy, 18: 35-51.

Hambrick D, Fredrickson J. 2005. Are you sure you have a strategy?

Academy of Management Executive, 19: 51-62.

103
Harper N, Schneider A. 2004. Private equity‟s new challenge. McKinsey
Quarterly, Summer, 1-6.

Haunschild P. 1993. Interorganizational imitation: the impact of interlocks

on corporate acquisition activity. Administrative Science Quarterly,
38: 564-592.

Haunschild P, Beckman C. 1998. When do interlocks matter? Alternate

sources of information and interlock influence. Administrative
Science Quarterly, 43: 815-844.

Heckman J. 1979. Sample selection bias as a specification error.

Econometrica, 47:153-161.

Hendry J. 2002. The principal‟s other problems: honest incompetence and

the specification of objectives. Academy of Management Review, 27:
98-113.

Holderness C, Kroszner R, Sheehan D. 1999. Were the good old that

good? Changes in managerial stock ownership since the great
depression. Journal of Finance, 54: 435-470.

Holmstrom B, Kaplan S. 2003. Corporate governance and takeovers in the

US: making sense of the „80s and „90s. Journal of Economic
Perspectives, 15: 121-144.

Holmstrom B, Kaplan S. 2001. The state of US corporate governance:

what’s right and what’s wrong? Working paper.

Holmstrom B, Milgrom P. 1991. Multitask principal agent analyses:

Incentive contracts, asset ownership and job design. Journal of Law,
Economics and Organization, 7: 24-52.

Hoskisson R, Hitt M. 1994. Downscoping: how to tame the diversified

firm. Oxford University Press: New York, NY.

104
Hubbard G. 2007. The history, impact and future of private equity.
Presented at the American Enterprise Institute for Public Policy
Research, Washington DC, November 27-28.

Ilinitch A, Zeithaml C. 1995. Operationalizing and testing Galbraith‟s

center of gravity theory. Strategic Management Journal, 16: 401-410.

Jensen M. 1986. Agency costs of free cash flow, corporate finance and
takeovers. American Economic Review, 76: 323-329.

Jensen M. 1989. Eclipse of the public corporation. Harvard Business Review,

67: 61-74.

Jensen M. 1991. Corporate control and the politics of finance. Journal of

Applied Corporate Finance, 4: 13-33.

Jensen M. 1993. The modern industrial revolution, exit, and the failure of
internal control systems. Journal of Finance, 48: 831-880.

Jensen M, Kaplan S, Ferenbach C, Feldberg M, Moon J, Hoesterey B,

Davis C, Jones A. 2006. Morgan Stanley roundtable on private equity
and its import for public companies. Journal of Applied Corporate
Finance, 18:8-37.

Jensen M, Meckling W. 1976. Theory of the firm: managerial behavior,

agency costs, and ownership structure. Journal of Financial Economics,
3: 305-360.

Jensen M, Murphy K. 1990. Performance pay and top-management

incentives. Journal of Political Economy, 98: 225-264.

Kahneman T, Slovic P, Tversky A. 1982. Judgment under uncertainty:

heuristics and biases. Cambridge, UK: Cambridge University Press.

Kaplan S, 1989. The effects of management buyouts on operating

performance and value. Journal of Financial Economics, 24: 217-254.

105
Kaplan S. 1991. The staying power of leveraged buyouts. NBER working
paper #3653.

Kaplan S. 1997. The evolution of US corporate governance: we are all

Henry Kravis now. Working paper.

Kaplan S. 2007. Private equity: past, present and future. Journal of Applied
Corporate Finance, 19: 8-16.

Kaplan S, Schoar A. 2005. Private equity performance: returns, persistence,

and capital flows. Journal of Finance, 50: 1791-1823.

Kaplan S, Stein J. 1993. The evolution of buyout pricing and financial

structure in the 1980s. Quarterly Journal of Economics, 108: 313-357.

Kaplan S, Zingales L. 1997. Do investment-cash flow sensitivities provide

useful measures of financing constraints? Quarterly Journal of
Economics, 112: 159-216.

Kerr J, Bettis R. 1987. Boards of directors, top management compensation,

and shareholder returns. Academy of Management Journal, 30: 645-
664.

King G, Langche Z. 1999. Logistic Regression in Rare Events Data.

Department of Government, Harvard University, available from
[Link] .

Kosnik R. 1987. Greenmail: a study of board performance in corporate

governance. Administrative Science Quarterly, 32: 163-85.

Kruger J. 1999. Lake Wobegon be gone! The „below-average effect‟ and

the egocentric nature of comparative ability judgments. Journal of
Personality and Social Psychology, 77:221–232.

106
Lambert R, Larcker D. 1985. Executive compensation, corporate decision-
making, and shareholder wealth: a review of the evidence. Midland
Corporate Finance Journal, 2: 6-22.

Lee L. 1983. Generalized econometric models with selectivity.

Econometrica, 51:507-512.

Lehn K, Poulsen A. 1989. Free cash flow and stockholder gain in going
private transactions. Journal of Finance, 44: 771-788.

Leiblein M, Reuer J, Dalsace F. 2002. Do make or buy decisions matter?

The influence of organizational governance on technological
performance. Strategic Management Journal, 23: 817-833.

Leland H, Pyle D. 1977. Informational asymmetries, financial structure,

and financial intermediation. Journal of Finance, 32: 371-387.

Lichtenberg F, Siegel D. 1991. The effects of leveraged buyouts on

productivity and related aspects of firm behavior. Journal of Financial
Economics, 27: 165-194.

Liebeskind J, Opler T. 1994. The causes of corporate refocusing: evidence

from the 1980s. Working paper.

Liebeskind J, Wiersema M, Hansen G. 1992. LBOs, corporate restructuring,

and the incentive-intensity hypothesis. Financial Management, 21: 73-
88.

Long W, Ravenscraft D. 1993a. The financial performance of whole

company LBOs. Center for Economic Studies discussion paper, 93-16.

Long W, Ravenscraft D. 1993b. LBOs, debt and R&D intensity. Strategic

Management Journal, 14: 119-136.

Lorsch J, MacIver E. 1989. Pawns or potentates: The reality of America’s

corporate boards. Harvard University Press: Boston, MA.

107
Lowenstein L. 1985. Management buyouts. Columbia Law Review, 85: 730-
784.

Lubatkin M, Chatterjee S. 1991. The strategy-shareholder value

relationship: Testing temporal stability across market cycle. Strategic
Management Journal, 12: 251-270.

Mace M. 1971. Directors: Myth and reality. Harvard University Press:

Boston, MA.

Mallette P, Fowler K. 1992. Effects of board composition and stock

ownership on the adoption of poison pills. Academy of Management
Journal, 35: 1010-1035.

Malmendier U, Tate G. 2005. Does overconfidence affect corporate

investment? CEO overconfidence measures revisited. European
Financial Management, 11:649–659.

Malone S. 1989. Characteristics of smaller company leveraged buyouts.

Journal of Business Venturing, 4: 349:359.

Manne H. 1965. Mergers and the market for corporate control. Journal of
Political Economy, 73: 110-120.

March J, Shapira Z. Managerial perspectives on risk and risk taking.

Management Science, 33:1404-1418.

Markides C, Williamson P. 1996. Corporate diversification and

organizational structure: A resource-based view. Academy of
Management Journal, 39: 340-367.

Marris R. 1964. The economic theory of managerial capitalism. Free Press:

Glencoe, IL.

108
McGahan, A. 2007. Academic research that matters to managers: on
zebras, dogs, lemmings, hammers, and turnips. Academy of
Management Journal, 50:748-753.

Meerkatt H, Rose J, Brigl M, Liechtenstein H, Prats M, Herrera A. 2008.

The advantage of persistence. The Boston Consulting Group and
IESE University of Navarra working paper.

Mizruchi M. 1983. Who controls whom? An examination of the relation

between management and board of directors in large American
corporations. Academy of Management Review, 8: 426-435.

Montgomery C. 1994. Corporate diversification. Journal of Economic

Perspectives, 8: 163-178.

Montgomery C, Wernerfelt B. 1988. Diversification, Ricardian rents, and

Tobin‟s Q. Rand Journal of Economics, 19: 623-632.

Morck R, Shleifer A, Vishny R. 1989. Alternative mechanisms for

corporate control. American Economic Review, 79: 842-852.

Muscarella C, Vetsuypens M. 1990. Efficiency and organizational

structure: a study of reverse LBOs. Journal of Finance, 45: 1389-
1414.

Myers S. 1984. The capital structure puzzle. Journal of Finance, 39: 575-
592.

Neter J, Wasserman W, Kutner M. 1985. Applied Linear Statistical Models

(2nd edition), Homewood, IL: Irwin.

Nilikant V, Rao H. 1994. Agency theory and uncertainty in organizations:

An evaluation. Organization Studies, 15: 649-672.

Nisbett R, Ross L. 1980. Human Inference: Strategies and Shortcomings of

Social Judgment. Englewood Cliffs NJ: Prentice-Hall.

109
Opler T. 1992. Operating performance in leveraged buyouts: evidence from
1985-1989. Financial Management, 21: 27-34.

Opler T, Titman S. 1993. The determinants of leveraged buyout activity: free

cash flow vs. financial distress costs. Journal of Finance, 48: 1985-
1999.

Palepu K. 1985. Diversification strategy, profit performance, and the entropy

measure. Strategic Management Journal, 6: 239-255.

Palepu K. 1990. Consequences of leveraged buyouts. Journal of Financial

Economics, 27: 247-262.

Penrose E. 1959. The theory of the growth of the firm. M. E. Sharpe: White
Plains, NY.

Perry T, Zenner M. 2000. CEO compensation in the 1990s: shareholder

alignment or shareholder expropriation? Wake Forest Law Review, 35:
123-152.

Pfeffer J. 1981. Power in organizations. Pitman: Boston, MA.

Porter M. 1985. Competitive advantage. Free Press: New York, NY.

Phan P, Hill C. 1995. Organizational restructurings and economic

performance in leveraged buyouts: an ex post study. Academy of
Management Journal, 38: 704-739.

Rajan R, Wulf J. 2006. Are perks purely managerial excess? Journal of

Financial Economics, 79: 1-33.

Rappaport A. 1990. The staying power of the public corporation. Harvard

Business Review, 68: 96-104.

Rickertsen R. 2001. Buyout. AMACOM Books: New York, NY.

110
Rumelt R. 1974. Strategy, structure, and economic performance. Harvard
University Press: Cambridge, MA.

Rumelt R. 1984. Towards a strategic theory of the firm. In R. Lamb (ed.),

Competitive strategic management. Prentice-Hall: New Jersey, NJ.

Schendel D, Hofer C. 1979. Strategic management: a new view of business

policy and planning. Little Brown: Boston, MA.

Servaes H. 1991. Tobin‟s Q and the gains from takeovers. Journal of

Finance, 46: 409-419.

Servaes H. 1994. Do takeover targets overinvest? Review of Financial

Studies, 7: 253-278.

Shleifer A, Summers L. 1988. Breach of trust in hostile takeovers, in

Corporate takeovers: causes and consequences, A. Auerbach (ed.).
University of Chicago Press: Chicago, IL.

Shleifer A, Vishny R. 1990. The takeover wave of the 1980s. Science, 249:
745-749.

Shleifer A, Vishny R. 1997. A survey of corporate governance. Journal of

Finance, 52: 737-783.

Simon H. 1957. Models of man: social and rational. Wiley: New York,
NY.

Simon H. 1991. Organizations and markets. Journal of Economic

Perspectives, 5: 25-44.

Singh H. 1990. Management buyouts: distinguishing characteristics and

operating changes prior to public offering. Strategic Management
Journal, 11: 111-129.

Smith A. 1990. Corporate ownership structure and performance: the case

of management buyouts. Journal of Financial Economics, 27: 143-164

111
Stein J. 2003. Agency, information and corporate investment, in Handbook
of the Economics of Finance, G. Constantinides, M. Harris & R. Stulz
(eds.). Elsevier: Amsterdam.

Swensen D. 2000. Pioneering portfolio management. Free Press: New York,

NY.

Teece D. 1980. Economies of scope and the scope of the enterprise. Journal
of Economic Behavior and Organization, 1: 223-247.

Turner S, Mitchell W, Bettis R. 2008. Responding to rivals and

compliments: how market concentration shapes generational product
innovation strategy. Working paper.

The Wall Street Journal. 2006. Hot topic: going private. Editorial, June 3.

Tobin J. 1969. A general equilibrium approach to monetary policy. Journal

of Money, Credit and Banking, 1: 15-29.

Vance S. 1983. Corporate leadership: Boards, directors and strategy.

McGraw Hill: New York, NY.

Van Mourik G. 2007. Information note for practitioners: Research

opportunities and issues in the measurement and determination of
organizational effectiveness. Monash Business Review, 3(3): 1-9.

Walsh J, Seward J. 1990. On the efficiency of internal and external

corporate control mechanisms. Academy of Management Review, 15:
421-58.

Weinstein N. 1980. Unrealistic optimism about future life events. Journal

of Personality and Social Psychology, 39:806–820.

Weir C, Laing D, Wright M. 2005. Undervaluation, private information,

agency costs and the decision to go private. Applied Financial
Economics, 15: 947-961.

112
Whisler T. 1984. Rules of the game. Dow Jones – Irwin: Homewood, IL.

Wiersema M, Liebeskind J. 1995. The effects of leveraged buyouts on

corporate growth and diversification in large firms. Strategic
Management Journal, 16: 447-460.

Williamson O. 1964. The economics of discretionary behavior:

managerial objectives in a theory of the firm. Prentice-Hall:
Englewood Cliffs, NJ.

Williamson O. 1985. The economic institutions of capitalism. Free Press:

New York, NY.

Zahra S, Fescina M. 1991. Will leveraged buyouts kill US corporate research

& development? Academy of Management Executive, 5: 7-21.

Zahra S. 1995. Corporate entrepreneurship and financial performance: the

case of management leveraged buyouts. Journal of Business Venturing,
10: 225-247.

113
Appendix: Tables
Table 1. Sample Description by Industrial Sector

2-digit Number of
Description
SIC Buyouts
13 Oil and gas extraction 1
20 Food and kindred products 2
23 Apparel and other textile products 1
27 Printing and publishing 2
28 Chemicals and allied products 2
33 Primary metal industries 1
35 Industrial machinery and equipment 1
36 Electrical and electronic equipment 5
37 Transportation equipment 3
38 Instruments and related products 3
39 Miscellaneous manufacturing industries 1
48 Communications 2
49 Electric, gas, and sanitary services 2
50 Wholesale trade of durable goods 2
51 Wholesale trade of nondurable goods 1
53 General merchandise stores 3
54 Food stores 1
56 Apparel and accessory stores 1
57 Furniture, home furnishings and equipment stores 1
58 Eating and drinking places 2
59 Miscellaneous retail 6
70 Hotels, rooming houses, camps, and other lodging 2
73 Business services 9
75 Automotive repair, services, and parking 1
80 Health services 4
82 Educational services 2
87 Engineering and management services 4
Total 65

114
Table 2. Sample Description by Year

Year Number of Buyouts

1998 0
1999 4
2000 8
2001 1
2002 1
2003 1
2004 3
2005 5
2006 12
2007 30
Total 65

115
Table 3. Variables and Measures

Variable Type Measure

Probability of a buyout (by any type of buyout firm) occurring
Buyout Dependent
at time t, given that the event has not occurred prior to that
Board Percentage of board members that are independent and
Independent
Independence therefore not affiliated directors according to the IRRC
Dummy variable coded as one if the firm‟s CEO also serves as
CEO Duality Independent
chairperson at the same firm
Insider Equity Percentage of equity controlled by the firms‟ directors and
Independent
Ownership officers as provided by the IRRC
Construct obtained from factor analysis of four inefficiency
measures: liquidity inefficiency (cash and equivalents #1
scaled by net sales #12); working capital inefficiency
(working capital #179 scaled by net sales #12); fixed-asset
Asset
Independent inefficiency (property, plant and equipment #7 scaled by net
Inefficiency
sales #12); and total asset inefficiency (total assets #6 scaled
by net sales #12). The factor analysis yielded a one-factor
weighted average with an eigenvalue of 3.38 that explained
84.5% of total variance
Free cash flow divided by firm sales. Free cash flow is
operating income before depreciation (#13), minus total
income taxes (#16), minus the change in deferred taxes from
the previous year to the current year (change in #35), minus
Free Cash Flow Control gross interest expenses on short and long-term debt (#15),
minus the total amount of preferred dividend requirement on
cumulative preferred stock and dividends paid on non-
cumulative preferred stock (#19), minus the total dollar
amount of dividends declared on common stock (#21)
Unrelated Jacquemin-Berry entropy measure of diversification at the 2-
Control
Diversification digit SIC level
Market value divided by book value of assets. Market value is
the firm‟s market value of common stock (#24 * #25), plus
Tobin‟s Q Control
book value of assets (#6), less book value of common equity
(#60), less deferred taxes on the balance sheet (#74)
Leverage Control Total debt (#9 plus #34) divided by total equity (#216)
Taxes Control Income tax expenses (#16) scaled by net sales (#12)
Year Fixed Dummy variables that equal one if the buyout was completed
Control
Effects in the given year, and zero otherwise
Industry Fixed Dummy variables that equal one if the buyout was completed
Control
Effects in the given 2-digit SIC industrial sector, and zero otherwise

116
Table 4. Robustness: Two-Stage Cox Regression (Selection Correction)
Stage II: Stage II: Stage II: Stage II: Stage II:
Independent
Stage I Model I Model II Model III Model IV Model V
Variables
(Controls) (H1-H3) (H4) (H5) (H6)
-.010 2.233 2.436 2.037 2.297 2.619
Free cash flow
(.225) (2.294) (2.307) (2.333) (2.314) (2.353)
Unrelated .124*** .056 .049 .051 .063 .044
diversification (.035) (.151) (.152) (.152) (.153) (.151)
-.235*** -1.431*** -1.442*** -1.504*** -1.501*** -1.461***
Tobin’s Q
(.357) (.373) (.381) (.353) (.385) (.383)
-.024 -.021 -.025 -.019 -.026 -.021
Leverage
(.033) (.129) (.127) (.128) (.128) (.129)
.051* .335 .333 .436† .406 .352
Taxes
(.023) (.227) (.250) (.252) (.249) (.249)
.014 .013 .013 .014 .015
Lambda (λ) ---
(.030) (.030) (.031) (.031) (.031)
Year fixed
1933.31*** 4.71 4.58 4.78 4.56 4.56
effects
Industry fixed
79.50*** 64.78*** 63.70*** 63.99*** 64.43*** 64.11***
effects
Board .095 .090 .089 .098
--- ---
independence (.154) (.154) (.154) (.159)
.026 .028 .039 .027
CEO duality --- ---
(.127) (.127) (.128) (.127)
Insider equity .124 .138 .116 .086
--- ---
ownership (.124) (.126) (.123) (.133)
-.062 -.231 -.141 -.074
Asset inefficiency --- ---
(.479) (.493) (.493) (.456)
Asset inefficiency
-.482*
* board --- --- --- --- ---
(.244)
independence
Asset inefficiency .404
--- --- --- --- ---
* CEO duality (.353)
Asset inefficiency
-.369
* insider equity --- --- --- --- ---
(.415)
ownership
Log Likelihood,
-10787.72 -375.91 -375.37 -373.85 -374.77 -375.05
L(β)
*** *** *** *** ***
Model χ 2 1850.98 132.18 133.26 136.30 134.47 133.89***
Number of
37939 8631 8631 8631 8631 8631
observations
Number of firms 8422 1459 1459 1459 1459 1459
Number of
1459 65 65 65 65 65
events
†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

117
Table 5. Robustness: Cox Regression with Employee Inefficiency

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
2.235 2.449 2.416 2.435 2.502
Free cash flow
(2.295) (2.361) (2.360) (2.358) (2.372)
Unrelated .076 .068 .070 .068 .069
diversification (.144) (.146) (.146) (.146) (.146)
-1.462*** -1.496*** -1.494*** -1.501*** -1.511***
Tobin’s Q
(.368) (.378) (.377) (.378) (.378)
-.029 -.031 -.031 -.032 -.029
Leverage
(.129) (.128) (.128) (.128) (.131)
.337 .369 .365 .369 .372
Taxes
(.227) (.251) (.241) (.251) (.252)
Year fixed effects 30.21*** 28.08*** 27.88*** 28.02*** 27.47***
Industry fixed effects 64.81*** 54.76*** 55.09*** 54.56*** 56.43***
.097 .089 .097 .085
Board independence ---
(.154) (.157) (.158) (.155)
.024 .024 .016 .020
CEO duality ---
(.127) (.128) (.129) (.127)
Insider equity .126 .127 .128 .131
---
ownership (.124) (.123) (.123) (.123)
-.065 -.066 -.062 -.065
Asset inefficiency ---
(.481) (.482) (.481) (.482)
.024 .046 .055 .060
Employee inefficiency ---
(.159) (.176) (.129) (.144)
Employee inefficiency * -.037
--- --- --- ---
board independence (.136)
Employee inefficiency * .289
--- --- --- ---
CEO duality (.820)
Employee inefficiency * .178
--- --- --- ---
insider equity ownership (.174)
Log Likelihood, L(β) -376.03 -375.47 -375.44 -375.35 -374.94
Model χ 2 131.95*** 133.06*** 133.13*** 133.24*** 134.13***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65
†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

118
Table 6. Robustness: Cox Regression with Overheads Inefficiency

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
2.235 2.817 2.782 2.798 2.856
Free cash flow
(2.295) (2.285) (2.284) (2.283) (2.302)
Unrelated .076 .098 .100 .098 .098
diversification (.144) (.148) (.148) (.148) (.148)
-1.462*** -1.581*** -1.579*** -1.586*** -1.591***
Tobin’s Q
(.368) (.383) (.382) (.383) (.383)
-.029 -.025 -.024 -.025 -.022
Leverage
(.129) (.132) (.132) (.131) (.135)
.337 .449† .445† .449† .449†
Taxes
(.227) (.246) (.245) (.245) (.248)
Year fixed effects 30.21*** 27.98*** 27.76*** 27.88*** 27.36***
Industry fixed effects 64.81*** 56.81*** 57.01*** 56.55*** 58.28***
.084 .076 .086 .074
Board independence ---
(.155) (.157) (.155) (.155)
.017 .017 .008 .014
CEO duality ---
(.127) (.128) (.122) (.127)
Insider equity .121 .122 .123 .125
---
ownership (.125) (.125) (.125) (.124)
-.291 -.293 -.288 -.283
Asset inefficiency ---
(.517) (.516) (.517) (.518)
.007 .032 .052 .045
Employee inefficiency ---
(.158) (.177) (.188) (.146)
.370 .371 .371 .359
Overheads inefficiency ---
(.238) (.238) (.239) (.239)
Overheads inefficiency * .039
--- --- --- ---
board independence (.136)
Overheads inefficiency * .055
--- --- --- ---
CEO duality (.130)
Overheads inefficiency * -.169
--- --- --- ---
insider equity ownership (.177)
Log Likelihood, L(β) -376.03 -374.32 -374.28 -374.24 -373.86
Model χ 2 131.95*** 135.36*** 135.44*** 135.54*** 136.29***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65
†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

119
Table 7. Robustness: Cox Regression with Dry Powder and Credit Spread
Model I Model II Model III Model IV
(Controls with (Controls with (H1-H3 with (H1-H3 with
Independent Variables
year fixed dry powder and year fixed dry powder and
effects) credit spread) effects) credit spread)
2.235 2.201 2.481 2.467
Free cash flow
(2.295) (2.273) (2.355) (2.342)
Unrelated .076 .079 .067 .074
diversification (.144) (.144) (.146) (.145)
-1.462*** -1.490*** -1.493*** -1.525***
Tobin’s Q
(.368) (.369) (.377) (.379)
-.029 -.019 -.031 -.019
Leverage
(.129) (.132) (.128) (.132)
.337 .348 .370 .393
Taxes
(.227) (.225) (.251) (.251)
Year fixed effects 30.21*** --- 28.14*** ---
.007** .007**
Dry powder --- ---
(.002) (.002)
-1.096* -1.092*
Credit Spread --- ---
(.477) (.476)
***
Industry fixed effects 64.81 57.27*** 63.79*** 56.92***
.095 .085
Board independence --- ---
(.154) (.152)
.023 .037
CEO duality --- ---
(.127) (.127)
Insider equity .126 .121
--- ---
ownership (.123) (.123)
-.059 -.119
Asset inefficiency --- ---
(.479) (.483)
Log Likelihood, L(β) -376.03 -379.89 -375.48 -379.35
Model χ 2 131.95*** 124.22*** 133.03*** 125.31***
Number of observations 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459
Number of events 65 65 65 65
†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

120
Table 8. Robustness: Cox Regressions without Industry Fixed Effects

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
-.001 -.206 -.302 -.244 -.234
Free cash flow
(.233) (.572) (.606) (.833) (.771)
Unrelated -.029 -.021 -.033 -.018 -.016
diversification (.234) (.134) (.135) (.134) (.135)
-.954** -.978** -1.019*** -.998** -.969**
Tobin’s Q
(.307) (.309) (.312) (.314) (.309)
-.069 -.070 -.069 -.070 -.070
Leverage
(.118) (.117) (.121) (.117) (.118)
.091 .105 .195 .136 .077
Taxes
(.078) (.079) (.124) (.132) (.103)
Year fixed effects 34.87*** 33.07*** 32.91*** 33.09*** 33.22***
-.015 -.014 .017 -.017
Board independence ---
(.147) (.146) (.147) (.137)
.081 .082 .090 .081
CEO duality ---
(.121) (.121) (.122) (.121)
Insider equity .148 .152 .147 .137
---
ownership (.112) (.112) (.112) (.116)
-.161 -.317 -.232 -.123
Asset inefficiency ---
(.378) (.409) (.418) (.375)
Asset inefficiency * -.466*
--- --- --- ---
board independence (.234)
Asset inefficiency * .230
--- --- --- ---
CEO duality (.328)
Asset inefficiency * -.147
--- --- --- ---
insider equity ownership (.318)
Log Likelihood, L(β) -415.13 -413.71 -412.14 -413.44 -413.59
Model χ 2 53.73*** 56.59*** 59.72*** 57.13*** 56.82***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65

†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

121
Table 9: Robustness: Correlation Table for Asset Inefficiency Measures

(1) (2) (3) (4) (5) (6) (7) (8)

1. Free cash flows
2. Board
-.01
independence
3. Insider equity
-.02† -.41***
ownership
4. CEO duality -.01 -.11*** .04***
5. Asset
-.79*** .01 .00 -.01
inefficiency (POE)
6. Cash
-.84*** .01 .02 .01 .96***
inefficiency
7. Working capital
-.82*** .00 .02* .02 .93*** .99***
inefficiency
8. Fixed-asset
-.82*** .03** -.05*** -.01 .72*** .68*** .64***
inefficiency
9. Total-asset
-.82*** .01 -.01 -.01 .98*** .94*** .92*** .81***
inefficiency

n = 8631. † p < .10, * p < .05, ** p < .01, *** p < .001

122
Table 10. Robustness: Cox Regression with Liquidity Inefficiency

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
2.235 .715 .066 .227 -.010
Free cash flow
(2.295) (2.599) (2.524) (2.067) (2.417)
Unrelated .076 .044 .052 .050 .049
diversification (.144) (.145) (.145) (.146) (.145)
-1.462*** -1.441*** -1.466*** -1.445*** -1.459***
Tobin’s Q
(.368) (.379) (.375) (.378) (.376)
-.029 -.015 -.001 -.011 -.001
Leverage
(.129) (.121) (.120) (.121) (.119)
.337 .556* .618** .575* .646
Taxes
(.227) (.261) (.250) (.252) (.244)
Year fixed effects 30.21*** 28.04*** 28.66*** 28.30*** 28.94***
Industry fixed effects 64.81*** 60.45*** 61.42*** 60.99*** 62.02***
.103 .114 .102 .016
Board independence ---
(.154) (.155) (.154) (.158)
.035 .026 .079 .018
CEO duality ---
(.127) (.127) (.136) (.128)
Insider equity .126 .244† .126 .136
---
ownership (.122) (.133) (.122) (.123)
-2.013† -2.208* -2.122† -1.989†
Liquidity inefficiency ---
(1.022) (1.095) (1.143) (1.088)
Liquidity inefficiency * 1.233†
--- --- --- ---
board independence (.669)
Liquidity inefficiency * .632
--- --- --- ---
CEO duality (.678)
Liquidity inefficiency * -1.186*
--- --- --- ---
insider equity ownership (.510)
Log Likelihood, L(β) -376.03 -373.62 -372.55 -373.27 -371.41
Model χ 2 131.95*** 136.76*** 138.90*** 137.46*** 141.18***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65

†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

123
Table 11. Robustness: Cox Regression with Working Capital Inefficiency

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
2.235 .324 -.694 .372 -.585
Free cash flow
(2.295) (2.332) (1.223) (2.644) (1.385)
Unrelated .076 .038 .052 .040 .041
diversification (.144) (.146) (.146) (.146) (.146)
-1.462*** -1.450*** -1.494*** -1.469*** -1.469***
Tobin’s Q
(.368) (.373) (.358) (.380) (.359)
-.029 -.007 -.010 -.010 -.001
Leverage
(.129) (.109) (.111) (.110) (.110)
.337 .577* .641*** .550* .633**
Taxes
(.227) (.224) (.174) (.243) (.189)
Year fixed effects 30.21*** 27.60*** 27.12*** 27.68*** 28.00***
Industry fixed effects 64.81*** 57.92*** 57.53*** 57.29*** 59.36***
.096 .109 .093 .030
Board independence ---
(.154) (.155) (.154) (.161)
.045 .041 .022 .039
CEO duality ---
(.127) (.127) (.148) (.127)
Insider equity .148 .262* .146 .146
---
ownership (.123) (.126) (.123) (.122)
Working capital -2.852** -2.789** -2.711** -2.610**
---
inefficiency (.923) (.853) (.903) (.926)
Working capital
1.357*
inefficiency * board --- --- --- ---
(.588)
independence
Working capital
-.618
inefficiency * CEO --- --- --- ---
(.661)
duality
Working capital
-.702
inefficiency * insider --- --- --- ---
(.612)
equity ownership
Log Likelihood, L(β) -376.03 -370.96 -369.12 -370.55 -370.43
Model χ 2 131.95*** 142.09*** 145.77*** 142.91*** 143.15***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65

†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

124
Table 12. Robustness: Cox Regression with Fixed-Asset Inefficiency

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
2.235 1.427 1.402 1.484 1.162
Free cash flow
(2.295) (2.001) (2.129) (2.200) (2.266)
Unrelated .076 .084 .085 .075 .097
diversification (.144) (.126) (.146) (.146) (.146)
-1.462*** -1.393*** -1.392*** -1.449*** -1.456***
Tobin’s Q
(.368) (.362) (.365) (.376) (.377)
-.029 -.041 -.039 -.052 -.036
Leverage
(.129) (.122) (.124) (.116) (.121)
.337 .232 .228 .243 .208
Taxes
(.227) (.201) (.211) (.221) (.231)
Year fixed effects 30.21*** 29.75*** 29.91*** 29.49*** 28.67***
Industry fixed effects 64.81*** 65.27*** 64.90*** 62.22*** 54.14***
.090 .089 .096 .053
Board independence ---
(.153) (.155) (.156) (.163)
.022 .021 .049 .024
CEO duality ---
(.126) (.126) (.131) (.128)
Insider equity .141 .120 .135 .144
---
ownership (.124) (.132) (.124) (.126)
.920*** .889** .486 .087
Fixed-asset inefficiency ---
(.234) (.247) (.439) (.543)
Fixed-asset inefficiency -.155
--- --- --- ---
* board independence (.343)
Fixed-asset inefficiency .376
--- --- --- ---
* CEO duality (.264)
Fixed-asset inefficiency
-.535*
* insider equity --- --- --- ---
(.260)
ownership
Log Likelihood, L(β) -376.03 -371.32 -371.22 -369.90 -368.18
Model χ 2 131.95*** 141.37*** 141.57*** 144.20*** 147.64***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65

†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

125
Table 13. Robustness: Cox Regression with Total-Asset Inefficiency

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
2.235 2.345 2.485 2.226 2.066
Free cash flow
(2.295) (2.344) (2.358) (2.265) (2.355)
Unrelated .076 .064 .063 .081 .062
diversification (.144) (.146) (.144) (.147) (.148)
-1.462*** -1.461*** -1.468*** -1.489*** -1.583***
Tobin’s Q
(.368) (.381) (.381) (.382) (.385)
-.029 -.034 -.032 -.041 -.024
Leverage
(.129) (.127) (.129) (.125) (.127)
.337 .331 .326 .363 .420
Taxes
(.227) (.249) (.248) (.245) (.257)
Year fixed effects 30.21*** 28.06*** 28.55*** 28.35*** 28.17***
Industry fixed effects 64.81*** 63.96*** 64.51*** 64.61*** 62.68***
.095 .097 .082 .089
Board independence ---
(.154) (.155) (.154) (.156)
.026 .027 .047 .038
CEO duality ---
(.127) (.127) (.128) (.127)
Insider equity .126 .096 .117 .150
---
ownership (.124) (.128) (.123) (.126)
.121 .083 .082 -.162
Total-asset inefficiency ---
(.446) (.406) (.481) (.519)
Total-asset inefficiency * -.327
--- --- --- ---
board independence (.345)
Total-asset inefficiency * .424
--- --- --- ---
CEO duality (.338)
Total-asset inefficiency * -.534*
--- --- --- ---
insider equity ownership (.210)
Log Likelihood, L(β) -376.03 -375.46 -374.96 -374.70 -372.78
Model χ 2 131.95*** 133.09*** 134.07*** 134.61*** 138.44***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65

†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

126
 Table 14. Robustness: Descriptives with FCF Adjusted for Capital
Expenditures

Independent
Mean S.D. (1) (2) (3) (4) (5) (6) (7) (8) (9)
Variables

1. Buyout .008 .086

2. Adjusted free
-.030 2.101 .00
cash flow
3. Unrelated
.145 .242 -.01 .01
diversification

4. Tobin’s Q 2.107 1.701 -.01 -.01 -.14***

5. Leverage .775 3.151 .00 .00 .05*** -.01

6. Taxes .032 .084 .01 -.01 -.02† .17*** -.02

7. Board
independence 67.239 16.754 .00 -.01 .07*** -.04*** .00 .00

8. CEO duality .320 .466 .02 -.01 -.11*** .02* -.04** .00 -.11***

9. Insider
equity 10.733 14.342 .01 -.02† -.06*** .01 .02† -.02† -.41*** .04***
ownership
10. Asset
2.562 6.782 .00 -.89*** -.03** .04** .00 -.09*** .01 -.01 .00
inefficiency

†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

127
Table 15. Robustness: Cox Regressions with FCF Adjusted for Capital
Expenditures

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
4.231 4.476† 3.963 4.163 4.771†
Adjusted free cash flow
(2.581) (2.659) (2.624) (2.648) (2.686)
Unrelated .065 .056 .058 .071 .051
diversification (.144) (.146) (.147) (.147) (.144)
-1.567*** -1.598*** -1.637*** -1.633*** -1.607***
Tobin’s Q
(.373) (.354) (.386) (.382) (.385)
-.028 -.028 -.022 -.030 -.027
Leverage
(.131) (.129) (.129) (.129) (.132)
.523* .557* .613* .575* .553*
Taxes
(.252) (.277) (.277) (.275) (.277)
Year fixed effects 30.25*** 28.24*** 28.83*** 28.57*** 28.70***
Industry fixed effects 65.48*** 64.57*** 64.86*** 65.15*** 64.99***
.103 .096 .095 .106
Board independence ---
(.154) (.154) (.154) (.154)
.021 .025 .034 .024
CEO duality ---
(.127) (.127) (.128) (.127)
Insider equity .129 .138 .117 .086
---
ownership (.123) (.125) (.123) (.132)
-.067 -.222 -.124 -.078
Asset inefficiency ---
(.482) (.496) (.492) (.459)
Asset inefficiency * -.476†
--- --- --- ---
board independence (.247)
Asset inefficiency * .367
--- --- --- ---
CEO duality (.357)
Asset inefficiency * -.397
--- --- --- ---
insider equity ownership (.409)
Log Likelihood, L(β) -375.08 -374.54 -373.09 -374.04 -374.00
Model χ 2 133.84*** 134.93*** 137.83*** 135.94*** 136.01***
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65
†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

128
Table 16. Robustness: Logit Regressions

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
.361 .403 .110 .388 .430
Free cash flow
(.936) (1.157) (.743) (1.039) (1.186)
Unrelated -.131 -.113 -.126 -.110 -.111
diversification (.131) (.134) (.137) (.134) (.137)
-.832** -.859** -.891*** -.871** -.857***
Tobin’s Q
(.242) (.248) (.255) (.253) (.248)
-.103 -.091 -.091 -.088 -.091
Leverage
(.071) (.068) (.074) (.068) (.068)
.121 .147 .231 .174 .144
Taxes
(.118) (.137) (.161) (.169) (.141)
.156 .152 .154 .155
Board independence ---
(.154) (.151) (.154) (.154)
.172 .174 .180 .173
CEO duality ---
(.119) (.119) (.122) (.118)
Insider equity .200* .196† .197† .196†
---
ownership (.100) (.101) (.100) (.096)
-.199 -.355 -.181 -.173
Asset inefficiency ---
(.376) (.507) (.485) (.118)
Asset inefficiency * -.463†
--- --- --- ---
board independence (.278)
Asset inefficiency * .208
--- --- --- ---
CEO duality (.378)
Asset inefficiency * -.054
--- --- --- ---
insider equity ownership (.238)
Wald χ 2 18.84*** 23.54*** 23.89*** 24.20*** 24.31***
Pseudo R2 .018 .025 .029 .026 .025
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65

†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

129
Table 17. Robustness: Rare-Events Logit Regressions

Model I Model II Model III Model IV Model V

Independent Variables
(Controls) (H1-H3) (H4) (H5) (H6)
-.449 -.285 -2.027** -.773 -.304
Free cash flow
(.936) (1.157) (.743) (1.039) (1.186)
Unrelated -.119 -.099 -.113 -.093 -.099
diversification (.131) (.134) (.137) (.134) (.137)
-.770** -.794** -.756** -.782** -.794**
Tobin’s Q
(.242) (.248) (.255) (.253) (.248)
-.162* -.142* -.146* -.136* -.141*
Leverage
(.071) (.068) (.074) (.068) (.068)
.091 .101 .095 .091 .093
Taxes
(.118) (.137) (.161) (.169) (.141)
.148 .139 .142 .149
Board independence ---
(.154) (.151) (.154) (.154)
.173 .177 .184 .173
CEO duality ---
(.119) (.119) (.122) (.118)
Insider equity .216* .207* .212* .237*
---
ownership (.100) (.101) (.100) (.096)
.054 .251 .060 .119
Asset inefficiency ---
(.376) (.507) (.485) (.389)
Asset inefficiency * -.534†
--- --- --- ---
board independence (.278)
Asset inefficiency * .293
--- --- --- ---
CEO duality (.378)
Asset inefficiency * .060
--- --- --- ---
insider equity ownership (.238)
Pseudo R2 .018 .025 .029 .026 .025
Number of observations 8631 8631 8631 8631 8631
Number of firms 1459 1459 1459 1459 1459
Number of events 65 65 65 65 65
†
p < .10, * p < .05, ** p < .01, *** p < .001
Wald chi-square statistics for the null hypothesis of equal year and industry effects.

130