Manager Divestment in Leveraged Buyouts
James Ang≠
Florida State University
Irena Hutton*
Florida State University
Mary Anne Majadillas**
University of New Mexico
Keywords: Leveraged buyouts, private equity, agency costs, earnings management, buyout pricing,
buyout premium, buyout performance
≠
Department of Finance, College of Business, Florida State University, Tallahassee, FL 32306. Phone: (850) 644-8208,
email: jang@cob.fsu.edu
*
Department of Finance, College of Business, Florida State University, Tallahassee, FL 32306. Phone: (850) 645-1520,
email: ihutton@cob.fsu.edu
**
Anderson School of Management, University of New Mexico, Albuquerque, NM 87131. Phone: (505) 277-6471, email:
maj@unm.edu
1
ABSTRACT
We examine changes in managers’ investment in the firm around leveraged buyouts and find evidence
of agency costs opposite to of those described in the extant literature. In the majority of leveraged
buyouts during 1997-2008, managers divested a portion of their pre-LBO share holdings while
maintaining an ownership stake in the post-LBO firm. We find that such divestment opportunities
encourage managers to behave in a way that benefits existing shareholders but is costly to new investors.
Specifically, we provide evidence of a positive relation between management’s divestment and pre-LBO
upward accrual-based and real earnings management, market timing, and more favorable buyout pricing.
We also examine whether Although we find evidence of subpar post-buyout performance among these
firms, and low effort and find that however, the involvement of private equity funds mitigates it. those
tendencies.
2
Leveraged buyouts (LBOs) have received a lot of attention in academic literature as a unique
organizational form that is effective in reducing agency costs of managerial discretion. In fact, at the
peak of buyout activity in the 1980s, the argument for this new organizational form was so convincing
that Jensen (1989) famously wrote about the “eclipse of the public corporation.” The reduction in
agency costs stems from three changes to corporate governance. First, it improves managers’ incentive
to perform by their greater commitment of personal wealth in the post buyout firm. managers are
encouraged to invest in the post-buyout firm to improve their incentives. Second, it reduces free cash
flow by committing the firm to demanding debt schedule under the new leverage-heavy capital
structure. facilitates financial discipline by diverting free cash flow to debt payments. Third, it creates
the incentive and mechanism for close and active monitoring due to equity concentration in the hands of
private investors represented by private equity funds’ not only creates incentives for close monitoring
but also facilitates more active monitoring through their representation dominant role on the board of
directors. This effort is supported by yet another group of active stakeholders – the creditors. In other
words, these changes in firm governance simultaneously strengthen the alignment of managers’ and new
investors’ objectives.
However, the success of this new governance structure implicitely takes as given that , in part,
depends on the common assumption that managers commit substantial personal wealth to acquire
ownership in the post-buyout firm. This gives assurance to the outside investors and creditors that
managers’ objectives are aligned with theirs and ensures capable management during the first two to
three years following an LBO, which are considered to be the high risk period of the deal. However, in
firms where managers already hold a significant equity stake, the buyout may serve as a divestment
opportunity and allow insiders to sell their pre-LBO equity at a sizeable premium and then reinvest a
3
fraction of that amount in the post-LBO firm1. The recent wave of buyout activity has afforded firm
managers many lucrative opportunities. For example, one of the largest divestments took place during
the 2007 buyout of Aramark, when Joseph Neubauer, the firm’s CEO, received nearly$940 million for
his 23 percent stake in the firm, and, after reinvesting $250 million, netted out $690 million. In other
words, his personal wealth committed to the firm decreased from $940 million before buyout to $250
after buyout, or a reduction of 73%. This and other similar cases violates one of the underlying
rationales justifying buyout.
The rise in buyout-aided divestment motivated buyout over the last three decades warrants a reexamination of the relation between changes in managers’ personal wealth, effectiveness of leveraged
buyouts in resolving traditional agency problems and emergence of new agency problems. According to
Kaplan and Stein (1993), during the early phase of the 1980s buyout wave, managers reinvested more
than half of their cashed-out equity back into the firm, which worked well to align the interests of
managers and post-buyout shareholders. As the buyout wave of the 1980s progressed, the amount of
reinvested equity decreased and so did the incentive for sound deals.
We argue that manager divestment in leveraged buyouts creates a different agency problem.
Since The managers’ objective is to maximize their personal wealth from the sum of buyouts allow
managers to benefit from selling a portion of their pre-buyout equity stake, and the potential
appreciation in their remaining post-buyout equity stake or both, Their actions are expected to may be
influenced by their relative wealth to be derived from these two sources. of a larger expected wealth
gain. In the ‘traditional’ buyout structure, managers commit more of their personal wealth, sometimes
through personal leverage, in the post buyout firm.
When managers’ wealth gain depends mainly on
post-buyout performance, they invest additional of personal equity which gives the buyout a more This
is the “traditional” structure that has been extensively documented in the literature. . In this type of
1
This practice of divesting while raising an ownership stake has been criticized in the financial press [Davidoff (2011)].
4
buyout the agency problem leads to an is the incentive to take actions in the pre buyout firm to depress
firm value and thus, to cause less buyout premium to be paid. Their bias is to favor of future not current
shareholders. , trading off current gains for future gains.2 However, in a buyout where managers divest
much of their shareholdings and current gains outweigh are more attractive then future gains, there is an
incentive to increase pre-buyout firm value, i.e., their bias is instead to favor current shareholders. .
Moreover, firm value may be affected not only by manipulating pre-buyout financials or timing the
market as has been previously shown, but also by the method of sale. Thus, while the agency costs of
investment buyouts are likely to be detrimental to the firm’s existing shareholders, whereas the agency
costs of high divestment buyouts are likely to harm new investors. In these cases, self interested
managers in pre buyout firms could use various means to affect buyout prices, such as, manipulating
earnings and other financials, real asset accruals, market timing and method of sale.
Aside from the agency problem at around buyout, managers, who divested the bulk of their
personal wealth but still play a large role in the management of the post buyout firms, may actually
deliver less effort as they may choose to divert their effort to manage and consume their much larger
outside wealth. This is the second agency problem. . Bitler, Moscowitz and Vissing-Jorgensen (2005)
find that while there is a positive relation between an entrepreneur’s ownership and effort, personal
wealth has a negative effect on effort levels. Elitzur, Halpern, Kieschnick and Rotenberg (1998)
theoretically model the effect of reduction in managers’ wealth in the post-LBO firm on the structure of
a buyout and manager's efforts in the post-buyout firm. Their model suggests a negative relation
between managers’ divestment and post-LBO performance.
Then, there is the third agency problem. Managers, who diverted bulk of their wealth which are
no longer correlated with the value of the buyout firm, may actually have greater incentive to take
2
See, for example, Fischer and Louis (2008) and Perry and Williams (1994).
5
excessive risks, financial or/and operational. This is because, having achieved financial security, they
could now afford to regard their remaining personal stake as and Furthermore, managers may use their
remaining equity in the firm as a cheap call option, in which they can improve its upside potential by
making risky investments [Jensen and Meckling (1976)]. Moreover, the new post-buyout compensation
structure that is also equity-heavy may actually augment the value of the option and to may encourage
investment in riskier projects [Smith and Stulz (1985)].
We find that, contrary to the assumption in the ‘traditional’ theoretical model of buyout, in 81%
of LBO deals, instead of committing more personal wealth, the management team cashes out a portion
of their wealth at the time of the LBO. More surprisingly, , in 46% of LBOs, managers divested more
than 50% of their pre-LBO holdings. In dollar terms, the total value of such divestments is $6.1 billion
dollars averaging $34 million per firm3. Our analyses demonstrate significant agency costs inherent in
of such divestments. First, we observe a positive relation between pre-buyout accruals, real earnings
manipulation and extent of managerial divestment. We also find evidence of market timing by managers
in that high divestment buyouts are preceded by stock run-ups. Second, we find that the buyout process
and pricing are also affected by the divesting managers’ incentives. Divesting managers are more likely
to sell the firm through an auction process or conduct a market check, if the initial offer is unsolicited, to
maximize obtain more favorable buyout pricing. Third, following the buyout, firms with managerial
divestment perform slightly worse than firms with no divestment.
However, the difference in
performance is attenuated with the presence participation of private equity investors consistent with their
monitoring and disciplining roles.
The ability of managers to significantly increase the amount and liquidity of their wealth through
an LBO and the effect of such divestment on the financial performance of the firm are relatively
unexplored. While the following three studies empirically document divestment and even its effect on
3
These figures do not include severance payments for departing executives.
6
some buyout characteristics, they do not comprehensively examine performance around the buyout, its
pricing and method of sale in the context of manager divestment. Crawford (1987) analyzes 30 deals
completed over 1981-1985 and finds that managers both realize large cash-outs and continue to maintain
control after retaining an inexpensive equity stake in an over-levered buyout firm. Kaplan and Stein
(1993) find that managerial divestment increased during the LBO wave of the 1980s and that it
positively affected the likelihood of a firm’s subsequent financial distress. Frankfurter and Gunay (1992)
suggest that most buyouts are motivated by insiders’ need for personal liquidity and find that the level of
insiders’ divestment leads to wealth gains for pre-buyout shareholders in the form of buyout
announcement returns.
This Our study makes a contributes to the literature by reporting several new results. First, we
report the level of divestment by managers during the recent wave of LBOs. Second, we document
perverse incentives of divestment and relates it to pre-buyout earnings management and updates
previous evidence on earnings management in LBOs since the first buyout wave of the 1980s. Since
then, two securities litigation acts were passed by Congress in 1995 and 1998. Whether they have
reduced the likelihood of litigation and increased the incentives to manage earnings is an empirical
question. Third, we document that management divestment influences the method and timing of firm
sale, the bidding process and buyout pricing. Lastly, we provides new evidence relating divestment in
buyouts to value creation by examining post-buyout performance.
The remainder of this study is structured as follows. Section 1 reviews the relevant literature and
develops the hypotheses, Section 2 discusses the sample, Section 3 describes the data and methodology,
Section 4 discusses the results, and Section 5 concludes the paper.
1. Previous Research and Hypothesis Development
7
Changes in managers’ wealth due to a leveraged buyout create a range of agency problems, as
the interests of the management become aligned with either the interests of the buyout team in the case
of investment or with the interests of the selling shareholders in the case of divestment. In this section
we develop hypotheses about the effects of management’s divestment on pre-buyout performance,
method of sale, attributes of the bidding process, buyout pricing and post-buyout performance. We lay
the foundation for our empirical analyses in Table 1, by formally comparing agency problems that arise
prior to, during, and following the buyout process.
[Insert Table 1 here]
Pre-buyout earnings management
In the context of leveraged buyouts, Lowenstein (1985) and Schadler and Karns (1990) argue
that managers may employ accounting and non-accounting techniques to affect pre-announcement share
price. One example of such manipulation, earnings management, has been explored in several empirical
studies.
The early literature on accrual-based earnings management prior to leveraged buyouts draws
evidence from the sample of management buyouts in which managers often roll their investment over or
commit additional personal funds. DeAngelo (1986) argues that, in this setting, managers have an
incentive to report poor earnings to reduce the buyout price, but finds no downward earnings
management. Perry and Williams (1994) examine a similar hypothesis and find negative accruals in the
year preceding the buyout. In a more recent study, Fischer and Louis (2008) suggest that earnings
management prior to management buyouts is affected by two conflicting objectives, namely,
maximizing the value of a personal gain from the buyout while securing buyout financing and reducing
its cost. Their study finds significantly negative accruals in the fiscal year preceding the buyout
8
announcements. However, managers who depend on the external funds the most report less negative
accruals prior to the buyout. Cornelli and Li (2006) theoretically argue that managers may take actions
to depress the offer price prior to increasing their ownership stakes. Moreover, they go as far as stating
that “nobody pays attention to the ex-ante perverse effects” associated with a change in ownership
structure over the LBO episode.4
We argue that the gradual shift from investment-heavy to divestment-heavy buyouts over time,
changed the nature of earnings manipulation. Actions depressing short term firm value are likely to be
observed when managers buy into the firm; however, steps to increase short term value are more likely
in the case of buyouts with significant managerial divestment. The net result of such reporting incentives
for divestment firms will be higher earnings manipulation measures. 5 Such earnings manipulation
activities may also correlate with high stock returns. First, managers may time the divestment buyout to
a period of high stock returns. Second, earnings management may either be necessary to sustain or
justify such high market values. Alternatively, more aggressive earnings management may even
facilitate an increase in market value if investors translate higher reported earnings cash flows (My
note: cash flow is not appropriate here, as the idea behind the calculation of accruals is to assume that
cash flow is correct but earnings are manipulated) into higher stock prices. Thus, we formulate our first
hypothesis.
Hypothesis 1: The size of intended divestment by managers affects their incentives for upward
pre-LBO earnings management.
4
Firms may resort to other options to depress pre-buyout firm value. Restoration Hardware, for example, made a very public
show of firing more than 100 employees out of its corporate offices. According to insiders, the move was short lived and the
company promptly hired replacements [Meagher (2008)].
5
The issue of whether pre-LBO run-up can be advantageous to insiders is an open one. Schwert (1996) finds that bidders are
likely to interpret run-ups as an increase in the target’s value. However, Betton, Eckbo and Thornburn (2008) suggest that
while pre-offer run-up is costly to the buyer, it is likely to substitute for the intended merger premium.
9
Method of sale and buyout pricing
The majority of current empirical evidence relates buyout pricing to various sources of wealth
gains.6 While the evidence on the method of sale, bidding competition, bid revisions and the role of prebuyout run-up in buyout pricing is scant, anecdotal evidence suggests significant variation in these
variables across different divestment levels.
Most buyouts are carried out via two methods: a negotiated sale or an auction. In a negotiated
sale the LBO firm contacts the buyer directly (or is contacted by an unsolicited bidder) and negotiates
the sale. An auction process typically begins with the firm or its investment bank contacting multiple
potential strategic and financial acquirers. Contacted parties then indicate their interest and submit
several rounds of bids until the winner emerges. Often, firms receive unsolicited bids from their own
management teams, strategic bidders or private equity firms. The independent committee evaluating
fairness of the merger may then elect to conduct a “market check” or solicit indications of interest from
additional bidders, which resembles an auction sale. Fidrmuc, Roosenboom, Paap and Teunissen (2012)
show that firms choose a method of sale (i.e., auctions, controlled sales and private negotiations) to fit
their firm and deal characteristics. Moreover, the method of sale is crucial to the success of the selling
process. Anilowski, Macias and Sanchez (2009) demonstrate that firms sold via auctions rather than
negotiated sales or hostile takeovers have higher wealth gains. This suggests that managers may choose
the sales method that is most likely to maximize their personal gain.
Managers in buyouts with high divestments have the incentive to use an auction sale to generate
bidder competition. On the other hand, in investment-heavy buyouts, managers may purposefully select
potential acquirers with little low expected synergy? interest in the firm thereby limiting the amount of
6
See, for example, Kaplan (1989), Lehn and Poulsen (1989), Kieschnick (1998).
10
competition. A recent study by Subramanian (2008) demonstrates that in management-led buyouts that
are likely to have investments and rollovers of personal equity bidder competition is quite low, even if
the buyout agreement allows for a market-check. The use of such cosmetic provisions is likely to reduce
litigation risk, but despite following the letter of the law, violates the spirit of the Revlon rule.7
Moreover, Lowenstein (1985) and Easterwood, Singer, Seth and Lang (1994) report that the premium in
buyouts with competing bidders is significantly larger than the premium in deals with no competing
bids.
The effect of an auction sale and bidder competition on bid revisions is likely to be an empirical
matter. Betton and Eckbo (2000) find significant bid revisions (13%) following the initial bid in tender
offers. However, Betton, Eckbo, and Thorburn (2008) report that the initial offer premium is higher in
takeovers without? competing bidders. This is consistent with bidders increasing their first bid to deter
competition [Fishman (1988)] and may result in fewer and lesser revisions from the initial to final bid.
These two parts read like contradictory.
Consistent with these studies we expect that divesting managers will attempt an auction sale by
soliciting bids from a group of potential bidders or, in the event of an unsolicited bid, perform a “market
check.” An additional observable outcome of this sales choice will be a more competitive bidding
process. A combination of pre-buyout earnings management and a more competitive bidding process
will lead to more favorable buyout pricing. Thus, we formulate our next hypothesis.
Hypothesis 2. Likelihood of an auction sale, bidder competition and buyout premium are
positively related to managers’ divestment.
Shareholders of Restoration Hardware won the class action lawsuit against Catterton Partners and the firm’s management.
The sales process favored the management-led group and the firm went to the low private equity-backed bid rather than the
higher bid from Sears Holdings. Shareholders were awarded an extra $0.19 per share in addition to the $4.50 merger
consideration.
7
11
Post-buyout performance
A substantial body of empirical work supports value creation in leveraged buyout deals. Due to
the limited availability of post-LBO data, most studies either focus on different value-related aspects of
leveraged buyouts or change in firm value from pre-buyout to a later corporate event, such as an IPO or
second LBO. Fewer papers have examined the effect of divestment-related managerial incentives on
post-buyout performance.
Kaplan and Stein (1993) study the post-buyout performance incentives of managers that owned
a large portion of pre-LBO equity and “cashed out” through the LBO. They find that the degree of
divestment by the firm’s management is positively associated with the probability of default. MY
NOTE: we can also cite this as being consistent with/ supportive of the option / risk increasing incentive.
Nikoskelainen and Wright (2007) examine gains from exited UK buyouts and conclude that the
governance mechanisms of buyouts do not solve agency problems associated with the post buyout? free
cash flow. Instead, their study supports the heterogeneous view This term is not clear or defined earlier.
of buyouts, particularly whether it is driven by insider or outsider management. Elitzur, Halpern,
Kieschnick, and Rotenberg (1998) develop a theoretical model in which management's net dollar
investment in the post-buyout firm is an important factor in both the buyout structure and post-buyout
performance.
We expect that in firms where management has significantly increased personal wealth through
divestment, post-LBO effort is diminished. Additionally, managers can engage in risk-taking to
maximize the value of their remaining ownership in the firm [Jensen and Meckling (1976) and Smith
and Stulz (1985)] or become entrepreneurial, innovative and more tolerant of failures [Wright,
Hoskisson and Busenitz (2001) and Ferreira, Manso and Silva (2011)]. Thus, we formulate our last
hypothesis.
Hypothesis 3. Post LBO performance is negatively related to managers’ divestment.
12
Is there a corollary to this hypothesis that says large investors, like private equity funds, mitigate
managerial shirking and excessive risk taking as discussed above?
2. Data and Variable Construction
We obtain our initial sample of LBO deals from SDC’s Mergers and Acquisitions database. We
start with LBOs that became effective over 1997-20088 and further restrict the sample to public targets
with available deal size. Additionally, we require that the firm is taken private by a financial firm or the
firm’s management team which produces a sample of 338 transactions. We then screen all target firms
for the availability of Compustat and CRSP data and access each firm’s pre-LBO, LBO and post-LBO
filings to hand-collect our key analysis variables: reasons for the buyout, management ownership,
change in managers’ investment in the firm, sales method and timeline, number and magnitude of
buyout bids, and the presence of competing bidders. We are then left with a sample of 179 observations
with reasonably complete information. The sample size compares favorably to other studies analyzing
LBOs, which typically rely on 100-200 observations.
Variable construction and descriptive statistics
Our key analysis variable is Net Dollar Divestment. Alternatively, we refer to this variable as
manager cash-out. It is defined as the dollar amount received by management for their shares valued at
the LBO offer price less the amount reinvested in the firm. In cases where executive stock options are
terminated at the time of the LBO, we incorporate the payoff by computing the difference between the
purchase price and the strike price.9 The value of options represents a small fraction of managers’
wealth, which is consistent with other studies that examine changes in insider wealth around tender
offers and mergers [Cotter and Zenner (1994), Hartzell, Ofek and Yermack (2004)]. We then construct
8
9
This time period has reliable coverage of LBO filings in the SEC Edgar database.
This applies only to in-the-money options. Out-of-money options are cancelled and no payment is made.
13
a measure of Relative Divestment by dividing Net Dollar Divestment by the dollar value of managers’
pre-LBO holdings. We define pre-LBO holdings as managerial ownership by all named executive
officers obtained from the most current LBO filing, 10-K or proxy statement. In nine deals, managers
invest personal equity in the firm and in one of these deals the amount of invested equity exceeds their
original investment. Does this mean this manager doubled his/her $ investment? Not clear.
Table 2 shows the annual distribution of buyout activity and divestment by managers for our data
set. The number of LBOs exhibits an increasing trend up to 2000; it then declines and rebounds by
2003-2006. This is consistent with the hot merger and LBO market of the late 1990s and early-to-mid
2000s. NOTE TOO: that buyout is cyclical is supportive of divesting motive than investing, which should
be counter cyclical, i..e, more (less) buyouts in period of low (high) share prices.
The total amount of personal wealth taken out by executives during our 11-year sample period is
$6.1 billion, which is large in economic terms. Moreover, the size of divestment relative to pre-LBO
ownership, which includes firms with rollovers and net investments, averages at 42% of managers’
personal wealth in the pre-buyout firms. In 81% of deals managers realized some divestment, while in
nearly 46% of the deals they cashed out more than 50% of their pre-LBO ownership. There is no
discernable detectable time trend in either of the divestment variables.
Although our sample includes only 179 firms, we believe that these buyouts are representative of
the serve as a common exit strategy for managers/owners. In the process of selecting a useable sample,
we discard nearly five hundred LBOs of standalone private firms carried out by private equity funds
which do not meet the criteria for our analyses due to data availability. Since managers of private firms
hold a large fraction of their firms’ equity, buyouts of these firms are even more likely to result in a
sizeable divestment.
[Insert Table 2 here]
14
In Table 3 we provide summary statistics pertaining to the buyout reasons stated in the pre-LBO
filings as well as deal- and firm-specific characteristics of firms in our sample. In Panel A, we report
buyout reasons for three groups of deals. The first group contains 34 investment and rollover deals
(Investment/Rollover) in which managers contribute additional personal equity or reinvest all of their
cashed-out equity. The other two groups contain divestments split into Low Divestment (63 firms with
below 50% Relative Divestment) and High Divestment (82 firms with above 50% Relative Divestment).
It is apparent that buyouts in the Investment/Rollover group are motivated by low liquidity,
undervaluation, poor operating performance, miscellaneous costs of maintaining public status, and the
insiders’ need for control. In contrast, buyouts in the High Divestment group are more influenced by the
insiders’ desire to diversify their holdings, favorable buyout price and market conditions. In all but two
cases (low growth potential and costs of Sarbanes Oxley Act), the differences in reasons stated by
Investment/Rollover versus High Divestment firms are statistically significant in two-sample tests of
proportions.
In Panel B, we summarize buyout characteristics. In the Investment/Rollover group, on average,
executives put an additional $0.432 million of funds into the firm. Executives in the Low Divestment
group divest $13.513 million, while executives in the High Divestment group divest over $64.194
million. Relative Divestment increases from -13.7% to 18.6% to 83.3% across these three groups. The
deal size of the average Investment/Rollover firm is also smaller than that of the average High
Divestment firm ($177 million vs. $1,916 million in deal size) with buyout prices per share showing
similar differences ($9.1 vs. $21.8).
The management team in the Investment/Rollover group owns 32.2% of the firm prior to the
buyout, which is significantly higher than management ownership in High Divestment deals (17.2%).
Furthermore, consistent with the stated preference for control, managers in Investment/Rollover and
15
Low Divestment deals increase their percentage ownership after the buyout. The post-LBO ownership
doubles from its pre-LBO levels for Investment/Rollover deals (65.6%) and almost doubles for Low
Divestment deals (56.2%). For High Divestment buyouts, due to infusion of large amount of new debt,
proportional management ownership drops slightly to 15.4% allowing managers to maintain their
percentage ownership while significantly reducing their dollar investment. One caveat is in order: postLBO managerial ownership is not available for about 1/3 of our sample.10
Outside private equity involvement varies in its intensity and ranges from no private equity or
management-led deals (Management Buyout) to multiple private equity firms (Club Buyout). It is
important to note that even in the deals that are not led or co-led by the firm’s management, managers
often continue to be employed by the firm. As expected, Investment/Rollover buyouts are mainly led by
these firms’ management teams (64.7%), while the buyouts of High Divestment firms are led by either a
single private equity firm or a consortium of firms (52.4% and 29.3%). All differences between
Investment/Rollover and High Divestment buyouts are statistically significant, with the exception of
deals led by private equity with management. These results are not surprising in that firms with high
manager ownership are taken private by managers seeking to maintain or increase their control of the
firm. First, managers with much control of the firm may not agree to the acquisition by a private equity
firm. Second, firms with high managerial ownership tend to be smaller and may be able to obtain
financing without the help of a private equity sponsor those advantage is the in large scale financing.
In the last Panel of Table 3 we report pre-buyout financial characteristics of sample firms. The
book value of total assets is significantly smaller for Investment/Rollover firms than for High
Divestment firms, consistent with differences in deal values reported earlier. All three groups of firms
have comparable leverage measured by total liabilities divided by total assets. The operating margin
10
The information on post-buyout ownership is often concealed from the public and can, in some cases, be obtained from
subsequent litigation filings.
16
(EBITDA/sales) is higher for the High Divestment group, but the difference from the
Investment/Rollover group is significant only in a medians test. Differences in market-to-book and stock
returns between the Investment/Rollover and High Divestment groups are large and statistically
significant possibly indicating market timing. The market-to-book ratio is calculated as stock price four
weeks prior to the announcement divided by book value per share. The stock return is measured by
monthly buy-and-hold abnormal returns (BHARs) adjusted by the value-weighted CRSP index over the
fiscal year preceding the announcement. In the Rollover/Investment deals, stock returns are very
negative (-18.6%), consistent with the motivation to obtain the lowest buyout price, compared to the
High Divestment deals (5.1%). In Figure 1, we plot monthly BHARs for all three groups. The plots
provide additional information about stock price behavior prior to LBOs in that the differences between
the groups increase as returns of firms with divesting managers trend up and returns of firms with
investing managers steeply trend down over the year leading up to the buyout announcement.
Lastly, we find that the volatility of monthly stock returns over one year leading up to the buyout
announcement is largest for Investment/Rollover firms. Additionally, these firms have lower liquidity
measured by the average daily share turnover over the pre-announcement year. Overall, these findings
are consistent with the stated reasons for buyouts in that Investment/Rollover firms are influenced by
poor operating performance, undervaluation, and lack of liquidity and High Divestment firms
opportunistically time the buyout to a period of high valuations.
[Insert Table 3 here]
[Insert Figure 1 here]
3. Main Results
17
In this section we test our hypotheses about the relation between managers’ divestment and firm
performance around the LBO. In Table 4, we examine the determinants of manager divestment. This
analysis is important because it helps us understand the drivers of the decision to divest and address
potential reverse causality issues in subsequent analyses. One potential criticism of testing the
relationship between managerial divestment and managerial actions is that divestment itself may be
determined by successful earnings management or buyout pricing.
To examine the determinants of divestment, we model the decision to divest as a function of
managers’ and firm characteristics.
Dit  0  1 Ageit 1   2 FamilyDummyit 1  3 FounderDummyit 1   4Top3Influenceit 1  5 ManOwnershipit 1 
 6 IndustyMar ket  to  Book it 1   7 IndustyMer gerActivit yit 1  8 Industy PrivateEquit yActivity it 1 
9 MarketRet urnit 1  10 LogTotalAs sets it 1   it
(1)
In (1), the dependent variable Dit is managers’ Relative Divestment and we use a Tobit model
with censoring at 1 to estimate the regression equation. This censoring scheme reflects the fact that
divestment cannot exceed 100%. The model is not censored at -1 because net investment can be greater
than 100% percent. The independent variables are measured at one-year lags relative to the dependent
variable. The effect of manager characteristics on divestment may help us show that it is heavily driven
by personal preferences and it may not be fully dependent on firm performance.
Since personal characteristics of managers are likely to affect the decision and amount of
divestment, we control for the average age of top 3 officers (Age), as proximity to retirement can trigger
divestment. We also include dummy variables for a family firm (Family Dummy) and whether the
founder serves as a top 3 executive officer (Founder Dummy). Founders have strong incentives to pass
the firm to their heirs, making divestment less likely [Anderson and Reeb (2003)]. Additionally, in most
family firms, family members serve as the firm’s CEO or top management to maintain control; there is
18
an additional incentive to preserve family shareholdings and protect family managers [Schulze,
Lubatkin, Dino, and Buchholtz (2001)]. Moreover, controlling families are generally not willing to lose
their control of the firm [Gomez-Mejia, Nunez-Nickel, and Gutierrez (2001)].
Since the CEO or one of the top executives may initiate a buyout, we control for the influence of
the firm’s top 3 executive officers (Top3 Influence) by computing the ratio of their stock ownership to
that of the management team. We also include management ownership (Management Ownership) as
entrenched managers may be less likely to divest. Lastly, favorable market conditions and especially
activity in the merger and buyout market may affect the likelihood of divestment. Our measures of
market conditions are the one year return on the CRSP index (Market Return), as well as the volume of
merger activity in the buyout firm’s industry relative to total merger volume (Industry Merger Activity)
and the volume of private equity transactions in the buyout firm’s industry relative to the total merger
volume (Industry Private Equity Activity). Lastly, we control for industry growth opportunities with the
industry market-to-book ratio, and firm size with the logged value of total assets. In the second model,
we introduce debt (Leverage), operating performance (EBITDA/Sales) and growth opportunities
(Market-to-Book) to assess the incremental effect of firm performance. Our results in Table 4 indicate
that managerial characteristics are important determinants of the decision to divest. We find that the age
of top 3 officers has a positive and significant coefficient, suggesting that the desire to hold more liquid
and diversified assets associated with retirement may motivate divestment. These results are consistent
with Frankfurter and Gunay (1992) who report that the need for liquidity is a driver of many buyouts.
The coefficients of founder dummy, influence of the top 3 officers and managerial ownership are
negative and significant, highlighting the importance of control in these types of firms. Managers in
industries with high market-to-book ratios are also less likely to divest in order to realize future growth
opportunities. Additionally, we find that market return is a positive and significant determinant of
19
managerial divestment consistent with market timing. Lastly, managers of large firms are likely to divest
more.
In the second model that incorporates firm financials we find that firm profitability is positively
related to divestment, thus indicating the importance of earnings management. However, the incremental
effect of including firm characteristics is small raising the pseudo R2 of the model from 0.301 to 0.319.
These results suggest that that the divestment decision is primarily influenced by managers’
characteristics, market conditions and to a lesser extent by pre-buyout financials.
[Insert Table 4 here]
To test our first hypothesis of whether pre-LBO earnings management is positively related to
manager divestment, we examine both accrual-based and real earnings management. While the prior
literature on leveraged buyouts has focused on accrual-based manipulation, real earnings manipulation
may be preferred to accrual manipulation as it is harder to detect. To capture accrual-based earnings
management, we use the modified cross-sectional Jones model as implemented by Teoh, Welch and
Wong (1998) to compute discretionary current accruals (DCA). This variable picks up abnormal
changes in current accruals due to accelerated recognition of revenues and delayed recognition of
expenses. Specifically, we run the following regression by industry and by year
CurrentAccruals j ,t
TotalAssets j ,t 1


 Sales j ,t

1
 0 
 1 


 TotalAssets j ,t 1 
 TotalAssets j ,t 1  j ,t




(2)
where current accruals are calculated as the difference between the change in non-cash current assets
and the change in current operating liabilities; and j and t indicate industry and year, respectively. We
then use the coefficients from the industry-year regressions and apply them to the LBO-firm to calculate
expected accruals that arise from the firm’s ordinary course of business and are not subject to
managerial discretion, i.e., non-discretionary current accruals (NDCA). Discretionary current accruals
represent the difference between a firm’s observed current accruals and NDCA.
20
To calculate real earnings management, we follow Roychowdhury (2006). This real
manipulation measure is based on the premise that firms try to minimize reporting losses in three ways.
First, they attempt to increase sales by speeding them up or generate additional sales by offering price
discounts and relaxed credit terms. These strategies will temporarily increase sales volume but may
decrease cash flows. Second, firms may reduce cost of goods sold by lowering per-unit fixed costs
through increased production. However, firms can still incur other production costs, which can also
lower cash flows for a given level of sales. Third, firms can aggressively reduce aggregate discretionary
expenses. This strategy can boost current earnings and potentially result in higher current period cash
flows if the firm previously paid for such expenses in cash.
These measures of real earnings management are calculated similar to discretionary current
accruals as the difference between observed and predicted values computed using coefficients from
industry-year regressions. Specifically, Equation 3 was used in industry-year regressions of operating
cash flow.
OperatingCashFlows j ,t
TotalAssets j ,t 1




 Sales j ,t

Sales j ,t
1
 0 
 1 
 2 



 TotalAssets j ,t 1 
 TotalAssets j ,t 1 
 TotalAssets j ,t 1  j ,t






(3)
Equation 4 was used to fit discretionary expenses defined as the sum of advertising, R&D, and SG&A in
industry-year regressions.
DiscretionaryExpenses j ,t
TotalAssets j ,t 1


 Sales j ,t 1 
1
 0 
 1 
  j ,t

 TotalAssets j ,t 1 
 TotalAssets j ,t 1 




(4)
Lastly, Equation 5 was used to estimate production costs defined as cost of goods sold and change in
inventories in industry-year regressions.
21
ProductionCosts j ,t
TotalAssets j ,t 1




 Sales j ,t

Sales j ,t
1
 0 
 1 
 2 


 TotalAssets j ,t 1 
 TotalAssets j ,t 1 
 TotalAssets j ,t 1 






 Sales j ,t 1 
 3 

 TotalAssets j ,t 1  j ,t


(5)
After constructing these measures of real earnings management, we combine them into one
comprehensive measure. We construct this measure (Real Earnings Management Proxy), following
Cohen, Dey and Lys (2007) by adding abnormal discretionary expenses and abnormal production costs
to abnormal cash flows. We modify their formula by multiplying both production costs and
discretionary expenses by -1 so that higher values of this composite variable indicate greater real
manipulation.
In Panel A of Table 5, we report our measures of accrual-based and real earnings management.
Discretionary current accruals is negative for Investment/Rollover deals (-0.014) and large and positive
for High Divestment deals (0.045), consistent with the incentives to depress share prices in the former
and boost prices in the latter case. This difference is statistically significant at the 1% level. The real
earnings management proxy provides weaker evidence for greater earnings manipulation in the sample
of High Divestment firms relative to the sample of Investment/Rollover firms, as the difference in mean
values (0.306 vs. 0.133) is statistically significant only at 10%. The difference in medians is not
significant. Since earnings manipulation can begin several years prior to the buyout, we examine both
measures at year t=-2. We find that earnings management is absent at t=-2 and picks up at t=-1 in
preparation for the buyout. Taken together, this evidence provides some support for our hypothesis that
insiders manage earnings upward more aggressively if they plan to cash out. Lastly, as an additional
support for deliberate earnings management, we find that the number of days lapsed between the first
discussion of strategic alternatives by the firm and buyout announcement is a little under a year, which
gives firms ample opportunity to manage earnings. However, we do not find any statistically significant
22
difference in this variable between Investment/Rollover firms and High Divestment firms which
indicates that both types of firms have the same amount of time to prepare for the buyout.
In Panel B of Table 5, we test the predictions of our first hypothesis for robustness in a
multivariate framework by modeling pre-LBO accrual-based and real earnings manipulation as a
function of managerial divestment and control variables that are likely to affect the dependent variable.
Eit 1   0  1 Re lativeDivestmentit 2   2 LogTotalAssetsit 2   3 Leverageit 2   4 EBITDA / salesit 2 
 5 Market  to  Book it 2   6Volatility it 2   7 ManagementOwnership it 2   it
(6)
Specification (6) is used to fit two pairs of OLS models. In the first pair, the dependent variable
Eit-1 is discretionary current accruals (DCA); in the other pair, the dependent variable is a measure of real
earnings management (Real Earnings Management Proxy). The independent variables are measured at
one-year lags relative to the dependent variable and represent the information available to managers
prior to earnings manipulation.
In addition to observed insider divestment we use predicted divestment in the second model of
each regression pair. Predicted divestment values are estimated using the first model from Table 4 and
applying t=-2 lags to the independent variables to alleviate reverse causality concerns as these predicted
values are not affected by pre-buyout financials. The remaining set of control variables includes the log
of total assets (LogTotal Assets), as larger firms tend to be more transparent and better monitored,
making earnings management and other value-enhancing actions more difficult. Following Dechow,
Sloan and Sweeney (1996), we use Leverage to proxy the closeness to debt violations as managers are
more likely to select income-increasing accounting policies, the closer a company is to violating its
accounting-based debt covenant [Sweeney (1994)]. Additionally, we test whether firms with poor
operating performance (EBITDA/Sales) are more likely to engage in earnings management. The effect of
the market-to-book ratio on earnings management is an empirical matter. On one hand, firms with
23
higher valuations may have less incentive to manage earnings. On the other hand, greater information
asymmetry of growth firms may make earnings management difficult to detect and temporarily high
market valuation may encourage earnings management to conceal or prolong it. Dechow, Sloan, and
Sweeney (1996) find that firms subject to SEC accounting enforcement actions tend to have high prior
market-to-book ratios; and Richardson, Tuna, and Wu (2002) associate prior market-to-book ratios with
earnings restatements. Firms with high levels of uncertainty, measured by stock price volatility, are also
more likely to commit accounting fraud due to lower monitoring and less likely detection [Erickson,
Hanlon, Maydew (2006)]. We use the standard deviation of monthly stock returns over one fiscal year as
our measure of volatility. We also include ownership of the management team (Management
Ownership) as managers with more power may have more opportunities and incentives for earnings
management. Although our measure of managerial equity holdings is collected from LBO filings rather
than t=-2 filings, we believe it to be a reasonably good proxy for managerial holdings prior to the
decision to manage accruals. First, according to LBO filings, most firms in our sample are illiquid,
which makes any large pre-LBO dispositions unlikely. Second, Harlow and Howe (1993) find no insider
trading prior to buyouts led by private equity firms, but do report evidence of share accumulation prior
to management-led buyouts. This accumulation is due to non-selling rather than acquisition of additional
shares. We also conduct a quick check of pre-LBO insider trading activity by examining insider trading
data from Thomson Financial. We find that only 79 firms in our sample have some insider trading
during the pre-LBO year with net insider sales being small and not significantly different from zero.
In the first pair of DCA regressions of Table 5, managerial divestment is a positive and
significant (at the 1% level) predictor of pre-LBO accruals even when we use its predicted values. We
also find that market-to-book ratio and volatility are positively and significantly related to accrual
management consistent with our expectations. Managers are more likely to manage earnings to support
24
high valuation multiples and are able to do so more successfully when their firm is difficult to value and
monitoring is insufficient. In the second pair of Real Earnings Management Proxy regressions the
coefficients of both divestment measures are likewise positive and significant at the 5% level.. Similar
to the first model, Real Earnings Management Proxy is positively related to market-to-book ratio and
volatility. In the regressions reported in the Internet Appendix11, we include a Manufacturing Firm
dummy that is equal to 1 if the firm is in the single SIC code 2 or 3 since overproduction is more
relevant for manufacturing firms. Our results remain qualitatively unchanged. Overall, these results
provide support for our first hypothesis.
[Insert Table 5 here]
Our second hypothesis tests the effect of managerial divestment on the method of sale, bidder
competition and buyout pricing. In Panel A of Table 6, we report measures of LBO pricing and
characteristics of the buyout process that are likely to vary with managerial incentives. One measure of
buyout premium (Premium 4 wk) is calculated as the percentage difference between the LBO offer price
per share and stock price four weeks before the LBO announcement. The other measure of premium
(Premium 1 yr) is computed relative to the stock price 12 months before the announcement to mitigate
the effect of the stock price run-up prior to the LBO. Additionally, buyout filings indicate that some
buyouts are priced relative to the average stock price several months prior to the buyout to deliberately
reduce the influence of unusual price behavior on buyout pricing. We also construct a commonly-used
transaction multiple, deal value to revenues (Deal/Sales), which is independent of pre-LBO stock price
movements, but may be negatively affected by earnings management. We use a multiple based on
revenues rather than EBITDA because negative values of the latter render the multiple meaningless.
Overall, we find that the higher buyout premiums coincide with divestment by management:
Deal/Sales averages increase with divestment from Investment/Rollover to High Divestment (1.034 and
11
The Internet Appendix can be found at xxxxxx.com
25
1.427) and so does the premium relative to stock price 12 months prior to the announcement (0.275 and
0.566). The four week premium declines in both means and medians across divestment groups, which
indicates that buyout specialists may take into account pre-offer run-up in setting the buyout price. The
differences in all three variables are statistically significant.
We also analyze the method of sale and characteristics of the bidding process. First, we examine
whether the firm attempted an auction sale by soliciting offers from multiple potential bidders. Such
form of sale is more likely to translate into competing bids and higher LBO premium than single-bidder
negotiated bids. We find that, on average, 11.8% of all LBOs in the Investment/Rollover sub-sample try
to sell the firm via an auction. This contrasts sharply to Low Divestment and High Divestment deals,
where 46% and 64.6% of firms attempt an auction. Second, we examine whether an additional
solicitation of bids, a “market check”, is attempted. The fraction of deals with an auction and/or market
check varies from 35.3% to 60.3% and 86.6% across the three groups, consistent with prediction.
Lastly, we find that the effort to generate competition is likely to pay off in that a larger fraction of High
Divestment and Low Divestment deals experience competing bids (61% and 39.7%) than
Investment/Rollover deals (29.4%). These differences are statistically significant at the 1% level.
Although our evidence fail to provide support for the conjecture that firms with high managerial
divestment negotiate more aggressively,- . F (firms in the Investment/Rollover group require 2.765
revisions, while firms in the Low Divestment and High Divestment groups on average require 2.968 and
3.111 revisions),
however, these differences are not statistically significant. Lastly, the percent
revision from the winning bidder’s initial to offer price is surprisingly lower for the High Divestment
firms (0.069) than for Investment/Rollover firms (0.129). This suggests that since firms in the High
Divestment group are more likely to be auctioned off and generate more bidder interest, potential
bidders may increase their initial bids to discourage competition, consequently leading to fewer
26
revisions, as evidence by high final premium (true?). While this difference is economically significant it
does not meet statistical levels of significance. To summarize, the results of these univariate
comparisons support our first two hypotheses in that divesting managers take steps to increase proceeds
from the buyout by managing earnings and timing the market as well as structuring the sales process to
receive the most favorable pricing.
In Panel B, we present parameter estimates from two pairs of Logit regressions and two pairs of
OLS regressions that use the same set of independent variables as earnings management regressions.
Similar to the previous table, we use realized as well as predicted measures of divestment. In the first
pair of regressions, the dependent indicator variable equals 1 if the firm attempted an auction sale
and/or market check and 0 otherwise. In the second pair, we use an indicator variable for bidder
competition as the dependent variable. Since firms may attract very few buyers, or buyers that have low
levels of interest or ability to purchase the firm, the auction process or market check may not
necessarily result in a competitive sales process. Focusing on bidder competition allows for an
additional test of managerial incentives.
We find that the likelihood of an auction and/or market check increases with both realized and
predicted managerial divestment and growth opportunities/market valuation (Market-to-Book); it
decreases with managerial ownership. The measures of realized and predicted divestment are significant
at 1% and 5%, respectively. This suggests that divesting managers are more likely to attempt an auction
sale or market check as better growth opportunities may make the firm more attractive to bidders and
allow for a successful auction. Firms with high managerial ownership are less likely to use an auction
sale, possibly due to their interest in buying into the firm. We observe similar but somewhat weaker
results in the Bidder Competition models, where only realized divestment and growth opportunities are
positively related to competition in that divesting managers select genuinely interested buyers to
27
participate in the sale of the firm. Another possible explanation for these results is that strong operating
performance makes an auction more likely. Moreover, if the auction is successful, managers may chose
to divest if the auction outcome meets their reservation price. We argue that this flow of causality is
unlikely to drive our result and present additional empirical evidence in the Endogeneity section.
In the next set of Table 6 models, we regress a 1-year buyout premium and a transaction
multiple, Deal/Sales, on divestment and a set of control variables. We chose to focus on these variables
rather than measures capturing pricing relative to short-term benchmarks (i.e., 4-week premium) as our
univariate results indicate that bidders may take into account long-term stock price trends when placing
their initial bid. For example, firms with divesting managers may have a high 1-year premium, but a
lower 4-week premium.
We find a significant at 5% and positive effect of realized and predicted divestment on the 1-year
premium indicating that divesting managers succeed in obtaining better buyout pricing. In these two
models, the only other significant control variables are profitability and firm size, indicating that efforts
to manage earnings translate into higher buyout values and that large firms obtain lower premiums.
Lastly, our Deal/Sales regressions confirm a positive and also significant at 5% relation between buyout
pricing and divestment. Profitability enters with a positive coefficient indicating that earnings
management may affect buyout pricing; leverage and managerial ownership enter with negative
coefficients. Overall, we find that divesting managers are able to achieve more favorable buyout
pricing.
[Insert Table 6 here]
Thus far, we have shown that managers tend to act opportunistically prior to LBOs and their
actions appear to pay off. Now, we turn to post-LBO performance to examine whether significant
divestment dis-incentivizes reduces the incentives for managers after the buyout. For our primary
28
measure of post-LBO performance, we hand-collect select financial data for a subset of LBO firms.
After going private, most firms are not required to file their annual reports with the SEC. The filings are
only available for firms that file voluntarily, have outstanding public debt or back-filled financials after
going public again, and, in some cases, after being subsequently acquired by public firms12. We were
able to locate such filings and extract post-LBO financial data for 54 firms. For this subset of firms, we
examine operating performance measured by EBITDA/Sales in years t=1 and t=2 relative to the LBO.
The usefulness of data for t=3 is greatly limited by the available sample size.
We report post-LBO operating performance in Table 7. The results are provided for three groups
of firms based on the change in management’s investment similar to Table 3. The average
EBITDA/Sales ratios indicate that the post-LBO performance in the sample of Investment/Rollover deals
is somewhat better than in the sample of High Divestments in the first year after the buyout. The average
EBITDA/Sales of Investment/Rollover deals exceeds that of High Divestments by almost a factor of 3.3
(0.226 vs. 0.069) with the difference being statistically significant, but only at the 10% level. In the
second year, the difference remains rather large (0.150 vs. 0.045) although it lacks statistical
significance.
Lack of significance in means could still support our case that the divestment group takes more
risks, only if one of the two is true: a) high divestment group has greater risk, or earnings volatility. b)
divestment group has greater number of extreme negative returns (earnings), all due to greater risk
taking. Perhaps, we already have these numbers.
The medians, however, are quite similar. Additionally, we examine changes in operating
performance from year t=-1 to years t=1 and t=2. High divestment deals experience more negative
changes in operating performance with change to year t=1 being statistically significant at the 10% level.
12
Typically, IPO or merger filings contain two to three years of back-filled financial data for the issuer or the target firm.
Poor availability of financial data for post-LBO firms is a common problem for studies that examine value creation in LBOs.
29
The change to year t=2, although still negative, does not meet conventional levels of significance. We
also generate a measure of Excess EBITDA/Sales, which is adjusted for mean-reversion that is especially
likely to affect performance after earnings management. Excess EBITDA/Sales is the difference between
the LBO firm’s post-buyout EBITDA/Sales and the control firms’ EBITDA/Sales. Control firms are five
firms from the same industry and with similar operating performance in year t=-1. The results using
excess measures are similar to those using unadjusted measures.
These results suggest that firms where managers invest additional personal funds in the firm at
the time of the LBO perform slightly better after the LBO. However divesting managers perform almost
as well in year t=2. Such small differences in post-LBO performance suggest that changes in post-LBO
governance and active monitoring by the private equity sponsor may partially mitigate the agency
problem resulting in poor post-LBO performance and greater risk taking by managers. All but four firms
in the subsample with post-LBO data have some degree of private equity backing, which reduces the
active influence of managers on firm performance and its risk. However, more data is needed to answer
the question conclusively. As an additional test reported in the Internet Appendix, we examine the postbuyout bankruptcy and financial distress rates of firms in our sample. The data were collected by
performing extensive Lexis-Nexis and web searches. We find that 17.1 percent of high divestment firms
are likely to default after the LBO, compared to 11.8 percent of firms with managerial investment and
rollovers. However, the difference in proportions is not statistically significant. The latter finding is
somewhat consistent with increased risk taking incentives of divesting managers, although more
evidence is necessary.
[Insert Table 7 here]
Endogeneity
30
One issue that could affect the robustness of our results is potential reverse causality between
firm performance and managerial divestment in that strong performance can lead divestment. Moreover,
they can be determined simultaneously as a function of the same firm characteristics. We have partially
addressed this issue by demonstrating that the decision to sell the firm is made almost a year prior to the
buyout announcement, and that deliberate earnings management takes place well before the buyout.
Moreover, we have shown in Table 4 that the decision to invest or divest is consistent with managerial
characteristics associated with the need (or lack of it) for control. Proximity to retirement and firm
founder’s need to maintain control of the firm suggest that the causality is likely to flow from the need
to divest to the actions affecting the buyout. We also show that the majority of our multivariate results
hold when we measure divestment by its expected value which is a function of manager characteristics,
market returns, market for corporate control and firm size prior to earnings management.
Our method of sale results and buyout pricing results are also prone to endogeneity. Since firms
experience strong operating performance before divestment, the likelihood of an auction sale can be
driven by performance rather than the need for divestment. Moreover, strong operating performance can
improve success of the bidding process, leading to higher divestment. We show that our results hold
when we use expected divestment fitted with variables independent of firm performance such as
executives’ age and need for control as well as availability of exit and investment opportunities (Model
1 of Table 4). This approach is similar to an instrumental variable (IV) approach since several variables
in this regression are related to divestment but are unlikely to have a direct effect on pre-buyout firm
performance. In other words, they can serve as instruments. We also implement the traditional IV
approach and employ financial performance controls from the second stage in addition to the
instrumental variables (specification used in Model 2 of Table 4); the fitted value of divestment
estimated in the first stage is used as an independent variable in the second stage along with control
31
variables. The results reported in the Internet Appendix continue to hold, although in the auction
regression the t-statistic of expected divestment coefficient declines from 2.4 to 1.7, remaining
statistically significant at the 10% level.
Since our IV regressions confirm potential for endogenetity, we further examine a setting where
managers are initially indifferent to divestment, but if the firm is sold through an auction and competing
buyers may bid up the offer price above the managers’ reservation price, managers will divest. This will
induce a positive relation between the likelihood of an auction and divestment, albeit with the opposite
causality. We argue that this scenario does not fully drive our result. First, an auction sale is a costly and
deliberate decision rather than a random event. It requires identification of multiple strategic and
financial buyers, distribution of confidential financial information and substantial marketing to elicit
buyers’ interest. This effort is likely to pay off when managers want to maximize the value of their
divested shares. Managers who want to gain control of the firm would choose a process where they can
restrict the buyout price. They would avoid an auction and privately contact a few large sponsors such
as private equity firms. Thus the method of sale is chosen after the decision to divest or invest is made.
Fidrmuc, Roosenboom, Paap and Teunissen (2012) confirm that firms choose a method of sale to fit
their firm and deal characteristics including private information. Moreover, they show that the method of
sale is the primary decision in the selling process and determines the acquirer type. Second, in cases
where managers were already planning a large divestment, there may be little room for a divestment
increase regardless of the auction outcome. Therefore, a change in divestment is more likely in buyouts
where managers originally planned a small divestment, but increased it after observing a favorable
auction outcome.
We further explore the magnitude of this effect. The reservation prices are more likely to be met
in an auction with large upward bid revisions. Therefore, we may see high divestment levels in auctions
32
with high revisions. However, if there is no relation between revisions and divestment, the need for
divestment is likely driven by factors other than favorable pricing. In the Internet Appendix, we present
regressions in which we re-estimate divestment models from Table 4 on a subsample of auctions and/or
market checks and, more generally, deals with bidder competition. We add an independent variable
measuring the percent revision from the initial bid to the offer price. In the auction/market check
subsample the coefficient on revision is positive but statistically insignificant with a t-statistic of 1.0. In
the regression limited to deals with bidder competition, the coefficient on revision is negative with a
t-statistic of -1.2. In both regressions, measures such as CEO age, relative power of the CEO, and
founder dummy or family-owned dummy continue to be significant. These results suggest that although
some managers’ divestment may be positively affected by auction pricing, however, other factors like
their personal characteristics or need for divestment dominate.
Club deals
Club deals have been the subject of recent controversy regarding the ability of private equity
funds to engage in collusive practices to purchase target firms at more favorable terms. The Department
of Justice started inquiring into private equity's bidding practices as early as 2003 and stepped up the
effort in 2006, which was promptly followed by several civil suits by the shareholders of the companies
acquired by private equity clubs.
The academic literature has not yet reached a consensus on whether club deals are associated
with lower buyout prices. Officer, Ozbas and Sensoy (2008) find that existing shareholders receive 10%
less in club deals. Cao (2008) comes to the opposite conclusion that club deals increase the wealth of the
existing shareholders. Guo, Hotchkiss and Song (2009) trace the sources of pricing differences to the
private equity firms’ ability to identify best performers rather than the ability to eliminate the
33
competition. A relatively large proportion of our sample (23%) is made up of club deals, which allows
us to test for the effect of club deals on buyout pricing. In the Internet Appendix, we present regressions
where we add a Club dummy variable to the buyout pricing models in Table 6. The variable is not
statistically significant in all regressions, indicating similar pricing for club and non-club buyouts.
5. Conclusions
Using a hand-collected data set of LBO transactions, we find that in the last two decades the
average management team reduced its dollar investment in the firm while maintaining a significant
ownership stake in the post-buyout firm. This leads to an agency problem opposite of that predicted for
a traditional LBO in which managers commit substantial personal wealth. We show that managers’
divestment is positively associated with pre-LBO upward earnings management and market timing.
Moreover, divesting managers are more likely to stage an auction sale to obtain higher buyout pricing,
which erodes private equity returns. Post-buyout performance for firms with divesting managers is also
slightly weaker for at least one year following the buyout. Such evidence questions the ability of private
equity funds to detect such manipulation, i.e., if private equity investors are aware of manipulation, why
are they willing to pay higher prices? Could it be that the most optimistic private equity funds bid or
winning bidding? Or, those are careless, or least capable bid?
One explanation has been advanced by Kaplan and Stein (1993) in an effort to make sense of the
abrupt decline in buyout activity in the early 1990’s. They suggested that the success of the 1980s
buyout wave attracted a large inflow of funds and by the end of the 1980s “too much financing was
chasing too few good deals,” which led to many overpriced and poorly structured transactions. Our data
34
demonstrates that the increase in merger activity and market returns positively correlate with larger
divestments, which is consistent with pressure on private equity firms to allocate excess funds.
35
References:
Anderson, R. and D. Reeb, 2003,” Founding-family ownership and firm performance: evidence from the
S&P 500”, Journal of Finance 58, 1301– 1328.
Anilowski, C., Macias, A., and J. Sanchez, 2009, “Target firms earnings management and the method of
sale: evidence from auctions and negotiations”, Working paper.
Betton, S. and E.Eckbo, 2000, “Toeholds, bid jumps, and expected payoffs in takeovers”, Review of
Financial Studies 13, 841-882.
Betton, S., Eckbo, E., and K.Thorburn, 2008, “Markup pricing revisited”, Working paper.
Bitler, M, Moskowitz, T., and A.Vissing-Jorgensen, 2005, “Testing agency theory with entrepreneur effort
and wealth”, Journal of Finance 60, 539-576.
Cao, J., 2008, “An empirical study of LBOs and takeover premium”, Working paper.
Cohen,D, Dey, A., and T.Lys, 2007, “Real and accrual-based earnings management in the pre- and postSarbanes Oxley periods”, Working paper.
Cornelli, F. and D.Li, 2006, “Ex ante effects of ex post managerial ownership”, Working paper.
Cotter, J. and M. Zenner, 1994, “How managerial wealth affects the tender offer process”, Journal of
Financial Economics 35, 63-97
Crawford, E., 1987, “A management guide to leveraged buyouts: A case study digest”, John Wiley &
Sons, Inc.
Davidoff, S.,”A Cautionary Tale as Private Equity Exits Two Deals”, The New York Times, Feb. 8, 2011
DeAngelo, L., 1986, “Accounting numbers as market valuation substitutes: A study of management
buyouts of public stockholders,” Accounting Review 61, 400 – 420.
Dechow, P., Sloan, R., and A. Sweeney, 1996, “Causes and consequences of earnings manipulations: An
analysis of firms subject to enforcement actions by the SEC”, Contemporary Accounting Research
13, 1-36.
Elitzur, R., Halpern, P., Kieschnick, R., and Rotenberg, W., 1998, “Managerial incentives and the
structure of management buyouts,” Journal of Economic Behavior & Organization 36, 347 – 367.
EricksonM. , M. Hanlon and E. Maydew, 2006, “Is there a link between executive
equity incentives and accounting fraud?”, Journal of Accounting Research 44, 113-143.
Easterwood, J., Singer, R., Seth, A. and D. Lang, 1994, “Controlling the conflicts of interest in
management buyouts”, Review of Economics and Statistics 76, 512-522.
Ferreira D., G.Manso and A.Silva , 2011, “Incentives to innovate and the decision to go public or
private”, Working paper.Fishman, M, 1988, “A theory of preemptive bidding”, Rand Journal of
Economics 19, 88-101.
Fidrmuc J., P. Roosenboom, R.Paap and T.Teunissen, 2012, “One size does not fit all: Selling firms to
private equity versus strategic acquirers”, Journal of Corporate Finance, forthcoming.
Fischer, P., and H. Louis, 2008, "Financial reporting and conflicting managerial incentives: The case of
management buyouts". Management Science 54, 1700 – 1714.
Frankfurter, G. and E. Gunay, 1992, “ Management buyouts: the sources and sharing of wealth between
outside and inside shareholders”, Quarterly Review of Economics and Finance 32, 82-95.
Gomez-Mejia, L., Nunez-Nickel, M., and I. Gutierrez, 2001, “ The role of family ties in agency
contracts”, Academy of Management Journal 44, 81-95.
Guo, S., Hotchkiss, E., and W. Song, 2009, “Do buyouts (still) create value?”, forthcoming in Journal of
Finance.
Harlow, W. and J. Howe, 1993, “Leveraged buyouts and insider nontrading”, Financial Management
22, 109-118.
36
Hartzell, J., Ofek, E. and D.Yermack, 2004, “What’s in it for me? CEOs whose firms are acquired”,
Review of Financial Studies 17, 37-61.
Jensen, M, and Meckling, W., 1976, “Theory of the firm: Managerial behavior, agency costs and
ownership structure”, Journal of Financial Economics 3, 305 – 360.
Jensen, M., 1989, “Eclipse of the public corporation”, Harvard Business Review 67, 61-74.
Kaplan, S., 1989, “Management buyouts: evidence on taxes as a source of value”, Journal of Finance
44, 611-632.
Kaplan, S., and J.Stein, 1993, “The evolution of buyout pricing and financial structure in the 1980s,”
Quarterly Journal of Economics 108, 313 – 357.
Kieschnick, R., 1998, “Free cash flow and stockholder gains in going private transactions revisited”,
Journal of Business Finance and Accounting 25, 187-202.
Lakonishok, J. and I. Lee, 2001, “Are insider trades informative?”, Review of Financial Studies 14, 79111.
Lehn, K., and A. Poulsen, 1989, “Free cash flow and stockholder gains in going private transactions”,
Journal of Finance 44, 771-787.
Lowenstein, L., 1985, “Management buyouts”, Columbia Law Review 85, 730-784.
Meagher, B., “Who Will Win the Battle for Restoration Hardware?”, Northbay Biz, Feb. 2008
Nikoskelainen, E. and M. Wright, 2007, “The impact of corporate governance mechanisms on value
increase in leveraged buyouts”, Journal of Corporate Finance 13, 511-537.
Officer M., O. Ozbas and B. Sensoy, 2010, “Club deals in leveraged buyouts", Journal of Financial
Economics 98, 214-240.
Perry, S. and T.Williams, 1994, “Earnings management preceding management buyout offers”, Journal
of Accounting and Economics 18, 157 – 179.
Richardson, S., I. Tuna, and M. Wu. 2002. Predicting earnings management: The case of earnings
restatements. Working Paper, University of Pennsylvania
Roychowdhury, S., 2006, “Earnings management through real activities manipulation”, Journal of
Accounting and Economics 42, 335-370.
Schadler, F. and J. Karns, 1990, “The unethical exploitation of shareholders in management buyout
transactions”, Journal of Business Ethics 9, 595-602.
Schwert, W.,1996, ”Markup pricing in mergers and acquisitions”, Journal of Financial Economics 41,
153-192.
Schulze, W., Lubatkin, M., Dino, R., and A. Buchholtz, 2001, “Agency relations in family firms: theory
and evidence”, Organizational Science 12, 99-116.
Smith, C., and R. Stulz, 1985, “ The determinants of firms’ hedging policies”, Journal of Financial and
Quantitative Analysis 20, 391-405.
Subramanian, G., 2008, “Go-Shops vs. No-Shops in Private Equity Deals: Evidence and Implications”,
Working paper.
Sweeney, A., 1994, “Debt covenant violations and managers’ accounting responses”, Journal of
Accounting and Economics 17, 281-308.
Teoh, S., Welch, I., and T. Wong, 1998, “Earnings management and the underperformance of seasoned
equity offerings,” Journal of Financial Economics 50, 63 – 99.
Wright M., R. Hoskisson and L.Busenitz , 2001, “Firm rebirth: Buyouts as facilitators of strategic
growth and entrepreneurship”, Academy of Management Executive 15, 111-125.
37
Figure 1. Pre-LBO Buy-and-Hold Abnormal Returns
This figure demonstrates monthly buy-and-hold abnormal returns (BHAR) over the twelve months prior to announcement.
The BHARs are computed by subtracting compound return to the value weighted CRSP index from the compound return of
the LBO firm over the same period. The sample is divided into three groups according to the managers’ divestment.
Investment/Rollover are deals to which managers contributed additional personal equity or fully reinvested (rolled over) their
pre-LBO equity. Divestments are split into two groups: Low (relative divestment < 50%) and High (relative divestment
50%<).
38
Table 1. Summary of Agency Costs by LBO Type
LBO Types
Investment/Rollover, Low Divestment :
High Divestment:
Buyouts in which managers act as buyers
Buyouts in which managers act as
and contribute additional personal wealth
selling shareholders and convert part of
or re-invest 100% of their pre-LBO firm
their pre-LBO shareholdings to cash.
dollars in the post-LBO firm.
The extent of such divestment varies
from low to high.
Alignment of interests
a. Managers vs. shareholders
Not aligned
Aligned
b. Managers vs. new investors
Aligned
Not aligned
Agency problems/managers actions
a. Pre-LBO
Incentive to minimize purchase price as
Incentive to maximize buyout price as
managers are buyers. May take actions to
managers are selling shareholders.
decrease short term firm value: decrease
May take actions to increase short term
earnings via negative accruals or real
firm value: increase earnings via
manipulation. Go private during periods of
positive accruals or real manipulation.
low market values.
Go private during periods of high
market values.
b. At LBO
c. Post-LBO
Negotiated sale, low competition, low
Auction sale, high bidding competition,
bidding activity, low price revisions, low
high bidding activity, high price
premium.
revisions, high premium.
High performance incentives.
Low effort incentives.
39
Table 2. Annual Distribution of LBOs and Managerial Divestment
This table reports annual distribution of LBOs, net dollar divestment and relative divestment. Net dollar divestment is the
amount received by management for the shares valued at LBO price less the amount reinvested in the firm. Relative
divestment is net dollar divestment scaled by the dollar value of pre-LBO management team ownership.
Year
Number of
LBOs
Total Net
Dollar
Divestment
Relative
Divestment
% Deals with
Divestment
% Deals with
50%+ Divestment
1997
7
527.571
0.601
0.857
0.857
1998
14
191.676
0.593
0.857
0.643
1999
16
238.090
0.280
0.813
0.313
2000
27
321.004
0.364
0.630
0.370
2001
17
60.313
0.201
0.706
0.235
2002
7
9.385
0.203
0.714
0.286
2003
20
67.649
0.021
0.700
0.100
2004
15
406.559
0.389
0.733
0.333
2005
14
493.263
0.553
0.929
0.571
2006
19
1437.880
0.711
1.000
0.737
2007
17
1983.329
0.704
1.000
0.765
2008
6
363.748
0.701
1.000
0.667
Total
179
6100.466
0.421
0.810
0.458
40
Table 3. Sample Characteristics by Buyout Type
This table reports stated reasons for buyouts as well as divestment and buyout characteristics classified according to
management divestment. Investment/Rollover are deals to which managers contributed additional personal equity or fully
reinvested (rolled over) their pre-LBO equity. Divestments are split into two groups: Low (relative divestment < 50%) and
High (relative divestment 50%<). Net dollar divestment is the amount received by management for the shares valued at LBO
price less the amount reinvested in the firm. Relative divestment is net dollar divestment divided by pre-LBO dollar
management team ownership. Deal value is the dollar amount paid by the acquirer for the target. Price per share is the price
paid by the acquirer for each share of the target. Management ownership is the percentage of shares owned by all executive
officers in the pre-LBO proxy statements. Top 3 ownership is the percentage of shares owned by the President, CEO, and
Chairman of the Board. Management team post-LBO ownership is the percentage of shares owned by the management team
in the post-buyout firm. Management deals are transactions where management is the sole acquirer. Management with private
equity deals are transactions where management teams up with private equity to make the acquisition. Single private equity
deals are deals carried out by a single private equity fund. Club deals are buyouts carried out by two or more private equity
funds. Total assets is the book value of assets. Leverage is total liabilities divided by total assets. EBITDA is operating
income before depreciation. Market-to-book is stock price divided by book value per share. Stock return is a buy-and-hold
abnormal return (BHAR (-12,-1)) computed by subtracting the compound return to the value weighted CRSP index from the
compound return of the LBO firm over months (-12,-1) relative to announcement. Volatility is the standard deviation of
monthly returns over the same horizon. Liquidity is the average daily turnover over the same time horizon. T-statistics and zstatistics are reported for the difference in means and medians (Wilcoxon) tests between Investment/Rollover and High
Divestment sub-samples. Tests of proportions were conducted where appropriate. Values significantly different from zero at
the 10%, 5% and the 1% level are marked *,** and *** respectively.
Divestment
Investment/Rollover
N.
%
Low < 50%
N.
%
High 50%<
N.
%
diff. prop.
z-stat
Panel A. Reasons for LBO
Low stock liquidity
25
0.735
26
0.413
13
0.159
6.0***
Undervaluation
24
0.706
37
0.587
29
0.354
3.5***
Limited growth potential
9
0.265
24
0.381
21
0.256
0.1
Poor performance
16
0.471
20
0.317
21
0.256
2.3**
Low institutional ownership
8
0.235
8
0.127
1
0.012
4.1***
Poor access to capital
7
0.206
10
0.159
6
0.073
2.1**
Cost of being public
18
0.529
19
0.302
11
0.134
4.5***
Cost of Sarbanes Oxley Act
4
0.364
11
0.175
5
0.061
1.0
Market pressure
8
0.235
16
0.254
8
0.098
2.0**
Mgmt/block. wants control
6
0.177
5
0.079
6
0.073
1.7*
Mgmt/block. wants divers.
0
0.000
3
0.048
11
0.134
-2.2**
Good offer price
3
0.088
7
0.111
25
0.305
-2.5**
Opportune time to sell firm
0
0.000
4
0.063
21
0.256
-3.3***
Obs.
34
63
82
41
Divestment
Investment/Rollover
Mean
Median
Low <50%
Mean
Median
High 50%<
Mean
means
Median t-stat
medians
z-stat
Panel B. Divestment and LBO characteristics
Net dollar divestment
-0.432
0.000
13.513
3.172
64.194
24.523
-6.0***
-8.5***
Relative divestment
-0.137
0.000
0.186
0.201
0.833
0.841
-12.9***
-8.5***
177.415
28.907
803.067
70.996
-3.3***
-5.2***
Price per share
9.136
6.000
15.523
10.250
21.825
17.750
-5.8***
-4.7***
Management ownership
0.322
0.202
0.297
0.262
0.172
0.089
2.9***
3.0***
Top 3 ownership
0.289
0.161
0.259
0.186
0.137
0.055
3.13***
3.5***
Post-LBO management ownership
0.656
0.950
0.562
0.665
0.154
0.065
5.8***
4.5***
Management
0.647
1.000
0.476
0.000
0.073
0.000
6.6***
6.5***
Management with private equity
0.147
0.000
0.127
0.000
0.110
0.000
0.6
0.6
Single private equity
0.118
0.000
0.190
0.000
0.524
1.000
-4.1***
-4.0***
Club
0.088
0.000
0.190
0.000
0.293
0.000
-2.4**
-2.1**
-2.4**
-3.5***
Deal value
1916.140 332.478
Panel C. Pre-LBO Firm Characteristics
Total assets
342.950
81.726
684.015
145.407
Leverage
0.567
0.578
0.520
0.509
0.540
0.542
0.5
0.5
EBITDA/sales
0.092
0.097
0.082
0.088
0.126
0.110
-1.5
-1.6*
Market-to-book
1.454
0.874
1.809
1.122
2.941
1.831
-2.6***
-4.5***
Stock return
-0.186
-0.285
-0.092
-0.216
0.051
-0.037
-2.4***
-2.1**
Volatility
0.210
0.150
0.163
0.149
0.123
0.111
2.1**
3.5***
Liquidity
3.046
3.536
5.539
4.144
6.275
5.162
-3.9***
-2.7***
Obs.
34
63
1245.511 264.856
82
42
Table 4. Determinants of Managerial Divestment
The table reports results from Tobit regressions censored at 1. Relative divestment is net dollar divestment divided by preLBO dollar management team ownership. EBITDA/sales is operating income before depreciation divided by net sales.
Leverage is total liabilities divided by total assets. Market-to-book is stock price divided by book value per share.
Management ownership is % of shares owned by all named executive officers. Age is the average age of top 3 officers.
Founder dummy equals 1 if the firm founder is an executive officer and 0 otherwise. Family dummy equals 1 if the firm is
family owned and 0 otherwise. Top3 influence is the ratio of Top 3 ownership to management ownership. Industry market-tobook is the average market-to-book ratio of firms in the same 2-digit SIC code as the LBO firm. Industry merger activity is
the relative volume of merger activity in the same 2-digit SIC code as the LBO firm. Industry private equity activity is the
relative volume of merger deals carried out by private equity firms in the same 2-digit SIC code as the LBO firm. Market
return is one-year compound return to the value weighted CRSP index. All financial variables are as of t=-1 relative to the
year of LBO announcement. Robust t-statistics with year clustering are reported in ( ). Values significantly different from
zero at the 10%, 5% and the 1% level are marked *,** and *** respectively.
Relative Divestment
Constant
Age
Founder dummy
Family dummy
Top3 influence
Management ownership
Industry market-to-book
Industry merger activity
Industry priv. equity activity
Market return
Log total assets
-0.274
(-0.9)
0.017***
(3.7)
-0.231***
(-2.7)
-0.102
(-1.5)
-0.483***
(-2.8)
-0.391**
(-2.0)
-0.010***
(-4.0)
0.721
(0.4)
0.000
(0.3)
0.474**
(2.5)
0.053***
(3.1)
-0.257
(-0.9)
0.017***
(4.3)
-0.218***
(-2.8)
-0.091
(-1.4)
-0.473***
(-3.0)
-0.410**
(-2.2)
-0.010***
(-4.7)
0.517
(0.3)
0.000
(0.4)
0.419***
(2.6)
0.039**
(2.0)
0.013
(1.4)
0.493**
(2.2)
-0.089
(-0.8)
179
179
0.301
0.319
Market-to-book
EBITDA/sales
Leverage
Obs.
Pseudo R2
43
Table 5. Pre-LBO Earnings Management
Panel A presents summary statistics for pre-buyout earnings management. Relative divestment is net dollar divestment divided
by pre-LBO dollar management team ownership. Investment/Rollover are deals in which managers contributed additional
personal equity or fully reinvested (rolled over) their pre-LBO equity. Divestments are grouped into Low (relative divestment <
50%) and High (relative divestment 50%<). Discretionary current accruals (DCA) follow the modified Jones methodology. Real
earnings management follows Roychowdhury (2006). Strategic evaluation (days) is from the decision to sell to buyout
announcement. Variables in panel A are as of t=-1and t=-2 relative to the year of LBO announcement. T and z statistics are
reported for the difference in means and medians (Wilcoxon) between investment/rollover and high divestment deals. Panel B
reports results from four OLS regressions. Expected relative divestment is fitted following Model 1 in Table 4. EBITDA/sales is
operating income before depreciation divided by net sales. Leverage is total liabilities divided by total assets. Market-to-book is
stock price divided by book value per share. Volatility is the standard deviation of monthly returns over the fiscal year.
Management ownership is % of shares owned by all executive officers. All financial variables in Panel B are as of t=-2 relative
to the year of LBO announcement. Robust t-statistics with year clustering are reported in ( ). Values significantly different from
zero at the 10%, 5% and the 1% level are marked *,** and *** respectively.
Panel A. Pre-LBO Financial Manipulation
Divestment
Investment/Rollover
Low < 50%
High 50%<
means medians
t-stat
z-stat
Mean
Median
Mean
Median
Mean
Median
DCA
-0.014
-0.013
0.011
0.007
0.049
0.018
-3.0***
-2.5**
Real earn. management
0.133
0.156
0.153
0.122
0.306
0.158
-1.7*
-0.9
DCA t=-2
-0.002
0.001
0.009
0.009
0.007
-0.001
-0.3
-0.5
Real earn. management t=-2
0.260
0.207
0.145
0.144
0.265
0.172
-0.1
0.7
Strategic evaluation (days)
326.500
219.500
349.540
262.000
293.720
259.000
0.6
-0.3
Panel B. Pre-LBO Financial Manipulation Regressions
Discretionary Current Accruals (DCA)
Intercept
Relative divestment
-0.027
(-0.5)
0.064***
(4.9)
EBITDA/sales
Leverage
Market-to-book
Volatility
Management ownership
Obs.
R2
-0.015
(-0.2)
0.310**
(2.2)
-0.069
(-0.5)
-0.005
(-0.5)
0.024
(0.5)
-0.014
(-0.4)
0.005***
(2.8)
0.188**
(2.2)
0.044
(0.9)
0.070***
(3.7)
-0.005
(-0.5)
0.026
(0.5)
-0.026
(-0.9)
0.006**
(2.7)
0.203**
(2.6)
0.058
(1.0)
0.003
(0.1)
0.058
(0.1)
-0.240
(-1.5)
0.016*
(1.8)
0.905***
(5.0)
0.164
(1.1)
0.378**
(2.2)
0.001
(0.1)
0.096
(0.2)
-0.290
(-1.6)
0.019*
(1.8)
1.004***
(4.8)
0.269
(1.0)
167
167
147
147
0.071
0.047
0.122
0.091
Exp. relative divestment
Log total assets
-0.033
(-0.6)
Real Earnings Management Proxy
44
Table 6. Choice of Sale Method and Target Wealth Effects
Panel A presents summary statistics for method of sale, sale process and buyout pricing. Relative divestment is net dollar divestment divided by pre-LBO dollar
management team ownership. Investment/Rollover are deals to which managers contributed additional personal equity or fully reinvested (rolled over) their preLBO equity. Divestments are split into two groups: Low (relative divestment < 50%) and High (relative divestment 50%<). Buyout premium 4 wk (1yr) is the
percent change from the stock price four weeks (1 year) before buyout announcement to buyout price. Deal/sales is a transaction multiple computed as deal value
divided by net sales. Auction is a dummy variable equal to 1 if the target firm attempts an auction sale and 0 otherwise. Auction and/or market check is a dummy
variable that equals 1 if the firm attempts an auction, or, in case of an unsolicited bid, performs a “market check”, and 0 otherwise. Bidder competition is a
dummy variable that equals 1 if there is more than one bidder and 0 otherwise. Number of bid revisions is the number of bid revisions by the winning bidder.
Offer revision is percent change from initial bid to offer price. T-statistics and z-statistics are reported for the difference in means and medians (Wilcoxon) tests
between investment/rollover and high divestment sub-samples. Tests of proportions were conducted where appropriate. Panel B reports results from four Logit
regressions estimating the likelihood of auction and/or market check and bidder competition and four OLS regressions estimating buyout premium (1 yr) and
deal/sales. Expected relative divestment is fitted following Model 1 in Table 4. EBITDA/sales is operating income before depreciation divided by net sales.
Leverage is total liabilities divided by total assets. Market-to-book is stock price divided by book value per share. Management ownership is % of shares owned
by all executive officers. All financial variables are as of t=-1 relative to the year of LBO announcement. Pseudo R2 is reported for the Logit regressions and R2
for the OLS regressions. Robust t-statistics with year clustering are reported in ( ). Values significantly different from zero at the 10%, 5% and the 1% level are
marked *, ** and *** respectively.
Panel A. Summary Statistics for LBO Pricing and Buyout Process Characteristics
Divestment
Investment/Rollover
Low < 50%
High 50%<
Mean
Median
Mean
Median
Mean
Median
means
t-stat
Buyout premium (4wk)
0.514
0.471
0.412
0.322
0.313
0.275
2.8***
2.7***
Buyout premium (1yr )
0.275
0.138
0.218
0.143
0.566
0.442
-2.0*
-2.2**
Deal/sales
1.034
0.336
0.850
0.520
1.427
0.868
-4.4***
-5.1***
Auction
0.118
0.000
0.460
0.000
0.646
1.000
-5.2***
-5.2***
Auction /market check
0.353
0.000
0.603
1.000
0.866
1.000
-5.6***
-5.5***
Bidder competition
0.294
0.000
0.397
0.000
0.610
1.000
-3.1***
-3.1***
Number of bid revisions
2.765
3.000
2.968
3.000
3.110
3.000
-1.2
-1.5
Offer revision
0.129
0.077
0.067
0.063
0.069
0.048
1.3
1.5
Obs.
34
63
medians
z-stat
82
45
Panel B. LBO Pricing and Buyout Process Regressions
Auction and/or
Market Check
Intercept
Relative divestment
0.390
(0.9)
2.162***
(4.2)
-0.047
(-0.5)
1.414
(0.6)
-0.917
(-1.6)
0.344*
(1.7)
-1.711*
(-1.9)
1.116**
(2.4)
-0.017
(-0.2)
1.471
(0.7)
-0.941
(-1.5)
0.456*
(1.9)
-1.815*
(-2.0)
179
0.236
Exp. relative divestment
Log total assets
EBITDA/sales
Leverage
Market-to-book
Management ownership
Obs.
Pseudo R2/R2
0.419
(0.8)
Bidder Competition
-0.640
(-1.1)
1.212***
(4.3)
-0.467
(-0.9)
0.030
(0.3)
-1.344
(-0.8)
-0.618
(-1.1)
0.224***
(2.8)
-0.592
(-0.9)
0.484
(1.0)
0.057
(0.5)
-0.942
(-0.6)
-0.686
(-1.2)
0.250***
(2.6)
-0.838
(-1.2)
179
179
0.168
0.106
Buyout Premium (1yr)
0.162
(0.6)
0.347**
(2.6)
0.128
(0.6)
Deal/Sales
0.530*
(2.1)
0.410**
(2.7)
0.514*
(2.1)
-0.050
(-1.3)
1.274**
(2.2)
0.297
(1.0)
0.021
(0.7)
0.027
(0.1)
0.579**
(2.9)
-0.073**
(-2.2)
1.439*
(2.1)
0.221
(0.7)
0.026
(0.9)
0.212
(0.8)
0.086
(1.6)
2.976**
(2.6)
-0.900*
(-2.1)
0.064*
(1.8)
-0.519**
(-2.8)
0.567**
(2.4)
0.072
(1.6)
3.131**
(2.8)
-0.941*
(-2.2)
0.069*
(2.0)
-0.387*
(-1.8)
179
175
175
178
178
0.075
0.127
0.120
0.338
0.329
46
Table 7. Post-LBO Performance
This table reports post-LBO operating performance for the first two years after the LBO. Excess EBITDA/sales is adjusted for mean reversion and industry
effects. The sample is divided into three groups according to the managers’ relative divestment. Relative divestment is net dollar divestment divided by pre-LBO
dollar management team ownership. Investment/Rollover are deals to which managers contributed additional personal equity or fully reinvested (rolled over)
their pre-LBO equity. Divestments are split into two groups: Low (relative divestment < 50%) and High (relative divestment 50%<).
Divestment
Investment/Rollover
Low <50%
High 50%<
Mean
Median
Obs.
Mean
Median
Obs.
Mean
Median
Obs.
means
t-stat
medians
z-stat
EBITDA/sales
t=1
0.226
0.144
10
0.137
0.129
13
0.069
0.106
31
1.8*
0.7
EBITDA/sales
t=2
0.150
0.136
6
0.013
0.126
11
0.045
0.118
24
1.2
0.1
∆EBITDA/sales -1,1
0.069
0.006
10
-0.012
-0.020
13
-0.089
-0.014
31
1.9*
1.3
∆EBITDA/sales
0.004
-0.003
6
-0.146
-0.027
11
-0.114
-0.012
24
1.5
0.3
Excess EBITDA/sales t=1
0.089
0.012
10
-0.002
0.000
13
-0.085
-0.002
27
1.8*
0.7
Excess EBITDA/sales t=2
0.038
0.012
6
-0.132
0.007
11
-0.023
0.011
21
0.8
0.0
-1,2
47