Post on 21-Apr-2020
Page 1 of 34
CEO Turnover and Compensation: An Empirical Investigation
Rachel Graefe-Anderson Krannert Graduate School of Management
Purdue University West Lafayette, IN 47907
rdiana@purdue.edu
September 2009
Abstract
Because CEO turnover events provide a the board of directors with a unique opportunity to potentially completely restructure CEO compensation packages, changes to CEO compensation following a turnover event could prove to inform the ongoing debate regarding CEO compensation. This paper investigates what happens to CEO compensation when a turnover event occurs. Specifically, I examine CEO compensation levels and pay-performance sensitivity for incoming and outgoing CEOs involved in turnover events at public companies in the United States. My main findings are as follows: 1) incoming CEOs are paid as much as or more than those they replace, 2) outsider replacements are paid more than their predecessors even after controlling for education and skills, and 3) CEOs who are forced out are not paid differently from those who replace them, while CEOs who leave voluntarily are paid significantly less than their replacements. Further analysis reveals that proxies for managerial power including CEO tenure, CEO centrality, founder status, and high CEO ownership cannot explain these results. Overall, these findings are difficult to reconcile with the view that managerial power is the primary determinant of CEO compensation.
Page 2 of 34
1. Introduction
CEO turnover events provide a unique opportunity for boards of directors to restructure
CEO compensation packages. As such, adjustments to CEO compensation packages occurring
around such events can provide information regarding the ongoing debate concerning CEO
compensation. This paper investigates the determinants of CEO compensation by exploring
what happens to CEO compensation when turnover events take place. Specifically, I examine
CEO compensation levels, pay-performance sensitivity, and “excess pay” (measured as the
residual from a standard model of compensation) for a sample of 1,232 incoming and outgoing
CEOs involved in turnover events in U.S. public companies between 1993 and 2006.
Executive compensation has been a widespread and hotly debated topic for over a
decade. Nonetheless, no consensus has been reached regarding whether CEO compensation
reflects an effective mechanism through which managers’ interests are aligned with those of their
shareholders. The two main views of executive compensation -- managerial discretion and
optimal contracting -- offer competing explanations for the dramatic rise in executive pay over
the past several decades. Under optimal contracting, contracts are the efficient result of arms’
length bargaining between the board of directors and the CEO and the rise in compensation
simply reflects appropriate adjustments in response to market factors. A large literature supports
this view, finding corroborative evidence by looking for compensation behavior that would be
consistent with market forces such as outside employment opportunities or a shift in the types of
qualities desired.1 Alternatively, under the managerial discretion view, managers exercise power
over the board to expropriate from shareholders by overpaying themselves and the overall rise in
executive pay reflects an increase in their power to do so. An equally large literature documents
support for the “overpay” hypotheses by identifying behavior that appears to be inconsistent with
efficiency such as asymmetric benchmarking or stock option re-pricing.2
This paper contributes to the ongoing debate in its evaluation of a relatively unique event.
Because the board of directors is contracting with the presumably less powerful new CEO for the
1 See, for example, Fama (1980), Abowd and Kaplan (1999), Murphy (1999), Core and Guay (1999), Core, Guay, and Larcker (2003), Himmelberg, Hubbard, and Palia (1999), Core, Guay, and Thomas (2005); Rajgopal, Shevlin, and Zamora (2006), Oyer (2004), Himmelberg and Hubbard (2000); Murphy and Zabojnik (2007) 2 See, for exampe, Morck, Shleifer, and Vishny (1988), Crystal (1991), Jensen (1993), Bebchuk, Fried, and Walker (2000); and Mullainathan (2001), Garvey and Milbourn (2004); Brenner, Sundaram, and Yermack (2000), Chance, Kumar, and Todd (2000), Pollock, Fisher, and Wade (2002)
Page 3 of 34
first time, turnovers give the board the option to structure their CEO’s compensation package as
they see fit. In a sense, turnovers provide boards with an opportunity to “start over” regarding
the ways in which they compensate managers and align managers’ incentives with those of
shareholders. Thus, turnovers have the potential to have a dramatic impact on the level and
structure of compensation packages.
This paper focuses on the following question: what is the impact of a turnover event on
CEO compensation? My primary purpose is to describe what we observe when a turnover event
occurs. To a lesser degree, I also try to address whether or to what extent the empirical evidence
allows me to distinguish between competing views of executive compensation. I investigate
these questions by examining the compensation packages of incoming and outgoing CEOs. In
particular, I examine differences in levels of cash pay (salary and bonus), total compensation
(total of salary, bonus, Black-Scholes value stock options granted, restricted stock, payouts from
long-term plans, and benefits), pay-for-performance sensitivities, and “excess pay” for a sample
of 1,232 paired incoming and outgoing CEOs around turnover events occurring between 1993
and 2006. Following the approach used by Hartzell, Ofek, and Yermack (2004), I measure
“excess pay” as the residual from a standard model of CEO pay as a function of firm size, firm
performance (3-year stock price performance), industry, and year.
Overall, I find that incoming CEO levels of pay and pay-for-performance sensitivities are
greater than or do not differ significantly from that of the CEOs they replace. The difference in
total compensation is positive and greater for outsider CEO replacements than for insider CEO
replacements. This is consistent with Murphy (2002) and Murphy and Zabojnik (2007), who
look at just the incoming CEOs and find that outsider CEO replacements receive higher
compensation than insider CEO replacements. I further find that when the CEO is forced out,
there is no significant difference between the total compensation of the incoming and outgoing
CEOs. In contrast, when the CEO leaves voluntarily, the incoming CEO is paid substantially
more than his predecessor.
In distinguishing between the competing perspectives on CEO compensation, the views
discussed above offer predictions for what we expect to observe when an incumbent CEO leaves
and is replaced. For instance, under the assumption that an outgoing CEO’s long-term
involvement as the highest executive within a firm represents enhanced power (over that of his
replacement), the managerial discretion view would predict that, in the event of a turnover, the
Page 4 of 34
outgoing CEO’s pay and “excess pay” should be higher than that of his replacement while his
pay-performance sensitivity should be lower. Furthermore, managerial discretion hypotheses
predict that these effects will be augmented (dampened) for situations in which we expect a
CEO’s bargaining power to be high (low). The overall results listed above appear to contradict
this hypothesis. That is, since I expect outgoing CEOs to have more power than their
replacements, on average, I expect their compensation levels to be higher than those of their
replacements, but find the opposite. When the replacement CEO is an insider, I presume that
since he has already been an executive at the firm for at least a few years, he would have an
existing relationship with the board of directors. This relationship, under the managerial
discretion view, should thus lead to a higher bargaining power for insider replacements than for
outsiders. Thus the results regarding outsiders are also opposite of what is expected. Further, I
interpret a forced turnover based on poor performance as representative of a situation in which
the outgoing CEO’s power over the board has been undermined. Thus, in comparison to
outgoing CEOs leaving voluntarily, the forced out CEOs have relatively less power.
Digging a little deeper, I try to determine whether the baseline results hold because
incoming CEOs are systematically higher quality CEOs. I also try to determine whether
outgoing CEOs’ power over the board has an impact on observed changes in CEO compensation
surrounding a turnover. I find that outsider replacements are paid more than their predecessors
even after controlling for skills using relative educational backgrounds (that is, the education
background of the incoming CEO compared to that of his predecessor) and previous employers’
return on assets (ROA). I further examine differences between incoming and outgoing CEO pay
and pay-performance sensitivities based on outgoing CEO founder status, ownership levels, CEO
centrality3, CEO tenure, board composition, and the firm’s governance index.4
3 Measured as in Bebchuk, Cremers, and Peyer (2007) as the percentage of total top-five executive pay received by the CEO.
I find that
incoming CEOs are paid more in option-based pay and total compensation regardless of whether
the outgoing CEO was the founder and that the differential is larger when the outgoing CEO was
the founder. Incoming CEOs are also paid more in total compensation and option-based pay
when the outgoing CEO’s tenure is greater than 4 years, when his ownership levels are high
(above average, above 5%, and above 10%), and when the board composition indicates fewer
4 The Gompers, Ishii, and Metrick (2003) G-index
Page 5 of 34
than average independent outsiders on the board.5
Taken together, the results are difficult to reconcile with the managerial power view of
executive compensation. On the one hand, we have a number of inconsistencies with the
managerial discretion hypotheses as we partition the data on the basis of managerial power.
However, as we further partition the data to classify additional distinctions of power, some
results tend to become consistent with those hypotheses while others do not. If, as is purported
by the managerial power view, a CEO’s power over the board is a primary driver of both levels
and structures of compensation, much stronger results in line with managerial power hypotheses
would be needed.
However, when the outgoing CEO’s measure
of centrality is above average and when the firm has a very high G-index, the results move in the
opposite direction – with the outgoing CEO’s pay moving to equal or exceed his replacement’s.
As yet, I am aware of only limited research that has explicitly and comprehensively
examined compensation contracts around turnover events. Rather, most papers examine
turnover in the context of its being a potential component of the overall governance package that
is designed to align managers’ incentives and curb rent extraction – that is, the threat of turnover
is typically viewed as playing a disciplinary role in the context of the manager-shareholder
agency problem. These studies typically examine the occurrence of management turnover in
relation to firm performance and other firm or market characteristics.6
In spirit, this paper resembles a varied set of studies regarding CEO compensation.
Gilson and Vetsuypens (1990) examine the nature of compensation packages for financially
distressed firms and include a discussion of changes observed when a turnover event occurs
within this set of firms. They find that, within a small sample of financially distressed firms,
when a turnover occurs, insider replacement CEOs are paid substantially less than their
predecessors, but outsider replacement CEOs are paid substantially more. Murphy (2002)
compares levels of CEO pay for insider and outsider replacement CEOs, finding that outsider
replacements are typically paid more than insider replacements. This study expands upon
Murphy (2002) in two major ways. First, I examine differences between the incoming and
outgoing CEOs in turnover events, rather than differences between replacement CEOs’ pay
5 I use 4 years as a threshold based on evidence found by Gregory-Smith, Thompson, and Wright (2009) that probabilities of CEO dismissal decline after the 4th year and this appears to be due to entrenchment effects. 6 See, for example, Denis, Denis, and Sarin (1997); Coughlan and Schmidt (1985); Murphy and Zimmerman (1993); Huson, Parrino, and Starks (2001); Goyal and Park (2001); Lehn and Zhao (2006)
Page 6 of 34
based on whether they are insiders or outsiders. Second, in an attempt to explain observed
changes in compensation surrounding turnover events, I extend the investigation to incorporate
CEO skills and additional measures of a CEO’s power over the board. Two other papers directly
expand upon Murphy (2002). Blackwell, Dudney, and Farrell (2007) examine changes in
compensation structure following turnover events and relate those changes to firm performance.
Theyfind that incoming CEOs’ compensation is comprised of significantly more equity-based
pay and a positive association between post-turnover performance and new stock grants. Elsaid
and Davidson (2009) examine differences between incoming and outgoing CEO pay and the
percentages that salary and “pay-at-risk” (i.e. stock and option grants) contribute to total
compensation surrounding turnover events.
This paper complements and extends these papers along several lines. First, I examine a
much more comprehensive set of potential CEO compensation determinants and attempt to
differentiate between CEO, turnover, and firm characteristics that may contribute to managerial
power in a positive way and those that may contribute to some types of “inappropriate”
managerial power (i.e. power over the board based on relationships and/or entrenchment effects).
Second, a large part of the focus in both papers is on changes in the structures of compensation
packages. This paper examines compensation structure as well as changes in levels of pay and
relates them to relative managerial power. Finally, I seek to distinguish between competing
views of executive compensation within the literature, as discussed above. Thus, there is some
emphasis on managerial power and examining its impact on executive compensation.
The remainder of this paper is organized as follows: Section II describes the sample;
Section III contains the preliminary results; Section IV contains additional analysis regarding
what we happens to CEO compensation when a turnover event occurs; and Section V concludes.
II. Sample Selection and Summary Statistics
II.1 Sample Selection
The initial sample is collected from the Standard & Poor’s ExecuComp database, which
provides information on firms in the S&P 500, the Midcap 400, and the Smallcap 600, between
1993 and 2006. Data collected includes CEOs’ cash pay, total compensation including the value
of option grants and other forms of pay, tenure as CEO, and CEO percent equity ownership in
the firm. ExecuComp provides information regarding the years during which an executive
becomes the CEO and leaves office as CEO. Outgoing CEOs are identified by the year in which
Page 7 of 34
they leave office as CEO. Incoming CEOs are identified by the year in which they become
CEO. The sample of turnovers is then constructed by matching outgoing and incoming CEOs on
firm and year. The sample is then limited to those observations for which salary, bonus, and
total compensation data is available for the last full year of pay of the outgoing CEO and the first
full year of pay of the incoming CEO. This results in 1,232 turnover events over the 13 year
period.
II.1.1 Nature of Turnover
To identify the nature of the sample turnovers, I search the Factiva and LexisNexis
databases for news articles around the time period and announcements of CEO appointments.
Forced turnovers are identified by searching through news around the time period and
announcements of CEO appointments. A turnover is classified as forced if an article is found
describing an ouster of the outgoing CEO or a pressured resignation amidst any of the following:
federal or state investigations, shareholder lawsuits, or poor firm performance. Overall,
approximately 15% are forced turnovers, either as direct firings or pressured resignations. The
rest of the turnovers’ natures breakdown as follows: approximately 70% represent planned
retirements or normal successions, approximately 5% represent resignations that were not forced
(for various reasons, such as “to pursue personal interests”), 2.3% represent deaths or
incapacitating illnesses, 4% represent CEOs leaving for to take new positions, 2% represent
CEOs leaving due to a merger or spinoff. The remainder represent various other reasons with
very few cases, such as the firm deciding to separate the positions of CEO and chairman.
It is quite possible that within the 75% of turnovers listed above as being either planned
retirements, normal successions, or resignations not classified as forced, some have been mis-
classified. As discussed by Denis and Denis (1995) and previously noted by Warner, Watts, and
Wruck (1988) and Weisbach (1988), press releases often do not specifically state that a CEO has
been fired or ousted even when this is the case. Thus I further analyze the sample of forced
turnovers to determine whether they compare with those classified as voluntary in a way that
would be expected. First, prior literature has classified forced turnovers as a percentage of all
sample turnovers in percentages ranging from 7.3% to 24%. It is important to note here that this
rather significant variation stems from alternative specifications of forced turnovers. Here, the
classification of a forced turnover is relatively conservative and results in a sample for which I
am confident that the change is forced. This classification is consistent with Denis and Denis
Page 8 of 34
(1995). Others classify turnovers as forced on slightly looser criterion: Parrino (1997), Parrino,
Sias, and Starks (2003), and others classify additional turnovers as forced if the CEO is under 60
years of age and the news announcements regarding the turnover do not specify the cause as
being death, illness, or departure for a new position. This typically results in a slightly higher
sample representation for forced turnovers.
To further check my sample of forced turnovers, I compare firm performance,
prevalence of outsider replacements (specification of these is discussed below), and CEO age for
forced vs. voluntary turnovers to ensure that the differences are consistent with expectations. For
forced turnovers, we expect to observe significantly worse firm performance, more outsider
replacements, and younger outgoing CEOs. The differences are as expected: Average 3 years
stock price performance of firms involved in forced turnovers is approximately -5.3% while that
for firms involved in voluntary turnovers is approximately 9.6%; on average, outsiders are
appointed as CEO 44.7% of the time when the turnover is forced and approximately 30% of the
time when the turnover is voluntary; and the average age of forced-out CEOs is 56 years of age
in comparison to 63 years of age for voluntarily leaving CEOs. The differences between these
values for forced vs. voluntary turnovers are all significant at the .01 level.
II.1.2 Insider/Outsider Status
Data regarding the CEOs specifically are collected from Bloomberg People Search.
Bloomberg People Search provides profiles including educational history, career history, and
board memberships. CEOs are classified as insiders if they have a prior employment history
with the firm in which they become CEO. Specifically, where available, if the Bloomberg
People Search career history indicates employment with the firm in positions other than CEO
prior to the appointment, they are considered insiders. If the career history indicates that the
individual has worked for the firm for less than 3 years and that they came in at a high-level
executive rank, he is considered a “recent insider”. Otherwise, he is considered an outsider.
“Recent insiders” deserve a brief discussion here. Where possible, these are verified
through news announcements of the appointment. They represent cases in which the
replacement CEO was hired away from another firm with the likely outcome that he would take
over as CEO, but start off as COO or CFO. Since they were almost all hired specifically for the
purpose of fulfilling a normal succession plan and since, prior to their hire, they had no
employment through the firm, in the subsequent analysis, I combine the set of “recent insiders”
Page 9 of 34
with outsiders. Where Bloomberg People Search career histories are unavailable, Execucomp
data is used to determine insider/outsider status. This occurs in few cases, but represents
approximately 5% of the sample of incoming CEOs. In this case, CEOs are considered insiders
if 1) the first year of the CEO’s employment at the firm (according to ExecuComp) differs from
the year in which he becomes CEO and 2) the executive appears in the ExecuComp database for
that firm for the year prior to the year in which he becomes CEO.
II.2 Summary Statistics
II.2.1 Turnover Characteristics
Table 1 and Figure 1 present a time profile of the turnover events. With the exception of
1993, each year represents at least 5% of the sample of turnovers, with proportional
representation rising to over 10% in 2000. Overall, there were more turnovers during the
subperiod 2000-2005 than the earlier subperiod 1993-1999. The percentage of turnovers that
were forced in any given year is typically between 10% and 15%.
II.2.2 Compensation Characteristics
Table 2 presents summary statistics regarding the compensation of the CEOs in the
turnover sample. Overall, mean (median) CEO cash pay and total pay are, respectively, $1.36
million ($957,526) and $4.39 million ($2.24 million). Median CEO ownership of the firm for
outgoing CEOs is 1.4%, almost double the median for their replacements of .76%. At a glance,
the figures in Panels B and C do not appear, other than ownership, to be significantly different
from each other, including option grants. Figure 2 presents the percentage of total CEO
compensation that is option-based. As a form of equity-based pay whose use has increased
dramatically over the last several decades (both domestically and internationally), stock options
are a main component of incentive-based pay and a large contributor to pay-performance
sensitivities7
III. Baseline Results
. The data in Figure 2 show that incoming CEO option-based pay as a percentage of
total compensation is higher in every year.
III.1 Univariate Analysis: Differences in Pay
Table 3 presents the differences from paired CEOs surrounding the sample turnover
events. The outgoing CEO’s last full year of pay is subtracted from the incoming CEO’s first
7See Jensen and Murphy (1990); Yermack (1995); Hall and Liebman (1998); Hall (2003); Bebchuk and Grinstein (2005); Murphy (2003); Bryan, Nash, and Patel (2006); Bechman and Jorgensen (2004
Page 10 of 34
full year of pay. Univariate analysis is then run on the differences. The first column shows the
differences for the entire sample of turnovers. Overall, there does not appear to be a significant
difference between incoming and outgoing CEO pay levels. Both mean and median cash pay as
well as mean total pay are not significantly different.8
Consistent with Murphy (2002), the second two columns show that outsider replacement
CEOs, at the median, typically make $335,360 more than their predecessors while insiders are
typically paid only $126,156 more than their predecessors. Table 3 also shows that there is no
significant difference between incoming and outgoing CEO total compensation when the
turnovers are forced. However, we observe significantly higher total compensation for incoming
CEOs when the turnovers are voluntary. Again, this contradicts the managerial discretion
hypotheses, presuming that the poor performance resulting in a forced turnover has undermined
the outgoing CEO’s power over the board.
However, median total pay differences
indicate that replacements make a statistically significant $159,560 more in total compensation
than those whom they are replacing. Both option-based pay and its relative contribution to total
compensation are significantly higher for replacement CEOs than for those whom they are
replacing. Incoming CEOs are paid, at the median, $296,130 more in options than the CEOs
they are replacing. Furthermore, the percentage of total compensation that is option-based for
incoming CEOs is almost 8% higher than that for their predecessors.
The last two columns of Table 3 show differences between incoming and outgoing CEO
pay based on whether the outgoing CEO received “excess pay”. Using an approach similar to
Hartzell, Ofek, and Yermack (2004), I calculate “excess pay” as the residual from a basic
regression model. That is, using the entire universe of ExecuComp data, I run the following
regression:
CEO Pay = log(MktCap) + 3-year stock returns + Industry Dummies + Year Dummies.
CEO Pay represents total compensation including restricted stock, payouts from long-
term plans, benefits, and stock options valued at the grant-date using ExecuComp’s modified
Black-Scholes methodology. Log(MktCap) is the log of number of shares outstanding
8 Total pay and total compensation are terms used interchangeably throughout this analysis and represent the total compensation figures provided by ExecuComp including total cash pay, stock options, restricted stock, payouts from long-term plans, and benefits. Stock options are valued at the grant-date using ExecuComp’s modified Black-Scholes methodology.
Page 11 of 34
multiplied by the year-end stock price. The residuals for the sample of turnover CEOs are
retained and used as an estimate of “excess pay”.
Interestingly, Table 3 shows that when the outgoing CEO’s “excess pay” ≤ 0, his
replacement is much more consistently paid more than he was and vice versa. That is, when the
outgoing CEO’s “excess pay” ≤ 0, the median replacement CEO is paid $490,343.50 more than
his predecessor. When the outgoing CEO’s “excess pay” > 0, his replacement is paid $275,269
less than his predecessor. This suggests that “excess pay” may very well be a good proxy for
managerial bargaining power and also provides some support for the managerial power theory.
Alternatively, this could be a function of mean reversion in levels of compensation.
Core, Guay, and Larcker (2008) document “strong mean reversion in pay among the general
population of highly paid CEOs.” Examining non-turnover years relative to “excess pay” does
reveal some evidence of mean reversion. That is, approximately 67% of CEOs who receive
“excess pay” > 0 in a given year subsequently see a drop in their pay the following year. When a
CEO receives “excess pay” > 0, the median change in pay for the following year is $-158,408.
When a CEO receives “excess pay” ≤ 0, approximately 50% subsequently see a drop in the pay
the following year, but the median change in pay for the following year is $298,174.50. For the
entire sample of 15,996 CEO-year observations, I observe that approximately 36% receive
“excess pay” > 0. Of those observations, approximately 40% receive “excess pay” ≤ 0 the
following year while the remaining 60% continue to receive positive “excess pay”. So, it is clear
that, at best, “excess pay” can be a rough proxy for managerial power. Consequently, I do not
focus additional analysis on “excess pay”, but rather try to examine other potential proxies for
managerial power in subsequent sections.
III.2 Multivariate Analysis: Levels of Pay
Table 4 presents multivariate analysis of CEO compensation surrounding the turnover
event. The outgoing CEOs’ last full year of pay and the incoming CEOs’ first full year of pay
are used. The regression follows similar lines and uses determinants identified in several prior
studies including Core, Holthausen, and Larcker (1999), Brick, Palmon, and Wald (2006), and
Graham, Li, and Qiu (2008). The regressions specification is as follows:
sYearDummiemmiesIndustryDuExcessNewceoInsiderNewceoNewceoevceoNummtgs
InterlockTenureipCEOOwnershExecDirAssetsCEOPay
++++++
++++++=**Pr
)log(109876
543210
βββββββββββ
Page 12 of 34
Here log(assets) is the log of total assets, ExecDir is a dummy which equals 1 if the CEO
is on the board of directors. Ownership is the shares owned by the CEO. Tenure is the number
of years the CEO has been active in his current position. Interlock is a dummy which equals 1 if
the CEO is listed as a member of the executive compensation committee. Num Mtgs is the
number of board meetings during the year. Prevceo is a dummy variable which is equal to 1 if
the CEO has previously been a CEO at another firm. Newceo is a dummy variable equal to 1 if
the CEO is an incoming CEO, Insider is a dummy variable equal to 1 if the CEO is an incoming
CEO who came from within the firm, and Excess is a dummy variable equal to 1 if the outgoing
CEO was paid “excess pay” > 0. As we expect, the coefficient on log(Assets) is positive and
significant, consistent with the literature. Tenure has a positive and significant impact on cash
pay, but is insignificant in the total compensation regressions. Interlock is negative and
significant, indicating the CEO participation in the executive compensation committee actually
has a negative impact on both his cash pay and his total compensation. The coefficient on
Prevceo is positive and significant in the total compensation regressions, indicating that prior
experience at the level of CEO results in higher subsequent levels of pay. Results here, as above
in the univariate analysis, also indicate that there is little of significance in the difference
between replacement CEO pay and outgoing CEO pay. The coefficient on Newceo is significant
for only two of the cash pay regressions and not significant for total compensation. The
coefficient on Newceo*Insider is negative and significant for both cash pay and total
compensation. Thus, it appears that insider status has a negative impact on levels of pay.
III.3 Multivariate Analysis: Pay-Performance Sensitivities
Thus far, I have examined pay-performance sensitivities by comparing option-based pay,
as in Table 3, and found that they are consistently higher for replacement CEOs than for
outgoing CEOs. However, as pay-performance sensitivities broadly incorporate any adjustments
made to pay in response to firm performance, option-based pay is not the only way through
which pay-performance sensitivities can be impacted. Virtually all components (salary, bonus,
stock grants, and stock holdings in addition to option grants) of a CEO’s compensation can be at
least partially tied to firm performance. Furthermore, though compensation is potentially tied to
various accounting measures of firm performance, the ultimate goal of a compensation package
is to ensure that managers behave in the shareholders’ best interests. Thus, the most appropriate
Page 13 of 34
way to link pay to performance is often considered to tie compensation to stock price
performance. One way to examine pay-performance sensitivities is to quantify the impact that
changes in shareholder wealth have on changes in CEO pay from year to year. Jensen and
Murphy (1990) and a multitude of subsequent papers quantify this impact by regressing changes
in CEO pay on changes in shareholder wealth.
Table 5 presents multivariate analysis of pay-performance sensitivities measured in this
way surrounding the turnover event. Change in pay between the last two full years of pay for the
outgoing CEO and the first two full years of pay for the incoming CEO are used to perform the
following regression:
InsiderNewceorWealthShareholdeNewceoExcessrWealthShareholdeNewceorWealthShareholdeipCEOOwnershExecdirrWealthShareholdeCEOPay
*****
6
54
3210
∆+∆+∆+
++∆+=∆
βββ
ββββ
If anything, the overall evidence in Table 5 indicates pay-performance sensitivities
measured in this manner appear to be lower for replacement CEOs. The coefficients on
Newceo*ΔShareholder Wealth is negative and significant for changes in cash pay and not
significant for changes in total pay. However, insider replacement CEOs do appear to be
subjected to higher pay-performance sensitivities, as evidenced by the significant and positive
coefficient on Newceo*Insider*∆Shareholder Wealth.
IV. Additional Analysis
IV.1 A Closer Look at CEOs
The initial results described above show mixed results with regards to what happens to
CEO compensation when a turnover event occurs. We have seen that levels of total
compensation are higher for incoming CEOs overall, that the differential between incoming and
outgoing CEO pay is greater for outsider CEO replacements than for insider CEO replacements,
and that there is no significant difference between incoming and outgoing CEO pay when the
turnover is forced. These results are difficult to reconcile with the managerial discretion view of
CEO compensation. As discussed in previous sections, I expect an outgoing CEO to have more
managerial power than his replacement simply by virtue of his standing position and relationship
with the board of directors. Furthermore, insiders (who have served as other high level
executives of the firm and are also expected to have a relationship with the board of directors)
Page 14 of 34
are expected to have higher managerial power than outsiders and outgoing CEOs who are forced
out are expected to have lower managerial power than outgoing CEOs who leave voluntarily.
Thus, managerial discretion hypotheses predict the opposite result to those listed above.
However, the initial results do not address the possibility that observed differences in
CEO pay between outgoing and incoming CEOs are driven by CEO characteristics not yet
examined. That is, I posit that there are two sources of managerial power: 1) managerial skill,
ability, and effort (skill-based power) and 2) power over the board of directors (board-based
power). Both sources of managerial power are expected, under the managerial discretion view,
to cause increases in executive compensation. However, only the second results in inefficient
levels of pay or inefficient compensation structures. Thus far, the analysis has made no
distinction between the two sources of managerial power. It may be possible, however, that
managerial skill overwhelms managerial power over the board of directors in many cases. If this
is the case, then what we observe above reflects systematically higher quality replacement CEOs
(than the managers they replace), that outsider replacements are systematically higher quality
replacements than insider replacements, and that firms recruit higher quality CEOs when their
own CEOs leave voluntarily than when they fire their CEOs
In this section, I more closely examine CEO characteristics to see whether they can
explain what we observe in the preliminary analysis section. In an effort to determine which
contributor to CEO compensation (skill-based power or board-based power) seems to drive my
results, I examine characteristics expected to be tied to CEO skill alongside characteristics
expected to be tied to the CEO’s relative power within the organization (i.e. power over the
board). In particular, for skill-based power, I examine the impact of education and the
insider/outsider specification and prior firm performance where possible on the pay of incoming
and outgoing CEOs surrounding a turnover. For board-based power, I examine the impact of
founder status, tenure, high equity ownership, board composition, the GIM-index, and CEO
centrality.
These are all considered throughout the literature to be associated either directly with
managerial power (as in founder status, tenure, and CEO centrality) or indirectly via their impact
on the quality of governance mechanisms. Donaldson and Lorch (1983), Finkelstein (1992), and
Adams, Almeida, and Ferreira (2005) are just a few of the papers that discuss the increased
influence of ownership and founder status on the board of directors. Tenure as CEO
Page 15 of 34
undoubtedly is positively associated with a CEO’s influence over the board. Furthermore,
Gregory-Smith, Thompson, and Wright (2009) find that the likelihood of CEO departure drops
dramatically after a CEO’s fourth year in office and that this appears to be due to CEO
entrenchment.
Bebchuk, Cremers, and Peyer (2007) find that CEO centrality (measured as the
percentage of top five executive total pay received by the CEO9
The GIM index constructed by Gompers, Ishii, and Metrick (2003) has largely been used
to measure the quality of corporate governance. It is constructed by summing indicators for
various 24 charter provisions, bylaw provisions, and other firm-level rules associated with hostile
bidders, voting rights, director/officer protection, other takeover defenses, and state laws. A
higher index score represents greater managerial power and weaker shareholder rights.
) is negatively associated with
firm value and correlated with a broad set of firm behavior and performance including higher
“pay-for-luck” (either in the form of a “lucky” option grant at the lowest price of the month or in
a lower tendency for the firm to filter out industry and/or market-wide returns before rewarding
the CEO for performance) and lower likelihood of CEO turnover. Independent (or outside)
board members are expected to be farther removed from the CEO (and have a more tenuous
relationship with the CEO) and thus are expected to have a negative impact on managerial
power. However, Cyert, Kang, and Kumar (1997) find a positive association between CEO
compensation and percentage of outside directors on the board.
I also examine the compensation of both outgoing CEOs who obtain subsequent positions
as high level executives at other firms as well as outsider replacement CEOs who had recent
prior high level executive positions at other firms.
IV.1.1 Insider/Outsider and Founder Summary Statistics
Before examining the impact of certain CEO characteristics on CEO pay, I first examine
the composition of my set of CEOs and their characteristics. 25.4% of insider replacements have
an MBA, 4.3% have participated in an executive management program, 4.2% are CPAs, 10.3%
have master’s degrees, 4.4% have PhDs, and 7% have law degrees. On average, it appears that
the education levels are higher for outsider replacement CEOs. 27.9% of outsider replacements
have an MBA, 6.1% have participated in an executive management program, 4.1% are CPAs,
9 As in Bebchuk, Cremers, and Peyer (2007), CEO centrality is measured by aggregating the total compensation of the top 5 executives at each firm. CEO centrality is then calculated as the total compensation of the CEO divided by the aggregated total of the top 5 executives.
Page 16 of 34
16.8% have master’s degrees, 7.4% have PhDs, and 3% have law degrees. Outsider CEOs also
appear to have more board memberships, both at the time of the turnover (1.8 vs 1.5) and over
their lifetimes (4.8 vs 3.5). Where available, “excess pay” is determined for outsider CEOs at
their previous firms. Approximately 36% of these CEOs received “excess pay” > 0 at their
previous jobs. The average return on assets (ROA) at the previous companies the outsider CEOs
were employed at is 4.82%. On average, insider CEO replacements own a higher percentage of
the firm than outsider CEO replacements in the year following the appointment (1.2% vs. .83%),
but measure approximately the same CEO centrality as outsiders (~24%).
Founder status is identified via a combination of Bloomberg People Search, news
searches, online company histories, and company web-sites. Specifically, when Bloomberg
specifies that a CEO is a founder, this is used. If the CEO is not identified as a founder, the news
search regarding the turnover event is double-checked for any background information on the
individuals. This allows for a designation of both founder status and family status. If founder
status is still not found through either of these sources, I search for company histories online
using Google. If founder status is still not found, I go directly to the company’s web-site.
Regarding CEO characteristics of founder vs. non-founder outgoing CEOs, only 8.8% of
founders have MBAs, less than 1% are CPAs, 9.9% have Master’s degrees, 7% have earned a
PhD, and 4.1% have earned law degrees. On average, they hold 1.7 board seats at the time of the
turnover and 3.6 board seats over their careers. Average tenure for founders is 12 years, a
statistically significant (at the .01 level) 5 years greater than for non-founders. Founders also
own a significantly larger proportion of the firm in comparison to non-founders (3.98% vs. 1.02).
This difference is also significant at the .01 level. CEO centrality is also higher for founders, but
this difference is only significant at the .1 level and is only 1%. Only 25.7% of founders have
“excess pay” > 0. In contrast, among non-founders, 36.5% have “excess pay” > 0. Almost twice
as many non-founders have MBAs. More non-founders also hold master’s degrees, participated
in an executive management program, or are CPAs. Non-founder CEOs also hold more board
memberships both at the time of the turnover (2.3) and over their careers (4.9).
IV.1.2 Univariate Analysis: Skill-Based Managerial Power
IV.1.2.1 Education
Table 6 shows additional univariate analysis of differences in pay based on
insider/outsider status and education. Prior literature has established a potential shift in the
Page 17 of 34
relative desirability of general managerial skills and specific managerial skills. Murphy and
Zabojnik (2007) argue that such a trend has occurred, making the “skills critical in leading a
complex modern corporation but not specific to any organization” more important over recent
years than the “skills, knowledge, contacts, and experience valuable only within the
organization.” In support of their argument, they document a parallel between rises in CEO pay,
increases in CEOs holding MBA degrees, and decreases in managerial firm-specific experience.
Thus, these two categorizations are made to distinguish between skills expected to be useful for
all high-level management regardless of firm and those that firm- or industry-specific.
Educational background is used as a rough proxy for these types of skills. As in Murphy
and Zabojnik (2007) and Palia (2000), I consider whether the CEO has obtained an MBA or law
degree. I additionally consider whether the CEO has completed an executive management
program and whether he is a certified public accountant to be representative of general
managerial skills. Firm- or industry-specific skills are proxied for by more specific educational
attainment including professional certifications (such as Professional Engineer) and master’s
degrees and PhDs in specific fields (such as engineering or computer science.
To aggregate the educational information acquired through Bloomberg PeopleSearch, I
construct two variables: general_skills and specific_skills. These variables are the sums of
dummy variables to indicate whether the CEO has completed these various degrees and/or
programs. Specifically,
• General_skills = mba + cpa + lawyer + executive management program
• Specific_skills = masters degree + phd + certification
These variables are constructed for each of the CEOs. Then, the value for incoming
CEOs is subtracted from that for outgoing CEOs to determine the relative education levels of the
CEOs involved in the turnover event.10
Panel A of Table 6 shows the differences in pay levels between incoming and outgoing
CEOs on the basis of “general” and “specific” skills. The first two columns show differences on
the basis of whether the replacement CEO has a higher level of “general_skills” than the
outgoing CEO. Interestingly, though the differential is higher for replacement CEOs with a
10 The literature handles education in a multitude of ways, none of which are identical to this method. There is some support for examining education in this way, however. Gottesman and Morey (2006) assume that higher levels of education equate with higher quality educations (that is, CEOs with graduate degrees have had a higher quality education than those without)
Page 18 of 34
higher level of “general_skills”, even those replacement CEOs who have lower or equal levels of
“general_skills” are paid more (at the median) than those they replace. Also interesting is that
only around 12.8% of the replacement CEOs have higher general skills/education than those they
replace. Options granted are higher for replacement CEOs regardless of relative
“general_skills”, but option-based pay (as a percentage of total compensation) is only higher for
CEOs with “general_skills” lower than or equal to the outgoing CEO.
The other columns in Panel A of Table 6 show the differences based on “specific_skills”.
It is interesting here to note that when the replacement CEO has higher “specific_skills” than the
CEO he is replacing, there is no difference in pay between the two, though he is still paid more
than the outgoing CEO if his “specific_skills” are lower than or equal to the outgoing CEOs.
This counterintuitive result suggests that the parameters contributing to the “specific_skills”
variable do not drive compensation levels. Again, it is interesting to note that only
approximately 13% of the replacement CEOs have higher specific skills/education than those
they replace. Regarding option-based pay, we see a similar result to above regarding
“general_skills”. Overall, Panel A shows that neither differences in “general_skills” nor
differences in “specific_skills” can explain the overall positive differential between incoming
and outgoing CEO pay. That is, Table 6 shows not only that is it relatively rare for the incoming
CEO to possess higher “general” or “specific” skills than his predecessor, but that even when he
does not, he is still paid more than his predecessor. Next, I examine the impact of managerial
skill on the differentials we observe when replacement CEOs are insiders vs when they are
outsiders.
Though it appears to be inconsistent with the managerial discretion view of CEO
compensation that outsider replacement CEOs would be paid more than the CEOs they replace
and that the differential would be greater for outsider replacements than for insider replacements,
other forces could be driving this result. In particular, it could be the fact that outsider
replacements are systematically higher quality CEOs (than both those they replace and potential
insider candidates). If this is the case, we would expect to see some different results regarding
CEO characteristics and insider/outsider status. For instance, if general managerial skills are
highly valued and this drives the result regarding outsider replacements, then we would expect to
see both that outsider replacements have higher general managerial skills than those whom they
replace and that the differential results we have seen above hold only for the group of outsider
Page 19 of 34
replacements who have higher general managerial skills. We would also expect to see that there
are relatively less insider replacements with higher general managerial skills than outsider
replacements.
Panel B of Table 6 shows the differences in pay levels on the basis of insider/outsider
status and education levels. Interestingly, outsiders are paid more than their predecessors
regardless of whether their “general_skills” are higher than those of their predecessors, though
the differential is larger when the incoming CEO has a more comprehensive general education.
Furthermore, only 62 of the 332 outsider replacements have higher general skills/education than
those they replace. This represents approximately 15% of the outsider CEOs, slightly higher
than the 12.8% found overall for replacement CEOs, but is still remarkably low. One interesting
and counterintuitive result is that insider replacements are paid more than their predecessors only
if their “general_skills” or “specific_skills” are less than those of their predecessors. This seems
to indicate that skills (as proxied by education) does not drive pay differentials between
incoming and outgoing CEOs when the replacement is an insider.
IV.1.2.2 Outsider Replacements’ Prior Performance
In contrast to proxying for skills using relative education, an alternative is to examine the
prior performance of the outsider replacement CEOs. Though this allows for a limited analysis
of a much smaller subset of the turnovers, it may be able to shed light on whether, at least for
these outsider replacements, our differences in pay between the replacements and their
predecessors are driven by skill. I first obtain information about outsider replacement CEOs
prior compensation and compare that to the compensation they receive one year after the
turnover event. The prior compensation data is obtained through Execucomp. There are
approximately 160 outsider replacement CEOs for whom this data is available. On average,
these incoming CEOs receive a pay increase of $860,954 in total cash compensation (salary +
bonus) and a raise of $2,203,280 in total compensation. Approximately 37% of them received
“excess pay” > 0 at their previous jobs.
Panel A of Table 7 shows differences for just these outsider CEO replacements based on
the performance of the firm at which they were previously employed and based on whether they
received “excess pay” at the firm at which they were previously employed. On average, the
Page 20 of 34
ROA of these firms (at which outsider CEO replacements were previously employed) is 4.8%. I
thus use 4.8% as the threshold for above average performance.11
The first columns show that when the outsider replacements come from firms at which
the ROA > 4.8%, they are paid substantially more than the CEOs they are replacing, but when
their previous employer’s ROA < 4.8%, only cash compensation and option-based pay are higher
than the CEOs they are replacing, while total compensation is not significantly different. Thus, it
appears that prior firm performance for outsider replacements does indeed explain part of the
differential. However, the results regarding the CEOs whose prior performance was average or
worse are still inconsistent with the managerial power hypotheses. These CEOs still get paid as
much as their predecessors who should have more managerial power over the board and thus be
paid more. Furthermore, their cash pay is still higher than their predecessors’. Since a CEO’s
preference is, wherever possible, going to be certain cash pay over uncertain equity-based pay,
managerial power hypotheses predict that, as power increases, cash pay will increase and pay-
performance sensitivities will decline. Thus, under managerial power, we would expect this
subset to be paid less cash pay and more option-based pay than his predecessor. While option-
based pay is, indeed, higher for the replacements in this scenario, the differential is not
substantially different from the replacements whose prior firm performance is above average (at
the median, option-based pay for those whose prior firm performance is average or worse is
11.9% higher than their predecessors vs. 11% for those whose prior firm performance is above
average).
The latter columns show that when the outsider replacements did not receive “excess
pay” at their previous firm, they are paid significantly more than the CEOs whom they replace,
but when they did receive “excess pay” previously, only their cash compensation is higher than
the CEOs they replace. This is a somewhat puzzling result, particularly for the managerial
discretion hypotheses. Bebchuk, Fried, and Walker (2004) argue that outsider replacements who
are previously CEOs at other firms have already been exercising their discretion at these other
firms to extract rents and pay packages reflect this. They then further argue that, in order to
entice these executives to change employers, the rents extracted at the prior firm must be met by
the new firm. The expectation, then, is that those outsider CEO replacements who received
“excess pay” at their prior firms would be paid more than those who did not. This would then
11 However, the results listed below hold if industry-adjusted ROA is used instead with a threshold of 0%
Page 21 of 34
imply that differentials between incoming and outgoing CEO pay for these CEOs would be
greater than for those not receiving “excess pay” at their prior job. What we observe is the
opposite.
The final univariate check with regard to CEO skill-based power that I perform reflects
univariate differences in pay based on whether the outgoing CEO achieves subsequent
employment as a CEO elsewhere. This set may reflect outgoing CEOs with above average
abilities and thus may be able to shed some light on what we observe with regard to changes in
CEO pay surrounding turnover events. I first gather information about the compensation of
outgoing CEOs who subsequently obtain employment at other firms as CEOs. The subsequent
employment data is gathered through Execucomp. There are 80 CEOs for whom I find definite
new employment as a CEO within 3 years of the turnover event. Of these, subsequent
compensation data is available for half (40 outgoing CEOs). On average, the differences
between the total cash compensation and total compensation for the outgoing CEOs at the new
firm and at the turnover firm are $152,559.50 and $5,184,886. Approximately 35% of these
outgoing CEOs received “excess pay” > 0 at the turnover firms. Approximately 55% of them
receive “excess pay” > 0 at the subsequent employers. Panel B of Table 7 shows differences
between incoming and outgoing CEO pay surrounding the turnover event based on whether the
CEOs obtain subsequent employment. When the outgoing CEO does find subsequent
employment within 3 years of the turnover, there is no significant difference between his cash or
total compensation and that of his replacement. However, when he does not find subsequent
employment, the results more closely mirror our overall results (higher total compensation and
option-based compensation for the incoming CEO). We would have expected the opposite
results. Thus it appears that subsequent employment as a proxy for outgoing CEO skill also
cannot explain our overall results.
IV.1.3 Univariate Analysis: Board-Based Managerial Power
We have seen that the preliminary results regarding differences in pay can, at best, only
be partially explained by our proxies for CEO skill. Next we turn to whether they can be
explained by proxies for managerial power. For this analysis, we examine the impact of board
composition, founder status, CEO centrality, CEO tenure, CEO firm ownership, and GIM-index
on compensation surrounding the turnover events.
Page 22 of 34
Panel A of Table 8 shows differences between incoming and outgoing CEO pay when the
outgoing CEO is the founder and when he is not the founder. We would expect founders to have
more power over their boards. However, while in both categories outgoing CEOs are paid less
than their replacements, but the differential is larger when the outgoing CEO is the founder
rather than when he is not the founder. Whatever board-based power a CEO may derive via his
status as founder of the firm does not drive our results regarding overall differences in pay.
Thus, we again see results in direct opposition to what we would expect under the managerial
discretion view.
Another proxy for board-based managerial power is CEO centrality, as discussed above.
Panel A of Table 8 also shows univariate differences results based on whether CEO centrality is
above or below average. When CEO centrality is average or below average, the incoming CEO
makes significantly more in total cash pay, option-based pay, and total compensation. On the
other hand, when CEO centrality is above average, the outgoing CEO makes more in total cash
pay and slightly less option-based pay. Total compensation does not significantly differ between
incoming and outgoing CEOs within this set of turnovers. These results move in the direction
that the managerial discretion view would expect. However, only cash pay results match
predictions that would stem from the managerial discretion view. That is, here managerial
discretion would predict higher cash and total pay for the outgoing CEO based on his enhanced
power over the board (as reflected by his high value for CEO centrality). Furthermore, since any
manager will prefer riskless pay to risky pay, all else equal, we would expect the option-based
pay differential here to increase. That is, the managerial discretion view would also predict that
the incumbent CEO would exercise his power over the board to reduce his exposure to risk,
increasing any non-performance based pay and decreasing the performance-based pay. This
gives ambiguous predictions on levels of options granted, but not on option-based pay as a
percentage of total pay. But what we observe is that, while for average to low CEO centrality,
the difference in option-based pay is 9.5%, for above average CEO centrality, it is a mere .3%.
If founder status and CEO centrality cannot explain our results, perhaps a simple look at
tenure or ownership can. Table 8 displays univariate differences results for tenure and above and
below average levels of CEO ownership. Tenure-based results are based on whether or not CEO
tenure is greater than 4 years. I use 4 years as a cutoff based on Gregory-Smith, Thompson, and
Wright (2009), who find evidence of entrenchment effects and lower probabilities of forced
Page 23 of 34
turnovers after the fourth year in office. When CEO tenure is 4 years or less at the time of the
turnover, his replacement is paid more total cash pay and more options and option-based pay.
His total compensation does not significantly differ from that of his replacement. On the other
side, when tenure is greater than 4 years (and he is presumably entrenched), his replacement
makes slightly less in cash pay, more in options and option-based pay, and significantly more in
total compensation. If this enhanced power over the board results in managerial manipulation of
pay levels and structure, we would expect to see that the outgoing CEOs whose tenures were
longer than 4 years would be paid more than their replacements.
We observe similar results regarding CEO ownership in the firm. Using average
outgoing CEO firm ownership as a threshold, table 8 shows that incoming CEOs are paid more
than their predecessors regardless of whether the outgoing CEO has below average or average
levels of firm ownership or whether he has above average levels of firm ownership. The
differential, however, increases for the group for which CEO firm ownership is above average.
That is, if CEOs own 1.4% of the firm or less, their replacements make a statistically significant
$128,210 more than they do while if CEOs own more than 1.4% of the firm, their replacements
make a statistically significant $343,480 more. This, too, is inconsistent with managerial
discretion hypotheses. Presuming that higher levels of ownership result in higher managerial
power, we would expect to observe the opposite results under managerial discretion.
Admittedly, ownership is a slightly more complicated issue since it is also shown through
the literature to effectively align managers’ incentives with those of shareholders in a manner
similar to what compensation packages attempt to do. Thus, it is quite possible that
compensation levels are lower for CEOs with high levels of ownership in the firm because the
need to align their incentives with those of shareholders is reduced by their ownership levels.
Morck, Shleifer, and Vishny (1988), Hermalin and Weisbach (1987), and McConnell and
Servaes (1990) demonstrate a non-monotonic relationship between firm value and insider
ownership that is believed to reflect the trade-off between the incentives and entrenchment
effects of ownership. They all show that concentrated ownership first enhances firm value, but
then, at differing thresholds, starts to detract from firm value. It is important to note, however,
that none of this detracts from the expectations that higher levels of CEO firm ownership result
in higher managerial power over the board. However, in untabulated results, the analysis is run
Page 24 of 34
with higher thresholds (5% and 10% firm ownership).12
It could be that differences in pay are driven by firm-level governance characteristics
rather than CEO governance characteristics. To explore this possibility, I examine the impact of
the firms’ governance indices and board compositions. The governance index first introduced by
Gompers, Ishii, and Metrick (2003), commonly referred to as simply the Gindex, is acquired
through the RiskMetrics Governance database, which provides data regarding various
governance characteristics of firms, including the constructed gindex. Board size and
composition (measured as the percentage of directors who are independent) are obtained through
the RiskMetrics Directors database, which provides detailed data regarding board provides
detailed data regarding board members of a large number of firms.
The results from above hold and, in
fact, are even strong, reflected an even larger positive differential between replacement CEO pay
and outgoing CEO pay when outgoing CEO firm ownership is greater than 5% or 10%.
Panel B of Table 8 shows univariate differences results based on these two governance
proxies as well as based on financial distress (as discussed below). G-index results are based on
whether or not firm G-index is in the 75th percentile or higher (this equates to a G-index of 11 or
greater). When the G-index is below the 75th percentile, incoming CEO total compensation and
option-based compensation are both significantly higher than for the incumbent CEO. In
contrast, when the G-index is at or above the 75th percentile, only options and option-based pay
are higher for the incoming CEO. Neither cash nor total compensation are significantly
different. These results are similar to those regarding CEO centrality. That is, the differences in
pay between incoming and outgoing CEOs moves in the direction the managerial discretion
theories would predict based on increased incumbent CEO power over the board. But they do
not reach the predicted difference being significant and positive on the outgoing CEO side.
Board composition results are based on the percentage of independent outsiders on the
board of directors being above or below average. Here, we see that in both categories, total cash
and total compensation are both higher for the incoming CEO. However, the differential is
greater and significant when the percentage of independent directors is below average. So, again
here, we see that managerial power cannot be explaining our differentials. That is, an incumbent
12 Morck et al (1988) find that firm value rises as insider ownership increases up to 5%, then declines until insider ownership reaches 25%, then rises again. Thus, 5% appears to be one appropriate threshold to use. Further, 5% allows for the subsample for high CEO ownership to allow for meaningful statistical analysis (there are 117 outgoing CEOs whose ownership levels are above 5% and 64 CEOs whose ownership levels are above 10%).
Page 25 of 34
CEO can be expected to have more power over his board in firms where there are less
independent board members. If this is the case, then we would expect the outgoing CEO in these
firms to be paid more than his successor, but not necessarily so in firms where the board
composition represents a relatively higher proportion of independent directors. Table 8 shows
that this does not appear to be the case. Specifically, it shows differences in pay based on the
percentage of the board that is comprised of independent directors. When the board composition
indicates lower percentages of independent directors, the outgoing CEO’s total compensation is
significantly less than that of his replacement. When the board composition indicates higher
percentages of independent directors, the outgoing CEO total cash compensation and total
compensation are not significantly different from those of his replacement.
The final univariate check I perform related to board-based managerial power tests
univariate differences in pay on the basis of financial distress. Following Asquith, Gertner, and
Scharfstein (1994), I classify firms as being in financial distress based on interest coverage
ratios. A firm is in financial distress in the year of the turnover if its earnings before interest,
taxes, depreciation, and amortization (EBITDA) is less than 80 percent of its interest expense
(i.e. if the interest coverage ratio is less than .8). The average interest coverage ratio for all firms
surrounding the turnover event is 5.9. Approximately 15% of the firms are in financial distress
using this classification. In later sections, two alternative measures of financial distress are also
used: Altman’s Z scores and a measure based on firm performance and leverage ratio. The
average Altman’s Z-score13
Here, a similar argument can be made to the one for forced turnovers. If the firm is in
financial distress, it is likely that the manager is at a relative bargaining disadvantage (to
is 3.6 and the percentage of firms in danger of bankruptcy (with a Z-
score less than 1.8) is approximately 23%. As the final alternative measure, I use stockholder
returns, return on assets (ROA), and leverage ratios similar to Opler and Titman (1994). This
measure classifies firms with leverage ratios in deciles 8-10 for all firms’ leverage ratios and
negative stock holder returns and ROA as financially distressed firm. Using this measure,
approximately 4.5% of firms are in financial distress.
13 Calculated as Z-Score = A x 3.3 + B x 0.99 + C x 0.6 + D x 1.2 + E x 1.4 A = EBIT/Total Assets B = Net Sales/Total Assets C = Market Value of Equity/Total Liabilities D = Working Capital/Total Assets E = Retained Earnings/Total Assets
Page 26 of 34
managers of firms that are not in financial distress) due to the poor performance that
accompanies financial distress. Furthermore, the financial distress may be constraining firm
expenditures and have an additional restraining effect on CEO compensation. Thus, we expect
the compensation of incumbent CEOs of financially distressed firms to be lower than those of
non-financially distressed firms. We would further expect that the relative bargaining
disadvantage that financial distress should mean that CEOs of non-financially distressed firms
would be paid more than their replacements, but that this would not necessarily be the case for
CEOs of financially distressed firms. Panel B of Table 8 shows univariate differences on the
basis of financial distress (as indicated by the interest coverage ratio threshold discussed
above14
IV.2 Multivariate Analysis: Education, Insider/Outsider, Founder
). What we observe is similar to what we observe for voluntary vs. forced turnovers.
When we expect the relative managerial power to be lower (i.e. when turnovers or forced or
when firms are in financial distress), there is no significant difference between cash or total
compensation while when we expect relative managerial power to be higher, the incoming CEO
makes significantly more than his predecessor.
The final analysis performed is multivariate analysis of levels of CEO pay and
differences between incoming and outgoing CEO pay. I first examine CEO compensation levels
surrounding the turnover event incorporating CEO skill and board-based power characteristics.
The outgoing CEOs’ last full year of pay and the incoming CEOs’ first full year of pay are used.
The regressions are variations on the following:
sYearDummiemmiesIndustryDuprevExcessprevceoprevROAprevceorshipsBoardMembershipsBoardMembeistressFinancialD
sitionBoardCompoFounderillsSpecificSkllsGeneralSkiInsiderNewceoNewceoprevceoNummtgsInterlockTenureipCEOOwnershExecDirAssetsCEOPay
ct
+++++++
+++++++++
++++=
**
*)log(
1918
171615
14131110
98765
43210
βββββ
βββββββββ
βββββ
Here, the specifications for the first 10 variables are as listed above in the Baseline
Results section. General skills are the constructed variable from above based on general
educational achievements, specific skills are the constructed variable from above based on
specific educational achievements, board composition is the percentage of board members who
are independent, financial distress is a dummy for whether the firm is in financial distress as 14 Results are robust to alternative measures of financial distress as discussed above (Altman’s Z-Score and Opler and Titman’s measure of financial distress).
Page 27 of 34
defined above, board membershipst is the number of board seats held at the time of the turnover,
board membershipsc is the number of board seats held over the CEO’s career, PrevROA is the
ROA for the firm at which the incoming CEO was previously employed (if he is an outsider),
and PrevExcess is a dummy for whether the incoming CEO received “excess pay” from his
previous employer.
In untabulated results, there is very little change in coefficients or interpretation from the
previous multivariate analysis. The one major difference is that the almost all coefficients of
Newceo and Newceo*insider become insignificant. The adjusted R2 for the new specifications
are higher than in the simpler specification, but do not represent a dramatic increase in
explanatory power (typical increase is around .5%). The analysis does find positive and
significant coefficients for both general and specific education. It also finds that founders
receive less cash and total compensation. Firms that are in financial distress pay their CEOs less
cash. Board seats held at the time of the turnover does not appear to impact compensation
around this time, but board seats held over the course of a career do positively impact total
compensation. Firm performance for previous CEOs at their prior firms has a positive and
significant coefficient for total compensation, but whether they received “excess pay” does not.
Finally, to confirm univariate results regarding differences in CEO pay, I regress the
incoming CEO’s compensation minus the outgoing CEO’s compensation on my proxies for
managerial power and skill. The specification is the following:
Here, cutoffs for the independent variables are defined as above in the univariate results.
In untabulated results, I find only two significant coefficients for this regression and a very low
adjusted R2 of .0155, indicating (as the univariate results suggest) that these proxies contribute
very little to differences in compensation when a turnover occurs. The two exceptions appear to
be that, financial distress and high CEO centrality for the outgoing CEO both have a positive
impact on outgoing CEO pay relative to that of his replacement. These results, as well as the
non-significance of the remaining coefficients, are consistent with my univariate findings.
V. Conclusion
Page 28 of 34
In this study, I have empirically examined what happens to CEO compensation when
turnover occurs. Because turnover events involve the negotiation of a new contract for the newly
appointed CEO, they inherently provide opportunity for modification of compensation. Thus,
we may expect to find some dramatic changes in pay levels and/or structures surrounding a
turnover event. Turnovers can also be further instructive to the compensation debate because
they involve CEOs whose expected power over the board differ considerably. Thus, in this
study, I have examined changes in CEO compensation levels and structures to determine whether
or not boards of directors capitalize on the opportunity provided by a turnover event to change
the ways in which they pay their CEO. I further undertake to determine whether firm
characteristics or CEO characteristics have an impact on what is observed regarding
compensation surrounding turnover events.
Overall, I observe that option-based pay and total compensation increase from outgoing
CEO to incoming CEO. I further observe that the positive differential between replacement and
incumbent pay is larger for outsider replacements and when the turnover is voluntary (rather than
forced). Incorporating various measures of CEO skill/abilities and potential sources of influence
over the board, I find that these overall differences in pay between incoming and outgoing CEOs
largely cannot be explained by either of these factors.
To determine whether CEO skill can explain my overall results, such that the incoming
CEO receives more than his predecessor more so when measures of skill are higher, I examine
CEO education and outsider CEO prior firm performance and “excess pay”. I find that, whether
or not the incoming CEO has achieved a higher level of education relative to his predecessor, he
is still paid the same as or more than his predecessor. The same holds true for outsider
replacements regardless of prior firm performance or prior “excess pay”. I expect that all of
prior firm performance, “excess pay”, and relative education levels indicate higher levels of skill
for the incoming CEO. Thus skill cannot explain our overall results.
To determine whether an outgoing CEO’s power over the board can help to explain my
results, such that the outgoing CEO’s compensation is less than that of his replacement only
when his power over the board appears to be weak, I examine firm and CEO characteristics that I
expect to impact a CEO’s power over the board. Specifically, I propose that an outgoing CEO’s
power over the board is weakened when the turnover is forced, when the firm is in financial
distress, when the CEO is not the founder or a co-founder of the firm, when CEO centrality is
Page 29 of 34
below average, when his “excess pay” ≤ 0, when the outgoing CEO has been at the firm for 4
years or less, when his firm ownership levels are low, and when the firm’s Gindex is low. I find
that replacement CEOs are paid more total compensation than their predecessors regardless of
whether the turnover is forced or voluntary, but the difference is only statistically significant
when the turnover is voluntary. Replacement CEOs are also paid more in option-based pay
regardless of the nature of the turnover and regardless of financial distress. Cash and total
compensation differences between incoming and outgoing CEOs are not significantly different
when firms are in financial distress.
When the outgoing CEO is the founder or a co-founder of the firm, I find that his
replacement also makes more than he did and that these differences are greater within this group
of outgoing CEOs than within the turnovers involving non-founder CEOs. Similar results hold
regarding tenure. That is, regardless of a tenure of 4 years or less or a longer tenure, the
incoming CEO receives more than his predecessor, but the difference is greater and more
significant when the outgoing CEO’s tenure is longer than 4 years. When the outgoing CEO has
relatively higher levels of firm ownership, this differential also lies on the incoming CEO side
and is larger than when his ownership levels are lower. All of these results contradict the notion
that managerial power over the board can explain observed differences in CEO pay in a way
predicted by the managerial discretion view of CEO compensation. Though results regarding
CEO centrality and the Gindex move more in the direction expected by the managerial discretion
view, they do not come inline with direct predictions. That is, hypotheses predict a clear positive
differential between outgoing and incoming CEOs within those for whom outgoing CEO
centrality is high or the firm’s Gindex is high. However, the results simply show no statistically
significant difference between the two CEOs’ compensation within these firms. Taken
altogether, the analysis suggests that managerial power is not driving observed changes in
compensation, but provide limited opportunity for explaining what is the driving force.
Page 30 of 34
References:
Abowd, John M. and David S. Kaplan, 1999, Executive Compensation: Six Questions that Need Answering. Journal of Economic Perspectives 13, 145-168. Acharya, V. V., K. John, and R. K. Sundaram, 2000, On the optimality of resetting executive stock options, Journal of Financial Economics, 47, 65-101. Agarwal, Naresh C., 1981, Determinants of Executive Compensation, Industrial Relations, 20, 1, 36-46. Aggarwal, R. and A. Samwick, 1999, The other side of the tradeoff: The impact of risk on executive compensation, Journal of Political Economy 107, 65-105. Almazan, Andres, Jay C. Hartzell, and Laura T. Starks, 2005, Active Institutional Shareholders and Costs of Monitoring: Evidence from Executive Compensation, Financial Management 34, 5-34. Asquith, Paul, Robert Gertner, and David Scharfstein, 1994, Anatomy of financial distress: An examination of junk-bond issuers, Quarterly Journal of Economics, August, 625-258. Bainbridge, Stephen M., 2005, Executive Compensation: Who Decides?, Texas Law Review 83, 1615-1662. Bebchuk, Lucian Arye, Cremers, Martijn and Peyer, Urs C., 2007, "Pay Distribution in the Top Executive Team". Harvard Law and Economics Discussion Paper No. 574 Available at SSRN: http://ssrn.com/abstract=954609 Bebchuk, Lucian A., Jesse M. Fried, and David I. Walker, 2002, Managerial Power and Rent Extraction in the Design of Executive Compensation, University of Chicago Law Review 69, 751-846. Bebchuk, Lucian A. and Jesse M. Fried, 2004, Pay Without Performance: The Unfulfilled Promise of Executive Compensation, Cambridge: Harvard University Press, Pp 304. Bertrand, Marianne, and Sendhi Mullainathan, 2001, Are CEOs rewarded for luck? The ones without principals are. Quarterly Journal of Economics 116, 901-931. Black, David W., Donna M. Dudney, and Kathleen Farrell, 2007, Changes in CEO compensation structure and the impact of firm performance following CEO turnover, Review of Quantitative Financial Accounting, 29, 315-338. Brick, Ivan E., Oded Palmon, and John K. Wald, 2006, CEO Compensation, director compensation, and firm performance: Evidence of cronyism? Journal of Corporate Finance 12 402-423.
Page 31 of 34
Brickley, J. A., J. L. Coles, and J. Linck, 1999, What happens to CEOs after they retire? Evidence on career concerns and CEO incentives, Journal of Financial Economics, 52, 341-377. Chalmers, Keryn, Ping-Sheng Koh, and Geof Stapledon, 2006, The determinants of CEO compensation: Rent extraction or labour demand, Working Paper. Chance, D. M., R. Kumar, and R. B. Todd, 2000, The ‘repricing’ of executive stock options, Journal of Financial Economics, 57, 129-154. Coles, Jeffrey L., Naveen D. Daniel, and Lalitha Naveen, 2005, Option repricing and incentive realignment, Working Paper. Coles, Jeffrey L., Naveen D. Daniel, and Lalitha Naveen, 2006, Managerial incentives and risk taking, Journal of Financial Economics 79, 431-468. Core, John E., R. W. Holthausen, and D.F. Larcker, 1999, Corporate Governance, Chief Executive Officer Compensation, and Firm Performance, Journal of Financial Economics, 51, 371-406. Core, John E., Wayne R. Guay, and Randall S. Thomas, 2005, Is U.S. CEO Compensation Inefficient Pay Without Performance?, Michigan Law Review 103, 1142-1185. Coughlan, A. T., and R. M. Schmidt, 1985, Executive compensation, management turnover, and firm performance: An empirical investigation, Journal of Accounting and Economics, 7, 43-66. Dahya, Jay, John J. McConnell, and Nickolaos G. Travlos, 2002, The Cadbufy Committee, Corporate Performance, and Top Management Turnover, Journal of Finance, 57, 461-483. Denis, D. J., and D. K. Denis, 1995, Performance changes following top management dismissals, Journal of Finance, 50, 1029-1057. Denis, D. J., D. K. Denis, and A. Sarin, 1997, Ownership structure and top executive turnover. Journal of Financial Economics, 45, 193-221. Denis, D., Hanouna, P. & Sarin, A. (2006). Is There a Dark Side to Incentive Compensation?, Journal of Corporate Finance, (12), 467-488. Dow, Janes, and Clara C. Raposo, 2005, CEO compensation, change, and corporate strategy. Journal of Finance, 60, 2701-2727. Elsaid, E. and W. Davidson, 2008, What happens to CEO compensation following turnover and succession? Quarterly Review of Economics and Finance, in press. Garvey, Gerald, and Todd Milbourn, 2003, Incentive Compensation when executives can hedge the market: Evidence of RPE in the cross-section. Journal of Finance 58, 1557-1581.
Page 32 of 34
Garvey, Gerald, and Todd Milbourn, 2004, Asymmetric benchmarking in compensation: Executives are rewarded for good luck, but not penalized for bad. Journal of Financial Economics Gaver, J. and K. Gaver, 1993, Additional evidence on the association between the investment opportunity set and corporate financing, dividend and compensation policies, Journal of Accounting and Economics, 16, 125-160. Gibbons, R., and K. J. Murphy, 1990, Relative performance evaluation for chief executive officers, Industrial and Labor Relations Review, 43, 30-51. Gibbons, R., and K. J. Murphy, 1992, Optimal incentive contracts in the presence of career concerns: theory and evidence, Journal of Political Economy 100, 468-505. Gilson, Stuart C. and Michael R. Vetsuypens, 1993, CEO Compensation in Financially Distressed Firms: An Empirical Analysis, Journal of Finance 48, 425-458. Gompers, Paul A., Joy L. Ishii, and Andrew Metrick, 2003, Corporate governance and equity prices, Quarterly Journal of Economics 118, 107-155. Goyal, Vidhan K., and Chul W. Park, 2002. “Board leadership structure and chief executive turnover”. Journal of Corporate Finance 8, 49-66. Gregory-Smith, Ian, Steve Thompson, and Peter W. Wright, 2009, Fired or retired? A competing risks analysis of chief executive turnover, The Economic Journal, 119, 463-481. Hambrick, D.C., and S. Finkelstein, 1995, The effects of ownership structure on conditions at the top: The case of CEO pay raises, Strategic Management Journal, 16, 175-193. Hartzell, Jay C., Eli Ofek, and David Yermack, 2004, What’s in it for me? CEOs whose firms are acquired. Working Paper. Hermalin, B., and M. Weisbach, 2000, Boards of directors as an endogenously determined institution: A survey of the economic literature. University of California at Berkeley and University of Illinois, Working Paper. Himmelberg, Charles, and R. Glenn Hubbard, 2000, Incentive pay and market for CEOs: An analysis of pay-for-performance sensitivity. Working paper, Columbia University Himmelberg, Charles, R. Glenn Hubbard, and D. Palia, 1999, Understanding the determinants of managerial ownership and the link between ownership and performance, Journal of Financial Economics, 53, 353-384. Huson, Mark R., Robert Parrino, and Laura T. Starks, 2001, Internal Monitoring Mechanisms and CEO Turnover: A Long-Term Perspective. Journal of Finance Vol. LVI, No. 6.
Page 33 of 34
Huson, Mark R., P. H. Malatesta, and Robert Parrino, 2004, Managerial succession and firm performance, Journal of financial Economics, 74, 237-275. Hwang, Byoung-Hyoun and Seoyoung Kim, 2009, It pays to have friends, Journal of Financial Economics, in press Jensen, Michael, and Kevin J. Murphy, 1990, Performance pay and top-management incentives. Journal of Political Economy 98, 225-264. Kaplan, Steven N., 1994, Top executive rewards and firm performance: a comparison of Japan and the United States, Journal of Political Economy, 102, 3, 510-546. Kaplan, Steven N. and Minton, Bernadette A., 2006, "How has CEO Turnover Changed? Increasingly Performance Sensitive Boards and Increasingly Uneasy CEOs" NBER Working Paper No. W12465 Available at SSRN: http://ssrn.com/abstract=924751 Mehran, H., 1995, Executive compensation structure, ownership, and firm performance, Journal of Financial Economics, 38, 163-184. Murphy, Kevin J., 2002, Explaining Executive Compensation: Managerial Power vs. the Perceived Cost of Stock Options, University of Chicago Law Review 69, 847-870. Murphy, Kevin J. and Jan Zabonjnik, 2004, CEO Pay and Appointments: A Market-Based Explanation for Recent Trends, The American Economic Review 94, 192-196. Murphy, Kevin J. and Zabojnik, Jan, 2007, "Managerial Capital and the Market for CEOs". Available at SSRN: http://ssrn.com/abstract=984376 Murphy, Kevin J., and Jerold L. Zimmerman, 1993, Financial performance surrounding CEO turnover, Journal of Accounting and Economics, 16, 273-315. Opler, Tim C., and Sheridan Titman, 1994, Financial Distress and Corporate Performance, Journal of Finance 49, 1015-1040. Oyer, Paul, 2004, Why do firms use incentives that have no incentive effects? Journal of Finance 59, 1619-1649. Palia, D., 2001, The endogeneity of managerial compensation in firm valuation: a solution, Review of Financial Studies 14, 735-764. Parrino, R., 1997, CEO turnover and outside succession: A cross-sectional analysis, Journal of Financial Economics, 46, 165-197. Perry, T., and M. Zenner, 2001, Pay for performance? Government regulation and the structure of compensation contract, Journal of Financial Economics, 62(3), 453-488.
Page 34 of 34
Rajgopal, Shivaram, Terry Shevlin, and Valentina Zamora, 2006. CEOs’ Outside Employment Opportunities and the Lack of Relative Performance Evaluation in Compensation Contracts. The Journal of Finance Vol. LXI, No. 4 Ryan, Harley E. and Roy A. Wiggins, 2004, Who is in whose pocket? Director compensation, board independence, and barriers to effective monitoring, Journal of Financial Economics 73, 497-524. Schwab, Stewart J. and Randall S. Thomas, 2005, An Empirical Analysis of CEO Employment Contracts: What Do Top Executives Bargain For. Vanderbilt University Law & Economics Research Paper Series Working Paper No. 04-12. Cornell Law School Legal Studies Research Paper Series No. 04-024. Available at SSRN:http://ssrn.com/abstract=529923 Shivdasani, A., and D. L. Yermack, 1999, CEO involvement in the selection of new board members: An empirical analysis, Journal of Finance, 54, 1829-1853. Smith, C., and R. Watts, 1992, The investment opportunity and corporate financing, dividend, and compensation policies, Journal of Financial Economics, 32, 263-292. Weisbach, Michael S., 1988, Outside directors and CEO turnover, Journal of Financial Economics 20, 431-460 Yermack, D., 1995, Do corporations award CEO stock options effectively? Journal of Financial Economics, 39, 237-269. Yermack, D., 1997, Good timing: CEO stock option awards and company news announcements, Journal of Finance LII, 2, 449-476. Zajac, Edward J., 1990, CEO selection, succession, compensation, and firm performance: a theoretical integration and empirical analysis, Strategic Management Journal, 11, 3, 217-230. Zajac, E., and J. Westphal, 1996, Who shall succeed? How CEO/board preferences and power affect the choices of new CEOs, Academy of Management Journal 39, 64-90.
Table 1
Turnovers by Year and Characteristics 1993-2005
This table presents frequencies and percentages of turnovers by year and turnover characteristics.
Outsiders indicates that the replacement CEO came from outside the firm. Forced indicates that
the outgoing CEO was fired or pressured out by the board of directors.
All Turnovers Outsiders Insiders Voluntary Forced
N % N % N % N % N %
1993 17 1.38% 7 41.18% 10 58.82% 16 94.12% 1 5.88%
1994 69 5.60% 17 24.64% 52 75.36% 57 82.61% 12 17.39%
1995 93 7.54% 20 21.51% 73 78.49% 85 91.40% 8 8.60%
1996 78 6.33% 17 21.79% 61 78.21% 66 84.62% 12 15.38%
1997 94 7.62% 25 29.60% 69 73.40% 84 89.36% 10 10.64%
1998 92 7.46% 27 29.35% 65 70.65% 84 91.30% 8 8.70%
1999 102 8.27% 26 25.49% 76 74.51% 92 90.20% 10 9.80%
2000 128 10.38% 45 35.16% 83 64.84% 108 84.38% 20 15.63%
2001 111 9.00% 39 35.14% 72 64.86% 98 88.29% 13 11.71%
2002 99 8.03% 33 33.33% 66 66.67% 86 86.87% 13 13.13%
2003 102 8.27% 36 35.29% 66 64.71% 84 82.35% 18 17.65%
2004 107 8.68% 47 43.93% 60 56.07% 95 88.79% 12 11.21%
2005 140 11.28% 54 35.85% 85 61.15% 118 84.89% 21 15.11%
Total 1232 100% 393 31.95% 839 68.05% 1073 87.10% 159 12.90%
Figure 1
Turnovers by Year and Characteristics 1993-2005
This figure presents percentages of the sample turnovers based on turnover characteristics.
Outsiders indicates that the replacement CEO came from outside the firm. Forced indicates that
the outgoing CEO was fired or pressured out by the board of directors.
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
45.00%
50.00%
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Outsiders
Forced
Table 2
CEO Compensation Surrounding Turnover Events 1993-2005
This table presents the summary statistics for CEO compensation of incoming and outgoing
CEOs when a turnover event occurs acquired through the Compustat ExecuComp database. Total
cash compensation represents salary + bonus. Shares owned is the number of shares owned by the
CEO. CEO ownership is the number of shares owned by the CEO multiplied by the stock price
(fiscal year-end closing price). Tenure is the number of years the CEO has been in his current
position. Total compensation includes total cash pay, stock options, restricted stock, payouts
from long-term plans, and benefits. Stock options are valued at the grant-date using
ExecuComp’s modified Black-Scholes methodology.
Mean Median
Number of
Observations
Panel A: All CEOs
Salary $ 664,119.3 $ 600,000 2464
Bonus $ 691,178.4 $ 345,750 2464
Total Cash Compensation $ 1,355,300 $ 957,526.5 2464
Value of Options Granted $ 3,502,000 $ 1,086,680 2348
Shares Owned (%) 3.75% 1.08% 2336
CEO Ownership $ 39,029,820 $ 3,583,000 2336
Total Compensation $ 4,390,370 $ 2,236,840 2464
Panel B: Incoming CEOs
Salary $ 655,253.1 $ 600,000 1232
Bonus $ 701,236.5 $ 354,500 1232
Total Cash Compensation $ 1,356,490 $ 950,699.5 1232
Value of Options Granted $ 3,772,060 $ 1,121,910 1117
Shares Owned (%) 3.97% 1.125% 1176
CEO Ownership $ 19,730,910 $ 2,208,480 1176
Total Compensation $ 4,400,050 $ 2,439,790 1232
Panel C: Outgoing CEOs
Salary $ 672,985.5 $ 600,000 1232
Bonus $ 681,120.4 $ 326,560 1232
Total Cash Compensation $ 1,354,110 $ 964,555 1232
Value of Options Granted $ 3,231,400 $ 1,015,440 1231
Tenure 9.1 7 1232
Shares Owned (%) 4.83% 1.4% 1160
CEO Ownership $ 58,594,920 $ 5,850,450 1160
Total Compensation $ 4,380,700 $ 2,107,070 1232
Table 3
CEO Compensation Around Turnover Events: Differences in Pay 1993-2005
This table presents the differences between incoming and outgoing CEOs around turnover events. P-values are in parentheses. Total cash
compensation represents salary + bonus. Total compensation includes total cash pay, stock options, restricted stock, payouts from long-term
plans, and benefits. Stock options are valued at the grant-date using ExecuComp’s modified Black-Scholes methodology. Options as a percentage
of total pay are measured by dividing Options Granted by Total Compensation.
Median Differences in CEO Pay:
Incoming CEO – Outgoing CEO
All CEOs Replacement
Insiders
Replacement
Outsiders
Forced
Turnovers
Voluntary
Turnovers
Excess Pay ≤ 0 Excess Pay > 0
Total Cash Pay ($) 0 -35,256* 96,557** 79,640* -8,370 33,789* -47,279**
Total Compensation ($) 159,560*** 125,156** 335,360** 134,324 163,560*** 490,343*** -275,269***
Options ($) 296,130*** 79,230*** 1,103,390*** 723,600** 247,900*** 331,220*** 163,340*
Options (%) 7.91%*** 2.3%** 2.1%*** 12.5%* 7.4%*** 4%*** 13.3%***
Table 4
Multivariate Analysis of CEO Compensation Levels
This table presents multivariate analysis of CEO compensation levels around a turnover event. The logs
of total cash pay and total compensation are the dependent variables. Log(assets) is the log of total assets.
ExecDir is a dummy which equals 1 if the CEO is on the board of directors. Ownership is the shares
owned of the CEO. Tenure is the number of years the CEO has been active in his current position.
Newceo is a dummy which equals 1 if the CEO is a replacement CEO. Interlock is a dummy which
equals 1 if the CEO is listed as a member of the executive compensation committee. Num Mtgs is the
number of board meetings during the year. Excess is a dummy which equals 1 if the outgoing CEO has
Excess Pay > 0. Insider is a dummy which equals one if the replacement CEO is an insider. Prevceo is a
dummy variable which is equal to 1 if the CEO has previously been a CEO at another firm. T-stats are
clustered at the firm level and are in parentheses.
Total Cash Pay Total Compensation
(1) (2) (3) (4) (5) (6)
Intercept 4.82 4.80 4.81 4.68 4.67 4.63
(9.91) (9.89) (9.90) (13.27) (13.25) (13.18)
Log(Assets) .3167 .31986 .31909 .42987 .43192 .43664
(22.90) (23.10) (22.82) (42.88) (43.01) (43.16)
ExecDir .1884 .19842 .19653 -.01829 -.00645 .00519
(.94) (1.00) (.99) (-.09) (-.04) (.04)
Ownership -.00002 -.00002 -.00002 -.000009 -.000009 -.000009
(-7.33) (-7.28) (-7.27) (-4.2) (-4.15) (-4.19)
Tenure .0199 .01959 .0196 -.00014 -.00033 -.00038
(5.30) (5.22) (5.22) (-.05) (-.12) (-.14)
Interlock -.25187 -.24383 -.24371 -.25715 -.25197 -.25274
(-2.61) (-2.53) (-2.53) (-3.68) (-3.61) (-3.63)
Num Mtgs -.00758
(-1.21)
-.00832
(-1.33)
-.00813
(1.29)
.00327
(.72)
.0028
(.62)
.00165
(.36)
Prevceo .10392
(1.46)
.07621
(1.06)
.07712
(1.07)
.18745
(3.62)
.16962
(3.25)
.16398
(3.15)
Newceo .09405 .28066 .29179 -.02613 .09392 .02515
(1.77) (3.43) (3.37) (-.68) (1.58) (.4)
Newceo*Insider -.23437 -.23648 -.15077 -.13775
(-2.99) (-3.01) (-2.66) (-2.43)
Newceo*Excess
-.02616
(-.40)
.16154
(3.40)
Industry Yes Yes Yes Yes Yes Yes
Year Yes Yes Yes Yes Yes Yes
Adj-R2
.2349 .2372 .2370 .5039 .5050 .5070
p-value (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
N 2695 2695 2695 2695 2695 2695
Table 5
Multivariate Analysis of Pay-Performance Sensitivities
This table presents multivariate analysis of pay-performance sensitivity around a turnover event. Total
cash pay and total compensation are the dependent variables. ExecDir is a dummy which equals 1 if the
CEO is on the board of directors. Ownership is the ownership level of the CEO. Newceo is a dummy
which equals 1 if the CEO is a replacement CEO. Resigned is a dummy which equals 1 if the CEO was
forced to resign the following year. Insider is a dummy which equals one if the replacement CEO is an
insider. T-stats are in parentheses.
∆Total Cash Pay ∆Total Compensation
(1) (2) (3) (4) (5) (6)
Intercept -44.98 -44.92 -40.79 2822.12 2819.71 2837.21
(-.22) (-.22) (-.20) (1.16) (1.16) (1.17)
∆Shareholder Wealth .03075 .03075 .03075 .0451 .0451 .0451
(13.30) (13.30) (13.39) (1.63) (1.63) (1.63)
ExecDir 83.485 83.475 78.735 -3679.1 -3678.9 -3711.8
(.41) (.41) (.39) (-1.5) (-1.50) (-1.51)
Ownership .00497 .00497 .00474 -.1178 -.11763 -.11858
(1.46) (1.46) (1.40) (-2.82) (-2.82) (-2.84)
Newceo*∆Shareholder Wealth -.01072 -.01054 -.02936 .04459 .03774 -.04349
(-3.62) (-3.43) (-6.94) (1.26) (1.02) (-.84)
Newceo*Excess*∆Shareholder
Wealth
-.00111
(-.22)
-.00701
(-1.38)
.04156
(.69)
.01755
(.29)
Newceo*Insider*∆Shareholder
Wealth
.02708
(6.43)
.11532
(2.24)
Adj-R2
.0885 .0883 .1003 .0082 .0080 .0093
p-value (0.000) (0.000) (0.000) (0.000) (0.000) (0.000)
N 3025 3025 3025 3001 3001 3001
Table 6
CEO Compensation Around Turnover Events: Incoming - Outgoing 1993-2005
Education and Insider/Outsider Status
This table presents the differences between incoming and outgoing CEOs around turnover events. P-values are in parentheses. Total cash
compensation represents salary + bonus. Total compensation includes total cash pay, stock options, restricted stock, payouts from long-term
plans, and benefits. Stock options are valued at the grant-date using ExecuComp’s modified Black-Scholes methodology. Options as a percentage
of total pay are measured by dividing Options Granted by Total Compensation.
Panel A: General and Specific Education
Median Differences in CEO Pay:
Incoming CEO – Outgoing CEO
Incoming CEO General
Education Lower or Equal
Incoming CEO General
Education Greater
Incoming CEO Specific
Education Lower or Equal
Incoming CEO Specific
Education Greater
Total Cash Pay ($) 1,667 -52,057 0 -6,861
Total Compensation ($) 155,960*** 260,524* 185,270*** 2,632
Options ($) 284,390*** 532,400* 296,130*** 219,190**
Options (%) 8.5%*** 3.8% 9.3%*** 1.3%
Panel B: Education and Insider/Outsider Status
Median Differences in CEO Pay:
Incoming CEO – Outgoing CEO
Incoming CEO General
Education Lower or Equal
Incoming CEO General
Education Higher
Incoming CEO Specific
Education Lower or Equal
Incoming CEO Specific
Education Higher
Insider
Replacement
Outsider
Replacement
Insider
Replacement
Outsider
Replacement
Insider
Replacement
Outsider
Replacement
Insider
Replacement
Outsider
Replacement
Total Cash Pay ($) -26,343* 101,471** -87,425 22,401 -25,250* 90,000** -83,800 255,887**
Total Compensation ($) 126,156** 241,350* 113,999 961,030* 127,450*** 348,840** 27,193 -9,272
Options ($) 109,003*** 863,310*** 0 1,566,540** 92,905** 1,103,390*** 0 947,000
Options (%) 2.4%** 27.5%*** -2.5%*** 2.4% 3.3%** 23.3%*** 1.8% 3.6%
Table 7
CEO Compensation Around Turnover Events: Incoming - Outgoing 1993-2005
Prior Firm Performance and Pay
This table presents the differences between incoming and outgoing CEOs around turnover events. P-values are in parentheses. Total cash
compensation represents salary + bonus. Total compensation includes total cash pay, stock options, restricted stock, payouts from long-term
plans, and benefits. Stock options are valued at the grant-date using ExecuComp’s modified Black-Scholes methodology. Options as a percentage
of total pay are measured by dividing Options Granted by Total Compensation.
Panel A: Incoming CEO prior firm performance and pay
Median Differences in CEO Pay:
Incoming CEO – Outgoing CEO
ROA > 4.8% ROA < 4.8% Prior Excess Pay > 0 Prior Excess Pay ≤ 0
Total Cash Pay ($) 256,616* 168,330* 229,890* 203,376**
Total Compensation ($) 693,800* 452,236 487,680 472,550*
Options ($) 1,630,690** 1,658,500** 3,145,000*** 774,570**
Options (%) 11%* 11.9%** 46.4%* 9.9%*
Panel B: Outgoing CEO Subsequent Employment
Median Differences in CEO Pay:
Incoming CEO – Outgoing CEO
Outgoing CEO does not find a new job within 3 years Outgoing CEO does find a new job within 3 years
Total Cash Pay ($) -10,357 59,282*
Total Compensation ($) 194,070*** 66,824
Options ($) 299,390*** 223,250**
Options (%) 8.35%*** 5.8%
Table 8
CEO Compensation Around Turnover Events: Incoming CEO – Outgoing CEO 1993-2005
Managerial Power
This table presents the differences between incoming and outgoing CEOs around turnover events. P-values are in parentheses. Total cash
compensation represents salary + bonus. Total compensation includes total cash pay, stock options, restricted stock, payouts from long-term
plans, and benefits. Stock options are valued at the grant-date using ExecuComp’s modified Black-Scholes methodology. Options as a percentage
of total pay are measured by dividing Options Granted by Total Compensation.
Panel A: CEO-based managerial power proxies
Median Differences in CEO Pay:
Incoming CEO – Outgoing CEO
Founder Non-Founder CEO
Centrality
High
CEO
Centrality
Low
CEO Tenure
> 4
CEO Tenure ≤
4
CEO Firm
Ownership >
1.4%
CEO Firm
Ownership <
1.4%
Total Cash Pay ($) 57,888* -10,178 -105,810*** 24,208* -27,809* 81,254** 26,917 -9,502
Total Compensation ($) 405,290*** 128,360** -48,449 204,140*** 207,330*** 55,660 343,480** 128,210**
Options ($) 431,440*** 243,360*** 63,120* 370,330*** 214,300*** 337,355*** 232,410** 261,060***
Options (%) 8.4%* 7.7%*** .3%* 9.5%*** 4.8%*** 12.3%*** 14.4%** 5.3%***
Panel B: Firm-based managerial power proxies
Median Differences in CEO Pay:
Incoming CEO – Outgoing CEO
Gindex ≥ 11 Gindex < 11 Board Composition
< 50th
Percentile
Board Composition
> 50th
Percentile
Firm not in Financial
Distress
Firm in Financial Distress
Total Cash Pay
($)
-52,627 9,134 -464 2,072 -7,802 40,000
Total Compensation
($)
-17,600 214,260*** 250,810** 130,073 184,540*** -31,521
Options ($) 106,638* 327,570*** 188,470*** 297,750*** 190,530*** 1,120,890***
Options (%) 4.4%** 9.3%*** 6.4%*** 8.25%*** 5.7%*** 49%***