Journal of Corporate Finance 6 2000. 141–164
www.elsevier.comrlocatereconbase
Corporate governance and recent consolidationin the banking industry
Yaron Brook a,), Robert J. Hendershott a, Darrell Lee ba Santa Clara UniÍersity, 500 El Camino Real, Santa Clara, CA 95053, USA
b Kennesaw State UniÍersity, Kennesaw, GA, USA
Abstract
Using the universe of publicly traded banks at year-end 1993, we find that target banks’outside directors, but not inside directors, tend to own more stock than their counterparts inother banks. Having an 留學生dissertation網outside blockholder is also associated with banks becoming targets.In contrast to existing research on industrial firms, board structure does not help determinewhich sample banks sell. Neither the fraction of outsiders on a bank’s board nor having anoutside-dominated board differentiate the target banks in our sample. Instead, outsidedirectorsrshareholders and blockholders appear to be primarily responsible for encouragingbank managers to accept an attractive merger offer q2000 Elsevier Science B.V. All rightsreserved.
JEL classification: G21; G34
Keywords: Takeovers; Corporate governance
Financial economists have long recognized that the widespread separation ofownership and control in large US corporations creates the potential for costlyagency conflicts. Dispersed shareholders’ limited incentive to monitor the behaviorand performance of the agents hired to run their firm can give managerssubstantial freedom to pursue their own interests at the expense of shareholderwealth. Absent mechanisms to control managerial behavior, usually called ‘‘corpo-) Corresponding author.0929-1199r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
PII: S0929- 119900.00011-0
142 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164rate governance structures,’’ wealth maximization will not exclusively motivatecorporate decision-making.
A company’s board of directors is one of the mechanisms within the firm tomonitor and control managerial behavior.1 The board hires, fires, and compensatesa firm’s top management. However, board members are subject to their ownagency problems. A firm’s chief executive officer CEO. is generally a memberof, and is often the chairman of, the board of directors. CEOs typically also play amajor role in selecting new nominees for the board Lorsch and MacIver, 1989.. Ifboard members are strongly influenced by or beholden to the firm’s top officers, itis unclear whether the board can successfully align shareholder and managerinterests.
The banking industry’s ongoing consolidation offers an excellent experimentalsetting for examining board effectiveness. During the last decade, technologicaladvances in communication and information technologies have reduced the costsof having a geographically dispersed banking organization. Simultaneously, lawsand regulations that had previously fragmented the banking industry have beenrelaxed or, in some cases, eliminated.2 The joint effect has been a dramatic surgein merger activity that has sharply reduced the number of US banks Holland et#p#分頁標題#e#
al., 1996..Similar to acquisitions by industrial firms, the lion’s share of takeover gains inbank mergers goes to target shareholders see Jensen and Ruback 1983. for asummary of industrial takeover research.. For example, in a sample of 153 bankacquisitions between 1985 and 1991, Houston and Ryngaert 1994. find that targetbanks earn positive average abnormal returns of 14.4% and bidder shareholderssuffer negative average abnormal returns of y2.3%. In contrast to target shareholders’gains, a target bank’s managers tend to find themselves out of a job afterbeing acquired while the bidding bank’s managers preside over the newly mergedinstitution. Hadlock et al. 1999. find that more than one-half of the top executivesin their sample of target banks are not employed by the buying bank 2 years afterthe acquisition. Because shareholders tend to benefit from being acquired whilemanagers generally lose both future compensation and prestige if their firm isacquired, takeovers epitomize how shareholder and managerial interests cancollide.
The natural divergence of target shareholder and manager incentives in responseto a merger bid makes takeovers a model experiment for exploring the firmcharacteristics that are associated with managers acting in shareholders’ interests.However, past research attempting to discern target firms’ distinctive character-
1 Shleifer and Vishny 1986. show how outside blockholders can also discipline managers.
2 For example, in 1994 the Interstate Banking and Branching Efficiency Act eliminated interstatebank merger restrictions and virtually eliminated restrictions on banks’ ability to consolidate theirsubsidiaries today, only Texas continues to require bank holding companies to maintain independentbanks within their state..
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 143istics have met with limited success see, for example, Hasbrouck 1985. andPalepu 1986... One potential problem is that sample size requirements generallydictate using acquisitions from many different industries over a wide time period.Having this heterogeneous sample will likely decrease the power of any statisticalcomparison, despite the authors best efforts to control for differences by, forexample, industry-adjusting financial characteristics.. Having a heterogeneoussample is a particular problem when evaluating corporate governance characteristicsbecause it is prohibitively expensive to gather data on the universe of allpossible targets in order to control for industry or temporal variation.
This paper exploits the banking industry’s recent consolidation to explore whatcorporate governance characteristics are associated with managers acting in shareholders’
best interests. Banks provide a useful experiment because the burst ofrecent merger activity in this historically fragmented industry allows us to study a
reasonably large sample of very homogeneous firms. Additionally, because theneed for regulatory approval makes hostile bank takeovers particularly difficult,the target bank’s cooperation is generally a prerequisite for a bank mergerProwse, 1997.. Therefore, governance structures that amplify an institution’sconcern for shareholder wealth may play a greater role in determining the targetsof bank acquisitions than the targets of industrial takeovers.Our study uses all publicly traded banks with data available at year-end 1993that were not in the process of being acquired. Sixty-one 19%. of these 321potential target banks were subsequently acquired during the next 3 years. We findthat higher director and officer D&O. ownership greatly increases the likelihood#p#分頁標題#e#
that a sample bank sells. Upon further examination, however, we find that outsidedirector ownership drives this result: targets’ outside directors, but not insidedirectors or officers, tend to own more stock than their counterparts in other banks.This holds whether we compare target banks to all other banks, compare targets tomatched samples of other banks, or measure the marginal impact of directorownership in a logistic regression while controlling for other bank characteristics.In addition to outside director ownership, we find that having an outsideblockholder also makes a bank more apt to sell. Other commonly studiedgovernance variables, however, do not differentiate the target banks in our sample.The fraction of outsidersrinsiders on a bank’s board is unrelated to a bank’swillingness to sell. Additionally, target banks are no more likely than other banksto have an outside-dominated board. There is also no consistent relation betweenour insider ownership variables and which banks become targets. It appears thatoutside directorsrshareholders and outside blockholders are primarily responsiblefor encouraging bank managers to accept a merger offer.Our results suggest that board ownership, particularly outside director ownership,can play a crucial role in effective corporate governance. Although outside
directors bear some personal costs in a takeover most of a target’s outside
directors do not join the merged company’s board., these costs are small compared
to those borne by the firm’s top officers. This appears to give outside directors’
144 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164equity ownership a far greater marginal impact than the equity ownership ofinsiders. Additionally, insider ownership is a double-edged sword. Although it canalign manager and shareholder interests, it can also entrench managers, insulatingthem from external discipline. There is no corresponding downside to outsidedirector ownership.
We also examine the cross-sectional relation between target governance and thegains from a takeover measured using the target’s announcement return.. Consis-tent with previous research on industrial takeovers Cotter et al., 1997., having anoutside-dominated board is associated with greater target gains. D&O ownershipand target takeover gains are weakly negatively related in our sample, althoughthis result is driven by inside director ownership. Oddly, having an outsideblockholder is associated with lower announcement returns.
Many industries have experienced substantial consolidation during the 1990s.How efficiently these industries evolve depends in part on managerial incentives.Our study of the banking industry in the 1990s describes how corporate governancehas affected consolidation in this particular sector, and, hopefully, provides
a foundation for exploring other industries. It is important to recognize, of course,that our focus on the highly regulated banking industry may limit the generality ofour results.#p#分頁標題#e#
The rest of the paper is organized as follows. The first section reviews therelevant existing literature. The second presents our sample and data. The third
explores the differences between the sample banks that do and do not becometargets. The fourth section analyzes the cross-sectional distribution of targetannouncement returns. The fifth section concludes.
1. Background
The potential for agency conflicts inherent in the separation of ownership andcontrol makes it imperative that publicly traded corporations have effective
corporate governance systems. The most immediate governance mechanism is thefirm’s board of directors. The power to hire, fire, and compensate top managersgives well-functioning boards the ability to greatly reduce any incentive conflictsbetween managers and shareholders.
There is ample evidence that boards have an impact on corporate decision-makingand performance. In particular, outside directors appear to improve a board’seffectiveness.3 Using Forbes 500 firms between 1974 and 1983, Weisbach 1988.finds a significant relationship between firm performance and CEO turnover onlywhen at least 60% of the board is made up of outside directors. Rosenstein and
3 Outside directors are essentially all directors without existing or likely future business relationshipswith the firm. Inside directors currently work for the firm. Past employees and parties with a current orlikely business relationship are categorized as gray directors.
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 145Wyatt 1990. find a positive stock price response to 1251 outside director
appointment announcements between 1980 and 1985. Byrd and Hickman 1992.use 128 tender offers between 1980 and 1987 to show that bidding firms withoutside-dominated boards have significantly less negative abnormal returns aroundtheir takeover announcements than firms without outside-dominated boards. Simi-larly, Brickley et al. 1994. use a sample of 247 poison pill announcements
between 1984 and 1986 to show that the market reaction to poison pills is positivewhen a firm has an outside dominated board and negative when the majority of a
firm’s directors are insiders. Finally, Cotter et al. 1997. use a sample of 169tender offer targets to show that target shareholder gains are larger when amajority of the target’s board are outsiders.
To date, most results involving outside directors has emphasized the importanceof their presence, but not their ownership stake in the firm.4 Byrd and Hickman1992. and Cotter et al. 1997. explicitly test for the impact of outside directorownership on bidder and target returns, respectively, during takeover contests.
However, neither paper finds a significant effect.In this paper we examine the relation between corporate governance, particularlyboard ownership, and consolidation in the banking industry between 1994
ahttp://www.400190.com/dissertation_writing/Finance/nd 1996. The substantial number of mergers in this large, historically fragmentedindustry allows us to compare a relatively large sample of banks that become#p#分頁標題#e#
targets over a short period of time to all of the remaining institutions in theindustry. Doing so avoids the need to mix takeovers from many disparate
industries over a long time period in our analysis of corporate governance efficacy.
Additionally, although there are certainly disciplinary bank takeovers, theconsolidation boom in the 1990s is in a large part the result of technological andregulatory shocks that have shifted the optimal number of banks downwardHolland et al., 1996.. Large numbers of banks need to disappear, but it is notobvious which should.5 Because target bank shareholders receive a substantial
premium in a bank merger while target managers risk losing their jobs, managerand shareholder incentives tend to diverge on the question of whether to sell.
Shareholders generally prefer their bank become a target while managers prefertheir bank be one of the survivors.6 Which banks sell could well reveal whichcorporate governance mechanisms most effectively tilt corporate decision-making
4 An exception is Shivdasani 1993. who reports that outside director ownership makes a disci-plinary takeover less likely presumably because it is unnecessary..
5 Harris 1994. discusses purely synergistic takeovers in which the merger gains do not depend onwhich firm buys which.
6 Because the vast majority of bank mergers are tax-free stock-for-stock transactions, the takeoverpremium means that target shareholders receive a larger share of the merged firm than the banks’
pre-merger values would justify. Target shareholders who receive the bidder’s stock are free toparticipate in any additional post-merger value increases by continuing to hold the merged bank’s
stock.146 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
toward shareholder interests. This is not to say that economic factors have no
influence on which banks buy and which banks sell, only that the board’s concern
for shareholder wealth should have an impact at the margin.
Our study is closely related to Hadlock et al. 1999. who compare 84 banks
acquired between 1982 and 1992 to a sample of carefully matched control banks.
These authors find that banks with substantial inside ownership both director and
affiliated blockholder. are less likely to be acquired. In particular, these banks are
less likely to be acquired and see their top management fail to remain employed
by the buying bank.
Our study differs from Hadlock et al. 1999. in two important ways. First, our
period of study 1994–1996. comes during a time when takeover restrictions have
been dramatically relaxed, making virtually all banks potential targets. This
window also covers a consistently strong period for the banking industry, largely
eliminating takeovers in which weak targets were forced to merge under either
explicit or implicit threats from regulators. Second, using the universe of public#p#分頁標題#e#
banks may give truer estimates of the relation between governance and whether a
bank sells the regression coefficients. than are possible using a matched-sample
approach.
2. Sample and data
2.1. Sample
Our initial sample consists of all firms with three-digit SIC codes of 602 or 671
that are active on CRSP at year-end 1993 SIC code 602 indicates a financial
institution and SIC code 671 indicates a holding company.. These firms are
matched with banks in the Sheshunoff Bank Database, published by Sheshunoff
Information Services of Austin, Texas.7 This database provides a comprehensive
collection of detailed financial data, compiled from bank call reports, for 11,656
banks as of year-end 1993. It is necessary to match the CRSP firms to Sheshunoff
banks by name and location. Sixteen ambiguous matches are dropped from the
sample. Using CRSP delistings and newspaper searches, we eliminate all banks
that were in the process of being acquired at year-end 1993. This produces a
sample of 426 publicly traded banks that entered 1994 as potential takeover
targets.
The Sheshunoff database provides all of the financial data for our tests. A
bank’s market value of equity is calculated using CRSP data as the number of
shares outstanding times the closing stock price at year-end 1993. The number of
7 Throughout we use ‘‘bank’’ to refer to both banks and bank holding companies.
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 147
US offices branches. owned by each bank is taken from The Banking Organization
Quarterly, published by Sheshunoff. Data on corporate governance structures
ownership and board structures. are compiled from 1993 proxy statements. If a
1993 proxy is unavailable, the 1994 proxy is used, provided it is filed during the
first quarter of 1994. Because our study focuses on corporate governance characteristics,
only the 321 banks with an appropriate proxy available are included in
our final sample.
2.2. GoÍernance Íariables
From the proxy statements, we identify equity ownership of all officers and
directors as a fraction of total shares outstanding. Directors are then classified as
insiders, outsiders, or gray defined as in Hermalin and Weisbach, 1988., and the
different categories’ aggregate fractional ownership are recorded. For our purposes,
investment bankers are always classified as gray directors while other
professionals lawyers, accountants, consultants, etc.. are classified as gray only if
they are reported as having a business relationship with the bank. Affiliated block
ownership is included in our measures of director ownership. Undistributed ESOP
shares and trust shares when the bank is trustee. are assigned to the bank’s CEO,
unless the proxy indicates control belongs elsewhere e.g., in a few cases an#p#分頁標題#e#
outside Chairman is the ESOP trustee.. If an individual director controls another
entity that is reported as owning shares of the bank, the individual director is
assigned ownership of those shares.
We define a board as ‘‘outside-dominated’’ if the fraction of outside directors
exceeds 0.5. We also calculate aggregate outside block ownership unaffiliated
owners reported as holding more than 5% of the bank’s shares. as a fraction of
total shares outstanding. Our expectation is that greater director ownership and
having an outside-dominated board will improve the board’s ability and incentives
to monitor and discipline managers. If so, banks with these governance characteristics
should be more likely, at the margin, to accept an attractive merger offer.
Similarly, we expect that banks with substantial outside block ownership will be
more apt to become targets.
2.3. Control Íariables
Research showing that acquiring firms tend to be larger than target firms
Stevens 1973., among others. suggests it may be important to control for bank
size in our tests. We measure size using the bank’s total assets. We also control for
banks’ average branch size total assetsrnumber of branches.. Branch size is
intended to capture two effects. Banks with many small offices may make
attractive targets because eliminating redundant branches offers greater consolidation
benefits. Banks with many small branches may also be attractive targets
because they allow an acquirer to enter a new market quickly.
148 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
Finally, we control for banks’ past performance. Hasbrouck 1985. finds lower
q-ratios in 86 non-financial takeover targets than in either size- or industry-matched
control firms. Craig and Dos Santos 1996. show that banks actively acquiring
other banks typically have performed better than the banks they acquire. Controlling
for performance is important to reduce omitted variable biases. Because board
members will generally have exceptional information about a bank’s future
profitability, board ownership may vary with performance. When prospects are
good bad., board members have an incentive to accumulate more less. stock.
Therefore, failing to control for prior performance could lead to a spurious
correlation between director ownership and the likelihood of a merger.
We use two measures of banks’ prior performance, a q-ratio and return on
assets ROA.. q is the ratio of the market value to the replacement cost of a firm’s
assets. Because banks’ tangible assets are primarily loans and liquid financial
assets, market values and replacement costs are identical in many cases. However,
if q is interpreted as the ratio of the bank’s value as an ongoing concern to its#p#分頁標題#e#
liquidation value, q can be estimated as total assetsqmarket value of equityy
book value of equity.rtotal assetsqmarket value of investment securitiesybook
value of investment securities.. ROA is calculated as the ratio of 1993 net income
to book value of total assets. Table 1 provides our sample’s summary statistics as
of year-end 1993.
Although over 10 thousand banks existed at the end of 1993, our sample of 321
banks covers most of the industry’s assets. Because the vast majority of banks at
Table 1
Sample banks’ summary statistics
Year-end 1993 summary statistics for our sample of 321 publicly traded banks. Total assets and
average branch size are in $ millions. q is calculated as total assetsqmarket value of equityybook
value of equity. divided by total assetsqmarket value of securitiesybook value of securities.. Return
on assets is calculated as net income divided by book value of assets. Average branch size is calculated
as total assets divided by number of branches. Inside directors are all directors that are also officers of
the bank. Outside directors are all directors that are neither current nor past employees of the bank in
any capacity other than as a director. Outside blockholders are all 5% owners with no affiliation to the
bank other than their equity ownership.
Number Mean Median Standard deviation
Total assets 321 8344 934 24,436
Return on assets 321 0.009 0.011 0.009
q-ratio 306 1.074 1.025 0.382
Number of branches 317 110 24 246
Average branch size 317 360 45 3839
Percent outside directors on board 315 0.621 0.650 0.177
Ownership by all directors and officers 321 0.156 0.101 0.154
Ownership by inside directors 319 0.069 0.025 0.125
Ownership by outside directors 319 0.053 0.028 0.066
Ownership by outside blockholders 310 0.064 0.000 0.101
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 149
this time were small private institutions the same is true today., our sample’s
aggregate assets of $2.7 trillion represent over 70% of the banking industry’s total
assets. Fortunately, despite its tilt toward larger institutions, our sample contains a
wide variety of banks. For example, total assets ranges from $19 million to $217
billion, market value of equity ranges from $3 million to $17 billion, and the
number of bank branches ranges 1 to 2057.
3. Which banks sell?
CRSP delistings and newspaper searches allow us to identify all sample banks
that were acquired in 1994, 1995 or 1996. A bank is classified as a target if it
agrees to a merger after 12r31r93 and the merger is successfully completed
before 1r1r97. For each acquisition, we also identify the bidding bank and the
merger’s announcement date. Table 2 reports the number of acquisitions announced
each year. Sixty-one sample banks were acquired: 23 in 1994, 27 in#p#分頁標題#e#
1995, and 11 in 1996.8
The optimal time period over which to categorize targets is not obvious. One
extreme would be to require that a bank be acquired in 1994. However, the
volume of merger activity in the banking industry over 15% of our sample banks
were bought within 2 years. makes it unlikely that all willing targets could have
sold themselves for an attractive price in 1994. If so, a 1 year window would miss
many banks that should be classified as targets because they were about to be
acquired. Because misclassifying targets will reduce our tests’ power, we believe it
appropriate to categorize target banks based on a longer horizon. However,
extending the window indefinitely does not make sense — agreeing to a merger is
supposed to indicate the bank’s willingness to sell as of year-end 1993. Admittedly,
our requirement that targets be acquired in a merger both announced and
completed between 1r1r94 and 12r31r96 is somewhat ad-hoc. Later in the
paper we show that our results are not specific to this particular window.
Table 2 also provides market-adjusted announcement returns adjusted using
the CRSP equally weighted index. for the bidder and target banks.9 The marketadjusted
returns are calculated over 2 days, the day before and day of the merger
announcement, unless the target bank announces it is for sale before the specific
merger is negotiated and reported.10 In these cases, the target bank’s market-ad-
8 The number of 1996 acquisition announcements is limited by our requirement that the takeover be
successfully completed by 12r31r96 and the fact that mergers in 1994 and 1995 left the remaining
sample somewhat smaller at the beginning of 1996 than at the beginning of 1994.
9 Our results are insensitive to both using different market proxies and different methods for
calculating abnormal returns Brown and Warner, 1985..
10 In seven cases the bank’s imminent sale was reported an average of 4 trading days before the
detailed merger announcement.
150 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
Table 2
The distribution of bank takeovers and announcement returns
Takeovers of banks in our sample of 321 publicly traded institutions announced during 1994, 1995, and 1996. A sample bank is deemed to be acquired if the
takeover is both announced and completed between 1r1r94 and 12r31r96. Announcement returns are calculated net-of-market using the CRSP
value-weighted index. for the day before and day of the merger announcement. A sign test determines whether the percent of positive announcement returns
differs significantly from 50%.
Significant at the 0.10 level.
Year Number of takeovers Target announcement return Bidder announcement returns
Mean median. t-statistic % positive Mean median. t-statistic % positive
1994 23 0.159 0.105. 4.631) ) ) 82.6) ) ) y0.034 y0.039. y3.375) ) ) 23.2) ) )#p#分頁標題#e#
1995 27 0.215 0.184. 6.183) ) ) 96.3) ) ) y0.014 y0.015. y1.126 44.0
1996 11 0.163 0.125. 3.981) ) ) 100) ) ) y0.009 y0.006. y0.70 35.7
Total 61 0.184 0.160. 8.624) ) ) 91.8) ) ) y0.021 y0.022. y3.063) ) ) 33.8) )
))Significant at the 0.05 level.
)))Significant at the 0.01 level.
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 151
justed return is calculated from the day before the bank is reported as being for
sale to the day the actual merger is announced.
We find that only target bank shareholders directly benefit from mergers.
Target banks’ market-adjusted announcement returns average 18.4% medians
16.0%. in our sample. In contrast, bidder announcement returns average y2.1%
mediansy2.2%. in our sample. These returns are comparable to the 20%
average announcement return reported in Houston and Ryngaert 1994.. Both the
bidder and target returns in our sample differ from zero and from each other. at
better than the 0.01 significance level.
Although these takeovers clearly benefit target shareholders, target managers
will view them as a mixed blessing at best. We examine the bidding bank’s proxy
statement 1 year following the takeover to discover whether the target’s top
Table 3
Univariate comparison of target banks to other banks
Univariate comparison of the banks in our sample of 321 publicly traded institutions that do and do not
become targets. A sample bank is deemed to be a target if the institution is acquired in a takeover that
is both announced and completed between 1r1r94 and 12r31r96. All characteristics are measured as
of year-end 1993. Total assets and average branch size are in $ millions. q is calculated as total
assetsqmarket value of equityybook value of equity. divided by total assetsqmarket value of
securitiesybook value of securities.. Return on assets is calculated as net income divided by book
value of assets. Average branch size is calculated as total assets divided by number of branches. Inside
directors are all directors that are also officers of the bank. Outside directors are all directors that are
neither current nor past employees of the bank in any capacity other than as a director. Outside
blockholders are all 5% owners with no affiliation to the bank other than their equity ownership. The
t-statistic and z-statistic are calculated using a standard t-test and a Mann–Whitney test, and indicate
differences between means and medians, respectively.
Targets 61. Other banks 260. t-statistic z-statistic
Total assets 4722 976. 9193 907. 1.288 0.601
Return on assets 0.008 0.010. 0.009 0.011. 0.213 1.067
q-ratio 1.017 1.018. 1.086 1.027. 1.199 1.833)
Number of branches 91 23. 114 25. 0.657 0.558
Average branch size 179 47. 400 43. 0.397 0.062
Percent outside 632 0.655. 0.619 0.650. y0.527 0.259#p#分頁標題#e#
directors on board
Ownership by all 0.175 0.144. 0.152 0.094. y1.072 1.492
directors and officers
Ownership by inside 0.059 0.018. 0.072 0.026. 0.683 0.613
directors
Ownership by outside 0.079 0.042. 0.047 0.027. y3.514) ) ) 2.418) )
directors
Ownership by outside 0.078 0.060. 0.060 0.000. y1.178 1.840)
blockholders
)Significant at the 0.10 level.
))Significant at the 0.05 level.
)))Significant at the 0.01 level.
152 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
managers retain a high level position in the bidding bank after the takeover. Only
5% of target CEOs hold positions in the bidding bank that are senior enough to
warrant mention in the bidder’s proxy statement. Board representation is some-
what higher 16%., but it is clear that many, if not most, target CEOs do not
maintain their status after an acquisition. The chairmen of target banks’ boards
fare similarly — only 23% retain a board seat after the merger.
3.1. A uniÍariate comparison
Table 3 provides a univariate comparison of the sample banks that were
acquired and the rest of the sample. The only highly significant difference is that
target banks’ outside directors’ equity ownership is almost double that in the other
banks: targets’ outside directors own an average median. of 7.9% 4.2%. of their
bank’s stock while other banks’ outside directors own an average median. of
4.7% 2.7%.. Targets also have slightly lower median q-ratios and slightly higher
median unaffiliated block ownership.
Table 4 documents the percent of sample banks that have 1. an outside-
dominated board, 2. at least one outside blockholder, or 3. outside director
ownership that exceeds inside director ownership. Outside blockholders are more
common in targets than in other banks 56% of the targets have an outside
blockholder versus 41% of the other banks, ps0.04.. It is also more common for
a target bank’s outside directors to own more equity than the bank’s inside
directors in 66% of the target banks outside directors own more than inside
directors versus 47% of the other banks, ps0.01.. There is not, however, a
significantly greater frequency of outside-dominated boards in target banks 78%
Table 4
The relative frequencies of corporate governance characteristics
The frequency of an outside-dominated board, an outside blockholder, and outside director ownership
exceeding inside director ownership in the 61 sample banks that become targets and the 260 sample
banks that are not acquired. A sample bank is deemed to be a target if the institution is acquired in a
takeover that is both announced and completed between 1r1r94 and 12r31r96. Board and ownership
structures are measured as of year-end 1993. An outside-dominated board is defined as the majority of#p#分頁標題#e#
directors being outsiders. Inside directors are all directors that are also officers of the bank. Outside
directors are all directors that are neither current nor past employees of the bank in any capacity other
than as a director. Outside blockholders are 5% owners with no affiliation to the bank other than their
equity ownership. The x 2-test statistic indicates whether the frequency in the targets exceeds the
frequency in other banks.
Frequency in
Targets Other banks x 2
Outside-dominated board 78% 75% 0.310
Outside blockholder 56% 41% 4.305) )
Outside director ownership) 66% 47% 6.883) ) )
insider director ownership
))Significant at the 0.05 level.
)))Significant at the 0.01 level.
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 153
Table 5
Matched sample univariate comparison of target banks to other banks
Comparison of the 61 sample banks that become targets and a matched sample of 61 banks that are not
acquired. A sample bank is deemed to be a target if the institution is acquired in a takeover that is both
announced and completed between 1r1r94 and 12r31r96. Each target bank is matched to the control
banks from the same Federal Reserve Bank district that is closest in size to the target bank. All
characteristics are measured as of year-end 1993. q is calculated as total assetsqmarket value of
equityybook value of equity. divided by total assetsqmarket value of securitiesybook value of
securities.. Return on assets is calculated as net income divided by book value of assets. Average
branch size is calculated as total assets divided by number of branches, and is reported in $ millions.
Inside directors are all directors that are also officers of the bank. Outside directors are all directors that
are neither current nor past employees of the bank in any capacity other than as a director. The
t-statistic and z-statistic are calculated using a standard t-test and a Mann–Whitney test, and indicate
differences between means and medians, respectively.
Targets Matched sample t-statistic z-statistic
Return on assets 0.008 0.010. 0.008 0.010. 0.754 0.562
q-ratio 1.017 1.18. 1.102 1.023. y1.384 1.539
Number of branches 91.2 23. 7.7 31. 0.900 0.704
Average branch size 179 48. 142 47. 0.212 1.514
Percent outside directors on board 0.632 0.655. 0.617 0.643. 0.310 0.151
Ownership by all directors and officers 0.175 0.144. 0.148 0.107. 0.923 1.379
Ownership by inside directors 0.059 0.018. 0.065 0.017. y0.229 0.00
Ownership by outside directors 0.079 0.042. 0.044 0.026. 2.885) ) ) 2.740) ) )
Ownership by outside blockholders 0.078 0.060. 0.075 0.000. 0.211 0.551
)))Significant at the 0.01 level.
of the targets have an outside-dominated board versus 75% of the other targets,
ps0.60..
3.2. A matched sample comparison#p#分頁標題#e#
The univariate comparison’s value is limited by the possibility that the takeover
sample is biased toward a particular type of institution, creating a spurious
correlation between becoming a target and a particular corporate governance
characteristic. For example, directors’ percentage equity ownership tends to vary
inversely with size. Therefore, if targets tended to be relatively small institutions,
size differences could account for greater director ownership in target banks.
Similarly, sample biases could mask important relations between other corporate
governance structures and becoming a target.
To control for size and geographic differences across banks we match each
target bank to the non-target bank in the same Federal Reserve Bank district that is
closest in size as measured by total assets..11Table 5 compares targets to their
11 The United States is divided into districts covered by the 12 Federal Reserve banks headquartered
in Boston, New York, Philadelphia, Cleveland, Richmond, Atlanta, Chicago, St. Louis, Minneapolis,
Kansas City, Dallas, and San Francisco.
154 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
matches. Consistent with the univariate tests, target banks have greater outside
director ownership than control banks — the median difference and the statistical
significance both increase slightly. Additionally, target banks’ median q-ratio is
somewhat lower than the control banks’. This suggests that the targets’ greater
outside director ownership cannot be explained by director incentives. If anything,
targets’ relatively poor historical performance suggests directors have had an
incentive to buy less stock.
Additional matching criteria produce similar results results not reported in a
table.. In two performancerregion-matched comparisons, using banks’ q-ratios
and ROA to measure performance, each target bank is matched to the non-target
bank in the same Federal Reserve Bank district that has the closest q ROA..
Consistent with the sizerregion match, target banks have consistently higher
outside director ownership than the performancerregion-matched sample.
Two performancersize-matched comparisons provide an additional robustness
check. Non-target banks are divided into four quartiles based on q ROA. and
each target bank is matched to the non-target bank in the same performance
quartile that is closest in size. Again, targets’ outside director ownership is
significantly higher than outside director ownership in the control sample.
3.3. A multiÍariate comparison
Table 6 presents a series of logistic regression models in which the dependent
variable is a dummy indicating whether the bank was acquired. Using a logistic
regression allows us to estimate the marginal impact of various corporate governance#p#分頁標題#e#
attributes on the likelihood that a bank becomes a target while controlling
for other bank characteristics. Specifically, we control for bank size, prior performance,
and the bank’s average branch size.12
3.3.1. Director ownership
The second logistic regression model shows that higher D&O ownership is
associated with a bank becoming a target coefficients2.013, ps0.072.. However,
when we replace total D&O ownership with outside director ownership and
inside director ownership, only outside director ownership receives a significant
coefficient the coefficients range from 7.442 to 9.481 in the various regression
models, with p-values between 0.005 and 0.0002.. Inside director ownership
receives an insignificant coefficient in all of the regression models.
Based on the third logistic model, the implied probability of a bank becoming a
takeover target is e z ir1qe z i ., where z s11.768q0.285 lnTotal assets .y i i
12 In the regressions, bank and average branch size are measured as lntotal assets. and lntotal
assetsrnumber of branches., respectively. Although we only report regressions controlling for prior
performance using q, our findings are similar using ROA except that ROA is not significant in the
regressions..
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 155
Table 6
Logistic regressions explaining which banks become targets within 3 years
Logistic regression models estimating the likelihood that a sample bank becomes a target. The dependent variable is set to one if a bank is acquired in a
takeover that is both announced and completed between 1r1r94 and 12r31r96. All characteristics are measured as of year-end 1993. q is calculated as total
assetsqmarket value of equityybook value of equity. divided by total assetsqmarket value of securitiesybook value of securities.. Average branch size is
calculated as total assets divided by number of branches. Inside directors are all directors that are also officers of the bank. Outside directors are all directors
that are neither current nor past employees of the bank in any capacity other than as a director. The outside board dummy is set to one if the majority of
directors are outsiders. The blockholder dummy is set to one if there is a 5% owner with no affiliation to the bank other than their equity ownership.
Coefficient p-values are in parentheses.
Model 1 Model 2 Model 3 Model 4 Model 5
lntotal assets. 0.055 0.589. 0.162 0.169. 0.285) ) 0.026. – 0.302) ) 0.019.
q-ratio y9.429) ) 0.032. y8.313) 0.059. y9.704) ) 0.044. – y10.205) ) 0.038.
lnaverage branch size. y0.456) 0.086. y0.645) ) 0.033. y0.733) ) 0.024. – y0.688) ) 0.035.
Outside board dummy – 0.033 0.931. y0.473 0.250. y0.12 0.754. y0.624 0.143.
Blockholder dummy – 0.655) ) 0.049. 0.630) 0.064. 0.639) ) 0.036. 0.697) ) 0.044.#p#分頁標題#e#
Ownership by all directors and officers – 2.013) 0.072. – – –
Ownership by inside directors – – 0.664 0.727. y1.77 0.332. 0.768 0.684.
Ownership by outside directors – – 9.481) ) ) 0.000. 7.442) ) ) 0.000. 7.654) ) ) 0.005.
Outside director ownership) – – – – 0.729) 0.096.
insider director ownership
Number of observations 305 292 291 304 291
Pseudo-R2 3.6% 5.8% 10% 6.3% 11%
)Significant at the 0.10 level.
))Significant at the 0.05 level.
)))Significant at the 0.01 level.
156 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
9.704 q y0.733 lnAverage branch size .y0.473 Outside board dummy q0.630 i i i
Outside blockholder dummy y0.664 Inside director ownership q9.481 Outside i i
director ownership . This suggests that a bank with median characteristics as i
reported in Table 1. has a 0.089 probability of becoming a target. Raising outside
director ownership to 10% from the median level of 2.8%. increases the implied
probability to 0.161.
Outside director ownership being strongly related to becoming a target is
consistent with substantial equity ownership giving outside directors an incentive
to carefully monitor managers. It is also consistent with substantial equity ownership
giving outside directors the ability to insist that corporate decision-making
maximize shareholder wealth. These two effects are related, because there is little
incentive for outside directors to monitor managers if they lack the power to
discipline, but are also distinct. Whether outside directors’ financial incentives or
‘‘clout’’ is more important has implications for what corporate governance
characteristics are optimal.
The clout associated with outside director ownership likely depends on both
outside and inside director ownership — outside directors that own 10% of a
bank’s stock are unlikely to wield much influence if inside directors own 35% of
the bank’s stock. To explore the relative importance of outside directors’ direct
financial stake and their clout, we introduce a dummy variable that indicates
whether outside directors own more equity than inside directors. When included in
the logistic regression along with the level of outside director ownership, this
dummy variable should capture at least part of the importance of equity ownership
giving outside directors additional clout. At the same time, the level of outside
director ownership continues to measure these directors’ financial incentive because
outside directors’ financial incentives do not depend on inside directors’
relative stake in the firm. If clout is an essential ingredient for outside directors to
force their bank to accept a lucrative takeover bid, the dummy variable should#p#分頁標題#e#
receive a significant positive coefficient in the logistic regression. Additionally, if
providing clout is the primary reason outside director ownership is strongly related
to banks becoming targets, including the dummy variable should greatly attenuate
the coefficient received by the level of outside director ownership.
Regression model 5. in Table 6 reports the results including both the level of
outside director equity ownership and the clout dummy. Both receive coefficients
that are significant at the 0.10 level, although the clout dummy barely meets this
threshold coefficients0.729, ps0.096.. The coefficient received by outside
director ownership falls slightly when the clout variable is included, and continues
to be significant at better than the 0.01 level. This is in contrast to the clout
variable’s coefficient  p-value. falling from 2.805 0.005. to 0.729 0.096. when
the outside director ownership variable is included regression with clout dummy
alone not reported.. We interpret this as evidence that outside directors’ direct
financial stake in the firm plays the dominant role. However, clout also appears
important: the implied probability of a takeover when outside directors own 10%
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 157
and inside directors own 20% is 0.167 while the implied probability of a takeover
when outside directors own 10% and inside directors own 5% is 0.270. This
dramatic difference in implied probabilities is despite the fact that once the clout
dummy is included in the regression, inside ownership receives a positive coefficient.
It is somewhat surprising that higher inside director ownership is not associated
with a bank becoming a target. Although it is unlikely manager and shareholder
interests can be perfectly aligned, because insider ownership makes managers
shareholders, it should have some impact on managerial incentives. To see
whether the lack of a relation between insider ownership and a bank’s willingness
to sell is sensitive to how insider ownership is measured, we repeat the logistic
regression models using a variety of alternative metrics. However, insider owner-
ship’s unimportance is unchanged if we replace inside director ownership with 1.
top officer ownership D&O ownership minus outside and grey director owner-
ship., 2. top officer ownership plus grey director ownership D&O ownership
minus outside director ownership., or 3. CEO ownership regressions not reported
but available upon request..
It is possible that managerial ownership’s importance is not captured by our
linear regression specification. Morck et al. 1988. propose that managerial
ownership has two effects. Because it makes managers bear part of the costs and
reap part of the benefits. of their decisions, ownership gives managers an incentive#p#分頁標題#e#
to maximize value. However, greater ownership also gives managers more direct
control over the firm, increasing their ability to resist external discipline. This
entrenchment could allow managers to take value-destroying actions without
risking replacement. The net impact of these two effects is unclear.13 Both Morck
et al. 1988. and McConnell and Servaes 1990. provide empirical evidence
suggesting that the marginal effect of increased managerial ownership depends on
the current level. Similarly, Rosenstein and Wyatt 1997. show that how investors
react to inside board member appointments depend on the level of insider
ownership.
Introducing the square of insider ownership into the logistic regression allows
us to explore whether insider ownership has a nonlinear impact on the likelihood
that a bank sells. The quadratic specification produces a significantly negative
coefficient on insider ownership and a significantly positive coefficient on the
variable’s square regression not reported but available upon request.. However,
this result is difficult to interpret. On January 31, 1995, Firstar completed a merger
with First Colonial Bankshares. First Colonial’s inside directors and CEO owned
85% and 69% of the bank’s equity, respectively. This single takeover with very
13 For example, although high insider ownership gives managers financial incentives to act in
shareholders’ best interests, both Mikkelson and Partch 1989. and Song and Walkling 1993. find that
takeover targets tend to have low insider holdings.
158 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
high insider ownership accounts for the significant coefficients when using the
quadratic insider ownership regression specification — without this takeover
neither insider ownership nor its square receive a significant coefficient. Although
there is no reason to delete the First Colonial observation, we are reluctant to
interpret a result that is driven by a single extreme data point.
3.3.2. Other goÍernance characteristics
The logistic regressions also indicate that having an outside blockholder
increases the likelihood that a bank becomes a target the outside block dummy
receives coefficients between 0.630 and 0.697 with p-values between 0.036 and
0.064.. Consistent with Shleifer and Vishny 1986., this suggests that outside
blockholders can pressure managers to accept a takeover offer that benefits
shareholders.14 Using median values in Table 6’s third regression model, having
an outside blockholder increases the implied probability of a takeover from 0.089
to 0.154. Having both inside director ownership of 10% and an outside blockholder
increases the implied probability to 0.265. These results are consistent with
Shivdasani’s 1993. study of hostile takeovers, but inconsistent with Mikkelson#p#分頁標題#e#
and Partch’s 1989. evidence on corporate control events. However, it is difficult
to contrast our results with past research on industrial firms for two reasons. First,
regulatory burdens specific to banks make blocks above 10% rare in our sample.
Second, the dearth of hostile takeovers in banking suggests that how blockholders
influence managers differs between banks and industrial firms. Bank-blockholders
likely depend more on ‘‘jawboning’’ while industrial-blockholders are more likely
to explicitly sponsor a often hostile. takeover.
Having an outside-dominated board is not associated with banks becoming
targets this dummy variable receives insignificant coefficients between y0.624
and 0.033.. This suggests that outside directors mere existence may be insufficient
to create a good monitoring system. It appears to be crucial that outside directors
have their own wealth at stake.
In light of past research e.g., Byrd and Hickman 1992. and Brickley et al.
1994., among others., outside director ownership playing a more important role
than having an outside-dominated board is puzzling. We can think of two possible
explanations. The sharp drop in hostile takeover activity during the 1990s has
moved the corporate governance focus to inside control mechanisms. Perhaps this
has caused outside director equity ownership to increase to the point where it has a
meaningful and observable. impact on director incentives.
Alternatively, outside directors’ incentives may be more important in banks
than in industrial firms. It is likely more difficult for bank directors to be truly
independent — virtually everyone is a potential bank customer. Presumably, the
14 This is consistent with Brook et al. 1998. who find that banks with blockholders responded more
positively to the Interstate Banking and Branching Efficiency Act’s passage in 1994.
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 159
likelihood of a director’s loan application being approved is quite high. Therefore,
a bank’s outside directors may value their position more than the typical industrial
firm’s outside directors. If so, when faced with an attractive takeover bid, the
direct financial stake provided by outside directors’ equity ownership may play a
relatively important role by offsetting these directors’ incentives to continue as a
member of the bank’s board.
3.3.3. SensitiÍity to how targets are defined
To discover whether our results are sensitive to the window used to define
takeover targets, Table 7 repeats our Table 6 logistic regression models 1.
through 3. using an alternative window. In these regressions the dependent
variable indicates whether the bank becomes a target of a takeover both announced#p#分頁標題#e#
and completed in 1994 or 1995. Reducing the takeover window lowers our results’
statistical significance, but does not change their basic tenor. Outside director
ownership continues to be strongly related to a bank becoming a target. The
existence of an outside blockholder continues to be associated with banks being
acquired. Insider ownership and having an outside-dominated board continue to be
unrelated to banks becoming targets.
Table 7
Logistic regressions explaining which banks become targets within 2 years
Logistic regression models estimating the likelihood that a sample bank becomes a target in a takeover
that is both announced and completed between 1r1r94 and 12r31r95. All characteristics are
measured as of year-end 1993. q is calculated as total assetsqmarket value of equityybook value of
equity. divided by total assetsqmarket value of securitiesybook value of securities.. Average branch
size is calculated as total assets divided by number of branches. Inside directors are all directors that
are also officers of the bank. Outside directors are all directors that are neither current nor past
employees of the bank in any capacity other than as a director. The outside board dummy is set to one
if the majority of directors are outsiders. The blockholder dummy is set to one if there is a 5% owner
with no affiliation to the bank other than their equity ownership. Coefficient p-values are in
parentheses.
Model 1 Model 2 Model 3
Log of total assets. 0.106 0.339. 0.21) 0.099. 0.273) )
q-ratio y7.34 0.111. y6.1 0.168. y6.578 0.159.
Log of Average branch size. y0.548) 0.071. y0.756) ) 0.026. y0.799) ) 0.024.
Outside board dummy y0.091 0.825. y0.391 0.368.
Blockholder dummy 0.608) 0.093. 0.569 0.116.
Ownership by all directors 1.784 0.143. –
and officers
Ownership by inside directors – 0.25 0.881.
Ownership by outside directors – 6.075) ) 0.020.
Number of observations 305 292 291
Pseudo-R2 3.10% 4.90% 6.20%
)Significant at the 0.10 level.
))Significant at the 0.05 level.
160 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
As an additional sensitivity test, we repeat the Table 6 logistic regressions using
only the geographically matched sample of 122 banks reported in Table 5 detailed
regression results available upon request.. All of our important results maintain,
except that the q-ratio is no longer consistently significant it continues to
consistently receive a negative coefficient.. In particular, greater outside director
ownership continues to be associated with target banks at better than the 0.01
significance level.
4. The cross-section of target takeover gains
Target takeover gains provide an additional opportunity to explore the relation
between board ownership and firm value. Cotter et al. 1997. show that target#p#分頁標題#e#
shareholder gains are larger when a majority of the target’s board are outsiders,
but find no relation between outside director ownership and takeover gains. Table
8 provides the cross-sectional relation between target announcement returns and
various corporate governance characteristics in our sample of bank mergers. All
variables are measured at the end of the year prior to the takeover announcement
e.g., if a bank is bought in 1996, we measure governance and financial character-
istics as of the end of 1995.. Consistent with Cotter, Shivdasani and Zenner
Table 8
Cross-sectional regressions explaining announcement day target returns
OLS regression models relating target announcement returns to target characteristics for the 61 sample
acquisitions announced and completed between 1r1r94 and 12r31r96. All characteristics are
measured at the end of the year before the takeover announcement. Adjusted ROA is the difference
between the target bank’s return on assets and the full sample’s aggregate return on assets for the same
year. Outside directors are all directors that are neither current nor past employees of the bank in any
capacity other than as a director. The outside board dummy is set to one if the majority of directors are
outsiders. The blockholder dummy is set to one if there is a 5% owner with no affiliation to the bank
other than their equity ownership. Coefficient t-statistics’ absolute values are in parentheses.
Model 1 Model 2
Intercept 0.325) 1.76. 0.310 1.62.
lntotal assets. y0.011 0.81. y0.011 0.76.
Adjusted ROA y0.109 0.45. y0.100 0.39.
Outside board dummy 0.102) 2.01. 0.119) ) 2.28.
Blockholder dummy y0.103) ) ) 2.75. y0.102) ) 2.68.
Ownership by all directors and officers y0.200 1.57. –
Ownership by inside directors – y0.229 1.65.
Ownership by outside directors – y0.231 0.88.
Number of observations 55 55
F-statistic 3.20) ) 2.78) )
Adjusted R2 16.4% 15.5%
)Significant at the 0.10 level.
))Significant at the 0.05 level.
)))Significant at the 0.01 level.
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 161
CSZ., target banks with an outside-dominated board respond more positively to
the takeover announcement than target banks without an outside-dominated board.
Although this variable’s coefficient is smaller than in the CSZ regressions
0.10–0.12 versus 0.19–0.20., it is proportionally similar the average announcement
return in our sample is less than one half the announcement return reported
in CSZ. and statistically significant at the 0.05 level.
Ownership by all officers and directors receives a marginally significant
negative coefficient in the Table 8 regressions. Replacing total ownership with
inside and outside director ownership, however, shows that although both owner-#p#分頁標題#e#
ship variables receive a negative coefficient, only insider ownership is weakly.
statistically significant.
The unaffiliated blockholder dummy receives a significant negative coefficient
that implies having a blockholder is associated with roughly a 10 percentage point
lower announcement return. It is unclear why outside blockholders would be
associated with lower announcement returns. Given the blockholder variable’s
significance in the logistic regressions, it is possible that bids for banks with
blockholders are partially anticipated, leading to lower announcement returns
Malatesta and Thompson, 1985.. If so, outside blockholders do not reduce
takeover gains, their presence simply causes part of the gains to be incorporated
into the bank’s stock price before the actual takeover announcement.
This anticipation argument could also explain why outside director ownership is
not associated with higher announcement returns — the positive impact of outside
director ownership on the takeover premium would be offset by investor’s
increased anticipation of a takeover. Under this interpretation, the lack of a
negative relation between outside director ownership and announcement returns
implies that outside director ownership increases the target’s takeover premium,
although the relation is masked by the fact that investors recognize outside director
ownership’s importance. This causes part of the takeover gains to be incorporated
into the bank’s stock price before the actual takeover announcement, and, consequently,
no statistical relation between outside director ownership and target
announcement returns. If this is the case, our results suggest outside ownership
increases the likelihood of a takeover while outside directors, through both their
presence and ownership, increase the target’s takeover premium. However, further
research on the relation between investors’ anticipation of a takeover, corporate
governance characteristics, and announcement returns is necessary before we can
confidently interpret the cross-sectional distribution of announcement day returns.
5. Conclusion
The banking industry’s recent consolidation provides an excellent opportunity
to evaluate which governance structures tilt corporate decision-making toward
shareholder wealth maximization. Technological advances and deregulation have
shifted the optimal number of banks downward, creating pressure to merge. Bank
162 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
shareholders receive a substantial premium if their bank is acquired. Conversely,
announcement returns suggest that a typical bank merger hurts the acquiring
http://www.400190.com/dissertation_writing/Finance/#p#分頁標題#e#banks’ shareholders. Bank managers, however, have an incentive to make their
bank one of the surviving institutions. Because shareholder and manager interests
diverge, which banks end up becoming targets provides valuable evidence about
corporate governance efficacy.
Using the universe of available publicly traded banks at year-end 1993, we find
that almost one-fifth are acquired in the next 3 years defined as the merger being
announced and completed between 1r1r94 and 12r31r96.. Target shareholders
receive a substantial premium in these takeovers: targets rise an average of over
18% at the merger announcement net of market.. Bidders, however, drop an
average of more than 2% at the merger announcement.
Using a variety of empirical methods, we document a strong and consistent link
between outside director equity ownership and banks becoming takeover targets.
Because agreeing to a merger systematically benefits target shareholders, outside
director ownership appears to focus decision-making on shareholder wealth maximization.
We also find a greater frequency of outside blockholders in the banks
that become targets, suggesting that large non-director shareholders can also
encourage banks to act in shareholders’ best interests.
Our results provide an interesting contrast to the existing literature. Although
outside ownership is associated with banks being acquired, neither insider ownership
nor outside-dominated boards are consistently associated with banks becoming
targets. Although past research has had limited success predicting successful
takeovers using ownership variables, the bulk of the evidence suggests that
substantial insider ownership impedes takeovers. In our sample, inside ownership
inhibits takeovers only when insider director ownership exceeds outside director
ownership.
Past research using industrial firms tends to find that having an outsidedominated
board improves corporate decision-making, but little evidence that
greater outside director ownership is associated with shareholder wealth maximization.
We find the reverse in our sample of bank mergers. At the same time, we are
able to replicate Cotter’s et al. 1997. result that firms with outside-dominated
boards tend to respond more positively to takeover announcements. This divergence
of results suggests that outside ownership and presence of outside directors
may play discrete roles — outside ownership helps put the firm ‘‘in play’’ while
outside directors make sure shareholders receive the largest premium possible
once merger negotiations begin. Future studies will need to explore these potentially
different roles more carefully.
Acknowledgements
We are grateful to the editor, Michael Ryngaert, and two anonymous referees
for comments that greatly improved the paper. We also appreciate comments by#p#分頁標題#e#
Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164 163
seminar participants at the 1999 FMA meetings. This research was supported by
the Leavey School of Business at Santa Clara University and the Dean Witter
Foundation.
References
Brickley, J., Coles, J., Terry, R., 1994. Outside directors and the adoption of poison pills. Journal of
Financial Economics 35, 371–390.
Brook, Y., Hendershott, R., Lee, D., 1998. The gains from takeover deregulation: evidence from the
end of interstate banking restrictions. Journal of Finance 53, 2185–2204.
Brown, S., Warner, J., 1985. Using daily stock returns: the case of event studies. Journal of Financial
Economics 14, 3–31.
Byrd, J., Hickman, K., 1992. Do outside directors monitor managers? Evidence from tender offer bids.
Journal of Financial Economics 32, 195–222.
Cotter, J., Shivdasani, A., Zenner, M., 1997. Do independent directors enhance target shareholder
wealth during tender offers? Journal of Financial Economics 43, 195–218.
Craig, B., Dos Santos, J., 1996. Performance and asset management effects of bank acquisitions.
Federal Reserve Bank of Cleveland working paper.
Harris, E., 1994. Why one firm is the target and the other the bidder in single-bidder, synergistic
takeovers. Journal of Business 67, 263–280.
Hadlock, C., Houston, J., Ryngaert, M., 1999. The role of managerial incentives in bank acquisitions.
Journal of Banking and Finance 23, 221–249.
Hasbrouck, J., 1985. The characteristics of takeover targets. Journal of Banking and Finance 9,
351–362.
Hermalin, B., Weisbach, M., 1988. The determinants of board composition. Rand Journal of Economics
19, 589–606.
Holland, D., Iscoe, D., Waldrop, R., Kuta, W., 1996. Interstate banking — the past, present, and
future. The FDIC Banking Review 9, 1–17.
Houston, J., Ryngaert, M., 1994. The overall gains from large bank mergers. Journal of Banking and
Finance 18, 1155–1176.
Jensen, M., Ruback, R., 1983. The market for corporate control. Journal of Financial Economics 11,
5–50.
Lorsch, J., MacIver, E., 1989. Pawns or Potentates? The Reality of America’s Corporate Boards.
Harvard Business School Press, Boston, MA.
Malatesta, P., Thompson, R., 1985. Partially anticipated events: A model of stock price reactions with
an application to corporate acquisitions. Journal of Financial Economics 14, 237–250.
McConnell, J., Servaes, H., 1990. Additional evidence on equity ownership and corporate value.
Journal of Financial Economics 27, 595–612.
Mikkelson, W., Partch, M., 1989. Managers’ voting rights and corporate control. Journal of Financial
Economics 25, 263–290.
Morck, R., Shleifer, A., Vishny, R., 1988. Management ownership and market valuation: an empirical#p#分頁標題#e#
analysis. Journal of Financial Economics 20, 293–316.
Palepu, K., 1986. Predicting takeover targets: a methodological and empirical analysis. Journal of
Accounting and Economics 8, 3–35.
Prowse, S., 1997. Alternative methods of corporate control in commercial banks. Journal of Financial
Research 20, 509–527.
Rosenstein, S., Wyatt, J., 1990. Outside directors, board independence, and shareholder wealth. Journal
of Financial Economics 26, 175–192.
164 Y. Brook et al.rJournal of Corporate Finance 6 (2000) 141–164
Rosenstein, S., Wyatt, J., 1997. Inside directors, board effectiveness, and shareholder wealth. Journal of
Financial Economics 44, 229–250.
Shivdasani, A., 1993. Board compensation, ownership structure, and hostile takeovers. Journal of
Accounting and Economics 16, 167–198.
Shleifer, A., Vishny, R., 1986. Large shareholders and corporate control. Journal of Political Economy
94, 461–488.
Song, M., Walkling, R., 1993. The impact of managerial ownership on acquisition attempts and target
shareholder wealth. Journal of Financial and Quantitative Analysis 28, 439–457.
Stevens, D., 1973. Financial characteristics of merged firms: a multivariate analysis. Journal of
Financial and Quantitative Analysis 8, 149–158.
Weisbach, M., 1988. Outside directors and CEO turnover. Journal of Financial Economics 20,
431–460.