Do independent directors enhance target shareholder wealth during tender offers?

Journal

oi Financial

Economics

43 (1997)

195

218

Do independent directors enhance target shareholder wealth during tender offers? James F. Cotter*“,

(Ruxtvcd

b&y wwdr. InIxtitoting

Corporate

JEL chsijkdrm:

We thank

Aaup

Anil Shhdasanib.

November

govzmaxc

1994: find

version

Marc Zennef

Iuxncd

Board ofdirectors;

Taxkr

May

I9964

o&n;

Takawcrs

Board

G32 GW J33; Ll4

Agnwd,

Oyvind

Botum

John

Ajay Kborrar. Gordoa philtip Robmtahnano. Hawi Save lknnis Sk&m. !hoil Wahah partmpats at the AFA and FMA meetings 8nd the Univmity of Glgafy. Duke University. the Claivcfsity d Iowa Midtigan State sdmoldMuupacaSPurduc University, the Uaivcnity UT North Chdim. the Nwqim Univasity. cht !kxkboh zidtool of Eumomb, tk Univasdy d Tiih and tbc Umivmity d ViI-gihtkUKZ siadursForMpfld-Riurdraftsdtbispaper~entitkd7htclkccd bad compositim and iocmtiva on tht tender o&r pmad.

Laura IA. Ken Martin. Wayae Mitkdson

0304-4OSX/97/slS.O0 jf: PII SO304-405X(96)00886-0

1997

Elsevim

Byrd

(the

rdrrrt~

David

Den&

(he cdilor). Ndl Minow. Tod fkwy.

Scima

SA.

All tights

rcscrvd

196

J.F.

Cc trer et al. /Journal

o/Financial

Economics

43 (1997)

!95

21X

1. IntNuluetioa Corporate boards are charged with overseeing and rewarding managers. Several authors have suggested that outside directors, particularly those with no ties to the company other than their dif-tutorship, play an important rok in corporate governance.’ When certain events, such as a firm’s response to a tender offer, create a conflict of interest between shareholders and managers, the ovcnight proby outside directors assumes paramount importance for shareholders. We examine the rok played by independent outside directors when firms become the target of tender offers. Tender offers can have very different eNects on the wealth of shareholders and managers. While target shareholder gains are usually large in successful tender off& target firm managers can sufler significant losses in compensation and other control benefits and thus often try to defeat such takeoverrz Tender o&s are also one of the few corporate events that threaten the tenure of outside directors. Outside directors might be r&ctant to accept 9 ten&r offer if its success would jeopardize their position on the boq 6.3 Thus. exnmining tnder ,tTer~can provide evider.x on whether outside directors enhance shareholder gains even when their directorship is threat&. Finally. by examining the teuder o&r process, we can provide direct evidence on the sharebolder-wealth amsequcnces of takeover resistance, and can evaluate whctber the value oftbow resi6tana 6trategk6 varies according to the composition of the target’s board of directors Our analysis focuses on ir&pe&mt outside directors, whom we define as nonemployee directors without existing or poteutial ties to the firm other than their directorship. We consider a board to be indepe&ot when imkpen&nt directors occupy at least one-half of the board seats. If imkpemknt outside directors perfom a monitoring role. as argued by Fama (1980) and Fama and Jensen (1983). independent boards are mote likely to make decisions consistent with sharehokkr-wealth maximization. Thus, we expect that shareholder gains from tender o&s will be greater for targets with indqet&nt boards than for other targets.

‘Sa Brkkky ad J~QCS (1987). Wcisbmch Bricey. Cole% sod Terry (19%

(19B8).

Byrd

md

Hickman

(1992).

Shivdasani

'Jcmm d Rubwk (1983) aal Jarrdl, Bhckky. and Netter (1988) provide suaxshltenrkralEarEvidemxonohrcbddcrloPdain cvi&occoo -pinrio umocecsful otkm is pmwidcd by Brdky. Desai, sod Kim (1983). Mikkdsom DMO rod DcAo&o (1966). Rubock (wm). and cotter alMl2!cnm (1994L (1oM~Ma~inaadMcConadl(l99l~Cot~znd~m(l994).mdAgrawal~ doalmuuEbuul~bwesiot8kcwvcramcatr ‘KG.

Kmcaw.

and Mim board

tion of tk ~rgc

(1995) docummc io the r&&on

or

tit

suaxssful

takcovcrs

typwxlly

d the number Otoutside directors

(1993).

an ovawicw

and d

the

and Rub& (1985). W&ling and Long Walkling(l!FM) tuuh

in tbc dimina-

J.F. Cotter et al. IJoumal of Financial Economics 43 (1997) 195-218

197

We study 169 tender offers over 1989 through 1992. Controlling for target firm and tender ofier characteristics, we find that targets with independent boards experience higher shareholder gains from the inception of the offer to its resolution than do other targets. The regmssion estimates indicate that the target shareholder pins from tender offers are about 20 percentage points greater when the board is independent. suggesting that independent outside directors perform a statistically and economically significant value-enhancing rok during tender ofkrs. Further. our results suggest that these higher target sharchokkr gains come at the expense of the returns to bidder shamhoMers. We further d ocumeat that the larger taqijet shambolder returns associated with independent boards are a result of both higher initial tender offer premiums and greater r,Gsions in the initial bid premium. We ako find that target s&a&older returns in resisted oft&s am jpater when the board is independent than when it is not. In addition, oKers to targets that have both poison pills and independent boards lead to substantially higher shambolder gains than offers to targets with pcison pills but without independent boards. These results corroborate evkknce presented by Brickky. Cda. and Terry (1994) who find that poison pills adopted by imkpemknt boards are associated witb a mom favorabk stock-price reaction than are poison piJk adopted by nonimkpemknt boards. Cdkctively. our results indicate that &en the board is indepcndcnL takeover resistance ax.d poison pills are likdy to be used to enhance shareholder mums, rather than to entrench target managm. The paper is organize.t ds Mows. We discuss our data collection procedure and variabk sekction in .Section 2 We present results on the relation between board composition and shareholder ,ains in !kction 3. Section 4 examines the relation between tender otTer characteristics and board composition and Section 5 concludes.

2.1. SornprP consmmion

We scamh the Dow Jones News Retrieval database of Wall Street Journal (WSJ) abstracts during 1989-1992 for articks on tefder offers Our search yields a sample of229 tender offers some of which are reported as early as 1988 or compktcd as late as 1993. Certain observations are elimisated. We ehminate observations in which data on tender offer characteristics or the target are unavailable from the WSJ or proxy statements. We eliminate utilitrcs hecause their board composition. ownership structure, and other aspects of the tender offer process are a&ted by regulation. Thus takeovers of utilities are not comparable to takeovers of industrial firms. We eliminate foreign firms because information on the target firm is usually iackinp and firms in financiai distress because directors,

198

J.F.

Cotter

et al. IJournal

of

Financial

Economirs

43 (1997)

195.-2IX

managers, and other stockholders have less influence on the tender offer process when the target is in financial distress. We also eliminate firms that announce publicly that they are for sale because the role of their outside directors during tender offers may not be comparable to directors’ role in tender offers that are not initiated by the target firm. This selection procedure yields a sample of 169 tender offer targets that are traded on the NYSE, AMEX, or NASDAQ. We use the WSJ to determine the first report of any takeover activity related to the tender offer (the rumor date), the first announcemen t of the off&, the resolution date. and other tender offer characteristics. We consider that a tender offer is resisted if the WSJ reports that the target firm will resist the offer. We cc,nsider that an offer is successful if the WSJ reports that it has been completed or if the target is subsequently delisted. Most of the tender offers in our sample are for 100% of the target’s outstanding shares.

2.2. Te&r

o@r premiums and shareholder gains

We compute three different tender cffcr premiums. First, the initial tender 9ih ;~-en!:l;n 15 txI;.,:urcc 2:. th, uutial tender ol%r price minus the pm-tender offer stock price diGdal by the pre-tender oger stock price. The pre-tender o&r stock price is the target’s price 30 calendar days prior to the rumor date if arumorisprrocn&or~Ocskndardrrppriortothefirstdayolthctcndero&r announcement if there is no rumor. The choice of a 30day pm-tender of!& period attempts to account for the market anticipation of the tender of?& tht hasbeend ocwnated in previous studies (e.g.. Jarrell and Paulsen, 1989). In three w we use a procedure outlined by Walkling (1985) to adjust the premium to refiect the fac4 that the bidder is not seeking all of the target’s shams outstanding. Second we compute the premium revision as the tender otfer price at the resolution of the contest minus the o&r price at the initial announcement. divided by the ofkr price at the initial announament. Finally, we compute the targel shareholder gain over tbe entire contest period. For successful tender ofI&, the tar@ shareholder gain is the final tender OR& price minus the pre-tender o&r stock price. divided by the pmtender o&r stock price. For unsuccessful offers, the shareholder gain is the stock p* 90 days after the announcemcn t that the offer &hasbeen withdrawn minus the prc-tender olkr stock price, divided by the pre-tender o&r stock price. We also calculate the target shareholder gain for unsucce&ul offers by using the stock price 180 days, 270 days, and 360 days after the resolution of the o&r. These tests yield qualitatively similar tesults and are not reported. 2.3. Measures of board composition

Our analysis focuses on the role of independent outside directors. We define independent outside directors as directors who are not current or past

J.F.

Cotter

et al. i~ournal

qf Financial

Economics

43 (1997)

195. 218

199

employees of the corporation, do not have substantial business or family ties with management (as indicated in the proxy statement), nor have potential business ties with the firm. The last criterion excludes directors who are employees of banks, law, and consulting firms. Following Brickley. Cola and Terry (S94). we consider boards to be independent if independent directors comprise at least 50% of the hoard. We define inside directors to include a!! directors who are full-tune employees of the firm. Finally, we de& as ‘gray’ al) directors who are either former employees of the firm or are affiliated with managers, &use of current or potential future business or family ties We also measure the financial incentives of imkpemknt outside directors by using their percentagcb of equity ownership. We expect higher independent outside director equity ownership to enhance shareholder wealth during tender offers. Shivdasani (1993). for example, documcn ts that imxeasd ownership by outside directors reduces the likelihood of bcouming the tar@ of a disciplinary takeover. Fama (19fSO).Fama and Jensen (1983). and Milgrom and Roberts (1992) argue that reputation e&ts car provide outside directors with incentives to monitor managers Kaplan and Reishus( 19yo) and Gikon (1990) provide evidence on the importance of reputation capital in the directorship market- Thus, we expect that directors who have reputation capital at stake wiU be better monitors We use the number of additional outside directorship hekf in other corporations as a proxy for the reputation cap&r of indqemknt outside directors We exclude directorships in the target firm’s s&sidiaries. in firms in which the it&per&n1 outside director is ako a full-time employee, and in pr&ss&tal associations We also determine if any of the bidders directors serve on the target’s board. The ptesence of such interiocking directorships can a&t target shareholder gains lor at feast two reasons. First., such directors have a Mtciary duty to both the shareholders of the target and the bii thus creating a conflict of interest. !kcond. interlocking directorships can reduce the information asymmetry hetween the bidder and target and discourage other potential bidders from making a bid We collect all data on board composition from the proxy statement immediately preceding the 6rst tender offer announcement.

To evaluate the impact of independent ouiside directors on brget shareholder wealth we estimate regre&ons with the target shareholder gain as the dependent variabk and the board composition measures as explanatory variables In each of these regressions. we include seven control trariabks that can influence shareholder gains regardless of b.-ard composition. The first control variable is the logarithm of the equity market value of the target firm 30 dap prior to the initial tender offer. This variable is included because larger firms tend to have larger boards, lower managerial ownership and lower fractions of inside

200

J.F. Goner

et al. fJownal

ojFinancial

Economics

43 (1997)

195-218

directors on the board. Larger firms can also have access to more resources to thwart takeover attempts. We control for the presence of poison pills because Malatesta and Walkling (1988) and Ryngaert (1988) show that these can increase the board’s ability to defeat a tender oger, resulting in lower shareholder gains. Alternatively, as suggested by Comment and Schwert (!995), poison pills can enhance the boatd’s bargaining power and increase the gains to target shareholder We also control for the pmaence of golden parachutea because they can mduce manag&al compensation loaues from takeovers and can afRct manawl incentives to resist an of&r and bargain for premium revisions. Data on the presence of golden parachutes and poison pills are collected from the proxy statements preceding the tender o&f, the WSJ, Lexis/Nexis, and Corporate Control Alert. We control for managerial stock ownership because it may afkct managem’ iuccntives to resist a tender offer. Mikkelson and Partch (1989) and Cotter aod Zenner (1994) show that, with greater share ownership, managerial gains from a tender o&r are Iargc; and mar4dgerial resistance less likely. We collect data on nwne: ,hip of inside directors i ncluc .ng options exercisable *Ahin 60 days, from the proxy statement prror to the tender o&r. Shkifer aod Vishny (1986) and Barclay and Ho&mess (WI) show that affiliated Mockholders typicaRy side with top management in corporate control contest& wbefeu ownership by tmaf5hated blo&hdm can facilitate changea in cootrol. Accorditt~~Iy, we control for the ownership of both atliliated and unaffilirted Moddrddcrs. We ddioe affiliated block ownemhip to include ownemhip by family trusts, company stock ownemhip plans+ and retilement plaw.IanotiicttoCthefirmisdeJcribedinthtproxystrtemcntasbciagaa officerortrurtetolablodr.tbeathc~isaLsoi~udedintheaftiEnted Mockholder category. Because block and nooblock ownership by inside direcran is cona&md separately, this category does not inch& inside director ownership. All remainiog Mockholders are clasai6ed as unafRhated blockholdtn. We colkct blockholder ownership data from the proxy statement prior tot&lendero&r. Hermalin and Wetsbach (19%8,1995) argue that poorly &orming firms are more likely to oomioate outside directors to the board. If takeovers of poor performers geoerate higher shaW gainr. tbeo failing to control for performalKx am mute a deceptive relation between sharehoMer gains and board indepcndebcc. Ahernativdy. Byrd and Hickman (1992) and Brick&y, Coka, and Terry(1994)suggeatthatbetter manaps may Corm boards with mote outsiders If better managem also generate higher takeover gaim, then using firm performance as a measure of managerial quality can help assure that the relation betwceo takeover gains and board imkpabe does not reflect an omitted ~qualityc&ct.Tocoatrdfortber#possiMee~,wtiadudetbe target’s industry-adjusted operating return on amets over the three years prior to the o&r as an indepenbt variable. We obtain similar results if we use operating returns on assets unadjusted for industry performance, compute

J.F. Coffer

er ol.~Jm~mal

qfhumcia~

Ecomaics

43 (1997)

195-218

201

performance over one year prior to the offer, or use the market to book ratio of equity as a measure of performance. Performance data are obtained from Compustat.

2.5. Lhcriptiw

statistics

Tabk I reports summary statistics for the 169 tender o&s in our ram* and compares the 47 targets with independent boa.& to the 122 tam without kkpodent boar& As shown io panel A. the sampk contains 79 (47%) tender offers tbat are initially resisted by the target, 47 (28%) o&m with muhipk bidders, aod 127 (‘;5%) ofhs that eventually sucoead. Of the targets with an depeodat board 60% resist the tender ofkr. comparai to only 42% of the targels witbout an iodqdeat board. Statirticsoathe!endcrdkrprrmiumsarrrrportediopanclB.TaMeI.Whik the initial premium is 53.7% for targets with an i&pendent board and 44 8% for targets without an independent board, the di&cncc between the two subsets is oat statistisally significeot. Targats with an imkpemknt hrd experiauz higbupmiumrevisiocs,however.tbaotargelswilhoutaniodepcbdcatboard (129% versus 6.4% with a pvalue of 0.02) Finally. the shareholder gain from the inception of the ofkr tti its resolution is 623% for m with an indcpardeot board ampand to 40.9% for targets without an indqemknt board This diacrcmx is also statistically igaificrot, with a p-value of 0.02 Thus, targets withanin&pemkntboardobtainbigherpranium revkiomaodexpcricwt largershareholdergainsthantar@switboutanimkpankntboard. PllaelC,Tablelreportsthatthtamagcequityrrmrtavalucoltht~ firmsisS60tmillion.~medianmaflrtlvrlutdequityol~withul imkpas&nt board is :!ighdy larger (5 168 million) Tao other jiynpk firms ($124 million). Furthermore, t& pmsutcc of poison pills and golden parachutes differs sigoiBauJtly buweco targets with and wilbout an irbpedmt hoard. of i!E largets without an iodepemknt board 48% have poisoo pills and 57% have go&k0 parachutes, versus 68% and 77%. rcspectivdy. far the targets with an iodqmdmt board. hide diree!ors d taqpted dnns owo 129% of outstanding shams in the full sampk. However, in targets with iodcpadenr boards they own only 4.7% compared to 16.0% in targets without au idqencknt board. Both t-testsand Wikoxoo tests indicate that this difkrence is highly signiscaat. We do not 6nd a sign&ant differcna in Mock ownership and brdder toeholds between firms with and without indepc&nt boards. Board characteristics are reported in panel D. Tabk 1. The average (median) board size for the whole sampk is 8.8 (8.0) Targets with an indqemknt board have larger boards with a mean (median) of9.6 (9.0) directors compared to 8.4 (7.5) for firms without an independent board. lmkpen&nt outside directors have a larger equity stake in the target when the board is imkpcmknt than when it is not (3.6% versus 0.7%). 10 addition, the median number of additional

of the tender

otEr

samplz

and comparison

ul targels

with

and without

an ~~~depcndent

board

o&r

0lTer

._

_ _.

prrmhuns

.__

C:

Ownership

sfrucfure

ad

_

other

/Irm

__~ ____._ - .__._. __. -Market value of common equity’ (SMJ Tar$ct firm has a poison pill in place 0’0) Target firm exocutivo have a @den parachute 0wncnhip or inside directors’ (%) Ownership of afliliaccd blockholder+ (%) Ownership of unattlbatcd blockholders’ (%I Prior owncrahip of biddof(%)

Pond

_

und ,shurrholdrt

_ ___..__ -- . __ -. .Initial bid pnmiumb (%) Revision ol initial premium’ (%I Target shareholder gaind (%) .-__.__-___-_..._. -.

~___. Multiple bidders Resisted o&n !Gthxcmful tender -._-.-.Panel 8: Ton&r

.

(%I

.

t hurwrtridcs

145.4

601.9 53.3 62.7 12.9 2.0 14.1 11.0 i.3 0.0 10.3 0.0

Median

41.6 0.0 43.9

47.3 8.2 46.9

Mean

Mcdinn

Mean

weadrh @CIS

28 47 73

47 79 127

.-

%

No.

Full sample

55:.2 47.5 57.4 16.0 2.1 14.4 11.9

Mean

44.8 6.4 40.9

Mean

32 31 90

No.

Indqc!d dirators fn = 122) --

7.7 0.0 Il.9 0.0

t ‘3.8

Median

39.7 0.0 .tn. I

Mcdm

3 t: ,!

‘”

‘I( otltsidc M%

13.3 8.7

i.8

Mean - ----. 733.6 68.1 76.6 4.7

53.7 12.9 62.3

Mean

15 28 37

NO.

Median _--_____ 167.8 -. I.9 0.0 9.7 0.0

48.5 0.1 63.6 _.. ---.-._-_-_

Median

32 60 79

%

independent outside directors 3 50% In = 47!

0.42 0.0 I 0.01 0.00 0.76 0.65 0.31

t-ksl p-value

0.15 0.02 0.02

pvalue

0.48 0.04 0.50

I-tea pvaluc

0.02 0.02 0.00 0.46 0.76 0.50

0.06

pvalue

Wikoxon

p-value _..-. 0.20 0.0 I 0.03

0.46 0.04 0.50

Wikoxon p-value

_

Characteristics of tbc ten&~ of&r :ampk and compartson d targets with and without dn iI.1 ~cfmdcnt board. Panel A presents general characteristics of the ten&r otkrs. Panels B and C provide dercri@vc srulistics on tarpt lirm chnractcr&n~ panel II provides information on board composition and mrsbip of the target firms The sample consists of 169 tender o&s announcal in the W. 0’ Street Journal over the I%1992 period with 122 targets without an independent board and 47 targets with an i-1 board.’ _- _._---.-----------______.. __-_---.-w-M -

Tabk 1 &race&iltil.r

0: Board

characrerbrics

sampk

cxcluda

uGliCcs.

foreign

firms,

Median

shareholder bid price

-hip

of inside

The

dircctors.

Inside

directors

own

more

who

are not full-Gmc

number

of additional

outsidc

dircctnnhips

prior

moiurion

Mean -_. .-9.6 10.6 19.8 60.0 3.6 1.3 - ..---

.--

Median _..-_ -.- ._. 9.0

fhm

of Ihc target

ts the avcragc

number

cmployecs of the target to the proxy slt~tcment.

employem

of the contest

employed

o&r.

(Jirccror

cr,rporatc

firm and have no family of additional

3Odays

stock d&cd

minus

dircclorslnfH

rolallnnshlps

or mana@r)

the Wall Lrt-rt

firm.

heforc

the tender

ofkr

Wilcoxon pvaluc _--_.. 0.0 I 0.67 0.00 0.00 0.00 0.09 data are unavailable.

price

end quaI the managcmcnl

managers.

target

60 days

or

dixxtor.

or potcnlial outside

or former

IO 0 otherwise.

ts l bout the tcndcr offer.

target

held hy each indcpcndcnt

with

with

with

within

of&r

Lo the final

at rhe initial

of the tender

and erc effiliatcd

announcemcn of the bidder.

Jcti

firm e.

Optionsexcrcisebk

the first rumor

efflliatcd

of \hc tar&t in footnote

by the tergct

prta

and ore not currently or from

WI

3% of the common insider ownership

is en insider firm.

t-test ,-value .0.01 0.65 0.00 0.00 all 0.39

from 30 days before the first tender o&r rumor of the tender oRcr if the bid is not mful.

the stock

~~ICZ &angc the resolution

the

of the ,arpl

to the tcndcr

*to&

of the target

stetcmmt

1% of the common rhc prory

one dir&or

than

firm ;s from

10 I if at kast

bidding

equal

of directors

variabk

ofthc

blockholders

Median

of the tar& firm for the period from 30 deys kfon the first tender offer t&-i ofkt. to the initial tender oRer bid price. All premium calculations ‘9RV.

arc currently

blockhold~-rs ate individuals or corporations who own more than throufi femily or business rclaltons. This varinblc does not inch&

consists of the ow~rship in th’r pcrant;ryc.

Mean

is qua1 to ~hc numhcr ofshercs outsfand;ns time t ,>I tho tender offer if no rumor is prcscnt

‘Pemntagc ofdircctors who arc not full-tune future business tin with the firm xcordinfg

‘Percentage

‘Indicator

‘Prior

LUnafRlietcd

sAffilietcd manaBet

‘This variabk arc included

p&ice at

eein over the entirc ccndcr o&r contsl is the perccntep if the bid is successful. or 10 the rta( k price 90 days after

of the initial bid pnmium is equal to the tender off& divided by the tcndcr offer price a1 the nn.?ounamcnl.

The merket veluc ofcommon quity 30 days before the initial annou-

+The target tender olkr

The revision ennouncernen~.

‘The initial ~KJ premium is computed PI the pfruntage increase in the stock w rumor or, if FO rumor is prcscnt. 3Odays beforo the initial nnnouncement oftha are adjusted for the pcrcentagc of shares sought by the Cddcr (see Walklinp.

The

Mean

8.tl 8.0. 8.4 73 ! 2.4 13.1 6Y.6 71.4 65.7 66.7 00.0 36.0 33.3 26.8 28.6 %J I.J 0.1 0.7 0.1 0.2 I.2 I.0 1.2 I.0 I.3 -_ _. .__ --__-.---_ ._. -.. .- _. . Arms in finamal distress, those theI publicly announce t!lsy are for sak. and firms for which

-__-_-.. -. _. _. _ _ Number of direclors on the target firm‘s board interlocking directorship f%Y Pcrcentagc of outside direclon (%b’ Percentage of independent outside directors (%)I Ownership of independent outside dimc?on f%) Number of addilional outsidc directorships’

Panel

204

J.F.

Cotter

et al. iJournal

of Financial

Econvwms

43 (1997)

I95

218

directorships is marginally higher for targets with independent boards than for targets without an independent board (1.3 compared to 1.0). Thus. independent boards are larger and have independent directors with more reputation capital and larger equity stakes. Twenty-one (12.4%) of the targets have an outside director who is also a director of the bidding firm. In most of these cases, the bidder’s toehold in the target firm is substantial. The mean (median) toehold is 56.6% (57%) when the bidder is represented on the target board. compared to only 4.5% (0%) for the remainder of the sample. In 17 of the 21 cases the bidder’s toehold exceeds 50%. whereas in another three cases it exceeds 20%. Thus, interlocking directorships appear to he the result of substantial bidder toeholds in the target firm.

We e- amine the effect of boa-r! co qn+tion on shareholdf,r wealth by using (HO mam m&l spccitk;ations. In the first specification. we include two indicator variables for the presence of an independent board and an interlocking directorship. In the second specScation. we include the equrty ownership and ;ht: number of additional directorships of i&epen&nt outside directors To assure that our results are not driven by model specification errors we mnducta~oCspccification~.Firstntest~IhepracnceoCbetetoskedasti&y. Chi-square tests do not, however, reject the assumption of bomoskedasticerror terms throughout our regrc&on specScations second toexamine whether our results are driven by influartial observations we conduct the diagnoslicsoutlinadin&bly,Kuh.aodWctscb(l~~TbesttcstsdobotFmaltht presence of signScant outliers. Third, to investigate whether collinearity afkcts our results+ we examioe correlation coe&ients among i&pemknt variables. Whik these coe&kots are generally not large, the correlation do.46 betweco firm size and tbe number of outside directorships held by in&pen&t outside directors is an exception. We estimate our results exdudiog these variables, aod obtain results similar to those reported. Further. inspection of variance inflatioo factors suggests that multicollinearity is unlikely to inflate the standard errors.

3. I. Multitwtiate

0nalysi.s o/target

We estimate the following

Target skuehoidkr

sharehobr

multivariate

walth gains

regression:

gains

We report the results in Table 2. The most striking result in modei (i) is that the coefficient on the independent board variable is positive and highly

J.F.

Cotter

er al. lJoumal

aJ Financial

tcmomrcs

43 (1997)

195

2/X

205

significant. The point estimnte suggests that shareholder gains are approxim.ate!y 20 percentage poi,tts higher for targets with an independent board than they are for other targets. Thus, board independence lends to statistically significant and economically higher shareholder gains from tender offers. For all models. we find that the cocticnt on the interlocking directorship variable is negative and marginally significant, suggesting that target shareho&r gains are lower when an insider of the bidder serves on the target’s board. White the coe&ient on the poison pill indicator is consistently posrtive. its statistical signiticance is weak. None of the other variables signScantly a5&s &a&older gains. In model (ii) we include the equity ownership and the additional directorships of independent directors as expianatory variabks. Neither variabk is significantly related to shareholder gains The independent board variable remains however, positive and statistically significant. Using data on stock-price reactions of poison pill adoptions. Brickley. Cola and Terry (1994) argue that in&pcmknt boards are more likely to use poison pills to enhance sharchclldcr wealth from takeovers. To test this hypothesis dimctly. we include in mockI (iii) an iotcraction variable between the indcpen&nt board and poison pill JartaMes With this spcci5ation. the coc5icien1 on the poisoo pill dummy is insignificant. However. the intcra&on between the poison pill and in+andent board dummies is positive and significant. suggcsting that i&pendant boards .,re more likely to use poison pills than nonindqco&nt boards to enhance shareholder gains from tender ofkn. To see if the results on poison pills can be generahi to LoLaover resistam strategies, we include in model (iv) a dummy variabk that is equal to one if the o&r is resisted and an interaction variable between the resisted o5cr dummy and the in&per&at board dummy. The marginally significant negative a.nSdent For resisted o&rs and the positive co&cient on the interaction variable suggest that if the board is not i&pcndcnt. resisted 05&s result in lower target shareholder gains However, if the target’s board is indqcndcat. the shareholder gains in nsisted o&rs are larger than for resisted o&s where the target’s board is not indcpmdent. We provide additional evidence on gains in resisted 05crs in Section 4.2 In model (v) we estimate tb regression by using a dummy that is equal to one iftbc tender 05er is suwxssful, and an interaction variabk betwam the successful offer dummy and the independent board dummy. Consistent with prior studies (e-g, Bradley, Desai and Kim, 1983). successful o5ers are associated with target gains that are 55 percentage points higher than unsuccessful offers. More interestingly, however. for targets with independent boards, the shamholder gains in successful ten&r 05ers are an additional 22 percentage points higher. Overall, the positive and signif#ilnt c-oellicient on the independent board dummy variable and on the interactions hetwecn this variable and board resistance, the presence of a poison pill. and tender o&r suazss indicate that i&pcndent boards enhance shareholder wealth during tender o5ers.

sharcholdcr

gains

lndcpcmknt

board

board

offer’

x Poison

x Successful

of afRliolcd

of unal%liated

Ownership

parachute’

Golden

Ownership

pill’

Poison

of market

Logarithm

of equuyl

outside

pill’

offcf

oRcr‘

blockholders”

blockholdcrs”

value

of indcpcndcnt

Ownership directors’

Number of additional outside diraztonhips per independent director’

lndepcndcnt

suaxssful

x ResIsted

board

Indepcndcnl

0%:’

boardb

Independent

Resisted

directorship’

lnmriocking

Interapt

Explanatory variables _ ._-_.- ___. - . .._.

__-.---. -...- --.. ..__. _ _

(0.03

0.20

(0.95) (0.59)

- 0.15

(0.99)

(0.14)

- 0.04

0.001

0.13

co.sn,

(0.12)

-. 0.19

- 0.015

(0.09)

(i)

0.58

Model

0.13

0.02

(0.55)

0.17

-

(0.87)

(0.13)

(0.42)

(0.44)

(0.03)

(0.15)

(0.07)

(0.95)

(ii)

- 0.04

- 0.01

-

(0.03)

0.19

0.18

0.66

Model

-0.10

- 0.08

O.MI

0.07

- 0.02

0.24

-- 0 I !

0.61

Mode’

(0.71)

(0.89)

(0.95)

(0.43)

(0.45)

(0.02)

IO 16)

(0.04)

!ii)

-0.20

- 0.02

- 0.001

0.16

- 0.03

(0.49)

(0.97)

(0.99)

(0.09)

(0.40)

(0.06) (0.09)

0.22

(0. I I )

(0.04)

(iv)

- 0.15

- 0.20

0.76

Model

- 0.25

0.04

0.004

0.12

- 0.01

Model

(0.27)

(0.94)

(0.95)

(0.10)

(0.18)

(v)

-

Regression of the target sharcholdcr gains on the mndsr offer characteristics. the targel firm and board. The targcl shareholder gain is cne percmtctge price change from 30 days before the fh’st tender Olkr rumor IO Ihc final trndcr ot%x bid pr%z If I) I~. bid is sucxxssful. or IO the sloc& price 90 days after the resoWon of the ten&r o&r if the bid is not wc~asful. The sampk consists of 169 tender off& announcrd in the Wall fjlm Journal ovrr Iha l9gg 1992 period. fhe pvalucl are in parcnthcscs and lba v~?i~bk dchnnions JIV con~aincd in the foolnd :cs to this table.

Tabk 2 Analysis of the target

of inside

variable

41ndicator

equal

opcrsting

return

0.06

WI4

- 0.16

0. I?

IO I if a~ Icast one dtrcctor

directors”

of the target

10.361

IO.651

lirm

is an ins&r

0. IO

0.04

- 0.18 (0.291

0.17 10.551

ldtrcctor

variable

owncrshrp

dircctorrhips

‘Averugc

industry-ndjustcd

ownership

blockholdcrs

“Pcrccntagc share in this percentage.

“Un&‘iliated

blockholdcn arc indrviduah through famrly or hummers

“Affihnwd mrnagers

opcraung

return

drrcctors

rno1.c than

of insi&

own

firm firm

t.rrpct\

ducctorr.

au indcpcndcnt

who vnriahlc

on :IJWI*

paruchutc

in place.

of common the initial

cxcrcisabk

and

60 days

m place.

stock

or tk

ycan

arc currently

target

prccodmg

cmployrd tk

tirm

with

tender

directors

IO tk number of ~hc tender

prctnlsgc.

attiliutcd lirm. 0ptrnnscxcrcts;thlc

within

target

60days

managers.

and arc affiliated with

hrm o.

(0.57l

target

are included

with

the stock

director.

by

_.

IO.1 1)

employed

0.35 0.00

of shares outstanding times olkr it no rumor is present.

in this

stock of the target defined m lootnote

o&r.

pill in place. held by each independent

a poison

0.13 - 0.23

outside

und 0 otherwise. are not currently

and 0 otherwise.

is successful.

arc included

and arc not currently by ~hc trrpt

0. IS

0.03

(0.80) (0.22)

0.07 - 0.22

of the bidder,

over.

and 0 othcrwisc.

and 0 othcrwisc.

cquny is equal announutmcnt

within

ogcr taken

dircctonhips

and targets

corporate

board

tcndcr

olkrs.

Indcp;ndent

,tnd rnistcd

IJ .~c~sfuIIy

board

board

own more than 5% ol’thc common dcrs not include insider ownership

hjr the three

tltrcctors

5% or the common Ins&-

pill defense

lirm

of additional

an indcpcndcnt

Options

numkr

with

that III. ~hc target

WII~I an indcpcndcnt

and 0 othcrwisc.

have it golden

has a potson munagcrs

or corporation\ rclatron. Thtc

IO I if the target

cq11.11 IO I if the target

variable

variabk

equal

c.trgcts

is thr *hrcragc ,ufstdc

with

the oger.

rarptr

is su~~~ful.

vanahlc*:

olkr

varuthlc\:

rnivta

variublcs: firms

vuluc n* common equity. Tk market value rumor ~4 the tender offer or JO days kforc

of mdcncndent

outside

‘Indicator

of the market betorc II,C lint

the tender

IWO mdicator

if

IWO indicator

‘indicator

The logarithm price 30 days

‘Percentage

share

of additional

number

“The

hctwccn

to I

hctwccn

equal

IUO tndtcator

IO I II the target

between

equal

variahlc

variable

variable

variable

@Interaction

‘Indicator

‘Interaction

“Indicator

‘Interaction

(0.27)

10.65)

or manager)

O.lM 0.09

- 0.19

0.13

‘Targets with an mdcpcndcnt hoard have OI lcasr 5Wb indcpcndcnt outside directors on the board. the target tirm and do not have current or potcnti.rl future busincsc tics with the target tirm.

p-vahc

R-yuarc

Rcgrebsion

Adjusted

Industry adjusted on awW

Owncnhip

208

J.F.

Cotter

et 01. ikumol

3.2.

The source @target takeover gains

of Financial

shareholder

gains

Economics

43 (1997)

and the combined

195-218

bidder

and target

To evaluate whether the gains to shareholders of target firms with independent boards result from higher total takeover gains, we compute the combined gains to shareholders of both target and bidding firms for 48 successful transactions for which stock price data are available on both firms. We follow Bradley, Desai, and Kim (1988) computing the total takeover gain as the market-adjusted return to the value-weighted portfolio of the target and the bidder from 30 days prier to the first takeover announcement to the final offer date. The mean (median) total takeover gain is 6.57% (3.59%). For 1 I of these 48 transactions, the target’s board is independent. These takeovers result in a mean (median) combined takeover gain of 2.48% ( - 0.75%). This return is not significantly dilferent from the mean (median) takeover gain of 7.78% (3.89%) for the 37 takeovers in which the !arge:-s board is not independent. \Ve use rb came windn .N to comr !I te * Tmnrket-adjusted retu:drs for the 48 SU~IUI budders. t’or these bidders, the mean (median) bidder return is - 5.18% ( - 5.22%). Bidder returns are - 18.65% ( - 15.39%) for the II transactions in which the target’s board is independent, compared to - 1.17% ( - 4.0%) for the 37 uses in which the target’s board is not imkpemknt. Using both t-teats and Wilcoxon tests, these returns are significantly different at the 5% levd. Overall, these results suggest that the higher returns to target shareholders with indepe&nt boards do not result from higher total takeover gains, but in fact come at the expense of the bidder’s returns. 3.3. Sjm+jicution

checks

3.3. I. Altemative measures o/board mmpo~irion Our definition of imkpemknt outside directors does not include directors from banks, consulting firms, or law firms because such directors are likely to have potential business ties with the firm. This classification scheme is consistent with Brickky, Lease, and Smith (1988) and Van Nuys (1993). who find that banks, insurance firms, and nonbank trusts tend to support managers during antitakeover amendment proposals. When we classify such directors as independent outsiders, it reduces the significance of the independent board variable. Brickky, Coles, and Terry (1994) also report that their results are sensitive to the way in which independent outside directors witb potential business ties are classified. We also estimate our tests using alternative specifications to measure board composition. First., we replace the imkpemknt board variable with tbe percentages of independent outside and gray ditectors on the board. We find that the percentage of gray directors is unrelated, while the percentage of independent outside directors is positively related to shareholder gains. though its stat&A

J.F.

Goner

et ol.lJournal

of FinancW

Econonia

43 (1997)

195. 218

209

significance is weaker than the independent board dummy. Second, we follow Weisbach (1988) and classify our sample into three categories: Firms with an outsiderdominated hoard (i.e., independent outsiders constitute more than 60% of the board), firms with a mixed board (i.e., independent outsiders constitute between 40% and 60% of the board), and firms with an insiderdominated board (i.e., insiders constitute more than 60% of the board). With this speciticatiort, we obtain qualitatively similar results for the outsiderdominated board variable, whereas the mixed board variab& fails to enter our models signiticantly. In all tests, he significance of the other variables in the regmssion is unaltered. Brickley, Coles, and Terry (1994) suggest that among all independent outside directors, directors who are retired decision tnskets or who list ‘director’ as their occupation are associated with higher abnormal tetums at the announcement of poison pill adoptions. To examine whether such directors drive our results on independent boards. we add four variahks to measure the fraction of the board members who arc (i) Ja;ision makers in firms other than banks consulting or law firma (ii) private investors in the firm. (iii) educator54 govemmerit otI?cials or clergy. and (iv) who list ‘director* as their occupation or who are retired decision make * We find that none of these measures are related to the target shareholder return. and that the inclusion of these variabies does not alter tbe statistical or econ-rmic significance of the independent board VWihk.

3.3.2. Twlrdds Bidder toeholds can facilitate the tender o&r and reduce the efie&eness of managerial resistance. Conversely. Ho&,-ness and Sheehan (1985) sbow that the accumulation of a toehold in a potential tatget is often the first step in an unsolicited bid and can trigger resistance by the target board. Because of the high correlation between bidder toeholds and interlocking directorships, we do not include the toehold in our basic models. To evaluate the et%@ oftoehoids on shareholder gains, we estimate our regressions by @acing the interlocking ditectorships with tbe bidder toehoid. The co&i&~ on toehold are similar to those on the interlocking directorships variable, but they are gerletally not statistically signi&ztnt. When both variables arc ~nchtded in the regressions. neither is significant. whereas the coefficients on other variables remain qualitatively unchanged. We also estimate our results for the subsample of firms with no

%oard members who codtu:c about lY% 23% (private inwstonl. executives or directors repnxnt independent dent hoards

arc decision maker in firms 1ha1 are MU bank+ law firms, or awlsuIting firms of the board ior the full sampIe. Other occupational atcgorics ca~stitute t otkialo~. and 7.2% (Mired 4.5% (educators. defgy. 8od gowmmm who list ‘dira~or’ as their occupation) dthe board. Baatsc these atcgwies outs&e directors. lhese p~~~~~tages arc higher tbr tar@ firms with i&pen-

21(;

J. F. Carter

er ni. lJournal

o/humciat

Economics

43 II 997)

I95 - 2 IX

bidder toeholds and obtain results similar to those reported. Thus our results are not driven by firms with large bidder toeholds.

3.3.3. Ahematioe measures of directvrs ‘financial incenrices lnslead of the percentage share ownership by independent outside directors, we use a wealth-gain variable similar to that used by Cotter and Zenner (1994) as a proxy for the financial incentives of directors. This variable is the gain on the equity ownership from the tender offer (number of shares times the dollar premium OR&~) minus the present value of the potential losses in compensation (the annual director retainer and eight times the meeting fees). Using this measure yields results that are qualitatively similar to using the perctntagc of independent outside director equity ownership. Thus, the results appear robust lo the measurement of directors’ financial incentives.

.3.4. 71re components qf target shureholdrr gains *t ;i&pcndxt boards iad lo higher target shareholder gains from tender offers. In this section, we investigate the source of these gains by examining how board indepemkna affects the initial ten&r offer premium and the bargaining process after the initial bid. For targets that are suax~~&lly aquirad the initial premium and the subsequent revisions constitute returns eamed by shareholders. Shareholders of targets that are not successfully acquired typically expe6ence negative returns at the o&r’s withdrawal. They may also expcriencc negative Riums subsequent to the ofkr’s resolution. Thus we examine target shareholder returns for hilcd offers over longer post-takeover event windows. Our

-1. .‘lliS

s.J~,:~l

3.4. I. ZAC initial Ie&r

II..

ofl&r pwmium An important decision facing potential bidder3 is the choia of the initial t&r o&r premium. Fishman (1988) shows that, if revisions are costly, bidders will structt;. t the initial bid to reflect expected bargaining by the board and the likelihood of competing bids Thus if in&pen&nt outside dincton are mom likdy to maximize shareholder value, we expect initial bid premiums to be higher for i&pen&at boards The initial bid premium can aJso be the result of intensive bargaining by the target firm’s board preceding the first official tender o&r announcement. Accordingly. if independent boards engage in more intensive or more dkctive pre-bid bargaining. then the initial bid premium can be larger for targets with independent boards. We estimate regrasions similar to those in Table 2 and report the results in mode&(i) and (ii) of Table 3. In both modek the initial bid premium is positively and significantly related to board independence4 and negatively and significantly related to the presence of a biis dieor on the target’s board. The point estimates suggest that the presena of an independent board inrrcascs the initial

J.F.

Cotter

Tabk 3 Analysis of the initial

el al. !JoumuI

lender

o&r

q/Financial

premium

and

Economics

of the post-bid

43 (1997)

IYS

218

premium

revision

211

Regression dthc initial tender o&r premium and the post-bid premium revision on the ~cndcr olTcr characteristia. the target firm and its board The initial tender ofkr pmnium is tk percentage prise change from 30 days before the first tender offer rumor lo the first tender ofler bid pria. The post-bid premium revision is tk percentage pria change from the firxt tender olkr bid plcc to the bid pria o&red at the end d the context. The nmpk consists oll69 tender otTcts announced in the Wall !Strcct Journal over the 1988 .I992 period. The pvalua arc in parentheses and the varixbk dctinitioax are umtaintd in tk footnotcx to Tabk 2 Mrdd:

Initial

or poshi

tender

= / (hoard . harm -.____-

~crisbcs.

o&r kc

pmuinm /irm

characrtistics.

omership stmcruw) _____-___-

hdcpcdmt

variahk

Initial

tender

offer

Explanrtov

bariabks

Model

(i)

Modd

lntcrupt IntcrIockintt

directorshIp

Indcpcn&~

hwrd

Numhcr d additional dirstorshipspcf i&pcndcnt diruztor

Po&on Go&n

Post-M

(Ii)

Model

___ -...-...-_-.

premtum nnl

Modd

0.58 10.01)

0.07 to.481

to.391

- 0.21 to.01 1

- 0.20 lO.02)

0.002 10.95)

0.002 t0.W

0.10 IO.061

0.10 l0.W

0.05 (0.05)

0.05 (0.04

ourside

0.03 IO.221

0.01 to.571 - 0.06 (0.731

0.29

value

- 0.006

Nl.75)

pill

0.06 to.201

parachute

CMncrship of alliluled blodholdcrs

IIVI

0.09

(0.431

of market

rewston

a49 (0.03 I

OIllmhipdiadcpcnJcnt oumdc Lraztors Logarithm of quity

premium

--002 IO.43

- 0.003 (0 721

- 0.001

to.591

006

0.07 W-W

tO.291

0.07 to.01 )

- 0.01 m.7x1

._ c.004 to.x5,

- 0.01 (0.78)

- 0.3x 10.34~

0.02 (0911

- 003 tti.851

Ownmhip of unallilutcd Mockholdcrs

Cl.15 w *I

0.12 (0 w

-0.1; t0, . .19,. .

- 0.12 IO.1 5)

Ownership

0.X (0.09)

0.36 (0.06)

- 0.03 ~O.?Ol

- 0.02 (0.77)

- 0.12 to.261

- 0.15 (O.Wl

- 0.04 (0.45)

- 0.03 to.521

O.Ot 0.01 ___ --___--..----_

0.10 0.01

00% 0.01

of inside

director%

Industry return

adjusted opcratmp on assets

Adjusted Rcgre~ion -_--

R-square P-value

_--__

_-_

._.... - --... -_.

_._ ._-.__.

_.-.._-. .-.-...--. _.

.--_----

. _--

0.07 0.01 -- .---

212

J.F.

Cotter

et ol.,JJounutl

ofFinancial

Economics

43 (1997)

195 -218

premium by IO percentage points. whereas interlocking directorships reduce it by about 20 percentage points. In addition, the initial bid premium increases with the ownership of the target’s inside directors. Variables representing owncrship stakes and reputation incentives of independent outside directors are, however. unrelated to the initial bid premium. Overall, these results suggests that target firms with independent boards receive higher initial premiums. 3.4.2. The rwision of the tender oflet- premium We estimate the same multivariate regression mode’s as in our previous analyses to examine the additional premium tbat is obtained from the bidder(s) during the bargaining process using the revision in the bid premium as the dependent variable. These results are reported in models (iii) and (iv), Table 3. In both models the coetlicient on the independent board variable is positive and statistically significant. The point estimates suggest that the revision in the initial premium is five percentage points larger when the board is independent. In addition. the presence of a poison pill increases the bid premium revision by sch :i px.yr’ !A; L ,>i:>ts. i&r untqorted tests, we also estimate models with interaction variables between poison pills (or resisted ofi&) and an i&pen&t board. We find that the premium revision is higher when the target has a poison pill (or resists the offer)and has an indepe&ent board than when its board is not indcpeodent. None of the other variables are related to the premium revision. Thus, controlling for other factors, independent boards mhancc target shareholder gains over the entire tender ofkr contest. through a higher initial premium and a higher bid premium revision. 3.4.3.

Sharehokkr

retwns

and target jim

actions follnn

ing unsuccess-I

oflkrs

We also evaluate whether the returns experienced by targets of unsuccessful offers depend on board composition. Of the 42 tender offers that fail. IO have an independent board and 32 do not On averagc target shareholder gains (based on the price 90 days after the failure of the o&r) are 6.0% for firms with independent boards and 3.3% for firms without an independent board. These gains are not significantly diikrent from zero. or from each other. We also examine cumulative stock returns from 30 days prior to the initial announcement (or the rumor date) to 180. 270. 360. and 720 trading days after the resolution of the olTer. and do not find dilTerenccs between targets with and without i&pendent boards. Thus. shareholder returns in unsuccessful o&s are not related to board composition. We also follow the unsuccessful targets for three years after the resolution of the o&r. Of the 42 targets. 37 are still traded three yean later, while four have merged or been taken over, and one is delisted after bankruptcy. Of the 32 targets without an independent board, two are eventually taken over. four engage in asset sales, four make acquisitions, six experience CEO turnover, four make seasoned equity o&rings, and five repurchase their own shares. Of the ten

J. F. Cotter

et ul. )IJottnml

of Finattciul

Economics

43 (1997)

I95 - 21X

213

targets with an independent board, two are eventually taken over, three engage in asset sales, and two make seasoned equity offerings; none make acquisitions, experience CEO turnover, or repurchase their own shares. Thc.~ results suggest that unsuccessful targets with independent boards are more likely to sell assets or be acquired later, and less likely to make acquisitions or replace managers. The small size of the unsuccessful tender offer sample, however. precludes a statistical analysis of the differences in subsequent actions between targets with and without independent boards We conclude that the difference in shareholder gains between targets with and without indepentient boards are driven by the higher initial premiums and premium revisions for targets with independent boards, and not by ditferences in tetums for targets of unsuccessful tender offers.

Once a firm receives a tender offer, its board can resist or accept the offer. seek additional suitors and attempt to influence the outcome of the o&r. In this section, we discuss regrc sion results that examine whether board composition a&cts the initial target reaction to the tender offer, the Iikdihood of multiple bidders for the target and thc ultimate outcome of the offer. We also compare the shareholder wealth et%% of resistance by indepeaknt and nonindependent hoards. For brevity. we do not tabulate the regmssion results. 4. i.

T&e fargel

firm ‘s it&id

rpcction

IO he lemdkr ogler bid

The decision by the target firm’s board to resist the initial tender offer is an important characteristic of a tender offir. Takeover msistance can reduce the likelihood that the o&r will succeed. but can also increase the final takeover premium for successful offers Using regression models similar to those in previous tests, we estimate the relation between board composition and the probability that the target msists an offer. We do not find a significant relation between board indepcndtnce and the likelihood of target resistance. However. we find that resistance is significantly fess likely in the presenaz of interlocking directorships Perhaps in these cases close relations between bidder aird tqet have already heen established. paving :he way for a friendly takeover. ;Cvcn though a lower initial premium is off&d in the presence of interlocking directorships. these targets are less likely to resist. Tbe regressions also show that targets with poison pills are more likely to resist an offer. It is possibk that targets with poison pills areentrenched and that their boards value the benefits ofcontrol more than other targets. or that targets with a poison pill are better equipped to bargain with the bidder. The latter hypothesis is consistent with the resuhs in Tabia 2 an3 2. which suggest that

214

J. F. Cottw

et al. IJournal

of Financial

EC ononics

4j (1997)

195 -. 218

targets with a poison pill and an independent board obtain higher shareholder gains, and that targets with a poison pill extract higher bid premium revisions. 4.2. Comparing

resistance

by independent

and nonindependent

boar&

We find that independent and nonindependent boards are equally likely to resist an offer. Given the positive relation between board independence and shareholder wealth, it is possible that targets with independent boards are more eflective in using resistance strategies. This interpretation is supported by the results in Table 4, in which we compare the premiums and eventual outcomes of resisted oilers. For the 79 resisted ofl’ers in the sample, the average initial premium is 50% and the premium revision is 12%. Fifty-four percent of the offers are eventually successful. The target shareholder gain for all resisted offers is 48%. Consistent with the results in Table 3. we find significant differences between firms with and without independent boards. On average, the premium revision for resisted offers is 16.6% for independent boards compared to 9.4% when the ;*Jrd ix 01.t: iti,JC~
of multiple

hi&t&

Another aspect of the tender otTer process that can impact shareholder gains during tender o&s is the presem~ of multiple bidden. Although the presem~ of multiple bidders depends in part on industry characteristics that are beyond the control of the board directors can actively seek other bidders We evaluate the impact of board composition on the likelihood of multiple bidders and find that the coefb&nt on the indepe&nt board dummy is insignificant. We do. however. find that interlocking directorships significantly decrease the likelihood of multiple bidders Sina the presence of an interlocking directorship reduces the probability that the bid will be resisted. this result can reflect the friendly. negotiated nature of these transactions. Alternatively. an interlocking directorship. combined with a higb bidder toehold can be a deterrent to other potential bidders. 4.4. l%e final

outcome

of the tender

0g.k

To examine the effect of independent directors on tender offer outcome, we estimate logistic regression results in which the dependent variable equals one if

.__...

_.__

_-...

. .._

mistana.

otkr

prcmwmr

‘O.:\% 12.0% 48.0% 544%

Mean

(n = 79) - __

ORCfl

All rcsirlcd

44.2% 9.7% Jll.O%

M&an

._ --. _ _ ^

tender

._ -..- _.__ __ ..-_-

itnd tender

-. -.- .-. .I -

48.3% 9.4% 39 7% 49%

Mean

-.

41.6% I *I % W.ll%

Mcdtan

.-

the twption

Independent ou~ridc directors < 50% (n - SI) ._ ____-____

of the initial premium. bid revision. target nhurcholder gain from classified h) independent and nonAndcpndent target boards

Initial premium Bid revision Target sharehotir gem Ten&r oRcr NKZC& .- ._

_- ___ -. ..- _

and medians tender o&r,

Means rcrirtcd

Tabk 4 Board independence.

S3.R% 16.6% 63.1% 64%

Mean

49.9% 14.0% 75.2%

.-

050 0.04 0.09 0.19

.-. .

r-teal p-value

of the ofier and succcw~ul

Median

Independent outride directon 3 50% (I) - 28)

to the resolution

._

0.51 0.04 0.15 0.20

Wilcoxon p-value

.

outcome

fur

216

J.F.

Cotter

et al. fJoumal

of Financial

Economics

43 (1997)

195.-218

the target is successfully acquired, and zero otherwise. Across all models, neither the board composition

and incentives variables nor any of the other control

variables are statistically significant. WC also estimate these regressions controlling for the final tender offer premium and for resistance by the board, and obtain similar results. Thus, independent directors enhance shareholder wealth from tender offers by obtaining higher premiums, not by affecting the likelihood of success.

5. coachlsh Tender offers can result in important conflicts of interest between the managers and shareFolders of target firms. We examine whether independent outside directors perform an important role in controlling he-se co&As and enhance shareholder wealth dgring tender olTers. We document that the target shamholder gains over the entire contest period are higher when the target’s I~NxI is in dependent. Ir evaluatin; thr l XT of these gains, we 2nd that targets kdil Independent boards extract. both higher initial tender offer premiums and higher bid premium revisions thao do targets without independent boards. We find that targets with i&pendent boards are not more likely to be successfully taken over and that the target shareholder returns of unsuccessful tender o&s do not di&r between targets with and without independent boards. These results suggzst that the higkr sharrbolder gains for targets with indqer&nt boards are driven by higher initial premiums and bid premium revisions We also find that in resisted o&rs and of&n to targets with poison pills+ target shareholder gains arc higher when the board is i&per&n1 than when it is not. Our evidem~ adds to the literaiurc documenting the importance of outside direaors during specific corporate events. Weisbach (1988) reports that boards dominated by outsiders are more likely to force resignation of poorly petforming CEOs. In the same vein, Byrd and Hickman (1992) find that bidding firms with a majority of in&pen&nt outside directors earn higher announcement abnormal returns than do firms without a majority of independent directors, Brickky, Co& and Terry (1994) suggest that the proportion of outside board members rciates positively to the abnormal returns at the announcement of poison pill adoptions. Roseostein and Wyatt (1990) report that the market reacts positively to the anno uncement of outside director appointments We conclude that ir&pe&ent outside directors play an important role during tender o&s and that they enhance shareholder wealth during tender offers. Our evidence aiso suggests that bid premium revisions and target shareholder gains 812 higher in resisted o&s when the target’s board is independent than when it is not. This suggests that targets with independent boards are likely to resist so that they can extract a higher premium for target shareholders, not to entrench incumbent managers.

J.F.

Cotter

et al. r Journal

of Financial

Economics

43 ll947)

195

218

217

Agmwal. Anup and Ralph Walkling 1994. Erecuttve careers and compensation surrounding takeover bids. Journal of Finance 49.9851014. Barclay. Michael J. and Clifford G. Ho&ncss. 1991. Negotiated block trades and corporate contrd. Journal of Finana 46, 861.-878. Bdsky. David A.. Edwin Kuh and Roy E. Wdrh. 1980. Rcgrcssion diaptics: Identifytng influcntutl datr and souras of collinearity (Wiky. New York. NY L Bradley. Michad. Anand ti. and Han Kim. 1983. The ntionak bdtind interfinn tender offers Information qr synergy?. Journal of Financial Economics I I. 183-206. Bdky. Mid& Atmod hi. and Han Kim, 1988. SW gains from aqxxatc acquisitions and tkir diviricbn bctwtm the stockhddcrs of tarpt and acquiring timu Journal d Financial Econotnics2l.~40. Brickky. Jam A. and Christopher James 1987. The takcovcr ma&et. corporate board compostion and ownership structutz Tbc asc d banking Joumal of Law and Eunmtnics 30. 161 180. Bddey. James A.. J&q I- Coks and Rory L. Terry. 1994. The boanl of dimctm and the enactment of poison pill Journal of Financial Eamom k 35.371 390. Brickky. James A, Ronald C. lcln and ClifTixd W. Smith 1988. Owncnlup strtuztum and votmg 00 atttitakcowr ammdmmi. Journal of Financial Eamomics 20.267 31. Ryrd. John and Ken1 Hickman. 142. Do outs& dtrmors monitor managers? Evidence from tender dTer bids. Jaurld of FiNlKiar Ecomtaks 32.195 221. Commmt. Robert and G. Wiltiam Eh~m. 1995. Poison or plaabo? EG&ncz on tk deterrent dTcctr of mo&m antitakeover m sums+ Journal of Finatxial EumonGs 29.3 -43. Cotter. James F. and Marc Zmner. 1994. How managerial realth alVats tk ta&r o&r pra Journal of Finamzial Eamotn its 35.63 97. Dann Larry Y. and Harr) DeAngdo. 1% Corporate financial polity aod axpontc amtroi: A study of ddemiw 0djustmtnt.s in asxt and owner&p structure. Journal of Financial EamoaGa#L87-127. Fanu Eugene. 1980. Agency probkms and tk tm of tk firm. Journal of Pohtwsll Eamomy 88. 28%307. Fama. Eugene and Michad knscn. 1983. Scparat, r, downership and control Journal of Law and Econom& 26,301 325. Fishman. MichadJ.. 1988.A thccryofprcemptivctaien+erbidding Rand JoumalofFamonucs 19. 88 101. C&km. Quart. 1990. Bankruptcy. boardr banks aaJ blockhdks E\dmot m changes on cotpomtc ownership and control when firms dda&. Journal of Ftnatxial Fconomics 2?. 355 387. Han&n. Benjamin E. and Michael S. Wcisbach. 1988. ‘I& dctcnninants of hoard composition. Rand Journal of Emnonks 19.589 NW Hermali& Benjamin E. and Michael S. Wci&ach. IWS. End~~gcnously cboscn boards and their monitoring of tk CEO. Working paper (Univcratr oi Cahfomta. Berkck!. <:A). Ho&mess, Chfird and Dennis Sheehan. 1985. Raiders or sanorr? The evidence front six controvcrstal invotors Journal of Financial Eumomn% 14. 555 579. t of tender o&5: Jarrell. Gregg and Annate Poulscn. 1989. Stock tradtng bebrc tk asnoutxxxnen lnsidcr trading or market anticipation’!. Journal of Law. F.zolls,mics and Orpntiation 5. 225 248. JarrdL Grcep A.. James A. Brickky. aad Jeffrey M. Netter. 1988. The market for cotporatc control: The empirical &dcna since 1980, Jounral of Economic Pcrspc&vcs 149.68. Jensen. Michael and Richard S. Ruback. 1983. Tk market for corporate control: The scientific evidena. Journal of Financial Economics I I. S-50.

218

J.F.

Cotter

et al. lJouma1

o/Finuncial

Economics

43 (1997)

195 -218

Kaplan, Steven and David Reishus, 1990. Outside dixctorshlps and corporate performance. Journal of Financial Economics 27.389 410. Kini, Omsh. William A. Kracaw, and Shehzad Mian. 1995. Corporate takeovers, firm performance. and board composition, Journal of Corporate Finance 1.383412. Malatesta. Paul H. and Ralph Walkling, 1988. Poison pi’! securities Stockholder wealth, profitability, and ownership structure, Journal of Financial Economics 20. 347.-376. Martin, Kenneth and John J. McConnell. 1991. Corporate performaria. corporate takeovers and management turnover, JoumaJ of Finance 4b. 671 687. Mikkebon, Wayne H. and M. Mwn Part&. 1989, Managers’ voting ri&s and corporate control Journal ol Financial Economies 25263-290. Mikkekon. Wayne H. and Richard S. Ruback. 1985. An empiric01 analysis of the mtetlirm a&y invatmen~ pruecs Journal of Financial Economics 14.523-553. Milgmrn. Paul and John Roberts. 1992 Ecorwnicr. organization and mawqxmen I (Prentia Hail, Engkrood Cli& NJL Rosensaein. Stuart and JelTrey G. Wyatt, 199O.Owide directors, board independena. and shaxholder wealth, Journal of Financial Economics 26, 175-192. Rub&. Richard S. 1988. Do car@ sharrhddm lose in unsuccessful control contests?. in: Alan J. Auerbach, Corporate takeovers Gusa and consequenas IIJniversty of Chicago Press Chicago. ILL Ryngaert. Michael 1948. m eiTect al potson pdl securities on shareholder walth, Journal d I’lcar %I Fawoti 3. 777 417. L ~&UIII, .&nil. I’&, &oaru wmposition. owwrshlp r(nnxure and hostik takeovers Journal of Aawnong and EooMlmicr 16 l67- 198. Shkikr, And* ood Robert Vishny. 1986. Lars sharrholden awl aqxmte control :oud d PululUI &oaonly w. .&Jr 488. Van Nuys. Kum, 1993. Corpora~c wna through I& proxy proas Evidens Imm the 1989 Hoaywcll proxy solicitatina. Journal of Financial Eamomics 34. IO1 I32 WalkRalph, : -c Pltdid0g*eodrrdklSUWUS A lo&ic anal* Journal Ot Financial and Qumiwive Analysis 20.461-478. wdklin& Ralph 8od Michael IAm& 1984. m theory. malmgefial *dram and takeover bid resistma. Rand Journal ai Economies IS, % 68. Weishaeh Michael S, 1988.Ou1side dirutors and CEO turnover. Journal of Financial Economics 20.431 460.