Journal of Business & Leadership (2005-2012) Volume 4 Number 1 Journal of Business & Leadership 1-1-2008

CEO COMPENSATION, BACKDATED STOCK OPTIONS, AND COMPENSATION COMMITTEES Steven A. Frankforter Winthrop University

Bret Becton University of Southern Mississippi

Follow this and additional works at: http://scholars.fhsu.edu/jbl Recommended Citation Frankforter, Steven A. and Becton, Bret (2008) "CEO COMPENSATION, BACKDATED STOCK OPTIONS, AND COMPENSATION COMMITTEES," Journal of Business & Leadership (2005-2012): Vol. 4: No. 1, Article 9. Available at: http://scholars.fhsu.edu/jbl/vol4/iss1/9

This Article is brought to you for free and open access by FHSU Scholars Repository. It has been accepted for inclusion in Journal of Business & Leadership (2005-2012) by an authorized administrator of FHSU Scholars Repository.

Article 9

Frankforter Becton: CEO COMPENSATION, BACKDATED STOCK OPTIONS, AND COMPENSATION COMMI Journa l or Bu siness & Leaders hip : Research . Practice and Teac hing I ' '"'~fo n c r andand 13ccloll 2008 , Vo l. 4. No . I , 64-69

CE O CO 1PENSATION, B ACK DATED STOCK OPTIONS, AND COMPENSATION C Ol\ll\IITTEES Steve n A. Frankfo rter, Winthrop Uni ve rs it y 13re t lk c ton, Uni ve rsit y o f So uthern lvl iss iss ippi

CEO compensation in U. S. based companies has undergon e considerable scrutiny in recent years. Among th e common ohserl'(lfion s are that U. S . executives are hig hly paid relative to th ose of oth er co untries and that th e di~parities in compensa tion are in creasing over tim e. In this stttr~l' · H'e investigate th e effec ts that backdated stock options, contpensation commi11ee stru cture and process, and own ership fa ctors h ave on le vels of executive compensation. Co mbining ag ency and orgrmi ~ (l(ional th eory p erspectives, we find CEO compensation positively associated with th e presen ce of backdated stock options, few large-block stockholders, and sm all compensation committees. Ex ecut ive co mpensat io n is a n intense ly debated and researched a rea . Wi th argum ent s that CEOs are O\'c rco mpcnsatcd, wit h prob lemati c acco untab ilit y to shareho lders, ca lls for re form frequ entl y appear in the popular press. W1th low acco untab ility to stakeho lders, C FO co mpensati on tends to in crease as the firm ado pt s lown ~ k co mpe nsa ll on schcmc;s (Wcm er, Tos 1, & Go mez- Mej ia, 2005). Further, \\'eak boa 1·d co mm ittee gove ma nce controls may llllli catc s ituati o ns whereby board o f directo r co mpensati On co mmittees arc co-opt ed by C EOs , resultin g 111 1nfl a ted co mpensa ti on (Va feas , 1999 ), o r in in novat ive co mpensat iOn sc hemes, s uch as ba ckdated stoc k opti ons ( I 1e, 2005 ). Accord1n g to age ncy th eo ry, exec uti ves e mp loy pos iti on po wer 111 pursuit o f th eir eco no mi c sc i!- in teres ts even when 11 ma y co nfl1 ct \~ 1111 the we lfa re o f th eir co rporati ons (J ensen & l\1 ec kl1ng, 1976 ). One remedy to the age ncy problem is to usc Inte rest a li gnment mec han 1sms to lin k C EO co mp e n ~at i o n to shJ reho ldc r-bc ne fi cial results. A co mmo nl y e mp loyed app ro ::~ c h IS the awa rdin g o f stoc k opt io ns so that CEOs benefit finan c ia ll y whe n stoc k pri ces esca late . ll owever, in the case o f ba ckdat ed stock opti ons, co mpensati o n beco mes g uarant eed , instead o r co ntingent o n fi rm perform ance. 13 ac kd atin g g u a r ::~ nt ccs an o ft en-genero us pro fit to exccut 1vcs at th e time of iss ue. A seco nd co mpo nent o f age nt co ntro l is board monit o rin g o f CEOs. \V 1th directo r o ve rsight , an oppo rt un isti c CEO is sup posed ly unable to usurp corpora te asse ts to th eir own bene fit. l lowe\'c r. Incenti\'C a li gnm e nt mcc lwni sms :md agent mo ni ttm ng by the board ma y int era ct, res ultin g in co mpro m1 scd co ntro ls. Age nt opportuni s m is curbed only \\' he n bo th arc prese nt and effec tua l. Fo r ex amp le, if a finn' s boa rd uses stock o pti o ns bu t b o ::~ rd monito ring is \leak, ex ecuti ves may nw n1pu l ~l t e fi na nc ial info rmati on to reap unc:1n1ed finan cial rewa rds. Such problems \\CIT adu1 esscd th rough the adop t io n of til e Sarbancs-O x Icy Act in

co ntrary, we ex pect to find hi gher CEO co mpensa ti on on ly with those firm s that e mpl oy b::tckda ted stock op ti ons and whose co mpe nsatio n co mmittee structure and processes mi ght be relati vely weak. II Y POT III~ S ES Inquiring into th e determinants o f CEO co mpensation, tile hypotheses will exa mine one CEO co mpe nsa tion mcchani . m; backd at ed stoc k options. We also include two co mpensati on co mn ttee fac to rs; th e number of tim es it meets, a nd its size. Lastl y, we in ves ti ga te two ownership fa ctors; the num ber o f large- bloc k owners o f shares, and CEO s h ::~ r e ownership . IJa ckdat ed S to c k Options 13 Jckda tin g s toc k opt ions permits rec ipi ents to selec t a da te where the exercise price is low, which in variab ly guarant ees a profit as soo n as the options are granted. T hi s prJctice is wides pread and it h ::~s attracted the prosecutori al scru tin y o f the Securities and Exchange Co mmi ss ion and the Justi ce Department. Li e (2005) repo rted low stoc k retums before co rporate exec uti ve stock opt ion grants and unu suall y hi gh return s imm ed iately aft er. He co nclu ded that, unl ess executi ves had so me unu sual abi lit y to forecast the future, they were m o~ t likely backda ting their stoc k optio n grants. T herefore, we pred ic t th at us ing bac kdat ed stoc k opti ons wi ll tend to in crease C EO co mpcns:J tio n. Hypoth esis J. The use o f backda ted stoc k opti o ns will be pos iti ve ly assoc iat ed with CEO co mpensati on. T he Numh er of Five Percent Ow ners When large, co ncentrated block o f shares are owned by significant num ber o f gro ups o r indi viduals, it increases th e li kelih ood o f oppos iti o n to po licies and ac ti ons that are not in th e int erests o f shareholders. II ighl y concentrated ownershi p ma kes it eas ier and re lati vely less ex pensive for shareholders to coo rdinate among th emse lves and take acti on aga inst firm s they percei ve as not serving sharehold er wishes (Fa ma & Jensen, 1983 ). Such large- b lock owners :1

2002.

In th is stud y we; cxa m1n c th e effects o f backda ted stoc k op ti Ons, ownership conce ntrat io n, and co mpensa ti o n Ul lllllli ttcc structure and processes o n C EO co mpe nsati o n. \\'c dn not assume that all C I-' Os a rc overpaid . To the

Published by FHSU Scholars Repository, 2008

64

1

Fra rtkf011 er a nd Becto n

uma l4o f[2008], Bus t nessNo. & Lc3d ers htp9 Rcse3rch, Pt-:tCitcc a nd Teac hing Journal of Business & Leadership (2005-2012),JoVol. 1, Art.

2008 . Vo l 4. No I . 6-l -69

avo id firm s w ith co rp o rate gove m a nce p rob lems. Fo r insta nce, instituti o na l o wners h ip a t co mp a ni es w hose manageme nt is s uspec ted o f e nt re nc h me nt w ill tend to be lower (F rankforte r, Bem1 an, & Jo nes, 2000) . Beca use of the diffi cult o f p ro fit abl y se lling large b loc ks o f sha res w it hout m akin g a ma rke ts price drop , m aj o r inves to rs w ill te nd to oppose manage me nt po li c ies no t in the ir bes t in terest ra the r than atte mp t to se ll the m . Shareho lde rs possess ing s ig nifi ca n t eq uity ho ld in gs above a fi ve perce nt level mu s t ide nt ify the mse lves b y sub mittin g a sec ti o n 13 (d) filin g w it h th e Sec urit ies a nd Excha nge Co mmi ss io n. T hese owne rs may have s ignifi ca nt influe nce because th e be ne fit s o f the ir in vo l ve mcnt in mo nit oring a co mpa ny's ma nageme n t o ut we igh th e cos ts (De mse tz, 198 3) . Large-b loc k owne rs have sig nifi ca nt influe nce ove r co rpo ra te po l icy (De mse tz & Lehn , 1985), board comp os iti o n (Po un d , 1992), a nd he lp ensure th at th e finn 's exec uti ves ac t to fu rth e r th e share ho lder int e res ts (Be the l & Liebes kind , 1993). T he g rea te r the numbe r of fi ve pe rcent o wners, th e g rea ter th e ir co mb ined ab il ity to mo nit o r the fi rm , thu s, c urbin g age ncy iss ues, ho ld ing CEOs mo re acco unt ab le (B ri c kley, Lease, & S mith , 1988).

qui ckly th a n la rger g ro up s, a nd indiv idua ls tend to he ighte n th e ir perfo rm a nce in s ma ll gro up s (Seij ts & La tham, :WOO). W hil e thi s ev idence suppo rts sma ll gro up s ize sma ll in so me co nt ex ts, o th e r e m pirica l evidence ind ica tes sma ll group size ma y e nab le CEOs to ma nipul3t e compe nsation co mmi tt ees fo r perso na l ga in . Severa l s tud ies im ply that sma lle r group s tend to be mo re coo pe rative w ith CEOs a nd more suscep ti b le to tit - fo r-ta t s trateg ies. An exa min atio n o f socia l d il emm a resea rch revea ls th at coo perat io n drops as group size inc reases (A ll iso n, McQ uee n, & Schaerfl , 1992 ; L ieb rand. Mess ic k, & W il ke, 1992; M ess ic k & Brc\\'er. 1983, Scij ts & La th am , 2000). T wo possib le ex planations fo r this e ffec t arc ; ti t-fo r- tat s trategies thJI promo te mutua l coo pe rat io n a re less e ffec ti ve with large r g roups, and the feeling o f a group me mbe r 's sense o f respo nsibi li ty to the gro up d ec reases as its size inc reases (Se ij ts & Lat ham, 2000). l ie nee, larger gro ups are less like ly to accept the s ta tu s q uo, ac t in th e self- in te res t o f the g ro up, and cha ll e nge dec is io ns. T hi s s ize o f boards and co mmi tt ees ha s J lso been the subj ec t o f mu c h resea rch O\'Cr th e yea rs. Dai ly and Dalto n ( 1993) fo und that g rea ter numb ers of tota l direc to rs to be pos iti ve ly assoc iated w it h fi rm pe rfo rma nce . CEO do min at io n o f boa rd s is mo re di ffi cult as boards inc rease in s ize beca use th e re are more po te nti a l oppone nts to mana ger ia l do m in ation (R ose nste in, 1987) . Beca use o f a hi g he r po te ntia l fo r ho mogene it y, group co hes ion, and mo re inte nse co mm un ica ti o n, s m a ll er co mpensa ti o n co mmittees are mo re like ly to yie ld to CEO w ishes, inc reas ing exec utive co mp e nsa tio n.

H ypoth esis 2 . T he nu mb er of fi ve pe rce nt own ers w ill be nega ti ve ly assoc ia ted w ith CEO co mpe nsa ti o n. The Numbe r of C omp en sa tion C o mmittee Mee tin gs A mo ng the ir o th er duti es in m anag ing exec utive co mp ensa ti o n. co rpo ra te co m pe nsa tio n co mmittees ad ministe r s ha re ho lde r-a pp roved stock o p tio n p lans_ T hese co mmittees d e te rmin e the size a nd tim ing of s toc k op tio n grants. Ho weve r, it usuall y occurs w ith CEO in vo lve me nt ( Lie , 2005) . O ft e n, CEOs p ro pose g rant te rm s th a t co mpe nsa ti o n co mmit tees a lm os t in va ri ab ly ra ti fy (Yerm ack, 1997), prov idi ng evide nce ofCEOs ma nip ul at ion of those interes t a li g nm e nt a nd mo nit o ring m echa ni sms th a t mi ght o th erw ise tend to curb agency p ro b le ms. Boa rd o f d irec tor ru bbe r-s ta mpin g ma nageri al dec is io ns has bee n a n oft-leve led c riti c is m (F le isc he r, Haza rd , & Klippe r, 1988) . O ne so luti o n is to inc rease th e frequ ency o f meetings , a llo w ing fo r add itio na l co llabo rati o n (iv!cG rat h, 199 1), whi ch can ha ve a pos iti ve imp ac t o n team pe rfom1a nce fo r g roups suc h as boa rd s o f direc tors and its co mmittees. A n ex am ple o f imp roved g ro up perfo rman ce is Va feas ( 1999), w ho foun d boa rd mee ting frequ e ncy re la ted to firm va lue. In lig ht o f the evid e nce, we pred ic t th at m ee tin g freq uenc y w ill be in ve rsely re la ted to CEO pa y.

H ypot hes is 4 . Co mpe nsat ion com mit tee s ize will be nega ti ve ly assoc iated with CEO co mp ensat ion. S har e O w ne r ship by th e CEO \V hen exec uti ves have littl e ownership in th e firm , th ey have dimi ni shed ince nti ves to pro mo te shareholder wea lth. a nd ca n be expec ted to be mo re self- serving (J\la lates ta & Wa lk! in? 1988). Wh en CEO stoc k ow ners h ip is he ighte ned , a lig nm ent wi th th e fi nan c ia l int eres ts of the fir m and its shareho lde rs increases (E ise nh ard t, 1988; Je nse n & Meck ling, 1976). CEOs possess in g sub stant ia l eq uit y in th e ir fi m1s ha\'e risk a nd rewa rd pe rcep tions linke d to those o f th e sha re ho lde rs , and w ill be more like ly to ac t in the share ho lde rs' interests (Da lto n ct a!. , 2003). Hi gher CEO share owne rship is associated with reduced instances of co mp e nsa ting exec uti ves w ith stoc k op ti o ns ( Brya n, Whan g. & Lili e n, 2000) , d imin ishe d usc o f go lden pa rachut es (Si ng h & 1-l ::~r i a nt o, 1989), res istin g ta keove r atte mpts (St ulz, 19 88) , and adopti ng po iso n p ill takeove r defenses (Malates ta & Wa lkling, 1988).

Hypoth esis 3. T he num be r o f co mpe nsa ti o n co mmittee m ee tin gs w ill be nega ti ve ly as soc ia ted with C EO co mpensa ti on.

H y po th es is 5. CEO share ownership wi ll be nega ti ve ly associated w ith CEO c mpe nsat io n.

Compe nsation C ommittee S ize Gro up s ize a ffec ts th e be havio r o f g ro up s. tea ms, and co mmittees. In gene ra l. s ma ll e r g ro ups co mp lete tasks mo re

http://scholars.fhsu.edu/jbl/vol4/iss1/9

65

2

Jouma l o f Bu siness AND & Leaders hip : Researc h, Practi ce and Teac hing Frankforter and Becton: CEO COMPENSATION, BACKDATED STOCK OPTIONS, COMPENSATION COMMI

F ra nkfon er a nd Becton

2008, Vol. 4, No. I , 64 -69

stoc k op ti o ns, the numb er o f co mpensation comm ittee mee tin gs and the s ize o f th e co mmittee , the numb e r of largebl oc k ow ners, a nd th e pro po rti o n o f the firm's shares the CEO ow ned. W e introdu ced two co ntro l va ri ab les ; firm s ize ( log o f e mpl oyees) , a nd re turn o n equit y. Fi rm s ize may influe nce th e fo nn o f powe r a nd govem ance s tructures. For examp le , Finke lste in a nd D ' Ave ni ( 1994 ) re po rt ed that organi zati on s ize tend ed to a ffec t th e usc o f du a l s truc tures, th e po wer of its exec uti ves, a nd firm p erfo nn ance. We co ntro lle d for firm s ize b y co mputin g the log o f th e to ta l numb er o f e mpl o yees (Frankfo rt er, Dav is, & Vo llra th , 200 I ) to redu ce he teroscedas ti c it y ( Ke rlin gcr, 19 73) . N ex t, the re a re links be twee n pro fitabilit y and gove m ance s tru c ture. Direc to rs o f unde rpe rfonnin g firm s exerc ise th e ir a uth o rit y m o re readi ly , ho ldin g mana gement to he ig ht e ned stand a rd s o f acco unt a bi lit y (A lderfer , 19 86 ; Mi zruc hi , 1983). Da vis, Sc hoo nn a n , and Dona ldso n ( 199 7) pred ic ted th at firm s w ith hi g h degrees o f a li gnm ent be tween th e CEO phil oso ph y a nd th e firm ' s gove rn a nce stru cture wo uld tend to have hi g he r p ro fits. W e mea sured pro fi tab ilit y a s th e firm 's re turn o n e quit y. W e o bta ined tota l CEO co mpe nsati o n fr o m Exec uCo m p, de fin ed as th e s um o f s a la ry, bo nu s, o the r annu a l co mp e nsati on to ta l va lue o f restri c ted stoc k grant ed , to ta l va lue o f stoc k o p ti o ns g ra nt ed , lo ng- te rm ince nti ve pa yo u ts, a nd a ll o the r to ta l co mpe nsa ti on. Firm s ize and re turn o n eq uit y d a ta we re o b tai ned fr o m Resea rch lns1ght. A ll o th er da ta we re co ll ec ted fro m proxy s ta te ments. W e e m p loyed multipl e reg ressio n a na lys is to tes t our mo de l.

DES IGN, MEASURES, AND METHODS Design W e se lec ted 166 fim1 s ide ntifi ed by th e W a ll S tree t Jo urn a l as be ing und er in ves ti ga ti o n fo r e mpl oyin g backd ated stoc k optio ns (P e rfec t Payd ay Op ti o ns Scoreca rd . W a ll S tree t Jo urn a l. (O n-lin e). http ://o n line. wsj .co m/ public/reso urces/doc um e nts/ infoo pti o nssco re06- full.html) . The finn s so lis ted we re subj ec t to e ither Securiti es and Exc ha nge Co nuni ss io n a nd/o r U.S. Ju sti ce Departme nt inves ti gati o n. Whi le the fo rn1 a nd c ircum stances fo r in ves ti ga ti o ns vari ed , we co ns ide red th e initia ti on o f a fed era l in ves ti ga ti o n a s ig nifi cant s ig n that wro n gdo in g may have occ utTed . Acco rding ly, we decided th at named co mpa ni es mi ght be pro ne to agency p ro bl e ms a nd co rp o ra te gove rn ance iss ues. W e de fin ed th e co nt ro l g ro up as firm s w ithin the sa me indus tri es th a t we re no t un de r suc h in ves ti ga ti o n. W e e mpl oyed case-co ntro l p rocedures desc ri bed by Seabri g ht , Levintha l, and Fic hm an ( 1992). Case-co ntro l des ig ns exam ine re la ti ve ly ra re eve nts. Co m pa ny s ize wa s se lec ted as a ma tc hin g va ri abl e beca use it co uld po tenti a lly co nfo un d res ult s, b eca use CEO co m pe nsa ti o n te nd s to inc rease w ith th e s ize o f th e firm . T he re fo re, we se lec ted the two co ntro l gro up firm s nea res t in to ta l asse ts w ithin the sa me 2-di g it S ICs to eac h firm in th e ex pe ri me nt a l g ro up (S ing h & Ha ri ant o, 1989). The res u lt was th e initia l se lec ti o n o f 332 co ntro l firm s. T he reaso n we se lec ted a ge n e r o u ~ numb e r o f co ntro l firm s wa s beca use we a nti c ipa ted th a t th e use of multi p le dat a so urces wo ul d res ul t in a s ig nifi ca nt redu c ti o n o f fi rm s re ma ini ng fo r s tati stica l ana lys is beca use o f mi ssi ng da ta . T his d es ig n a ll ows us to mo re c lea nl y in ves ti ga te phe no me na beca use we are no t se lec tin g o r so rtin g firm s acco rd in g to th e d epe nd e nt vari3b lc (CEO co mp e nsa ti o n). M iss in g o r inco m p le te proxy s tatemen t da t3 red uced th e numb e r o f co mpa ni es b y I 39. Five fi rm s lacked Resea rc h Ins ig ht da ta . Fina ll y , we e limin3ted I 3 I co mp:lll ics d ue to m iss in o CEO co mp e nsa ti o n da ta. T hus, th e num be r of firm s in o ur ; tud y was redu ced to 223; 73 we re in th e ex p erime nt a l g ro up a nd 150 in th e co ntro l g ro up. We perfo rm ed a t- tes t to dc tc rn1ine w he th e r asse t s ize d iffe red betwee n th e gro ups and found a n ins ig ni fi ca nt !-s ta ti s ti c o f .OJ. Acco rd in g ly, we co nc lud ed th e finn s re mai n in g in o ur s tud y we re of s imi lar s ize. W e se lec ted 2000 as th e exam in a ti o n yea r bec ause bac kda ted opti o ns we re in va ri ab ly e nac ted b e fo re thi s date. Fu rther mo re, th e passage o f th e Sa rb3 ncs Ox ley Ac t in 2002 g rea tl y dim in ished th e freq ue ncy o f bac kda ted s toc k o ptio ns a nd a lso led to w idesp read c hanges in boa rd stru c tures and processes th at wo ul d like ly c l ud o ur in ves ti ga ti o n.

R ESU LTS Tab le 1 repo rts d esc ri p ti ve s tati s ti cs, va ri ati o n infl ati o n f::t c to rs, a nd th e co rre la ti o n m 3tr ix . We addressed muti co llincarit y co nce rns b y exa minin g co rre lati ons a nd va ri ati o n infl a ti o n fa c to rs. N o co rre lati o n coe ffi c ie nt exceeded .30 . Ad d iti ona ll y, none o f th e va ri a ti on infl ati on fa c to rs s urpa ssed 1. 18 , fa r fro m th e c riti ca l limit o f 10 (N etter, W as se rm a n, & Kutn e r, 1989) These res ult s suppo rt ed o ur co nc lus io n th a t muti co llinea rit y did not threa te n to co nt a min a te o ur res ult s. Tab le 2 d isp lays th e d esc ri pti ve st3ti sti cs a nd mu lti pl e reg ressio n resu lts. O ur mod e l revea ls s ig nifi ca nt m a in effec ts fo r th e use of bac kd a ted s toc k o pti o ns, th e numb er o f five perce nt ow ne rs, a nd co mmitt ee s ize o n to ta l CEO co mp e nsa ti o n. Ho wever, th e res ult fo r th e numb er o f . co mp e nsa ti o n co mmittee mee tin gs was pos iti ve ly assoc tate d w ith CEO co m pe nsati o n , w hi c h was in the o ppos ite direc ti o n to th a t w hi c h was predi c ted . There fo re, hypo theses 1 2 a nd 4 we re s upp o rt ed . I loweve r, hyp o th eses 3 a nd 5 ,~ere no t s upp o rt ed . Ove ra ll , o ur mo de l had good predi c ti ve va lu e, w ith a s ig nifi c a nt F= 3.79 (p < .00 1) a nd an adjus ted R2 = .08 .

l\1ca s ur es a nd Method s T he in dcp(' nd e nt va ri a b le was CEO to ta l co mp ensati o n. The d epe nd e nt va ria bles we re th e prese nce of b3c kd a ted

Published by FHSU Scholars Repository, 2008

66

3

Frankforter and Becto n

Journal of Business & Leadership (2005-2012),Journal Vol. 4o[2008], 1, Art. 9p: ResC3 rch, Prn cucc and T eac hin g f Bu sinessNo. & LcJdcrshi 2008, Vol 4, No. I , 6~ -69

Tab le 1: Descriptive S tat istics, Variation Infl a ti o n Fac to r s (V IFs), a nd t he C on-elatio n Matrix

I. CEO co mpe nsa ti o n 2. Backdated s toc k opti ons 3. Number o f fi ve percent owne rs 4. Co mmittee s ize 5. Co mmittee mee tings 6. CEO owners hip 7. Firm s ize 8 . Return o n equit y

Mean I 0529 .06

S. D. 4209 .66

.33

47

1. 0 8

. 10

2.45

14 8

1. 03

-. 15*

3.09 3.46

.96 2.5 4

I . 18 I. 14

-.07 .22***

-.2 1** -. 17**

-09 -. 10

.04 .64 8.33

.07 .68 85.27

I. I I 1.1 2 1. 0 I

-.02 . 10 .0 1

.07 -.08

. 12 -.08 - 03

7

8

4 4 . Co mmittee size 5. Co mmittee meetin gs 6. CEO own e rshi p 7. Firm s ize 8. Return o n equit y * p < .05 **p