Does cross listing in the U.S. really enhance the value of emerging market firms?

ECONOMICS DEPARTMENT WORKING PAPERS SERIES N184/12/07 Does c oss lis ing in he U.S. eally enhance he alue o eme ging ma ke i ms? by Thomas G. O’Conno Depa men O Economics Na ional Uni e si y o I eland, Maynoo h NATIONAL UNIVERSITY OF IRELAND, MAYNOOTH MAYNOOTH CO. KILDARE A. IRELAND h p://www.nuim.ie/academic/economics/ h p://ideas. epec.o g/s/nuim/mayecw.h ml h p://ep in s.nuim.ie/ 1 Does c oss lis ing in he U.S. eally enhance he alue o eme ging ma ke i ms? Thomas G. O’Conno ♣ June 25 h, 2007. Abs ac In his pape , I s udy he alua ion e ec s o c oss lis ing in he U.S. o a panel o eme ging ma ke i ms o e he pe iod om 1990 o 2003. In line wi h K is ian-Hope e al. (2007), I ind ha only hose i ms om high disclosu e egimes gain om Le el 2/3 lis ing in he U.S. The gains a e no immedia e, bu ma e ialize once he i m has lis ed in he U.S. o a leas i e yea s. I also documen long- e m, bu no immedia e alua ion gains o Le el 1 o e - he- coun e issues. In con as o Le el 2/3 issues, he gains a e concen a ed amongs i ms om low-disclosu e egimes. I ind no posi i e alua ion e ec s o Rule 144a p i a e placemen s. The esul s sugges ha he decision on he pa o he majo i y o i ms om low-disclosu e egimes no o lis as exchange aded deposi a y eceip s is wa an ed. JEL Classi ica ion: G15, G34, G35. Key Wo ds: C oss lis ing, co po a e alua ion, eme ging ma ke s. Acknowledgemen s: I would like o acknowledge he eceip o inancial suppo om he Maynoo h Finance Resea ch G oup (MFRG) and Ins i u e o In e na ional In eg a ion S udies a T ini y College Dublin. Help ul and in aluable commen s we e g a e ully ecei ed om Tom Fla in, Donal O'Neill, Denis Conni e, Ge ald Dwye Jn ., Ad iana and Pio Ko czak, Vicen iu Co ig and om pa icipan s a he IIIS Minicon e ence on In e na ional Financial In eg a ion, Maynoo h No embe 2004, he Maynoo h Pos g adua e Colloquium, Maynoo h Ma ch 2005, he Eme ging Ma ke s Finance Con e ence, CASS Business School, May 2005, and he Global Finance Con e ence, Dublin, June 2005. All emaining e o s a e en i ely my own. 1. In oduc ion ♣ Depa men o Economics, NUI Maynoo h, Maynoo h, Co. Kilda e, I eland. Con ac in o ma ion: Tel: +353 1 7083452 E-mail Add ess: [email p o ec ed] This pape was p esen ed a he Money Mac o Finance (MMF) Annual Con e ence (2005) unde he heading “C oss lis ing in he U.S: co ela ed and causal e ec s”. This pape is a e ised e sion o Chap e 5 o my PhD hesis a NUI Maynoo h. 2 Up o ecen ly, he p e ailing wisdom dec eed ha he g ea es gains o exchange-c oss-lis ing in he U.S. should acc ue o hose i ms domiciled in coun ies whe e in es o s a e poo ly p o ec ed. G ounded in wha is commonly e e ed o as he ‘legal inance’ li e a u e, an exchange-c oss-lis ing in he U.S. p o ides a emedy o i ms o o e come hei inancing cons ain s a home, and hus acili a e hei hi he o s agna ed g ow h (See La Po a e al., 1998 and Deme guc-Kun and Maksimo ic, 1998 o he legal inance iew, and Co ee, 1999, 2002 and Lins e al., 2005 o a gumen s speci ic o c oss-lis ing). Al hough his op ion is no cheap, and is g ea es o eme ging ma ke i ms (i.e. cos o U.S. G.A.A.P. compliance), he belie was ha he po en ial bene i s om lis ing, would be mo e han su icien o mee hese addi ional on-going cos s. In ac , in hei heo e ical model, Doidge e al. (2004), using alua ion me ics heo ize ha he g ea es alua ion gains o lis ing, wha hey e e o as a ‘c oss lis ing p emium’, should acc ue o i ms domiciled in eme ging ma ke s. B is and B isley (2006) each simila heo e ical conclusions. In suppo , using s anda d e en s udy analysis, Mille (1999) and Se a (1999) ind empi ical indings in a ou o his iew. Howe e , his iew has been ecen ly challenged. Fi s , in con as o Doidge e al. (2004), K is ian-Hope e al. (2007), also using alua ion me ics conclude ha exchange- aded i ms om low-disclosu e egimes ecei e a lowe alua ion han hei coun e pa high- disclosu e domiciled i ms. They heo ize ha he bene i s om lis ing may no be su icien o mee he sizable cos s (o la ge enough o ensu e ha low disclosu e egime i ms gain mos ). As a esul , exchange- aded i ms om high disclosu e egimes gain mos om lis ing1. Second, and consis en wi h his iew, he numbe o i ms ha ha e de-lis ed om deposi a y eceip p og ams has in ensi ied in he las ew yea s (See Wi me (2006) and Ma osi and Massoud (2006)). In many ins ances, i ms ha e ci ed he cos s associa ed wi h SEC compliance as he p ima y eason o delis ing, which ha e in ensi ied since he imposi ion o he Sa banes-Oxley Ac in 20022. Taken oge he , bo h sugges ha he bene i s om lis ing may no be su icien o mee he cos s o lis ing o eme ging ma ke i ms. This sugges s ha he alua ion gains, documen ed by Mille (1999) and Se a (1999) may be sho - li ed. In his pape , I examine he alua ion e ec s o lis ing in he U.S. o a sample o 583 c oss-lis ed eme ging ma ke i ms. Like Doidge e al. (2004) and K is ian-Hope e al. (2007), I use alua ion me ics, namely Tobin’s q. Howe e , unlike hem, I abs ac om hei c oss-sec ional app oach and examine he alua ion e ec s o lis ing wi hin a panel se ing. This app oach allows me o examine he alua ion e ec s in calenda (as Doidge e al., 2004 and K is ian-Hope e al., 2007 do), bu also in e en ime. The bene i o he la e is ha i allows me o examine he e ec s o lis ing on i m alue in he sho and long un3. The p ima y d awback wi h he c oss-sec ional 1 Fi ms om high-disclosu e egimes ha e also ci ed he cos s o SEC complinace as hei p ima y eason o delis ing. Fo example, Skyepha ma, a U.K. i m delis ed om he Nasdaq in 2007 due o he “expense and bu den associa ed wi h main aining compliance wi h SEC and Nasdaq ules”. Heal hca e Finance, Tax & Law Weekly. 2 The g ea e cos s o lis ing pos -Sa banes-Oxley a e no he only easons o olun a y i m delis ings. Chaplinsky and Ramchand (2007) show ha a sizable p opo ion o hose ha olun a ily c oss-delis a e low quali y i ms i.e. i ms wi h low a e age p o i abili y, asse s, and ma ke capi aliza ion, poo s ock p ice pe o mance (50% decline since lis ing), and no analys co e age (60%). In connec ion, Hos ak e al. (2006) show ha hose i ms ha olun a ily c oss-delis ha e weake co po a e go e nance sys ems. 3 I could, o cou se ha e examined he long- e m alua ion e ec s o lis ing using he e en -s udy app oach o Mille (1999) and Se a (1999). Howe e , his app oach has wo majo d awbacks. Fi s , Ko ha i and Wa ne (2005) highligh he limi a ions o long-ho izon e en s udy me hods. Second, e en s udies may no adequa ely con ol o sel -selec ion bias. These conce ns a e oiced by Heidle (2003) and Mi oo (2003). In his synopsis o Mi oo (2003), Heidle (2003, pg. 1664) concludes, “As wi h all e en s udies, he analysis in his pape su e s om a po en ial sel -selec ion bias”. In ac Mi oo (2003, pg. 1659) explici ly acknowledges his sho coming in he conclusion, “…long- e m pe o mance is gene ally di icul o measu e and ou esul s should be in e p e ed wi h some cau ion because o se e al limi a ions o ou 3 app oach o Doidge e al. (2004) and K is ian-Hope e al. (2007) is ha he e ec o lis ing is es ima ed using a c oss-sec ional o i ms wi h a ying deg ees o exposu e o lis ing. Thus, in e ec , he c oss-sec ional es ima e assumes ha gains om lis ing a e homogenous in each pos -lis ing e en yea . Howe e , he e is ample e idence o sugges ha his is no he case. Using a se ies o pooled o dina y leas squa es and ea men e ec eg essions, I ind suppo in a ou o he p edic ions o K is ian-Hope e al. (2007). Fo eme ging ma ke Le el 2/3 issues, he alua ion gains acc ue only o i ms om coun ies whe e in es o s a e well p o ec ed. Howe e , he gains a e no immedia e. I ind ha he gains o lis ing ma e ialize, once he i m has lis ed in he U.S. o a leas i e yea s. I also documen long- e m, bu no immedia e alua ion gains o Le el 1 o e - he-coun e issues. In con as o Le el 2/3 issues, he gains a e concen a ed amongs i ms om low-disclosu e egimes. I ind no posi i e alua ion e ec s o Rule 144a p i a e placemen s. The pape p oceeds as ollows. In he nex sec ion, I ou line he sample o eme ging ma ke i ms. In Sec ion 3, I p esen uni a ia e s a is ics and p oceed o ou line and discuss he eg ession analysis employed (pooled o dina y leas squa es and ea men e ec s). Sec ion 4 concludes. 2. Da a I begin by sou cing a ull lis o eme ging ma ke coun ies wi h i ms c oss-lis ed in he Uni ed S a es. F om each, I iden i y hose i ms wi h a c oss lis ing in he U.S. All in o ma ion on c oss-lis ed i ms is sou ced om he Bank o New Yo k, and c oss- e e enced wi h in o ma ion om Deu sche Bank, JP Mo gan, he New Yo k S ock Exchange, and Nasdaq. F om my c oss-lis ed sample o i ms, I classi y i ms acco ding o hei i s c oss lis ing, and classi y simul aneous Le el 1/Po al ‘lis ings’ as Le el 1 issues. My inal sample (Table 1) is comp ised o 4,563 non-c oss-lis ed, non- inancial i ms and 583 c oss-lis ed i ms. The c oss-lis ed sample is comp ised o 260 Le el 1 i ms, 142 exchange-lis ed Le el 2/3 issues and 181 i ms ha ade unde Rule144a. I supplemen my o iginal sample o 4,563 non-c oss-lis ed non- inancial i ms, wi h an addi ional 1,031 inancial i ms o ensu e app op ia e ma ches o ou inancial c oss-lis ed i ms. I do no include hese inancial i ms in my ixed- e ec s, pooled o dina y leas squa es, and ea men -e ec s models since he alua ion a ios o inancial i ms a e no compa able o hose o non- inancial i ms. Thus, he elimina ion o inancial i ms acili a es a g ea e compa ison o i ms ac oss coun ies. Finally, I only include i ms wi h a e age o al asse s g ea e han 100 million U.S. dolla s. This la e app oach acili a es a g ea e compa ison be ween c oss-lis ed and non c oss-lis ed i ms. In Table 1 I ou line by coun y he numbe o non-c oss-lis ed i ms, and he numbe o c oss-lis ed i ms lis ed in he Uni ed S a es by deposi a y eceip le el. I p o ide he pe cen age ha each coun y (i.e. numbe o i ms) con ibu es o he o al numbe o i ms in each deposi a y eceip le el and adop an iden ical app oach o my non-c oss-lis ed sample. Taken oge he Sou h Ko ean and Malaysian i ms comp ise almos 28% o he non- c oss-lis ed sample: Colombian i ms con ibu e jus o e hal o 1%. Hong Kong i ms p o ide he g ea es numbe o Le el 1 i ms (37.31%), while A gen ina p o ides no i m. B azil and Mexico equally p o ide he g ea es sha e o exchange Le el 2/3 issues, while India and Taiwan supply he majo i y o i ms ha ade in he me hodology. Fi s , benchma king pe o mance wi h ma ke indexes as done in ou s udy could lead o se ious biases and measu emen p oblems”. I seek o con ol o sel -selec ion bias using i m- ixed e ec s and ea men e ec s eg essions. They a e ou lined la e in he ex . 4 U.S. unde Rule 144a on he Po al. An in e es ing ea u e e iden om Table 1 is ha ac oss and wi hin coun ies he e exis s signi ican ly di e ing p e e ences o he di e en ypes o deposi a y eceip lis ings. Fo example, he majo i y o i ms om Hong Kong ade o e - he-coun e as Le el 1 issues. This con as s no ably wi h he p e e ence o Indian and Taiwanese i ms o a Rule 144a issue. Is aeli i ms ha a e p edominan ly high- ech i ms e eal a s ong p e e ence o exchange-lis ed (Nasdaq) deposi a y eceip s. Al hough now seen as de eloped na ions, I include Hong King and Singapo e, as hey we e deemed ‘eme ging’ du ing a sizable po ion o he sample pe iod. I ollow Doidge e al. (2004, 2006, 2007), and K is ian-Hope e al. (2007) and employ Tobin’s q o measu e i m alue, whe e Tobin’s q is de ined book alue o deb + ma ke capi aliza ion book alue o asse s ⎛⎞ ⎜⎟ ⎝⎠ whe e book alue o deb is calcula ed as book alue o al asse s less he book alue o equi y. All a iables a e exp essed in local cu ency, sou ced om Wo ldscope and a e collec ed on he 31s o Decembe in each yea om 1990 o 2003. I employ he ollowing i m-le el con ol a iables in my empi ical speci ica ions: I use he a e age sales g ow h o e he las wo yea s (geome ic a e age) and Global Indus y q o accoun o i m and indus y g ow h, espec i ely. Based upon p ima y s anda d indus y classi ica ions, he (yea ly) mean Global Indus y q is calcula ed as he a e age q o all global i ms wi hin each classi ica ion. I employ o e 15,000 in e na ional i ms om he Wo ldscope da abase o calcula e he mean Global Indus y q o each yea . I use he log o o al asse s ($) o con ol o i m size. To emo e he in luence o ou lie s, I emo e he op 1% o obse a ions o Tobin’s q, and wo-yea a e age sales g ow h, and o al asse s. I include La Po a e al. (1998) coun y-le el go e nance a iables in o de o examine he alua ion e ec s o lis ing ac oss di e en go e nance egimes. I employ legal o igin (English Common, F ench, Scandina ian and Ge man Ci il Law), an i-di ec o igh s index, an equally weigh ed index o 6 di e en sha eholde igh s, which anges om a low o 0 o a high o 5, and judicial e iciency, which anges om 0 o 10. A highe a ing implies g ea e judicial e iciency. The coun y-le el go e nance a iables a e ou lined by coun y in Appendix 2. In Appendix 3, I p esen co ela ion coe icien es ima es o all i m and coun y a iables employed in he analysis. By and la ge, he co ela ion coe icien s a e o he co ec sign, and s a is ically di e en om ze o. Tobin’s q is inc easing in Global q, and an i-di ec o igh s. La ge i ms a e wo h less. In he emaining column, I p esen a iance in la ion ac o s. Mul icollinea i y is o no conce n in his s udy. 3. Empi ical esul s In his sec ion, I ou line he main esul s on c oss-lis ing and i m alue. I begin wi h uni a ia e compa isons, whe eby I compa e he alue o c oss-lis ed o non-c oss-lis ed i ms in calenda and e en ime. I hen p oceed o panel eg ession es ima es. 3.1 Yea -by-yea alua ion compa isons 5 In Table 2 (Panel A) I compa e he alue o c oss-lis ed i ms o non-c oss-lis ed i ms in each yea om 1990 o 2003. Fo each subse o c oss-lis ed i ms, I ou line he alue o he mean and median i m in each yea . In he emaining columns, I calcula e he mean alua ion di e ence [Mean Di ] be ween he a e age c oss-lis ed i m, and non-c oss-lis ed i m. Numbe s in bold signi y ha he mean di e ence is s a is ically signi ican a con en ional le els. To complemen hese numbe s, I p esen in Figu es 1-3, he mean and median alue o c oss- lis ed i ms, and he mean alue o non-c oss-lis ed i ms. The mean di e ence is calcula ed using only la ge i ms i.e. i ms wi h a e age o al asse s g ea e han one hund ed million U.S. dolla s. This acili a es a g ea e compa ison be ween c oss-lis ed and non-c oss-lis ed i ms. The median alua ion di e en ials a e a ailable om he au ho upon eques . In he inal column, I ou line he A e age E ec o he T ea men on he T ea ed [ATT] using p opensi y sco e ma ching. The ATT, calcula ed as CL NCL E(q q |D 1, X) − =, is he di e ence in alue be ween each c oss- lis ed i m, and a ma ched sample o non-c oss-lis ed i ms, wi h an almos iden ical p obabili y o lis ing i.e. p opensi y sco e. C oss-lis ed i ms a e ma ched o non-c oss-lis ed i ms based upon size ( o al asse s), g ow h ( wo-yea sales g ow h), legal o igin, and indus y g oup using p opensi y sco e ma ching. Li and Zhao (2006) adop an iden ical app oach in hei s udy o seasoned equi y o e ings. In Appendix 1, I ou line in g ea e de ail he calcula ion o he es ima ed p opensi y sco es and he ATT es ima es. The summa y measu es p esen ed in Table 2 (Panel A) a e somewha consis en wi h he indings o Doidge e al. (2006, 2007). Speci ically, he ma ching es ima es sugges ha exchange aded i ms expe ience he la ges c oss-lis ing p emia. Nex I ind ha he c oss-lis ing p emium ends o a y o e ime. Fo example, eme ging ma ke Le el 2/3 i ms a e wo h mo e, bu no s a is ically so in e e y pe iod. The c oss-lis ing p emium is g ea es o hese i ms in 1994. In con as , o Le el 1 and Rule 144a i ms, he alua ion di e ence ends o a y om discoun o p emium o e ime. 3.2 E en ime alua ion compa isons I compa e in Table 2 (Panel B), he alue o c oss-lis ed i ms o non-c oss-lis ed i ms in e en ime. I deno e he lis yea as ‘0’, and compa e c oss-lis ed o non-c oss-lis ed i ms o he i e yea s be o e o i e yea s a e lis ing. I ou line he mean alue o c oss-lis ed i ms, and calcula e he mean di e ence [Mean Di ] be ween bo h se s o i ms. The (un epo ed) median di e en ials a e simila , and a e a ailable om he au ho upon eques . In he inal column, I calcula e he ATT in e en ime ( CL NCL E(q q |D 1, X) − =). Speci ically, in he yea p eceding he lis yea , I ma ch c oss-lis ed o non-c oss-lis ed i ms based upon size ( o al asse s), g ow h ( wo-yea sales g ow h), legal o igin, and indus y g oup. I also include ime dummies o ensu e ha he ma ches a e gene a ed p io o lis ing. Fo each g oup o i ms, I ou line he ATT o each yea (including he lis yea ) up o i e yea s pos -lis ing. In Appendix 1a, I ou line he co esponding i s -s age p obi es ima es4. The column labeled ‘Ma ches’ e e s o he numbe o c oss-lis ed o non-c oss-lis ed ma ches. In Appendix 1b, I p esen al e na i e p obi speci ica ions. The ATT 4 In hei s udy, Li and Zhao (2006, pg. 358) es ima e sepa a e p opensi y sco e models o each yea . I ca y ou a simila exe cise in Table 2 (Panel A). They e ain om es ima ing a pooled p opensi y sco e model o e he en i e pe iod because o he yea -by-yea analysis p o ides a “ lexible speci ica ion o business cycle”. Al hough I am awa e o he limi a ions o he pooled speci ica ion o adequa ely accoun o business cycle e ec s, I am p ima ily mo i a ed in his pape o examine he alua ion e ec s in e en ime, and no in calenda ime. 6 es ima es a e simila , i espec i e o he p obi speci ica ion employed. Finally, I also p esen in Figu es 1-3, he ime se ies beha iou o Tobin’s o he a e age and median c oss-lis ed i m, om en yea s p io o lis ing o en yea s a e lis ing. Panel B sugges s he ollowing. Fi s , Le el 1 i ms lis a e a pe iod o poo pe o mance (i.e. alling alue). The absolu e and ela i e alue o Le el 1 i ms, alls in he p e-lis ing pe iod, and con inues o all pos - lis ing. The g ea es all-o in alue occu s in he p e-lis ing pe iod. Fo example, in he i e-yea pe iod immedia ely p io o lis ing, he a e age Le el 1 i m loses jus less han 25% o i s alue. A e lis ing in he U.S., alue con inues o all, bu a a much educed pace. Speci ically, a e i e yea s o lis ing, a e age alue alls by 14.53%. The e idence om Figu e 1 sugges s ha a e i e yea s o lis ing, mean and median alue ends o le el o . The all in absolu e alue o e his pe iod esul s in Le el 1 i ms losing hei alua ion p emium o e non- c oss-lis ed i ms. The a e age Le el 1 i m is wo h mo e han domes ic i ms p e-lis ing, bu on a pa wi h hese i ms, pos -lis ing. In con as o Le el 1 i ms, i ms ha ade unde Rule 144a appea o ‘ ime’ hei decision o lis in he U.S; o he a e age (and un epo ed median) i m, alue inc eases d ama ically in he yea s p io o lis ing, ollowed by a co esponding all o he ea e . This end is depic ed g aphically in Figu e 3. In essence, hese i ms ake ad an age o a ou able ma ke condi ions, and hus aise equi y, ia p i a e placemen s du ing boom ma ke s. Finally, he uni a ia e summa y measu es o Le el 2/3 i ms highligh ha Le el 2/3 i ms a e no wo h mo e han non-c oss-lis ed i ms in e e y yea up o i e yea s pos -lis ing. This is in con as o he calenda yea c oss lis ing p emia om Panel A. Thus, he ques ion emains; a e he esul s econcilable wi h one ano he ? I belie e so. F om Panel A, I ind ha he c oss lis ing p emia mani es in he la e hal o he sample pe iod, wi h he excep ion o 1994. In he la e yea s o he sample, he majo i y o he sample o Le el 2/3 i ms has been lis ed in he U.S. o i e yea s o mo e. The calenda yea c oss lis ing p emia may he e o e be a esul o long- e m alua ion gains om lis ing, which ha e ye o ma e ialize a e i e yea s o lis ing. I e u n o his in he nex sec ion. 3.3 Reg ession analysis In his sec ion I examine he e ec o c oss lis ing on i m alue. I begin wi h he ollowing speci ica ion, whe eby I model i m alue as a unc ion o i m cha ac e is ics: (1) i i 1 i 2 i 3 i i q X Le el 1 Le el 2/3 Rule 144a u=α+ β+δ +δ +δ + Whe e is a se o exogenous obse able cha ac e is ics o he i m, i X i Le el 1 , , i Le el 2/3 i Rule 144a a e s anda d dummy a iables ha ake he alue o 1 i he i m ades in he Uni ed S a es as a Le el 1, Le el 2/3, o unde Rule 144a on Po al, espec i ely. is a s anda d idiosync a ic dis u bance e m, and is a ec o o pa ame e s o be es ima ed. i u123 {,, , , }αβδ δ δ I explici ly acknowledge he non- andomness o he c oss-lis ed sample, and model hei decision o c oss lis as ollows: 7 * i i i * i i * i i CL Z CL 1 i CL 0 CL 0 i CL 0 = γ+η = > = < (2) Whe e ( * i CL i i i i Le el 1 ,Le el 2/3 ,Rule 144a CL ∈ ) is an unobse ed la en a iable, is a se o obse able i m-le el cha ac e is ics ha de e mine he decision o c oss-lis in he Uni ed S a es, and is a dis u bance e m. Selec ion bias a ises because o he co ela ion be ween i Z i η i i i Le el 1 ,Le el 2/3 ,Rule 144a i and u. This co ela ion can a ise in wo ins ances i.e. (1) selec ion on obse ables5 which a ises h ough co ela ion be ween i Z and i u, o (2) h ough selec ion on unobse ables i.e. co ela ion be ween i η and i u . Bo h ins ances ende o dina y leas squa es es ima es o he e ec o c oss lis ing on alue, biased. In my analysis, I es ima e he e ec o lis ing on i m alue using wo app oaches. Fi s , I es ima e i m- ixed e ec eg essions. In his speci ica ion, I assume ha he unobse ables a e ime-in a ian . Thus, he inclusion o i m- ixed e ec s is su icien o adequa ely model, and hus con ol o unobse ables. Unlike ma ching es ima es, I do no assume away unobse ables; I jus assume ha I can adequa ely con ol o hem. Second, I mus assume ha he unobse ables, in addi ion o being ime-in a ian , ha e no causal e ec in p ecipi a ing c oss lis ing (See Li and P abhala (2005) o a discussion). Nex , I explici ly model o unobse ables by p oxying o hem. To do so I es ima e a ea men e ec s model, whe eby I augmen he second s age equa ion wi h a selec ion co ec ion e m namely he in e se mills a io, om a i s -s age p obi model. The in e se mills a ios a e gene a ed on a yea -by-yea basis (using yea ly p obi models), hus esul ing in a se ies o ime- a ian unobse ables in he second s age equa ion. Nex , I ou line bo h me hods in g ea e de ail. Fi m- ixed e ec s. I begin wi h a s anda d ixed-e ec s speci ica ion. I augmen Eq. (1) wi h ime- ixed e ec s and es ima e he ollowing wo-way ixed e ec s model6: i i i 1 i 2 i 3 i i q X Le el 1 Le el 2/3 Rule 144a=α +β +δ +δ +δ +α +υ (3) α a e ime- ixed e ec s ha accoun o con empo aneous co ela ion, and i α a e i m speci ic ixed e ec s, which e lec di e ences ac oss i ms ha a e cons an , bu unobse ed o e ime. Nex , I es ima e a pooled e sion gi en my conce ns o e iola ions o s ic exogenei y7. I speci y he indi idual speci ic e ec s as Mundlak (1978) co ec ions: T i ii i s1 1 Xa, whe e X X T= α= ζ+ = ∑i . Subs i u ing in o Eq. (3) yields he ollowing: 5 I I assume selec ion on obse ables, I mus assume ha unobse ables (p i a e in o ma ion) do no in luence he decision o lis and/o in luence pos -lis ing alue. The signi icance o he ‘in e se-mills a io’ in la e ables sugges s ha his is no he case. Howe e , in Table 2 I ou line he a e age e ec o he ea men on he ea ed (ATT) o c oss-lis ed i ms, in calenda and e en ime. The ATT is he di e ence in alue be ween he c oss-lis ed and ma ched non-c oss-lis ed i m. These p opensi y sco e ma ching es ima es assume ha he decision o c oss- lis is a unc ion o obse able ac o s only. 6 The esul s om bo h he s anda d Hausman (1978) es , and Mundlak (1978) auxilia y eg ession speci ica ion con i m ha in his ins ance a andom e ec s speci ica ion is no app op ia e. 8 i i 1 i 2 i 3 i i i q X Le el 1 Le el 2/3 Rule 144a X = α+ β+δ +δ +δ + ζ+μ (4) T ea men E ec s In his sec ion I ou line a s anda d ea men e ec s model, whe eby I co ec o he p obabili y o lis ing based upon unobse able ac o s. This app oach is simila , bu no iden ical o he s anda d Heckman (1979) wo- s age es ima ion p ocedu e8. I begin by e e ing o Eq. (2). Now I assume ha he decision o c oss-lis in he Uni ed S a es is a unc ion o unobse able cha ac e is ics. Campa and Kedia (2002), Colak and Whi ed (2005), and Villalonga and Ami (2006) es ima e simila ‘pooled Heckman’ models. Thus, he impac on i m alue condi ional on being c oss-lis ed in he Uni ed S a es as: (5) i i i 1 1 i i i E(q |CL 1) X CL E( |CL 1)==α+ β+δ +υ = Gi en Eq. (2) and assuming ha he e o s e ms om bo h Eq. (1) and (2) a e bi a ia e no mal, he unobse able componen om Eq. (2), he gene alized esidual om he p obi model is de ined as: i i 1 i E(q |C 1) ( Z ) υ = =ρσ λ β (6) Whe e: i 1i i (Z) (Z) (Z) ϕ β λβ = φβ (7) The la e is commonly e e ed o as he In e se Mills Ra io, and is a se ies o ime-speci ic ‘in e se mills a ios’. In he second-s age, I add his selec ion-co ec ion e m, yielding he ollowing: (8) i i 1 1 i 1 2 i i qXC c=α+ β +δ +λβ + +υ In addi ion, I speci y he unobse ed he e ogenei y as in Mundlak (1978) i.e. T i ii i s1 1 cX a, whe e X X T= =ζ+ = i ∑ , and es ima e he ollowing: i i 1 1 i 1 2 i i qXC X=α+ β +δ +λβ + δ+υ (9) In hei pooled ‘Heckman’ speci ica ion, Dewen e e al. (2005) con ol o unobse ed he e ogenei y by es ima ing leas squa ed dummy a iable model, whe eby, as he name sugges s hey include a dummy- a iable o each i m9. Gi en he disad an age o using his app oach in la ge samples, I speci y he unobse ed he e ogenei y by including Mundlak (1978) co ec ion e ms as an addi ional se o eg esso s in Eq. (9). I es ima e ea men e ec s models o each se o c oss-lis ed i ms sepa a ely. The coe icien es ima es o Eqs. (3, 4, 9) a e p esen ed in Table 3 and sugges he ollowing. Fi s , and in line wi h Doidge e al. (2004, 2006, 2007) I ind a s a is ically signi ican c oss lis ing p emium o Le el 2/3 issues. 7 I o mally es o his possibili y, ollowing Woold idge (2002), by inse ing he one-yea o wa ded c oss-lis ing a iables as independen a iables and es ing whe he hei coe icien s a e join ly equal o ze o. 8 Technically, he Heckman (1979) wo-s age p ocedu e is no a ea men e ec s model. In addi ion o he s anda d Heckman (1979) model, a ea men e ec s model includes, unlike he Heckman (1979) model, he selec ion indica o om he i s s age p obi as a eg esso in he second- s age eg ession. 9 I would like o hank Ka h yn Dewen e and Wal e No aes o cla i ying o me hei es ima ion p ocedu e. 9 Re e ences 1. Blundell, R., and Cos a Dias, M., 2000. E alua ion me hods o non-expe imen al da a. Fiscal S udies 21, 427-468. 2. B is, A., and B isley, N., 2006. A heo y o op imal exp op ia ion, me ge s and indus y compe i ion. Yale wo king pape . 3. Campa, J.M., and Kedia, S., 2002. Explaining he di e si ica ion discoun . Jou nal o Finance, 4, 1731-1762. 4. Chaplinsky, S., and Ramchand, L., 2007. F om lis ing o delis ing: o eign i ms en y and exi om he U.S. Uni e si y o Vi ginia wo king pape . 5. Co ee, J., 1999. The u u e as his o y: The p ospec s o global con e gence in co po a e go e nance and i s implica ions. No hwes e n Uni e si y Law Re iew 93, 641-708. 6. Co ee, J., 2002. Racing owa ds he op: The impac o c oss-lis ings and s ock ma ke compe i ion on in e na ional co po a e go e nance. Columbia Law Re iew, 102, 7, 1757-1832. 7. Colak, G., and Whi ed, T.M., 2005. Spin-o s, di es i u es, and conglome a e in es men . Re iew o Financial S udies (Fo hcoming). 8. Dehejia, R., 2005. P ac ical p opensi y sco e ma ching: a eply o Smi h and Todd. Jou nal o Econome ics, 125, 355-364. 9. Demi guc-Kun , A., and Maksimo ic, V., 1998. Law, inance, and i m g ow h. Jou nal o Finance, 53, 2107-2137. 10. Dewen e , K., Kim, C., Lim, U., and No aes, W., 2005. Commi ing o p o ec in es o s in eme ging ma ke s: can local exchanges p o ide alue-enhancing bonding mechanisms? Uni e si y o Washing on wo king pape . 11. Doidge, C., Ka olyi, G.A., S ulz, R.M., 2004. Why a e o eign i ms lis ed in he U.S. wo h mo e? Jou nal o Financial Economics, 71, 205-238. 12. Doidge, C., Ka olyi, G.A., and S ulz, R.M., 2006. The alua ion p emium o non-U.S. s ocks lis ed in U.S. ma ke s. NYSE Resea ch Pape . 13. Doidge, C., Ka olyi, G.A., and S ulz, R.M., 2007. Has New Yo k become less compe i i e in global ma ke s? E alua ing o eign lis ing choices o e ime. Fishe College o Business, Ohio S a e Uni e si y wo king pape . 14. Hausman, J.A., 1978. Speci ica ion es s in econome ics. Econome ica, 46, 1251-1271. 15. Heckman, J.J., 1979. Sample selec ion as a speci ica ion e o . Econome ics 47, 153-161. 16. Heidle, H.G., 2003. Discussion o ‘Globaliza ion and he alue o a U.S. lis ing: Re isi ing Canadian e idence’. Jou nal o Banking and Finance 27, 1663-1665. 17. Hos ak, P., Lys, T., and Yang, Y., 2006. Is he Sa banes-Oxley Ac sca ing away lemons o o anges? An examina ion o he impac o he Sa banes-Oxley Ac on he a ac i eness o U.S. capi al ma ke s o o eign ma ke s. No hwes e n Uni e si y wo king pape . 18. Ko ha i, S.P., Wa ne , S.B., 2005. Econome ics o e en s udies. Handbook o Co po a e Finance: Empi ical Co po a e Finance, B. Espen Eckbo, ed., Else ie /No h-Holland. 16 19. K is ian Hope, O., T. Kang, and Zang, Y., 2007. Bonding o he imp o ed disclosu e en i onmen in he Uni ed S a es: Fi ms’ lis ing choices and hei capi al ma ke consequences. Uni e si y o To on o wo king pape . 20. La Po a, R., F. Lopez-de-Silanes, Shlei e , A., and Vishny, R., 1998. Law and inance. Jou nal o Poli ical Economy, 106, 1113-1155. 21. La Po a, R., Lopez-de-Silanes, F, Shlei le , A, and Vishny, R., 2002. In es o p o ec ion and co po a e alua ion. Jou nal o Finance 57, 1147-1170. 22. Li, K., and P abhala, N.R., 2005. Sel -selec ion models in co po a e inance. B. Espen Eckbo (ed.) Handbook o Co po a e Finance: Empi ical Co po a e Finance, Chap e 1 2005. 23. Li, X., and Zhao, X., 2006. P opensi y sco e ma ching and abno mal pe o mance a e seasoned equi y o e ings. Jou nal o Empi ical Finance, 13, 351-370. 24. Lins, K., D, S ickland, and Zenne , M., 2005. Do non-U.S. i ms issue equi y on U.S. s ock exchanges o elax capi al cons ain s? Jou nal o Financial and Quan i a i e Analysis, 40, 109-133. 25. Ma osi, A., and Massoud, N., 2006. You can en e bu you canno lea e – U.S. secu i ies ma ke s and o eign i ms. Uni e si y o Albe a wo king pape . 26. Mille , D., 1999. The ma ke eac ion o in e na ional c oss lis ing: e idence om deposi a y eceip s. Jou nal o Financial Economics, 51, 103-123. 27. Mi oo, U.R., 2003. Globaliza ion and he alue o U.S. lis ing: e isi ing Canadian e idence. Jou nal o Banking and Finance, 27, 1629-1661. 28. Mundlak, Y., 1978. On he pooling o ime se ies and c oss sec ion da a. Econome ica, 1, 69-85. 29. Reese, W., and Weisbach, M., 2002. P o ec ion o mino i y sha eholde in e es s, c oss-lis ings in he Uni ed S a es, and subsequen equi y o e ings. Jou nal o Financial Economics 66, 65-104. 30. Sa kissian, S., and Schill, M.J., 2007. A e he e pe manen alua ion gains o o e seas lis ings? E idence om ma ke sequencing and selec ion. Re iew o Financial S udies (Fo hcoming). 31. Se a, A.P., 1999. Dual-lis ings on in e na ional exchanges: he case o eme ging ma ke s’ s ocks. Eu opean Financial Managemen 5, 165-202. 32. Villalonga, B., and Ami , R., 2006. How do amily owne ship, con ol and managemen a ec i m alue? Jou nal o Financial Economics, 80, 2, 33385-417. 33. Wi me , J., 2006. Why do i ms c oss-delis ? An examina ion o he de e minan s and e ec s o c oss delis ing. Bank o Canada wo king pape . 17 Table 1: Sample Desc ip ion Coun y NCL SIC 6 % Le el 1 % Le el 2/3 % Rule 144a % To al CL Sample A gen ina 60 7 1.31 0 0.00 11 7.75 5 2.76 16 76 B azil 246 29 5.39 26 10.00 25 17.61 3 1.66 54 300 Chile 113 35 2.48 2 0.77 17 11.97 2 1.10 21 134 China 89 4 1.95 8 3.08 12 8.45 4 2.21 24 113 Colombia 27 6 0.59 1 0.38 1 0.70 4 2.21 6 33 Hong Kong 540 167 11.83 97 37.31 7 4.93 1 0.55 105 645 Hunga y 23 4 0.50 2 0.77 1 0.70 9 4.97 12 35 India 278 23 6.09 5 1.92 9 6.34 50 27.62 64 342 Is ael 83 16 1.82 1 0.38 8 5.63 0 0.00 9 92 Ko ea 636 74 13.94 4 1.54 7 4.93 20 11.05 31 667 Malaysia 638 153 13.98 12 4.62 0 0.00 0 0.00 12 650 Mexico 71 14 1.56 18 6.92 25 17.61 11 6.08 54 125 Pe u 45 8 0.99 3 1.15 1 0.70 1 0.55 5 50 Philippines 110 70 2.41 5 1.92 1 0.70 6 3.31 12 122 Poland 56 15 1.23 1 0.38 1 0.70 11 6.08 13 69 Singapo e 407 67 8.92 19 7.31 1 0.70 1 0.55 21 428 Sou h A ica 313 151 6.86 37 14.23 8 5.63 3 1.66 48 361 Taiwan 404 60 8.85 0 0.00 6 4.23 42 23.20 48 452 Thailand 296 98 6.49 14 5.38 0 0.00 1 0.55 15 311 Tu key 128 30 2.81 5 1.92 1 0.70 7 3.87 13 141 To al 4,563 1,031 100% 260 100% 142 100% 181 100% 583 5,146 This able ou lines he inal sample. To enable ma ching o inancial c oss-lis ed i ms, I include a se o non-c oss-lis ed inancial i ms (ou lined in column 3). These i ms a e no included in he alua ion eg essions. All i ms a e ob ained om he Wo ldscope Coun y Lis s. All in o ma ion on i ms c oss-lis ed in he U.S. a e ob ained om he Bank o New Yo k, and c oss- e e enced wi h da a p o ided by Deu sche-Bank, JP Mo gan and Ci ibank. Rule 144a ADRs ade on Po al; Le el 1 ADRs ade o e - he-coun e as pink shee issues, and Le el 2/3 ade on he NYSE o NASDAQ. 18 Table 2: Valua ion compa ison o c oss-lis ed and non-c oss-lis ed i ms in calenda and e en ime. Le el 1 Le el 2/3 Rule 144a CL Med CL Mean Mean Di PS ATT CL Med CL Mean Mean Di PS ATT CL Med CL Mean Mean Di PS ATT Panel A Calenda Time 1990 1.40 1.49 (0.22) (0.20) 1.42 1.42 (0.29) - - - - - 1991 1.30 1.40 (0.45) (0.57) 1.57 1.66 (0.19) - 1.17 1.17 (0.68) - 1992 1.40 1.78 (0.05) (0.06) 1.86 1.81 (0.02) - 1.76 1.82 (0.01) - 1993 1.67 1.83 (0.06) (0.31) 2.20 2.11 0.22 - 1.71 1.77 (0.12) 0.64 1994 1.70 1.81 (0.28) (0.26) 1.99 2.05 (0.04) 0.70 2.40 2.61 0.52 0.67 1995 1.56 1.75 (0.11) 0.20 1.64 1.68 (0.18) 0.03 1.84 1.96 0.10 0.19 1996 1.43 1.73 (0.10) 0.32 1.63 1.79 (0.04) 0.19 1.51 1.70 (0.13) (0.18) 1997 1.54 1.79 0.00 0.51 1.70 1.96 0.17 0.32 1.53 1.76 (0.03) 0.16 1998 1.20 1.42 (0.03) 0.33 1.28 1.44 (0.01) 0.26 1.26 1.50 0.05 0.18 1999 1.27 1.53 0.00 (0.07) 1.36 1.61 0.08 0.22 1.35 1.60 0.07 0.00 2000 1.29 1.52 0.01 0.01 1.39 1.70 0.19 0.41 1.31 1.64 0.13 0.10 2001 1.21 1.38 0.01 (0.05) 1.25 1.42 0.05 0.16 1.21 1.37 0.00 0.19 2002 1.23 1.41 (0.01) 0.03 1.20 1.38 (0.04) 0.37 1.24 1.43 0.01 0.09 2003 1.36 1.51 (0.01) (0.02) 1.29 1.51 (0.01) 0.32 1.38 1.54 0.02 0.21 Panel B E en Time Le el 1 Le el 2/3 Rule 144a Mean CL Mean Di PS ATT Mean CL Mean Di PS ATT Mean CL Mean Di PS ATT -5 2.31 0.65*** - 1.81 0.15 - 1.57 (0.09) - -4 1.89 0.23*** - 1.61 (0.05) - 1.66 0.00 - -3 1.78 0.12** - 1.61 (0.05) - 1.91 0.25*** - -2 1.72 0.06* - 1.73 0.07 - 2.24 0.58*** - -1 1.77 0.11** - 1.73 0.07 - 2.10 0.44*** - 0 1.72 0.06 (0.12) 1.70 0.04 0.32 2.18 0.52*** 0.42** 1 1.63 (0.03) (0.02) 1.60 (0.06) 0.15 1.96 0.30*** 0.22 2 1.55 (0.11) 0.07 1.57 (0.09) 0.12 1.72 0.06 0.21 3 1.61 (0.05) 0.14 1.62 (0.04) 0.18 1.64 (0.02) 0.29 4 1.60 (0.06) (0.04) 1.56 (0.10) 0.06 1.50 (0.16) (0.03) 5 1.47 (0.19) 0.14 1.48 (0.18) 0.15 1.44 (0.22) (0.07) E en Time (Be o e-A e Value) Mean Median Mean Median Mean Median Be o e 1.87 1.59 1.66 1.45 1.89 1.43 A e 1.56 1.33 1.59 1.36 1.62 1.34 Di e ence (0.31)*** (0.26)*** (0.07) (0.09) (0.27)*** (0.09) This able compa es he mean pe o mance o c oss-lis ed (Le el 1, Le el 2/3, and Rule 144a) o non-c oss-lis ed i ms in each yea om 1990 o 2003, and in e en ime (5 yea s p e-lis ing o 5 yea s pos -lis ing). Fi m alue is p oxied using Tobin’s q, whe e. Fo each subse o c oss-lis ed i ms, I also ou line he a e age e ec o he ea men on he ea ed [] using p opensi y sco e ma ching [PS ATT], in calenda and e en ime. In he Appendix I ou line exac ly how I calcula e he ATT. Fo c oss-lis ed i ms, I also ou line hei median alue. ***, **, * Rep esen s signi icance a he 1%, 5%, and 10% le el, espec i ely. CL NCL E(q q |D 1, X)−= 19 Table 3: Reg ession es ima es o he impac o c oss-lis ing on i m alue. Le el 1 Le el 2/3 Rule 144a POLS FE TE POLS FE TE POLS FE TE Le el 1 0.04 [0.55] -0.07 [1.55] 0.07 [0.85] Le el 2/3 0.24 [2.89]*** 0.16 [2.07]** 0.15 [1.57] Rule 144a 0.03 [0.43] 0.08 [1.23] 0.04 [0.57] Global q 1.13 [8.05]*** 0.60 [7.56]*** 1.10 [7.75]*** 1.12 [8.11]*** 0.60 [7.58]*** 1.05 [7.55]*** 1.13 [8.07]*** 0.60 [7.62]*** 1.12 [7.96]*** Sales G ow h 0.54 [4.83]*** 0.45 [7.21]*** 0.44 [3.90]*** 0.53 [4.74]*** 0.45 [7.29]*** 0.50 [4.43]*** 0.53 [4.80]*** 0.45 [7.26]*** 0.42 [3.70]*** To al Asse s -0.10 [5.95]*** -0.26 [14.40]*** -0.10 [6.25]*** -0.26 [14.64]*** -0.10 [5.75]*** -0.26 [14.54]*** An i-Di ec o 0.12 [8.81]*** 0.12 [9.21]*** 0.12 [9.03]*** Lambda () λ - - 0.01 [1.17] - - 0.04 [5.49]*** - - 0.07 [5.62]*** Time Dummies No Yes Yes No Yes Yes No Yes Yes # Obs 7,001 7,426 7,426 7,001 7,426 7,426 7,001 7,426 7,426 2 R 0.14 0.07 0.09 0.14 0.08 0.10 0.14 0.08 0.10 P F> 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 P F(Mundlak)> 0.000 - - 0.000 - - 0.000 - - This able p esen s coe icien es ima es om panel eg essions. I p esen h ee se s o panel da a es ima es; pooled o dina y leas squa es wi h Mundlak (1978) co ec ions [POLS], i m- ixed e ec s [FE], and T ea men E ec s [TE]. The T ea men E ec s eg essions a e es ima ed as h ee sepa a e eg essions based upon he di e en ADR sub-sample o i ms. Fo each ADR le el, we es ima e a i s -s age p obi model whe e he decision o lis is de e mined in e ms o size (Log (To al Asse s)), and Legal O igin (F ench, Ge man). To sa is y he exclusion es ic ions, hese a iables a e excluded in he second-s age eg essions. The i s s age p obi es ima es a e a ailable om he au ho upon eques . Fi m alue is p oxied using Tobin’s q. The independen a iables a e de ined in he ex . A ull se o yea speci ic dummy a iables a e epo ed (excep in he case o he pooled o dina y leas squa es es ima es) bu no epo ed. I epo -s a is ics in pa en heses. The pooled o dina y leas squa es -s a is ics a e calcula ed using s anda d e o s clus e ed a he i m le el. # Obs is he numbe o obse a ions and is he coe icien o de e mina ion (I epo he o e all o he i m- ixed e ec es ima es). I epo wo F-S a s: (join signi icance o all RHS a iables) and which es s he join signi icance o he included (un epo ed) Mundlak (1978) ime-a e aged co ec ion e ms. ***, **, * Rep esen s signi icance a he 1%, 5%, and 10% le el, espec i ely. 2 R 2 RP F> P F(Mundlak)> 20 Table 4: Reg ession es ima es based on coun y legal cha ac e is ics. An i-Di ec o Righ s Index Judicial E iciency English Common Law All Abo e Median Below Median Abo e Median Below Median Common Law Ci il Law Le el 1 -0.02 [0.20] 0.25 [1.78]* -0.03 [0.24] 0.26 [3.01]*** 0.04 [0.39] 0.08 [0.61] 0.28 [2.62]*** Le el 2/3 0.20 [1.67]* 0.16 [1.39] 0.39 [2.15]** 0.12 [1.25] 0.51 [3.04]*** 0.15 [1.74]* Rule 144a -0.08 [0.70] 0.24 [3.09]*** -0.03 [0.25] 0.18 [2.29]** -0.10 [0.84] 0.25 [3.42]*** -0.03 [0.40] Global q 0.79 [4.10]*** 1.38 [7.49]*** 0.91 [4.66]*** 1.26 [6.82]*** 0.84 [4.77]*** 1.36 [6.99]*** 1.13 [8.05]*** Sales G ow h 0.11 [0.75] 0.87 [5.87]*** 0.11 [0.66] 0.76 [5.40]*** 0.31 [1.97]** 0.73 [5.03]*** 0.44 [3.74]*** To al Asse s -0.10 [3.62]*** -0.12 [6.57]*** -0.10 [3.64]*** -0.11 [5.93]*** -0.10 [3.54]*** -0.12 [6.38]*** -0.11 [6.35]*** Le el 1 * Sales G ow h * An i -0.21 [2.83]*** Rule 144a * Sales G ow h * An i 0.16 [1.30] Time Dummies No No No No No No No # Obs 3,863 3,563 3,259 4,167 3,835 3,591 7,027 2 R 0.12 0.17 0.13 0.14 0.12 0.16 0.11 P F> 0.000 0.000 0.000 0.000 0.000 0.000 0.000 P F(Mundlak)> 0.000 0.000 0.000 0.000 0.000 0.000 0.000 This able p esen s coe icien es ima es om pooled o dina y leas squa es eg essions (wi h s anda d e o s clus e ed a he le el o he i m) based on legal cha ac e is ics. These cha ac e is ics a e An i-Di ec o Righ s Index, Judicial E iciency, and Legal O igin (English common law o no ). All h ee a iables a e aken om and de ined in La Po a, Lopez-de-Silanes, Shlei e , and Vishny (1997). Fi m alue is p oxied using Tobin’s q. The independen a iables a e de ined in he ex . . # Obs is he numbe o obse a ions and is he coe icien o de e mina ion (I epo he o e all o he i m- ixed e ec es ima es). I epo wo F-S a s: (join signi icance o all RHS a iables) and which es s he join signi icance o he included (un epo ed) Mundlak (1978) ime-a e aged co ec ion e ms. ***, **, * Rep esen s signi icance a he 1%, 5%, and 10% le el, espec i ely. 2 R2 R P F>P F(Mundlak)> 21 Table 5: C oss-lis ing and he e olu ion o Tobin’s q. Le el 1 Le el 2/3 Rule 144a All Com Ci il All Com Ci il All Com Ci il Lis yea 0.03 [0.37] -0.02 [0.18] -0.20 [2.58]*** 0.38 [1.83]* 0.66 [1.73]* 0.33 [1.43] 0.34 [2.33]** 0.60 [1.46] -0.20 [2.58]*** 1 yea a e lis 0.08 [0.99] 0.06 [0.55] -0.18 [2.31]** 0.21 [1.42] 0.49 [1.46] 0.20 [1.40] 0.23 [1.83]* -0.07 [0.32] -0.18 [2.31]** 2 yea s a e lis 0.07 [0.73] 0.01 [0.07] -0.02 [0.21] 0.02 [0.14] 0.35 [1.37] 0.06 [0.42] 0.12 [1.31] -0.16 [0.93] -0.02 [0.21] 3 yea s a e lis 0.03 [0.30] -0.06 [0.47] 0.05 [0.38] 0.10 [0.86] 0.41 [1.82]* 0.10 [0.85] 0.08 [0.93] -0.25 [1.94]* 0.05 [0.38] 4 yea s a e lis 0.17 [1.50] 0.16 [1.07] 0.04 [0.22] 0.06 [0.56] 0.56 [1.61] -0.02 [0.23] 0.06 [0.80] -0.03 [0.17] 0.04 [0.22] 5 yea s a e lis 0.13 [1.16] 0.08 [0.57] 0.09 [0.52] 0.02 [0.23] 0.52 [4.98]*** -0.02 [0.23] 0.01 [0.05] -0.24 [2.52]** 0.09 [0.52] > 5 yea s a e lis 0.22 [1.69]* 0.11 [0.67] 0.32 [1.60] 0.21 [1.27] 1.09 [4.62]*** 0.04 [0.41] -0.06 [0.87] -0.17 [1.39] 0.32 [1.60] Global q 1.16 [8.24]*** 0.86 [4.79]*** 1.39 [6.93]*** 1.16 [8.28]*** 0.82 [4.75]*** 1.38 [6.95]*** 1.15 [8.25]*** 0.85 [4.72]*** 1.39 [6.93]*** Sales g ow h 0.41 [3.75]*** 0.33 [2.15]** 0.73 [5.02]*** 0.39 [3.52]*** 0.31 [1.97]** 0.71 [4.83]*** 0.39 [3.52]*** 0.30 [1.94]* 0.73 [5.02]*** To al Asse s -0.10 [6.53]*** -0.09 [3.36]*** -0.10 [5.33]*** -0.10 [6.50]*** -0.10 [3.62]*** -0.10 [5.46]*** -0.10 [6.22]*** -0.08 [3.12]*** -0.09 [5.33]*** Time Dummies No No No No No No No No No # Obs 7,452 3,853 3,599 7,452 3,853 3,599 7,452 3,853 3,599 2 R 0.11 0.11 0.15 0.11 0.12 0.15 0.11 0.11 0.15 P F> 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 P F(Mundlak)> 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 This able p esen s coe icien es ima es om pooled o dina y leas squa es eg essions (wi h s anda d e o s clus e ed a he le el o he i m) o he ull sample and by legal o igin. ‘Com’ is English common law. Ci il is ci il law. Fi m alue is p oxied using Tobin’s q. The independen a iables a e de ined in he ex . The single yea c oss-lis ing dummies equal one in he e e ed yea , and ze o o he wise. The ‘> 5 yea s a e lis ing’ dummy equals one a e he i h yea o lis ing and ze o be o e. . # Obs is he numbe o obse a ions and is he coe icien o de e mina ion (I epo he o e all o he i m- ixed e ec es ima es). I epo wo F-S a s: (join signi icance o all RHS a iables) and which es s he join signi icance o he included (un epo ed) Mundlak (1978) ime-a e aged co ec ion e ms. ***, **, * Rep esen s signi icance a he 1%, 5%, and 10% le el, espec i ely. 2 R2 R P F>P F(Mundlak)> 22 Figu e 1 Value o Le el 1 i ms in calenda and e en ime 23 Figu e 2 Value o Le el 2/3 i ms in calenda and e en ime 24 Figu e 3 Value o Rule 144a i ms in calenda and e en ime 25