An empirical study of the performance of South African conglomerates

A leck-G a es, J. F.; Bu , G. H.; Cleasby, J. M. A icle An empi ical s udy o he pe o mance o Sou h A ican conglome a es Sou h A ican Jou nal o Business Managemen P o ided in Coope a ion wi h: Uni e si y o S ellenbosch Business School (USB), Bell ille, Sou h A ica Sugges ed Ci a ion: A leck-G a es, J. F.; Bu , G. H.; Cleasby, J. M. (1989) : An empi ical s udy o he pe o mance o Sou h A ican conglome a es, Sou h A ican Jou nal o Business Managemen , ISSN 2078-5976, A ican Online Scien i ic In o ma ion Sys ems (AOSIS), Cape Town, Vol. 20, Iss. 1, pp. 1-6, h ps://doi.o g/10.4102/sajbm. 20i1.932 This Ve sion is a ailable a : h ps://hdl.handle.ne /10419/218009 S anda d-Nu zungsbedingungen: Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen Zwecken und zum P i a geb auch gespeiche und kopie we den. Sie dü en die Dokumen e nich ü ö en liche ode komme zielle Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich machen, e eiben ode ande wei ig nu zen. So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen (insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en, gel en abweichend on diesen Nu zungsbedingungen die in de do genann en Lizenz gewäh en Nu zungs ech e. Te ms o use: Documen s in EconS o may be sa ed and copied o you pe sonal and schola ly pu poses. You a e no o copy documen s o public o comme cial pu poses, o exhibi he documen s publicly, o make hem publicly a ailable on he in e ne , o o dis ibu e o o he wise use he documen s in public. I he documen s ha e been made a ailable unde an Open Con en Licence (especially C ea i e Commons Licences), you may exe cise u he usage igh s as speci ied in he indica ed licence. h ps://c ea i ecommons.o g/licenses/by/4.0/ S.A .J.Bus.Mgm .1989,20(1) An empi ical s udy o he pe o mance o Sou h A ican conglome a es J.F. A leck-G a es, G.H. Bu and J.M. Cleasby G adua e School o Business, Uni e si y o Cape Town, P i a e Bag, Rondebosch, 7700 Republic o Sou h A ica Accep ed 13 June 1988 ~xi~ ~n inancial heo y is unable o explain whe he on agg ega e conglome a ion is bene icial o ei he md1 1dual s?a eholde s o . ~ he econom . Bo h ad an ages and disad an ages can be lis ed o he conglome a on p~ocess and_ , s hus an empmcal ques ion as o whe he o no sha eholde s eally do bene i om c~nglome a on. In h s pape he long- e m p o i abili y o conglome a es is examined in an a emp o de e mine whe he o no such sha eholde s ea n supe io e u ns on agg ega e. This is done by con as ing he s ock ma ke pe o mance o a sample o Sou h A ican (SA) conglome a es o e a six-yea pe iod wi h he pe o mance o he o e all ma ~e . In .addi ion, hei pe o mance is con as ed wi h ha o a andom po olio o non-congl~me a e companies. Fmally, a pseudo-conglome a e po olio was cons uc ed o each conglome a e m such a way ha each po olio had he same asse s uc u e as i s ma ched conglome a e. The pe o mance o he conglome a es was hen con as ed wi h ha o he pseudo-conglome a e po olio using ma ke e u ns, e u n on asse s, and e u n on equi y. The esul s indica e ha on agg ega e, he conglome a es signi ican ly unde pe o m non-conglome a es. This is consis en wi h he iew ha conglome a ion is in he in e es o managemen a he han ha o he sha eholde s. Beskikba e inansiele eo ie kan nie e duidelik o die o ming an konglome a e, agg egaa beskou, 'n posi iewe byd ae lewe aan die indi iduele aandeelhoue o die ekonomie as geheel nie. Beide oo dele en nadele an konglome asie kan genoem wo d en di is dus 'n empi iese aag o aandeelhoue s we klik oo dcel ek ui die p oses. Hie die s udie onde soek die lang e myn winsgewendheid an konglome a e in 'n poging om as e s cl o aandeclhoue s we be e on angs e e k y. Di wo d gedoen deu die p es asie an 'n mons e an Suid-A ikaanse konglome a e op die aandelema k e kon as ee me die ma k as geheel, oo 'n pc iode an ses jaa . Ve de wo d hulle p es asie e gelyk me 'n ewekansige po olio an nie-konglome aa -maa skappye. Laas ens, 'n pseudo-konglome aa -po olio is saamges el i elke konglome aa , op so 'n wyse da elke po olio diesel de ba es uk uu as die e gelykende konglome aa gehad he . Die p es asie an die konglome a e is e gelyk deu middel an ma kopb engs, opb engs op ba es en opb engs op eienaa sbelang. Die esul a e oon da agg egaa beskou, konglome a e nie so goed aa soos nie-konglome a e nie. Di be es ig die beeld da konglome a e in die belang an bes uu en nie die an aandeelhoue s is nie. * Co espondence add ess: G.D.I. Ba , Depa men o Ma hema ical S a is ics, Uni e si y o Cape Town, P i a e Bag, Rondebosch, 7700 Republic o Sou h A ica In oduc ion Me ge s and acquisi ions ha e always played a majo ole in he g ow h o bo h indi idual i ms and he economy as a whole. By enabling success ul co po a ions o g ow apidly, me ge s and acquisi ions assis in he op imal alloca ion o economic esou ces o hose manage s and i ms bes able o u ilize he esou ces in a p oduc i e manne , he eby esul ing in bene i s o all o socie y. I is, he e o e, no su p ise ha in a g owing economy such as he Sou h A ican economy o e he pe iod 1960 o 1985, a conside able deg ee o me ge ac i i y has been e iden . The op imal alloca ion o esou ces and he abili y o ce ain i ms o u ilize hei asse s mo e e icien ly a e clea and ob ious easons o he exis ence o bo h ho izon al and e ical me ge s. in addi ion, a gumen s such as di e en ial e iciencies, syne gy, ma ke powe , and s a egic ealignmen (Copeland & Wes on, 1983:562-568) all p o ide alid jus i ica ion o his ype o me ge ac i i y. Howe e , a hi d ype o me ge ac i i y has been dominan in Sou h A ica in ecen yea s, namely conglome a ion, which occu s when wo un ela ed business uni s combine o me ge. Many possible ad an ages which can esul om conglome a ion ha e been sugges ed. These include in e alia, he inancial gains esul ing om imp o ed ea nings pe sha e (EPS) o he acqui ing i m (B igham & Gapenski, 1985: 857), syne gy in he gene al managemen unc ion (Copeland & Wes on, 1983: 569), isk educ ion h ough di e si ica ion (B ealey & Mye s, 1985: 706), he educed isk o inancial dis ess wi h i s accompanying bene i o cheape deb (Lewellen, 1971: 522), and easie access o he capi al ma ke s esul ing om he inc eased size o he co po a ion (Le y & Sa na , 1970: 798). Subsequen esea ch has shown, howe e , ha many o hese jus i ica ions may no be uni e sally alid and indeed ha in many conglome a e me ge s hey may ac ually p o e disad an ageous (B ealey & Mye s, 1985: 706). Fo example, Mye s (1976:679) has shown ha imp o emen in EPS wi hou any o he e iciency o syne gis ic bene i s will no esul in imp o ed sha eholde e u ns. Also, Galai & Masulis (1976) ha e shown ha in he absence o syne gy, he educ ion in inancial isk and cheape deb do no e en ua e. Indeed, hey show ha he equi y is mo e isky as, in essence, each se o sha eholde s mus now gua an ee he o he i m's deb in addi ion o hei own deb . Thus, hey a gue ha i is he bondholde s and o he deb p o ide s and no he equi y holde s who ecei e he bene i s o he educ ion in inancial isk. Shas i (1982) has ex ended he wo k o Galai & Masulis o show ha 2 unde some condi ions he alue o equi y in he conglome a e i m may be la ge han he sum o he equi ies in he wo p e-me ge i ms, e en in he _absence o syne gy. Howe e , Shas i's esul s also con i m ha he e a e many ci cums ances when his will no be he case and, in he absence o syne gy, he alue o he combined i m will be less han he sum o he p e- me ge alues o he componen i ms. Addi ional a gumen s agains he bene i s o conglome a e me ge s o he sha eholde s o he acqui ing company can be made in e ms o Agency Theo y (Jensen & Meckling, 1976). In e ms o his heo y, an agency p oblem a ises when manage s own only a ac ion o he i m. This may cause hem o ac in hei own in e es s a he han in hose o he sha eholde s. While compensa ion a angemen s in he ma ke o manage s may o e come some o hese agency p oblems (Fama, 1980), incen i es s ill exis o manage s o en e conglome a e me ge s in he absence o syne gis ic bene i s. Fo example, Muelle (1969) has a gued ha manage s a e mo i a ed o inc ease he size o he co po a ion since hei compensa ion is a unc ion o size. While his has no been empi ically alida ed in he US (Lewellen & Hun sman, 1970), Naude & Hipkin (1985) showed ha in he SA con ex , size was a mo e signi ican p edic o o managemen compensa ion han was p o i abili y. Addi ional a gumen s in a ou o managemen 's desi e o conglome a ion would include he desi abili y o di e si ica ion in o de o educe hei indi isible pe sonal employmen isk. F om he abo e, i is appa en ha heo e ical a gumen s can be made bo h o and agains he bene i s o conglome a ion om he o dina y sha eholde 's pe spec i e. I hus emains an empi ical ques ion o asce ain whe he , on agg ega e, conglome a ion has, ex pos , esul ed in bene i s o he o dina y sha eholde . This pape con ibu es o he deba e by p o iding e idence on he p o i abili y o conglome a e companies in he SA con ex o e he pe iod 1977 o 1983. Backg ound Nume ous s udies o he pe o mance o conglome a es ha e been conduc ed using US da a. These s udies can be oughly di ided in o wo g oups: hose ha examined he pe o mance o conglome a es in e ms o hei accoun ing pe o mance and o he ope a ing cha ac e is ics, and hose ha examined conglome a e pe o mance in e ms o s ock ma ke pe o mance. S udies in he o me g oup ha e ocused on a iables such as e u n on asse s, ea nings pe sha e, o al asse s, deb capaci y, e c. Fo example, Reid (1968) showed ha in he US, conglome a ion did no esul in inc eased EPS al hough o al asse size was inc eased; Wes on & Masingha (1971) showed ha conglome a ion esul ed in inc eased e u n on asse (ROA), while Meliche & Rush (1974) demons a ed ha conglome a e acqui e s made g ea e use o la en deb capaci y han did non-conglome a e acqui e s. On he o he hand, Mason & Goudzwaa d (1976) ound ha hei sample o 22 conglome a es had signi ican ly S. -A . Tydsk . Bed y sl .1989 ,20( 1) poo e pe o mance in e ms o bo h ROA and e u n on equi y (ROE) when compa ed o unmanaged po olios o simila indus y in es men s. The second g oup o s udies examined he pe o mance o conglome a es in e ms o s ock ma ke pe o mance. Fo example, Wes on, Smi h & Sch ie es (1972) compa ed he pe o mance o conglome a es wi h ha o mu ual unds o e he pe iod 1960 -1969. They concluded ha on a isk-adjus ed basis he conglome a es ea ned highe e u ns. This esul was con i med in a la e s udy (Smi h & Wes on, 1977). On he o he hand, Meliche & Rush (1973) compa ed he pe o mance o conglome a es wi h a ma ched sample o non-conglome a es o e he pe iod 1965 -1971, concluding ha he e we e no signi ican di e ences in he e u ns ea ned by sha eholde s in ei he g oup. In a simila ein, Joehnk & Nielsen (1974) ound no signi ican di e ences when compa ing 21 conglome a es wi h a sample o 23 non-conglome a es o e he pe iod 1962 -1969. Finally, Mason & Goudzwaa d (1976) ound ha hei sample o 22 conglome a es unde pe o med a g oup o andomly selec ed po olios ha ing simila asse s uc u es. As can be seen om he abo e, he e idence om he US is unclea as o he ac ual bene i s esul ing om conglome a ion. E idence ha conglome a es ha e bo h highe accoun ing e u ns and s ock ma ke e u ns has been p esen ed as has e idence o he con a y (i.e., lowe accoun ing e u ns and lowe s ock ma ke e u ns). In his s udy, an a emp is made o examine he pe o mance o conglome a es in he SA con ex , he eby enabling compa ison o be made be ween he SA and US en i onmen s. In addi ion, he pape con ibu ed by es ablishing a well-ma ched pseudo- conglome a e po olio which pe mi s an unambiguous e alua ion o he po en ial syne gis ic bene i s o conglome a ion. The me hodology by which his is accomplished is discussed in he ollowing sec ion. Finally, i should be poin ed ou ha his s udy examines he long- e m e ec s o conglome a ion. I is he e o e no conce ned wi h ei he he sho - e m e ec s o ake-o e s o wi h me ge s in gene al. The as body o li e a u e dealing wiih hese wo aspec s is well summa ized in Jensen & Ruback (1983) and is no discussed u he in his pape . Fo esul s on he sho - e m eac ion in he Sou h A ican con ex , he eade is e e ed o A leck-G a es, Flach & Jacobson (1988). Me hodology and Da a The hypo hesis es ed in his s udy can be s a ed as: Ho: Managemen con ol con e s ad an ages o he conglome a e i m and hence e u n o sha eholde s will be highe o conglome a e i ms han o non- conglome a e i ms. This hypo hesis is es ed o e he pe iod Janua y 4, 1977 o Janua y 4, 1983. This pe iod was selec ed as i is easonably ep esen a i e o he di e en phases o he business cycle. The s udy was es ic ed o companies quo ed on he Johannesbu g S ock Exchange (JSE) o e he en i e pe iod. S.A .J.Bus.Mgm .1989,20(1) Biso o (1980) iden i ied 31 companies quo ed on he JSE as conglome a es. These 31 companies (see Appendix A) comp ised he sample o conglome a es s udied in his pape and a e e e ed o as he CONGLOMERATE g oup. The pe o mance o each conglome a e was compa ed o ha o h ee benchma k po olios, namely: 1. A andomly selec ed po olio o single o dominan p oduc companies (called he RANDOM po olio); 2. The o e all ma ke as ep esen ed by he JSE Ac ua ies Indus ial Index (called he MARKET po olio) ; and 3. A g oup o po olios whe e each po olio was chosen o ma ch one o he conglome a e po olios in e ms o asse composi ion ( called he PSEUDO- CONGLOMERA TE po olio). The andomly selec ed po olio comp ised 20 secu i ies chosen a andom om he Indus ial Sec o o he JSE. The only es ic ions imposed on hese wen y secu i ies we e ha hey we e con inuosly aded e e he six yea s o he s udy and ha hey we e single o dominan p oduc companies. The wen y selec ed secu i ies a e lis ed in Appendix B. This po olio was cons uc ed assuming equal and in es men s in each o he wen y secu i ies. The o e all ma ke was ep esen ed by he JSE Ac ua ies Indus ial Index and p o ides an indica ion o he pe o mance o he conglome a es ela i e o he ma ke as a whole. I should be no ed, howe e , ha his does no p o ide an objec i e es o he hypo hesis as he index includes bo h conglome a e and non- conglome a e companies. Indeed, since i is a alue- weigh ed index and i is easonable o assume ha he conglome a e companies a e, on a e age, la ge han he non-conglome a e companies, i ollows ha he JSE Indus ial Index will be biased owa ds he pe o mance o he conglome a e po olio. The pseudo-conglome a e po olio was cons uc ed as ollows. Fo each o he 31 conglome a es an a emp was made o de e mine he ela i e asse sizes o he di isions wi hin he conglome a e. Because o di e ences in epo ing s anda ds, his was only possible o 15 o he 31 conglome a es. Fo each o hese 15 conglome a es a pseudo-conglome a e po olio was cons uc ed by choosing single o dominan p oduc i ms o ep esen he di e en di isions o he conglome a e. The weigh s assigned o each o hese single p oduc i ms was se equal o he p opo ion o ha di ision's asse s o he o al asse s o he conglome a e. Thus, 15 pseudo-conglome a e po olios we e cons uc ed o ma ch he 15 conglome a e i ms o which asse size pe di ision was known. These 15 pseudo-conglome a e po olios indica e he e u ns ha an in es o could ha e achie ed had he pe o med he di e si ica ion himsel a he han elying on he conglome a e o pe o m he di e si ica ion on his behal . Compa isons beween he conglome a e po olio and he ma ke and he andom po olio we e done on bo h 3 an unadjus ed and a isk-adjus ed basis. The isk adjus men was pe o med using he Sha pe (1966) isk- adjus ed pe o mance measu e: whe e SP= he Sha pe isk-adjus ed pe o mance measu e; Rp = he a e age annual e u n on he po olio in he six-yea pe iod; and SDP = he s anda d de ia ion o he annual e u ns on he po olio o e he six-yea pe iod. Compa isons be ween he conglome a e and pseudo- conglome a e po olios we e also made on a isk- adjus ed basis once again using he Sha pe isk- adjus men p ocedu e. The compa isons we e made by examining he di e ences be ween he isk-adjus ed e u n ea ned by a sha eholde in he conglome a e and he isk-adjus ed e u n ha would ha e been ea ned by an in es o in he ma ched pseudo-conglome a e po olio. This was done o each o he 15 pai s analysed and he hypo hesis es ed using he amilia pai ed es o es we he he mean di e ence was di e en om ze o. In addi ion o compa ing di e ences in s ock ma ke e u ns, he conglome a e and pseudo-conglome a e po olios we e also examined o di e ences in wo accoun ing measu es o e u n, namely e u n on asse s (ROA) and e u n on equi y (ROE). Fo he pseudo- conglome a e po olios, ROA and ROE we e aken as he weigh ed a e age o he ROA and ROE o he componen secu i ies whe e he weigh s we e once again assigned on he basis o he di isional asse s o he conglome a e ela i e o he conglome a e's o al asse s. Once again, he null hypo hesis o no di e ence be ween conglome a es and pseudo-conglome a es was es ed using he pai ed es . Resul s In his sec ion, he pe o mance o he conglome a e i ms is con as ed wi h ha o he h ee benchma ks discussed in he p e ious sec ion, namely he ma ke po olio, he andom po olio, and inally he pseudo- conglome a e po olios. We begin by compa ing he a e age annual e u n on each conglome a e wi h he a e age annual e u n on he ma ke (JSE Ac ua ies Indus ial Index) and a e age annual e u n on he andom po olio. Summa y s a is ics a e p esen ed in Table 1. F om Table 1 i is appa en ha o e he pe iod Janua y 1977 -Decembe 1983, he conglome a es yielded lowe e u ns on a e age han ei he he ma ke po olio o he andom po olio. Howe e , i is impo an o emphasize he ollowing poin s: 1. The compa ison is being made be ween indi idual conglome a es and po olios o secu i ies. This accoun s o he highe a e age s anda d de ia ion in he case o he conglome a es. 2. In no case a e he di e ences s a is ically signi ican a he 5% le el. In addi ion, when compa ing each 4 Table 1 Compa ison o conglome a es, andom po olios and he ma ke A e age e u n S anda d de ia ion Po olio (% P.A.) o e u n s a is ic Conglome a es Random po olios Ma ke 26,98 28,89 27,54 59,77 25,04 22,61 -0,16 -0,05 Table 2 Risk adjus ed compa ison o conglome a es Random po olio Ma ke # Conglome a es # Conglome a es Pe o mance wi h la ge wi h la ge measu e e u ns (EX 31) z e u ns (EX 31) z Non- isk adjus ed 12 -1,26 17 0,54 Sha pe Index 4 -4,13* 6 -2,16* • Signi ican a he 5% le el indi idual conglome a e wi h he andom po olio, a signi ican di e ence was ound in only one case ou o he 31 which is no mo e han would be expec ed by chance. 3. Risk has been igno ed in he abo e es s. I is pe cei able ha conglome a es a e less isky han indi idual p oduc i ms and ha in es o s a e p epa ed o accep he lowe a e age e u n because o he lowe isk in ol ed. Thus, he esul s in Table 1 should be in e p e ed wi h he u mos cau ion. Table 2 p o ides a compa ison be ween he conglome a e i ms and he andom and ma ke po olios on a isk-adjus ed basis, using he Sha pe isk-adjusmen p ocedu e. In his able, he numbe o conglome a es ( ou o 31) ha ing g ea e isk- adjus ed e u n han he andom po olio and he ma ke po olios was coun ed. Unde he null hypo hesis o no di e ence be ween conglome a es and non-conglome a es, he coun s would all be expec ed o be 15,5. Thus, he signi icance o he esul s can be easily es ed using he no mal app oxima ion o he binomial, i.e., Z = (P -50)/(P(lOO -P)/n)112 whe e Z= a s anda d no mal (0;1) andom a iable; P = he pe cen age o conglome a es wi h e u n g ea e han he compa ison po olio ( ei he andom o ma ke ); n = he sample size (i.e. 31). The esul s in Table 2 con i m he esul s in Table 1 in ha , on an unadjus ed basis, he di e ences be ween he conglome a es and he andom and ma ke po olios a e no s acis ically signi ican . Howe e , on a isk- adjus ed basis i is appa en ha he conglome a es' pe o mance is s a is ically di e en om ha o bo h he andom po olio and he ma ke po olio. Mo eo e , he nega i e sign indica es ha his di e ence S. -A . Tydsk . Bed y sl. 1989 ,20( ) is in he di ec ion o unde pe o mance. Thus, i is concluded ha he esul s p esen ed in Tables 1 and 2 indica e ha , on a e age, conglome a es unde pe o m in compa ison o he non-conglome a es on a isk- adjus ed basis. While he abo e esul is o in e es because i con i ms much o he empi ical e idence om o he exchanges such as New Yo k and London, he es s a e weak in ha hey con as he pe o mance o indi idual conglome a es wi h ha o po olios. Mo eo e , he esul s a e subjec o a possible selec ion bias in e ms o he andom po olio wi h espec o, o example, he di e si ica ion o his po olio and he size o he sample companies. Fo hese easons i was decided o eexamine he pe o mance o he conglome a es in compa ison o pseudo-conglome a es. This was done o each conglome a e sepa a ely and p o ides a s onge s a is ical es o he ollowing easons. Fi s ly, i o e comes many o he selec ion p oblems since each pseudo-conglome a e po olio has a simila asse cons uc ion o i s ma ched conglome a e. This o e comes he p oblems o selec ion bias wi h espec o indus ial sec o s. Secondly, i enables a di ec compa ison o he e u ns a sha eholde could ha e achie ed had he pe o med his own di e si ica ion a he han le he di e si ica ion o he managemen o he conglome a e. I should be no ed ha his esul s in a sligh bias in a ou o he conglome a es in ha i assumes ha he deg ee and ype o di e si ica ion he in es o desi ed we e exac ly he same as hose ac ually implemen ed by he managemen o he conglome a e. Finally, he use o a ma ched pai design allows o he mo e powe ul pai ed es app oach o be used in he signi icance es s. The esul s o he compa ison be ween he conglome a es and he pseudo-conglome a e po olios a e summa ized in Table 3. This able p o ides a compa ison o he pe o mance o conglome a es on bo h a ma ke e u ns basis and on he basis o wo accoun ing measu es o e u n, e u n on asse s and e u n on equi y. The esul s p esen ed in Table 3 indica e ha (a he 10% le el o signi icance), he pseudo-conglome a e po olios p o ide g ea e ma ke e u n han hei Table 3 Compa ison o conglome a es wi h pseudo- conglome a es Pe o mance measu e Ma ke e u n Re u n on asse s Re u n on equi y • Signi ican a 10% le el b Signi ican a 1 % le el alue Di ec ion o di e ence -1,80" Pseudo-conglome a es supe io -5 ,01 b Pseudo-conglome a es supe io -3 ,85b Pseudo-conglome a es supe io S.A .J .Bus.Mgm .1989,20(1) ma ched conglome a es. Since ma ke -de e mined e u n is he p ima y conce n o sha eholde s, his p o ides s ong e idence agains he bene i s o conglome a ion. As such, hey se iously ques ion he alue o conglome a ion in he SA con ex om he o dina y sha eholde s' pe spec i e. I is some imes a gued ha s udies such as he abo e do no p o ide an adequa e es because many sha eholde s may ha e a longe ime ho izon han, o example, he six yea s s udied in his s udy. P oponen s o his iew a gue ha i one examines he longe e m bene i s o conglome a ion, one will e en ually ind highe e u ns due o he mo e e icien use o asse s and he g ea e deb capaci y o conglome a es. These should be e lec ed by highe e u n on asse s (u iliza ion e iciency) and e u n on equi y. Al hough we do no ag ee wi h he a gumen s o many easons which a e well documen ed in he li e a u e ( o example, B ealey and Mye s, 1985), he abo e me hodology p o ides a eady es o hese asse ions. Acco dingly, he ma ched conglome a es and pseudo-conglome a es we e compa ed on a e u n on asse s and e u n on equi y basis. These esul s a e p esen ed in he second and hi d ows o Table 3. As can be seen, he hypo hesis o no di e ence be ween conglome a es and pseudo- conglome a es is e en mo e s ongly ejec ed han in he case o ma ke e u ns. Once again, he di ec ion o ejec ion indica es ha he pseudo-conglome a es p o ide e u ns supe io o hose o he conglome a e companies on he basis o ROA and ROE. Conclusions In his pape , he long- e m p o i abili y o conglome a es has been examined bo h in e ms o accoun ing and ma ke - ela ed measu es o pe o mance. The esul s p esen ed clea ly indica e ha conglome a es in he Sou h A ican con ex ha e signi ican ly unde pe o med in compa ison o non- conglome a e companies in e ms o bo h bases o measu emen . I is, he e o e, concluded ha conglome a ion pe se is no in he long- e m in e es s o Sou h A ican sha eholde s. Why hen, does conglome a ion con inue o be a ea u e o he SA capi al ma ke s? The esea ch p esen ed in his pape does no allow a de ini i e answe o his ques ion as i me ely p o ides e idence o subop imal pe o mance by he conglome a e g oup. Howe e , hese esul s a e consis en wi h he agency heo y a gumen s ha conglome a ion, while no in he o e all in e es s o he sha eholde s, is in he in e es s o managemen . By he p ocess o conglome a ion, manage s inc ease he asse base unde hei con ol, he eby possibly p o iding non- inancial bene i s such as s a us, powe , e c., o hemsel es. In addi ion, Naude & Hipkin (1985) ha e shown ha managemen compensa ion may be mo e highly co ela ed wi h i m size han wi h p o i abili y. Consequen ly, conglome a ion can esul in bo h non- inancial and inancial bene i s o manage s. Thus, manage s ha e clea mo i es o inc easing he size o hei companies. 5 The esul s in his pape indica e ha hey may do so o he de imen o he common s ockholde s. I mus be s essed, howe e , ha his is only one possible explana ion o he empi ical e idence p esen ed in his pape . Se e al o he explana ions a e possible. In pa icula , i is possible ha he pe iod s udied was oo sho o p o ide a de ini i e es o he alue o conglome a ion. Al e na i ely i is easible ha because i ms compe e o e icien manage s in he o e all pe sonnel ma ke , sha eholde s may be p epa ed o pay addi ional compensa ion o he mos e icien manage s in o de o a ac hem o he co po a ion. Pa o his compensa ion may be in he o m o allowing he i m size o inc ease (Fama, 1980). O cou se, his doesn' ully explain why such sha eholde s should be p epa ed o accep lowe e u ns han hey could ea n by doing hei own di e si ica ion. The only a ional explana ion would be ha hey belie ed ha in he longe e m (g ea e han he six yea s co e ed in his s udy) hey would indeed be compensa ed in e ms o s ock ma ke e u ns as a esul o hei in es men in such highly e icien manage s. Finally, i is ob ious ha he esul s p esen ed in his pape only p o ide an ini ial examina ion o he alue o conglome a ion in he SA con ex . Much addi ional esea ch is necessa y be o e many o he ques ions aised abo e can be answe ed. Acknowledgemen The au ho s wish o acknowledge he inancial assis ance ende ed by he Human Sciences Resea ch Council owa ds he cos s o his esea ch. Opinions exp essed and conclusions d awn a e hose o he au ho s and a e no o be ega ded as hose o he HSRC. The au ho s also wish o acknowledge he assis ance and aluable commen s o G.D.I. Ba and R.C. an den Hone in he p epa a ion o his pape . Re e ences A leck-G a es, J.F., Flach, T. & Jacobson, A. 1988. The E ec o Me ge Announcemen s on he Sha e P ices o he Acqui ed and Acqui ing Companies. S. A . J. Bus. Mgm ., ol. 19, 141-146. Biso o, F.E. 1980. Sou h A ican Conglome a es: A P o ile. Unpublished Technical Repo . G adua e School o Business, Uni e si y o Cape Town. B ealey, R. & Mye s, S. 1984. P inciples o Co po a e Finance. Second Edi ion. Singapo e: McG aw Hill, Inc. B igham, E.F. & Gapenski, L.C. 1985. In e media e Financial Managemen . Chicago: The D yden P ess. Copeland, T.E. & Wes on, J.F. 1983. Financial Theo y and Co po a e Policy. Second Edi ion. Reading: Addison Wesley. Fama, E.F. 1980. Agency P oblems and he Theo y o he Fi m. J. Pol. Econ., ol. 88, 288--307. Galai, D. & Masulis, R.W. 1976. The Op ion P icing Model and he Risk Fac o o S ock. J. Fin. Econ., ol.3, 5>--82. Hipkin, LB. & Naude, P. 1985. A Ma ke P icing Model o Execu i e Remune a ion. 1PM Jou nal, 14-21. 6 Jensen, M.C. & Meckling, W. 1976. Theo y o he Fi m: Manage ial Beha io , Agency Cos s and Owne ship S uc u e. J. Fin. Econ., ol. 3, 305-360. Jensen, M. C. & Ruback, R.S. 1983. The Ma ke o Co po a e Con ol: The Scien i ic E idence. J. Fin. Econ., ol.11, 5-50. Joehnk, M.D. & Nielsen, J.F. 1974. The E ec s o Conglome a e Me ge Ac i i y on Sys ema ic Risk. J. Fin. Quan . Anal., ol. 9, 215-225. Le y, H. & Sa na , M. 1970. Di e si ica ion, Po olio Analysis and he Uneasy Case o Conglome a e Me ge s. J. Fin., ol. 25, 795-802. Lewellen, W.G. 1971. A Pu e Financial Ra ionale o he Conglome a e Me ge . J. Fin., ol.26, 521-537. Lewellen, W.G. & Hun sman, 8. 1970. Manage ial Pay and Co po a e Pe o mance. Am. Econ. Re ., ol.60, 710---720. Mason, R.H. & Goudzwaa d, M.B. 1976. Pe o mance o Conglome a e Fi ms: A Po olio App oach. J. Fin., ol. 31, 39---48. Meliche , R.W. & Rush, D.F. 1973. The Pe o mance o Conglome a e Fi ms: Recen Risk and Re u n Expe ience. J. Fin., ol.28, 381-388. Meliche , R.W. & Rush, D.F. 1974. E idence on he Acquisi ion-Rela ed Pe o mance o Conglome a e Fi ms. J. Fin., ol.29, 141-149. Muelle , D.C. 1969. A Theo y o Conglome a e Me ge s. Qua . J. Econ., ol. 84, 643-659. Mye s, S.C. 1976. A F amewo k o E alua ing Me ge s. In: S.C. Mye s {Ed.) Mode n De elopmen s in Financial Managemen . New Yo k: F wdw ick Paege , Inc. Reid, S.R. 1968. Me ge s, Manage s and he Economy. New Yo k: McG aw-Hill. Sha pe, W.F. 1966. Mu ual Fund Pe o mance. J. Bus., ol.39, 119-138. Shas i, K. 1982. Valuing Co po a e Secu i ies: Some E ec s o Me ge s by Exchange O e s. Uni e si y o Pi sbu gh, Wo king pape #517. Smi h, K.V. & Wes on, J.F. 1977. Fu he E alua ion o Conglome a e Pe o mance. J. Bus. Res., ol. 5, 5-14. Wes on, J.F. & Masinghka, S.K. 1971. Tes s o he E iciency Pe o mance o Conglome a e Fi ms. J. Fin., ol.26, 919-936. Wes on, J.F., Smi h, K.V. & Sch ie es, R.E. 1972. Conglome a e Pe o mance Using he Capi al Asse P icing Model. Re . Econ. S a ., ol. 54, 357-363. Appendix A The sample o conglome a es Abe com In es men s L d. Ad anced Holdings L d. Adcock Ing am L d. Amalgama ed In es men Co po a ion L d. Anchusa Holdings L d. Anglo Ame ican Indus ial Co po a ion L d. S.-A .Tydsk .Bed y sl.1989,20(1) Appendix A Con inued Anglo-T ans aal Indus ies L d. Ba low Rand L d.• Bonusko Bpk. • Blue Ci cle L d.• B omain Holdings L d.• Ca an L d.•• Cu ie Finance Co po a ion L d. Di oyal In es men s L d. Fede ale Volksbeleggings Bpk. • Hule 's Co po a ion L d.• Indus ial and Comme cial Holdings G oup L d. Lucem Holdings L d.• O ens one In es men s L d. Pica di Beleggings Bpk. • P emie Indus ies L d. P o ea Holdings L d. The Remb and G oup L d. Rennies Consolida ed Holdings L d. Ren mees e beleggings Bpk. Sea del In es men Co po a ion L d. C.G. Smi h L d. The Sou h A ican B ewe ies L d.• Suide land De elopmen Co po a ion L d. The Tongaa G oup L d.• The Unisec G oup L d.• W. & A. In es men Co po a ion L d. • Indica es inclusion in he pseudo-conglome a e compa ison • • Included in pseudo-conglome a e compa ison bu no in o he compa isons Appendix B The andom po olio sample I. Al ech 2. B adlows 3. Claude Neon 4. Cu ie Mo o s 5. Elle ine 6. F ase s 7. Gene al Op ic 8. Gen ec 9. G oup 5 10. Gypsum 11. I.C.S. 12. Kanhym 13. Lion Ma ch 14. Me al Box 15. O.K. Bazaa s 16. Piccan 17. P e o ia Po land Cemen 18. Sa icon 19. T.E.J. 20. Willem Ba ends