Long-run Volatility and Risk around Mergers and Ac

Long-runVolatilityandRiskaroundMergersandAcquisitions¤SreedharT.BharathUniversityofMichiganGuojunWuUniversityofMichiganThisversion:January31,2005y¤SreedharT.BharathisatRossSchoolofBusiness,UniversityofMichigan,DepartmentofFinance,AnnArbor,MI48109,U.S.A.Phone:[phone].e-mail:[e-mail]Business,UniversityofMichigan,DepartmentofFinance,AnnArbor,MI48109,U.S.A.Phone:[phone].e-mail:[e-mail]sions,andseminarparticipantsatUniversityofMichiganforcomments.1Long-runVolatilityandRiskaroundMergersandAcqusitionsAbstractInthispaperwestudythechangesinvolatilityandriskofacquirersaroundmergersandacquisitionsandseektounderstandthedeterminantsofthosechanges.Wefindthatthereisastrongrun-upinvolatilityandriskbeginningfouryearsbeforethemerger.Thispre-mergerrun-upisconsistentwiththehypothesisthatM&Asarearesponsetoindustryshocks.Wefindthatforaperiodofaboutoneyearafterthemergerthecross-sectionalaverageofthevolatilitymeasurescontinuetoincrease.Beyondthatthesystematicvolatilityandbetabegintodecline.However,idiosyncraticvolatilitycontinuestoincreaseforthenexttwoyears.Thevolatilitypatternsuncoveredisalsoconsistentwiththeriskofpost-mergerintegrationoftheacquirerandthetargetfirmsthatgetsresolvedslowlyovertime.Ourfindingsmayhaveimportantimplicationsforunderstandingseveralissues,includingtheannouncementeffectofmergers,thediversificationdiscount,andthelong-rununder-performanceofacquirersinM&Atransactions.Thekeyinsightisthatasweunderstandthevolatilityandriskdynamicsbetter,wewillbeabletocomputerisk-adjustedreturnsmoreaccurately,potentiallychangingtheconclusionsofpreviousstudiesontheseissues.Keywords:MergersandAcquisitions,Volatility,Risk,Diversification,Long-rununderperformanceJ.E.L.ClassificationCode:G12G34.1IntroductionCorporateMergersandAcquisitions(M&A)activityreachedunprecedentedlevelsinthelate1990s.Intheyear2000,thedollarvolumeofworldwideM&Atransactionsreachedapproximately$3.2trillionthroughover3,000transactions.OftheseapproximatelyhalfinvolvedU.S.partiesandU.SmergeractivityasapercentageofU.S.GNPwasashighas18%(Bruner2004).Inthispaperwestudythechangesinlong-runvolatilityandriskofacquirersaroundamergerandseektounderstandthedeterminantsofthosechanges.AlargebodyofempiricalliteraturehasstudiedthevaluationeffectsofM&Aannouncementsontargetsandacquirersusingeventstudymethods[SeeMoeller,SchlingemannandStulz(2003)forrecentevidence].Thegeneralconsensusisthattargetfirmshareholdersenjoysignificantlyandmateriallypositiveabnormalreturnswhileacquiringfirms’shareholdersexperiencesignificantlynegativeabnormalreturnsaroundtheannouncementdates.Theevidenceonlong-runstockreturnsfollowingM&A(overa5-yearperiod)pro-videdbyLoughranandVijh(1997)suggeststhatacquirershavesignificantlynegativereturns.WhileresearchershavelearnedalotaboutstockreturnsaroundanM&Atransactiontherehasbeenverylittlestudyonlong-runchangesinvolatilityandriskoftheacquiringfirmaroundamergerevent.1Studyingvolatilityandriskchangesmaybeimportantinunderstandingtheannouncementeffectandthelong-runreturnsfollowingmergers.Manyoftheexistingexplanationsofnegativeannouncementreturnsforacquirersfocusonrelativefirmvaluesorequityreturns.2Totheextentthatacquisitionsmightincreasethevolatilityofcashflowsofthecombinedfirm,thenegativeannouncementeffectsdocumentedintheliteraturemightalsobeconsistentwiththeviewthatmarketsrecognizetheincreaseinriskofthepostmergerfirm.Iftheincreaseinvolatilityriskispriced,investorswoulddemandhigherexpectedreturnsfromtheacquirer.Thiscanresultinanimmediatedropinthestockprice.Similarly,changesinlong-runreturnsmightalsoberelatedtolong-runchangesinvolatilityandriskoftheacquiringfirm.Inanycase,itisworthwhiletoinvestigatethechanges1Jayaraman,MandelkerandShastri(1991)isanotableexceptionwhichstudiestheshort-runchangesinvolatility(-160+2days)oftargetsaroundamergerusingasmallsampleof27firms.2Forexample,thehubrishypothesisofRoll(1986)suggeststhatmanagersofacquiringfirmssufferfromhubris,sotheyoverpay.Jensen’s(1986)freecashflowhypothesisarguesthatempirebuildingmanagementwouldrathermakeacquisitionsthanincreasepayoutstoshareholders.ShleiferandVishny(2003)andRhodes-KropfandViswanathan(2003)suggestthatbuyingfirmstendtoacquirewithstockwhentheybelievethecompany’ssharesareovervalued.Theyfindthatmergeractivity,theuseofstocksasformofpayment,andwhobuyswhomaredrivenbytherelativevaluationsofthepairsoffirms.Thus,theoftenobservedpostmergerdeclineinstockpriceisarguedtobenotareflectionofthesuccessorfailureofthemerger,butratheracorrectioninthemarket’valuationoftheacquirer.1involatilityandriskoftheacquirersaroundmergersandtounderstandthedeterminantsofthesechanges.WebeginbystudyingthechangesinvolatilityandriskoftheequityofallU.S.acquiringfirmsthatundertookamergertransactionbetweenthefourthquarterof1995andthethirdquarterof2002.Oneimportantinnovationinourstudyisthatratherthanusingvolatilitycomputedfromrealizedreturns,weusetheimpliedvolatilityof30-dayat-the-moneycalloptionsoftheacquirersineventtime(measuredinquartersrelativetothemergerevent).3Thetotalvolat