A Review of IPO Activity,Pricing,and Allocations

AReviewofIPOActivity,Pricing,andAllocationsJAYR.RITTERandIVOWELCH*ABSTRACTWereviewthetheoryandevidenceonIPOactivity:whyfirmsgopublic,whytheyrewardfirst-dayinvestorswithconsiderableunderpricing,andhowIPOsperforminthelongrun.Ourperspectiveisthreefold:First,webelievethatmanyIPOphenomenaarenotstationary.Second,webelieveresearchintoshareallocationissuesisthemostpromisingareaofresearchinIPOsatthemoment.Third,wearguethatasymmetricinformationisnottheprimarydriverofmanyIPOphe-nomena.Instead,webelievefutureprogressintheliteraturewillcomefromnon-rationalandagencyconflictexplanations.Wedescribesomepromisingsuchalternatives.From1980to2001,thenumberofcompaniesgoingpublicintheUnitedStatesexceededoneperbusinessday.Thenumberofinitialpublicofferings~IPOs!hasvariedfromyeartoyear,however,withsomeyearsseeingfewerthan100IPOs,andothersseeingmorethan400.TheseIPOsraised$488billion~in2001dollars!ingrossproceeds,anaverageof$78millionperdeal.Attheendofthefirstdayoftrading,theirsharestradedonaverageat18.8percentabovethepriceatwhichthecompanysoldthem.Foraninvestorbuyingsharesatthefirst-dayclosingpriceandholdingthemforthreeyears,IPOsreturned22.6percent.Still,overthreeyears,theaverageIPOunder-performedtheCRSPvalue-weightedmarketindexby23.4percentandun-derperformedseasonedcompanieswiththesamemarketcapitalizationandbook-to-marketratioby5.1percent.Inanutshell,thesenumberssummarizethepatternsinissuingactivity,underpricing,andlong-rununderperformance,whichhavebeenthefocusofalargetheoreticalandempiricalliterature.Wesurveythisliterature,focus-ingonrecentpapers.SpaceconstraintsforceustotakeaU.S.-centricpointofviewandtoomitadescriptionoftheinstitutionalaspectsofgoingpublic.TheinterestedreadercanconsultEllis,Michaely,andO’Hara~2000!,*RitterisfromtheWarringtonCollegeofBusinessAdministrationattheUniversityofFlor-ida;WelchisfromtheYaleSchoolofManagementandtheNBER.Thissurveywaspresentedatthe2002AtlantaAFAmeetings.WethankChrisJames,RoniMichaely,AnnSherman,Dong-hangZhang,andespeciallyTimLoughranandMaureenO’Haraforcomments,andKennethFrenchforsupplyingfactorreturns.TheauthorsmaintainamoreextensivebibliographyofIPO-relatedworkathttp:00VOL.LVII,NO.4AUGUST20021795Foerster~2000!,JenkinsonandLjungqvist~2001!,andKillian,Smith,andSmith~2001!fordescriptionsoftheinstitutionalprocess.Ritter~2002!andespeciallyJenkinsonandLjungqvistgiveextensivecoveragetointer-nationalpatternsandpractices.Averageshidethetimetrendsandyear-by-yearvariationinthesephe-nomena,asshowninTableI.The1980ssawmodestIPOactivity~about$8billioninissuingactivityperyear!.Inthe1990s,issuingvolumeroughlydoubledto$20billionperyearduring1990to1994,doubledagainfrom1995to1998~$35billionperyear!,andthendoubledagainfrom1999to2000~$65billionperyear!,beforefallingto$34billionin2001.Averagefirst-dayreturnsshowasimilarpattern,increasingfrom7.4percentinthe1980sto11.2percentintheearly1990s,to18.1percentinthemid-1990s,andto65.0percentin1999and2000,beforefallingbackto14.0percentin2001.Thelong-runperformanceofIPOsalsovariesovertime.Three-yearmarket-adjustedbuy-and-holdreturnsarenegativeineverysubperiod,butnotforeverycohortyear.Style-adjustedbuy-and-holdreturnsarenotasreliablynegative,withmanycohorts,andsomesubperiods,havingpositivestyle-adjustedbuy-and-holdreturns.Ourarticleseekstoreviewdifferentexplanationsforthesepatternsinissuingactivity,underpricing,andlong-rununderperformance.Butitalsoweighsinwithourpersonalperspectiveonissuesthatarestillcontentious.Webelievethatthetime-variationinthesephenomenadeservesmoreem-phasis.Forexample,thelong-runperformanceofIPOsisnotonlysensitivetothewidelydebatedchoiceofeconometricmethodology,butalsotothechoiceofsampleperiod,asshowninTableI.Further,wearguethatasym-metricinformationtheoriesareunlikelytobetheprimarydeterminantoffluctuationsinIPOactivityandunderpricing,especiallytheexcessesoftheInternetbubbleperiod.Instead,webelievethatspecificnonrationalexpla-nationsandagencyexplanationswillplayabiggerroleinthefuturere-searchagenda.Indiscussingtheoriesofunderpricing,wedevotesignificantattentiontothetopicofshareallocationsandsubsequentownership.Inourview,howIPOsharesareallocatedisoneofthemostinterestingissuesinIPOresearchtoday.Theremainderofthisarticleisorganizedasfollows.InSectionI,wesurveyIPOactivity.SectionIIcoversIPOpricingandallocation.SectionIIIpresentsevidenceandanalysisonthelong-rununderperformanceofIPOs,andSectionIVconcludes.I.IPOActivity:ChoosingtoGoPublicThefirstquestionmustbe“whydofirmsgopublic?”Inmostcases,theprimaryansweristhedesiretoraiseequitycapitalforthefirmandtocreateapublicmarketinwhichthefoundersandothershareholderscanconvertsomeoftheirwealthintocashatafuturedate.Nonfinancialrea-sons,suchasincreasedpublicity,playonlyaminorroleformostfirms:1796TheJournalofFinanceTableINumberofIPOs,First-dayReturns,GrossProceeds,AmountofMoneyLeftontheTable,andLong-runPerformance,byCohortYear,1980to2001Theequallyweighted~E