Disappearing dividends changing firm characteristi


WeacknowledgethecommentsofJohnGraham,DouglasHannah,AnilKashyap,TobiasMoskowitz,G.WilliamSchwert(theeditor),AndreiShleifer,JaniceWillett,PaulZarowin,andseminarparticipantsatHarvardUniversity,theUniversityofChicago,theNationalBureauofEconomicResearch,theUniversityofRochester,andVirginiaPolytechnicalInstitute.Thereferees,HarryDeAngeloandReneHStulz,wereparticularlyhelpful.*Correspondingauthor.E-mailaddress:[e-mail](K.R.French).JournalofFinancialEconomics60(2001)3}43Disappearingdividends:changingrmcharacteristicsorlowerpropensitytopay?
EugeneF.Fama
,KennethR.French
*
GraduateSchoolofBusiness,UniversityofChicago,Chicago,IL60637,USASloanSchoolofManagement,MassachusettsInstituteofTechnology,Cambridge,MA02142,USAReceived7January2000;accepted17August2000AbstractTheproportionofrmspayingcashdividendsfallsfrom66.5%in1978to20.8%in1999,dueinparttothechangingcharacteristicsofpubliclytradedrms.Fedbynewlistings,thepopulationofpubliclytradedrmstiltsincreasinglytowardsmallrmswithlowprotabilityandstronggrowthopportunities}characteristicstypicalofrmsthathaveneverpaiddividends.Moreinteresting,wealsoshowthatregardlessoftheircharacteristics,rmshavebecomelesslikelytopaydividends.Thislowerpropensitytopayisatleastasimportantaschangingcharacteristicsinthedecliningincidenceofdividend-payingrms.
2001PublishedbyElsevierScienceS.A.JELclassixcation:G35;G32Keywords:Dividends;Payoutpolicy;Stockrepurchases;Newlistings0304-405X/01/$-seefrontmatter
2001PublishedbyElsevierScienceS.A.PII:S0304-405X(01)00038-11.IntroductionDividendshavelongbeenanenigma.Sincetheyaretaxedatahigherratethancapitalgains,thecommonpresumptionisthatdividendsarelessvaluablethancapitalgains.Inthisview,rmsthatpaydividendsareatacompetitivedisadvantagesincetheyhaveahighercostofequitythanrmsthatdonotpay.Thefactthatmanyrmspaydividendsisthendi$culttoexplain.UsingCRSPandCompustat,westudytheincidenceofdividendpayersduringthe1926}99period,withspecialinterestintheperiodafter1972,whenthedatacoverNYSE,AMEX,andNASDAQrms.Thepercentofrmspayingdividendsdeclinessharplyafter1978.In1973,52.8%ofpubliclytradednon-nancialnon-utilityrmspaydividends.Theproportionofpayersrisestoapeakof66.5%in1978.Itthenfallsratherrelentlessly.In1999,only20.8%ofrmspaydividends.Thedeclineafter1978inthepercentofrmspayingdividendsraisesthreequestions.(i)Whatarethecharacteristicsofdividendpayers?(ii)Isthedeclineinthepercentofpayersduetoadeclineintheprevalenceofthesecharacteristicsamongpubliclytradedrms,or(iii)havermswiththecharacteristicstypicalofdividendpayersbecomelesslikelytopay?Weaddressthesequestions.Weuselogitregressionsandsummarystatisticstoexaminethecharacteristicsofdividendpayers.Bothapproachessuggestthatthreecharacteristicsa!ectthedecisiontopaydividends:protability,investmentopportunities,andsize.Largerrmsandmoreprotablermsaremorelikelytopaydividends.Dividendsarelesslikelyforrmswithmoreinvestments.Thesummarystatisticsprovidedetailsonthenatureofdividendpayers,formerpayers,andrmsthathaveneverpaid.Formerpayerstendtobedistressed.Theyhavelowearningsandfewinvestments.Firmsthathaveneverpaiddividendsaremoreprotablethanformerpayersandtheyhavestronggrowthopportunities.Dividendpayersare,inturn,moreprotablethanrmsthathaveneverpaid.Butrmsthathaveneverpaidinvestatahigherrate,domoreR&D,andhaveahigherratioofthemarketvalueofassetstotheirbookvalue(
/A
,aproxyforTobin'sQ)thandividendpayers.Theinvestmentsofdividendpayersareontheorderofpre-interestearnings,buttheinvestmentsofrmsthathaveneverpaidexceedearnings.Finally,payersareabout10timesaslargeasnon-payers.Thedeclineafter1978inthepercentofrmspayingdividendsisdueinparttoanincreasingtiltofpubliclytradedrmstowardthecharacteristicsofrmsthathaveneverpaid}lowearnings,stronginvestments,andsmallsize.Thistiltinthepopulationofrmsisdrivenbyanexplosionofnewlylistedrms,andbythechangingnatureofthenewrms.Thenumberofpubliclytradednon-nancialnon-utilityrmsgrowsfrom3,638in1978to5,670in1997,beforedecliningto5,113in1999.Newlylistedrmsalwaystendtobesmall,withextraordinaryinvestmentopportunities(highassetgrowthratesandhigh
/A
).What4E.F.Fama,K.R.French/JournalofFinancialEconomics60(2001)3}43changesafter1978istheirprotability.Before1978,newlistsaremoreprot-ablethanseasonedrms.In1973}77,theearningsofnewlistsaverageahefty17.79%ofbookequity,versus13.68%forallrms.Theprotabilityofnewlistsfallsthroughoutthenext20years.Theearningsofnewlistsin1993}98average2.07%ofbookequity,versus11.26%forallrms.Thedeclineintheprotabilityofnewlistsisaccompaniedbyadeclineinthepercentofnewliststhatpaydividends.During1973}77,one-thirdofnewlylistedrmspaydividends.In1999,only3.7%ofnewlistspaydividends.Thesurgeinnumbersandthechangingnatureofnewlistsproduceaswellinggroupofsmallrmswithlowprotabilitybutlargeinvestmentsthathaveneverpaiddividends.Thisgroupofrmsisabigfactorinthedeclineinthepercentofrmspayingdividends.Itisperhapsobviousthatinvestorshavebecomemorewillingtoholdthesharesofsmall,relativelyunprotablegrowthcompanies.Buttheresultingtiltofthepubliclytradedpopulationtowardsuchrmsisonlyhalfofthestoryforthedeclininginciden