capitalization

GarenMarkarian,LorenzoPozza,AnnalisaPrencipe,InternationalJou

rnalofAccounting2008-3爱思唯尔期刊

CapitalizationofR&DCostsandEarningsManagement：Eviden

cefromItalianListedCompanies

ABSTRACT:Thecapitalizationofrearchanddevelopment

(R&D)costsisacontroversialaccountingissuebecauofth

econtentionthatsuchcapitalizationismotivatedbyincenti

nasampleofItalianlist

edcompanies,thisstudyexamineswhethercompanies'decision

stocapitalizeR&Dcostsareaffectedbyearnings-management

rovidesanaturalcontextfortestingou

rhypothesizedrelationshipsbecauItalianGAAPallowsfor

thecapitalizationofR&Tobitregressionmod

eltotestourhypothes,weshowthatcompaniestendtou

othesisthatfirmscapitalizeR&Dcoststoreducetheriskof

DS:Earnin

gsmanagement,Costcapitalization,R&Daccounting,Earnings

smoothing,Debtcovenants,Italiancompanies

1IntroductionInthecurrenteraofglobalization,ahigh

lyrelevantissuefacingregulators,academics,andpractitio

nersisthedeterminationofanappropriateaccountingtreatm

entforrearchanddevelopment(R&D)ationalA

ccountingStandardsdiscussaccountingforR&DcostsinIASN

o.38“IntangibleAsts”(IASB,2004;IASB,2004).Paragrap

h54ofthisstandardstatesthatnointangibleastarising

fromrearch(orfromtherearchphaofaninternalpro

ject)shallberecognizedasanast;andthatrearchexpe

nsshallbeexpendintheincomestatementwhentheyare

ningdevelopmentcosts,paragraph57statest

hatanintangibleastarisingfromdevelopment(orfromthe

developmentphaofaninternalproject)shallberecognize

dif,andonlyif,anentitycandemonstrateallofthefollo

wing:(a)thetechnicalfeasibilityofcompletingtheintangi

bleastsothatitwillbeavailableforuorsale;(b)i

tsintentiontocompletetheintangibleastanduorll

it;(c)itsabilitytouorlltheintangibleast;(d)

howtheintangibleastwillgenerateprobablefutureecono

micbenefits;(e)theavailabilityofadequatetechnical,fin

ancial,andotherresourcestocompletethedevelopmentandt

ouorlltheintangibleast;(f)itsabilitytomeasur

ereliablytheexpenditureattributabletotheintangibleass

ghIAS38allowscompaniest

ocapitalizedevelopmentcosts,theinherentsubjectivityof

thevalidationprocesspermitsmanagementtoexercidiscret

ionindecidingwhethertheconditionsofIAS38havebeensa

nce,IAS38givesmanagementconsiderablefle

xibilityregardingthetreatmentofdevelopmentcosts.

.2

—AccountingforRearchandDevelopmentCosts(FASB,1974)

—requiresthatallR&Dexpendituresbeexpendinthecurre

yexceptiontothefullexpensingruleiss

eptionrelatestothecapitaliza

tionofsoftwaredevelopmentcosts(FASB,1985).Attheinter

nationallevel,certainnationalaccountingstandards(e.g.,

thoofItaly)allowflexibilityforthecapitalizationofR

&recondit

ionssimilartothorequiredbyIAS.

ThecapitalizationofR&Dcostshasalwaysbeenacontrove

tersofcapitalizationreportr

esultssuggestingthatR&Disalong-livedastthatinfluen

cesfutureprofitability(e.g.,BublitzandEttredge,1989,J

anuary;Sougiannis,1994,January;Ballesteretal.,2003).A

lso,R&Dcostsarepositivelyrelatedtomarketvalue(Hirsch

eyandWeygandt,1985,Spring;Shevlin,1991,January;Sougia

nnis,1994,January)andyieldvalue-relevantinformationto

investors(e.g.,AboodyandLev,1998;LevandZarowin,1999;

Healyetal.,2002;Monahan,2005).Supportersofexpensing

ressthelackofreliableevidenceoffutur

eeconomicbenefits(e.g.,FASB,1974;AssociationforInvest

mentManagementandRearch,1993;Kotharietal.,2002)or

refertothebenefitsofconsistencyandcomparability,point

ingoutthatsuchbenefitstrumpthecostsidentifiedbythe

onally,reliabilityandt

heriskofearnings-managementpoliciesareunderscoredbysu

rticular,expensingispreferabletocapitalizationbecaui

,

iteliminatestheopportunityformanagerstocapitalizecos

tsofprojectsthathavelowprobabilitiesofsuccessortod

elayimpairmentofR&Dasts(Nelsonetal.,2003;Schilit,

2002).Thedebatesurroundingthemosteffectiveaccounting

methodforR&Dcostssupplementsotherliteraturethatexamin

esthetrade-offbetweenrelevance(i.e.,thepredictiveabil

ity)andreliability(i.e.,thereprentativefaithfulness)

ofaccountinginformation(FASB,1980;AICPA,1994;IASB,200

4;IASB,2004).Thusfar,empiricalrearchonR&Dcostshas

focudmainlyontherelevancesideofthetrade-off,while

littlehasbeenwrittenaboutthereliabilitysidethatis,

thepossibilitythatR&Dcostsaresubjecttoearningsmanage

ment.

However,afewstudieshaveindeedshownthatR&Dexpendit

t,this

meansthatcompaniescuttheirR&Dinvestmentsinordertoac

hievetheirearningsgoals(e.g.,PerryandGrinaker,1994;B

ushee,1998;Mandeetal.,2000).Butthereisstillapaucit

yofrearchthatexploresthemotivesbehindtheaccounting

treatmentofR&Dcostswithinattingwhereflexibilityis

gwhethercompaniengageinearningsmanage

mentthroughR&Dcostaccountingcansignificantlycontribute

debatehasrecentlybeenraidwithintheconvergenceproje

ctbyUSGAAPandIAS/ratingthatR&Dcostcapita

lizationismotivatedbyincentivestomanipulateearningswo

sition,whichdoesnotal

ontrary,showin

gthatcompaniesdonotuR&Dcostaccountingforearnings-

managementpurposwouldsupporttheapproachnowstatedby

IAS/IFRS,inwhichcapitalizationisallowedundercertainco

udycontributestothisdebatebyproviding

empiricalevidenceonthemotivationsforR&Dcostcapitaliza

thesizethatthedecisiontocapitalizeR&Dexpe

ndituresisrelatedtotwoprimarymotivations:incomesmooth

ourhypothesusingasamp

ariate

resultsindicatethatfirmsucapitalizationofR&Dcostst

osmoothearnings,whilethereisnosupportforthedebt-cov

esultsarerobustwithinavarietyo

ffirmcharacteristics,suchasfirmsize,risk,opportunitie

sforgrowth,profitability,governancecharacteristics,indu

strialmembership,erproceedsasfo

n2introducesaccountinginItalyandtheinsti

tutionalbackgroundrelatingtoR&n3dis

n4prentsthehypoth

esand

n

6prentstheresultsandSection7concludesthestudy.2R

&DaccountinginItalyItalianaccountingregulationhasalwa

ysallowedforsomeflexibilityinthecapitalizationofR&D

tingf

orintangibles,includingR&Dcosts,isregulatedbyPrincipi

oContabilen.24(AccountingStandardNo.24).Thisstandard

distinguishesthreedifferenttypesofR&Dcostsasfollows:

1)“Basicrearch,”whichconsistsofstudies,surveys,and

experimentsthatdonotrefertoaspecificproject;thisty

peofR&Dcostisnormallycarriedoutforthegeneralutilit

yofacompany(e.g.,marketrearch,updating,etc.);

2)“Appliedrearch,”whichconsistsofstudies,surveys,

andexperimentsthatrefertospecificprojects;

3)“Development,”whichconsistsoftheapplicationofre

archresultstospecificmaterials,tools,products,andpr

ocessprecedingproduction.

Thecostsforbasicrearcharetobeexpendintheinc

r,costsrelatedtoappliedR&Dcanbec

apitalizedifthefollowingconditionsaremet:a)thecosts

refertoaprojectfortherealizationofaclearlydefinedp

roductorprocess;b)thecostsareidentifiableandmeasurab

le;c)theprojecttowhichthecostsreferistechnicallyfe

asible;d)thecompanyownsthenecessaryresourcestocomple

teandexploittheproject;ande)thecostsarerecoverable

ividentthattheconditionsstatedbytheItalianaccounti

ngstandardsaresimilartothostatedbyIASfordevelopme

,thedefinitionofappliedrearchunderI

talianstandardsalsofitsintothedefinitionofdevelopment

lianstandardsdifferfrom

IASinthattheydonotrequireR&Dcapitalizationwhenthea

bovementionedconditionsoccur,leavingflexibilityintheha

r,thisdifferenceismoreformal

hesubjectivityinasssingtheoc

currenceofsomeoftheconditions,itemsthat,evenunder

IAS,companiesthatpreferimmediateexpensingcaneasilyju

stifythisapproach—evenwhentheaforementionedconditions

ningtheamortizationofR&Dcosts,theItalia

naccountingstandardsrequirethattheamortizationbecarri

edoutoveraperiodofnolongerthanfiveyearsbeginningf

romthemomenttheoutcome(productorprocess)isreadytob

lianCivilCode(art.2426)statesthattheca

pitalizationofR&Dcostsshallbeauthorizedbythecollegio

sindacale(statutoryauditors)andthatitisnotpossiblet

opaydividendsuntilthereareenoughretainedearningstoc

overthecarryingamountofthecapitalizedR&

tipulationlimitstheincentivetocapitalizeR&Dcostsfort

vilCodealsorequiresthatR&Dactivitiesbediscusdinth

erelazionesullagestione(managementdiscussionandanalysi

sction);however,thereisnoclearrequirementastowhat

quantitativeorqualitativedisclosuresshouldberelayedwi

thregardtothecapitalizationofR&y,theCi

vilCodestatesthatinformationregardingtheamortizations

chedulesofsuchR&Dcostsbeprovidedintheexplanatorynot

esofthefinancialstatements.3Earningsmanagementandspe

cificaccruals

Earningsmanagementisdefinedasa“purpofulinterventi

onintheexternalfinancial-reportingprocess,

withtheintentofobtainingsomeprivategain”(Schipper,

1989,p.92).Ingenerallyacceptedterms,earningsmanageme

ntoccurs“whenmanagersujudgmentinfinancial-reporting

andinstructuringtransactionstoalterfinancialreportsto

eithermisleadsomestakeholdersabouttheunderlyingeconom

icperformanceofthecompanyortoinfluencecontractualout

comesthatdependonreportedaccountingnumbers”(Healy&Wa

hlen,1999,p.368).Thelargeamountofrearchcarriedout

thusfarindicatesthatmanagerxercidiscretionandman

ageearningsusingawidevarietyofmethods,rangingfromca

rryingoutspecialtransactions(so-calledrealearningsmana

gement)lofthemain

incentivesforearningsmanagementincludedebtcovenants,bo

nusplans,t-covenanthypothesis

suggeststhatmanagershaveanincentivetomanageearnings

inordertoavoidviolatingcovenantsindebtcontracts,whic

haretypicallystatedintermsofaccountingnumbersorrati

us-planhypothesissuggeststhatmanagersmanagee

(1985)shows

thatmanagerstendtoreduceearningsiftheyfalleitherab

rast,theytendtoinc

y,

theincome-smoothinghypothesissuggeststhatfirmsaspiret

oreduceearningsfluctuations.

Empiricalearnings-managementstudiesfindsupportforthe

the

studiestesttherelationshipbetweenaggregateaccrualsa

ndincentivesforearningsmanagement(e.g.,Healy,1985;DeA

ngelo,1986;Dechowetal.,1995).Asanalternativeapproach,

otherstudiesfocusonsingleitems,suggestingthatincome

fromspecificaccrualsisrelatedinasystematicwaytoearn

helatterstudies,McNic

holsandWilson(1988)showthatcompaniesmanagetheirbad-d

ebtprovisionsaccordingtothebonus-planhypothesis(Healy,

1985).ZuccaandCampbell(1992)examinediscretionaryast

write-downs,showingthatcompaniesutheaccrualithe

rfor“bigbath”

is,Hanna,andVincent(1996)confirmthatearnings-managemen

tincentivesplayasignificantroleinexplaininggoodwillw

tudiesfocusona

llowancesfordeferredtaxes(e.g.,MillerandSkinner,1998;

SchrandandWong,2003).Thestudiesprovidemixedresults.

Finally,DowdellandPress(2004)analyzethein-processR&D

write-offs,buttheydonotfindevidencetosupporttheirb

withtheaforementionedstudies

onearningsmanagementandspecificaccruals,thisstudyaim

sattestingwhetherthedecisiontocapitalizeortoexpen

R&Dcosts(whenflexibilityexists)isaffectedbyearnings-

managementmotives.4HypothesdevelopmentPreviousrearc

hinvestigatesthreemainincentivesforearningsmanagement:

earningssmoothing,debtcovenant,andbonus-planincentives.

Inthisstudy,wefocusonthefirsttwosincedisclosureof

dataontheexistenceandstructureofbonusplansbyItalia

ome-smoothinghypothesissugge

ststhatamanager'saccountingdiscretionisdrivenbyhiso

rherdesiretoreduceincome-streamvariability(Fudenberga

ndTirole,1995).Theprocessofsmoothingrvestomoderate

year-to-yearfluctuationsinincomebyshiftingearningsfro

ocesslowers

thepeaksandmakearningsfluctuationslessvolatile(Cope

land,1968).

Incomesmoothinghasbeenviewedbothasapositivestrate

gy,wherebymanagerstransmitprivate

informationtoinvestors(e.g.,Gordon,1964,April;Beidl

eman,1973;RonenandSadan,1981;TuckerandZarowin,2006),

andasamanipulativepracticedrivenbyopportunisticaims

(Gordon,1964,April;Imhoff,1977,Spring;KaminandRonen,

1978).Inthisstudy,wedonotintendtoargueforeithero

,wetestwhetherR&Dcostcapit

alizationisudforpurposofearningssmoothing.