Beruflich Dokumente
Kultur Dokumente
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
degree of association. In other words, they have value relevance. Finally, this study suggests that basic trading strategies
will improve if investors revert to the accounting fundamentals.
Keywords:accountingfundamentals;bookvalue;earningsyield;growthopportuni
ties; shortrun and longrun investment scalabilities; trading strategy;
valuerelevance
JEL Classification: M41(accounting);G12(assets pricing; interest rate);G14(information and
market efficiency);G15(international financial markets)
Introduction
Chen and Zhang (2007) present
thelatestreturnmodelthatrelatesthe
fundamentalfirmvaluetothevariation
instockprice.Theyalsoprovidetheo
reticalandempiricalevidencethatstock
returnisafunctionofaccountingvari
ables, namely earnings yield, equity
capital, the change in profitability,
growthopportunities,anddiscountrate.
ChenandZhang(2007)arguethatfirm
valueembracesinformationonpoten
tialfutureassetsandgrowthopportuni
ties. This argument is supported by
MillerandModigliani(1961).Inasimple
explanation,bothstudiesinferthatstock
priceisafunctionoffutureassetsor
capitalscalability.1Earningscouldbe
determinedbytheadaptationconcept
whenthefirmsinvestedresourcesare
modifiabletogeneratefutureearnings
(Wright1967).
The association between stock
returnandfundamentalfirmvaluehas
Scalabilityisactuallyafirmsscaleofoperations.Thisstudyshortensitintoscalability.Itrefers
tothemeasureofincreasingordecreasingscaleofoperationsinaratioorproportion.Inthisstudy,
the ratios denominator is the previous years assets.
190
ingsnumbersasanadditionalpredic
tivevalue,whichiscalledthevaluation
link(Ou1990).
This complementaryanalysis re
liesonthefollowingreasons.First,the
limitation of Ohlsons (1995) model
(FelthamandOhlson1995,1996).This
weaknessliesinitsassumptionsthat:
(i)futureearningscouldbedetermined
using consecutive previous earnings
and (ii) earnings could be predeter
mined stochastically. Second, earn
ings is a noise when measuring eco
nomicearningsandequityvalue(Kolev
etal.2008;Collinsetal.1997;Givoly
and Hayn 2000; and Bradshaw and
Sloan2002).Third,highvalueisrel
evant when eliminating earnings
(Bradshaw and Sloan 2002; and
Bhattacharyaetal.2003).Therefore,
this study provides complementary
measurement of earnings. Addition
ally,thisstudyisfocusedontheadap
tation theory in which assets on the
statement of financial position are a
determinant of equity value
(BurgstahlerandDichev1977).
Ourmainresearchobjectiveisto
design a new return model. It also
examinesthedegreeofassociationin
this model. Not only does this new
return model associate stock return
with four cashflowrelated factors,
namely earnings (Easton and Harris
1991; Burgstahler and Dichev 1997;
Collinsetal.1999),equitycapital(Jan
andOu1995,andCollinsetal.1999),
profitabilityandgrowthopportunities
(Ohlson 1995; Feltham and Ohlson
1995,1996;Zhang2003,andChenand
Zhang2007),anddiscountrate(Zhang
191
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
2003,andChenandZhang2007),butit
alsoinvestigatesfurtherbyfactoringin
theshortrunandlongruninvestment
scalabilities.Thisstudyexaminesthe
new theoretical return model using
empirical data. Furthermore, robust
nesschecksareconductedtoconfirm
theconsistencybetweenthenewmodel
anditspredecessors,includingtheas
sociationbetweeneachconstructand
stockreturn.
Thisstudybenefitsbothinvestors
and managers. From the investors
pointofview,thisstudyprovidesmore
comprehensive,realistic,andaccurate
parametersforpredictingpotentialfu
ture cash flows since the new model
extractsmoreinformationthandocur
rently available models. From the
managers point of view, this study
gives incentives to managers to dis
close more information publicly as
mandatedbySFACNo.5,paragraph
24(FASB1984).Finally,thenewre
turnmodelcanleadinvestorsandman
agement to assess comprehensively
theinformationconveyedinfinancial
statements.
Thisstudycontributestoaccount
ingliteraturebyprovidingmorecom
pleteandrealisticreturnmodel. This
study has advantages compared with
themodelsofEastonandHarris(1991),
LiuandThomas(2000),Zhang(2003),
Copelandetal.(2004),ChenandZhang
(2007), and Weiss et al. (2008), ex
plainedasfollows.First,thismodelis
morecomprehensiveduetoitsbroader
coverage,specificallytheinclusionof
assets scalability to generate future
cash flows.
192
Second,byincludingscalability,
thismodelisexpectedtobecloserto
the economic reality as firms should
reasonably choose future investment
projectsthatwillcontributepositivenet
cash inflow. Cash inflow magnifies
earningsanditsvariability.Thesecond
advantage is labeled as the earnings
capitalizationmodelbyOhlson(1995),
whoexplainsthatearningsanditsvari
abilityareaffectedbycurrentprojects.
Third,thenewreturnmodelcre
atesamorecomprehensiveandaccu
ratepredictoroffuturecashflowsto
estimate potential future earnings by
extracting multiple relevant informa
tion(Liuetal.2001).Multipleinforma
tioncouldimprovemodelaccuracyas
long as it is aligned with increasing
valuerelevance.Eventually,thisstudy
offers considerable contribution by
improvingthedegreeofassociationof
returnmodelasitismorecomprehen
sive,realistic,andaccurate.Thiscon
tributionisreflectedbyhigherR2and
adj-R2thanthepreviousmodels.
This study assumes that, firstly,
the association between accounting
fundamentalsandstockpricevariabil
ityislinear.Accountinginformationis
positivelyproportionaltoearningsyield,
investedequitycapital,profitability,and
growthopportunities,andisnegatively
proportionaltodiscountrate.Secondly,
investors payattention to accounting
information comprehensively, mean
ingthatinvestorsuseaccountingfun
damentalsforbusinessdecisionmak
ing. Thirdly, investors comprehend a
firmsprospectbasednotonlyoneq
uitycapitalanditsgrowth,butalsoon
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Investment Scalability
The first limitation of Ohlsons
(1995)modelliesinits assumptions.
Continued by Feltham and Ohlson
(1995;1996),itstillassumesthatfuture
earningsisdeterminedbyconsecutive
previousearnings.However,investors
mayhavedifferentinsightsbyobserv
ing future potential earnings.
BurgstahlerandDichev(1977)clearly
revealthatequityvalueisnotaffected
bypreviousearningsonly,butcouldbe
determinedbytheadaptationtheory,2
which is the firms invested capital
whenitsresourcesaremodifiablefor
otherutilizations.Furthermore,theother
2
utilizationsmaygeneratefuturepoten
tialearnings.Thisconceptisbasedon
Wright (1967), who argues that the
adaptation value is derived from the
role of financial information on the
balance sheet, and the role primarily
comes from assets.
The second limitationofOhlsons
model(Ohlson1995;andFelthamand
Ohlson1995,1996)liesinitsearnings
assumption.Earningsisassumedtobe
predetermined stochastically. This
concept is based on Sterling (1968),
assumingthat firms areinstationary
condition.Theconceptbasicallypostu
lates that afirmcontinues tooperate
basedonitspaststrengthandperfor
mance.Infact,thefirmsstrengthand
performancemaychangeduetotech
nology,mergerandacquisition,take
over,liquidation,bankruptcy,restruc
turing,managementturnover,andnew
investedcapital.
Ohlson(1995;2001)himselfad
mitted to the limitations, citing that
therewasotherinformationnotedasa
mysterious variable. This variable
makesthestockmarketsfailtoreflect
bookvalue,orlessenstheinformation
content. Further research has been
attempting to replace the mysterious
variable(e.g.,Beaver1999;Hand2001),
althoughbothofthosestudiesaremerely
aninterpretativecommentaryorevalu
ativereviewoftheOhlsonsmodel.
Later research has left Ohlsons
conceptandtriedtocomplementitwith
Apart of the adaptation theory, another approach to determining firm equity value is the
recursiontheory.Usingtherecursionapproach,equityvalueisadiscountedfutureexpectedearnings
underthe assumption that thefirm merely applies currentbusiness technology into thefuture.
194
195
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Changes in Growth
Opportunities
Ohlsons(1995)modelmaintains
thecleansurplustheorywhichrelates
accountinginformationtothefollowing
premises: (1) stock market value is
basedondiscountedfuturedividendsin
whichinvestorshaveaneutralposition
againstrisk,(2)accountinginformation
issufficienttocalculatecleansurplus,
and(3) future earnings is stochastic,
predetermined by consecutive previ
ousearnings.However,investorsmay
respond differently to minimum or
maximumprofitability.Hence,growth
factors,ashavebeenincludedbyother
research, may affect earnings.
Rao and Litzenberger (1971),
LitzenbergerandRao(1972),andBao
andBao(1972)concludethatgrowth
anditschangeincreasefirmcompeti
tiveness.Consequently,thehigherthe
efficiency,thehighertheproductivity
andaccordinglythehigherthestock
holder and country wealth. Rao and
Litzenberger(1971)andLitzenberger
and Rao (1972) specifically disclose
thatgrowthopportunitiesaredirectly
associatedwithlongrunprospectwithin
oneindustry.Thosestudiesarebased
onMillerandModigliani(1961),con
cludingthatgrowingfirmisafirmthat
has apositiverateofreturnforeach
invested capital. It also means that
every invested resource has a lower
cost of capital than that within the
industry.
Liu et al. (2001), Aboody et al.
(2002), and Frankel and Lee (1998)
showaperspectivethatafirmsintrin
196
sicvalueisdeterminedbygrowthand
futurepotentialgrowth.Currentgrowth
drivestheincreaseinpotentialfuture
earnings, whereas future potential
growth reduces the models residual
errortoimprovethedegreeofmodel
association. Lev and Thiagarajan
(1993),AbarbanellandBushee(1997),
andWeissetal.(2008)suggestthatthe
growthininventories,grossprofit,sales,
accountsreceivable,etc.improvesfu
tureearningsgrowth.Moreover,their
researchconcludes that market value
adapts to all the growth factors.
DanielsonandDowdell(2001)exam
inedgrowingfirms,andfindthatthey
havebetterfinancialperformancethan
dootherfirms.Theirstudyalsoshows
thattheP/Bratioofgrowingfirmsis
greaterthanthatofothercompanies.
ChenandZhang(2007)findevi
dence that firm value completely de
pends on growth opportunities. The
growth opportunities per se are the
functionofassetsoperationscale,and
affect the potential to grow continu
ously.Theinclusionofgrowthopportu
nitiesisbasedontheperspectivethat
earningsandbookvaluearenotsuffi
cienttoexplainstockpricemovement.
Therefore,theanalysisoncurrentand
futureearningscouldbeenhancedwhen
externalenvironment,industry,andin
terestratearetakenintoaccount.
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Vt = kEt(Xt+1)+Bt.P(qt)+
Bt.gt .C(qt).....................(1)
while,gtisearningsgrowthopportuni
ties.ChenandZhang(2007)formulate
equityvalueasfollows.
Model(1)formulatesthatequity
value(Vt)isassociatedwithexpected
future earnings from invested assets
(Et(Xt+1),earningscapitalizationfactor
(k), the probability of abandonment
option(P(qt)), andthe probability of
continuationoption(C(qt)).Thismodel
indicatesthatequityvalueisequalto
thecontinuationofcurrentoperations
(qt)addedbyfirmgrowthopportuni
ties,eitherpositiveornegative(gt).
BasedonthemodelbyChenand
Zhang(2007),thisstudyexpandstheir
model by complementing and trans
forming it into a detailed form. This
transformation is supported by Ou
(1990)whoimpliesthatnonearnings
accountingvaluecanbeusedascur
rent and future earnings predictors.
Nonearningsinformationmaygivean
additionalpredictivevaluereflectedin
stockprice.Therefore,thisstudyadds
thenonearningsvaluesaspredictors.
The transformation is based on
the rationale that qt Xt/Bt-1 may be
specifiedbysrtandlrt.Shortrunin
vestment scalability is srt = (Asrt Lsrt)/(Asrt-1-Lsrt-1), where Ais assets
andLisliabilities;andlongruninvest
ment scalability is lrt = (Alrt - Llrt)/
(Alrt-1-Llrt-1). The transformation re
sultsinacompleteformulaexpressed
inModel(2)asfollows.
198
Vt = kEt(Xt+1)+Bt(P(srt)+
P(lrt))+Bt.gt(C(srt)+
C(lrt))............................(2)
Bytransformingqtintosrtandlrt,
this study develops a logical frame
workasfollows.Parameterqtasearn
ingsiscapitalinflowtothefirmfromits
operatingactivities.Thus,Model(1)is
basedonthecapitalcashflows.Itis
formulatedinthisstudythatearningsis
measuredbyassets,symbolizedassrt
andlrt.Inordertosynchronizewiththe
flow form, this study transforms the
stockformintotheflowformbymea
suring the changes,namely by (AsrtLsrt)and(Alrt-Llrt),andthennormal
izesthemonthebasisofpriorperiod
(Asrt-1-Lsrt-1) and (Alrt-1-Llrt-1). Sec
ondly,Zhang(2003)positsthatearn
ingsincreasesduetothefirmsexpan
sion. This study formulates that the
increaseinearningsisnotonlycaused
bythefirmsexpansion,butalsobythe
scalability of their productive assets.
Assetsrefertoallresourcesmanaged
to generate earnings. Therefore, the
net difference between assets and li
abilitiescouldbeusedtomeasurethe
firms earnings power. Additionally,
thetransformationofqtintosrtandlrt
is based on Rao and Litzenberger
(1971),suggestingthatthebookvalues
ofassetsandliabilitiescouldincrease
ordecreasethepotentialfutureearn
ings(Smith1973).
ThenextstepisModel(2)simpli
fication.Earningsgrowthusuallyfol
lows the random walk, meaning that
earningsgrowthdepends onprevious
yearsobservedearnings.Withqt+1=
qt + et+1,withet+1beingthemeanerror
close to zero, then E t (X t+1 )=
Et(Btqt+1)= Btqt, and with k = 1/rt.
Assets growthusedtogenerateearn
ingsfollowsthesamepatternasdoes
earningsgrowth.Transformationofqt
intosrtandlrtresultsinthefollowing
equation.
Et (Xt+1)= Et(Btqt+1)= Btqt=
Bt((srt) + (lrt))........(3)
P(lrt)+gt(C(srt)+C(lrt))
......................................(6)
To show the change in each re
latedfactor,thedifferentialequationis
......................................(4)
developed as follows. v1
AccordingtoEquation(4),anad
ditionofoneunitofassetsoroneunitof
investedcapitalintothefirmsequity
(v)couldincreasewithacertainmag
nitudecurrentequityvalue.Itsformu
lationinEquation(5)isasfollows.
(srt)+(lrt)
Vt = Btv+P(sr
)+
t
rt
v2
dv
dv
, and v3
d (lrt 1 )
drt 1 , with
dv
C ( srt 1 ) (lrt 1 ) .
dg t 1
IfthefirmpaysdividendDtduring
periodt,thenetcontributionforcurrent
return(Rt)isasfollows.
P(lrt)+gt(C(srt)+C(lrt))
......................................(5)
dv
d (srt 1 ) ,
Rt
DVt+Dt
.........(7)
199
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Bt
Bt1
Rt=v+v1srt
Vt1
Vt1
Bt1
v2lr
+(C(Srt)+
t
Vt1
Bt1
C(lrt))gt+
Vt1
Bt1
Dt
v3rt+
Vt1
Vt1
rentperiod,orreferredtoastheclean
surplus relation, then Bt = Xt Dt.
ThisequationisreversedintoDt = Xt
Bt.Ifthisequationissubstitutedinto
Equation(10),itresultsinthefollowing
equation.
Xt
Bt1
Rt=v+v1srt
Vt1
Vt1
Bt1
B t Bt1
v2lrt+ 1
+
Vt1
Vt1 Bt1
..........................................(8)
Substituting Equation (7) into
Equation(6),anequationtocalculate
stockreturnduringcurrentperiod(Rt)
isasfollows.
Bt Bt
v
Becauseof,
V t 1 B t 1
substituting it into Equation (9) will
obtainEquation(9)asfollows.
Bt
Bt1
Rt=+v1srt
Vt1
Vt1
Bt1
(C(Srt)+C(lrt))gt+
Vt1
Bt1
Dt
v3rt+
Vt1
Vt1
..........................................(9)
Assumingthatbookvaluegrowth
isequaltoearningsduringcurrentpe
riodsubtractedbydividendduringcur
200
Bt1
(C(srt))+C(lrt))gt+
Vt1
Bt1
v3rt
Vt1
..........................................(10)
Equation (10) shows that stock
return is a function of the following
factors:(1)earningsyield(Xt/Vt-1), (2)
thechangeinearningsfromshortrun
invested assets (srt), (3)the change
inearningsfromlongruninvestedas
sets (lrt), (4) the change in book
equityvalue(Bt/Bt-1),(5)thechange
ingrowthopportunities(gt),and(5)
thechangeindiscountrate(rt).
Hypotheses Development
Earnings Yield
Earningsyield(Xt)showsanaddi
tionalvaluegeneratedsincethebegin
ning of invested capital (henceforth,
dRt
1
dX t Vt 1 ,
and1/Vt-1isalwaysgreaterthanzero,
thendRt/dXtisalwayspositive.There
fore, our hypothesis is stated as fol
lows.
HA1: Earnings yield is positively related to stock return
B
B
dRt
v1 t 1 C t 1 g t .
d () srt
Vt 1
Vt 1
inreinre-
201
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
B 1
dRt
1 t 1
dBt Vt 1 Bt 1
Bt 1
1
,andB
/Bt-1was
t-1
Bt 1 Vt 1 Bt 1
greater than 1/(Vt-1Bt-1), then dRt/dBt
is always positive and greater than
zero.Thisassociationisstatedinthe
followinghypothesis.
Changes in Growth
Opportunities
Thefirmsbookvaluedependson
thechangeingrowthopportunities(gt).
In other words,stock return depends
on whetheror not the firmgrows. A
firmiscalledanoptiontogrowifitcan
increase its book value and, in turn,
increase its stock price. Similarly, a
firmiscalledanoptiontoexpandwhen
itcouldgeneratefutureearningsfrom
itsassets.Thegrowthconceptisalso
inspiredbythefirmsabilitytogener
ate future earnings from multiplied
shortrun and longrun assets
(C((srt )+(lrt )). It infers that assets
growthmaybedifferentfromthegrowth
ofbookvalue.Therefore,growthop
portunities(gt),afterbeingadjusted
byBt-1/Vt-1andconsideringthemulti
pliereffectofC((srt)+(lrt)), arecon
jecturedtohaveapositiverelationwith
stock price variation (Rao and
Litzenberger 1971; Litzenberger and
Rao1972;BaoandBao1989;Weisset
al.2008;Ohlson1995;Abarbanelland
Bushee 1997; Lev and Thiagarajan
1993; Danielson and Dowdell 2001;
andAboodyetal.2001).
Thechangeinbookvalue,which
increasesproportionallywiththegrowth
of beginning shortrun and longrun
investedassets,supportsthispositive
dR
V t 1
202
association.With dg t C ( sr t ) C
B
C ( lrt ) t 1 ,whenBt-1/V/t-1isgreater
dRt
dg t isgreater
thanzero.Thehypothesisisstatedas
follows.
H A5 : The change in growth opportunities is positively associated
with stock return
t
t 1
With d r v 3 V ,whenBt-1/
t
t 1
Vt-1isgreaterthanzero,andv3isone
unitinvestment,because rt
V t 1
B t 1
1
,then
k
becomessmallerthanzero.
Hence,ournexthypothesis is as
follows.
HA6: The change in discount rate is
negatively associated with
stock return
Research Methods
Data
All cashflowrelated factors de
terminingthereturnmodelinthisre
search(earningsyield,expectedearn
ingsyield,shortruninvestmentassets
andexpectedshortruninvestmentas
sets, longrun investment assets and
expected longrun investment assets,
thechangeincapital,andthechangein
growthopportunitiesandthechangein
expected growth opportunities) are
gathered from financial statements.
Dataonexpectedvaluesandfinancial
statementsprospectusescanbefound
inthenotestofinancialstatements.All
dataareobtainedfromOSIRISdata
base.Thechangeindiscountratedata
are obtained from the central banks
websiteofeachcountry,eventhough
the financial statements of each firm
alsocontainlongtermliabilitiesorob
ligationinterestrate.Pureinterestrate
isproxiedbythelongtermobligation
interest rate enacted by the central
bankineachcountry.Thisstudy,then,
extracts stock price and return for
each firm from the stock markets in
everycountrydirectly.
Thisstudysobservationembraces
allAsiaPacificcountriesandtheU.S.,
along with their stock markets and
centralbanks.Thisstudyemploysdata
during20022009,excluding2003and
2008becauseoffinancialcrisisonall
stock markets. However, these years
arestillincludedtobethebaseyearfor
calculating the expected value com
paredtopreviousyears.
203
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
<
204
Rit = +xit+qit+bit+
git+rit+eit
........................................(11)
<
Thisstudyusesthepurposivesam
plingwhereasetofsamplearechosen
undercriteria suitedfor researchob
jectives. The criteria are as follows.
Firstly,sampleiscomprisedofmanu
facturingandtradingfirms.Secondly,
iteliminatesfirmswithnegativebook
values at the beginning and the end
(Bit-1 <0; Bit <0). This exclusion is
based on the logical reasoning that
firmswithnegativebookvaluestendto
abandonoperationsowingtotheirshort
run and longruncapacities. In other
words,thosefirms are inclinedtogo
broke.Thirdly, sampleconsistsoffirms
whosestocksaretradedactively.Sleep
ing stocks are excluded as they can
compromise this researchs validity.
This study also selects sample with
liquidity(LQ-n)accordingtoeachstock
market.
Thisstudyisaimedatimproving
ChenandZhangs(2007)model.There
fore,thisresearchiscarriedoutthrough
the following stages. Firstly, we ex
amineChenandZhangs(2007)model.
Secondly,this studyexamines a new
model using Equation (11). Thirdly,
thisstudycomparestheresultsofex
aminations(1)and(2).
Thefirstexaminationislinearre
gressionasfollows.
<
Sampling Method
<
withRitisannualstockreturnforfirm
iduringperiodt,measuredinoneyear,
one year and three months, one year
andsixmonths,andoneyearandnine
months. The calculation begins from
thefirstdayofthebeginningyeartothe
endofthemonthduringperiodt;xitis
earnings generated by firm i during
period t, calculated by earnings ac
quiredbycommonstockholdersduring
period t (Xit) divided by the opening
marketvalueofequityincurrentperiod
(Vit-1 ); qit (qit qit 1 ) Bit 1 / Vit 1
isthechangeinprofitabilityoffirmi
duringperiodt,deflatedbytheopening
bookvalueofequityincurrentperiod.
Profitabilityiscalculatedusingthefor
mulaqit= Xit / bit-1; b [( B
it
it
openingbooktomarketequityratioin
currentperiod; g it ( g it g it 1 ) B
) Bit 1 / Vit 1 isthechangeingrowthis
thechangeingrowthopportunitiesfor
firmiduringperiodt; rit (rit rit 1 ) B
<
<
<
<
qit+bit+git +
<
<
growthopportunitiesforfirmiduring
periodtmeasuredbyconsideringthe
Rit = +xit+qit+qit+
git + git +
<
<
multipliereffectofgrowthopportuni
tiesagainstshortrunandlongrunin
vestedassets.Itisthenadjustedbythe
openingbooktomarketequityratioin
currentperiod;othervariablesareiden
tical.
It shouldbe notedthat Rit in re
gressionmodel(13)representsvarious
return periods, namely one year, one
year and three months, one year and
six months, and one year and nine
months. This study applies multiple
periods because by inducing invest
mentscalability,currentshortrunand
longrun assets are considered to be
utilizedtogeneratecurrentandfuture
earnings. Therefore, different return
periodsrefertocurrentreturn(Rit)and
potentialfuturereturn(Ri,t+1).Never
theless,itisstillnotatedasRit.
rit + eit
........................................(13)
205
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
206
Table1.Sample Data
No.
Decrease
Number
%
Note
Sample
Number
%
1
2
3
4
5
6
7
Population
Stockpricedataincomplete
8,939
Earningsdataunavailable
661
Expecteddataunavailable
8,038
Lossingcompanyexclusion
167
Extremevalueexclusion
120
Inabilitytocalculateabnormalreturn
38
37.10
2.74
33.36
0.69
0.50
0.16
24,095
15,156
14,495
6,457
6,290
6,170
6,132
100.00
62.90
60.16
26.80
26.11
25.61
25.45
Total
74.55
17,963
Min.
Max.
Mean
Ri1
0.9954
9.8966 0.8463
Ri2
0.9964
8.0000 0.4600
Ri3
0.9966
9.0000 0.1627
Ri4
0.9939
6.6310 0.0528
Xit
0.0000
46.2025 0.2092
q it
55.1125
58.8148 0.0571
b it
54.3503
33.3750 0.0873
g it
10.6073
54.4328 0.1977
r it
29.9957
28.9790 0.1362
sr it 506.3845 202.6165 0.0336
lrit 250.0161 289.1262 0.2959
p it
54.3503
33.3750 0.0873
PB it
0.0026
70.4000 1.0362
Vit
0.0100 6,843.3600 39.3251
Bit
0.0200 4,601.1500 29.8525
AR i1
2.6632
8.9513 0.0000
AR i2
2.3542
7.1236 0.0000
AR i3
1.8951
8.5445 0.0000
AR i4
1.3450
6.2174 0.0000
0.1667 1.2500
0.0151 0.7500
0.1981 0.3689
0.2450 0.2186
0.0532 0.1959
0.0313 0.0772
0.0608 0.0553
0.0056 0.1976
0.4694 0.0301
0.1125 0.4198
0.0368 0.2572
0.0608 0.0553
0.3594 1.2095
1.1600 16.3400
0.5400 10.6200
0.5655 0.3361
0.4069 0.2438
0.3150 0.1953
0.2785 0.1558
207
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Thisstudyperformsdataanalysis
toinvestigateinitialdatatendency.The
descriptive statistics are presented in
Table 2. Return for one year period
(Ri1)is 0.8463, whichthendecreases
overtimeandplungesto0.0528forRi4.
Thedecreasesoccurinalllevelswithin
25thpercentile(from0.1667to0.2450)
and 75th percentile (from 1.2500 to
0.2186). These findings indicate that
market value in the longer period is
closer to real firms intrinsic value.
Withthistendency,thefirmsfunda
mentalvaluecalculatedusingaccount
ing information is expected to be re
flectedinthefirmsmarketvalue.
Focusingonearnings aftertaxes
(xit), this study only employs profit
firms. Earnings minimum value is
0.0000, with mean 0.2092, median
0.0968,andstandarddeviation0.9104.
The median lies on the left from its
mean,signalingthatsomefirmshave
extremely great earnings, and so the
meanispushedupward.However,itis
notaproblemasthestandarddeviation
islessthanone.Thealignedmovement
between return and earnings shows
thattheyarelikelytoberelated.The
change in earnings power (qit), the
changeingrowthopportunities(git),
andlongrun assets scalability (lrit)
showrelativelythesamepatternasthe
variationofearnings.Meanwhile,the
change in discount rate (rit ), the
changeinshortrunassets scalability
(srit),andthechangeinprofitability
(pit) show otherwise. However, the
changeindiscountrateisnotexpected
tobealigned.Nevertheless,thechange
inshortrunscalabilityandthechange
208
sequently,thisstudyconcludesthatthe
basic model is adequately substanti
ated,exceptforearningspower.How
ever,thebasicmodelanalysisshowsa
sufficientdegreeofassociationwithF
valueof35.5187,whichissignificantat
1 percentlevel.Thebasicmodelhas
R2of2.82 percentforRi1,andlower
forothertypesofreturn.Thedegreeof
association with the adjusted level is
notsignificantlydifferent,withadj-R2
of2.74 percent.
Theresultsoftheinitialinvestiga
tion are interesting. The rejection of
earningspower(qit)behooves us to
changethebasicmodel.Theresultsof
thebasicmodelanalysisimplythatthe
relationbetweenaccountinginforma
tion and stock return is not flexible
enough with respect to the forms of
stock market efficiency, economic
uncertainty,andthereflectionoffirms
fundamentalvaluepertainingtodebtor
capitalconcentration.Theresultsneed
totransformthebasicmodelintoanew
modelwhichismoredetailedandable
toexplainthechangeinearningspower.
Furthermore,thetransformationdoes
notconsiderthechangeinpureinterest
rate(rit),whichisactuallyservesas
aliftforthechangeinearningspower.
The change in pure interest rate has
beenproveninconsistentlybyprevious
studies.Thisstudyconjecturesthatthe
changeinpureinterestrateshouldbe
morereflectedwhenitisspecifiedinto
shortrunorlongrunearningspowers.
209
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Xit
q it
b it
g it
r it
F-value
R2
Adj-R2
Ri1
Pred.
Coef.
t-value
Sig.
Coef.
t-value
?
+
+
+
+
-
0.8096
0.1452
0.0002
0.0450
0.0770
0.0370
61.3526
6.7848
0.0228
4.7703
7.0549
3.9584
35.5187
2.82%
2.74%
0.0000 ***
0.0000 ***
0.9818
0.0000 ***
0.0000 ***
0.0001
0.0000 ***
0.4447
0.0518
0.0071
0.0277
0.0438
0.0158
44.4938
3.1938
1.0400
3.8822
5.2991
2.2393
13.5133
1.09%
1.01%
Var(s)
Xit
q it
b it
g it
r it
F-value
R2
Adj-R2
Ri2
Ri3
Sig.
0.0000
0.0014
0.2984
0.0001
0.0000
0.0252
0.0000
***
***
***
Ri4
Pred.
Coef.
t-value
Sig.
Coef.
t-value
Sig.
?
+
+
+
+
-
0.1548
0.0203
0.0084
0.0191
0.0246
0.0000
19.5395
1.5765
1.5582
3.3806
3.7618
0.0070
6.0406
0.49%
0.41%
0.0000 ***
0.1150
0.1192
0.0007 ***
0.0002 ***
0.9944
0.0000 ***
0.0419
0.0397
0.0019
0.0256
0.0248
0.0017
6.0803
3.5517
0.4119
5.2008
4.3416
0.3432
10.9147
0.88%
0.80%
0.0000
0.0004
0.6805
0.0000
0.0000
0.7315
0.0000
***
***
***
***
***
***
***
Analysis of Investment
Scalability Model
The second analysis transforms
thebasicmodelanalysis,inwhichwe
includethechangeinearningspower
(qit)intoamodelusingthechangein
shortrun earnings power (srit) and
longrun earnings power (lrit). This
modelisalsocalledtheshortrunand
longruninvestmentscalabilityinduc
ingmodel(Model13).Themodelspeci
fiestheearningspowerintomorede
tailedformstoinvestigatetheirasso
ciations with the variation of stock
price. Table 4 presents the analysis
results.
TheresultsofModel12showthat
earningsyield(xit),thechangeinbook
value (bit), the change in shortrun
earningspower(Dsrit), thechangein
longrun earnings power (lrit), the
changeingrowthopportunities(git),
andthechangeindiscountrate(rit)
areassociatedwithstockpricemove
ment. Consequently, HA1 , HA4 , and
HA5 are confirmed at 1 percent level
for return models Ri1 Ri4. HA4 is
partiallysupportedat10percentlevel
onlyforRi1 returntypewithtvalueof
1.7644.HA3issupportedforRi1return
typeaswellasforRi2 withtvalueof
1.7466,whichissignificantat10per
cent level. The results of Model 13
examinationshowanadequatedegree
ofassociationwithFvalueof31.3601,
whichissignificantat1percentlevel.
ThemodelhasR2of2.98percentfor
Ri1 type, and lower for other return
types. The model has adj-R2 of 2.89
percent.
The analysis results show that
Model13isabletoexplaintherelation
betweenthechangeinearningspower
(qit)andstockreturnvariationafter
specifyingitintomoredetailedforms,
i.e.,shortrun(srit)andlongrun(lrit)
investment scalabilities. HA2 and HA3
are confirmed for both Ri1 and Ri2
returntypes.HA2 isalsosupportedfor
Ri2 return type. The findings suggest
thattheeffectofearningspoweronthe
aggregate value is actually weak.
Therefore,splittingtheearningspower
intomoredetailedformsisnecessary.
Therefore,itsassociationwiththevaria
tion of stock return becomes more
comprehensible. Model 13 is better
than the basic modelin its degree of
association with adj-R2 of 2.89 per
cent, which is better than that of the
basicmodel(2.74%).
211
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Ri1
Pred.
X it
srit
lrit
p it
g it
r it
Fvalue
R2
AdjR2
?
+
+
+
+
+
Coef.
t-value
0.8075
0.1447
0.0030
0.0035
0.0461
0.0833
0.0374
31.3601
2.98%
2.89%
61.4695
7.9547
2.6663
1.7644
4.9185
7.5241
4.0118
0.0000
Var(s)
Sig.
Coef.
t-value
Sig.
1.13%
1.03%
44.5037
4.3603
1.7446
0.4149
4.0283
5.4937
2.2068
0.0000
0.0000
0.0000
0.0811
0.6782
0.0001
0.0000
0.0274
***
Coef.
t-value
Sig.
0.95%
0.85%
6.0579
4.3937
1.3158
1.6076
5.2351
4.6801
0.3181
0.0000
0.0000
0.0000
0.1883
0.1080
0.0000
0.0000
0.7504
***
Ri3
Pred.
X it
srit
lrit
p it
g it
r it
Fvalue
R2
AdjR2
Ri2
?
+
+
+
+
+
***
***
*
***
***
Ri4
Coef.
t-value
0.1535
0.0305
0.0008
0.0013
0.0200
0.0250
0.0004
5.0317
0.49%
0.39%
19.4414
2.7868
1.1375
1.0701
3.5407
3.7516
0.0790
0.0000
Sig.
***
***
***
***
Notes:Numberofobservation(N):6,132.Rit:stockreturn,firmiduringperiod1 (ayear),2(ayearandthreemonths),3(a
yearandsixmonths),and4(ayearandninemonths); xit:earnings,firmiduringperiodt; srit:changeofshortrunassets
scalability,firmiduringperiodt; lrit:changeoflongrunassetsscalability,firmiduringperiodt; pit:changeof
profitability,firmiduringperiodt;git:changeofgrowthopportunities,firmiduringperiodt; rit:changeofdiscountrate,
firmiduringperiodt;***significantat1%level,**significantat5%level,*significantat10%level.Linearitytestfor
thismodel13showsthat:(1)WithKolmogorovSmirnovtestshowstvalue9.035andpvalue0.000,andJarqueandBerra
showstvalue15,202.42andchisquare0.000,itmeansthattheresidualsarenotdistributednormally.However,normality
testisignorableforlargedatasamplethatis6,132.Ittendstofollowcentrallimittheorem(Gudjarati2003).(2)Glejsers
testforheteroscedasticityshowsthatallvariableshavesignificanceabove0.05,withtvalue(sig.)xitamountto0.045
(0.964);sritamountto0.045(0.964);lritamountto0.035(0.972);pitamountto0.000(0.990);gitamountto0.067
(0.946);andritamountto0.000(0.990).Thetestshowsthatthedataisfreefromheteroscedasticityproblem.(3)
MulticolinearitytestshowsthatallvariableshaveVIFaboutonewhichmeansthatthereisnocolinearityamongvariables,
VIFvalueforeachvariableis,xitamountto1.731;sritamountto1.086;lrit amountto1.014;pitamountto1.650;git
amountto1.257;andritamountto1.008.
212
Sensitivity Analysis 1:
Categorical Arrangement
Sensitivity Analysis 2:
P/B Partitioning
Subsequently,thisstudyanalyzes
themodelbasedoncategoricaldiffer
entiation.Thisanalysisservestofinda
morefavorabledegreeofassociation.
Model14shouldhaveahighergood
nessoffitwhen,afterdifferentiation,it
hasahigherdegreeofassociationand
isstillconsistentwiththemainvari
ables. The results of categorical ar
rangement for the basic model are
presentedinTable5.
Thisanalysispurportstoidentify
the incremental explanatory power.
Moreover,thecategoricalarrangement
servestoidentifytheinitialsensitivity
such that hypotheses examination is
supported in accordance with the
theory. The categorical arrangement
for Model 14 exhibits that there are
positivedifferences(greaterthanzero)
forthechangesinearningspowerand
growthopportunities. HA1-HA5areac
cordinglysupported,asareModel13
above.Indetails,thechangeinearn
ingspowerforhighgroup(Hqit)has
agreaterdegreeofassociationwitht
valueof16.2990,whichissignificantat
1 percent level, compared to that of
mediumgroup(Hqit).Similarresults
areapplicabletogrowthopportunities.
Model 14 shows a better degree of
associationwithR2of12.34 percent
and adj-R2 of 12.21 percent for Ri1
returntype.Accordingly,Model14has
beenabletobetterexplaintheassocia
tionpowerrelativetothebasicmodel.
Therefore, the ratio between market
valueandbookvalueserveswellwithin
thenextanalysis.
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Ri1
Pred.
?
Xit
+
q it
+
Mqit H>M>0
Hqit H>M>0
b it
+
g it
+
Mgit H>M>0
Hgit H>M>0
r it
F-value
R2
Adj-R2
Ri2
Coef.
t-value
0.6058
0.1219
0.0188
0.0174
0.4895
0.0363
0.0453
0.1477
0.1975
0.0493
95.7330
12.34%
12.21%
18.7617
5.9680
2.1794
0.5442
16.2990
4.0447
4.2684
4.1981
5.5108
5.5458
0.0000
Var(s)
Sig.
Coef.
t-value
Sig.
8.60%
8.46%
4.5000
3.3264
1.7297
8.4532
17.2896
3.1501
2.1477
2.0283
8.7095
3.6413
0.0000
0.0000
0.0009
0.0837
0.0000
0.0000
0.0016
0.0318
0.0426
0.0000
0.0003
***
Coef.
t-value
Sig.
4.48%
4.34%
9.8938
4.1175
1.5931
7.0670
11.2394
4.7548
2.0831
5.8089
7.7714
1.1422
0.0000
0.0000
0.0000
0.1112
0.0000
0.0000
0.0000
0.0373
0.0000
0.0000
0.2534
***
Ri3
Pred.
?
Xit
+
q it
+
Mqit H>M>0
Hqit H>M>0
b it
+
g it
+
Mgit H>M>0
Hgit H>M>0
r it
F-value
R2
Adj-R2
***
***
***
***
***
**
**
***
Ri4
Coef.
t-value
0.1311
0.0297
0.0071
0.2334
0.3096
0.0161
0.0105
0.0978
0.1315
0.0050
46.4409
6.39%
6.25%
6.6248
2.3692
1.3465
11.9242
16.8177
2.9241
1.6150
4.5328
5.9864
0.9099
0.0000
Sig.
***
***
***
***
***
**
***
***
Notes:Numberofobservation(N):6,132.Rit:stockreturn,firmiduringperiod1 (ayear),2(ayearandthreemonths),3(a
yearandsixmonths),and4(ayearandninemonths); xit:earnings,firmiduringperiodt; qit:changeofprofitability,firm
iduringperiodt(inbasicmodelnotatedaspit); bit:changeofbookvalue,firmiduringperiodt;git:changeofgrowth
opportunities,firmiduringperiodt;rit:changeofdiscountrate,firmiduringperiodt;***significantat1%level,**
significantat5%level,*significantat10%level.Categoricalarrangementofprofitabilityandgrowthopportunitieswith
conditions,consecutively,gH>gM>0, andwH>wM>0servestoexaminetheassociationdegreerelatedtoprofitabilityand
growthopportunitiescharacteristics.
214
Discussion
Overall,ouranalysisprovidesevi
dence that six cashflowrelated fac
torsofaccountinginformationarere
lated to stock price variability with
directionsashypothesized.Thisstudy
interprets the accounting information
variables one by one, and suggests
someresearch findings.
Earnings Yields
Earningsyieldispositivelyrelated
tofirmvalue.Theresultsofthisstudy
supporttheclassicalconcept(Ohlson
1995),alongwiththederivativestudies
by Lo and Lys (2000), Francis and
Schipper (1999), Meyers (1999),
Bradshawetal.(2006),CohenandLys
(2006), Bradshaw andSloan (2002),
Bhattacharyaetal.(2003),Collinset
al. (1997), Givoly and Hayn (2000),
Kolev,MarquadtandMcVay(2008),
andWeissetal.(2008).Eventhough
Ohlson(1995)hasaflawwhereearn
ingsisanoisywhenmeasuringmarket
equityvalue,thisstudyconcludesthat
earningsistheprimarydeterminantof
the firms market value. Therefore,
thisstudydenotesthatearningsisthe
measures of value added in account
ing.Furthermore,itsmeasurabilityis
alwaysreflectedinthemarketvalue.
Correspondingwiththeevidence
of earnings being reflected in stock
pricevariability,thisstudyshowsthat
earnings is the fundamental signal
(Ohlson 1995; Feltham and Ohlson
1995,1996).Thisstudycomprehends
thatthefundamentalsignalisdigested
fromitscharacteristicswhichserveas
aliftfor firmperformance.Earnings
serves as a lift for operation perfor
manceaswellasforstockpricevari
ability.Earningsisperceivedbyfinan
cialusersastheprimarydeterminant
of the firms equity value. In other
words,thisstudysupportstheconcept
of recursion theory (Sterling 1968),
suggestingthatfirmvalueisidentified
frombookvalueandearnings.Conse
quently,wesuggestthatthevariation
ofstockpricefullyreflectsbookvalue
andearnings.Finally,thisstudycon
cludes that the association between
accountingearningsandstockpriceis
undeniable.
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
turepotentialearnings.Thisstudycon
cludesthatitistheroleofinformation
on financial positionespecially the
roles ofassets andliabilities,but not
equity capitalthat may become a
determinantofstockpricevariability.
Thomas2000;Liuetal.2001;Weisset
al.2008;ChenandZhang2007;Ohlson
1995; Feltham and Ohlson 1995;
FelthamandOhlson1996;Bradshaw
etal.2006;andAbarbanellandBushee
1997).
Growth Opportunities
This study supports Rao and
Litzenberger(1971),Litzenbergerand
Rao(1972),andBaoandBao(1972)
thatgrowthopportunitiesincreasefirm
competitiveness. Consequently, the
higher the efficiency, the higher the
productivityis.MillerandModigliani
(1961)suggestthatgrowingfirmsal
wayshaveapositiverateofreturnfor
eachinvestedasset,meaningthatev
eryinvestedresourcehasalowercost
ofcapitalthanthatofotherfirmsinthe
industry.
Thisstudypositsthatfirmvalueis
determinedbygrowthandfuturepo
tentialgrowthopportunities(Liuetal.
2001;Aboodyetal.2002;andFrankel
andLee1998).Currentgrowthdrives
theincreaseinfutureearnings,while
future potential growth reduces the
modelserrortoimprovetheassocia
tion degree of the return model. Lev
and Thiagarajan (1993), Abarbanell
and Bushee (1997), and Weiss et al.
(2008) conclude that the growth of
inventories,grossprofit,sales,accounts
receivable,etc.improvesfutureearn
ings growth.Simultaneously, this re
217
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Model
This study conducts five model
examinationswithtwosensitivitytests.
The results of investment scalability
analysisshowthatModel13hasadjR2 of 2 percent3 percent, which is
higherthanthatofModel12(2%).This
study shows that the newly designed
modelhasabetterdegreeofassocia
tion,andcouldexplainthereturnasso
ciation by 1 percent increase. Next,
this study examines the models by
categoricalarrangementbasedonP/B
ratio.Theanalysisresultsdemonstrate
that adj-R2 is within the range of 6
percent11percent.Thesefindingsin
dicatethatwhensampleisdifferenti
ated categorically into subsamples,
thedegreeofassociationofthereturn
model increases.It is alsonoted that
theincrementalexplanatorypoweris
around9percentcomparedtothatof
thebasicmodel.Theanalysisbasedon
P/B ratio partition confirms that the
modelshowsahighdegreeofassocia
tion with adj-R2 of approximately 5
percent38percent,whichisapproxi
mately 10 percent20 percent higher
thanthoseofthetwopreviousanaly
ses.Uptothisstage,thisstudyisable
toshowabetterdegreeofassociation
ofthereturnmodel.Thus,thismodelis
more comprehensive, realistic, and
accurate.
Research Findings
Theresultsofoverallanalysiscon
firmthetheory,andprovidesomeem
piricalevidence.First,allaccounting
218
fundamentals,assuggestedbytheory,
areconfirmedtoberelatedwithstock
pricevariability.Allcashflowrelated
variables,i.e.,earningsyield,shortrun
andlongruninvestmentscalabilities,
bookvalue,andgrowthopportunities
are positively associated with stock
pricevariability.Meanwhile,thechange
indiscountrateorpureinterestratehas
a negative relation with stock price
variability.Second,thechangeinearn
ingspowerinasinglemeasureisfound
toweaklyexplainstockpricevariabil
ity.Untilrecently,someempiricalevi
dencemeasurestheearningspoweras
asingleunit.Thisstudysplitsthismea
sureintoshortrunandlongruninvest
mentscalabilities,andfindsthatboth
measuresarepositivelyassociatedwith
annualstockreturn.Theexamination
usingP/Bratiopartitionshowsconsis
tentresultsforsubsampleswithlowto
mediumhighP/Bratio.
Third,thisstudycouldsynergize
the adaptation theory (Wright 1967)
with the recursion theory (Sterling
1968). Earnings has explained stock
pricevariabilityforhalfacentury,show
ing that the recursion theory is still
valid.Ontheotherhand,thefindingon
shortrun and longrun investment
scalabilitiesimpliesthattheadaptation
theory is also prevalent. This study
combinesboththeoriesintoonemodel,
andfindsthatboththeoriesindeedhold,
andeventhemodelhasabetterdegree
of association. The recursion theory
thatreliesonearningsandbookvalue
as showninOhlsonsmodel(Ohlson
1995;FelthamandOhlson1995;1996)
iscalledtheorthodoxparadigm.This
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
stockprice.Inotherwords,stockprice
adjusts to growth opportunities. The
change in discount rate is negatively
relatedto annualstock return, which
stems fromthe use ofcheap alterna
tiveresources orlowerinterestrates.
All examination results confirm the
hypothesized directions. In addition,
thesensitivitytestbasedonP/Bratio
showssimilarresults.Thisstudydeliv
ers abetterdegree ofreturnassocia
tion.Althoughthisparticularfindingis
comparablewiththatofpreviousstudy,
whichshowsalowdegreeofrelation,
this studycontributes an incremental
explanatorypower.
Theassociationbetweenaccount
ingfundamentalsandthevariationof
stockpricecategorizedbyP/Bratiois
confirmed as suggested by theory.
Specifically,highandmediumhighP/
Bratioscouldexplainstockpricevari
abilitybetterthandoeslowerP/Bratio.
Withinthetheoreticallevel,thisstudy
findsempiricalevidenceofthesynergy
betweentheadaptationtheoryandthe
recursiontheory.Therefore,investors
shouldnotmerelyuseinformationre
latedtoearningsandbookvalue,but
theyshouldanalyzethecharacteristics
of investment scalability or invested
resources.
This study documents a higher
degree of association between stock
pricevariabilityandaccountingfunda
mentalsthandopreviousstudies.The
relationhasmoresignificantresultsin
thesubsamplepartition,especiallywith
P/Bratio.Overall,thefindingsleadto
conclusionthattherelationofaccount
inginformationtostockpricevariabil
220
Limitations
Thisstudyhassomelimitationsas
follows. First, it uses a large data
samplesothatitsAdj-R2islowdueto
thelawoflargedatasample.Second,
thisstudyhasasurvivorship biasinits
sample.Ofall24,095firmyears,this
study only uses 6,132 (25.45%) be
cause the remainders are un
analyzable.Third,thisstudydoesnot
employfirmswithnegativebookval
uesandnegativeearningsaftertaxes,
asitusespurposivesamplingcriteria.
Future researchers should consider
employingthemasthecontrolgroup.
Because of their unavailability, this
studyfailstoconductrobustnesschecks
forthisgroup.Fourth,thereisabias
duetotheblendingofallstockmarkets,
from semistrong to weak forms of
efficiency.Althoughthislimitationis
deniable by the marketwide regime
concept,thisstudyignoresthecharac
teristicsofeconomies,regulations,trad
ing mechanisms, and cultures across
countries. In fact, those factors may
affectthereturnmodel.
Fifth,thisstudyusesearningsaf
tertaxes,andsoitdisregardsearnings
quality, which may affect the return
model.Nevertheless,thisissueis not
influentialasthesampletendstoshow
alow P/Bratio. This means that the
sampleusuallyhasgoodearningsqual
ity.Sixth,thisstudydoesnotconsider
conservatisminthepublishedfinancial
References
Abarbanell,J.S.,andB.J.Bushee.1997.Fundamentalanalysis,futureearnings,andstock
returns.Journal of Accounting Research35(1):124.
Aboody, D.;J. Hughes,and J.Liu. 2001. Measuring valuerelevance in a (possibly)
inefficientmarket.Journal of Accounting Research40(4):965986.
Baginski,S.P.,andJ.M.Wahlen.2003.Residualincomerisk,intrinsicvalue,andshareprice.
The Accounting Review78(1):327351.
Bao,BH.,andDaH.Bao.1989.Anempiricalinvestigationoftheassociationbetween
productivityandfirmvalue.Journal of Business Finance and Accounting16(5):699
717.
Bhattacharya,N.,E.Black,T.Christensen,andC.Larson.2003.Assessingtherelative
informativenessandpermanenceofproformaearningsandGAAPoperatingearn
ings.Journal of Accounting and Economics36:285319.
Beaver,W.H.1999.Commentson:Anempiricalassessmentoftheresidualincome
valuationmodel.Journal of Accounting and Economics26:3442.
Beaver,W.,R.Lambert,andD.Morse.1980.Theinformationcontentofsecurityprices.
Journal of Accounting and Economics2(March):328.
Bradshaw,M.T.,S.A.Richardson,andR.G.Sloan.2006.Therelationbetweencorporate
financingactivities,analysisforecastandstockreturn. Journal of Accounting and
Economics42:5385.
Bradshaw,M.,andR.Sloan.2002.GAAPversusthestreet:Anempiricalassessmentoftwo
alternativedefinitionsofearnings.Journal of Accounting Research40:4166.
Burgstahler,D.,andI.Dichev.1997.Earnings,adaptation,andequityvalue.The Accounting Review73:187215.
Chen,P.,andG.Zhang.2007.Howdoaccountingvariablesexplainstockpricemovements?
Theoryandevidence.Journal of Accounting and Economics43:219244.
Cohen,D.A.,andT.Z.Lys.2006.Weighingtheevidenceontherelationbetweenexternal
corporatefinancingactivities,accrualandstockreturn.Journal of Accounting and
Economics42:87105.
Collins,D.W.,E.L.Maydew,andI.Weiss.1997.Changesinvaluerelevanceofearnings
andbookvaluesoverthepastfortyyears.Journal of Accounting and Economics24
(1):3967.
221
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Collins,D.W.,M.Pincus,andH.Xie.1999.Equityvalueandnegativeearnings:Therole
ofbookvalueofequity.The Accounting Review74(1):2961.
Copeland,T.,A.Dolgoff,andA.Moel.2004.Theroleofexpectationsinexplainingthecross
sectionofstockreturns.Review of Accounting Studies9:149188.
Conrad,J.,B.Cornell,andW.R.Landsman.2002.When is bad news really bad news?
Journal of Finance57(6).
Danielson,M.G.,andT.D.Dowdell.2001.Thereturnstagesvaluationmodelandthe
expectationswithinafirmsP/BandP/ERatios.Financial Management30(2)
(Summer):93124.
David,A.,1997.Fluctuatingconfidenceinstockmarket:Implicationforreturnandvolatility.
Journal of Financial and Quantitative Analysis32(December):427482.
Dogramaci,A.1981.Perspectivesonproductivity.InA.Dogramaci(MartinusNijholf1981),
Productivity Analysis: 123.
Drucker,P.1986.Ifearningsarentthedialtoread.Wall Street Journal30(October):132.
Easton,P.D.,andT.S.Harris.1991.Earningsasanexplanatoryvariableforreturns.Journal
of Accounting Research29(1)(Spring):1936.
Feltham,G.,andJ.Ohlson.1995.Valuationandcleansurplusaccountingforoperatingand
financingactivities.Contemporary Accounting Research11:689731.
Feltham,G.,andJ.Ohlson.1996.Uncertainresolutionandthetheoryofdepreciation
measurement.Journal of Accounting Research34:209234.
FinancialAccountingStandardsBoard(FASB).1978.Statement of Financial Accounting
Concepts No. 1: Objectives of Financial Reporting by Business Enterprises.Stam
ford,CT.
FinancialAccountingStandardsBoard(FASB).1984.Statement of Financial Accounting
Concepts No. 5: Recognition and Measurement in Financial Statements of Business
Enterprises.Stamford,CT.
Francis,J.,andK.Schipper.1999.Havefinancialstatementlosttheirrelevance.Journal of
Accounting Research37(2)(Autumn):319352.
Frankel,R.,andC.M.C.Lee.1998.Accountingvaluation,marketexpectationandcross
sectionalstockreturn.Journal of Accounting and Economics25(3):283319.
Givoly,D.,andC.Hayn.2000.Thechangingtimeseriespropertiesofearnings,cashflows
andaccruals:Hasfinancialreportingbecomemoreconservative?Journal of Accounting and Economics29:287320.
Gujarati,D.N.2003.Basic Econometrics (4thed.).Boston:McGrawHillFirms,Inc.
Hand,J.R.2001.Discussionofearnings,bookvalues,anddividendsinequityvaluation:
Anempiricalperspective.Contemporary Accounting Research18(1):121130.
Hatsopoulos.G.1986.Productivitylagisrealtradebarrier.Wall Street Journal14(May):
30.
Ho,YK.,andJ.M.Sequeira.2007.Earningssurprises,asymmetryofreturns,andmarket
levelchanges:Anindustrystudy.Journal of Accounting, Auditing and Finance20
(February):2955.
222
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Sterling,R.1968.Thegoingconcern:Anexamination.The Accounting Review(July):481
502.
Veronesi,P.1999.Stockmarketoverreactiontobadnewsingoodtimes:Arational
expectationsequilibriummodel.Review of Financial Studies12(Winter):9751007.
Warfield,T.D.,andJ.J.Wild.1992.Accountingrecognitionandtherelevanceofearnings
asanexplanatoryvariableforreturns.The Accounting Review67(4):821842.
Weiss,D.,P.A.Naik,andCL.Tsai.2008.Extractingforwardlookinginformationfrom
securityprices:Anewapproach.The Accounting Review83(4):11011124.
Wright,F.K.1967.Capacityforadaptationandtheassetmeasurementproblem.Abacus
(August):7479.
Zhang,G.2003.Accountinginformation,capitalinvestmentdecisions,andequityvalua
tion:Theoryandempiricalimplications.Journal of Accounting Research38:271295.
224
APPENDIX 1
The Results of Inducing the Change in Investment Scalability Analysis
Ri1
o w
P B
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
Ri2
Coef.
t-value
Sig.
0.9262
3.6746
0.0002
0.0306
0.0414
0.7296
1.9473
26.3673
15.2294
0.0892
2.0497
2.6972
9.9661
9.4720
56.8679
21.86%
21.47%
0.0000
0.0000
0.9289
0.0406
0.0071
0.0000
0.0000
0.0000
Coef.
***
***
***
***
***
0.8079
0.8236
0.0006
0.0198
0.0293
0.0601
1.2520
25.2705
3.7505
0.3446
1.4573
2.1012
0.9018
6.6911
11.7171
5.45%
4.98%
Ri1
P B
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
t-value
Sig.
0.0000
0.0002
0.7305
0.1453
0.0358
0.3673
0.0000
0.0000
***
***
**
***
***
Ri2
Coef.
t-value
Sig.
0.4863
0.5175
0.0008
0.0187
0.0175
0.0061
0.8871
9.2309
4.34%
3.87%
18.5310
2.8714
0.5967
1.6717
1.5289
0.1121
5.7763
0.0000
0.0000
0.0042
0.5509
0.0948
0.1265
0.9107
0.0000
***
Coef.
***
***
***
t-value
Sig.
3.43%
225
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
L o w - M e d i u m
P B
L o w - M e d i u m
P B
226
Ri1
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
Ri2
Coef.
t-value
Sig.
0.9273
0.1362
0.0055
0.0041
0.0517
0.7051
0.0479
13.7341
6.33%
5.87%
27.1602
2.2081
1.0154
0.4881
1.0571
8.4779
0.7671
0.0000
0.0000
0.0274
0.3101
0.6256
0.2907
0.0000
0.4432
***
***
**
***
Coef.
t-value
0.4944
0.1212
0.0022
0.0046
0.0740
0.4950
0.0464
13.7660
6.35%
5.88%
20.1531
2.7338
0.5594
0.7620
2.1054
8.2823
1.0332
0.0000
Ri1
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
Sig.
0.0000
0.0064
0.5760
0.4462
0.0355
0.0000
0.3017
***
***
***
**
***
Ri2
Coef.
t-value
Sig.
0.2299
0.1147
0.0035
0.0000
0.0808
0.2773
0.0938
9.3563
4.40%
3.93%
11.6663
3.2218
1.1290
0.0072
2.8596
5.7771
2.6034
0.0000
0.0000
0.0013
0.2591
0.9942
0.0043
0.0000
0.0093
***
Coef.
***
***
***
***
***
t-value
Sig.
4.88%
M e d i u m- M e d i u m
P B
Ri1
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
Ri2
Coef.
t-value
Sig.
0.5106
1.0372
0.0072
0.0060
0.0226
0.8992
0.0213
44.6507
18.01%
17.60%
17.1578
13.2985
1.1376
0.7466
1.0891
10.0492
0.7259
0.0000
0.0000
0.0000
0.2555
0.4554
0.2763
0.0000
0.4680
***
Coef.
***
***
0.1834
0.5999
0.0062
0.0123
0.0227
*** 0.7072
0.0118
40.0357
16.45%
16.04%
M e d i u m- M e d i u m
P B
Ri1
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
t-value
8.9008
11.1093
1.4053
2.2334
1.5800
11.4168
0.5816
0.0000
Sig.
0.0000
0.0000
0.1602
0.0257
0.1144
0.0000
0.5609
***
***
***
***
Ri2
Coef.
t-value
Sig.
0.0236
0.3527
0.0040
0.0046
0.0052
0.3908
0.0421
23.7174
10.45%
10.01%
1.5134
8.6174
1.1825
1.0887
0.4738
8.3219
2.7392
0.0000
0.1304
0.0000
0.2372
0.2765
0.6357
0.0000
0.0062
***
Coef.
t-value
Sig.
11.28%
227
Gadjah Mada International Journal of Business, May-August 2010, Vol. 12, No. 2
Ri1
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
Coef.
?
+
+
+
+
+
0.2714
1.6294
0.0010
0.0030
0.0258
0.2448
0.0279
127.7231
38.60%
38.30%
M e d i u m - H i g h
Coef.
228
Ri2
t-value
Sig.
11.2791
21.2164
0.4065
1.7680
1.9758
4.8097
1.8392
0.0000
0.0000
0.0000
0.6844
0.0773
0.0484
0.0000
0.0661
***
Coef.
***
***
*
**
***
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
t-value
Sig.
25.28%
Ri1
P B
M e d i u m - H i g h
P B
Ri2
Coef.
t-value
Sig.
0.1066
0.5053
0.0008
0.0008
0.0059
0.0192
0.0331
34.7146
14.59%
14.17%
8.0359
11.9339
0.5618
0.8013
0.8180
0.6838
3.9548
0.0000
0.0000
0.0000
0.5744
0.4231
0.4135
0.4942
0.0001
***
Coef.
***
***
***
t-value
Sig.
8.58%
g h
P B
Coef.
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
Ri2
Coef.
t-value
Sig.
0.4335
0.0938
0.0053
0.0047
0.0359
0.0688
0.0247
19.6192
8.81%
8.36%
21.6449
4.1453
4.5853
1.4485
2.2089
6.8601
3.2440
0.0000
0.0000
0.0000
0.0000
0.1477
0.0274
0.0000
0.0012
***
Coef.
P B
h
g
i
H
Pred.
X it
srit
lrit
pit
git
rit
Fvalue
R2
AdjR 2
?
+
+
+
+
+
Coef.
Sig.
***
***
***
4.81%
Ri3
Coef.
t-value
Ri4
t-value
Sig.
0.1275 12.4631
0.0259
2.2394
0.0019
3.1430
0.0007 0.4333
0.0162
1.9544
0.0237
4.6342
0.0013 0.3272
5.7043
0.0000
2.73%
2.25%
0.0000
0.0253
0.0017
0.6649
0.0509
0.0000
0.7435
***
Coef.
***
**
***
*
***
t-value
Sig.
2.94%
Additional Notes:Numberofobservation(N)forLowPB:1,227,LowMediumPB:1,226,
MediumMediumPB:1,227,MediumHighPB:1,226,HighPB:1,226.ThelimitsofeachPB:
LowPB<0.3065;LowMediumPB<0.5462;MediumMediumPB<0.8505;MediumHigh
PB<1.3687,HighPB>1.3687.
229