intellectual capital investments: evidence from panel var analysis
TRANSCRIPT
Intellectual capital investments:evidence from panel VAR analysis
Iuliia Naidenova and Petr ParshakovDepartment of Financial Management,
National Research University Higher School of Economics, Perm, Russia
Abstract
Purpose – Investments in intellectual capital (IC) are often linked to competitive advantages thatimprove economic profit and increase the value of a company. However, this effect is reciprocal:companies that generate higher cash flow can invest more in IC. The purpose of this paper is toanalyze a dynamic relationship between IC components and economic profit, with a special emphasison industry specific effects in pharmaceutical, retail, steel, telecommunications, and service sectors.Design/methodology/approach – Panel vector autoregression was used to deal with the mutualinfluence of IC components, the lag effect, and heterogeneity. The data were taken from Compustatdatabase and covers the period from 2001 to 2010.Findings – The paper proves that there is interaction between investments in the IC componentsand company performance. However, there are sectoral differences: there is a positive impactof economic profit on human capital in retail; in the steel industry a mutual influence is revealed.Moreover, interaction between different IC components is detected in telecommunication andsteel industries.Originality/value – This is the first study to present clear evidence of the effects of performance onIC investment decisions. The time lag in the effects of IC investments was estimated and compared fordifferent industries. On the methodological side, the paper presents a rather simple method capable ofyielding results consistent with other studies and yet rich enough to be applied to an analysisof sectoral differences in dynamic IC investment decisions.
Keywords Intellectual capital, Vector autoregression, Financial performance, Interaction effect
Paper type Research paper
1. IntroductionNowadays the increasing in the importance of the intellectual resources of companiesis confirmed by empirical studies and is recognized in the business world (Lev, 1999).Consequently, the financial performance of a company and its equity attractiveness forpotential shareholders are also largely driven by its intellectual capital (IC). However,the means by which investment in IC actually allows companies to achieve success is adebatable question. So, if a company invest some funds in intellectual resourcesand expects them to create a competitive advantage, managers should measure thereturn on them.
IC theory considers such key companies’ resources as employee knowledge,information systems, relationships with suppliers and customers, and management.It combines existing achievements in different areas, such as intangible assetevaluation, theory of competitive advantage, resource-based approach to the theory ofthe firm, and human capital. It also disseminates approaches to human capital analysis
The current issue and full text archive of this journal is available atwww.emeraldinsight.com/1469-1930.htm
Received 30 January 2013Revised 7 March 20134 April 201316 April 2013Accepted 19 April 2013
Journal of Intellectual CapitalVol. 14 No. 4, 2013pp. 634-660r Emerald Group Publishing Limited1469-1930DOI 10.1108/JIC-01-2013-0011
The authors would like to thank Elena Shakina, Maria Molodchik, Anna Bykova, MarinaOskolkova, and Dmitri Vinogradov for the valuable advice. This study comprises researchfindings from the “The changing role of companies’ intangibles over the crisis” carried outwithin The National Research University Higher School of Economics’ Academic Fund Programin 2013, grant No. 13-05-0021.
634
JIC14,4
to other types of IC. There are three levels of IC analysis – macro level,sectoral level, and micro level – and in the present paper IC is considered at thecompany level.
Nowadays, the role of IC in corporate management is being actively investigated.A number of empirical studies show that the contribution of intellectual resources ina company’s value is significant (Bontis et al., 2000; Chen et al., 2004; Tseng and Goo,2005; Huang and Hsueh, 2007; Kamukama et al., 2010; Chang and Hiesh, 2011).A company’s IC is heterogeneous and is usually divided into several components. Sothe interaction between these components is also of interest to researchers. A featurethat should be noted is a theoretically possible delay between the impacts of oneintellectual component on performance or others components (Tseng and Goo, 2005;Kaplan and Norton, 1992; Chen et al., 2004). Although this delay is recognizedtheoretically, empirical investigations usually ignore it in regression analysis.
Thus, researchers recognize a company’s intellectual resources as being crucial forsurvival and for successful competition in the knowledge economy. However, most ofthese studies are based on the assumption that the impact is completely exhausted inone period, while theoretical assumptions run to the contrary. Moreover, these studiesignore the following inverse direction of relationship: company performance (earningsor economic profit) is closely connected with a company’s ability to invest, which inturn causes the growth of intellectual resources. Consequently, in models that do nottake into account the reverse effect, there is some endogeneity, which biases the model’sevaluation results toward giving a higher importance to IC. In the present paper it isattempted to solve this problem as reasonable a way as possible, and in addition to thatto take into account the mutual influence of intellectual resources and their relationshipwith a company’s tangible resources.
The remainder of the paper is organized as follows. Section 2 provides thetheoretical background on IC, its structure and features. Section 3 describes sampleselection, variables, research model, and method of analysis. Section 5 presentsthe results of this research that are discussed further in Section 6.
2. Literature reviewThere is no precise agreement on the definition of IC, but the majority of authors pointsout that it helps a company to compete. Some authors include this feature justin definition of IC: “Intellectual capital is intellectual material-knowledge, information,intellectual property, experience, that can be put to use to create wealth” (Stewart,1997). So the existence of a relationship between IC and a company’s wealthseems to be obvious. Although it is still a point at issue how IC can be converted intocompany wealth.
This wealth creation process can be illustrated by the input-output-outcome concept(Cheng et al., 2010; Molodchik et al., 2012). IC is treated as part of the resources (input)that a company invests in order to gain competitive advantage and to improveperformance (output) which then causes an increase in company value (output).However, papers based on company value miss a lot of information about companieswhich are not listed on a stock exchange. This study is focussed on the first stage, thatis, the transformation of intellectual resources into company performance.
IC is generally recognized as not being a one and indivisible company resource,and for investigation purposes it is usually subdivided into several components.Following previous studies in this field, three main structural components of IC areidentified: human capital, structural capital, and relational capital (Bontis et al., 2000;
635
Intellectualcapital
investments
Bontis, 1996, 1998, 1999; Roos et al., 1998; Stewart, 1991, 1997; Sveiby, 1997;Edvinsson and Malone, 1997; Edvinsson and Sullivan, 1996; Moon and Kym, 2006;Nazari, 2010):
. Human capital, in the concept of IC, is treated as knowledge that belongs toemployees, for example skills, abilities, and experience (Stewart, 1997; Lee, 2011).
. Structural capital is defined as “the knowledge that doesn’t go home at night”(Stewart, 1997). Although it is created by employees, it can also be separatedfrom them. First of all, structural capital includes organizational procedures,strategies, patents, manuals, and databases (Nazari, 2010). In other words,structural capital is determined by human capital, but at the same time it isindependent.
. Relational capital is a company’s ability to interact successfully with its externalstakeholders in order to develop the potential of value creation by enhancinghuman and structural capital.
Theoretical models of IC such as Scandia Navigator (Edvinsson and Malone, 1997)describe the interaction between various IC components. Empirical research usuallyconfirms the idea that dimensions of IC are interrelated (Chang and Hiesh, 2011; Huangand Wu, 2010; Cabrita et al., 2007; Kamukama et al., 2010; Bontis et al., 2000; Tseng andGoo, 2005).
Table I combines the hypotheses and findings of previous papers on the relationshipsbetween IC components and company performance. The common idea is that ICcreates competitive advantage for a company so it should contribute to the company’sperformance (Chang and Hiesh, 2011; Huang and Wu, 2010; Cabrita et al., 2007;Kamukama et al., 2010; Bontis et al., 2000; Tseng and Goo, 2005; Dıez et al., 2010).It should be noted that although the research question is roughly the same, the results arecontradictory. For example, some authors found that there is direct influence of humancapital on performance (Chang and Hiesh, 2011; Huang and Wu, 2010; Cheng et al., 2010),while other papers have contrary results (Bontis et al., 2000; Cabrita et al., 2007). Thusfurther research is required.
Some authors also assume that there is a certain inertia which delays total andimmediate use of benefits derived from IC investments. This is what Tseng and Goo(2005) call “effect of time delay”. This means that investments in human, relational,and structural capital take time to be fully implemented: in general, this propertyis common to all types of investment, but taking it into consideration, it shows the needto consider the return on investment in human capital with a certain lag. Norton andKaplan emphasize this feature in the construction of the balanced scorecard (Kaplanand Norton, 1992). They call operational indicators, such as customer satisfaction andrelationships with suppliers, “leading” or “factors of activity,” while they considerfinancial indicators to be “lag indicators.” Chen et al. (2004) also mentioned this effect,explaining the relatively low correlation between two IC components.
According to the literature on the relationship between investments andcompany performance, earnings are the determinant of investments (Love andZicchino, 2006; Eklund, 2010). Consequently inverse direction on influence should exist:company performance affect on investments in intellectual resources. Moreover ininvestment decision-making process management take into consideration thestability of earnings. So company performance in current years may affect on futureinvestments.
636
JIC14,4
Stu
dy
Cou
ntr
yan
dp
erio
dD
epen
den
tv
aria
ble
sH
yp
oth
eses
Res
ult
s
Ch
ang
and
Hie
sh(2
011)
Tec
hn
olog
yco
mp
anie
s
list
edin
Tai
wan
Sto
ck
Ex
chan
ge
(200
1-20
07)
HC
–h
um
anca
pit
al
SC
–st
ruct
ura
lca
pit
al
CE
–so
cial
cap
ital
TC
–te
chn
olog
ical
cap
ital
HC
,S
C,
CE
,T
C–
VA
ICm
odel
com
pon
ents
MP
–m
ark
etp
erfo
rman
ce(T
obin
Q)
1:H
C,
SC
,C
Ean
dT
Cp
osit
ivel
yaf
fect
MP
2:H
Cm
edia
tes
the
rela
tion
ship
sb
etw
een
SC
,
CE
,T
C,
and
MP
3:H
Cm
oder
ates
the
rela
tion
ship
sb
etw
een
SC
,
CE
,T
C,
and
MP
Sta
tist
ics
sup
por
tsa
med
iati
ng
/
mod
erat
ing
effe
ctof
HC
onC
Ean
d
SE
ofor
gan
izat
ion
s
Hu
ang
and
Wu
(201
0)
Tai
wan
bio
tech
nol
ogy
ind
ust
ry(T
BI)
and
Tai
wan
ese
ph
arm
aceu
tica
l
man
ufa
ctu
rers
(200
5)
HC
–h
um
anca
pit
al
OC
–or
gan
izat
ion
alca
pit
al
SC
–so
cial
cap
ital
KP
–k
now
led
ge
pro
du
ctiv
ity
HC
,O
C,
SC
,K
P–
dat
aob
tain
ed
from
qu
esti
onn
aire
s
1:T
he
gre
ater
HC
,th
eh
igh
erth
eK
P
2:T
he
gre
ater
OC
,th
eh
igh
erth
eK
P
3:T
he
gre
ater
SC
,th
eh
igh
erK
P
4:T
he
gre
ater
SC
,th
est
ron
ger
the
infl
uen
ce
ofH
Con
KP
5:T
he
gre
ater
SC
,th
est
ron
ger
the
infl
uen
ce
ofO
Con
KP
All
dim
ensi
ons
ofIC
pos
itiv
ely
and
sig
nif
ican
tly
infl
uen
ceK
P
Th
est
ud
yp
rov
esth
ere
are
inte
ract
ive
effe
cts
bet
wee
nth
eco
mp
onen
tsof
IC
and
KP
Car
bit
a
etal.
(200
7)
53b
ank
s,al
laf
fili
ated
mem
ber
sof
the
Por
tug
ues
eB
ank
ers
Ass
ocia
tion
HC
–h
um
anca
pit
al
SC
–st
ruct
ura
lca
pit
al
RC
–re
lati
onal
cap
ital
OP
–or
gan
izat
ion
alp
erfo
rman
ce
HC�
SC�
RC
–in
tera
ctio
nte
rm
HC
,R
C,
SC
,O
P–
dat
aob
tain
edfr
om
qu
esti
onn
aire
s
1:H
Cis
pos
itiv
ely
asso
ciat
edw
ith
SC
2:H
Cis
pos
itiv
ely
asso
ciat
edw
ith
RC
3:S
Cis
pos
itiv
ely
asso
ciat
edw
ith
RC
4:S
Cis
pos
itiv
ely
asso
ciat
edw
ith
OP
5:R
Cis
pos
itiv
ely
asso
ciat
edw
ith
OP
6:T
he
rela
tion
ship
bet
wee
nH
Can
dO
P
isp
osit
ivel
ym
oder
ated
by
the
inte
ract
ion
bet
wee
nS
Can
dR
C
Sig
nif
ican
tef
fect
sin
dic
ate
dir
ect
and
ind
irec
tre
lati
onsh
ips
bet
wee
nIC
com
pon
ents
and
OP
SC
and
RC
pos
itiv
ely
mod
erat
esth
e
rela
tion
ship
bet
wee
nH
Can
dO
P
Kam
uk
ama
etal.
(201
0)
65U
gan
dan
mic
rofi
nan
ce
inst
itu
tion
s(2
009/
2010
)
HC
–h
um
anca
pit
al
SC
–st
ruct
ura
lca
pit
al
RC
–re
lati
onal
cap
ital
FP
–fi
nan
cial
per
form
ance
HC
,R
C,
SC
,F
P–
dat
aob
tain
ed
from
qu
esti
onn
aire
s
1:H
Cp
osit
ivel
yaf
fect
sF
P
2:S
Cp
osit
ivel
yaf
fect
sF
P
3:R
Cp
osit
ivel
yaf
fect
sF
P
4:H
Cin
flu
ence
sF
Pif
itin
tera
cts
wit
hR
C
5:H
Cin
flu
ence
sF
Pif
itin
tera
cts
wit
hS
C
6:S
Cin
flu
ence
sF
Pif
itin
tera
cts
wit
hR
C
Th
em
agn
itu
de
effe
ctof
HC
onF
P
dep
end
son
any
ofS
Cor
RC
;hen
ceth
e
assu
mp
tion
ofn
onad
dit
ivit
yis
met
.
No
sig
nif
ican
tin
tera
ctio
nef
fect
sw
ere
esta
bli
shed
bet
wee
nR
Can
dS
C
(con
tinu
ed)
Table I.Prior studies investigating
the relationship betweenof intellectual capital (IC)
components andperformance
637
Intellectualcapital
investments
Stu
dy
Cou
ntr
yan
dp
erio
dD
epen
den
tv
aria
ble
sH
yp
oth
eses
Res
ult
s
Bon
tis
etal.
(200
0)
107
Mal
aysi
anfi
rms
from
serv
ice
and
non
serv
ice
ind
ust
ries
HC
–h
um
anca
pit
al
SC
–st
ruct
ura
lca
pit
al
CC
–cu
stom
erca
pit
al
FP
–fi
nan
cial
per
form
ance
HC
,R
C,
SC
,F
P–
dat
aob
tain
ed
from
qu
esti
onn
aire
s
1:H
Cp
osit
ivel
yas
soci
ated
wit
hC
C
2:H
Cas
soci
ated
wit
hS
C
3:C
Cp
osit
ivel
yas
soci
ated
wit
hS
C
4:S
Cp
osit
ivel
yas
soci
ated
wit
hF
P
HC
isim
por
tan
tre
gar
dle
ssof
ind
ust
ry
typ
e;H
Ch
asa
gre
ater
infl
uen
ceon
how
ab
usi
nes
ssh
ould
be
stru
ctu
red
inn
onse
rvic
ein
du
stri
esco
mp
ared
to
serv
ice
ind
ust
ries
;CC
has
asi
gn
ific
ant
infl
uen
ceov
erS
Cir
resp
ecti
ve
of
ind
ust
ry;
the
dev
elop
men
tof
SC
has
a
pos
itiv
ere
lati
onsh
ipw
ith
bu
sin
ess
FP
reg
ard
less
ofin
du
stry
Tse
ng
and
Goo
(200
5)
500
larg
est
(in
term
s
ofsa
les
rev
enu
es)
Tai
wan
ese
man
ufa
ctu
rers
INC
–in
nov
atio
nca
pit
al
OC
–or
gan
izat
ion
alca
pit
al
RC
–re
lati
onsh
ipca
pit
al
HC
–h
um
anca
pit
al
V–
corp
orat
ev
alu
e:M
/B,
Tob
inQ
,an
dV
AIC
HC
,O
C,
RC
,IN
C–
dat
aob
tain
ed
from
qu
esti
onn
aire
s
1:IN
Cp
osit
ive
affe
cts
V
2:O
Cp
osit
ivel
yaf
fect
sV
3:R
Cp
osit
ivel
yaf
fect
sV
4:H
Cp
osit
ivel
yaf
fect
sIN
C
5:H
Cp
osit
ivel
yaf
fect
sO
C
6:H
Cp
osit
ivel
yaf
fect
sR
C
7:O
Cp
osit
ivel
yaf
fect
sIN
C
8:O
Cp
osit
ivel
yaf
fect
sR
C
9:IN
Cp
osit
ivel
yaf
fect
sR
C
HC
and
OC
ind
irec
tly
and
pos
itiv
ely
infl
uen
ceV
INC
and
RC
dir
ectl
yan
dp
osit
ivel
y
imp
act
V
HC
has
ah
igh
deg
ree
ofin
tera
ctio
n
wit
hot
her
thre
ety
pes
ofIC
OC
also
pos
itiv
ely
infl
uen
ces
onIN
C
and
RC
INC
pos
itiv
ely
affe
cts
RC
Eff
ect
ofIC
onen
han
cin
gV
inh
igh
-
tech
com
pan
ies
mor
eth
anin
non
-hig
h-
tech
com
pan
ies
Dıe
zet
al.
(201
0)21
1S
pan
ish
firm
s(2
007)
HC
–h
um
anca
pit
al
SC
–st
ruct
ura
lca
pit
al
V–
corp
orat
ev
alu
e:sa
les
gro
wth
,
pro
du
ctiv
ity,
and
retu
rnon
asse
ts
HC
,S
C–
dat
aob
tain
edfr
om
qu
esti
onn
aire
s
1:H
Cp
osit
ivel
yaf
fect
sV
2:S
Cp
osit
ivel
yaf
fect
sV
Th
est
ud
yco
nfi
rms
the
pos
itiv
e
rela
tion
bet
wee
nth
eu
seof
HC
and
SC
ind
icat
ors,
and
Vm
easu
red
by
sale
sg
row
th
Hig
her
lev
els
ofth
eV
AIC
,in
par
ticu
lar
for
the
com
pon
ent
that
refe
rsto
the
sum
ofth
eco
effi
cien
tof
HC
and
SC
,
are
also
rela
ted
toim
pro
vem
ents
inV
Pro
du
ctiv
ity
and
retu
rnon
asse
tsd
o
not
show
edsi
gn
ific
ant
resu
lts
(con
tinu
ed)
Table I.
638
JIC14,4
Stu
dy
Cou
ntr
yan
dp
erio
dD
epen
den
tv
aria
ble
sH
yp
oth
eses
Res
ult
s
Ch
eng
etal.
(201
0)
224
sam
ple
sfr
om56
UK
hea
lth
care
com
pan
ies
span
nin
gth
ep
erio
d
(200
2-20
05)
INN
–in
nov
ativ
eca
pac
ity
CU
S–
mai
nta
inab
lecu
stom
er
rela
tion
ship
cost
s
HU
M–
hu
man
val
ue
add
ed
PR
O–
effi
cien
top
erat
ing
pro
cess
es
PE
R–
corp
orat
ion
’sp
erfo
rman
ce
H1a
INN
has
ap
osit
ive
corr
elat
ion
wit
hC
US
H1b
INN
has
ap
osit
ive
corr
elat
ion
wit
hH
UM
H2
PR
Ore
du
ceC
US
H3
HU
Mis
pos
itiv
ely
corr
elat
edw
ith
CU
S
H4a
Th
eh
igh
erC
US
,th
eh
igh
erth
ep
osit
ive
effe
cton
the
PE
R
H4b
Th
eh
igh
erth
eH
UM
,th
eh
igh
erth
e
pos
itiv
eef
fect
onP
ER
Sig
nif
ican
tre
lati
onsh
ipb
etw
een
ICan
dco
mp
any
per
form
ance
.
Th
ese
resu
lts
also
sug
ges
tth
atIN
N
and
pro
cess
refo
rmat
ion
shal
lb
e
con
sid
ered
firs
t,an
dth
rou
gh
the
HU
Mof
hu
man
cap
ital
,fi
rms
can
imp
rov
eP
ER
Table I.
639
Intellectualcapital
investments
Considering all of the above, the following hypotheses are proposed:
H1. Investments in IC components are associated with each other.
H2. Investments in IC components positively affect company performance for aperiod of more than one year.
H3. Company performance positively affects investments in IC components.
Measurement is the principal issue in empirical studies on IC. Sveiby (2001) suggestsfour categories of measurement methods to IC: direct intellectual capital (DIC) methods,market capitalization methods, return on assets (ROA) methods, scorecard (Scorecard)methods. ROA methods are based on performance indicators such as economic valueadded or earnings. So, they can be used to assess the return on IC investments (outputmeasures), but are inappropriate for measuring investments in IC. DIC methods measurethe values of intellectual resources by identifying various components. Scorecard methodsimply the creation of a system of indices and indicators that can report company’s IC ina scorecard or graph. Following this classification DIC methods are relevant formeasuring input factors (investments in IC).
In empirical research three main approaches to measure IC components can bedistinguished. First, authors generally use questionnaires to estimate the amountof IC (Huang and Wu, 2010; Cabrita et al., 2007; Kamukama et al., 2010; Bontis et al.,2000; Tseng and Goo, 2005; Dıez et al., 2010). Second, components of VAIC model arefrequently used (Firer and Williams, 2003; Chen et al., 2004; Chan, 2009). Third, ICcomponents are approximated with proxy indicators (Huang and Liu, 2005; Chenget al., 2010; Shakina and Barajas, 2012). The latter approach is similar to the first one tosome extent: the measurement is based on evaluation of separated intellectualresources.
The output of investments in IC also can be measured in several ways. As for the ICcomponents, authors obtain indicators of company performance through interviewingemployees and top managers (Huang and Wu, 2010; Cabrita et al., 2007; Kamukamaet al., 2010; Bontis et al., 2000; Dıez et al., 2010). Also one of the common approachesin performance measurement is the using of financial performance indicators: ROA(Firer and Williams, 2003; Chen et al., 2004; Chan, 2009; Cheng et al., 2010), returnon equity (Cheng et al., 2010), and asset turnover (Firer and Williams, 2003; Chen et al.,2004; Chan, 2009).
Also it should be noted that the terms “intellectual capital” and “intangibles” are nottreated as synonymous in this paper. For the purposes of the current study the authorsuse the term “intangibles” for balance sheet item “Intangibles” only in order to avoidmisunderstanding.
3. DataIn order to evaluate the impact of IC on a company’s performance, it is necessary todetermine an indicator for each IC component. At the same time, since it is necessaryto take into account the delayed impact of investment in IC, indicators inherent indynamics are needed. The measure of human capital (HC) will be the number ofcompany employees (Edvinsson and Malone, 1997; Bajburina and Golovko, 2008;Garanina, 2009; Sullivan, 2000; Wang and Chang, 2005; Zickgraf et al., 2007).This indicator characterizes the number of “carriers” of this component of IC and it is
640
JIC14,4
known to external investors. On the other hand, this indicator has a disadvantagein that it ignores differences in employee knowledge that depend on their positionin the company.
The proxy indicator of a company’s structural capital (SC) can be represented by thebalance sheet value of its intangible assets (Shakina, 2011; Shakina and Barajas, 2012),since this is an estimate of the value of its patents and licenses. In this case, intangibleasset value fails to include many elements of structural capital, such as manuals,know-how, and corporate culture because the information about them is available onlyto internal stakeholders. However, in this paper it is assumed that the intensity of usefor these structural capital elements is proportional to the intangible assets value, asspecified in the balance sheet. In this case, the intangible assets value can be a proxyindicator for structural capital owned by the company.
In empirical studies relational capital (RC) is usually measured by an indicatorcalculated on the basis of a company’s revenue (Bassi and Van Buren, 1999; Brennanand Connell, 2000; Tsan, 2004; Chen, 2004; Marr and Adams, 2004) or advertisingexpenditures (Edvinsson and Malone, 1997; Tsan, 2004; Wu, 2004; Chen, 2004).However, these indicators characterize a company’s relationship only with customersand do not reflect the part of a company’s value created through relationships withsuppliers. A company’s contribution to the development of relational capital can beestimated as the excess of accounts receivable over accounts payable. The larger thisvalue is, the greater deferral a company offers to its clients and the quicker it pays forsupplier goods and services.
Also it is necessary to consider that tangible assets (TA) are required for acompany’s activity and may affect financial performance.
In order to make a cross-industry comparison, a group of major industries that playan important role in the US economy was selected. Each industry can be characterizedby some specific features that are related to the role of IC:
. The pharmaceutical industry is mainly characterized by significant structuralcapital value, represented by licenses and patents.
. Consulting and educational services were attributed to the service industry.Thus, the key resource for this industry is human capital and establishedrelationships with customers.
. The steel industry is generally classified as one of the traditional industries.Steel companies have minor amounts of IC.
. The retail industry is characterized by small amounts of tangible assets.The most important role here is the attraction of customers, which is viewed asrelational capital.
. The telecommunications industry, like the retail trade, is mostly focussed onthe attraction and retention of customers. Furthermore, the development ofcommunication facilities and customer databases requires investments instructural capital.
Table I contains information on the value of different components of IC relative totangible assets. Selected proxy indicators reflect the expected economic characteristicsof the various industries (Table II).
Empirical studies regarding the influence of IC on a company’s activity use differentmeasures of performance. Short-term indicators measure performance over one period,
641
Intellectualcapital
investments
such as ROA (Firer and Williams, 2003; Chen et al., 2004; Shiu, 2006; Ting and Lean,2009), return on equity (Chen et al., 2004), operating profit, or economic profit (EVA)(Huang and Wang, 2008). First, one of the traditional measures that could be deriveddirectly from a company’s reports – net operating profit (NOPAT). Second, EVA is usedin accordance with the value-based management concept, and it helps to take intoaccount not only explicit costs, but also opportunity costs. Moreover, it is believed thateconomic value added better reflects company value creation than operating profit(Stern et al., 2001). EVA is calculated as follows:
EVAi;t ¼ NOPATi; t �WACCi; t � ICi; t�1 ð1Þ
where ICi, t�1 is the sum of the book values of equity and long-term debt capital, andWACCi, t is the weighted average cost of capital:
WACCi; t ¼ kei; t �Ei; t
Vi; tþ kdi; t �
Di; t
Vi; t� ð1� taxi; tÞ ð2Þ
where Di, t is the book value of long-term debt; Ei, t is the book value of equity; taxi, t isthe effective tax rate of the company; kdi; t is the cost of debt determined based on thecompany’s synthetic credit rating as the sum of the risk-free rate and the default spread(found according to S&P ratings); and kei; t is the cost of equity:
kei; t ¼ RFt þ RPt � bi; t ð3Þ
RFt is the risk-free rate, calculated as the geometric mean of the yield of US treasurybills for 30 years. RPt is the risk premium associated with changes in the stock market.It is calculated as the difference between the geometric mean for the stock returns inthe US stock market for 40 years and the risk-free rate. bi, t is the b coefficientcharacterizing the company’s risks associated with whole economy risks (systematicrisks). The unlevered b coefficient ðbui; tÞis obtained from the Damodaran database andthen adjusted according the Hamada formula (Damodaran, 2004):
bi; t ¼ bui; t � 1þ Di; t
Ei; t� ð1� taxi; tÞ
� �ð4Þ
The source of the data is Compustat database. The sample contains information aboutAmerican companies working in pharmaceutical, retail, steel, telecommunications, andservice industries for 2001-2010. The choice of US companies is justified by the factthat the role of IC in the USA is a significant one, according to the KnowledgeEconomy Index of the World Bank and the National Intellectual Capital Model
Number of observations HC/TA SC/TA RC/TA
Pharmaceutical 1,035 0.09 10.47 �7.20Retail 1,080 0.06 1.82 �1.68Services 1,222 0.36 6.82 �6.85Steel industry 390 0.02 0.28 �0.50Telecommunications 1,727 0.04 3.33 �1.19
Table II.A comparison ofindustries on therelative size of intellectualcapital components
642
JIC14,4
(Lin and Edvinsson, 2008), which estimate the overall level of a country’s developmentand the efficiency of its use of knowledge.
4. Empirical methodologyIt is challenging to estimate the relationship between components of IC and acompany’s performance measures (NOPAT, EVA) because selected variables maysimultaneously affect each other. Thus, it is difficult to mark out the impact of oneselected factor over another, abstracting from the effects of other variables.
In the present research vector autoregression (VAR) is used for this purpose. Theuse of orthogonal response functions (impulse response function, hereinafter IRF) helpsto solve the problem described above. The reaction of one variable (e.g. NOPAT) to ashock in another variable (e.g. HC) can be evaluated, leaving all other variables ofstructural model constant and without considering their changes (shocks).
Panel VAR (panel vector autoregression, hereinafter pVAR) is estimated. It helps tocombine the advantages of VAR (all variables can be assumed as endogenous) and theadvantages of using panel data, allowing the taking into account of the individualheterogeneity of companies (Love and Zicchino, 2006).
pVAR of a second order is used:
zi; t ¼ aþ b1 � zi; t�1 þ b2 � zi; t�2 þ fi þ eit ð5Þ
where zi, t is a vector of four variables {NOPAT, hc, rc, sc}. Another specification wherezi, t¼ {EVA, hc, rc, sc} is used in order to consider the differences in the relationshipsbetween IC and accounting (NOPAT) and economic (EVA) indicators; fi is a company’sindividual features; hc is the increase in the number of employees, measured inthousands; rc is the increase of the difference between accounts receivable andaccounts payable, measured in millions of US dollars; sc is the increase in the value ofintangible assets, also measured in millions of US dollars; and NOPAT is the netoperating profit, calculated on the basis of a company’s effective tax rate. NOPAT andEVA are measured in millions of US dollars.
To analyze the impact of one variable shock to another variable an IRF is applied.However, since the empirical variance-covariance matrix is diagonal, in order to isolateshocks it is necessary to decompose residuals in the model so that they will beorthogonal (this procedure is known as Choleski decomposition) (Love and Zicchino,2006). Building an IRF is crucial for analysis, it permits to focus on how a shockedvariable (e.g. human capital) impacts another variable (e.g. EVA) keeping other shocks(e.g. capital expenditures, structural and relational capital and their lags) constant.
Using VAR for panel data, it is necessary to set a limit on the data structure: therandom structure of cross-sectional data is generally assumed. This assumption is notconsistent with the empirical data: companies are heterogeneous; each has its ownfeatures. Thus, for a correct analysis, it is necessary to allow for individualheterogeneity within a cross-section, i.e. to use a statistic model with “fixed effects”(fixed effect model).
Since fixed effects are correlated with dependent variables (the problem arisesbecause of the usage of lags), mean differencing cannot be used. To avoid this problemof endogeneity, the Helmert procedure is used. This procedure removes the averagevalue, calculated on future values, and keeps transformed and lagged regressorsorthogonal. Thus, they can be used as instrument variables in evaluating the resultingsystem of equations using the method of moments. In this case, the model is not
643
Intellectualcapital
investments
super-identified, so using the method of moments is equivalent to using the two-stepOLS. To analyze IRF, it is necessary to assess confidence intervals. Since the IRFmatrix is based on estimated pVAR coefficients, it is necessary to take standarderrors into account. A Monte-Carlo simulation is used in order to generate confidenceintervals, since the distribution of standard errors is not parameterized: a bootstrapapproach is used to approximate the distribution empirically.
5. ResultsThere is no empirical evidence that investments in IC contribute to the performanceof pharmaceutical companies (Figure 1). However, some interesting findings can bederived. First, an increase of capital expenditures leads to an increase of human capitalor a reduction of structural capital. Consequently, structural capital and tangible assetscould be substituted one for another, to some extent. In this case, no positive returnon investment in tangible assets is observed. Second, the relationship betweencapital investments and operating profit could be explained by the fact that capitalinvestments respond to shocks in operating profit. Thus, in the pharmaceuticalindustry, an investment strategy whereby companies consistently invest a certainamount of operating profit into tangible assets can be noted. Also, investments in ICaffect investments in tangible assets.
However, the graph shows a negative relation between economic profit and increasein human capital (Figure 2). Moreover, this response to current shock may become evenstronger over the following several years. The negative relationship with capitalinvestments could be explained by the fact that investments are rising, but there is noreturn on these investments during the analyzed period. Also capital expendituresdistract from a company’s economic profit or their contribution is close to zero.As far as they do not have a sustainable effect on operating profit, it can be resumedthat pharmaceutical companies invest in tangibles without consideration of the costof capital.
Thus, it can be assumed that, in the pharmaceutical industry, investments intangible and intellectual assets (human and structural capital) have a return onlywithin a period longer than four years. However, consideration of a model with longertime intervals requires consideration of more factors.
In the retail industry, no direct influence of IC on financial performance wasobserved (Figures 3 and 4). However, investments in structural capital frequently leadto a decrease in economic value added. Additionally human capital grows with anincrease in operating profit (or economic profit). So, retail companies try to hire newemployees when their profits go up, but graphs show that this policy do not leads to arise in profits. Also, as in the pharmaceutical industry, the capital expenditures isinfluenced by the shocks in structural and human capital and shocks in operatingprofit. Investments in structural capital and tangible assets are interrelated positively.In the retail industry they are probably complementary resources for a company.
For the telecommunications industry, a positive impact of investments in the humanand structural components of IC and investments in tangible assets on economic valueadded is typical (Figure 5). At the same time the value of NOPAT has no linearrelationship to IC (Figure 6). Also, it should be noted that there is a complexrelationship between investments in IC and investments in tangible assets.
Consulting and educational services have no clear linear relationship betweeninvestments in tangible and intellectual assets and a company’s financial results(Figures 7 and 8). In this case, making a profit does not increase investments. It can be
644
JIC14,4
Res
pons
e of
NO
PA
T to
NO
PA
T s
hock
s
(p 2
0) N
OP
AT
NO
PA
T
(p 8
0) N
OP
AT
(p 2
0) N
OP
AT
NO
PA
T
(p 8
0) N
OP
AT
(p 2
0) N
OP
AT
NO
PA
T
(p 8
0) N
OP
AT
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
(p 2
0) N
OP
AT
NO
PA
T
(p 8
0) N
OP
AT
06
6
0.00
00
568.
4094
Res
pons
e of
NO
PA
T to
hc
shoc
ks
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
06
–1.5
e+02
95.1
923
Res
pons
e of
NO
PA
T to
rc
shoc
ks
(p 2
0) r
crc rc
(p 8
0) r
c
(p 2
0) r
c(p
80)
rc
rc(p
20)
rc
(p 8
0) r
c
rc(p
20)
rc
(p 8
0) r
c
rc(p
20)
rc
(p 8
0) r
c
06
–8.3
429
7.41
85
Res
pons
e of
NO
PA
T to
sc
shoc
ks
(p 2
0) s
csc sc
(p 8
0) s
c
(p 2
0) s
c(p
80)
sc
sc(p
20)
sc
(p 8
0) s
c
sc(p
20)
sc
(p 8
0) s
c
sc(p
20)
sc
(p 8
0) s
c
0
–1.5
e+02
132.
9741
Res
pons
e of
NO
PA
T to
CA
PE
X s
hock
s
(p 2
0) C
AP
EX
CA
PE
X
CA
PE
X
(p 8
0) C
AP
EX
(p 2
0) C
AP
EX
(p 8
0) C
AP
EX
CA
PE
X(p
20)
CA
PE
X(p
80)
CA
PE
X
CA
PE
X(p
20)
CA
PE
X(p
80)
CA
PE
X
CA
PE
X(p
20)
CA
PE
X(p
80)
CA
PE
X
06
–2.1
e+02
331.
9985
Res
pons
e of
hc
to N
OP
AT
sho
ck
s0
66
–0.6
971
1.45
41
Res
pons
e of
hc
to h
c sh
ock
s0
6
–1.1
291
3.67
49
Res
pons
e of
hc
to r
c sh
ock
s0
6
–0.0
535
0.05
10
Res
pons
e of
hc
to s
c sh
ock
s0
–1.1
408
0.59
84
Res
pons
e of
hc
to C
AP
EX
sho
ck
s0
6–1
.243
4
1.27
39
Res
pons
e of
rc
to N
OP
AT
sho
ck
s0
6
–0.2
090
0.18
33
Res
pons
e of
rc
to h
c sh
ock
s0
6
–0.1
992
0.18
76
Res
pons
e of
rc
to r
c sh
ock
s0
6
–2.5
837
4.44
42
Res
pons
e of
rc
to s
c sh
ock
s0
6
–0.0
105
0.01
53
Res
pons
e of
rc
to C
AP
EX
sho
ck
s0
6
–0.0
141
0.00
73
Res
pons
e of
sc
to N
OP
AT
sho
cks
06
–1.6
e+03
1.5e
+03
Res
pons
e of
sc
to h
c sh
ock
s0
6
–2.1
e+03
4.4e
+03
Res
pons
e of
sc
to r
c sh
ock
s0
6
–1.5
e+02
133.
5779
Res
pons
e of
sc
to s
c sh
ock
s
sc
0
–1.8
e+03
3.3e
+03
Res
pons
e of
sc
to C
AP
EX
sho
cks
06
s0
6s
06
s0
6s
0s
06
–2.6
e+03
1.5e
+03
Res
pons
e of
CA
PE
X to
NO
PA
T s
hock
–12.
0074
190.
9449
Res
pons
e of
CA
PE
X to
hc
shoc
k
–10.
6022
134.
1533
Res
pons
e of
CA
PE
X to
rc
shoc
k
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
–5.8
329
4.84
07
Res
pons
e of
CA
PE
X to
sc
shoc
k
sc
–1.7
e+02
21.9
014
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
–2.3
080
396.
8294
Figure 1.Impulse responses for2 lag VAR of NOPAT,
human, relational,structural capital, and
tangible assets forpharmaceutical companies
645
Intellectualcapital
investments
Res
pons
e of
EV
A to
EV
A s
hock
s
(p 2
0) E
VA
EV
A(p
80)
EV
A
06
–4.0
e+02
754.
7198
Res
pons
e of
EV
A to
hc
shoc
ks
(p 2
0) h
chc
(p 8
0) h
c
06
–1.1
e+03
31.7
592
Res
pons
e of
EV
A to
sc
shoc
ks
(p 2
0) s
csc
(p 8
0) s
c
06
–2.0
e+02
140.
5886
Res
pons
e of
EV
A to
rc
shoc
ks
(p 2
0) r
crc
(p 8
0) r
c
06
–19.
7555
19.5
425
Res
pons
e of
EV
A to
CA
PE
X s
hock
s
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
–1.2
e+03
20.7
744
Res
pons
e of
hc
to E
VA
sho
cks
(p 2
0) E
VA
EV
A(p
80)
EV
A
06
–0.8
589
1.66
71
Res
pons
e of
hc
to h
c sh
ock
s
(p 2
0) h
chc
(p 8
0) h
c
06
–1.3
983
5.56
08
Res
pons
e of
hc
to s
c sh
ock
s
(p 2
0) s
csc
(p 8
0) s
c
06
–1.0
648
0.45
60
Res
pons
e of
hc
to r
c sh
ock
s
(p 2
0) r
crc
(p 8
0) r
c
06
–0.0
484
0.04
09
Res
pons
e of
hc
to C
AP
EX
sho
cks
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
–0.9
022
2.04
98
Res
pons
e of
sc
to E
VA
sho
ck
s
(p 2
0) E
VA
EV
A(p
80)
EV
A
06
–1.7
e+03
1.8e
+03
Res
pons
e of
sc
to h
c sh
ock
s
(p 2
0) h
chc
(p 8
0) h
c
06
–2.0
e+03
5.0e
+03
Res
pons
e of
sc
to s
c sh
ock
s
(p 2
0) s
csc
(p 8
0) s
c
06
–1.6
e+03
3.4e
+03
Res
pons
e of
sc
to r
c sh
ock
s
(p 2
0) r
crc
(p 8
0) r
c
06
–51.
0824
41.3
645
Res
pons
e of
sc
to C
AP
EX
sho
ck
s
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
–9.5
e+02
2.5e
+03
Res
pons
e of
rc
to E
VA
sho
ck
s
(p 2
0) E
VA
EV
A(p
80)
EV
A
06
–0.2
634
0.29
46
Res
pons
e of
rc
to h
c sh
ock
s
(p 2
0) h
chc
(p 8
0) h
c
06
–0.3
213
0.27
68
Res
pons
e of
rc
to s
c sh
ock
s
(p 2
0) s
csc
(p 8
0) s
c
06
–0.2
846
0.33
16
Res
pons
e of
rc
to r
c sh
ock
s
(p 2
0) r
crc
(p 8
0) r
c
06
–5.4
059
7.96
26
Res
pons
e of
rc
to C
AP
EX
sho
ck
s
ss
ss
s
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
–0.0
173
0.01
24
Res
pons
e of
CA
PE
X to
EV
A s
hock
06
–19.
9955
49.4
271
Res
pons
e of
CA
PE
X to
hc
shoc
k
Not
e: E
rror
s are
5%
on
each
side
gen
erat
ed b
y M
onte
-Car
lo w
ith 5
00 re
ps
06
–9.5
470
115.
6343
Res
pons
e of
CA
PE
X to
sc
shoc
k0
6
–9.2
210
38.7
715
Res
pons
e of
CA
PE
X to
rc
shoc
k0
6
–4.0
880
4.39
33
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
06
–0.1
958
179.
3978
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) h
chc
(p 8
0) h
c(p
20)
sc
sc(p
80)
sc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
Figure 2.Impulse responses for2 lag VAR of EVA, human,relational, structuralcapital, and tangibleassets for pharmaceuticalcompanies
646
JIC14,4
Res
pons
e of
NO
PA
T to
NO
PA
T s
hock
s
(p 2
0) N
OP
AT
NO
PA
T
NO
PA
T
NO
PA
T
(p 8
0) N
OP
AT
06
–6.2
588
219.
3983
Res
pons
e of
NO
PA
T to
hc
shoc
k
s
(p 2
0) h
chc
(p 8
0) h
c
06
–15.
3322
53.4
725
Res
pons
e of
NO
PA
T to
rc
shoc
k
s
(p 2
0) r
crc
(p 8
0) r
c
06
–2.9
140
2.41
18
Res
pons
e of
NO
PA
T to
sc
shoc
k
s
(p 2
0) s
csc
(p 8
0) s
c
06
–19.
7271
86.8
934
Res
pons
e of
NO
PA
T to
CA
PE
X s
hock
s
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
–47.
6632
44.1
928
Res
pons
e of
hc
to N
OP
AT
sho
cks
06
–0.2
540
2.00
48
Res
pons
e of
hc
to h
c sh
ock
s0
6–0
.466
0
8.21
04
Res
pons
e of
hc
to r
c sh
ock
s0
6–0
.071
4
0.04
23
Res
pons
e of
hc
to s
c sh
ock
s0
6–1
.020
0
0.79
78
Res
pons
e of
hc
to C
AP
EX
sho
cks
06
–0.8
771
1.27
77
Res
pons
e of
rc
to N
OP
AT
sho
cks
06
–0.0
058
0.00
42
Res
pons
e of
rc
to h
c sh
ock
s0
6–0
.003
9
0.00
43
Res
pons
e of
rc
to r
c sh
ock
s0
6–0
.128
1
0.11
24
Res
pons
e of
rc
to s
c sh
ock
s0
6–0
.000
8
0.00
06
Res
pons
e of
rc
to C
AP
EX
sho
cks
06
–0.0
012
0.00
09
Res
pons
e of
sc
to N
OP
AT
sho
cks
06
–58.
6737
131.
8430
Res
pons
e of
sc
to h
c sh
ock
s0
6–7
2.52
16
312.
9201
Res
pons
e of
sc
to r
c sh
ock
s0
6–1
2.40
47
13.0
871
Res
pons
e of
sc
to s
c sh
ock
s0
6–1
.0e+
02
361.
7500
Res
pons
e of
sc
to C
AP
EX
sho
cks
06
–26.
1536
114.
9918
Res
pons
e of
CA
PE
X to
NO
PA
T s
hock
s0
6–1
.509
9
120.
6269
Res
pons
e of
CA
PE
X to
hc
shoc
k
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
s0
60.
0000
63.3
014
Res
pons
e of
CA
PE
X to
rc
shoc
ks
06
–5.5
197
2.79
18
Res
pons
e of
CA
PE
X to
sc
shoc
ks
06
–3.5
249
54.6
775
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
s0
6–1
4.63
10
123.
5674
(p 2
0) N
OP
AT
(p 8
0) N
OP
AT
(p 2
0) h
chc
(p 8
0) h
c(p
20)
rc
rc(p
80)
rc
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) N
OP
AT
(p 8
0) N
OP
AT
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) r
crc
(p 8
0) r
c
(p 2
0) r
crc
(p 8
0) r
c
(p 2
0) r
crc
(p 8
0) r
c
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Figure 3.Impulse responses for2 lag VAR of NOPAT,
human, relational,structural capital, and
tangible assets for retailcompanies
647
Intellectualcapital
investments
Res
pons
e of
EV
A to
EV
A s
hock
s
(p 2
0) E
VA
EV
A(p
80)
EV
A
06
–2.4
642
183.
3500
Res
pons
e of
EV
A to
hc
shoc
ks
(p 2
0) h
chc
(p 8
0) h
c
06
–33.
5667
30.5
799
Res
pons
e of
EV
A to
sc
shoc
ks
(p 2
0) s
csc
(p 8
0) s
c
06
–54.
4724
4.22
04
Res
pons
e of
EV
A to
rc
shoc
ks
(p 2
0) r
crc
(p 8
0) r
c
06
–0.9
900
1.26
20
Res
pons
e of
EV
A to
CA
PE
X s
hock
s
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
–40.
4744
44.6
731
–0.1
146
2.88
83
Res
pons
e of
hc
to h
c sh
ock
s0
6–0
.967
1
7.68
61
Res
pons
e of
hc
to s
c sh
ock
s0
6–1
.497
7
0.23
02
Res
pons
e of
hc
to r
c sh
ock
s0
6–0
.068
1
0.03
97
Res
pons
e of
hc
to C
AP
EX
sho
cks
06
–1.0
620
1.28
54
Res
pons
e of
sc
to E
VA
sho
cks
06
–1.1
e+02
206.
8695
Res
pons
e of
sc
to h
c sh
ock
s0
6–6
5.07
90
283.
9501
Res
pons
e of
sc
to s
c sh
ock
s0
6–6
7.02
60
362.
8218
Res
pons
e of
sc
to r
c sh
ock
s0
6–4
.364
5
2.16
15
Res
pons
e of
sc
to C
AP
EX
sho
cks
06
–23.
9217
132.
7474
Res
pons
e of
rc
to E
VA
sho
cks
06
–0.0
047
0.00
43
Res
pons
e of
rc
to h
c sh
ock
s0
6–0
.003
7
0.00
44
Res
pons
e of
rc
to s
c sh
ock
s0
6–0
.003
7
0.00
39
Res
pons
e of
rc
to r
c sh
ock
s0
6–0
.125
3
0.11
34
–0.0
014
0.00
08
Res
pons
e of
CA
PE
X to
EV
A s
hock
s0
60.
0000
85.6
449
Res
pons
e of
CA
PE
X to
hc
shoc
ks
06
–2.0
065
54.5
671
Res
pons
e of
CA
PE
X to
sc
shoc
ks
06
–18.
7166
43.8
465
Res
pons
e of
CA
PE
X to
rc
shoc
ks
06
–6.6
456
2.15
70
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
s0
6–2
.588
6
129.
4921
Res
pons
e of
EV
A to
EV
A s
hock
s0
6
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) r
crc
(p 8
0) r
c
(p 2
0) r
crc
(p 8
0) r
c
(p 2
0) r
crc
(p 8
0) r
c
(p 2
0) r
crc
(p 8
0) r
c
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
s0
6
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
Figure 4.Impulse responses for 2lag VAR of EVA, human,relational, structuralcapital, and tangibleassets for retail companies
648
JIC14,4
Res
pons
e of
NO
PA
T to
NO
PA
T s
hock
s
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
06
–2.3
e+02
1.3e
+03
Res
pons
e of
NO
PA
T to
CA
PE
X s
hock
s
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
–2.0
e+02
159.
3653
Res
pons
e of
NO
PA
T to
rc
shoc
k
s
(p 2
0) r
crc
(p 8
0) r
c
06
–33.
7301
11.2
696
Res
pons
e of
NO
PA
T to
sc
shoc
k
s
(p 2
0) s
csc
(p 8
0) s
c
06
–2.7
e+02
99.0
638
Res
pons
e of
NO
PA
T to
hc
shoc
k
s
(p 2
0) h
chc
(p 8
0) h
c
06
–1.0
e+02
53.5
631
Res
pons
e of
CA
PE
X to
NO
PA
T s
hock
s0
60.
0000
150.
0618
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
s0
60.
0000
437.
7981
Res
pons
e of
CA
PE
X to
rc
shoc
k
s0
6–1
1.30
97
14.0
537
Res
pons
e of
CA
PE
X to
sc
shoc
k
s0
60.
0000
262.
3232
Res
pons
e of
CA
PE
X to
hc
shoc
k
s0
6–4
.029
0
89.9
224
–0.0
189
0.02
82
–0.0
316
0.01
21
–0.0
767
0.80
02
–0.0
059
0.00
13
-0.0
142
0.00
24
–3.8
e+02
777.
6888
–1.2
e+03
361.
9001
–2.9
e+02
165.
8400
–8.8
e+02
4.6e
+03
–3.1
e+02
316.
2474
–0.4
919
0.38
76
–1.3
350
1.75
38
–0.3
007
0.18
61
–0.5
283
2.88
79
–2.4
853
7.95
30
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
Res
pons
e of
rc
to N
OP
AT
sho
ck
s0
6
Res
pons
e of
rc
to C
AP
EX
sho
ck
s0
6
Res
pons
e of
rc
to r
c sh
ock
s0
6
Res
pons
e of
rc
to s
c sh
ock
s0
6
Res
pons
e of
rc
to h
c sh
ock
s0
6
Res
pons
e of
sc
to N
OP
AT
sho
cks
06
Res
pons
e of
sc
to C
AP
EX
sho
cks
06
Res
pons
e of
sc
to r
c sh
ock
s0
6
Res
pons
e of
sc
to s
c sh
ock
s0
6
Res
pons
e of
sc
to h
c sh
ock
s0
6
Res
pons
e of
hc
to N
OP
AT
sho
cks
06
Res
pons
e of
hc
to C
AP
EX
sho
cks
06
Res
pons
e of
hc
to r
c sh
ock
s0
6
Res
pons
e of
hc
to s
c sh
ock
s0
6
Res
pons
e of
hc
to h
c sh
ock
s0
6
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) h
chc
(p 8
0) h
c
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) h
c
hc(p
80)
hc
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) h
chc
(p 8
0) h
c
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
Figure 5.Impulse responses for2 lag VAR of NOPAT,
human, relational,structural capital, and
tangible assets fortelecommunications
companies
649
Intellectualcapital
investments
Res
pons
e of
EV
A to
EV
A s
hock
s
(p 2
0) E
VA
EV
A(p
80)
EV
A
06
0.00
00
6.5e
+03
Res
pons
e of
EV
A to
hc
shoc
ks
(p 2
0) h
chc
(p 8
0) h
c
06
0.00
00
489.
0822
Res
pons
e of
EV
A to
sc
shoc
ks
(p 2
0) s
csc
(p 8
0) s
c
06
–33.
2391
344.
1633
Res
pons
e of
EV
A to
rc
shoc
ks
(p 2
0) r
crc
(p 8
0) r
c
06
–70.
1196
120.
2645
Res
pons
e of
EV
A to
CA
PE
X s
hock
s
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
06
0.00
00
710.
6732
–0.1
749
0.88
09
–2.0
828
8.48
25
–0.7
013
1.60
20
–0.1
718
0.06
37
–1.2
470
1.38
41
–1.2
e+02
485.
6896
–6.1
e+02
1.6e
+03
–7.9
e+02
4.4e
+03
–2.6
e+02
41.2
893
–9.8
e+02
100.
9016
–0.0
225
0.02
10
–0.0
319
0.01
15
–0.0
118
0.03
18
–0.0
759
0.77
60
–0.0
158
0.00
11
0.00
00
55.9
916
0.00
00
246.
4486
–10.
2396
255.
4200
–15.
5555
12.0
900
0.00
00
410.
3901
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) r
crc
(p 8
0) r
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
Res
pons
e of
hc
to E
VA
sho
cks
06
Res
pons
e of
hc
to h
c sh
ock
s0
6
Res
pons
e of
hc
to s
c sh
ock
s0
6
Res
pons
e of
hc
to r
c sh
ock
s0
6
Res
pons
e of
hc
to C
AP
EX
sho
cks
06
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) r
crc
(p 8
0) r
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
Res
pons
e of
sc
to E
VA
sho
cks
06
Res
pons
e of
sc
to h
c sh
ock
s0
6
Res
pons
e of
sc
to s
c sh
ock
s0
6
Res
pons
e of
sc
to r
c sh
ock
s0
6
Res
pons
e of
sc
to C
AP
EX
sho
cks
06
Res
pons
e of
rc
to E
VA
sho
cks
06
Res
pons
e of
rc
to h
c sh
ock
s0
6
Res
pons
e of
rc
to s
c sh
ock
s0
6
Res
pons
e of
rc
to r
c sh
ock
s0
6
Res
pons
e of
rc
to C
AP
EX
sho
cks
06
Res
pons
e of
CA
PE
X to
EV
A s
hock
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
s0
6
Res
pons
e of
CA
PE
X to
hc
shoc
ks
06
Res
pons
e of
CA
PE
X to
sc
shoc
ks
06
Res
pons
e of
CA
PE
X to
rc
shoc
ks
06
Res
pons
e of
CA
PE
X to
CA
PE
X s
hoc k
s0
6
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) r
crc
(p 8
0) r
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) r
crc
(p 8
0) r
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
Figure 6.Impulse responses for2 lag VAR of EVA, human,relational, structuralcapital, and tangibleassets fortelecommunicationscompanies
650
JIC14,4
Res
pons
e of
NO
PA
T to
NO
PA
T s
hock
s
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T
06
0.00
00
89.9
270
(p 2
0) h
chc
(p 8
0) h
c
–15.
3816
23.7
387
(p 2
0) r
crc
(p 8
0) r
c
–2.8
575
0.77
84
(p 2
0) s
csc
(p 8
0) s
c
–14.
9298
11.7
771
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
–45.
3609
8.93
79
–0.3
382
0.56
01
–0.5
632
3.02
75
–0.0
137
0.08
96
–0.0
669
0.83
16
–0.2
614
1.21
56
–0.4
081
0.36
39
–0.9
017
0.15
29
–1.0
829
5.97
57
–0.2
901
0.11
54
–0.8
004
0.13
15
–59.
9490
5.99
51
–1.2
e+02
52.2
835
–3.9
464
16.1
596
–20.
1231
283.
7227
–24.
4588
94.4
399
–1.5
281
4.36
81
–0.5
133
10.0
763
–0.1
328
1.81
08
–3.7
970
2.35
13
–0.9
731
18.9
780
Res
pons
e of
NO
PA
T to
hc
shoc
k
s0
6
Res
pons
e of
NO
PA
T to
rc
shoc
k
s0
6
Res
pons
e of
NO
PA
T to
sc
shoc
k
s0
6
Res
pons
e of
NO
PA
T to
CA
PE
X s
hock
s0
6
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
hc
to N
OP
AT
sho
ck
s0
6
Res
pons
e of
hc
to h
c sh
ock
s0
6
Res
pons
e of
hc
to r
c sh
ock
s0
6
Res
pons
e of
hc
to s
c sh
ock
s0
6
Res
pons
e of
hc
to C
AP
EX
sho
ck
s0
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
rc
to N
OP
AT
sho
ck
s0
6
Res
pons
e of
rc
to h
c sh
ock
s0
6
Res
pons
e of
rc
to r
c sh
ock
s0
6
Res
pons
e of
rc
to s
c sh
ock
s0
6
Res
pons
e of
rc
to C
AP
EX
sho
ck
s0
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
sc
to N
OP
AT
sho
ck
s0
6
Res
pons
e of
sc
to h
c sh
ock
s0
6
Res
pons
e of
sc
to r
c sh
ock
s0
6
Res
pons
e of
sc
to s
c sh
ock
s0
6
Res
pons
e of
sc
to C
AP
EX
sho
ck
s0
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
CA
PE
X to
NO
PA
T s
hock
s0
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
6
Res
pons
e of
CA
PE
X to
hc
shoc
ks
06
Res
pons
e of
CA
PE
X to
rc
shoc
ks
06
Res
pons
e of
CA
PE
X to
sc
shoc
ks
06
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
s0
6 66 6
Figure 7.Impulse responses for2 lag VAR of NOPAT,
human, relational,structural capital, and
tangible assets for servicecompanies
651
Intellectualcapital
investments
–4.1
e+03
2.7e
+04
–2.5
e+04
5.1e
+03
–8.3
e+03
6.9e
+03
–1.0
e+04
2.3e
+04
–2.1
e+04
7.2e
+03
–1.5
e+03
123.
5560
–92.
8804
3.0e
+03
–6.7
e+02
73.9
354
–2.8
e+03
84.9
502
–88.
9072
2.5e
+03
–5.8
e+05
5.5e
+04
–4.2
e+04
1.3e
+06
–2.5
e+05
5.5e
+04
–1.3
e+06
5.2e
+04
–3.9
e+04
9.5e
+05
–7.1
e+02
5.9e
+03
–1.2
e+04
461.
6634
–5.0
e+02
2.3e
+03
–5.2
e+02
1.2e
+04
–7.9
e+03
434.
1233
–1.7
e+03
9.1e
+03
–1.1
e+04
2.2e
+03
–3.2
e+03
3.4e
+03
–2.8
e+03
1.2e
+04
–8.4
e+03
3.0e
+03
(p 2
0) h
chc
(p 8
0) h
c(p
20)
rc
rc(p
80)
rc
(p 2
0) s
csc
(p 8
0) s
c
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
s0
6s
06
s0
6s
06
s0
6
(p 2
0) h
chc
(p 8
0) h
c(p
20)
rc
rc(p
80)
rc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
s0
6s
06
s0
6s
06
s0
6
(p 2
0) h
chc
(p 8
0) h
c(p
20)
rc
rc(p
80)
rc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
s0
6s
06
s0
6s
06
s0
6
(p 2
0) h
chc
(p 8
0) h
c(p
20)
rc
rc(p
80)
rc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
s0
6s
06
s0
6s
06
s0
6
(p 2
0) h
chc
(p 8
0) h
c(p
20)
rc
rc(p
80)
rc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
CA
PE
XC
AP
EX
(p 8
0) C
AP
EX
s0
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
6s
06
s0
6s
06
s0
6
Res
pons
e of
EV
A to
EV
A s
hock
Res
pons
e of
EV
A to
hc
shoc
kR
espo
nse
of E
VA
to s
c sh
ock
Res
pons
e of
EV
A to
rc
shoc
kR
espo
nse
of E
VA
to C
AP
EX
sho
ck
Res
pons
e of
hc
to E
VA
sho
ckR
espo
nse
of h
c to
hc
shoc
kR
espo
nse
of h
c to
sc
shoc
kR
espo
nse
of h
c to
rc
shoc
kR
espo
nse
of h
c to
CA
PE
X s
hock
Res
pons
e of
sc
to E
VA
sho
ckR
espo
nse
of s
c to
hc
shoc
kR
espo
nse
of s
c to
sc
shoc
kR
espo
nse
of s
c to
rc
shoc
kR
espo
nse
of s
c to
CA
PE
X s
hock
Res
pons
e of
rc
to E
VA
sho
ckR
espo
nse
of r
c to
hc
shoc
kR
espo
nse
of r
c to
sc
shoc
kR
espo
nse
of r
c to
rc
shoc
kR
espo
nse
of r
c to
CA
PE
X s
hock
Res
pons
e of
CA
PE
X to
EV
A s
hock
Res
pons
e of
CA
PE
X to
hc
shoc
kR
espo
nse
of C
AP
EX
to s
c sh
ock
Res
pons
e of
CA
PE
X to
rc
shoc
kR
espo
nse
of C
AP
EX
to C
AP
EX
sho
ck
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) E
VA
EV
A(p
80)
EV
A
(p 2
0) E
VA
EV
A(p
80)
EV
AFigure 8.Impulse responses for2 lag VAR of EVA,human, relational,structural capital, andtangible assets forservice companies
652
JIC14,4
assumed that the relationships between assets and a company’s activity in the processof value creation are more complex than simple linear ones.
In the steel industry, the key factors of value increase are human capital andtangible assets (Figures 9 and 10). It should be noted that, in the steel industry,investments have a weak impact on reported financial performance. Also, in the steelindustry, capital investments depend only on the value of capital investments inprevious periods. This justifies the assumption that IC does not play a significant rolein the steel industry, which is commonly regarded as being a traditional one.
6. ConclusionIn this paper, the authors analyzed the mutual influence of investments on differenttypes of IC and company performance using pVAR. The results suggest that there is arelationship between investments in intellectual assets and financial performance.There is also a reverse relationship. The significance of shocks in the impact offinancial performance on investments in intellectual assets indicates the need to takethis into account in future research on the contribution of IC to company performance.
Insofar as growth in IC is regarded as an investment, it should contribute tocompany performance for a long period of time (more than one year). The resultsevidence that such investments can give a positive return over the course of six years.However, return on IC is frequently close to zero, or fluctuates with changing sign, oreven is negative for a long period. Also, empirical research shows that managementtakes into consideration company performance when making investment decisions oncapital expenditures, but not on IC. Moreover, there is a significant difference betweenthe influence of investments on operating profit and on economic profit. Consequentlythe cost of capital should be taken into account. Contrary to the theoretical assumptionthat economic profit can not be earned from investments in tangible assets, researchresults show evidence that the influence of capital expenditures on EVA is significant.Furthermore the dependence between investments in IC components supports the ideathat they could substitute or complement each other.
In addition, the authors identified a number of sectoral peculiarities regarding therelationship between IC and company performance:
. In industries with long production cycles, such as pharmaceuticals, investmentsin human and tangible resources could have a positive return after a longperiod of time, such as five years and more. For short-term investment returns, itis advisable to select companies with less current investments in tangible assetsand human capital.
. In the retail industry, major investments in structural capital often do not lead tocorresponding increases in operating profit, meaning that economic value addeddecreases.
. In the telecommunications industry, investments in human and structuralcapital and tangible assets contribute to the creation of economic value added. Inthis case, the most important roles are played by investments in human capitaland tangible resources of the previous year, and, as for structural capital,investments from the two previous years. Relational capital has no direct effecton company performance. Consequently, the efficiency of this type of capital isdetermined by the efficiency of other resources. Thus, telecommunicationcompanies with large investments in human, structural, and tangible assetsshould be treated as preferred objects for investing.
653
Intellectualcapital
investments
–7.9
e+02
1.0e
+03
–1.8
e+02
397.
3874
–70.
3967
85.2
013
–8.1
e+02
562.
1942
–3.1
e+02
499.
0095
–3.4
436
0.88
49
–0.2
940
10.3
069
–0.0
478
0.50
90
–2.2
578
0.80
77
–2.0
062
0.93
42
–0.0
101
0.00
41
–0.0
052
0.00
95
–0.0
386
0.13
03
–0.0
034
0.00
38
–0.0
031
0.00
29
–4.5
e+02
716.
8349
–91.
3227
617.
2422
–1.2
e+02
76.5
498
–4.9
e+02
1.2e
+03
–2.4
e+02
612.
8962
–1.3
e+02
259.
3378
–73.
2961
147.
9819
–28.
5783
19.0
132
–2.3
e+02
104.
3617
–18.
2859
241.
7079
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
NO
PA
T to
NO
PA
T s
hock
s0
6
Res
pons
e of
NO
PA
T to
hc
shoc
ks
06
Res
pons
e of
NO
PA
T to
rc
shoc
ks
06
Res
pons
e of
NO
PA
T to
sc
shoc
ks
06
Res
pons
e of
NO
PA
T to
CA
PE
X s
hoc k
s0
6
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
hc
to N
OP
AT
sho
cks
06
Res
pons
e of
hc
to h
c sh
ock
s0
6
Res
pons
e of
hc
to r
c sh
ock
s0
6
Res
pons
e of
hc
to s
c sh
ock
s0
6
Res
pons
e of
hc
to C
AP
EX
sho
cks
06
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
rc
to N
OP
AT
sho
cks
06
Res
pons
e of
rc
to h
c sh
ock
s0
6
Res
pons
e of
rc
to r
c sh
ock
s0
6
Res
pons
e of
rc
to s
c sh
ock
s0
6
Res
pons
e of
rc
to C
AP
EX
sho
cks
06
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
sc
to N
OP
AT
sho
cks
06
Res
pons
e of
sc
to h
c sh
ock
s0
6
Res
pons
e of
sc
to r
c sh
ock
s0
6
Res
pons
e of
sc
to s
c sh
ock
s0
6
Res
pons
e of
sc
to C
AP
EX
sho
cks
06
(p 2
0) N
OP
AT
NO
PA
T(p
80)
NO
PA
T(p
20)
hc
hc(p
80)
hc
(p 2
0) r
crc
(p 8
0) r
c(p
20)
sc
sc(p
80)
sc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
CA
PE
X to
NO
PA
T s
hock
s0
6
Res
pons
e of
CA
PE
X to
hc
shoc
ks
06
Res
pons
e of
CA
PE
X to
rc
shoc
ks
06
Res
pons
e of
CA
PE
X to
sc
shoc
ks
06
Res
pons
e of
CA
PE
X to
CA
PE
X s
hock
s0
6
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
Figure 9.Impulse responses for 2lag VAR of NOPAT,human, relational,structural capital, andtangible assets for steelcompanies
654
JIC14,4
0.00
00
1.2e
+03
–2.3
162
1.1e
+03
–8.3
e+02
488.
0940
–37.
9098
64.1
053
–19.
1801
487.
0750
–0.5
320
11.6
171
–4.2
931
12.7
518
–7.4
893
2.23
99
–0.2
555
4.78
080.
5647
–0.7
543
–3.2
e+02
662.
4446
–3.9
e+02
784.
6743
–5.7
e+02
1.2e
+03
–59.
0072
51.0
309
–1.0
e+02
299.
9732
–0.0
135
0.01
27
–0.0
179
0.01
37
–0.0
122
0.00
86
–0.0
368
0.13
12
–0.0
117
–1.7
e+02
269.
9638
–2.3
e+02
210.
2328
–2.9
e+02
168.
8970
–28.
3366
17.6
214
0.00
00
266.
8825
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
rc
rc(p
80)
rc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
EV
A to
EV
A s
hock
s0
6
Res
pons
e of
EV
A to
hc
shoc
ks
06
Res
pons
e of
EV
A to
sc
shoc
ks
06
Res
pons
e of
EV
A to
rc
shoc
ks
06
Res
pons
e of
EV
A to
CA
PE
X s
hock
s0
6
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
rc
rc(p
80)
rc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
hc
to E
VA
sho
cks
06
Res
pons
e of
hc
to h
c sh
ock
s0
6
Res
pons
e of
hc
to s
c sh
ock
s0
6
Res
pons
e of
hc
to r
c sh
ock
s0
6
Res
pons
e of
hc
to C
AP
EX
sho
cks
06
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
rc
rc(p
80)
rc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
sc
to E
VA
sho
cks
06
Res
pons
e of
sc
to h
c sh
ock
s0
6
Res
pons
e of
sc
to s
c sh
ock
s0
6
Res
pons
e of
sc
to r
c sh
ock
s0
6
Res
pons
e of
sc
to C
AP
EX
sho
cks
06
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
rc
rc(p
80)
rc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
rc
to E
VA
sho
cks
06
Res
pons
e of
rc
to h
c sh
ock
s0
6
Res
pons
e of
rc
to s
c sh
ock
s0
6
Res
pons
e of
rc
to r
c sh
ock
s0
6
Res
pons
e of
rc
to C
AP
EX
sho
cks
06
(p 2
0) E
VA
EV
A(p
80)
EV
A(p
20)
hc
hc(p
80)
hc
(p 2
0) s
csc
(p 8
0) s
c(p
20)
rc
rc(p
80)
rc
(p 2
0) C
AP
EX
CA
PE
X(p
80)
CA
PE
X
Res
pons
e of
CA
PE
X to
EV
A s
hock
Not
e: E
rror
s are
5 p
erce
nt o
n ea
ch si
de g
ener
ated
by
Mon
te-C
arlo
with
500
reps
s0
6
Res
pons
e of
CA
PE
X to
hc
shoc
ks
06
Res
pons
e of
CA
PE
X to
sc
shoc
ks
06
Res
pons
e of
CA
PE
X to
rc
shoc
ks
06
Res
pons
e of
CA
PE
X to
CA
PE
X s
hoc k
s0
6
0.00
34
Figure 10.Impulse responses for 2
lag VAR of EVA, human,relational, structuralcapital, and tangible
assets for steel companies
655
Intellectualcapital
investments
. In the steel industry, investments in human capital and tangible resources playthe most significant role, and their positive impact on value created remainsduring a long period of time with no significant weakening. Thus, choice ofthe investment object should be based on a maximum amount of investmentsaccumulated over time.
. Value creation in companies providing consulting and educational services is aless definite process. To analyze the efficiency of investments in this industry,more detailed proxy indicators of investments or an individual study of everycompany from this sector is necessary.
The results of this research support the idea that it is necessary to take into account themutual dependence of IC components and performance – otherwise incorrectconclusions may be arrived at, due to the endogeneity problem. Another implication ofthis study is that it is essential to use dynamic models when modeling IC outputs.The authors consider the chosen methodology (orthogonal response functions) to besuitable for the analysis of such a complex process as the transformation of IC intovalue. Another important practical implication is that, surprisingly, IC does notsignificantly affect performance (or even have negative impact) in pharmaceutical,retail, and service industries. This paper suggests a relatively easy way of estimatingthe time delay between IC investment and value creation. Taken together, thisinformation can be used to develop a portfolio strategy based on IC of companies.
The findings in this research are subject to at least three limitations. First, currentlyused proxies seem to be suitable, but it is interesting to collect more relevant data,especially for knowledge-intensive industries. Second, the authors consider only linearmodels; however, instinctively there are reasons to suppose nonlinearity in such asystem. It is necessary to say that linear approximation is still correct, but a nonlinearmodel can have more descriptive power. Third, for practical implications EVA can beeasily replaced by market capitalization for portfolio-building purposes. However, insuch a case much more control variables are required – market capitalization seemsto be driven by many factors, not only IC investments.
References
Bajburina, J.R. and Golovko, T.V. (2008), “Empirical study of the intellectual value of largeRussian companies and the factors of its growth”, EJournal of Corporate Finance, Vol. 2No. 6, pp. 56-78.
Bassi, L. and van Buren, M. (1999), “Valuing investments in intellectual capital”, InternationalJournal of Technology Management, Vol. 18 No. 5, pp. 414-432.
Bontis, N. (1996), “There’s a price on your head: managing intellectual capital strategically”,Business Quarterly, Vol. 60 No. 4, pp. 40-47.
Bontis, N. (1998), “Intellectual capital: an exploratory study that develops measures and models”,Management Decision, Vol. 36 No. 2, pp. 63-76.
Bontis, N. (1999), “Managing organizational knowledge by diagnosing intellectual capital:framing and advancing the state of the field”, International Journal of TechnologyManagement, Vol. 18 Nos 5-8, pp. 433-462.
Bontis, N., Keow, W. and Richardson, S. (2000), “Intellectual capital and business performance inMalaysian industries”, Journal of Intellectual Capital, Vol. 1 No. 1, pp. 85-100.
Brennan, N. and Connell, B. (2000), “Intellectual capital: current issues and policy implications”,Journal of Intellectual Capital, Vol. 1 No. 3, pp. 206-240.
656
JIC14,4
Cabrita, M., Landeiro de Vaz, J. and Bontis, N. (2007), “Modelling the creation of value fromintellectual capital: a Portuguese banking perspective”, International Journal of Knowledgeand Learning, Vol. 3 Nos 2-3, pp. 266-280.
Chan, K.H. (2009), “Impact of intellectual capital on organisational performance: an empiricalstudy of companies in the Hang Seng Index (part 2)”, Learning Organization, Vol. 16 No. 1,pp. 22-39.
Chang, W.S. and Hiesh, J.J. (2011), “Exploring a human capital driven intellectual capitalframework: evidence from information technology industry in Taiwan”, European Journalof Social Sciences, Vol. 21 No. 3, pp. 392-404.
Chen, J., Zhu, Z. and Xie, H.Y. (2004), “Measuring intellectual capital: a new model and empiricalstudy”, Journal of Intellectual Capital, Vol. 5 No. 1, pp. 195-212.
Chen, M.C. (2004), “Intellectual capital and competitive advantages: the case of TTY”, Journalof Business Chemistry, Vol. 1 No. 1, pp. 14-20.
Cheng, M.-Y., Lin, J.-Y., Hsiao, T.-Y. and Lin, T.W. (2010), “Invested resource, competitiveintellectual capital, and corporate performance”, Journal of Intellectual Capital, Vol. 11No. 4, pp. 433-450.
Damodaran, A. (2004), Investment Valuation: Tools and Techniques for Determining the Value ofAny Asset, Alpina Business Books, Moscow, (in Russian).
Dıez, J.M., Ochoa, M.L., Prieto, M.B. and Santidrian, A. (2010), “Intellectual capital and valuecreation in Spanish firms”, Journal of Intellectual Capital, Vol. 11 No. 3, pp. 348-367.
Edvinsson, L. and Malone, M. (1997), Intellectual Capital: Realising Your Company’s True Valueby Finding its Hidden Brainpower, Harper Collins, New York, NY.
Edvinsson, L. and Sullivan, P. (1996), “Developing a model for management intellectual capital”,European Management Journal, Vol. 14 No. 4, pp. 187-199.
Eklund, J.E. (2010), “Q-theory of investment and earnings retentions – evidence fromScandinavia”, Empirical Economics, Vol. 39 No. 3, pp. 793-813.
Firer, S. and Williams, S.M. (2003), “Intellectual capital and traditional measures of corporateperformance”, Journal of Intellectual Capital, Vol. 4 No. 3, pp. 348-360.
Garanina, T.A. (2009), “The role of intangible assets in value creation of the company: thetheoretical and practical aspects”, EJournal of Corporate Finance, Vol. 4 No. 12,pp. 79-96.
Huang, C. and Wang, M.-C. (2008), “The effects of economic value added and intellectual capitalon the market value of firms: an empirical study”, International Journal of Management,Vol. 25 No. 4, pp. 722-731.
Huang, C.F. and Hsueh, S.L. (2007), “A study on the relationship between intellectual capital andbusiness performance in the engineering consulting industry: a path analysis approach”,Journal of Civil Engineering and Management, Vol. 13 No. 4, pp. 265-271.
Huang, C.J. and Liu, C.J. (2005), “Exploration for the relationship between innovation, IT andperformance”, Journal of Intellectual Capital, Vol. 6 No. 2, pp. 237-252.
Huang, Y.C. and Wu, Y.-C.J. (2010), “Intellectual capital and knowledge productivity: the Taiwanbiotech industry”, Management Decision, Vol. 48 No. 4, pp. 580-599.
Kamukama, N., Ahiauzu, A. and Ntayi, J.M. (2010), “Intellectual capital and performance: testinginteraction effects”, Journal of Intellectual Capital, Vol. 11 No. 4, pp. 554-574.
Kaplan, R.S. and Norton, D.P. (1992), “The balanced scorecard – measures that driveperformance”, Harvard Business Review, Vol. 70 No. 1, pp. 71-79.
Lee, J.Y. (2011), “Incremental innovation and radical innovation: the impacts of human,structural, social, and relational capital elements”, PhD dissertation, Michigan StateUniversity, East Lansing, MI.
657
Intellectualcapital
investments
Lev, B. (1999), “Seeing is believing – a better approach to estimating knowledge capital”, CFOMagazine, February, pp. 29-37.
Lin, Y.Y. and Edvinsson, L. (2008), “National intellectual capital: comparison of the Nordiccountries”, Journal of Intellectual Capital, Vol. 9 No. 4, pp. 525-545.
Love, I. and Zicchino, L. (2006), “Financial development and dynamic investment behavior:evidence from panel VAR”, The Quarterly Review of Economics and Finance, Vol. 46 No. 2,pp. 190-210.
Marr, B. and Adams, C. (2004), “The balanced scorecard and intangible assets: similar ideas,unaligned concepts”, Measuring Business Excellence, Vol. 8 No. 3, pp. 18-27.
Molodchik, M., Shakina, E. and Bykova, A. (2012), “Intellectual capital transformation evaluatingmodel”, Journal of Intellectual Capital, Vol. 13 No. 4, pp. 444-461.
Moon, Y.J. and Kym, H.G. (2006), “A model for the value of intellectual capital”, Canadian Journalof Administrative Sciences, Vol. 23 No. 3, pp. 253-269.
Nazari, J.A. (2010), “An investigation of the relationship between the intellectual capitalcomponents and firm’s financial performance”, PhD Thisis, Calgary, Alberta.
Roos, J., Roos, G. and Edvinsson, L. (1998), Intellectual Capital – Navigating the New BusinessLandscape, Macmillan Press Ltd, London.
Shakina, E.A. (2011), “Intellectual capital: value drivers and role of the institutions in Europe andRussia”, conference proceedings, EACES, Milocer, available at: www.hse.ru/data/2011/09/29/1270040956/Paper_ShakinaE_BarajasA.pdf (accessed December 29, 2012).
Shakina, E.A. and Barajas, A. (2012), “The relationship between intellectual capital quality andcorporate performance: an empirical study of Russian and European companies”,Economic Annals, Vol. LVII No. 192, pp. 79-98.
Shiu, H. (2006), “The application of the value added intellectual coefficient to measure corporateperformance: evidence from technological firms”, International Journal of Management,Vol. 23 No. 2, pp. 356-364.
Stern, J.M., Shiely, J.S. and Ross, I. (2001), The EVA Challenge: Implementing Value Added Changein an Organization, Wiley & Sons Inc, New York, NY.
Stewart, T.A. (1991), “Brainpower: intellectual capital is becoming corporate America’s mostvaluable asset and can be its sharpest competitive weapon; the challenge is to find whatyou have – and use it”, Fortune, Vol. 123 No. 11, pp. 44-60.
Stewart, T.A. (1997), Intellectual Capital: The New Wealth of Organization, Nicholas Brealey, London.
Sullivan, P.H. (2000), Value-Driven Intellectual Capital. How to Convert Intangible CorporateAssets into Market Value, Wiley, New York, NY.
Sveiby, K.-E. (1997), “The ‘invisible’ balance sheet”, available at: www.sveiby.com/articles/InvisibleBalance.html (accessed November 12, 2007).
Sveiby, K.-E. (2001), “Methods for measuring intangible assets”, available at: www.sveiby.com/articles/IntangibleMethods.htm (accessed December 29, 2012).
Ting, I.W.K. and Lean, H.H. (2009), “Intellectual capital performance of financial institutions inMalaysia”, Journal of Intellectual Capital, Vol. 10 No. 4, pp. 588-599.
Tsan, W.N. (2004), “The measurement of intellectual capital for IT industry in Taiwan – systemtheory perspective”, presented at IC Congress, Amsterdam.
Tseng, C.-Y. and Goo, Y.-J.J. (2005), “Intellectual capital and corporate value in an emergingeconomy: empirical study of Taiwanese manufacturers”, R&D Management, Vol. 35 No. 2,pp. 187-199.
Wang, W.-Y. and Chang, C. (2005), “Intellectual capital and performance in causal models:evidence from the information technology industry in Taiwan”, Journal of IntellectualCapital, Vol. 6 No. 2, pp. 222-236.
658
JIC14,4
Wu, C.C. (2004), “An empirical examination of relevance between intellectual capital and equityvalue”, presented at IC Congress, Amsterdam.
Zickgraf, S., Mertins, K., Will, M., Wang, W.-H. and Meye, J. (2007), “C.InCaS: intellectual capitalstatement”, European ICS Guideline, available at: www.psych.lse.ac.uk/incas/page114/files/page114_1.pdf (accessed 30 May 2013).
Further reading
Brynjolfsson, E., Hitt, L.M. and Yang, S. (2002), “Intangible assets: computers and organizationalcapital”, Brookings Papers on Economic Activity, Vol. 1, pp. 137-198.
Appendix
VariableNumber of
observations Average SD Minimum Maximum
Pharmaceutical industryInvestments in human capital (hc) 927 0.33 3.53 �20.66 63.35Investments in structural capital (sc) 998 437.63 4,252.49 �8,886.00 71,206.00Investments in relational capital (rc) 868 �0.20 15.52 �276.59 307.00Investments in tangible assets (CAPEX) 1,160 205.50 569.53 0.00 3,106.19NOPAT 1,170 763.63 2,264.92 �1,200.18 15,911.69EVA 1,162 625.09 1,977.94 �4,264.83 12,693.93Retail industryInvestments in human capital (hc) 971 0.74 8.12 �84.77 139.00Investments in structural capital (sc) 963 9.80 424.14 �5,783.00 6,656.37Investments in relational capital (rc) 1,005 0.01 0.21 �2.74 4.17Investments in tangible assets (CAPEX) 1,128 210.47 505.32 �0.08 4,010.00NOPAT 1,131 214.36 569.24 �3,590.83 5,986.91EVA 1,121 150.00 480.40 �3,419.14 5,550.54Service industryInvestments in human capital (hc) 1,189 0.55 3.38 �35.00 32.00Investments in structural capital (sc) 1,327 33.52 217.84 �990.70 4,337.70Investments in relational capital (rc) 1,117 0.20 20.02 �327.10 351.60Investments in tangible assets (CAPEX) 1,515 24.55 65.22 0.00 1,342.00NOPAT 1,515 67.84 218.21 �1,384.12 2,524.13EVA 1,125 42.04 166.77 �395.96 2,524.13Steel industryInvestments in human capital (hc) 351 1.10 10.50 �37.88 149.19Investments in structural capital (sc) 400 128.27 1,017.72 �6,249.43 13,463.59Investments in relational capital (rc) 404 �0.01 3.87 �64.71 31.43Investments in tangible assets (CAPEX) 457 344.75 1,179.32 0.00 14,157.90NOPAT 460 500.65 1,718.62 �1,427.34 19,800.47EVA 459 336.05 1,349.93 �3,057.31 17,030.97TelecommunicationsInvestments in human capital (hc) 1,519 0.55 9.78 �99.49 196.91Investments in structural capital (sc) 1,625 305.99 4,538.37 �37,429.86 104,839.00Investments in relational capital (rc) 1,664 0.25 8.25 �40.05 317.76Investments in tangible assets (CAPEX) 1,896 1,074.78 2,616.27 0.00 20,478.00NOPAT 1,904 845.28 2,538.20 �14,961.28 32,009.89EVA 1,892 318.85 2,224.89 �27,376.30 31,227.48
Table AI.Descriptive statics
of data used
659
Intellectualcapital
investments
About the authors
Iuliia Naidenova is an Assistant Professor at NRU HSE. Her research interests are related tointellectual capital evaluation and strategic financial management. The research results arepublished in Russian scientific journals and international conferences. For further informationsee: http://www.hse.ru/en/org/persons/14538835. Iuliia Naidenova is the corresponding authorand can be contacted at: [email protected]
Petr Parshakov is an Assistant Professor at NRU HSE. His field of research interest includestime series econometrics, asset pricing, performance evaluation and forecasting. The studyresults are published in different Russian scientific journals and international conferences.For further information see: http://www.hse.ru/en/org/persons/14678557
To purchase reprints of this article please e-mail: [email protected] visit our web site for further details: www.emeraldinsight.com/reprints
660
JIC14,4