Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
In o ma ion sha ing in supply chains wi h he e ogeneous e aile s
Robe o Domingueza,*, Sal a o e Cannellab, Ana P. Ba bosa-Pó oaa, Jose M. F aminanc
aCen e o Managemen S udies, Ins i u o Supe io Técnico (CEG-IST), Technical Uni e si y o Lisbon,
Po ugal
bDICAR, Uni e si y o Ca ania, Ca ania, I aly
cIndus ial Managemen & Business Adminis a ion Depa men , School o Enginee ing, Uni e si y o
Se ille, Spain
E-Mails: obe o.dominguez@ ecnico.ulisboa.p , cannella@unic .i , apo[email p o ec ed]a.p ,
[email p o ec ed]
*Co esponding au ho : Robe o Dominguez, Cen e o Managemen S udies (CEG-IST), Ins i u o
Supe io Técnico, Technical Uni e si y o Lisbon, A e. Ro isco Pais 1, 1049-001, Lisbon, Po ugal.
Abs ac
This wo k analyses pa ial in o ma ion sha ing in ol ing e aile s wi h di e en
ope a ional con igu a ions. Re aile s a e cha ac e ized by ou ope a ional ac o s, i.e.,
demand a iance, lead ime a e age, o ecas ing pe iod and in en o y policy. The
indings show ha he pe o mance imp o emen based on in o ma ion sha ing depends
on e aile s’ ope a ional ac o s. Consequen ly, pa ial in o ma ion sha ing s uc u es
need o be ca e ully designed in o de o achie e a subs an ial pe o mance
imp o emen . The esul s also se e o p o ide inno a i e ecommenda ions o supply
chain manage s in o de o e icien ly implemen in o ma ion sha ing mechanisms a
e aile s.
Keywo ds: Supply chain managemen ; pa ial in o ma ion sha ing; he e ogeneous
e aile s; bullwhip e ec ; mul i-agen sys ems; dynamic pe o mance.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
2
1 INTRODUCTION
1.1 Con ex
Globaliza ion and a high olume o ou sou cing has esul ed in decen alized Supply
Chains (SCs), shi ing om a sequen ial linea SC o an inc easingly complex global
supply ne wo k (see e.g., Me zi onluoglu 2015, Anna elli and Nonino 2016, Li and
Zhen 2016). SC pa ne s ha e a highe au onomy, as hey a e pa o many pa allel
chains a he same ime (Zissis e al. 2015, Thomas e al. 2016). This ac ein o ced he
p esence o con lic ing objec i es wi hin he SC whe e compe i ion exis s o common
esou ces and decisions a e aken on indi idually based local incen i es (Rached e al.
2016). The complexi y o SCs has isen sha ply in ecen decades (Ca doso e al. 2015,
Gue le and Spinle 2015), o en leading o a lack o coo dina ion among SC membe s.
In his con ex , SCs om wes e n economies o low-cos coun ies ha e been
expe iencing unp edic able and in ensi e de e io a ion o pe o mance (Ch is ophe and
Holweg 2017). Addi ionally, he se e e and synch onized ade collapse has ampli ied
ine iciencies wi hin he SCs, and subsequen ly led o de imen al phenomena such as
he bullwhip e ec (see e.g. Al omon e e al. 2012, Duan e al. 2015, Osadchiy e al.
2015). To o e come hese ine iciencies, esea che s and p ac i ione s ha e been
wo king on obus solu ions. Among hese, SC collabo a ion p ac ices ha e been
ad oca ed as some o he mos e ec i e app oaches (see e.g. Dejonckhee e e al. 2004,
Chen and Lee 2009, T ape o e al. 2012, Li and Zhang 2015, among o he s). A he co e
o collabo a ion p ac ices lies in o ma ion sha ing (IS), a collabo a i e mechanism in
which he supplie may ob ain and u ilize he demand and in en o y s a us o he e aile
(Huang e al. 2016).
Du ing he las decade, he bene i s o IS in decen alized SCs ha e been deeply
esea ched wi h empi ical s udies o eal cases (see e.g. Huo e al. 2014, Bian e al.
2016, Ren 2017), analy ical me hods (Chen and Lee 2009, T ape o e al. 2012, Ali e al.
2017), and simula ion (Da a and Ch is ophe 2011, Ramana han 2014, Dominguez e
al. 2015b, Cannella e al. 2017). In gene al, ega dless o he adop ed me hodologies
and he explo ed aspec s o IS (e.g. easons o sha ing, wha in o ma ion o sha e wi h
whom, how o sha e, as well as p e- equisi es, d i e s and ba ie s o IS, see Kemb o e
al. 2014), he majo i y o he li e a u e ag ees on he pi o al ole bene i s o IS p ac ices
in SC pe o mance (Maghsoudi and Pazi andeh 2016). The expec ed e enues (e.g. a
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
3
educ ion in in en o y holding cos , Hosoda e al. 2008) ha e been cap u ing he
a en ion o SC p ac i ione s (Kemb o and Sel ia idis 2015). As an example, a ecen
su ey ound ha 61% o Chinese i ms belie e ha IS is essen ial o business success
(Bian e al. 2016). Basically, IS has been and con inues o be a majo opic in mode n
SC managemen and, con a y o popula belie , he e is s ill signi ican need o mo e
esea ch ega ding IS in SC (Kemb o e al. 2014, Cos an ino e al. 2015).
1.2 P oblem S a emen
Despi e he po en ial bene i s o IS in SC, i s p ac ical implemen a ion p esen s ele an
di icul ies (Fawce e al. 2011, Spekman and Da is 2016). Full coo dina ion among
SC membe s, while desi able, is o en imp ac ical, since i is deemed o be oo cos ly o
oo isky (Geunes e al. 2016). Making in o ma ion a ailable o o he en e p ises and
managing he in o ma ion equi es in es men in In o ma ion Technology (IT) and
en ails signi ican esou ce in es men s, which could esul in a nega i e cos –bene i
analysis (Chan and Chan 2010, Kemb o e al. 2014). Addi ionally, companies need o
bea he isk ha in o ma ion may be leaked in en ionally o unin en ionally by
supplie s (Kong e al. 2013, Huang e al. 2016). Finally, esul ing bene i s o IS may be
di icul o alloca e in a easonable way among SC pa ne s (Shih e al. 2015).
E idence o hese ba ie s o achie e ull collabo a ion among SC membe s can be
ound in p ac ice. Acco dingly, he Re aile -Di ec Da a Repo o he G oce y
Manu ac u e s Associa ion (GMA) poin ed ou ha e aile s may no ha e an incen i e
o sha e da a wi h supplie s (GMA 2009, Shang e al. 2016). Addi ionally, a s udy
pe o med by Fo es e Resea ch on 89 e aile s in 2006 epo ed ha only 27% o
e aile s sha ed POS da a (Shang e al. 2016). In his con ex , achie ing a ull IS (i.e., all
SC membe s pa icipa e in IS, e e ed o as FIS in he ollowing) is no always
possible. Thus, in p ac ice, pa ial IS is ound o be p e alen (Shnaide man and
Oua dighi 2014, Xu e al. 2015). Howe e , in he scien i ic li e a u e, pa ial IS has been
a ely analysed because he majo i y o s udies dealing wi h IS assume a ull
collabo a ion p ac ice among all membe s (Holms őm e al. 2016). In ligh o hese
conside a ions, s udying he dynamics o SC in scena ios whe e FIS canno be achie ed
ep esen s a challenge o esea che s and may b ing po en ial bene i s o indus y.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
4
To he bes o he au ho s’ knowledge, up o now, pa ial IS has been add essed in
li e a u e in i e ele an s udies. Ganesh e al. (2014a,b) and Cos an ino e al. (2014)
analyse he impac o di e en deg ees o collabo a ion on SC pe o mance (i.e.,
in en o y holding and sho age cos s, bullwhip e ec and/o cus ome se ice le el) in a
se ial SC, while Lau e al. (2004) analyse pa ial IS in mo e complex SCs, in pa icula
in h ee di e gen SCs. Finally, Huang and I a ani (2005) ocus on one capaci a ed
manu ac u e and wo e aile s unde a (Q,R) in en o y policy, whe e he o me
ecei es demand and in en o y in o ma ion om only one o he e aile s.
The abo e-men ioned wo ks ha e signi ican ly con ibu ed o he unexplo ed opic o
pa ial IS by showing wo no el insigh s:
(1) Re aile s should be he i s membe s o be in ol ed in IS (Ganesh e al. 2014a,b,
Cos an ino e al. 2014, Lau e al. 2004), since hey epo he highes pe o mance
imp o emen o he SC.
(2) The ope a ional ac o s (OFs) o e aile s, such as ma ke sha es and o de sizes,
may ha e a signi ican impac on he bene i s p o ided by he IS p ac ice unde
pa ial collabo a ion (Huang and I a ani 2005).
The o me insigh easse s he cen al ole o e aile s o he e icacy o IS, while he
la e sugges s ha SCs cha ac e ized by he e ogeneous e aile s (i.e., e aile s wi h
di e en OFs such as lead imes, o de policies, ma ke demand, e c.), may pe o m
di e en ly unde he same IS p ac ice. Bo h insigh s open in e es ing challenges o
esea che s and ad oca e impo an implica ions o indus y, as hey poin ou he
ele ance o explo ing he e iciency o pa ial IS a e aile s when hese a e
he e ogeneous. Acco ding o hese insigh s, we o mula e he ollowing esea ch
ques ions: how e aile s wi h di e en OFs may impac on SC pe o mance when hey
sha e in o ma ion abou cus ome demand? Which e aile s’ OFs a e mo e ele an in
o de o conside a e aile as a po en ial pa ne o he IS scheme and a wha ex en ?
1.3 Objec i e
Mo i a ed by he abo e conside a ions, in his pape we aim o con ibu e o he exis ing
li e a u e by assessing how he e ogeneous e aile s, cha ac e ized by di e en c i ical
OFs (i.e., demand a iabili y, a e age lead ime, o ecas ing pe iod and in en o y
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
5
policy), may imp o e SC pe o mance by sha ing (o no ) ue demand in o ma ion. We
assume ha due o he decen alized na u e o mode n SCs, companies and, speci ically,
he e aile s, a e au onomous membe s who wo k o hei own goals and in e es s and
hus, e aile s’ OFs a e conside ed as exogenous ac o s. In his manne we aim o
p o ide ecommenda ions o SC manage s on how o p ope ly exploi he bene i s o
implemen ing IS p ac ices wi h e aile s by iden i ying which e aile s p o ide a highe
con ibu ion o SC pe o mance.
To ul il he esea ch objec i e, we ocus on a ou echelon SC (i.e., Fac o y,
Dis ibu o , Wholesale and Re aile ) in which each echelon is cha ac e ized by one
membe wi h he excep ion o he Re aile ’s echelon, which is cons i u ed by ou
membe s. We compa e di e en pa ial IS scena ios (some e aile s may sha e demand
in o ma ion, while some o he s may no sha e in o ma ion) unde wo di e en
hypo hesis: (1) homogeneous e aile s and (2) he e ogeneous e aile s. Unde he o me
hypo hesis we analyse he SC pe o mance when iden ical e aile s a e in ol ed in IS
one by one, on a ie y o SC con igu a ions. Unde he la e hypo hesis we assess he
impac on SC pe o mance o in ol ing e aile s wi h di e en OFs in IS. SC
pe o mance is measu ed using a se o sys em le el me ics (i.e., Bullwhip Slope,
In en o y Slope and Sys emic In en o y Le el), which p o ide a clea , comp ehensi e
and s uc u ed assessmen o he SC pe o mance a sys emic le el and he “in e nal
p ocess e iciency”, as well as p o ide in o ma ion on he po en ial bene i s o
pa ne ships, collabo a ion and in o ma ion p oduc i i y o SC membe s (Cannella e al.
2013).
Due o he explo a o y na u e o his esea ch, we adop an app op ia e and s uc u ed
me hodology o s udying he dynamic o SCs, i.e., compu e simula ion (Oli ei a e al.
2016), and mo e speci ically he Mul i-Agen Sys ems (MAS) modelling app oach
(Cha ield e al. 2006, Rahmandad and S e man 2008). MAS has been ecognized as a
use ul me hodology o pe o m complex p ospec i e SC analysis, and indings ob ained
wi h i s p ope adop ion ha e been signi ican ly con ibu ing o unde s and he
dynamics in SC (see e.g., Swamina han e al. 1998, Long and Zhang 2014, Hille o h e
al. 2016 o Pon e e al. 2017). In o de o pe o m a sys ema ic simula ion analysis we
adop easonable assump ions and da a inpu s o simula ions ob ained om di e en
cases o emula e eal-wo ld logis ic sys ems (Rabino ic and Cheon 2011, Cannella e al.
2017).

Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
6
The esul s ob ained e eal new insigh s on he impac o IS in SC by showing he need
o indi idually es ima ing he po en ial alue o e aile s’ in o ma ion p io o he
implemen a ion o IS. When e aile s a e homogeneous, hei collabo a ion may p o ide
equal po en ial bene i s o SC pe o mance ( hey con ibu e he same o imp o e SC
pe o mance when hey a e in ol ed in IS). Unde his hypo hesis, bene i s o IS
inc ease wi h he numbe o e aile s in ol ed and a ull IS app oach is ecommended.
On he con a y, when e aile s a e he e ogeneous hey ha e di e en po en ial alue
depending on hei ope a ional con igu a ion. Unde his hypo hesis, pe o mance
achie ed by di e en pa ial IS s uc u es wi h he same numbe o e aile s migh be
signi ican ly di e en (e.g., we ound ha in ol ing hal o he o al numbe o e aile s
in o IS may lead o ob ain o e 70% o he o al bene i s o a FIS unde he bounda y
condi ions). In ac , e aile s wi h (1) highe demand a iance, (2) lowe o ecas ing
pe iod, and (3) highe a e age lead ime, a e po en ially he mos bene icial pa ne s
when implemen ing IS.
The emainde o his pape is as ollows: Sec ion 2 desc ibes he SC model and
me hodology. Sec ion 3 p esen s he design o expe imen s and pe o mance me ics.
Sec ion 4 analyses he esul s ob ained. Sec ion 5 p esen s manage ial implica ions.
Finally, Sec ion 6 d aws he conclusions, limi a ions o he s udy and u u e esea ch
lines.
2 SC MODEL AND METHODOLOGY
In o de o analyse he pa ial IS scena ios, we de elop a SC model o conduc he
expe imen s. In SC dynamics li e a u e, he mos used SC model is he ou -echelon
se ial SC (see e.g. S e man 1989, Cha ield e al. 2004, C oson e al. 2014, Cannella e
al. 2015). Echelons a e e e ed as Fac o y (i=1), Dis ibu o (i=2), Wholesale (i=3),
and Re aile (i=4). In o de o analyse scena ios whe e only some o he e aile s
pa icipa e in IS ( e e ed as pa ial IS) we ex end his SC model by inc easing he
numbe o e aile s o ou , hus esul ing a di e gen SC (Lau e al. 2004, Dominguez e
al. 2015a, Rached e al. 2016), as shown in Figu e 1.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
7
Figu e 1. SC con igu a ion.
In addi ion, we make he ollowing assump ions ega ding IS:
1. As we ocus ou analysis o pa ial IS a e aile s’ s age, only e aile s sha e
in o ma ion on cus ome ’s demand.
2. Assuming ha , due o some ba ie s (as desc ibed in Sec ion 1) each en e p ise
is willing o sha e i s local in o ma ion only o i s immedia e ups eam en e p ise
(see Lau e al. 2004, Kemb o and Sel ia idis 2015, o simila assump ions),
only he wholesale will be able o ecei e in o ma ion om e aile s.
2.1 Supply Chain model
The SC gene al model has been adap ed om Cha ield e al. (2004) so as o model a
gene ic di e gen SC (Dominguez e al. 2015a,b, Cannella e al. 2017) and o include
pa ial IS (i.e., any node a any echelon o he SC may sha e in o ma ion wi h an
ups eam linked node). The no a ion is desc ibed in Table 1. This gene al model is
adap ed in Sec ion 3.1 o he SCs unde s udy wi h speci ic pa ame e s alues and
expe imen al ac o s.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
8
Table 1. No a ion.
i
Echelon posi ion in he SC
𝜏𝑖𝑗
Fo ecas ing pe iod o 𝑛𝑖𝑗
j
Node posi ion wi hin a gi en echelon
𝑂𝑖𝑗
𝑡
O de s placed by 𝑛𝑖𝑗 a ime
𝑛𝑖𝑗
Node a posi ion j in echelon i
𝐼𝑖𝑗
𝑡
In en o y on hand o 𝑛𝑖𝑗 a ime
E
To al numbe o echelons
𝑊𝐼𝑃𝑖𝑗
𝑡
Wo k in p og ess o 𝑛𝑖𝑗 a ime
𝑁𝑖
To al numbe o nodes in echelon i
𝐵𝑖𝑗
𝑡
Backlog o 𝑛𝑖𝑗 a ime
𝑁𝐶
To al numbe o cus ome s
𝑆ℎ𝐷𝑖𝑗
𝑡
Sha ed demand o 𝑛𝑖𝑗 a ime
𝐶𝑗
Cus ome a posi ion j
𝛿𝑖𝑗
𝛿𝑖𝑗=1 i 𝑛𝑖𝑗 is in ol ed in IS, 0 o he wise.”
Cu en simula ion ime
𝐼𝑃𝑖𝑗
In en o y policy o 𝑛𝑖𝑗
T
To al simula ion ime (excluding wa m-up)
𝑉𝑖𝑗
Se o downs eam linked pa ne s o 𝑛𝑖𝑗
𝐷𝐶𝑗
𝑡
Demand placed by cus ome 𝐶𝑗 a ime
𝑠𝑂𝑖𝑡
2
Es ima ed a iance o o de s placed by echelon i
𝜇𝐷𝐶𝑗
A e age demand placed by 𝐶𝑗
𝑂
𝑖𝑡
Es ima ed a e age o o de s placed by echelon i a
ime
𝐷
𝐶𝑗
𝑡
Es ima ed a e age demand placed by 𝐶𝑗 a
ime
𝜎𝑂𝑖𝑗
2
Va iance o o de s placed by 𝑛𝑖𝑗
𝜎𝐷𝐶𝑗
2
Va iance o demand placed by 𝐶𝑗
𝑠𝑂𝑖𝑗
𝑡
2
Es ima ed a iance o o de s placed by 𝑛𝑖𝑗
𝑠𝐷𝐶𝑗
𝑡
2
Es ima ed a iance demand placed by 𝐶𝑗
𝑠𝐼𝑖𝑡
2
Es ima ed a iance o in en o y a echelon i
𝐷𝑖𝑗
𝑡
Demand aced by 𝑛𝑖𝑗 a ime
𝐼𝑖𝑡
Es ima ed a e age o in en o y a echelon i a
ime
𝐷
𝑖𝑗
𝑡
Es ima ed a e age demand aced by 𝑛𝑖𝑗 a
ime
𝑠𝐼𝑖𝑗
𝑡
2
Es ima ed a iance o in en o y a 𝑛𝑖𝑗
𝑠𝐷𝑖𝑗
𝑡
2
Es ima ed a iance demand aced by 𝑛𝑖𝑗 a
ime
𝐼𝑖𝑗
𝑡
Es ima ed a e age o in en o y a 𝑛𝑖𝑗 a ime
𝐿𝑖𝑗
𝑡
Lead ime o 𝑛𝑖𝑗 a ime
𝐷
𝐶
𝑡
Es ima ed a e age demand placed by cus ome s a
ime
𝜇𝐿𝑖𝑗
A e age lead ime o 𝑛𝑖𝑗
𝜋𝑖
Posi ion o he i- h echelon
𝐿
𝑖𝑗
𝑡
Es ima ed a e age lead ime o 𝑛𝑖𝑗 a ime
𝑂𝑅𝑉𝑟𝑅𝑖
O de Ra e Va iance Ra io echelon i
𝜎𝐿𝑖𝑗
2
Va iance o he lead ime o 𝑛𝑖𝑗
𝐼𝑛𝑣𝑉𝑟𝑅𝑖
In en o y Va iance Ra io echelon i
𝑠𝐿𝑖𝑗
𝑡
2
Es ima ed a iance o he lead ime o 𝑛𝑖𝑗 a
ime
𝐼𝑛𝑣𝐴𝑣𝑖
In en o y A e age a echelon i
R
In en o y e iew pe iod
BwSl
Bullwhip slope
𝑆𝑖𝑗
𝑡
Desi ed le el o s ock o 𝑛𝑖𝑗 a ime
In Sl
In en o y slope
z
Sa e y ac o o he OUT policy
SysIn A
Sys emic in en o y a e age
Gene al Modelling Assump ions
 A pe iod , each cus ome 𝐶𝑗 places an independen s ochas ic demand
𝐷𝐶𝑗
𝑡 ollowing a no mal dis ibu ion wi h mean 𝜇𝐷𝐶𝑗, es ima ed by 𝐷
𝐶𝑗
𝑡, and
a iance 𝜎𝐷𝐶𝑗
2, es ima ed by 𝑠𝐷𝐶𝑗
𝑡
2. Cus ome s do no ill o de s.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
9
 The ac o y places o de s o an ou side supplie wi h unlimi ed capaci y.
 S ocking and anspo a ion capaci ies a e unlimi ed.
 The demand ecei ed by node 𝑛𝑖𝑗 (𝐷𝑖𝑗
𝑡), wi h mean es ima ed by 𝐷
𝑖𝑗
𝑡 and
a iance es ima ed by 𝑠𝐷𝑖𝑗
𝑡
2, equals he o al o de s ecei ed by downs eam
(linked) pa ne s (deno ed by 𝑉𝑖𝑗), i.e., 𝐷𝑖𝑗
𝑡=∑𝑂𝑖+1,𝑟
𝑡
𝑟∈𝑉𝑖𝑗 . Demand ecei ed by
e aile s is cus ome demand 𝐷𝐸𝑗
𝑡=𝐷𝐶𝑗
𝑡.
 When he s ock is no enough o ill an o de comple ely he e is a s ock-ou
si ua ion and pa ial eplenishmen is used (Cha ield e al. 2004).
 I a s ock-ou si ua ion a he e aile s’ echelon occu s, we assume ha
backo de ing is no allowed and un illed demand is los . Howe e , ue demand
ecei ed a e aile s is eco ded (𝐷𝐶𝑗
𝑡), and sha ed wi h he ups eam pa ne in
case o pa icipa ing in IS (see a de ailed desc ip ion o IS below) (Cha ield e
al. 2004, Ag awal e al. 2009, Choudha y and Shanka 2015). Ups eam
membe s o he SC a e allowed o backo de .
 We assume ha e u ns o excess in en o y o ups eam pa ne s a e no
pe mi ed since he allowance o e u ns, al hough a common assump ion in he
bullwhip e ec li e a u e, may no be ealis ic and may o e es ima e he
bullwhip e ec (Cha ield and P i cha d 2013, Dominguez e al. 2015b).
Lead Times
Lead imes (𝐿𝑖𝑗
𝑡) a e de ined as he ime elapsed be ween o de and eceip , and may
include manu ac u ing ime, shipmen o po , ship ansi ime, unloading, ans e o
ail and/o uck, e c. (Disney e al. 2016). We assume s ochas ic lead imes, which a e
s a iona y, independen , and iden ically dis ibu ed. In line wi h p e ious li e a u e
wo ks and indus ial da a se s, lead imes a e assumed o ollow a Gamma dis ibu ion
(Cha ield e al. 2004, Kim e al. 2006, Hayya e al. 2011, Cha ield and P i cha d 2013,
Bischak e al. 2014, Dominguez e al. 2015b, Cannella e al. 2017, among o he s) wi h
mean 𝜇𝐿𝑖𝑗 and a iance 𝜎𝐿𝑖𝑗
2. Since we use a pe iodic O de -Up-To (OUT) eplenishmen
policy (see below), and his policy ope a es on a disc e e ime basis, lead imes mus be
in ege s (Disney e al. 2016, Wang and Disney 2017). The e o e, alues ob ained om
he Gamma dis ibu ion a e disc e ized. Consequen ly, each ime an o de is gene a ed,
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
16
ex eme alues o he ac o s (Cos an ino e al. 2014, Cannella e al. 2017). These alues
a e chosen acco ding o wo p inciples:
(1) In o de o sa is y he hypo hesis o he e ogeneous e aile s, OFL and OFH need
o be signi ican ly di e en .
(2) In o de o p oduce compa able esul s, OFL and OFH need o adop alues
om o he simila s udies in SC dynamic li e a u e.
OFH alues o 𝜎𝐷𝐶𝑗
2, 𝜏𝑖𝑗, and 𝜇𝐿𝑖𝑗, can be ound in Cha ield e al. (2004), Cha ield e
al. (2013), Cos an ino e al. (2014) and Dominguez e al. (2015b). OFL alues o hese
ac o s a e ob ained by signi ican ly educing he OFH alues. Fo IPij, OFH is se o S2
(see e.g. Cha ield e al. 2004, Nach mann e al. 2010, Cha ield e al. 2013, Dominguez
e al. 2015b), while OFL is se o S1 (see e.g. Cha ield e al. 2004, Dominguez e al.
2014, Cos an ino e al. 2014). This is an a bi a y choice wi hou impac in he esul s.
These alues can be ound in Table 2.
Table 2. Ope a ional ac o s, model pa ame e s, simula ion pa ame e s and pe o mance me ics.
OFs
Re aile s
Ups eam
Membe s
Low (OFL)
High (OFH)
Demand a iance (𝜎𝐷𝐶𝑗
2)
100 (𝜎𝐷𝐶𝑗=10)
400 (𝜎𝐷𝐶𝑗=20)
N.A.
Fo ecas ing pe iod (𝜏𝑖𝑗)
5
15
15
Lead ime a e age (𝜇𝐿𝑖𝑗)
2
4
2
In en o y policy (IPij)
S1
S2
S1
Gene al model pa ame e s
Value
Simula ion pa ame e s
Value
Demand a e age (𝜇𝐷𝐶𝑗)
50
Simula ion ime (T)
4000
Lead ime c. . (𝜎𝐿𝑖𝑗/𝜇𝐿𝑖𝑗)
0.50
Wa m-up
1000
Re iew pe iod (R)
1
Numbe o eplica ions
20
Sa e y ac o (z)
2
Pe o mance Me ics
Echelon posi ion (i)
i=1…4
BwSl
In Sl
SysIn A
Node posi ion in echelon i (j)
j=1 ∀i<4
j=1…4 ∀i=4
IS (𝛿𝑖𝑗)
𝛿𝑖𝑗=0 ∀i<4
𝛿𝑖𝑗=0,1 ∀i=4

Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
17
Ups eam membe s o he SC (i.e., Wholesale , Dis ibu o and Fac o y) a e no subjec
o analysis in his wo k. Hence we simpli y he DoE by main aining he ope a ional
con igu a ion o hese membe s ixed in all expe imen s (Table 2).
The pa ame e s o he gene al model –summa ised in Table 2– a e chosen as usual
alues used in SC dynamics li e a u e (see, e.g., Cha ield 2013, Cha ield and P i cha d
2013, Cos an ino e al. 2014, Dominguez e al. 2015a). The alue o he sa e y ac o
(z=2) co esponds wi h a cus ome se ice le el o 97.72% when using he no mal
app oxima ion.
In o de o adap he model p esen ed in Sec ion 2.1 o he di e gen SC unde s udy
(Figu e 1), we es ablish he bounda ies o 𝛿𝑖𝑗 and subsc ip s i and j, as in Table 2.
3.2 Simula ion pa ame e s
Unce ain y is inhe en o many o he SC’s p ocesses (Heckmann e al. 2015). In o de
o accoun o andomness, mul iple eplica ions o he expe imen s we e pe o med,
and he simula ion ou pu s we e s a is ically analysed. Acco ding o Kel on e al. (2007),
when he hal -wid h o con idence in e al is smalle han a use -speci ied alue (e.g.
wi hin 10% o he mean, Yang e al. 2011), he numbe o eplica ions is accep able o
s a is ical analysis. As sugges ed by hese au ho s, simula ions we e i s conduc ed wi h
10 eplica ions. Due o he use o sys emic pe o mance me ics (see Sec ion 3.3), we
ob ained esul s wi h e y low a iances, and hus he hal wid h was below 10% o he
a e age in all cases. E en hough, in o de o inc ease p ecision o esul s, we ha e
pe o med 20 eplica ions o each expe imen (see e.g. Nai and Vidal 2011, Yang e al.
2011).
To al simula ion ime (T) was se o 4,000 pe iods o ensu e ha a s eady s a e o he
sys em is eached. Also, he i s 1,000 pe iods we e emo ed om he esul s, as a
wa m-up ime, o elimina e sys em’s ini ializa ion e ec s.
3.3 Pe o mance me ics
In o de o cap u e he dynamics o he SC, we adop a s uc u ed non- inancial
pe o mance measu emen sys em, gi en by h ee common me ics, namely: O de
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
18
Va iance Ra io, In en o y Va iance Ra io and In en o y A e age (see e.g. Cannella e
al. 2013, Cos an ino e al. 2014, Wang and Disney 2016, among o he s). These me ics
a e compu ed a echelon’s le el. Due o he high numbe o SCs ha esul om he
DoE (see Sec ion 3.4), we ocus ins ead on he global pe o mance o he SC, allowing
o an easy compa ison among he di e en SCs (Cannella e al. 2017). To do so, we
use sys emic me ics (i.e., SC-le el me ics), which a e compu ed om hei
co esponding echelon’s me ics, i.e., Bullwhip Slope, In en o y Slope, and Sys emic
In en o y A e age, espec i ely. A educ ion o his se o me ics e lec s imp o ed
cos e ec i eness o membe s’ ope a ions. They p o ide a comp ehensi e and
s uc u ed assessmen o he in e nal p ocess e iciency o he SC a sys emic le el and
p o ide in o ma ion on he po en ial bene i s o pa ne ships, collabo a ion and
in o ma ion p oduc i i y o SC membe s (Cannella e al. 2013). A de ailed desc ip ion
o each me ic is p o ided below.
3.3.1 O de Ra e Va iance Ra io - Bullwhip Slope
A echelon’s le el, O de Ra e Va iance Ra io (𝑂𝑅𝑉𝑟𝑅𝑖) accoun s o o de a iance
ampli ica ion ups eam in he SC. In he long- e m un i is compu ed as 𝑂𝑅𝑉𝑟𝑅𝑖=
𝑠𝑂𝑖𝑇
2/𝑠𝐷𝐶
𝑇
2 (Chen e al. 2000, Cha ield e al. 2004, Dejonckhee e e al. 2004). In o de o
apply his me ic o a di e gen SC, we use agg ega e measu es o each echelon
(Dominguez e al. 2015b). The e o e, assuming ha all cus ome s’ demands a e
independen and ha each node places o de s independen ly, we can agg ega e o de
a iances a each echelon and hus 𝑂𝑅𝑉𝑟𝑅𝑖 o a di e gen SC can be w i en as in
Equa ion (9):
𝑂𝑅𝑉𝑟𝑅𝑖=∑𝑠𝑂𝑖𝑗
𝑇
2
𝑁𝑖
𝑗=1
∑𝑠𝐷𝐶𝑗
𝑇
2
𝑁𝐶
𝑗=1
(9)
A sys em’s le el we use he Bullwhip Slope (BwSl) (Cannella e al. 2013, Dominguez
e al. 2015b). BwSl is compu ed as he slope o he linea in e pola ion o he se o
𝑂𝑅𝑉𝑟𝑅𝑖 alues o a gi en SC (Equa ion (10)), whe e 𝜋𝑖 is he posi ion o he i- h
echelon in Dejonckhee e’s e al. cu e. This me ic measu es he magni ude o he
bullwhip p opaga ion ac oss he SC and allows o a concise and holis ic compa ison
be ween di e en SCs. A high alue o BwSl indica es a as p opaga ion o he
bullwhip e ec h ough he SC, whe eas a low alue indica es a smoo h p opaga ion.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
19
BwSl p o ides in o ma ion on po en ial unnecessa y cos s o supplie s, such as los
capaci y o oppo uni y cos s, and on all o he unexpec ed cos s gene a ed by he
bullwhip e ec (Cannella e al. 2013, T ape o and Ped egal 2016).
𝐵𝑤𝑆𝑙=𝑡𝑔𝜗𝑂𝑅𝑉𝑟𝑅 =𝐸∑𝜋𝑖𝑂𝑅𝑉𝑟𝑅𝑖−
𝐸
𝑖=1 ∑𝜋𝑖
𝐸
𝑖=1 ∑𝑂𝑅𝑉𝑟𝑅𝑖
𝐸
𝑖=1
𝐸∑𝜋𝑖2𝐸
𝑖=1 −(∑ 𝜋𝑖
𝐸
𝑖=1 )2
(10)
3.3.2 In en o y Va iance Ra io - In en o y Slope
A echelon’s le el, he In en o y Va iance Ra io (𝐼𝑛𝑣𝑉𝑟𝑅𝑖) (Disney and Towill 2003),
assesses he s abili y deg ee o he in en o y and i can be associa ed wi h he a ia ion
and he po en ial inc emen o he holding cos s pe uni (Cannella e al. 2015). I is
compu ed as he a io be ween he in en o y a iance a echelon i and he cus ome
demand a iance: 𝐼𝑛𝑣𝑉𝑟𝑅𝑖=(𝑠𝐼𝑖𝑇
2/𝐼𝑖𝑇)/(𝑠𝐷𝐶
𝑇
2/𝐷
𝐶
𝑇). Following he same p ocedu e as
wi h 𝑂𝑅𝑉𝑟𝑅𝑖, we de i e 𝐼𝑛𝑣𝑉𝑟𝑅𝑖 o a di e gen SC, esul ing he exp ession shown in
Equa ion (11).
𝐼𝑛𝑣𝑉𝑟𝑅𝑖=∑𝑠𝐼𝑖𝑗
𝑇
2
𝑁𝑖
𝑗=1 /∑𝐼𝑖𝑗
𝑇
𝑁𝑖
𝑗=1
∑𝑠𝐷𝐶𝑗
𝑇
2
𝑁𝐶
𝑗=1 /∑𝐷
𝐶𝑗
𝑇
𝑁𝐶
𝑗=1
(11)
A sys em’s le el we use he In en o y Slope (In Sl) (Cannella e al. 2013). This me ic
is simila o BwSl (Equa ion (12)), bu accoun s o in en o y ins abili y p opaga ion
ac oss he SC. An inc eased In Sl esul s in highe holding and backlog cos s, in la ing
he a e age in en o y cos s pe pe iod (Disney and Lamb ech 2008), inc easing holding
uni cos s, missing p oduc ion schedules, job sequencing and esou ce e-alloca ion
(Cannella e al. 2013, Duong e al. 2015).
𝐼𝑛𝑣𝑆𝑙=𝑡𝑔𝜗𝐼𝑛𝑣𝑉𝑟𝑅 =𝐸∑𝜋𝑖𝐼𝑛𝑣𝑉𝑟𝑅𝑖−
𝐸
𝑖=1 ∑𝜋𝑖
𝐸
𝑖=1 ∑𝐼𝑛𝑣𝑉𝑟𝑅𝑖
𝐸
𝑖=1
𝐸∑𝜋𝑖2𝐸
𝑖=1 −(∑ 𝜋𝑖
𝐸
𝑖=1 )2
(12)
3.3.3 In en o y A e age - Sys emic In en o y A e age
A echelon’s le el, In en o y A e age (𝐼𝑛𝑣𝐴𝑣𝑖) can be associa ed o he a e age
holding cos o e he obse a ion ime (Cannella e al. 2013), and i is commonly used
in p oduc ion-dis ibu ion sys ems analysis o assess concise in o ma ion on in en o y
in es men (Cannella and Ciancimino 2010, Ganesh e al. 2014a). I can be iewed as a
me ic complemen a y o 𝐼𝑛𝑣𝑉𝑟𝑅𝑖. Fo a di e gen SC his me ic can be exp essed as
ollows:
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
20
𝐼𝑛𝑣𝐴𝑣𝑖=∑ ∑ 𝐼𝑖𝑗
𝑡
𝑇
𝑡=1
𝑁𝑖
𝑗=1 𝑇
(13)
A sys em’s le el we use he Sys emic In en o y A e age (SysIn A ) (Cannella e al.
2013). This me ic accoun s o he a e age in en o y o he whole SC. As i is common
o model holding cos s as linea ly dependen om s ock le els (Sha ma 2010, Cachon
and Oli a es 2010), his me ic quan i ies he a e age holding cos s ac oss he
obse a ion ime (Cannella e al. 2013). Since all SCs unde analysis ha e he same
numbe o nodes, we can use he ollowing exp ession:
𝑆𝑦𝑠𝐼𝑛𝑣𝐴𝑣=∑ ∑ ∑ 𝐼𝑖𝑗
𝑡
𝑇
𝑡=1
𝑁𝑖
𝑗=1
𝐸
𝑖=1 𝑇
(14)
3.4 Expe imen s
We pe o m wo se s o expe imen s. In he i s one we assume homogeneous e aile s,
and in ends o assess he con ibu ion o each e aile in ol ed in IS on imp o ing SC
pe o mance when all o hem ha e iden ical ope a ional con igu a ions. In o de o
inc ease he gene ali y o esul s, we conside a wide ange o possible ope a ional
con igu a ions o he e aile s by analysing he ull ac o ial se o he OFs. Since each
OF has wo le els, we analyse 24 di e en e aile s’ ope a ional con igu a ions. Then,
each e aile s’ ope a ional con igu a ion is e alua ed unde i e IS s uc u es: (1) no IS
(NIS), (2) 1 e aile sha es in o ma ion (1 e IS), (3) 2 e aile s sha e in o ma ion
(2 e IS), (4) 3 e aile s sha e in o ma ion (3 e IS), and (5) 4 e aile s sha e in o ma ion
(FIS). The e o e, we analyse a o al o 5x24=80 SCs in his se o expe imen s.
The second se o expe imen s is pe o med unde he hypo hesis o he e ogeneous
e aile s, and in ends o assess he con ibu ion o each e aile in ol ed in IS on
imp o ing SC pe o mance when hey ha e di e en ope a ional con igu a ions, and
how e aile s’ OFs may in luence o hei con ibu ion. To his aim, o each o he ou
OFs, we model a se o SCs whe e he e a e wo g oups o wo e aile s. The wo
e aile s in each g oup ha e he same OF alue (OF=OF* om now on), bu he OF is
di e en among he wo g oups. Mo e speci ically, he i s pai o e aile s ha e he
OFL alue and he second pai ha e he OFH alue (e.g. i OF*=𝜎𝐷𝐶𝑗
2, hen he i s pai
o e aile s will ha e 𝜎𝐷𝐶𝑗
2=100 and he second pai o e aile s will ha e 𝜎𝐷𝐶𝑗
2=400). The
o he h ee OFs emain he same o all he e aile s. To inc ease he gene ali y o he
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
21
esul s, o a gi en OF=OF* we analyse he ull ac o ial se o he o he h ee OFs
(OF≠OF*). The e o e, we ha e a o al o 4 (OF=OF*) x 23 (OF≠OF*) = 32 e aile s’
ope a ional con igu a ions. Finally, each e aile s’ ope a ional con igu a ion is e alua ed
unde ou IS s uc u es: (1) NIS, (2) i s pai o e aile s sha e in o ma ion ( e e ed as
OFLIS), (3) second pai o e aile s sha e in o ma ion ( e e ed as OFHIS), and (4) FIS.
We analyse a o al o 4x32=128 SCs in his se o expe imen s. A summa y o he DoE
is p esen ed in Table 3.
Table 3. Summa y o expe imen s (DoE).
Full ac o ial se o he OFs
Re aile s’
ope a ion
al
con igu a
ions
IS
s uc u es
Analysed
SCs
Pe o mance
Me ics
Homogeneous
Re aile s
𝜎𝐷𝐶𝑗
2(OFL, OFH)
𝜏𝑖𝑗 (OFL, OFH)
𝜇𝐿𝑖𝑗(OFL, OFH)
IPij(OFL, OFH)
24=16
NIS
1 e IS
2 e IS
3 e IS
FIS
5x24=80
BwSl
In Sl
SysIn A
(OF=OF*)
(2 e aile s - OFL,
2 e aile s - OFH)
Full ac o ial se o
OF≠OF*
(same o all e aile s)
He e ogeneous Re aile s
𝜎𝐷𝐶𝑗
2
[𝜏𝑖𝑗(OFL, OFH),
𝜇𝐿𝑖𝑗(OFL, OFH),
IPij(OFL, OFH)]=
=23=8
4x23=32
NIS
4x32=128
BwSl
In Sl
SysIn A
𝜏𝑖𝑗
[𝜎𝐷𝐶𝑗
2(OFL, OFH),
𝜇𝐿𝑖𝑗(OFL, OFH),
IPij(OFL, OFH)]=
=23=8
2 e IS
(OFLIS)
𝜇𝐿𝑖𝑗
[𝜎𝐷𝐶𝑗
2(OFL, OFH),
𝜏𝑖𝑗(OFL, OFH),
IPij(OFL, OFH)]=
=23=8
2 e IS
(OFHIS)
IPij
[𝜎𝐷𝐶𝑗
2(OFL, OFH),
𝜏𝑖𝑗(OFL, OFH),
𝜇𝐿𝑖𝑗(OFL, OFH)]=
=23=8
FIS
The simula ions we e pe o med on an In el Co e 2 Duo P8600 2.40GHz compu e wi h
2GB RAM. The e ec i e simula ion ime was 14 hou s and 49 minu es o he se o
homogeneous e aile s (1,600 simula ion uns), and 26 hou s and 4 minu es o he se
o he e ogeneous e aile s (2,560 simula ion uns).

Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
22
4 ANALYSIS OF RESULTS
This sec ion p esen s he esul s ob ained by he simula ions pe o med wi h SCOPE
acco ding o he DoE p esen ed in Sec ion 3. We also de i e meaning ul indings on he
implemen a ion o IS on a SC wi h se e al e aile s.
4.1 Homogeneous e aile s
He ein we p esen he esul s ob ained o he se o expe imen s unde he hypo hesis o
homogeneous e aile s. Table 4 shows a legend, labelling he 16 e aile s’ ope a ional
con igu a ions om #1 o #16. Fo each e aile s’ ope a ional con igu a ion, he SC is
analysed unde i e IS s uc u es (Table 3). The me ics ob ained om all scena ios a e
a e aged o e he 20 eplica ions, and esul s a e plo ed in Figu e 3. Fo cla i y, esul s
ob ained o each pe o mance me ic a e di ided in 4 plo s. Also, hey a e displayed
om he highes alue o he me ic o he lowes alue o he me ic.
Table 4. Re aile s’ ope a ional con igu a ions.
H= OFH
L= OFL
#1
#2
#3
#4
#5
#6
#7
#8
#9
#10
#11
#12
#13
#14
#15
#16
𝜇𝐿𝑖𝑗
H
H
H
H
H
H
H
H
L
L
L
L
L
L
L
L
IPij
H
H
H
H
L
L
L
L
H
H
H
H
L
L
L
L
𝜏𝑖𝑗
H
H
L
L
H
H
L
L
H
H
L
L
H
H
L
L
𝜎𝐷𝐶𝑗
2
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
Due o he use o sys em’s me ics, he esul s ob ained o e he 20 eplica ions a e e y
close o he a e age wi h e y low a iances. To ensu e he signi icance o esul s
ob ained we pe o med an ANOVA o each me ic and each e aile s’ ope a ional
con igu a ion. All es s we e signi ican a he 95% con idence le el.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
23
BwSl
In Sl
SysIn A
Figu e 3. SC pe o mance unde he hypo hesis o homogeneous e aile s.
0
5
10
15
20
NIS 1 e IS 2 e IS 3 e IS FIS
#2
#6
#14
#1
0
2
4
6
8
10
12
NIS 1 e IS 2 e IS 3 e IS FIS
#10
#5
#8
#9
0
1
2
3
4
5
6
7
8
NIS 1 e IS 2 e IS 3 e IS FIS
#16
#4
#13
#3
0
1
2
3
4
5
NIS 1 e IS 2 e IS 3 e IS FIS
#12
#7
#11
#15
0
5
10
15
20
25
30
35
40
NIS 1 e IS 2 e IS 3 e IS FIS
#2
#10
#6
#4
0
5
10
15
20
25
NIS 1 e IS 2 e IS 3 e IS FIS
#14
#12
#8
#16
0
5
10
15
20
NIS 1 e IS 2 e IS 3 e IS FIS
#1
#9
#5
#3
0
2
4
6
8
10
NIS 1 e IS 2 e IS 3 e IS FIS
#13
#11
#7
#15
0
200
400
600
800
1000
1200
1400
1600
NIS 1 e IS 2 e IS 3 e IS FIS
#1
#5
#9
#3
0
100
200
300
400
500
600
700
800
NIS 1 e IS 2 e IS 3 e IS FIS
#2
#13
#11
#7
0
100
200
300
400
500
600
NIS 1 e IS 2 e IS 3 e IS FIS
#6
#10
#15
#4
0
50
100
150
200
250
300
350
400
450
NIS 1 e IS 2 e IS 3 e IS FIS
#14
#8
#16
#12
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
24
Resul s show a quasi-linea pe o mance imp o emen in BwSl om NIS o FIS, wi h all
cu es yielding a coe icien o de e mina ion (R2) o e 0.99. In ac , since e aile s a e
iden ical and ansmi demand in o ma ion, i is expec ed ha he impac o IS on
educing demand a iabili y would be linea wi h he numbe o e aile s. Howe e ,
cu es ela ed o he in en o y me ics (In Sl and SysIn A ) a e no s ic ly linea , wi h
67% o all cu es yielding a coe icien o de e mina ion o e 99%, and he es o he
cu es showing small de ia ions om linea i y, wi h 0.90< R2<0.99. This esul
sugges s ha he impac o ansmi ing demand in o ma ion on in en o y pe o mance
imp o emen is linea wi h he numbe o e aile s in mos cases, bu i may p esen
some non-linea i y.
Pe o mance cu es show di e en slopes depending on e aile s’ ope a ional
con igu a ions. Thus, bene i s o inco po a ing a e aile o IS may depend on cu en
e aile s’ ope a ional con igu a ion. In o de o app ecia e his phenomenon, we
compu e he pe cen age o pe o mance imp o emen o each me ic om NIS o 2 e IS
and om NIS o FIS o each o he 16 e aile s’ ope a ional con igu a ions and plo he
esul s in Figu e 4. A gene ic o mula ion o his measu e is shown in Equa ion (15),
whe e ‘me ic’ can be ei he BwSl, In Sl o SysIn A , and A,B ep esen any o he IS
s uc u es.
∆𝑚𝑒𝑡𝑟𝑖𝑐𝐴→𝐵(%)=(𝑚𝑒𝑡𝑟𝑖𝑐𝐴−𝑚𝑒𝑡𝑟𝑖𝑐𝐵)
𝑚𝑒𝑡𝑟𝑖𝑐𝐴∗100
(15)
F om Figu e 4 i can be seen ha he bene i s ob ained in e ms o BwSl educ ion a e
less dependen on e aile s’ ope a ional con igu a ion han hose ela ed o In Sl and
SysIn A . In ac , pe o mance imp o emen in e ms o BwSl is e y simila o all
scena ios. This esul indica es ha he expec ed bullwhip educ ion om adding
e aile s o IS weakly depends on e aile s’ ope a ion. Ne e heless, he pe o mance
imp o emen ela ed o In Sl and SysIn A show a s onge dependence on e aile s’
ope a ional con igu a ion. Addi ionally, BwSl educ ion is highe han In Sl and
SysIn A educ ions: he e is an a e age BwSl educ ion o a ound 20%-25% o 2 e IS
and a ound 40%-50% o FIS, while a e age In Sl and SysIn A educ ions a e a ound
5%-16% o 2 e IS and 10%-32% o FIS.
We can summa ize he abo e indings as ollows:
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
25
(1) The imp o emen in BwSl in a SC wi h homogenous e aile s ob ained by
in o ma ion sha ing is linea wi h he numbe o e aile s sha ing in o ma ion.
Ne e heless some (weak) non-linea i y appea s o In Sl and SysIn A me ics.
(2) The imp o emen in BwSl in a SC wi h homogeneous e aile s ob ained by
in o ma ion sha ing:
a. I is highe han o In Sl and SysIn A me ics.
b. I is less dependen on e aile s’ ope a ional con igu a ion han o In Sl
and SysIn A me ics.
Figu e 4. SC pe o mance imp o emen o all he e aile s’ ope a ional con igu a ions.
0
10
20
30
40
50
60 #1 #2
#3
#4
#5
#6
#7
#8
#9
#10
#11
#12
#13
#14
#15
#16
ΔBwSl_NIS->2 e IS(%)
ΔBwSl_NIS->FIS(%)
0
5
10
15
20
25
30 #1 #2
#3
#4
#5
#6
#7
#8
#9
#10
#11
#12
#13
#14
#15
#16
ΔIn Sl_NIS->2 e IS(%)
ΔIn Sl_NIS->FIS(%)
0
5
10
15
20
25
30
35 #1 #2
#3
#4
#5
#6
#7
#8
#9
#10
#11
#12
#13
#14
#15
#16
ΔSysIn A _NIS->2 e IS(%)
ΔSysIn A _NIS->FIS(%)
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
32
pe o mance imp o emen by pa ial IS. In he case o OF*=
IPij
i can be seen ha he
esul s a e highly dependen on e aile s’ ope a ional con igu a ion and i is no clea
which e aile s a e mo e a ou able.
OF*=𝜎𝐷𝐶𝑗
2
OF*=
IPij
OF*=𝜇𝐿𝑖𝑗
OF*=𝜏𝑖𝑗
Figu e 6. SC pe o mance inc ease (BwSl) unde pa ial IS and FIS o he e ogeneous e aile s.
4.2.1 Sensi i i y analysis on e aile s’ demand a iance and o ecas ing pe iod
In o de o enhance he simula ion models and o ex end he applicabili y o he esul s
ob ained, we pe o m a sensi i i y analysis (Kleijnen 2008) wi h espec o he mo e
ele an OFs (i.e., OF*=𝜎𝐷𝐶𝑗
2 and OF*=𝜏𝑖𝑗). Since he alues assumed by he OFs in
0
10
20
30
40
50
60 #1
#2
#3
#4
#5
#6
#7
#8
0
10
20
30
40
50
60 #1
#2
#3
#4
#5
#6
#7
#8
0
10
20
30
40
50 #1
#2
#3
#4
#5
#6
#7
#8
0
10
20
30
40
50
60 #1
#2
#3
#4
#5
#6
#7
#8

Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
33
Table 2 a e di e en in o de o ensu e a he e ogeneous scena io, in his u he analysis
we aim o add ess he ollowing ques ion: how much he esul s will change i he
di e ences be ween e aile s’ OFs a e educed? To do so, we analyse (1) wo new
a ian s o demand a iance, wi h he OFH alue educed o 𝜎󰇗𝐷𝐶𝑗=17.5 (𝜎󰇗𝐷𝐶𝑗
2=306.25)
in he i s a ian and o 𝜎󰇘𝐷𝐶𝑗=15 (𝜎󰇘𝐷𝐶𝑗
2=225) in he second a ian ; and (2) wo new
a ian s o he o ecas ing pe iod, wi h he OFL alue inc eased o 𝜏󰇗𝑖𝑗=7 in he i s
a ian and o 𝜏󰇘𝑖𝑗=9 in he second a ian . Fo each new a ian we analyse he ull
ac o ial combina ion o he OF≠OF*, which main ains he o iginal alues, as in Table
7. The e o e we analyse a o al o 4 ( a ian s) x 8 ( ull ac o ial OF≠OF*) x 4 (IS
s uc u es) = 128 SCs (2560 simula ion uns).
Following he same p ocedu e ca ied ou in Sec ion 4.2, we compu e Equa ions (16)
and (17) and show he a e age alues o each me ic in Table 8. As i could be
expec ed, he ad an ages o disad an ages ob ained om OFLIS o OFHIS s uc u es
(i.e., di e ences be ween ∆𝑚𝑒𝑡𝑟𝑖𝑐𝑂𝐹𝐿𝐼𝑆 𝐹𝐼𝑆
⁄(%) and ∆𝑚𝑒𝑡𝑟𝑖𝑐𝑂𝐹𝐻𝐼𝑆 𝐹𝐼𝑆
⁄(%)) a e lowe
as he di e ences be ween OFs dec ease.
Table 7. DoE o he sensi i i y analysis on 𝜎𝐷𝐶𝑗
2 and 𝜏𝑖𝑗.
OFs
𝜎𝐷𝐶𝑗
2 sensi i i y
𝜏𝑖𝑗 sensi i i y
1s a ian
2nd a ian
1s a ian
2nd a ian
𝑂𝐹𝐿
󰇗
𝑂𝐹𝐻
󰇗
𝑂𝐹𝐿
󰇘
𝑂𝐹𝐻
󰇘
𝑂𝐹𝐿
󰇗
𝑂𝐹𝐻
󰇗
𝑂𝐹𝐿
󰇘
𝑂𝐹𝐻
󰇘
𝜎𝐷𝐶𝑗
2
100
306.25
100
225
100
400
100
400
𝜏𝑖𝑗
5
15
5
15
7
15
9
15
𝜇𝐿𝑖𝑗
2
4
2
4
2
4
2
4
𝐼𝑃𝑖𝑗
S1
S2
S1
S2
S1
S2
S1
S2
Fo OF*=𝜎𝐷𝐶𝑗
2, di e ences be ween ∆𝑚𝑒𝑡𝑟𝑖𝑐𝑂𝐹𝐿𝐼𝑆 𝐹𝐼𝑆
⁄(%) and ∆𝑚𝑒𝑡𝑟𝑖𝑐𝑂𝐹𝐻𝐼𝑆 𝐹𝐼𝑆
⁄(%)
smoo hly dec ease as he OFH alue dec eases. In he i s a ian , whe e he c. . o he
demand aced by OFH e aile s changes om 0.4 o 0.35 and he c. . o demand aced
by OFL e aile s emains he same (i.e., c. .=0.2), he bene i s ob ained om OFHIS
s ill ep esen o e 70% o he bene i s o a FIS o he h ee me ics. In he second
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
34
a ian , whe e he c. . o he demand aced by OFH e aile s is educed o 0.30, bene i s
ob ained om OFHIS a e s ill signi ican ly highe han bene i s ob ained om OFLIS.
Fo OF*=𝜏𝑖𝑗, he di e ence be ween ∆𝑚𝑒𝑡𝑟𝑖𝑐𝑂𝐹𝐿𝐼𝑆 𝐹𝐼𝑆
⁄(%) and ∆𝑚𝑒𝑡𝑟𝑖𝑐𝑂𝐹𝐻𝐼𝑆 𝐹𝐼𝑆
⁄(%)
dec eases as he OFL alue inc eases. I is known ha a high alue o 𝜏𝑖𝑗 p oduces a
mo e s able o ecas , while a low alue o 𝜏𝑖𝑗 p oduces a mo e ne ous o ecas .
The e o e, as 𝜏𝑖𝑗 inc eases o he OFL e aile s, o ecas pa e ns o bo h pai s o
e aile s become mo e aligned, and he ad an ages ob ained by choosing he OFLIS
s uc u e a e consequen ly educed.
Table 8. Resul s o he sensi i i y analysis.
OF*=𝜎𝐷𝐶𝑗
2
OF*=𝜎󰇗𝐷𝐶𝑗
2
OF*=𝜎󰇘𝐷𝐶𝑗
2
OF*=𝜏𝑖𝑗
OF*=𝜏󰇗𝑖𝑗
OF*=𝜏󰇘𝑖𝑗
∆𝐵𝑤𝑆𝑙𝑂𝐹𝐿𝐼𝑆 𝐹𝐼𝑆
⁄(%)
24,88
31,02
37,88
71,71
65,47
58,26
∆𝐵𝑤𝑆𝑙𝑂𝐹𝐻𝐼𝑆 𝐹𝐼𝑆
⁄(%)
74,24
71,18
65,40
31,94
38,80
49,75
∆𝐼𝑛𝑣𝑆𝑙𝑂𝐹𝐿𝐼𝑆/𝐹𝐼𝑆(%)
21,94
24,92
29,75
77,60
74,91
66,81
∆𝐼𝑛𝑣𝑆𝑙𝑂𝐹𝐻𝐼𝑆/𝐹𝐼𝑆(%)
85,62
71,46
56,20
26,03
30,53
46,56
∆𝑆𝑦𝑠𝐼𝑛𝑣𝐴𝑣𝑂𝐹𝐿𝐼𝑆/𝐹𝐼𝑆(%)
20,33
27,91
35,35
70,71
63,79
55,06
∆𝑆𝑦𝑠𝐼𝑛𝑣𝐴𝑣𝑂𝐹𝐻𝐼𝑆/𝐹𝐼𝑆(%)
73,79
72,33
65,31
26,36
35,45
45,85
5 DISCUSSION AND MANAGERIAL INSIGHTS
In his sec ion we discuss he manage ial implica ions de i ed om ou wo k. We ocus
on how SC mange s may success ully implemen IS a e aile s’ s age, since i epo s
highe bene i s o he SC (Lau e al. 2004, Ganesh e al. 2014a, Cos an ino e al. 2014).
In his way we p o ide p ac ical insigh s o he es ima ion o he po en ial alue o each
e aile (i.e., es ima ing he po en ial con ibu ion o each e aile i hey join IS o
imp o e SC pe o mance).
The e a e wo possible app oaches when implemen ing IS: a FIS app oach, o a pa ial
IS app oach. The o me always esul s in a highe imp o emen o SC pe o mance
han he la e , as we ha e seen in Sec ion 4. Howe e , implemen ing IS in a SC cos s
ime (nego ia ions and physical ins alla ion o IT) and cash (IT is expensi e), and
e aile s may ask o a la ge discoun o sha e hei in o ma ion (Huang and I a ani
2005). The e o e, in e ms o he ne bene i s ou wo k highligh s he need o conside
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
35
e aile s’ ope a ional cha ac e is ics in o de o decide on which ype o IS could be
adop ed.
I e aile s a e simila in e ms o a iance o cus ome demand, in en o y policy,
o ecas ing pe iod and lead ime a e age, he po en ial alue o all o hem is also
simila when hey a e in ol ed in IS. The e o e, a FIS app oach should be pu sued,
since bene i s inc ease wi h he numbe o e aile s in ol ed. Addi ionally, since all
e aile s ha e simila po en ial alue, manage s may s a nego ia ions wi h hose ha
a e mo e p one o collabo a e.
In case ha a e aile , o g oup o e aile s, signi ican ly di e s om he o he s in one o
mo e o he a o emen ioned OFs, he bene i s achie ed by pa ial IS may signi ican ly
depend on he pa icipan e aile /s. As shown in Sec ion 4, a pa ial IS s uc u e is able
o achie e a signi ican pa o he o al bene i s ob ained by FIS i e aile s a e
signi ican ly di e en (e.g., we ound ha in ol ing hal o he o al numbe o e aile s
in o IS may lead o ob ain o e 70% o he o al bene i s o a FIS unde he bounda y
condi ions). Assuming a linea inc ease o cos s wi h he numbe o e aile s in ol ed in
IS, a cos -bene i analysis may e eal ha a pa ial IS app oach is mo e bene icial o
he SC han a FIS app oach, hus sa ing cos s ela ed o he in ol emen o addi ional
e aile s. On he o he hand, an e oneous choice o he pa ial IS s uc u e may esul in
a e y low pe o mance inc ease, unde mining all e o s and in es men s.
Consequen ly, a p io e alua ion o he po en ial alue o e aile s may help manage s o
e icien ly selec a pa ial IS s uc u e, cons i u ed by he mos bene icial e aile s. To do
so, e aile s should be e alua ed in his o de o impo ance: (1) (highe ) demand
a iance; (2) (lowe ) o ecas ing pe iod; (3) (highe ) a e age lead imes. Na u ally,
hese esul s a e less signi ican as he di e ences be ween e aile s’ OFs dec ease.
Once he pa ial IS s uc u e o be adop ed has been decided, manage s may s a
implemen ing IS acco ding o e aile s’ po en ial alue. By doing so, he bene i s
ob ained by each new e aile in ol ed in IS a e maximal and hus an e icien
implemen a ion o IS can be achie ed.
E en hough i has been shown ha conside ing e aile s’ ope a ion du ing he
implemen a ion o pa ial IS may p o ide impo an bene i s o he SC, ob aining such
in o ma ion may p esen some di icul ies. Only he lead ime a e age o each e aile
could be accessed ( h ough he wholesale ). Howe e , esul s ob ained in his pape
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
36
show ha e aile s’ demand a iance and e aile s’ o ecas ing pe iod a e he mos
signi ican OFs. Since hese ac o s a e e aile s’ p i a e in o ma ion, a p e-
collabo a ion s a egy o sha e hese da a needs o be de eloped wi h e aile s p io o
he implemen a ion o pa ial IS. In his case, SC manage s should s a by de eloping
channels o us and/o e enue con ac s.
To sum up, we sugges manage s o implemen pa ial IS a he e ogeneous e aile s
using he ollowing s eps:
1. Analyse e aile s’ ope a ional cha ac e is ics.
2. I hey a e signi ican ly di e en in one o mo e OFs, es ima e he po en ial
alue o each e aile s’ collabo a ion in IS and ank hem acco dingly.
3. Es ima e cos s o in ol ing e aile s in o IS and pe o m a cos /bene i analysis.
4. Decide he bes IS s uc u e using esul s om 3) and p oceed in ol ing
e aile s acco ding o 2).
6 CONCLUSIONS AND FUTURE RESEARCH
This wo k p esen s an explo a o y s udy on pa ial in o ma ion sha ing a e aile s le el,
i.e., some e aile s may no pa icipa e in in o ma ion sha ing. We analyse he po en ial
con ibu ion o he pa icipa ion o each indi idual e aile in in o ma ion sha ing unde
wo di e en hypo hesis: (1) e aile s a e homogeneous (i.e., hey ha e iden ical
ope a ional con igu a ion), and (2) e aile s a e he e ogeneous (i.e., hey ha e di e en
ope a ional con igu a ion). Using a Mul i-Agen Sys ems simula ion app oach, we
model a ou echelon supply chain wi h ou e aile s, wi h s ochas ic demands and lead
imes, using wo common O de -Up-To in en o y policies. Re aile s’ ope a ion is
cha ac e ized by ou ope a ional ac o s: demand a iance, o ecas ing pe iod, lead
ime a e age, and in en o y policy. We measu e he supply chain pe o mance using
sys emic supply chain me ics: Bullwhip Slope, In en o y Slope and Sys emic
In en o y A e age. Supply chain pe o mance is measu ed o di e en pa ial
in o ma ion sha ing s uc u es and di e en e aile s’ ope a ional con igu a ions.
The esul s o ou s udy emphasize he need o indi idually es ima ing he po en ial
alue o e aile s’ in o ma ion p io o he implemen a ion o in o ma ion sha ing, and
p o ides he ollowing insigh s:
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
37
- When e aile s a e homogeneous hey ha e equal po en ial alue i hey a e
in ol ed in in o ma ion sha ing (i.e., hey iden ically con ibu e o imp o e
supply chain pe o mance). Thus, a ull in o ma ion sha ing app oach is
ecommended.
- When e aile s a e he e ogeneous hey ha e di e en po en ial alue i hey a e
in ol ed in in o ma ion sha ing, depending on hei ope a ional con igu a ion.
As a consequence
o The pe o mance imp o emen achie ed by di e en pa ial in o ma ion
sha ing s uc u es wi h he same numbe o e aile s migh be
signi ican ly di e en .
o A pa ial in o ma ion sha ing s uc u e in ol ing e aile s wi h high
po en ial alue may cap u e a subs an ial pa o he bene i s o ull
in o ma ion sha ing.
o Assuming a linea inc ease o cos s wi h he numbe o e aile s in ol ed
in in o ma ion sha ing, a cos -bene i analysis may e eal ha a pa ial
in o ma ion sha ing app oach is mo e bene icial o he supply chain
han a ull in o ma ion sha ing app oach, hus sa ing cos s ela ed o he
in ol emen o addi ional e aile s.
- Re aile s’ ope a ion need o be ca e ully examined in o de o de elop an
e icien implemen a ion o in o ma ion sha ing. In his o de o impo ance,
e aile s wi h (1) highe demand a iance, (2) lowe o ecas ing pe iod, and (3)
highe a e age lead ime, a e po en ially mo e bene icial pa ne s o
implemen ing in o ma ion sha ing.
Due o he complex ela ionships be ween e aile s’ ope a ional ac o s and e aile s’
po en ial alue when become pa icipan s o in o ma ion sha ing, i was no possible o
come up wi h a single and p ecise ule o iden i ying he mos bene icial in o ma ion
sha ing s uc u e. In ac , p ope ly balancing each ope a ional ac o is s ill an issue.
Ne e heless, he indings epo ed in his pape should help manage s o be e
unde s and he oppo uni ies o pa ial in o ma ion sha ing and pu hem in a s onge
posi ion in hei nego ia ions abou es ablishing in o ma ion sha ing links.
The p esen s udy has some limi a ions ha may c ea e oom o imp o emen and
u he esea ch. Also, due o he explo a o y na u e o his wo k, he e a e many ways
o possible ex ensions:

Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
38
- Deepening he analysis o each ope a ional ac o by inc easing he numbe o
in e media e alues, and conside ing o he di e en se ups o supply chain (i.e.,
di e en demand o lead ime dis ibu ions, di e en o ecas me hods, e c.)
would p o ide addi ional esul s ha migh be use ul o p ecisely balance he
impo ance o each ope a ional ac o on a wide a ie y o condi ions and o look
o a single ule o choosing he bes in o ma ion sha ing s uc u e.
- Analysing scena ios whe e e aile s may di e in mo e han one ac o a he
same ime would p o ide mo e ealis ic esul s.
- This wo k analyses ei he homogeneous o he e ogeneous e aile s. The “g ey
zone” ha alls in he middle o bo h scena ios has been b ie ly analysed h ough
a sensi i i y analysis. De e mining he limi s be ween bo h scena ios would be a
signi ican con ibu ion in his line o esea ch.
- A simila analysis o ha conduc ed in his wo k on o he ope a ional ac o s
(e.g. lead ime a iance, o ecas me hod, sa e y ac o , e c.) would p o ide a
wide pe spec i e o his p oblem o supply chain manage s.
- Resul s o his wo k a e scalable o highe o lowe numbe o e aile s.
Howe e , i could be in e es ing o analyse how he ups eam pa o he supply
chain may impac on he esul s ob ained. Mo e speci ically, i should be
add essed how he ups eam supply chain s uc u e and ups eam membe ’s
ope a ional con igu a ion may impac on he implemen a ion o in o ma ion
sha ing a e aile s. Addi ionally, a simila esea ch o ha p esen ed in his wo k
could be pe o med on he ups eam echelons o he supply chain, in o de o
come up wi h a mo e gene al o e iew on how o e icien ly implemen
in o ma ion sha ing in supply chain.
ACKNOWLEDGEMENTS
This esea ch was suppo ed by he Po uguese Founda ion o Science and Technology
[G an SFRH/BPD/108491/2015], by he I alian Minis y o Educa ion, Uni e si y and
Resea ch (Ri a Le i Mon alcini ellow), and by he Spanish Minis y o Science and
Inno a ion, unde he p ojec PROMISE wi h e e ence DPI201680750P.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
39
REFERENCES
Ag awal, S., Sengup a, R.N., Shanke , K. 2009. Impac o in o ma ion sha ing and lead ime on bullwhip
e ec and on-hand in en o y. Eu opean Jou nal o Ope a ional Resea ch, 192(2), 576-593.
Ali, M. M., Babai, M. Z., Boylan, J. E., Syn e os, A. A. 2017. Supply chain o ecas ing when in o ma ion
is no sha ed. Eu opean Jou nal o Ope a ional Resea ch, 260(3), 984–994.
Al omon e, C., Mau o, F. D., O a iano, G., Rungi, A., Vica d, V. 2012 Global alue chains du ing he
g ea ade collapse: a bullwhip e ec ?. Eu opean Cen al Bank. Wo king Pape Se ies 1412.
Anna elli, A., Nonino, F. 2016. S a egic and ope a ional managemen o o ganiza ional esilience:
Cu en s a e o esea ch and u u e di ec ions. Omega, 62, 1-18.
Bian, W., Shang, J., Zhang, J. 2016. Two-way in o ma ion sha ing unde supply chain compe i ion.
In e na ional Jou nal o P oduc ion Economics, 178, 82-94.
Bischak, D.P., Robb, D.J., Sil e , E.A., Blackbu n, J.D. 2014. Analysis and managemen o pe iodic
e iew, O de -Up-To le el in en o y sys ems wi h o de c osso e . P oduc ion and Ope a ions
Managemen , 23 (5), 762-772.
Cachon, G.P., Oli a es, M. 2010. D i e s o inished-goods in en o y in he U.S. au omobile indus y.
Managemen Science, 56 (1), 202-216.
Cannella, S., Ba bosa-Po oa, A.P., F aminan J.M., Rel as S. 2013. Me ics o bullwhip e ec analysis.
Jou nal o he Ope a ional Resea ch Socie y, 64, 1-16.
Cannella, S., Ciancimino, E. 2010. On he bullwhip a oidance phase: supply chain collabo a ion and
o de smoo hing. In e na ional Jou nal o P oduc ion Resea ch, 48 (22), 6739-6776.
Cannella, S., Dominguez, R., F aminan, J.M. 2017. In en o y eco d inaccu acy – The impac o
s uc u al complexi y and lead ime a iabili y. Omega, 68, 123-138.
Cannella, S., López-Campos, M., Dominguez, R., Ashaye i, J., Mi anda, P.A. 2015. A simula ion model
o a coo dina ed decen alized supply chain. In e na ional T ansac ions in Ope a ional Resea ch, 22
(4), 735-756.
Ca doso, S.R., Ba bosa-Pó oa, A.P., Rel as, S., No ais, A.Q. 2015. Resilience me ics in he assessmen
o complex supply-chains pe o mance ope a ing unde demand unce ain y. Omega, 56, 53-73.
Chan, H.K., Chan, F.T.S. 2010. A Re iew o Coo dina ion S udies in he Con ex o Supply Chain
Dynamics. In e na ional Jou nal o P oduc ion Resea ch, 48 (10), 2793-2819.
Chase, R. B., Jacobs, F. R., Aquilano, N. J. 2004. Ope a ions managemen o compe i i e ad an age
(10 h ed.). Bos on, MA: I win/McG aw-Hill.
Cha ield, D. C., Ha ison, T. P., Hayya, J. C. 2006. SISCO: An objec -o ien ed supply chain simula ion
sys em. Decision Suppo Sys ems, 42(1), 422-434.
Cha ield, D.C. 2013. Unde es ima ing he bullwhip e ec : A simula ion s udy o he decomposabili y
assump ion. In e na ional Jou nal o P oduc ion Resea ch, 51 (1), 230-244.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
40
Cha ield, D.C., Hayya, J.C., Cook, D.P. 2013. S ockou p opaga ion and ampli ica ion in supply chain
in en o y sys ems. In e na ional Jou nal o P oduc ion Resea ch, 51 (5), 1491-1507.
Cha ield, D.C., Kim, J.G., Ha ison, T.P., Hayya, J.C. 2004. The bullwhip e ec - Impac o s ochas ic
lead ime, in o ma ion quali y, and in o ma ion sha ing: A simula ion s udy. P oduc ion and
Ope a ions Managemen , 13 (4), 340-353.
Cha ield, D.C., P i cha d, A.M. 2013. Re u ns and he bullwhip e ec . T anspo a ion Resea ch Pa E:
Logis ics and T anspo a ion Re iew, 49 (1), 159-175.
Chen, F., D ezne , Z., Ryan, J.K., Simchi-Le i, D. 2000. Quan i ying he bullwhip e ec in a simple
supply chain: The impac o o ecas ing, lead imes, and in o ma ion. Managemen Science, 46 (3),
436-443.
Chen, L., Lee, H.L. 2009. In o ma ion sha ing and o de a iabili y con ol unde a gene alized demand
model. Managemen Science, 55 (5), 781-798.
Choudha y, D., Shanka , R. 2015. The alue o VMI beyond in o ma ion sha ing in a single supplie
mul iple e aile s supply chain unde a non-s a iona y (Rn, Sn) policy. Omega, 51, 59-70.
Ch is ophe , M., & Holweg, M. 2017. Supply Chain 2.0 e isi ed: a amewo k o managing ola ili y-
induced isk in he supply chain. In e na ional Jou nal o Physical Dis ibu ion & Logis ics
Managemen , 47 (2), 2-17.
Cos an ino, F., Di G a io, G., Shaban, A., T onci, M. 2014. The impac o in o ma ion sha ing and
in en o y con ol coo dina ion on supply chain pe o mances. Compu e s and Indus ial Enginee ing,
76, 292-306.
Cos an ino, F., Di G a io, G., Shaban, A., T onci, M. 2015. The impac o in o ma ion sha ing on
o de ing policies o imp o e supply chain pe o mances. Compu e s & Indus ial Enginee ing, 82,
127-142.
C oson, R., Donohue, K., Ka ok, E., S e man, J. 2014. O de s abili y in supply chains: coo dina ion isk
and he ole o coo dina ion s ock. P oduc ion and Ope a ions Managemen , 23 (2), 176-196.
Da a, P. P., Ch is ophe , M. G. 2011. In o ma ion sha ing and coo dina ion mechanisms o managing
unce ain y in supply chains: a simula ion s udy. In e na ional Jou nal o P oduc ion Resea ch, 49(3),
765-803.
Dejonckhee e, J., Disney, S. M., Lamb ech , M. R., Towill, D. R. 2004. The impac o in o ma ion
en ichmen on he Bullwhip e ec in supply chains: A con ol enginee ing pe spec i e. Eu opean
Jou nal o Ope a ional Resea ch, 153 (3), 727-750.
Disney, S.M., Lamb ech , M.R. 2008. On eplenishmen ules, o ecas ing, and he bullwhip e ec in
supply chains. Founda ions and T ends in Technology, In o ma ion and Ope a ions Managemen , 2
(1), 1–80.
Disney, S.M., Mal z, A., Wang, X., Wa bu on, R.D.H. 2016. In en o y managemen o s ochas ic lead
imes wi h o de c osso e s, Eu opean Jou nal o Ope a ional Resea ch, 248, 473-486.
Dominguez R., Cannella S., Pó oa A.P., F aminan J.M. 2017. In o ma ion sha ing in supply chains wi h
he e ogeneous e aile s. Omega. DOI: h ps://doi.o g/10.1016/j.omega.2017.08.005
41
Disney, S.M., Towill, D.R. 2003. On he bullwhip and in en o y a iance p oduced by an o de ing
policy. Omega – The In e na ional Jou nal o Managemen Science, 31 (3), 157-167.
Dominguez, R., Cannella, S., F aminan, J.M. 2014. On bullwhip-limi ing s a egies in di e gen supply
chain ne wo ks. Compu e s and Indus ial Enginee ing, 73 (1), 85-95.
Dominguez, R., Cannella, S., F aminan, J.M. 2015a. The impac o he supply chain s uc u e on bullwhip
e ec . Applied Ma hema ical Modelling, 39 (23-24), 7309-7325.
Dominguez, R., Cannella, S., F aminan, J.M. 2015b. On e u ns and ne wo k con igu a ion in supply
chain dynamics. T anspo a ion Resea ch Pa E: Logis ics and T anspo a ion Re iew, 73, 152-167.
Dominguez, R., F aminan, J.M. 2013. A decisión managemen ool: modelling he o de ul ilmen
p ocess by mul i-agen sys ems. In e na ional Jou nal o Managemen and Decision Making, 12 (3),
240-258.
Duan, Y., Yao, Y., Huo, J. 2015. Bullwhip e ec unde subs i u e p oduc s. Jou nal o Ope a ions
Managemen , 36, 75-89.
Duong, L.N., Wood, L.C., Wang, W.Y. 2015. A mul i-c i e ia in en o y managemen sys em o
pe ishable & subs i u able p oduc s. P ocedia Manu ac u ing, 2, 66-76.
E e s, P. T., & Wan, X. 2012. Sys ems analysis using simula ion. Jou nal o Business Logis ics, 33(2),
80-89.
Fawce , S.E., Wallin, C., All ed, C., Fawce , A.M., Magnan, G.M. 2011. In o ma ion echnology as an
enable o supply chain collabo a ion: A dynamic-capabili ies pe spec i e. Jou nal o Supply Chain
Managemen , 47 (1), 38-59.
Ganesh, M., Raghuna han, S., Rajend an, C. 2014a. Dis ibu ion and equi able sha ing o alue om
in o ma ion sha ing wi hin se ial supply chains. IEEE T ansac ions on Enginee ing Managemen , 61
(2), 6573365, 225-236.
Ganesh, M., Raghuna han, S., Rajend an, C. 2014b. The alue o in o ma ion sha ing in a mul i-p oduc ,
mul i-le el supply chain: Impac o p oduc subs i u ion, demand co ela ion, and pa ial in o ma ion
sha ing. Decision Suppo Sys ems, 58 (1), 79-94.
Geunes, J., Romeijn, H.E., Van Den Heu el, W. 2016. Imp o ing he e iciency o decen alized supply
chains wi h ixed o de ing cos s. Eu opean Jou nal o Ope a ional Resea ch, 252 (3), 815-828.
GMA - G oce y Manu ac u e s Associa ion. 2009. Re ail-di ec da a epo . Repo , GMA, Washing on,
DC, h p://www.gmaonline .o g/downloads/ esea ch-and- epo s/WP-Re aile -DDR09-6.pd .
Gue le , B., Spinle , S. 2015. When does ope a ional isk cause supply chain en e p ises o ip? A
simula ion o in a-o ganiza ional dynamics. Omega, 57, 54-69.
Hayya, J.C., Bagchi, U., Kim, J.G., Sun, D. 2008. On s a ic s ochas ic o de c osso e . In e na ional
Jou nal o P oduc ion Economics, 114 (1), 404-413.