Choi, Sunghee
A icle
Es ima ing spillo e e ec om in e na ional oil ma ke
o s ock ma ke : E idence om Ko ean po olio-le el
analysis
Economies
P o ided in Coope a ion wi h:
MDPI – Mul idisciplina y Digi al Publishing Ins i u e, Basel
Sugges ed Ci a ion: Choi, Sunghee (2024) : Es ima ing spillo e e ec om in e na ional oil ma ke
o s ock ma ke : E idence om Ko ean po olio-le el analysis, Economies, ISSN 2227-7099, MDPI,
Basel, Vol. 12, Iss. 4, pp. 1-14,
h ps://doi.o g/10.3390/economies12040092
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/329018
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Ci a ion: Choi, Sunghee. 2024.
Es ima ing Spillo e E ec om
In e na ional Oil Ma ke o S ock
Ma ke : E idence om Ko ean
Po olio-Le el Analysis. Economies 12:
92. h ps://doi.o g/10.3390/
economies12040092
Academic Edi o : Angeliki
N. Menegaki
Recei ed: 12 Ma ch 2024
Re ised: 4 Ap il 2024
Accep ed: 10 Ap il 2024
Published: 15 Ap il 2024
Copy igh : © 2024 by he au ho .
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
economies
A icle
Es ima ing Spillo e E ec om In e na ional Oil Ma ke o
S ock Ma ke : E idence om Ko ean Po olio-Le el Analysis
Sunghee Choi
Depa men o In e na ional Comme ce, Facul y o Economics and Comme ce, Keimyung Uni e si y,
Daegu 42601, Republic o Ko ea; [email p o ec ed]
Abs ac : Using a diagonal BEKK model, his pape es ima es a spillo e e ec om he in e na ional
c ude oil ma ke o he Ko ean s ock ma ke . Empi ical esul s sugges ha shocks and ola ili y
in Dubai oil p ices a e signi ican ly ansmi ed in o wen y po olios o he Ko ean s ock ma ke .
Also, i was ound ha hese spillo e e ec s d ama ically ose du ing he yea 2020, when he
h ea o COVID-19 was he mos se ious. Mo e speci ically, oil-o ien ed po olios, such as he
powe and gas i ms’ po olio and chemical i ms’ po olio, had a g ea e spillo e e ec om he
in e na ional c ude oil ma ke a he han o he po olios. Fu he , compa ed o la ge -capi aliza ion
i m po olios, small-capi aliza ion i m po olios had a ela i ely g ea e spillo e e ec . Se e al
implica ions and impo an a enues o u he esea ch a e iden i ied.
Keywo ds: spillo e e ec ; in e na ional oil ma ke ; Ko ean s ock ma ke ; a diagonal BEKK model;
po olio-le el analysis
1. In oduc ion
The ela ionship be ween in e na ional oil p ices and s ock ma ke s has been b ough
o he a en ion o in es o s, policymake s, and e en he gene al public in ecen yea s
because oil p ices ha e been showing d ama ic ola ili y du ing he pas decade. As Figu e 1
displays, he weekly o ecas ed p ices o in e na ional c ude oils mo ed d ama ically
be ween 2014 and 2022. The oil p ices ha had been abo e 100 USD pe ba el in he i s
hal o 2014 began o all below 30 USD in Decembe 2015. As o he yea 2016, he oil p ices
g adually ose and eco e ed o 80 USD in July 2018 bu immedia ely plunged abou 60 USD
and hen emained in he ange o 60–70 USD un il 2019. In he yea 2020, when COVID-19
occu ed, he oil p ice collapsed o 20 USD and hen con inued o ise and eached 120 USD.
The oil p ice declined again and mo ed o a ound 80 USD in Decembe 2022.
These su ges and alls in oil p ices a e expec ed o a ec s ock ma ke s because oil is a
c ucial sou ce o he mode n economic sys em. Howe e , he e ec does no necessa ily
exhibi a s able pa e n. As Figu e 1shows, in he case o he Republic o Ko ea (he ea e
Ko ea), he pa e n o he oil/s ock ma ke ela ionship a ies wi h ime. The e we e
pe iods o posi i e o nega i e ela ionships, and e en seemingly no ela ionship, be ween
he oil p ices and he KOSPI index, which is a p oxy o he Ko ean s ock ma ke . Fo he
i s “no co ela ion” box, which co esponds o he pe iod om he 1s week o June 2014
o he 4 h week o May 2015, he co ela ion coe icien be ween he oil p ices (ei he Dubai
o B en ) and KOSPI is calcula ed as 0.096. The second “no co ela ion” box wi h s able
mo emen s in KOSPI, al hough oil p ice mo emen s luc ua ed om he 1s week o Ap il
2016 o he 4 h week o Decembe 2016, has qui e a low co ela ion, which is 0.19. The hi d
and ou h box shows he nega i e and posi i e ela ionship be ween oil p ices and Ko ean
s ock p ices, espec i ely, which suppo s he a gumen ha he oil/s ock ela ionship is
uns able o e ime.
Economies 2024,12, 92. h ps://doi.o g/10.3390/economies12040092 h ps://www.mdpi.com/jou nal/economies
Economies 2024,12, 92 2 o 14
Economies 2024, 12, x FOR PEER REVIEW 2 o 14
Figu e 1. C ude oil p ices and he Ko ean s ock ma ke index: 2014–2022.
Mo i a ed by he obse a ion o d ama ic changes in oil p ices and inconclusi e pa -
e ns o oil/s ock ela ionships o e ime, his pape aims o in es iga e a ime- a ying
spillo e effec om oil ma ke ola ili y o s ock ma ke ola ili y wi h an indus y-spe-
ci ic da ase , especially o he case o Ko ea. The aim o his pape is also meaning ul o
dealing wi h h ee esea ch ques ions ha ha e no been esol ed in he oil/p ice esea ch
ield. Fi s , he Ko ean case helps disco e clea pa e ns o oil/s ock ela ionships due o
i s highe economic ulne abili y o changes in in e na ional oil p ices. Acco ding o he
Minis y o Fo eign Affai s o Republic o Ko ea (2021), Ko ea is known as he 8 h la ges
ene gy consume in he wo ld, and 94.8% o he la ge ene gy consump ion depends on
impo ed oil. Second, po olio-speci ic analysis is use ul o de ec ing diffe ences in he
oil/p ice ela ionship ac oss po olios, unlike he agg ega ed analysis by mos p e ious
s udies. Thi d, ime- a ying analysis, ins ead o ime-s a ic analysis, is sui able o han-
dling he dynamic ela ionship pa e n o e ime. Mo eo e , he ime span o ime- a y-
ing analysis includes he yea s o he COVID-19 pandemic, which con ibu es o sugges -
ing signi ican e idence o he oil/s ock ma ke linkage o he u bulen ime o in e na-
ional commodi y and inancial ma ke s.
As he scope and he a ge o esea ch, his pape es ima es a weekly ime- a ying
spillo e effec om 2014 o 2022 by using a diagonal BEEK model, which is sugges ed as
an app op ia e model by exis ing s udies. The main esul s a e as ollows: Fi s , o he
en i e sample pe iod, his pape inds ha shocks and ola ili y in he Dubai oil ma ke
we e ansmi ed in o he Ko ean s ock ma ke . In pa icula , hese spillo e effec s d a-
ma ically ose du ing he yea 2020, when he h ea o COVID-19 was he mos se ious.
Second, among a a ie y o classi ied po olios, his pape inds ha he spillo e effec
du ing he yea 2020 was he g ea es o he ene gy-o ien ed po olios, such as elec ici y,
gas, and chemicals, a he han o he indus ial po olios. Fu he , i is ound ha small-
capi aliza ion po olios had a g ea e spillo e effec om he oil ma ke han la ge-capi-
aliza ion po olios, which suppo s he in ui ion ha small i ms ha e lowe ope a ional
and inancial capabili y o manage ex e nal isk and unce ain y han la ge i ms.
The emainde o his pape is o ganized as ollows. Sec ion 2 b ie ly e iews he
ela ed li e a u e and he limi a ions o be o e come, and Sec ion 3 explains he da a and
me hodology, espec i ely. Sec ion 4 p o ides he empi ical esul s. Sec ion 5 summa izes
he s udy and sugges s key implica ions o u he esea ch.
Figu e 1. C ude oil p ices and he Ko ean s ock ma ke index: 2014–2022.
Mo i a ed by he obse a ion o d ama ic changes in oil p ices and inconclusi e
pa e ns o oil/s ock ela ionships o e ime, his pape aims o in es iga e a ime- a ying
spillo e e ec om oil ma ke ola ili y o s ock ma ke ola ili y wi h an indus y-speci ic
da ase , especially o he case o Ko ea. The aim o his pape is also meaning ul o
dealing wi h h ee esea ch ques ions ha ha e no been esol ed in he oil/p ice esea ch
ield. Fi s , he Ko ean case helps disco e clea pa e ns o oil/s ock ela ionships due o
i s highe economic ulne abili y o changes in in e na ional oil p ices. Acco ding o he
Minis y o Fo eign A ai s o Republic o Ko ea (2021), Ko ea is known as he 8 h la ges
ene gy consume in he wo ld, and 94.8% o he la ge ene gy consump ion depends on
impo ed oil. Second, po olio-speci ic analysis is use ul o de ec ing di e ences in he
oil/p ice ela ionship ac oss po olios, unlike he agg ega ed analysis by mos p e ious
s udies. Thi d, ime- a ying analysis, ins ead o ime-s a ic analysis, is sui able o handling
he dynamic ela ionship pa e n o e ime. Mo eo e , he ime span o ime- a ying
analysis includes he yea s o he COVID-19 pandemic, which con ibu es o sugges ing
signi ican e idence o he oil/s ock ma ke linkage o he u bulen ime o in e na ional
commodi y and inancial ma ke s.
As he scope and he a ge o esea ch, his pape es ima es a weekly ime- a ying
spillo e e ec om 2014 o 2022 by using a diagonal BEEK model, which is sugges ed
as an app op ia e model by exis ing s udies. The main esul s a e as ollows: Fi s , o
he en i e sample pe iod, his pape inds ha shocks and ola ili y in he Dubai oil
ma ke we e ansmi ed in o he Ko ean s ock ma ke . In pa icula , hese spillo e
e ec s d ama ically ose du ing he yea 2020, when he h ea o COVID-19 was he
mos se ious. Second, among a a ie y o classi ied po olios, his pape inds ha he
spillo e e ec du ing he yea 2020 was he g ea es o he ene gy-o ien ed po olios,
such as elec ici y, gas, and chemicals, a he han o he indus ial po olios. Fu he , i is
ound ha small-capi aliza ion po olios had a g ea e spillo e e ec om he oil ma ke
han la ge-capi aliza ion po olios, which suppo s he in ui ion ha small i ms ha e
lowe ope a ional and inancial capabili y o manage ex e nal isk and unce ain y han
la ge i ms.
The emainde o his pape is o ganized as ollows. Sec ion 2b ie ly e iews he
ela ed li e a u e and he limi a ions o be o e come, and Sec ion 3explains he da a and
me hodology, espec i ely. Sec ion 4p o ides he empi ical esul s. Sec ion 5summa izes
he s udy and sugges s key implica ions o u he esea ch.
Economies 2024,12, 92 3 o 14
2. Li e a u e Re iew
The ela ionship be ween in e na ional oil p ices and he s ock ma ke has long been
an in iguing opic in economic esea ch. In pa icula , since 2008 when a ious global
economic c ises equen ly eme ged and hi bo h he oil ma ke s and s ock ma ke s, he
esea ch in e es s in he oil/s ock ela ionship ha e been g owing ema kably. The numbe
o pape s in es iga ing he oil/s ock ela ionships published in op- anked jou nals, on a -
e age, was abou 1.9 du ing 1997–2007 bu inc eased o 13.1 du ing 2008–2017 (Degiannakis
e al. 2018), which shows a mo e han six imes a e o g ow h be o e and a e 2008.
In o de o a clea e p esen a ion, we p o ide he s eps o ou li e a u e e iew wo k
in Figu e 2. As he undamen al wo k, he i s s ep is o e iew he heo e ical backg ound
and channel o explaining he e ec o oil p ices on s ock p ices. The ea lies empi ical
analyses wi h he heo e ical esul s we e mos ly abou agg ega ed es ima ion. The e o e,
he i s s ep o he e iew ocuses on summa izing he majo indings as well as he
limi a ions o empi ical s udies by agg ega ed app oaches. Then, o he second s ep, he
ocus o ou e iew mo es na u ally o he disagg ega ed empi ical analyses because he
need o disagg ega ed analyses was b ough by he limi a ions o he agg ega ed analysis
in he i s s ep. In addi ion o he empi ical s udies wi h disagg ega ed app oaches, we
e iewed he li e a u e on he ime-speci ic analyses o he hi d s ep because he need o
speci ying he ime sample was eme ging and g owing oo. Consequen ly, hese s eps o
ou li e a u e e iew lead o a speci ic esea ch opic ha may con ibu e o he exis ing
li e a u e on he oil/p ice ela ionship, which is he ime- a ying spillo e e ec o oil
p ices on he Ko ean s ock po olio-le el analysis.
Economies 2024, 12, x FOR PEER REVIEW 3 o 14
2. Li e a u e Re iew
The ela ionship be ween in e na ional oil p ices and he s ock ma ke has long been
an in iguing opic in economic esea ch. In pa icula , since 2008 when a ious global
economic c ises equen ly eme ged and hi bo h he oil ma ke s and s ock ma ke s, he
esea ch in e es s in he oil/s ock ela ionship ha e been g owing ema kably. The num-
be o pape s in es iga ing he oil/s ock ela ionships published in op- anked jou nals,
on a e age, was abou 1.9 du ing 1997–2007 bu inc eased o 13.1 du ing 2008–2017
(Degiannakis e al. 2018), which shows a mo e han six imes a e o g ow h be o e and
a e 2008.
In o de o a clea e p esen a ion, we p o ide he s eps o ou li e a u e e iew wo k
in Figu e 2. As he undamen al wo k, he i s s ep is o e iew he heo e ical backg ound
and channel o explaining he e ec o oil p ices on s ock p ices. The ea lies empi ical
analyses wi h he heo e ical esul s we e mos ly abou agg ega ed es ima ion. The e o e,
he i s s ep o he e iew ocuses on summa izing he majo indings as well as he limi-
a ions o empi ical s udies by agg ega ed app oaches. Then, o he second s ep, he ocus
o ou e iew mo es na u ally o he disagg ega ed empi ical analyses because he need
o disagg ega ed analyses was b ough by he limi a ions o he agg ega ed analysis in
he i s s ep. In addi ion o he empi ical s udies wi h disagg ega ed app oaches, we e-
iewed he li e a u e on he ime-speci ic analyses o he hi d s ep because he need o
speci ying he ime sample was eme ging and g owing oo. Consequen ly, hese s eps o
ou li e a u e e iew lead o a speci ic esea ch opic ha may con ibu e o he exis ing
li e a u e on he oil/p ice ela ionship, which is he ime- a ying spillo e e ec o oil
p ices on he Ko ean s ock po olio-le el analysis.
Figu e 2. The s eps o exis ing li e a u e e iew.
Fo he i s s ep, we e iew he heo e ical amewo k o he oil/s ock ela ionship
and ind ha he esul s a e well-documen ed. The esul s sugges se e al ansmission
mechanisms wi h mac oeconomic pe spec i es. Th ee p ima y mechanisms a e summa-
ized in his e iew, which a e he mone a y channel, he agg ega e ou pu channel, and
he unce ain y channel. These channels explain he nega i e e ec o oil p ices on s ock
p ices, in gene al. The speci ic explana ion o each channel is sugges ed as ollows:
Fi s , he mone a y channel ocuses on in la ion and in e es a es. Rising oil p ices
a e expec ed o cause in la ion because p oduc ion cos s and e ail p ices a e unde p es-
su e o ise (Gaglia done and Ge le 2023; Kilian and Zhou 2022; Hamil on 1988, 2003).
Such in la ion expec a ion in luences mone a y policy-make s o decide o aise sho -
e m in e es a es unde he assump ion ha hei policy a ge is o s abilize he highe
p ice le els by ollowing he Tayle ’s Rule (Bashe and Sado sky 2006). The Tayle ’s Rule
is designed o app oxima e he esponse o sho - e m nominal in e es a es, as hese a e
se by he cen al bank when economic condi ions change (Tayle 1993). As a esul , aised
in e es a es inc ease he bo owing cos s o he i m, which de e io a es he i m’s cash
Figu e 2. The s eps o exis ing li e a u e e iew.
Fo he i s s ep, we e iew he heo e ical amewo k o he oil/s ock ela ionship
and ind ha he esul s a e well-documen ed. The esul s sugges se e al ansmission
mechanisms wi h mac oeconomic pe spec i es. Th ee p ima y mechanisms a e summa-
ized in his e iew, which a e he mone a y channel, he agg ega e ou pu channel, and
he unce ain y channel. These channels explain he nega i e e ec o oil p ices on s ock
p ices, in gene al. The speci ic explana ion o each channel is sugges ed as ollows:
Fi s , he mone a y channel ocuses on in la ion and in e es a es. Rising oil p ices a e
expec ed o cause in la ion because p oduc ion cos s and e ail p ices a e unde p essu e
o ise (Gaglia done and Ge le 2023;Kilian and Zhou 2022;Hamil on 1988,2003). Such
in la ion expec a ion in luences mone a y policy-make s o decide o aise sho - e m
in e es a es unde he assump ion ha hei policy a ge is o s abilize he highe p ice
le els by ollowing he Tayle ’s Rule (Bashe and Sado sky 2006). The Tayle ’s Rule is
designed o app oxima e he esponse o sho - e m nominal in e es a es, as hese a e se
by he cen al bank when economic condi ions change (Tayle 1993). As a esul , aised
in e es a es inc ease he bo owing cos s o he i m, which de e io a es he i m’s cash
low, which is he mos essen ial ac o o s ock p ices o i ms. Second, he ou pu channel
Economies 2024,12, 92 4 o 14
ocuses on he household’s income and agg ega e ou pu . Rising oil p ices lead o lowe
disc e iona y income o households due o highe e ail p ices o commodi ies and se ices
in mos economies (S ensson 2005,2006). The nega i e ela ionship be ween oil p ices and
income holds o mos o he oil-impo ing coun ies in he wo ld. Howe e , in he case o
oil-expo ing coun ies, he income inc eases when he oil p ice ises due o inc eased oil
e enue. Since his pape deals wi h he case o Ko ea, which is an oil-impo ing coun y,
he li e a u e e iew on he ou pu channel ocuses on he oil-impo ing case. Such lowe ed
income leads o lowe consump ion and lowe agg ega e ou pu , which de e io a es he
cash low o i ms, which is he ac o o s ock p ices in alue (Hamil on 1983,2003;Kilian
2008). Thi d, he unce ain y channel poin s ou he highe unce ain y gene a ed by highe
oil p ice ola ili y because highe unce ain y de e io a es he u u e cash lows o i ms by
pos poning hei decision-making on in es men and pu chasing (Be nanke 1983;Pindyck
2003;Edels ein and Kilian 2009).
Al hough hese heo e ical amewo ks suppo o nega i e oil/s ock ela ionship,
many o he analyses epo ed posi i e ela ionships. The backg ound o he posi i e
oil/s ock ela ionship is he iscal ansmission channel. The iscal channel a gues ha
highe oil p ices may lead o an inc ease in he weal h o oil-expo ing coun ies, which
allows o inc eases in go e nmen pu chases o he coun ies. Conside ing ha such
pu chases a e no subs i u es bu complemen s, hese can lead o inc eases in household
consump ion as well as agg ega e demand. Responding o he expec a ion o highe
agg ega e demand, p i a e i ms will make mo e ac i e p oduc ion and in es men , which
con ibu es o highe cash low o i ms and s ock p ices in he wo ld. In addi ion, Be nanke
(2016) sugges s an addi ional iew o he posi i e oil/s ock ela ionship h ough he lens o
agg ega e demand. He a gues ha oil p ices and s ock p ices mo e oge he owa d he
agg ega e demand mo emen s in he eal wo ld. Fo example, he all ( ise) in in e na ional
oil p ices leads o an expec a ion o a downwa d (upwa d) global economy and lowe
(highe ) agg ega e demand in he wo ld. In such cases, p i a e i ms become less (mo e)
ac i e in p oduc ion and in es men , which pushes he s ock p ices o lowe (highe ) le els.
Summing up he i s s ep o he li e a u e e iew on heo e ical amewo ks and ag-
g ega ed mac oeconomic pe spec i es, i is no simple o conclude he sign o he oil/s ock
p ice ela ionship. Because highe oil p ices may lead o ei he lowe s ock p ices h ough
some pe spec i es o highe s ock p ices h ough o he pe spec i es, agg ega ing hese
wo opposi e signs leads o he inconclusi e esul s on he oil/p ice ela ionship. Mo e
speci ically, i he agg ega e demand e ec is g ea e han o he ac o e ec s, hen he
agg ega ed oil/s ock p ice ela ionship is posi i e. O he wise, he ela ionship becomes
ei he nega i e o unexplained. The agg ega ed analysis, which combines all he e ec s
o oil p ice changes on s ock ma ke s, has limi a ions in yielding a conclusi e ela ionship
be ween oil and s ock.
To o e come he limi a ions in he agg ega e oil/s ock analysis, a lo o esea ch has
been ex ended o disagg ega e analysis, such as indus y-speci ic analysis, i m-speci ic
analysis, and pe iod-speci ic analysis. Fo he second s ep, we e iew he disagg ega ed
analyses and ind ha he con ibu ion o he disagg ega ed analyses was o de ec he e o-
genei y in oil/s ock ela ionships ac oss indus ies and i ms. Fo example, he indus y-
speci ic sec o al index analyses e eal ha he e a e indeed he e ogeneous esponses o
he oil p ice changes by di e en indus ial sec o s (El-Sha i e al. 2005;Boye and Filion
2007;A ou i and Nguyen 2010;A ou i 2012;B oads ock e al. 2012,2014). Fo example,
highly oil- ela ed sec o s, such as oil and gas, elec ici y, oil e ine s, e c., espond posi i ely
o oil p ice changes, whe eas less oil- ela ed sec o s, such as ood, Chemical, compu e ,
inancial, gene al se ice, medical, eal es a e, e c., ha e nega i e o no esponse o oil
p ice changes o he case o global indus y sample (Nandha and B ooks 2009), US sample
(Na ayan and Sha ma 2011;Elyasiani e al. 2011), Eu opean sample (A ou i and Nguyen
2010;Shol ens and Yu se e 2012), and De eloped Eu ope and G7 sample (Nandha and
B ooks 2009). Fu he , i m-speci ic esea ch enables p o iding addi ional insigh s in o he
oil/s ock ela ionship in he sense ha i m-le el cha ac e is ics de e mine he ela ionship.
Economies 2024,12, 92 5 o 14
Howe e , i m-speci ic analyses a e limi ed because hey p o ide sca ce e idence poin ing
o he di ec ion ha e en hough he e ogenei y among and wi hin s ock ma ke sec o s
does exis ; oil- ela ed s ocks end o be highe in alue when oil p ices ise.
In addi ion o i m- o indus y-speci ic analysis, ime-speci ic analysis has been
eme ging in he oil/s ock nexus. Especially since 2008, when he global inancial c isis
occu ed, he numbe o publica ions dealing wi h he ime- a ying ela ionship has been
g owing pe sis en ly. As Figu e 3shows, acco ding o Degiannakis e al. (2018), he
numbe o pape s inding ime- a ying ela ionships be ween oil and s ock ma ke s has
been g owing, al hough all o he pape s dec eased a e 2012. In pa icula , om 2013 o
2017, he numbe o a icles ha disco e ed he ime- a ying ela ionship was g ea e han
he numbe o ones on all o he opics. The ime- a ying ela ionship analysis mainly aims
a iden i ying di e ences be ween he demand-side shock and he supply-side shock o
oil ma ke s. Fo example, Kang e al. (2015) iden i ied ha he oil shocks on he demand
side o oil ma ke s, such as he economic boom, exe cised a posi i e e ec du ing 1973–
2012, whe eas he oil shocks on he supply side o he oil ma ke , such as oil supply
capaci y, ha e a nega i e e ec du ing he ea ly pe iod, 1973–1980. Also, An onakakis
e al. (2013) examined whe he he di e en oil p ice shocks ha e spillo e e ec s on s ock
ma ke s a di e en imes. They ound ha he agg ega e-demand shocks played a ole as
a ne ansmi e o shocks in o s ock ma ke s du ing pe iods cha ac e ized by economic
u bulence, while he supply-side and p ecau iona y-demand shocks we e ne ansmi e s
o shocks du ing geopoli ical u bulence pe iods.
Economies 2024, 12, x FOR PEER REVIEW 5 o 14
G7 sample (Nandha and B ooks 2009). Fu he , i m-speci ic esea ch enables p o iding
addi ional insigh s in o he oil/s ock ela ionship in he sense ha i m-le el cha ac e is ics
de e mine he ela ionship. Howe e , i m-speci ic analyses a e limi ed because hey p o-
ide sca ce e idence poin ing o he di ec ion ha e en hough he e ogenei y among and
wi hin s ock ma ke sec o s does exis ; oil- ela ed s ocks end o be highe in alue when
oil p ices ise.
In addi ion o i m- o indus y-speci ic analysis, ime-speci ic analysis has been
eme ging in he oil/s ock nexus. Especially since 2008, when he global inancial c isis oc-
cu ed, he numbe o publica ions dealing wi h he ime- a ying ela ionship has been
g owing pe sis en ly. As Figu e 3 shows, acco ding o Degiannakis e al. (2018), he num-
be o pape s inding ime- a ying ela ionships be ween oil and s ock ma ke s has been
g owing, al hough all o he pape s dec eased a e 2012. In pa icula , om 2013 o 2017,
he numbe o a icles ha disco e ed he ime- a ying ela ionship was g ea e han he
numbe o ones on all o he opics. The ime- a ying ela ionship analysis mainly aims a
iden i ying diffe ences be ween he demand-side shock and he supply-side shock o oil
ma ke s. Fo example, Kang e al. (2015) iden i ied ha he oil shocks on he demand side
o oil ma ke s, such as he economic boom, exe cised a posi i e effec du ing 1973–2012,
whe eas he oil shocks on he supply side o he oil ma ke , such as oil supply capaci y,
ha e a nega i e effec du ing he ea ly pe iod, 1973–1980. Also, An onakakis e al. (2013)
examined whe he he diffe en oil p ice shocks ha e spillo e effec s on s ock ma ke s a
diffe en imes. They ound ha he agg ega e-demand shocks played a ole as a ne ans-
mi e o shocks in o s ock ma ke s du ing pe iods cha ac e ized by economic u bulence,
while he supply-side and p ecau iona y-demand shocks we e ne ansmi e s o shocks
du ing geopoli ical u bulence pe iods.
Figu e 3. Numbe o pape s on he oil/p ice ela ionship o e yea s.
Despi e he nume ous s udies exis ing in he li e a u e on he oil/s ock ela ionship,
no empi ical e idence showed a possible he e ogeneous ela ionship be ween he wo
ma ke s o e diffe en ime pe iods un il 2010. Fo example, oil ma ke ola ili y was he
main ansmi e o ola ili y shocks o s ock ma ke s, a he han he e e se. Fu he ,
hese spillo e effec s we e mo e appa en du ing he inancial c isis and no dis inc ion
be ween oil impo e s and expo e s exis ed. (A ou i e al. 2011; Awa ani and Maghye eh
2013; Boldano e al. 2016; Maghye eh e al. 2016).
O e all, by conside ing he indings, insigh s, and ends o he exis ing li e a u e on
he oil/p ice ela ionship, ou s udy is expec ed o make a signi ican con ibu ion o he
Figu e 3. Numbe o pape s on he oil/p ice ela ionship o e yea s.
Despi e he nume ous s udies exis ing in he li e a u e on he oil/s ock ela ionship,
no empi ical e idence showed a possible he e ogeneous ela ionship be ween he wo
ma ke s o e di e en ime pe iods un il 2010. Fo example, oil ma ke ola ili y was he
main ansmi e o ola ili y shocks o s ock ma ke s, a he han he e e se. Fu he ,
hese spillo e e ec s we e mo e appa en du ing he inancial c isis and no dis inc ion
be ween oil impo e s and expo e s exis ed. (A ou i e al. 2011;Awa ani and Maghye eh
2013;Boldano e al. 2016;Maghye eh e al. 2016).
O e all, by conside ing he indings, insigh s, and ends o he exis ing li e a u e
on he oil/p ice ela ionship, ou s udy is expec ed o make a signi ican con ibu ion o
he exis ing li e a u e ega ding h ee aspec s. Fi s , his pape deals wi h he oil/p ice
ela ionship opic ha has been b ough o academic a en ion o equi ing he need o use
a ime- a ying spillo e e ec be ween oil and s ock ma ke s. Second, his pape is expec ed
o p o ide a conclusi e esul on he oil/p ice ela ionship unlike mos p io s udies using
Economies 2024,12, 92 6 o 14
a la ge open economy case. Speci ically, he case o Ko ea wi h i s high dependence on
impo ed oil is mo e help ul o cla i ying he oil/p ice ela ionship due o i s s uc u al
ulne abili y o in e na ional oil p ice ola ili y. Thi d, po olio-speci ic analysis enables a
be e unde s anding o he he e ogeneous e ec s o oil p ices on disagg ega e indus ial
po olios a he han agg ega e s ock ma ke analysis. Hence, his pape is expec ed o ill
his gap and con ibu e o he exis ing s udies by ca ying ou a Ko ean po olio-speci ic
in es iga ion o ime- a ying ola ili y spillo e e ec s om in e na ional oil p ices o
s ock ma ke sec o al indexes. Also, we shed ligh on long- e m pe iods expanding o
Decembe 2022, which enables checking a signi ican change in he oil/p ice ela ionship
a e COVID-19 occu ed because i is a e o ind exis ing oil/p ice li e a u e including he
pandemic pe iod.
3. Me hodology: Da a and Es ima ion Model
The da ase is composed o weekly da a on Dubai c ude oil o ecas p ices and s ock
ma ke po olio indexes o Ko ea, spanning he pe iod om he i s F iday o June (da e:
13 June) in 2014 o he las F iday o Decembe (da e: 23 Decembe ) in 2022. The ime span
and obse a ions a e su icien enough compa ed wi h many ecen wo ks co e ing e en
less han 8 yea s, such as Li e al. (2023), Mo e al. (2023), Zal agha i e al. (2020), and so on.
Also, his pape examines 20 po olio indexes ha a e o icially classi ied by he Ko ea Ex-
change Ma ke (KRX). These po olio index da a we e collec ed om he KisValue, which
is known as a p ima y sou ce o Ko ean s ock ma ke in o ma ion. The 20 po olios a e as
ollows: KOSPI, la ge-capi aliza ion (he ea e called as “La ge-cp”), small-capi aliza ion
(he ea e “Small-cp”), ood and be e ages (“F&B”), ex ile and clo hing (“T&C”), pa-
pe and woods (“P&W”), chemicals (“CHEM”), pha maceu ical (“PHRM”), non-me allic
mine als (“NMM”), s eel and me allic mine als (“SMM”), machine y (“MACH”), elec ic-
i y and elec onics (“E&E”), p ecision medicine (“PM”), anspo a ion and equipmen
(“T&E”), dis ibu ions (“DIST”), powe and gas (“P&G”), cons uc ion (“CON”), wa e-
house (“WRH”), elecommunica ion (“TEL”), inancial (“FIN”) index. In addi ion, his
pape uses he Dubai oil u u es p ices (Pla s) om he CME g oup because he Dubai p ice
ep esen s he Asian oil ma ke s whe e Ko ea depends highly on he impo s o c ude oil.
We ans o med he le el-based da a in o he e u n-based da a by using he o mula,
ln P
P −1
, whe e
P
deno es p ices (indexes) o Dubai ( he po olio) and is he weekly ime.
The o al sample size o he e u ns du ing he sample pe iod is 418, and he desc ip i e
s a is ics o all he weekly e u n se ies a e epo ed in Table 1. Acco ding o he J.B. and
ADF es s, all he a iables a e no mally dis ibu ed and s a is ically s a iona y.
In o de o es ima e a spillo e e ec om in e na ional oil ma ke s o Ko ean s ock
ma ke po olios, his pape employs a diagonal BEKK model (Choi 2021;Zal agha i e al.
2020). Acco ding o he sugges ion ha a diagonal BEKK is a mo e eliable model han
DCC and ull BEKK o es ima ing he spillo e s be ween oil and s ock ma ke s (McAlee
2019a,2019b;Zal agha i e al. 2020), his pape employs a diagonal BEKK app oach o
expec ing mo e eliable e idence on he oil/s ock linkage han p io s udies. This pape
examines he ela ionship be ween wo a iables, oil and s ock log- e u ns, esul ing in a
bi a ia e app oach as ollows:
= 1,
2, ,
whe e
is he ec o o log e u ns a ime , subsc ip ion 1 deno es he oil ma ke , and
subsc ip ion 2 deno es he s ock ma ke . can be decomposed as ollows:
=µ +α ,
whe e
µ =E( |F −1)
and
α =H
1
2
ϵ
. The i s e m,
µ
, is he expec ed e u n o ime
gi en he a ailable in o ma ion a he p e ious ime a
−
1. The in o ma ion se is deno ed
as F. The second e m implies he unexpec ed pa o he e u n wi h
H
and
ϵ
.
ϵ
a e
Economies 2024,12, 92 7 o 14
independen ly iden ically dis ibu ed (known as iid) andom ec o s wi h a mean equal o
ze o. H is he co a iance o α , which is de ined as ollows:
Co (α |F −1)=H =h11, h12,
h21, h22,
Table 1. Summa y s a is ics o he weekly e u n se ies om June 2014 o Decembe 2022.
Dubai KOSPI La ge-cp Small-cp Indus y-Speci ic Po olios
F&B T&C P&W CHEM PHRM NMM
Mean −0.007 0.0004 0.0004 0.0006 −0.0002 0.0010 0.0001 0.0006 0.0028 0.0019
S d. 0.051 0.025 0.025 0.031 0.028 0.037 0.037 0.033 0.045 0.043
Max. 0.303 0.128 0.147 0.111 0.107 0.177 0.150 0.145 0.179 0.178
Min. −0.276 −0.141 −0.132 −0.203 −0.134 −0.183 −0.241 −0.162 −0.183 −0.182
Skew. −0.032 −0.559 −0.186 −1.557 −0.280 0.033 −1.058 −0.300 −0.235 −0.129
Ku osis 8.661 6.136 5.904 8.344 2.065 3.231 7.209 3.519 2.375 3.042
J.B. 1306.42 677.60 609.62 1381.45 79.75 181.88 983.27 222.00 102.08 162.36
ADF. −6.12 −6.66 −6.62 −6.83 −6.28 −6.04 −6.72 −5.71 −6.42 −7.06
Indus y-Speci ic Po olios
SMM MACH E&E PM T&E DIST P&G CON WRH TEL FIN
Mean −0.0002 −
0.0001
0.0017 0.0002 −0.0007 −0.0009 −0.0008 −0.0013 −0.0004 0.0005 −0.0004
S d. 0.037 0.038 0.034 0.048 0.038 0.028 0.034 0.041 0.036 0.029 0.029
Max. 0.150 0.106 0.196 0.140 0.219 0.110 0.134 0.108 0.138 0.095 0.153
Min. −0.192 −0.226 −0.129 −0.238 −0.220 −0.148 −0.176 −0.189 −0.180 −0.129 −0.167
Skew. −0.259 −0.913 0.233 −0.456 −0.025 −0.477 −0.132 −0.356 −0.172 −0.572 −0.453
Ku osis 2.338 3.741 2.744 1.795 6.818 3.354 2.742 1.174 2.827 1.831 6.442
J.B. 99.86 301.74 134.90 70.64 809.75 211.76 132.17 32.83 141.27 81.22 737.07
ADF. 6.28 −7.12 −6.67 −6.29 −7.20 −6.81 −6.59 −7.44 −6.19 −6.74 −6.71
The ma ix
H
is symme ic since
H =H′
, and i is a posi i e de ini e a all . Thus, a
gene al ull and diagonal BEKK model wi h a
n×n
pa ame e s ma ix and es ic ions o
scala aand bis exp essed as ollows:
H =C·C′+
p
∑
i=1
K
∑
k=1
Aik·ϵ −iϵ′
−i·A′
ik +
q
∑
j=1
K
∑
k=1
Bjk·H −j·B′
jk
whe e Aand Ba e pa ame e s ma ices in aspi a ion o ob aining a posi i e H, and C is
a lowe iangula ma ix. Because p=q= 1, he inal model o a BEKK o his pape
becomes as ollows:
H =M+A·ϵ −1ϵ′
−1·A′+B·H −1·B′
The backg ound eason o p=q= 1 is ha he ime-se ies oil p ices and s ock p ice
da a show s a is ically signi ican (squa ed) s anda dized esiduals by using GARCH
(1, 1)
es s h ough o de up o 36 lags acco ding o Choi (2021). Also, Min he abo e model
is an inde ini e ma ix because
M=C·C′
. Thus, he abo e model allows his pape o
es ima e he ime- a ying co ela ion be ween he unexpec ed componen s o oil and s ock
ma ke log e u ns:
Co (e ,s ock,e ,oil
h ough he quasi-maximum likelihood es ima o s
(QMLE). McAlee e al. (2008) p o ed ha he QMLE o he pa ame e s o he diagonal
BEEK models we e consis en and asymp o ically no mal. The elemen s o he ma ices A
and B e lec he impac o shocks in indi idual ma ke s and he pe sis ence o shocks in
each indi idual ma ke , espec i ely. In speci ic, he ma ix Ap o ides in o ma ion abou
“di ec ” shock spillo e s and “indi ec ” shock spillo e s. The “di ec ” e ec means ha he
condi ional a iance in an indi idual ma ke eac s o i s own lagged shocks and/o lagged
shocks in o he ma ke s, and he “indi ec ” e ec means ha he condi ional a iance in an
indi idual ma ke eac s o any combina ion o he lagged shocks c oss- e ms. The ma ix
Bno only ells he “di ec ” ola ili y spillo e s, meaning ha he condi ional a iance in
an indi idual ma ke esponds o i s own lagged ola ili y and/o o lagged ola ili y in
Economies 2024,12, 92 8 o 14
he o he ma ke s bu also he “indi ec ” ola ili y spillo e s, meaning ha he condi ional
a iance in indi idual ma ke esponds o any lagged co a iance. Ma ke ola ili y, which
is in e p e ed as he ma ke unce ain y isk, is he key a iable o he spillo e e ec
analysis (Ju ado e al. 2015;Haddad e al. 2023).
4. Empi ical Resul s
Fi s ly, by es ima ing he ma ices Aand Bin he diagonal BEKK model, his pape
disco e s he di ec and indi ec spillo e e ec om he Dubai oil ma ke o he Ko ean
s ock ma ke . As shown in Table 2, he co a iance coe icien s o ma ices A and B o all he
po olios a e es ima ed signi ican ly a 1% and 5% le els, excep o wo coe icien s. These
esul s ensu e he exis ence o di ec shock spillo e e ec s om Dubai oil p ices o Ko ean
s ock po olio e u ns, as well as indi ec ola ili y spillo e s om Dubai oil p ices o
Ko ean s ock po olio e u ns. I is impo an in o ma ion o in es o s and policymake s
o pe cei e he exis ence o bo h shock and ola ili y spillo e e ec s om he in e na ional
oil ma ke o he Ko ean s ock ma ke because app op ia e hedging s a egies o oil ma ke
isks should be p epa ed.
Table 2. Va iance coe icien s o he ma ices Aand Bin a diagonal BEKK model.
Po olio Index A(1, 1) A(2, 2) B(1, 1) B(2, 2)
KOSPI 0.2818 (0.10) ** 0.8213 (0.35) * 0.9414 (0.02) ** 0.6549 (0.14) **
La ge-cp 0.6861 (0.28) ** 0.2750 (0.09) ** 0.6841 (0.14) ** 0.9404 (0.02) **
Small-cp 0.7571 (0.27) ** 0.2637 (0.19) ** 0.4670 (0.22) * 0.9374 (0.02) **
F&B 0.4897 (0.02) ** 0.2458 (0.05) ** 0.1125 (0.61) 0.9361 (0.02) **
T&C 0.5714 (0.16) ** 0.2389 (0.05) ** 0.6667 (0.14) ** 0.9412 (0.02) **
P&W 0.5704 (0.18) ** 0.2639 (0.07) ** 0.8284 (0.07) ** 0.9384 (0.02) **
CHEM 0.2823 (0.08) ** 0.6026 (0.21) ** 0.9286 (0.03) ** 0.7808 (0.09) **
PHRM 0.3368 (0.13) ** −0.1625 (0.10) 0.9426 (0.02) ** 0.9812 (0.01) **
NMM 0.2774 (0.07) ** 0.6140 (0.22) ** 0.9380 (0.02) ** 0.8489 (0.05) **
SMM 0.2361 (0.05) ** 0.1826 (0.07) ** 0.9470 (0.02) ** 0.9835 (0.01) **
MACH 0.2792 (0.08) ** 0.3892 (0.16) ** 0.9298 (0.03) ** 0.7888 (0.16) **
E&E 0.3080 (0.10) ** −0.6067 (0.28)* 0.9425 (0.02) ** 0.2032 (0.59)
PM 0.2320 (0.05) ** 0.5491 (0.18) ** 0.9403 (0.02) ** 0.5657 (0.23) **
T&E 0.2348 (0.05) ** 0.4376 (0.12) ** 0.9435 (0.02) ** 0.8761 (0.05) **
DIST 0.2594 (0.07) ** 0.5560 (0.20) ** 0.9396 (0.02) ** 0.7021 (0.15) **
P&G 0.2656 (0.07) ** 0.4072 (0.19) ** 0.9344 (0.02) ** 0.6662 (0.27) *
CON 0.2123 (0.04) ** 0.4402 (0.12) ** 0.9479 (0.02) ** 0.9144 (0.04) **
WRH 0.2509 (0.07) ** 0.5621 (0.18) ** 0.9354 (0.02) ** 0.7656 (0.11) **
TEL 0.2577 (0.05) ** 0.2163 (0.09) ** 0.9278 (0.02) ** 0.9554 (0.05) **
FIN 0.2833 (0.09) ** 0.3245 (0.12) ** 0.9365 (0.02) ** 0.9371 (0.03) **
* and ** mean signi icance le els a 5% and 1%, espec i ely. Numbe s in pa en heses a e s anda d e o s.
Fo a simple obus ness check, we e es ed a ola ili y spillo e e ec o he Dubai
oil ma ke on each po olio e u n by using an ex ended GARCH (1, 1) as Choi (2021)
sugges ed. The GARCH (1, 1) ola ili y equa ion is
h2
i, =ω+δiu2
i, −1+θih2
i, −1+ρ1,jh2
oil. +
ρ2,jh2
oil, −1
, showing ha
h2
i,
is he a iance o each i’s po olio e u ns as a GARCH
(1, 1)
p ocess in which he condi ional a iance equa ion depends on a se o a iables: i s
own pas squa ed esiduals (
u2
−1
), he lagged a iance e m (
h2
−1
), he a iance o oil
ma ke e u ns (
h2
oil.
), and he lagged a iance o oil ma ke e u ns (
h2
oil. −1
). To a oid
edundancy, Table 3only shows he es ima ed esul s on
ρ1
and
ρ2
, which implies he
cu en ola ili y spillo e and lagged ola ili y spillo e e ec , espec i ely. We ind ha ,
ou o he 20 po olio indexes, 12 po olios ha e s a is ically signi ican lagged ola ili y
spillo e e ec s (
ρ2
) om he oil ma ke , while only a couple o po olios, ex ile, clo h, and
chemical, ha e signi ican cu en ola ili y spillo e e ec s. The esul s by an ex ended
GARCH (1, 1) a e unable o pe ec ly check he obus ness o he BEKK esul s, bu i
is success ul in suppo ing he BEKK esul s in he pe spec i e ha mos o he Ko ean
s ock ma ke po olios a e exposed o he oil ma ke ola ili y. Fu he mo e, he esul s
