Çiçek, Se kan; Yıldı ım, Aynu
A icle
The impac o domes ic and global ac o s on indi idual
public, domes ic and o eign bank pe o mances in
Tü kiye
Cen al Bank Re iew (CBR)
P o ided in Coope a ion wi h:
Cen al Bank o The Republic o Tu key, Anka a
Sugges ed Ci a ion: Çiçek, Se kan; Yıldı ım, Aynu (2024) : The impac o domes ic and global ac o s
on indi idual public, domes ic and o eign bank pe o mances in Tü kiye, Cen al Bank Re iew
(CBR), ISSN 1303-0701, Else ie , Ams e dam, Vol. 24, Iss. 1, pp. 1-16,
h ps://doi.o g/10.1016/j.cb e .2023.100139
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/297968
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/
CBREV 24 (2024) 100139
A ailable online 17 Decembe 2023
1303-0701/© 2023 The Au ho s. Published by Else ie B.V. on behal o Cen al Bank o The Republic o Tu key. This is an open access a icle unde he CC BY
license (h p://c ea i ecommons.o g/licenses/by/4.0/).
The impac o domes ic and global ac o s on indi idual public, domes ic
and o eign bank pe o mances in Tü kiye
Se kan Çiçek, Aynu Yıldı ım
*
Mu˘
gla Sı kı Koçman Uni e si y, Fe hiye Business Facul y, Depa men o Economics and Finance, Mu˘
gla, Tu key
ARTICLE INFO
JEL classi ica ion:
C32
G21
Keywo ds:
Vola ili y spillo e
Diagonal BEKK-GARCH
In e es a e isk
Fo eign exchange isk
ABSTRACT
The Tu kish economy has encoun e ed signi ican shocks in in e es a es and o eign exchange along wi h global
isks in ecen yea s. These shocks had an impac no only on he eal sec o bu also on he banking sec o ’s
e u ns, depending on he owne ship s uc u e. This s udy examines he sensi i i y o banking sec o s ock
e u ns o he exchange a e, in e es a e, and VIX index using da a om Janua y 4, 2005 o Ma ch 28, 2023.
Using mul i a ia e diagonal BEKK-GARCH me hodology, he s udy ound ha (i) hal o p i a e banks expe i-
enced a mean spillo e om he in e es a e o hei e u ns, bu no om he exchange a e and VIX index, (ii)
he e u ns o public banks, on he o he hand, did no espond o any a iable in he mean equa ions, (iii) he
explana o y powe o exchange a e and in e es a e isks is highe han he powe o he changes in hese
a iables, (i ) he spillo e o global isk in co a iance equa ions is highe compa ed o exchange and in e es
a e isks, ( ) he mean equa ions do no ha e an asymme ic s uc u e, bu he co a iance equa ions exhibi
s uc u al b eaks. These indings sugges ha in he las decade, he in e es a e policy has become he main
a iable a ec ing he s ock e u ns in Tü kiye, o eign exchange has become a sa e ha en due o his policy, and
he ela ionship be ween he exchange a e and s ocks ha exis ed in he pas has been dis up ed.
1. In oduc ion
In he las wo decades, he e has been an imp o emen in he pe -
o mance o he banking sec o in Tü kiye due o swi es uc u ing,
ad ancemen s in echnology and a succession o libe alisa ion e o ms.
The 2001 banking c isis, which caused bank up cy o 24 banks and had a
deep impac on he Tu kish economy, was he key eason behind he
es uc u ing. The BRSA (Banking Regula ion and Supe ision Agency)
moni o s and egula es he banking sec o ’s ac i i ies closely ollowing
he c isis-d i en legisla ion o es uc u e he sec o .
The p og ess in echnology wo ldwide has signi ican ly con ibu ed
o enhancing he banking sec o in Tü kiye by enabling banks o in e-
g a e hese ad ancemen s in o hei business a eas (Kasman, 2012). The
expansion o inancial se ice ac i i ies has simul aneously inc eased
compe i ion in he banking sec o and s eng hened in eg a ion be ween
banks. Whils he inc ease in in eg a ion be ween banks inc eased
compe i ion in he sec o and esul in highe p oduc i i y (Tan and
Flo os, 2018), i equally exposed all banks o simila ad e se shocks
a ec ing he sec o (Dungey e al., 2020). Mo eo e , he inc ease in he
deg ee o in eg a ion among banks has led o a banking sys em ha is
mo e ulne able o he spillo e o shocks (Fo bes and Rigobon, 2002;
Ab eu and Mendes, 2010; Assidenou, 2011; Coeu dacie and Guibaud,
2011; Mun and B ooks, 2012; Dungey e al., 2020). Since shocks o he
banking sys em ha e a p o ound impac on he compe i i eness o
banks, i is o g ea impo ance o in es iga e how he shocks ha hi he
banking sys em a ec he pe o mance o banks ope a ing in he sec o .
Possible impac s on he banking sys em can a ise bo h om o eign
and domes ic sou ces. The nega i e impac o global de elopmen s leads
o an inc ease in non-pe o ming loans, which ad e sely may a ec he
eal sec o . The olume o c edi may con ac , while he cos o c edi
ex ended by banks ises. Non- inancial sec o s ha canno use bank
loans may be ad e sely a ec ed, leading o a con ac ion in he eal
economy. Shocks o he banking sec o may spill o e o he non-
inancial sec o h ough wo channels: he inancial channel and he
demand channel (Tong and Wei, 2015). In imes o ecession, he de-
mand channel a ises due o he low le el o consump ion, while he
inancial channel is ela ed o he dis up ion in he low o c edi du ing
a banking c isis (Peek e al., 2003; Lae en and Valencia, 2013).
Pee e iew unde esponsibili y o he Cen al Bank o he Republic o Tu key.
* Co esponding au ho .
E-mail add esses: [email p o ec ed] (S. Çiçek), [email p o ec ed] (A. Yıldı ım).
Con en s lis s a ailable a ScienceDi ec
Cen al Bank Re iew
jou nal homepage: www.jou nals.else ie .com/cen al-bank- e iew/
h ps://doi.o g/10.1016/j.cb e .2023.100139
Recei ed 18 Ap il 2023; Recei ed in e ised o m 3 No embe 2023; Accep ed 29 No embe 2023
Cen al Bank Re iew 24 (2024) 100139
2
Empi ical esea ches indica e ha he pe o mance o he banking in-
dus y is signi ican ly in luenced by o eign shocks, as e ealed in p e-
ious s udies by Ce o elli and Goldbe g (2011), Balcila and Demi e
(2015), and Tiwa i e al. (2022).
1
Se e al s udies add essed pa icula
shocks, such as Ce o elli and Goldbe g (2011), who con ended ha
global banks played a signi ican ole in dissemina ing he impac s o
global shocks, pa icula ly du ing he global inancial c isis. Howe e , a
signi ican numbe o s udies concen a ed on he in luence o unce -
ain y o ola ili y indices on he s ock e u ns o banks (Hajilee and
Nasse , 2017; Alja ayesh e al., 2018; Fe ei a Ma ins e al., 2021).
Ul ima ely, analyses by bo h g oups demons a ed ha global impac s
play a signi ican ole in he e u ns on banks’ s ocks.
Domes ic issues, in addi ion o global e en s, exe a signi ican
impac on he banking sec o . In ecen yea s, in Tü kiye, he policy a e
has been se signi ican ly below he in la ion a e, and addi ional mea-
su es ha e also been implemen ed o minimize he excessi e luc ua-
ions in exchange a es. In addi ion, he Banking Regula ion and
Supe ision Agency (BRSA) has issued se e al egula ions o bo h he
bond and o eign exchange ma ke s, limi ing he scope o banks in he
economy. These ecen de elopmen s ha e led o inc eased economic
unce ain y since May 2018 (Demi alp and Demi alp, 2019). Gi en ha
sudden luc ua ions in in e es a es and exchange a es can se e ely
ha m banks’ pe o mance (Lloyd and Shick, 1977; Flanne y and James,
1984; Hancock, 1985; G amma ikos e al., 1986: 671; Demi guc-Kun
and Huizinga, 1999; Den Haan e al., 2007; Zei un e al., 2007; Kasman
e al., 2011; Hajilee and Al Nasse , 2014; Aydemi and Demi han, 2009),
s udying and analysing he spillo e e ec s o shocks caused by hese
a iables on banks’ s ock e u ns is c ucial.
Apa om he na ional de elopmen s, bank owne ship could also
lead o a ia ions in he indus y. Ou o he banks ope a ing in Tü kiye,
h ee a e public banks, and wo o hem ha e publicly aded sha es.
Public banks, unlike p i a e banks, a e no p o i -o ien ed ins i u ions
by na u e. Consequen ly, hese banks may eac di e en ly o in e es
a e and o eign exchange ma ke luc ua ions compa ed o p i a e
banks, i.e. he exchange a e and in e es a e isks aced by public banks
may be di e en om he isks aced by o he banks. Thus, i is c ucial o
e alua e public banks sepa a ely. On he o he hand, i is o u mos
impo ance o examine he esponse o o eign banks o changes in he
exchange a e and in e es a e in he economy, as o eign banks a e
mo e likely o ha e access o ex e nal inancing han p i a e and public
banks.
This s udy in es iga es he e ec o in e es a es, exchange a es,
and global ac o s on indi idual bank e u ns in Tü kiye. We ocus on
public banks, p i a e domes ic banks, and p i a e o eign banks since
hese ac o s ha e seen signi ican changes ecen ly. We use a ou -
a iable diagonal BEKK-GARCH me hod o examine he mean and
a iance spillo e coe icien s. By doing so, we also examine whe he
exchange a e and in e es a e a iables as well as global isk ac o s
ha e spillo e e ec s on he pe o mance o banks in Tü kiye.
The numbe o s udies in es iga ing he e ec s o exchange a es and
in e es a es on he Tu kish economy is qui e limi ed and does no co e
ecen de elopmen s in he analysis pe iods. Kandil G¨
oke and Uysal
(2020) in es iga ed he e ec s o in e es and exchange a e isks on he
Tu kish ou ism indus y’s e u ns, whe eas Kasman e al. (2011)
ocused on banks aded in BIST and examined he e ec o changes in
he in e es a e and he exchange a e on he s ock e u ns o Tu kish
banks. In he s udy o Kasman e al. (2011), he au ho s used he uni-
a ia e GARCH me hods o analyse he e ec o hese a iables. Ekinci
(2016a) used he same uni a ia e me hodology and in es iga ed he
e ec o hese ac o s on he banking sec o , in addi ion o he indus ial
and se ices sec o s. In ano he s udy, Ekinci (2016b) also in es iga ed
he e ec o c edi and ma ke isk on bank pe o mance, again using he
uni a ia e GARCH app oach.
Ou s udy di e s om hese s udies in i e ways. Fi s , unlike hese
s udies, we ocused on bo h mean and a iance spillo e coe icien s
while hei main ocus was only on ola ili y spillo e . Second, we
ocused on he ola ili y spillo e e ec o indi idual bank e u ns o
he Tu kish economy by using a diagonal BEKK-GARCH me hod, as i is
widely s a ed in he li e a u e ha he mul i a ia e me hods o ana-
lysing he spillo e e ec p o ide be e pe o mance and mo e signi -
ican esul s. Thi d, we also included he global ea index (VIX) in ou
s udy, al hough hese s udies did no use a a iable ha akes in o ac-
coun he e ec s o global ac o s. In his way, we ha e made i possible
o ou equa ions o gene a e mo e obus e o e ms. And he ou h
ac o is he pe iod o analysis o he s udy. Pa allel o eme ging ma ke s,
he u moil expe ienced by he Tu kish economy a e May 2013 due o
Fed ape ing announcemen s and a e he second hal o 2018 due o
geopoli ical ensions, inc eased isk p emiums had de e io a ed inan-
cial s abili y. An unde s anding o he impac o hese ac o s on he
pe o mance o banks is ex emely impo an o in es o s in he sec o .
As ou da a co e he pe iod in ques ion, ou s udy has he po en ial o
shed ligh on a numbe o issues. And he las one is he g ouping o he
banks ha a e conside ed in he s udy. In ou s udy, he indi idual banks
ha a e pa o he XBANK index in B˙
IST we e g ouped in o h ee
di e en ca ego ies: domes ic p i a e banks, domes ic public banks and
o eign p i a e banks, and i was analysed whe he he esul s o hese
ca ego ies we e di e en .
In he s udy, we i s buil a model based on he undamen als o
economic heo y ha sea ched o he ola ili y spillo e be ween in-
di idual bank pe o mances. In his model, we used (i) exchange a e
changes, (ii) in e es a e changes and (iii) he VIX index as ac o s
a ec ing indi idual bank e u ns. Then, based on he ola ili y se ies o
he a iables conside ed as in luencing indi idual bank e u ns, he
impac o (i) exchange a e isk, (ii) in e es a e isk a iables and (iii)
global isk a iables was examined. Since he spillo e e ec may be
empo a ily s eng hened o weakened, depending on a ious easons,
2
we ocused on whe he he exchange a e, in e es a e, and global isk
ha e a mo e subs an ial impac , exceeding a speci ic h eshold. The e-
o e, in he hi d s ep, aking in o accoun he asymme y ha isk
a iables a e mo e e ec i e abo e a ce ain h eshold, we used dummy
a iables in he mean equa ion and ep oduced he e o e ms.
Only a ew s udies in es iga ing he spillo e e ec ha e inco po-
a ed s uc u al b eaks in hei models. The exis ence o s uc u al
b eaks is uni e sally acknowledged as a ecu ing issue in daily asse
se ies - pa icula ly o less esilien eme ging ma ke s. The ICSS algo-
i hm was de ised by Sanso e al. (2004) as an app op ia e me hod o
de ec ing s uc u al b eaks in uncondi ional a iance, ecei ing
endo semen as such om Kuma and Maheswa an (2013), Kang e al.
(2009), and Mensi e al. (2014). Zi ko e al. (2015) inco po a ed
s uc u al b eaks in o he model and assessed ola ili y spillo e in he
Czech Republic, Hunga y, Poland, and Russia using da a om 2002 o
2014. The s udy employed he ICSS algo i hm o de ec mul iple s uc-
u al b eaks. The s udy obse ed ha he p esence o s uc u al b eaks in
GARCH models may lead o biased es ima es o he ola ili y spillo e
e ec . Addi ionally, he s udy disco e ed e idence suppo ing he ex-
is ence o long- e m ola ili y pe sis ence in a iance when s uc u al
b eaks a e p esen . Mensi e al. (2014) disco e ed ha s uc u al b eaks
ha e signi ican e ec s on he pe sis ence o ola ili y wi hin he da a o
1
The e u ns o he banking sec o may be a ec ed by sudden changes in
some commodi y p ices, such as oil, na u al gas, coal, gold, and sil e , o by
global ac o s such as he con ac ion o impo and expo olumes, esul ing
om an economic ecession expe ienced in o eign ade pa ne s, which
nega i ely a ec s o eign ade ela ions (Claessens e al., 2001; Kim e al.,
2000).
2
Se e al s udies ha e analysed whe he spillo e e ec s di e be o e and
a e c isis pe iods (Bekae e al., 2005; King e al., 1990; Rigobon, 2003;
Kaminsky and Reinha , 2000; Jawadi e al., 2015; Apos olakis e al., 2021) and
ound ha spillo e e ec di e s du ing hese pe iods.
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
3
Saudi A abia. Kang e al. (2009) assessed s uc u al b eaks and he
pe sis ence o ola ili y in Japanese and Ko ean s ock ma ke s o he
pe iod o 1986–2008 by employing he ICSS algo i hm and iden i ied
ha e ec i ely con olling sudden changes educes ola ili y pe sis-
ence. The e o e, in he las s ep, we included he Fou ie app oxima ion
o he co a iance equa ions, conside ing he possibili y o s uc u al
b eaks depending on he ecen de elopmen s in he Tu kish economy.
Con a y o he li e a u e on he Tu kish banking sec o , ou models
e ealed a mean spillo e om changes in he in e es a e o he pe -
o mance o he banking sec o o hal o he p i a e domes ic and
o eign banks, bu no om changes in he exchange a e and he VIX
index o he e u ns o he banking sec o o all banks in he sec o . We
we e able o ob ain simila esul s when we used he isks o he ex-
change a e, he in e es a e and he VIX index in he equa ions. We
ha e also ound ha he a e age e u ns o he public banks do no eac
o any o he a iables. Rega ding he ola ili y spillo e e ec , we
ound ha a shock o exchange a e, in e es a e and global isk a i-
ables igge ed he shock o s ock e u ns and he e ec is qui e pe sis-
en . Speci ically, we ound ha he ola ili y spillo e is mos ela ed o
global isk and leas ela ed o in e es a es. The asymme y analysis
e ealed no asymme ic s uc u e in he mean equa ions, bu s uc u al
b eaks in he co a iance equa ions, al hough he p esence o s uc u al
b eaks did no change he esul s.
The s uc u e o he pape is as ollows. Sec ion 2 b ie ly e iews he
main s udies in he ele an li e a u e. Sec ion 3 is an in oduc ion o he
econome ic me hodology o VAR-DBEKK-GARCH. Sec ion 4 p o ides a
desc ip ion o he da a and some desc ip i e s a is ics. The empi ical
esul s and indings a e p esen ed in Sec ion 5. And he inal ema ks a e
p o ided in sec ion 6.
2. Me hodology
In he inancial economics li e a u e, economis s ha e gene ally
p e e ed o use mul i a ia e GARCH me hods o model he spillo e o
con agion e ec . The ad an age o mul i a ia e GARCH me hods is ha
hey speci y equa ions ega ding he mo emen o a iances and co-
a iances o unde lying asse s o e ime (Musunu u, 2014). The e o e,
i is possible o ind se e al mul i a ia e GARCH me hods in he li e a-
u e, such as VECH-GARCH, BEKK-GARCH, hei diagonal e sions,
DCC-GARCH and CCC-GARCH me hods.
The VECH-GARCH me hod canno ensu e ha he condi ional
a iance-co a iance ma ix is posi i e semide ini e. The e o e, i is no
widely used in empi ical applica ions. To sol e he non-posi i i y
p oblem, Engle and K one (1995) p oposed he BEKK speci ica ion,
which can ensu e ha he condi ional a iance-co a iance ma ix is
posi i e semide ini e. Subsequen ly, he diagonal BEKK me hod was
de eloped o educe he numbe o pa ame e s o be es ima ed. Since he
CCC-GARCH me hod makes an un ealis ic cons an co ela ion
assump ion, he DCC-GARCH me hod is p e e ed in he li e a u e.
Howe e , Capo in and McAlee (2010) a gue ha he BEKK me hod is
he op imal me hod o es ima ing condi ional co a iances and condi-
ional co ela ions. The easons why his me hod is mo e use ul can also
be explained as ollows. Fi s , his model acili a es he analysis o
mul idimensional ela ionships. Many esea che s in he li e a u e ha e
es ima ed he ola ili y analysis be ween se e al ma ke s using he
VAR-BEKK-GARCH model (Awa ani and Maghye eh, 2013; Du and He,
2015; Jouini, 2013; among o he s). One o he ad an ages o his model
is ha i simul aneously es ima es he in e ac ion be ween he a iables
we use in he s udy. The second is ha he numbe o es ima ed co-
e icien s in he model is smalle han in o he models, such as VECH,
and he e o e he pa ame e s es ima ed by he model a e mo e s able.
Fo example, Sch eibe and Mülle (2012), S elze (2008), and Ca -
pan ie and Samkha adze (2013) ound suppo ing e idence ha
VAR-BEKK-GARCH is mo e e icien han i s coun e pa s. Following
hese a gumen s, we decided o use he ou - a ia e diagonal
BEKK-GARCH me hod p oposed by Engle and K one (1995) o
in es iga e he ola ili y link be ween e u n se ies.
2.1. The models
We buil ou - a ia e diagonal BEKK-GARCH(1,1) models, since he
main cha ac e is ics o he inancial e u n se ies a e ha hey ha e a
ails, s ong ku osis and ola ili y clus e ing, as indica ed by he wo k o
Bauwens and Lub ano (2002). The models ha we ha e es ima ed in his
s udy a e p esen ed in he ollowing.
Model 1: The Model wi h Changes in he Exchange Ra e, Changes in
he In e es Ra e and VIX index in he Mean Equa ion
In he s udy, we i s examined he e ec o he changes in he ex-
change a e (Δexc) and he changes in e es a e (Δin ) in o de o
cap u e he domes ic de elopmen s, and he e ec o he VIX index ( ix)
o ha e he global e ec s.
Y =κ0+κ1Y −1+ϵ (1)
In Eq. (1), he 4x1 dependen a iable ma ix is Y = { i
,Δexc ,Δin ,
ix }, i
is he s ocks e u ns o i bank ins i u ion unde in es iga ion
whe e i={1,2,…,12}, he 4x1 cons an ma ix is κ0={k10,k20,k30,k40},
he lagged alues ma ix is Y −1= { i
−1,Δexc −1,Δin −1, ix −1}, he 4x4
coe icien ma ix o lagged alues is κ1,
3
he e o e m ma ix is ϵ =
{
ε
1, ,
ε
2, ,
ε
3, ,
ε
4, }and U |Ω −1∼N(0,H ), he 4x4 condi ional a iance-
co a iance ma ix is H = {h1, ,h2, ,h3, ,h4, }and Ω −1 is he in o ma-
ion in he ma ke a ime −1. The diagonal BEKK-GARCH speci ica-
ion we employed in he s udy is shown in Eq. (2).
H =C
′
C+A
′
ε
−1
ε
′
−1A+B
′
H −1B(2)
In Eq. (2), H is he a iance-co a iance ma ix o Eq. (1), C is an
uppe iangula ma ix, A and B a e diagonal 4x4 pa ame e ma ices. I
we sol e he ma ices p esen ed in Eq. (2), we can ob ain he ollowing
co a iance equa ions unde in es iga ion.
4
h12, =c11c12 +a11 a22
ε
1, −1
ε
2, −1+b11b22 h12, −1
=m12 +a12
ε
1, −1
ε
2, −1+b12h12, −1(3)
h13, =c11c13 +a11 a33
ε
1, −1
ε
3, −1+b11b33 h13, −1
=m13 +a13
ε
1, −1
ε
3, −1+b13h13, −1(4)
h14, =c11c14 +a11 a44
ε
1, −1
ε
4, −1+b11b44 h14, −1
=m14 +a14
ε
1, −1
ε
4, −1+b14h14, −1(5)
whe e c11c1s=m1s, a11ass =a1s and b11bss =b1s when s={2,3,4}. The
condi ional a iance and co a iance ma ices canno be de ined nega-
i ely by i s na u e. To sa is y he mean e e ing condi ion, he co-
e icien s in Eqs. (3)–(5) a e expec ed o be less han uni y as ollowing.
[a1s+b1s]<1 whe e s={2,3,4}(6)
3
The VAR ype o coe icien s in he κ1 ma ix deno es he mean spillo e .
The mean spillo e e ec is assumed o in luence he alues o a iables unde
analysis based on hei and o he a iables’ pas alues, guided by economic
heo y. Economic heo y guided he c ea ion o he κ1 ma ix, suppo ing ou
assump ion o in e dependence be ween s ock e u n, exchange a e changes,
and in e es a e changes, all a ec ed by he VIX index, while he la e emains
una ec ed by hem.
4
The coe icien s o axy in he H ma ix (whe e x∕= y) pe ain o he ans-
mission o ola ili y. Vola ili y spillo e is p emised on he impac o a iance
changes in one a iable on he a iance changes in o he a iables.
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
4
Model 2: The Model wi h Squa e Roo o he Exchange Ra e Vola-
ili y, Squa e Roo In e es Ra e Vola ili y and VIX Index in he Mean
Equa ion
The second objec i e o he s udy is o measu e he impac o ex-
change a e isk, in e es a e isk and global isk on he e u ns o in-
di idual banks. To his end, we i s calcula ed he ola ili y o he
exchange a e and in e es a e a iables using he GARCH(1,1) me hod,
and hen de i ed he exchange a e isk (
Δexc ol
√) and in e es a e isk
(
Δin ol
√) a iables by aking he squa e oo o hese a iables. We
ha e also included he global isk a iable ( ix) in he analysis:
Y =κ0+κ1Y −1+ϵ (7)
whe e
Y ={ i
,
Δexc ol
√,
Δin ol
√, ix}
Model 3: The Model wi h Squa e Roo o he Exchange Ra e Vola-
ili y, Squa e Roo o In e es Ra e Vola ili y, VIX Index and Dummy
Va iables in he Mean Equa ion
Ano he objec i e o ou s udy is o de e mine whe he indi idual
bank e u ns espond asymme ically o isk ac o s. The e a e a numbe
o s udies, such as Kaminsky and Reinha (2000), Rigobon (2003) and
Bekae e al. (2005), which sugges ha inancial a iables espond o
isk ac o s in an asymme ic manne . Acco dingly, we belie e ha once
he isk a iables exceed a ce ain h eshold, he likelihood o indi idual
bank e u ns esponding o isk ac o s would be high. Fo his pu pose,
we de e mined he h esholds a e calcula ing he isk a iables as
desc ibed abo e. The h esholds we e de e mined by adding one s an-
da d de ia ion o he mean o each isk. Dummy a iables we e hen
c ea ed by assigning a alue o 1 o he alues whe e he isk alue
exceeds his h eshold and a alue o 0 o he alues below. Finally, he
dummy a iables a e included in he mean equa ions as in e ac ion
a iables.
Y =κ0+κ1Y −1+
τ
1D Y −1+ϵ (8)
whe e
τ
1=
⎡
⎢
⎢
⎢
⎢
⎢
⎣
0 12 13 14
0 0 0 0
0
0
0
0
0 0
0 0
⎤
⎥
⎥
⎥
⎥
⎥
⎦
and D =
⎡
⎢
⎢
⎢
⎢
⎢
⎣
0
dum
Δexc ol
√
dum
Δin ol
√
dum ix
⎤
⎥
⎥
⎥
⎥
⎥
⎦
Model 4 and 5: Models Conside ing S uc u al B eaks in he Va iance
Equa ions
Some s udies, such as Mensi e al. (2019) and Begiazi and Ka siampa
(2019), ha e s a ed ha he e a e likely o be s uc u al b eaks in he
a iance equa ions o inancial a iables, and ha hese b eaks ha e a
se ious impac on he signi icance le el o he es ima ed coe icien s.
The e o e, we e-es ima e models 1 and 2 by adding he Fou ie
app oxima ion o he a iance equa ions, a he han using dummy
a iables o cap u e exogenous s uc u al b eaks.
The use o he Fou ie app oxima ion has se e al ad an ages. Fi s , i
allows he s uc u al b eak o be es ima ed endogenously. Second, i
allows us o cap u e mul iple unknown s uc u al b eak da es. And hi d,
i allows us o a oid he di icul p ocedu e o es ima ing he b eak da e
(Bane jee e al., 2017). Becke e al. (2006) show ha he Fou ie
app oxima ion wo ks qui e well in he p esence o unknown b eaks in
he se ies. The a iance-co a iance equa ions associa ed wi h he
Fou ie app oxima ion a e as ollows.
H =C
′
C+A
′
ε
−1
ε
′
−1A+B
′
H −1B
+∑
n
k=1
D
′
cos(2
π
k /T)D+∑
n
k=1
E
′
sin(2
π
k /T)E(9)
Whe e n<T/2, k is he pa icula equency, is he end and T is he
numbe o obse a ions unde in es iga ion. D and E ma ixes a e
diagonal.
Model 6: Model Conside ing S uc u al B eaks in he Va iance
Equa ions and Dummy Va iables in he Mean Equa ion
Finally, we simul aneously analysed he e ec o s uc u al b eaks
and he asymme ic esponse o indi idual bank e u ns o isk ac o s.
In his way, we ha e ied o ensu e ha he e o e ms in he s udy
e lec bo h he asymme ic e ec and he s uc u al b eak oge he . In
his analysis, he mean equa ion is de e mined as in Equa ion (8), and
he a iance equa ion is de e mined as in Equa ion (9).
3. Da a
The s ock e u ns o indi idual banks ( i) a e calcula ed by aking he
loga i hmic di e ence o he closing p ice (P) a ime [ i, = (ln Pi, −
ln Pi, −1)*100]. The indi idual banking indices used in he s udy a e
p i a e domes ic banks [Akbank (AKBNK), Tü kiye ˙
Is¸bank A O d Shs
(ISATR), Tü kiye ˙
Is¸bank B O d Shs (ISBTR), Tü kiye ˙
Is¸bank C O d Shs
(ISCTR), S¸eke bank (SKBNK), Tü kiye Sınai Kalkınma Bank (TSKB), and
Yapı K edi Bank (YKBNK)], public domes ic banks [Halkbank (HALKB)
and Vakı bank (VAKBN)] and p i a e o eign banks [Alba aka Tü k
Bank (ALBRK), Ga an i Bank (GARAN), ICBC Tü kiye Bank (ICBCT)].
Change in he nominal exchange a e (Δexc) compu ed as he log di -
e ence o baske a e (exc) o e he p e ious day ob ained by weigh ing
50%–50% o USD/TRY and EUR/TRY a es [Δexc = (ln exc −
ln exc −1)*100].
5
The in e es a e is he in e es a e o a 2-yea go -
e nmen bonds (TR2) and he changes in in e es a e (Δin ) a e calcu-
la ed by he i s di e ence o e he p e ious wo king day (Δin =in −
in −1) The da a used in he s udy a e in daily equency and co e s he
pe iod be ween Janua y 4, 2005 and Ma ch 28, 2023. Howe e , he
sample pe iod a ies o some o he a iables due o da a a ailabili y.
6
Da a a e ob ained om he elec onic da a deli e y sys em o he Cen al
Bank o he Republic o Tü kiye (CBRT) and Bloombe g.
4. Es ima ion esul s
4.1. Calcula ing he condi ional co a iance equa ion
This sec ion p esen s he es ima ion esul s o he abo e models. In
o de o make he ollowing ables easie o ollow, he e u n o a bank
(AKBNK) has been aken as an example and de ailed in o ma ion on one
o he p edic ed models has been p esen ed. In he i s s ep, we se up
he mean equa ion as in equa ion (1) and es ima ed he coe icien s o
he VAR(1,1) model wi h he κ1 ma ix. In he second s ep, we pe o med
he ou - a ia e diagonal BEKK-GARCH(1,1) me hod o ob ain he di-
agonal coe icien s. In he hi d s ep, we mul iplied he diagonal co-
e icien s as in equa ions (3)–(5) o ob ain he ola ili y spillo e
coe icien s o sea ch o he exis ence o a spillo e e ec be ween bank
e u ns and he a iables unde in es iga ion. The signi icance o he
coin eg a ion coe icien s is es ed using he Wald es . Table 1 shows he
5
Al e na i e exchange a es o he han 50%–50% we e also used in he
s udy, bu since i was de e mined ha he esul s did no di e , he esul s
ega ding o he exchange a es we e no epo ed. These esul s a e a ailable
upon eques .
6
Desc ip i e s a is ics and uni oo es esul s a e p esen ed in Appendix B.
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
5
es ima ed esul s ob ained om he VAR(1,1)-DBEKK-GARCH(1,1)
me hod using he e u ns o AKBNK.
The e a e h ee pa s o Table 1. The i s pa shows he coe icien s
o he mean equa ion. Since one o he objec i es o he s udy is o
de e mine he e ec o exchange a e, in e es a e and VIX index on he
e u n o indi idual banks, he coe icien s k12, k13 and k14 in he mean
equa ion p o ide c ucial in o ma ion abou hei e ec s. In Table 1, he
alues o k12, k13 and k14 a e 0.0213, −0.5732 and −0.0034 espec-
i ely. As can be seen om he able, changes in in e es a es ha e a
nega i e e ec on he e u ns o AKBNK, which is signi ican a he 1%
le el, while changes in in e es a es and he global isk ac o ha e no
e ec on he e u ns o he indi idual banks s udied. On he o he hand,
we ound ha , in line wi h inancial heo y, AKBNK’s e u ns ha e a
s ong nega i e e ec on changes in exchange a es and in e es a es.
Howe e , since ou aim in his s udy is only o y o de e mine he e -
ec s o exchange a es and in e es a es on s ocks, he e ec s o s ock
e u ns on exchange a es and in e es a es a e no men ioned in he
ollowing pa s o he s udy.
We used he DBEKK-GARCH (1,1) me hod o collec he coe icien s
o he M, A and B ma ices om he second pa o Table 1, since he aim
o ou s udy is o examine he ola ili y spillo e . A e he analysis o
he coe icien s, we ound ha all he coe icien s in he M, A and B
ma ices a e s a is ically signi ican , wi h he excep ion o m12. This
indica es ha hese coe icien s can be used o ob ain he equa ions o
a iance and co a iance, which a e shown in equa ions (4)–(6). The
calcula ed ARCH coe icien s o 0.0496, 0.0463 and 0.0638 a e s a is-
ically signi ican and imply a spillo e e ec in he e u n o AKBNK and
changes in he exchange a e, in e es a e and VIX index, espec i ely.
Fu he mo e, he calcula ed GARCH coe icien s o 0.9206, 0.9405 and
0.8579 a e s a is ically signi ican and demons a e a s ong pe sis ence
in he GARCH e ms. In o he wo ds, he GARCH e ms sugges ha he
ola ili y spillo e pe sis s e en a e he ini ial spillo e pe iod.
The e ec o he ola ili y spillo e las s longe when he GARCH
e m ge s close o 1. The s abili y condi ion holds equal signi icance o
he alues and signi icance o he co a iance coe icien s. In Table 1, we
ound ha he sum o ARCH and GARCH e ms a e less han one (0.
9702, 0.9867, 0.9216) which indica es ha he p o ided condi ional
co a iance coe icien s a e gua an eed o be s a iona y. Fig. 1 demon-
s a es he co a iance g aphs de i ed om ou model.
4.2. E ec s o he exchange a e changes, in e es a e changes and VIX
index on indi idual bank e u ns
4.2.1. Mean spillo e
Table 2 also con ains he alues om Table 1, making i easie o
eade s o unde s and he unde lined numbe s. Bold coe icien s indi-
ca e s a is ical signi icance. As pe Table 2, Model 1 shows mean
spillo e e ec s om in e es a e changes o he e u ns o i e speci ic
banks. Al hough he expec ed sign o se en o he banks is in ques ion,
no signi ican ela ionship could be iden i ied. An assessmen , di e -
en ia ing domes ic p i a e, domes ic public and o eign p i a e banks,
has ound ha only domes ic public bank e u ns emain una ec ed by
in e es a e changes. Sepa a ing he wo a o emen ioned banks, hal o
he emaining banks we e ound o espond o in e es a e changes.
The si ua ion becomes e en mo e in iguing when changes in ex-
change a es a e aken in o accoun . The analysis e ealed ha bank
e u ns, excep o ISCTR, do no espond o changes in exchange a es.
This implies ha in es o s who in es in s ocks do no conside exchange
a e changes as heo ized. The possible eason o his could be ha
policymake s ha e implemen ed a ious measu es o a ec he ola ili y
and changes in exchange a es. Hence, some economic agen s ha e
u ned o al e na i e in es men ins umen s such as eal es a e o gold.
I is c ucial o compa e he esul s o ou s udy wi h p e ious s udies
in he li e a u e. Kasman e al. (2011) s udied he e ec s o changes in
exchange a es and in e es a es on he indi idual banks’ pe o mance
in Tü kiye. The s udy ound ha he e ec o he exchange a e change
was mo e signi ican han he e ec o he change in he in e es a e. In
o he wo ds, bo h a iables a ec he e u ns, bu he e ec o exchange
a e changes is mo e p onounced. Howe e , ou s udy ound ha he
e ec o changes in exchange a es does no exis , and he e ec o
changes in in e es a es is no ques ionable o all indi idual banks.
Acco ding o Ekinci (2016a and 2016b), who used he same me hodol-
ogy and analysed he mo e s able pe iod o 2002–2015, he exchange
a e had a signi ican impac on he banking sec o ’s pe o mance,
whe eas changes in in e es a es had no no able e ec s. When we
e alua ed he eason o he di e gence be ween he indings, we
conside ed ha he key ac o was he ise in geopoli ical ensions and
isk p emiums in Tü kiye. Pa ially s a ing in 2013, and mo e e iden ly
a e 2018, inc eased in isk p emiums esul ed in in e es a e decisions
a ec ing se e al economic a iables in addi ion o s ock e u ns. As a
esul , hese de elopmen s ha e a signi ican impac on banks’ p o i -
abili y, wi h changes in in e es a es ha ing a p onounced e ec on
bank e u ns.
Upon examining he impac o he global isk ac o , i has been
ound ha he e is no mean spillo e e ec on he VIX index e u ns.
This indica es ha he e u ns o he banks included in he banking index
in Tü kiye a e wholly de e mined by domes ic ac o s. As p e ious
li e a u e did no accoun o he global isk ac o o Tu kish indi idual
banks, i was no easible o ma ch hese indings wi h hose o o he
s udies.
4.2.2. Vola ili y spillo e
The main objec i e o ou esea ch is he in es iga ion o ola ili y
spillo e be ween a iables. Analysis o Table 2 de e mined he p esence
Table 1
The esul s o VAR(1,1)-DBEKK-GARCH(1,1) model o he e u ns o AKBNK.
Coe icien s in Mean Equa ion i= {1,2,3,4}and j= {1,2,3,4}
κ1 κ0
ki1 ki2 ki3 ki4 ki0
AKBNK k1j −0.0156 0.0213 ¡0.5732*** −0.0034 0.1575**
Δexc k2j ¡0.0435*** 0.0619*** 0.9416*** 0.0022*** −0.0152
in k3j ¡0.0058*** 0.0066*** −0.0089 ¡0.0005** 0.0035
ix k4j – – – 0.9798*** 0.3149***
Coe icien s o Co a iance Equa ions s= {2,3,4}S abili y Condi ions s= {2,3,4}
m1s a1s b1s a1s+b1s
h12 0.0001 0.0496*** 0.9206*** 0.9702***
h13 ¡0.0025*** 0.0463*** 0.9405*** 0.9867***
h14 ¡0.0399*** 0.0638*** 0.8579*** 0.9216***
*, ** and *** indica es signi icance a 10%, 5% and 1% le el, espec i ely.
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
6
o ola ili y spillo e in all h ee co a iance equa ions. The coe icien s
we e s a is ically signi ican and s ong. The s udy ound ha SKBNK’s
indi idual bank e u ns had he s onges ola ili y spillo e (0.2123,
0.1668, 0.2431) compa ed o he change in he exchange a e, he
change in he in e es a e and he VIX index, whe eas ISCTR’s e u ns
had he weakes spillo e (0.0416, 0.0386, 0.0535). Focusing on he
a e age o he ola ili y spillo e , we obse ed ha he ola ili y spill-
o e o he global isk ac o was s onge (0.1107) han he change in
he in e es a e (0.0788).
7
An addi ional no able inding pe ains o he
exchange a e a ia ion. While he shi in exchange a e lacks s a is ical
signi icance in he mean equa ion o local banks, spillo e o ola ili y
has been ound o ha e a subs an ial impac . This sugges s ha he
ola ili y o he exchange a e is mo e signi ican han changes in he
exchange a e conce ning s ock ansac ions. All he indings a e
consis en wi h he esea ch o Kasman e al. (2011) and Ekinci (2016a,
2016b), indica ing ha bank e u ns a e impac ed by exchange a es.
The indings demons a e ha a mo emen in he exchange a e, in e es
a e o global isk ac o has a spillo e e ec on he e u ns o all banks
wi hin he Tu kish economy. Focusing on pe sis ence in GARCH co-
e icien s, we no e ha he igu es a e qui e high, sugges ing ha he
ola ili y spillo e akes longe among he changes in he a iables and
indi idual bank e u ns.
4.3. E ec s o he exchange a e, in e es a e and global isks on
indi idual bank e u ns
The p e iously conduc ed model esul s p o ide us wi h in o ma ion
on he ola ili y spillo e be ween indi idual bank pe o mances and
c i ical economic a iables. Clea ly, he isk ac o s associa ed wi h he
in es iga ed a iables, as well as hei a ia ions, a ec he indi idual
banks’ pe o mance. Hence, we es ablished he mean equa ion shown in
Eq. (7) and p oceeded o e-es ima e he model. The es ima ion ou -
comes a e p esen ed in Table 3.
While Table 2 con i med ha changes in he in e es a e had a
pa ial impac on he a e age e u ns o he bank, Table 3 e ealed ha
his e ec was no signi ican o he in e es a e. The si ua ion changes
o ola ili y spillo e . The co a iance equa ions e ealed ha all h ee
explana o y a iables a e esponsible o igge ing ola ili y spillo e .
The indings exhibi ha a change in o eign exchange, bond, and s ock
ma ke s leads o ac ion in associa ion wi h one ano he , bu i is no
subs an ial enough o impac he mean e u ns. These indings a e
consis en wi h hose o p e ious s udies conduc ed by Kasman e al.
(2011) and Ekinci (2016a, 2016b), which had shown a obus ola ili y
spillo e among a iables unde e alua ion. I is also essen ial o no e
ha as in he ea lie model, he coe icien s o ola ili y pe sis ence a e
high.
4.3.1. Asymme y analysis
As men ioned abo e, he asymme ic pa e n o inancial da a is well
documen ed in he li e a u e (Razzaq e al., 2021; Lee and Lee, 2022,
among o he s). The lack o add essing asymme y in p e ious s udies on
he Tu kish economy was conside ed a signi ican d awback. The second
model was edesigned using dummy a iables o se he h eshold pe
his eali y. We assumed ha i he isk exceeds a speci ic h eshold, i
will ha e a mo e signi ican impac on indi idual bank e u ns. Table 4
shows he esul s ob ained om he model de ined in Equa ion (8).
Due o space limi a ions, Table 4 omi s he κ1 coe icien s and ins ead
p esen s he coe icien s o he h eshold a iables (
τ
1= { 12, 13, 14}).
The coe icien s ob ained om he analysis we e no signi ican ly
di e en om hose ob ained in he Model 2 o Table 3. In o he wo ds,
he model’s indings do no indica e signi ican changes in he mean
equa ions, as coe icien s o dummy a iables a e s a is ically insig-
ni ican (excep ISCTR which was al eady signi ican in Table 3).
8
These
indings indica e ha no asymme ical e ec is p esen in he mean
equa ion when aking isk ac o s in o accoun . Howe e , examining he
co a iance equa ions e ealed ha he e is s ill a s ong spillo e e ec
be ween indi idual bank e u ns and he a iables examined. The
spillo e o ola ili y be ween indi idual bank e u ns, exchange a e,
Fig. 1. Co a iance o AKBNK and se ies unde in es iga ions.
7
The a e age ola ili y spillo e coe icien s ha e been calcula ed o all
banks.
8
Sea ching o he asymme ic e ec o he model including exchange a e
changes, in e es a e changes and VIX index, we eached he same esul : no
asymme ic e ec . We conduc ed al e na i e calcula ions while de i ing he
dummy a iables, bu none o hem showed signi ican e ec . We didn’ epo
hese indings due o space limi a ions, bu we can p o ide hem upon eques .
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
7
Table 2
Es ima ion esul s o he model ha includes changes in exchange a es, changes in in e es a es and ola ili y index.
MODEL 1 Changes in Exchange Ra es Changes in In e es Ra es VIX Index S abili y Condi ions
Mean eq. Va iance eq. (h12 ) Mean eq. Va iance eq. (h13 ) Mean eq. Va iance eq. (h14 )
k12 m12 a12 b12 k13 m13 a13 b13 k14 m14 a14 b14 a12 +b12 a13 +b13 a14 +b14
Domes ic P i a e Banks in XBANK Index
AKBNK 0.0213 0.0001 0.0496 0.9206 ¡0.5732 ¡0.0025 0.0463 0.9405 −0.0034 ¡0.0399 0.0638 0.9206 0.9702 0.9868 0.9844
(0.0346) (0.0012) (0.0019) (0.0026) (0.0981) (0.0003) (0.0018) (0.0015) (0.0038) (0.0058) (0.0023) (0.0026)
ISATR −0.0455 0.0003 0.1100 0.8379 −0.1589 0.0002 0.0850 0.9057 −0.0010 −0.0031 0.1231 0.8379 0.9479 0.9907 0.9610
(0.0199) (0.0024) (0.0019) (0.003) (0.1171) (0.0003) (0.0013) (0.0009) (0.0026) (0.0045) (0.0019) (0.003)
ISBTR 0.0501 0.0012 0.0757 0.9063 −0.3411 −0.0011 0.0713 0.9246 0.0031 −0.0172 0.0973 0.9063 0.9820 0.9959 1.0036
(0.0305) (0.0017) (0.002) (0.0026) (0.0921) (0.0003) (0.0018) (0.0015) (0.0038) (0.0059) (0.0023) (0.0026)
ISCTR 0.1234 0.0013 0.0416 0.8774 ¡0.4793 ¡0.0087 0.0386 0.8946 −0.0018 ¡0.0513 0.0535 0.8774 0.9190 0.9332 0.9309
(0.0143) (0.0022) (0.0033) (0.008) (0.0993) (0.0008) (0.0031) (0.0079) (0.0046) (0.0072) (0.004) (0.008)
SKBNK 0.0029 0.0006 0.2123 0.7423 −0.1718 −0.0002 0.1668 0.7994 −0.0079 ¡0.0063 0.2431 0.7423 0.9546 0.9662 0.9854
(0.0223) (0.0023) (0.0047) (0.0044) (0.0769) (0.0006) (0.0032) (0.0037) (0.0026) (0.0072) (0.0048) (0.0044)
TSKB −0.0202 −0.0001 0.0500 0.9235 −0.3614 ¡0.0019 0.0473 0.9401 0.0006 ¡0.0401 0.0643 0.9235 0.9735 0.9874 0.9878
(0.0338) (0.0011) (0.002) (0.0025) (0.1066) (0.0003) (0.0018) (0.0014) (0.0037) (0.0053) (0.0024) (0.0025)
YKBNK −0.0103 0.0004 0.0489 0.9198 ¡0.5879 ¡0.0029 0.0463 0.9380 −0.0003 ¡0.0461 0.0636 0.9198 0.9687 0.9843 0.9834
(0.0331) (0.0013) (0.0021) (0.0027) (0.0931) (0.0003) (0.002) (0.0017) (0.0037) (0.0057) (0.0025) (0.0027)
Domes ic Public Banks in XBANK Index
HALKB 0.0038 0.0010 0.0475 0.9212 −0.2956 ¡0.0032 0.0444 0.9381 0.0058 ¡0.0556 0.0634 0.9212 0.9687 0.9825 0.9846
(0.0364) (0.0015) (0.0022) (0.0028) (0.1383) (0.0003) (0.002) (0.002) (0.0043) (0.0081) (0.0027) (0.0028)
VAKBN 0.0186 0.0015 0.0491 0.9215 −0.2868 ¡0.0031 0.0465 0.9381 −0.0015 ¡0.0484 0.0637 0.9215 0.9706 0.9846 0.9852
(0.0376) (0.0013) (0.0021) (0.0026) (0.1151) (0.0003) (0.002) (0.0017) (0.004) (0.0062) (0.0026) (0.0026)
Fo eign P i a e Banks in XBANK Index
ALBRK −0.0006 0.0006 0.0871 0.8979 ¡0.4859 ¡0.0024 0.0879 0.9071 −0.0042 ¡0.0356 0.1218 0.8979 0.9850 0.9950 1.0197
(0.0321) (0.0016) (0.0026) (0.0028) (0.0975) (0.0003) (0.0025) (0.0023) (0.0036) (0.0073) (0.0033) (0.0028)
GARAN 0.0551 0.0006 0.0460 0.9189 ¡0.5603 ¡0.0025 0.0411 0.9440 −0.0035 ¡0.0438 0.0585 0.9189 0.9649 0.9851 0.9774
(0.0387) (0.0013) (0.0018) (0.0027) (0.1051) (0.0003) (0.0016) (0.0013) (0.0037) (0.006) (0.0022) (0.0027)
ICBCT 0.0181 0.0020 0.0783 0.7659 −0.3760 ¡0.0127 0.0737 0.7799 −0.0063 ¡0.0781 0.1003 0.7659 0.8442 0.8536 0.8662
(0.0569) (0.0063) (0.004) (0.0144) (0.1821) (0.0016) (0.0037) (0.0145) (0.0066) (0.0177) (0.0048) (0.0144)
(i) a1s indica es ola ili y spillo e whe e a11*ass and (s=2,3,4)(ii) b1s indica es pe sis ence in ola ili y spillo e whe e b11*bss and (s=2,3,4)(iii) Values in pa en heses gi e s anda d e o s., (i ) Bold in mean a iance
equa ions means ha he coe icien is signi ican a 1% le el. ( ) Bold in s abili y condi ions mean he s abili y condi ion is hold.
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
8
Table 3
Es ima ion esul s o he model ha includes exchange a e isk, in e es a e isk and global isk.
MODEL 2 Exchange Ra e Risk In e es Ra e Risk Global Risk S abili y Condi ions
Mean eq. Va iance eq. (h12 ) Mean eq. Va iance eq. (h13 ) Mean eq. Va iance eq. (h14 )
k12 m12 a12 b12 k13 m13 a13 b13 k14 m14 a14 b14 a12 +b12 a13 +b13 a14 +b14
Domes ic P i a e Banks in XBANK Index
AKBNK 0.0013 −0.0007 0.0488 0.9253 −0.0220 ¡0.0026 0.0461 0.9401 −0.0024 ¡0.0390 0.0624 0.8623 0.9741 0.9862 0.9247
(0.0606) (0.0012) (0.0019) (0.0021) (0.2047) (0.0003) (0.0018) (0.0015) (0.0038) (0.0055) (0.0023) (0.0033)
ISATR −0.0280 0.0003 0.0780 0.9093 −0.0153 −0.0001 0.0673 0.9297 0.0038 0.0060 0.0945 0.8533 0.9873 0.9970 0.9478
(0.0399) (0.0016) (0.0013) (0.0017) (0.2091) (0.0003) (0.0011) (0.0009) (0.003) (0.0048) (0.0015) (0.003)
ISBTR 0.0499 0.0011 0.0763 0.9106 0.0513 −0.0011 0.0714 0.9242 0.0020 −0.0150 0.0966 0.8509 0.9869 0.9956 0.9475
(0.0856) (0.0017) (0.002) (0.0021) (0.2456) (0.0003) (0.0018) (0.0015) (0.0038) (0.0056) (0.0023) (0.0031)
ISCTR 0.4739 −0.0006 0.0526 0.8677 −0.7255 ¡0.0094 0.0492 0.8814 −0.0046 ¡0.0530 0.0669 0.8112 0.9203 0.9306 0.8781
(0.0194) (0.0025) (0.0033) (0.0077) (0.1913) (0.0008) (0.0031) (0.0078) (0.0045) (0.0076) (0.0039) (0.0073)
SKBNK −0.0975 0.0007 0.2015 0.7580 0.1971 −0.0001 0.1673 0.7995 −0.0075 −0.0062 0.2429 0.7086 0.9595 0.9668 0.9515
(0.052) (0.0022) (0.0043) (0.0042) (0.1538) (0.0006) (0.0033) (0.0038) (0.0025) (0.0069) (0.0048) (0.0044)
TSKB −0.0940 0.0002 0.1404 0.8274 0.2126 ¡0.0002 0.1165 0.8744 −0.0019 −0.0014 0.1661 0.7761 0.9678 0.9909 0.9422
(0.0658) (0.0017) (0.004) (0.0034) (0.161) (0.0004) (0.0031) (0.0025) (0.0027) (0.0051) (0.0042) (0.0037)
YKBNK −0.0102 −0.0011 0.0487 0.9226 ¡0.0432 ¡0.0030 0.0454 0.9385 0.0004 ¡0.0452 0.0619 0.8588 0.9713 0.9839 0.9207
(0.0605) (0.0012) (0.0021) (0.0023) (0.1976) (0.0003) (0.0019) (0.0017) (0.0037) (0.0054) (0.0024) (0.0034)
Domes ic Public Banks in XBANK Index
HALKB −0.0548 −0.0005 0.1607 0.7846 0.1064 0.0002 0.1321 0.8308 0.0018 −0.0006 0.1986 0.7203 0.9453 0.9629 0.9189
(0.0623) (0.0027) (0.0053) (0.0051) (0.1802) (0.0005) (0.0042) (0.0049) (0.003) (0.0086) (0.0062) (0.0054)
VAKBN −0.0764 0.0007 0.0492 0.9236 −0.1270 ¡0.0031 0.0460 0.9383 0.0000 ¡0.0471 0.0625 0.8579 0.9728 0.9843 0.9204
(0.0772) (0.0013) (0.0021) (0.0023) (0.212) (0.0003) (0.002) (0.0018) (0.0041) (0.006) (0.0026) (0.0035)
Fo eign P i a e Banks in XBANK Index
ALBRK −0.0425 −0.0008 0.0842 0.9029 0.0009 ¡0.0023 0.0865 0.9083 −0.0030 ¡0.0353 0.1172 0.8165 0.9871 0.9948 0.9337
(0.0737) (0.0016) (0.0024) (0.0025) (0.2083) (0.0003) (0.0024) (0.0022) (0.0036) (0.007) (0.0031) (0.0043)
GARAN −0.0138 −0.0004 0.0450 0.9250 0.0858 ¡0.0026 0.0410 0.9436 −0.0030 ¡0.0422 0.0571 0.8621 0.9700 0.9846 0.9192
(0.068) (0.0012) (0.0018) (0.0021) (0.2047) (0.0003) (0.0016) (0.0014) (0.0037) (0.0057) (0.0022) (0.0033)
ICBCT −0.0288 −0.0023 0.0816 0.7607 0.0700 ¡0.0135 0.0762 0.7713 −0.0077 ¡0.0773 0.1022 0.7121 0.8423 0.8475 0.8143
(0.1038) (0.0062) (0.0041) (0.0148) (0.3069) (0.0016) (0.0039) (0.015) (0.0067) (0.0178) (0.0049) (0.0139)
(i) a1s indica es ola ili y spillo e whe e a11*ass and (s=2,3,4)(ii) b1s indica es pe sis ence in ola ili y spillo e whe e b11*bss and (s=2,3,4)(iii) Values in pa en heses gi e s anda d e o s., (i ) Bold in mean a iance
equa ions means ha he coe icien is signi ican a 1% le el. ( ) Bold in s abili y condi ions mean he s abili y condi ion is hold.
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
15
Table 6
Es ima ion Resul s o Risks wi h S uc u al B eak
Table 7
Es ima ion Resul s o Risks wi h S uc u al B eak and Dummy
Re e ences
Ab eu, M., Mendes, V., 2010. Financial Li e acy and Po olio Di e si ica ion. h ps
://www. esea chga e.ne /publica ion/227351484_Financial_Li e acy_and_Po olio_
Di e si ica ion. (Accessed 21 Oc obe 2019).
Alja ayesh, N.I.A., As ou , L.K., Al-Abdallah, S.Y., 2018. In e es a es ola ili y and i s
consequences on s ock e u ns: he case s udy om Amman S ock Exchange, Jo dan.
J. Econ. Lib. 5 (2), 149–160.
Apos olakis, G.N., Flo os, C., Gkillas, K., Woha , M., 2021. Financial s ess, economic
policy unce ain y, and oil p ice unce ain y. Ene gy Econ. 104, 105686.
Assidenou, K.E., 2011. Coin eg a ion o majo s ock ma ke indices du ing he 2008
global inancial dis ess. In . J. Econ. Finance 3 (2), 212–222.
Aydemi , O., Demi han, E., 2009. The ela ionship be ween s ock p ices and exchange
a es: e idence om Tü kiye. In . Res. J. Financ.Econ. 1 (23), 207–215.
Balcila , M., Demi e , R., 2015. E ec o global shocks and ola ili y on he d beha io in
an eme ging ma ke : e idence om Bo sa Is anbul. Eme g. Ma k. Finance T ade 51
(1), 140–159.
Bane jee, P., A ˇ
cabi´
c, V., Lee, H., 2017. Fou ie ADL coin eg a ion es o app oxima e
smoo h b eaks wi h new e idence om c ude oil ma ke . Econ. Modell. 67, 114–124.
Bauwens, L., Lub ano, M., 2002. Bayesian op ion p icing using asymme ic GARCH
models. J. Empi . Finance 9 (3), 321–342.
Becke , R., Ende s, W., Lee, J., 2006. A s a iona i y es in he p esence o an unknown
numbe o smoo h b eaks. J. Time Anal. 27 (3), 381–409.
Begiazi, K., Ka siampa, P., 2019. Modelling UK house p ices wi h s uc u al b eaks and
condi ional a iance analysis. J. R. Es a e Finance Econ. 58, 290–309.
Bekae , G., Ha ey, C.R., Lundblad, C., 2005. Does inancial libe aliza ion spu g ow h?
J. Financ. Econ. 77 (1), 3–55.
BRSA, 2023. Do we eally need bo h BEKK and DCC? A ale o wo mul i a ia e GARCH
models. J. Econ. Su . 26 (4), 736–751. Tu kish Banking Sec o Key Indica o s
Repo , Ma ch 2023, Capo in, M., and McAlee , M. (2010).
Ce o elli, N., Goldbe g, L.S., 2011. Global banks and in e na ional shock ansmission:
e idence om he c isis. IMF Econ. Re . 59 (1), 41–76.
Claessens, S., Do nbush, R., Pa k, Y.C., 2001. Con agion: why c ises sp ead and how his
can Be s opped", ss. 19-41. In: Claessens, S., Fo bes, K.J. (Eds.), In e na ional
Financial Con agion. Kluwe Academic Publishe s, Bos on/Dod ech /London.
Coeu dacie , N., Guibaud, S., 2011. In e na ional po olio di e si ica ion is be e han
you hink. J. In . Money Finance 30, 289–308.
Demi alp, S., Demi alp, S., 2019. E osion o cen al bank independence in Tü kiye. Tu k.
S ud. 20 (1), 49–68.
S. Çiçek and A. Yıldı ım
Cen al Bank Re iew 24 (2024) 100139
16
Demi guc Kun , A., Huizinga, H., 1999. De e minan s o comme cial bank in e es
ma gins and p o i abili y: some in e na ional e idence. Wo ld Bank Econ. Re . 13
(2), 379–408.
Den Haan, W.J., Sumne , S.W., Yamashi o, G.M., 2007. Bank loan po olios and he
mone a y ansmission mechanism. J. Mone a y Econ. 54 (3), 904–924.
Dungey, M., Fla in, T.J., Lagoa-Va ela, D., 2020. A e banking shocks con agious?
E idence om he eu ozone. J. Bank. Finance 112, 105386.
Ekinci, A., 2016a. Faiz ile ku daki ge i i e de˘
gis
¸kenli˘
gin bankacılık, sanayi e hizme le
sek ¨
o ü pe o mansına e kisi. Bankacıla De gisi 97, 73–82.
Ekinci, A., 2016b. The e ec o c edi and ma ke isk on bank pe o mance: e idence
om Tü kiye. In . J. Econ. Financ. Issues 6 (2), 427–434.
Ende s, W., Lee, J., 2012. A uni oo es using a Fou ie se ies o app oxima e smoo h
b eaks. Ox . Bull. Econ. S a . 74 (4), 574–599.
Engle, R.F., K one , K.F., 1995. Mul i a ia e simul aneous gene alized ARCH. Econom.
Theo . 122–150.
Fe ei a Ma ins, A.D.F., Ma ia Bo olon, P., Mo he Maia, V., 2021. On he ela ionship
be ween implied ola ili y index and s ock e u n index. Re . Con emp. Con ab. 18
(49).
Flanne y, M.J., James, C.M., 1984. The e ec o in e es a e changes on he common
s ock e u ns o inancial ins i u ions. J. Finance 39 (4), 1141–1153.
Fo bes, K., Rigobon, R., 2002. No con agıon, only in e dependence: measu ıng s ock
ma ke Co-mo emen s. J. Finance 57 (5), 2223–2261.
G amma ikos, T., Saunde s, A., Swa y, I., 1986. Re u ns and isks o U.S. Bank o eign
cu ency ac i i ies. J. Finance 41 (3), 671–682.
Hajilee, M., Al Nasse , O.M.A., 2014. Exchange a e ola ili y and s ock ma ke
de elopmen in eme ging economies. J. Pos Keynes. Econ. 37 (1), 163–180.
Hajilee, M., Nasse , O.M.A., 2017. The impac o in e es a e ola ili y on s ock ma ke
de elopmen : e idence om eme ging ma ke s. J. De elop. A ea. 51 (2), 301–313.
Hancock, D., 1985. Bank p o i abili y, in e es a es, and mone a y policy. J. Money
C edi Bank. 17 (2), 189–202.
Jawadi, F., Louhichi, W., Idi Che ou, A., 2015. In aday bidi ec ional ola ili y spillo e
ac oss in e na ional s ock ma ke s: does he global inancial c isis ma e ? Appl.
Econ. 47 (34–35), 3633–3650.
Kaminsky, G.L., Reinha , C.M., 2000. On c ises, con agion, and con usion. J. In . Econ.
51 (1), 145–168.
Kandil G¨
oke , ˙
I.E., Uysal, B., 2020. Tu izm sek ¨
o ünde aiz o anı e d¨
o iz ku unun
kˆ
a lılık üze indeki e kisi. ¨
Ome Halisdemi Üni e si esi ˙
I˙
IBF De gisi 13 (1), 159–170.
Kang, S.H., Cho, H.G., Yoon, S.M., 2009. Modeling sudden ola ili y changes: e idence
om Japanese and Ko ean s ock ma ke s. Phys. S a . Mech. Appl. 388, 3543–3550.
Kasman, A., 2012. Cos e iciency, scale economies, and echnological p og ess in
Tu kish banking. Cen al Bank Re . 2 (1), 1–20.
Kasman, S., Va da , G., Tunç, G., 2011. The impac o in e es a e and exchange a e
ola ili y on banks’ s ock e u ns and ola ili y: e idence om Tü kiye. Econ.
Modell. 28, 1328–1334.
Kim, S.H., K¨
ose, M.A., Plumme , M.G., 2000. Unde s anding he Asian Con agion: “An
In e na ional Business Cycle Pe spec i e”, ols. 2000–14. G adua e School o
In e na ional Economics and Finance B andeis Uni e si y. Wo king Pape Se ies.
King, M.A., Sen ana, E., Wadhwani, S., 1990. Vola il iy and Links be ween Na ional
S ock Ma ke s. NBER Wo king Pape s 3357. Na ional Bu eau o Economic Resea ch,
Inc.
Kuma , D., Maheswa an, S., 2013. De ec ing sudden changes in ola ili y es ima ed om
high, low and closing p ices. Econ. Modell. 31, 484–491.
Lae en, L., Valencia, F., 2013. The eal e ec s o inancial sec o in e en ions du ing
c ises. J. Money C edi Bank. 45 (1), 147–177.
Lee, W.S., Lee, H.S., 2022. Asymme ic ola ili y ansmission ac oss No heas Asian
s ock ma ke s. Bo sa Is anbul Re . 22 (2), 341–351.
Lloyd, W.P., Shick, R.A., 1977. A es o s one’s wo-index model o e u ns. J. Financ.
Quan . Anal. 12 (3), 363–376.
Mensi, W., Hammoudeh, S., Yoon, S.M., 2014. S uc u al b eaks and long memo y in
modeling and o ecas ing ola ili y o o eign exchange ma ke s o oil expo e s: he
impo ance o scheduled and unscheduled news announcemen s. In . Re . Econ.
Finance 30, 101–119.
Mensi, W., Al-Yahyaee, K.H., Kang, S.H., 2019. S uc u al b eaks and double long
memo y o c yp ocu ency p ices: a compa a i e analysis om Bi coin and
E he eum. Finance Res. Le . 29, 222–230.
Mun, M., B ooks, R., 2012. The oles o news and ola ili y in s ock ma ke co ela ions
du ing he global inancial c isis. Eme g. Ma k. Re . 13, 1–7.
Musunu u, N., 2014. Modeling p ice ola ili y linkages be ween co n and whea : a
mul i a ia e GARCH es ima ion. In . Ad . Econ. Res. 20 (3), 269–280.
Peek, J., Roseng en, E.S., Too ell, G.M.B., 2003. Iden i ying he mac oeconomic e ec o
loan supply shocks. J. Money C edi Bank. 35 (6), 931–946.
Razzaq, A., Wang, Y., Chup adi , S., Suksa an, W., Shahzad, F., 2021. Asymme ic in e -
linkages be ween g een echnology inno a ion and consump ion-based ca bon
emissions in BRICS coun ies using quan ile-on-quan ile amewo k. Technol. Soc.
66, 101656.
Rigobon, R., 2003. On he measu emen o he in e na ional p opaga ion o shocks: is he
ansmission s able? J. In . Econ. 61 (2), 261–283.
Sanso, A., A ago, V., Ca ion, J.L., 2004. Tes ing o change in he uncondi ional a iance
o inancial ime se ies. Re . Econ. Financ. 4, 32–53.
Tan, Y., Flo os, C., 2018. Risk, compe i ion and e iciency in banking: e idence om
China. Global Finance J. 35, 223–236.
Tiwa i, A.K., Jena, S.K., T abelsi, N., Hammoudeh, S., 2022. Condi ional ansmission o
global shocks o eme ging s ock ma ke s: e idence om he quan ile connec edness
ne wo k analysis. Appl. Econ. 54 (31), 3621–3634.
Tong, H., Wei, S.-J., 2015. The mis o une o non- inancial i ms in a inancial c isis:
disen angling inance and demand shocks. In: Measu ing Weal h and Financial
In e media ion and Thei Links o he Real Economy (Ed. C.R. Hul en & M.B.
Reinsdo ). Publishe : Uni e si y, pp. 349–376.
Zei un, R., Tian, G., Keen, S., 2007. Mac oeconomic de e minan s o co po a e
pe o mance and ailu e: e idence om an eme ging ma ke he case o Jo dan.
Co p. Owne sh. Con ol 5 (1), 179–194.
Zi ko , D., Njegic, J., Milen, I., 2015. Bidi ec ional ola ili y spillo e e ec be ween he
exchange a e and s ocks in he p esence o s uc u al b eaks in selec ed eas e n
Eu opean economies. J. Econ. Finance 65 (6).
S. Çiçek and A. Yıldı ım
