Ma o-Ahenko ah, Da id; As a o , Richa d Ko i; Asa e, Nicholas
A icle
The linkage be ween banking c isis and so e eign deb
c isis: E idence om Ghana
Resea ch in Globaliza ion
P o ided in Coope a ion wi h:
Else ie
Sugges ed Ci a ion: Ma o-Ahenko ah, Da id; As a o , Richa d Ko i; Asa e, Nicholas (2025) : The
linkage be ween banking c isis and so e eign deb c isis: E idence om Ghana, Resea ch in
Globaliza ion, ISSN 2590-051X, Else ie , Ams e dam, Vol. 10, pp. 1-10,
h ps://doi.o g/10.1016/j. esglo.2025.100280
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/331202
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/
The linkage be ween banking c isis and so e eign deb c isis: E idence
om Ghana
Da id Ma o-Ahenko ah
a,*
, Richa d Ko i As a o
b
, Nicholas Asa e
c
a
Finance Business Pa ne , Uni e sal Me chan Bank L d, Acc a, Ghana
b
Economics Depa men , Ghana Communica ion Technology Uni e si y, Ghana
c
Depa men o Accoun ing, Uni e si y o Ghana, Business School, Ghana
ARTICLE INFO
Keywo ds:
Banking c isis
So e eign deb c isis
Mac oeconomic a iables
G ange Causali y
Doom loop phenomenon
Ghana
ABSTRACT
This s udy in es iga es he e ec o he ecen Ghanaian banking c isis on he coun y’s ongoing so e eign deb
c isis. An Au o eg essi e Dis ibu ed Lag Model (ARDL) is adop ed o es ima e he immedia e e ec s o Ghana’s
banking c isis on public deb le els using seconda y da a spanning he pe iod o 1997–2022. The esul s show
ha causali y lows om banking c isis o so e eign deb and no ice e sa. The banking c isis has a posi i e
signi ican e ec on Ghana’s so e eign deb le els. I also shows a signi ican nega i e ela ionship be ween Real
GDP g ow h and public deb le els while p i a e sec o c edi , and cu ency dep ecia ion ha e a signi ican
posi i e ela ionship wi h public deb le els in bo h he sho un and long un. This s udy con ibu es o he
doom loop li e a u e by o e ing empi ical and heo e ical e idence o he linkage be ween banking c isis and
so e eign deb c isis in a on ie economy. The s udy ecommends p omo ing expo di e si ica ion, s eng h-
ening inancial sec o egula ions, enhancing ins i u ional amewo ks, and implemen ing e ec i e deb man-
agemen s a egies o ensu e economic s abili y.
1. In oduc ion
Banking c isis can ha e se ious consequences o an economy’s
gene al heal h, especially as i pe ains o so e eign deb . Gene ally,
so e eign deb e e s o he accumula ed deb o a na ion ha includes
bo h ex e nal and domes ic obliga ions. Globally, s udies posi ha
widesp ead bank ailu es o liquidi y issues ha shly s ess he inancial
sys em o an economy (Lae en & Valencia, 2020; Royo, 2020; Taskin-
soy, 2023). This leads o go e nmen in e en ions h ough deposi
gua an ees o bailou s o s abilize he banking sec o esul ing in in-
c eases in public deb le els. The unsus ainabili y o he public deb s
usually leads o a so e eign deb c isis which in ends lead o a wo sened
banking c isis as go e nmen bo owing cos s ise and banks’ balance
shee s decline esul ing in inancial ins abili y (James, 2024; Mi chene
& T ebesch, 2021; Lewis & Dange ield, 2021; Nguyen e al., 2022;
Mo elli e al., 2022). Mos s udies on he linkage o banking c isis and
so e eign deb c isis ha e cen ed on de eloped and eme ging econo-
mies (B unne meie e al., 2016; Hu e al., 2021; Gilch is e al., 2022;
Bal eanu & E ce, 2018; A ellano e al., 2024). Like he de eloped and
eme ging economies, on ie economies like ha o Ghana ha e hei
own peculia economic and banking indus y cha ac e is ics
(Quisenbe y, 2018), hence, he need o examine he linkage be ween
banking c isis and so e eign deb c isis in such a se ing.
Aside he global pe spec i e, coun ies in A ica, such as Ghana, ha e
expe ienced no able majo banking c ises (A yee ey & Ackah, 2011;
Boako & Alagidede, 2018; Koh e al., 2020). Fo ins ance, p io o 2017,
Ghana aced a se e e banking c isis because o economic mismanage-
men and ex e nal deb bu dens in 1983, whils be ween 1999–2000 he
banking c isis led o he collapse and es uc u ing o se e al inancial
ins i u ions (Abo ebuno Akolgo, 2023; As a o , e al., 2023). The mos
no ewo hy o he banking c ises was he 2017–2019 banking c isis
which led o a majo inancial sec o clean-up exe cise. The inancial
sec o clean-up exe cise ini ia ed by he cen al bank ( he Bank o
Ghana, BoG), had he p ima y objec i e o sa egua ding he s abili y o
he inancial sec o whils suppo ing Ghana’s economic g ow h. As
indica ed by Dwamena and Yuso (2022), Ghana’s banking sec o c isis
has shaken he Ghanaian economy and plagued he inancial sec o in o
a s a e o unce ain y.
Banking and so e eign deb c ises ha e p o ound consequences on
indi iduals, businesses, and he na ion. Fo indi iduals, hese c ises
o en lead o job losses, income ins abili y, and a educ ion in household
weal h, exace ba ing po e y and inequali y. Businesses ace inc eased
* Co esponding au ho .
E-mail add esses: [email p o ec ed] (D. Ma o-Ahenko ah), [email p o ec ed] (R.K. As a o ), [email p o ec ed] (N. Asa e).
Con en s lis s a ailable a ScienceDi ec
Resea ch in Globaliza ion
jou nal homepage: www.sciencedi ec .com/jou nal/ esea ch-in-globaliza ion
h ps://doi.o g/10.1016/j. esglo.2025.100280
Recei ed 28 Janua y 2024; Recei ed in e ised o m 15 Ma ch 2025; Accep ed 16 Ma ch 2025
Resea ch in Globaliza ion 10 (2025) 100280
A ailable online 25 Ma ch 2025
2590-051X/© 2025 The Au ho (s). Published by Else ie L d. This is an open access a icle unde he CC BY license (
h p://c ea i ecommons.o g/licenses/by/4.0/ ).
bo owing cos s, lowe consume demand, and dis up ions o supply
chains, which can lead o bank up cies and a hal in in es men . On a
na ional scale, he banking and so e eign deb c ises ha e esul ed in
economic con ac ion, in la ion, and educed public se ices due o he
aus e i y measu es implemen ed by he go e nmen o Ghana (GoG).
Addi ionally, he so e eign deb c ises ha e e oded he na ional c edi
a ings, limi access o in e na ional capi al ma ke s, and unde mine
in es o con idence, u he deepening he economic u moil (Mu ize &
Nkhalamba, 2021). Fo ins ance, he in e na ional a ings agency, Fi ch
Ra ings, downg aded Ghana’s c edi wo hiness o u he junk s a us in
2024.
Consequen ly, he clean-up esul ed in 420 inancial ins i u ions’ li-
cences been wi hd awn. The 420 ins i u ions included 9 comme cial
banks, 23 sa ings and loan/ inance house companies, 347 mic o inance
companies, 39 mic oc edi companies, 1 emi ance company, and 1
leasing company. The o al asse s aken o e o he 420 dissol ed in-
s i u ions was GHS 26.05bn (7.45 % o GDP). The GoG paid GHS
18.99bn (5.49 % o GDP) o eimbu se impac ed deposi o s’ money,
which included he cons uc ion o a b idge bank, he Consolida ed Bank
Ghana Limi ed.
1
The inance minis e con i med in Ma ch 2025 ha
o e GHS 30.3bn (USD 2.06bn) had been spen by end o 2024 o
add ess he sec o ’s challenges. The inance minis e s a ed ha addi-
ional GHS 10.5bn will be needed o comple e he clean-up.
2
As o he
end o 2022, Ghana’s o e all public deb s ood a GHS 88.8 % o GDP, a
signi ican inc ease om 57 % o GDP in 2017 (IMF, 2023) wi h he
expendi u e om he inancial clean-up signi ican ly con ibu ing o he
coun y’s deb bu den.
Acco ding o As a o e al. (2023), he p eca ious and inc easing
deb bu den (bo h ex e nal and domes ic deb s) has led Ghana o be
classi ied as weak sus ainable deb coun y. The au ho s a gued ha i is
he e o e no su p ising ha he GoG launched he Domes ic Deb Ex-
change P og am (DDEP) on 5 Decembe 2022 o add ess he unsus-
ainable deb le els. Ne e heless, he IMF (2023) no ed ha DDEP
poses a signi ican challenge o he s abili y o he inancial sec o in
Ghana due o i s impac on he exposu e o inancial ins i u ions o
go e nmen deb . P io o he DDEP, banks, including s a e-owned
banks, had subs an ial holdings o go e nmen secu i ies, accoun ing
o 30 % o 50 % o hei o al asse s. Howe e , because o coupon e-
duc ions and ma u i y ex ensions unde he DDEP, he alue o hese
asse s is expec ed o dec ease o app oxima ely 70 % o hei ace alue.
3
This e alua ion signi ican ly shocked he balance shee s o hese
inancial ins i u ions. As a esul , he 2022 audi ed inancial s a emen s
published by 22 ou o 23 comme cial banks show ha impai men
cha ges o inancial asse s amoun ed o GHS 19.53bn (~USD 2.28bn),
4
ha ing a conside able e ec on he p o i abili y o he banks. This in-
dica es he se e e impac o he DDEP on he banking indus y. Based on
he published inancials, he indus y eco ded a loss be o e ax o GHS
7.33bn (~USD 854.71m).
5
Ou o he 22 banks, i e (5) did no mee he
p uden ial equi emen o a minimum capi al adequacy a io (CAR) o
13 %.
6
Likewise, he cen al bank epo ed a loss o GHS 60.8bn (~USD
7.09bn) and a nega i e equi y posi ion o GHS 55.1bn (~USD 6.42bn) in
2022.
7
The inance minis e s a ed in Ma ch 2025 ha he Bank o
Ghana is asking o a bailou o abou GHS 53bn o add ess hei nega-
i e equi y posi ion.
8
The ecen Ghanaian deb c isis as well as hose o o he A ican
coun ies has p o oked a signi ican numbe o publica ions. Empi ical
s udies, such as hose o As a o e al (2023) and Abo ebuno Akolgo
(2023) ha e la gely in es iga ed he e-eme gence o he indeb edness
ac oss he A ican con inen . Despi e his, a pa icula aspec o he
Ghanaian deb which has no been explo ed empi ically is he link be-
ween he coun y’s 2017–2019 banking sec o c isis and he cu en
so e eign deb dis ess. The empi ical e idence om Ghana, he e o e,
o e s a compelling case s udy o un a el hese in ica e dynamics be-
ween bank c isis and so e eign deb c isis. To close his gap in he
exis ing li e a u e, he objec i e o he s udy is o in es iga e he ela-
ionship be ween banking sec o c ises and so e eign deb c ises in
Ghana. The ques ions ha equi e he a en ion o u he esea ch and
empi ical in es iga ion a e:
Wha is he di ec ion o causali y be ween banking c isis and public
deb ?
Wha is he impac o he banking c isis on public deb ?
Wha is he ela ionship be ween selec ed mac oeconomic a iables
and he public deb le el?
The esea ch is signi ican because i con ibu es o he inancial
c isis and so e eign deb c isis li e a u e in di e se ways: Fi s , i is
among he e y ew empi ical a emp s o explain he win e ec o he
banking c isis and so e eign deb le el in Sub-Saha an A ica and Ghana
o be speci ic. In his s udy, we empi ically examine he ela ionship
be ween so e eign deb and mac oeconomic indica o s ha ha e been
epo ed in ex an li e a u e o in luence a coun y’s deb posi ion
(As a o e al., 2023; Eij inge and Ka a as¸, 2023), inco po a ing hese
mac oeconomic a iables as con ol ac o s in ou analysis. Also, his
s udy p o ides empi ical e idence on how dis up ions in he banking
sec o can p ecipi a e a so e eign deb c isis, he eby a ec ing he
o e all economic heal h o he na ion. Gi en Ghana’s his o y o banking
sec o ulne abili ies and ising public deb s, his esea ch is signi ican
in highligh ing he in e dependencies be ween hese wo inancial
phenomena. The indings will o e aluable insigh s o policymake s,
inancial egula o s, and s akeholde s o de elop s a egies ha can be
adop ed o mi iga e he isks associa ed wi h banking and so e eign deb
c ises. Addi ionally, his s udy will con ibu e o he exis ing body o
knowledge by p o iding a con ex ual analysis speci ic o Ghana, he eby
o e ing a ounda ion o u he esea ch and policy de elopmen in
simila on ie economies. The emainde o his pape is s uc u ed in
his manne : Sec ions 2 and 3 explain he li e a u e e iew and ma e ial
& me hods, espec i ely. Analysis o esul s a e p esen ed in Sec ion 4
and conclusions a e gi en in Sec ion 5.
1
h ps://mo ep.go .gh/ad e s/2022–09-27/ eques - o -exp essions-o -in-
e es -gh-mo - sd-315018-cs-ind .
2
Go e nmen o Ghana. (2025). The budge s a emen and economic policy o
he go e nmen o Ghana o he 2025 inancial yea (p. 20). Minis y o Finance.
3
The es ima ed educ ion in Ne P esen Value (NPV) is de i ed om using a
discoun a e o 16% −18%, which is he a e ha mos banks plan o adop
a e consul a ions wi h he Ins i u e o Cha e ed Accoun an s, Ghana (ICAG).
4
Fo mos o he banks, he impai men cha ges epo ed in he P&L we e no
seg ega ed in o impai men cha ges on loans & ad ances and ha on bonds
a ising om DDEP. Fo con ex , impai men cha ges on loans and ad ances in
2021 we e ~GHS 1.3bn, accoun ing o 6.8% o he agg ega e impai men s
epo ed in 2022. Wi h he ising NPLs ac oss he indus y, hypo he ically,
impai men cha ges on loans and ad ances in 2022 could accoun o 10% −
20% o he agg ega e impai men cha ges epo ed in 2022.
5
Based on he May 2023 MPC P ess Release, he indus y pos ed be o e- ax
losses o GHS 8bn in 2022. This includes all he 23 comme cial banks wi h one
unpublished inancial da a included.
6
In calcula ing he Capi al Adequacy Ra io (CAR), banks a e pe mi ed o
alloca e a po ion o he losses incu ed om he Deb Exchange P og am (DDE)
o e a maximum pe iod o ou yea s, wi h a qua e o he losses eco ded each
yea .
7
Bank o Ghana. (2022). F equen ly asked ques ions (FAQs) on Bank o
Ghana’s 2022 published annual epo and inancial s a emen s (p. 1).
8
h ps://www.myjoyonline.com/bog-seeks-gh%E2%82%B553bn-bailou - o-
add ess- inancial-challenges-d - o son/.
D. Ma o-Ahenko ah e al.
Resea ch in Globaliza ion 10 (2025) 100280
2
2. Li e a u e e iew
2.1. Theo e ical li e a u e e iew
The heo e ical explo a ion o he linkage be ween banking c ises
and so e eign deb c ises is spa se. Acco ding o he heo e ical li e a-
u e, he “doom loop phenomenon
9
” bes explains he ela ionship be-
ween he banking and so e eign deb c ises. The heo y con ends ha
inancial ins abili y in banks exace ba es go e nmen budge a y p ob-
lems, and ice e sa. This cycle happens when he go e nmen p o ides
conside able iscal suppo such as liquidi y suppo and deposi gua -
an ee o alle ia e a inancial c isis. The consequen inc ease in public
deb s may aise bo owing a es and diminish he p ice o go e nmen
bonds held on banks’ balance shee s, igge ing mo e banking p oblems
in a icious cycle. Fu he mo e, he cycle can en ail iscal con ac ion,
esul ing in lowe economic ac i i y, mo e nonpe o ming loans, wo s-
ening banking posi ion, and highe iscal cos s o bank ecapi aliza ion.
(James, 2024; Capponi e al., 2022; G´
omez-Puig & Sos illa Ri e o,
2024).
None heless, ewe heo e ical pape s ha e iden i ied mechanisms
h ough which so e eign dis ess can lead o inancial dis ess o ice-
e sa. Fo example, Acha ya e al. (2014) o e a model demons a ing
ha when a so e eign becomes o e bu dened, he alue o public gua-
an ees i has issued declines, exace ba ing inancial dis ess due o he
linkages be ween he go e nmen and he inancial sec o .
Acco ding o Eij inge and Ka a as¸ (2023), ano he impo an heo y
ega ding so e eign deb c isis is he “o iginal sin” hypo hesis which
posi s ha a cu ency de alua ion can igge a so e eign de aul when a
subs an ial po ion o he deb is in o eign cu ency (Eicheng een &
Hausmann, 2005; Jeanne, 2005). Acco ding o he “o iginal sin” heo y,
de eloping coun ies a e equen ly unable o issue deb in hei own
cu ency, equi ing hem o bo ow in o eign cu encies. This causes a
cu ency misma ch, esul ing in suscep ibili y du ing cu ency dep e-
cia ion, which can lead o so e eign deb c ises owing o he inc easing
weigh o o eign-denomina ed deb .
Also, an ini ial cu ency c isis is linked o a so e eign deb c isis
h ough ac o s such as o e alued exchange a es (Jahjah & Mon iel,
2003), a all in he c edi a ings (Reinha , 2002), o a ise in in e na-
ional in e es a es (D ehe e al., 2006). Equally, an ini ial so e eign
de aul can lead o a cu ency c isis i he cen al bank implemen s
expansiona y mone a y policies o a e a ecession ollowing he
wi hd awal o o eign capi al.
2.2. Ghana so e eign deb s si ua ions
Abo ebuno Akolgo (2023) a gues ha Ghana’s deb s mus be un-
de s ood in he con ex o i s poli ical-economic his o y since indepen-
dence, emphasizing s uc u al ac o s and ex e nal limi a ions. These
ac o s include he decline o de elopmen al e o s in he 1970s, he
implemen a ion o S uc u al Adjus men P og ammes in he 1980s, and
a global lending sys em whe e Wes e n c edi o s hold signi ican sway.
Recen iscal miss eps by Ghana’s go e nmen ha e exace ba ed hese
challenges.
As o 2022, he o e all public deb o Ghana s ood a GHS 88.8 % o
he coun y’s GDP, compa ed o 57 % o GDP in 2017 (IMF, 2023).
Inc easing in es o isk a e sion due o heigh ened deb dis ess, ac-
co ding o Fi ch (2022), caused Ghana’s sp eads o expand. Bloombe g
(2022) inds ha he e was a ligh om he Ghanaian cu ency and
bond ma ke s, making he cedi he poo es pe o me in he wo ld as i
declined by mo e han 55 % in he i s 10 mon hs o 2022. Th oughou
he yea 2022, in es o s consis en ly equi ed a highe e u n on
in es men o holding Ghana’s deb compa ed o he sa e U.S. T ea-
su ies. This yield sp ead, also known as he p emium, emained abo e
1,000 basis poin s o mos o he yea . By Decembe 2022, i had
exceeded 3,000 basis poin s. Mos o Ghana’s Eu obonds aded a a
discoun anging om 60 % o 65 % o hei ace alue in 2022, based on
Chicago Boa d o T ade Bond T ading p icing compiled by Bloombe g
(Fig. 1).
10
Following he build-up o he coun y’s deb , he go e nmen
announced he launch o DDEP on 5 Decembe 2022 p emised on a USD
3bn IMF bailou condi ion. Acco ding o he go e nmen , he b oade
objec i e o he domes ic deb es uc u ing is o add ess he issue o
public deb sus ainabili y. Impo an ly, he comple ion o he i s
anche o DDEP p o ided he go e nmen wi h subs an ial deb elie ,
amoun ing o nea ly GHS 50bn in 2023 (IMF, 2023). Mo eo e , he
go e nmen implemen ed addi ional deb exchanges o domes ic deb
denomina ed in US dolla s and Cocobills issued by he Ghana Cocoa
Boa d. Al hough hese ypes o deb make up a small po ion, app oxi-
ma ely 5 %, o he o al domes ic deb , hey ca ied a signi ican deb
se ice obliga ion o a ound GHS 15bn in 2023. Mo eo e , he Ghanaian
go e nmen announced on 19 Decembe 2022 ha i would empo a ily
hal all deb se ice paymen s o some o i s ex e nal deb s o allow o a
sys ema ic es uc u ing o hose obliga ions. The suspension applied o
se e al ypes o deb , such as Eu obonds (USD 13.2bn), comme cial e m
loans, and mos o i s bila e al deb (USD 30bn o al ex e nal deb s
a ec ed). Howe e , i does no include paymen s o mul ila e al deb ,
new deb s, o deb s ela ed o sho - e m ade acili ies.
11
Consequen ly, in Janua y 2023, he Eu ope, Middle Eas and A ica
(EMEA) De i a i es Commi ee, a commi ee a ilia ed wi h he In e -
na ional Swaps and De i a i es Associa ion, de e mined ha Ghana’s
decision o hal deb paymen s on i s Eu obonds, comme cial e m loans,
and bila e al deb quali ied as a “po en ial epudia ion/mo a o ium.”
Subsequen ly, in Ma ch 2023, he commi ee con i med he occu ence
o he c edi e en in Ghana, igge ing he paymen o c edi de aul
swaps as insu ance co e age o he coun y’s so e eign deb . This
uling necessi a ed he disbu semen o insu ance p o ec ion, wi h
c edi -de aul swaps co e ing a o al amoun o USD 66.4m g oss and
USD 34.4m ne as o 10 Feb ua y 2023, acco ding o da a om he U.S.
Deposi o y T us & Clea ing Co po a ion.
In Janua y 2024, he go e nmen ag eed in p inciple o es uc u e
USD 5.4bn in loans wi h o icial bila e al c edi o s. As o Ap il 2024, he
go e nmen has sugges ed o mos bondholde s o ex e nal deb s a one-
hi d educ ion in he ace alue o hei deb , oge he wi h es uc u ed
coupon paymen s o 5 % o e he ollowing h ee yea s and 7.5 %
he ea e , un il he new deb expi es in he 2030s. In June 2024, Ghana
and wo bondholde g oups ag eed in p inciple o es uc u e he
coun y’s USD 13bn deb wi h a po en ial 37 % p incipal hai cu . Unde
he ag eemen , he bondholde s will ha e o o go claims wo h
app oxima ely USD 4.7bn and p o ide liquidi y elie o abou USD
4.4bn wi hin he scope o he exis ing IMF eme gency aid plan ha will
expi e in 2026.
12
As indica ed by Abo ebuno Akolgo (2023), he ecen accoun s o
A ica’s deb dis ess, hough a he immedia e can be a ibu ed o he
COVID-19 pandemic, Russia-Uk aine wa , and iscal indiscipline, he
oo s lie in economic subo dina ion. While deb elie is c ucial,
comp ehensi e economic and inancial e o ms a e indispensable mo -
ing o wa d.
9
The e m "doom loop" was i s used by he Bank o England in 2009 o
desc ibe he po en ial o a icious ci cle o nega i e eedback be ween he
inancial sys em and he eal economy.
10
h ps://www.bloombe g.com/news/a icles/2023–03-03/ghana-s-missed-
in e es -paymen -spa ks-de aul -insu ance-payou #xj4y7 zkg.
11
Suspension o Paymen s on Selec ed Ex e nal Deb s o he Go e nmen o
Ghana (Decembe 2022), Minis y o Finance, Ghana.
12
h ps://www.myjoyonline.com/go e nmen -bondholde s- each-ag eemen
-in-p inciple- o- es uc u e-13bn-deb -sou ces-say/.
D. Ma o-Ahenko ah e al.
Resea ch in Globaliza ion 10 (2025) 100280
3
2.3. Empi ical li e a u e e iew
Gene ally, mos empi ical s udies ha e epo ed a posi i e ela ion-
ship be ween bank c isis and so e eign deb c isis. Fo ins ance, Capasso
e al. (2022), in es iga e he ans e o bank-so e eign isk be ween
eu ozone coun ies. By employing a Global Vec o Au o eg essi e
(GVAR) app oach o compu e he c oss-coun y “dis ance” in he deb -
o-GDP a io, he au ho s ind a so e eign-bank loop as a shock in one
coun y’s c edi de aul swap (CDS) sp eads o o he eu ozone coun ies
exace ba ing he so e eign isks. Despi e he insigh , howe e , he ocus
o he au ho s was on Eu ozone coun ies limi ing he gene alizabili y o
he indings o eme ging and on ie economies, whe e inancial sys-
ems a e o en mo e agile, and go e nmen in e en ions a e mo e
equen .
Mo ing a s ep u he by including cu ency c isis, Eij inge and
Ka a as¸ (2023) examine he linkages be ween so e eign deb , cu ency
c ises, and banking c ises om 1985 o 2020. Thei indings e eal ha
banking c ises ypically occu be o e so e eign de aul s, and no ice
e sa. Also, hey ind ha sho - e m ex e nal deb incu ed du ing
banking c ises, combined wi h misaligned exchange a es, unde lined by
cu ency c ises, ends o ele a e he p obabili y o u u e so e eign de-
aul s. While his conclusion aligns wi h he li e a u e, hei s udy does
no explo e he e e se causali y, ha is, whe he so e eign deb c ises
can lead o banking c ises, pa icula ly in de eloping economies like
Ghana.
In ano he s udy, Kalemli-¨
Ozcan, e al. (2016) in es iga e he
banking and so e eign deb c ises by ex ending he s udy o ocus on
ou (4) coun ies, pa icula ly in G eece, I eland, Po ugal, and Spain.
The pape by explo ing he impac o bank losses, iscal de ici s, and
an icipa ed deb hai cu s on bank sol ency inds ha ma ke -implied
de aul p obabili ies di e en ia e be ween coun ies based on i e
c i e ia. Howe e , his analysis p ima ily a ge s ad anced economies,
and he impac o simila ac o s in Sub-Saha an A ica emains unde -
explo ed. Hoque e al. (2015) analyze whe he egula ion educed isk
du ing he c edi c isis and he so e eign deb c isis o a c oss-sec ion o
global banks. By employing da a om he Wo ld Bank su ey, he e-
sea che s ind ha egula o y es ic ions, o icial supe iso y powe ,
capi al s ingency, along wi h p i a e moni o ing explain bank isk in
bo h c ises. Addi ionally, he esea che s ind mo e isk accele a ed he
so e eign deb c isis. E ce (2012) analyses 11 cases o so e eign de aul s
c oss di e se coun ies and inds ha liquidi y p essu es, banking sys em
obus ness, and co po a e sec o eliance on capi al ma ke s explained
obse ed disc imina ion pa e ns. Thus, E ce (2012) shows ha so e -
eign de aul s can be designed o minimize hei impac on domes ic
banks. Al hough his s udy emphasizes key s uc u al de e minan s o
so e eign de aul s, he ela ionship be ween hese ac o s and banking
c ises is no well-de ined, pa icula ly in he con ex o he banking
sec o s in de eloping economies.
In e ms o o he mac oeconomic a iables, As a o e al. (2023)
using da a om 1994 o 2018 in Ghana e eal a subs an ial c owding-in
e ec be ween domes ic deb and p i a e sec o c edi and in es men .
Mba e (2013) posi s ha domes ic deb has a nega i e elas ici y o 0.3
pe cen o GDP, and c owding ou was pe cei ed as an obs acle o he
accumula ion o capi al and he g ow h o he p i a e sec o and he
economy. Memon e al (2015) examine how i m-speci ic, mac oeco-
nomic ac o s and i m he e ogenei y a ec deb le els in Pakis an’s
non- inancial lis ed i ms. Using s a ic panel da a models, including
pooled OLS and ixed e ec s, hey ind ha p o i abili y, angibili y,
size, in e es a es, and in la ion signi ican ly impac deb le els.
Despi e he aluable insigh s om hese s udies, a gap exis s in he
empi ical li e a u e ega ding he bidi ec ional ela ionship be ween
banking c ises and so e eign deb c ises in on ie economies, such as
Ghana. While p e ious esea ch has p edominan ly ocused on ad anced
economies o speci ic c ises in isola ion, he e is limi ed unde s anding
o how hese c ises in e ac in a de eloping economy wi h a agile
inancial sys em. This s udy aims o ill his gap by empi ically es ing
he linkages be ween banking c ises and so e eign deb c ises in Ghana,
wi h a ocus on hei bidi ec ional causali y and ansmission
mechanisms.
3. Ma e ials and me hods
3.1. Sou ces o da a
The p ima y da a sou ces o his s udy a e he Bank o Ghana, he
In e na ional Mone a y Fund (IMF), Bloombe g and he Wo ld Bank.
Fig. 1. Ghana’s Eu obonds plumme compa ed o pee s. No e. F om “Explaine : Who holds Ghana’s deb and wha es uc u ing is planned?” by K. S ohecke , 2022.
(h ps://www. eu e s.com/wo ld/a ica/who-holds-ghanas-deb -wha - es uc u ing −is-planned- 2022–12-09/). Copy igh 2023 by Reu e s.
D. Ma o-Ahenko ah e al.
Resea ch in Globaliza ion 10 (2025) 100280
4
The s udy employs seconda y da a co e ing a pe iod o 26 yea s, om
1997 o 2022. Speci ically, da a on public deb - o-GDP (deb _gdp) was
ob ained om he IMF, while in o ma ion on banking c ises (bnk_c is)
was sou ced om he Bank o Ghana. The In la ion (in _a g) and he eal
GDP g ow h ( eal_gdp) da a we e e ie ed om he IMF. The banking
indus y’s loan- o-deposi a io (ld) da a was sou ced om he Bank o
Ghana, and eal sec o g ow h (ciea) was measu ed using he Bank o
Ghana’s Composi e Index o Economic Ac i i y (CIEA). Da a on cu ency
dep ecia ion a e (USD/GHS) (cu _dep) was sou ced om Bloombe g.
The p i a e sec o c edi - o-GDP (psc_gdp) and he claims on he cen al
go e nmen (% GDP) (ccg_gdp) da a we e sou ced om he Wo ld Bank.
3.1. De ini ion o a iables
To examine he ela ionship be ween banking c ises and so e eign
deb in Ghana, he dependen a iable used is he public deb - o-GDP
a io (deb _gdp), while he independen a iable is banking c isis
(bnk_c is). The independen a iable, banking c isis (bnk_c is), is
measu ed as a dummy a iable (1 i a c isis occu ed, 0 o he wise) and is
expec ed o ha e a posi i e sign wi h public deb , as banking c ises
ypically lead o inc eased deb . In la ion (in _a g) is measu ed as he
a e age in la ion a e o e he pe iod and is expec ed o ha e a posi i e
sign, as highe in la ion may d i e highe so e eign deb due o
inc eased go e nmen bo owing. Real GDP G ow h ( eal_gdp),
measu ed as annual pe cen age g ow h in eal GDP, is expec ed o ha e a
nega i e sign, as highe g ow h educes bo owing needs. Banking
Indus y’s Loan- o-Deposi Ra io (ld) is measu ed as he a io o loans o
deposi s and is expec ed o ha e a posi i e sign, indica ing ha highe
a ios may signal inc eased banking sec o isk, leading o highe so -
e eign deb . Real Sec o G ow h (ciea), measu ed by he Bank o Ghana’s
Composi e Index o Economic Ac i i y, is expec ed o ha e a nega i e
sign, as s onge g ow h educes go e nmen bo owing. Cu ency
Dep ecia ion Ra e (USD/GHS) (cu _dep), measu ed as he annual
dep ecia ion o he Ghanaian cedi agains he US dolla , is expec ed o
ha e a posi i e sign, as dep ecia ion inc eases he cos o o eign deb .
P i a e Sec o C edi - o-GDP (psc_gdp), measu ed as he a io o p i a e
sec o c edi o GDP, is expec ed o ha e a nega i e sign, as a well-
de eloped p i a e c edi ma ke educes go e nmen bo owing.
Finally, Claims on he Cen al Go e nmen (% GDP) (ccg_gdp),
measu ed as he a io o claims on he cen al go e nmen o GDP, is
expec ed o ha e a posi i e sign, as highe claims indica e inc eased
go e nmen bo owing, aising so e eign deb .
3.2. Model speci ica ion and es ima ion s a egies
3.2.1. G ange -Causali y es
To answe he i s esea ch ques ion o he di ec ion o causali y, we
es he G ange causali y be ween bank c ises and so e eign deb c ises
by speci ying a Vec o Au o eg ession (VAR) model. This enables us o
examine he dynamic ela ionship be ween hese wo a iables o e
ime. The i s model speci ied is he so e eign deb c isis as speci ied in
equa ion (1):
deb gdp =∝0+∑p
i=1∝ideb gdp −1+∑q
j=1βjbnk c is −1+
ε
(1)
The second au o eg essi e model speci ica ion o he banking c isis is
p esen ed in equa ion (2):
bnk c is =γ0+∑p
i=1γibnk c is −1+∑q
j=1δjdeb gdp −1+∊ (2)
Whe e deb gdp and bnk c is ep esen s he so e eign deb c isis indi-
ca o and he bank c isis indica o a ime , espec i ely; ∝0 and γ0 a e
he cons an e m; ∝i and γi a e coe icien s o he lagged alues o
so e eign deb and bank c isis, espec i ely; βj and δj a e he coe icien
o he o so e eign deb and bank c isis, espec i ely; and
ε
and ∊ a e
he e o e ms o so e eign deb and bank c isis, espec i ely. We es
o G ange causali y by e alua ing he signi icance o he lagged co-
e icien s. Speci ically, by pe o ming he F- es s o es he null hy-
po heses:
H0,bnk c is →deb gdp :β1=β2=⋯=βq=0
H0,deb gdp→bnk c is :β1=β2=⋯=βq=0
Rejec ing H0,bnk c is →deb gdp indica es ha bank c ises G ange -cause
so e eign deb c ises, while ejec ing H0,deb gdp→bnk c is indica es ha
so e eign deb c ises G ange -cause bank c ises.
3.2.2. Impac o banking c isis, selec ed mac oeconomics a iables on
so e eign deb
To examine he impac o banking c isis and selec ed mac oeconomic
a iables on so e eign deb c isis, we de eloped and es ima e wo main
models, he i s model is wi hou co a ia es (selec ed mac oeconomic
a iables) whe eas he second model is es ima ed wi h selec ed co a-
ia es. The es ima ed econome ic model(s) a e p esen ed in equa ion
(3) and (4) as ollows:
deb gdp =β0+β1bnk c is +
ε
(3)
deb gdp =β0+β1in a g +β2 eal gdp +β3psc gdp +β4ccg gdp
+β5Id +β6ciea +β7cu dep +β8bnk c is +
ε
(4)
The a iable de ini ion in Equa ion (1) has been p esen ed in Sec ion
3.1. Acco ding o he empi ical s udies o Bal eanu and E ce (2018), he
banking sec o ’s claims on go e nmen , c edi o he p i a e sec o ,
go e nmen deb , eal GDP g ow h, and po olio capi al in lows in lu-
ence so e eign deb s in eme ging ma ke s.
To examine his ela ionship, he s udy employs he au o eg essi e
dis ibu ed lag model (ARDL) sugges ed by Pesa an e al. (2001), o co-
in eg a ion in es iga ion ( ime se ies da a) and e o co ec ion (sho
un) analysis. The choice o he ARDL was because he ini ial uni oo
es showed di e en o de s, hus, some o he a iables we e in eg a ed
in o de 1 (I(1))whe eas o he a iables we e no . Thus, he ARDL model
is pa icula ly use ul when dealing wi h non-s a iona y ime se ies da a
a di e en le els (As a o & Fonu, 2021).
Fu he mo e, he choice o he ARDL was because i allows o he
inclusion o bo h lagged alues o he dependen a iable and lagged
alues o he independen a iables. Excep o he bank c ises a iable,
he s udy employed a log ans o ma ion o he a iables and p esen s
he equa ion as ollow (equa ion (5) and (6):
log(deb gdp ) =β0+β1(bnk c is ) +
ε
(5)
log(deb gdp ) = β0+β1log(in a g ) + β2log( ealgdp ) + β3log(pscgdp )
+β4log(ccggdp ) + β5log(Id ) + β6log(ciea )
+β7log(cu dep ) + β8bnk c is +
ε
(6)
4.1. P esen a ion o esul s
4.1.1. Uni oo es
The i s es pe o med was he uni oo es using he Augmen ed
Dickey Fulle (ADF), Phillip Pe on (PP) and he Kwia kowski-Phillips-
Schmid -Shin (KPSS) es . Upon conduc ing a ho ough analysis o he
a iable pa ame e s p esen ed in Equa ion (1), i becomes appa en ha
he e is a s ong indica ion o he p esence o a uni oo in he le el o
some o he a iables whe eas o he s we e s a iona y a le els. Fo
ins ance, he ADF shows ha Ciea and cu _dep was signi ican a le els
(in eg a ed a he le el I (0)) whils he PP es shows ha he banking
c isis was signi ican a le els. Howe e , a e employing he echnique
o i s di e ences (Table 1), i is obse ed ha mos o he a iables
exhibi s a iona i y.
D. Ma o-Ahenko ah e al.
Resea ch in Globaliza ion 10 (2025) 100280
5
Since he model’s a iables consis o a combina ion o i s di e -
ence and le el se ies, he Au o eg essi e Dis ibu ed Lag (ARDL)
me hodology is app op ia e o long- un analysis (Pesa an e al., 2001).
4.1.2. G ange causali y es
Table 2 shows ha wi h a lag leng h o 4, he Pai wise G ange
Causali y Tes s e eal ha he e is no signi ican e idence ha he
so e eign deb c isis (deb - o-GDP a io) does no g ange -cause banking
c ises (BNK_CRIS), as indica ed by he p obabili y alue (0.9385).
Con e sely, wi h a p obabili y alue o 0.0026, he second null hy-
po hesis indica es ha banking c ises (BNK_CRIS) do g ange -cause
changes in he so e eign deb (deb gdp ). The inding sugges s a sig-
ni ican G ange causali y om banking c ises o so e eign deb bu no
ice e sa. Speci ically, banking c ises a e signi ican p edic o s o
u u e changes in so e eign deb .
4.1.3. ARDL bound es
Based on he alues o he ARDL Bounds Tes p esen ed in he
Table 3, he F-s a is ic alues o 5.443916 and 3.986221 o bo h model
wi h co a ia es and wi hou co a ia es a e signi ican a 1 % and 5 %
espec i ely. This implies a ejec ion o he null hypo hesis and con i -
ma ion o he long un coin eg a ion and signi ies ha model equa ions
a e coin eg a ed a 1 % and 5 % espec i ely.
The uni oo es and ARDL bounds es ing p ocedu es con i m ha
he s udy a iables exhibi signi ican coin eg a ing ela ionships. As a
esul , he s udy model is sui able o analyzing bo h long- un e-
la ionships and sho - un dynamics h ough e o co ec ion.
Table 1
Uni Roo s and S a iona i y es s.
In Le el Fi s Di e ence
deb _gdp D i −0.6521 (2)
[0.52100]
−2.9697 (2)
[0.00757]**
ADF T end −0.8357 (3)
[0.41300]
−3.4893 (3)
[0.00245]**
None 0.6612 (2) [0.51500] −2.9650 (2)
[0.00739]**
PP Cons an −0.5783 (1)
[0.53100]
−4.2842 (1)
[0.60400]
T end −0.8541 (2)
[0.34900]
−4.3831 (2)
[0.38700]
KPSS Cons an 0.30060 (2)*** 0.15210 (8)
T end 0.2070 (2) 0.26820 (8)***
in _a g D i −2.2190 (2)
[0.03760]
−3.7059 (2)
[0.001400]**
ADF T end −1.8925 (3)
[0.07300]
−3.6981 (3)
[0.001530]**
None −0.5692 (2)
[0.57500]
−3.8714 (2)
[0.000883]***
PP Cons an −3.1337 (1)
[0.14000]
−6.1191 (1)
[0.09200]
T end −2.8995 (2)
[0.27970]
−6.1936 (2)
[0.08160]
KPSS Cons an 0.3424 (2)*** 0.2345 (2)***
T end 0.1415 (2)** 0.0750 (2)***
eal_gdp D i −2.4502 (2)
[0.02310]
−3.8890 (2)
[0.000912]***
ADF T end −2.3802 (3)
[0.02740]
−3.8453 (3)
[0.001090]**
None −0.9303 (2)
[0.36200]
−3.9856 (2)
[0.000673]***
PP Cons an −3.3617 (1)
[0.10621]
−6.9891 (1)
[0.00000]
T end −3.2844 (2)
[0.11302]
−6.8749 (2)
[0.00000]
KPSS Cons an 0.1241 (2)*** 0.0668 (2)***
T end 0.1239 (2)** 0.0521 (2)***
psc_gdp D i −2.4790 (2)
[0.02170]
−3.7578 (2)
[0.002140]**
ADF T end 2.3802 (3) [0.02740] −4.0300 (3)
[0.000715]***
None 0.1281 (2) [0.89900] −3.8879 (2)
[0.000849]***
PP Cons an −2.8645 (10)
[0.007188]**
−7.2209 (10)
[0.047930]
T end −2.3575 (10)
[0.020210]*
−8.1943 (10)
[0.01490]*
KPSS Cons an 0.4424 (2)** 0.3057 (2)***
T end 0.1973 (2) 0.0513 (2)***
ld D i −1.9332 (2)
[0.06680]
−2.8521 (2)
[0.009850]**
ADF T end −2.4613 (3)
[0.02310]
−3.0200 (3)
[0.007040]**
None −0.4182 (2)
[0.68000]
−2.9098 (2)
[0.008380]**
PP Cons an −1.9897 (2)
[0.10900]
−3.7747 (1)
[0.27460]
T end −2.2266 (1)
[0.03950]
−4.2058 (2)
[0.16990]
KPSS Cons an 0.2474 (2)*** 0.3422 (2)***
T end 0.2027 (2) 0.0461 (2)***
ciea D i −2.6690 (2)
[0.01440]*
−5.0829 (0)
[0.000057]***
ADF T end −2.4475 (3)
[0.02370]*
−5.2498 (1)
[0.000046]***
None −1.5910 (2)
[0.12600]
−5.2142 (0)
[0.000036]***
PP Cons an −3.8921 (2)
[0.42890]
−8.6452 (1)
[0.02056]
T end −3.7181 (2)
[0.72940]
−8.9714 (2)
[0.01840]
KPSS Cons an 0.1838 (2)*** 0.1625 (2)***
T end 0.1788 (2) 0.0560 (2)***
cu _dep D i −1.0577 (2) [0.7158] −3.7354 (0) [0.0101]
**
Table 1 (con inued)
In Le el Fi s Di e ence
ADF T end −2.4771 (3) [0.3353]
**
−3.6817 (1) [0.0436]
**
None −0.2184 (2) [0.5975] −2.0939 (0) [0.0372]
**
PP Cons an −1.0737 (1) [0.7097] −3.7207 (1) [0.0104]
**
T end −2.3921 (2) [0.3742] −3.6677 (2) [0.0448]
**
ccg_gdp D i −1.0302 (2)
[0.31500]
−3.4585 (2)
[0.002480]**
ADF T end 0.6916 (3) [0.49700] −3.9093 (3)
[0.000943]***
None 0.2985 (2) [0.76810] −3.5137 (0)
[0.002060]**
PP Cons an −2.0967 (1)
[0.00363]
−6.9579 (1)
[0.06577]
T end −1.8652 (2)
[0.00339]
−7.3708 (2)
[0.04100]*
KPSS Cons an 0.1701 (2)*** 0.2304 (2)***
T end 0.1700 (2) 0.0747 (2)***
bnk_c sis ADF D i n.a. n.a.
T end n.a. n.a.
None n.a. n.a.
PP Cons an −2.0805 (1)
[0.00914]**
−3.7540 (1) [1.0000]
T end −2.9200 (2)
[0.09520]
−3.6458 (2)
[0.79750]
KPSS Cons an 0.3426 (8)*** 0.1014 (2)***
T end 0.1150 (2)*** 0.0429 (2)***
No e. ***, ** o * deno e s a is ical signi icance a 1%, 5% o 10%; In () he
numbe o lags o bandwid h; In [ ] he p- alue.
Table 2
Pai wise g ange causali y es s.
Null hypo hesis Obs F-s a is ic P ob.
DEBT_GDP does no G ange Cause BNK_CRIS 22 0.19168 0.9385
BNK_CRIS does no G ange Cause DEBT_GDP 7.28600 0.0026
D. Ma o-Ahenko ah e al.
Resea ch in Globaliza ion 10 (2025) 100280
6
4.1.4. Es ima ed long- un e ec s
The long un analysis o he co-in eg a ion ela ionship be ween
so e eign deb and bank c isis shows ha wi hou o he co a ia es,
he e is a posi i e ela ionship be ween bank c isis and so e eign deb s
(0.329993) a e he hi d in eg a ion (Table 4). The Akaike in o ma ion
c i e ion (AIC) explains app oxima ely 89.4 % o he a iance in he
so e eign deb s whils he model explaining powe (adjus ed R-squa ed)
was 83.1 %.
The long- un co-in eg a ion ela ionship be ween so e eign deb and
he explana o y a iables wi h co a ia es using ARDL(1, 0, 1, 1, 1, 1, 0,
1, 0)) using he end and cons an bound es , shows ha he Akaike
in o ma ion c i e ion (AIC) explains app oxima ely 96.70 % o he
a iance o so e eign deb , wi h an adjus ed R-squa ed o 92.78 %. Mo e
impo an ly, he bank c isis (BNK_CRIS) has a signi ican posi i e e ec
on so e eign deb s by 36.90 %. Also, he lag o so e eign deb s (LOG
(DEBT_GDP)) was posi i e and signi ican indica ing ha abou 67.42 %
o he cu en yea ’s deb s we e explained by he p e ious yea ’s deb .
O he co a ia es, such as p i a e sec o c edi (LOG(PSC_GDP)), and he
lag o Bank o Ghana Composi e Index o Economic (CIEA(−1)) had a
posi i e signi ican impac on so e eign deb s. Table 4 show u he
shows a long- un nega i e ela ionship be ween eal g oss domes ic
p oduc (LOG(REAL_GDP)) and he lagged alue o eal g oss domes ic
p oduc (LOG(REAL_GDP)), cu ency dep ecia ion (LOG(CURR_DEP)),
and banking indus y loan- o-deposi a io (LD).
4.1.5. E o co ec ion model (ECM)
The e o co ec ion e m (ECT) signi ies he a e a which equilib-
ium is es o ed in he dynamic model. The ECM e m mus mee wo
c i e ia, ha is, i mus be s a is ically signi ican and ha e a nega i e
sign (As a o & Fonu, 2021).
Table 5 p esen s he sho - un coe icien es ima es om he e o
co ec ion ECM e sion o he ARDL model. In he model, he nega i e
signi ican alue o ECT o 0.459 sugges ha he speed o adjus men o
he long un o so e eign deb is app oxima ely 46 % o he model
wi hou co a ia es (bank c isis and so e eign deb ). The in oduc ion o
o he a iables o co a ia es dec eased he speed o adjus men o he
long un o 0.3258. Thus, sugges s ha he speed o co ec ing long- un
de ia ions is app oxima ely 32.58 pe cen in subsequen pe iods. Also,
he coe icien s in he sho - un model (wi h and wi hou co a ia es)
sugges ha BNK_CRIS signi ican ly in luences so e eign deb (LOG
(DEBT_GDP)) a 5 % signi ican le el. Also, a iables such as CURR_DEP,
CCG_GDP, and REAL_GDP we e signi ican in he model wi hou
co a ia es.
Table 3
ARDL Bounds Tes .
Tes S a is ic Value Signi . I(0) I(1)
Model wi hou co a ia es
F-s a is ic
3.986221 10 % 3.02 3.51
K 1 5 % 3.62 4.16
2.5 % 4.18 4.79
Model Wi h Co a ia es
F-s a is ic 5.443916 10 % 1.66 2.79
K 8 5 % 1.91 3.11
2.5 % 2.15 3.4
1 % 2.45 3.79
F-Bounds Tes Null Hypo hesis: No le els ela ionship
Table 4
Long- un Es ima es o banking c isis and so e eign deb c isis.
Es ima e wi hou Co a ia e Es ima e wi h Co a ia e
Va iable Coe icien S d. E o Coe icien S d. E o
LOG(DEBT_GDP(−1)) 0.771075*** 0.127925 0.674166*** 0.127355
LOG(DEBT_GDP(−2)) 0.108104
LOG(DEBT_GDP(−3)) 0.020695
LOG(DEBT_GDP(−4)) −0.358875*** 0.191426
BNK_CRIS 0.092145 0.140741 0.369082*** 0.118797
BNK_CRIS(−1) 0.106339
BNK_CRIS(−2) −0.026491
BNK_CRIS(−3) 0.329993**
LOG(CCG_GDP) 0.217694** 0.093995
LOG(CCG_GDP(−1)) −0.223350* 0.108293
CIEA 0.005507 0.004436
CIEA(−1) 0.013105*** 0.004054
LOG(CURR_DEP) 1.047314 0.553649
LOG(CURR_DEP(−1)) −1.518300** 0.523769
LOG(INF_AVG) 0.019446 0.108850
LOG(INF_AVG(−1)) −0.141484 0.103536
LOG(PSC_GDP) 1.112767*** 0.196687
LOG(REAL_GDP) −0.047684*** 0.012895
LOG(REAL_GDP(−1)) −0.057776*** 0.016592
LD −0.015134** 0.004510
C 1.657160*** 0.515479
R-squa ed 0.894242 0.966930
Adjus ed R-squa ed 0.830786 0.927847
S.E. o eg ession 0.171065 0.111705
Sum squa ed esid 0.438951 0.137257
Log likelihood 15.05458 29.58623
Mean dependen a 3.825761 3.825761
S.D. dependen a 0.415858 0.415858
Akaike in o c i e ion −0.404367 −1.246898
Schwa z c i e ion 0.083184 −0.564328
Hannan-Quinn c i e . −0.269141 −1.057582
Du bin-Wa son s a 1.852315 2.409044
***, ** o * deno e s a is ical signi icance a 1 %, 5 % o 10 %
D. Ma o-Ahenko ah e al.
Resea ch in Globaliza ion 10 (2025) 100280
7
4.2. Discussions o esul s
4.2.1. Causali y be ween banking c isis and Ghana’s public deb
The indings show ha Ghana’s banking c ises a e signi ican p e-
dic o s o so e eign deb le els which implies ha banking sec o
ins abili y leads o inc ease in Ghana’s so e eign deb . This causali y
highligh s a c i ical ulne abili y wi hin he Ghanaian economy, ha is,
banking c ises necessi a e go e nmen in e en ion, o en h ough
bo owing, which in u n exace ba es he so e eign deb bu den. This
inding con i ms ea lie indings by esea che s such as Eij inge and
Ka a as¸ (2023) and Reinha and Rogo (2009). Thus, he inding in
Ghana, sugges ha Ghana’s banking c ises usually he ald and agg a a e
so e eign deb c ises.
4.2.2. Impac o banking c isis and selec ed mac oeconomic a iables on
Ghana’s public deb
The mos impo an a iable o he s udy was he banking c isis and
i s e ec on Ghana’s so e eign deb le els. We epo ha Ghana’s
banking c isis inc eases Ghana’s so e eign deb . Du ing he inancial
c isis o Ghana, he go e nmen p o ided inancial suppo o bailou s o
banks, h ough liquidi y suppo and deposi gua an ees, o p e en a
o al collapse o he banking sec o . This in e en ion necessi a ed ha
he go e nmen o Ghana makes subs an ial public unds a ailable o
hese inancial ins i u ions leading o inc eased go e nmen spending.
Also, he banking c isis led o he loss o in es o con idence which
implies a highe cos o bo owing and highe in e es a es on go -
e nmen bonds. The go e nmen o Ghana, he e o e, aced inc eased
bo owing cos s as in es o s demanded highe p emiums o he
pe cei ed isk especially a e he a ing agencies downg aded he
Ghanaian economy o a junk s a us. The e was also se ious capi al ligh
as some in es o s wi hd ew hei capi al om he coun y and mo ed
hei businesses o neighbou ing coun ies. This equi ed u he
bo owing o s abilize he economy. Fu he mo e, Ghana’s banking
c isis igge ed an economic down u n, as businesses s uggled, unem-
ploymen le els inc eased, and consume spending d opped, all leading
o a educ ion in ax e enues. This inding is in line wi h hose epo ed
by Eij inge and Ka a as¸ (2023).
4.2.3. Mac oeconomic a iables and public deb le el
This pape shows ha g ow h o eal GDP g ow h esul s in a decline
o Ghana’s public deb . The inc ease in Ghana’s eal GDP implies an
inc ease in economic g ow h which ypically esul s in highe go e n-
men e enue h ough ax collec ion and educ ion in go e nmen ’s
bo owing. This inding is simila o hose epo ed by de Soy es e al.
(2022) highligh s a nega i e ela ionship be ween eal GDP and so e -
eign deb s.
Also, he p i a e sec o c edi inc eases Ghana’s so e eign deb s
since he go e nmen some imes gua an ees a po ion o p i a e sec o
loans. Hence, an inc ease in Ghana’s p i a e sec o c edi inc eases
Ghana’s go e nmen con ingen liabili ies. This inding con i ms hose
epo ed by As a o e al. (2023) which shows ha an inc ease in public
deb le els does no c owd ou p i a e-sec o c edi s. Fu he mo e, a ise
in he banking indus y loan- o-deposi a io (LD) helps educe so e -
eign deb by s imula ing p i a e-sec o c edi and boos ing economic
g ow h. Inc eased business lending enhances p oduc i i y, gene a ing
highe ax e enues and educing eliance on go e nmen bo owing.
P io i ising p i a e-sec o loans o e go e nmen secu i ies suppo job
c ea ion and in es men , s eng hening Ghana’s economic
undamen als.
Cu ency dep ecia ion in Ghana ease so e eign deb bu dens o
local-cu ency-denomina ed obliga ions bu inc eases he local cu ency
cos o ex e nal deb . When we ocus on cedis denomina ed deb s, hen
ou answe poin o he ac ha dep ecia ion o en leads o in la ion,
educing he eal alue o money and lowe ing he e ec i e cos o
epaying domes ic deb . When i comes o in e na ional ade, he cu -
ency dep ecia ion enhances expo compe i i eness, po en ially
inc easing o eign exchange ea nings and ax e enues, which can be
di ec ed owa d deb se icing. Howe e , dep ecia ion inc eases he
local cu ency cos o ex e nal deb , po en ially wo sening epaymen
challenges. E ec i e deb managemen and expo p omo ion a e c i -
ical o maximizing hese po en ial bene i s.
5.1. Diagnos ics and obus ness checks
To enhance he eliabili y o ou indings, we conduc ed addi ional
es s a e es ima ing he model. Fi s among hese is he Ramsey RESET
es , p esen ed in Table 6, which assesses he co ec ness o he model
speci ica ion. The e is no subs an ial empi ical suppo o posi he
p esence o omi ed a iables o unc ional o m misspeci ica ion in he
model, as pe he esul s o he Ramsey RESET es . Also, he e is
insu icien compelling e idence o indica e he p esence o he e o-
scedas ici y in he esiduals, as de e mined by he B eusch-Pagan
(Table 6).
The p esence o se ial co ela ion wi hin he esiduals o he
eg ession model is conside ed lacking using he B eusch-God ey Se ial
Co ela ion LM es .
The ecu si e CUSUM is p esen ed in Figs. 2 and 4depic he esul s
o he CUSUM o models wi hou co a ia es and wi h co a ia es,
espec i ely whils Figs. 3 and 5p esen he CUSUM Squa e es s o
Table 5
E o Co ec ion Model.
Model wi hou Co a ia e Model wi h co a ia e
Va iables Coe . S d. E o Va iables Coe . S d. E o
DLOG(DEBT_GDP(−1)) 0.230075 0.178528 DLOG(DEBT_GDP(−1)) 0.325834** 0.127355
DLOG(DEBT_GDP(−2)) 0.33818* 0.179312 D(BNK_CRIS) 1.132729** 0.479032
DLOG(DEBT_GDP(−3)) 0.358875* 0.175092 DLOG(CCG_GDP) 0.217694*** 0.051999
D(BNK_CRIS) 0.092145 0.115205 D(CIEA) 0.005507** 0.001917
D(BNK_CRIS(−1)) −0.3035** 0.143086 D(DEP_CURR_LOG) 1.047314*** 0.209038
D(BNK_CRIS(−2)) −0.32999** 0.140104 DLOG(INF_AVG) 0.019446 0.045742
Coin Eq(−1)* −0.459 0.123566 D(REAL_GDP) −0.047684** 0.007112
D(LD) −0.046447** 0.018909
Coin Eq(−1)* −0.325834*** 0.035419
***, ** o * deno e s a is ical signi icance a 1 %, 5 % o 10 %
Table 6
Diagnos ic es and Sensi i i y analysis.
Tes s a is ic Value Model
wi hou
Co a ia e
Model wi h
Co a ia e
Func ional o ms Ramsey
Rese
F-S a is ics
(P ob.)
1.1708
(0.30100)
0.4712
(0.79000)
He e oskedas ici y B eusch-
Pagan-
God ey Tes
F-S a is ics
(P ob.)
0.2729
(0.96400)
0.8631
(0.61020)
Se ial Co ela ion B eusch-
God ey
F-S a is ics
(P ob.)
0.6466
(0.54260)
0.5583
(0.39085)
D. Ma o-Ahenko ah e al.
Resea ch in Globaliza ion 10 (2025) 100280
8
