Jung, Yongseung
A icle
Limi ed Financial Ma ke Pa icipa ions and Shocks in
Business Cycles in Ko ea
Eas Asian Economic Re iew (EAER)
P o ided in Coope a ion wi h:
Ko ea Ins i u e o In e na ional Economic Policy (KIEP), Sejong-si
Sugges ed Ci a ion: Jung, Yongseung (2024) : Limi ed Financial Ma ke Pa icipa ions and Shocks in
Business Cycles in Ko ea, Eas Asian Economic Re iew (EAER), ISSN 2508-1667, Ko ea Ins i u e o
In e na ional Economic Policy (KIEP), Sejong-si, Vol. 28, Iss. 2, pp. 245-273,
h ps://doi.o g/10.11644/KIEP.EAER.2024.28.2.436
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/316632
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/
PISSN 2508-1640 EISSN 2508-1667 an open access jou nal
Eas Asian Economic Re iew ol. 28, no. 2 (June 2024) 245-273
h ps://dx.doi.o g/10.11644/KIEP.EAER.2024.28.2.436
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Limi ed Financial Ma ke Pa icipa ions and Shocks in
Business Cycles in Ko ea
Yongseung Jung†
Kyung Hee Uni e si y
[email p o ec ed]
This pape se s up a small open new Keynesian economy model wi h cons ained
households and incomple e ma ke s o add ess he d i ing o ces o business cycles in
Ko ea. I shows ha he e exis s a subs an ial ac ion o cons ained households who
canno ha e access o inancial ma ke . Fu he mo e, he es ima ed model e eals ha
a TANK model is be e han a RANK model in explaining business cycles in Ko ea.
The e ec o domes ic p oduc i i y shock on Ko ean economy has domina ed in he
a ia ions o ou pu , while he con ibu ion o he o eign p oduc i i y shock o he
a ia ions o ou pu and in la ion has inc eased a e he Asian inancial c isis. The
mone a y policy shock has domina ed he a ia ion o in la ion a sho and medium
ho izons.
Keywo ds: Business Cycles, H M, Ko ea, Maximum Likelihood Es ima ion, TANK
JEL Classi ica ion: E32
I. In oduc ion
A bu geoning o li e a u e on he he e ogeneous agen New Keynesian (HANK)
model has con ibu ed o unde s anding he ansmission o mone a y and iscal policy
o e he business cycle. Ma ke incomple eness and he e ogenei y ha e been u ilized
o add ess he in e ac ion be ween inequali y and iscal and mone a y policy. Kaplan
e al. (2018) show ha he gene al equilib ium e ec o an in e es a e cu , ope a ing
h ough an inc ease o household income associa ed wi h he labo demand expansion,
domina es he di ec e ec associa ed wi h he in e empo al subs i u ion. Aucle e
al. (2019) a gue ha he iscal mul iplie s depend on he in e ac ion o an in e empo al
ma ginal p opensi y o consume, i.e., iMPC and de ici - inanced iscal policy and
ID
† Depa men o Economics, Kyung Hee Uni e si y, Kyunheedae- o 26, Dongdaemunku, Seoul,
02447, Sou h Ko ea. I would like o hank o help ul commen s om he pa icipan s in nume ous
ins i u es and anonymous e e ees. All e o s a e mine.
246 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
only he HANK model can gene a e he empi ical iMPC. McKay e al. (2016) add ess
how he HANK model can sol e he o wa d guidance puzzle a ising om he
ep esen a i e agen new Keynesian (RANK he ea e ) model.
The quan i a i e HANK models ha explici ly ake in o accoun he e ogenei y and
he eedback e ec s om equilib ium dis ibu ions o agg ega es a e success ul in
deli e ing he gene al equilib ium e ec o exogenous shocks compa able o he one
in he da a. Howe e , i is di icul o ack he weal h dis ibu ion, as i is necessa y o
use non i ial compu a ional echniques o sol e he equilib ium in he HANK models.
The ea lie li e a u e on wo-agen models o add ess he business cycle has
emphasized he he e ogenei y in shaping he ansmission o mone a y and iscal
policy. The wo-agen new Keynesian (TANK) model wi h a minimal he e ogenei y
and analy ical ac abili y has been u ilized o unde s and and quan i y he implica ions
o he e ogenei y in households. In he canonical TANK model, he e is a cons an
ac ion o cons ained o hand- o-mou h (H M he ea e ) households who canno ha e
access o he inancial ma ke and ha e o consume hei cu en income. O he ac ion
o households, called uncons ained o Rica dian households who can ha e access o
inancial ma ke sa is y he Eule equa ion. Though he simple TANK model does no
allow any idiosync a ic shock and endogenous ac ion o cons ained households,
Debo oli and Galí (2019) show ha he ac able TANK model can app oxima e he
dynamics o he HANK model unde compa able edis ibu ion schemes.1
The open economy TANK model is isomo phic o he open economy ep esen a i e
agen new Keynesian (RANK) model. Howe e , he e a e s a k di e ences be ween a
TANK model and a RANK model in open economy. Fi s , he agg ega e demand
equa ion, i.e., he uncons ained household’s Eule equa ion in he TANK model is
di e en om he one in he RANK model in ha he o me depends on he agg ega e
demand as well as he ac ion o H M households in he economy. Second, only
consump ion o uncons ained households ma e s o he isk-sha ing in an open
economy as H M households canno pa icipa e in inancial ma ke s. Finally, he
NKPC and goods ma ke clea ing condi ion depends on consump ion and income
inequali y be ween uncons ained households and H M households.
The ques ion abou he impo an ole o inancial ic ions in shaping he business
cycle in Ko ea has been ac i ely deba ed in Ko ean academia and policy make s since
1 Bilbiie (2019) ex ends a simple TANK model o a ac able wo agen he e ogeneous agen model
wi h idiosync a ic shock o add ess mone a y and iscal policy e ec on consump ion and ou pu .
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 247
ⓒ 2024 Eas Asian Economic Re iew
he Ko ean go e nmen has adop ed an expo d i en economic g ow h s a egy om
he 1960s. Some c i ics ha e been skep ical abou he sus ainabili y o he economic
g ow h s a egy in Ko ea whe e a subs an ial ac ion o economically neglec ed
households exis s. They ha e c i icized he s uc u e o he Ko ean economy, since he
economy which hea ily depends on he es o he wo ld is oo agile o sus ain i s
s able economic g ow h.
In his pape , we add ess he ollowing ques ions wi h a TANK model wi h
incomple e ma ke s. How impo an ha e he economically neglec ed households, i.e.,
he H M households been in shaping he business cycle in Ko ea o e ime?
Speci ically, does he TANK model pe o m be e han he RANK model in
explaining business cycles in Ko ea? Wha kind o shock has been he main d i ing
o ce in business cycles in Ko ea? Wha ac ion o H M households ha e been in
Ko ea? Has he ac ion o H M households who canno ha e access o inancial
ma ke s inc eased in Ko ea o e ime du ing 1997 Asian inancial c isis and he
ongoing G ea Mode a ion pe iods?
Fo his pu pose, we se up a small open economy TANK model wi h a simple
he e ogenei y along he lines o Bilbiie (2008) and Debo oli and Galí (2019).
Speci ically, we se up a small open economy TANK model wi h domes ic and o eign
p oduc i i y shocks, p e e ence (o demand), and mone a y shocks. We es ima e he
key pa ame e s o he model wi h qua e ly da a spanning om 1970 o 2018 by
employing maximum likelihood. In pa icula , we es ima e he sha e o H M
households in Ko ea by di iding he sample pe iods in o h ee subsample pe iods o
look a how he sha e o H M households has a ied be o e and a e 1997 Asian
inancial c isis and he ongoing G ea Recession pe iods. Then, we examine,
quan i a i ely and wi h he help o o mal econome ic me hods, he impo ance o
H M households wi hin he speci ied amewo k. Finally, we e alua e he ela i e
impo ance o each shock and he ele ance o inancial ic ions o e he business
cycle.
Th ee impo an indings come om his pape .
Fi s , he ac ion o H M households in Ko ea inc eased o e ime in Ko ea.
Fu he mo e, he es ima ed model e eals ha he TANK model pe o ms be e han
he RANK model in explaining business cycles in Ko ea. The likelihood a io s a is ic
shows ha he model wi hou H M households is ejec ed by da a. The es ima ed sha e
o H M households is abou 0.2 in he i s subsample pe iod, 1976:3Q - 1996:3Q
be o e he Asian inancial c isis. Howe e , i inc eased o abou 0.4 du ing he second
248 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
subsample pe iods. The low es ima e o he H M households be o e he Asian inancial
c isis seems o echo a high sa ing a e du ing a high economic g ow h e a. Households
who ha e been e y op imis ic abou he u u e o he economy we e willing o sa e
hei income o hei child en in e ms o o ced sa ings. When he high economic
g ow h e a has come o an end wi h he Asian inancial c isis, a li e ime wo kplace has
also disappea ed, inc easing he sha e o H M households. The es ima ed sha e o H M
households oughly ma ches he es ima ed ac ion o he H M households in Jung and
Kim (2019) who ound using KLIPS (Ko ea Labo Ins i u e Panel Su ey) om 2001
o 2018.
Second, he domes ic p oduc i i y shock has domina ed in explaining he a ia ions
o ou pu a all ho izons, while he o eign p oduc i i y shock and he p e e ence ha e
played an impo an ole in he a ia ion o in la ion a he sho and medium ho izons
a e he Asian inancial c isis as he Ko ean economy has libe alized he capi al
mo emen s. The mone a y policy shock has played a mino ole in he a ia ion o
ou pu in he whole sample pe iod. The o eign p oduc i i y shock has been he mos
impo an ac o in he a ia ions o he in e na ional ela i e p ice in he whole sample
pe iod.
Finally, he mone a y policy shock has been he mos impo an ac o in he
a ia ion o in la ion a sho and medium ho izons, while he o eign supply shock has
hea ily con ibu ed o he luc ua ion o in la ion a medium and long ho izons a e
he G ea Recession. Du ing he G ea Recession pe iods, he mone a y shock has
domina ed in explaining he a ia ions o in la ion as he mone a y au ho i y ies o
s imula e he economy by manipula ing i s policy a e.
The ou line o he pape is ollows. In sec ion 2, we speci y a simple TANK model.
In sec ion 3, we discuss an equilib ium and he implica ions o he model ela ed o
eal ac i i ies and p ices. In sec ion 4, we p esen he quan i a i e implica ions o he
model. Finally, concluding ema ks a e gi en in sec ion 5.
II. Model
This sec ion se s up a canonical TANK model wi h incomple e ma ke s. In he home
coun y, a sha e o 1-λ o households, i.e. uncons ained households ha e access o
inancial ma ke s, while he emaining sha e λ o he households, i.e. cons ained o
H M households do no ade any asse and simply consume hei cu en labo income.
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 249
ⓒ 2024 Eas Asian Economic Re iew
1. Households
(1) Uncons ained households
Uncons ained households can ha e access o in e na ional inancial ma ke s. They
seek o maximize
𝒲,=𝐸∑𝛽exp (𝑣)𝑈𝐶,−,
,0<𝛽<1, (1)
whe e 𝑈𝐶,=,
o σ≠1, and 𝑈𝐶,=𝑙𝑛𝐶, o σ=1.
𝐸 deno es he expec a ion ope a o o e all possible s a es o na u e on his o y 𝑠
and 𝐶, and 𝑁, ep esen he uncons ained household’s consump ion and labo
hou s in pe iod , espec i ely. He e 𝑣 is a p e e ence shock which ollows an AR(1)
p ocess as 𝑣=(1−𝜌)𝑣+𝜉,, 0<𝜌<1, whe e ξ, is an i.i.d., no mally
dis ibu ed p ocess wi h mean 0 and a iance 𝜎. 𝐶, is a composi e consump ion
index de ined by
𝐶,=𝜃𝐶,
+(1−𝜃)𝐶,
, 𝜂>0 (2)
He e 𝐶, and 𝐶, a e indices o domes ic uncons ained households’ home
and o eign consump ion goods. No e ha η measu es he subs i u abili y be ween
domes ic and o eign goods, and θ∈[0,1] measu es he deg ee o openness in goods
ma ke , i.e. he sha e o domes ic consump ion alloca ed o domes ic goods. 𝐶,
and 𝐶, ake he ollowing CES agg ega o :
𝐶,=[
𝐶(𝑖),
𝑑𝑖]
,𝐶,=[
𝐶(𝑖),
𝑑𝑖]
, ϵ>1 (3)
whe e 𝜖 deno es he elas ici y o subs i u ion among goods wi hin each ca ego y.
Domes ic uncons ained households a e subjec o a sequence o budge cons ain s.
We assume incomple e asse ma ke s whe e only one-pe iod nominal iskless bonds
250 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
denomina ed in home and o eign cu ency a e aded in he in e na ional inancial
ma ke s. Domes ic uncons ained household’s budge cons ain is gi en by
𝑃𝐶,+𝑅𝐵,+𝑅∗𝐵,
∗𝐹(𝜀𝐵,
∗𝑃)+Θ𝑉≤𝐵,+𝐵,
∗+
𝑊𝑁,+Θ𝑉+𝑃𝐷,+𝑃𝑇𝑅,, (4)
whe e 𝑃 and ℰ a e he home consume p ice index (CPI) and he nominal
exchange a e in pe iod . He e Θ, 𝑉, and 𝐷, deno e domes ic sha e holdings,
a e age ma ke alue o he co esponding sha es, and he eal di idends a ime ,
espec i ely. 𝐵, and 𝐵,
∗ a e one-pe iod domes ic and o eign cu ency
denomina ed iskless nominal bonds wi h he co esponding in e es a es 𝑅 and
𝑅∗, espec i ely. Since he nons a iona i y o he incomple e ma ke s wi h iskless
bonds complica es he ask o app oxima ing equilib ium dynamics, we assume ha
he in e na ional ade o o eign cu ency denomina ed bonds is subjec o
in e media ion cos s as in Benigno (2009) and Schmi -G ohé and U ibe (2003).
Speci ically, he in e es a e 𝑅∗𝐵,
∗𝐹(ℰ𝐵,
∗𝑃) aced by domes ic uncons ained
households is inc easing in domes ic coun y’s a e age o eign deb ℰ𝐵,
∗𝑃. Tha
is, F′(.) > 0, and 𝐹𝑩,
∗=1 in he s eady s a e whe e 𝑩,
∗≡ℰ𝐵,
∗𝑃.
In simila , he budge cons ain o he ep esen a i e o eign households can be
w i en as
𝑃∗𝐶∗+𝐵,
∗𝑅∗≤≤𝐵,
∗+𝑊∗𝑁∗+Γ, (5)
whe e Γ is he in e media ion p o i s om loans o he small coun y. Fi s o de
condi ions o he household implies ha he equilib ium eal exchange a e 𝑄 is
de e mined by
𝐸[
∗
∗ ∗
∗]=𝐹(𝐵,
∗)𝐸[𝑒𝑥𝑝(𝑣−𝑣),
,ℰ
ℰ]. (6)
The a io o he consume p ice index 𝑃 ela i e o he domes ic p ice index 𝑃,
is ela ed o he e ms o he e ms o ade 𝑆=𝑃,𝑃,
as ollows
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 251
ⓒ 2024 Eas Asian Economic Re iew
,
,=[(1−𝜃)+𝜃𝑆]
≡K(𝑆). (7)
Finally, no ice ha in e empo al condi ion o domes ic bond holdings
β𝐸[𝑒𝑥𝑝(𝑣−𝑣),
,
()]=1 (8)
also holds, whe e 𝜋≡
−1 is he CPI in la ion a e a ime .
The isk-sha ing condi ion in incomple e ma ke can be log-linea ized a ound he
s eady-s a e as ollows
𝐸[𝑞]−𝑞=𝜎𝐸[𝑐,−𝑦
∗]−σ(𝑐,−𝑦∗)+𝜁𝑏,+(1−𝜌)𝑣, (9)
whe e 𝑏,≡𝑩,
and ζ is a isk p emium o bo owing p emium in he in e na ional
ma ke . The small le e 𝑥 deno es he log-linea iza ion o he co esponding
a iable 𝑋 a ound i s s eady s a e X, 𝑥=𝑙𝑛(𝑋/𝑋).
(2) Cons ained households
The H M o cons ained households who do no ha e any asse s wo k o 𝑁,
hou s and consume hei income de e mined in each pe iod:
𝑃𝐶,=𝑊𝑁,, (10)
whe e 𝐶, is H M household's consump ion in pe iod .
H M households seek o maximize hei empo al u ili y unc ion (𝑈,) subjec o a
budge cons ain (10):
𝑈,≡𝑒𝑥𝑝(𝑣)[,
−,
]. (11)
H M household’s op imiza ion condi ions a e gi en by
𝑁,
𝐶,
=𝑤 (12)
252 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
and he budge cons ain (10). He e 𝑤≡
is he eal wage a ime .
2. Domes ic Fi ms
The domes ic i ms’ p oblem is s anda d. Each good is p oduced by a
monopolis ically compe i i e i m indexed by i∈[0,1] using a linea echnology
𝑌(𝑖)=𝑒𝑥𝑝(𝑧)𝑁(𝑖). He e 𝑧 is an AR(1) p ocess echnology shock in home
coun y a pe iod , i.e. 𝑧=(1−𝜌)𝑧+𝜉,, 0<𝜌<1, whe e ξ, is an i.i.d.,
no mally dis ibu ed p ocess wi h mean 0 and a iance 𝜎. 𝑌(𝑖) and 𝑁(𝑖)
ep esen he co esponding i m's ou pu and o al labo inpu , espec i ely.
Since he labo ma ke is pe ec ly compe i i e, he cos minimiza ion implies ha
(1 −τ)𝑤=𝑚𝑐𝑒𝑥𝑝(𝑧)𝐾(𝑆), (13)
whe e τ is an employmen subsidy o a ain he e icien and equi able s eady s a e and
𝑚𝑐(≡
,) is domes ic i m's eal ma ginal cos a ime . No e ha he labo hou s
o each household can be exp essed in e ms o he e ms o ade as
𝐶,
𝑁,
=(1−𝜏)𝑚𝑐𝑒𝑥𝑝(𝑧)𝐾(𝑆), (14)
whe e i=U, K.
Nex , we in oduce Cal o- ype s icky p ices along he lines o Yun (1996). Each
domes ic i m i in equen ly adjus i s op imal p ice 𝑃,(i) wi h p obabili y (1-α) in
any gi en pe iod, aking 𝑃, and he agg ega e demand as gi en. Since 𝑃,(𝑖) is
he same o he eop imizing i ms, i.e., 𝑃,(i) = 𝑃,, he op imal p ice se ing
equa ion can be w i en as
𝐸∑𝛼Ξ,,
,𝑌[𝑚𝑐−𝑀 ,
,]
=0, (15)
whe e Ξ,≡𝛽(𝐶.
P)(𝐶.
𝑃)
⁄ and M=
is he a e age ma kup in
he home goods ma ke .
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 259
ⓒ 2024 Eas Asian Economic Re iew
𝜎,𝜎∗, and 𝜎 in he TANK and RANK models show ha he con ibu ion a o eign
p oduc i i y shock o business cycles in Ko ea declines in TANK model han in he
RANK model be o e he Asian inancial c isis.
Table 1. Maximum Likelihood Es ima es and S anda d E o s (1976:3Q-1997:2Q)
Pa ame e
TAN
K
Model RAN
K
Model
Es ima e S anda d E o Es ima e S anda d E o
Α 0.7026 0.0416 0.5506 0.0098
Λ 0.2659 0.0085 0 -
𝜌 0.9113 0.0786 0.9957 0.0185
ρ
∗ 0.9321 0.0456 0.9996 0.0244
𝜌 0.8859 0.0474 0.9816 0.0917
𝜌 0.6462 0.0467 0.6630 0.0072
𝑎 1.4768 0.0539 1.5335 0.0471
𝑎
0.0948 0.0088 0.0000 0. 0095
𝜎 0.0383 0.0030 0.0312 0.0004
𝜎
∗ 0.0544 0.0043 0.0727 0.0011
𝜎 0.0063 0.0006 0.0062 0.0005
𝜎 0.0250 0.0013 0.1314 0.0615
Η 1.2488 0.0304 1.5172 0.0122
𝐿 -1017.82 𝐿 -1011.99
No e: 𝐿 and 𝐿 deno e he maximized alue o he TANK and RANK models’ log-likelihood unc ion,
espec i ely.
Nex , Table 2 displays he decomposi ion o o ecas e o a iances in de ended
ou pu , in la ion, he nominal in e es a e, and he eal exchange a e in o componen s
o each o he model's ou o hogonal dis u bances. The able shows ha he domes ic
p oduc i i y shock has domina ed in he a ia ions o ou pu a all ho izons by
accoun ing o mo e han 85 pe cen o uncondi ional a iance o ou pu , while he
o eign p oduc i i y and mone a y policy shocks ha e played a mode a e ole in ou pu
a ia ions a sho ho izon. The mone a y policy shock has been by a he dominan
ac o in he luc ua ions o in la ion a sho and medium ho izons by accoun ing o
mo e han 40 pe cen o he uncondi ional a iance o he in la ion a e a he
co esponding ho izons. The domes ic p oduc i i y and demand shocks ha e played
an impo an ole in he a ia ions o in la ion a all ho izons by accoun ing o 20
pe cen o he uncondi ional a iance o in la ion a e a he co esponding ho izons.
260 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Table 2. Fo ecas E o Va iance Decomposi ions (1976:3Q-1997:2Q)
Qua e s
Ahea
d
Domes ic P od.
Shoc
k
Fo eign P od
Shoc
k
Policy
Shoc
k
P e e ence
Shoc
k
Ou pu
1 54.2 27.9 14.6 3.3
4 85.1 9.2 4.7 1.0
8 88.2 8.4 2.8 0.6
12 88.2 8.2 2.2 0.5
20 89.4 7.9 2.2 0.5
40 89.3 8.0 2.2 0.5
In la ion
1 2.2 6.2 68.0 23.6
4 18.9 11.4 47.8 21.9
8 24.3 11.9 42.5 21.3
12 26.0 11.7 41.3 21.0
20 26.3 12.1 40.8 20.8
40 25.1 17.8 37.5 19.6
In e es Ra e
1 0.1 6.9 18.4 74.6
4 10.5 11.7 5.8 72.0
8 12.8 14.6 3.9 68.7
12 12.9 14.4 3.7 69.0
20 13.8 16.3 3.6 67.3
40 14.6 35.7 2.4 47.3
Exchange a e
1 21.9 74.7 3.4 0.0
22.0 76.9 0.8 0.3
8 21.2 78.0 0.5 0.3
12 20.4 78.9 0.4 0.3
20 19.6 79.7 0.4 0.3
40 19.3 80.0 0.4 0.3
The demand shock has domina ed in he beha io o he policy a e du ing a high
economic g ow h e a in Ko ea. The o eign p oduc i i y shock has domina ed in he
beha io o an in e na ional ela i e p ice by accoun ing o mo e han 75 pe cen o
he uncondi ional a iance o he in e na ional ela i e p ice a all ho izons, while he
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 261
ⓒ 2024 Eas Asian Economic Re iew
con ibu ion o he mone a y policy shock o he a ia ion o he eal exchange a e is
nil.
Table 3. Maximum Likelihood Es ima es and S anda d E o s (1998:1Q-2007:2Q)
Pa ame e
TAN
K
Model RAN
K
Model
Es ima e S anda d E o Es ima e S anda d E o
Α 0.3820 0.1551 0.1644 0.0012
Λ 0.4424 0.0061 0 -
𝜌 0.7825 0.0286 0.8277 0.0113
ρ
∗ 0.7341 0.0424 0.8955 0.1022
𝜌 0.8471 0.1073 0.9869 0.0319
𝜌 0.0000 0.1669 0.1854 0.0049
𝑎 2.0827 0.1048 2.6509 0.0087
𝑎
0.0004 0.0409 0.0000 0.0122
𝜎 0.0284 0.0071 0.0298 0.0061
𝜎
∗ 0.1076 0.0137 0.0452 0.0094
𝜎 0.0077 0.0014 0.0090 0.0008
𝜎 0.0140 0.0017 0.1222 0.0204
Η 0.9470 0.0643 0.04880 0.0202
𝐿 -458.7445 𝐿 -434.7973
No e: 𝐿 and 𝐿 deno e he maximized alue o he TANK and RANK models’ log-likelihood unc ion,
espec i ely.
Table 3 p esen s maximum likelihood es ima es o he deep pa ame e s in he
second subsample pe iods, i.e., a e he Asian inancial c isis, bu be o e he G ea
Recession, 1998:1Q-2007:2Q. To es o e he heal h o Ko ean economy hi by he
Asian inancial c isis, Ko ea go e nmen has implemen ed some es uc u ing polices
o allow mo e lexibili y in labo ma ke as well as domes ic and in e na ional inancial
ma ke s. The ma ke o ien ed economic policies ha e inc eased he sha e o non-
egula o pa - ime wo ke s and made he housing ma ke as well as he c edi ma ke
uns able. To deal wi h he inc ease in housing p ices and c edi booms, he go e nmen
in e ened in he housing ma ke wi h mac op uden ial ools such as LTV and DTI o
he i s ime o cool down he ma ke . The go e nmen ’s e o o cool down he
housing ma ke and he s uc u al change in he labo ma ke ha e subs an ially
inc eased he ac ion o households in he second subsample pe iod. The la ge
262 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
es ima e o λ echoes he p e alence o he nega i e e ec o he unp eceden ed Asian
inancial shock in e wined wi h go e nmen ’s p uden ial policy on households.
The la ge es ima e o 𝑎 in he in e es a e ule shows ha he mone a y au ho i y
implemen ed an in la ion a ge ing ule o s abilize p ices in Ko ea a e he Asian c isis.
Since he Ko ean go e nmen was mo e willing o libe alize capi al lows by adop ing
a lexible exchange a e egime, he Ko ean economy has been mo e closely connec ed
wi h he es o he wo ld han be o e. Table 3 indica es ha 𝐿=458.75 and
𝐿=434.80, implying ha LR=47.9. Since he 0.1 pe cen c i ical alue o LR is 10.8,
he null hypo hesis is ejec ed.
Table 4. Fo ecas E o Va iance Decomposi ions (1998:1Q-2007:2Q)
Qua e s
Ahead
Domes ic P od.
Shock
Fo eign P od
Shock
Policy
Shock
P e e ence
Shock
Ou pu
1 89.9 9.7 0.2 0.2
4 81.8 17.8 0.1 0.3
8 82.0 17.5 0.1 0.4
12 82.1 17.4 0.1 0.4
20 82.1 17.4 0.1 0.4
40 82.1 17.4 0.1 0.4
In la ion
1 0.8 23.2 59.4 16.6
4 1.4 35.4 37.1 26.1
8 1.5 36.3 33.4 28.8
12 1.6 36.5 32.6 29.3
20 1.6 36.8 32.4 29.2
40 1.6 37.3 32.0 29.1
In e es Ra e
1 2.0 57.1 0.0 40.9
4 2.3 56.3 0.0 41.4
8 2.3 54.5 0.0 43.2
12 2.4 54.2 0.0 43.4
20 2.4 54.5 0.0 43.1
40 2.5 54.8 0.0 42.7
Exchange a e
1 4.3 95.5 0.0 0.2
4.8 95.0 0.0 0.2
8 5.0 94.8 0.0 0.2
12 5.1 94.7 0.0 0.2
20 5.1 94.7 0.0 0.2
40 5.1 94.7 0.0 0.2
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 263
ⓒ 2024 Eas Asian Economic Re iew
Table 4 displays he decomposi ion o o ecas e o a iances in ele an a iables
in o componen s a ibu able o each o he model’s o hogonal dis u bances. The able
shows ha he domes ic p oduc i i y has hea ily con ibu ed o ou pu a ia ions a all
ho izons, and he con ibu ion o he mone a y policy shock o ou pu luc ua ions is
nil du ing he second subsample pe iod. Table 4 also shows ha he e ec o he
o eign coun y on he Ko ean economy has subs an ially inc eased as Ko ea has
mo ed om a managed o pegged exchange a e egime o he lexible exchange a e
egime wi h an in la ion a ge ing ule a e he Asian inancial c isis. The o eign
p oduc i i y shock has con ibu ed hea ily o he a ia ions o in la ion a e and
in e es a e by accoun ing o mo e han 30 pe cen o he uncondi ional a iance o
in e es a e and he exchange a e a all ho izons. In addi ion o he dominan ole o
a mone a y policy shock, he p e e ence shock has also con ibu ed o he luc ua ion
o in la ion du ing he second subsample pe iods. Since he mone a y policy has been
conduc ed o s abilize he p ice, he e ec o a mone a y shock on he key
mac oeconomic a iables excep in la ion is nil as in Table 4.
Table 5. Maximum Likelihood Es ima es and S anda d E o s (2007:3Q-2018:4Q)
Pa ame e
TAN
K
Model RAN
K
Model
Es ima e S anda d E o Es ima e S anda d E o
α 0.1227 0.0044 0.3189 0.0002
Λ 0.2559 0.0018 0 -
𝜌 0.8964 0.0035 0.9129 0.0089
ρ
∗ 0.9026 0.0110 0.8993 0.0006
𝜌 0.8565 0.0415 0.7625 0.0016
𝜌 0.3009 0.0120 0.8102 0.0001
𝑎 1.0000 0.0001 1.0102 0.0001
𝑎
0.0025 0.0302 0.1144 0.0006
𝜎 0.0173 0.0025 0.0199 0.0043
𝜎
∗ 0.0714 0.0153 0.0366 0.0039
𝜎 0.0017 0.0002 0.0007 0.0001
𝜎 0.0017 0.0001 0.0311 0.0003
Η 0.8877 0.0045 0.4703 0.0005
𝐿 -695.69 𝐿 -645.49
No e: 𝐿 and 𝐿 deno e he maximized alue o he TANK and RANK models’ log-likelihood unc ion,
espec i ely.
Table 5 p esen s maximum likelihood es ima es o he deep pa ame e s du ing he
G ea Recession, 2007:3Q-2018:4Q. A i s glance, he es ima e o λ which is
compa able o he one in he i s sub-sample pe iod migh signal ha Ko ea has
264 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
success ully o e come he Asian inancial c isis. Bu i migh be he esul o
households’ p ecau iona y beha io in he G ea Recession. The highe unce ain y
abou wha is going on can o ce households o cu consump ion and sa e mo e. The
small es ima e o he nominal p ice igidi y α implies ha he mone a y policy can be
ine ec i e in inc easing ou pu a he cos o in la ion. Also no ice ha he mone a y
policy coe icien 𝑎 implies ha he e is a one- o-one ela ionship be ween a
nominal policy a e and he eal in e es a e nea he ze o-lowe bound. Table 5
indica es ha 𝐿=695.59 and 𝐿=645.49, implying ha LR=100.2. Since he 0.1
pe cen c i ical alue o LR is 10.8, he null hypo hesis ha he e is no H M household
in he economy is ejec ed by he da a.
Table 6. Fo ecas E o Va iance Decomposi ions (2007:3Q-2018:4Q)
Qua e s
Ahea
d
Domes ic P od.
Shoc
k
Fo eign P od
Shoc
k
Policy
Shoc
k
P e e ence
Shoc
k
Ou
p
u
1 87.5 12.5 0.0 0.0
4 80.5 19.5 0.0 0.1
8 79.4 20.6 0.0 0.1
12 79.0 21.0 0.0 0.1
20 78.8 21.2 0.0 0.1
40 78.7 21.2 0.0 0.2
In la ion
1 0.3 0.3 87.2 12.2
4 0.4 19.0 66.1 14.5
8 0.4 20.1 64.2 15.3
12 0.6 21.1 63.3 15.0
20 1.2 30.6 54.6 13.6
40 2.1 53.3 34.3 10.3
In e es Ra e
1 3.3 1.6 0.6 94.5
4 0.7 54.9 0.1 44.3
8 0.9 55.3 0.1 43.7
12 1.7 56.3 0.1 42.9
20 3.2 67.9 0.1 28.8
40 3.5 82.4 0.0 14.1
Exchan
g
e a e
1 3.6 96.4 0.0 0.0
3.6 96.4 0.0 0.1
8 3.5 96.5 0.0 0.1
12 3.5 96.5 0.0 0.1
20 3.5 96.5 0.0 0.1
40 3.4 96.6 0.0 0.2
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 265
ⓒ 2024 Eas Asian Economic Re iew
Table 6 displays he decomposi ion o o ecas e o a iances in ele an a iables
in o componen s a ibu able o each o he model’s o hogonal dis u bances. Fi s , no e
ha nei he a mone a y policy shock no a p e e ence shock is ele an o he a ia ions
in ou pu . The domes ic p oduc i i y shock has been he dominan ac o in he
a ia ions o ou pu by explaining abou 80 pe cen o he uncondi ional a iance o
ou pu a all ho izons. The o eign p oduc i i y shock has subs an ially con ibu ed o
he luc ua ions o ou pu by explaining abou 20 pe cen o ou pu a ia ions. As he
mone a y au ho i y has ied o boos he agg ega e demand by manipula ing i s policy
a e, he e ec o a mone a y policy shock on he uncondi ional a iance o in la ion is
la ge in he hi d subsample pe iods han he ones in he i s and second subsample
pe iods a sho and medium ho izons. The o eign p oduc i i y shock has played an
impo an ole in he a ia ions o in la ion and in e es a es, in addi ion o he
luc ua ion o he in e na ional ela i e p ice. No ice ha he con i ion o a mone a y
policy shock o he a ia ions o o he ele an a iables is nil, implying ha he
mone a y policy is ine ec i e o boos he economy wi h e y low in e es a e, i.e.,
nea he ze o-lowe bound.
Figu e 2. Impulse Response Func ion o a Domes ic P oduc i i y Shock
No es: The lines wi h ci cles, he lines wi h s a s, and he do ed lines display he esponse o selec ed
a iables o a one-s anda d de ia ion o posi i e domes ic p oduc i i y inno a ion in he i s sub-
pe iods (1976:3Q-1997:2Q), he second sub-pe iods (1998:1Q-2007:2Q), and he hi d sub-pe iods
(2007:3Q-2018:3Q).
266 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
Figu e 2 displays he impulse esponse unc ion o some selec ed a iables o a
posi i e domes ic p oduc i i y shock. The ci cle lines ( ), he s a lines ( ), and
he do ed lines ( ) deno e he esponse o ele an a iables o he shock in he i s ,
second, and he hi d subsample pe iods, espec i ely. The eal exchange a e
dep ecia es o he posi i e domes ic p oduc i i y shock wi h he expansion o domes ic
ou pu . The s ong inc ease in ou pu en ails a all o in la ion a e o he shock in he
i s subsample pe iod, which induces he mone a y au ho i y o cu i s in e es a e o
s abilize he p ice as in Figu e 1. Howe e , a mode a e expansion o ou pu o he
posi i e domes ic p oduc i i y shock gene a es mu ed in la ion, which induces he
mone a y au ho i y o mildly adjus i s policy a e in he second and hi d subsample
pe iods. No ice ha he e is a e y mild a ia ion du ing he hi d sub-pe iod, whe ein
he policy a e is nea he ze o-lowe bound du ing he G ea Recession pe iod.2
Figu e 3. Impulse Response Func ion o a Fo eign P oduc i i y Shock
No es: The lines wi h ci cles, he lines wi h s a s, and he do ed lines display he esponse o selec ed
a iables o a one-s anda d de ia ion o posi i e o eign p oduc i i y inno a ion in he i s sub-
pe iods (1976:3Q-1997:2Q), he second sub-pe iods (1998:1Q-2007:2Q), and he hi d sub-pe iods
(2007:3Q-2018:3Q).
2 To ge some in ui ion on he ele ance o he TANK model o e he business cycle in Ko ea, I ha e
added impulse esponse unc ions o he RANK model in he appendix.
Limi ed Financial Ma ke Pa icipa ions and Shocks in Business Cycles in Ko ea 267
ⓒ 2024 Eas Asian Economic Re iew
Figu e 3 p esen s he impulse esponse unc ion o some selec ed a iables o he
o eign p oduc i i y shock in he i s , second, and hi d subsample pe iods. The e ec
o o eign p oduc i i y shock on ou pu is milde han he e ec o domes ic
p oduc i i y shock, bu i s e ec on he eal exchange a e is much la ge han he e ec
o he domes ic p oduc i i y shock since he o eign ou pu en ails a p opo ional
change in he in e na ional ela i e p ice h ough he isk-sha ing condi ion. A e he
Asian inancial c isis wi h he inancial libe aliza ion in Ko ea, he e ec o o eign
p oduc i i y shock on he Ko ean economy has been s onge han be o e he Asian
inancial c isis.
Figu e 4. Impulse Response Func ion o a P e e ence Shock
No es: The lines wi h ci cles, he lines wi h s a s, and he do ed lines display he esponse o selec ed
a iables o a one-s anda d de ia ion o posi i e domes ic p e e ence inno a ion in he i s sub-
pe iods (1976:3Q-1997:2Q), he second sub-pe iods (1998:1Q-2007:2Q), and he hi d sub-pe iods
(2007:3Q-2018:3Q).
Figu e 4 displays he impulse esponse unc ion o some selec ed a iables o he
domes ic p e e ence p oduc i i y shock in he ele an subsample pe iods. The
posi i e impac o domes ic demand shock on ou pu is expansiona y du ing he i s
subsample pe iod, while i s e ec on domes ic economy ac i i y is mild in he second
268 Yongseung Jung
ⓒ Ko ea Ins i u e o In e na ional Economic Policy
and hi d sample sub-pe iods whe ein he boos ing e ec o domes ic demand shock
has been limi ed wi h he inancial libe aliza ion in Ko ea.
Figu e 5. Impulse Response Func ion o an In e es Ra e Shock
No es: The lines wi h ci cles, he lines wi h s a s, and he do ed lines display he esponse o selec ed
a iables o a one-s anda d de ia ion o nega i e domes ic in e es inno a ion in he i s sub-
pe iods (1976:3Q-1997:2Q), he second sub-pe iods (1998:1Q-2007:2Q), and he hi d sub-pe iods
(2007:3Q-2018:3Q).
Figu e 5 shows he impulse esponse unc ion o an in e es a e shock o he selec ed
a iables. The e a e s onge esponses o ele an a iables o he shock in he i s
subsample pe iod han in he second and hi d subsample pe iods. The mu ed esponse
o ou pu and in la ion associa ed wi h a mild inc ease o he policy a e du ing he
second and hi d subsample pe iods displays ha he mone a y au ho i y has
success ully conduc ed i s policy o s abilize he economy wi h an adop ion o he
in la ion a ge ing ule a e he Asian inancial c isis.
V. Concluding Rema ks
This pape speci ies a simple wo-agen small open economy new Keynesian model
wi h incomple e inancial ma ke , and hen in es iga es he ole o H M households in
