34 2017, XX, 4
Ekonomie
DOI: 10.15240/ ul/001/2017-4-003
In oduc ion
In his a icle, we discuss he ela ionship o
banks, o loans p o ided by hem, and economic
de elopmen . We decided o in es iga e his
ela ionship as banks in oday’s economies
play a signi i can ole as a i al ins i u ion in he
i nancial ma ke s, whe e he e is a dis ibu ion
o mone a y unds om su plus en i ies o
de i ci en i ies. A necessa y p econdi ion o
a unc ioning economy, in i s p esen mos ly
mixed o m, is a unc ioning and s able banking
sys em. Cu en ly, unco e ed money is o
a g ea ex en he money gene a ed by p i a e
banks, mainly in he o m o loans.
In hese u bulen imes, when economies
a e g owing and declining a as e in e als
han we e cus oma y in p e ious decades, his
is e y much a con empo a y issue. I is due
o he ac ha he a e o g ow h/decline in
lending, due o he impo ance and size o he
i nancial ma ke s and he o m o he unco e ed
money issued, signi i can ly in l uences he
economic cycle. As epo ed by Če noho ský
(2015), banks p o ide loans o businesses
and households o hei consump ion and
in es men and hus suppo he economy.
Wha is impo an is he du a ion o a loan, as
pa icula ly long- e m in es men s con ibu e o
long- e m economic g ow h. Fo his eason, we
decided o examine he impac o o al loans, as
well as di iding hem in o loans o non- i nancial
businesses, loans o households, mo gage
loans and consume loans. The impo ance o
c edi access o banks is also compounded by
he i nancial and economic c isis which mos o
he de eloped coun ies expe ienced in ecen
yea s. The consequence o his c isis oday is
an abno mal si ua ion on he i nancial ma ke s,
which is e l ec ed in nega i e in e es a es,
o eign exchange in e en ion and quan i a i e
easing by cen al banks. They a e ying o
use hese uncon en ional mone a y policies
o es o e he impai ed c edi channel o he
mone a y policy ansmission mechanism.
The main idea o his a icle is exp essed
by he hypo hesis ha he de elopmen o
a ious ypes o bank lending has a posi i e
e ec on economic de elopmen . We will
examine he alidi y o his hypo hesis using
selec ed s a is ical and ma hema ical me hods
as p esen ed below.
The aim o his a icle is o assess he impac
o he de elopmen o di e en ypes o loans in
he banking sec o on economic de elopmen ,
based on he example o he Czech Republic.
In achie ing his se goal, we shall begin wi h
he hypo hesis ha economic pe o mance
inc eases wi h he g ow h o he a e o a ious
ypes o loans.
1. Theo e ical Backg ound
In he pas , he ela ionship o he i nancial
sys em and economic g ow h was examined
by a numbe o enowned economis s.
Schumpe e (1912) emphasised he s ong
in l uence o banks on economic g ow h by
encou aging inno a ion. P o iding loans o
hese inno a ions and in es men s leads
o he g ow h o business ope a ions and
hus o economic g ow h. In con as , Lucas
(1988), in mo e mode n imes, e e s o an
excessi e in l uence o banks on he economy
in a nega i e sense. Robinson (1952) conside s
ha banks ha e a passi e impac on economic
de elopmen . As can be seen, e en in he pas
hese dis inguished economis s held di e en
iews on he impac o he banking sec o on
economic de elopmen . I is s ill he same
oday, as e idenced by subsequen esea ch.
Ou pape is based on he ansmission
mechanism o mone a y policy as i is
unde s ood bo h in economic heo y and
applied in p ac ice in he en o cemen o
mone a y policy by cen al banks. In pa icula ,
TYPES OF BANK LOANS AND THEIR IMPACT
ON ECONOMIC DEVELOPMENT: A CASE
STUDY OF THE CZECH REPUBLIC
Jan Če noho ský
EM_4_2017.indd 34EM_4_2017.indd 34 13.12.2017 12:53:2813.12.2017 12:53:28
35
4, XX, 2017
Economics
we ollow he logic o he c edi channel o he
mone a y policy ansmission mechanism.
This is based on he change in he in e es
a es se by he cen al bank, which a ec
in e bank in e es a es and in he end also he
ma ke in e es a es o e ed o clien s. I we
conside declining in e es a es, his esul s
in a highe demand o loans om banks by
companies and households. These loans a e
used o co po a e in es men and household
consump ion (o om a mac oeconomic poin
o iew also o in es men s h ough he
pu chase o eal es a e). A he same ime, he
amoun o money in ci cula ion is g owing. The
inc ease in consump ion and in es men hus
con ibu es o he g ow h o he economy. In he
case o inc easing in e es a es, he change
in he gi en a iables is he opposi e o i may,
o example, esul no only in a dec ease,
bu also in a decline in he g ow h a e, o he
gi en quan i ies. A de i ni e ac o in suppo o
his p ocess is he cen al banks moni o ing
and s i ing o in l uence he a e o lending o
a ce ain ex en in o de o suppo economic
g ow h o o ensu e ha he economy doesn’
ge o e hea ed.
The e ec o he amoun o money issued
as bank loans on economic de elopmen
is highligh ed by he main p oponen o
mone a ism, Mil on F iedman (1968). As well,
F iedman and Schwa z (1963) came o he
conclusion ha he co ela ion coe i cien s
be ween he change in money and he nominal
ou pu ange om 0.79 o 0.92 pe su ey
pe iod. They quan i y he ime delays in he
e ec i eness o mone a y policy in a ange
o 12-24 mon hs, wi h he maximum g ow h
in he amoun o money being in ad ance o
18 mon hs ahead o he peak o economic
g ow h. The minimum amoun o money g ow h,
acco ding o his calcula ions, will be e l ec ed in
he economy in he o m o a ecession ea lie ,
wi h an in e al o app oxima ely 12 mon hs.
Also, he money supply is unde s ood as
an au onomous exogenous quan i y gi en
by he cen al bank, which a ec s o he
mac oeconomic a iables.
The logic o his app oach is suppo ed
by Miskhkin (2016), who, in addi ion o he
c edi channel, also de i nes o he channels o
mone a y policy ac ion. Kau mann and Kugle
(2010) also es ima e eal GDP on he basis o
he de elopmen o M3, including he aspec o
coin eg a ion o he gi en a iables. This idea
is suppo ed by Hol emölle (2004), who se s
he ime delay o he mone a y policy ools on
p oduc changes a six qua e s. Fo his he
uses in eg a ion and coin eg a ion analysis.
Cu en ly, he connec ion be ween bank
pe o mance (measu ed, o example, in he
o m o lending a e) and economic pe o mance
is much close . This is due o he eno mous
scale o globalised and also local i nancial
ma ke s due o he size o he economies
and hei impac on business ac i i ies. Today
he e is a highe deg ee o in e connec i i y o
i nancial ma ke s and economic de elopmen .
The i nal p oo is ce ainly he i nancial c isis in
he USA. I de eloped p ima ily in he banking
sec o and spilled o e in o a public i nance c isis
and an economic down u n in he economically
impo an coun ies in he wo ld and Eu ope.
The e o e examining he ela ionship be ween
bank lending and economic de elopmen
has gained impo ance. Among he a ious
wo ks a ious indica o s o lending a e used o
measu e economic pe o mance.
These con ibu ions can be di ided in o
h ee basic g oups. The mos signi i can in
e ms o numbe s is he g oup o economis s
who belie e in he posi i e impac o bank loans
on economic de elopmen . Le ine and his co-
economis s in hei wo ks (Le ine & Ze os,
1998; Beck, Le ine, & Loayza, 2000; Beck &
Le ine, 2004) examined a ious combina ions
o he e ec s o he liquidi y o s ock ma ke s
and banks (collec i ely, i nancial in e media ion)
on economic g ow h, capi al accumula ion and
inc eased p oduc i i y. All he abo e, acco ding
o he au ho s, is posi i ely in l uenced by he
ac i i ies o banks. A meanu e al. (2015) es ed
he e ec s o c edi expansion on sus ainable
economic g ow h. They see a g ea e e ec
wi h loans o legal en i ies a he han o na u al
pe sons. The impo ance o loans o legal en i ies
(companies) ac s o e a longe pe iod, because
hei in es men s lead o u he g ow h. Banu
(2013) ocused on he ques ion o whe he an
economy, speci i cally he Romanian economy,
would be capable o economic g ow h in he
absence o lending. Wi hou he loans p o ided
o he p i a e sec o , he Romanian economy
would no g ow, as no new p ojec s would
a ise. Con e sely, e y low dependence was
ound be ween loans o he public sec o and
economic g ow h. Kelly e al. (2013) began wi h
a ange o da a om 10 yea s o he economy
o I eland, which was signi i can ly a ec ed
EM_4_2017.indd 35EM_4_2017.indd 35 13.12.2017 12:53:2913.12.2017 12:53:29
36 2017, XX, 4
Ekonomie
by he i nancial c isis speci i cally because o
he banking sec o . Despi e his signi i can
l uc ua ion, Kelly i nds a posi i e impac o
lending ac i i ies on g ow h in he economy.
As well, E misoglu e al. (2013) in es iga ed
whe he da a on loans would be an app op ia e
o ecas o he de elopmen o g oss domes ic
p oduc (GDP). They s essed he impo ance
o loan da a in e ms o a minimum delay. Again,
hey ound a posi i e e ec ; i.e., hey s a e ha
using he a iable “c edi incen i es” in GDP
p edic ion models inc eases hei accu acy.
The esul s o a u he s udy by Ce o elli
and Gambe (2001) show ha he banking
sec o acili a es access o c edi o “young”
i ms, he eby suppo ing he pace o economic
g ow h, as in es men s by new i ms a e mo e
likely o be in ol ed in inno a i e echnologies.
Benci enga and Smi h (1993) conclude ha
he banking sec o can also educe excessi e
c edi limi a ion h ough educed moni o ing
cos s and hus ensu e accele a ed economic
g ow h in a coun y. Le ine (2005) shows he
link be ween he ope a ion o he i nancial
sys em and economic g ow h.
On he o he hand, he e a e s udies
ha show a nega i e ela ionship be ween
bank loans and economic de elopmen as
measu ed by GDP g ow h. Lei ao (2012) came
o his conclusion based on an analysis o
mac oeconomic a iables (economic g ow h,
ade balance and in l a ion) and bank loans. He
concluded ha in l a ion is nega i ely co ela ed
wi h economic g ow h. The main idea behind
he s udy is ha excessi e c edi g ow h ends
o weaken a banking sys em and inc ease
in l a iona y p essu es, he eby unde mining
economic g ow h. Mian e al. (2015) based hei
s udy on an analysis o he ela ionship be ween
household deb and GDP. Acco ding o hei
esul s, he g ow h o household deb in ela ion
o GDP p edic s a lowe g ow h in p oduc ion
and highe unemploymen a es in he medium
e m. As well, an inc ease in household deb will
esul in consump ion g ow h and wo sening
cu en accoun balances as a esul o he
inc eased impo o consume goods. Koi u
(2002) published a s udy based on da a om 25
ansi ional economies in he yea s 1993-2000.
In his wo k, he concluded ha an inc ease in
lending does no accele a e economic g ow h.
The causes a e a se ies o banking c ises in
hese economies and i scal es ain . He also
s essed ha hese esul s a e non-s anda d
wi h economic i ndings p ima ily due o speci i c
condi ions in ansi ion economies. Ibáñez-
He nández e al. (2015) came o he conclusion
ha he high g ow h in lending leads o ins abili y
in he i nancial sec o and hus nega i ely
a ec s economic de elopmen .
The e a e also s udies ha do no indica e
any signi i can ela ionship be ween he loans
p o ided and economic g ow h. Fo example,
Taka s and Uppe (2013) in es iga ed he
e ec o bank loans on economic g ow h a e
he i nancial c isis on he basis o da a om
39 i nancial c ises ha had been p eceded
by a c edi boom. They ound ha a declining
amoun o bank lending o he p i a e sec o
does no necessa ily hinde economic eco e y
a e a i nancial c isis. In hese c ises, changes
in he a e o bank lending, ei he in eal e ms o
in ela ion o GDP, do no co ela e wi h g ow h
du ing he i s wo yea s o eco e y. In he
hi d and ou h yea , he ela ionship becomes
s a is ically signi i can , bu s ill emains
insigni i can in economic e ms. De G ego io
and Guido i (1995) ound a posi i e co ela ion
be ween he g ow h a e o bank loans o he
p i a e sec o and he g ow h o GDP, bu he
impac a ies in di e en coun ies. In La in
Ame ican coun ies, he ela ionship is ac ually
nega i e. Thei a ionale was he ecen i nancial
libe alisa ion in hese ma ke s combined wi h
a poo le el o egula o y amewo k. They also
emphasise ha he main me hod whe eby he
g ow h o lending a ec s economic g ow h is
p ima ily ha hese loans mus be p o ided o
e ec i e p ojec s; i.e., no a c i ical amoun o
hese loans.
Based on he lis men ioned abo e, i is
clea ha s udies a e p e alen which con i m
he logic o he c edi channel o he mone a y
policy ansmission mechanism and show
a posi i e ela ionship be ween he g ow h
o lending and he g ow h o he economy.
This co esponds o he cu en p e ailing
heo e ical knowledge o bank con ibu ions
h ough money issuance by p o iding bank
loans o g ow he economy.
As well, we a e awa e o he in e dependence
o he e ec s o bank loans and he de elopmen
o he economy in bo h di ec ions (i.e., ac ing
as a mul iplie and accele a o ). Howe e , in
his a icle we ha e ocused on he impac o
bank loans on he de elopmen o he economy.
A wo-way ela ionship is also aken in o accoun
in he discussion o he esul s achie ed.
EM_4_2017.indd 36EM_4_2017.indd 36 13.12.2017 12:53:2913.12.2017 12:53:29
37
4, XX, 2017
Economics
2. Me hods
This a icle ocuses on examining he
ela ionship be ween wo a iables – loans
g an ed and economic de elopmen . I is clea
ha in economic p ac ice, he e a e a numbe
o ac o s which a ec economic de elopmen .
We ha e d a ed ou analysis on he basis o
he ce e is pa ibus condi ion, which simpli i es
he eal ela ionship, bu is s ill sui able o
examining he ela ionship o wo a iables. In
he i nal discussion, we also de i ne he ac o s
ha will o he wise de i ni ely ha e an impac on
he de elopmen o he economy.
In his wo k we ha e decided o use
coin eg a ion analysis; i.e., a me hod ha
dis inguishes sho and long e m ela ionships
o ime se ies. This is a ela i ely mode n
me hod used in many o he s udies men ioned
abo e and s udies o cen al banks. The esul o
his is whe he he ime se ies a e coin eg a ed
o no . Coin eg a ion means ha he de ia ion
in he di ec ions o he de elopmen o he ime
se ies can only be sho - e m, and he e is a limi
beyond which he de ia ion may no con inue.
The ime se ies a e hen in equilib ium and
ha e a long- e m ela ionship be ween hem;
i.e., hey ha e a common elemen ha can be
examined (A l & A l , 2007). The ad an age o
his me hod o e adi ional s a is ical me hods
is ha i iden i i es any appa en eg ession.
The analysis model selec ed is designed
in acco dance wi h p o essional analyses
and based on he speci i c cha ac e is ics o
he ime sequence. The model is c ea ed
o es ing delays o he dependen a iable
o GDP and s a iona i y es ing, including
necessa y adjus men s o da a by di e encing.
Coin eg a ion is hen es ed and he i nal es is
o pe o m G ange causali y.
The i s s ep is he need o es he ime
sequence on he op imum o de o delays o he
dependen a iable GDP. To de e mine he delay,
we used a calcula ion using Akaike’s in o ma ion
c i e ion (AIC) in he equa ion below:
(1)
whe e M de i nes he numbe o pa ame e s in
he model, is he esidual a iance, and T is
he numbe o obse a ions. The bes ange o
delay is he one whe e he in o ma ion c i e ion
eaches he lowes alues. The es ou pu s
o he bes ange o delay a e applied in he
ollowing es s.
An impo an p e equisi e be o e es ing
coin eg a ion is o e i y he s a iona i y o
he ime sequence being inpu o he model.
S a iona i y o he ime se ies is equi ed in
o de o es ima e he eg ession model. In he
case o non-s a iona y da a and modelling
using he leas squa es me hod, he analysis
could ha e dis o ed ou comes and could
p esen an appa en eg ession. In he case
o non-s a iona y ime sequence, adjus men
should be made using di e en ia ion o he
o iginal da a. A s ochas ic p ocess is a ime-
o de ed se o andom a iables, which in
heo y may be iewed as a unc ion o mean
alue, a iance, co a iance and co ela ion
unc ions. A s ochas ic p ocess is hus e e ed
o as s a iona y i he cha ac e is ics o he
andom a iable a e ime cons an .
These condi ions a e o mally w i en as
ollows (A l e al., 2007):
 Mean alue unc ion:
(2)
 Va ia ion unc ion:
(3)
 Co a iance unc ion:
(4)
 Co ela ion unc ion:
(5)
whe e X is he dependen a iable, E(X ) is he
mean alue D(X ) is he a iance. S a iona i y
e i i ca ion is pe o med using he ex ended
Dickey-Fulle es (ADF es ) o es he
hypo hesis o he exis ence o a uni oo . The es
is based on eg ession o he i s di e ences o
he ime sequence based on hei own delayed
alues, o he delayed di e ences. In p ac ice
he e a e h ee o ms o he ADF es s: wi hou
a cons an , wi h a cons an , and las is wi h a
cons an and a end. The selec ion is made
h ough he lowes Akaike c i e ia. E alua ion is
based on an assessmen o he null hypo hesis
when i is es ed a a signi i cance le el o 0.05, i
he ime sequence has a uni oo . Then we may
say ha he ime sequence is non-s a iona y.
EM_4_2017.indd 37EM_4_2017.indd 37 13.12.2017 12:53:2913.12.2017 12:53:29
38 2017, XX, 4
Ekonomie
Ve i i ca ion o he null hypo hesis is e alua ed
based on he calcula ed p- alues. In es ing,
we assume ha he gene a ing p ocess has he
o m (A l e al., 2007):
(6)
whe e we es ha Ø = 0 ( a iable con ains
a uni oo ), X is he dependen a iable, p is
a delay and e is a esidual componen . In he
e en ha i is a non-s a iona y ime sequence,
i is necessa y o adjus he ime sequence by
using he i s di e ence. Based on he new
alues o he ime sequence we decide on i s
s a iona i y.
I he inpu ime sequences a e non-
s a iona y and a e adjus men by di e en ia ion,
hey acqui e s a iona i y o he same o de ,
i is possible o pe o m a coin eg a ion
analysis. When he abo e condi ions a e
me , he Engle-G ange es (EG es ) will be
applied on he ime sequence o de e mine he
coin eg a ion o he ime sequence. This es
is based on es ing he es ima ed esidues o
he coin eg a ing eg ession o he p esence
o a uni oo . Coin eg a ion eg ession will be
pe o med using he smalles squa es me hod.
In acco dance wi h he Engle-G ange heo ies
in he nex s ep, andom componen s a e es ed
using he ADF es o he p esence o uni oo s.
E alua ion o his es is iden ical o he
ADF es men ioned p e iously, including he
selec ion o he ype o eg ession model by
he lowes AIC. We will es he null hypo hesis
ha he ime sequences a e no coin eg a ed a
a signi i cance le el o 0.05. I he p- alue o
he esidues es ed is highe han he le el o
signi i cance o 0.05, we will no ejec he null
hypo hesis and he es ed ime sequences a e
no coin eg a ed. The a iables a e hen es ed
o a possible mu ual causal link be ween he
es ed se ies on he basis o G ange causa ion.
The coe i cien o de e mina ion obse ed,
o he co ec ed (adjus ed) coe i cien o
de e mina ion desc ibes he closeness o
he connec ion. The esul ing alue can be
in e p e ed in e ms o pe cen age, while
indica ing wha pe cen age he changes in he
esponse a iables a e dependen on changes
in he explana o y a iables. The coe i cien
o de e mina ion indica es he quali y o he
eg ession model; exp essed mo e p ecisely,
i indica es wha pe cen age o a iance o he
esponse a iables is explained by he model
and how much emains unexplained.
The i nal es is o es he causal link
be ween he ime sequences, using he
G ange causali y analysis. G ange de i ned
he concep o causali y in he p ac ical use o
ec o au o eg ession models (VAR models)
o es ic ed and un es ic ed eg ession.
The basic idea is ha when a se ies X a ec s
a se ies Y, hen he se ies X should imp o e
p edic ions o he se ies Y (Hendl, 2012). VAR
models a e based on a compa ison o esidues
o indi idual models di e ing in he numbe o
delays. The mos sui able model is chosen o
a ype ha has a minimum alue o AIC. Fo
he G ange causali y es we will use he null
hypo hesis ha he a iable X does no a ec he
a iable Y unde G ange ’s condi ions. The basic
models ake he ollowing o m (Hušek, 2007):
(7)
(8)
whe e αi and βi a e he coe i cien s o he
a iables, X and Y a e ime sequence a iables,
p is he delay and u is he andom componen .
The i s equa ion es ima es he dependen
a iable based on i s own delayed alues, he
second equa ion adds o i s own delayed alues
he delayed alues o he i s a iable. The
es is conduc ed using VAR models in which
an in e ac ion o up o eigh delays is es ed.
We ejec he null hypo hesis i he p- alue is
less han he signi i cance le el o 0.05. Hnízdo
(2015) u he explains he issue in de ail.
3. Da a
Da a on loans a e aken om he Czech
Na ional Bank (ČNB) da abase. These a e
loans o non- i nancial businesses, loans o
households, mo gage loans, consume loans
and o al loans. All a iables a e in he o m o
ela i e annual change in a qua e ly equency.
These da a a e seasonally adjus ed o he ime
pe iod 2004-2015. The ime se ies is based on
i nancial ma ke de elopmen s and changes in
he Czech economy. One eason o se ing his
se ies is he ac ha be o e 2004 he e we e
signi i can changes in he banking sec o in he
Czech Republic, which changed he owne ship
s uc u e, and he go e nmen had in e ene o
EM_4_2017.indd 38EM_4_2017.indd 38 13.12.2017 12:53:2913.12.2017 12:53:29

39
4, XX, 2017
Economics
s abilise he banking sec o and clea he deb s
o declining la ge banking companies. The da a
a e shown in he ollowing able. De elopmen
o he economy is measu ed using s anda d
indica o s o g oss domes ic p oduc in he
Czech Republic, epo ed in he s a is ics o
he Czech S a is ical O i ce and also e e ed
o in he da abase o he Czech Na ional
Bank. Again, o compa a i e pu poses, hese
a iables a e in he o m o ela i e annual
changes in a qua e ly equency. These da a
a e shown in he ollowing i gu e (Fig. 1).
Fig. 1 shows he s ong g ow h o all
componen s o loans g an ed in 2005-2008,
household loans and mo gages om an ea lie
pe iod. This is ela ed o he apid g ow h o
he economy, including expo s, which a e,
among o he hings, d i en by in es men
ac i i y. This is widely unded by loans
p o ided o businesses as well as by mo gage
loans. In addi ion, in e es a es ha e allen
sha ply a his ime, con ibu ing o a g ea e
willingness pa icula ly among households o
incu deb s. Ano he ac o is undoub edly he
demog aphic de elopmen , whe e a signi i can
po ion o he popula ion in his pe iod deal
wi h hei housing needs. A e 2008, on he
o he hand, he e ec s o he i nancial c isis
in de eloped coun ies begin o show. These
ha e mani es ed hemsel es in he Czech
Republic in he economic c isis and he decline
in in es men and c edi ac i i y esul ing om
a c isis o con i dence; i.e., due o he cau ion o
banks in g an ing loans.
Va iables a e examined in absolu e alues
and in in e -annual changes. Tab. 1 shows
desc ip ions o he a iables used.
Fig. 1: De elopmen o GDP and selec ed ypes o loans (annual change in %)
Sou ce: Czech Na ional Bank (2016)
EM_4_2017.indd 39EM_4_2017.indd 39 13.12.2017 12:53:2913.12.2017 12:53:29
40 2017, XX, 4
Ekonomie
4. Resul s
As men ioned abo e, we a e in e es ed in
whe he he e is coin eg a ion among he
selec ed a iables. Tha is, whe he he gi en
ime se ies e ol e simila ly o e he long e m.
This is, om an economic poin o iew, he
unde lying idea o he c edi channel o he
mone a y policy ansmission mechanism. In
he sho e m, based on he in e p e a ion o
he coin eg a ion analysis, some disc epancies
may occu .
Based on he model chosen, an op imum
o de o delay is es ed, as well as da a
s a iona i y, coin eg a ion es and subsequen ly
he G ange causali y es is pe o med.
Fo his model, we i s de e mined he
op imum delay o de on he basis o AIC
(acco ding o Fo mula 1) o GDP in absolu e
e ms as well as in e -annual changes. The
esul s a e shown in he Tab. 2.
Based on he lowes alue o AIC, we
can conclude ha o he dependen a iable
GDP, he op imum delay is ha o he second
o de . The semi-annual delay iden i i ed will be
e l ec ed in subsequen es s. We will p oceed
o e i y he s a iona i y o he ime sequence.
The i ndings o s a iona i y in he ime se ies will
be made using he ex ended Dickey-Fulle es
(see o mula 6). A null hypo hesis is used o
he ADF es when he ime sequences es ed
a e no s a iona y. S a iona i y es esul s o
he se ies in absolu e alues and in e -annual
changes o GDP and o al loans a e shown in
he Tab. 3.
The esul s o he ADF uni oo es indica e
ha he o iginal da a o all ime sequences
a e non-s a iona y. Non-s a iona i y o he ime
se ies means ha appa en co ela ion could
occu o he co ela ion analysis. S a iona i y
o all ime sequence was achie ed only a e
hei di e en ia ion and he ime sequences a e
he e o e in eg a ed in s age I (1); see Pa 2
o he Tab. 3.
Based on he esul s shown abo e, we
can p oceed o he coin eg a ion es . The
coin eg a ion es is pe o med using he
Engle-G ange es (see o mula 2). This es
equi es non-s a iona i y o he o iginal ime
sequence and he same deg ee o in eg a ion.
Bo h condi ions a e shown in Tab. 3. The null
hypo hesis o coin eg a ion is ha he ime
sequences es ed a e no coin eg a ed. The
ype o es chosen is based on he lowes alue
o he Akaike c i e ion.
Tes ing coin eg a ion ela ionships o
a iables in absolu e alues is pe o med using
he EG es , whe e he model is chosen wi h a
cons an and end based on he lowes alue o
Va iable Mac oeconomic alue Uni Sou ce
Y G oss domes ic p oduc bn. CZK ČNB
Uc To al loans bn. CZK ČNB
∆Y GDP g ow h % ČNB
∆Uc To al loan g ow h % ČNB
Sou ce: own
O de o Delays AIC o ∆Y
13.41229
22.94612
32.99098
43.03142
Sou ce: own
Tab. 1: De i ni ion o he a iables used in he analyses
Tab. 2: Resul s o op imum o de o delay
EM_4_2017.indd 40EM_4_2017.indd 40 13.12.2017 12:53:3013.12.2017 12:53:30
41
4, XX, 2017
Economics
AIC, which amoun s o 1,121.736. The esul ing
calcula ed alues o de e mining coin eg a ion
a e lis ed in he Tab. 4.
Based on es s, we will no ejec he null
hypo hesis o non-coin eg a ion o he ime
sequence, as he calcula ed all p- alues a e
highe han he speci i ed signi i cance le el o
0.05. Fo his eason, in bo h cases, he es
se ies is non-coin eg a ed. In he ollowing
es , causali y is pe o med using a VAR model
(see o mula 3 and 4). Fo he G ange es s o
ime causali y, a null hypo hesis is se , ha he
de elopmen o bank loans does no a ec he
economic cycle and he e o e has no impac on
he o ecas s o he GDP. Tes s a e pe o med
o eigh qua e ly delays, whe e any causali y
can be assumed. Tes esul s o in e -annual
changes o qua e ly alues a e gi en in he
ollowing able (Tab. 5).
Causal ela ions o he de elopmen o
loans o GDP a e shown a a signi i cance
le el o 0.05 a wo, ou and eigh qua e ly
delays. Based on he calcula ed p- alue, i is
possible in h ee cases, o decide o ejec he
null hypo hesis a a signi i cance le el o 0.05.
F om he economic poin o iew i means ha
he de elopmen o bank loans in he o m
o ela i e annual changes ( o o al loans)
Model p- alue E alua ed esul o ADF es
Tes wi h cons an
∆Y0.6492 Time sequence is non-s a iona y
∆Uc 0.7910 Time sequence is non-s a iona y
∆Up 0.5196 Time sequence is non-s a iona y
∆Ud 0.8450 Time sequence is non-s a iona y
∆Uh 0.4006 Time sequence is non-s a iona y
∆Us 0.8225 Time sequence is non-s a iona y
Tes wi h cons an
Fi s di e ence ∆Y0.01756 Time sequence is s a iona y
Fi s di e ence ∆Uc 0.00427 Time sequence is s a iona y
Fi s di e ence ∆Up 0.00007 Time sequence is s a iona y
Fi s di e ence ∆Ud 0.00010 Time sequence is s a iona y
Fi s di e ence ∆Uh 0.00068 Time sequence is s a iona y
Fi s di e ence ∆Us 0.00006 Time sequence is s a iona y
Sou ce: own
Model Va iable AIC p- alue H0:
2 delays wi h cons an ∆Y 239.4507 0.0693 No ejec ed
1 delay wi h cons an and end ∆Up 232.1227 0.4894 No ejec ed
1 delay wi h cons an and end ∆Ud 231.2742 0.2772 No ejec ed
1 delay wi h cons an and end ∆Uh 228.4799 0.1836 No ejec ed
1 delay wi h cons an ∆Us 237.2326 0.1728 No ejec ed
Sou ce: own
Tab. 3: Resul s o ADF s a iona i y es o o al loans and GDP
Tab. 4: Resul s o he E-G coin eg a ion es in annual changes
EM_4_2017.indd 41EM_4_2017.indd 41 13.12.2017 12:53:3013.12.2017 12:53:30
42 2017, XX, 4
Ekonomie
causally ac wi hin he meaning o G ange
causali y on he de elopmen o GDP, wi h a
ce ain ime lag.
The ollowing able (Tab. 6) shows he esul s
o he G ange causali y in qua e ly in e -annual
changes be ween de elopmen s in loans o
non- i nancial businesses and GDP g ow h.
Causal ela ions o he de elopmen o loans
o non- i nancial businesses o GDP a e shown
a a signi i cance le el o 0.05 a wo, ou and
eigh qua e ly delays. Based on he calcula ed
p- alue, i is possible in h ee cases, o decide
o ejec he null hypo hesis a a signi i cance
le el o 0.05. F om he economic poin o iew
i means ha he de elopmen o bank loans
in he o m o ela i e annual changes ( o
loans o non- i nancial businesses) causally ac
wi hin he meaning o G ange causali y on he
de elopmen o GDP, wi h a ce ain ime lag.
The Tab. 7 shows he esul s o he G ange
causali y in qua e ly in e -annual changes
be ween de elopmen s in loans o households
and GDP g ow h.
Causal ela ions o he de elopmen o loans
o households o GDP a e shown a a signi i cance
le el o 0.05 a one, wo, ou and six qua e ly
delays. Based on he calcula ed p- alue, i is
possible in ou cases, o decide o ejec he null
hypo hesis a a signi i cance le el o 0.05. F om
he economic poin o iew i means ha he
de elopmen o bank loans in he o m o ela i e
annual changes ( o loans o households) causally
ac wi hin he meaning o G ange causali y on he
de elopmen o GDP, wi h a ce ain ime lag.
The Tab. 8 shows he esul s o he G ange
causali y in qua e ly in e -annual changes
be ween de elopmen s in mo gage loans and
GDP g ow h.
Null hypo hesis Delay p- alue H0:
∆Uc does no causally ac on ∆Y1 0.1843 No ejec ed
∆Uc does no causally ac on ∆Y2 0.0070 Rejec ed
∆Uc does no causally ac on ∆Y3 0.2816 No ejec ed
∆Uc does no causally ac on ∆Y4 0.0086 Rejec ed
∆Uc does no causally ac on ∆Y5 0.3925 No ejec ed
∆Uc does no causally ac on ∆Y6 0.4210 No ejec ed
∆Uc does no causally ac on ∆Y7 0.1402 No ejec ed
∆Uc does no causally ac on ∆Y8 0.0176 Rejec ed
Sou ce: own
Null hypo hesis Delay p- alue H0:
∆Up does no causally ac on ∆Y1 0.5110 No ejec ed
∆Up does no causally ac on ∆Y2 0.0039 Rejec ed
∆Up does no causally ac on ∆Y3 0.4321 No ejec ed
∆Up does no causally ac on ∆Y4 0.0172 Rejec ed
∆Up does no causally ac on ∆Y5 0.6775 No ejec ed
∆Up does no causally ac on ∆Y6 0.8680 No ejec ed
∆Up does no causally ac on ∆Y7 0.0769 No ejec ed
∆Up does no causally ac on ∆Y8 0.0458 Rejec ed
Sou ce: own
Tab. 5: Resul s o G ange causali y in annual changes – o al loans
Tab. 6: Resul s o G ange causali y in annual changes – loans o non- i nancial
businesses
EM_4_2017.indd 42EM_4_2017.indd 42 13.12.2017 12:53:3013.12.2017 12:53:30