Stress testing: assessing possible impacts of COVID-19 pandemia on the credit default in Portugal
Abstract
The current health crisis is shaking the economic and financial world and Portugal was no exception. The present dissertation main goal is to assess the credit risk impact by the current situation due to COVID-19 pandemic in Portugal. The key objective is to evaluate credit inherent risk, in order to be able to intervene in advance and to mitigate possible risks, as well as predict likely defaults. In order to predict possible serious effects in terms of credit default having, the present dissertation has as its object the study of the impact of macroeconomic variables and their influence on credit default. An exploratory quantitative approach was used in the empirical study, complemented with a qualitative approach, focused fundamentally on the description of the results obtained with the SPSS software. Linear regression models were tested, which were defined as independent variable o credit default. As dependent variables the indicators of credit risk management; UR, LR, LMC, LCC, LBC, EUR, GDP, PSI, DIG, CPI, ER and CP. Among all these indicators of credit risk management used, the ones that had the greatest impact were the LR, LMC, UR, GDP, CPI and CP have more significant effect. However, the variable CP does not suggest the existence of a direct and reliable relationship between the independent variable, but recent studies refer to it as one of the most important to be considered. As noted in world history, the impacts of financial disasters, reinforce the need for systematic analysis and effective financial stability instruments. In order to forecast those situations, stress testing will be used to predict possible scenarios of induced financial crisis by the current healthy crisis, such as credit risk, in specific. To be able to foresee, a country macroeconomic analysis is needed, by merging several credit components, as established by the Basel agreements. The data herein as reference has taken from Banco de Portugal, INE, Stooq and oecd, since the 2003 to 2020, in order to be provisions regarding economic.
Full text
i
STRESS TESTING: ASSESSING POSSIBLE IMPACTS
OF COVID-19 PANDEMIA ON THE CREDIT
DEFAULT IN PORTUGAL
Rena a Baião Se a Be na do Da Sil a
Disse a ion p esen ed as pa ial equi emen o ob aining
he Mas e ’s deg ee in In o ma ion Managemen
ii
NOVA In o ma ion Managemen School
Ins i u o Supe io de Es a ís ica e Ges ão de In o mação
Uni e sidade No a de Lisboa
STRESS TESTING: ASSESSING POSSIBLE IMPACTS OF COVID-19 PANDEMIA ON THE CREDIT DEFAULT
IN PORTUGAL
by
Rena a Baião Se a Be na do Da Sil a
Disse a ion p esen ed as pa ial equi emen o ob aining he Mas e ’s deg ee in S a is ics and
In o ma ion Managemen , wi h a specializa ion in Risk Analysis and Managemen
Ad iso : P o esso Ca los Ra ael San os B anco
Decembe 2021
iii
DEDICATION
To my babies boys,
A u and F ancisco.
To my husband Tomé,
Fo uncondi ional suppo and dedica ion.
And God!
i
ACKNOWLEDGEMENTS
Disse a ion is a long jou ney, which includes a se e al challenges, unce ain ies, g ea happiness, and
i is a soli a y p ocess ha who p oposes is des ined. Despi e o his jou ney was no done en i ely
alone, he e was some people o suppo me, gi e me ene gy and s eng h, whom I should
acknowledge.
I would like o app ecia e my ad iso , PhD. Dou o Ca los Ra ael San os B anco, o suppo ,
coo dina ed and his c i ical and imely iew.
I should like o hank my husband, Tomé, o his uncondi ional psychological and physical suppo o
ake ca e o ou babies.
A supe hank you o pa en s o my husband, o helping us o ake ca e o ou babies, in his e y
oubled pandemic ime.
He e I mus exp ess my deepes g a i ude o Ma a Hei o , Che na Gokaldas, Ri a Figuei edo, Sónia Jin,
João Ca alho, Hen ique Sil a, Ped o Ba bosa, o linguis ic e ision o pa s o his disse a ion.
I would like o con ey my hanks o P o esso Te esa Ve dasca o he comp ehensi e linguis ic e iew
o his p ojec .
Also, I hank my babies, pa en s and sis e o being pa o my li e.
Finally, las bu no leas , o hank you o my Colleagues, Ri a Ma ins, Luz Dias and head o my
Depa men , Ma ia Te esa Paulo, o encou aging me o inish he p esen disse a ion.
ABSTRACT
The cu en heal h c isis is shaking he economic and inancial wo ld and Po ugal was no excep ion.
The p esen disse a ion main goal is o assess he c edi isk impac by he cu en si ua ion due o
COVID-19 pandemic in Po ugal.
The key objec i e is o e alua e c edi inhe en isk, in o de o be able o in e ene in ad ance and
o mi iga e possible isks, as well as p edic likely de aul s.
In o de o p edic possible se ious e ec s in e ms o c edi de aul ha ing, he p esen disse a ion
has as i s objec he s udy o he impac o mac oeconomic a iables and hei in luence on c edi
de aul . An explo a o y quan i a i e app oach was used in he empi ical s udy, complemen ed wi h a
quali a i e app oach, ocused undamen ally on he desc ip ion o he esul s ob ained wi h he SPSS
so wa e.
Linea eg ession models we e es ed, which we e de ined as independen a iable o c edi de aul .
As dependen a iables he indica o s o c edi isk managemen ; UR, LR, LMC, LCC, LBC, EUR, GDP,
PSI, DIG, CPI, ER and CP. Among all hese indica o s o c edi isk managemen used, he ones ha
had he g ea es impac we e he LR, LMC, UR, GDP, CPI and CP ha e mo e signi ican e ec .
Howe e , he a iable CP does no sugges he exis ence o a di ec and eliable ela ionship be ween
he independen a iable, bu ecen s udies e e o i as one o he mos impo an o be
conside ed.
As no ed in wo ld his o y, he impac s o inancial disas e s, ein o ce he need o sys ema ic
analysis and e ec i e inancial s abili y ins umen s. In o de o o ecas hose si ua ions, s ess
es ing will be used o p edic possible scena ios o induced inancial c isis by he cu en heal hy
c isis, such as c edi isk, in speci ic. To be able o o esee, a coun y mac oeconomic analysis is
needed, by me ging se e al c edi componen s, as es ablished by he Basel ag eemen s. The da a
he ein as e e ence has aken om Banco de Po ugal, INE, S ooq and oecd, since he 2003 o 2020,
in o de o be p o isions ega ding economic.
KEYWORDS
C edi De aul , S ess Tes ; Scena io; C edi Risk; COVID-19
i
INDEX
1. In oduc ion .................................................................................................................. 1
1.1. Backg ound ............................................................................................................ 1
1.1.1. P oblem Jus i ica ion ...................................................................................... 2
1.1.2. P oblem .......................................................................................................... 3
1.1.3. Objec i es ....................................................................................................... 3
1.1.4. Main objec i e ................................................................................................ 3
1.1.5. Speci ic objec i es .......................................................................................... 4
1.1.6. S udy ele ance and impo ance ................................................................... 4
2. Li e a u e e iew .......................................................................................................... 5
2.1. SECTION I Regula ion, C edi Risk and mac oeconomic componen s .................. 5
2.1.1. F amewo k his o y and egula ion - Basel Commi ee on Banking Supe ision
5
2.1.2. C edi Risk ....................................................................................................... 8
2.1.3. C edi Risk Componen s ................................................................................. 9
2.1.4. Mac oeconomic componen s a e in luenced c edi isk ............................. 10
2.2. SECTION II S ess Tes .......................................................................................... 13
2.2.1. A b ie his o y o s ess es ing on banks ..................................................... 13
2.2.2. Banking sys em o e iew ............................................................................. 14
2.2.3. How o use s ess es esul s ...................................................................... 15
2.2.4. Main objec i e o S ess Tes ....................................................................... 16
2.2.5. Subjec i e s objec i e p obabili ies ............................................................ 17
2.2.6. Imp o emen s ess es o assess he esilience o he bank ..................... 17
2.2.7. Scena io analysis .......................................................................................... 18
2.2.8. S ess es ing and Value a Risk .................................................................... 20
2.3. SECTION III COVID-19 Pandemic ......................................................................... 22
2.3.1. F amewo k COVID-19 Pandemic .................................................................. 22
2.3.2. Recommenda ions by Eu opean Cen al Bank ............................................ 22
2.3.3. Recommenda ion by Eu opean Baking Au ho i y ........................................ 24
2.3.4. Mac o- inancial models used o calib a e shocks and p oduce scena ios .. 27
2.4. Clima e s ess es ing pilo exe cise .................................................................... 28
3. Me hodology .............................................................................................................. 30
3.1. Resea ch Me hodology ....................................................................................... 30
3.2. Hypo hesis ........................................................................................................... 30
ii
3.3. Resea ch Ques ion............................................................................................... 31
3.4. Sample and Da a Collec ion ................................................................................ 31
3.5. Model Type .......................................................................................................... 32
3.6. Linea Reg ession Model ..................................................................................... 33
3.7. Au o eg essi e Dis ibu ed Lag Model ................................................................ 34
3.7.1. The maximum Lag ........................................................................................ 34
3.8. Speci ic es .......................................................................................................... 35
3.8.1. Du bin-Wa son and B eusch-God ey .......................................................... 35
3.8.2. Tes - and Tes -F ........................................................................................... 36
3.9. Model Va iables ................................................................................................... 36
3.9.1. S ess Tes ..................................................................................................... 36
3.9.2. Unemploymen a e ..................................................................................... 36
3.9.3. Loans ............................................................................................................. 37
3.9.4. TBA – Eu ibo in e es a e ........................................................................... 38
3.9.5. Nominal GDP g ow h a e ............................................................................ 38
3.9.6. PSI 20 ............................................................................................................ 38
3.9.7. Disposable Income G ow h .......................................................................... 38
3.9.8. Consume P ice Index ................................................................................... 38
3.9.9. Exchange a e ............................................................................................... 39
3.9.10. Ca bon P ice .......................................................................................... 39
3.10. Ini ial Model .................................................................................................. 39
3.11. Scena io analysis ........................................................................................... 40
4. Resul s and discussion ................................................................................................ 41
4.1. Resul s Analyses .................................................................................................. 41
4.1.1. P ac ical example ......................................................................................... 41
4.1.2. Va iables ....................................................................................................... 41
4.2. Selec ion and analysis o a iables ...................................................................... 42
4.2.1. Dependen a iable ...................................................................................... 42
4.2.2. Independen a iable ................................................................................... 43
4.2.3. S a iona y Tes .............................................................................................. 44
4.2.4. Da abase desc ip ion .................................................................................... 45
4.2.5. Model pe o mance e alua ion ................................................................... 46
4.2.6. Speci ica ion o model a iables .................................................................. 47
4.3. Robus Tes .......................................................................................................... 49
4.3.1. Co ela ion Ma ix ........................................................................................ 49
iii
4.3.2. E alua ion o he mul icollinea i y ............................................................... 50
4.4. Scena io analysis.................................................................................................. 50
4.5. Final Model Quali y ............................................................................................. 51
4.6. Tes ing The S udy Hypo hesis ............................................................................. 53
5. Conclusions ................................................................................................................. 61
6. Limi a ions and ecommenda ions o u u e wo ks ................................................. 63
7. Bibliog aphy ................................................................................................................ 64
8. Appendix ..................................................................................................................... 67
8.1. Appendix A – S a iona y es ............................................................................... 67
8.2. Appendix B – Desc ip i e S a is ics ..................................................................... 73
8.3. Appendix C – Model Summa y ............................................................................ 73
8.4. Appendix D – ANOVA .......................................................................................... 73
8.5. Appendix E – Coe icien s .................................................................................... 74
8.6. Appendix F -Co ela ions ..................................................................................... 74
8.7. Appendix G – Desc ip i e S a is ics (Model 2) .................................................... 75
8.8. Appendix H – Model Summa y (Model 2) ........................................................... 75
8.9. Appendix I – ANOVA (Model 2) ........................................................................... 75
8.10. Appendix J – Coe icien s (Model 2) ............................................................. 75
8.11. Appendix L – Co ela ions (Model 2) ............................................................ 76
ix
LIST OF FIGURES
Figu e 1 Global Economic P ospec s - June 2020 (Sou ce: The Wo ld Bank (Le )); Pu chasing
Manage s Index – (Sou ce: IHS Ma ki e Re ini i . (Righ )) .............................................. 27
4
1.1.5. Speci ic objec i es
In o de o achie e he gene al objec i e, he speci ic objec i es a e as ollows:
I. Li e a u e e iew ega ding c edi isk, s ess es s and known mac oeconomic ac o s;
II. P opose a model o de aul a es;
III. Tes he model;
IV. E alua e he esul s.
1.1.6. S udy ele ance and impo ance
P esen s udy may con ibu e o policies and ma ke in o de o p e en an a ainable c isis, iden i ying
he main ac o s o de aul a he mac oeconomic le el.
5
2. LITERATURE REVIEW
2.1. SECTION I REGULATION, CREDIT RISK AND MACROECONOMIC COMPONENTS
2.1.1. F amewo k his o y and egula ion - Basel Commi ee on Banking Supe ision
The e a e same impo an c isis in ou his o y. The i s one is G ea Dep ession o 1929. The second
one is Subp ime C isis, when in Sep embe 2008, US au ho i ies decided no o sa e in es men bank
Lehman B o he s, wi nessing u he de e io a ion economic ac i i y and he inancial sys em. In 2008
was ma ked by he c isis in he in e na ional inancial ma ke s and by he economic slowdown global
le el, ansla ing in o a pa icula ly ad e se amewo k o he pe o mance o he ac i i y o banks.
Acco ding o he In e na ional Mone a y Fund he high le e age, le els insu icien capi al and
inadequa e con ingency plans we e ac o s ha con ibu ed o se ious inancial p oblems.
The global p opo ions ha he inancial c isis had assumed, he Eu opean De elopmen Fund and
Eu opean Cen al Bank began o exe cise mo e ac i ely hei ole as he highes inancial supe iso y
au ho i y, p omo ing measu es o con ain he nega i e e ec s and ese e he ecessi e scena io
c ea ed in o de o p e en his c isis om assuming sys emic p opo ions.
In he 1970s, he e was g ea ins abili y in he in e na ional ma ke s, he e o e he G10
1
Cen al Banks
es ablished a Basel Commi ee on Banking Supe ision (BCBS). This was a wo ldwide o um o
discussion and coope a ion on p uden ial banking egula ion. The main goal was o ha monize he
banking supe ision s anda ds o egula o y capi al assessmen .
Cu en ly BCBS consis s o 19 na ions, 32 economies, known as he G20
2
. I s main pu pose is o c ea e
s anda d ules o all in e na ional economies belonging o he g oup, ne e heless each coun y ends
up egula ing acco ding o i s p inciples. The Basel acco ds aim o educe he likelihood o sys ema ic
isk, wi h a ocus on liquidi y, c ea ing de ense mechanisms agains possible inancial c ises.
(Vasconcelos, Pe es M., & C is o, 2017)
The e a e 3 Basel Acco d in exis ence: Basel I, Basel II and Basel III ag eemen s. Basel I app o ed a se
o egula o y p oposals o he wo ld banking sec o in 1988, he In e na ional Con e gence o Capi al
Measu emen and Capi al S anda ds (ICCMCS). The ag eemen made i possible o measu e he
equi emen capi al o co e addi ional c edi isks in he inancial sys em and o minimize he
bank up cy isk. F om his ag eemen highligh s he es ablishmen o minimum egula o y capi al
equi emen s o ensu e he sol ency o inancial ins i u ions and he sys em obus ness p omo ion,
1
The GAB was o med in 1962, when he go e nmen s o eigh IMF membe s—Belgium, Canada, F ance, I aly, Japan, he
Ne he lands, he Uni ed Kingdom, and he Uni ed S a es—and he cen al banks o Ge many and Sweden, ag eed o make
esou ces a ailable o he IMF.
2
The G20 membe s a e A gen ina, Aus alia, B azil, Canada, China, F ance, Ge many, India, Indonesia, I aly, Japan, Mexico,
Republic o Ko ea, Russia, Saudi A abia, Sou h A ica, Tu key, he Uni ed Kingdom, he Uni ed S a es and he Eu opean Union
(EU)
6
h ough he c ea ion o a uni ied s anda d o all in e na ional banks, hus educing he compe i i e
imbalance be ween ins i u ions. (Vasconcelos, Pe es M., & C is o, 2017)
The p og ess made wi h Basel I in egula o y amewo k was undeniable. Howe e , in iew o he
sho comings in he p e ious acco d, ega ding o capi al a bi age
3
he new Basel Acco d, known as
Basel II, was published in 2004. I de ined how o apply he new capi al equi emen s, which is mo e
sensi i e o c edi and ma ke isks. Unde he New Acco d, c edi isk measu emen app oaches a e
classi ied in o wo ypes: s anda dized and based on In e nal Ra ing Based (IRB). The in oduc ion o
IRB me hodologies, in e nal a ings, a e he main inno a ion o his acco d, aiming o make capi al
equi emen s mo e dynamic and sensi i e o isk. Fu he mo e, o imp o e c edi isk managemen ,
Basel II in oduces he ope a ional isk as a majo minimum capi al equi emen isk componen
(Vasconcelos, Pe es M., & C is o, 2017).
BCBS ag eemen is s uc u ed on h ee pilla s:
Pilla I: de e mina ion o minimum capi al equi emen s: opics ela ed o he de e mina ion o
minimum egula o y capi al equi emen s o co e c edi , ma ke and ope a ional isks a e add essed.
Wi h he implemen a ion o his pilla , i is expec ed ha banking ins i u ions will be able o use hei
own me hodologies and mo e sensi i e isk, ul ima ely, hey can bene i om lowe isks han hose
using s anda d me hodologies.
Pilla II: supe iso y e iew: quali a i e supe ision will be ca ied ou by egula o y en i ies o in e nal
con ol o bank isk, equi ing om banking au ho i ies’ s a egies o main ain sui able le els o capi al.
In addi ion o he ein o cemen o he ex e nal supe ision, Pilla II seeks o es ablish a se o in e nal
isk managemen p ocedu es, so ha inancial en i ies could be able he capi al’s sui abili y and
su iciency. This se o p ocedu es was called he In e nal Capi al Adequacy Assessmen P ocess
(ICAAP), which includes s ess es s, which should be ca ied ou wi h he p ope equency.
Pilla III: ma ke discipline: new condi ions o public disclosu e o inancial epo s wi h in o ma ion
abou isks, wi h he pu pose o imp o ing banking isk managemen .
The Basel III: Ag eemen is a esponse o he main ulne abili ies p esen ed by he banking sec o
du ing he 2008 inancial c isis. The Basel Commi ee in oduces a se ies o changes ela ed o Basel II,
wi h emphasis on he capi al s uc u e o inancial ins i u ions, expansion o banks' esilience and
s eng h.
In his ag eemen , a se ies o s udies abou he causes and impac s o he c isis began:
• Iden i ies he main laws in he egula o y models in o ce in he Basel II ag eemen .
• Reshapes he ope a ional amewo k in o de o imp o e he abili y o inancial ins i u ions o
abso b shocks p o ided by he inancial sys em o some sec o o he economy.
• Reduces he isk o con agion in he inancial sec o .
3
A bi age is he pu chase and sale o an asse in o de o p o i om a di e ence in he asse 's p ice be ween ma ke s. I is
a ade ha p o i s by exploi ing he p ice di e ences o iden ical o simila inancial ins umen s in di e en ma ke s o in
di e en o ms.
7
A e he c isis in 2009, he Basel Commi ee made a se o measu es equi ing banks o in eg a e
ma ke isk calcula ions based on VaR models and accompanied by " igo ous and ex ensi e" s ess
es s.
In May 2009, he Basel Commi ee published he inal e sion o i s ecommenda ions on s ess es ing
p ac ices and how s ess es ing should be supe ised by egula o s (Supe ision, May 2009). The
ecommenda ions emphasize he impo ance o s ess es ing in de e mining how much capi al is
needed o abso b losses ( egula o y capi al).
These ecommenda ions highligh ed he impo ance o op managemen and he in ol emen o he
boa d. Bea ing in mind ha op managemen and he boa d o di ec o s mus be in ol ed in he
es ablishmen o s ess es objec i es, scena ios de ini ion, he discussion o he s ess es s esul s
and he e alua ion o po en ial ac ions as well as decision making. He says ha he banks ha wen
well in he mid-2007 inancial c isis we e hose in which senio managemen as a whole was ac i ely
in e es ed in he de elopmen and ope a ion o s ess es s, con ibu ing hese esul s o he s a egic
decision making. The s ess es mus be ca ied ou in all a eas o he bank as a hole.
The Basel ecommenda ions s a e ha many o he scena ios chosen be o e 2007 we e based on
his o ical da a and he e o e less se e e han wha ac ually happened. Speci ic ecommenda ions o
he banks a e:
I. The s ess es mus be an in eg al pa o he bank's cul u e and isk managemen , and so, i
mus a ec decision making;
II. The bank mus ha e a s ess es p og am ha in ol es he p omo ion and he iden i ica ion
o isk con ol, which imp o es he company's managemen on capi al and liquidi y as well as
in e nal and ex e nal communica ion;
III. I mus be aken in o accoun he s andpoin o he en i e o ganiza ion and co e a a ie y o
pe spec i es and echniques;
IV. Policies and p ocedu es mus be placed in o pape by he bank;
V. Flexible in as uc u e o accommoda e changes in s ess es s;
VI. Indi idual componen s mus be equen ly e alua ed;
VII. Co e isks o a eas o nego ia ion - in o de o p o ide a comple e isk image o he
en i e company;
VIII. The s ess es s mus co e a a ie y o p ospec i e scena ios and mus conside
in e ac ions in he sys em and i s eedback e ec ;
IX. To include a se ies o losses, aking in o accoun ha i s epu a ion mus be
main ained among he sha eholde s;
X. Conside simul aneous ma ke p essu es and ac i e asse inancing;
XI. To be e ec i e in isk mi iga ion echniques;
XII. They mus include complex and pe sonalized p oduc s;
XIII. They mus co e isks o pipelines and wa ehouses;
XIV. Imp o e s ess es ing me hodologies o cap u e he e ec o in e na ional isk;
XV. E alua e he ulne abili ies o he asse s classes.
The ecommenda ions o op managemen o senio bank manage s a e as ollows:
I. They mus egula ly e alua e he s ess es p og am;
8
II. Take co ec i e measu es;
III. Assess he se e i y o he scena ios, sensi i i y o po olios o pa ame e s;
IV. In e nal capi al assessmen and liquidi y isk;
V. Implemen s ess exe cises in common base scena ios;
VI. Iden i y sys ema ic ulne abili ies.
2.1.2. C edi Risk
The main ac i i y o banks is he g an ing o c edi , o inance consump ion o in es men by he
popula ion. The bank c edi is a ele an ins umen o le e aging economies, c i ical o possible new
p oduc i e combina ions, and hese will be a he basis o he economies dynamics. The e a e in e nal
s udies o banks in o de o p edic si ua ions o de aul . Basel I came o de ine wha c edi isk was,
de ining he p obabili ies o de aul . BCBS de ined c edi isk ope a ions, which include gua an ees and
in es men s in bank secu i ies.
“Banking c edi is a igh ha he Bank acqui es, h oughou an ini ial cash deli e y
( eal o po en ial) o a cus ome , o ecei e om ha cus ome he amoun due, on
u u e da es, one o mo e cash ins allmen s whose o al alue is equal o he ini ial
deli e y plus he p ice ixed o ha se ice (in e es and commissions). (APB, 2020)
Acco ding o Tabo da e al. (2004), ci ed by (Veloso, 2016), when we e e o bank c edi , we mus ake
in o accoun he ollowing six elemen s:
Pu pose: “wha will be acqui ed wi h he amoun p o ided by he bank and i s use” (Nunes, 2009 -
quo ed by (Veloso, 2016);
Te m: acco ding o Tabo da e al. (2004), e e enced by (Veloso, 2016), can be classi ied as:
▪ Sho e m: ma u i y <1 yea ;
▪ Medium e m: 1 yea <ma u i y <5 yea s;
▪ Long e m: ma u i y> 5 yea s.
P ice: in e es and commissions. Among o he a iables, he g ea e he isk in ol ed in he ope a ion,
he highe he p ice o be cha ged (Veloso, 2016);
Amoun : de e mined acco ding o he clien s' needs and he alue o he asse o be acqui ed (Nunes,
2009 - ci ed by (Veloso, 2016);
Po en ial loss esul ing om he c edi ope a ion (Veloso, 2016);
9
Gua an ees: hey a e a way o compensa ing he c edi o o a possible loss in he business o one
(Ama al e al. (1997), ci ed by (Veloso, 2016).
The de ini ion o c edi isk also includes o he de ini ions, as men ioned (Veloso, 2016):
Coun e pa y Risk - unde s ood as he possibili y o non-compliance, by a ce ain coun e pa y, wi h
obliga ions ela ed o he se lemen o ansac ions ha in ol e he ading o inancial asse s,
including hose ela ed o he se lemen o de i a i e inancial ins umen s;
Coun y Risk - unde s ood as he possibili y o losses associa ed wi h non-compliance wi h inancial
obliga ions unde he e ms ag eed upon by a policyholde o coun e pa loca ed ou side he coun y,
as a esul o ac ions aken by he go e nmen o he coun y whe e he policyholde o coun e pa y
is loca ed, and he isk o ans e , unde s ood as he possibili y o he occu ence o obs acles in he
cu ency con e sion o he amoun s ecei ed;
Commi men isk - unde s ood as he possibili y o disbu semen s o hono su e ies, gua an ees, co-
obliga ions, c edi commi men s o o he ope a ions o a simila na u e;
In e ene isk - unde s ood as he possibili y o losses associa ed wi h o he non-compliance wi h
inancial obliga ions unde he e ms ag eed by an in e media y o con en ion o c edi ope a ions;
Concen a ion isk - unde s ood as he possibili y o c edi losses esul ing om signi ican exposu es
o a coun e pa y, a isk ac o o o g oups o ela ed coun e pa ies h ough common cha ac e is ics.
2.1.3. C edi Risk Componen s
The Basel II ag eemen came o p o ide guidelines o measu ing capi al isk, using in e nal
pa ame e s.
The minimum equi emen s o calcula ing c edi isk capi al (IRB App oach) a e he ollowing:
De aul P obabili y (PD), pe cen age ha co esponds o he long- e m expec a ion o de aul a es,
wi h a ime ho izon o 1 (one) yea o bo owe s o a ce ain le el o c edi isk o homogeneous g oup
o isk;
Exposu e a De aul (EAD), co esponds o he alue o he ins i u ion's exposu e, whe he e ec i e
o con ingen , o he bo owe o coun e pa y a he ime o he e en o de aul , g oss o p o isions
and possible pa ial educ ions a loss;
Loss Gi en De aul (LGD), co esponds o he pe cen age, in ela ion o he obse ed EAD pa ame e ,
o he economic loss esul ing om he spli , conside ing all ele an ac o s, including discoun s
g an ed o c edi eco e y and all di ec and indi ec cos s associa ed wi h cha ging he use;
E ec i e Ma u i y Te m (M), co esponds o he emaining e m o he ope a ion weigh ed by he
cash lows o each u u e pe iod.
Financial ins i u ions, o quan i y c edi isk, a e au ho ized o use one o h ee app oaches o
me hodologies: s anda d me hodology, In e nal Ra ing Based (IRB) Founda ion me hodology (o
simple in e nal me hod) and IRB Ad anced (o in e nal me hods ad anced). (Pe ei a, 2019)
10
In he s anda d me hodology, he isk weigh s a y acco ding o he a ing gi en by he a ing agencies
ce i ied by he supe iso y au ho i ies. These a ings a y wi h he deg ee o compliance o he deb o
and by i s na u e, namely in e na ional o ganiza ion, ins i u ion, so e eigns, companies o eal es a e.
(Pe ei a, 2019)
In he IRB me hodology, he bank uses in e nal c edi isk es ima es o measu e he capi al
equi emen s. Howe e , o use he in e nal models, he ins i u ion is subjec o he app o al o he
egula o , and i is necessa y o comply wi h a se o equi emen s.
In summa y, wi hin he IRB, banks can choose a basic in e nship (Founda ion) o an ad anced
in e nship (Ad anced). In he i s , banks use in e nal es ima es o he de aul p obabili y associa ed
wi h bo owe 's ca ego y, and supe iso s p o ide o he inpu s ( isk componen s). In he second,
banks a e allowed o de elop an in e nal capi al alloca ion p ocess conside ing in e nal es ima es o
he isk componen s. (Pe ei a, 2019)
In he IRB app oach, inancial ins i u ions can assess he Expec ed C edi Losses (ECL) o hei c edi
exposu e. The pe cen age o EL is calcula ed by he p oduc o he PD wi h he LGD; when mul iplied
by EAD, he expec ed loss (EL) is ob ained in absolu e alue - acco ding o Equa ion:
𝐸𝐶𝐿𝑡=𝐿𝐺𝐷𝑡 × 𝑃𝐷𝑡 × 𝐸𝐴𝐷𝑡
The e a e h ee a iable ha can be de eloped o LGD and PD. Fo p esen disse a ion will ocus on
he PD calcula ing me hod, bu i is impo an o unde s and he concep .
The p obabili y o de aul is he p obabili y o coun e pa y de aul ing, i is meaning ha he e no be
able o make paymen s p e iously ag eed he comple ion o he c edi assignmen ag eemen . The
LGD e lec s he pe cen age o exposu e he bank expec s o lose4 i he coun e pa y de aul , and
he EAD e lec s an es ima e o ou s anding amoun in he case o he coun e pa y de aul .
The pa ame e s desc ibed abo e depend an idiosync a ic and sys ema ic componen . The sys ema ic
componen can be explained by ex e nal ac o s, such as mac oeconomic componen s.
2.1.4. Mac oeconomic componen s a e in luenced c edi isk
A pe spec i e o o ecas ing and de eloping s ess scena ios, mo i a ed by ac o s ex e nal, in o de
o espond o he challenges se ou abo e (s ess es ), banks we e o ced o de elop models in
which he de aul a iable, i can po en ially be explained by mac oeconomic ac o s, exogenous o
he dependen a iable. The o mula ion o hese econome ic models aims o y o p edic he
p obabili y o de aul , bu also ha o he a ious ansi ions be ween isk classes, ep esen ed
ypically by a ma ix o mig a ions.
Mac oeconomic s abili y is a undamen al de e minan in he banking sys em (Cha es, 2017).
Mac oeconomic a iables in luence he banking sys em, as well as a coun y's GDP g ow h, in la ion
a e, he in e es a e, loans g an ed, unemploymen a e, household disposable income, exchange
a e was in luence economy.
The e is a la ge li e a u e on c edi de aul isk and economic ac i i y, ha is e e ed o a con agious
e ec cycle, he main ac o s a e c edi isk, i s de aul and economic ac i i y. The mos s udied
hypo hesis emphasizes he ela ionship be ween he de aul a e, income and e enues being
in e connec ed, in expansiona y phases hese a iables ha e opposi e signs, low de aul a e and high
yield. Likewise, when he economy is ecession, he cha ge o inc eases and he emaining a iables
dec ease. (Cha es, 2017)
11
GDP is he business es ima e o all goods and se ices c ea ed in a e i o y in a gi en pe iod o ime.
GDP g ow h is seen as an image o p og ession, composed o he o ali y o p i a e and public
consump ion and in es men , public and p i a e. The symbols o a coun y's economic g ow h is GDP
g ow h, his is a a iable in insically linked o he de elopmen o he inancial sys em. GDP was
cha ac e ized by he business es ima e o goods and se ices c ea ed in a e i o y, du ing a pe iod o
ime, usually one yea . GDP g ow h is seen as an image o he coun y's p og ession, adding p i a e
and public consump ion, wi h public and p i a e in es men . (Yahaya & Oni, 2016)
Repullo & Salas (2011) sugges he need o a capi al cushion when he economy is high, in o de o
le e age he economy when i p esen s lowe han expec ed GDP esul s and a ecession end,
happens a he momen when deb o s ail o epay c edi s, leading o high de aul a es and majo
bank losses.
The e a e s udies ha deno ed GDP g ow h a e has a nega i e e ec on p oblema ic loans, imes o
ecession, c edi de aul s inc ease. (Cha es, 2017)
Ano he simila s udy by Amuakwa-Mensah and Boakye, ci ed by (Cha es, 2017) e e s a signi ican
empi ical e idence o a nega i e ela ionship be ween eal GDP g ow h and c edi de aul . The
nega i e ela ionship is ela ed o he ac ha wi h he inc ease in eal GDP, i gene ally ansla es
in o an inc ease in disposable income, he e may be a dec ease in cha ge o .
A heal hy economic en i onmen means an inc ease disposable income and consequen ly a educ ion
o unemploymen . Acco ding o se e al au ho s who s udied he ela ionship be ween c edi a iables
and unemploymen , hey ound a s ong posi i e co ela ion be ween c edi and unemploymen . Tha
means ha wi h he lack o employmen he bo owe s ail he paymen s. (Cha es, 2017)
Disposable income is he amoun o money ha households ha e a ailable o spending and sa ing
a e income axes. Disposable income is closes o he concep o income as gene ally unde s ood in
economics. Household disposable income is he sum o wages and sala ies, mixed income, ne p ope y
income, ne cu en ans e and social bene i s. (OECD, 2020)
In la ion occu s when he e is a pe sis en inc ease o p ices o goods and se ices wi hou an inc ease
le el o p oduc ion. I means cu ency uni is able o buy less goods and se ices om ano he
economy. This ac o is associa ed wi h an inc ease cu ency in economy. (Yahaya & Oni, 2016)
When his si ua ion occu s, o ising in la ion, mone a y egula o s y o keep up wi h he inc ease
in e es a e, in o de o con ol in la ion. This measu e has consequences on he le el o c edi ,
leading o an inc ease in bo owing cos s. In la ion also has an in luence on households' pu chasing
powe . (Cha es, 2017) I can mean an inc ease in c edi de aul . The e o e, i is expec ed ha he e
will be a posi i e ela ionship be ween in la ion and c edi de aul .
The exchange a e is alue o a na ional cu ency, in uni s o o eign cu ency. The nominal exchange
a e is exp essed in mone a y uni s, on he o he hand, he eal exchange a e exp esses he pu chasing
powe o a na ional cu ency in ol ed in o eign ansac ions o goods o be acqui ed in a coun y in
exchange o he same se in he ou side. (Yahaya & Oni, 2016)
12
The exchange a e has an e ec on he compe i i eness o domes ic p oduc s ab oad, so he
de alua ion inc eases he compe i i eness o domes ic p oduc s ab oad, which may inc ease expo s,
p oduc ion and employmen , and he e may be an inc ease in domes ic p ices causing in la ion.
(Cha es, 2017)
A de alua ion o he alue o domes ic cu ency can make impo ed goods e y expensi e, so banks
a e a high isk o de aul ing on c edi . (Cha es, 2017)
Eu ibo is meaning Eu o In e bank O e ed Ra e. The Eu ibo Ra es a e based on he a e age in e es
a es a which a la ge panel o Eu opean banks bo ow unds om one ano he . The e a e di e en
ma u i ies, anging om one week o one yea . The Eu ibo a es a e conside ed o be he mos
e e ence a es in Eu opean money ma ke . (Eu ibo , 2020)
The ela ionship be ween he de aul and he in e es a e is h ough he cos o bo owe s, when
in e es a e inc ease de aul will be also inc ease.
Ci ed by Cha es, he in e es a es on loans o he e olu ion o c edi de aul a es a e ela i ely small,
since hey only ha e an impac on he sho - e m policies es ablished by cen al banks.
13
2.2. SECTION II STRESS TEST
2.2.1. A b ie his o y o s ess es ing on banks
Be o e he inancial c isis, banks pe o med s ess es s o assess hei in e nal isks, bu i was a simple
exe cise wi h ew impac s on policies and decision-making. A e he inancial c isis, s ess es s
became an impo an ool o suppo isk managemen and assumed a leading ole in egula o y
ac i i y.
The c isis in he 1990s and he cu en inancial ins abili y ha e inc eased he impo ance o a be e
unde s anding o possible ulne abili ies in he inancial sys em o egula o s and banke s. Thus,
se e al quan i a i e echniques ha e been de eloped o assess po en ial isks in inancial ins i u ions.
These echniques a e called s ess es s. Basel Commi ee on banking supe ision sugges ed es ing in
he inancial sec o (Chowdhu y, 2010).
The i s s ess es s based on scena io analysis appea ed in he mid-1990s. Manage s began o assess
hei po olios h ough s a is ical analysis based on his o ical and hypo he ical scena ios. His o ical
scena ios a e based on pas e en s wi h a iew o he u u e. Al hough his is an essen ial ool, he
p oblem is ha pas e en s may no be epea ed in he u u e, hence he need o in oduce
hypo he ical es s, based on se ious bu plausible e en s. These hypo he ical e en s a e based on
changes in he ou look o economic g ow h. A he ime, he e was no pa e n o he use o s ess
es s, he e we e banks ha used he es s o quan i y he maximum loss o a bank in a gi en po olio,
while o he banks sough limi s on ading o quan i ied he amoun o capi al o inance a speci ic
po olio (Wes wood, Den , & Sego iano , 2016)
In 1996, he e was a change o he in e na ional egula o y capi al egime o he ma ke isk, ha is,
he isk o losses in posi ions due o changes in ma ke p ices. A e his change, banks began o use
hei in e nal models o quan i y ma ke isk o de e mine he capi al equi ed in he de aul ’s e en .
In 1999, he Basel Commi ee on Banking Supe ision (BCBS) saw li le p og ess in he echniques and
hei applica ion on he ma ke , he e o e i de eloped echniques o implemen he s ess es applied
o c edi isk.
In 2004, Basel II, a ises he i s s eps o assess c edi and ma ke isks a ise in o de o de e mine he
necessa y capi al, known as egula o y capi al. A e his s anda d, banks we e equi ed o assess hei
c edi isk s ess es s. Despi e his, Basel II had no been implemen ed by all economies be o e he
s a o he inancial c isis. Be o e he c isis, e en banks ha had he s ess es model in place we e
unable o p edic he 2009 c isis. The pu pose o hese es s is o p o ide a quan i a i e measu e o
he ulne abili y o a coun y's inancial sys em o di e en mac o- inancial scena ios and o
complemen he ideas collec ed by o he componen s o he e alua ion. This includes quali a i e
ulne abili y assessmen s and a e iew o egula o y and c isis managemen in a coun y. (Wes wood,
Den , & Sego iano , 2016)
In 2007, he Uni ed S a es expe ienced he wo s inancial c isis since he 30’s. The c isis apidly sp ead
om he Uni ed S a es o o he coun ies and om inancial ma ke s o he eal economy. Some
inancial ins i u ions ailed. Many mo e had o be bailed ou by na ional go e nmen s. The e can be
no ques ion ha he i s decade o he wen y- i s cen u y was disas ous o he inancial sec o .
Risk managemen has now assumed a much g ea e impo ance in inancial ins i u ions (Hull, 2018)
20
A undamen al aspec o any c edi isk VaR model is he c edi co ela ion. The s anda ds o di e en
companies do no happen independen ly o each o he . Du ing an economic down u n, mos
companies a e ad e sely a ec ed and become mo e likely o s op paying. When he economy is doing
well, hey a e a o ably a ec ed and less likely o de aul . This ela ionship be ween de aul a es and
economic ac o s is one o he main easons o he c edi co ela ion. I he c edi co ela ion inc eases
(as i usually does unde s essed economic condi ions), he isk o a inancial ins i u ion wi h a
po olio o c edi exposu es inc eases. (Hull, 2018)
To de e mine he VaR o ES o c edi sp ead-dependen ins umen s, he e a e wo app oaches:
To collec his o ical da a on companies' c edi sp ead a ia ions and use a his o ical simula ion
me hodology.
To model co po a e a ing ansi ions and mo emen s in a e age c edi sp eads associa ed wi h
di e en a ing ca ego ies. (Hull, 2018)
2.2.8. S ess es ing and Value a Risk
Be kowi z (2000) sugges s ha he s ess es will ha e be e esul s i hey a e in eg a ed in o he
VaR calcula ion. This can be done by assigning a p obabili y o each s ess scena io conside ed.
Suppose a inancial ins i u ion has conside ed s ess scena ios and he o al p obabili y a ibu ed o
s ess scena ios is p. Assume he e a e VaR scena ios gene a ed using his o ical simula ion in he usual
way.
Un o una ely, human beings a e no good a es ima ing a subjec i e p obabili y ha a a e e en will
occu . To make he ask o he s ess es commi ee easible, one app oach is o ask he s ess es
commi ee o alloca e each scena io o ca ego ies wi h p e-assigned p obabili ies. The ca ego ies can
be:
P obabili y = 0.05%. Ex emely unlikely. One chance ou o 2,000.
P obabili y = 0.2%. Ve y unlikely, bu he scena io mus be p esen ed in he same way as he
500 scena ios used in he his o ical analysis o he simula ion.
P obabili y = 0.5%. Unlikely, bu he scena io should be gi en mo e weigh han he 500
scena ios used in his o ical simula ion analysis.
S ess es ing in ol es e alua ing he impac o ex eme, bu plausible, scena ios ha
a e no conside ed by VaR o expec ed sho all (ES) models. I he e is one lesson o be lea ned om
he ma ke u moil ha s a ed in he summe o 2007, i is ha mo e emphasis should be placed on
s ess es ing and less emphasis should be placed on he mechanis ic applica ion o VaR and ES models.
VaR and ES models a e use ul, bu hey a e ine i ably backwa d looking (Hull, 2018).
S ess es s meaning VaR. VaR is hough o be a c i ical ool, and i can be a isk o a i m’s po olio
on a day le el and pe o mance o indi idual business uni s. VaR has been ound o be o limi ed use
o ex eme ma ke e en s. I happens because by de ini ion, such e en s occu oo a ely o be
cap u ed by empi ically d i en s a is ical models. The e a e co ela ion pa e ns be ween inancial
21
p ices ( he co ela ion ha would be es ima ed using da a om o dina y imes). S ess es s o e a
way o measu ing and moni o ing he po olio consequence o ex eme p ice mo emen s o his ype.
The bes app oach is he in eg a ion o objec i e and subjec i e p obabili ies. This means ha on he
one hand, Va and ES models mus be in eg a ed h ough his o ical and e ospec i e simula ions bu
on he o he hand, subjec i e p obabili ies mus be in eg a ed in he c ea ion o scena ios h ough
senio manage s, whe e he e is an analysis o a ious ma ke a iables, such as he economic
en i onmen and analysis o di e si ied ma ke scena ios. (Hull, 2018).
22
2.3. SECTION III COVID-19 PANDEMIC
2.3.1. F amewo k COVID-19 Pandemic
The COVID-19 pandemic is b inging a lo o ins abili y o he wo ld, and coun ies also ha e di icul y
on choosing he bes op ions o apply in hei coun y, because many o hese measu es, such as he
case o social con inemen , will ha e e y signi ican economic consequences. The mos ele an
eason is ha small and medium-sized companies and indi iduals a e being a ec ed by his public
heal h c isis, which has economic consequences, in which hey may su e om liquidi y p oblems and
di icul ies on paying hei inancial commi men s. Consequen ly, his will ha e an impac on inancial
ins i u ions and on hei c edi obliga ions. A delay in he paymen o c edi obliga ions will lead o a
la ge numbe o de aul s and an inc ease in und equi emen s o hese same ins i u ions.
The pandemic c isis has a signi ican impac on wo king, ei he o easons o illness o wo ke s and
need o suppo hei amilies o heal h secu i y and manda o y dis ance. The in e ac ion o hese
ac o s clea ly esul s a dec ease p oduc ion on companies, despi e he he e ogenei y obse ed
be ween sec o s o ac i i y. In pa allel wi h he con ex o unce ain y unde lying he e olu ion o he
pandemic all con idence o economic agen s, his led o a signi ican dec ease in he demand o goods
and se ices. A e he g adual educ ion in con ainmen measu es, and despi e some imp o emen
obse ed, he main enance o an un a o able mac oeconomic en i onmen and high unce ain y,
esul ed in an inc ease in unemploymen and a d op income, loss o con idence. Leading o an inc ease
in sa ings o easons o p ecau ion and o he weakening o consump ion and in es men .
In his con ex , in an a emp o minimize he medium and long- e m economic impac s o con ain he
pandemic COVID-19, he coun ies o he Eu opean Union ha e implemen ed a wide ange o suppo
measu es. These measu es a e in he o m o paymen s o c edi obliga ions, in o de o suppo he
ope a ional and liquidi y pa o he c edi deb o s.
2.3.2. Recommenda ions by Eu opean Cen al Bank
ECB (Eu opean Cen al Bank) has implemen ed a se ies o measu es o help he banks o deal wi h he
cu en si ua ion we a e li ing nowadays due o COVID-19, which ha e led many companies, banks and
o ganiza ions o limi hei p ocesses, ac i i ies and human esou ces. I has c ea ed an app oach called
SREP
7
- supe iso y e iew and e alua ion p ocess. This app oach aims o ensu e an e icien and
s uc u ed assessmen o he banks. Th oughou 2020, his p ocess will ocus on he banks' abili y o
7
I is a se o p ocedu es led annually by he supe iso au ho i ies ha ensu es ha each c edi ins i u ion has s a egies,
p ocesses, capi al and liquidi y app op ia e o he isks ha i may be exposed. This p ocess makes he Basel II ope a ional
in he Eu opean and Na ional egula iza ion. In he SREP i is also e alua ed he isk ha each ins i u ion is o he inancial
sys em
23
espond he cu en challenges and isks ela ed o he c isis and manage he impac in he upcoming
mon hs. Supe iso s will assess he ou undamen al Pilla :
I. Scena io design: consis s o he design o he mac o- inancial scena ios o be imposed on he
banking sec o ;
II. Top-down sa elli e models: consis s o he modules used o ansla e he scena ios in o
a iables a ec ing he alua ion o bank balance shee componen s and banks’ loss
abso p ion capaci y;
III. Balance shee module: akes he p ojec ed p o i and losses de i ed om he sa elli e
models o indi idual bank balance shee s wi h he aim o calcula ing he esul ing impac on
each bank’s sol ency posi ions;
IV. Feedback modules: akes he analysis beyond he i s - ound impac on bank capi aliza ion o
assess wha could be he de i ed second- ound e ec s o he ini ial bank sol ency impac in
e ms o con agion wi hin he inancial sys em and in e ms o eedback e ec s o he eal
economy.
The main ocus will be o assess he banks' abili y o deal wi h he c isis and manage he impac in he
coming mon hs.
ECB adop ed an ex ao dina y package in o de o help so e eign deb yields. The i s announcemen
was he pu chase o asse s in he ini ial amoun o 750 billion eu os, inc easing o 1350 billion eu os,
which will con inue un il he end o June 2021 (Pandemic Eme gency Pu chase P og am).
In e nal capi al adequacy assessmen p ocess (ICAAP
8
) and In e nal Liquidi y Adequacy Assessmen
P ocess (ILAAP
9
) a e impo an sou ces o in o ma ion o unde s anding how banks manage capi al
and liquidi y in his e y challenging ime. Supe iso s a e collec ing e idence o bank’s assessmen o
he cu en p ocesses o managing capi al and liquidi y, including he decision making, he abili y o
upda e liquidi y and inancial planning, as well as s ess es scena ios.
The ECB published a p ess elease s a ing ha banks could empo a ily ope a e below he le el o
capi al de ined by Pilla 2, capi al conse a ion and he liquidi y co e age a io.
Ano he measu e ecommended by he ECB is ha banks should no dis ibu e di idends o a iable
emune a ions and hey should use hei capi al o suppo he economy. I also poin s ou ha banks
should con inue o apply solid unde w i ing s anda ds, seek app op ia e policies in he ela ion o he
ecogni ion and co e age o non-p oduc i e exposu es, conduc a solid capi al and liquidi y planning
and ha e a s ong isk managemen .
8
The igo ous e alua ion and de e mina ion o he le el o in e nal capi al unde lying he isk p o ile o a c edi ins i u ion o
in es men company a e essen ial condi ions o he implemen a ion o sus ainable business s a egies, on he assump ion
ha hey a e suppo ed by app op ia e con ols. In pa icula , he planning o he e olu ion o in e nal capi al is conside ed
undamen al o ensu e i s adap a ion on a pe manen basis o he isk p o ile o he ins i u ions, pa icula ly in he ace o
c isis o ecession.
9
The ecen inancial c isis has shown he undamen al impo ance o liquidi y o c edi ins i u ions, gi en ha hei
insu iciency ep esen s an immedia e h ea o hei con inui y. One o he main lessons lea ned is ha liquidi y isk
managemen has o ensu e he abili y o c edi ins i u ions o mee hei paymen obliga ions a all imes, e en unde ad e se
condi ions.
24
Banks will also be able o implemen Pilla 2
10
equi emen s pa ially wi h an in e io capi al (as
addi ional ins umen s o Le el 1 o Le el 2). This an icipa es a measu e in he CRD5 ha should ha e
aken e ec in Janua y 2021 (A icle 104a) - EU banks will be able o se ice he P2R wi h 56.25% CET1
and he emaining wi h AT1 and Tie 2 capi al.
The ECB also wa ns o he need o banks o p epa e o possible nega i e e ec s caused by he sp ead
o he co ona i us.
In he eu o a ea, he ECB's announcemen a se o ex ao dina y measu es has pa ially e e sed he
inc ease in so e eign deb yields. The measu es o be highligh ed by he new eme gency asse
pu chase p og am (Pandemic Eme gency Pu chase P og am, PEPP), ini ially amoun ed o 750 billion
eu os, ollowed by he inc ease o 1350 billion eu os, which should be main ained un il he end o June
2021. In any case, he expec a ion o a longe economic con ac ion is e lec ed in a di e en alua ion
o isk p emiums o he sec o s o ac i i y and ma ke segmen s mos ulne able o he e ec s o he
COVID-19 pandemic. Con inemen measu es and limi a ions on in e na ional mobili y had a
pa icula ly signi ican impac on ai lines and ou ism- ela ed ac i i ies, whe e de alua ions o mo e
han 40% we e obse ed du ing Ma ch 2020, wi h a mo e ecen limi ed eco e y.
2.3.3. Recommenda ion by Eu opean Baking Au ho i y
EBA is one o he h ee en i ies ha a e pa o Eu opean Sys em o Financial Supe ision (ESFS).
The e mo e wo en i ies: Eu opean Secu i ies and Ma ke s Au ho i ies (ESMA) and Eu opean
Insu ance and Occupa ional Pensions Au ho i y (EIOPA). The da e o ounda ion o EBA is 1s o
Janua y o 2011.
The EBA has c ea ed guidelines o ace he c isis c ea ed by he pandemic COVID-19. The measu es a e
limi ed and a e applied only o he economic si ua ion caused by COVID-19. The bene icia ies o hese
guidelines a e sho - e m bo owe s, which include small and medium-sized en e p ises, mo gage
loans o indus ial sec o s a ec ed by his c isis. The measu es a e no applied o deb o s iden i ied
be o e he COVID-19 pandemic ou b eak, nei he o deb o s whose ma u i y has been unchanged.
New loans g an ed a e he ou b eak o COVID-19 a e also no co e ed by hese measu es.
The condi ions o hese guidelines a e o change he paymen schedule in o de o add ess he
sho age o liquidi y. Consequen ly, paymen s a e suspended, pos poned o educed, wi hin a limi ed
pe iod. This will a ec he en i e paymen schedule and may lead o an inc ease in paymen s a e he
10
Eu opean banking law de ines h ee elemen s o own unds. Common Equi y Tie 1 capi al (CET1) is he highes quali y o
own unds and is mainly composed o sha es and e ained ea nings om p e ious yea s. Addi ional Tie 1 capi al (AT1) and
Tie 2 capi al can be equi y o liabili y ins umen s and a e o lowe quali y. Pilla 2 capi al consis s o wo pa s. One is he
Pilla 2 Requi emen o P2R, co e ing isks which a e unde es ima ed o no su icien ly co e ed by Pilla 1. The o he is he
Pilla 2 Guidance o P2G, which indica es o banks he adequa e le el o capi al o be main ained in o de o ha e su icien
capi al as a bu e o wi hs and s essed si ua ions, in pa icula as assessed on he basis o he ad e se scena io in he
supe iso y s ess es s. Unde he new Capi al Requi emen s Di ec i e V (CRDV) banks can ul il Pilla 2 Requi emen s wi h
a minimum 56.25% CET1 as a gene al p inciple. The emaining P2R can be illed wi h Addi ional Tie 1 and Tie 2 ins umen s.
This law was ini ially scheduled o come in o e ec in Janua y 2021 as pa o he la es e ision o he CRDV.
25
pe iod o he special loan condi ions, in pa icula he in e es a e, in o de o a oid majo losses by
he inancial ins i u ion, and he impac on ne p esen alue o be neu alized. Howe e , his change
should no a ec in e es a es.
A inal poin o conside is ha inancial ins i u ions mus con inue o bo ow o new cus ome s.
Howe e , he new loan mus ollow no mal c edi policies and be based on an assessmen o
cus ome s' c edi wo hiness. These new loans mus be assessed agains cu en paymen capabili ies
and indi idual condi ions mus be assessed on a case-by-case basis.
The main objec i e is ha inancial ins i u ions con inue o apply isk policies, paying special a en ion
o he assessmen o deb o s mos likely o ace paymen di icul ies. EBA also ecognizes ha policies
and p ac ices ega ding he assessmen o imp obabili y o paymen may di e depending on he
po olio and he ype o deb o , aking in o accoun he a ailabili y o in o ma ion. These checks a e
expec ed o con inue o he du a ion o he paymen s and a hei comple ion.
Likewise, i is expec ed ha co po a e cus ome s, whe e he po en ial imp obabili y o paymen is
assessed manually as pa o he egula moni o ing p ocess, based on he deb o s' inancial
s a emen s and o he in o ma ion, should con inue h oughou he du a ion o he paymen s un il i s
e m.
Ins i u ions mus iden i y possible economic losses ha may su e due o he applica ion o he
guidelines, including h ough addi ional cha ges o impai men .
Financial ins i u ions mus con inue o assess he likelihood and de aul o paymen s. All go e nmen
measu es o p o ide addi ional suppo o deb o s du ing he cou se o he COVID-19 pandemic should
be e iewed, and be a ailable, so ha i he deb o is in c edi wo hiness, hey should be conside ed
in non-paymen s. Thus, he inancial ins i u ion mus mi iga e i s c edi isks, using non-paymen
assessmen s.
The isk o a gene alized downg ade o lowe le els o in es men assumes pa icula ele ance o
economic agen s who, in a p olonged con ex o low in e es a es, adjus ed he composi ion o hei
po olios looking o g ea e sea ch- o -yield. In he eu ozone, in es men was mos ly in in es men
unds, inc easing i s exposu e o lowe c edi quali y asse s and highe isk o coun e pa y. Gi en he
signi ican size o in es men unds in he Eu opean con ex and he high p opo ion o secu i ies a
he h eshold o he le el o in es men in hei po olios may imply an ab up de alua ion in he
alue o asse s, ep esen ing a sou ce o isk in Eu ope.
EBA w o e guidelines o help comba he c isis caused by COVID-19:
▪ C i e ia o c edi paymen s no o igge o he classi ica ions;
▪ P uden ial equi emen s in he con ex o de aul ;
▪ Ensu e consis en ea men o measu es and calcula ion o capi al equi emen s.
EBA also has a guidelines o s ess es , ha he main isk o assess a e:
• C edi isk;
• Ma ke isk;
• Secu i iza ion;
• Cos o unding.
26
C edi isk is meaning a possibili y o loss esul ing om a bo owe ’s ailu e o epay a loan o
ag eemen be ween wo pa s. I e e s o he isk ha a lende may no ecei e he owed p incipal
and in e es , which esul s in an in e up ion o cash low and inc eased cos s.
Ma ke isk is he isk associa ed o he possibili y ha a bank, as in es o , su e s losses due o ac o s
ha a ec ed nega i ely he pe o mance o he inancial ma ke s. This ype o isk is also called
sys ema ic.
Secu i iza ion is a complex p ocess ha uses inancial enginee ing o ans o m an illiquid asse o
g oup o illiquid asse s in o secu i ies. The isks associa ed o his p ocess inc ease when he
complexi y o he ins umen s inc ease, making ha de he analysis o he u u e secu i y pe o mance.
Secu i iza ion o e s oppo uni ies o in es o s and ees up capi al o o igina o s, bo h o which
p omo e liquidi y in he ma ke place.
Cos o unds is a e e ence o he in e es a e paid by inancial ins i u ions o he unds ha used by
business. The sp ead be ween he cos o unds and he in e es a e cha ged is one o he mos
impo an sou ces o p o i o he inancial ins i u ions. The isk in his ype o p ocess comes when
he cos o unding inc eases. I he cos o unding inc eases, he bank has h ee op ions. Fi s , cha ge
a bigge a e on loans, inding new cos ume s and making p o i in hese new loans. Second, keep he
same a e, inding new cos ume s and losing money in his new loans. Thi d, cha ge a bigge a e on
loans, bu did no ind new cos ume s, losing also money
The measu es a e aimed a suspending o pos poning paymen s o deb o s a ec ed by he COVID-19
pandemic, in which hey saw a educ ion in hei budge , allowing hem o la e esume egula
paymen s when he pandemic si ua ion is al eady sol ed.
The EBA ecommends ha inancial ins i u ions should con inue o assess and iden i y si ua ions, in
which sho - e m paymen s a e commi ed, and in he long e m ha could lead o insol ency. The
eal impac o he economic shock can only be assessed on he basis o isk a ings and measu emen s.
(Final epo - Guidelines on legisla i e and non-legisla i e mo a o ia on loan epaymen s applied in
he ligh o he COVID-19 c isis, 2020)
I should be no ed ha since he p e ious in e na ional inancial c isis, egula o y s anda ds and
supe iso y p ac ices ha e been adop ed, a na ional and in e na ional le els, in o de o inc ease
esilience in he inancial sec o . In he pa icula case o he banking sec o , hese esul ed in a
s eng hening o capi al a ios and liquidi y posi ion. This ac o is pa icula ly ele an in he cu en
pandemic con ex , since he ulne abili ies o he economies, mi iga ing he economic impac o
shocks. (Bank, 2020)
The i s e ec s o he pandemic c isis and he i s measu es adop ed o mi iga ion poin o a e y
signi ican con ac ion in wo ld economic ac i i y in 2020. Con a y o p e ious c isis, i s o igin was
ou side he inancial sec o . The i s economic impac was e lec ed in he dis up ion o dis ibu ion
and p oduc ion chai s a a global le el. The con inemen measu es adop ed by go e nmen au ho i ies
and he d op in he con idence o economic agen s, associa ed wi h unce ain y abou he e olu ion
o he pandemic c isis, caused, in a second phase, a demand shock. In his way, he cu en c isis s ands
ou om p e ious c isis due o simul aneous shocks on he supply and demand side. The con idence
o economic agen s eached his o ic lows and below ma ke expec a ions.
27
Figu e 1 Global Economic P ospec s - June 2020 (Sou ce: The Wo ld Bank (Le )); Pu chasing
Manage s Index – (Sou ce: IHS Ma ki e Re ini i . (Righ ))
As he sp ead o he SARS-CoV-2 i us ook on a global dimension, he unce ain y and he expec a ion
o economic e ec s we e e lec ed in inc ease isk p emiums and an ab up de alua ion o inancial
asse s. This e ec was especially isible in highe isk asse s, namely sha es and deb secu i ies o lowe
c edi quali y, as he p essu e o sell hese ins umen s and he sca ci y o liquidi y was e lec ed in
he inc ease in ola ili y o his o ic highs.
2.3.4. Mac o- inancial models used o calib a e shocks and p oduce scena ios
Mac oeconomic scena io is exogenous shocks, should e lec he unde lying sys emic isks o be
analysed. A mapping sys emic isks in o exogenous shocks has a simula ion ools a e employed o
de e mine he ele an shock sizes and p o iles. The shocks can be used by da a base on his o ical
dis ibu ions and o compu e a 1 % Value -a -Risk measu e he en i e y o ma ke s unde sc u iny.
In wha conce ns COVID-19 i is s ill ha d o p edic wha will happen since i is a new si ua ion, a new
and an unknown i us. The go e nmen has o ake measu es o p e en a U o L b eak and y o ha e
he V shape, hence he impo ance o analyzing scena ios.
In gene al, he e ec s o he pandemic and he consequen ab up and signi ican educ ion in
economic ac i i y a in e na ional and na ional le els, pose addi ional challenges o non- inancial
companies in Po ugal. The dec ease o ac i i y, and he alling o e enues ake o consequen
meaning o insol ency si ua ion and de aul obliga ions in he sho e m. The con inemen measu es
and limi a ions on in e na ional mobili y had a signi ican impac on ai lines and ou ism ac i i ies.
Consequen ly, he e we e b eaks in he o de o 40% du ing Ma ch 2020.
A he inancial le el, unlike o he sec o s o ac i i y, i con inues o eco d a e y s ong de alua ion,
app oxima ely 40%, since he beginning o he COVID-19 pandemic in Eu ope. In e es a es on he
lowes isk asse s a e egis e ed, combined wi h he mos ecen de alua ion o he asse s o which i
is exposed. A possible consequence is an inc ease in he de aul o c edi g an ed o he non- inancial
sec o .
Ins i u ions expec a conside able educ ion in he demand o c edi by indi iduals, ha ing al eady
seen a d op in he i s qua e , a dec ease in consume con idence and a de e io a ion in expec a ions
o he e olu ion o housing p ices. The un a o able e olu ion o hese ac o s could be accen ua ed
28
in a s ong economic ecession, inc easing unemploymen and a e y unce ain economic eco e y.
The i s e ec o he pandemic on he inancial ma ke s was an inc ease in isk p emiums, and a
de alua ion o hese secu i ies.
The pandemic c isis cons i u es a mac oeconomic shock wi h subs an ial impac s no only on he
economic c isis, bu also wi h long- e m nega i e e ec s. The cu en c isis will end o imply
pe manen losses in e ms o p oduc i e capaci y. This esul will be associa ed wi h he des uc ion
o less accumula ion o physical and human capi al and he in e up ion o comme cial and knowledge
ne wo ks.
The long- e m mac oeconomic shock calls o join Eu opean Union esponse measu es in o de o
a oid an asymme ic economic eco e y, and possible asymme ic consequences o na ional banking
sec o s.
In sho , he pandemic c isis has c ea ed a si ua ion o exace ba ed unce ain y, which is pa icula ly
challenging o inancial s abili y a na ional and in e na ional le els. The cu en pandemic c isis is a
es o he inancial sec o 's esilience. In u n, he na u e and implica ions o he pandemic c isis
equi e a coo dina ed esponse a Eu opean le el.
2.4. CLIMATE STRESS TESTING PILOT EXERCISE
Eu opean Cen al Bank and Bank o England will conduc clima e s ess es ing exe cises on he basis
o he same scena ios p o ided by Ne wo k o G eening he Financial Sys em (NGFS), while he Basel
Commi ee is ad ancing on i s s a egy o in eg a e clima e inancial isks in o banking p uden ial
egula ion.
The e a e some limi a ions due o he pionee ing na u e o his exe cise which o aising awa eness
on clima e- ela ed inancial isk. The u u e is o es a clima e s ess ha banks and insu e s will
ha e o gain expe ise in clima e scena io modeling, unde s anding clima e isk ansmission
channels and ely on be e o wa d-looking da a and imp o ed me hodologies. Clima e change is
expec ed o ha e agg ega e inancial impac s ha a e pe asi e in na u e. The main impo an is o
build a exposu e analysis, which in ol es mapping millions o i m and add ess-le el clima e isk
d i e s o inancial balance shee s. On inancial sys em losses could inc ease by almos 10% in he
e en o c edi a ing downg ades o high-emi ing om apid ises in he ca bon p ice o ensu e
alignmen wi h Pa is Ag eemen le els.
The p esen exe cise in ol es s essing he balance shee s banks and insu e s wi h a se o common
hypo heses:
• A long ime ho izon (2050), he use o ansi ion and physical isk scena ios b oken down in
sec o s and a dynamic balance shee ;
• Pa icipan s we e subjec o h ee ansi ion isk scena ios: ze o emission, once physical isk
scena io and based on a 30 yea s ime ame.
Physical isks lead o a signi ican inc ease in sinis ali y and insu e s balance shee s. The exposu es
o banks and insu e s o ansi ion isk is deemed as mode a e, wi h an inc ease cos o isk be ween
30% and 40% o co id-19 c isis inc eased i by 100%.
Physical isk in he banking sys em a e i ms exposed o high o inc easing isk. A ound 10,6% o bank
c edi exposu es o high o inc easing lood isk, 1,4% o coas al loods/ sea le el ise, 11,2% o hea
29
s ess, 12,2% o wa e s ess and 4,8% o wild i es. The dis inc ion be ween haza d p obabili y and
in ensi y becomes impo an when ansla ing haza d in o economic damage. (ECB/ESRB, 2021)
The banking sec o is exposed o i ms ansi ion isk by c edi and ma ke isk. The main analysis
sugges s ansi ion isks o he banking sys em a e p edominan ly om c edi isk. These exposu es
a e ela ed o 14% in o al balance shee a e isks o inancial s abili y appea b oadly manageable.
The eu o a ea banking sys em may be exposed o ail isk in he e en o sudden changes in ca bon
p ices i i ms do no educe hei emissions.
Scena ios ely on a decoupling be ween emissions and GDP g ow h, since GDP g ow h each
emissions neu ali y. This implies ha a apid educ ion o emissions could be achie ed wi hou a
signi ican impac on GDP ha s ess es conduc ed by he EBA equi e scena ios based on h ee
consecu i e yea s o GDP con ac ion. Consequen ly, banks c edi exposu es a e subjec o a ia ions
in p ices om he ajec o y o a ca bon ax and p oduc i i y o ac o s.
Scena io analysis equi es hypo he ical bu plausible scena ios o highligh he impac o clima e isks
on he inancial sys em. I could help cen al banks and supe iso s o in eg a e clima e isk in o
inancial s abili y moni o ing. Mi iga ion policies can be in oduced ei he immedia ely, la e on, o
emain insu icien and include a numbe o echnological assump ions, o ins ance ega ding he
a ailabili y o ca bon dioxide emo al echnologies. (ECB/ESRB, 2021)
The banking sec o is a in e media ing unds o co po a es and is he eby exposed o i ms ansi ion
isk. Bank loan exposu es o clima e sec o s, a ound hal o o al loans. Eu opean Union Membe
S a es exposu es o he housing (36%) and ene gy-in ensi e sec o s amoun (8%) o o al loans
ac oss he eu o a ea.
Mainly concen a ed sec o s o exposu es in he eu o a ea a e manu ac u ing, elec ici y,
anspo a ion and cons uc ion sec o s, con ibu e o a ound wo- hi ds o bank loan-weigh ed
emissions in ensi y.
36
The eg ession es o DW es is augmen ed by explana o y a iables o he o iginal model
𝑋𝑡,1,𝑋𝑡,2,…,𝑋𝑡,𝑘. The s a is ic es is ob ained by 𝑅2 o auxilia eg ession.
3.8.2. Tes - and Tes -F
The simples e sion signi icance es s a e es - ( o indi idual signi icance) and es -F ( o global
signi icance). The es -F is adap a ion o he es - . Conside ing he Linea eg ession equa ion:
𝑦=𝛽1+𝛽2𝑥2+𝛽3𝑥3+⋯+𝛽𝑘𝑥𝑘
In his case, hypo hesis null o es - can be 𝐻0: 𝛽𝑖=0 and hypo hesis null o es -F can be 𝐻0: 𝛽𝑖=
0,𝑤ℎ𝑒𝑟𝑒 𝑖 =1,…,𝑘. The s a is ic es o an indi idual signi icance es can be gi en by:
𝑡 = 𝛽
𝜎𝛽
√𝑛
⁄
Whe e 𝛽
is model OLS es ima o , 𝜎𝛽
is he s anda d de ia ion es ima e o he OLS es ima o and 𝑛 is
numbe o obse a ions.
In he case o es -F, i is necessa y o es ima e an auxilia eg ession, imposing by es ic ions es
and all coe icien s is equal o ze o. The s a is ic es can be gi en by ollowing equa ion:
𝐹 = 𝑅𝑈𝑅
2−𝑅𝑅
2
𝑞
1−𝑅𝑈𝑅
2
𝑇−𝑝
Whe e 𝑅𝑈𝑅
2 is measu e by 𝑅2un es ic ed model, 𝑅𝑅
2 is he model wi h imposed es ic ions, 𝑞 is he
numbe o es ic ions o be es and 𝑇−𝑝 numbe o deg ees o eedom
3.9. MODEL VARIABLES
3.9.1. S ess Tes
The dependen a iable, c edi de aul , o he model is he s ess es s esul s. C edi de aul is a
dummy a iable and i he esul is equal o 1, c edi de aul will be ail in pe iod . Howe e , i he
esul s is equal o 0, c edi de aul in Po ugal will no be ail, so deb o s can pay hei mo gages o
loans.
3.9.2. Unemploymen a e
The unemploymen a e is he pe cen age o he o al labo o ce ha ’s unemployed and ac i ely
seeking wo k. The unemploymen a e is an impo an measu e o economic heal h. I goes up du ing
a c isis pe iod o economic when demand o goods and se ices is low. The unemploymen a e is
one o he de e minan s o he non-pe o ming loans, because households unemployed ha e less
income and he e o e he p obabili y is goes up o hei loans paymen s, so de aul c edi is inc ease
and economic will ail.
37
When unemploymen a e is highe , he e o e de aul c edi is highe . So, i is expec ed ha 𝛽2𝑈𝑅𝑡
will has a posi i e sign, since he unemploymen a e is posi i ely co ela ed wi h he dependen
a iable.
3.9.3. Loans
C edi is an ag eemen be ween a c edi ins i u ion (lende ) and a cus ome (deb o ), who is obliged
o eplay he amoun an ag eed pe iod, wi h in e es cha ge and o he cos s.
Loans o deposi s a io is he a io be ween he o al loans and he o al deposi s o he bank.
Deposi s as deb issuance and sha eholde s equi y a e main sou ces o unds o a bank, being
deposi s he mos s able sou ce o unding. Bigge loans o deposi s a io, means ha deposi s,
inc easing he c edi isk. Also, Sil a ci ied Ape gis and Payne ha i is expec ed ha 𝛽3𝐿𝑅𝑡 p esen s
a posi i e sign, since loans o deposi s a io is posi i ely co ela ed wi h he dependen a iable.
Only c edi ins i u ions and ce ain inancial companies egis e ed in Banco de Po ugal can g an
c edi . The h ee mos impo an ca ego ies used by cus ome s a e: mo gage c edi , consume
c edi and business c edi .
3.9.3.1. Mo gage c edi
Mo gage c edi is an ag eemen be ween wo pa s ha he main pu pose is a acquisi ion o
cons uc ion o pe manen , seconda y o en al house. I also includes c edi con ac s o he
acquisi ion o main enance o p ope y igh s. This ype o c edi is a long- e m loan, he mo gage is
gi en as a gua an ee o epaymen .
C edi con ac s ha no co esponding o a mo gage loan, a e gua an eed by a mo gage o by
ano he equi alen gua an ee commonly used on eal es a e, such as consolida ed c edi o c edi in
which pu pose is no de ined. I is expec ed ha 𝛽4𝐿𝑀𝐶𝑡 p esen s a posi i e sign, since Mo gage
C edi o deposi s a io is posi i ely co ela ed wi h he dependen a iable.
3.9.3.2. Consume c edi
Consume c edi is a speci ic ype o pe sonal c edi in which specialized inancial ins i u ion a ailable
o a indi idual p i a e, ce ain amoun o goods o se ices. Also included consume c edi ca ego y
a e c edi ca ds and sala y accoun s. Rega ding he paymen e ms o consume c edi s a e gene ally
in a sho - e m, a mos up 5 o 6 yea s. I is expec ed ha 𝛽5𝐿𝐶𝐶𝑡 p esen s a posi i e sign, since
Consume C edi o deposi s a io is posi i ely co ela ed wi h he dependen a iable.
3.9.3.3. Business C edi
Business c edi is a speci ic c edi o a business’s accoun s and ac i i y o a business c edi . Business
c edi is used o Small Business Adminis a ion, business’s insu ance p emiums, c edi limi om
endo s and supplie s, o aise money om in es o s and con ac s wi h o he o ganiza ions.
38
I is expec ed ha 𝛽6𝐿𝐵𝐶𝑡 p esen s a posi i e sign, since Business C edi o deposi s a io is posi i ely
co ela ed wi h he dependen a iable.
3.9.4. TBA – Eu ibo in e es a e
TBA – EURIBOR in e es a e is he combina ion o he wo ds Eu o In e bank O e ed Ra e. Eu ibo
a es a e based on he a e age in e es a e on eu o in e bank loans by a la ge numbe o p ominen
Eu opean banks ( he Banking panel). Fo he de e mina ion o Eu ibo a es, 15 pe cen o bo h he
highes and he lowes epo ed pe cen ages a e excluded. E e y business day, a 11:00 am Cen al
Eu opean Time, Eu ibo in e es a es a e eleased and ansmi ed o all pa icipa ing pa ies and
he p ess. I is expec ed ha 𝛽7𝐸𝑈𝑅𝑡 p esen s a nega i e sign, since TBA – Eu ibo in e es a e
co ela ed wi h he dependen a iable.
3.9.5. Nominal GDP g ow h a e
GDP g ow h is one o he kay indica o s o economic heal h. I he economy is g owing, he e is less
isk o a bank ailu e. GDP ep esen s he sum (in mone a y alues) o all goods and se ices
p oduced in egion, o e a pe iod o ime. In coun ing GDP, goods and se ices a e conside ed,
excluding in e media e consump ion goods, a oiding he p oblem o double coun ing. Nominal GDP
e e s o he alue calcula ed in e ms o cu en p ices, ha is, in he yea in which he p oduc was
p oduced and ma ke ed.
In 2018, Sil a ci ed Mayes and S emmel ha ound GDP g ow h has a good p edic i e powe and a
nega i e e ec when p edic ing banks dis ess. So, i is expec ed ha 𝛽8𝐺𝐷𝑃𝑡 p esen s a nega i e
sign, since GDP g ow h is nega i ely co ela ed wi h he dependen a iable.
3.9.6. PSI 20
P ice S ock Index 20 is a index agg ega es he la ges 20 companies lis ed on Eu onex Lisbon. The
main pu pose o PSI 20 is an indica o o he e olu ion o Po uguese s ock ma ke and o suppo
he nego ia ion o u u es and op ions. The PSI 20 index has a cha ac e is ics ha i was a good
indica o o Po uguese s ock ma ke and consequen ly he Po uguese economy. I is expec ed ha
𝛽9𝑃𝑆𝐼𝑡 p esen s a nega i e sign, since he PSI 20 index is nega i ely co ela ed wi h he dependen
a iable.
3.9.7. Disposable Income G ow h
Rep esen s he amoun o money ha households ha e a ailable o spending and sa ing a e
income axes. Gene ally, a dec ease in disposable income, consequen ly he e is less money a ailable
o he same hings. I is expec ed ha 𝛽10𝐷𝐼𝐺𝑡 p esen s a nega i e sign, since he disposable income
3.9.8. Consume P ice Index
The Consume P ice Index measu es he a e o in la ion by de e mining p ice changes o a
hypo he ical baske o goods, such as ood, housing, clo hing, medical ca e, appliances, au omobiles,
and so o h, bough by a ypical household. The CPI a e used o assess p ice changes associa ed wi h
in la ion o de la ion and i is a majo concep in mac oeconomics. Banks gene ally p o i in
39
en i onmen s o in la ion, so p obabili y o a bank ailu e dec ease. I is expec ed ha 𝛽11𝐶𝑃𝐼𝑡
p esen s a nega i e sign, since he in la ion a e is nega i ely co ela ed wi h he dependen a iable.
3.9.9. Exchange a e
The exchange is he ope a ion o change one cu ency o ano he , he exchange a e is he p ice a
which his exchange is made, ha is, he alue o a cu ency in mone a y uni s o ano he cu ency. I
is expec ed ha 𝛽12𝐸𝑅𝑡p esen s a nega i e sign, since he exchange a e is nega i ely co ela ed
wi h he dependen a iable.
3.9.10. Ca bon P ice
The banking sys em is exposed o ail isks in he e en o sudden changes in ca bon p ices i i ms do
no educe hei emissions, bu he impac would be con ained wi h mo e g adual o e icien
emissions educ ions by i ms. This a iable is o analysis he exposu es o loans in Po ugal. I ca bon
p ice changes i will be impac on i ms, consequen ly his i ms ha e loans and u n an impac on
p obabili ies o de aul . I is expec ed ha 𝛽13𝐶𝑃𝑡 p esen s a posi i e sign, since he ca bon p ice a e
is posi i ely co ela ed wi h he dependen a iable.
3.10. INITIAL MODEL
Acco ding o he heo e ical basis men ioned abo e, he ollowing equa ion se es as he basis o
he model:
𝐶𝐷=𝛽1+𝛽2𝑈𝑅𝑡+𝛽3𝐿𝑅𝑡+𝛽4𝐿𝑀𝐶𝑡+ 𝛽5𝐿𝐶𝐶𝑡+ 𝛽6𝐿𝐵𝐶𝑡+ 𝛽7𝐸𝑈𝑅𝑡+ +𝛽8𝐺𝐷𝑃𝑡+𝛽9𝑃𝑆𝐼𝑡
+𝛽10𝐷𝐼𝐺𝑡+𝛽11𝐶𝑃𝐼𝑡+𝛽12𝐸𝑅𝑡+ 𝛽13𝐶𝑃𝑡
Whe e:
: ep esen he qua e ly da a pe iod de = 01/2003 ... 12/2020
𝛽1: Cons an ha ep esen s he o dina e a he o igin o he model.
𝛽2,…,𝛽13: ep esen he coe icien s associa ed wi h he espec i e a iables.
CD: C edi De aul coe icien is he dependen a iable ep esen ed by he speci ic
de aul a e index
UR: Rep esen s he unemploymen a e in Po ugal.
LR: Rep esen s he a e o loans g an ed in Po ugal
LMC: Rep esen s he a e o loans – Mo gage C edi g an ed in Po ugal
LCC: Rep esen s he a e o loans – Consume C edi g an ed in Po ugal
LBC: Rep esen s he a e o loans – Business C edi g an ed in Po ugal
40
EUR: I ep esen s he a e age in e es a e on in e bank loans in he eu o a ea.
GDP: Rep esen s he nominal alue o GDP.
PSI: Rep esen s he a e o he alue o he 20 la ges companies in Po ugal.
DIG: Rep esen s he nominal alue a e disposable income g ow h.
CPI: Rep esen s he a e o in la ion alue.
ER: Rep esen s he exchange a e agains he US dolla .
CP: Ca bon P ice a e.
3.11. SCENARIO ANALYSIS
The main objec i e o scena io analysis is o o e a lexible me hodological amewo k which can
ake in o accoun he o wa d-looking isks. I can p o ide a sys ema ic way o making s uc u ed
assump ions abou di e en possible u u es o explo e he isks. The p esen s udy, he scena io
analysis equi es hypo he ical bu plausible scena ios i clima e isks ha e impac on inancial
ins i u ions and ano he a iable.
The i s elease o clima e scena ios was in oduced in June 2020, include elemen s o bo h
ansi ion and physical isks. The p esen s udy i included ca bon dioxide emo al echnologies.
The i s phase o scena io analysis consis ed on mo o iza ion o mac oeconomic a iables s udied, as
s a ed in Pilla II, which consis s o s eng hening ex e nal supe ision o help mi iga e he isks
inhe en in he banking sys em.
The scena io analysis was based on he linea eg ession model p e iously s udied, in o de o p edic
he ou pu a iable o c edi de aul . Banco de Po ugal and ECB ha e p ojec ions o some o he
a iables s udied, and hese made p ojec ions will be used o he analysis o di e en scena ios, Banco
de Po ugal and ECB ha e some in o ma ion a ailable ha he explo e does no ha e, he asymme ic
in o ma ion. Va iables ha e no a p ojec ion in Banco de Po ugal we e p ojec ed acco ding o
his o ical da a, undamen ally in main aining he beha io and ajec o y obse ed in he a iables
o iginally epo ed by Banco De Po ugal.
41
4. RESULTS AND DISCUSSION
4.1. RESULTS ANALYSES
In p esen chap e he case s udy is p esen ed wi h he applica ion o he me hodology in he p e ious
chap e , o ganized in wo pa s: a i s pa ha consis s on measu ing he a iables ha d aw he
model, a second pa ha consis s on he de elopmen o he applica ion o he esul s o ob aining
base and s ess scena ios in he con ex o he cu en pandemic.
All da a ex ac ion and ea men we e pe o med in SPSS and Mic oso Excel.
4.1.1. P ac ical example
The model was aken om se e al sou ces o in o ma ion, such as Banco de Po ugal, S ooq and OECD.
The model o be buil is based on he coun y's his o ical da a om 2003 o he las qua e o 2020,
in o de o be able o co e a majo c isis and hus be able o make u u e p ojec ions o he cu en
si ua ion we li e due o he e ec o COVID -19.
The main goals o he analysis is he occu ence o c edi de aul by cus ome s and i s e ec on he
economy. As he e is a possible con agion e ec o c edi isk, he lack o supply (labo ) may lead o a
up u e in demand, o in he cu en si ua ion, bo h b eak simul aneously, he analysis co e s all
impo an c i e ia o he analysis, such as Unemploymen , nominal GDP, nominal disposable income
g ow h, CPI in la ion a e, psi20 index, Loans, Exchange a e, TBA - 3-mon h Eu ibo and ca bon p ice
a e.
The dis ibu ions o losses es ima ed using he model will be e lec ed in mac oeconomic e ms, in
o de o assis in he cons uc ion o policies and s a egies o isk mi iga ion in ad e se
ci cums ances.
4.1.2. Va iables
Due o he pandemic si ua ion ha Po ugal and he wo ld a e expe iencing, he clima e change , i is
necessa y o eso o some mac oeconomics a iables, in o de o obse e which o hese a iables
ha e he g ea es impac on un a o able economic en i onmen ha could lead o an inc eased isk
o c edi de aul , leading o a e y ulne able si ua ion. In his way, i becomes pe inen o analyze
he selec ed a iables, assess how hey may a ec he c edi de aul and wha a e he bes policies o
implemen .
42
The p esen s udy has a pu poses e idence o he ela ionship be ween c edi isk (dependen a iable)
and mac oeconomics a iables ha is explana o y a iables.
4.2. SELECTION AND ANALYSIS OF VARIABLES
4.2.1. Dependen a iable
The dependen a iable is he de aul c edi a e, whose alue has been aken by Banco de Po ugal’s
da abase. The de aul c edi a e is a a io be ween he amoun o c edi ha e been aken by
companies and pa icula s, bu he o al amoun o c edi a ibu ed o hese wo ca ego ies o
economic agen s a e epaymen is doub ul.
Figu e 4 p esen s he his o ic e olu ion o c edi de aul a e in Po ugal. The las in e na ional inancial
c isis was o igina ed in he USA, de i ed om he p oblems wi h housing loans, he well-known
subp ime c isis, in he summe o 2007. The Ame ican economy eached a c i ical poin in 2008 wi h
he bank up cy o Lehman B o he s, one o he main banking ins i u ions in he coun y. The e ec s o
he ecession in he Ame ican economy ha e sp ead globally, causing a sha p d op in c edi de aul .
Po ugal was s uggling wi h a s agnan economy and a ising unemploymen . Al hough Po ugal ell
in o a se ious ecession, e lec ing a c i ical poin be ween 2012 o 2015 in c edi de aul . Po ugal ell
in o a se ious ecession, e lec ing a c i ical poin be ween 2012 o 2015 in c edi de aul .
43
Figu e 2 – E olu ion o c edi de aul since 2003Q1 o (sou ce: p epa ed by he au ho wi h da a base
o Banco de Po ugal.)
4.2.2. Independen a iable
All a iables men ioned abo e a e impo an o cons uc ion model, as hey all in luence he de aul
a e in Po ugal. In his way, he model uses all a iables, such as Unemploymen Ra e (UR), Loans Ra e
(LR), Mo gage C edi (LMC), Consume C edi (LCC), Business C edi (LBC), Th ee Mou h TBA –
EURIBOR (EUR), Nominal GDP G ow h Ra e (GDP), Po ugal S ock Index 20 (PSI), Disposable Income
G ow h Ra e (DIG), Consume P ice o In la ion Ra e (CPI), Exchange Ra e (ER) and Ca bon P ice (CP).
0,00%
5,00%
10,00%
15,00%
20,00%
25,00%
30,00%
35,00%
Qua e
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
4
2
yea 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
C edi De aul Ra e [%]
C edi De aul Ra e [%]
44
4.2.3. S a iona y Tes
The i s s ep is o analyse whe he he se ies is s a iona y o no , in o de o choose absolu e alues
o lag alues. A s a iona y se ies ci cula es be ween wo bands (a maximum and a minimum) loa ing
uni o mly a ound a midpoin – p ocess I(0). A non-s a iona y se ies has no band, since he p obabili y
o 1 exceeds i and does no luc ua e a ound a midpoin – p ocess I(1). The e a e wo cha ac e is ics
o a s a iona y se ies: hei p ope ies a e cons an and ha e ze o mean and cons an a iance. Thus,
i is necessa y o dis inguish p ocesses I(0) and I(1). S a iona y es s a e used o unde s and he ex en
o which he se ies has a endency, and hus emo e hem in o de o obse e he eal luc ua ions o
he a iable. The main objec i e is o obse e he eal luc ua ions o he a iable. The e a e se e al
es s o s a iona i y such as he Augmen ed Dickey-Fulle (ADF) and Philips-Pe on (PP) es s, Ellio ’s
ADF-GLS es s, Ro henbe g and S ock es , M es s o Ng and Pe on, KPSS es s, among o he s.
(Rod igues, 2013)
The chosen es s we e he ADF es s, de eloped by Dickey and Fulle , p io o he uni oo es s, o
ind ou whe he he obse ed ime se ies p esen a s ochas ic end o e he ime o , i on he o he
hand, he se ies p esen a de e minis ic end, wi h s a iona y luc ua ions o e ime. (Fe ei a, 2013)
The ADF model is illus a ed using he ollowing model:
𝑌𝑡=𝜙∗𝑌𝑡−1 +𝜀𝑡
whose, 𝜀𝑡~𝑁(0,𝜎2).
The main objec i e is o es ima e he alue o 𝜙. I 𝜙 =1, i means ha he se ies ha e a uni a y oo ,
he e o e is no s a iona y. Nex s ep is o manipula ed se ies in o de o e e se s a iona y and i is
consis ing o ha di e ences is manda o y. I 𝜙 ≠1 means ha se ies is le el s a iona y, i can be used
wi hou ans o ma ion o es ablish a linea eg ession wi h ano he s a iona y a iable. The e a e wo
al e na i e hypo heses o checking s a iona y:
𝐻0:𝜙=1, he e a e uni a y squa e, means, se ies is no s a iona y
𝐻1:𝜙<1, he e a e no uni a y squa e, means, se ies is s a iona y.
The p esen s udy was based on he p obabili y’s alues calcula ed in Mic oso Excel a signi icance
le el o 5%, in o de o each a conclusion ega ding he s a iona i y o he se ies. I he alue is less
han 5%, he hypo hesis 𝐻0 is ejec ed and hypo hesis 𝐻1 o se ies s a iona i y is accep ed. Appendix
on able A p esen s all esul s o ADF s a iona y es o all a iables in absolu es alues and di e en
lags. In same way o ADF es , PP es s consis s o e i y i null hypo hesis ha e a uni a y oo (𝐻0:𝜙=
1). The decision c i e ion de ines ha , i 𝑡1<𝑟𝑐𝑟𝑖𝑡, he null hypo hesis will be ejec ed, we used
S anda d C i ical Values o signi ican s a is ical le el o 5%, wi h = - 1,65.
The esul s ob ained show ha a iable Business C edi has lagged wi h ρ=0,0001, Nominal GDP has
lagged wi h ρ<0,0001, Nominal Disposable Income has lagged wi h ρ<0,0001, CPI in la ion has lagged
wi h ρ<0,0001, PSI20 has lagged wi h ρ<0,0001, Exchange Ra e has lagged wi h ρ=0,001735 and Ca bon
p ice wi h ρ<0,0001, meaning ha hese a iables ejec ed null hypo hesis and accep hypo hesis 𝐻1
se ies is s a iona i y, all absolu es alues will be used on linea eg ession. All o he a iables p esen ed
only e eal o be s a iona y a di e en lagged, and he e u n alues a e used in he es ima es.
45
Wi h he se ies ans o med in o s a iona y se ies and using le els ha ha e al eady been shown o
be s a iona y, he eg ession is es ima ed be ween he a ious endogenous and explana o y a iables.
4.2.4. Da abase desc ip ion
The da abase used s a s in Janua y 2003 and o 2021, e alua ed by qua e s.
Analyzing he se ies o he collec ed da a, i should obse e he main cha ac e is ics o cen al
endency and a iables o he s udy is dispe sion. Appendix B shows alues o he desc ip i e s a is ics
o all a iables collec ed by he absolu e alues and lagged di e ences.
Rega ding s anda d de ia ion can be obse ed ha he e is no nega i e alue, as expec ed and all
a iables ha e a low s anda d de ia ion alue, his means ha alues a e a ound he a e age wi h ew
de ia ions. The a iables wi h he highes s anda d de ia ion a e ER (Exchange a e), ha ha e a
s anda d de ia ion o 0,1244, and PSI (Po uguese S ock Index) wi h a s anda d de ia ion o 0,1111,
al hough i is no a cause o isk.
Skewness is a measu e o he symme y o a dis ibu ion. In an asymme ical dis ibu ion, a nega i e
skew indica es ha he ail on he le side is longe han on he igh , con e sely a posi i e skew
indica es he ail on he igh side is longe han on he le . Asymme ic dis ibu ions occu when
ex eme alues lead o a dis o ion o he no mal dis ibu ion. (Encyclopedia, 2021) F om he s a is ical
skewness alues ob ained, he one ha is closes o he symme ical alues o he no mal dis ibu ion
is he c edi de aul ( he alue is 0,307). As c edi de aul indica ed LR (loans a e) ( he alue is 0,281),
LMC (Mo gage c edi ) he alue is 0,307, LBC (Business C edi ) he alue is 0,331, EUR (TBA – 3 mon hs
Eu ibo ) he alue is 0,852, UR (Unemploymen a e) he alue is 0,796, DIG (Disposable Income
G ow h) he alue is 0,153, CPI (Consume P ice Index) he alue is 0,018, ER (exchange a e) he alue
is 0,1217, he a iables ha e a longe ail on he igh han on he le side. I means ha he e a e
mo e a ia ions in alues o e a pe iod o ime, e en hough hey a e low. The a iables wi h mo e
a ia ions a e EUR and UR. On he o he hand, LCC (Consume C edi ) he alue is -0,147, CP (Ca bon
P ice) is a alue o -0,673, PSI (Po ugal S ock Index) he alue is -0,693 and GDP (Nominal GDP a e) is
alue is -2,564, his means ha i has a longe le ail han he igh . This measu e indica es ha he
obse ed alues ha e emained homogeneous o e a long pe iod o ime like, o ins ance, he GDP.
Ku osis measu es whe he a se ies has peaks (peaked) o is cons an ( la ) ela i e o he alues o he
no mal dis ibu ion. In he e en ha a a iable has a e y high ku osis alue, i ends o ha e a e y
high peak alue a ound he mean and i dec eases apidly, wi h e y hea y ails. I a se ies shows lowe
ku osis alues, i ends o ha e a la op close o he mean ins ead o an acu e peak. The expec ed
alues o ku osis in a s anda d no mal dis ibu ion a e alues ha end o o a e close o ze o, in his
case all a iables a e cons an CD(-1,308), LR(-0,959), LMC(-1,119), LBC(0,280), LCC(-0,775), EUR(-
0,263), UR(-0,557), DIG(-0,044), CPI(-1,072) PSI (0,017) and ER(-0,585), wi h he excep ion o GDP
which has a peak alue o 8,439 and CP he alue is 5,104.
52
The c edi isk assessmen exe cises a e based on h ee mac oeconomic scena ios, hey include a
baseline ha e lec s he expec ed ajec o y o c edi de aul o wo s uggling scena ios.
Baseline scena io: This is he scena io ha has he e e ence and aims o cap u e he expec ed
e olu ion o economic ac i i y wi h eg ession s udied and p ojec ions by EBC and Banco de Po ugal,
called by baseline.
Scena io 1: This scena io is c ea ed causing a dec ease by 0,5 pe cen age poin s o 4 consecu i e
qua e s, alls 1 pe cen age poin o 2 consecu i e qua e s and goes down o 2 pe cen age poin s in
he las qua e o 2023 on Unemploymen a e, loans and Ca bon P ice, called by pessimis scena io.
Scena io 2: This scena io is c ea ed causing a wo sening in all 5% pe cen age poin s o 4 consecu i e
qua e s, alls 5 pe cen age poin s in he las qua e o 2023 in he same a iables o scena io 1, called
by ex eme pessimis scena io.
G aphic 1 - Analyses Scena ios - combina ion mac oeconomics ac o s wi h Cha ge O (sou ce:
p epa ed by he au ho )
The p ojec ed scena ios con i m ha he e may be a la ge inc ease in c edi de aul , he e o e
con i ming he need o apply imp o emen ac ions. G aphic 1 shows he base, pessimis and ex eme
pessimis p o ision o he 3 scena ios wi h he combina ion o he calcula ed mac oeconomic a iables
wi h he coe icien s ound esul ing in he C edi De aul alue.
F om he p e ailing in o ma ion ha is cu en ly in o ce, ega ding c edi de aul , he lag in he
escala ion o he a iables was applied and e lec s a possible inc ease in c edi de aul a e he end
o mo a o ium c edi . The ex eme pessimis scena io ale s o le els o c edi de aul ne e eached
in Po ugal be o e.
The ollowing g aphics p o ide o ecas s o he a ious mac oeconomic a iables analyzed combined
wi h he planned C edi De aul . C edi De aul is calcula ed using all indica o s pe o med in he
53
analysis o scena ios, o his eason, CD will be he same in all g aphics p esen ed, and is be ween 25%
o 26% in he upcoming yea s. Cha ge o is o conside as a loss a deb ha will p obably no be paid.
Re u ning o he pandemic si ua ion, elabo a ed p o isions a e o an inc ease o unemploymen in
Po ugal in he upcoming yea s. The cu en ly unemploymen a e has no eached e y high le els
due o policies applied in Po ugal, such as he layo , so he e is a sho all in he impac o he
employmen c isis, wi h i s peak expec ed in 2022.
Layo is he empo a y suspension o pe manen e mina ion o employmen o an employee o a
g oup o employees o a business easons, in his case, cause by a pandemic si ua ion. The e is a
signi ican dec ease o wo k in any o ganiza ion, in some cases he e is no business.
Measu es o c edi de aul ha e been implemen ed, i leads o an ex ended paymen en , in a pe iod
o ime, o banks. The e o e, he e is no sha p d op in c edi de aul . Banks con inue o g an loans,
helping o p omo e p i a e consump ion and in es men .
4.6. TESTING THE STUDY HYPOTHESIS
The p esen disse a ion had as he mean goals o s udy he c edi de aul in Po ugal and sough o
answe he esea ch ques ion: Wha is he in luence o mac oeconomic ac o s on c edi de aul in
Po ugal, du ing he managemen o he pandemic caused by COVID-19 and he clima e change c isis?
To es he hypo heses o he s udy, a conside a ion was gi en o he analysis o linea eg ession,
which was c i ical o he accep ance o ejec ion o he hypo heses p oposed in he s udy. In o de o
op imize he model we op ed o elimina ion same explana o y a iable which i is no s a ically
signi ican . Th ough his p ocedu e, he inal models a e composed o six explana o y a iables, such
as LR, LMC, UR, GDP, CPI and CP.
Hypo heses 1: The loans a e in luence nega i ely c edi de aul a e.
The p esen analyses sugges ha 96,4% o C edi De aul a ia ion may be explained by second model
o a linea ela ionship. The esul s sugges s ha coe icien s we e ob ained by linea eg ession LR
had β=-0,215 ha means loans con ibu e nega i ely o impac on ou pu a iable, mo e speci ic a
inc ease on LR o dec ease 21,5% on CD. Pea son co ela ion is con i med wi h ρ=-0,914 and he e is
no mul icollinea i y wi h VIF bellow 10. Thus alida ing he hypo hesis 1, he nega i ely in luence o
loans on C edi De aul .
The esul s collec ed go agains ou ini ial expec a ions, we can obse e ha he whole pe iod unde
analysis he momen s in which he e was a low a e o c edi de aul , he e was a g ea e supply o
c edi h ough banking ins i u ions, and in pe iods o high de aul , like he case be ween 2012 and
2016 ha o igina ed he subp ime c isis in he USA, which was e lec ed in Po ugal yea s la e , was
pa icula ly ma ked by a con ac ion in c edi supply. This pe iod was pa icula ly ma ked by a high
a e o c edi de aul , on he o he hand, his pe iod was also ma ked by a con ol on he bank's pa
in he p o ision o loans, as can be seen in g aph s. The g an ing o c edi has ad e sa ial signs on he
de aul o c edi .
54
The ECB allowed he ins i u ions o supe ise he ope a ion empo a ily wi h a lowe le el o own unds
guidance (PILLAR 2) and ha o he combined own unds ese e, and wi h liquidi y le els below he
liquidi y co e age equi emen . In his way, i is possible o ee inancial esou ces om economic
ac i i y, hus p ese ing inancial s abili y. I should be no ed ha he ECB and he Bank o Po ugal
ha e decided o ecommend banks no o dis ibu e di idends o he yea s 2019 and 2020, in o de
o p omo e hei pe o mance in inancing he economy and he capaci y o abso b po en ial losses.
(Po ugal, Financial S abili y Repo , June 2020)
Hypo heses 2: The mo gage c edi in luence de aul c edi a e.
Simila ly wi h hypo heses 1, mo gage c edi is s a is ically signi ican on model 1 and his a iable is
included on model 2. Mo gage C edi has an e ec nega i e on C edi De aul (CD) wi h a coe icien
on model 2 [𝛽3=-0,715; =-2,122; ρ=0,038], ha is meaning o each uni inc ease in Mo gage C edi ,
he e is an dec ease o 71,5 % in C edi De aul . The s a is ical is signi ican o CD is <0,001, his sugges s
ha o he a ia ion in C edi De aul can be p edic ed 96,4% by Mo gage C edi , his sugges s ha
he a iable is s a is ically signi ican . S anda dized coe icien s e lec s ha he a iable wi h he
g ea es impac on c edi de aul was Mo gage C edi . Pea son coe icien s was con i med in he same
way, wi h 𝜌 = -0,929 and p- alue=<0,0001 and he e is no mul icollinea i y wi h VIF bellow 10.
Mo gage C edi is he a iable acco ding o he model s udied, wi h he g ea es in luence on c edi
de aul , and o his eason i is an ale a iable. The compe en en i ies ha e also his a iable as
c ucial one in he o m o isk mi iga ion, o his eason hey ha e implemen ed some p e en i e
measu es. Thus alida ing he hypo hesis 2, he nega i ely in luence o loans on C edi De aul .
The Po uguese go e nmen , like o he Eu opean coun ies, de ined Dec ee-Law no. 10-J / 2020, o
Ma ch 26, wi h a e m ha will emain in o ce un il Ma ch 31, 2021, es ablishes a se o ex ao dina y
measu es o p o ec deb o s and payees:
1. The ex ension o a pe iod equal o he e m o he mo a o ium measu e, o c edi s wi h
paymen o capi al a he end o he con ac .
2. The suspension o c edi s wi h pa ial epaymen o capi al o wi h pa ial ma u i y o o he
cash bene i s, du ing he pe iod in which he measu e o capi al paymen , en s and in e es
wi h e ec i e ma u i y is in e ec un il he end o ha pe iod.
3. The p ohibi ion on he e oca ion o c edi lines con ac ed and loans g an ed in he amoun s
con ac ed a he da e o en y in o o ce o he Dec ee-Law.
Mo a o iums ep esen a undamen al measu e, which aims o econcile he con inui y o business
ac i i y a e he immedia e impac o he heal h c isis, a oiding cash low p oblems om u ning in o
insol encies, and on he o he hand, he sa egua ding o he economy's inancing capaci y by he
banking sys em, minimizing i s consump ion o capi al. (Po ugal, Financial S abili y Repo , June 2020)
The ques ion ha a ises is wi h he end o hese a ea s, he de aul o c edi can inc ease quickly, and
o ha eason, he possible impac was s udied in he s ess es . Th ough he a ec ed s udy, he
a iable ha can mi iga e hese isks is o ac on he unemploymen a iable, in o de o pa icula s o
con inue ul illing hei obliga ions.
55
Hypo heses 3: The consume c edi in luence c edi de aul a e.
Rega ding hypo hesis 3, he esul s o he s udy sugges ha he co ela ion be ween C edi De aul
and he LCC is no s a is ically signi ican . No alida ing hypo hesis 3, he in luence o consume
c edi on c edi de aul . Al hough, his esul is no consis en wi h p e ious s udies ha e e ed o
he exis ence o ela ionship be ween a iables, (Nunes, 2009 - quo ed by (Veloso, 2016), is, howe e
consis en wi h s udies ha do no con i m he exis ence o a di ec ela ionship.
Hypo heses 4: The business c edi in luence de aul c edi a e.
Rega ding hypo hesis 4, in he same way, he esul s o he s udy sugges ha he co ela ion
be ween C edi De aul and he LBC is no s a is ically signi ican . No alida ing hypo hesis 4, he
in luence o consume c edi on c edi de aul . Al hough, his esul is no consis en wi h p e ious
s udies ha e e ed o he exis ence o ela ionship be ween a iables, (Nunes, 2009 - quo ed by
(Veloso, 2016), is, howe e consis en wi h s udies ha do no con i m he exis ence o a di ec
ela ionship.
Hypo heses 5: The unemploymen a e in luence posi i ely c edi de aul a e.
The esul s ob ained om he p esen s udy he hypo heses 5 is con i ming he ini ial expec a ions o
he posi i e signals be ween a iables unemploymen and c edi de aul . Linea eg ession model
β=0,452 p esen s a g ea weigh in he ou pu a iable, c edi de aul . In he same way S anda dized
Be a on model 2, wi h all a iables, con i med ha unemploymen a e had a g ea in luence on he
c edi de aul wi h S anda dized β=0,697. The es Pea son co ela ion ca ied ou e eals a s ong
co ela ion o 𝜌 = 0,797 wi h p- alue=0,0001, his means ha he inc ease in he unemploymen a e
is ela ed o c edi de aul . This leads o an u gen need o mi iga e he possible isks o high
unemploymen and o con ol high le el o c edi de aul is o ake p e en i e measu es o con ol he
unemploymen a e.
P esen li e a u e men ion he unemploymen a e is ano he wa ning signal a iable o he inancial
si ua ion o a coun y. When unemploymen a e inc eases ansla es in o a dec ease in expec ed
ea nings, as he e a e ewe people con ibu ing o a coun y's weal h. Consequen ly, he e will be less
agg ega e demand, esul ing in a c edi de aul o indi iduals who ha e hei condi ional income and
o companies. The expec ed sign be ween he c edi de aul a iables and unemploymen a e a e
posi i e. Wi h he de e io a ion o economic and inancial condi ions, inancial agen s condi ion o
access o c edi .
The Po uguese go e nmen , like o he go e nmen s, has p omo ed layo schemes h ough, which i
seeks o main ain employmen links and con ibu ing o a p og essi e esump ion o economic ac i i y.
In his way i is possible o ealize ha his lay-o scheme may ha e been a good, adop ed measu e
o he p og essi e e u n o economic ac i i y.
Hypo heses 6: The 3-mon h Eu ibo in e es a e in luence c edi de aul a e.
Rega ding hypo hesis 6, On model 1 loans β=0,201 wi h p- alue=0,247, i is meaning ha i is no
s a is ically signi ican o le el 1%. Pea son co ela ion ein o ces he same idea, wi h a low posi i e
56
co ela ion be ween TBA-Eu ibo 3 mon hs and De aul C edi . The esul s o he s udy sugges ha
he co ela ion be ween C edi De aul and he EUR is no s a is ically signi ican . No alida ing
hypo hesis 6, he in luence o 3-mon h Eu ibo in e es a e on c edi de aul . Al hough, his esul is
no consis en wi h p e ious s udies ha e e ed o he exis ence o ela ionship be ween a iables,
Ci ed by Cha es, is, howe e consis en wi h s udies ha do no con i m he exis ence o a di ec
ela ionship.
The in e es a e has an essen ial ole, mone a y au ho i ies can manipula e he isk o s abiliza ion
pe o mance. The in e es a e is inc eased, he g ea e he isk o de aul , as bo owe s will ha e o
pay mo e o he same loan, esul ing in a disposable income dec ease.
The coun ies ha adhe ed o Eu opean cu ency, known as he eu o a ea, as in he case o Po ugal,
he issuance o bankno es is he esponsibili y o he Eu opean Cen al Bank, which is esponsible o
mone a y policy, in pa icula he con ol o he g ow h o money supply. The Cen al Bank inc eases
o dec eases he in e es a e as a measu e o con ol o he money supply, when in e es a e
inc ease, i is meaning c edi is mo e expensi e, consequen ly c edi o s ind i mo e di icul o mee
hei obliga ions. On he o he hand, when in e es a e dec eases economic agen s look o o he
al e na i es o he applica ion o hei deposi s. Fo his eason, he ECB con ols he in e es a e
acco ding o he o he economic a iables o esea ch a balance in economy.
The esul s ob ained go agains ini ial expec a ions, e lec ing an opposi e e ec on he ela ionship
be ween TBA-Eu ibo and c edi de aul . This opposi e e ec is explained by he p e en i e measu es
applied by he Eu opean Cen al Bank, since he way o mi iga e he isks o de aul is h ough
measu es o educe he in e es a e.
Hypo heses 7: Nominal GDP g ow h a e in luence nega i ely c edi de aul a e.
The esul s sugges s ha nominal GDP has lagged wi h ρ<0,0001, meaning ha hese a iables ejec ed
null hypo hesis and accep hypo hesis 𝐻1 se ies is s a iona i y, di e en lag is used on linea eg ession.
The esul s sugges s ha coe icien s we e ob ained by linea eg ession GDP had β=-0,149 ha means
GDP con ibu e nega i ely o impac on ou pu a iable. Pea son co ela ion is con i med wi h ρ=-
0,053, indica ing a weak co ela ion and he e is no mul icollinea i y wi h VIF bellow 10. The p esen
analyses sugges ha 96,4% o C edi De aul a ia ion may be explained by second model o a linea
ela ionship. On model 2, goes agains ini ial expec a ions wi h con adic o y signals. Rein o cing he
idea con eyed in he sub-chap e speci ica ion o he model a iable, his can be explained by he
e ec o c edi de aul on la e yea s, in which he e was a g ea e de aul o c edi , be ween 2012 and
2016, p og ess in economic GDP s a ed o make p og ess. Since, acco ding o he his o ical da a
analyzed, he e ec s o a shock can ake abou 2 yea s o be e lec ed. The GDP had a long pe iod o
ime wi h nega i e alues, be ween 2008 (con agious e ec o he subp ime c isis) and 2013. The
e ec on he c edi de aul a iable only wo sened be ween 2012 and 2016. This may explain he
posi i e signs be ween he a iables, despi e ha , GDP has li le impac on he ou pu a iable due o
he ime lag ha he impac on GDP has on he a iable dependen on c edi de aul .
The dec ease in GDP is due o he pandemic si ua ion COVID-19, cha ac e ized by he in e ac ion o
supply and demand shocks, oge he wi h he high deg ee o unce ain y abou how he economic
si ua ion will e ol e, accompanied by a all in p oduc i i y, due o he closu e pa ial o o al o
companies. A he same ime, consump ion and in es men le els we e also a ec ed, wi h nega i e
consequences on job sea ch.
57
Con ex ualizing he economic heo y ha ansla es in o a phase o expansion, he e is a dec ease
c edi de aul . This expansion phase is cha ac e ized by a endency o economic agen s o bo ow,
because he e is mo e disposable income a p esen . Bu his expansiona y phase usually b ings a
consequen ecession phase, called economic cycles. Second phase o ecession is e i ied by a end
owa ds a c edi de aul inc ease. Fo his eason, he heo y explains ha he e is a nega i e
associa ion be ween c edi isk and he GDP g ow h a e.
In he coming yea s, GDP is expec ed o eco e due o he e olu ion o global p oduc i i y, leading o
an inc ease in expo s. This scena io is based on he non-adop ion o con ainmen measu es.
Thus alida ing he hypo hesis 7, he nega i ely in luence o GDP on C edi De aul .
Hypo heses 8: The P ice S ock Index o he bes wen y companies in Lisbon in luence c edi de aul
a e.
Rega ding hypo hesis 8, he esul s o he s udy sugges ha he co ela ion be ween C edi De aul
and he PSI is no s a is ically signi ican . No alida ing hypo hesis 8, he in luence o P ice S ock Index
on c edi de aul . The Since he coe icien ob ained h ough linea eg ession was nega i e, wi h
β=0,013 o model 1 and ρ=0,047, i means ha PSI and CD had a weak co ela ion. Al hough he e is
a weak co ela ion be ween PSI 20 and ou pu a iable, he s ock p ice index allows o know he
in luence o he inancial condi ions o companies lis ed on he Lisbon s ock exchange on c edi isk.
An inc ease in he s ock ma ke index can ansla e in o good inancial heal h o lis ed companies,
which can con ibu e o educing he isk o c edi de aul . The heo e ical poin o iew is no he
expec ed ela ionship o he a iables, o a posi i e sign be ween c edi de aul and PSI20.
Hypo heses 9: The nominal Disposable Income g ow h in luence nega i ely c edi de aul a e.
Rega ding hypo hesis 9, he esul s o he s udy sugges ha he co ela ion be ween C edi De aul
and he DIG is no s a is ically signi ican . The es o he co ela ion be ween he a iables con i ms
ha he e is a nega i e weak co ela ion be ween he wo ac o s o 𝜌 = -0,090, meaning when inc ease
disposable income will lead o a dec ease in C edi De aul . DIG had a β=0,019. The e o e, when he e
a e le els o sa ings on he pa o indi iduals and companies, in si ua ions wi hou liquidi y agen s can
use sa ings o ul il obliga ions, o his eason, an inc ease in he inc ease in DIG has a dec easing
impac on c edi de aul .
The disposable income o amilies is also expec ed o all, because o he go e nmen measu es
applied, speci ically he lay-o sys em in which he wo ke o e s an income o 100%, as well as, o he
sel -employed wo ke s, in o de o main ain nominal disposable income le el o amilies. O he cases
a e COVID pa ien s, who, in con as o indi iduals in p ophylac ic isola ion who o e hei income in
he o de o 100%, COVID pa ien s ha e hei sala y in he o de o 55%, in he i s 30 days o he
disease (Social, 2020). Consequen ly, he e is a d op in pu chasing powe . Likewise, he e is no
p o ision o any go e nmen policy o deal wi h he loss o income.
Disposable income is limi ed by bo owing, since pu chasing powe is los o mee commi men s.
Disposable income exp esses he alue a ailable, his means ha a high alue ansla es in o g ea e
sa ings o consume s. Lowe le els o sa ings mean g ea e isks o de aul .
58
Hypo heses 10: The Consume P ice Index in luence posi i ely c edi de aul a e.
The p esen analyses sugges ha 96,4% o C edi De aul a ia ion may be explained by second model
o a linea ela ionship. The esul s sugges s ha coe icien s we e ob ained by linea eg ession CPI
had β=-0,762 ha means CPI con ibu e nega i ely o impac on ou pu a iable, mo e speci ic a
dec ease on CPI o inc ease 76,2% on CD. Pea son co ela ion is con i med wi h ρ=-0,395 and he e is
no mul icollinea i y wi h VIF bellow 10. Thus alida ing he hypo hesis 10, he nega i ely in luence o
CPI on C edi De aul .
Rega ding he in la iona y impac on c edi isk, om a heo e ical poin o iew, in la ion occu s when
he e is a pe sis en inc ease in he p ices o goods and se ices wi hou a co esponding inc ease in
he le el o p oduc ion. When he e is an inc ease in in la ion in he economy, mone a y egula o s
accompany he inc ease in in e es a es, as a con ac iona y measu e o con ol in la ion. This
measu e, as a gene al ule, leads o an inc ease in he cos o bo owing. (Yahaya & Oni, 2016) Despi e
his, Cas o ci ed by (Rod igues, 2013) says ha a highe in la ion a e can make deb se ice easie by
educing he eal alue o loans in o ce. The signal be ween he CPI and he c edi de aul isk is
inde e mina e om a heo e ical poin o iew, in his pa icula case he signal is nega i e.
Hypo heses 11: The exchange a e in luence c edi de aul a e.
Rega ding hypo hesis 11, he esul s o he s udy sugges ha he co ela ion be ween C edi De aul
and he ER is no s a is ically signi ican . No alida ing hypo hesis 11, he in luence o exchange a e
on c edi de aul . Al hough, his esul is no consis en wi h p e ious s udies ha e e ed o he
exis ence o ela ionship be ween a iables, he e ec i e eal exchange a e is he measu e o he
Po uguese economy's compe i i eness han a ou side wo ld. Theo e ically, a dec ease in he
exchange a e ep esen s a dep ecia ion o he local cu ency, ansla ing in o mo e compe i i e goods
and se ices p oduced in Po ugal, consequen ly he e is g ea e income o ul il obliga ions.
Al hough, on model 1 i can be e i ied wi h heo ical li e a u e, he inc ease o 1,4% on exchange a e
dec ease on CD, wi h β=-0,014. Despi e o hypo hesis being e i ied; i is concluded i is no s a is ically
signi ican o c edi de aul .
Hypo heses 12: The e is a ela ionship be ween c edi and mac oeconomics a iables.
Rega ding o hypo heses 12, he esul s o model 2 sugges an imp o emen in he quali y i same
measu es is o applied wi h de aul ing cus ome s, such as LR, LMC, UR, GDP, CPI and CP.
Hypo heses 13: The ca bon p ice index in luence c edi de aul a e.
Rega ding hypo hesis 13, he esul s o he s udy sugges ha he co ela ion be ween C edi De aul
and he CP is no s a is ically signi ican . No alida ing hypo hesis 13, he in luence o ca bon p ice on
c edi de aul . Al hough, his esul is no consis en wi h p e ious s udies ha e e ed o he
exis ence o ela ionship be ween a iables, he clima e change is e y impo an (ECB, 2021). Fo his
eason, CP was es ima ed o model 2. The esul s sugges ha inc ease on Ca bon P ice, will be a
inc ease on 1,7% C edi De aul .
59
In iew o he hypo heses p esen ed, i is necessa y o analyze he main objec i e o he
in es iga ion. Check h ough he assessmen and managemen model o c edi isk, Wha is he
in luence o mac oeconomic ac o s on c edi de aul in Po ugal, du ing he managemen o he
pandemic caused by COVID-19 and he clima e change c isis?
A e he s udy ca ied ou , i was ime o e lec ion and discussion, in which he ocus o he wo k
de eloped is ela ed o he economic impac on inancial sec o , speci ically in c edi de aul , bu
which s a ed in heal h, wi h he con inemen measu es o con ol he ansmission o he i us.
The e ha e been o he economic luc ua ions in his o y, bu he cu en c isis equi es new
ins umen s and ac ions, since in Po ugal we had wo con inemen s. The companies had o eadjus ,
and hose ha ailed o do so ha e high le els o indeb edness, consequen ly, he insol ency o he
companies. In companies he e is a lo o human capi al ha has no been ealloca ed, some
ine i ably wen o he layo , o he s we e en olled in he unemploymen und. S ill, he e a e sec o s
such as ou ism and ai anspo ha ha e g ea e di icul y in eadjus ing business, bu he e a e
c ea i e solu ions o he sec o s in ques ion. Fo example, eadjus ing ooms o ecei e
elecommu ing employees a mo e a o dable p ices, ensu ing daily meals, ei he o gues s o wi h
wo ke s om a ce ain company, making ooms a ailable o people on he on lines, examples o
ill he weak demand.
As i was e i ied, he measu es o he i s con inemen led o a d op in GDP o a ound 16% in he
qua e co esponding o he i s con inemen . a e y se ious eason o ale ing he inancial sec o .
The e a e s udies ha indica e ha he wo ld economic c isis is mo e se ious han he g ea
dep ession o 1930. The measu es o social con inemen , he es ic ions o ci cula ion and he loss
o wo king hou s a e limi ing he global p oduc ion, he chains global ade and se ices sec o .
The Eu opean Cen al Bank app o ed a p og am called he Pandemic Eme gency Pu chase P og am
ha was launched o comba he economic e ec s o he pandemic, such as he case o loan de aul .
The cu en economic c isis was ma ked by he all in agg ega e demand, bu which has an impac on
he supply side. Consequen ly, i d agged he co po a e ab ic, leading o signi ican billing losses,
inc easing liquidi y p oblems. These losses on he supply side a e e lec ed in unemploymen , loss o
income, e ac ion in demand, bank up cies and closu es, inancial di icul ies and he sha p ise in
unce ain y and isk.
The go e nmen 's esponse was, in he i s ins ance, o comba he heal h eme gency, hen by he
deploymen o social suppo aimed a mi iga ing he i s ele an impac s on income and
employmen and, om hen on, by he mobiliza ion o esou ces aimed a sus aining ma ke s,
companies and he economic in e en ion o he S a e i sel . This new con ex equi es a lo o
esilience om he business ab ic, as well as adap ing o his new way o being, in which ace- o- ace
wo k gi es way o elewo king. Realloca ing human capi al in ol es p o iding mo e se ices,
p o iding aining, s eng hening he business o digi al comme ce and, inally, eaching he inal
consume h ough home deli e ies.
One way o p e en c edi de aul is o ensu e liquidi y on he pa o companies and jobs, on he
pa o indi iduals, as s udied, is one o he a iables wi h he g ea es impac acco ding o ou
model. The ac ha he e is a accine agains he new co ona i us has inc eased con idence le els,
bu since he u u e is no known, i his will become he new no mal o e e yday li e, he e is an
eme gency o adap a ion. I his adjus men is no made, we will be acing a e y se ious c isis, wi h
e ec s on c edi de aul s ne e el be o e.
60
61
5. CONCLUSIONS
In his disse a ion, a s udy was execu e o he possible ela ionship be ween c edi isk and
a iables mac oeconomics in Po ugal esul ing om he pandemic. Risk is pa o he g an ing o
c edi , bu his should be minimized as much as possible, so ha lending ins i u ions ake ca e.
Howe e , i is necessa y o ake in o accoun c edi policies and mac oeconomic a iables. The
s a is ical in e ence based on linea eg ession model, highligh ed a se ies o esul s ha allow us o
d aw a conclusion abou he e olu ion o c edi isks in ela ion o he o he mac oeconomic
a iables in he Po uguese economy, whe e in he pe iod om 2003Q1 o 2021Q3, wo eg essions
we e es ima ed o he dependen a iable unde s udy.
Wi h he p esen esea ch wo k i was possible o answe he scien i ic p oblem ini ially men ioned,
he gene al and speci ic objec i es, as well as he hypo hesis. E alua ing each speci ic objec i e and
he p oposed main one:
The main objec i e p oposed was o iden i y he mac oeconomic a iables ha a ec c edi de aul .
The objec i e was ul illed. A he mac oeconomic le el, he main de e minan s o economic ac i i y
measu es s and ou : loans, mo gage c edi , unemploymen a e, GDP, consume p ice index and
ca bon p ice we e he a iables used in he analysis o he scena io, as hey a e he a iables wi h he
g ea es impac on he dependen a iable and s a is ically signi ican .
The p esen wo k had as i s i s speci ic objec i e o s udy he exis ing li e a u e ega ding c edi
isk, s ess es ing, iden i ica ion o he main mac oeconomic a iables ha a ec he economy. In
his way, we we e able o p opose a me hodology o building a s ess es in he Po uguese
economy, wi h a g ea e ocus on p e en ing c edi de aul . This was he main ocus, since all he
p e ious inancial c ises s a ed wi h c edi de aul and had a con agious e ec on he es o he
economy. Likewise, he s udied li e a u e con ibu ed o he cons uc ion o he s ess es , so ha
he me hodology would ake in o accoun he Basel ag eemen s in he connec ion be ween
mac oeconomic and c edi isk ac o s.
The e is an inc easing need o he c ea ion o isk con ol and mi iga ion ins umen s in an economy,
in o de o unde s and, p edic , manage and measu e isks in i . As we a e acing a si ua ion ne e
be o e expe ienced, COVID-19, hese ins umen s a e aluable in he pe o mance o co ec i e
measu es in u u e e en s.
The second speci ic objec i e was o c ea e a model ha de ec s he de aul a e. The objec i e was
achie ed h ough he explo a o y desc ip i e esea ch model wi h he quan i a i e app oach as a
esea ch s a egy, i was possible o collec mac oeconomic da a and om he desc ip i e s a is ics,
namely he analysis o mul iple linea eg ession, i was possible o achie e his objec i e.
The cu en si ua ion did no s a wi h c edi de aul , bu h ough he analysis o he a ious
mac oeconomic a iables, he objec i e is o p e en his si ua ion om happening. Thus, he s udy
made i possible o conduc a policy analysis on he e ec o changes in mac oeconomic isk ac o s.
The hi d objec i e was o es he model and check i s a iables. I was concluded ha he model
has a iables acco ding o he s udied li e a u e, such as he loans, mo gage c edi , unemploymen
a e, GDP, consume p ice index and ca bon p ice a e ela ed o c edi de aul . The emaining
mac oeconomic a iables s udied canno be concluded.
68
SUMMARY OUTPUT LCC
Reg ession S a is ics
Mul iple R 0,53479
R Squa e 0,286
Adjus ed R Squa e 0,2656
S anda d E o 0,013782
Obse a ions 73
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,005326 0,002663 14,01964 7,58E-06
Residual 70 0,013296 0,00019
To al 72 0,018622
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,000495 0,001803 0,274334 0,784636 -0,0031 0,004091 -0,0031 0,004091
Lagged Consume C edi -0,04639 0,024733 -1,87582 0,06485 -0,09572 0,002934 -0,09572 0,002934
Di e en Lagged Consume C edi 0,473545 0,10124 4,677461 1,37E-05 0,271629 0,675462 0,271629 0,675462
SUMMARY OUTPUT LBC
Reg ession S a is ics
Mul iple R 0,62883
R Squa e 0,395427
Adjus ed R Squa e 0,369141
S anda d E o 0,005891
Obse a ions 49
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,001044 0,000522 15,04336 9,4E-06
Residual 46 0,001596 3,47E-05
To al 48 0,00264
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep -0,00064 0,000845 -0,75455 0,454362 -0,00234 0,001064 -0,00234 0,001064
Lagged Business C edi -0,15291 0,036634 -4,17406 0,000132 -0,22665 -0,07917 -0,22665 -0,07917
Di e en Laged Business C edi 0,41214 0,123997 3,323797 0,001749 0,162548 0,661733 0,162548 0,661733
69
Reg ession S a is ics UR
Mul iple R 0,659047
R Squa e 0,434342
Adjus ed R Squa e 0,418181
S anda d E o 0,005243
Obse a ions 73
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,001478 0,000739 26,87488 2,18E-09
Residual 70 0,001924 2,75E-05
To al 72 0,003402
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,002298 0,001961 1,172056 0,245148 -0,00161 0,006209 -0,00161 0,006209
Lagged Unemploymen -0,02442 0,019249 -1,26842 0,208849 -0,06281 0,013975 -0,06281 0,013975
Di e en e Lagged Unemploymen 0,465432 0,064718 7,191682 5,58E-10 0,336356 0,594508 0,336356 0,594508
SUMMARY OUTPUT GDP
Reg ession S a is ics
Mul iple R 0,872981
R Squa e 0,762096
Adjus ed R Squa e 0,755201
S anda d E o 0,017412
Obse a ions 72
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,067013 0,033506 110,5167 3,06E-22
Residual 69 0,020919 0,000303
To al 71 0,087932
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,001685 0,002053 0,820877 0,414543 -0,00241 0,00578 -0,00241 0,00578
Lagged GDP -0,37325 0,055931 -6,67337 5,17E-09 -0,48483 -0,26167 -0,48483 -0,26167
Di e en e Lagged GDP 0,62154 0,044767 13,88397 1,17E-21 0,532232 0,710847 0,532232 0,710847
70
SUMMARY OUTPUT DIG
Reg ession S a is ics
Mul iple R 0,980487
R Squa e 0,961354
Adjus ed R Squa e 0,960234
S anda d E o 0,01801
Obse a ions 72
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,55673 0,278365 858,2272 1,8E-49
Residual 69 0,02238 0,000324
To al 71 0,57911
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,045443 0,005533 8,212873 8,16E-12 0,034404 0,056481 0,034404 0,056481
Lagged Nominal Disposible Income -0,53214 0,059734 -8,90852 4,38E-13 -0,65131 -0,41298 -0,65131 -0,41298
Di e en e Lagged Nominal Disposible Income
0,619524 0,018397 33,67536 1,46E-44 0,582823 0,656225 0,582823 0,656225
SUMMARY OUTPUT CPI
Reg ession S a is ics
Mul iple R 0,674985
R Squa e 0,455605
Adjus ed R Squa e 0,440051
S anda d E o 0,004399
Obse a ions 73
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,001134 0,000567 29,29153 5,71E-10
Residual 70 0,001355 1,94E-05
To al 72 0,002488
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,001681 0,000761 2,209063 0,030445 0,000163 0,003199 0,000163 0,003199
Lagged Consume P ice Index -0,13774 0,037863 -3,63781 0,000522 -0,21325 -0,06222 -0,21325 -0,06222
Di e en e Lagged Consume P ice Index 0,477734 0,075664 6,313867 2,17E-08 0,326826 0,628641 0,326826 0,628641
71
Reg ession S a is ics PSI
Mul iple R 0,942222
R Squa e 0,887782
Adjus ed R Squa e 0,884432
S anda d E o 0,050868
Obse a ions 70
ANOVA
d SS MS F
Signi icance F
Reg ession 2 1,371541 0,68577 265,0257 1,5E-32
Residual 67 0,173367 0,002588
To al 69 1,544907
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep -0,00501 0,006112 -0,81938 0,415477 -0,01721 0,007191 -0,01721 0,007191
Lagged PSI20 -0,55533 0,061111 -9,08711 2,67E-13 -0,6773 -0,43335 -0,6773 -0,43335
Di e en Lagged PSI20 0,594371 0,025817 23,0228 1,55E-33 0,542841 0,645901 0,542841 0,645901
SUMMARY OUTPUT LR
Reg ession S a is ics
Mul iple R 0,505989
R Squa e 0,256024
Adjus ed R Squa e 0,234768
S anda d E o 0,008214
Obse a ions 73
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,001625 0,000813 12,04455 3,2E-05
Residual 70 0,004723 6,75E-05
To al 72 0,006348
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,000344 0,001059 0,324959 0,746181 -0,00177 0,002456 -0,00177 0,002456
Lagged Loans -0,04996 0,022095 -2,26097 0,026871 -0,09402 -0,00589 -0,09402 -0,00589
Di e en Lagged Loans 0,464117 0,113512 4,088686 0,000114 0,237723 0,69051 0,237723 0,69051
72
SUMMARY OUTPUT ER
Reg ession S a is ics
Mul iple R 0,651539
R Squa e 0,424504
Adjus ed R Squa e 0,408061
S anda d E o 0,039336
Obse a ions 73
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,079894 0,039947 25,81707 3,99E-09
Residual 70 0,108312 0,001547
To al 72 0,188207
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,033173 0,010756 3,084165 0,002921 0,011721 0,054625 0,011721 0,054625
Lagged Exchange Ra e -0,12639 0,038237 -3,30549 0,001498 -0,20266 -0,05013 -0,20266 -0,05013
Di e en Lagged Exchange Ra e 0,470453 0,077809 6,046243 6,49E-08 0,315268 0,625639 0,315268 0,625639
SUMMARY OUTPUT TBA
Reg ession S a is ics
Mul iple R 0,490641
R Squa e 0,240729
Adjus ed R Squa e 0,219036
S anda d E o 0,002909
Obse a ions 73
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,000188 9,39E-05 11,09685 6,51E-05
Residual 70 0,000592 8,46E-06
To al 72 0,00078
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep -5,3E-05 0,000425 -0,12371 0,901902 -0,0009 0,000795 -0,0009 0,000795
Lagged TBA - 3 mon hs Eu ibo -0,03049 0,021996 -1,38634 0,170044 -0,07436 0,013376 -0,07436 0,013376
Di e en Lagged TBA - 3 mon hs Eu ibo 0,477281 0,110621 4,314559 5,15E-05 0,256654 0,697908 0,256654 0,697908
73
8.2. APPENDIX B – DESCRIPTIVE STATISTICS
8.3. APPENDIX C – MODEL SUMMARY
8.4. APPENDIX D – ANOVA
SUMMARY OUTPUT CP
Reg ession S a is ics
Mul iple R 0,966597
R Squa e 0,93431
Adjus ed R Squa e 0,932378
S anda d E o 0,026918
Obse a ions 71
ANOVA
d SS MS F
Signi icance F
Reg ession 2 0,700789 0,350395 483,5819 6,24E-41
Residual 68 0,049272 0,000725
To al 70 0,750061
Coe icien s
S anda d E o
S a P- alue
Lowe 95%
Uppe 95%
Lowe 95,0%
Uppe 95,0%
In e cep 0,007775 0,003249 2,392771 0,019489 0,001291 0,014259 0,001291 0,014259
Lagged Ca bon P ice -0,70389 0,055409 -12,7035 1,27E-19 -0,81446 -0,59333 -0,81446 -0,59333
Di e en Lagged Ca bon P ice 0,625814 0,020207 30,97026 8,09E-42 0,585491 0,666136 0,585491 0,666136
74
8.5. APPENDIX E – COEFFICIENTS
8.6. APPENDIX F -CORRELATIONS
75
8.7. APPENDIX G – DESCRIPTIVE STATISTICS (MODEL 2)
8.8. APPENDIX H – MODEL SUMMARY (MODEL 2)
8.9. APPENDIX I – ANOVA (MODEL 2)
8.10. APPENDIX J – COEFFICIENTS (MODEL 2)
76
8.11. APPENDIX L – CORRELATIONS (MODEL 2)
Page | i