The impact of macroeconomic variables on the evolution of the credit risk rate

Transcription

1 Vol. 17, No. 1, pp , 2018 DOI: The impact of macroecoomic variables o the evolutio of the credit risk rate Lumiița Gabriela Istrate a1 ad Bogda Ștefa Ioescu a a Bucharest Uiversity of Ecoomic Studies, Romaia Abstract:The dyamics of the real ecoomy is a major driver of the evolutio of arrears at the level of the pool of loas grated to o-fiacial compaies, completed by the fiacial pressure iduced by the moetary coditios. Ledig allows o the oe had providig resources for compaies that eed fiacig for ivestmet projects, o the other had, it supports the fud holders to place resources for obtaiig profit. The role of the ledig policy i the activity of commercial baks is very importat, as it may ifluece both the cost of credits ad the loa portfolio quality i the future. The purpose of this research is to fid the macroecoomic variables that sigificatly ifluece credit risk ad to develop a statistical model for predictig the doubtful ad o-performig loas rate. Thus, it is evisaged the research of mechaisms by which the dyamics of the real ecoomy ad the moey market coditios ifluece the evolutio of the credit risk i differet busiess sectors. Keywords:credit risk, itermediate cosumptio, gross value added, foreig exchage rate, iterest rate JEL codes: E44, G21, O1 1 Correspodig author: Lumiiţa Gabriela Istrate, Departmet of Accoutig ad Maagemet Iformatio Systems, Bucharest Uiversity of Ecoomic Studies, 6, Romaă Square, Bucharest, Romaia, Telephoe: (+40) , istrate_lumiita@yahoo.com

2 The impact of macroecoomic variables o the evolutio of the credit risk rate 1. Itroductio The sesitivity of o-performig (bad) loas depedet of the evolutio of macroecoomic idicators varies from oe sector to aother, the ecoomic sectors the most exposed to this process beig those for which the bakig system is fiacig a importat part of the workig capital, but also the demad for products / services provided. Accordig to Derelioğlu ad Gurge (2011) ad Boss et al. (2009) the credit risk aalysis aims to reduce future losses by estimatig the potetial risk ad elimiatig the ewly proposed credit if the risk is greater tha a defied tolerace value. I this respect, it is essetial to idetify the mai factors determiig this risk i order to maage it effectively. The ledig activity is ot without risks ad ca geerate o-performig loas, resultig from risk mismaagemet. No-performig loas occur whe there are delays i the reimbursemet ad recovery of arrears, ad the paymet of assumed liabilities is ot secured or is partially secured. Europea supervisory authorities cosider that a loa is o-performig whe it has bee more tha 90 days without the debtor payig the agreed istalmets. The level of o-performig loas should be at a level as low as possible due to the impact o bak profitability. This goal ca be achieved oly by a coservative ledig policy practiced i lie with the ecoomic coditios ad the fiacial eviromet ad through a appropriate risk classificatio. Of course, the origi of o-performig (bad) loas should ot be sought exclusively i macroecoomic variables. The policies adopted by idividual baks, particularly those relatig to the icrease i efficiecy ad the risk maagemet, sigificatly ifluece the evolutio of o-performig loas. The study aims to outlie a framework for highlightig the macroecoomic iflueces o the credit risk. The paper examies the macroecoomic coditios that have a sigificat impact o credit risk, proposig to familiarize the iteratioal public with the uderstadig of the aspects proposed i the aalysis for the case of Romaia. Thus, the paper aalyzes the aggregate rate of the credit risk ad the macroecoomic data of Romaia for the period uder review. As the study shows, the most sigificat ifluece o credit risk is the exchage rate, but the iterest rate ad the gross domestic product also have a sigificat impact. It should be stressed that a icrease i the iterest rate could be determied by a decisio take by the Natioal Bak of Romaia to icrease the moetary policy iterest rate i order to moderate the iflatioary pressure. Vol. 17, No

3 Takig ito accout the results of the statistical tests, the macroecoomic variables uderlyig the aalysis play a importat role i the fiacial stability of the moetary ad bakig system, the severe deterioratio of oe of them ifluecig the credit risk. The paper proposes both a statistical ad a qualitative study. The expected outcome of the qualitative aalysis is to show the ifluece of the credit risk o the capital supportig a bak s activity. A loss from a bad credit is idirectly traslated ito a capital deficit through provisioig. Uder these circumstaces, capital icreases have as their primary motivatio the icrease i credit risk due to o-performig loas ad the recogitio of loss. The study may be a startig poit for research i the field of accoutig sice the credit risk rate has a idirect ifluece o the available capital, which implies that a icrease i the level of o-performig loas leads to a higher level of credit risk specific provisios, thus reflectig a balace sheet expese ad a loss i the profit ad loss accout ad available capital. Thus, research results ca be used to model the dyamics of some elemets of the balace sheet of commercial compaies through macroecoomic variables. The research results have both practical applicability, as a basis for stress tests ad for forecastig purposes, ad also implicatios i ecoomic policies, outliig the effects of credit risk through the macroecoomic eviromet. Further to the study i this article, a related aalysis will look at idetifyig a causal relatioship betwee macroecoomic variables ad the capital of fiacial istitutios. The study is structured o several sectios, edig with the mai coclusios ad future directios of aalysis. After the itroductory cocepts, the paper presets, i the secod sectio, the review of the specialized literature, ad i sectio o. 3 describes the methodology of the research ad presets the used data, detailig both the ecoomic reasoig ad the statistical records which led to the selectio of explaatory variables. I the fourth sectio we sythesized the mai aspects of the empirical research, this sectio dealig with data aalysis ad the presetatio of the results obtaied. The paper eds with the presetatio of the coclusios ad of the future directios of research withi sectio o Literature review Most empirical studies examie the ifluece of macroecoomic idicators o the o-performig loas. Mcwayizei (2016) studied the impact of macroecoomic variables o o-performig loas, cocludig that variables such as GDP, iterest rate, exchage rate, public debt, sovereig debt, uemploymet, the volume of imports, the volume of exports, have a sigificat ifluece o the dyamics of o-performig loas i emergig ecoomies, but do ot have a otable ifluece i the case of developed ecoomies. 124 Vol. 17, No. 1

4 The impact of macroecoomic variables o the evolutio of the credit risk rate Gagais et al.(2010) show that there are four mai criteria of bakig stability: regulatios, other attributes of the bakig ad fiacial sector, the istitutioal framework ad the macroecoomic coditios. Thus, the GDP growth ot oly reduces the rate of o-performig loas, but it may delay the emergece of bakig crises. Figlewski et al.(2012) state that the macroecoomic factors have a sigificat impact o the solvecy of compaies. These authors group the macroecoomic risk factors ito three large classes: 1. Factors related to geeral macroecoomic coditios (uemploymet rate, iflatio, etc.) 2. Factors related to the directio i which the ecoomy is evolvig (a real GDP growth, chages i the cosumer behavior, etc.) 3. Factors relatig to coditios i the fiacial markets (iterest rates, listig / delistig, etc.). As far as the category of macroecoomic idicators with the greatest impact o credit risk is cocered, the views are divided. There are studies that claim that the dyamics of the real ecoomy has the greatest impact o credit risk, while other authors have come to the coclusio that moetary factors have the greatest impact o the likelihood of o-reimbursemet. Thus, Makri ad Papadatos (2016) suggest that the macroecoomic eviromet (public debt, uemploymet, iflatio) appears to be closely liked to the repaymet of loas i arrears, ad Makri et al. (2014) cocluded that the macroecoomic variables have a sigificat ifluece o o-performig loas i the Euro area. Followig the same directio, Makri (2015) argues that i the aalysis carried out i Greece, the probability of o-reimbursemet at maturity of loas is iflueced by uemploymet, public debt, GDP, iflatio, outstadig loa, capital adequacy, liquidity ad profitability. Skala (2014) studied the factors ifluecig the credit policy of 356 commercial baks i Polad i the period from 2006 to 2012, highlightig the fact that the macroecoomic idicators have a strog impact o the o-performig credits. If a bak ca establish a lik betwee the macroecoomic eviromet ad systematic credit risk factors, this kowledge may help i assessig ad maagig the portfolio credit risk over time ad may prove useful i dyamic credit risk maagemet circumstaces i which default scearios ca occur over a variety of ecoomic coditios. Alessi et al. (2014) aalyzed the settig up of provisios for loa impairmet i the case of Italia commercial baks for the period from 2006 to 2012, revealig that the provisios for impairmet are positively associated with the o-performig loa rate ad the loa/assets ratio. Vol. 17, No

5 Some researchers (Diaz & Olivero, 2011; Duica & Pop, 2015; Faistei & Novikov, 2011; Festic et al., 2011; Fiori et al., 2007; Leitao, 2012; Mileris, 2012, 2014; Virolaie, 2004; Yeyati, 2010) have foud relatios betwee the macroecoomic factors ad o-performig loas i the bakig system. The amout of past o-performig loas iteract with the macroecoomic chages ad idicate that baks with more o-performig loas are more affected by the macro shocks. The results suggest that o-performig loas icrease across baks as macroecoomic risks rise. However, the iteractio terms ad the macroecoomic factors are highly correlated, so it is difficult to disetagle the idepedet effects of each macroecoomic variable. Geerally, baks with a higher exposure to macroecoomic risk are expected to experiece larger o-performig loas. Istrate et al. (2016) cosider that the umber of debtors to credit is iflueced by the iterest rate, meaig that a icrease i iterest rates o the iterbak market determies a icrease i the probability of default at maturity of loas cotracted, while the destiatio of loas cotracted (equipmet, treasury) does ot sigificatly ifluece PD. O the other had, Castro (2013) determied based o a model of pael data that the credit risk is sigificatly iflueced by the macroecoomic eviromet. Thus, the credit risk icreases with the decrease i the GDP, the share price, but decreases with the icrease i the uemploymet rate or the real exchage rate appreciatio. Berge ad Boye (2007) state that as determiats of o-quality of ledig ad, implicitly, of bak loas, respectively of recordig of NPLs by creditor baks, should be cosidered the very lack of quality or performace i the activities of the credit cotract parties. Most of bakig crisis theories are based o chages i ecoomic fudametals as a atural cosequece of busiess cycles, with credit growig procyclically. Credit grows rapidly whe the ecoomy is boomig, as ivestors tur more optimistic about the future ad ledig stadards deteriorate. Whe the ecoomic growth slows dow, this causes a collapse i creditig. The macroecoomic origis of bakig crises lie i usustaiable macro policies ad global fiacial coditios. Louzis et al. (2012) examied the determiats of o-performig loas i the Greek bakig sector, differetiated for each category of loas (cosumer, corporate, mortgage loas). Assumig that both the macroecoomic idicators ad the credit coditios affect the credit quality, the authors cocluded that for all categories of loas the o-performig credit rate ca be explaied by the evolutio of macroecoomic variables (GDP, uemploymet, public debt, etc.) ad the quality of maagemet, mortgage credits beig less sesitive to chages i macroecoomic idicators. Derbali s study (2011) revealed a positive associatio betwee iflatio ad bak profitability. High iflatio is geerally associated with high iterest rates, ad 126 Vol. 17, No. 1

6 The impact of macroecoomic variables o the evolutio of the credit risk rate therefore, high reveue. However, if iflatio is ot aticipated ad baks are slow i adjustig iterest rates, there is the possibility of a rapid growth of expediture compared to bakig reveue ad therefore bak profitability ca be egatively affected. Buch et al. (2014) aalyzed the relatioship betwee the macroecoomic variables ad the credit policy i the US, ivestigatig how baks uder aalysis perceive aalyzed macroecoomic shocks, cocludig that baks perceive shocks differetly, depedig o size, capitalizatio, liquidity. The scietific literature aalysis cofirmed the importace of cosiderig the systemic risk factors o credit risk i baks, because the loa portfolio quality is ofte strogly related to macroecoomic coditios. Pestova ad Mamoov (2014) foud that the most egative ifluece over the credit quality i Russia have had the macroecoomic coditios. The results obtaied have highlighted the importace of macro-prudetial regulatios for improvig the fiacial stability of the Russia bakig system. A prelimiary coclusio draw from most studies usig the GDP growth rate as primary idicator of macroecoomic coditios ad sustaiability of the debt of a large group of borrowers idicates that a GDP rate growth traslated by higher reveue determies a higher probability of repaymet of loas by borrowers. Although the studies refer to the iteratioal eviromet, we believe that the macroecoomic factors ifluece the credit risk rate for Romaia as well. However, the iformatio available for the Romaia bakig eviromet is ot sufficietly well sythesized give that the Natioal Bak of Romaia does ot disclose all prudetial idicators. 2. Cosideratios o research methodology The methodological solutios applied to idetify a simplified fuctioal form of macroecoomic models of credit risk measuremet are based o multivariate regressios, estimated idepedetly or simultaeously i equatios. This research aims to assess the impact of the macroecoomic idicators uder aalysis o the deterioratio of the loa portfolio. Work methodology is based o the aalysis ad iterpretatio of statistical results, the variables cosidered beig aalyzed usig the multiple liear regressio. The empirical aalysis is based o a set of aual data that allows the assessmet of the process of repaymet of loas cotracted by o-fiacial corporatios, with the aim of explaiig the ifluece that idepedet variables have o the credit risk (the depedet variable).the empirical aalysis is based o a set of data havig a aual frequecy, which allows the assessmet of the process of reimbursemet of loas cotracted by o-fiacial compaies. Vol. 17, No

7 To capture the global ecoomic coditios o the reimbursemet behavior of compaies i various ecoomic sectors, the sesitivity of bak loa oreimbursemet rates to the dyamics of macroecoomic variables was studied distictively i four distict major busiess sectors: agriculture, costructio, trade ad services. Busiess sectors retaied i the aalysis are relevat to the atioal ecoomy. I order to carry out the research, the followig questio was formulated: Ca the credit risk be predicted based o the evolutio of macroecoomic variables? The coceptual model of the credit risk rate predictio system assumes that the fuctioal relatioship betwee the edogeous variable ad the set of explaatory idicators is described by the logit fuctio. The aalysis covers aual frequecy data from the period from 2009 to 2015, available o the websites of the Natioal Istitute of Statistics ad the Natioal Bak of Romaia. For research purposes, we used: o The credit risk rate depedet variable as it appears i the aggregated idicators for credit istitutios o the website of the Natioal Bak of Romaia, expressed as a proportio of gross exposure o loas ad iterest classified uder doubtful ad loss i total loas ad classified iterest (excludig off balace sheet items) - Y. To quatify the credit risk, the Natioal Bak of Romaia used the credit risk rate to defie the ratio betwee the gross exposure of o-bak loas classified as loss or doubtful ad the total amout of loas classified as o-bak loas excludig off-balace sheet items. Accordig to the prudetial regulatios of the Natioal Bak of Romaia, oly loas for which miimum capital requiremets for credit risk are categorized, at the idividual level, accordig to the stadard approach, the remaider of the portfolio beig exempted from classificatio (for baks usig the iteral ratig models).the credit risk rate is a more restrictive idicator of the loa portfolio quality tha the o-performig loas share. If a loa is cosidered operformig if the debt service exceeds 90 days, the credit risk rate icludes i additio the loas/iterest with a outstadig debt service of less tha 90 days, ad the classificatio criteria also iclude the fiacial performace of the debtor ad the priciple of dowgradig by cotamiatio. 2 o 1. Variables of the real ecoomy: gross value added X 1 itermediate cosumptio X 2 brach level productio X 3 The idepedet variables that were divided ito two groups: 2 Accordig to the Natioal Bak of Romaia 128 Vol. 17, No. 1

8 The impact of macroecoomic variables o the evolutio of the credit risk rate 2. Moetary variables: EUR/RON exchage rate X 4 ROBOR with a maturity of six moths X 5 The aalyzed data are show i Table 1: Table 1. Database with aalyzed variables Real ecoomy Moetary factors Year Productio Itermediate cosumptio Gross value added EUR/RON exchage rate ROBOR 6M Credit risk rate** (% p.a.) (% p.a.) (% p.a.) (% p.a.) (% p.a.) X1 X2 X3 X4 X5 Y Agriculture Costructio Trade Vol. 17, No

9 Trasport Source: ad author s processig 4. Data aalysis based o empirical research I order to study the impact of the macroecoomic idicators evolutio o the credit risk, usig the pael data methodology, there were estimated some equatios for determiig the credit risk depedig o the variables of real ecoomy ad moetary factors. The period for which the aalysis is coducted is , usig series of aual frequecy for the selected idicators. To facilitate iterpretatio of the coefficiets obtaied from the regressio, the ecoometric aalysis is performed o logarithmic series. The idepedet variable coefficiet idicates the percet by which the Y depedet variable chages whe the X jt idepedet variables chage by 1 percet. I the selectio of specificatios of the credit risk aalysis models, we started from estimatig the equatio system. Give the complexity of the researched pheomeo ad the objective limits o the coverage over time of the available iformatio, the multivariate procedure of selectio of the determiats aimed at buildig a reasoable umber of models that ecompass a diversity as high as possible of the idepedet variables. I order to study the impact of idepedet variables o the credit risk, empirical correlatios have bee made betwee the depedet variable ad the idepedet variables, the ecoometric aalysis beig performed by logarithmic series, for facilitatig iterpretatio of obtaied results. The matrix of correlatio of the aalyzed variables (Table 2) shows that high values of credit risk are associated with exchage rate movemets i all four idustry sectors uder aalysis, as evideced by the positive correlatio betwee the rate of credit risk ad the exchage rate. O the other had, the egative values of the correlatio betwee the credit risk rate ad productio i the case of eterprises workig i agriculture, respectively productio, itermediate cosumptio, gross value added i the case of trade, 130 Vol. 17, No. 1

10 The impact of macroecoomic variables o the evolutio of the credit risk rate costructio ad trasport, show that the high values of the idepedet variables are associated with a low credit risk, while the low values of the idepedet variables are associated with a high credit risk. AGRICULTURE Table 2. Matrix of correlatio of the aalyzed variables Productio Itermediat e cosumptio Gross value added Exchag e rate ROBOR 6M Productio 1 Itermediate cosumptio Gross value added Exchage rate ROBOR 6M Credit risk rate , TRADE Productio Itermediat e cosumptio Gross value added Exchag e rate ROBOR 6M Productio 1 Itermediate cosumptio Gross value added Exchage rate ROBOR 6M Credit risk rate CONSTRUCTION Productio Itermediat e cosumptio Gross value added Exchag e rate ROBOR 6M Productio 1 Itermediate cosumptio Gross value added Exchage rate ROBOR 6M Credit risk rate Credi t risk rate Credi t risk rate Credi t risk rate Vol. 17, No

11 TRANSPORT Productio Itermediat e cosumptio Gross value added Exchag e rate ROBOR 6M Productio 1 Itermediate cosumptio Gross value added Exchage rate ROBOR 6M Credit risk rate , Credi t risk rate I order to show the depedece betwee the selected variables i the chose time frame, we estimated the regressio model i Excel. The multiple correlatio betwee variables shows that there is a strog lik betwee the credit risk rate ad the aalyzed macroecoomic variables i all four busiess sectors (Table 3). Table 3. Estimate of the regressio model Agriculture Trade Costructio Trasport SUMMARY OUTPUT Regressio Statistics Multiple R R Square Adjusted R Square Stadard Error years Takig ito accout the results of the statistical tests, the macroecoomic variables uderlyig the aalysis play a importat role i the fiacial stability of the moetary ad bakig system, the severe deterioratio of oe of them ifluecig the credit risk rate. O the other had, cosiderig the specificity of the Romaia market, a raised R 2 is ot etirely uusual. Lookig at the results, we see that the rate of credit risk is iflueced i more tha 99% of cases by the macro variables aalyzed i all sectors uder study (R Square > 99%). Of course, the results obtaied are ot surprisig at all. As preseted i Sectio o. 2, authors such as McwayizeiThwala(2016), Makri et al. (2014), Skala (2014), or Figlewski et al. (2012), cocluded that macroecoomic idicators have a sigificat impact o credit risk.studies o the impact of macroecoomic variables o the evolutio of the credit risk rate are covergig; the coclusio draw from most studies is that macroecoomic variables have a sigificat ifluece o the 132 Vol. 17, No. 1

12 The impact of macroecoomic variables o the evolutio of the credit risk rate evolutio of the credit risk. Of these, the most sigificat impact o credit risk is the exchage rate, but also the variables such as iterest rate ad GDP. The values of the statistical tests show that the obtaied model meets the requiremets of a good ecoometric performace. The dataset used was proved to be correct both from the poit of view of the validatio of the correlatio coefficiets ad the compliace with the sigificace threshold (the probability associated with the t test is lower tha the most restrictive relevace level (1%) ad the probability associated with the F test shows that at least oe of the regressio coefficiets is sigificat from the statistical poit of view).next, we carried out the aalysis of data series usig the EViews software applicatio to determie the correlatio betwee the credit risk ad the aalyzed variables, the aalysis result beig preseted i Table 4. Table 4. Research results i relatio to the idepedet variables Agriculture Trade Costructio Trasport Std. Coefficiet Std. Error Coefficiet Error Std. Coefficiet Error Std. Coefficiet Error Real ecoomy Productio Itermediate cosumptio VAB Moetary coditios Exchage rate ROBOR 6M The research results highlight the fact that the dyamics of the credit risk is iflueced maily by the evolutio of moetary factors. Thus, a icrease i the exchage rate determies a icrease i the credit risk by 54.59% for eterprises operatig i agriculture, 72% i trade, 62.55% i costructio ad 73.40% i the case of trasport compaies. Istead, for eterprises operatig i the trasport field, the credit risk rate is strogly iflueced by the dyamics of the real ecoomy, as demostrated by the coefficiets calculated for this busiess sector. Ulike the results obtaied i this research, Moiescu ad Codirlaşu (2013) said i the study etitled "Sectoral dyamics of the bad loas rate: ecoomic ad moetary ladmarks", that real ecoomy dyamics is the mai vector of the evolutio of arrears at the level of the loa portfolio grated to o-fiacial corporatios, the moetary pressure iduced by moetary coditios beig a secodary elemet, with the exceptio of the costructio sector. O the other had, Treca ad Beyovszki (2008) usig a credit risk model to determie the ifluece of macroecoomic factors (GDP growth rate, real iterest Vol. 17, No

13 rate, exchage rate) o default rates of the Romaia corporate sector (idustry, services, costructio, agriculture) over the period , cocluded that the iterest rate has a major impact o credit risk. Both the depedet variable ad the idepedet variables are expressed i atural logarithms, idicatig the percet by which the depedet variable chages to a 1% icrease i the idepedet variables. Thus, a 1% icrease i productio shall cause a decrease i the credit risk by 4.09% for eterprises operatig i agriculture, while i the case of eterprises operatig i the other sectors, the icrease of productio shall ot determie the decrease of the credit risk rate. At the same time, the icrease i the itermediate cosumptio determies the decrease i the credit risk rate i all busiess sectors, while the icrease i the gross value added reduces the credit risk rate for eterprises i trade, costructio ad trasport sector, but ot for those i the agricultural sector. As per the moetary factors, it ca be oticed that the credit risk icreases with the icrease i the exchage rate ad the iterest rate i all aalyzed sectors. 5. Coclusios ad directios for future research The ecoometric results suggest that the real ecoomy ad the moetary factors ifluece the evolutio of the credit risk rate i all four aalyzed idustry sectors. As future lies of research, we suggest itroducig more macro variables i the aalysis, variables that ifluece the probability of default. The credit risk is oe of the most importat risks that the baks are facig, which is why i recet years focus was set o the adoptio of bakig busiess patters, i order to eable fiacial istitutios to develop a effective framework for risk assessmet, but without puttig profitability at risk. The developmet ad applicatio of techiques for maagig the credit risk is a old cocer that has evolved from techiques such as exposure assessmet, to limit excessive focus o a sigle borrower, busiess sector or idustry brach, to moder techiques such as trasactios with swaps ad optios, tailored to this risk. Baks should idetify ad assess the degree of risk that accompaies each credit applicatio, ad where the risk is high it will be reflected i the credit coditios (high iterest rate).the degree of risk associated with a firm or a credit grated for that compay is determied both by iteral factors ad exteral factors amog which the compay operates (macroecoomic factors).the global ecoomy evolutio, but also the curret developmets i the Romaia ecoomy commit to a policy of careful risk measuremet to achieve acceptable returs i the future, give the ifluece that the bak may suffer from competig exteral factors. 134 Vol. 17, No. 1

14 The impact of macroecoomic variables o the evolutio of the credit risk rate Refereces Alessi, M., Di Colli, S. & Lopez, J.S. (2014) Loa loss provisioig ad relatioship bakig i Italy: Practices ad empirical evidece, Joural of Etrepreeurial ad Orgaizatioal Diversity, vol. 3, o. 1: Berge, T.O. & Boye, K.G. (2007) A aalysis of bak s problem loas, Norges BakEcoomic Bulleti, vol.78: Boss, M., Fez, G., Pa, J., Puhr, C. Scheider, M. & Ubl E. (2009) Modelig credit risk through the Austria busiess cycle: A update of the OeNB Model, OeNB Fiacial Stability Report, vol. 17: Buch, C., Eickmeier, S. & Prieto, E. (2014) Macroecoomic factors ad micro level bak behavior, Joural of Moey, Credit ad Bakig, vol. 46, o. 4: Castro, V. (2013) Macroecoomic determiats of the credit risk i the bakig system: The case of the GIPSI, Ecoomic Modellig, vol. 31: Derbali, A. (2011) Determiats of bakig profitability before ad durig the fiacial crisis of 2007: The case of Tuisia baks, Iterdiscipliary Joural of Cotemporary Research i Busiess, vol.3: Derelioglu, G. & Gurge, F. (2011) Kowledge discovery usig eural approach for SME s credit risk aalysis problem i Turkey, Expert Systems with Applicatios, o.38: Diaz, R. A. & Olivero, M. P. (2011) The cyclicality of price cost margis i bakig: a empirical aalysis of its determiats, Ecoomic Iquiry, vol. 49, o. 1: Duica, E.R.& Pop, A. (2015) The implicatios of credit activity o ecoomic growth i Romaia, Procedia Ecoomics ad Fiace, o.30: Faistei, G. & Novikov, I. (2011) The comparative aalysis of credit risk determiats i the bakig sector of the Baltic States, Review of Ecoomics & Fiace, o. 3: Festic, M., Kavkler, A. & Repia, S. (2011) The macroecoomic sources of systemic risk i the bakig sectors of five ew EU member states, Joural of Bakig & Fiace, o. 35: Figlewski, S., Frydma, H. & Liag, W. (2012) Modellig the effect of macroecoomic factors o corporate default ad credit ratig trasitios, Iteratioal Review of Ecoomics ad Fiace, vol. 21: Fiori, R., Foglia A. & Iaotti S. (2007) Estimatig macroecoomic credit risk ad sectoral default rate correlatios for the italia ecoomy, Workig Paper, Bak of Italy Gagais, C., Pasiouras, F., Doumpos, M. & Zopouidis, C. (2010) Modellig bakig sector stability with multicriteria approaches, Optimizatio Letters, vol. 4: Vol. 17, No

15 Istrate, L.G., Ioescu, B.Ș. & Haralambie, M. (2016) Aspects of the impact of iterest rate developmet o the probability of default, Audit Fiaciar, vol. 14, o. 142: Leitao, N.C. (2012) Bak, credit, ad ecoomic growth. A dyamic pael aalysis, The Ecoomic Research Guardia, vol. 2(2): Louzis, D., Vouldis, A. & Metaxas, V. (2012) Macroecoomic ad bak-specific determiats of o-performig loas i Greece: A comparative study of mortgage, busiess ad cosumer loa portfolios, Joural of Bakig & Fiace, vol. 36, o. 4: Makri, V., Tsagkaos, A. & Bellas, A. (2014) Determiats of o-performig loas: The case of eurozoe,paoecoomicus, vol. 61, o.2: Makri, V. (2015) What triggers loa losses? A empirical ivestigatio of Greek fiacial sector, SPOUDAI Joural of Ecoomics ad Busiess, vol.65, o. 2: Makri,V. & Papadatos, K. (2016) Determiats of loa quality: Lessos from Greek cooperative baks, Review of Ecoomic & Busiess Studies, vol. 9, o.1: Mcwayizei Thwala, C. (2016) The sesitivity of bak credit risk idicators to macroecoomic variables, A dissertatio submitted to the Faculty of Commerce, Law ad Maagemet Uiversity of the Witwatersrad Busiess School, Johaesburg Mileris, R. (2012) The effects of macroecoomic coditios o loa portofolio credit risk ad bakig system iterest icome, Ekoomika, vol. 91, o. 3: Mileris, R. (2014) Macroecoomic determiats of loa portfolio credit risk i baks, Izierie Ekoomika-Egieerig Ecoomics, vol. 23, o. 5: Moiescu, B. & Codirlașu, A., (2013) Ledig, ecoomic growth ad operformig loas: empirical evideces from the ew EU member states, Workig Paper Skała, D. (2014) Credit policy i small polish baks Is there room for icome smoothig?, Collegium of Ecoomic Aalysis Aals, Warsaw School of Ecoomics, o.34: Treca, I. & Beyovszki, A. (2008) Credit risk, a macroecoomic model applicatio for Romaia, Fiace Challeges of the Future, o. 7: Virolaie, K. (2004) Macro stress testig with a macroecoomic credit risk model for Filad, Bak of Filad, Discussio Papers No.18 Yeyati, E. L., Peria, M. S. M.& Schmukler, S. L. (2010) Depositor behavior uder macroecoomic risk: evidece from bak rus i emergig ecoomies, Joural of Moey, Credit ad Bakig, vol. 42, o 4: Vol. 17, No. 1