An Empirical Analysis of Credit Risk Factors
ofthe Slovenian Banking System
Boštjan Aver
The study presents the results of an analysis of credit risk factors of
the Slovenian banking system. The objective of the empirical analysis
is to establish which macroeconomic factors influence the systematic
credit risk of the Slovenian banking loan portfolio. The research results
have confirmed the main hypothesis that certain macroeconomic fac-
tors have a major influence on the examined credit risk. We could con-
clude that the credit risk of the loan portfolio depends on the employ-
ment or unemployment rate in Slovenia, on short and long-term inter-
est rates of Slovenian banks and the Bank of Slovenia, and on the value
of the Slovenian stock exchange index. We cannot claim that the exam-
ined credit risk depends on the inflation rate in Slovenia, the growth
of gdp (industrial production), eur and usd exchange rates or the
growth of Slovenian import and export.
Key Words: Slovenian banking system, credit risk factors,
loan portfolio, Bank of Slovenia, macroeconomic factors
jel Classification: gii, g21
Introduction
Empirical studies on credit risk factors have shown that the factors that
influence risk in all sorts of investments and cause credit risk are mainly
different macroeconomic factors (Saunders 1997; Crouhy, Galai, and Mark
2000). Changes in economic policies, political changes and the goals of
leading political parties also influence the range of the investment credit
risk (Saunders 1997; Belkin, Suchower, and Forest 1998). Since these fac-
tors are difficult to examine, there is no point in including them in 'the
research model'. Our objective is therefore to establish which macroeco-
nomic factors influence the systematic credit risk of the Slovenian bank-
ing system.
Foreign researchers, such as Saunders (1997), Asarnow (1996), Crouhy,
Galai, and Mark (2000), Carty and Lieberman (1996), as well as Carty
and Fons (1993), have determined that the range of the credit risk of an
individual investment (e. g. loan) is influenced by risk factors, which can
Dr Boštjan Aver is President ofthe Management Board
of Vzajemna, d. v. z., Slovenia.
Managing Global Transitions 6 (3): 317-334
be divided into those influencing the risk of all investments and caus-
ing investment's systematic credit risk and those influencing the risk of
an individual investment and causing investment's unsystematic credit
risk. The factors influencing the systematic credit risk are macroeconomic
factors, changes in economic policies, political changes and the goals
of leading political parties. Macroeconomic factors include the inflation
rate, the employment rate, growth in gross domestic product, stock in-
dex and exchange rate movements, and conjuncture fluctuations in the
economy. Changes in economic policies are represented by changes in
monetary and tax policies, economic legislation changes, as well as im-
port restrictions and export stimulation (Saunders 1997; Temeljotov Salaj
2005) L The factors influencing the unsystematic credit risk are primarily
the factors of individual customers, such as their personality, their finan-
cial solvency and capital, credit insurance and general terms and condi-
tions. In the case of companies, specific factors of the industry sector and
the company are emphasised (Mramor 1996). Industry factors include
the structure and economic successfulness of the industry, maturity of
the industry and its stability, while company factors include factors such
as general characteristics of the company, management, financial posi-
tion, sources of funds and financial reporting.
When deciding about new bank customers it is thus important to con-
sider the credit risk of an individual customer, which can nowadays be
measured with the help of various modern credit models (Aver 2003).
However, in assessing the influence of an individual obligor on the bank
operation risk and the influence of capital requirements, the risk of the
entire bank portfolio (see Asarnow 1996) must also be taken into ac-
count. Customer risk and portfolio risk are specified by the expected
loss and standard deviation of the loss, which defines the unexpected
loss (Ong 2000). The latter is often assessed as a multiple of the standard
deviation of the loss (Bessis 1998).
According to CreditMetrics™ technical document (jp Morgan 1997),
company credit risk may also be influenced by systematic market risk,
which is reflected in the changes in interest rates, stock index, exchange
rates and unemployment level. The probability of default of a company
and the probability of its credit rating migration is thus related to market
risk. In an ideal world, the methodological tool for bank risk measure-
ment should therefore link market risk and credit risk, which banks are
exposed to in their operations.
Figure 1 presents the influence of the state of world economy on default
65 ^ Number of defaulted companies
15 20-
16
m
<
						
						
17				28		
27	30	29			21 14	17 14
16					7	
1985 1986 1987 1988 1989 1990 1981 1992 1993 1994 1995 1996 1997
figure 1 Number of defaulted companies and the amount of defaulted loans
(adapted from Crouhy, Galai, and Mark 2000, 319)
62
12
8
4
0
probabilities of companies or frequencies of default of foreign compa-
nies in the period from 1985 to 1997. In 1990 and 1991, when the world
economy was in recession, the frequency of company defaults rose sub-
stantially. On the other hand, during the period after 1991, which was
characterised by a growing economy, the number of defaults declined.
Actual default probabilities of companies vary over time, depending
on the state of economy. In 1996, Carty and Lieberman conducted a study
on average probabilities of default and their standard deviations for in-
dividual credit rating categories of bank obligors in the period from 1970
to 1995, the results of which are provided in table 1. Similar studies have
been conducted by Carty and Fons (1993) and Lucas and Lonski (1992).
Tom Wilson and McKinsey and Company (Paul-Choudhury 1998) de-
veloped CreditPortfolioView, a multifactorial model for credit risk mea-
surement, which can be useful for generating the distribution of default
probabilities and credit rating migration probabilities for different in-
dustry sectors and for each individual country. CreditPortfolioView takes
into account that default probabilities and probabilities of obligor credit
rating migrations depend on the state of the economy. When the econ-
omy is not doing well, the probability of default of companies and credit
downgrades increase, and vice versa when the economy is strong (Belkin,
Suchower, and Forest 1998). Slower economic growth thus causes more
frequent credit rating migrations, with lower credit rating classes having
table 1 One-year default probabilities in the period from 1970 to 1995
Credit rating category	One-year default probability	
	Average	Standard deviation
aaa	0.00%	0.00%
aa	0.02%	0.12%
a	0.01%	0.05%
baa	0.15%	0.30%
ba	1.22%	1.35%
b	6.32%	4.78%
note Adapted from Moody's Investor Service 1995.
figure 2 Markov transition matrix
Initial rating
Rating at year-end
aaa aa
a bbb bb
ccc Def.
aaa
aa
a
bbb
bb
b
ccc
Recession
Expansion
note Adapted from Wilson 1997.
a higher correlation with macroeconomic factors. Credit rating migra-
tions during the time of economic recession and expansion can be clearly
illustrated with the Markov transition matrix, as shown in figure 2.
As economic growth depends on macroeconomic factors, CreditPort-
folioView takes into account the links between macroeconomic variables
and default and credit rating migration probabilities of companies. The
model includes the unemployment rate, gross domestic product growth,
the level of long-term interest rates, foreign exchange rates, government
expenditures and the savings rate among macroeconomic factors (Belkin,
Suchower, and Forest 1998).
During a period of recession, default probabilities of speculative-grade
obligors are higher than their average default probability, as frequencies
of obligor migrations to lower credit rating categories are increasing and
frequencies of upward obligor migrations are decreasing. The contrary
b
holds true for the period of economic expansion. The probabilities of
obligor credit rating migrations thus depend on the credit cycle, the in-
dustry sector and the country, the credit cycle, however, being the most
important factor in the variability of company default probabilities (Ban-
gia, et al. 2002).
The Duffie and Singleton model has been generalised by Lando (1997)
and Jarrow and Turnbull (1998) with an assumption that the intensity
of default or the obligor default probability (A) depends on the vector
of various macroeconomic variables (Xt), such as risk-free interest rate,
stock index, etc. The obligor default probability is modelled in the form
of the Cox distribution, which has the characteristics of the Poisson dis-
tribution and depends on the vector of macroeconomic variables (Xt):
A(t) = A(Xt).	(1)
In 1997, Duffie and Singleton derived a risk-adjusted short-term inter-
est rate (Y):
7 (t) = r(t) + A(t)LGD + 1.	(2)
In order to simplify the model implementation, Jarrow and Turnbull
suggested the following:
1.	the risk-free interest rate r(t) is in accordance with the one-factor
Vasicek model from 1977,
2.	the obligor default probability A(t) depends on the change in the
risk-free interest rate r(t) and unexpected changes in the value of
the market index WM(t):
A(t) = A0 + Axr(t) + A2WM (t),	(3)
where A0, Ai and A2 are constants. WM(t) indicates unexpected changes
in the value of the market index according to the value of the stock index
M(t), for which standard lognormal distribution is presupposed:
dM(t) = [r(t)dt + aMdWM (t)]M(t).	(4)
1.	the level of loss given default (lgd) is constant,
2.	the liquidity premium (l) depends on the risk-free interest rate r(t),
the stock index M(t) and the variability of the daily value of the
stock index:
l(t) = lo + lir(t) + feM(t) + 13[Mh(t) - ML(t)]2,	(5)
where M (t) is the stock index value on a certain day, while MH (t) and
ML(t) indicate its highest and lowest daily value.
In the following sections, we explain the underlying conceptual and
methodological framework and the results of the analysis conducted ac-
cording to foreign findings on the example of the Slovenian banking sec-
tor.
Developing the Hypotheses
On the basis of conducted studies and analyses, Saunders (1997), Asar-
now (1996), Crouhy, Galai, Mark (2000), Carty and Lieberman (1996), as
well as Carty and Fons (1993) have determined that the range of the credit
risk of an individual investment is influenced by risk factors which influ-
ence the systematic credit risk and risk factors which influence the unsys-
tematic credit risk of an investment. The factors influencing the system-
atic credit risk are macroeconomic factors, changes in economic policies,
political changes and the goals of leading political parties. Macroeco-
nomic factors include primarily the inflation rate, the employment rate,
growth in gross domestic product, stock index and exchange rate move-
ments, and conjuncture fluctuations in the economy. Changes in eco-
nomic policies are represented by changes in monetary and tax policies,
economic legislation changes, as well as import restrictions and export
stimulation (Saunders 1997; Mramor 1996).
Based on the above-mentioned findings of foreign researchers and ex-
perts, particularly in terms of the macroeconomic factors which can in-
fluence the range of systematic credit risk, the main hypothesis can be
developed:
H1 The range of the credit risk of the Slovenian banking system depends
on specific macroeconomic factors.
The main hypothesis can be further divided into several subhypotheses.
The results of testing the latter represent the basis of confirming or re-
jecting the main hypothesis (h1). In order to confirm or reject individ-
ual subhypotheses, it is necessary to determine the influence of specific
macroeconomic variables on the credit risk of our banking system. The
subhypotheses (H1-1 to H1-7) and examined variables for testing them
are as follows:
H1-1 The systematic credit risk of the Slovenian banking system depends
on the inflation rate in Slovenia. Examined variables: revalor,
czp and dpc.
H1 - 2 The systematic credit risk of the Slovenian banking system depends
on the employment or unemployment rate in Slovenia. Examined
variables: zaposl and brezposl.
H1-3 The systematic credit risk of the Slovenian banking system de-
pends on the growth in Slovenian gross domestic product. Exam-
ined variable: rastindu.
h1-4 The systematic credit risk of the Slovenian banking system de-
pends on the short-term and long-term interest rate activities of
banks and the Bank of Slovenia. Examined variables: medbom,
lombom, zom, omtekpo, omstan, omos, ompotr, nom-
obs, nompotr, nomos, nomstan, nomdobs and nomdos.
H1-5 The systematic credit risk of the Slovenian banking system depends
on the movement of specific exchange rates. Examined variables:
eur and usd.
H1-6 The systematic credit risk of the Slovenian banking system depends
on the movement of the value of the Slovenian stock exchange in-
dex (sbi) or the movement of share trade on the organised securi-
ties market. Examined variables: sbi and promdeln.
H1-7 The systematic credit risk of the Slovenian banking system depends
on the Slovenian export and import growth. Examined variables:
izvoz and uvoz.
In examining the influence of 25 chosen macroeconomic variables on
the Slovenian banking system credit risk, factorial analysis is employed
in order to find new dimensions of macroeconomic factors, which rep-
resent common characteristics of some macroeconomic variables. Thus,
the second hypothesis can be developed:
H2 The method of principal components and choice of two factors will
provide us with good enough partial correlation coefficients between
individual macroeconomic variables and both factors, so that the first
factor and second factor will together account for more than 50% of
the overall credit risk variance.
Thus, the used empirical analysis of the credit risk factors of the Slove-
nian banking system portfolio is based on monthly data of 25 chosen
macroeconomic factors and the calculated indicator of the Slovenian
banking system portfolio credit risk for the period from 31 December
1995 to 30 November 2002. The evaluation of the initial hypothesis (h1):
The range of the credit risk of the Slovenian banking system depends on
specific macroeconomic factors, was based on the set research model ex-
plained below.
Research Methodology
data collection and operationalisation
of the variables
According to the findings of foreign researchers such as Saunders (1997),
Crouhy, Galai, and Mark (2000), Belkin, Suchower, and Forest, (1998),
Carty and Lieberman (1996), as well as Carty and Fons (1993), the data
on potential factors that influence systematic credit risk could be found
among different macroeconomic variables, e. g. inflation rate, employ-
ment or unemployment rate, the increase in gross domestic product (in-
dustrial production), the movement of short and long term interest rates
of banks and the central bank, the movement of exchange rates (e. g. eur
and usd), the movement of the value of the stock exchange index and
share trade, and other macroeconomic factors (e. g. import and export).
Thus, 25 different macroeconomic factors were chosen to test the stated
hypothesis.
The research model design is shown in figure 3. The figure presents cho-
sen factors which can influence the systematic credit risk and which rep-
resent the entry model variables used to determine which of the chosen
25 macroeconomic variables have a significant influence on the range of
the Slovenian banking system credit risk.
The following data sources were used in the analysis:
•	gvin - Gospodarski vestnik database,
•	sors - Statistical Office of the Republic of Slovenia database,
•	Bank of Slovenia database (the archive of financial data from the
Bulletins ofthe Bank of Slovenia),
•	monthly balance sheets and profit and loss accounts of all Slovenian
banks for the period from 31 December 1995 to 30 November 2002,
•	other publicly available materials of the Bank of Slovenia.
The conducted analysis of the Slovenian banking system portfolio is a
quantitative analysis, since the attributes used to describe the systematic
credit risk factors of the banking portfolio are numerical, not descrip-
tive. The influence of specific macroeconomic factors on the range of the
Slovenian banking system credit risk has been examined with the help
of spss software and specific statistical methods, such as multiple linear
regression and factorial analysis.
An Empirical Analysis of Credit Risk Factors 325
figure 3 Systematic credit risk factors
Inflation rate	revalor	Revalorisation clause
	czp	Consumer goods prices*
	dpc	Retail prices
Employment	zaposl	Employment rate
	BREzpoSL	Unemployment rate
gdp	rastindu	Industrial production
Interest rates	medbom	Interbank ir
	lombom	Lombard ir
	zom	Default ir
	omtekpo	Real ir on current business loans*
	omstan	Real ir on long-term home loans
	omos	Real ir on long-term fixed asset loans
	ompotr	Real ir on short-term consumption loans
	nomobs	Nominal ir on short-term current asset loans*
	nompotr	Nominal ir on short-term consumption loans
	nomos	Nominal ir on long-term fixed asset loans
	nomstan	Nominal ir on long-term home loans*
	nomdobs	Nominal ir on short-term foreign exchange loans for current assets
	nomdos	Nominal ir on long-term foreign exchange loans for fixed assets
Exchange	eur	EuR exchange rate
rates	uSD	usd exchange rate*
Stock exchan-	sbi	Slovenian stock exchange index
ge market	promdeln	Stock exch. trade and the trade on the free share market*
Export/	izvoz	Import of goods and services
import	uvoz	Export ofgoods and services
* Systematic credit risk.
THE sample
The established model is based on the data of the variety of macroe-
conomic credit risk factors, collected on monthly basis, as well as on
monthly data of the calculated indicator of the Slovenian banking system
portfolio credit risk for the period from 31 December 1995 to 30 Novem-
ber 2002. The chosen assessment period of the Slovenian banking system
after 1995 is the most appropriate one for the analysis, since the period
before the end of the year 1995 would mystify the results of the analysis
to a great extent.
In that period, the range of the Slovenian bank portfolio credit risk
was being contracted due to different system effects, among others, due
to the implementation of the rehabilitation programme in Ljubljanska
banka, d. d. and Kreditna banka Maribor, d. d. (in 1993), as well as in
lb Komercialna banka Nova Gorica, d. d. (in 1994). System measures are
also visible in the period from March 1994 to September 1996, namely in
the exchange of poor receivables with government bonds.
Taking into account the examined period from 31 December 1995 to
30 November 2002, when there were no significant system effects on the
range of the Slovenian bank credit risk, good results can be obtained on
the basis of a statistical analysis of the effect of macroeconomic factors
on the credit risk of the Slovenian banking system portfolio. At the same
time, monthly data for all independent variables and dependant variables
can be examined as individual cases and not as time series, thus neglect-
ing the time influence or the time components on the credit risk range.
The indicator of loan portfolio credit quality was first defined as a ratio
between the range of formed value corrections for credit risk of loans to
the non-banking sector2 (government excluded) and the range of gross
loans given to the non-banking sector (government excluded). The loans
given to the government were excluded. According to the methodology
of the Bank of Slovenia, the loans given to the government are included
in the loans given to the non-banking sector. Therefore, they influence
or have greater effect on the percentage of the formed value corrections,
since receivables to the government had better ratings. We analysed the
mentioned indicator of the credit quality of the Slovenian banking sys-
tem loan portfolio in the period from December 1995 to November 2002.
data analyses
We can establish which of the 25 chosen macroeconomic variables influ-
ence the range of the Slovenian banking system portfolio credit risk with
the method of multilinear regression. The following indicator (tveg1)
was defined for measuring the credit risk:
Value corrections for the credit risk of loans given to the non-banking sec-
tor/gross loans to the non-banking sector.
The multiple linear regression model can also be presented as (see fig-
ure 3):
An Empirical Analysis of Credit Risk Factors 327
tveg1 = ß1 + ß2 revalor + ß3 czp + ß4 dcp + ß5 zaposl
+ ß6 brezposl + ß7 rastindu + ß8 medbom
+ ß9 lobom + ß10 zom + ß11 omtekpo + ß12 omstan
+ ß13 omos + ß14 ompotr + ß15 nomobs
+ ß16 nompotr + ß17 nomos + ß18 nomstan
+ ß19 nomdobs + ß20 nomdos + ß21 eur + ß22 usd
+ ß23 sbi + ß24 promdeln + ß25izvoz
+ ß26 uvoz + u	(6)
The analysis of the influence of the 25 chosen macroeconomic vari-
ables on the credit risk of the Slovenian banking system could also be
conducted with a factorial analysis (method of principal components), by
finding new dimensions of macroeconomic factors that represent com-
mon characteristics for some of the macroeconomic variables.
Findings and Results
results of multiple linear regression model
As a result, we get a multiple linear regression model of seven statistically
significant variables, with R-square being equal to 0.863. This means that
86.3% variability of the systematic credit risk of the Slovenian banking
system portfolio is explained by linear dependency on seven macroeco-
nomic factors. It can be noticed that the increase in the real interest rate
for short-term consumption loans, the increase in the Slovenian stock
exchange index, a decreasing number of employees in Slovenia, an in-
creasing securities interest rate (interest rate of the Bank of Slovenia for
securities exchange) and the increase in the real interest rate on home
loans have the greatest influence on the increase in credit risk. In a way,
the mentioned results are logical. On the other hand, the two results, the
increase in the real interest rate for long-term loans for current assets and
the increase in the inter-banking interest rate, which have an effect on the
decrease of the Slovenian banking system portfolio credit risk, seem less
logical.
The obtained results are shown in table 2 and figure 4. Table 2 shows
partial regression coefficients for all 7 macroeconomic factors which have
a significant influence on the range of the examined credit risk, as well
table 2 Partial regression coefficients and T-test for testing their influence
Variable	(1)	(2)	(3)	(4)	(5)	(6)
Constant	17.62368	1.981	-	8.898	0.000	-
ompotr	0.45794	0.068	2.902	6.775	0.000	0.614
sbi	0.00050	0.000	0.551	6.046	0.000	0.570
omos	-0.66577	0.119	-2.681	-5.577	0.000	-0.539
zaposl	-0.00001	0.000	-0.463	-4.441	0.000	-0.454
medbom	-0.05264	0.014	-0.350	-3.684	0.000	-0.389
lombom	0.09185	0.027	0.179	3.343	0.001	0.358
omstan	0.12914	0.066	0.600	1.949	0.050	0.218
notes Column headings are as follows: (1) unstandardized coefficients: B, (2) stan-
dard error, (3) standardized coefficients: ß, (4) T-test: t, (5) significance, (6) partial cor-
relation coefficients.
Systematic credit risk
of the Slovenian banking
system
figure 4 Dependence of the Slovenian banking system credit risk on seven
macroeconomic variables on the basis of partial correlation coefficients (r)
and partial regression coefficients (ß); R2 = 0.863
as T-test results in relation to testing the characteristics of the indepen-
dent variables' influence on the credit risk. Moreover, figure 4 illustrates
an extended linear regression model, which fully explains the variabil-
ity of the systematic credit risk range of the Slovenian banking system
portfolio.
results of factorial analysis - method of principal
components
In examining the influence of 25 chosen macroeconomic variables on the
Slovenian banking system credit risk, factorial analysis was employed in
order to find new dimensions of macroeconomic factors that represent
common characteristics for some of the macroeconomic variables.
The method of principal components and choice of two factors pro-
vided us with partial correlation coefficients between individual macroe-
conomic variables and both factors. The first factor (various interest rates
of the Slovenian banking system) explains 58.8% of the credit risk vari-
ance, while the second factor (other macroeconomic factors) explains
10.8%. Thus, the entire model explains approximately 69.6% of the over-
all credit risk variance. The rest of the variance can be ascribed to specific
impacts that are not included in the model. The obtained result is rela-
tively good, although it should be borne in mind that there exist other
credit risk factors besides the examined ones.
The obtained communalities for individual macroeconomic variables
show what extent of the individual variable variance is accounted for by
both factors. Despite the fact that only the variables with a communal-
ity higher than 0.2 should be included in the analysis, the variables of
consumer goods prices (cpz) and retail prices (dpc) - their explained
variance with both factors being less than 20% - have also been included
in the analysis. The explained variance of the other 18 macroeconomic
variables with both factors is more than 20%. Therefore, they can be in-
cluded in the examined analysis without difficulty.
Thus, the results of the factorial analysis have shown that 20 macroe-
conomic variables are connected into 2 significant factors. The following
10 variables form the part of the first factor and they represent short and
long-term real and nominal interest rates:3 real interest rate on current
business loans, real interest rate on long-term current asset loans, real
interest rate on long-term home loans, real interest rate on short-term
home loans, nominal interest rate on short-term current asset loans,
nominal interest rate on short-term consumption loans, inter-bank in-
terest rate, nominal interest rate on long-term home loans, nominal in-
terest rate on long-term current asset loans and back interest rate.
The macroeconomic variables that form the second factor are the fol-
lowing: export of goods and services, import of goods and services, the
rate of industrial production, the number of all people employed in
Slovenia, retail prices, cost of goods sold, usd exchange rate, eur ex-
change rate, the value of the Slovenian stock exchange index, stock ex-
change trade and the trade on the free share market.
The first factor could be labelled as 'interest rates of the Slovenian
banking system', for it comprises different real and nominal (short and
long-term) interest rates of Slovenian banks. The second factor could be
labelled as 'other macroeconomic factors', for it comprises 10 different
macroeconomic variables, e. g. export (import) of goods and services,
employment rate, growth rate of production, exchange rate, develop-
ments on the securities market and the inflation rate in Slovenia.
It is possible to show the position of the above-average and below-
average risk of portfolios of our banking system regarding the first fac-
tor that represents the interest rates of the Slovenian banking system
and regarding the second factor representing other macroeconomic fac-
tors. The group of the 46 above-average risk portfolios of the Slovenian
banking system lie in the top right square,4 which means that the above-
average risk portfolios have high values of the first factor (high short and
long-term interest rates) and medium values of the second factor (va-
riety of macroeconomic factors). On the other hand, the below-average
risk portfolios of our banking system lie in the bottom left square,5 which
means that these portfolios have low values of short and long-term in-
terest rates and low values of the variety of macroeconomic factors. The
position of the above-average and below-average portfolios of our bank-
ing system regarding the first factor, which represents the interest rates
of the Slovenian banking system, and regarding the second factor, which
represents other macroeconomic factors, is shown by figure 5.
Conclusions and Implications
The results of the analysis of the credit risk of the Slovenian banking
system for the period from December 1995 to November 2002 show that
specific macroeconomic factors have important influence on the range of the
credit risk of the Slovenian banking system portfolio. The increase in credit
risk is highly influenced by the increase in short and long-term interest
rates of Slovenian banks, the increase in the value of the Slovenian stock
exchange index and the decrease in the number of employees.
In the examined period of 84 months, one of the most prominent fea-
tures of the portfolio has shown above-average risk in the period of high
real and nominal interest rates on different short and long-term loans,
during the period of high loan on securities interest rates and high inter-
	1 - above-average risk portfolios •
• 0 - below-average risk portfolios	
-1.0
-0.5
0.0	0.5
Interest rates of the banking system
1.0
figure 5 Position of the above-average and below-average portfolios of the Slovenian
banking system
bank interest rate, in the period of high unemployment (or low employ-
ment) and in the period of low eur and usd exchange rates. Taking into
account the theory of macroeconomic factors' influence on credit risk,
the given results are fairly logical and expected.
The examined macroeconomic variables are related by 2 characteristic
factors. The first is represented by the interest rate of the Slovenian bank-
ing system, which includes different real and nominal interest rates for
both short and long-term loans. The second can be described as the other
macroeconomic factors, as it consists of different macroeconomic vari-
ables. The high value of the first factor and the medium value of the
second factor is the most significant in the group of 46 above-average
risk portfolios of the banking system. On the other hand, the low value
of both factors is significant for the group of below-average risk of all
credit risk factors.
The analysis findings have shown that certain macroeconomic factors
have significant influence on the range of the Slovenian banking system
credit risk. Therefore, we can conclude that, among other factors, Asian
and Russian crises most probably influenced the increase of the range of
the loan portfolio credit risk in non-bank sectors in the period between
1997 and 1998. On the other hand, favourable movements of macroeco-
nomic government environment, the growth of consumption prior to
vat implementation, good business results in the real sector and growth
of the economies of the most important market partners influenced the
increase in the quality of credit portfolios of banking systems between
1999 and 2000. Thus, we can state with certainty that credit risk in Slove-
nian banks shows positive reactions and that banks have prompt reaction
time in cases of unfavourable market situation.
In spite of the analysis findings showing that macroeconomic factors
have substantial influence on the range of the credit risk in the Slovenian
banking system, we cannot overlook the influence of other factors that the
analysis does not explain - limitations ofthe research model. These are the
other factors that could have influenced credit risk in the banking system
in the examined period between December 1995 and November 2002:
•	the first case of bankruptcy in Slovenia: the bankruptcy of Komer-
cialna banka Triglav d. d. in 1996, due to compulsory liquidation,
•	stricter criteria for client categorisation since January 1997 (since
that date, the receivables insured by assets put in pledge could be
solely categorised into one group higher than the debtor's credit
rating);
•	the influence of the bank business results on provisions policy
(more conservative policy on estimating bank clients is typical for
the period of higher profits and preparation for new investments),
•	considerably higher net provisions for credit risk for the portfolio
of the bank skb d. d. in December 2001.
The results of the influence of the chosen macroeconomic factors on
the credit risk in the banking system have confirmed the main hypothesis
(h1), as certain macroeconomic factors have major influence on the ex-
amined credit risk. Further, we also confirmed some subhypotheses (H1-2,
H1-4 and H1-6), concluding that the credit risk portfolio of the Slove-
nian banking system depends on the employment or unemployment
rate in Slovenia (H1-2), on short and long-term interest rates activities
of Slovenian banks and the Bank of Slovenia (H1-4), and on the value
of the Slovenian stock exchange index (h1-6). Contrary to that, we can-
not claim that the examined credit risk depends on the inflation rate in
Slovenia (H1-1), the growth of gdp - industrial production (h1-3), eur
and usd exchange rates (H1-5) or the growth of Slovenian import and
export (H1-7).
On the basis of the obtained factorial analysis results we can also con-
firm the second hypothesis (h2) that more than 50% of the overall credit
risk variance of the examined Slovenian banking system portfolio can be
accounted for by the two factors together.
The conducted analysis and obtained results of the study of credit risk
factors in Slovenia can also be beneficial to banks in other countries in
transition, which do not yet apply the latest credit risk measurement and
management methods. Moreover, an important challenge is seen in recon-
ducting the study of the influence of macroeconomic factors on the range of
the credit risk of the Slovenian banking system portfolio with the data
which are yet to become available in the following years. It would also
be useful to undertake such studies in other countries, particularly in
the area of ex-Yugoslavia, where different findings about the influence
of individual macroeconomic factors on the credit risk can be expected
to arise due to a higher risk of their banking system portfolio and lower
economic development. The results of our analysis are also applicable
to other financial institutions such as insurance companies or pension
funds in risk managements of their financial investments. Finally, it will
be useful to elaborate the study of credit risk factors by employing new
and coming statistical tools such as structural equation modelling or
gravitational macroeconomic models, which can further increase the re-
liability of results.
Notes
1	In addition to the mentioned factors, the systematic credit risk of an
investment may also be influenced by the environment, e. g. the move-
ments of immovable property value, etc. (Temeljotov Salaj 2006).
2	Provisions and charges.
3	A high partial correlation coefficient associated with the first factor is
significant for the 10 variables mentioned.
4	The value of the first factor is 0.38 and the value of the second factor is
0.09.
5	The value of the first factor is -0.5 and the value of the second factor
is -0.12.
References
Asarnow, E. 1996. Best practices in loan portfolio management. Journal of
Lending and Credit Risk Management 78 (7): 14-24.
Aver, B. 2003. Modeli za oceno potencialnih kreditnih izgub banke. Bančni
vestnik 52 (9): 46-50.
Bangia, A., F. X. Diebold, A. Krohimus, C. Schagen, and T. Schuermann.
2002. Ratings migration and the business cycle, with application to
credit portfolio stress testing. Journal of Banking and Finance 26 (2-3):
445-74.
Belkin, B., S. Suchower, and L. Forest. 1998. The effect of systematic credit
risk on loan portfolio value-at-risk and loan pricing. CreditMetrics
Monitor, QK17-28.
Bessis, J. 1998. Risk management in banking. Chichester: Wiley.
Carty, L., and J. Fons. 1993. Measuring changes in credit quality. Special
report, Moody's Investors Service.
Crouhy, M., D. Galai, and R. Mark. 2000. Risk management. New York:
McGraw-Hill.
Jarrow, R. A., and S. M. Turnbull. 1999. The intersection of market and
credit risk. Journal of Banking and Finance 24 (1-2): 271-99.
jp Morgan. CreditMetrics™ technical document. 1997. jp Morgan, New
York.
Lucas, D. J., and J. Lonski. 1992. Changes in corporate credit quality 1970-
1990. The Journal of Fixed Income 1 (4): 7-14.
McDonough, W. 1998. Amending the accord. Risk Magazine 11 (9): 46-51.
Mramor, D. 1996. Analiza kreditne sposobnosti podjetja: Ocena kreditne
sposobnosti podjetja. Ljubljana: cisef.
Moody's Investor Service. 1995. Corporate bond defaults and default rates:
1970-1994. New York: Moody's Investor Service.
Ong, M. K. 2002. Internal credit risk models: Capital allocation and perfor-
mance measurement. London: Risk Books.
Paul-Choudhury, S. 1998. McKinsey's model alternative. Risk Magazine 11
(1): 7.
Saunders, A. 1997. Financial institutions management: A modern perspec-
tive. 2nd ed. New York: Irwin.
Temeljotov Salaj, A. 2005. Sinergicni ucinek opazovalca grajenega okolja.
Urbani izziv 16 (2): 48-54.
-. 2006. The quality of a built environment and legislation. Pa-
per presented at the enhr conference Housing in an Expanding Eu-
rope: Theory, Policy, Participation and Implementation, Ljubljana.
Http://enhr2006-ljubljana.uirs.si/publish/w08_TemeljotovSalaj.pdf.
Wilson, T. C. 1997. Measuring and managing credit portfolio risk: Mod-
elling systemic default risk. Journal of Lending and Credit Risk Man-
agement 79 (11): 61-72.