2. the banking system as a whole.

2.    Literature Review

 

It is undoubtedly that
in the last decades, there is an increased interest concerning the
non-performing loans and the various determinants affecting their development.
The raising interest emanate not only from the effort to detect financial and
credit exposure, as a result of the recent financial and banking crisis, but
also from the growing published data at bank-specific, country-specific, Eurozone
and even global banking system level. Several studies from all the world have
been conducted on NPL’s and default loans and the results disclose significant insights,
as far as it concerns the factors affecting the level of non-performing loans
globally, the quality of loan portfolios and generally the brittleness of banks
and the banking system as a whole.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

But in order for us to examine
and analyze all the related studies concerning the factors that affect the
appearance and growth of NPL’s, we should first make a brief analysis of
important definitions, such as the credit risk, the Basel Accord and the
non-performing loans, in order to help us deeply understand them and move on to
our study.

 

Credit Risk Assessment

 

Credit risk represents
one of the most critical risks associated with the banking sector, as it
entails the probability of a borrower not fulfilling the stipulations of their
loan contract. In this sense, default risk assessment is of paramount
importance for any financial institution, while it has direct impact on the
bank’s overall performance. The vital role of this risk assessment can be
identified though the Basel Accord, which is a recommendation framework of
banking bylaws and regulatory agreements, consisting of three different
evolutions of Basel Accord.

 

Basel was firstly
introduced in 1988 and it had recognized the importance of credit risk
assessment, proposing common rules for banks, which are highly correlated with
the default risk probability of a loan. The main propose of Basel I was to
strengthen the stability of international banking system and setup a fair and
consistent system, in order to decrease competiveness among banks. Thus, the
Basel I Accord imposed that banks must define their capital, core and
supplementary, determine the risk weighting of their assets, by place them in
five categories, from zero risk to high risk (0%, 10%, 20%, 50% and 100%), and finally
achieve capital adequacy by maintaining the minimum level of 8% in their capital
reserves. In 2004, Basel Accord has evolved, presenting Basel II or Revised
Capital Framework and its transformations; the introduction of the three
mutually reinforcing pillars. Firstly, the minimum requirements of bank’s own
capital, or else its capital adequacy ratio, must still be at least 8%, but its
assets to be weighted not only according to credit risk, as it was in Basel I, but
also according to market and operational risk. Secondly, supervisory process of
the bank’s activity must include internal assessment of equity, while
supervisory authority should review the assessment and a run rapid intervention
to maintain and prevent any decline of capital.

Lastly, the compulsory
use of disclosure for strengthening market discipline. As far as it concerns
the credit risk, according to Masuma Mehta (2017), the Basel II Accord proposes
three implementation options: The Standardized Approach, where external ratings
are used by the bank to define risk weights, similarly to Basel I, but with
different asset weightings to reduce capital requirements. The Foundation
Internal Rating Based approach, which allows a bank to use an internal rating
system, but in case of insolvency condition the recorded losses are given by
the supervisory institution. Finally, the Advanced Internal Rating Based
approach according to which a financial institution calculates its capital
requirements based on internal models, with the approval of the supervisory
institution. The latest Basel Accord, Basel III, was inducted in 2010 and it
reformed the global regulatory standards by setting stricter regulations on
capital ratios than the previous versions of Basel agreements, due to the
imminent European financial crisis. More specifically, it requires banks to
hold 4.5% of Common Equity Tier I (4% in Basel II) and 6% of Tier I capital of
risk-weighted-assets (RWA) (4% in Basel II), with the total capital adequacy
ratio remaining stable at 8%. The third accord also imposed a leverage ratio
capital buffer of 3% and introduced a new type of risk that must be taken under
consideration, the market liquidity risk with short term and long term
liquidity ratios, calculated with the help of Tier I, core capital, and Tier
II, supplementary capital, respectively. The impact of Basel III was essential
for the banking sector, as it putted more pressure on banks due to the increased
liquidity and capital costs; yet it created incentives for them to improve their
operating process in order to have lower costs and higher efficiency.

 

As of today,
institutions can access a wide range of methods for assessing credit risk, with
direct impact on their capital adequacy ratios. Approaches based on internal
rating models, as introduced by Basel II, allow banks to use their own methods
to quantify credit risk, essential to the risk-weighting of their assets and
therefore to the measuring of their capital requirements. Regulatory framework
specifies that banks should hold at least 8% of their risk-weighted assets as
reserve capital and underlines that different kinds of assets are weighted
corresponding to their conceivable risk. For example, shipping loans, have a
risk weighting of 100%, in contrast with insured mortgages that have only 20%
or even better cash that has 0% risk weighting.

 

NPLs Assessment

 

As a result of the raising regulatory
supervision of the Basel Framework, after the recognition of the importance of
the credit risk assessment, banks had to deal with the recognition, measurement
and therefore reduction of the level of their NPL’s in order to reduce their
credit risk and hold the required amount of capital. According to European
Central Bank, a brief definition of non-performing loan (NPL), or “bad debt” as
it is also called, is that a bank loan is considered as a non-performing when
more than 90 days pass without the borrower paying the agreed instalments or
interest. This unpleasant event for the bank could often happen when a borrower
experiences unforeseen financial difficulties, such as when a company faces serious
income difficulties, or when an individual bank customer loses cannot repay
their consumer loan as agreed for their personal reasons. Of course there is
the unpleasant possibility that, the borrower cannot fully repay the provided loan
and the bank needs to re-value the loan on its balance sheet, or even often
“write off” the loan. The European Central Bank is responsible for addressing
the non-performing loans within the European banking system.

 

Generally, the NPLs issue is of a paramount
importance, not only for an individual bank but for the whole European and
global banking system. Performing loans provide bank with the needed income to
make profit and grant new loans, while when a loan become non-performing the
bank must set aside more capital to cover the probability of the loan will not
be paid back. This reduces its capacity to provide new loans, so in order to be
successful in the long run, banks need to keep the level of NPLs at a low
level. Thus, if a bank has a great number of NPLs on its balance sheet, it will
face profitability and liquidity problems, for the reason that it will no
longer earn enough money from its credit business and it will need to increase
the required capital for regulatory purposes in case it needs to write off the loan.

 

Determinants of NPLs

 

The study and assessment of non-performing
loans, as well the determination of the factors that affect their growth has
become of paramount importance the last few decades, especially after the
financial and therefore banking crisis of 2008. Previous studies globally have
recognized two different sets of factors that affect the amount of NPLs over
time, the bank specific or internal factors and the macroeconomic specific or
external factors. Most of the empirical evidence supports that a combination of
both sets of determinants influence the level of NPLs in banks. The following
section will discuss briefly those internal and external factors that affect
NPLs, according to an overtime literature overview.

 

The literature begins in the middle of 1980’s,
with the introduction of some theoretical models and the macroeconomic factors
that could affect the level of NPLs. So, the theoretical models of King and
Plosser (1984), Williamson (1987) and Bernanke and Gertler (1989) highlighted
the strong relationship between business cycle phases and banking stability. As
expected, the economic phase of growth is described by a low level of non-performing
loans, as borrowers have the needed income to repay their debts on time, while,
on the contrary, during recession period, an increase in bad debts is noted, as
a result of high unemployment rates, mitigation of disposable income and
overall increased difficulties in paying back debt. The aforementioned studies pointed
out a strong negative relationship between macroeconomic determinants and bad
loans.

 

In 1987, Keeton and Morris presented a pioneer
study concerning the macroeconomic determinants of loan losses, using a sample
of approximately 2,500 USA banks for the 1979-1985 period, comparing the net
charge off rate with loan losses. They concluded to the adverse regional
economic and industry-specific conditions, as the main causes for loan losses
diversification, along with other minor reasons.

 

Lawrence (1995) was firstly to introduced the
probability of default in the life-cycle consumption model and that GDP,
unemployment rate and interest rate are the main factors affecting NPLs. In
this sense, a borrower with limited income usually has higher possibility to
default, as he/she has higher probability of risk of being unable to repay the
debt, due to unemployment or other reasons, while also, in equilibrium, the banks
charge higher rates to those debtors.

 

Gambera (2000) tried to detect an existing
relationship between macroeconomic determinants and bank failures, using
quarterly bank data of US past due loans for the period 1987-1999, in both
simple linear regression and VAR estimation model. The results showed that
national and regional macroeconomic factors can be often used as variables for
safe NPLs’ predictions; variables like bankruptcy listings, farm income (in
Primary Sector economies), state annual product, housing permits, unemployment.

 

The analysis of the linkages between macro-financial
vulnerabilities and non-performing loans is also applied by Nkusu (2011), in a sapper
working paper on behalf of International Monetary Fund; using two different,
but supplementary, approaches to achieve it. The sample dated from 1998 to 2009,
including global data for 26 advanced economies, such as Australia,
Switzerland, Eurozone countries and US, while the data for NPL variable is
modeled at the macroeconomic level from the consolidated balance sheet of each
country’s banking sector. In the first approach, the authors run several panel
regressions, like OLS, PCSE and GMM estimation method, suggesting that adverse
macroeconomic developments (GDP, unemployment and asset prices rates) are
associated with the increasing level of NPLs; while in the second one, they run
a panel vector autoregressive (PVAR) model, resulting in that the impulse
response functions (IRFs) describe the NPLs’ crucial connecting role of credit market
frictions and macroeconomic exposure. Lastly, they noted that a favorable
macroeconomic environment is linked with lowered level of NPLs, as it was
observed by the IMF in the run up to the 2008 crisis, while asset quality
inclines to strengthen the business cycle, ending up procyclical.

 

Louzis et al. (2012) scrutinized the
macroeconomic and the bank-specific determinants of Greek non-performing loans,
using panel data of the 9 largest banks in Greek banking system during the
period of economic development until the early years of recession, that is
2003-2009. In this study, the dependent variable (NPLs) has been examined
separately in three different categories, mortgage, business and consumer loans
respectively, in order to investigate if the different independent variables
have various effects among the different loan categories. The estimations
resulted in that, for all loan categories, Greek NPLs can be interpreted mostly
by macroeconomic variables, and specifically by GDP growth, unemployment,
interest rate and public debt. As far as it concerns the bank-specific
indicators, performance and efficiency have proven to be quite significant,
suggesting that regulators should concentrate on management’s quality and
performance for improving financial system’s stability. It is worth mentioning,
that there are obvious differences in the quantitative influence of macro-specific
determinants among loan categories, with mortgages’ NPLs being the least sensitive
to macroeconomic changes.

 

In 2013 Jouini and Messai researched the indicators
influencing the level of non-performing loans of the three major Mediterranean countries,
which namely are Italy, Greece and Spain and were facing major financial and
banking crisis, like mortgage and debt problems, in the beginning of the
recession in 2008. The sample of this model has been formed of 85 banks coming
from the above mentioned countries for six macroeconomic and bank-specific
variables accordingly, for the 2004-2008 period, using the method of Fixed
Effects estimation of panel data. The macroeconomic factors added in are the
rate of growth of GDP, the unemployment rate and the real interest rate, while
the specific ones for bank are the ROA ratio, the change in loans and the loan
loss reserves to total loans ratio (LLR/TL). Overall, the authors suggested
that bad loans correlate negatively with the GDP and ROA ratio, yet positively
with the unemployment rate, the loan loss reserves to total loans and the real
interest rate, underlining the importance of impaired loans during the period
of recession that these countries are facing the years after the examined
period.

 

Curak et al. (2013) presented an empirical
study about the determining factors, both internal as well as external ones, of
non-performing loans in banks of the Southeastern Europe, driven by fact that the
growth of NPLs implies an unfavorable impact of credit risk for financial
stability of banking system and the growth of economy in general. The data-set
for this model emanated from 69 banks in 10 countries for the 2003-2010 period
and the method used by the authors was the GMM difference estimator for dynamic
panel models. The results showed that lower economic growth and higher inflation
and interest rate have positive relationship with non-performing loans, while
also credit risk is affected by the bank-specific variables of bank size, ROA
ratio and solvency ratio.

 

Bellas et al. (2014) firstly introduced a study
similar to ours, striving to identify the NPLs’ determinants of Eurozone’s
banking system for a nine-year pre-crisis period of economic growth and
stability (2000-2008), as a preliminary attempt to query the proper functioning
of the European and Eurozone banking system in a whole. So, they performed the
GMM difference estimation method, using an unbalanced panel data of 14
countries with 120 observations in total for the aforementioned period. The
independent variables that are included in this econometric model are macro-specific,
as the GDP growth rate, the public debt as percentage of GDP, the unemployment
rate and the inflation rate and the bank-specific ones, as the loans to
deposits ratio, CA ratio, the ROE and ROA ratios. The final results of this
research denoted a strong correlation between the total level of NPLs and the
following macro and micro factors, public debt, unemployment, GDP, capital
adequacy ratio, ROE and the rate of NPLs of the previous year, proving that the
loan portfolio quality is an integral part of Eurozone’s economy.

 

In 2015, Polodoo et al., presented an
econometric analysis concerning the bank-specific and the macroeconomic factors
affecting Mauritian non-performing loans, using panel data from 10 existing
banks for the years 2000-2012. The variables included in this model were the
inflation rate, lending interest rates, growth of construction and tourism sector,
along with global variables such as the Euro zone’s GDP growth, while the
estimation techniques used were the Fixed Effects, the Random Effects, and the
differenced and System GMM, both in a static and dynamic framework. The findings
reveal the existence of a number of significant variables influencing NPL, in
which the decrease in the construction sector and the increase in cross border
loans proven to be the most crucial ones, yet the global indicators proven to
be statistically insignificant, indicating Mauritius’ elasticity in external
financial shocks.

 

According to Nikolov and Popovska (2016), non-performing
loans are one of the most vulnerable categories in the balance sheet of banks,
because their increase can affect
banks’ liquidity and solvency and this fact is of paramount importance for bank’s
performance and the financial system generally. So, they tried to investigate the significant bank-specific
and macro-specific determinants that define the NPLs growth in FYROM for the
2006-2015 period, with both descriptive and econometric analyses. The first
model presents the relationship between NPLs and only two macroeconomic
indicators, GDP growth and inflation rate, while the second one, the econometric
analysis, presents the correlation between NPLs and two bank indicators, the CA
and ROE ratios, including also the macro variable of inflation. The results
indicated that during periods with economic growth and high inflation, NPLs
remain constant and in low level, yet the raise of CA and ROE ratios mitigate
the level of NPLs.

 

Kjosevski and Petkovski (2016) analyzed the
relationship of macro-specific and bank-specific determinants of NPLs and
therefore their affection on the macroeconomic prosperity in Baltic Region, with
two complementary approaches. The first model presented external and internal
variables affecting NPLs with panel data from 27 Baltic banks in annual basis
for the period 2005–2014, indicating GDP growth, inflation, domestic credit to
the private sector, shareholder’s equity ratio, ROA, ROE and the growth of
gross loans to be the most significant factors. The other model scrutinized the
relationship between NPLs and its macroeconomic factors only, with the results
indicating that the real economy responds to NPLs and that there are strong
feedback effects from macroeconomic conditions, such as domestic credit to
private sector, GDP growth, unemployment and inflation rates.

 

Khan and Ahmad (2016)

 

Overall, the empirical academic literature indicates
a strong relationship between the NPLs and a number of macroeconomic
determinants, such as the GDP and the inflation growth in annual basis, the
real long-term interest rate and the real exchange rate, the loans growth, the
unemployment rate, the money supply (M2), and others. On the other hand, there
are also some bank specific indicators that affect the level of NPLs, such as