Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Three essays on the credit growth and banking structure of central and eastern European countries
(USC Thesis Other)
Three essays on the credit growth and banking structure of central and eastern European countries
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THREE ESSAYS ON THE CREDIT GROWTH AND BANKING STRUCTURE
OF CENTRAL AND EASTERN EUROPEAN COUNTRIES
by
Burcu Aydin
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSTIY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ECONOMICS)
August 2007
Copyright 2007 Burcu Aydin
i
Dedication
To
My parents Nesibe and Husamettin
And my brother Mirkan
ii
Acknowledgments
I would like to thank my advisors; Robert Dekle, John Ham, and Cheng
Hsiao for making valuable comments and suggestions. They regularly provided me
with encouragement and guidance for my research.
Also, I would like to thank Selahattin Imrohoroglu for accepting the role of
outside member of my dissertation committee, and for providing his valuable
comments throughout the writing of this dissertation.
I am greatly indebted to Wim Fonteyne from the International Monetary
Fund. It would not have been possible to write my thesis without his support,
guidance, and precious comments.
I would like to thank the participants in the seminars held at the International
Monetary Fund, Europe Department in July 2006, and the University of Southern
California in September 2006.
Last, I am grateful to my family for always supporting me. It would not have
been possible to complete this work, or any of my achievements, without feeling
their love and their faith in me.
iii
Table of Contents
Dedication .....................................................................................................................i
Acknowledgments........................................................................................................ii
List of Tables ...............................................................................................................v
List of Figures ............................................................................................................vii
Abbreviations ............................................................................................................viii
Abstract ........................................................................................................................x
Chapter 1: Introduction................................................................................................1
Chapter 2: Literature Review.....................................................................................12
Chapter 3: The Impact of Foreign Bank Ownership on Credit Growth in Central
and Eastern European Countries .......................................................................15
3.1 Stylized facts about Eastern European Countries ............................................19
3.2 Data ..................................................................................................................23
3.3 Empirical Model...............................................................................................31
3.3.1 Concerns about Endogeneity of the Foreign Ownership Variable............36
3.3.1 Concerns about Multicollinearity..............................................................40
3.4 Results..............................................................................................................42
3.4.1 Impact of Microeconomic Shocks ............................................................48
3.4.2 Impact of Macroeconomic Shocks............................................................54
3.6 Conclusion .......................................................................................................57
Chapter 4: The Managerial Impact of Parent Banks on their Affiliated Banks
Operating in Foreign Countries: A Case Study for the Central and Eastern
European Countries...........................................................................................59
4.1 Data ..................................................................................................................65
4.2 Empirical Model...............................................................................................73
4.2.1 Concerns about Multicollinearity..............................................................77
4.2.2 Do Parent Banks have different Managerial Strategies?...........................79
4.3 Results..............................................................................................................83
iv
4.4.1 Impact of Microeconomic Shocks ............................................................93
4.4.2 Impact of Macroeconomic Shocks............................................................97
4.4 Conclusion .....................................................................................................100
Chapter 5: The Lessons from the EMU Countries...................................................101
5.1 Are there any similarities between the CEE and the EMU-5 Countries? ......105
5.2 Data ................................................................................................................108
5.3 Econometric Model........................................................................................113
5.3.1 Concerns about Multicollinearity............................................................115
5.4 Results from the EMU-5 Banking Structure..................................................117
5.4.1 Impact of Microeconomic Shocks ..........................................................121
5.4.2 Impact of Macroeconomic Shocks..........................................................122
5.5 Conclusion .....................................................................................................123
Bibliography.............................................................................................................125
v
List of Tables
Table 1: Market Share of Foreign-Owned Banks in CEE Countries.........................20
Table 2: Summary Statistics for the CEE Countries..................................................27
Table 3: Pairwise Correlation Coefficients between Credit Growth and its
Lagged Value for the CEE Banks ......................................................................32
Table 4: Panel Data Probit Estimation Results for Foreign Ownership Variable......37
Table 5: Hausman Specification Test for Foreign Ownership Dummy.....................39
Table 6: Correlation Coefficient across Regressors for the CEE Countries ..............41
Table 7: Fixed Effect Estimator Results for the CEE Countries................................43
Table 8: Fixed Effect Estimator Results for the CEE Countries................................44
Table 9: Fixed Effect Estimator Results for the CEE Countries................................45
Table 10: Cross-Ownership Varying FE Estimators for the CEE Countries.............46
Table 11 : Cross-Ownership Varying FE Estimators for the CEE Countries............47
Table 12: List of Parent Banks and their Affiliated Companies ................................67
Table 13: Summary Statistics for Parent Banks and their CEE Subsidiaries ............70
Table 14: Pairwise Correlation Coefficient between Credit Growth and its
Lagged Value for Each CEE Subsidiary............................................................74
Table 15: Correlation Coefficients across Variables of Interest ................................78
Table 16: FE regression results for cross-sectional varying coefficients...................81
Table 17: Test of Different Coefficient Estimates across Affiliated Banks...............82
vi
Table 18: FE Estimator for Credit Growth in the CEE Subsidiaries .........................85
Table 19: FE Estimator for Credit Growth in the CEE Subsidiaries .........................86
Table 20: FE Estimator for Credit Growth in the CEE Subsidiaries .........................87
Table 21: FE Estimator for Credit Growth in the CEE Subsidiaries .........................88
Table 22: FE Estimator for Credit Growth in the CEE Subsidiaries .........................89
Table 23: FE Estimator for Credit Growth in the CEE Subsidiaries .........................90
Table 24: FE Estimator for Credit Growth in the CEE Subsidiaries .........................91
Table 25: FE Estimator for Credit Growth in the CEE Subsidiaries .........................92
Table 26: Summary Statistics for EMU-5................................................................109
Table 27: Pairwise Correlation Coefficient between Credit Growth.......................114
Table 28: Correlation Coefficient across Regressors for the EMU-5 Countries .....116
Table 29: Fixed Effect Estimator Results for EMU-5 .............................................119
Table 30: Fixed Effect Estimator Results for EMU-5 .............................................120
vii
List of Figures
Figure 1: Credit as a Percentage of GDP ...................................................................22
Figure 2: Number of Banks per Ownership Category ...............................................26
Figure 3: Credit Growth across Bank Ownership Types in CEE Countries..............49
Figure 4: Comparison of the transition indicators between the CEE and EMU-5
Countries. .........................................................................................................107
Figure 5: Mean and Standard Deviation of Banking Variables across Years in
EMU Countries ................................................................................................110
viii
Abbreviations
AIC Akaike Information Criteria
AR(2) Second-order Autoregressive Process
BIC Bayesian Information Criteria
CEE Central and Eastern Europe
ECB European Central Bank
EMU Economic and Monetary Union
EMU-5 Greece, Ireland, Italy, Portugal and Spain
EU European Union
FE Fixed Effects
GAAP Generally Accepted Accounting Principles
GDP Gross Domestic Product
IFRS International Financial Reporting Standards
IV Instrumental Variable
M3 A Measure of Money Supply
MIGA Multilateral Investment Guarantee Agency
n.a. Data not available
NIM Net Interest Margin
OLS Ordinary Least Squares
r Real Interest Rate
ix
ROA Return on Average Assets
ROE Return on Average Equity
TA Total Assets
TCR Total Capital Ratio
TSLS Two-Stage Least-Squares Estimator
US United States of America
x
Abstract
This thesis is comprised of three essays on the credit growth and banking
structure of the Central and Eastern European Countries (CEE). The first essay
studies the importance of bank ownership on this growth, the second explores the
influence of the managerial impact on credit growth and the third one draws
projections for the future EMU membership of the CEE countries from the past
experiences of the five current EMU states.
The first essay studies the impact of foreign bank ownership on credit growth
for the CEE countries. The analysis is based on an unbalanced panel for the 72
largest banks from these countries. The results indicate foreign bank ownership is
significant in raising credit. Although foreign banks have higher credit growth
compared to domestic banks, further analysis shows that private domestic banks in
fact also have high credit growth. Except for their means of raising funds, foreign
and private domestic banks share similar characteristics in terms of bank
performance and efficiency, whereas significant differences exist between private
domestic and state-owned banks.
The second essay studies the managerial impact on credit growth for foreign-
owned banks operating in the CEE countries which own around 70 percent of the
total CEE banking assets. The analysis is based on an unbalanced panel of 16 years
and 59 banks, comprised of 18 parent banks and their 41 CEE subsidiaries. This
study indicates that the size of the parent bank, the flow of funds between affiliated
xi
banks, interest margins, and the macroeconomic situation in the host market play
important roles for credit growth in CEE banks owned by foreigners.
The third essay draws projections on the future Economic and Monetary
Union (EMU) membership of the CEE countries based on the experiences of the five
current member states of the EMU (EMU-5) which formerly had similar
macroeconomic dynamics to those of the CEE countries. The analysis is based on an
unbalanced panel for the 44 largest banks operating in the EMU-5 countries. The
results show that the liquidity structure of the EMU-5 banks change as well as the
macroeconomic conditions after the introduction of the Euro.
1
Chapter 1
Introduction
Around 1990 and onwards, Europe experienced major changes both
politically and economically. The most important change in Central and Eastern
European (CEE) countries was the transition from a centrally planned economy to a
market based system. During the same years Western Europe also experienced major
changes: introduction of the single market and then the move towards a common
currency area. The following decade for the CEE countries was marked by their
membership to the European Union (EU) and their next step will be the entrance to
the common currency system.
The 1990s transition in the ten CEE countries: Bulgaria, Romania, Czech
Republic, Hungary, Poland, Slovakia, Slovenia, Estonia, Latvia, and Lithuania
caused major changes in their economic environment. The CEE countries liberalized
their institutions, went through a major privatization process and opened their
markets toward a free-market economy. As a product of these privatization and
liberalization processes the CEE countries experienced an important change in their
2
financial markets, namely a significant rise in the share of foreign ownership in their
banking structure, followed by high credit growth levels in the domestic economy.
This thesis studies the economic implications of foreign bank ownership and the
credit growth pattern in the CEE countries.
The rapid credit growth in the CEE countries is of interest for two reasons.
First, domestic credit as a share of GDP has been increasing continuously in the CEE
countries from 1990 onwards and this rapid credit growth is a reflection of economic
growth and financial integration toward more developed countries. Second, rapid
credit growth, if identified as a credit boom, can have undesirable effects on an
economy - such as systematic banking crises and economic recessions
1
- and
standard economic theory does not differentiate between rapid credit growth and
credit boom.
The financial markets of the CEE countries are heavily dominated by the
banking sector and their capital markets are not well developed. Therefore the rapid
credit growth in the CEE countries will come through a rise in bank loans. The
banking system in this region experienced notable adjustments throughout the
transition period from centrally planned economies to market economies. During this
transition the CEE countries privatized a significant portion of their state-owned
banks and foreign investors, directly and indirectly, acquired the majority of banking
1
There is a high correlation between credit booms and currency or systematic banking crises
(Terrones and Mendoza (2004)).
3
assets from these sales. Additionally, during this period foreign banks also started to
open subsidiaries and branches in the CEE countries. After accounting for all the
changes in banks’ ownership structure and greenfield investments, foreign investors
became the major share holders in the CEE banking sector.
The first essay in this thesis, The Impact of Foreign Bank Ownership on
Credit Growth in Central and Eastern European Countries, studies the importance
of bank ownership structure for the growth of credit in the CEE countries. The
foreign bank ownership for the credit growth pattern of the CEE countries is of
interest for several reasons. First, foreign banks bring the know-how to the CEE
countries which increases the efficiency of credit distribution. Under the command
economy, banks operating in the CEE countries did not act as financial
intermediaries but rather as record keepers. However, throughout the transition and
after that, banks needed to engage in credit evaluation and risk management
activities. Foreign acquisition of the CEE banks channeled the network of experience
and better technology from financially advanced countries to the CEE banks. For
example, the banking system indicators - such as higher information technology,
better risk management, and credit evaluation - improved in the CEE countries
between the 1990s and the 2000s [ECB (2005A)] during the foreign acquisition of
the CEE banks. The enhanced banking system in the CEE countries provides a more
efficient credit channeling and improved financial intermediation, and hence it
increases the credit growth in the CEE countries.
4
Second, foreign banks are not restricted by domestic market conditions and
this allows the foreign banks to increase the credit line much faster than the domestic
banks. A majority of the foreign banks are owned by very large multinational banks,
and the financial needs of these CEE subsidiaries can be met by funds coming from
the parent bank. Additionally, a foreign bank has more credibility than a private
domestic bank due to the reputation of its parent bank. Therefore a foreign bank can
also raise more credit than the private domestic bank by borrowing from other banks.
The easier access to interbank lending provides more financial resources to foreign
banks, which in turn lets them make more loans.
Third, foreign banks increase the amount of competition in the CEE banking
system which leads to an outward shift in the supply of credit
2
. This shift in the
supply of credit reduces the equilibrium spread between bank lending and borrowing
rate and increases equilibrium borrowing in the country. Greater competition in the
banking system is helpful in improving financial intermediation and efficiency.
Financial intermediation in the CEE countries gets better as both the number of
banks operating in these countries and the financial services that they offer increase
3
.
The efficiency in the CEE countries’ financial markets increases as the shift in the
supply curve decreases the cost of borrowing. The increase in financial
2
The number of banks operating in the CEE countries and the share of foreign owned assets increased
from 1990 and onwards in the CEE countries.
3
M3 as a ratio of GDP is a measure of financial development, and this ratio increased between the
1990s and 2000s in the CEE countries.
5
intermediation and efficiency allows the CEE firms and households to satisfy their
financial needs via bank credit.
This essay explicitly studies the different behavior of foreign bank ownership
with respect to domestic ownership types: private and state ownership in chapter 3.
The analysis in chapter 3 uses a panel dataset covering the years 1988 to 2005 for the
largest 72 banks in the ten CEE countries. The results show that there is a structural
break in the dataset by the 2000s in terms of bank level and macroeconomic
indicators due to the transition of these CEE countries and the banking and currency
crises that happened during the 1990s. Further results show that high interest margins
and economic growth are the two driving forces for foreign banks to enter to the
CEE markets. In terms of credit growth, foreign-owned banks on average have
higher loan growth rates than domestically-owned banks. However, by the 2000s the
estimation results indicate that the average credit growth rate of foreign banks is
higher than the state owned banks, but significantly lower than that of domestically
owned private banks. The long-term customer relations between the private domestic
banks and the CEE companies may be the underlying reason for this result. In terms
of funding the bank loans, foreign owned banks show significant differences from
domestic ownership types. By the 2000s, interbank lending becomes the major
source of funding for foreign owned banks, whereas domestically owned banks still
rely on domestic market conditions. By the 2000s, cost and efficiency indicators lose
importance on the growth of credit for private banks, but still remain important for
6
state owned banks. The insignificance of these indicators for private banks shows the
importance of the ‘know-how’ effect that foreign banks bring to the CEE countries.
The second essay, The Managerial Impact of Parent Banks on their Affiliated
Banks Operating in Foreign Countries: A Case Study for the Central and Eastern
European Countries, analyzes the impact of parent banks on the credit growth
pattern of their CEE subsidiaries. The transition in the CEE banking system and the
change in bank ownership caused the majority of the banks to move from domestic
to foreign ownership by late the 1990s and 2000s
4
. After the change in bank
ownership, the majority of the foreign owned CEE banks were acquired by Western
Europeans who are in close geographical proximity to the CEE subsidiaries. For
example; Austrian, German, and Italian banks acquired the banks in central Europe,
and Scandinavian banks owned the banks operating in Baltic countries. This high
degree of foreign bank ownership in the CEE banking system motivates this
investigation of the impact of the parent banks on the credit growth pattern of the
CEE countries. The interest for the fourth chapter of this thesis is based on the
following factors.
First, the credit growth pattern in the CEE countries may be affected by the
financial situation of the controlling bank or by the domestic market conditions that
the parent bank faces. A majority of these parent banks own several banks operating
4
In the CEE countries, around 70 percent of the total banking assets are controlled by foreigners the
2000s (ECB, eurostat database).
7
in various CEE countries. Therefore a change in the firm level or market level
conditions that these parent banks face may cause a change in the credit growth
pattern of their subsidiaries, which constitute the majority of the banking assets in
the CEE countries. The profitability of the parent bank has an influence on the
magnitude of operations that its branches will engage. A strand of the economic
theory suggests that a financial downturn may force a multinational bank to reduce
its operations company wide. For the macroeconomic conditions that the parent bank
faces at its headquarters or home country, a branch of economic theory suggests that
an economic downturn in the home market will cause the multinational bank to
reduce its operations company wide
5
. According to this strand of theory, the credit
line at the CEE subsidiaries may decline if the financial situation of the parent bank
and/or the macroeconomic conditions in the home country worsens.
On the other hand, the financial diversification theory suggests that the
profitability of the subsidiary or the growth prospects in the CEE countries are more
important for the multinational bank’s decision whether or not to extend the credit
line in the CEE countries. The profitability of the subsidiary will motivate the parent
bank to extend its operations at the subsidiary when the multinational company’s
prospects at home are declining. For the host country’s macroeconomic conditions,
the regional or global diversification policy of these multinational banks suggests
that the multinational banks increase their operations in regions with high economic
5
De Haas and Lelyveld (2003) discuss that the macroeconomic conditions that the parent bank faces
plays and important role in reducing the credit line of the subsidiaries operating abroad.
8
growth potentials. Low growth rates in the home market may stimulate the parent
banks to extend their operations in the CEE countries. In the future, a falling growth
rate in the CEE countries due to economic convergence or low interest margins due
to increased competition may lead the parent bank to limit their operations in the
CEE markets.
Secondly, all of the CEE countries are targeting a join with the EMU in the
near future. Both the macroeconomic indicators and the institutional framework in
these countries are being aligned to meet the EMU standards. Therefore the
European monetary policy, in particular the Euro area, has an important impact on
the credit growth pattern of the CEE countries. Due to the financial integration of the
CEE countries towards the Economic and Monetary Union (EMU) the spread
between domestic and Euro area interest rates lose importance over time and the
economic integration of the CEE countries increases the importance of EMU on the
CEE macroeconomic indicators. The EMU monetary policy affects the credit growth
in the CEE country and this impact is sounder in the 2000s. A monetary expansion in
the CEE countries (an increase in M3) and a decline in interest rates may cause a rise
in the CEE credit growth.
The analysis in this chapter is based on a panel dataset covering the period
1990 to 2005 for 59 banks, comprised of 18 parent banks and their 41 CEE
subsidiaries. The results indicate that the size of the parent bank matters for the
growth of credit: CEE banks can increase their credit line further as the parent bank
gets larger. Highlighting the impact of inter-bank lending, interbank liabilities of the
9
parent bank are significant and negative for the credit growth of foreign owned banks
operating in the CEE countries. This may indicate that parent banks may need to
borrow externally to finance the credit supply of its subsidiaries. The spreads in the
home market also influences the growth pattern of credit. As home market spreads
decline, the parent banks increase their CEE lending. Finally, Euro area monetary
policy has an impact on the CEE credit growth in terms of interest rate policy and
money growth: a monetary expansion in the EMU is followed by a rise in credit
growth in the CEE countries.
The third essay in this thesis is Lessons from the EMU Countries. This essay
is provided in chapter 4, and it draws projections on the EMU membership of the
CEE countries based on the experiences of the five current member states of the
EMU; Greece, Ireland, Italy, Spain, and Portugal (EMU-5). The post transition
period, the 2000s, was marked by the EU membership of the CEE countries. The
next transition for the CEE countries will be their membership to the monetary
union. On the way to their next destination, i.e. the monetary union, the CEE
countries have been experiencing similar macroeconomic conditions as that of the
EMU-5. The CEE countries have been converging from high inflation and interest
rates to the EMU level; they have been running current account deficits,
experiencing capital inflows, rising asset prices and robust domestic demand growth.
These macro conditions are very similar to those experienced by the EMU-5
countries before entering the euro area. Most importantly, the current credit growth
10
pattern in the CEE countries resembles that of the EMU-5 before their entrance to
the common currency area.
After the introduction of the Euro, domestic demand and economic growth
declined sharply in the EMU-5 markets which ended up in a credit crunch for all the
EMU-5 countries except Ireland a few years after joining the monetary union
6
. Even
though the economic growth slowed down in these countries, the credit crunch did
not cause any systematic risks to the banking sector. Considering the possibility that
the CEE countries may experience a similar stagnation in the macroeconomic
environment, this chapter draws projections for the CEE countries from the credit
growth pattern of the EMU-5 countries before and after their transition to the
monetary union. Even though the banking structure in the EMU-5 and CEE markets
have differences, a road map for the CEE countries can potentially be prepared by
drawing lessons from the past experiences of the EMU-5 countries and the transition
in their banking system.
This chapter uses a panel data from the banking and macroeonomic data of
the EMU-5 countries. The dataset covers banking sector indicators from the 44
largest banks operating in EMU-5 countries from 1988 to 2005. The paper studies
the data for two periods: 1988-2000 and 2001 to 2005. The time break in the dataset
is due to the change in economic environment in the EMU-5 countries before and
6
Italy and Portugal were in Euro Trap after the second year of joining the monetary union, and
further, Spain and Greece show the symptoms of getting caught in the same undesirable path. The
symptoms of the Euro trap are the loss of competitiveness, slow economic growth or recession,
increasing budget deficits and worsening capital account deficits (Mayer (2006)).
11
after the introduction of the Euro. The results from this chapter show that the
downturn of the economic boom in the 2000s caused a change in the domestic saving
behavior in the EMU-5 markets i.e.; savings increased and lending decreased. Even
though a majority of the banking indicators are pro cyclical, the change in the
economic environment and the saving behavior did not cause an increase in the
amount of risk EMU-5 banks were facing in the 2000s.
12
Chapter 2
Literature Review
The literature has thus far paid little attention to foreign bank ownership and
credit growth. Many articles focus either on the credit growth side or on the effects
of foreign ownership on banking performance. On the credit growth side, many
papers focus on the possible implications of such growth but not necessarily on the
role of bank ownership in this growth. Lambregts and Ottens (2006) study the causes
of banking crises in emerging market economies by using a multivariate logit model.
This work presents that a developing economy experiencing rapid lending growth
will face potential banking system fragilities. Cottarelli et al (2003) focus in
particular on Central and Eastern Europe in studying the risk structure of rapid credit
growth. In this paper, they estimate an equilibrium level of bank-credit-to-GDP ratio
in order to evaluate whether these developments in Central and Eastern Europe are
part of a convergence and financial deepening process and they present results that
are consistent with this hypothesis. Kraft and Jankov (2005) study the impact of
lending booms for banking and currency crises in Croatia. They argue that rapid loan
13
growth increases the probability of deteriorating credit quality and problems with the
current account and foreign debt.
The literature analyzing the impact of bank ownership on performance is
more extensive. Micco et al (2004) study the impact of foreign bank ownership for
119 countries covering both developing and industrialized countries, using a panel
data analysis. In their study the authors show that bank ownership is correlated with
bank performance only in developing countries and that the relationship is as
follows: state-owned banks are the least efficient and foreign bank entry to less
developed markets is important in improving efficiency. De Nicolo and Loukoianova
(2006) question the impact of bank ownership on the risk profiles of banks. They
introduce a banking industry equilibrium model with heterogeneous banks, which
predicts changes in risk profiles for different bank ownership types when there is a
change in market structure. They then implement their analysis empirically with
panel data on individual banks of 133 non-industrialized countries. In their
preliminary results they show that the risk profiles of different types are significantly
different across market structures and in particular, private domestic banks take more
risk if the markets are highly concentrated.
The literature on the implications of foreign bank ownership for credit growth
in the Central and Eastern European countries is quite limited. In this under-explored
area, De Haas and Lelyveld (2002) study foreign bank ownership for five of the
Central European economies from 1993 to 2000, looking at the implications for
private sector credit. They use quantitative data analysis, via interpretation of the
banking data presented in graphs and tables, to conclude that there is a positive
14
relationship between foreign banks and credit growth. In their 2003 paper, the same
authors analyze credit stability in Eastern Europe during business cycles by using a
panel data approach. In this paper, the authors use the same dataset as their 2002
paper and show that foreign banks do not cut credit in their host country during the
1990s business downturn, unlike domestic banks. However, foreign banks, for the
same period, reduce credit in the host country when their home country is in an
economic downturn.
15
Chapter 3
The Impact of Foreign Bank Ownership on Credit Growth
in Central and Eastern European Countries
This paper studies the economic implications of foreign bank ownership for
ten of the most recent members of the European Union: Bulgaria, Romania, Czech
Republic, Hungary, Poland, Slovakia, Slovenia, Estonia, Latvia, and Lithuania.
These countries had similar changeovers in their financial markets and
macroeconomic indicators during and after their transition from a centrally planned
to a market-based economy. In this shift, these post-command market countries
liberalized their institutions, went through a major privatization process, and opened
up their markets toward a free-market economy. As a product of these privatization
and liberalization processes, the new EU member states had experienced an
important change in their financial markets - a significant rise in the share of foreign
ownership in their banking structure - followed by high credit growth levels in the
domestic economy. This paper focuses on these two peculiar transitions.
The financial markets of the CEE countries are heavily dominated by the
banking sector and therefore the rapid credit growth in the CEE countries can be
16
explained by a rise in bank loans. The banking system in this region experienced
notable adjustments throughout the transition period from centrally planned to
market economies. During this transition phase, the CEE countries privatized a
significant portion of their state-owned banks and foreign investors, directly and
indirectly, acquired a majority of the total banking assets from these sales.
Additionally during this period, foreigners also started to open up subsidiaries and
branches in the CEE countries. After accounting for all the changes in banks’
ownership structure and greenfield investments, foreign investors achieved a market
share of 50 to almost 100 percent by 2005 in all of the CEE countries except for
Slovenia. This paper analyzes the impact of foreign bank ownership on the credit
growth in the CEE countries provided by the banking sector.
The analysis uses a panel dataset covering 18 years (1988 to 2005) for the
largest 72 banks of the ten CEE countries. These banks account for at least 60
percent of the total banking assets in all of the countries in the sample. After a
detailed study of the data, including possible endogeneity and multicollinearity
problems, a fixed effect within estimator model is employed for the econometric
analysis.
The results indicate that the banking systems in the CEE countries exhibit
significant differences over time in terms of profitability, efficiency, and liquidity
ratios. The sign and the significance level for a majority of micro and macro level
variables change between the 1990s and the 2000s. The reason for such a change is
related to the transition of these countries from centrally planned economies to a
17
market oriented system, and the banking and currency crises that happened during
this period.
In the CEE countries, foreign-owned banks have on average higher credit
growth rates than domestically-owned banks. However, after controlling for private
domestically owned banks, the estimation results indicate that the credit growth rate
of foreign-owned banks operating in the CEE countries is significantly lower than
that of domestically owned private banks by the year 2000. The long-term relations
of the private domestically owned banks with the domestic companies may be the
underlying reason for this result.
Further results show that the sources for funding bank loans changed over
time and across ownership types, in particular, after the 2000s. In the earlier periods,
foreign banks acted like private domestically owned banks, and customer deposits
constituted a very important source of funding for them to raise the credit line.
However, over time, foreign owned banks started to rely on lending from their parent
banks and/or other big banks in order to sustain their financial needs for making
bank loans. Hence the domestic market funding lost its restrictiveness for sustaining
the financial needs of foreign owned banks.
In terms of profitability, the entrance of foreign banks increased the amount
of competition in the CEE markets, and hence caused a decline in the profit margins.
However, the insignificance of the cost and efficiency variables on the growth of
credit for privately owned CEE banks shows that the foreign banks improved the
18
efficiency in the CEE markets by bringing in the ‘know-how’ from their home
countries. In this period, the main difference - in terms of changes in credit growth in
reaction to cost and efficiency variables - is between the state-owned and private-
owned banks.
Another important result of the paper concerns the domestic market
conditions. The paper shows that high economic growth and high domestic interest
rates are the underlying reason for foreign banks to extend their operations to the
CEE markets. The paper shows that foreign bank lending is significantly procyclical
by the second half of the dataset. This result shows that during a prosperous
economic environment foreign banks can increase credit more than domestic banks
through either external borrowing from the parent bank, or from other banks -
relying on the reputation of the parent bank.
The paper shows that the CEE markets are experiencing financial deepening
and economic integration into the Euro area in particular during the second half of
the dataset. The decline in the spread between the bank lending and borrowing rate,
and the spread between the domestic and Euro area interest rate, is influential in the
credit growth of the CEE countries.
This chapter of the thesis has the following structure. The following section
provides the characteristics of the transition in the CEE countries, and the
implications of such a transition to the CEE banking system. Section 3.2 Data
provides the data analysis. Section 3.3 Empirical Model outlines the panel data
19
estimation model. Further, this section discusses the peculiarities of the
microeconomic dataset, and the problems which may arise regarding the
endogenieity and multicollinearity issues. The results of the regression analysis
regarding the impact of the microeconomic and macroeconomic variables on the
credit growth pattern of the CEE counties, and the impact of foreign ownership on
this pattern, are provided in section 3.4 Results. The last section is 3.6 Conclusion.
3.1 Stylized facts about Eastern European Countries
The financial markets of the CEE countries experienced major structural
changes from the early 1990s onward as these countries started to open up their
economies internationally. Under communism, banks were designed to satisfy the
financial needs of a centrally planned economy. There were very few but very big
state-owned banks that used to perform some specific functions, such as financing
the agricultural sector. Further, state savings banks were a place for households to
deposit their savings.
During the transition from centrally planned to market economies, these
countries established a two-tier banking system: one central bank
7
and many other
individual commercial or specialty banks. In this transition period, banking crises
caused many of the small private banks (which were set up in the early 1990s) and
the state-owned banks to go bankrupt. The Central and Eastern European countries
began privatizing their state-owned banks and foreign investors acquired a majority
of the banking assets during these public offers. In addition to buying shares from
domestic banks, foreigners also started to open up greenfield subsidiaries and
branches.
Table 1: Market Share of Foreign-Owned Banks in CEE Countries
Foreign Branches and Subsidiaries % Market Share
2000 2001 2002 2003 2004 2005
Slovenia n.a. 15.0 16.6 18.5 20.0 22.5
Latvia n.a. 45.2 41.9 45.7 48.6 58.5
Hungary n.a. 59.21 60.1 56.7 59.0 58.8
Poland n.a. 68.9 67.3 67.7 67.7 67.1
Lithuania n.a. 75.6 84.4 84.1 83.9 84.1
Czech Republic n.a. 77.1 93.2 96.0 96.1 93.4
Slovakia 85.0 92.4 95.6 96.3 96.7 99.5
Estonia n.a. 97.9 97.4 97.4 98.1 99.2
Source: ECB
Table 1 provides information on the market share of foreign-owned banks in
eight of the ten CEE countries for the most recent years
8
. Looking at this table, one
7
The central banks established during the transition period of the CEE countries had a similar
structure as that of the central banks in developed countries. The CEE central banks were responsible
of pursuing the monetary policy, and supervising and monitoring the banking system.
8
Market share data for foreign bank ownership is not available for Bulgaria and Romania.
20
21
can see that foreign companies have come to dominate in the banking sectors of all
these countries, except for Slovenia, such that their banking assets cover more than
half of the total banking assets in these countries.
In the run-up to EU membership and since then, the CEE countries have been
experiencing rapid credit growth, current account deficits, capital inflows, rising
asset prices, and robust domestic demand growth. Figure 1 provides a cross-country
comparison of credit as a percentage of GDP for the CEE countries. The patterns in
this figure reflect the transition of these ten Central and Eastern European countries
from centrally-planned to market economies. As an example, the sharp credit drop in
Estonia is concurrent to its independence from the Soviet Union.
The earlier years in most of these countries are marked by high credit ratios.
The sharp drop from high Domestic-Credit-to-GDP rates to significantly lower ratios
is attributed to hyperinflation and banking crises in the 1990s
9
. As an example, in
1997 Bulgaria experienced a sharp drop in its Domestic-Credit-to-GDP ratio from
100 percent to 20 percent due to the skyrocketing inflation rates of more than 1000%
in 1997, in addition to the 1996-97 banking crises. High inflation rates may cause
drastic changes in Domestic-Credit-to-GDP ratios, since GDP is a flow variable and
credit is a stock variable which is measured at the end of the year. A hyperinflation
experienced during the year will inflate GDP much more than the end of year
9
As stated in Caprio and Klingebiel (2002), and De Haas and Van Lelyveld (2006) the banking crises
years for the CEE countries are as following: Czech Republic 1993-97, Estonia 1992-95, Hungary
1991-95, Latvia 1995-97, Lithuania 1996-96, Poland 1991-95, Romania 1998-99, Slovak Republic
1996-2000 and Slovenia 1992-94.
inflation rate inflating the domestic credit. Hence notable increases in inflation
during the year, reflects as a rise of the denominator, causing sharp declines in
Domestic-Credit-to-GDP ratios.
Figure 1: Credit as a Percentage of GDP
Source: International Financial Statistics, World Economic Outlook and Eurostat.
Eastern European Countries
0
20
40
60
80
100
120
140
1988 1990 1992 1994 1996 1998 2000 2002 2004
Bulgaria Romania
Baltic Countries
0
10
20
30
40
50
60
70
80
1988 1990 1992 1994 1996 1998 2000 2002 2004
Estonia Latvia Lithuania
Central European Countries
0
10
20
30
40
50
60
70
1988 1990 1992 1994 1996 1998 2000 2002 2004
Poland Slovenia
Central European Countries
0
20
40
60
80
100
120
1988 1990 1992 1994 1996 1998 2000 2002 2004
Czech Republic Hungary Slovak Republic
Moving to the second half of the 1990s, for most of the Central and Eastern
European countries, the trend in domestic credit as a percentage of GDP changes
significantly. These countries exhibit an upward trend in the Domestic-Credit-to-
GDP ratios. Macroeconomic stabilization, banking system regulations and better
performing financial systems, in part due to foreign entry, fueled a rapid rise in their
22
23
Domestic-Credit-to-GDP ratios. The Baltic countries, in particular Estonia and
Latvia, are the leaders in this rapid growth experience.
3.2 Data
The analysis in this paper is based on an unbalanced panel dataset comprised
of annual macro and micro level data. The macroeconomic dataset is obtained from
three main sources: International Financial Statistics, World Economic Outlook, and
Eurostat. Alternatively, the micro level data is obtained from; BankScope, the
Banker’s Almanac, Privatization Barometer, MIGA, the World Bank, European
central banks, and individual bank’s web sites.
The balance sheet data was gathered from BankScope for the largest banks of
that country
10
. The number of banks covered varies per country and the criterion in
this decision is to have as many banks as to account for at least 60 percent of the
country’s total banking sector assets
11
. In BankScope, balance sheet information is
reported under two different accounting standards. In order to standardize the
dataset, all the balance sheets that are reported under the International Financial
Reporting Standards (IFRS) are used when available, and Generally Accepted
10
The banking sector in the CEE countries are highly concentrated and the largest 4 or 5 banks hold
around 72 percent of the total banking sector assets in a CEE country [ECB (2005A)].
11
For simplicity, the total banking sector assets of a country is assumed to be equal to the total assets
of banks reporting in BankScope.
24
Accounting Principles (GAAP) are used otherwise
12
. Also in this dataset, balance
sheet statements are available under consolidated and unconsolidated accounts
13
. For
the CEE countries both consolidated and unconsolidated accounts are used according
to the availability of balance sheet information. Use of different statements for these
countries should not create any inconsistency, as most banks operating in Central and
Eastern Europe are small and domestically focused
14
. As a result, there is not much
difference between the consolidated and unconsolidated statements of these banks.
A major challenge for this study was to obtain bank ownership information.
In BankScope the ownership structure is available only with respect to the last
accounting year. It would be misleading to rely only on BankScope information
because many of the banks, in particular those operating in the CEE countries, had
experienced major changes in ownership. In order to have the full history of the
changes in ownership, several resources are used. Banker’s Almanac is utilized for
an individual bank’s ownership history and Privatization Barometer, MIGA and the
World Bank is used for any listed privatization transactions. If a bank is not listed in
12
GAAP is very similar to IFRS with respect to the components of financial systems such as balance
sheets, income statements and accounting standards. Therefore this paper assumes that use of GAAP,
where IFRS is not available, will not create any significant differences across the balance sheet data
[Thornton (2003)].
13
Consolidated statements contain the balance sheet information of all the companies affiliated with
that bank. If a bank has extended its operations abroad, then the consolidated statements capture the
foreign business transactions of that multinational bank. On the other hand unconsolidated statements
cover only the domestic business transactions of a bank.
14
Banks operating in the New Member States of the EU, namely the CEE countries, do not extend
their operations abroad [ECB (2005A)] – except a few banks operating in Hungary.
25
any of these datasets, central bank and/or individual bank web sites are used as
needed for change in ownership information.
The ownership of banks is classified under three categories: state, private
domestic, and foreign-owned. This paper differs from other papers in the literature in
terms of its determination of ownership. Many papers classify a bank as foreign if
the share of its capital controlled by all foreign shareholders combined is above a
certain threshold level. This paper classifies a bank as foreign-owned when there is a
single foreign investor holding the majority of shares. This distinction is important,
as this paper tracks the influence of the foreign owner on the behavior of its
subsidiary. State ownership is decided according to whether the state has a
controlling majority stake. All banks that are not classified as foreign or state-owned
are listed under the private domestic ownership category.
Figure 2 plots the number of banks in this dataset across different ownership
types over time per region, according to the aforementioned ownership classification.
This figure shows that the number of foreign-owned banks is increasing over time in
all CEE countries. In the earlier years of the dataset, there is a clear dominance of
state-owned banks. As privatizations and foreign greenfield investments take place
over time, the majority of large banks are classified as foreign in the CEE countries.
Lastly, looking at Figure 2, one can see that the number of banks in the CEE
countries is increasing over time. The drop in the total number banks in 2005 is due
to the lag in reporting balance sheet statements to the BankScope database.
Figure 2: Number of Banks per Ownership Category
Ownership Structure by Regions
Description: Y-axis shows the number of banks that belongs to a specific ownership structure in the dataset.
Bulgaria and Romania
0
2
4
6
8
10
12
14
16
18
1989 1991 1993 1995 1997 1999 2001 2003 2005
Baltic Countries
0
2
4
6
8
10
12
14
16
18
1989 1991 1993 1995 1997 1999 2001 2003 2005
Central European Countries
0
5
10
15
20
25
30
35
1989 1991 1993 1995 1997 1999 2001 2003 2005
Foreign State Total
Source: Various, indicated in the data section
26
Table 2 provides the summary statistics for all the variables of interest of the
panel dataset. This table covers balance sheet items from the 72 largest banks of the
ten CEE countries and their macroeconomic variables. The timeline starts from 1988
and lasts to 2005, which are respectively the earliest and latest years available for the
microeconomic data. Unfortunately, the number of variables covering the whole
sample period is limited, in particular for micro-level data, and therefore this is an
unbalanced panel with different number of observations across variables of interest.
Additionally, all variables which are related to the banking sector, and Domestic
Credit over GDP, are lagged one period in order to alleviate any endogeniety
concerns that may arise otherwise.
Table 2: Summary Statistics for the CEE Countries
Obs Mean Std. Dev Obs Mean Std. Dev Obs Mean Std. Dev <0 ≠0 >0
Net Loan Growth 648 29.85 43.72 361 26.51 49.06 287 34.06 35.54 0.99 0.03 0.01
Foreign Ownership Dummy 648 0.46 0.50 361 0.31 0.46 287 0.66 0.48 1.00 0.00 0.00
State Dummy 648 0.33 0.47 361 0.52 0.50 287 0.10 0.31 0.00 0.00 1.00
Total Assets over GDP (-1) 648 9.01 15.29 361 9.96 18.43 287 7.81 9.93 0.03 0.06 0.97
Customer Deposits / Total Assets (-1) 644 62.22 18.93 358 60.53 19.87 286 64.34 17.48 1.00 0.01 0.01
Interbank Liabilities / Total Assets (-1) 636 15.68 15.85 354 18.04 17.58 282 12.73 12.79 0.00 0.00 1.00
ROE (-1) 639 18.32 74.47 353 20.52 95.01 286 15.60 35.39 0.18 0.37 0.82
ROA (-1) 640 1.42 3.70 355 1.44 4.78 285 1.40 1.51 0.45 0.90 0.55
Cost to Income Ratio (-1) 625 64.30 38.04 341 61.81 48.05 284 67.29 20.01 0.97 0.06 0.03
Net Interest Margin (-1) 635 5.94 5.16 351 6.80 6.15 284 4.88 3.28 0.00 0.00 1.00
Loan Loss Provision/ Total Assets (-1) 614 1.22 2.11 334 1.80 2.62 280 0.53 0.85 0.00 0.00 1.00
Total Capital Ratio (-1) 434 16.94 10.24 210 16.14 9.48 224 17.69 10.87 0.94 0.11 0.06
Real GDP Growth Rate 634 4.05 3.77 352 3.04 4.44 282 5.32 2.11 1.00 0.00 0.00
Domestic Credit / GDP (-1) 628 38.85 23.49 341 41.69 28.31 287 35.48 15.37 0.00 0.00 1.00
Real Interest Rate 626 -0.24 9.42 339 -1.39 12.35 287 1.11 3.24 1.00 0.00 0.00
Spread 1 (Lending and Borrowing) 638 7.97 8.24 351 9.74 10.24 287 5.80 3.79 0.00 0.00 1.00
Spread 2 (Domestic and Euro) 626 9.95 17.26 339 14.70 21.22 287 4.33 7.80 0.00 0.00 1.00
Euro Area M3 Growth 657 6.01 2.11
Euro Area S-T Interest Rate (3 Month) 657 4.23 2.00
≠0: P-value of the two-sided t-test on the equality of means, where the data are not assumed to have equal variances.
<0: P-value of the one-sided t-test on the equality of means, where the data are not assumed to have equal variances.
>0: P-value of the one-sided t-test on the equality of means, where the data are not assumed to have equal variances.
S
1
: 1988-2000 S
2
: 2001-2005 Ha: S
2
-S
1
Whole Sample
27
28
As mentioned in the Stylized Facts about the CEE Countries section, the
1990s was the transition period for all the CEE countries. In this period, not only did
they have major changes in their banking structure, but all the CEE countries also
experienced banking crises. Therefore, Table 2 presents the statistics for the
variables of interest for the earlier and the later years of the dataset, in addition to the
whole sample. The first sample (S
1
) covers the period 1988 to 2000, and the second
sample (S
2
) is for the remaining years: 2001 to 2005. Looking at Table 2, one can see
that a majority of the statistics for the variables of interest vary significantly across
time in the CEE countries.
In this table, Net Loan Growth is calculated as the percentage change in net
loans of a bank outstanding. In order to eliminate outliers, any bank year with credit
growth rates greater than 200 percent is eliminated. Looking at Table 2, average Net
Loan Growth rate in the CEE countries is significantly higher in the 2000s compared
to its previous value. As mentioned earlier, this variable shows the amount of credit
growth in an economy, and a level higher than that of the previous period may
indicate financial deepening.
Table 2 shows that the CEE countries are dominated by foreign owned banks
in the second half of the dataset: in this period, 66 percent of the observations belong
to foreign owned banks. On the other hand, state ownership declined significantly
over time from 52 percent to 10 percent, showing that the majority of the
observations in the CEE countries now belong to privately owned banks.
Total Assets over GDP is a measure for the size of a bank, and as given in
Table 2, this statistic has a bigger mean for the earlier years of the sample. Since
29
most of the banks before the transition were state owned, a larger mean for the
earlier period of the sample reflects the characteristics of these large state-owned
enterprises: there were very few but very large banks during the centrally planned
economies.
The liquidity structure of the CEE banks also shows a change over time.
During the centrally planned economies banks did not act as financial intermediaries
but rather as entities for account keeping. Similarly for households, the state savings
banks did not constitute an investment opportunity but acted as an entity to collect
household credit. As the banking system in the CEE countries matured by providing
extended financial services, the percentage of Customer Deposits over Total Assets
increased. The higher mean value for this variable during the second half of the
dataset is a reflection of banks functioning as financial intermediaries. Additionally
the gradual decline in the average Interbank Liabilities over Total Assets is an
indication of banks becoming more independent from the government. The high
mean value of the interbank liabilities shows the state backing during the banking
crises of the CEE countries.
The banks operating in the CEE countries have high means and variance in
profitability, efficiency, and soundness measures. Return on Equity (ROE) and
Return on Assets (ROA) determine a bank’s efficiency in generating income by
using their capital and assets respectively. The higher the ratio, the more profitable or
30
efficient a bank is
15
. Net Interest Margin is a measure for profitability of a bank and
it is the difference, between interest income and interest expense, divided by the
Average Earning Assets. Even though the Net Interest Margin declined over time in
the CEE banks, looking at Table 2, one can see that there are no significant
differences over time across the ROE and ROA ratios of the CEE banks. Cost to
Income Ratio is a measure of a bank’s efficiency and it is given by Operating Costs
as percent of Total Income. Even though the mean value of this ratio is increasing
over time, looking at Table 2, one can see that the standard deviation is declining
over time. Loan Loss Provisions are the net allowances that banks make against bad
or impaired loans. Since the data for Nonperforming Loans is not available, this
variable divided by Total Assets is used as a proxy for the asset quality of a bank.
Total Capital Ratio is another measure for the soundness of a bank and it is given by
Total Capital divided by Risk Weighted Assets. This ratio gets smaller as the risky
assets of a bank get larger. Both soundness measures show improvement over time.
The macroeconomic indicators show that the CEE countries are having
higher real growth rates and financial deepening and integration over time. The Real
GDP Growth rate increased from 3 percent in the 1990s to 5 percent in the 2000s.
The decline in Spread 1 across the CEE banks’ lending and borrowing rate shows
signs of financial deepening. Last, the smaller mean value of Spread 2, across the
15
However, one should mention that a low ROE ratio could also indicate high capitalization for a
bank. Therefore a low ROE value may not necessarily indicate the inefficiency of a bank.
31
domestic and the Euro are nominal interest rates, is an indication of economic and
financial integration of the CEE countries towards the Euro area.
As provided in Table 2, the significant differences in the means and standard
deviations of both the micro level and the macroeconomic variables of interest over
time show indications of a structural break in the dataset across the 1990s and the
2000s. This characteristic of the dataset argues in favor of studying these two
samples separately rather than pooling them together.
3.3 Empirical Model
Panel data analysis is used in order to analyze the importance of bank
ownership on credit growth. Such a procedure is necessary considering the short time
dimension of the dataset which is limited to 18 years. Additionally, there are several
advantages to using panel data models. The first advantage is that by pooling the data
it improves the accuracy of parameter estimates and therefore the estimation
procedure has more degrees of freedom and sample variability. Second, panel
estimation provides possibilities for reducing estimation bias. Last, it allows the
specification of more complicated behavioral hypotheses. Additionally, these models
lets one relax the assumption that the regressors in this dataset are independent of the
country specific errors.
Table 3: Pairwise Correlation Coefficients between Credit Growth and its Lagged Value for the
CEE Banks
1-8 9-16 17-24 25-32 33-40 41-48 49-56 57-64 65-72
Correlation Coefficient 0.49 0.30 0.15 0.71 0.11 1.00 -0.72 0.47 -0.26
Significance Level 0.060.470.670.030.801.000.010.350.45
Number of Observations 15 8 11 9 8 2 11 6 11
Correlation Coefficient -0.07 0.02 -0.17 0.14 0.05 . 0.73 0.14 -0.33
Significance Level 0.890.950.710.740.891.000.160.720.39
Number of Observations 6 11 7 8 9 0 5 9 9
Correlation Coefficient -0.56 -0.29 -0.87 0.32 0.17 0.40 0.98 -0.22 -0.26
Significance Level 0.250.310.060.610.890.500.110.600.45
Number of Observations 6 14 5 5 3 5 3 8 11
Correlation Coefficient -0.16 0.35 -0.14 -0.22 0.19 0.25 0.20 0.06 -0.23
Significance Level 0.800.270.690.500.560.690.510.910.65
Number of Observations 5 12 10 12 12 5 13 6 6
Correlation Coefficient 0.09 0.83 0.48 -0.32 0.00 0.03 0.50 -0.50 0.00
Significance Level 0.800.000.080.361.000.940.120.121.00
Number of Observations 10 11 14 10 14 12 11 11 14
Correlation Coefficient -0.87 -0.41 0.14 -0.01 -0.26 0.27 0.00 -1.00 0.06
Significance Level 0.010.360.670.980.680.490.991.000.87
Number of Observations 7 7 11 11 5 9 13 2 10
Correlation Coefficient 0.46 . 0.53 -0.02 0.82 -0.41 0.37 -0.23 -0.59
Significance Level 0.131.000.140.960.090.500.200.650.03
Number of Observations 12 0 9 11 5 5 14 6 14
Correlation Coefficient -0.17 -0.70 -0.22 -0.02 0.87 0.71 0.05 -0.33 0.33
Significance Level 0.690.510.430.960.030.010.950.390.26
Number of Observations 8 3 15 12 6 13 4 9 14
Note: Correlation coefficients with p-values smaller than 0.05 are marked with bold letters.
Bank Number:
32
In this paper a static panel data model is utilized. Table 3 presents the pair-
wise correlation coefficients between credit growth and its lagged value for each
bank separately. Looking at this Table, only 5 out of 72 banks have significant
correlations at the five percent significance level. The dependent variable in this
model is a growth variable: it is the percentage change in credit over two consecutive
periods. Therefore it is not an unexpected result to obtain insignificant correlation
coefficients over time between two change variables. The insignificance of the
correlation coefficients for bank credit growth over time argues in favor of choosing
a static panel data estimation model over a dynamic one.
After deciding on a static model, three possible models are considered for the
regression analysis: pooled OLS, fixed effects, and random effects. The general
panel data model is specified as follows,
ijt jt ijt ijt s ijt f ijt
u Z X s f C + ′ + ′ + + + =
−
φ β γ γ µ
1
(3.1)
where
ijt i ijt
v u ε + =
The dependent variable is credit growth, measured by the growth rate in the
net loans of bank i in country j at time t. On the left hand side, µ is the common
constant, and , and X
ij,t-1
represent, respectively, foreign and state ownership
dummies, and bank-specific variables for bank i in country j at time (t-1). In order to
diminish any simultaneity problems between the balance sheet items and the
dependent variable, all balance sheet variables are lagged one period. Ownership
variables are time-specific: due to privatizations, mergers and acquisitions most of
the state ownership dummies change value from 1 to 0 and foreign ownership from 0
to 1. Last, Z
jt
is the matrix for country specific macroeconomic variables for country
j at time t - Domestic Credit over GDP is lagged one period because of simultaneity
issues.
ijt
f
ijt
s
The Hausman specification test shows that the fixed effects estimator better
captures the dataset. The significant correlation between
i
ν and the regressors
indicates that the random effects model will be biased and therefore should be
eliminated. The choice between the pooled OLS and the fixed effects model becomes
33
34
clearer after rejecting the random effects model: fixed effects model is more
efficient. Additionally, the rejection of the hypothesis that all bank specific dummies
are equal to zero and Akaike and Bayesian Information criteria yielding smaller
values for the fixed effects model, are in favor of choosing fixed effects as the panel
estimation model.
In panel data analysis, in order to have consistent coefficients on the dummy
variables, the time-space needs to be large. As this dataset does not have a large T-
dimension, a within estimator procedure will be necessary to eliminate the bank-
specific dummies. Additionally, the estimates of bank specific coefficients are not of
interest to this paper. The large cross-sectional dimension, on the other hand, will
yield consistent coefficient estimates for the time-varying regressors. Therefore, the
final panel data model is estimated from a fixed effect within estimator of Equation
(3.1). One concern with the fixed effects estimator is the elimination of time-
invariant ownership coefficients, that is the within estimator will not be capturing the
ownership dummies unless the bank changes ownership over time. However, this
will be a minimal concern in this analysis as the ownership structure for a majority of
the banks in this sample switch starting from the mid 1990s onwards (See Figure 2).
Next, the standard assumptions on error terms, independent and identical, are
relaxed by allowing for robust and clustered estimation procedures. Clustering the
error terms within each country assumes that the observations from the same country
may be correlated and robust standard errors allow for heterogeneity. However,
clustering and using a robust estimation method do not yield larger standard errors.
Therefore, standard assumptions on the error term will be used as the most efficient
model.
In order to control for any unspecified macro effect, time-specific-country
dummies are introduced to Equation (3.1). The introduction of these dummy
variables controls the impact of a year-specific macroeconomic shock which is
peculiar to each CEE country. Both Akaike and Bayesian information criteria
indicate that the fixed effect estimation model is improved after the introduction of
these time-specific-country dummies. However, the macroeconomic variables are
highly collinear with the time-specific country dummies. In order to observe the
coefficient of the macroeconomic variables, each regression is run once for the
macroeconomic variables and once with the time-specific-country dummies.
Last, Equation (3.1) is modified in order to allow the interaction of the
foreign and state ownership dummies with bank-specific and macroeconomic
variables. Equation (3.2) shows the modification of the coefficients of interest with
respect to the foreign and state ownership variables.
[ ][ ]+
′
+
′
+
′
+ + + =
− − − 1 1 1 ijt ijt s ijt ijt f ijt ijt s ijt f ijt
X s X f X s f C β β β γ γ µ
[ ] [ ]
ijt jt ijt jt ijt f jt
u Z s
s
Z f Z +
′
+
′
+ ′ + φ φ φ (3.2)
In Equation (3.2), the analysis controls for the impact of different ownership
types on various micro and macroeconomic variables. In order to observe the
differences across the two private ownership types -foreign and domestic-, state
ownership dummy is introduced to Equation (3.2) rather than the private domestic
35
36
ownership dummy. An affluent impact of ownership, if not studied separately, may
yield insignificant or misleading results for the coefficient estimates when the
observations are pooled together.
3.3.1 Concerns about Endogeneity of the Foreign Ownership Variable
In the literature it is largely accepted that the acquisition of state-owned
banks by foreigners is not a random process but that the best performing state banks
are purchased by foreigners. For the CEE case, it might also be argued that
governments are not privatizing state banks in a random fashion but that they
privatize the better performing banks. In order to overcome this issue several
procedures are used in this paper. First, the dataset contains the largest banks of the
CEE countries. If one were to consider that size of the bank is an important
characteristic for foreign investors’ acquisitions decisions, then there is less variation
across state-owned banks of the CEE countries. Second, a fixed effect within
estimator method is used for the panel data estimations. Applying fixed effects rather
than the random effect model eliminates possible concerns about correlation between
the foreign ownership dummy and any bank specific error component of the random
effects model. Third, with respect to the correlation between ownership dummy and
time-varying bank-specific variables, many important measures of bank profitability,
liquidity, size, and soundness are incorporated into the regression equation. Lastly,
an instrumental variable model is considered.
Table 4: Panel Data Probit Estimation Results for Foreign Ownership Variable
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Total Assets over GDP (-1) 0.03 0.03 0.01 0.02 0.02 0.01
[0.01]*** [0.03]** [0.35] [0.07]* [0.09]* [0.34]
Loan Loss Provision/Total Assets (-1) -0.13 -0.18 -0.12 -0.16 -0.2 -0.07
[0.03]** [0.01]*** [0.05]* [0.01]*** [0.00]*** [0.23]
Total Capital Ratio (-1) 0.03 0.04 0.06 0.04 0.05 0.05
[0.06]* [0.01]** [0.00]*** [0.01]** [0.00]*** [0.00]***
Spread bw Domestic and Euro Interest Rates -0.08 -0.06 -0.04
[0.00]*** [0.00]*** [0.02]**
ROA (-1) -0.11 -0.12 -0.08
[0.02]** [0.01]** [0.06]*
Net Interest Margin (-1) -0.23 -0.21
[0.00]*** [0.00]***
Spread between lending and borrowing Rate -0.23 -0.2 -0.21
[0.00]*** [0.00]*** [0.00]***
Constant -0.13 -0.12 0.74 0.84 0.69 1.65
[0.79] [0.81] [0.18] [0.11] [0.19] [0.01]***
AIC 342.79 337.24 323.31 336.15 331.78 320.85
BIC 367.07 365.56 355.65 360.53 360.21 349.28
Observations 423 422 421 430 429 429
Number of bankid 62 6262626262
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
The instrumental variable model is estimated by using a two-stage least-
squares (TSLS) generalization of the fixed effects panel data estimator. However, the
choice of instruments for foreign bank ownership is not a straight-forward decision.
In order to choose the instruments which best explain the variation in the foreign
ownership dummy a panel data probit estimation method is utilized
16
. Table 4
provides the estimation results of this model for various candidate instrumental
variables. Looking at this table, one can see that Model 6 has the smallest
16
A random effects probit model is employed for the estimations, and the model assumes a normal
distribution function for the bank-specific error term.
37
38
information criteria values. Therefore the regressors of this model are used as the
optimal instrumental variables.
After deciding the optimal instruments, a TSLS generalization of the fixed
effects panel data model is used to estimate the impact of the foreign bank ownership
on credit growth
17
. In the first stage, the foreign-ownership variable is regressed on
the variables of Model 6. In the second stage, the predicted values of this variable -
which were obtained from the first stage - and other possible regressors are used as
the independent variables of the credit growth estimation. The same model of the
second stage is then estimated without employing the instruments but the foreign
ownership dummy by itself and the other regressors, which were used in the second
stage of the TSLS regression. After obtaining the coefficient estimates from both
methods, a Hausman specification test is performed to see whether the coefficient
estimates of the two models are systematically different.
Table 5 provides the Hausman specification test results for two different
models. The fist model covers the whole sample and the results indicate that treating
the foreign ownership variable as exogenous, given the fixed effect model, does not
have any deleterious impact on the consistency of coefficient estimates. In section
3.2 Data, the analysis indicated that there may be a break in the time sequence of the
data set. Therefore a Hausman test covering the second part of the data set is also
17
The estimator assumes that the idiosyncratic error term in the credit growth equation is uncorrelated
with any of the regressors. However, since the bank-specific error term is assumed to be constant, this
variable may be correlated with the independent variables of that equation.
reported as well as the one for the entire sample. The result again supports that there
is no systematic difference across the estimated coefficients.
Table 5: Hausman Specification Test for Foreign Ownership Dummy
Model 1: Whole Sample
(b) (B) (b-B) [diag(V
b
-V
B
)]
1/2
IV No IV Diff. Standard Error
Foreign 90.31 -26.80 117.11 77.84
State 78.70 -31.12 109.82 74.30
ROE (-1) -0.03 -0.05 0.02 0.01
Real GDP Growth 3.54 2.92 0.62 0.37
B = inconsistent under Ha, efficient under Ho; obtained from xtreg
Test: Ho: difference in coefficients not systematic
chi2(4) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 2.87
Prob>chi2 = 0.5801
Model 2: For the Sample of 2001-2005
(b) (B) (b-B) [diag(V
b
-V
B
)]
1/2
IV No IV Diff. Standard Error
Foreign 173.74 33.21 140.53 126.70
Customer Dep./Total Assets (-1) 0.85 0.70 0.15 0.32
Interbank Liab./Total Assets (-1) 1.04 1.20 -0.16 0.18
Real GDP Growth 2.01 2.32 -0.31 0.99
B = inconsistent under Ha, efficient under Ho; obtained from xtreg
Test: Ho: difference in coefficients not systematic
chi2(4) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 2.15
Prob>chi2 = 0.71
Notes: IV: Uses the two-stage least-squares within estimator
No IV: Uses within estimator
Instrumented: Foreign
Instruments: Total Assets over GDP(-1), Net Interest Margin(-1),
Loan Loss Provn/TA(-1), Total Capital Ratio (-1)
and Spread
39
40
Regarding the test results provided in Table 5, the paper assumes that a
foreign investor’s acquisition decision is not shaped by the credit growth pattern of a
specific bank, conditional on the bank-specific fixed effect.
3.3.1 Concerns about Multicollinearity
A relevant issue in a multivariate regression analysis is the collinearity of
regressors with each other. In order to examine this problem, a correlation Matrix is
provided in Table 6 for the significance of the pair-wise correlation coefficients of
the micro and macro variables in the CEE countries. Looking at the matrix provided
in Table 6 one can see that most of the regressors are significantly correlated with
each other, even though the correlation coefficients in most cases are small
18
.
However, the coefficients across similar measurement variables are large in
magnitude, such as the variables for bank profitability and efficiency
19
or
macroeconomic variables
20
. In order to eliminate any multicollinearity problems, the
paper introduces these highly and significantly correlated variables into separate
regression equations rather than introducing all the variables at once.
18
The significant correlation coefficients, with an absolute value of an estimate smaller than 0.3, are
considered as small.
19
ROA, ROE and Cost-to-Income Ratio
20
Spread between Lending and Borrowing Rate, and Spread between Domestic and Euro Area
Interest Rate.
Table 6: Correlation Coefficient across Regressors for the CEE Countries
Spread
1
State Dummy -0.65
0.00
Total Assets over GDP -0.10 0.27
(TA/GDP)
0.01 0.00
Customer Dep. / Total Assets -0.03 0.09 -0.14 1
(Cust. Deps.)
0.50 0.02 0.00
Interbank Liab. / Total Assets 0.09 0.04 0.26 -0.68 1
(Inter. Liab.)
0.02 0.36 0.00 0.00
ROE 0.00 0.01 0.03 0.09 -0.02 1
0.96 0.79 0.48 0.03 0.64
ROA -0.05 -0.05 0.00 -0.02 -0.17 0.08 1
0.24 0.17 1.00 0.63 0.00 0.04
Cost to Income Ratio 0.04 -0.07 -0.16 0.05 -0.10 -0.13 -0.47 1
(C-to-I.)
0.27 0.08 0.00 0.22 0.02 0.00 0.00
Net Interest Margin -0.19 0.07 -0.15 0.09 -0.20 0.07 0.39 -0.12 1
(NIM)
0.00 0.10 0.00 0.02 0.00 0.07 0.00 0.00
Loan Loss Prov/ Total Assets -0.25 0.25 -0.04 -0.03 0.07 -0.14 -0.25 -0.08 0.30 1
(LLP/TA)
0.00 0.00 0.31 0.50 0.09 0.00 0.00 0.05 0.00
Total Capital Ratio (-1) -0.05 -0.17 -0.14 0.03 -0.18 -0.05 0.26 -0.09 0.38 -0.01 1
(TCR)
0.30 0.00 0.00 0.59 0.00 0.34 0.00 0.06 0.00 0.86
Real GDP Growth Rate 0.14 -0.22 -0.16 0.12 -0.13 -0.08 -0.08 0.15 -0.17 -0.24 -0.03 1
0.00 0.00 0.00 0.00 0.00 0.05 0.05 0.00 0.00 0.00 0.56
Real Interest Rate 0.08 -0.14 -0.24 0.03 -0.04 -0.13 -0.19 0.10 -0.07 -0.13 0.01 0.25 1
(r)
0.04 0.00 0.00 0.44 0.38 0.00 0.00 0.02 0.07 0.00 0.89 0.00
Domestic Credit / GDP 0.07 0.10 0.37 -0.19 0.30 0.02 -0.06 -0.18 -0.30 0.01 -0.36 -0.31 -0.24 1
(DC/GDP)
0.10 0.01 0.00 0.00 0.00 0.58 0.15 0.00 0.00 0.78 0.00 0.00 0.00
Spread 1 (Lend - Borrow) -0.18 0.20 0.05 -0.03 0.01 0.13 0.08 -0.06 0.31 0.28 0.22 -0.47 -0.48 0.10 1
(Spread
1
)
0.00 0.00 0.21 0.45 0.85 0.00 0.05 0.16 0.00 0.00 0.00 0.00 0.00 0.02
Spread 2 (Domestic - Euro) -0.17 0.19 0.14 -0.04 0.11 0.11 0.12 -0.18 0.35 0.24 0.19 -0.52 -0.10 0.17 0.78
0.00 0.00 0.00 0.32 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00
* Second row shows the p-values
Note: The correlation coefficiente with p-vaues smaller than 0.05 are marked with bold letters.
r
DC/G
DP NIM
LLP/
T.A. TCR GDP
Inter.
Liab. ROE ROA C-to-I Foreign State
TA/GD
P
Cust
Deps
41
42
3.4 Results
In the data analysis section, the mean and standard deviation for all variables
of interest are shown to be different across time for the CEE countries. Changes in
the significance level of variables due to sample selection indicate a structural break
in the second half of the dataset. Therefore the regression analysis is performed
separately for two samples: for Sample 1 from 1988 to 2000, and Sample 2 from
2001 to 2005. The results of the regression analysis for these samples are provided in
Table 7 to Table 9. As mentioned in section 3.3.1 Concerns about Multicollinearity,
most of the balance sheet items are significantly correlated with each other. Large
and significant correlation coefficients are symptoms of multicollinearity problem.
Therefore variables which may cause multicollinearity are introduced in separate
regressions equations. Following, the cross-ownership varying coefficients are
reported in Table 10 and Table 11.
In this section, results based on the regression analysis of the balance sheet
items on credit growth are reported under the section 3.4.2 Impact of
Macroeconomic Shocks and results derived from the macroeconomic variables are
provided under the section 3.4.2 Impact of Macroeconomic Shocks.
Table 7: Fixed Effect Estimator Results for the CEE Countries
Foreign (-1) -5.1 -1.76 -80.81 -84.52 1.64 1.11 -79.9 -87.24
[0.86] [0.95] [0.01]** [0.00]*** [0.96] [0.97] [0.01]** [0.00]***
State (-1) -2.22 -3.91 -113.59 -110.29 -0.91 -1.52 -109.86 -114.21
[0.94] [0.90] [0.00]*** [0.00]*** [0.98] [0.96] [0.00]*** [0.00]***
Total Assets over GDP (-1) 0.39 0.26 -1.21 -1.92 0.28 0.25 -1.77 -1.32
[0.07]* [0.21] [0.23] [0.11] [0.19] [0.23] [0.09]* [0.26]
Customer Deposits / Total Assets (-1) 0.49 0.31 0.67 0.55 0.12 0.13 0.67 0.66
[0.17] [0.37] [0.10] [0.17] [0.73] [0.70] [0.10] [0.10]
Interbank Liabilities / Total Assets (-1) -0.47 -0.52 1.06 1.06 -0.83 -0.72 1.22 1.17
[0.23] [0.14] [0.01]*** [0.01]** [0.02]** [0.03]** [0.00]*** [0.00]***
ROA (-1) 1.18 1.26 -3.34 -4.03
[0.06]* [0.03]** [0.07]* [0.02]**
ROE (-1) -0.01 -0.01 0 0.05
[0.73] [0.78] [0.99] [0.40]
Spread 1 (Lending - Borrowing) -0.07 -3.52
[0.81] [0.02]**
Spread 2 (Domestic - Euro) -0.38 -1.02
[0.03]** [0.08]*
Constant 8.59 16.43 63 85.46 33.36 31.13 75.37 57.98
[0.80] [0.62] [0.11] [0.03]** [0.31] [0.34] [0.06]* [0.13]
Country-Time Dummies No Yes No Yes No Yes No Yes
Observations 328 345 280 275 337 343 281 276
Number of bankid 63 63 66 65 63 63 66 65
R-squared 0.08 0.06 0.17 0.46 0.04 0.04 0.16 0.45
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
* significant at 10%; ** significant at 5%; *** significant at 1%
Model 1
1990s 2000s
Model 2
1990s 2000s
43
Table 8: Fixed Effect Estimator Results for the CEE Countries
Foreign (-1) 1.12 2.43 -79.22 -87.61 3.86 3.86 -76.93 -84.87
[0.97] [0.93] [0.01]** [0.00]*** [0.90] [0.90] [0.02]** [0.00]***
State (-1) 9.28 1.22 -109.2 -113.56 -0.36 -0.36 -114.3 -114.6
[0.76] [0.97] [0.00]*** [0.00]*** [0.99] [0.99] [0.00]*** [0.00]***
Total Assets over GDP (-1) 0.33 -0.03 -2.05 -1.86 0.03 0.03 -1.3 -1.21
[0.26] [0.90] [0.08]* [0.12] [0.90] [0.90] [0.23] [0.33]
Customer Deposits / Total Assets (-1) 0.1 0.04 0.7 0.64 0.07 0.07 0.79 0.67
[0.79] [0.92] [0.10] [0.13] [0.84] [0.84] [0.06]* [0.10]*
Interbank Liabilities / Total Assets (-1) -0.86 -0.79 1.11 1.25 -0.72 -0.72 1.11 1.18
[0.03]** [0.03]** [0.01]*** [0.00]*** [0.03]** [0.03]** [0.01]*** [0.00]***
Cost to Income Ratio (-1) 0.01 0 0.41 0.3
[0.88] [0.94] [0.01]*** [0.03]**
Net Interest Margin (-1) 0.29 0.29 -2.95 0.23
[0.58] [0.58] [0.08]* [0.90]
Euro Area M3 Growth -0.76 -0.76 1.79 1.93
[0.71] [0.71] [0.12] [0.49]
Domestic Credit over GDP (-1) -0.52 0.58
[0.01]*** [0.15]
Constant 49.59 38.95 8.94 45.64 36.42 36.42 44.32 36.9
[0.15] [0.25] [0.83] [0.24] [0.32] [0.32] [0.31] [0.46]
Country-Time Dummies No Yes No Yes No Yes No Yes
Observations 321 333 279 274 343 343 280 275
Number of bankid 63 63 66 65 63 63 66 65
R-squared 0.09 0.05 0.17 0.47 0.05 0.05 0.15 0.45
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
Model 3 Model 4
1990s 2000s 1990s 2000s
44
Table 9: Fixed Effect Estimator Results for the CEE Countries
Foreign (-1) -7.19 -9.17 -79.5 -84.1 -0.3 -0.3 -75.69 -91.33
[0.80] [0.74] [0.01]*** [0.00]*** [0.99] [0.99] [0.01]** [0.00]***
State (-1) -2.7 -6.13 -111.54 -112.82 25.61 25.61 -113.81 -117.05
[0.93] [0.84] [0.00]*** [0.00]*** [0.28] [0.28] [0.00]*** [0.00]***
Total Assets over GDP (-1) 0.14 0.09 -1.53 -1.26 0.12 0.12 -1.4 -0.72
[0.58] [0.74] [0.12] [0.28] [0.67] [0.67] [0.15] [0.54]
Customer Deposits / Total Assets (-1) -0.06 -0.09 0.69 0.72 0.49 0.49 0.34 0.67
[0.87] [0.78] [0.08]* [0.07]* [0.19] [0.19] [0.39] [0.13]
Interbank Liabilities / Total Assets (-1) -0.85 -0.81 1.19 1.15 0.4 0.4 0.97 1.39
[0.01]** [0.01]** [0.00]*** [0.01]*** [0.33] [0.33] [0.01]*** [0.00]***
Loan Loss Provision/Total Assets (-1) -4.01 -4.24 -0.91 3.02
[0.00]*** [0.00]*** [0.74] [0.25]
Total Capital Ratio (-1) 2.36 2.36 -0.42 0.69
[0.00]*** [0.00]*** [0.38] [0.23]
Euro Area S-T Interest Rate (3-Month) 4.56 4.56 0.68 2.91
[0.02]** [0.02]** [0.79] [0.65]
Real GDP Growth Rate 0.64 1.18
[0.32] [0.45]
Constant 50.55 57.77 45.95 69.6 -83.36 -83.36 79.01 35.03
[0.11] [0.07]* [0.24] [0.07]* [0.03]** [0.03]** [0.04]** [0.47]
Country-Time Dummies No Yes No Yes No Yes No Yes
Observations 321 327 271 270 206 206 224 224
Number of bankid 60 60 66 65 53 53 58 58
R-squared 0.11 0.1 0.16 0.46 0.23 0.23 0.19 0.48
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
* significant at 10%; ** significant at 5%; *** significant at 1%
Model 5 Model 6
1990s 2000s 1990s 2000s
45
Table 10: Cross-Ownership Varying FE Estimators for the CEE Countries
1990s 2000s 1990s 2000s
Foreign (-1) -29.28 -89.05 -13.99 -134.39
[0.44] [0.01]** [0.64] [0.03]**
State (-1) -48.74 -113.44 21.35 -230.82
[0.20] [0.00]*** [0.51] [0.01]***
T.A. over GDP (-1) -3.41 -1.51 0.06 -1.47
[0.00]*** [0.13] [0.84] [0.13]
Interbank Liabs/TA (-1) -2.22 -0.21 -1.05 1.18
[0.00]*** [0.69] [0.01]*** [0.00]***
Foreign 1.28 1.25 Customer Dep. / 0.09 0.61
[0.12] [0.06]*
TA (-1)
[0.81] [0.13]
State 2.01 1.31
[0.02]** [0.32]
ROA (-1) 3.34 -2.33 Cost/Income(-1) 0 -0.09
[0.03]** [0.58] [0.98] [0.87]
Foreign -2.83 -0.49 Foreign 0.19 0.37
[0.30] [0.92] [0.16] [0.48]
State -1.89 -18.86 State -0.14 1.39
[0.29] [0.02]** [0.31] [0.10]
Real Interest Rate 0.69 -3.42 Spread 1 0.78 -6.19
[0.57] [0.08]* [0.59] [0.03]**
Foreign 0.71 2.87 Foreign 1.43 4.99
[0.66] [0.20] [0.40] [0.12]
State -0.07 -2.81 State -0.89 1.4
[0.96] [0.43] [0.55] [0.66]
Constant 98.26 115.21 27.57 104.37
[0.00]*** [0.00]*** [0.42] [0.09]*
Observations 296 243 329 274
Number of bankid 57 58 63 65
R-squared 0.2 0.24 0.1 0.22
p values in brackets
Model 7 Model 8
46
Table 11 : Cross-Ownership Varying FE Estimators for the CEE Countries
1990s 2000s 1990s 2000s
Foreign (-1) 104.4 -32.09 -7.14 -95.74
[0.03]** [0.58] [0.82] [0.00]***
State (-1) 108.64 -101.03 -6.96 -148.31
[0.03]** [0.07]* [0.83] [0.00]***
T.A. over GDP (-1) 0.02 -1.77 0.18 -1.1
[0.95] [0.09]* [0.50] [0.23]
Customer Dep/TA (-1) 2.51 0.19 -0.02 0.64
[0.00]*** [0.73] [0.95] [0.08]*
Foreign -1.74 -0.46 Interbank Liab. / -0.87 1
[0.03]** [0.51]
TA (-1)
[0.01]** [0.00]***
State -2.26 -0.22
[0.00]*** [0.85]
Net Interest Margin (-1) -0.05 1.89 Loan Loss Prov / -5.5 5.33
[0.96]
[0.49]
TA (-1)
[0.03]** [0.39]
Foreign -6.19 -7.25 Foreign 0.25 -4.98
[0.01]*** [0.09]* [0.95] [0.47]
State -0.21 -2.37 State 1.83 10.48
[0.87] [0.61] [0.52] [0.27]
Spread 2 0.39 -2.42 Real GDP Growth 0.88 1.2
[0.45] [0.10]* [0.21] [0.41]
Foreign -0.78 3.09 Foreign -0.82 3.09
[0.19] [0.07]* [0.46] [0.09]*
State -1.16 0.78 State 0.37 6.23
[0.04]** [0.66] [0.61] [0.13]
Constant -90.85 106.13 50.11 48.87
[0.03]** [0.02]** [0.15] [0.19]
Observations 331 279 314 266
Number of bankid 63 65 60 65
R-squared 0.15 0.15 0.13 0.23
p values in brackets
Model 9 Model 10
47
48
3.4.1 Impact of Microeconomic Shocks
The results show major differences across sample periods. The t-tests,
performed for the mean comparison of variables of interest for the 1988-2000 and
2001-2005 samples in Table 2 (section 3.2 Data) support the statistical evidence for
the difference of these two samples. Also, by looking at the sign and the significance
of the coefficients in Table 7 to Table 11, one can see that there is clearly a time
break in the dataset. For example; looking at the coefficient estimate of Interbank
Liabilities in Table 7, one can see that the coefficient of this variable changes from a
significant negative value to a significant positive value. There are several reasons
for such a break in the dataset. First, the changes in the means and the variance of the
banking variables reflect the transition of the CEE banks from the centrally planned
to a market economy banking system. Second, the high standard errors during the
1990s (refer to Table 2) may be attributed to the banking crises and the
recapitalization programs that happened in this period. Third, robust economic
growth and rapid financial integration of the 2000s, shows in the significant changes
in the spreads and inflation rates of the CEE countries across these two samples.
Foreign bank ownership is positively correlated with credit growth in the
CEE countries. Figure 3 plots average credit growth across different ownership
types in the CEE countries from 1992 to 2005, for which the credit growth data
exists for all bank ownership types. This figure shows that foreign banks have higher
credit growth compared to domestically owned banks for all the years except three:
1992, 2003, and 2004.
Figure 3: Credit Growth across Bank Ownership Types in CEE Countries
-60
-40
-20
0
20
40
60
80
100
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Foreign State Private Domestic
Source: BankScope
Note: This Figure excludes the two years before 1992, since data for domestically owned
private banks does not exist for those two years.
Foreign banks have lower credit growth rates compared to private
domestically owned banks. Even though there is a significant correlation between
credit growth and foreign bank ownership, the paper further investigates whether this
statement is valid across different domestic ownership types. After differentiating the
credit growth across domestically owned banks i.e.; between private and state owned
banks, Figure 3 shows that credit grows slower in foreign banks compared to private
domestic banks. In particular, after controlling 4 peaks in the credit growth of foreign
banks, which happened in the earlier years of the dataset, private domestically owned
banks have consistently higher credit growth rates. Moving from the qualitative
49
50
analysis to the regression analysis, one can see that the estimation results yield the
same outcome as the aforementioned graphical analysis. The coefficient for foreign
ownership dummy in the CEE countries is insignificant during the first half of the
dataset and significant and negative for the second half of the dataset in all of the
regression results (See Table 7 to Table 9). Since these regressions control for state
ownership, a negative and significant coefficient for foreign ownership dummy
indicates that in the CEE countries foreign banks have, on average, lower credit
growth rates compared to the private domestically owned banks. There are several
explanations for private domestically owned banks to increase their credit line more
than foreign owned banks. First, some of the private domestically owned banks are
part of a big conglomerate group. The affiliated companies have more incentive to
raise their funding needs via the associated bank. Second, private domestically
owned banks have longer-term relationships with the domestic companies, which
puts them in a better position compared to the foreign owned banks that were opened
up as a Greenfield investment.
Customer Deposits were more important for the CEE banks’ financing
in the 1990s. Looking at the regression results provided in Table 7 through Table 9,
the coefficient of Customer Deposits over Total Assets do not yield any significant
results for the earlier years of the dataset
21
. However, the insignificance of this
coefficient may be due to the difference on the slope estimate of this variable across
21
The coefficient estimates of Customer Deposits over Total Assets yield a few significant results at
the 10 percent significance level during the second half of the dataset. However, the insignificance of
this coefficient across any of the ownership types, presented in and , leads the paper
to conclude that this variable is not influential on credit growth during 2000s.
Table 10 Table 11
51
different ownership types. Looking at Table 11, one can see that Customer Deposits
constitute a very important source of financing for the private banks in the earlier
years of the dataset and not so important for the state owned banks. This result has
two implications. First, it shows that private banks were restricted by the domestic
market conditions to increase their credit line and over time the importance of
Customer Deposits diminished for the credit growth of CEE banks. Second, the
foreign owned banks acquired the parent bank resources much later after their
acquisition and hence they acted like private domestically owned banks in the
1990s
22
.
Interbank Liabilities have different influences on the growth of credit
over the course of the time. Looking at Table 7 to Table 9, one can see that the
coefficient of Interbank Liabilities over Total Assets changed its sign over the two
periods for banks operating in the CEE countries. The coefficients of this regressor
have significant and negative coefficients for the earlier years, and significant and
positive estimates for the last five years. A negative estimate for the earlier years
reflects the impact of the recapitalization programs that occurred during the 1990s.
Looking at the cross-ownership estimates provided in Table 10, one can see that the
negative and significant estimates of Interbank Liabilities over Total Assets for
private banks support the importance of 1990 recapitalization programs and the
funding coming from the central banks for the bad or impaired loans contained in the
22
A majority of the foreign owned banks did not have a change in their managerial structure much
after their acquisition by foreigners (ECB (2005A)). The managerial structure of the foreign owned
was similar to the private domestically owned banks.
52
majority of the CEE banks at that time. Alternatively, a significant and positive
coefficient in the 2000s represents a change in the interpretation of this variable. It
shows that interbank lending became an important source for funding the bank loans
in the CEE countries
23
.
Interbank Liabilities increase the credit growth rate of foreign owned
banks only during the second half of the dataset. Cross-ownership varying
coefficients of Interbank Liabilities over Total Assets are reported in Table 10. The
interpretation of the Interbank Liabilities changes during the second half of the
dataset, and the results indicate that there is a significant difference across ownership
types. The coefficient of this variable is significant and positive only for the foreign
owned banks. This relationship supports two hypotheses: First, foreign banks are not
restricted by the domestic financing - such as Customer Deposits - but they can
easily increase their funding resources via transfers from their parent bank. Second,
foreign banks can also borrow from other banks using the credit worthiness of the
parent bank.
Financial Soundness is positively correlated to loan growth during the
earlier years in the CEE countries. The regression results in Table 9 show
significant and negative coefficients for Loan Loss Provision over Total Assets and
positive for Total Capital Ratio in the CEE countries. The first variable shows the
amount of bad or impaired loans that a bank expects to incur in that year. The latter
23
The share of cross-border lending as a percentage of GDP, coming from the Western European
countries to the banks operating in the CEE countries, increased from 31% in 2000 to 42% in 2003.
Cross-border lending grew around 4 percent each year during the second half of the dataset [ECB
(2005B)].
53
variable is the ratio of the total capital of a bank divided by its risk weighted assets.
It decreases in magnitude as the amount of its risk-weighted loans increases. This
implies that as a bank incurs riskier loans in the previous period it distributes fewer
loans during the following year unless its capital increases. These variables do not
yield any significant estimates for the last five years in the CEE countries. Looking
at Table 11, one can see that the soundness measure was significant for all bank
ownership types during the 1990s and insignificant during the 2000s. This can be
interpreted as the CEE banks less frequently facing systematic credit collapses,
unlike the several banking crises that they experienced during the mid-1990s.
Profitability and efficiency matter for all banks. The profitability and
efficiency ratios are reported in Table 7 and Table 8. Among these ratios, ROA has
time varying coefficient estimates. Table 7 reports that the coefficient estimate of
this variable changes from a significant and negative one to a significant and positive
one over the two periods considered in this study. ROE does not yield any significant
results. However, the insignificance of this variable may also indicate a change in the
capitalization behavior of the CEE banks. As reported in Table 8, the Cost to Income
Ratio and Net Interest Margin are insignificant for the first period and are significant
during the last five years, with a positive estimate for the Cost to Income Ratio and a
negative one for the Net Interest Margin. The inverse relationship between the credit
growth and the profitability ratios during the second half of the dataset is a reflection
of the supply side effect. The increase in the number of banks operating in the CEE
markets causes a decline in profits and an increase in the credit supply. Looking at
the cross ownership study in Table 10 and Table 11, one can see that foreign banks
54
are the driving force for this inverse relationship for the second half of the dataset.
The increased competition resulting from the entrance of these banks to the CEE
markets causes a decline in the interest margins, which is reflected by the significant
and negative coefficient estimate of the Net Interest Margin for the foreign owned
banks. The declining margins do not affect the efficiency and the cost structure of the
private banks operating in the CEE countries. The insignificant coefficient estimates
of the ROA and Cost to Income Ratio for the private banks show that efficiency does
not vary across private banks operating in the CEE countries. However, as shown in
Table 10 and Table 11, state owned banks are less efficient. The coefficient estimate
of the ROA is significant and negative and the coefficient estimate of the Cost to
Income Ratio is significant and positive during the second half of the dataset. This
indicates that state owned banks operating in the CEE countries need to incur higher
costs to increase their credit line.
3.4.2 Impact of Macroeconomic Shocks
Economic growth is the driving factor for credit growth of the foreign
owned banks in the 2000s. Looking at regression results in Table 9, one can see that
the coefficient of Real GDP Growth is insignificant for both periods in the CEE
countries. The paper investigates whether this result is due to the cross-ownership
varying results and shows that foreign ownership plays an important role on the
55
impact of business cycles on the growth of bank credit. The results reported on Table
11 indicate that economic growth was not influential in the first period. However, it
is significant and positive for the foreign owned banks in the second half of the
dataset. The significance of the coefficient of Real GDP Growth implies that credit
growth pattern of the foreign owned banks is procycylical: Foreign banks tend to
increase credit supply in good times and reduce it in bad times.
During the last five years, the spread between the bank lending and
borrowing rate shows signs for financial deepening in the CEE countries.
Looking at Table 7, one can see that the coefficient for the Spread between Lending
and Borrowing Rate is significant and negative in the CEE counties during the last
five years. This negative coefficient can be interpreted as a shift in the supply curve.
As more financial intermediaries operate in the CEE countries they increase the
amount of competition in the banking sector, cause a leftward shift in the supply
curve of bank credit, and respectively a decline in the spreads. Table 10 presents that
this negative spread has similar impacts on the growth of credit across different
ownership types in the CEE countries.
The spread between domestic and Euro area interest rates is negatively
related to loan growth. Table 7 presents negative and significant coefficient
estimates for the Spread between Domestic and Euro Area Nominal Interest Rates
for the CEE countries in both periods, with a larger estimate for the second half. A
negative coefficient for this variable can be seen as evidence of financial integration
with the Euro area. This implies that credit grows faster as the CEE countries
integrate more to the Euro Area.
56
Higher interest rates in the CEE countries are the driving reason for
foreign owned banks to increase their credit lines in this region. Looking at
Table 11, one can see that the impact of Spread between Domestic and Euro Area
Nominal Interest Rates on credit growth changes across ownership types over time.
In the first period, the coefficient estimate of this variable is significant only for the
state owned banks, and during the second period it is significant for all domestic
banks. Interestingly, during the second period the coefficient estimate of the Spread
between Domestic and Euro Area Interest Rates is significant and positive for the
foreign owned banks. This implies that as the interest rates are higher in the CEE
countries, foreign banks have more incentive to move to these markets and increase
their operations there.
Lower real interest rates increase credit growth in the CEE countries
during the last five years. Looking at Table 10, one can see that the Real Interest
Rate has a negative and significant coefficient for all bank types operating in the
CEE countries. This behavior indicates a supply-side effect in the CEE countries:
increased competition in this region forces a decline in the interest rates and as the
real interest rates fall, the credit growth supplied by the CEE banks increase.
57
3.6 Conclusion
This paper studies the impact of foreign bank ownership on the credit growth
pattern of the ten Central and Eastern European countries. The paper investigates the
role of foreign bank ownership in the growth of credit by employing four versions of
a fixed-effect-within estimator model: first two, via analyzing the data in two
different samples (1990s and 2000s), and the latter two by introducing the
macroeconomic indicators and the county-time dummies to each of these datasets
separately. The analysis on the CEE countries also includes tests on possible micro-
level data problems and then presents results on the role of bank ownership in credit
growth.
The results indicate that banks operating in the CEE countries have differing
banking structure over the course of the sample. The micro and the macroeconomic
variables have changing influence on the credit growth pattern in the 1990s and the
2000s, both in terms of changing sign and significance levels. In the earlier years, the
foreign banks acted like the private domestically owned banks. However by the
2000s, the foreign bank ownership became significant in the CEE countries. The
funding sources and the domestic market conditions showed major differences across
the foreign banks and the domestic banks. The analysis shows that interbank
liabilities constitute a major component for the increase in credit for the foreign-
owned banks. Additionally, the high economic growth and interest rates are the
driving factors for the foreigners to increase their credit lines in the CEE countries.
58
In the 2000s, the results show that private domestically owned banks have similar
profitability and efficiency indicators as that of the foreign owned banks. However,
the difference between bank ownership types, in terms of credit growth responses to
the profitability and efficiency indicators, are between the state-owned and the
privately owned banks: that the state owned banks are inefficient in generating credit
line.
59
Chapter 4
The Managerial Impact of Parent Banks on their Affiliated
Banks Operating in Foreign Countries: A Case Study for
the Central and Eastern European Countries
This paper studies the impact of managerial influence on the credit growth
pattern of the foreign owned banks operating in the Central and Eastern European
(CEE) countries. As stated in chapter 3 of this thesis, a majority of the banking assets
in the CEE countries are controlled by foreigners, and therefore, the credit growth
pattern of the foreign owned banks have a significant influence on the credit growth
pattern of the CEE countries. The management of the parent bank, head quartered in
a country outside the CEE region, will influence the credit growth pattern in their
CEE subsidiaries
24
, and this in turn may shape the trend in credit growth in these
24
This paper focuses only on the relationship between the CEE subsidiaries and their parent banks
operating in foreign countries. Foreign presence in the CEE countries is mostly in the form of
subsidiaries of banks which have their headquarters in a country outside the CEE region [ECB
(2005A)]. Since branches do not have a significant influence for the CEE banking system, this paper
considers only the foreign subsidiaries, rather than foreign branches, for the credit growth pattern of
the CEE countries.
60
countries. This paper analyzes the impact of these foreign parent banks on the credit
growth pattern of their affiliated banks operating in the CEE countries.
During the transition period of the CEE countries, foreign banks, mainly from
neighboring Western and Northern European countries, started extending their
operations in the CEE countries by buying equity stakes throughout the privatization
of the state- owned banks. A majority of these foreign-owned banks started to enter
the CEE market, in particular, for high profit opportunities. The high profit
opportunities in the CEE markets are realized via high interest margins between the
bank lending and borrowing rates. Additionally, the high economic growth in these
economies
25
, observed by a strong Real GDP Growth, provided another motivation
for foreigners to extend their operations to the CEE markets. High economic growth
stimulates increased financial activities for the domestic sector. Since the banking
sector is the major source of funding in the CEE countries
26
, growing domestic
demand will broaden the operations and consequently the major income source for
the banks operating in these markets. In addition to opportunities for profit, foreign
owners enter the CEE markets for long-term strategic reasons. Among these reasons,
global or regional diversification is a major motivation. One US and one Japanese
bank entering the CEE markets is an example of global diversification. Western
25
The CEE countries exhibited strong real GDP growth rates, on average around 5.3 percent, for the
period 2001 to 2005. This economic growth out pasted the EU-15 average, which was around 2.7
percent during the same period (source eurostat).
26
The financial sector in the CEE countries ise dominated by the banking sector and direct finance
does not have a significant importance in these markets [ECB (2005A)].
61
European banks operating in the neighboring CEE markets are examples of regional
diversification. Additionally for the European countries, strong historical and cultural
ties constitute another motivation for these banks to enter the CEE markets.
The analysis in this chapter is based on an unbalanced panel dataset which
covers 16 years (1990-2005) and 59 banks, comprised of 18 parent banks and their
41 CEE subsidiaries. The parent banks are from Western Europe, except one
Japanese bank and one US bank. After a detailed study of the data, including
possible multicollinearity problems, a fixed effect within estimator model is
employed for the econometric analysis.
After deciding the econometric model, this chapter investigates whether the
managerial influence of the parent banks vary across each owner. The results show
that parent banks do not vary significantly from each other in terms of their
managerial strategy that they employ in their CEE subsidiaries. A reason for such a
finding may be due to the fact that a majority of the largest banks in the Western
markets are using similar risk management and diversification techniques.
Additionally with respect to their management group, they are hiring employees with
similar backgrounds. Due to these similarities between the largest banks of the
financially advanced markets, parent banks do not indicate varying influence on the
credit growth pattern of their CEE subsidiaries.
The regression results, on the impact of parent banks to the credit growth
pattern of the CEE subsidiaries, indicate that various micro and macroeconomic
62
variables related to the home and host markets are important. With respect to the
micro-level analysis, the results of this paper support the extensions of the growth
theory to the banking structure in the CEE countries: entities grow slower as they get
bigger. However on the parent banks’ side, size has a different interpretation. The
size of the parent bank shows the amount of resources available for a CEE
subsidiary. Therefore the credit growth of the CEE subsidiaries is positively related
to the size of their parent bank.
The results in this paper show that Interbank Liabilities is an important
source of funding for the CEE subsidiaries. A subsidiary can increase the credit line
more, as it can borrow more from the parent bank. Over time, the CEE subsidiaries
rely more on this source, and this indicates additional borrowings from other banks
by relying on the reputation of the parent bank. Additionally, the results on the
Interbank Liabilities of the parent show a negative relation between the parent bank
and its subsidiary. This indicates that a parent bank may need to increase its external
borrowing to sustain the credit growth in its CEE subsidiary. However, one needs to
mention that domestic market borrowing via customer deposits still remains
important for the CEE subsidiaries, even though interbank financing dominates over
time.
The results with respect to the soundness measures of the CEE subsidiaries
show that the influence of these variables diminishes in the most recent years. For
the earlier years, the significance of the soundness measures may be related to the
high amount of bad or impaired loans that the banks collected in their accounts and
63
further, the banking crises that happened during the transition of the CEE banks
towards the market economy structure.
Profitability of the parent bank plays an important role on the growth of
credit for the CEE subsidiary. The results show that profitability of the multinational
bank is more influential on the credit growth pattern of the CEE subsidiary than the
profitability of the subsidiary by itself. The parent bank increases the credit line of
the CEE subsidiary when the operations of the multinational bank, on average, are
profitable. However, the results with respect to the last five years of the sample show
that reverse causality is not significant, and the parent does not cut the credit line of
the subsidiary when the multinational bank’s profitability is on decline.
The regression results relating to the efficiency show that the cost structure of
the parent bank is important. Parent banks tend to reduce the CEE credit line when
the operational costs in the entire conglomerate increase. Even though the parent
bank is cost concerned due to the tight profit margins it is facing, the CEE
subsidiaries incur higher costs to increase their credit line.
The efficiency of a CEE subsidiary in generating income is a major
motivation for a parent bank to increase the credit line at that subsidiary, even though
the CEE subsidiaries show declines in efficiency over time. Results show that the
efficiency of the whole company is more important for credit growth in the CEE
subsidiaries. The direct relation between the efficiency of the conglomerate in
generating income and the credit growth in the CEE subsidiaries shows that the
64
parent banks are seeking to exploit profit opportunities in the CEE markets.
However, as more foreign banks operate in the CEE markets, the magnitude of
competition increase. A rise in the supply of credit in the CEE markets causes an
outward shift of the supply curve, causing a lower equilibrium level of profitability
for the CEE subsidiaries.
The regression results related to the macroeconomic conditions show that the
host market conditions are more important than the home market conditions. Credit
growth in the CEE subsidiary is procyclical with domestic market conditions. With
respect to the interest rate spreads between the bank lending and borrowing rate,
however, the results show that declining spreads in the home markets is a motivation
for parent banks to increase credit line in the CEE subsidiary. Finally, Euro area
monetary policy has an impact on the CEE credit growth in terms of interest rate
policy and money growth.
This chapter has the following structure. Next section explains the dataset
used in this analysis. The econometric model is introduced in section, 4.2 Empirical
Model. The regression results are provided in section 4.3 Results. The last section is
4.4 Conclusion.
65
4.1 Data
The analysis in this paper is based on an unbalanced panel dataset comprised
of annual macro and micro level data. The macroeconomic data is obtained from
three main sources: International Financial Statistics, World Economic Outlook, and
Eurostat. Alternatively, the micro level data is obtained from; BankScope, the
Banker’s Almanac, Privatization Barometer, MIGA, the World Bank, European
central banks, and individual bank’s web sites.
The balance sheet data is gathered from the BankScope database. In
BankScope, balance sheet information is reported under two different accounting
standards. In order to standardize the dataset, all the balance sheets that are reported
under the International Financial Reporting Standards (IFRS) are used when
available and Generally Accepted Accounting Principles (GAAP) otherwise
27
. In this
dataset, balance sheet statements are available under consolidated and
unconsolidated accounts.
28
For the foreign owned banks operating in the CEE
countries, both consolidated and unconsolidated accounts are used according to the
availability of balance sheet information. Use of different statements for the CEE
27
The differences between the IFRS and GAAP accounting standards are assumed to be minimal, and
hence the use of these two standards together should not yield any mismeasurement problems.
28
Consolidated statements contain the balance sheet information of all the companies affiliated with
that bank. If a bank has extended its operations abroad, then the consolidated statements capture the
foreign business transactions of that multinational bank. On the other hand unconsolidated statements
cover only the domestic business transactions of a bank.
66
subsidiaries will hardly create any inconsistencies, since the business scope of these
banks is domestically focused. Similarly, balance sheet data for the parent banks is
collected from the Bank Scope database. However, consolidated statements are used
for parent banks rather than unconsolidated statements. Use of unconsolidated
statements would limit this study only to the domestic activities of the parent bank.
However, a parent bank’s managerial decision would not be solely based on its
domestic activities, considering that for a majority of the parent banks, a significant
amount of balance sheet items are due to the activities of the subsidiaries or branches
operating in different countries.
In order to capture the characteristics of ownership structure across the CEE
subsidiaries and their parent banks, it is useful to refer to Table 12. This is a table,
matching the parent banks and their affiliated banks operating in the CEE countries
as of the last accounting date of that subsidiary reported in the Bank Scope data base.
First of all, this table shows that there are 18 parent banks controlling the biggest
foreign owned banks in the CEE region: four Scandinavian banks control the
banking sector in the Baltic countries, and six Western European countries
dominating in the rest of the CEE countries. In particular Austrian and German
banks have a strong equity stake in the Central European Countries.
Table 12: List of Parent Banks and their Affiliated Companies
Parent Company Home
Country
CEE Subsidiary Host Country
Austria Ceska Sporitelna Czech Republic
Austria Erste Bank Hungary Rt Hungary
Austria Romanian Commercial Bank Romania
Austria Slovak Savings Bank Slovakia
Austria First Building Savings Bank Slovakia
Raiffeisen International Austria Raiffeisenbank EAD Bulgaria
Austria Raiffeisen Bank Zrt Hungary
Austria Tatra Banka Slovakia
Austria Raiffeisen Bank SA Romania
HVB Bank Germany HVB Bank Biochim ad Bulgaria
Germany HVB Bank Czech Republic Czech Republic
Germany HVB Bank Hungary Rt. Hungary
Germany Bank BPH SA Poland
Germany UniBanka Slovakia
Germany HVB Bank Slovakia Slovakia
SOCIÉTÉ GÉNÉRALE France Komercni Banka Czech Republic
France BRD-Groupe Societe Generale Romania
France SKB Banka DD Slovenia
KBC GROEP NV Belgium Ceskoslovenska Obchodni Czech Republic
Belgium K&H Bank Hungary
Banca Intesa SPA Italy CIB Közép Hungary
Italy Vseobecna Uverova Banka Slovakia
Unicredito Italiano SPA Italy Bulbank AD Bulgaria
Italy Bank Pekao SA Poland
Die Erste Österreichische
Spar-Casse
67
Table 12, Continued
Parent Company Home
Country
CEE Subsidiary Host Country
SAMPO PLC Finland AS Sampo Pank Estonia
Finland AB Sampo Bankas Lithuania
DNB NOR Bank ASA Norway AB DnB NORD Bankas Lithuania
Norway AS DnB NORD Banka Latvia
Sweden SEB Eesti Ühispank Estonia
Sweden SEB Latvijas Unibanka Latvia
Sweden SEB Vilniaus Bankas Lithuania
SWEDBANK AB Sweden HansaPank Estonia
Sweden Hansabanka Latvia
Sweden AB Bankas Hansabankas Lithuania
Commerzbank AG Germany BRE Bank Poland
ALPHA BANK AE Greece Alpha Bank Romania Romania
EFG Eurobank Ergasias Greece Bulgarian Post Bank Bulgaria
National Bank of Greece Greece United Bulgarian Bank Bulgaria
ING GROEP NV Netherlands ING Bank Slaski Poland
Nomura of Japan Japan Investicni a Postovni Banka Czech Republic
CITIGROUP INC US Bank Handlowy w Warszawie Poland
Skandinaviska Enskilda
Banken AB
68
69
Second, looking at Table 12, one can see that there is a clustering of foreign
bank ownership, caused by border or region effects. This clustering presents the
regional diversification strategy of the parent banks: a majority of the foreign banks
prefer to extend their operations in countries which are in relatively close proximity
to their headquarters. For example, one can see that Scandinavian countries operate
in Baltic countries, whereas Western European countries such as Austria, Germany,
and Italy have subsidiaries in the Central European countries. A motivation for
parent banks to extend their operations to neighboring countries is the sharing of
similar cultural and historical ties, such as the Austro-Hungarian Empire tie between
the Austrian parent banks and their subsidiaries operating in Hungary.
Third, clustering of the parent banks implies another important feature of
foreign bank ownership in the Central and Eastern Europe: the major foreign players
in the CEE banking system are Western Europeans
29
. This finding indicates that the
business cycles in Western Europe and the monetary policy of this region, in
particular the European Central Bank policy, may have a significant impact on the
credit growth pattern of the CEE countries.
The peculiarities of foreign bank ownership presented in Table 12 provide the
motivation for studying the managerial impact of foreign banks on the credit growth
pattern of the CEE countries. The fact that many CEE subsidiaries operating in
29
The only two exceptions in this sample are one Japanese bank operating in the Czech Republic and
the US Citigroup having subsidiaries in Poland. Such a foreign operation extension can be outlined
within the global diversification strategy of the US and Japan head quartered banks.
neighboring countries share a common owner is of particular interest. 18 parent
banks own the largest banks in the CEE region, and their subsidiaries constitute a
majority of the total banking assets in that region. This paper in particular studies the
impact of such a strong ownership influence coming from Western Europe and the
managerial impact of this ownership on the credit growth pattern of the CEE
countries.
Table 13: Summary Statistics for Parent Banks and their CEE Subsidiaries
Obs Mean Std. Dev Min Max Obs Mean Std. Dev Min Max
Net Loan Growth 271 36.15 39.04 -85.4 196.5 153 10.22 24.6 -100 94.09
Ln(Total Assets) (-1) 254 20.95 1.42 16.7 23.73 140 25.61 1.17 22.14 27.81
Customer Deposits / TA (-1) 270 61.16 15.43 0 87.65 127 41 15.95 20.03 85.36
Interbank Liabilities / TA (-1) 270 17.38 14.37 0 74.01 121 17.65 8.05 1.01 41.91
Loan Loss Provision/ TA (-1) 258 0.66 1.23 -3.27 8.57 125 0.42 0.48 -0.04 3.76
ROA (-1) 271 1.21 3.96 -44.1 23.13 139 0.66 0.79 -2.23 3.42
ROE (-1) 271 18.48 62.11 -199 601.3 139 12.23 12.04 -28.4 49.7
Net Interest Margin (-1) 271 4.82 2.36 -4.92 15.61 139 1.85 0.86 0.1 4.26
Cost to Income Ratio (-1) 269 65.95 34.97 11.67 464.8 139 66.73 20.78 38.81 265.8
Total Capital Ratio (-1) 200 16.39 7.83 0.00 65.2 108 11.16 1.94 8.2 17.3
Real GDP Growth 268 4.58 2.90 -11.9 17.14 158 2.18 1.55 -1.93 5.25
Domestic Credit / GDP (-1) 270 41.21 20.33 11.6 108.7 137 137.2 127.91 1.21 677.1
Spread
S
=Lending-Borrowing* 271 6.18 5.16 -1.37 48.8 158 4.47 1.9 0.44 9.23
Spread
S
-Spread
P
** 204 1.74 5.23 -8.15 41.35
Euro Area Real Int. Rate 271 1.35 1.33 -0.30 7.26
(3 Month)
Euro Area Nom. Int. Rate 271 3.51 1.45 2.11 10.92
(3 Month)
Euro Area M3 Growth 271 6.54 2.11 2.30 12
Domestic-Euro Area Int. Rate 270 6.30 11.97 -1.48 114.8
*
S
Denotes the Subsidiary
**
P
Denotes the Parent Bank
Parent Banks Subsidiary Banks
70
71
For this study, unbalanced panel data of 18 parent banks and their 41 CEE
subsidiaries, for the period 1990 to 2005, are used.
30
Table 13 provides the summary
statistics for parent banks and their subsidiaries operating in the CEE countries.
Since a majority of the parent banks have their headquarters in the EMU countries,
the last four statistics are not reported separately for parent banks.
Looking at this table, one can see that the statistics of interest vary
significantly across parent banks and their CEE subsidiaries. First of all, parent
banks are very big banks. Average Total Assets are significantly larger for parent
banks compared to what they are for their CEE subsidiaries.
Second, the parent bank grows slower: the mean value of Net Loan Growth in
the CEE subsidiary is considerably higher than that of the parent bank. This
characteristic is a reflection of the growth theory: big entities grow slower.
Third, the mean and variance in profitability and efficiency measures are
much smaller for parent banks - except for the mean value of Cost to Income Ratio -
than their CEE subsidiaries. Looking at Table 13, one can see that the CEE
subsidiaries, on average, have higher values for ROA, ROE, and Net Interest Margin:
three different indicators for measuring profitability -or efficiency in generating
income. For parent banks, a smaller mean and variance in their profitability ratio
may be regarded as a motivation to extend their operations in the CEE countries
where there is still room for profit. However, parent banks on average have higher
30
The years covered in this analysis is for the earliest and latest banking year data available for the
subsidiaries operating in CEE countries.
72
operational costs. Cost to Income Ratio, a measure for efficiency, is slightly lower
for the CEE subsidiaries, however still has a considerably high variance. The impact
of size on cost may be a factor underlying this result. Large banks may need to incur
additional operational expenses, such as labor costs, in order to sustain their large
scale of operations in various regions.
Fourth, even though the parent company predicts fewer bad or impaired loans
on average - a smaller value for the mean of Loan Loss Provision over Total Assets
31
- it holds riskier assets in its portfolio compared to its subsidiaries operating in the
CEE countries. Looking at Table 13, one can see that Total Capital Ratio
32
is smaller
for parent banks. This indicates that parent banks have riskier assets on average.
Parent banks can diversify risk more efficiently because they are much larger in scale
and operation, and better at risk management. Therefore, parent banks can afford to
hold riskier assets in their portfolios.
Last, the parent banks have their headquarters in more financially developed
countries. These home countries, on average, have a high Domestic Credit to GDP
ratio and a lower Spread between Lending and Borrowing rates. Due to the highly
competitive markets in their home countries that parent banks are operating in, they
function under very tight interest rate spreads, and therefore, have less room for
31
Loan Loss Provision: This is the amount of provisions which a bank accounts for its expected bad
loans. This amount does not reflect the realized amount of the bad or impaired loans, but only the
expectations of a bank on the amount of risky loans it will incur during that accounting year.
32
Total Capital Ratio is the total capital of a bank divided by the total amount of its risk weighted
assets. As the risky loans of a bank increase, so does the total amount of its risk weighted assets, and
therefore the Total Capital Ratio declines.
73
profit. Once again, high profit margins and rapidly growing economies are a source
of attraction for these banks to extend their operations toward the CEE countries.
4.2 Empirical Model
This paper pools the data in order to analyze the importance of bank
ownership on credit growth. Such a procedure is necessary considering the short time
dimension of the dataset which is limited to 18 years. For the model selection, the
paper employs a static panel data model.
Table 14 presents the pair-wise correlation coefficients between credit growth
and its lagged value for each CEE subsidiary separately. Looking at this Table, only
2 out of 41 subsidiaries have significant correlation coefficients at the five percent
significance level. Considering that the growth variable is the percentage change
within the two consecutive periods, the insignificance of the correlation coefficient
across the growth variables is not an unexpected result. The dominantly insignificant
correlation coefficients imply that the current value of the credit growth does not get
affected from its past value. Therefore, this paper uses a static panel data model
based on the results provided in Table 14.
Table 14: Pairwise Correlation Coefficient between Credit Growth and its Lagged Value for
Each CEE Subsidiary
1-8 9-16 17-24 25-32 33-40 41
Correlation Coefficient 1.00 -0.03 -0.22 0.17 - 0.50
Significance Level 1.00 0.96 0.43 0.89 1.00 0.40
Number of Observations 25 15305
Correlation Coefficient -0.47 -0.79 0.33 -0.12 0.55
Significance Level 0.29 0.06 0.79 0.80 0.63
Number of Observations 7 6 3 7 3
Correlation Coefficient -0.87 -0.23 0.66 0.00 -0.27
Significance Level 0.01 0.71 0.54 1.00 0.83
Number of Observations 7 5 3 7 3
Correlation Coefficient 0.81 -0.19 0.32 0.36 0.98
Significance Level 0.01 0.69 0.61 0.28 0.11
Number of Observations 9 7 5 11 3
Correlation Coefficient -0.17 -0.17 -0.63 - -0.07
Significance Level 0.69 0.71 0.18 1.00 0.93
Number of Observations 8 7 6 0 4
Correlation Coefficient - -0.87 0.14 0.03 0.50
Significance Level 1.00 0.06 0.79 0.96 0.12
Number of Observations 0 5 6 5 11
Correlation Coefficient -0.73 -0.14 -0.02 0.82 -0.43
Significance Level 0.27 0.69 0.96 0.09 0.47
Number of Observations 4 10 11 5 5
Correlation Coefficient -0.38 -0.04 0.50 - 0.88
Significance Level 0.45 0.92 0.25 1.00 0.12
Number of Observations 6 8 7 0 4
Bank Number:
After deciding on a static model, three possible estimation strategies are
considered for the regression analysis: pooled OLS, fixed effects, and random
effects. The general panel data model is specified as follows,
PH
ijt
H
t
PH
t
PH
jt
PH
ijt
PH
ijt
Z X Z X C υ φ β φ β µ + + + + + =
2 2 2 2 1 1 1 1
(4.1)
where,
it i
PH
ijt
u + = α υ
74
The dependent variable is credit growth in a CEE subsidiary, measured by the
percentage growth rate in the net loans of that bank. In Equation (4.1), subscript i is
for the CEE subsidiary, j denotes the country in which the subsidiary is operating –
also known as ‘the host country’, and t is the time subscript. Further, P is a
superscript denoting the parent bank of the CEE subsidiary, and H denotes the
country where the headquarters is located, or ‘the home country’. On the left hand
side of Equation (4.1), µ is the common constant; X
1
and X
2
, respectively are the
matrices for all the bank-specific variables for the CEE subsidiary and the parent
bank; β
1
and β
2
respectively are the vectors of coefficients for the banking variables
of the CEE subsidiary and the parent bank. In order to minimize simultaneity
problems between the balance sheet items and the dependent variable, all balance
sheet variables are lagged one period. Z
1
and Z
2
respectively are the matrices for
country specific macroeconomic variables and φ
1
and φ
1
are the vectors of
coefficients for these variables. Last, Domestic Credit over GDP is lagged one
period because of simultaneity issues.
The Hausman specification test indicates that the random effects model is
biased and the fixed effects estimator better captures the dataset. After choosing the
fixed effects model over the random effects, the choice between the pooled OLS and
the fixed effects model becomes straightforward: The fixed effects model will be
better performing. Additionally, the analysis indicates a rejection for the hypothesis
that all bank specific dummies are equal to zero and second; Akaike and Bayesian
Information criteria yield smaller values for the fixed effects model.
75
76
In panel data analysis, in order to have consistent coefficients on the dummy
variables, the time-space needs to be large. As this dataset does not have a large T-
dimension, a within estimator procedure will be necessary to eliminate the bank-
specific dummies. Additionally, the analysis is not interested in the bank specific
coefficient estimates. The redundancy of estimating the bank specific coefficient
estimates is in favor of using the fixed effects within estimator for the econometric
analysis.
After deciding on a fixed effects within estimator, the standard assumptions
on error terms, independent and identical, are relaxed by allowing for robust and
clustered estimation procedures. Robust standard errors allow for heterogeneity on
the error term. Clustering the error terms within a group assumes that the
observations from the same group may be correlated. Since the dependent variable is
the credit growth in a CEE subsidiary, the standard errors are clustered with respect
to each host country. However, clustering and using a robust estimation method does
not yield larger standard errors. Therefore, the standard assumptions on the error
term will be used as the most efficient classification.
Lastly, time-specific-host-country dummies and year dummies are introduced
to the fixed effects model. Introduction of these dummies will capture the impact of
time-specific macroeconomic shocks, which is peculiar to each host country, and
which is not captured by the macroeconomic indicators. Both Akaike and Bayesian
information criteria indicate that the model is improved after the introduction of
these time-specific-host-country dummies.
77
4.2.1 Concerns about Multicollinearity
A correlation Matrix is provided in Table 15 for both micro and macro level
data observed for the CEE subsidiaries and their parent companies. Looking at this
matrix, one can see that many of the regressors are significantly correlated with each
other, although the correlation coefficients are in most cases small (less than 0.3).
However, the coefficients across similar measurement variables are large in
magnitude, such as the variables for bank profitability and efficiency (ROA, ROE
and Cost-to-Income Ratio) or macroeconomic variables (Spread between Lending
and Borrowing Rate and Spread between Domestic and Euro Area Interest Rate). In
order to eliminate any multicollinearity problems we will introduce these highly and
significantly correlated variables separately into the regression equation.
Table 15: Correlation Coefficients across Variables of Interest
ln(TA)
S
ln(TA)
P
Cust
Deps
Inter.
Liab.
S
Inter.
Liab.
P
LLP
S
LLP
P
ROA
S
ROA
P
ROE
S
ROE
P
NIM
S
NIM
P
ln(Total Assets)(-1)
P
0.39 1
0.00
Customer Deposits / Total
Assets (-1)
S
0.15 -0.09 1
0.02 0.18
Interbank Liabilities / Total
Assets (-1)
S
-0.35 -0.05 -0.74 1
0.00 0.47 0.00
Interbank Liabilities / Total
Assets (-1)
P
0.26 0.32 0.24 -0.17 1
0.00 0.00 0.00 0.02
Loan Loss Provn/ Total
Assets (-1)
S
-0.12 0.10 -0.09 0.09 0.06 1
0.05 0.19 0.16 0.16 0.42
Loan Loss Provn/ Total
Assets (-1)
P
-0.07 0.14 0.07 0.06 0.06 -0.05 1
0.37 0.05 0.31 0.40 0.42 0.52
ROA (-1)
S
0.11 -0.06 -0.03 -0.24 -0.07 -0.39 0.17 1
0.09 0.41 0.67 0.00 0.31 0.00 0.02
ROA (-1)
P
-0.11 -0.44 -0.24 0.11 -0.53 0.05 -0.33 -0.09 1
0.15 0.00 0.00 0.12 0.00 0.50 0.00 0.21
ROE (-1)
S
-0.07 -0.04 0.09 0.06 -0.05 -0.04 0.01 -0.28 -0.06 1
0.24 0.53 0.16 0.30 0.52 0.54 0.90 0.00 0.37
ROE (-1)
P
-0.03 -0.40 -0.07 -0.09 -0.38 0.08 -0.48 -0.02 0.82 0.00 1
0.65 0.00 0.36 0.23 0.00 0.29 0.00 0.75 0.00 0.97
Net Interest Margin (-1)
S
-0.27 -0.16 0.01 -0.17 -0.08 0.14-0.080.100.130.02 0.24 1
0.00 0.03 0.88 0.01 0.25 0.02 0.24 0.11 0.07 0.79 0.00
Net Interest Margin (-1)
P
0.02 -0.31 0.10 -0.05 -0.31 -0.16 0.32 0.08 0.39 -0.02 0.30 0.09 1
0.82 0.00 0.15 0.48 0.00 0.03 0.00 0.30 0.00 0.83 0.00 0.23
Cost to Income Ratio (-1)
S
-0.16 -0.27 -0.01 0.11 -0.09 -0.01 -0.18 -0.58 0.23 -0.19 0.15 -0.09 0.05
0.01 0.00 0.93 0.07 0.24 0.87 0.01 0.00 0.00 0.00 0.03 0.12 0.50
Cost to Income Ratio (-1)
P
0.29 0.38 0.13 -0.10 0.58 -0.02 0.04 -0.05 -0.65 0.04 -0.62 -0.20 -0.35
0.00 0.00 0.08 0.15 0.00 0.78 0.61 0.53 0.00 0.59 0.00 0.01 0.00
Total Capital Ratio (-1)
S
-0.06 0.01 0.26 -0.33 -0.11 -0.13 0.26 0.25 0.03 -0.09 0.07 0.14 0.08
0.45 0.92 0.00 0.00 0.17 0.07 0.00 0.00 0.71 0.20 0.40 0.04 0.30
Total Capital Ratio (-1)
P
-0.10 -0.28 -0.06 0.04 -0.35 0.09 -0.25 -0.03 0.45 -0.31 0.32 0.23 0.16
0.20 0.00 0.42 0.64 0.00 0.28 0.00 0.68 0.00 0.00 0.00 0.00 0.04
Real GDP Growth
S
-0.07 -0.20 0.04 -0.01 -0.19 -0.24 -0.24 -0.04 0.14 -0.16 0.03 0.11 0.05
0.26 0.01 0.52 0.92 0.01 0.00 0.00 0.57 0.04 0.01 0.68 0.07 0.50
Real GDP Growth
P
-0.20 -0.49 -0.13 0.06 -0.34 0.04 -0.12 0.12 0.33 -0.01 0.35 0.36 0.27
0.00 0.00 0.07 0.38 0.00 0.56 0.12 0.08 0.00 0.89 0.00 0.00 0.00
Domestic Credit / GDP (-1)
S
0.15 0.19 -0.08 0.03 0.31 0.01 0.02 0.15 -0.19 0.09 -0.16 -0.28 -0.13
0.02 0.01 0.21 0.63 0.00 0.88 0.78 0.01 0.01 0.12 0.03 0.00 0.07
Spread
S
=Lending-Borrowing
Rate -0.27 -0.25 0.09 -0.04 -0.23 0.21 0.20 0.01 0.06 0.29 0.13 0.11 0.08
0.00 0.00 0.15 0.47 0.00 0.00 0.01 0.84 0.37 0.00 0.07 0.06 0.25
Spread
P
=Lending-Borrowing
Rate -0.16 0.31 0.08 0.07 0.15 -0.11 0.41 0.23 -0.48 0.19 -0.52 -0.11 -0.31
0.02 0.00 0.23 0.32 0.04 0.14 0.00 0.00 0.00 0.01 0.00 0.12 0.00
Spread
S
-Spread
P
-0.20 -0.35 0.04 -0.07 -0.28 0.12 0.07 0.13 0.22 0.36 0.30 0.22 0.18
0.01 0.00 0.60 0.33 0.00 0.09 0.34 0.07 0.00 0.00 0.00 0.00 0.01
Euro Area Short-Term Real
Int. Rate (3 Month) -0.34 -0.13 -0.22 0.16 0.12 0.22 0.03 0.06 0.01 0.10 0.18 0.36 -0.09
0.00 0.07 0.00 0.01 0.11 0.00 0.72 0.33 0.89 0.10 0.01 0.00 0.20
Euro Area Short-Term Nom.
Int. Rate (3 Month) -0.30 -0.09 -0.20 0.13 0.13 0.15 0.03 0.11 0.10 0.08 0.25 0.27 -0.07
0.00 0.20 0.00 0.03 0.08 0.02 0.65 0.07 0.18 0.17 0.00 0.00 0.35
Euro Area M3 Growth 0.17 0.12 0.06 -0.06 0.11 -0.07 -0.19 -0.010.110.000.12 -0.16 -0.03
0.01 0.08 0.34 0.32 0.14 0.24 0.01 0.81 0.11 1.00 0.09 0.01 0.64
Dom. - Euro Area Int. Rate -0.03 -0.06 -0.10 -0.02 -0.06 0.14 0.16 0.08 0.03 0.30 0.06 0.12 0.05
0.62 0.40 0.12 0.80 0.41 0.02 0.03 0.20 0.68 0.00 0.40 0.04 0.51
*
S
denotes subsidiary and
P
denotes the parent bank
Note: Correlation Coefficients Significant at the 5 percent level are marked in bold letters.
78
Table 15, Continued
Cost /
Inc.
S
Cost /
Inc.
P
TCR
S
TCR
P
Real
GDP
S
Real
GDP
P
Dom.
Credit
S
spread
S
spread
P
spread
S
-
spread
P
euro
real int.
euro
nom. int
euro
M3
Cost to Income Ratio (-1)
P
-0.07 1
0.36
Total Capital Ratio (-1)
S
-0.13 -0.16 1
0.08 0.04
Total Capital Ratio (-1)
P
0.35 -0.24 0.01 1
0.00 0.00 0.95
Real GDP Growth
S
0.22 -0.20 -0.09 0.38 1
0.00 0.01 0.24 0.00
Real GDP Growth
P
0.02 -0.34 0.10 0.24 0.10 1
0.78 0.00 0.22 0.00 0.14
Domestic Credit / GDP (-1)
S
-0.25 0.23 -0.29 -0.30 -0.44 -0.29 1
0.00 0.00 0.00 0.00 0.00 0.00
Spread
S
=Lending-Borrowing
Rate -0.06 -0.09 0.29 -0.22 -0.40 0.30 -0.09 1
0.32 0.23 0.00 0.00 0.00 0.00 0.13
Spread
P
=Lending-Borrowing
Rate -0.25 0.21 0.26 -0.40 -0.34 -0.23 0.13 0.21 1
0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.00
Spread
S
-Spread
P
-0.11 -0.16 0.18 -0.11 -0.33 0.38 -0.27 0.95 -0.11 1
0.13 0.03 0.02 0.17 0.00 0.00 0.00 0.00 0.12
Euro Area Short-Term Real
Int. Rate (3 Month) -0.18 -0.10 -0.01 -0.19 -0.43 0.32 0.29 0.27 0.17 0.32 1
0.00 0.16 0.85 0.01 0.00 0.00 0.00 0.00 0.02 0.00
Euro Area Short-Term Nom.
Int. Rate (3 Month) -0.17 -0.14 0.05 -0.11 -0.39 0.11 0.32 0.22 0.17 0.31 0.85 1
0.01 0.05 0.47 0.15 0.00 0.13 0.00 0.00 0.02 0.00 0.00
Euro Area M3 Growth 0.03 0.00 0.03 0.09 0.05 -0.37 -0.08 -0.18 -0.08 -0.16 -0.36 -0.07 1
0.63 0.98 0.67 0.22 0.40 0.00 0.22 0.00 0.24 0.02 0.00 0.22
Dom. - Euro Area Int. Rate -0.19 0.06 0.01 -0.36 -0.46 0.17 0.23 0.77 0.17 0.77 0.33 0.28 -0.22
0.00 0.40 0.86 0.00 0.00 0.02 0.00 0.00 0.01 0.00 0.00 0.00 0.00
*
S
denotes subsidiary and
P
denotes the parent bank
Note: Correlation Coefficients Significant at the 5 percent level are marked in bold letters.
4.2.2 Do Parent Banks have different Managerial Strategies?
In all the models that were considered in this section, we assumed common
coefficients across parent banks. Since the paper tracks the importance of the
managerial impact of parent banks on the credit growth of the CEE subsidiaries, a
relevant extension is to test whether different parent banks have different
79
coefficients. Therefore, Equation (4.1) is modified in order to capture parent bank
specific coefficients.
PH
ijt
P
H
t
P
P
PH
t
P
P
PH
jt
P
P
PH
ijt
P PH
ijt
Z X Z X C υ φ β φ β µ + + + + + =
∑ ∑ ∑ ∑
Θ ∈ Θ ∈ Θ ∈ Θ ∈
2 2 2 2 1 1 1 1
(4.2)
In Equation (4.2) all the variables of interest are allowed to vary across
different parent banks. After obtaining the estimation results, any difference in a
coefficient across parent banks will indicate the variation on the managerial impact
due to difference in ownership. As given in Table 12, parent banks own different
numbers of subsidiaries which operate in the CEE countries. The numbers of
affiliated banks that are owned by the same parent company vary from one to six and
also, the number of balance sheet years available for each bank changes from one to
sixteen. This characteristic of the dataset creates a restriction for estimating Equation
(4.2) for each parent bank separately. Therefore Equation (4.2) and the common
coefficient test are applied only to the groups where a parent bank owns at least three
subsidiaries. Among these parent banks, the minimum number of banking year
observations is 22 and the maximum is 43. Due to the limited number of banking
year observations, these models used only the three most important regressors for
credit growth: Total Assets over GDP, Interbank Liabilities over Total Assets, and
Real GDP Growth. These regressors control the size of a bank, its financing stream,
and also the business cycles in the country that it is operating in. Table 16 provides
the estimation results for the three variations of Equation (4.2) with these
aforementioned regressors.
80
Table 16: FE regression results for cross-sectional varying coefficients
Model 1 Model 2 Model 3
ln(Total Assets) (-1)
H
-23.34 -21.4
[0.00]*** [0.00]***
ln(Total Assets) (-1)
E
1.85 -16.22
[0.94] [0.59]
ln(Total Assets) (-1)
SW
-37.47 -24.94
[0.00]*** [0.00]***
ln(Total Assets) (-1)
SK
-48.78 -36
[0.00]*** [0.00]***
Real GDP Growth
H
3.99 -5.74
[0.25] [0.20]
Real GDP Growth
E
3.44 3.38
[0.55] [0.67]
Real GDP Growth
SW
7.39 0.72
[0.02]** [0.85]
Real GDP Growth
SK
3.58 -1.85
[0.14] [0.50]
Interbank Liabilities/ T.A. (-1)
H
-0.48 -0.71
[0.27] [0.27]
Interbank Liabilities/ T.A. (-1)
E
3.91 0.95
[0.19] [0.71]
Interbank Liabilities/ T.A. (-1)
SW
1.32 2.36
[0.19] [0.06]*
Interbank Liabilities/ T.A. (-1)
SK
-0.41 -1.1
[0.55] [0.20]
Constant 587.17 551.87 54.77
[0.00]*** [0.00]*** [0.00]***
Observations 99 99 103
Number of bankid 17 18 17
R-squared 0.4 0.37 0.09
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
H
denotes HVB Bank (Germany),
E
denotes Die Erst Bank (Austria)
SW
denotes Swedbank (Sweden),
SK
denotes Skandinavska bank (Sweden)
81
After obtaining the coefficient estimates from Table 16, a Wald test is
performed for each regressor for the null hypothesis that all coefficients are equal
across different parent banks. Table 17 reports the F statistics of the Wald test for
each regressor and for each model. Looking at this table, the test statistic fails to
reject the equality of the coefficient estimates for each of the three most important
regressors. Since the data restrictions make it difficult to apply the same test across
each parent bank in the dataset, the results obtained from these three parent banks are
assumed to be valid across other parent banks owning subsidiaries in the CEE
countries.
Table 17: Test of Different Coefficient Estimates across Affiliated Banks
Model 1
ln(Total Assets)(-1)
H
= ln(Total Assets)(-1)
E
= ln(Total Assets)(-1)
SW
= ln(Total Assets)(-1)
SK
F( 3, 74) = 1.79
Prob > F = 0.1565
Real GDP Growth
H
= Real GDP Growth
E
= Real GDP Growth
SW
= Real GDP Growth
SK
F( 3, 74) = 0.34
Prob > F = 0.7986
Model 2
ln(Total Assets)(-1)
H
= ln(Total Assets)(-1)
E
= ln(Total Assets)(-1)
SW
= ln(Total Assets)(-1)
SK
F( 3, 73) = 0.43
Prob > F = 0.7322
Interbank Liab/ T.A.(-1)
H
= Interbank Liab/ T.A.(-1)
E
= Interbank Liab/ T.A.(-1)
SW
= Interbank Liab/ T.A.(-1)
SK
F( 3, 73) = 1.60
Prob > F = 0.1976
Model 3
Real GDP Growth
H
= Real GDP Growth
E
= Real GDP Growth
SW
= Real GDP Growth
SK
F( 3, 78) = 0.55
Prob > F = 0.6484
Interbank Liab/ T.A.(-1)
H
= Interbank Liab/ T.A.(-1)
E
= Interbank Liab/ T.A.(-1)
SW
= Interbank Liab/ T.A.(-1)
SK
F( 3, 78) = 2.05
Prob > F = 0.1136
82
83
There are several reasons why parent banks do not vary in terms of their
managerial impact on the credit growth pattern of their CEE subsidiary. Among
these reasons, the first one is the commonality of the risk management rules that they
use and the second one is the pool of employees that they hire from. The largest
banks in the financially developed markets use similar risk management techniques.
In terms of the quality of their management, these banks hire employees with similar
backgrounds. All these commonalities between the largest banks in Western Europe
undermine the differences across their influence pattern on the growth of credit in
their CEE subsidiaries. Due to these results, this paper will use Equation (4.1) and its
variations, in order to analyze the managerial impact of the parent banks on the credit
growth pattern of their CEE subsidiaries.
4.3 Results
In this section results are derived from the fixed effect with estimator
regression analysis of Equation (4.1) for various banking and macroeconomic
variables. Regression results for various microeconomic variables are presented from
Table 18 to Table 25. As presented in Table 15, Real GDP Growth and Domestic
Credit over GDP does not indicate any signs of multicollinearity with the banking
sector variables. Therefore these two variables are included, in addition to the
microeconomic variables, to the regression analysis provided in these tables.
84
However, as presented in Table 18, Domestic Credit over GDP does not yield any
significant results. Therefore this variable will be excluded from the rest of the
analysis. In most of the regression results presented in Table 18 through Table 23,
the banking variables are first introduced only for the CEE subsidiary and then
together with the parent bank. Since many of the banking performance variables are
highly correlated between the parent bank and its subsidiaries, introducing them
together causes an insignificant coefficient for the subsidiary.
Next, Table 24 and Table 25 present the regression coefficients for various
macroeconomic variables. These tables report estimation results first for Equation
(4.1) for all the available banking-year data, and then it includes time dummies for
any macroeconomic shocks common to the CEE area.
Last, this paper considers these two methods for the last five years of the
dataset where there is less turbulence in terms of the variation of microeconomic
variables. Unlike the previous chapter, the smaller number of observations in the
earlier years restricts this chapter to apply the same analysis to the pre-2001 period.
In this section, results based on the impact of the balance sheet items of the
CEE subsidiary and the parent bank on the credit growth pattern of the subsidiary are
reported in Table 18 to Table 23, and they are analyzed in the section 4.4.1 Impact of
Microeconomic Shocks. Results based on the impact of the macroeconomic
indicators in the host and home countries on the credit growth of the CEE subsidiary
are reported in Table 24 and Table 25, and their interpretation is provided under the
section 3.4.2 Impact of Macroeconomic Shocks.
Table 18: FE Estimator for Credit Growth in the CEE Subsidiaries
ln(Total Assets) (-1) -26.65 -42.9 -33.67 -55.75 -21.52 -44.48 -34.3 -49.21
[0.00]*** [0.00]*** [0.00]*** [0.00]*** [0.00]*** [0.00]*** [0.00]*** [0.00]***
ln(Total Assets) of Parent (-1) 31.08 28.43 47.02 47.81
[0.02]** [0.03]** [0.00]*** [0.00]***
Customer Deposits/ T.A. (-1) 0.21 0.84 0.83 1.18 0.31 0.87 0.54 0.85
[0.66] [0.06]* [0.16] [0.05]* [0.53] [0.05]* [0.36] [0.16]
Interbank Liabilities/ T.A. (-1) 0.52 0.8 1.51 1.49 0.42 0.66 1.06 1.12
[0.23] [0.04]** [0.00]*** [0.00]*** [0.35] [0.09]* [0.04]** [0.03]**
Interbank Liabilities/ T.A. of Parent (-1) -1.03 -1.21 -1.91 -1.88
[0.08]* [0.02]** [0.00]*** [0.00]***
Real GDP Growth Rate 5.51 0.77 4.68 0.35 6.26 1.42 5.04 2.32
[0.00]*** [0.57] [0.01]*** [0.88] [0.00]*** [0.29] [0.00]*** [0.34]
Domestic Credit over GDP (-1) -0.28 -0.32 -0.14 -0.1
[0.26] [0.17] [0.75] [0.81]
Constant -243.11 179.76 -563.88 -120.78 452.71 957.04 725.12 1032.39
[0.40] [0.63] [0.12] [0.77] [0.00]*** [0.00]*** [0.00]*** [0.00]***
Time Dummies No Yes No Yes No Yes No Yes
Observations 186 186 138 138 180 180 134 134
Number of bankid 37 37 36 36 35 35 35 35
R-squared 0.25 0.49 0.29 0.35 0.25 0.5 0.29 0.33
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
Whole Sample 2001-2005 Period Whole Sample 2001-2005 Period
Model 1 Model 2
85
Table 19: FE Estimator for Credit Growth in the CEE Subsidiaries
ln(Total Assets) (-1) -21.45 -41.65 -35.86 -68.09 -15.83 -50.56 -28.53 -58.32
[0.00]*** [0.00]*** [0.00]*** [0.00]*** [0.01]*** [0.00]*** [0.01]** [0.00]***
Customer Deposits/ T.A. (-1) 0.53 1.08 1.13 1.42 0.68 0.83 0 0.67
[0.37] [0.05]** [0.08]* [0.03]** [0.38] [0.22] [1.00] [0.44]
Interbank Liabilities/ T.A. (-1) 0.66 0.98 1.46 1.56 1 1.04 1.12 1.26
[0.22] [0.04]** [0.01]*** [0.00]*** [0.12] [0.05]** [0.07]* [0.04]**
Loan Loss Provision/T.A. (-1)
♣
-10.18 -7.53 -4.27 -5.89
[0.02]** [0.05]* [0.30] [0.14]
Loan Loss Provision/T.A. of Parent (-1) -18.96 -10.96 -20.27 -20.8
[0.30] [0.55] [0.25] [0.25]
Total Capital Ratio (-1) -1.56 0.35 -1.32 -2.03
[0.48] [0.85] [0.59] [0.39]
Total Capital Ratio of Parent (-1) 1.09 1.52 0.16 1.14
[0.03]** [0.00]*** [0.82] [0.15]
Real GDP Growth Rate 4.73 0.61 4.26 -0.68 3.99 1.08 6.04 1.64
[0.00]*** [0.67] [0.02]** [0.77] [0.01]** [0.47] [0.01]*** [0.58]
Constant 434.76 870.25 695.86 1384.92 289.37 1048.77 607.98 1203.95
[0.00]*** [0.00]*** [0.00]*** [0.00]*** [0.06]* [0.00]*** [0.03]** [0.01]***
Time Dummies No Yes No Yes No Yes No Yes
Observations 165 165 124 124 129 129 99 99
Number of bankid 34 34 34 34 27 27 27 27
R-squared 0.24 0.51 0.25 0.38 0.18 0.55 0.23 0.33
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
♣
Values outside [-1,3] range are excluded.
Whole Sample 2001-2005 Period Whole Sample 2001-2005 Period
Model 4 Model 3
86
Table 20: FE Estimator for Credit Growth in the CEE Subsidiaries
ln(Total Assets) (-1) -16.78 -51.14 -27.07 -46.19 -13.61 -52.34 -26.15 -48.27
[0.00]*** [0.00]*** [0.00]*** [0.00]*** [0.01]** [0.00]*** [0.00]*** [0.00]***
Customer Deposits/ T.A. (-1) 0.26 0.73 0.61 0.95 0.28 0.71 0.61 0.99
[0.59] [0.09]* [0.33] [0.13] [0.56] [0.10] [0.33] [0.12]
Interbank Liabilities/ T.A. (-1) 0.53 0.71 1.42 1.42 0.63 0.78 1.42 1.44
[0.22] [0.06]* [0.01]*** [0.01]*** [0.16] [0.04]** [0.01]*** [0.01]***
Net Interest Margin (-1) 2.7 3.26 0.96 -2.02
[0.22] [0.11] [0.80] [0.62]
Net Interest Margin of Parent (-1) 18.28 18.45 9.75 9.76 16.78 14.86 9.59 10.71
[0.07]* [0.05]* [0.38] [0.42] [0.09]* [0.12] [0.39] [0.39]
Real GDP Growth Rate 4.7 1.55 4.48 0.94 4.53 1.11 4.57 0.69
[0.00]*** [0.25] [0.01]** [0.71] [0.00]*** [0.42] [0.01]** [0.79]
Constant 312.23 1061.88 512.42 911.44 233.16 1084.31 489.24 962.86
[0.00]*** [0.00]*** [0.01]*** [0.00]*** [0.06]* [0.00]*** [0.02]** [0.00]***
Time Dummies No Yes No Yes No Yes No Yes
Observations 183 183 136 136 183 183 136 136
Number of bankid 35 35 34 34 35 35 34 34
R-squared 0.18 0.46 0.21 0.27 0.19 0.47 0.21 0.27
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
Whole Sample 2001-2005 Period Whole Sample 2001-2005 Period
Model 5 Model 6
87
Table 21: FE Estimator for Credit Growth in the CEE Subsidiaries
ln(Total Assets) (-1) -11.37 -25.12 -6.74 -16.6 -17.15 -52.01 -26.1 -43.8
[0.00]*** [0.00]*** [0.34] [0.14] [0.00]*** [0.00]*** [0.00]*** [0.00]***
Customer Deposits/ T.A. (-1) 0.38 0.7 0.65 0.99 0.51 1 1 1.2
[0.37] [0.11] [0.21] [0.06]* [0.31] [0.03]** [0.12] [0.07]*
Interbank Liabilities/ T.A. (-1) 0.45 0.56 1.29 1.36 0.68 0.86 1.51 1.48
[0.25] [0.15] [0.01]*** [0.01]*** [0.13] [0.03]** [0.00]*** [0.01]***
Cost to Income Ratio (-1)
♣
0.24 0.27 0.43 0.52 0.14 0.18 0 0.05
[0.06]* [0.06]* [0.02]** [0.01]*** [0.46] [0.32] [0.98] [0.86]
Cost to Income Ratio of Parent (-1) -0.45 -0.57 -0.86 -0.7
[0.22] [0.07]* [0.03]** [0.09]*
Real GDP Growth Rate 3.36 2.97 4.39 2.22 4.99 1.51 5.74 2
[0.00]*** [0.00]*** [0.02]** [0.37] [0.00]*** [0.28] [0.00]*** [0.45]
Constant 211.35 450.56 63.38 249.4 352.49 1119.08 533.72 898.79
[0.00]*** [0.00]*** [0.68] [0.30] [0.00]*** [0.00]*** [0.01]*** [0.00]***
Time Dummies No Yes No Yes No Yes No Yes
Observations 246 246 162 162 182 182 135 135
Number of bankid 40 40 39 39 35 35 34 34
R-squared 0.11 0.17 0.14 0.19 0.17 0.47 0.26 0.29
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
♣
Values greater than 150 are excluded.
Whole Sample 2001-2005 Period Whole Sample 2001-2005 Period
Model 8 Model 7
88
Table 22: FE Estimator for Credit Growth in the CEE Subsidiaries
ln(Total Assets) (-1) -10.48 -25.65 -7.42 -17.13 -19.36 -54.96 -20.78 -47.88
[0.00]*** [0.00]*** [0.30] [0.14] [0.00]*** [0.00]*** [0.02]** [0.00]***
Customer Deposits/ T.A. (-1) 0.25 0.53 0.81 1.14 0.37 1.03 1.04 1.54
[0.55] [0.22] [0.11] [0.03]** [0.47] [0.02]** [0.11] [0.02]**
Interbank Liabilities/ T.A. (-1) 0.35 0.42 1.28 1.31 0.65 0.79 1.53 1.46
[0.38] [0.28] [0.01]*** [0.01]*** [0.15] [0.04]** [0.00]*** [0.01]***
ROE (-1)
♣
-0.14 -0.13 -0.29 -0.32 0.18 0.02 -0.17 -0.23
[0.22] [0.28] [0.10]* [0.07]* [0.21] [0.90] [0.48] [0.33]
ROE of Parent (-1) 0.08 0.59 0.78 1.16
[0.82] [0.09]* [0.03]** [0.01]***
Real GDP Growth Rate 3.29 2.95 4.56 2.51 4.53 -0.41 5.84 0.44
[0.00]*** [0.01]*** [0.01]** [0.30] [0.00]*** [0.79] [0.00]*** [0.86]
Constant 220.34 487.06 99.75 311.93 388.06 1163.12 352.03 954.14
[0.00]*** [0.00]*** [0.52] [0.22] [0.00]*** [0.00]*** [0.08]* [0.00]***
Time Dummies No Yes No Yes No Yes No Yes
Observations 246 246 163 163 180 180 135 135
Number of bankid 40 40 39 39 35 35 34 34
R-squared 0.09 0.15 0.15 0.18 0.16 0.48 0.24 0.32
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
♣
Values outside [-150,150] range are excluded.
Model 9 Model 10
Whole Sample 2001-2005 Period Whole Sample 2001-2005 Period
89
Table 23: FE Estimator for Credit Growth in the CEE Subsidiaries
ln(Total Assets) (-1) -11.16 -32.17 -6.74 -14.73 -19.58 -43.13 -22.67 -46.48
[0.00]*** [0.00]*** [0.34] [0.19] [0.00]*** [0.00]*** [0.01]** [0.00]***
Customer Deposits/ T.A. (-1) -0.12 0.17 0.78 1.13 0.4 1.01 0.95 1.27
[0.75] [0.68] [0.12] [0.03]** [0.43] [0.02]** [0.14] [0.05]*
Interbank Liabilities/ T.A. (-1) 0.04 0.19 1.23 1.27 0.6 0.78 1.48 1.42
[0.92] [0.62] [0.01]** [0.01]*** [0.18] [0.04]** [0.01]*** [0.01]***
ROA (-1) 1.24 1.49 -3.68 -4.15 0.83 -1.61 -0.63 -0.69
[0.05]* [0.03]** [0.04]** [0.02]** [0.35] [0.07]* [0.78] [0.77]
ROA of Parent (-1) 8.92 15.53 15.29 18.07
[0.27] [0.06]* [0.05]* [0.05]**
Real GDP Growth Rate 4.27 3.69 4.45 2.5 6.01 0.65 5.38 0.8
[0.00]*** [0.00]*** [0.01]** [0.29] [0.00]*** [0.63] [0.00]*** [0.75]
Constant 253.54 629.11 89.52 262.84 380.35 896.11 399.79 903.39
[0.00]*** [0.00]*** [0.56] [0.30] [0.00]*** [0.00]*** [0.05]** [0.00]***
Time Dummies No Yes No Yes No Yes No Yes
Observations 251 251 164 164 184 184 136 136
Number of bankid 40 40 39 39 35 35 34 34
R-squared 0.14 0.22 0.16 0.19 0.23 0.5 0.24 0.3
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
Whole Sample 2001-2005 Period Whole Sample 2001-2005 Period
Model 11 Model 12
90
Table 24: FE Estimator for Credit Growth in the CEE Subsidiaries
Whole Sample 2001-2005 Whole Sample 2001-2005
ln(Total Assets) (-1) -14.71 -15.84 -20.54 -24.28
[0.00]*** [0.07]* [0.00]*** [0.01]**
Customer Deposits/ T.A. (-1) 0.45 1.63 0.64 1.9
[0.37] [0.01]** [0.20] [0.00]***
Interbank Liabilities/ T.A. (-1) 0.31 1.94 0.3 1.9
[0.51] [0.00]*** [0.51] [0.00]***
Real GDP Growth Rate 5.51 4.67 3.03 4.84
[0.00]*** [0.04]** [0.05]* [0.02]**
Real GDP Growth Rate of Parent 1.81 0.76
[0.41] [0.81]
Spread
♣
-1.63 1.59
[0.07]* [0.46]
Spread Parent -6.38 -19.92
[0.19] [0.01]***
Constant 281.64 211.43 446.28 451.82
[0.01]*** [0.26] [0.00]*** [0.04]**
Observations 192 139 192 139
Number of bankid 39 38 39 38
R-squared 0.17 0.15 0.21 0.21
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
♣
Spread: Difference between lending and borrowing rate
Model 14 Model 13
91
Table 25: FE Estimator for Credit Growth in the CEE Subsidiaries
Whole Sample 2001-2005 Whole Sample 2001-2005 Whole Sample 2001-2005
ln(Total Assets) (-1) -21.35 -18.65 -22 -18 -12.96 -6.68
[0.00]*** [0.06]* [0.00]*** [0.05]* [0.00]*** [0.30]
Customer Deposits/ T.A. (-1) 0.48 1.69 -0.24 0.75 -0.08 0.58
[0.33] [0.01]** [0.28] [0.07]* [0.71] [0.15]
Interbank Liabilities/ T.A. (-1) 0.37 1.85 -0.03 1.35 -0.08 1.35
[0.43] [0.00]*** [0.90] [0.00]*** [0.74] [0.00]***
Real GDP Growth Rate 3.37 4.08 2.16 0.99 1.39 3.42
[0.03]** [0.08]* [0.00]*** [0.63] [0.02]** [0.05]**
Difference bt Subsidiary Parent Spread
♣
-1.8 3.89
[0.04]** [0.12]
Euro Area Real Interest Rate
♣♣
-2.98 -6.77
[0.50] [0.23]
Euro Area S-T Interest Rate (3 Month) -5.45 -10.93
[0.00]*** [0.05]*
Spread bw domestic & euro int. rate -0.64 0.53
[0.00]*** [0.44]
Euro Area M3 Growth 2.47 2.96
[0.00]*** [0.11]
Constant 438.99 273.96 496.19 346.39 302.8 97.34
[0.00]*** [0.20] [0.00]*** [0.09]* [0.00]*** [0.47]
Observations 192 139 558 269 551 269
Number of bankid 39 38 70 65 70 65
R-squared 0.19 0.18 0.12 0.08 0.11 0.07
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
♣
Spread: Difference between lending and borrowing rate
♣♣
Derived from expected inflation
Model 17 Model 15 Model 16
92
93
4.4.1 Impact of Microeconomic Shocks
Smaller banks grow faster. In all the regression results provided from Table
18 to Table 25, Total Assets of the CEE subsidiary has a significant and negative
coefficient. As a reflection of the growth literature, this presents that the growth rate
of an entity diminishes as it gets larger in magnitude.
Credit tends to grow faster the larger the size of parent bank. The size of
the parent bank, which is measured by the ln(Total Assets), is significant and positive
as it is presented in Table 18. This shows that as the parent bank grows there are
more resources for the CEE subsidiary to use in their host country. Since the CEE
subsidiary is much smaller in scale compared to its parent bank (refer to Table 13), a
big mother company allows the subsidiary to increase the credit line more easily via
the resources coming from the parent bank.
Interbank Liabilities is a more important source of funding for the CEE
subsidiaries. Looking at the regression results presented from Table 18 to Table 25,
coefficient estimates for Customer Deposits over Total Assets and Interbank
Liabilities over Total Assets present the importance of these two main sources of
financing for the credit growth of the CEE subsidiaries. Even though both means of
funding is significant for the growth rate of credit, looking at the regressions results
for the most recent five years, one can see that Interbank Liabilities over Total Assets
have larger coefficient estimates. This indicates that over time subsidiaries became
less dependent on the domestic market conditions, i.e. Customer Deposits, since they
94
can easily raise their funds either from their parent bank, or through borrowing from
other banks via the credibility of their parent company.
Interbank liability of the parent is negatively related to the credit growth
rate in the CEE subsidiary. The coefficient of Interbank Liabilities of the parent
bank is significant and negative as presented in Table 18. This may indicate that a
parent bank needs to borrow externally to get financing for its subsidiaries. The
intuition may also be drawn that as the parent bank lends more to its CEE subsidiary,
resulting in a decline in its Interbank Liabilities, it gets easier for the subsidiary to
increase its credit line.
Soundness Indicators lose importance for the CEE subsidiaries during
the last five years. Looking at the regression results provided in Table 19, one can
see that Loan Loss Provision over Total Assets and Total Capital Ratio of the parent
bank have significant coefficient estimates for the whole sample. A negative
coefficient for the Loan Loss Provision indicates that as the impaired loan
allowances increase, a bank tends to decrease its credit line. Similarly a positive
coefficient for Total Capital Ratio for the parent bank indicates that as the number of
risky loans increase in their portfolio (a larger denominator) they tend to increase
credit less in their subsidiaries. This result may be due to the fact that many banking
crises that the CEE countries experienced during the 1990s had a major impact on
their credit growth pattern. However, looking at the last five years of the dataset, one
can see that all the soundness indicators become insignificant, as an indication of a
stronger banking system.
95
Profit Margins lose importance over time. Looking at Table 20, one can
see that Net Interest Margin of the parent bank has a significant and positive
coefficient for the whole sample. Regression results in Table 20 first report the Net
Interest Margin of the parent bank, and then report the same variable by adding the
Net Interest Margin of the subsidiary. Results show that interest margins of the
parent bank is the leading factor for the CEE credit growth. A positive coefficient
shows that the parent bank increases the credit line more, when the home market
conditions are more welcoming. However, looking at the last five years of the
dataset, one can see that Net Interest Margin becomes insignificant. The amount of
competition in the financially developed markets increased much more in the 2000s.
Since saturated markets yield less profit, interest margins in the home markets
declined sharply. However, a decline in these margins does not create a similar effect
for the CEE credit growth. This indicates that the parent bank does not cut credit
when the home market conditions worsen.
CEE subsidiaries need to incur higher costs to increase their credit line.
As presented in Table 21, Cost to Income Ratio of the CEE subsidiary is positive and
significant through the whole sample and further, its coefficient becomes larger
during the last five years of the dataset.
The parent bank reduces the credit line in its subsidiary if it becomes
more costly to operate. As reported in Table 21, the Cost to Income Ratio of the
parent bank has a significant and negative coefficient. This indicates that parent
banks are more cost concerned, and they tend to decrease the credit line in the
subsidiary if they are incurring high operational costs. Parent banks have to be more
96
concerned about the operational costs because they operate under very tight interest
margins.
The ROE and ROA of a subsidiary is negatively related to its credit
growth. Table 22 and Table 23 respectively present the coefficient estimates for
ROE and ROA. Looking at these tables, one can see that both variables have
significant and negative coefficients for the last five years of the dataset. ROE and
ROA are variables measuring the efficiency of a bank in generating income,
respectively, per equity and per asset. A decline in these ratios reflects the increased
magnitude of competition in the CEE markets. Since more banks operate in the CEE
markets by the 2000s compared to the earlier years of the dataset, subsidiaries face
declining returns on average. However one should note that a declining ROE ratio
may also indicate higher capitalization for a subsidiary.
The ROE and ROA of the parent bank are positively related to the
subsidiary’s credit growth. Looking at Table 22 and Table 23, one can see that
ROE and ROA have significant and positive ratios for the parent bank. An important
reason for a parent bank to open branches in other than home markets is to exploit
profit opportunities in those markets, and positive estimates on both of these
profitability ratios support this goal of the parent banks.
97
4.4.2 Impact of Macroeconomic Shocks
Credit growth is procyclical with the host country conditions. In all the
regression results presented from Table 18 to Table 25, Real GDP Growth Rate has a
positive and significant coefficient. This indicates that business cycles in the host
country have a direct effect on the credit growth in the CEE subsidiaries: credit
grows more during the expansionary periods and less during recessions.
Credit growth is not responsive to the home country conditions. A major
concern over foreign ownership is the possibility that a parent bank may cut the
credit line in their subsidiaries if they face negative shocks in the home country.
However, as provided in Table 24, the coefficient estimates for Real GDP Growth
Rate in the home country are insignificant in both samples. Since a majority of
parent banks own subsidiaries which operate in other than home markets, a negative
shock at their headquarters has less influence in the growth of credit for these
affiliated banks.
Declining spreads indicate financial deepening for the CEE subsidiaries
in the earlier years. Table 24 reports the coefficient estimates for the host market
spreads. These spreads are significant and negative for the whole sample. Similar to
the efficiency ratios, a rise in the credit line shits the supply curve outward, leading
to lower equilibrium level of spread, and hence a negative coefficient estimate for
this variable. However, looking at the last five years of the dataset, one can see that
this significance vanishes as the CEE financial markets deepen.
98
Tight spreads in the home market is a motivation for parent banks to
increase the credit line in their subsidiaries operating in the CEE countries.
Looking at the estimation results in Table 24, for the last five years of the dataset, the
coefficient for the Spread between Lending and Borrowing Rate is significant and
negative for parent banks. The underlying result for a negative coefficient is because
all the home markets, except Germany, have very tight spreads in the second half of
the dataset. Since the home country offers less room for profit, these banks need to
move to other markets to earn profits from relatively higher spreads. Additionally,
the spread for the host market turns insignificant when jointly introduced to the
regression with home market spreads during the last five years of the dataset. This
implies that the home market has dominance in the credit growth pattern of the
subsidiary during these years.
Euro Area interest rates matter for the foreign banks. The coefficient for
Short-Term Euro-Area Nominal Interest Rates is significant and negative (refer to
Table 25) and the coefficient for the Short-Term Euro-Area Real Interest Rates is
insignificant. This indicates a supply effect coming from financial markets. The
Mundell-Flemming framework indicates that, financial markets will opt to only
nominal interest rate changes. The estimation results in this table show results in
favor of this financial markets theory. Additionally, a negative coefficient for the
Short-Term Euro-Area nominal interest rates show that as the interest rates are
falling in the EMU countries, which is the home market for a majority of the parent
banks, they increase the lending in their host countries which are operating under
higher interest rates.
99
Financial integration is a driving force for credit growth for the CEE
subsidiaries. The coefficient for the Difference between the Host Market and the
Home Market Spreads and the coefficient for the Difference between the Host
Country and Euro Area Interest Rates are significant and negative (Table 25). This
shows that as the CEE countries converge toward the Euro area financial markets
level, the credit line increases in these countries. Convergence may be interpreted as
the rise in the number of financial intermediaries operating in the CEE countries.
Over time there will be more institutions offering loans to the domestic sector, and
an increased credit line will cause a decline in the interest rate spreads. Looking at
the last five years of the dataset, one can see that both coefficients yield insignificant
results, since the spreads between the Euro area and the host market interest rates
become minimal after 2000.
Euro-area money growth is positively related to credit growth in the
CEE subsidiaries. The coefficient for Euro-area M3 Growth is positive and
significant (refer to Table 25). This indicates that a significant portion of the credit in
the CEE countries is financed through the EMU markets, and the CEE subsidiaries
owned by the EMU banks are the driving force behind these results.
100
4.4 Conclusion
This paper studies the impact of foreign bank ownership in credit growth for
ten Central and Eastern European countries. The paper uses panel data models via
four versions of a fixed-effect-within estimator model. The first two versions are for
the whole sample - one with country-time dummies and one without these dummies -
and the other two versions are for the last five years of the dataset. The analysis on
the CEE subsidiaries also includes tests on possible micro-level data problems, and
then presents results on the role of bank ownership in credit growth.
The analysis in the paper tracks the importance of parent banks in the credit
growth pattern of their CEE subsidiaries. The paper shows that the size of the bank is
an important variable: credit grows parallel to the size of the parent bank. The
significance of inter-bank lending is present also for the parent banks’ interbank
liabilities. Home country business cycles do not have a significant influence on the
credit growth pattern of the CEE subsidiaries. The supply side effect can be
highlighted through the spread between lending and borrowing rate in the home
market. As spreads are falling at the home market, the parent bank is more attracted
to lending through its CEE subsidiaries which operate under higher margins. The
paper shows that lending in the CEE countries is not only procyclical with respect to
the host market conditions, but also with respect to home market conditions. Last,
Euro area interest rates and money growth is significant for the credit growth of
foreign banks operating in the CEE countries.
101
Chapter 5
The Lessons from the EMU Countries
In the run-up to EU membership and since then, the Central and Eastern
European (CEE) countries have experienced rapid credit growth, current account
deficits, capital inflows, rising asset prices, and robust domestic demand growth.
These dynamics and their formerly high inflation rates resemble those of the older
member states of the European Monetary Union (EMU) in particular; Greece,
Ireland, Italy, Portugal, and Spain before and during the introduction of the Euro.
These similarities between the CEE countries and the five EMU member states
(EMU-5), especially in terms of credit growth, motivate this paper to draw a future
transition path for the CEE countries toward their EMU integration by studying the
experiences of the EMU countries.
During the 1990s, the EMU-5 countries experienced a convergence toward
the more developed EU countries
33
. In this convergence the EMU-5 countries
33
The convergence was in particular towards the average of the lowest three nominal interest rates in
the EU member states.
102
experienced continuous declines in the interest and inflation rates, as a sign of
financial integration and strong growth in domestic demand and real GDP. The rapid
credit growth in their domestic markets was fueled by expectation-based economic
growth, and caused an over-heating economy
34
.
A few years after the entrance to the Euro area, the EMU-5 countries
experienced an over-cooling economy with ceasing economic growth and a sluggish
demand side
35
. Even though the macroeconomic environment in the EMU-5
countries caused a reverse movement after entrance to the monetary union, strong
banking structure prevented any sort of systematic banking crises in these countries.
The CEE countries are so far experiencing a similar pattern of domestic
credit growth as did the EMU-5 countries before their entrance to the monetary
union. The financing of this credit demand in the CEE countries is mainly sustained
by banks, since the capital markets in this region are still underdeveloped. Future
EMU membership for the CEE countries and the possibility of an economic reversal
in the following years, may have undesired implications for the CEE banking
structure. First of all, EMU membership will minimize entrance barriers to the CEE
markets and it will further increase the amount of competition. With increased
competition, the excess capacity and profitability in the banking sector will decline
34
During the transition period of the Portuguese economy to the EMU, the growth rate of labor wages
were five times higher than the growth rate of labor productivity (European Commission (2004B)).
35
Even though Ireland had similar macroeconomic indicators in 1990s, during 2000s, the Irish
economy did not go through the over-cooling period as the other four EMU-5 countries. The positive
growth rate of the Irish economy as well as robust domestic demand was sustained also after the
introduction of the Euro (source eurostat).
103
sharply. The decline in bank profitability may force the inefficient domestic banks to
go out of business and may create risks to a vulnerable domestic banking system.
Second, EMU membership implies a loss of monetary policy for the CEE countries.
In case of a loss of competitiveness, the CEE countries may face economic
stagnation as did most of the EMU-5 countries. In an economic downturn the credit
demand declines and banks may face credit risk due to lending to more risky
customers.
The similarities between the current macroeconomic conditions in the CEE
countries and the EMU-5 (before their entrance to the monetary union) and the
possibility of a credit crunch that the CEE countries may face after their entrance to
the monetary union (as did the EMU-5 by 2001) motivate this chapter to analyze the
credit growth pattern in the EMU-5 countries before and after their entrance to the
EMU. In particular, strong banking system indicators in the EMU-5 during their
economic downturn are motivations to study the EMU-5 banking system in order to
provide a road map for the CEE countries of the possible risks that they may face
after their membership to the monetary union.
The analysis in this chapter is based on an unbalanced panel dataset which
covers 18 years (1988-2005) for the 44 largest banks operating in the EMU-5
countries. In this chapter, the paper utilizes a static panel data estimation technique,
very similar to the one described in the third chapter. However, in this section the
104
paper disregards the ownership variable since foreign bank ownership is not a
common type in the EMU-5 countries
36
.
The results obtained in this section show important notifications for the CEE
countries. First of all, the microeconomic analysis shows that the liquidity structure
of the EMU-5 banks changed over time. Customer Deposit had changed from a
positive to a negative influence over the course of time. The increase in customer
deposits causing lower credit growth rates may reflect a change in the consumption
and expenditure behavior. In transitioning to the monetary union, a significant and
positive estimate for the coefficient of customer deposits shows that internal markets
were an important source of funding for the EMU-5 countries. However, the
interpretation of this coefficient changed by the 2000s. After EMU membership,
EMU-5 countries engaged in saving behavior and refrained from the high
expenditures that they had incurred previously.
The macroeconomic analysis shows that high economic growth and financial
deepening, via the decline in interest rates, were very important factors which drove
high credit growth in the EMU-5 countries in the 1990s. These spreads lose
importance after the monetary union because they diminish over time. However, the
business cycle kept its importance and the results show that economic downturns
have a negative influence on the credit growth pattern. Therefore this is an indication
36
The total banking sector assets controlled by foreigners is around 15 percent in the EMU-5
countries, excluding Ireland, for the 1997 to 2005 period (source Bankscope).
105
that the CEE banks need to have prudential banking regulations and supervision to
control very risky bank lending when the business environment is not so welcoming.
This chapter has the following structure. The next section provides the
similarities between the EMU-5 countries before their membership to the EMU and
the CEE countries during their transition. The data is analyzed in section 5.2 Data.
The econometric model is introduced in section 5.3 Econometric Model. The
regression results are provided in section 5.4 Results from the EMU-5 Banking
Structure, and the last section is 5.5 Conclusion.
5.1 Are there any similarities between the CEE and the EMU-5 Countries?
The macroeconomic indicators in the CEE countries during the last five years
of the dataset follows a very similar pattern to that of the five current member states
of the EMU during their transition to the Euro area in the 1990s. Figure 4 provides a
plot of the transition indicators for the EMU-5 and the CEE countries. This figure
provides a selection of macroeconomic and banking indicators during the transition
of these regions to the monetary union.
In Figure 4, the lower x-axis is the timeline for the EMU-5 countries for the
years 1992 to 2000. The earliest year in this axis is the year of acceptance of the
Maastricht Treaty, when these countries agreed on the initials of the economic and
106
monetary union, and the time line ends at 2000, since all the EMU-5 counties, except
Ireland, showed symptoms of an overcooling economy after this period.
In Figure 4, the timeline starts in 2001 for the CEE countries, and it continues
until 2005, the last year of the dataset. As mentioned in the third chapter of this
thesis, the CEE countries show significant differences in their banking structure and
the macroeconomic environment by 2001. The earlier periods were considered to
have differences due to the transition of the CEE countries from a command
economy to a market based system. Since in the previous two chapters, the earlier
years of the sample are shown to have different characteristics for the CEE countries,
the graphical analysis in this section also starts from 2001 for the transition of the
CEE countries to the EMU area.
Figure 4 shows that the CEE countries exhibit a very similar trend in the
2000s as did the EMU-5 countries in the 1990s. The figure shows that the CEE
countries have been experiencing high economic growth, large current deficits
accounts, declining inflation rates, and interest rate spreads as did their preceding
member states in the late 1990s. Back in the 1990s, the high economic growth
prospects and the financial deepening in the EMU markets via the declining interest
rates spreads, and the integration towards the more developed EU level showed an
increasing trend of growth rate of domestic demand. The growth rate of Domestic-
Credit-to-GDP ratio shows an increasing trend in particular by 1996 for the EMU
countries. Since a majority of the domestic demand was financed by bank credit in
the EMU-5 countries, Figure 4 shows the steadily increasing trend in the loan growth
rate of the EMU-5 countries during the 1990s.
Figure 4: Comparison of the transition indicators between the CEE and EMU-5 Countries.
Note: The x-axis on the top is for the CEE Countries, and the one on the bottom is for the EMU countries.
107
108
The same trend in terms of financial deepening and high growth prospects
exists for the CEE countries after the 2000s. Figure 4 shows that the desirable
economic environment in the CEE countries is followed by a rapid domestic demand
growth. Again, similar to the EMU-5 experience, the CEE countries are financing the
domestic demand through the banking system and Figure 4 presents the high loan
growth trend of the banks operating in the CEE countries.
5.2 Data
A panel dataset is used for this analysis covering the years 1988 to 2005 for
the largest 44 banks operating in the EMU-5 countries. The microeconomic and
macroeconomic data for the EMU-5 countries are collected from the same databanks
as mentioned in the third chapter of this thesis. For the microeconomic dataset, the
banking structures of the CEE and EMU-5 countries contain differences with respect
to their size and the scope of their operations. Many banks of the EMU member
states are very large banks and their foreign operations constitute a major percentage
of their balance sheet items. Therefore the use of balance sheet statements is an
important choice for the EMU-5 countries, unlike the CEE banks where there is an
insignificant difference across the use of consolidated and unconsolidated accounts.
In order to focus on only the domestic activities of the EMU-5 banks, unconsolidated
balance sheet statements are used to collect the microeconomic data.
Table 26: Summary Statistics for EMU-5
Obs Mean Std. Dev Obs Mean Std. Dev Obs Mean Std. Dev <0 ≠0 >0
Net Loan Growth 463 16.44 22.86 322 18.11 22.03 141 12.63 24.31 0.01 0.02 0.99
Total Assets over GDP (-1) 463 12.46 11.25 322 10.61 9.10 141 16.69 14.21 1.00 0.00 0.00
Customer Deposits / Total Assets (-1) 463 50.66 21.19 322 51.53 22.35 141 48.67 18.21 0.07 0.15 0.93
Interbank Liabilities / Total Assets (-1) 459 21.63 14.36 320 21.85 14.13 139 21.12 14.92 0.31 0.63 0.69
ROE (-1) 445 13.45 13.11 310 13.06 13.45 135 14.35 12.31 0.84 0.32 0.16
ROA (-1) 446 0.75 0.91 311 0.75 1.05 135 0.74 0.49 0.47 0.93 0.53
Cost to Income Ratio (-1) 427 64.93 16.00 301 65.73 14.54 126 63.02 18.96 0.08 0.15 0.92
Net Interest Margin (-1) 429 3.04 1.36 303 3.30 1.44 126 2.42 0.90 0.00 0.00 1.00
Loan Loss Provision/ Total Assets (-1) 426 0.44 0.49 301 0.48 0.56 125 0.34 0.24 0.00 0.00 1.00
Total Capital Ratio (-1) 158 12.49 6.63 105 12.72 7.44 53 12.03 4.64 0.24 0.47 0.76
Real GDP Growth Rate 463 3.08 2.48 322 3.24 2.66 141 2.72 1.95 0.01 0.02 0.99
Domestic Credit / GDP (-1) 463 106.28 19.05 322 102.46 19.32 141 115.01 15.20 1.00 0.00 0.00
Real Interest Rate 463 2.96 3.06 322 4.32 2.65 141 -0.16 0.82 0.00 0.00 1.00
Spread 1 (Lending and Borrowing) 463 4.63 2.07 322 5.14 2.18 141 3.48 1.15 0.00 0.00 1.00
Spread 2 (Domestic and Euro) 456 2.28 3.52 315 3.30 3.81 141 -0.01 0.21 0.00 0.00 1.00
Euro Area M3 Growth 467 6.23 2.39
Euro Area S-T Interest Rate (3 Month) 460 5.31 2.76
≠0: P-value of the two-sided t-test on the equality of means, where the data are not assumed to have equal variances.
<0: P-value of the one-sided t-test on the equality of means, where the data are not assumed to have equal variances.
>0: P-value of the one-sided t-test on the equality of means, where the data are not assumed to have equal variances.
Whole Sample S
1
: 1988-2000 S
2
: 2001-2005 Ha: S
2
-S
1
The summary statistics for the balance sheet data and the macroeconomic
variables of the EMU-5 countries are provided in Table 26
37
. In this table, Net Loan
Growth outside the [-50, 150] range are marked as outliers and they are excluded
from the regression analysis. Similar to the CEE countries, the variables of interest
for the EMU-5 countries are considered for two samples: 1988 to 2000, and 2001 to
2005. After performing a t-test, Table 26 shows that the equality of means for a
majority of the variables of interest is rejected. Table 26 shows that the mean and the
variance of profitability, liquidity, and efficiency measures for the EMU-5 countries
37
For a detailed description of the micro and macroeconomic variables, refer to the Data section
provided in chapter 3.
109
are comparably smaller than the values of the CEE countries presented in the data
analysis section of chapter 3.
38
Figure 5: Mean and Standard Deviation of Banking Variables across Years in EMU Countries
0 20 40 60 80
Credit Growth
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
0 5 10 15 20
Bank Size
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
0 .5 1 1.5 2 2.5
ROA
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
0 1 2 3 4
Net Interest Margin
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
0 .5 1 1.5
Loan Loss Provisions
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
0 20 40 60 80
Cost to Income Ratio
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Mean Standard Deviation
38
For the explanation and relevance of the balance sheet items to this analysis, please refer to chapter
3, data analysis section.
110
111
The next destination for the CEE countries is the monetary union. Since the
EMU-5 states had experienced an overcooling after the introduction of the euro,
looking at the banking sector of these countries, a road map can be obtained for the
CEE countries for their way to the EMU.
Figure 5 plots the mean and standard deviation of various banking variables
for the EMU-5 countries over the course of the sample years.
39
This figure presents
that credit grows in the EMU-5 countries during the pre-EMU period i.e.; from the
Maastricht Treaty until entering the monetary union. Even though credit growth
peeks during the first two years of the monetary union, 1999 and 2000, EMU
countries face a steady decline in credit growth until 2004.
As analyzed in Figure 4, the CEE countries have been experiencing a steady
rise in credit growth, as of the pre-EMU pattern of these five countries. For the EMU
countries many of the banking sector variables show improvement before and during
the first two years of the monetary union but then show stagnation by 2002. Bank
size, which is measured as Total Assets of a bank divided by the GDP of the country
it is operating in, grows until 2000 and then stays around the 2000 level for the
following four years. Figure 5 also shows that profitability had an increasing trend
for these five EMU countries before they joined the monetary union. It peeked for
the first year of the union, and then fell back to the mid-1990s level. However, the
volatility of this variable was much higher than mid-1990s volatility (refer to Figure
39
There are only 3 banking year observations in year 2005 for the EMU countries. Therefore, this
year is excluded from this analysis.
112
5). Net Interest Margin had declined until the year 2000, and then it stayed at that
level for the rest of the sample years (refer to Figure 5). Similar to the previous
profitability variance, i.e.; ROA and Net Interest Margin also have a high variance.
Since the margins are also falling for the CEE countries, a high variance in
profitability may indicate more risk for the banks operating at smaller profitability
percentiles in the CEE countries. Even though efficiency ratios move less on
average, the variance moves in line with the profitability ratios. Looking at Figure 5
one can see that the standard deviation of the Cost to Income Ratio peeks at the same
years as ROA and Net Interest Margin: 1994, 1999, and 2002. A similar observation
is also valid for the variance of soundness measures in the EMU-5 countries; that
they peek in the same years as profitability and efficiency ratios. Looking at Figure
5, one can see that the three peeks of Loan Loss Provisions over Total Assets are
1992, 1999, and 2002. Even though the mean values of banking performance
indicators do not change much after the second year of EMU participation, there is a
high variance in these indicators. Considering that the CEE countries are following a
similar pattern, high variation in banking profitability and soundness measures may
indicate more risk for banks operating in the CEE countries. EMU countries
experienced a sharp decline in credit growth after the Euro participation. A similar
cooling down effect for the CEE countries may restrain more risk for the domestic
financial sector.
5.3 Econometric Model
In this section, the paper utilizes the same methodology as described in the
third chapter of the thesis. However, here the paper disregards the ownership
variable since foreign bank ownership is not a common type in the EMU-5
countries
40
. Therefore the panel data model is given as follows:
ijt jt ijt ijt
u Z X C + ′ + ′ + =
−
φ β µ
1
(5.1)
where
ijt i ijt
v u ε + =
The dependent variable is bank credit growth, measured by the growth rate in
the net loans of bank i in country j at time t. On the left hand side, µ is the common
constant, and X
ij,t-1
represents the bank-specific variables for bank i in country j at
time (t-1). In order to diminish any simultaneity problems between the balance sheet
items and the dependent variable, all balance sheet variables are lagged one period.
Last, Z
jt
is the matrix for country specific macroeconomic variables for country j at
time t - Domestic Credit over GDP is lagged one period because of simultaneity
issues.
40
Foreign bank ownership does not have a dominant role in the EMU-5 countries. Although Ireland
may seem an exception as foreign branches and subsidiaries hold more than 50 percent of its banking
assets, most foreign-owned banks in Ireland are engaged in off-shore operations and have a very
limited impact on the Irish economy. The EMU-5 countries show that private domestically-owned
banks are the major players in this region. Even though in the earliest years of the dataset, state-owned
banks constituted a major part of the total number of large banks in EMU, by the early 1990s, there is
a sharp rise in the number of private domestically-owned banks.
113
A static panel data estimation method is preferred over a dynamic model.
Table 27 provides the correlation coefficient between the current and the past value
of each EMU-5 bank’s credit growth rate. As provided in this table, only 2 out of 44
banks have significant correlation coefficients at the 5 percent level.
Table 27: Pairwise Correlation Coefficient between Credit Growth
and its Lagged Value for Each EMU-5 Bank.
1-8 9-16 17-24 25-32 33-40 41-44
Correlation Coefficient 0.87 0.13 0.19 -0.22 0.32 .
Significance Level 0.00 0.67 0.55 0.48 0.36 1.00
Number of Observations 14 14 12 13 10 0
Correlation Coefficient 0.52 0.27 1.00 0.20 -0.08 -1.00
Significance Level 0.06 0.82 1.00 0.55 0.90 1.00
Number of Observations 14 3 2 11 5 2
Correlation Coefficient 0.48 0.15 -0.54 0.42 -0.54 -0.23
Significance Level 0.33 0.69 0.11 0.26 0.64 0.65
Number of Observations 6 10 10936
Correlation Coefficient -0.04 0.01 -0.02 -0.05 1.00 0.06
Significance Level 0.98 0.96 0.96 0.92 1.00 0.87
Number of Observations 3 14 10 6 2 10
Correlation Coefficient 0.15 -0.19 -0.09 -0.37 0.33
Significance Level 0.62 0.57 0.81 0.19 0.26
Number of Observations 14 11 10 14 14
Correlation Coefficient -0.45 0.33 -0.43 0.89 0.03
Significance Level 0.11 0.24 0.47 0.30 0.96
Number of Observations 14 14535
Correlation Coefficient 0.23 0.35 -0.30 -1.00 .
Significance Level 0.42 0.44 0.62 0.02 1.00
Number of Observations 14 7530
Correlation Coefficient 0.12 0.05 -1.00 -0.03 .
Significance Level 0.68 0.86 1.00 0.91 1.00
Number of Observations 14 14 2 14 0
Note: Correlation coefficients with p-values smaller than 0.05 are marked
with bold letters.
Bank Number:
The choice of a static panel data model is considered across the fixed effects
and random effects model. The Hausman specification test indicates that the random
114
115
effects model is biased and indicating a correlation between the bank dummies and
the error term. Therefore the analysis utilizes the fixed effects within estimator as the
final model.
In order to control any unspecified macro effect, time-specific-country
dummies are introduced to Equation (5.1). The introduction of these dummy
variables controls the impact of a year-specific macroeconomic shock which is
peculiar to each EMU-5 country. However, the macroeconomic variables are highly
collinear with the time-specific country dummies. In order to observe the coefficient
of the macroeconomic variables, each regression is run once for the macroeconomic
variables and once with the time-specific-country dummies.
5.3.1 Concerns about Multicollinearity
A relevant issue in a multivariate regression analysis is the collinearity of the
regressors with each other. In order to examine this problem, a correlation Matrix is
provided in Table 28. The first row of this table shows the pair-wise correlation
coefficients of the micro and macro variables in the EMU-5 countries, and the
second row shows the significance level (p-value) of that correlation coefficient.
Table 28: Correlation Coefficient across Regressors for the EMU-5 Countries
Spread
State Dummy -0.18
0.00
Total Assets over GDP -0.07 0.12
(TA/GDP) 0.13 0.01
Customer Dep. / Total Assets 0.04 0.22 0.13 1
(Cust. Deps.) 0.41 0 0.01
Interbank Liab. / Total Assets 0.15 -0.10 0.06 -0.38 1
(Inter. Liab.) 0.00 0.03 0.23 0
ROE 0.02 -0.19 0.05 0.08 -0.09 1
0.70 0.00 0.31 0.09 0.05
ROA -0.17 -0.15 0.02 0.12 -0.21 0.58 1
0.00 0.00 0.75 0.01 0 0
Cost to Income Ratio -0.05 0.23 0.09 -0.09 0.06 -0.38 -0.45 1
(C-to-I.) 0.28 0 0.06 0.07 0.23 0 0
Net Interest Margin 0.01 -0.10 -0.32 0.15 -0.40 0.12 0.29 -0.30 1
(NIM) 0.92 0.04 0 0.00 0 0.01 0 0
Loan Loss Prov/ Total Assets 0.21 0.00 -0.11 0.00 0.00 -0.17 -0.48 -0.08 0.32 1
(LLP/TA) 0 0.95 0.02 0.98 0.97 0.00 0 0.09 0
Total Capital Ratio (-1) -0.17 -0.17 -0.15 -0.08 -0.17 0.02 0.25 -0.20 0.29 0.12 1
(TCR) 0.03 0.03 0.05 0.33 0.03 0.82 0.00 0.01 0.00 0.13
Real GDP Growth Rate -0.09 -0.12 0.26 0.10 0.04 0.11 0.09 -0.09 -0.14 -0.13 -0.05 1
0.05 0.01 0.00 0.04 0.45 0.02 0.06 0.05 0.00 0.01 0.55
Real Interest Rate -0.04 0.16 -0.28 0.12 -0.12 0.01 0.08 0.06 0.38 0.10 0.04 -0.28 1
(r) 0.35 0.00 0.00 0.01 0.01 0.83 0.07 0.19 0.00 0.03 0.65 0.00
Domestic Credit / GDP -0.04 -0.03 0.04 0.07 -0.12 0.08 0.05 0.06 -0.03 -0.17 0.00 -0.29 -0.06 1
(DC/GDP) 0.42 0.49 0.33 0.13 0.01 0.11 0.31 0.21 0.59 0.00 0.99 0.00 0.21
Spread 1 (Lend - Borrow) -0.05 0.41 0.01 0.40 -0.12 -0.04 0.08 0.10 0.16 0.06 0.11 -0.18 0.50 -0.04 1
(Spread
1
)
0.30 0.00 0.78 0.00 0.02 0.38 0.10 0.06 0.00 0.22 0.18 0.00 0 0.38
Spread 2 (Domestic - Euro) -0.08 0.21 -0.11 0.36 -0.33 0.08 0.14 0.06 0.31 0.11 0.19 -0.28 0.73 0.15 0.69
0.10 0.00 0.02 0.00 0.00 0.08 0.00 0.24 0.00 0.03 0.02 0 0 0.00 0
* Second row shows the p-values
Note: The correlation coefficiente with p-vaues smaller than 0.05 are marked with bold letters.
NIM
LLP/
T.A. TCR GDP r
DC/G
DP
Inter.
Liab. ROE ROA C-to-I Foreign State
TA/GD
P
Cust
Deps
Looking at the correlation matrix given in Table 28, one can see that most of
the regressors are significantly correlated with each other, even though the
correlation coefficients in most cases are small
41
. However, the coefficients across
41
The significant correlation coefficients, with an absolute value of an estimate smaller than 0.3, are
considered as small.
116
117
similar measurement variables are large in magnitude, such as the variables for bank
profitability and efficiency
42
or macroeconomic variables
43
. In order to eliminate any
multicollinearity problems, the paper introduces these highly and significantly
correlated variables into separate regression equations, rather than introducing all the
variables at once.
5.4 Results from the EMU-5 Banking Structure
In this study, the regression analysis is performed for two separate samples.
The first sample is from 1988 to 2000 and the second sample is from 2001 to 2005.
As mentioned in section 5.2 Data, the two samples exhibit significant differences
among each other in terms of the mean values of variables of interest. Therefore this
chapter studies these two samples separately. Another reason for such a distinction is
to analyze the impact of the monetary union on the credit growth of the EMU-5
banks.
The panel data model presented in Equation (5.1) is used for the EMU-5
regression analysis. As mentioned in the section 3.3.1 Concerns about
Multicollinearity, most of the balance sheet items are significantly correlated with
42
ROA, ROE and Cost-to-Income Ratio
43
Spread between Lending and Borrowing Rate, and Spread between Domestic and Euro Area
Interest Rate.
118
each other. Large and significant correlation coefficients are symptoms of a
multicollinearity problem. Therefore variables which may cause multicollinearity are
introduced in separate regressions equations.
The results are based on the regression estimates reported in Table 29 and
Table 30. In this section, results based on the impact of the balance sheet items on
the EMU-5 credit growth are reported under the section 3.4.2 Impact of
Macroeconomic Shocks, and results based on the impact of the macroeconomic
variables are provided under the section 5.4.2 Impact of Macroeconomic Shocks.
Table 29: Fixed Effect Estimator Results for EMU-5
Total Assets / GDP (-1) 0.08 0.18 -2 -2.53 0.04 0.07 -2.03 -2.49 0.17 0.14 -2.09 -2.63
[0.90] [0.80] [0.02]** [0.01]** [0.94] [0.91] [0.02]** [0.01]** [0.79] [0.85] [0.03]** [0.01]**
Customer Deposits/TA (-1) 0.13 0.14 -0.44 -1.14 0.09 0.13 -0.48 -1.09 0.13 0.15 -0.8 -1.11
[0.03]** [0.02]** [0.04]** [0.00]*** [0.14] [0.03]** [0.03]** [0.00]*** [0.03]** [0.01]** [0.00]*** [0.00]***
Interbank Liabilities/TA (-1) -0.16 -0.31 -0.63 -1 -0.13 -0.36 -0.76 -0.98 -0.11 -0.3 -0.34 -0.69
[0.41] [0.14] [0.07]* [0.01]*** [0.45] [0.07]* [0.02]** [0.01]*** [0.53] [0.13] [0.30] [0.08]*
ROA (-1) 1.39 1.28 5.7 -2.84
[0.22] [0.27] [0.19] [0.56]
ROE (-1) -0.04 -0.07 0.17 -0.14
[0.67] [0.42] [0.50] [0.57]
Cost to Income Ratio (-1) -0.08 -0.01 -0.43 -0.19
[0.45] [0.94] [0.00]*** [0.16]
Spread 1 (Lend - Borrow) -1.31 -2.84 3.39 1.72
[0.11] [0.09]* [0.24] [0.61]
Real GDP Growth Rate 2.2 2.3 3.1 10.94
[0.00]*** [0.00]*** [0.18] [0.38]
Euro Area M3 Growth 0.15 -0.8 0.8 -1.21
[0.74] [0.26] [0.39] [0.55]
Euro Area S-T Int. (3-Month) -1.16 -1.79 1.11 -0.4
[0.01]*** [0.00]*** [0.52] [0.87]
Constant 20.22 30.62 72.29 131.64 8.7 16.21 68.06 102.58 23.35 27.98 104.56 139.34
[0.01]** [0.00]*** [0.00]*** [0.00]*** [0.22] [0.08]* [0.00]*** [0.01]*** [0.04]** [0.05]* [0.00]*** [0.00]***
Country-Time Dummies No Yes No Yes No Yes No Yes No Yes No Yes
Observations 300 300 129 129 306 306 129 129 297 297 120 120
Number of bankid 37373434 37373434 3535 3232
R-squared 0.05 0.26 0.11 0.43 0.08 0.27 0.15 0.43 0.03 0.23 0.23 0.45
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
Model 1 Model 2 Model 3
1990s 2000s 1990s 2000s 1990s 2000s
119
Table 30: Fixed Effect Estimator Results for EMU-5
Total Assets / GDP (-1) 0.05 0.1 -2.67 -2.79 0.21 -0.08 -2.88 -2.86
[0.93] [0.89] [0.01]*** [0.01]*** [0.73] [0.91] [0.00]*** [0.01]***
Customer Deposits/TA (-1) 0.11 0.1 -0.52 -1.03 0.1 0.12 -0.5 -1.05
[0.06]* [0.10]* [0.02]** [0.00]*** [0.10] [0.03]** [0.02]** [0.00]***
Interbank Liabilities/TA (-1) -0.2 -0.35 -0.38 -0.71 -0.1 -0.33 -0.5 -0.77
[0.25] [0.08]* [0.26] [0.09]* [0.56] [0.10] [0.15] [0.06]*
Loan Loss Provision/TA (-1) -2.86 -1.93 1.45 0.18
[0.12] [0.31] [0.88] [0.99]
Real Interest Rate -0.9 -2.22 4.01 -3.03
[0.01]** [0.00]*** [0.02]** [0.62]
Real GDP Growth Rate 1.9 2.14 6.57 6.61
[0.00]*** [0.01]*** [0.00]*** [0.37]
Net Interest Margin (-1) -0.79 -1.07 5.31 2.54
[0.57] [0.55] [0.09]* [0.42]
Domestic Credit / GDP (-1) -0.09 -0.01 0.51 0.34
[0.18] [0.94] [0.11] [0.36]
Constant 22.36 20.61 -3.92 70.67 18.12 29.11 90.44 137.24
[0.05]* [0.25] [0.94] [0.26] [0.01]** [0.00]*** [0.00]*** [0.00]***
Country-Time Dummies No Yes No Yes No Yes No Yes
Observations 299 299 120 120 299 299 120 120
Number of bankid 35 35 32 32 35 35 32 32
R-squared 0.09 0.26 0.22 0.44 0.05 0.26 0.15 0.43
p values in brackets
* significant at 10%; ** significant at 5%; *** significant at 1%
Model 4 Model 5
1990s 2000s 1990s 2000s
120
121
5.4.1 Impact of Microeconomic Shocks
The monetary union has a significant impact on the macro and micro
structure of credit growth for the EMU-5 countries. Looking at Table 29, one can
see that the pre-EMU period and post-EMU period has a structural break type effect
on the EMU-5 countries. Further, the t-test performed on the equality of means for
the variables of interest on Table 26 indicates that these two samples have unequal
means. The two periods show differences across, in particular, the size and the
liquidity of the EMU-5 banks. The regression results on Table 29 show that size, as
measured by Total Assets over GDP, has a significant and negative coefficient only
for the last five years of the dataset. During the 1990s, size was an insignificant
factor for credit growth.
The liquidity structure in the EMU-5 Banks changed over time. Table 29
and Table 30 report the coefficient estimates for Customer Deposits over Total
Assets and Interbank Liabilities over Total Assets. Regarding the liquidity structure
of the EMU-5 countries; Customer Deposits over Total Assets change from a
significant and positive coefficient to a significant and negative one, and Interbank
Liabilities over Total Assets change from an insignificant coefficient to a significant
and positive one over the two periods. Higher Customer Deposits may be indicating
an increase in savings behavior for the EMU households and therefore reduce their
demand for bank loans.
122
The EMU-5 countries did not experience any changes over their
profitability, efficiency, and soundness measures after the introduction of the
Euro. Table 29 and Table 30 report the coefficient estimates of ROE, ROA, Loan
Loss Provision over Total Assets, and Cost to Income Ratio
44
. Except for a
statistically significant and negative value for the Cost to Income Ratio, the
profitability and soundness indicators do not yield any significant results for either
half of the dataset. Even though Net Interest Margin has a significant and positive
coefficient during the second half of the dataset (refer to Table 30), this ratio
becomes insignificant after the introduction of country-time dummies.
5.4.2 Impact of Macroeconomic Shocks
Economic growth is the driving factor for credit growth in the EMU-5
countries. The regression results reported in Table 29 and Table 30 show a positive
and significant coefficient for Real GDP Growth in the EMU-5 countries, in
particular, for the first half of the dataset. This finding supports the claim that strong
economic growth increases the expectation based growth and stimulates credit.
44
Total Capital Ratio is disregarded in the regression analysis for the EMU-5 countries due to the
large amount of missing variables.
123
The spread between lending and borrowing rate is significant and
negative in CEE countries during the last five years. Looking at Table 29, one can
see that the coefficient estimate for Spread between Lending and Borrowing Rate is
significant and negative only during the first period. This does not have any
influence over the EMU-5 credit growth after the monetary union because these
spreads diminished over time. However, financial deepening, expressed in the
negative coefficient estimate of this spread, had a very important influence on the
EMU-5 credit growth.
Higher real interest rates reduce credit during the transition years and
increase afterwards. Looking at Table 30, one can see that the Real Interest Rate
has a significant and negative coefficient during the transition years. This negative
coefficient is a reflection of financial deepening. As the real interest rates in the
EMU-5 countries are falling, the credit grows in these markets. On the other hand, in
the second half of the dataset, the coefficient estimate of this variable indicates a
supply side effect: when the price is higher, banks are more willing to increase their
credit line.
5.5 Conclusion
This chapter draws projections for the CEE countries for their future EMU
membership and from the experiences of the five current EMU states which had
similar macroeconomic indicators as that of the CEE countries. A possibility of
124
economic slowdown and credit crunch that the CEE countries may experience in the
future and the impact of this reversal on the CEE banking system motivate this
chapter to compare the credit growth pattern in the EMU-5 countries to that of the
CEE countries.
The analysis in this chapter uses a panel data analysis for the years 1988 to
2005. The data analysis indicates a structural break in the 2000s for the EMU-5
countries. Therefore the econometric analysis is performed separately for the 1990s
and the 2000s. The results show that the monetary union had important effects on the
liquidity structure of the EMU-5 countries. The impact of customer deposits changed
from a positive to a negative relation on the EMU-5 credit growth pattern indicating
a change in domestic expenditure and saving behavior: consumption and expenditure
declining, while savings increased. Further results show that economic integration
and growth were the driving factors for the EMU-5 credit growth in the 1990s.
125
Bibliography
Altunbas, Y., L. Evans and P. Molyneux, 2001, “Bank Ownership and
Efficiency,” Journal of Money, Credit and Banking, Vol. 33, No. 4, pp. 926-54.
Barajas, A., R. Steiner and N. Salazar, 2000, “The Impact of Liberalization
and Foreign Investment in Colombia’s Financial Sector,” Journal of Development
Economics, Vol. 63, pp. 157-196.
Baltagi, B. and H., 2003, “Econometric Analysis of Panel Data,” Wiley, Third
Edition.
Berger, A., R. DeYoung, H. Genay and F. Udell, 2000, “Globalization of
Financial Institutions: Evidence from Cross-Border Banking Performance,”
Brookings Papers on Economic Activity, Vol. 2, pp. 23-158.
Berger, A., S. Bonime, L. Goldberg and L. White, 2004, “The Dynamics of
Market Entry: The Effects of Mergers and Acquisitions on Entry in the Banking
Industry,” Journal of Business, Vol. 77, pp. 797-834.
Bonin, J.; P. Wachtel, 2003, “Financial Sector Development in Transition
Economies: Lessons from the First Decade,” Financial Markets, Institutions &
Instruments, Volume 12, Number 1, pp. 1-66(66).
Breyer, P., 2004, “Central and Eastern Europe — The Growth Market for
Austrian Banks,” Oesterreichische Nationalbank, Journal Monetary Policy & the
Economy, Issue 3, pp. 63-88.
Brzoza-Brzezina, M., 2005, "Lending Booms in the New EU Member States:
Will Euro Adoption Matter?" ECB Working Paper, No. 543.
Caprio, G. and D. Klingebiel, 2002, “Episodes of Systematic and Borderline
Banking Crises,” World Bank, Washington, DC, mimeo.
Cottarelli, C., G. Dell’Ariccia and I. V. Hollar, 2003, “Early Birds, Late
Risers, and Sleeping Beauties: Bank Credit Growth to the Private Sector in Central
and Eastern Europe and the Balkans,” IMF Working Paper, No. 03/213.
Claessens, S., A. Demirguç-Kunt and H. Huizinga, 2001, “How Does Foreign
Entry Affect Domestic Banking Markets?,” Journal of Banking and Finance, Vol. 25
pp. 891-911.
126
Clark, G., R. Cull, L. D’Amato and A. Molinari, 1999, “The Effect of
Foreign Entry on Argentina’s Domestic Banking Sector,” Unpublished paper.
Clark, G., R. Cull, M. Soledad Martinez Peria and S. M. Sanchez, 2003,
“Foreign Bank Entry: Experience, Implications for Developing Economies, and
Agenda for Further Research,” World Bank Research Observer, Vol. 18, No. 1, pp.
25-59.
De Gianni, N. and E. Loukoianova, 2006, “Bank Ownership, Market
Structure and Risk,” International Monetary Fund, First Preliminary Draft.
De Haas, R.T.A. and I. Van Lelyveld, 2006, “Foreign Banks and Credit
Stability in Central and Eastern Europe. A Panel Data Analysis”, Journal of Banking
and Finance, Elsevier, vol. 30(7), pp. 1927-1952.
De Haas, R.T.A. and I. Van Lelyveld, 2003, "Foreign Bank Penetration and
Private Sector Credit in Central and Eastern Europe," Netherlands Central Bank,
DNB Staff Reports (discontinued), Vol. 91.
Demirgüç-Kunt, A. and H. Huizinga, 1998, “Determinants of Commercial
Bank Interest Margins and Profitability: Some International Evidence,” Unpublished
paper.
Demirgüç-Kunt, A., L. Laeven, and R. Levine, 2003, “Regulations, Market
Structure, Institutions, and the Cost of Financial Intermediation.” National Bureau of
Economic Research, Inc, No. 9890.
Denizer, C., 1999, “Foreign Entry in Turkey’s Banking Sector, 1980-97,”
World Bank Policy Research Working Paper, Washington D.C., WPS 2462.
European Central Bank, 1999, “Possible Effects of EMU on the EU Banking
Systems in the Medium to Long Term,” European Central Bank, Frankfurt am Main.
European Central Bank, 2006, “Macroeconomic and financial stability
challenges for acceding and candidate countries,” European Central Bank,
Occasional Paper No. 48.
European Central Bank, 2005A, “Banking Structures in the New EU Member
States,” European Central Bank, Frankfurt am Main.
European Central Bank, 2005B, “EU Banking Structures,” European Central
Bank, Frankfurt am Main.
127
European Commission, 2004A, “Reviewing adjustment dynamics in EMU:
from overheating to overcooling,” European Commission Economic Papers,
European Economy, No. 198, Brussels.
European Commission, 2004B, “The Portuguese economy after the boom,”
European Commission Occasional Papers, European Economy, No. 8, Brussels.
Gormley, T. A., 2007, "Banking Competition in Developing Countries: Does
Foreign Bank Entry Improve Credit Access?". Available at Social Science Research
Network: http://ssrn.com/abstract=879244.
Gourinchas, P.O., R. Valdes and O. Landerretche, 2001, “Lending Booms:
Latin America and the World,” NBER Working Paper, No. 8249.
Gruben, W. C. and R. McComb, 1997, “Liberalization, privatization, and
crash: Mexico’s banking system in the 1990s,” Federal Reserve Bank of Dallas
Economic and Financial Policy Review, No. 9701, 21-30.
Gruben, W. C., and R. McComb, 2003, “Privatization, Competition, and
Supercompetition in the Mexican Commercial Banking System,” Journal of Banking
and Finance, Vol. 27, pp. 229-49.
Haber, S., 2005, “Banking with and without Deposit Insurance: Mexico’s
Banking Experiments, 1884-2004,” Mimeo. Stanford University.
Hanousek, J., E. Kocenda and J. Svejnar, 2004, “Ownership, Control and
Corporate Performance after Large-Scale Privatization,” William Davidson Institute
Working Paper, No. 652.
Hsiao, C., 2003, “Analysis of Panel Data” Cambridge University Press.
International Monetary Fund, 2006, “Financial Soundness Indicators:
Compilation Guide,” International Monetary Fund, ISBN/ISSN: 978-1-58906-385-
3.
Kraft, E. and L. Jankov, 2005 "Does Speed Kill? Lending Booms and their
Consequences in Croatia," Journal of Banking and Finance Vol. 29, No. 1, pp. 105-
121.
Kraft, E., 2006, "How Competitive is Croatia's Banking System?," Croatian
National Bank Working Papers, No. W-14, March.
Lambregts, E. and D. Ottens, 2006, “The Roots of Banking Crises in
Emerging Market Economics: A Panel Data Approach,” De Nederderlandsche Bank,
Monetary and Economic Policy Department, Working Paper, No. 084.
128
Lensink, R. and N. Hermes, 2004, “The Short-Term Effects of Foreign Bank
Entry on Domestic Bank Behaviour: Does Economic Development Matter?,”
Journal of Banking and Finance, No. 28, pp. 553-568.
Levine, R., 2002, “Denying Foreign Bank Entry: Implications for Bank
Interest Margins,” Banco Central de Chile, L. Ahumada and R. Fuentes eds., Bank
Competition.
Maechler, A., S. Mitra and W. DeLisle, 2006, “Exploring Financial Risks and
Vulnerabilities in New and Potential EU Member States,” forthcoming, IMF
Working Paper.
Martinez P., M. Soledad and A. Mody, 2004, “How Foreign Participation and
Market Concentration Impact Bank Spreads: Evidence from Latin America,” Journal
of Money, Credit and Banking, Vol. 36, No. 3 (June), pp. 511-537.
Mayer, T., 2006, “Beware of the EMU trap,” Focus Europe, Deutsche Bank
AG Global Markets Research , pp.8-15.
Mian, A., 2003, “Foreign, Private Domestic, and Government Banks: New
Evidence from Emerging Markets,” Mimeo, Graduate School of Business, University
of Chicago.
Micco, A. and U. Panizza, 2006, “Bank Ownership and Lending Behavior,”
Central Bank of Chile Working Papers, No. 369.
Micco, A., U. Panizza and M. Yañez, 2004, "Bank Ownership and
Performance," Inter-American Development Bank, Research Department Working
Papers, No. 1016.
Ottens D., E. Lambregts and S. Poelhekke, 2005, "Credit Booms in Emerging
Market Economies: A Recipe for Banking Crises?," DNB Working Papers 046,
Netherlands Central Bank, Research Department.
Sturm, J.-E. and B. Williams, 2004, “Foreign Bank Entry, Deregulation and
Bank Efficiency: Lessons from the Australian Experience,” Journal of Banking and
Finance, No. 28, pp. 1775-1799.
Tamirisa N. and D. Igan, 2006, “Credit Growth and Bank Soundness in New
Member States,” IMF Working Paper, Washington D.C.
Terrones, M. and Mendoza, E., 2004, “Are Credit Booms in Emerging
Markets a Concern?,” World Economic Outlook, IMF, Washington.
129
Thornton L., 2003, “Comparison of Key Standardards (IFRS, US GAAP and
Eurosystem) with Reference to Central Banks”, Central Banking Publications,
Extract from Accounting Standards for Central Banks.
Abstract (if available)
Abstract
This thesis is comprised of three essays on the credit growth and banking structure of the Central and Eastern European Countries (CEE). The first essay studies the importance of bank ownership on this growth, the second explores the influence of the managerial impact on credit growth and the third one draws projections for the future EMU membership of the CEE countries from the past experiences of the five current EMU states.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Three essays on distance: examing the role of institutional distance on foreign firm entry, local isomorphism strategy and subsidiary performance
PDF
Two essays on financial crisis and banking
PDF
Essays on inflation, stock market and borrowing constraints
PDF
Two essays on major macroeconomic shocks in the Japanese economy: demographic shocks and financial shocks
PDF
Essays on the economics of subjective well-being in transition countries
PDF
An empirical analysis of the quality of primary education across countries and over time
PDF
Three essays on linear and non-linear econometric dependencies
PDF
Essays on Japanese macroeconomy
PDF
Three essays on economics of early life health in developing countries
PDF
Essays on business cycle volatility and global trade
PDF
Three essays on the microeconometric analysis of the labor market
Asset Metadata
Creator
Aydin, Burcu
(author)
Core Title
Three essays on the credit growth and banking structure of central and eastern European countries
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Economics
Publication Date
07/28/2007
Defense Date
06/13/2007
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
bank ownership,CEE,credit growth,fixed effect,foreign bank,OAI-PMH Harvest,panel data
Place Name
Europe
(continents)
Language
English
Advisor
Dekle, Robert (
committee chair
), Ham, John C. (
committee member
), Hsiao, Cheng (
committee member
), Imrohoroglu, Selahattin (
committee member
)
Creator Email
aydin@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m712
Unique identifier
UC1445481
Identifier
etd-Aydin-20070728 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-536387 (legacy record id),usctheses-m712 (legacy record id)
Legacy Identifier
etd-Aydin-20070728.pdf
Dmrecord
536387
Document Type
Dissertation
Rights
Aydin, Burcu
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
bank ownership
CEE
credit growth
fixed effect
foreign bank
panel data