The Internationalization
of Banking:
A Micro-Macro
Approach with Focus on Romania
Daniel Ibrahim
and Peter Haiss0
EuropaInstitut
Vienna University
of Economics and Business
Paper submitted
for presentation at the EEFS Conference;
June 3 - 7, 2009; Udine, Warsaw
Abstract
The purpose
of this paper is twofold. On the macro side, we discuss the impact of
foreign bank penetration on the financial system by estimating a cross-country
panel data for 10 of Central and Eastern European countries (CEECs)
between 1997 and 2006. We hypothesize that increased foreign bank concentration
could lead to an increased supply of capital for foreign direct investment
and a higher credit supply for the domestic economy.
On the micro
side, we highlight the relationship between foreign bank penetration,
through influencing the local banking market structure, and the performance
of the banks operating in this market by measuring a cross-bank panel
data for 17 Romanian banks between 1999 and 2006. This estimation intends
to give evidence on the structure-conduct-performance paradigm. The
value of the paper lies in (1) providing banking sector data for CEECs
and specially Romania, (2) estimating the effect of foreign bank penetration
on a country and a sector level and (3) providing evidence on the SCP
paradigm.
Keywords:
Foreign bank penetration, financial sector foreign direct investment,
structure-conduct-performance, transitional economies, panel data analysis,
Romania
Daniel
IBRAHIM
Peter HAISS
Graduate
student, EuropaInstitut, Vienna University of Economics and Business
(WU - Wirtschaftsuniversitaet Wien), Vienna, Austria
Lecturer, EuropaInstitut,
Vienna University of Economics and Business, (WU - Wirtschaftsuniversitaet
Wien), Vienna, Austria
Althanstrasse
39-45/2/3
Althanstrasse 39-45/2/3
A-1090
Wien, Austria
A-1090 Wien, Austria
phone
++ 43/650 635 22 88
phone: ++43 (0)664 812 29
90
fax ++43(0)1
313 36- 758
fax ++43(0)1 313 36- 758
daniel.ibrahim@wu-wien.ac.at
peter.haiss@wu-wien.ac.at
Introduction
In the
last years, authorities are increasingly observing the development of
the financial markets and are implementing economic and legal measures
in order to accelerate financial integration of the European financial
markets. According to the vigorous effect of the banking sector on economic
growth, the European Union set the integration of this sector as one
of the main targets for the next years. A very important factor with
regard to integration and the development of the banking systems of
the Central and Eastern European countries was obviously the entry of
foreign banks via FDI. These banks were not only able to support the
transformation of the CEE economies from state-led to market-oriented
by introducing more sophisticated financial products and management
techniques, but also increased the access to capital for the domestic
economy.
Foreign financial
intermediaries also showed the capability of increasing the quality
and availability of financial services. Other positive effects have
been improved domestic financial policy and financial structure (see,
among others, Levine (1996), Gray and He (2001), Clarke et al. (2006),
Fink et al. (2007), Claessens et al. (1998, 2001)). The change of the
banking market structure, mainly caused by foreign bank entry, is characterized
by an increasing level of competition and a high level of bank concentration
(see, among others, Claessens and Laeven (2003), Barros (2007)). This
meaningful development of the banking industry shows deep impacts on
the microeconomic level of the credit institutes operating in this market,
especially on their performance and efficiency (see Claessens and Glaessner
(1999), Okuda and Rungsomboon (2006)). Even though domestic and foreign
banks may be positively affected by increased competition, a broad range
of literature discusses the negative impact of foreign bank entry on
domestic banks’ performance measures (see, among others, Stiglitz
(1994), Sabi (1995) and Claessens (2001)). Another extension of the
competition literature considers decreased bank concentration (e.g.
in Romania), which makes the Structure-Conduct-Performance hypothesis
very interesting (see, among others, Bain (1951), Gilbert (1984), Hannan
(1991), Lahusen (2004), Hahn (2006). In this case, the decreased level
of concentration seems to lead to a decreasing level of performance,
as the increased competition lowers the possibility of monopolistic
market structures.
Romania’s
banking market provides a very good example for investigating the relationship
between foreign bank entry, market structure and bank performance. Since
the beginning of the transition, Romania’s banking sector has experienced
reforms aimed at modernizing the obsolete financial system and making
it more competitive in comparison with Western Europe. An important
part of these reforms was the privatization of the major banks, which
began at the end of the nineties and still remains incomplete. The past
reforms and the EU accession at the beginning of 2007 led to a more
stable investment environment in Romania compared to the Commonwealth
of Independent States (CIS) countries, despite remaining risks. The
concentration in the Romanian bank market has been decreasing steadily
and has forced the participants to apply more competitive strategies.
While most
of the literature deals with either macro or micro aspects of the impact
of foreign bank involvement on the respective countries, we contribute
by
combining both aspects in our investigation. A lot of current research
focuses on risk aspects of foreign banks (see Arvai et. al. (2009),
Maechler and Ong (2009)), but we concentrate on performance and efficiency
aspects of foreign bank entry and presence. Rooted in the structure-conduct-performance
paradigm, we investigate both the impact of foreign banks on macroeconomic
performance of the financial sector and on the microeconomic performance
of the respective banks. Over the 1996 to 2007 period, we run a cross-country
panel regression for 10 New EU Member States (NMS) from Central and
Eastern Europe. In applying Levine’s (1996) dichotomy of foreign banks
influencing macro development via the volume channel (capital accumulation)
and/or the efficiency channel, we investigate both the relationship
between foreign bank penetration and capital flows (FDI, foreign loans,
foreign portfolio investment) and the relationship between foreign bank
penetration and efficiency aspects (share of non-performing loans, interest
rate spread, and loan availability) on the macro side.
Literature
Review
The Structure-Conduct-Performance
Paradigm
The performance-aspect
in the structure-conduct-performance paradigm can be interpreted in
two ways: macroeconomic performance of the financial sector, i.e. its
ability to support capital accumulation (volume channel) and to enhance
the productivity with which capital and labour are combined to produce
goods and services (the efficiency channel); and microeconomic performance
of the market participants (in our study: the banks), e.g. according
to accountancy based measures. Macro-performance deals with the relationship
between finance and growth, and we focus on the special role of foreign
banks in this relationship, given certain macroeconomic structures.
Micro-performance deals with various return measures of the market participants,
i.e. how well they reach their individual goals. Again we focus on the
special role of foreign banks related to market microstructures and
to the strategies of foreign banks.
Given
the significant influence of multinational banks upon their entry into
emerging markets, a lot of studies (see Gilbert (1984), Hannan (1991),
Lahusen (2004), etc.) focused on the relationship between profitability
and the increased level of competition. However, the turnaround effect
of a less concentrated market and a change in performance also gained
strong support by a broad selection of theories and studies to prove
this link. One of the most discussed theories among them seems to be
the Structure-Conduct-Performance
Hypothesis (SCP), which was first applied by Bain (1951). Hannan
(1991) developed a model of the SCP-approach, which finally demonstrates
the positive relation between market concentration and above-average
returns. The model points out that the monopolist is
able to vary output and price in order to maximize profit without considering
the behaviour of his competitors. The assumption of the supply monopole
(structure) determines the market behaviour (conduct) and the performance
of the company. This relation between market structure and performance
has been extended in the banking literature, allowing an analysis of
the connection of market concentration (or market share) and bank performance
(e.g. Gilbert (1984), Hahn (2005, 2006)).
A related theory
to the SCP approach is the Relative-Market-Power
Hypothesis (RMP), which asserts that, independent of the market
concentration, only companies which possess large market shares
and well-differentiated products can exercise market power in pricing
and therefore gain above-average profits (see Shepherd, 1982). The
Efficiency-Structure-Paradigm (ES) mainly has been developed by
researchers of the Chicago School like Demsetz (1973), Peltzman (1977)
and Brozen (1982) and reflects that not collusion but economies of
scope and scale lead to a higher level of profitability and performance.
Consequently, market concentration stems from highly efficient companies
which are able to increase their market share. Berger (1995) extends
the ES Hypothesis through the consideration of X-Efficiency and scale
efficiency and establishes the Efficiency-Structure-X-Efficiency
Hypothesis (ESX). His hypothesis states that firms with predominant
management and product technologies have the advantage of below average
costs and therefore usually gain higher profit (also see Berger (1995)).
The
Relationship between Finance and Growth
King
and Levine (1993a) were among the first who provided an endogenous theoretical
framework, linking financial development with the process of economic
growth. They argue that financial systems “evaluate prospective entrepreneurs
and fund the most promising ones”. Pagano (1993) similarly mentions
the positive impact of financial intermediation on economic growth,
whereby he alludes to the generic character of the term “financial
development.. Therefore, the term financial development in this
thesis always refers to an increase of physical capital accumulation
and efficiency, as explained by Levine (1996). On the one hand,
high-quality financial services can enhance growth by an increase in
the physical capital accumulation, which could result from a higher
national saving rate or from an increased inflow of foreign capital.
On the other hand, financial systems can alter the productivity and
efficiency with which capital and labour are combined in order to produce
goods and services. Regarding the efficient allocation of capital,
Wachtel (2001) lists four ways in which the financial industry improves
allocations. First and as mentioned earlier, the financial sector enhances
screening and monitoring of fund seekers. Second, mobilization
of savings may be encouraged due to innovative attractive instruments
and saving vehicles. Third, financial institutions can take advantage
of economies of scale and therefore reduce costs of project evaluation
and origination. Finally, financial intermediaries provide opportunities
for risk management and liquidity and develop instruments for risk sharing.
(also see Blum et al. (2002) and Haiss and Fink (2007)). Levine (2005)
states that the improvement of financial instruments, markets, and intermediaries
enhances the provision of the basic functions of financial markets,
and ultimately affects long-term economic development through the capital
accumulation or the efficiency channel. Merton and Bodie (1995) focus
in their conceptual framework rather on the functional perspective
than on the institutional aspect. They mention that over time financial
functions are assumed to be more stable and that “competition and
innovation among institutions ultimately result in greater efficiency
in the performance of financial system functions”. In the World Economic
Forum’s Financial Development Report 2008, Roubini and Bilodeau (2008)
holistically define financial development as “the factors, policies,
and institutions that lead to effective financial intermediation and
markets, and deep and broad access to capital and financial services”.
Levine (2008) specifies that countries with large and privately-owned
banks that funnel credit to enterprises and liquid stock exchanges tend
to grow faster than countries with lower levels of financial development
(also see Barisitz (2005), Eller et. al (2006), Hagmayr and Haiss (2007)).
Concurrently
Rajan (2006) mentions that a new deregulated and highly competitive
financial environment leads to an investment return sensitive compensation
system for bank managers, especially returns relative to their competitors.
Managers therefore have greater incentive to take risks and induce superior
performance. Kregel (2007) states that the deregulation wave that started
in 1980 led to a banking system that no longer serviced only business
lending but also emphasised more risky investment opportunities (also
see Winkler (2009))
The Special Role of Foreign Banks
Levine
(1996) developed a theoretical framework which highlights the role of
foreign banks in financial and economic development. He states that
if financial systems stimulate capital formation and enhance economic
efficiency, foreign banks play an essential role. He discusses the following
benefits from foreign banks’ market entries, on the one hand, and
potential costs deriving from the liberalization of foreign bank entry,
on the other.
Levine (1996)
comes to the following conclusion. First, a relaxation of restrictions
on foreign banks entry may increase the capital inflow. Second, economic
growth can be promoted by foreign banks since they contribute positively
to the improvement of the domestic banking system.
The effect capital formation is twofold as foreign banks are considered
to facilitate international capital inflows, and this capital may spur
economic development. Claessens et al. (1995) distinguish between short-term
and long-term capital flows while “the former are deemed as unstable
hot money and the latter are deemed as stable cold money” (also see,
Razin et al. (1998)). Daude and Fratzscher (2008) extend the study on
international capital flows and mention four categories of capital inflows
deducted from the financial account. These are FDI, foreign portfolio
investment (FPI), divided also into equity and debt investment, and
foreign loans (FL). They demonstrate empirically that countries which
are characterized by higher levels of corruption and a lower protection
of property rights receive more FDI and FL than foreign portfolio investments
(also see Gray and He (2001), Clarke et al. (2006), Fink et al. (2007)).
Regarding Levine
(1996), foreign banks enjoy the reputation of contributing directly
to the domestic market by providing new and better skills, management
techniques, training procedures, technology, and products. Moreover,
they may increase the level of competition in and contestability of
the domestic market, which will result in a downward pressure on other
domestic financial institutions to enhance the quality of their offered
financial services (also see Claessens and Glaessner (1998), Claessens
et al. (2001), Goldberg (2004), Bonin et al. (2005), Mehl et al. (2006),
Moshirian (2006) Beck and Martinez Peria (2008)).
Levine (1996)
states that financial arrangements rely strongly on the legal framework
provided by the government, which dictates the rules regarding property
rights, information disclosure, transparency and availability of financial
statements, and the supervisory and regulatory systems observing the
financial system. When all of the mentioned factors are governed in
order to support the financial market, the financial system itself will
provide better financial services. (also Fries et al. (2002), Passera
(2004), Sensarma (2006), Mihaljek (2006), Fink et al. (2007), Staikouras
et al. (2009)).
Performance related to Market structure and Foreign Competitors
The
consequences of foreign bank penetration on the individual banks
themselves, whether domestic or foreign, and on the banking market
in general is important to consider. Several theories approach the performance
measurement of banks in a certain market in different ways. Some studies
(see Claessens and Glaessner (1998), Okuda and Rungsomboon (2006), Yildirim
and Philippatos (2007), Staikouras et al. (2008a)) argue that the level
of foreign bank participation is of significant influence due to the
competition pressure from more efficient and experienced financial institutions.
Other works (see Stiglitz (1994), Sabi’s (1995), Claessens et al.
(2001),Staikouras et al. (2008b)) imply that the changing market structure
seems to be the main driver for performance of the financial intermediaries.
This chapter intends to give an overview of the existing studies on
the linkage between the competitive environment, or the market structure,
and the performance of banks.
Kantzenbach
(1967) developed a competition approach which explains the market structure
conditions for a functional competitive environment. In general, he
defines competitive intensity as the time which a market needs
to eliminate innovative advances by imitating competitors, and mentions
the important functions of competition. He concludes in his approach
that the deep impact of the entry of foreign banks, especially into
a less developed banking market, may occur in the form of a changed
environment, which can be a higher level of competition between the
operating banks. At the same time, the banking industry of a slightly
developed country like Romania moves from a government-led into a market-driven
financial market, which will witness an essential downward trend of
industry concentration (also see Classens and Laeven (2003), Barros
et al. (2007), Yeyati and Micco (2007).
Furthermore,
foreign banks tend to have a significant influence on the host banking
markets through the increased level of competition conduct, which has
a significant impact on the banking market’s profitability performance
(see Claessens and Glaessner (1998), Okuda and Rungsomboon (2006)).
Contrarily, Yildirim and Philippatos (2007) conclude in their study
of eleven Latin American countries that bank returns are negatively
linked to the increased degree of foreign bank participation which stimulates
the degree of competition in national banking markets (also see Staikouras
et al. (2008a)).
Apart from
the already discussed performance impact on the local financial market,
ownership of the operating financial institutions may play an important
role, since domestic banks and their foreign competitors will be affected
in different ways. Stiglitz (1994) discusses the additional costs to
domestic banks due to higher competition with large international banks
which enjoy a better reputation. Consequently, profitability and margins
may slow down and entrepreneurs could receive less access to loans (also
see Sabi (1995), Claessens et al. (2001), Staikouras et al. (2008b)).
Gilbert (1984),
describes the results of 44 studies which test the SCP relation, and
demonstrates that only 32 of these studies report some evidence of significant
association between market structure and measures of bank performance.
In seven of those 32 studies, the coefficients on measures of market
structure are not statistically significant in most of the equations
(also see Lahusen (2004). Hahn (2006) supports the SCP Hypothesis by
the analysis of the Austrian banking market, where the local banks seem
to possess market power due to a concentrated banking market. The significance
of the concentration variables with the dependent profitability indicators
shows the relevance of the SCP model and neglect the RMP and the ES
hypotheses (also seeKosmidou and Pasiouras (2006), Goldberg and Rai
(1995), Fu and Heffernan (2006), Samad (2008), Clayes and Vander Vennet
(2008)).
Macroeconomic
Empirical Evidence
Data
To investigate
how foreign bank participation affects the local financial system, a
data set is used which comprises data of 10 different Eastern European
countries (all of them are members of the European Union) from 1997
– 2007, which include the Czech Republic (CZ), Slovakia (SK), Hungary
(HU), Slovenia (SI), Poland (PL), Estonia (EE), Latvia (LV) and Lithuania
(LT), which entered the European Union in 2005, and Romania (RO) and
Bulgaria (BG), which entered the EU in the year 2007. Data stems mainly
from the database of the European Bank for Reconstruction and Development
(2008), the Vienna Institute for International Economic (WIIW) studies
and the Economist Intelligence Unit (EIU) Country Database.
Methodology and Variables
In order
to investigate the role of foreign banks in an economy, the relationship
between the levels of foreign bank penetration is executed through the
following model, which is applied to Levine’s (1996) hypotheses by
adjusting the model, using different dependent variables for each hypothesis.
For the regressions, the following model specification is used:
Πi, t = a0 + a1
FORi ,t + Σ aj Zij, t
+ εi, t
j = 6
where
Πi,t-1 denotes one of the dependent variables, which will
be in detail explained in the following paragraphs. Since this thesis
aims at examining the impact of foreign bank entry, FORi,t
represents the explanatory variable, which reflects a ratio that
measures foreign bank penetration ratio in country i at time t. Following
the research of Claessens et al. (2001) the degree of foreign bank penetration
is explained by the share of foreign banks in the total number
of banks and the share of foreign bank assets of total bank assets,
which are both ratios provided by the European Bank for Reconstruction
and Development. The asset share presence measure may be appropriate
if foreign banks have an impact on the pricing and profitability of
domestic banks only after gaining market share, while the second measure
is more appropriate when the number of foreign banks indicates competitive
conditions. Z describes a vector of additional control variables of
country i at time t, which could significantly influence the dependent
variable. Subscript i stands for cross-section units, i.e. countries
(i = 1…10), while t denotes time, i.e. years (t = 1997…2007) and
j describes a variable, i.e. control variables (j = 1…6). εi,
t is the error term, and a0, a1
and aj are parameters to be estimated. The explanatory
variable of this thesis possibly may induce international cross-country
externalities, which are hardly measurable, persistent over time and
different across countries. In order to include these externalities
in this panel regression, country-specific effects are included. Due
to the focus of this thesis on differences between specific CEE economies
and not on a randomly sampled pool of countries, the estimation procedure
takes fixed effects into account.
Gray and He
(2001) assume in their developed hypothesis that a higher presence of
foreign financial intermediaries can enhance the quality of the local
financial market and consequently promote inward FDI.
Therefore, the rate of inward FDI as percentage of GDP could be used
as a dependent variable (Π). But since the stock of
inward FDI in percentage of GDP
is considered to give more stable
evidence about the assumed coherence, it represents the first dependent
variable (also see Eller et al. (2008)) The second dependent variable
is foreign portfolio investment, which is deducted from a country’s
financial account under the heading “portfolio investments, liabilities”
expressed in percent of GDP. Another source of international capital
flows is the flow of foreign loans which in this thesis are considered
the third and last type of capital flows. Therefore, the foreign
liabilities of banks in percent of GDP will be used to reflect the
inflow of foreign loans. The latter two indicators were also applied
by Daude and Fratzscher (2008) in their investigation of the pecking
order of international investment capital flows.
In accordance
with the theoretical framework, the link between the degree of internationalization
and the development of domestic financial services will be analyzed
as follows. In this case Π measures the degree of the quality, pricing
and availability of domestic financial services. In accordance with
the investigation of Mehl et al. (2006) the availability of domestic
financial services, or financial depth, will be measured by the
monetization and the intermediation ratio (broad money ratio,
domestic private credit to private sector as percentage of GDP will
be used as a proxy for the intermediation ratio). They argue that these
two variables measure the financial system’s capacity to perform its
main functions and conclusively give evidence about the promotion of
capital accumulation and the pace of productivity growth. Pissarides
(2001) also uses the two mentioned variables in her study on financial
intermediation in South Eastern European countries. Moreover, the
credit to households as percentage of GDP could give empirical evidence
about the development of the retail banking sector on its own (also
see Mihaljek (2006)).
To measure
the quality of financial services, the author follows Barajas
et al. (2000), who use the Non-Performing-Loan
(NPL) ratio as indicator in their studies, intending a significant
proof that foreign banks may worsen loan quality of domestic banks.
They argue that the NPL-ratio among domestic banks increased, which
seems to reflect a fluctuation of higher quality clients to the entering
foreign financial institutions (also see Mihaljek (2006), Podpiera and
Weill (2008)). In addition, Barajas et al. (2000) state that concerning
pricing, the entry of foreign financial intermediaries into a banking
market lowered the intermediation spread1
(also see Claessens et al. (2001) and Clayes and Vander Vennet (2008)).
This thesis
also adapts a set of control variables that capture several basic
country characteristics and then conducts extensive sensitivity analysis
to verify the robustness of the panel analysis. The first control variable
is the market size measured through real GDP, according to Campos
and Kinoshita (2008), who state that generally countries with larger
market sizes draw more capital flows, whereas there might be exceptions.
They also mention that investors could also seek monopoly power instead
of domestic market size, as in Latin America for example. Second, they
also proved the relationship between capital flows and a low inflation
rate, where the latter reflects a stable macroeconomic environment
and is assumed to be a positive signal of credibility to foreign investors
(also see Claessens et al. (2001)). To examine the determinants of financial
depth the GDP growth rate is the third control variable, as applied
by Detragiache et al. (2006),
Lensink et al. (2004) and Koivu (2002) (also see Unite et al. (2003)
Barajas et al. (2000)). The fourth control variable is the Corruption
Perceptions Index (CPI) documented by Transparency International.
Since corruption can cause a slowdown for financial development, as
it causes instability, increased costs and insecure property rights
(see Detragiache et al. (2006)). According to Wachtel (2001), important
factors such as human capital investment and government consumption
may play a role for the development of financial markets. Eller et al.
(2006) also use government consumption in percent of GDP as a
control variable reflecting the size of the public sector in the CEE
region. Regarding educational attainment, this variable reflects
the level of human capital investment, and also is considered a positive
signal to attract potential investors, which is why it is added as the
sixth control variable (see Barro (2000)).
With respect
to Levine’s (1996) hypothesis, which reflects the influence of foreign
banks on the degree of financial regulation, the estimation of any coherence
pursuing the relationship between foreign bank entry and financial regulation
remains a broad topic of qualitative and descriptive research.
Estimation
results
The relationships
between the different variables are discussed after the analysis with
generalized least square regression taking into account fixed effects.
This regression method is used due to the results of tests for heteroscedasticity,
which shows the necessity to take account of the country-specific residuals
of the model and forces a gls-based regression in order to avoid heteroscedastic
effects. Through stepwise regression all insignificant control
variables are removed one by one.
We first discuss
the volume sight and then move on to efficiency. Regarding the link
between foreign bank penetration and FDI, the regression variables,
including all control variables, behave as expected: independently of
using the amount of foreign banks or the asset share of foreign banks
as the explanatory variable, foreign bank penetration is related
positively and highly significant to the inward stock of FDI. This
finding does not only undermine the theoretical approach of Levine (1996),
but is also in line with the empirical results of Grey and He (2001)
and the study of Haiss and Roessl (2008a).
Regarding the
relationship between FPI and the level of foreign bank penetration,
no significant link could be found, which is also in line with findings
of Haiss and Roessl (2008b). The analysis of the linkage between
foreign bank penetration and the flow of foreign loans shows at
the asset share level and the competition level different results. When
using the asset share of foreign banks as the explanatory variable,
no significant relationship can be found, however using the number of
foreign banks reflects a strong correlation with the foreign liabilities
of banks in the NMS, but only at the low significance level. The
absence of a strong significant correlation also supports Mihaljek’s
findings (2006), who also could not find any relationship that banks,
whether foreign or domestic, have an influence on the flow of foreign
loans. The differences in the results come from the fact that foreign
banks with minor activities in the host country (i.e. small in terms
of assets and equity in the host economy) may prefer providing loans
to host country clients directly from their home
country headquarters. Moving to the measurement of the quality of
financial services, as hypothesized in Levine’s (2006) second
assumption, the relationship between foreign bank penetration and
the rate of non-performing loans is significant. Using the foreign
asset share ratio as the independent variable, the negative and strong
significant correlation supports Mihaljek’s (2006) findings that foreign
bank entry improves the efficiency of the local banking market through
an increased level of competition, which results in a decreasing level
of non-performing loans. The other explanatory variable supports this
result with a higher coefficient but with a lower rate of significance.
The intermediation spread appeared to correlate significantly with
both of the explanatory variables, while the significance of the
number of foreign banks ratio was lower at the five percent level. This
underlines the empirical results of Barajas (2000), who “argue foreign
entry was found to be unambiguously beneficial for loan quality as well
as for competition and operative efficiency”. According to them, newly
entered banks had to charge lower spreads in order to gain market share
from their already established competitors. Concerning the influence
of foreign bank penetration on the monetization ratio, both explanatory
indicators demonstrate exactly the same results. The positive link
is significant at the ten percent level and proves the hypothesis
that foreign bank entry can improve the transfer of financial resources
from households and the corporate sector
to the financial sector. This proves the positive impact of foreign
intermediaries on the financial system’s ability to perform its main
functions. Analyzing the intermediation ratio only, FSFDI wields
a significant impact at the five percent level. This complies with
Mihaljek’s argument (2006) that foreign banks do not support the local
banking market with more loans but have a positive influence on its
efficiency. According to Pissarides (2001) low levels of bank credit
to the private sector are caused by a shortage of funds, lack of credit
skills and unavailability of good lending opportunities. Therefore,
through the entry of foreign banks the mentioned domestic financing
obstacles can be reduced and can, after a certain time period, lead
to increased financial intermediation. It seems that foreign banks entered
the NMS markets in the first years of transformation but did not extend
their loan ratios. Due to a lack of data on the household variable,
Romania and Poland had to be excluded from estimations. Regarding the
results, the two explanatory variables give different evidence. While
the asset share indicator demonstrates no significant correlation
with the amount of households in percent of GDP, the number variable
shows a positive and strong significant relationship. It appears
clear that before the entry of foreign banks, the retail sector of the
new member states hardly existed and that the former introduced this
business field to the CEE economies.
Microeconomic
Empirical Evidence
Data
In order
to examine the effect of the market structure on the banks operating
in the Romanian banking market, data for 17 commercial banks operating
in Romania was obtained. The used data includes information for the
period 1999 – 2006 and was drawn from Bankscope Database of
Bureau van Dijk’s company, the Database of the European Bank for Reconstruction
and Development, the National Bank of Romania and the Financial Structure
Dataset of the World Bank. Nine of the banks were foreign owned, five
owned by domestic shareholders and three of the banks in the dataset
were acquired by foreign shareholders during the period. This dataset
comprises the information of between fifty and sixty percent of the
Romanian banking market assets. As result, the author has to mention
that the used panel data may not be representative. Moreover, since
Citibank, the Romanian International Bank, OTP Bank and Volksbank Romania
lack data for the year 1999, the panel data analysis is considered unbalanced.
Methodology
and Variables
The theoretical
foundation for the estimation of the SCP-Modell in its easiest form
is defined as followed:
Πi, t = a0 + a1
Xi, t + Σ aj Zij, t
+ εi, t
J = 4
Since
this thesis aims at examining the impact of the existing market structure
and foreign bank entry, X represents the explanatory variable, which
reflects a ratio that measures the competition in the market or a foreign
bank penetration ratio. Regarding the measurement of the market structure,
a Concentration ratio which is composed of the asset share of
the five largest financial intermediaries is used, as following the
empirical surveys by Samad (2008) and Berger (1995). Another often used
market structure variable is the Herfindahl Index
(HERF), which also gives evidence about the level of concentration in
a banking industry. For a perfectly competitive market the index is
greater than 0 and 100 for a monopoly. Therefore, a low value of HERF
reflects a high amount of banks in the market. Moreover, the HERF takes
into account the number of banks and the inequality of market shares.
Various studies proved the relevance of this indicator in the measurement
of market concentration, such as Smirlock (1985), Berger (1995), Goldberg
and Rai (1996), Hahn (2006).
According
to Claessens et al. (2001), the degree of foreign bank penetration should
be explained through the share of number of foreign banks and
the share of foreign bank assets of total bank assets, where
we use the respective ratios provided by the EBRD. While the first one
can be appropriate when the number of foreign banks indicates competitive
conditions, the asset share presence measure may be appropriate if foreign
banks have an impact on the pricing and profitability of domestic banks
only after gaining market share also adopts these measures in order
to investigate whether market penetration or the presence of foreign
banks could be the reason for changes in bank performance (also see
Barajas (2000)). Similar to various studies (see Berger (1995), Goldberg
and Rai (1996) and Samad (2008)) this thesis uses usual profitability
indicators for measuring bank performance, since profitability is a
consolidated figure which takes into account all products and services
regarding profits and losses. This thesis uses three measures of profits,
which were also applied by Berger (1995), Goldberg and Rai (1996) and
Samad (2008). These variables are firstly the Return on Assets,
which is computed by the division of the total net profits through the
bank’s total assets, and secondly the Return on Equity,
which is calculated by dividing the bank’s total profits by the bank’s
total equity. Thirdly, the Net Interest Margin, relative to total
assets, should serve to show the pricing ability of a bank. Z describes
a vector of additional control variables of country i at time t, which
could significantly influence the dependent variable. Subscript i stands
for cross-section units, i.e. banks (i = 1…17), while t denotes time,
i.e. years (t = 1999…2006) and j describes a variable, i.e. control
variables (j = 1…4). εi, t is the error term, and a0,
a1 and aj
are parameters to be estimated. Because most of the previous studies,
such as Kosmidou and Pasiouras (2006) and Hahn (2006), apply the fix
effects as the more robust estimation method, this study also uses the
fixed effects estimator as the more appropriate choice.
We also apply
a set of control variables in order to underline the robustness
of the adopted model. Following the study of Pasiouras and Kosmidou
(2006), this thesis uses microeconomic variables to show the impact
of bank specific influences and also provides proof of the macroeconomic
impact on the banks’ profitability. The first applied microeconomic
control variable should be the total assets of the bank ass applied
by Goldberg and Rai (1996) and Pasiouras and Kosmidou (2006), who argue
that it reflects the cost difference regarding the bank’s size and
its ability to diversify (also see Bonin et al. (2005). Secondly, the
cost/income ratio demonstrates the efficiency of expense management,
which, according to to Pasiouras and Kosmidou (2006) is one of the main
drivers of bank’s performance and negatively related to it.
With view on
the macroeconomic impacts, the real GDP growth measures the total
economic activity within an economy and as a result should prove a positive
influence on the applied profitability variables, even if Pasiouras
and Kosmidou (2006) found different results for domestic and foreign
banks. Moreover, inflation tends to be related to banks’ performance
since through an anticipated inflation rate banks can timely adjust
interest rates so that revenues increase faster than costs and profitability
raises (see Staikouras et al. (2008a)).
Estimation results
The relationships
between the different variables are discussed after the analysis with
generalized least square regression taking into account fixed effects.
This regression method is used in order to avoid any heteroscedastic
tendencies in the results and to take account of the bank-specific residuals
of the model. The effect of the explanatory variables is measured
for both foreign and domestic banks separately. This should help
to investigate the effect of concentration and foreign bank penetration
on the profitability of each type of bank. To each variable which is
used to measure foreign bank characteristics, the sign d0 is added.
In addition, each indicator for domestic bank issues is marked by d1.
Table 10 summarizes the correlations between the level of banking concentration
and the three different measures of bank performance used in this study.
The results
suggest that neither the Return on Assets, nor the Return of Equity
is affected through level of concentration in the banking sector.
The results are the same for foreign as well as for domestic banks.
The author can only find a significant positive relationship between
the market structure and the net interest margins of foreign banks
at the 5 percent significance level. These findings confirm the study
of Kosmidou and Pasiouras (2006) Goldberg and Rai (1995), which found
no significant negative relation between market concentration and domestic
banks’ performance, but demonstrated a significant relationship between
concentration and foreign bank net interest margins.
Regarding the
first microeconomic control variable, it gives the impression
that only domestic banks are able to take advantage of economies of
scale in order to increase diversification and improve the Return ratio
and the net interest margin. This may come due to the fact that foreign
banks rather lend to big customers (at least initially) and avoid any
risk in newly entered developing countries, while domestic banks have
to expand more rapidly in order to compensate any competitive measures.
These findings rather would support the RMP hypothesis than the SCP-paradigmIt
is important to point out that the cost to income ratio mirrors only
the cost efficiency of a firm, while the efficiency of management and
risk monitoring skills can not be reflected by this simple ratio. Surprisingly,
and not in line with the study of Pasiouras and Kosmidou (2006), the
correlations are of positive nature, a fact which would reflect that
high cost efficiency leads to higher profits. This may be caused
by the fast growth of the Romanian banking market, as characterized
in the descriptive element of the microeconomic part. The fast growing
banking sector combined with a more competitive environment may force
banks to accept higher management expenses and costs in order to raise
the level of profits. Moreover, it looks like they focus rather on
economies of scale than on the improvement of cost efficiency. To
sum up, the significant relationship between cost efficiency and a bank’s
performance is given, although the direction of the relation depends
on the level of sector development and saturation. Considering the macroeconomic
control variables, only foreign banks seem to take advantage of a
growing economy particularly by improving their net interest margin.
Using inflation as an indicator of market development, in times of high
inflation and adverse market conditions foreign banks seem to be able
to improve their Return on Assets. In contrast, in circumstances with
lower inflation or better market conditions the net interest margin
seems to be raised due to the ability of foreign banks to price their
financial products more anti-competitively.
The estimation
results of the relationship between the Herfindahl index, used as a
proxy for the concentration ratio, and the performance measures of banks
in the Romanian financial market are given. The result disagrees with
the above mentioned hypothesis and shows a significant but negative
correlation of the index with the net interest margin of foreign banks.
This induces that concentration would hamper foreign banks in taking
advantage of their market power and pricing their financial products
anti-competitively. Since the two different concentration ratios
show different outcomes, the robustness of the relation should be
doubted. The relationships between banks’ size, cost efficiency and
the applied performance measures appear to be significant and also are
positive. Since the estimations seem robust, the same explanation of
the matter can be drawn, which leads the author to the conclusion
supporting the RMP and the ESX hypothesis.
Turning towards
the relationship between foreign bank penetration and the profitability
of banks operating in the Romanian banking sector, the only significant
relationship which can be derived is the positive correlation between
FSFDI and the net interest margin of foreign banks. This finding
is in line with the study of Claessens et al. (2001), which states
that foreign banks operating in developing countries have higher interest
margins and profitability than domestic competitors. This gives
the impression that foreign banks can price their financial products
better than their domestic competitors.
Changing the
foreign bank asset ratio for the number of foreign bank ratio does support
the positive correlation mentioned above, but with lower significance
at the 5 percent level. No essential change in the correlations of the
control variables can be addressed, only that the inflation rate does
not seem to have any significant influences on the performance of the
financial intermediaries operating the local banking sector. Additionally,
economic growth does not only spur the net interest margin of foreign
banks, but also the Return on Asset ratio is positively related at the
10 percent significance level.
Conclusion
This thesis
discusses the possible effects of foreign bank penetration on developing
countries in two ways. First, it assesses important influences of foreign
bank entry on international capital flows and the improvement of the
quality and availability of financial services. Methodically, it applies
descriptive data and panel regression analyses in order to measure the
impacts of foreign bank presence in the New Member States (NMS) from
Central and Eastern Europe (CEE) over the 1999 – 2007 period. Second,
it examines the essential competition pressures of foreign banks on
the profitability level of banks operating in the Romanian banking market.
Therefore it gives a detailed descriptive overview on the history of
the Romanian banking market and utilizes panel regression analyses
with the purpose of evaluating the impact of foreign banks on 17
large Romanian banks.
The literature
review proposes that financial development enhances economic development
by raising physical capital accumulation and efficiency. The improvement
of financial instruments, markets and intermediaries enhances the provision
of the basic functions of financial markets and therefore the appropriate
allocation of investment. In this thesis the focus lies on the role
of financial intermediaries, more specifically on the special role of
foreign banks. On the one hand, the main benefit of foreign bank entry
is the improvement of access to international capital, which can flow
in the form of foreign direct investment, foreign portfolio investment
or foreign loans. On the other hand, foreign bank penetration can improve
the quality, price and availability of financial services offered by
banks. Previous research on this topic found that more foreign capital
is channelled to transition countries which enjoy a higher degree of
foreign bank presence (See Gray and He (2001), Clarke (2006), Fink (2007)).
Addressing this issue, the literature suggests foreign banks as an important
factor raising the level of competition in the domestic financial sector.
This significantly influences the accomplishment of market functions
through the behaviour of market participants and finally ends in an
optimal competitive environment which is characterized by a certain
market structure. This development has been demonstrated in previous
studies, which argue that a larger presence of foreign banks leads to
a more competitive banking system (See Claessens and Laeven (2003),
Barros et al. (2007)). Furthermore, this increased level of competition
influences the financial companies operating in this environment significantly
by pressuring their performance. Several past theses stress this relationship
and found empirical evidence for it (See. Claessens and Glaessner (1998),
Okuda and Rungsomboon (2006)). The literature also highlights that this
may affect foreign and domestic banks in different ways. Domestic
banks may suffer from this new circumstance and their profitability
and margins may slow down (See Stiglitz (1994), Sabi (1995),Claessens
et al. (2001), Staikouras et al. (2008b)). The Structure-Conduct-Performance
Hypothesis discusses the same subject from the point of banking concentration,
stating a positive relation between market concentration and above-average
returns. Related but contrary theories are the Relative-Market-Power
Hypothesis, which assert large market shares are the reason for above-average
profits, and the Efficiency-Structure-Paradigm, which argues that rather
efficiency is the answer for increasing profits. These theories are
discussed and empirically tested by different researchers with varying
results (See Gilbert (1984), Berger (1995), Lahusen (2004), Hahn (2006),
Kosmidou and Pasiouras (2006), Goldberg and Rai (1995), etc.).
In the macroeconomic
empirical section, the thesis presents descriptive presentations of
the data and panel regressions with fixed-effects and generalized least
square-regressions. The panel analysis focuses on the New EU Member
States covering the years between 1997 and 2007. Based on the Levine
(1996) framework that foreign banks can stimulate financial development
via capital accumulation or via raising the efficiency of the financial
sector, we start with the volume channel.
Regarding the
link between FSFDI (measured by the asset share of foreign-owned
banks and the number of foreign banks) and FDI, the estimation
demonstrates a positive and significant relation between the
two indicators, inducing that a more attractive economic environment
to foreign investors due to an advanced banking market with a certain
amount of foreign competitors attracts FDI.
No significant
association between foreign bank presence and foreign portfolio investment
and foreign loans could be found. Following Levine (1996), we investigate
the impact of foreign banks on the efficiency of the financial sector
by approximating efficiency with non-performing loans, the interest
rate spread and availability measures. Also, concerning the impact of
foreign financial intermediaries on the amount of non-performing loans,
it seems that foreign bank entry significantly improves the efficiency
of the local banking market through an increased level of competition.
The intermediation spread appeared to correlate significantly with foreign
bank entry and gives conclusive evidence of the benefits for loan quality
as well as for competition and operative efficiency.
A positive and significant
relationship with the monetization ratio is given and proves the ability
of foreign banks to perform its main functions and to transfer of financial
resources from households and from the corporate sector to the financial
sector. (Seventhly,) the correlation with private credit to the domestic
private sector appears weak and therefore undermines the argument that
foreign financial intermediaries do not support the local economy with
more loans but have a positive influence on its efficiency. Finally,
there seems to be a significant relationship between the household ratio
and the number of foreign banks, confirming that retail banking in CEE
nearly did not exist before the entry of foreign banks.
In the microeconomic
empirical section, the thesis presents panel regressions with fixed-effects
and generalized least square-regressions. The panel analysis focuses
on 17 Romanian banks covering the years between 1999 and 2006
and will be carried out separately on domestic and foreign banks.
Due to a lack of data, the panel analysis is carried out unbalanced
and the sample might be considered as not representative.
Regarding the relationship
between banking market concentration and the two ratios Return on Assets
and Return on Equity, no significant relation can be found in this
thesis.
Using net interest
margin as the independent variable, a positive significant association
reflects that foreign banks can set their prices anti-competitively.
Moreover, the significance of the correlation between bank concentration
and the microeconomic variables bank size and cost efficiency demonstrates
the acceptance of the RMP and ESX hypothesis rather than the proof
of the SCP-paradigm.
Proving the impact
of foreign bank presence on the local banking market, the only positive
correlation observed was between foreign bank presence and net interest
margin. This result also proves that foreign banks can price their
financial products better than their domestic competitors.
To sum
up, the preliminary macroeconomic empirical results prove that the presence
of foreign banks can influence the banking market of a transition country
in a very positive way by allocating more international capital to the
local economy and also by enhancing the efficiency of the financial
sector in several ways. The tentative microeconomic results declare
only little support for the SCP paradigm, when foreign banks only influence
competition and not profitability.
Appendix
Table
1: Summarized estimation results with FSFDI
Table
2: Summarized estimation results with NRFB
Table
3: Banks included in panel analysis
Name
of bank
Dataset available
Reporting Strd.
Ownership
ABN Amro
1999 – 2006
IFRS unqual.
Foreign
Alpha Bank
1999 – 2006
IFRS unqual.
Foreign
Banca Romaneasca
1999 – 2006
IFRS gen. unqual.
Foreign since 2003
Bancpost
1999 – 2006
IFRS unqual.
Foreign since 2002
Banca Comerciala
Romana
1999 – 2006
IFRS gen. unqual.
Foreign since 2005
BRD
1999 – 2006
IFRS gen. unqual.
Foreign
CEC
1999 – 2006
IFRS mixed
Domestic
Emporiki
Bank
1999 – 2006
IFRS gen. unqual.
Domestic
Export-Import
Bank
1999 – 2006
IFRS unqual.
Domestic
Firenze Romania
1999 – 2006
IFRS gen. unqual.
Foreign
Libra Bank
1999 – 2006
IFRS gen. unqual.
Foreign
Piraeus Bank
1999 – 2006
IFRS gen. unqual.
Domestic
Tiriac Bank
1999 – 2006
IFRS mixed
Foreign
Citibank
Romania
2000 – 2006
IFRS mixed
Foreign
Romanian
International Bank
2000 – 2006
IFRS mixed
Domestic
OTP Bank
Romania
2000 – 2006
IFRS unqual.
Foreign
Volksbank
Romania
2000 – 2006
Local GAAP
Foreign
Source: Bankscope
Database Bureau van Dijk Company
Table 4:
Summarized estimation results with bank concentration ration
Table
5: Summarized estimation results with Herfindahl Index
Table
6: Summarized estimation results with FSFDI
Table
7: Summarized estimation results with number of foreign banks ratio
Bibliography
Asaftei, Gabriel /Kumbhakar, Subal C. (2008): Regulation and
efficiency in transition: the case of Romanian banks, Journal of Regulatory Economics,
33 (3): 253 – 282.
Arvai,
Zsofia /Driessen, Karl /Otker-Robe, Inci (2009): Regional Financial
Interlinkages and Financial Contagion Within Europe, IMF Working
Paper No. 09/6.
http://www.imf.org/external/pubs/cat/longres.cfm?sk=22569.0
Backe, Peter / Walko, Zoltan
(2006):
Bank Intermediation in Southeastern Europe: Depth and Structure,
OeNB Working thesis, Vienna: Österreichische Nationalbank.
BACA, Bank Austria CEE Quarterly
Report 2009, Vienna: Bank Austria.
Bain, Joe
S. (1951): Relation of Profit Rate and Industry Concentration: American
manufacturing, 1936 – 1940, in The Quarterly Journal of Economics,
65 (3): 293 – 324.
BankScope
(2008a). Bank Reports for BCR, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008b). Bank Reports for BRD, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008c). Bank Reports for Raiffeisen Bank, Brussels: Bureau van Dijk
Electronic Publishing.
BankScope
(2008d). Bank Reports for UniCredit Tiriac Bank, Brussels: Bureau van
Dijk Electronic Publishing.
BankScope
(2008e). Bank Reports for Banca Transilvania, Brussels: Bureau van Dijk
Electronic Publishing.
BankScope
(2008f). Bank Reports for Alpha Bank Romania, Brussels: Bureau van Dijk
Electronic Publishing.
BankScope
(2008g). Bank Reports for Bancpost, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008h). Bank Reports for ABN Amro, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008i). Bank Reports for Banca Romaneasca, Brussels: Bureau van Dijk
Electronic Publishing.
BankScope
(2008j). Bank Reports for CEC, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008k). Bank Reports for Citibank Romania, Brussels: Bureau van Dijk
Electronic Publishing.
BankScope
(2008l). Bank Reports for Emporiki Bank, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008m). Bank Reports for Export-Import Bank Romania, Brussels: Bureau
van Dijk Electronic Publishing.
BankScope
(2008n). Bank Reports for Firenze Romania, Brussels: Bureau van Dijk
Electronic Publishing.
BankScope
(2008o). Bank Reports for Libra Bank, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008p). Bank Reports for OTP Bank, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008q). Bank Reports for Piraeus Bank, Brussels: Bureau van Dijk Electronic
Publishing.
BankScope
(2008r). Bank Reports for Romanian International Bank, Brussels: Bureau
van Dijk Electronic Publishing.
BankScope
(2008s). Bank Reports for Volksbank Romania, Brussels: Bureau van Dijk
Electronic Publishing.
Barajas,
Adolfo /Steiner, Roberto /Salazar, Natalia (2000): The impact of
liberalization and foreign investment in Colombia’s financial sector,
in Journal of Development Economics, 63: 157 – 196.
Barisitz, Stephan (2005):
Der Transformationsprozess im rumänischen Finanz- und Bankensektor,
in Finanzmarktstabilitätsbericht 7 der OeNB, Vienna: Österreichische
Nationalbank 2005.
Barisitz, Stephan (2005): Banking in Central and Eastern Europe
since the Turn of the Millennium: An Overview of Structural Modernization
in Ten Countries,
OeNB Working Thesis, Vienna: Österreichische Nationalbank.
Barros, Carlos Pestana
/Ferreisa, Candida /Jonathan, Williams (2007): Analysing the determinants
of performance of best and worst European banks: A mixed logit approach,
in Journal of Banking and Finance, 31: 2189 – 2203.
Barro,
Robert J. / Lee, Jong-Wha (2000): “International Data on Educational
Attainment: Updates and Implications”. Working Thesis, Center
for International Development at Harvard University, No. 42, April 2000.
Beck, Thorsten
/Martinez Peria, Maria Soledad (2008): Foreign Bank Acquisition and
Outreach: Evidence from Mexico,
Policy Research Working Thesis 4467, Development Research Group, World
Bank.
Beck, Thorsten /Demirgüc-Kunt,
Asli /Laeven, Luc /Vojislav, Maskimovic (2006): The Determinants
of Financing Obstacles, in Journal of International Money and Finance
25: 932 – 952.
Berger,
Allen N. (1995): "The Profit-Structure Relationship in Banking
− Tests of Market-Power and Efficient-Structure Hypothesis",
Journal of Money, Credit, and Banking, 27 (2): 404-431.
Blum, David
/ Federmair, Klaus / Fink, Gerhard / Haiss, Peter (2002): The
Financial-real Sector Nexus: Theory and Empirical Evidence, EuropaInstitut
Working Paper Nr. 43, WU Wirtschaftsuniversität Wien, Sept. 2002.
http://ssrn.com/author=115752
Bonin,
John P. /Hasan, Iftekhar /Wachtel, Paul (2005): Bank performance,
efficiency and ownershi