Journal of Banking & Finance 36 (2012) 2949–2959

Contents lists available at SciVerse ScienceDirect

Journal of Banking & Finance

journal homepage: www.elsevier.com/locate/jbf

Foreign bank entry, credit allocation and lending rates in emerging markets:
Empirical evidence from Poland
Hans Degryse a, Olena Havrylchyk b, Emilia Jurzyk c,⇑, Sylwester Kozak d
a
University of Leuven, Tilburg University and CEPR, Department of Accounting, Finance and Insurance, Naamsestraat 69, 3000 Leuven, Belgium
b
CEPII, 113, rue de Grenelle, 75015 Paris, France
c
International Monetary Fund, 700 19th Street NW, Washington, DC 20431, USA
d
National Bank of Poland, ul. Swietokrzyska 11/21, 00-919 Warsaw, Poland

a r t i c l e i n f o a b s t r a c t

Article history: Earlier studies have documented that foreign banks charge lower lending rates and interest spreads than
Received 4 November 2010 domestic banks. We hypothesize that this may stem from the superior efficiency of foreign entrants that
Accepted 6 December 2011 they decide to pass onto borrowers (‘‘performance hypothesis’’), but could also reflect a different loan allo-
Available online 13 December 2011
cation with respect to borrower transparency, loan maturity and currency (‘‘portfolio composition hypoth-
esis’’). We are able to differentiate between the above hypotheses thanks to a novel dataset containing
JEL classification: detailed bank-specific information for the Polish banking industry. Our findings demonstrate that banks
G21
differ significantly in terms of portfolio composition and we attest to the ‘‘portfolio composition hypothe-
G28
G34
sis’’ by showing that, having controlled for portfolio composition, there are no differences in lending rates
L11 between banks.
Ó 2012 International Monetary Fund. Published by Elsevier B.V. All rights reserved.
Keywords:
Banks
Ownership
Credit allocation

1. Introduction
High foreign participation in the emerging banking markets has
led to a heated debate on the benefits and costs of foreign bank en-
try for the stability and efficiency of the banking systems, as well as
for borrowers. In this paper we focus on lending rates, which re-
flect the costs of using the financial system by borrowers. High
lending rates can become prohibitive to many safe customers,
blocking access to finance and hindering economic growth. Most
empirical studies find that foreign banks on average charge lower
lending rates and have lower spreads than domestic banks. For in-
stance, Martinez Peria and Mody (2004) study Latin American
countries during the late 1990s and find that foreign banks have
lower spreads than domestic banks and takeover banks have high-
er spreads than greenfield banks. Claeys and Hainz (2007) argue
that foreign banks can undercut domestic banks’ lending rates ow-
ing to their better screening skills.
⇑ Corresponding author. Tel.: +1 202 623 4223; fax: +1 202 589 4223.
E-mail addresses: hans.degryse@econ.kuleuven.be (H. Degryse), olena.havryl-
chyk@cepii.fr (O. Havrylchyk), ejurzyk@imf.org (E. Jurzyk), Sylwester.Kozak@-
mail.nbp.pl (S. Kozak).
The empirical findings quoted above can have two plausible
explanations. On the one hand, foreign banks could charge lower
lending rates because they have lower operating costs. There is a
considerable evidence that foreign institutions are able to over-
come cross-border disadvantages (distance, monitoring costs, dif-
ferences in institutional environment, language and culture) and
operate more efficiently than their domestic competitors in emerg-
ing economies (see e.g. Berger et al., 2000; Bonin et al., 2005; Chen
and Liao, 2011; Havrylchyk and Jurzyk, 2011; Weill, 2003). If for-
eign banks choose to pass on their higher efficiency to borrowers
in order to gain market share, this ‘‘performance’’ effect is expected
to be identical for all borrowers (‘‘performance hypothesis’’).1 It
should be stressed that if foreign banks are more efficient than other
banks but rely on this cost advantage to extract rents from borrow-
1
This hypothesis is related to the Berger et al. (2000) global advantage hypothesis.
Foreign entrants may also raise revenues through superior investment or risk
management skills, by providing better service quality/variety that some customers
prefer, or by obtaining diversification of risks that allows them to undertake higher
risk-higher expected return investments. Further, foreign banks in transition and
developing economies additionally benefit from their better access to international
capital markets and funding from their parent companies. This diminishes their cost
of funds, which in turn should be translated into lower lending rates, benefiting
borrowers. Moreover, foreign banks might enjoy lower cost of deposits due to their
superior reputation.

0378-4266/$ - see front matter Ó 2012 International Monetary Fund. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.jbankfin.2011.12.006
2950 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959
ers instead of charging lower lending rates, we would interpret this
as the rejection of the ‘‘performance hypothesis’’.
Alternatively, foreign banks could charge lower lending rates if
they supply credit to more transparent and less risky borrowers.
Dell’Ariccia and Marquez (2004) argue that foreign banks face large
information disadvantages and, thus, prefer to target more trans-
parent clients (relying on transactions-based lending), whereas
domestic banks are better placed to lend to firms based on soft
information (relationship lending) (see also Althammer and
Haselmann, 2011; Detragiache et al., 2008; Gormley, 2011). Seng-
upta (2007) shows that foreign entrants exploit their cost advan-
tage by offering collateralized loans to large transparent firms
whereas incumbent banks retain more risky borrowers. Even if for-
eign banks are able to collect soft information, they might have
more difficulties or be less willing to communicate it to their head-
quarters. Stein (2002) argues that organizations with more hierar-
chical structures are more likely to rely on hard information as
opposed to organizations with flatter structures. The reason is that
flatter organizations have better control and information on their
managers, and thus can afford to give them more discretion, which
allows them to use soft information. This modeling has been ex-
tended to large banks by Berger et al. (2005) and can further be ap-
plied to foreign banks, which belong to large multinational banking
groups, and where communication of soft information is
obstructed not only by the hierarchy, but also by cultural,
institutional and linguistic barriers. Numerous empirical studies
support these theories and find that foreign bank lend less to
opaque borrowers, such as small and medium enterprises, than
domestic banks (Clarke et al., 2005, 2006 for Latin America;
Berger et al., 2001, Gormley, 2010 for India; Giannetti and Ongena,
2009, 2012 for Central and Eastern Europe; Mian, 2006 for
Pakistan).
Credit allocation can also differ in other dimensions. Foreign
banks could extend more short-term loans to solve asymmetric
information problems as wells as to secure ‘‘hot’’ money that is
readily retracted during crises (Rodrik and Velasco, 2000; Popov
and Udell, 2010).2 In addition, foreign banks could supply more for-
eign currency loans because they rely less on domestic deposits and
have better access to the international capital markets and financing
from the parent banks (see Brown et al., 2008; Beer et al., 2010; ECB,
2006; Farnoux et al., 2004; Sorsa et al., 2007). Since loans to trans-
parent borrowers and in foreign currency have lower interest rates
than other types of loans, the observed lower lending rate of foreign
banks could be easily explained by a different loan allocation (‘‘port-
folio composition’’ hypothesis).3
In our paper, we distinguish between the ‘‘performance’’ and the
‘‘portfolio composition’’ hypotheses and, to the best of our knowl-
edge, this is the first to attempt to do this. We are able to differen-
tiate between the two hypotheses thanks to a unique database on
all Polish commercial banks that includes detailed information on
several dimensions of credit allocation – lending rates, amounts
of loans, and amounts of nonperforming loans to different borrower
types and in different currencies and maturities. We think that Po-
land provides an excellent testing ground for our hypotheses be-
cause it has the largest banking sector in Central and Eastern
Europe. Since the share of foreign investors in Polish banks amounts
to 74%, this provides us with a large number of foreign banks and at
the same time leaves us with a much higher number of domestic
banks than in most other Central and Eastern European countries
(CEECs). Furthermore, as we will argue below, the institutional
2
For example, Dooley and Shin (2000) argue that foreign creditors’ run from
Korean banks triggered the crisis in Korea in 1997.
3
Our two hypotheses are not mutually exclusive. Lower lending rates of foreign
banks might reflect both performance and portfolio composition effects.
environment of Poland is close to the average of other emerging
markets, and can therefore be considered as representative.
We differentiate between greenfield banks (foreign banks that
enter via greenfield investment) and takeover banks (foreign banks
that acquire an existing domestic institution), because theory sug-
gests that the impact of foreign banks’ behavior depends on their
mode of entry (see e.g., Claeys and Hainz, 2007). Indeed, both
hypotheses, ‘‘performance’’ and ‘‘portfolio allocation’’, could be
strongest for greenfield banks, because they have a larger effi-
ciency advantage and a larger informational disadvantage relative
to domestic and foreign takeover banks. The latter inherit ineffi-
cient and non-transparent organizational structure and could be
burdened by nonperforming loans, but they also obtain the person-
nel and access to loan information that may help them overcome
informational asymmetries.
Our findings can be summarized as follows. We find that port-
folio composition of banks in terms of borrower opacity, loan
maturity and currency differs significantly depending on bank
ownership and mode of entry. Most importantly, having controlled
for these differences, foreign banks do not charge lower lending
rates than their domestic counterparts. This result contrasts with
Martinez Peria and Mody (2004) and Claeys and Hainz (2007) that
suffered from omitted variable bias and, thus, were not able to con-
trol for the portfolio composition of banks. Finally, we confirm the-
oretical predictions of Dell’Ariccia and Marquez (2004) that the
entry of foreign banks via greenfield investment leads to the real-
location of lending by domestic banks to more opaque borrowers
increasing the riskiness of their portfolios.
The rest of this paper is organized as follows. Section 2 presents
our data. Sections 3 and 4 describe our empirical findings on port-
folio allocation and loan rates, respectively. Section 5 deals with
the impacts of foreign entry on domestic banks. Section 6
concludes.
2. Polish banking industry and data
We test our hypotheses using a novel dataset that was kindly
provided by the National Bank of Poland. It contains quarterly
information on 110 Polish banks4 between December 1996 and
December 2006. In addition to standard information from balance
sheets and income statements, it contains data on interest income,
amount of granted loans, and nonperforming loans for two borrower
types: private firms and individual entrepreneurs.5 This allows us to
examine lending volumes and lending rates that banks with differ-
ent ownership structure charge their transparent and non-transpar-
ent borrowers.6 Our data gives us a unique opportunity to construct
banks’ portfolio shares, interest rates, market shares, bank concen-
tration measures, and nonperforming loans for each borrower type
separately.
In principle, transparent firms should have reliable financial
statements, long credit history, and good collateral, all of which
helps the bank to evaluate borrower’s creditworthiness. Private
4
We define a bank as Polish if it is registered in Poland and the National Bank of
Poland collects information on it.
5
The distinction between private firms and individual entrepreneurs is grounded
in Polish law. A borrower is classified as a ‘‘private firm’’ if the firm is owned by
private investors (either entirely, or where the private share exceeds 50%), and is
either subject to commercial law or is subject to civil law and employs more than 9
workers. Additionally, such firms have to comply with accounting regulations that
require full bookkeeping. ‘‘Individual entrepreneurs (for short, entrepreneurs)’’, in
contrast, are small firms employing up to 9 workers, subject to civil law, and using
simplified accounting procedures.
6
We also have information on three other groups of borrowers: state-owned
enterprises, farmers and households. We decided not to use these three groups
because they cannot be unambiguously ranked in terms of transparency and, hence,
we have opted to rely on data for only two groups – large private firms and individual
entrepreneurs.
 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959
firms in our sample are closer to this definition than entrepreneurs
and, thus, we label them as transparent borrowers. Even though
they include all private firms with more than 9 employees, most
of them are rather large, and average statistics are driven by large
private firms. We classify entrepreneurs as opaque because they
are small, young entities with little available accounting informa-
tion, and that makes it more challenging for a bank to judge their
capacity and willingness to repay. Evaluating small firms presents
a particularly acute problem in emerging markets, where many
entrepreneurs do not properly report their profits and losses. It im-
plies that a firm might have larger turnover and assets than it de-
clares officially, but it also implies that the firm has unrecorded,
contingent senior liabilities to its employees (de la Torre et al.,
2010).
The Polish banking sector provides a good testing ground for
our hypotheses because, similar to other CEECs, it experienced
massive foreign direct investments into the banking sector (see
Farnoux et al., 2004). At the end of 2006, the share of foreign inves-
tors in Polish banking constituted 74%. This is less, however, than
in other major CEECs – for example in Hungary more than 80% of
banking assets is in foreign hands, and in the Czech Republic and
Slovakia foreign banks control more than 95% of assets. The rela-
tively smaller presence of foreign banks in Poland gives us addi-
tional strong argument to use Poland as a case study, since there
are still local private and state-owned banks present on the market
that can serve as a benchmark.
The institutional environment of Poland is representative of
many other emerging markets. For example, the restrictions on
FDI in the financial sector are close to the average of other emerg-
ing markets (Poland has a score of 0.33 versus an average of 0.3 for
the developing countries; see e.g. Koyama and Golub, 2006), and
the restrictions on permissible banking activities in Poland is close
to the OECD average (see e.g. Barth et al., 2004). Between 1992 and
1998, conditional licensing was applied to foreign banks, meaning
that a foreign bank could obtain a license only after agreeing to
rehabilitate a distressed Polish bank. The privatization process
started in 1993. Even though foreign investors were allowed to
participate, they were entitled only to minority shares. Restrictions
on foreign banks were removed in 1998 after the passing of new
laws on banking, which were in line with EU legislation. The high
minimum capital requirement of ECU 5 million accelerated the
involvement of foreign banks, since domestic banks faced difficul-
ties in raising these amounts on the local market. Banking supervi-
sion in Poland is executed by the Commission for Banking
Supervision and the National Bank of Poland. According to Barth
et al. (2004), their independence is high. Finally, the corporate
bond market was quite underdeveloped during our sample period.
While the corporate bond market was liberalized in 2000, the mar-
ket is still small compared with other important emerging markets.
Poland has a credit registry since 1997 but only for individuals.
During 1996–2006, there were a number of domestic mergers
and acquisitions in the Polish banking sector, hence we treat
merged institutions as two before the merger and as one after-
wards.7 For our estimations we have deleted the first four quarters
of operations for both greenfield and takeover banks in order to ex-
clude the initial setting-up and transformation period. Finally, dur-
ing our sample period, there was no greenfield domestic entry.
7
We also investigated the effect of domestic mergers on banks’ lending rates by
including a dummy that takes a value of one if the bank had undergone a domestic
merger, and zero otherwise. Our estimations revealed that this merger dummy was
never statistically significant and, therefore, we decided to exclude it from our final
results. Most of domestic mergers included either a merger of two domestic banks or
an acquisition of the domestic bank by a foreign bank. There were also a few cases of
mergers of two foreign banks which belonged to the same foreign financial group.
2951
Tables 1 and 2 report variable definitions and summary statistics,
respectively. First, we compare rates on loans to different types of
borrowers. Lending rates offered to private firms—the most trans-
parent borrowers—are significantly lower than lending rates
charged to entrepreneurs. The difference between them is 0.6%
points per quarter or 2.4% points per year. This is in line with the the-
oretical considerations that lending rates in the competitive markets
with the smallest informational asymmetries should be the lowest
whereas lending rates for opaque borrowers are the highest due to
high switching costs stemming from informational asymmetries.8
The share of large private firms (transparent borrowers) and
small entrepreneurs (opaque borrowers) amount on average to
48% and 11% of banks’ portfolios, respectively. It should be men-
tioned, however, that this average figure hides a large heterogene-
ity between banks: loans to transparent borrowers constitute 67%
of greenfield banks’ portfolios, while domestic private banks de-
vote one and a half times more of their portfolio to opaque borrow-
ers than other banks. The summary statistics reveal that short-
term loans amount to 41% of banks’ portfolios. Once again, loan
maturity differs according to bank ownership and mode of foreign
entry. Greenfield banks issue more short-term loans while take-
over banks issue less short-term loans than domestic private banks
(53%, 36% and 41%, respectively).
One of the characteristics of many CEECs is the high proportion
of loans in foreign currency. Their share ranges from around 10–
20% in Czech Republic and Slovakia, to 60–70% in the Baltic States
(see ECB, 2006). Our dataset shows that Polish banks extended 20%
of their loans in foreign currency, which is not very high in com-
parison to other countries. Still, this poses significant risks for the
banking sector, as many borrowers, especially the SMEs, are not
hedged against currency and interest rate risk.9 Foreign banks, par-
ticularly greenfield institutions, give more loans in foreign currency
than domestic banks, which is likely to be due to their better access
to international capital markets and parent companies. This appears
to be supported by disparities in average cross-border interbank lia-
bilities. Even though, on average, 12% of liabilities stem from foreign
banks, this figure is much higher for greenfield banks (15%) and
rather low for private and state-owned domestic banks (2%).
3. Empirical analysis: Portfolio allocation
3.1. Empirical model and strategy
We first study the impact of foreign bank ownership on the
allocation of loans with respect to borrower type, maturity and
currency. We model banks’ portfolio shares as a function of bank
ownership and mode of entry, as well as controls for bank charac-
teristics, and macroeconomic environment. In particular, we esti-
mate the following model:
 
Pit
ln ¼ a0 þ a1 Ownershipit1
1  Pit
þ a2 Bankcharacteristicsit1 þ a3 Macrot1
þ Quartert þ eit ; ð1Þ
8
While there was a general trend for all lending rates to decrease over 1996–2006,
the spread between lending rates to private firms and entrepreneurs has not changed
in a significant way.
9
A substantial fraction of foreign currency loans are extended in Swiss Francs, on
which lending rates are lower than on Euro loans (see ECB, 2006). This characteristic
is shared by other countries in CEECs, such as Hungary and Slovenia. This trend comes
from Austria where most of loans in foreign currency are denominated in Swiss Franc.
Originally this was constrained to regions bordering Switzerland where firms and
individuals had a natural hedge against currency risk since their income was often in
Swiss Franc. However, lending in Swiss Francs is now extended to other parts of
Austria and to CEECs where Austrian banks are active.
2952 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959

Table 1
Definition of variables.

Variable Definition
State-owned A dummy variable which takes the value of one if more than 50% of the bank is owned by the state, zero otherwise
Takeover A dummy variable which takes the value of one if more than 50% of the bank has been acquired by foreign investors. Before 2001 is a dummy
variable that equals one for years up to 2001, and zero otherwise
Greenfield A dummy variable which takes the value of one if foreign bank entered via greenfield investment and more than 50% of the bank is owned by
the foreign investor
Lending ratea The ratio of interest income to total loans. Calculated on a quarterly basis, unless explicitly stated otherwise
NPLa The ratio of nonperforming loans to total loans, measured as a deviation from the annual median for all banks divided by this median
Market sharea Share of loans of a bank in the total loans of banking sector in host country
Capitalization The ratio of risk adjusted capital
Costs The ratio of personnel and administrative costs to total assets
Herfindahl indexa Herfindahl index, calculated as the sum of the squared banks’ market shares of loans
Share privatea The ratio of loans to private firms (entrepreneurs) in bank’s portfolio
(entrepreneur)
Share greenfield The ratio of loans granted by greenfield (takeover) banks to total loans granted by all banks in quarter t
(takeover)
FX loansa The ratio of loans in foreign currency in bank’s loan portfolio
Interbank liabilities The ratio of foreign interbank liabilities in bank’s total liabilities
Short-term loansa The ratio of loans with maturity less than 1 year in a bank’s loan portfolio
Long-term loansa The ratio of loans with maturity over 5 years in a bank’s loan portfolio
GDP Real quarterly growth rate of GDP
Inflation Quarterly inflation rate
Real interest rate Quarterly real short-term interest rate, calculated using 3 month WIBOR (Warsaw interbank offered rate) interest rate and inflation rate by
Fisher equation
Trend A variable that varies from 1 to 41 depending on the quarter and is intended to capture a time trend. 1 refers to December 1996 and 41 to
December 2006
a
This variable is computed for all borrowers and for private corporations and entrepreneurs separately.

Table 2
Summary statistics.

Obs. Mean Std. dev. Min. Max.

Lending rate 2272 0.043 0.031 0.004 0.559
Private firms 2233 0.042 0.035 0 0.551
Entrepreneurs 1903 0.048 0.041 0 0.547
State-owned 2272 0.135 0.341 0 1
Takeover 2272 0.316 0.465 0 1
Greenfield 2272 0.315 0.464 0 1
Costs 2272 0.011 0.007 0.001 0.110
Capitalization 2272 0.188 0.263 3.569 4.189
GDP growth 2272 0.015 0.087 0.150 0.128
Inflation 2272 0.014 0.014 0.009 0.056
Real interest rate 2272 0.017 0.012 0.005 0.049
NPL 2272 0.356 1.233 1 10.110
Private firms 2233 0.416 1.410 1 10.800
Entrepreneurs 1903 0.394 1.426 1 9.012
Herfindahl index 2272 0.072 0.008 0.053 0.085
Private firms 2233 0.066 0.008 0.051 0.084
Entrepreneurs 1903 0.070 0.012 0.053 0.095
Market share 2272 0.017 0.030 0 0.184
Private firms 2233 0.017 0.029 0 0.225
Entrepreneurs 1903 0.020 0.032 0 0.218
Share private 2269 0.485 0.306 0 1
Share entrepreneurs 2269 0.113 0.115 0 0.640
FX loans 2272 0.203 0.214 0 0.999
Private firms 2233 0.198 0.205 0 1
Entrepreneurs 1903 0.157 0.218 0 1
Short-term loans 2272 0.415 0.255 0 1
Private firms 2233 0.523 0.254 0 1
Entrepreneurs 1903 0.408 0.254 0 1
Long-term loans 2272 0.232 0.212 0 1
Private firms 2233 0.179 0.193 0 1
Entrepreneurs 1903 0.136 0.154 0 1
Share greenfield 837 0.080 0.020 0.044 0.116
Share takeover 837 0.360 0.268 0.102 0.710
Interbank liabilities 617 0.124 0.177 0 0.972

where Pit is the share of loans (by borrower type, by maturity, or by
currency in loan portfolio of bank i at time t, respectively); Owner-
shipit—dummy variables that capture the effect of bank ownership
(State-owned) and mode of foreign bank entry (Takeover and Green-
field) for bank i at time t; Bankcharacteristicsit —variables that control
for Capitalization, Costs, the share of nonperforming loans (NPL) and
Market Share; Macrot —variables that control for Inflation, real short-
term interest rate (Real interest rate) and real GDP growth (GDP). We
lag all explanatory variables by one period and also include quar-
terly dummies.
 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959
Following Berger et al. (1998) and Clarke et al. (2006) we esti-
mate all share equations in log-odds logit form which allows us
to get around the problem that shares by definition are bounded
between 0 and 1. Since our data set is in panel version, initially
we have to choose between fixed and random effects estimation
methods. The former, however, presents significant problems be-
cause it does not allow us to estimate coefficients for one of our
variables of interest – time-invariant dummy for greenfield banks.
Consequently, we opt for Hausman and Taylor (1981) estimation
technique that allows some of the explanatory variables to be
correlated with the unobserved individual-level random effect
(hence, controlling for endogeneity of explanatory variables),
but also permits estimation of coefficients for exogenous time-
invariant variables.10 We treat State-owned and Takeover dummies,
as well as bank variables as time-variant and endogenous, a Green-
field dummy as time-invariant and exogenous, and country level
variables as time-variant and exogenous. Since these estimators
are based on instrumental variables, we compute Sargan-Hansen
test of overidentifying restrictions where the null hypothesis states
that the excluded instruments are valid instruments, i.e., uncorre-
lated with the error term and correctly excluded from the esti-
mated equation.
Our main variables of interest are two dummy variables that
correspond to modes of foreign bank entry (greenfield and take-
over), while private domestic banks serve the role of benchmark
for our estimation. To investigate convergence over time, we
interact each of our mode of entry variables (takeover and green-
field) with a time trend. We presume that the time period in
which foreign banks operates is more important than banks’
age. For example, a foreign bank that enters Poland in the later
period is likely to have already been present in another CEEC,
and hence already has a lot of experience in operating in a tran-
sition economy. Indeed, banks that were the first ones to enter
Polish banking markets in the early 1990s are also those that lend
most to large firms, which reflects the fact that they came to Po-
land following their clients and were not interested in the local
market.
As bank-level characteristics, we include Capitalization, Costs,
Market share and NPLs to capture financial health, market power
and performance of banks, which might be correlated with bank
strategies. The expected signs for these bank controls are unclear
because of conflicting hypotheses. Healthier banks with higher
profits, better loan quality and higher capitalization might be
able to grow faster over time and to expand to areas where it
takes time and effort to acquire know-how of the business, like
lending to SMEs. Similarly, healthier banks should be able to ex-
tend more foreign currency loans and loans at longer maturity,
because they have better access to foreign and long-term fund-
ing. On the other hand, banks that suffer from financial distress
can be required by government supervisors, depositors, and cap-
ital market investors to reduce their risk profile. This would lead
to a decrease in relationship lending to informationally opaque
small businesses because the risk of these loans cannot be easily
verified (Berger et al., 2001). Hainz (2005) additionally argues
that firms can more easily switch from distressed banks as
asymmetric information problems are lowered. Similarly, regula-
tors might ask distressed banks to reduce their lending in for-
eign currency to reduce their currency risk. All these
arguments lead us to suppose a positive relationship between
bank health and the share of loans to opaque clients, at longer
maturity and in foreign currency.
10
We also check the robustness of results by applying the fixed effect vector
decomposition procedure using three stage fixed effects methodology of Pluemper
and Troeger (2007) and our coefficients of interest remain robust.
2953
Contrary to the above hypothesis, failing banks might ‘‘gamble
for resurrection’’ and lend to riskier borrowers, such as SMEs
(Clarke et al., 2006). Moreover, large firms could be less willing
to continue their relationship with such banks, whereas opaque
clients face more difficulties switching to other lenders, and dis-
tressed banks might benefit from their ‘‘captured’’ clients. They
might also be more willing to lend in foreign currency disregarding
currency risk considerations.
While our bank-level variables, such as Costs and Capitalization
are calculated at the bank level, nonperforming loans (NPL) and
Market Share are calculated both at a bank and at a bank-portfolio
level, i.e., separately for entrepreneurs and private firms. In the first
case, we control for the general health and market power of the
bank, whereas in the second case the health and market power
in the particular part of a loan portfolio is controlled for. Since
the definition of NPL has been subject to changes during the ana-
lyzed period, we rely on a deviation of NPL from its annual median
for all banks divided by this median. To control for the macroeco-
nomic environment and the cost of capital, we include real short-
term interest rate (Real interest rate), inflation rate (Inflation) and
real GDP growth (GDP). In general, we expect that periods of high
economic growth should be more beneficial for riskier and opaque
clients, such as small entrepreneurs. Benign economic conditions
should also increase lending in domestic currency and at longer
maturity.
3.2. Portfolio allocation: results for borrower type
The results of estimating Eq. (1) are reported in Table 3: col-
umns 1–2 present findings with the share of private firms in
the bank portfolios, and columns 3–4 with the share of entrepre-
neurs as dependant variables. Our findings show that, controlling
for other factors, foreign banks that entered via greenfield invest-
ment devote 14% less of their portfolios to entrepreneurs (column
3) while they lend over 84% more to private firms (column 1)
than domestic private banks. This result points to a comparative
disadvantage of greenfield banks in lending to opaque clients
and is in line with earlier findings for other developing and
transition economies (Clarke et al., 2005, 2006; Berger et al.,
2001; Giannetti and Ongena, 2009, 2012; Gormley, 2010; Mian,
2006).
The results for takeover banks do not indicate a clear bias in
terms of borrower transparency, as these banks have relatively
less both large firms and entrepreneurs in their portfolios. As
it was mentioned earlier, this study does not consider lending
to state-owned firms, farmers and households, because they
are not easily classified in terms of transparency. Nevertheless,
it is important to mention that the above result is possible
because takeover foreign banks extend a lot of loans to
households.
To investigate convergence, we interact each of our takeover
and greenfield variables with a time trend (columns 2 and 4).
We find that there is some convergence between banks in terms
of portfolio composition as differences between greenfield and
domestic private banks in lending to opaque clients (column 4)
diminish with time. These findings reflect growing interest of
greenfield banks in lending to SME, brought by two main devel-
opments: 1) fierce competition in the credit market for large cor-
porations and 2) improved ability of foreign banks’ subsidiaries to
finance relatively opaque SMEs (De Haas and Naaborg, 2006).
Even though greenfield banks were still less able and less willing
to engage in relationship lending, they developed other technolo-
gies that helped them to overcome opaqueness of entrepreneurs,
such as small business credit scoring, asset-based lending, or
fixed assets lending. They were able to do this in the later period,
2954 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959

Table 3 3.3. Portfolio allocation: results for loans at different maturities

Borrower type: share of loans to private firms and entrepreneurs.

Private firms Private firms Entrepreneurs Entrepreneurs The results of share regressions for loan portfolios at different
(transparent) (transparent) (opaque) (3) (opaque) (4) maturities are reported in Table 4. We find that, controlling for
(1) (2) other factors, greenfield banks lend 68% more at short-term (up
Greenfield 1.644*** 1.676*** 1.802*** 2.654*** to 1 year) than domestic private banks, and this result is stable over
(0.579) (0.591) (0.547) (0.568) time. Interestingly, takeover banks grant more loans with a matu-
Takeover 0.615*** 0.092 0.743*** 0.601***
(0.111) (0.151) (0.111) (0.154)
rity of over 5 years than domestic banks, which might be due to sta-
Greenfield*trend 0.006 0.031*** ble and long-term funding due to well-developed depositor
(0.006) (0.008) network that they have inherited. This difference however becomes
Takeover*trend 0.028*** 0.005 less important as the negative coefficient on ‘‘takeover⁄trend’’ in
(0.006) (0.006)
the last column reveals. As to bank characteristics, we notice that
State-owned 0.336** 0.397** 0.549*** 0.581***
(0.155) (0.155) (0.155) (0.154) banks with a higher market share and healthy banks (measured
Costs 18.284*** 17.792*** 2.119 5.535 by capital ratios and nonperforming loan ratios) are also those that
(3.248) (3.269) (3.632) (3.686) lend long-term.
Capitalization 0.610*** 0.598*** 0.678*** 0.711***
(0.131) (0.131) (0.195) (0.195)
GDP growth 0.237 0.118 1.934* 1.655 3.4. Portfolio allocation: results for loans in foreign currency
(1.068) (1.062) (1.145) (1.141)
Inflation 2.028 1.819 14.410* 13.467*
Table 5 reports the results of our portfolio share regressions for
(6.869) (6.833) (7.495) (7.462)
Real interest 1.519 1.153 15.614** 14.729* loans in foreign currency. We find that greenfield banks offer a
rate higher share of loans in foreign currency, but there is a conver-
(7.373) (7.333) (7.895) (7.861) gence with domestic banks so that, in terms of statistical signifi-
Market share 11.126*** 11.175*** 14.767*** 12.429*** cance, the difference disappears after 2002 (model 2). There is no
(2.299) (2.301) (2.213) (2.255)
Herfindahl 11.148 10.513 0.129 0.677
difference between foreign takeover and domestic banks in their
index propensity to extend loans in foreign currency.
(7.646) (7.605) (5.738) (5.718) The higher share of foreign currency loans granted by greenfield
NPL 0.154*** 0.149*** 0.170*** 0.178*** banks might reflect their better access to international capital mar-
(0.020) (0.020) (0.023) (0.023)
kets either directly or via their parent banks. We test this hypoth-
Observations 2221 2221 1923 1923
Number of 107 107 101 101 esis by including the share of interbank liabilities with nonresident
banks banks to capture foreign funding (model 3).12 The share of inter-
Sargan-Hansen 0.643 0.639 0.231 0.215 bank liabilities is significantly positive, indicating that banks of
(p-value) any type that have higher share of their funding from nonresident
Table presents the results of Eq. (1). The dependent variable is the log-odds ratio of banks grant more loans in foreign currency as well. Importantly,
the share of loans to private firms and entrepreneurs in banks’ portfolios. The table the inclusion of this variable removes the significance of the green-
lists coefficients and standard errors (in parentheses) from a Hausman and Taylor field dummy, which suggests that greenfield banks grant more for-
(1981) regression. All independent variables except for ownership dummies are
eign currency loans only to the extent that they have a higher
lagged by one quarter. Regressions include seasonal dummies. Definitions of vari-
ables are provided in Table 1. share of foreign interbank liabilities.
*
Significance at 10% level. Among other variables, market share is significant in the share
**
Significance at 5% level. regressions, showing that larger banks have on average better ac-
***
Significance at 1% level.
cess to foreign currency liabilities, which they lend onto borrowers.
The decrease in the inflation and interest rates made lending in for-
eign currency less attractive for borrowers, diminishing the growth
as legal and accounting systems have improved, making some in these loans on average.
SMEs more transparent. 11
As to other bank-specific variables, we find that a bank de-
creases the share of loans to a specific borrower type if nonper- 4. Empirical analysis: loan rates
forming loans (NPL) in its loan portfolio for that particular
borrower type increase. Importantly, this effect is not found for 4.1. Empirical model and strategy
loans to private firms if we replace portfolio level NPLs with bank
level NPLs (for brevity, these results are not presented in the pa- Section 3 has clearly shown that the portfolio composition dif-
per). Thus, the result is driven not by the overall bank health, but fers across bank types. Therefore, we now look at the determinants
by the health of its different loan portfolios. Banks with a higher of bank lending rates to test whether there remain differences be-
market share in the Polish banking market allocate a larger fraction tween interest rates charged by banks with different ownership
of their portfolio to private firms and entrepreneurs. structure when we control for portfolio composition. Our strategy
to show that the portfolio composition effect is at work involves
three steps. First, we establish that greenfield foreign banks appear
11
New studies question the argument that large and foreign banks are not capable to charge lower lending rates if we do not control for portfolio
to lend to SMEs (Berger and Udell, 2006; de la Torre et al., 2010). The reasoning is that composition. Second, we show that Greenfield foreign banks do
latest advances in credit scoring methodologies coupled with enhanced computer
not charge different interest rates than domestic private banks
power and increased data availability make transaction lending technologies to be
well suited for funding small firms (Mester, 1997; Petersen and Rajan, 2002). This is once we control for portfolio composition. Third, we investigate
especially true when credit scores are based on the owner’s personal consumer data whether interest rates charged by Greenfield foreign banks differ
obtained from consumer credit bureaus, which is combined with data on the SME from domestic private banks within a more homogeneous set of
collected by the financial institutions. These studies still agree that small domestic borrowers, i.e. private firms and entrepreneurs.
banks have an advantage to gather and process soft information, but they argue that
large and foreign banks are also able to lend to SMEs, but using ‘‘hard’’ information-
12
based technologies. In this case, we should not observe differences in portfolio We can test this hypothesis only for a limited time period, as the data on
allocations of loans between different types of banks. interbank liabilities is reported only after 2001.
 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959 2955

Table 4 ter (or 4.8% points on an annual basis), whereas takeover banks do
Loan maturity: share of loans by maturity in banks’ portfolios. not charge less than domestic banks, which serve as a benchmark.
Up to Up to Over Over Among bank specific variables, the deviation from the median non-
1 year (1) 1 year (2) 5 years (3) 5 years (4) performing loans and market share are significant and have the ex-
Greenfield 0.781** 0.585 0.692 0.701 pected signs. Banks with higher market power and higher credit
(0.382) (0.398) (0.509) (0.525) risk are more likely to charge higher lending rates.
Takeover 0.176** 0.270** 0.290** 0.725*** In line with the literature that analyzes whether benefits of for-
(0.083) (0.116) (0.129) (0.181)
Greenfield*trend 0.004 0.005
eign ownership are constant over time, we interact our takeover and
(0.004) (0.007) greenfield variables with a time trend. Our findings (column 2)
Takeover*trend 0.021*** 0.024*** show that the impact of greenfield diminishes over time, which is
(0.004) (0.007) usually interpreted as convergence between banks of different
State-owned 0.265** 0.228* 0.604*** 0.633***
types of ownership due to competition or changes in portfolio com-
(0.119) (0.118) (0.179) (0.179)
Costs 2.112 3.316 5.278 5.003 position.13 Our results corroborate previous finding in the literature
(2.467) (2.488) (4.620) (4.674) and, hence, we are confident that our data reflect the situation in
Capitalization 0.499*** 0.497*** 0.843*** 0.844*** developing and transition countries and do not just deal with a partic-
(0.094) (0.094) (0.145) (0.146) ular Polish case (Martinez Peria and Mody, 2004; Claeys and Hainz,
GDP growth 1.105 1.255 1.440 1.332
2007).
(0.809) (0.803) (1.284) (1.281)
Inflation 2.729 3.297 9.174 8.715 As mentioned earlier, the disadvantage of the above empirical
(5.284) (5.241) (8.214) (8.195) specification is the possibility of an omitted variable due to the
Real interest rate 4.325 5.178 10.866 10.010 non-inclusion of information on borrower type. Such bias does
(5.642) (5.597) (8.775) (8.756)
not allow us to pinpoint the real reasons for lower lending rates
Market share 6.263** 7.801*** 20.020*** 18.079***
(2.504) (2.512) (3.718) (3.739) of greenfield banks, i.e., their superior performance or their portfo-
Herfindahl index 3.666 4.114 3.381 3.947 lio composition targeted to more transparent borrowers. Our first
(5.426) (5.383) (8.457) (8.439) step to remedy this omitted variable problem is to augment model
NPL 0.071*** 0.069*** 0.135*** 0.138*** (2) by the share of transparent and opaque borrowers in banks’
(0.018) (0.018) (0.028) (0.028)
portfolio into our baseline regression. We further control for matu-
Observations 2241 2241 2100 2100
Number of banks 108 108 108 108 rity and currency composition of banks’ portfolios as those may
Sargan–Hansen (p- 0.764 0.616 0.943 0.801 also determine average lending rates. The results, presented in col-
value) umn 3 of Table 6, clearly indicate that the impact of bank owner-
Table presents the results of Eq. (1). The dependent variable is the log-odds ratio of ship and foreign banks’ mode of entry disappears: average
the share of loans with maturity up to 1 year (models 1 and 2) and over 5 years lending rates between banks are no longer statistically different.
(models 3 and 4) in banks’ portfolios. The table lists coefficients and standard errors It seems, therefore, that the previous findings suffered from the
(in parentheses) from a Hausman and Taylor (1981) regression. All independent omitted variable bias, which rendered some of the ownership vari-
variables except for ownership dummies are lagged by one quarter. Regressions
include seasonal dummies. Definitions of variables are provided in Table 1.
ables significant.
*
Significance at 10% level. As expected, banks with a higher share of transparent borrow-
**
Significance at 5% level. ers and a lower share of opaque borrowers exhibit lower lending
***
Significance at 1% level. rates. Banks with a higher share of foreign currency loans also
demonstrate, on average, lower interest rates, whereas loans at
longer maturities are priced more expensively.14,15 Thus, the ob-
We expect to find that the lower loan rates of foreign banks, re- served earlier lower lending rates of greenfield banks are rather ex-
ported in Martinez Peria and Mody (2004) and Claeys and Hainz plained by their propensity to extend loans to transparent firms, in
(2007), are driven by the portfolio composition. To identify the
importance of portfolio composition on lending rates, and for com-
parability with previous studies, we first estimate a baseline lend- 13
We also use a Koyck transformation that allows computing long-term coefficients
ing rate model with specification similar to these earlier studies. In by assuming that in the distributed lag model that the effect of explanatory variable
this specification, we use lending rate for all borrowers as a depen- diminishes as the lag gets larger. The obtained result is that the long-term effect of
greenfield and takeover dummy variables is zero which is compatible with our
dent variable and examine the effect of bank ownership and mode results.
of entry on the lending rate for an average borrower, controlling for 14
The theoretical impact of loan maturity on cost of credit is ambiguous as it
bank characteristics, macroeconomic environment and market reflects two opposite effects. A borrower that issues short-term debt can face costly
structure. liquidations at expiration which motivates it to opt for longer-term debt. At the same
time, lenders prefer to give short term loans because of agency problems, such as
To be more formal, we estimate the following empirical model:
asset substitution and underinvestment. As a result, borrowers are willing to incur
and lenders demand higher lending rates for loans with longer maturity. Alterna-
Lit ¼ a0 þ a1 Ownershipit1 þ a2 Bankcharacteristicsit1 tively, lenders might ration credit to risky borrowers and force them to take short-
þ a3 Macrot1 þ a4 MarketStructureit1 þ Seasont þ eit ; ð2Þ term loans, which would decrease average lending rates on long-term loans.
Empirical evidence supports both hypotheses for corporate loans and bonds
(Gottesman and Roberts, 2004; Helwege and Turner, 1999). For an individual firm,
where the variables are as defined before (see Tables 1 and 2) ex-
the spread typically increases with maturity, reflecting greater uncertainty. At the
cept for Lit , the lending rate of bank i during quarter t, and same time, safer firms tend to issue longer-dated bonds or have access to long-term
MarketStructureit , which includes variables that control for market bank credit, which causes the average spread to decline with maturity.
15
concentration and market power (Herfindahl Index and Market Other explanatory variables have the expected sign. Riskier loans are more
Share). expensive, banks with a higher market share charge higher lending rates, macroeco-
nomic environment with higher inflation and interest rates leads to higher lending
rates. Banks with higher costs charge lower lending rates which appears to contradict
4.2. Results for the average bank lending rate earlier studies that find a positive impact of higher costs. It should be noted that
earlier studies considered cost-ratio to be exogenous, while we treat it as an
endogenous variable. If we consider cost-ratio to be exogenous, its impact is no longer
The results of estimating model (2) applied to the entire bank’s statistically significant. Thus, the previously found positive association is likely due to
portfolio are presented in the first column of Table 6. We find that the fact that loans with higher interest rates are also those that demand more
greenfield banks charge their borrowers 1.2% points less per quar- administrative costs.
2956 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959

Table 5 foreign banks is not an important determinant of lending rates

Foreign currency: share of loan portfolio in foreign currency. for both groups of borrowers. We additionally find that state-
Share of Share of Share of Share of owned banks offer lower lending rates to private firms and
foreign foreign foreign foreign entrepreneurs (columns 1 and 3). Annualized, the difference
currency currency currency currency amounts to 3.2% points, and is both statistically and economi-
loans (1) loans (2) loans (3) loans (4)
cally significant.16 Lower lending rates of state-owned banks are
Greenfield 0.985* 1.825*** 1.008 0.344 consistent with the idea that these banks are less influenced by
(0.519) (0.535) (2.569) (1.778)
Takeover 0.089 0.067 1.854 1.372
profit considerations than private banks and provide subsidized
(0.148) (0.194) (2.895) (1.939) loans. In line with our previous results, a higher share of foreign
Greenfield*trend 0.042*** 0.024 currency loans has a negative impact on average lending rates
(0.006) (0.019) for both private firms and entrepreneurs, whereas long-term loans
Takeover*trend 0.008 0.057***
are more expensive.
(0.007) (0.020)
State-owned 0.184 0.148 2.355 0.113 Our findings show that time does not play an important role
(0.191) (0.190) (2.907) (1.837) and even after many years of operations lending rates of foreign
Costs 15.881*** 20.793*** 9.321 4.980 banks are not different from those of domestic private banks. Our
(5.203) (5.189) (8.523) (8.620) results are robust to different specifications of variables accounting
Capitalization 0.538*** 0.400*** 0.808*** 0.830***
for time dynamics. For example, to check the stability of our results
(0.138) (0.138) (0.189) (0.189)
GDP growth 0.868 1.145 5.002* 4.723* we split the sample into early and later period, use bank age and an
(1.267) (1.250) (2.741) (2.821) interaction variable between age and foreign bank dummies. Our
Inflation 17.771** 16.892** 18.905 21.574 results remain the same and we do not find any age or time dy-
(8.194) (8.085) (12.575) (14.056)
namic effects. These results are not reported in the paper for brev-
Real interest 16.584* 15.778* 30.863** 33.155**
rate ity, but are available upon request.
(8.782) (8.665) (13.210) (13.882) Our results are contrary to the existing literature, which
Market share 12.283*** 11.974*** 9.266** 13.749*** shows that lending rates of greenfield banks converge in time
(3.380) (3.409) (4.419) (4.671) with lending rates of other banks, interpreted as convergence
Herfindahl 17.541** 18.139** 91.685*** 78.350***
in performance. In contrast, we attest to the changing portfolio
index
(8.480) (8.368) (21.600) (28.201) composition of foreign banks. As greenfield banks increase the
NPL 0.060** 0.080*** 0.038 0.029 share of their loans to entrepreneurs and domestic banks extend
(0.025) (0.025) (0.042) (0.044) more loans in foreign currency, we observe a convergence in
Interbank 1.539*** 1.465***
portfolio composition. Consequently, this has contributed to the
liabilities
(0.458) (0.455)
convergence in interest rates over the analyzed period, but this
Observations 1730 1730 617 617 is purely a portfolio composition effect. It should be noted that
Number of 92 92 47 47 we do not argue that greenfield banks have the same level of
banks efficiency as other types of banks. In fact, greenfield banks ap-
Sargan–Hansen 0.113 0.125 0.339 0.152
pear to be more efficient than other banks even after controlling
(p-value)
for portfolio composition (results not reported in the paper due
Table presents the results of Eq. (1). The dependent variable is the log-odds ratio of to space constraints). However, their superior efficiency is not
the share of loans in foreign currency in banks’ portfolios. The table lists coefficients
passed onto customers but rather gives these banks larger scope
and standard errors (in parentheses) from a Hausman and Taylor (1981) regression.
All independent variables except for ownership dummies are lagged by one quarter. to extract rents from borrowers, contributing to higher
Regressions include seasonal dummies. Definitions of variables are provided in profitability.
Table 1.
*
Significance at 10% level.
**
Significance at 5% level. 5. Findings for domestic banks
***
Significance at 1% level.
We have seen so far that foreign banks have a different portfolio
composition than domestic private institutions. However, the en-
foreign currency and their unwillingness to lend long-term. Our re- try of foreign banks influences the supply of loans not only directly
sults also hold if we account for the dynamic effects (column 4): we via lending by these banks, but also indirectly by influencing the
find no evidence of convergence in foreign bank’s interest rates. We behavior of domestic institutions. For example, higher competition
presume that convergence in lending rates observed in earlier stud- in the market for transparent borrowers would induce domestic
ies and in column 2 of Table 6 stems from convergence in portfolio banks to lend more to opaque clients. At the same time, if green-
composition. field banks cherry pick the best borrowers, the quality of loan port-
folios of domestic banks could deteriorate.
From the perspective of public policy, it is not important if for-
4.3. Lending rate: results for borrower types
eign banks prefer to target more transparent clients as long as
domestic banks continue to lend to SMEs (Detragiache et al.,
To further test the impact of portfolio composition on loan
2008). Dell’Ariccia and Marquez (2004) predict that the entry of
rates, we estimate model (2) separately for bank’s portfolios of pri-
foreign banks leads to a segregation, where foreign bank lend to
vate firms and entrepreneurs. This allows us to remove from our
transparent borrowers whereas domestic banks increase their sup-
estimations all portfolio composition effects and to observe the
ply of loans to opaque and risky clients in the wake of foreign bank
pure effect of bank ownership and mode of entry on lending rates
entry.
to those two groups of borrower types.
We present our results for transparent borrowers—private
16
firms—in columns 1 and 2 of Table 7 and for opaque The results presented in Table 6 show that the difference between lending rates
charged by state-owned banks and domestic private banks is even larger if portfolio
ones—entrepreneurs—in columns 3 and 4 of Table 7. We also composition is not accounted for. This earlier result is explained by a much larger
control for potential effects of maturity and foreign currency. propensity of state-owned banks to grant loans to state-owned firms at very low
Our results (columns 1 and 3) show that the mode of entry of interest rates.
 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959 2957

Table 6 Table 7
Average bank lending rate. Loan rates to borrower types – private firms and entrepreneurs.

Lending Lending Lending Lending Private firms Private firms Entrepreneurs Entrepreneurs
rate (1) rate (2) rate (3) rate (4) (transparent) (transparent) (opaque) (3) (opaque) (4)
(1) (2)
Greenfield 0.012* 0.015** 0.001 0.006
(0.007) (0.007) (0.007) (0.008) Greenfield 0.007 0.009 0.002 0.015
Takeover 0.003 0.004 0.002 0.002 (0.007) (0.007) (0.006) (0.010)
(0.002) (0.003) (0.002) (0.003) Takeover 0.005 0.008 0.006 0.004
Greenfield*trend 0.000*** 0.000 (0.003) (0.007) (0.004) (0.006)
(0.000) (0.000) Greenfield*trend 0.000 0.000
Takeover*trend 0.000** 0.000 (0.000) (0.000)
(0.000) (0.000) Takeover*trend 0.000 0.000
State-owned 0.027*** 0.028*** 0.026*** 0.027*** (0.000) (0.000)
(0.003) (0.003) (0.003) (0.003) State-owned 0.008* 0.007** 0.008* 0.016***
Costs 0.181** 0.250*** 0.288*** 0.338*** (0.004) (0.003) (0.005) (0.005)
(0.088) (0.089) (0.064) (0.064) Costs 0.031 0.022 0.145 0.015
Capitalization 0.019*** 0.018*** 0.021*** 0.017*** (0.094) (0.083) (0.119) (0.133)
(0.002) (0.002) (0.003) (0.002) Capitalization 0.010*** 0.009*** 0.028*** 0.026***
GDP growth 0.023 0.002 0.025 0.000 (0.003) (0.001) (0.006) (0.007)
(0.022) (0.022) (0.022) (0.021) GDP growth 0.005 0.006 0.014 0.034
Inflation 0.337*** 0.156*** 0.330*** 0.114** (0.028) (0.026) (0.038) (0.042)
(0.043) (0.047) (0.042) (0.045) Inflation 0.634*** 0.643*** 0.187 0.367
Real interest rate 0.441*** 0.282*** 0.525*** 0.338*** (0.182) (0.181) (0.248) (0.275)
(0.045) (0.048) (0.045) (0.047) Real interest 0.723*** 0.730*** 0.203 0.360
Market share 0.208*** 0.227*** 0.135* 0.140** rate
(0.074) (0.072) (0.072) (0.069) (0.195) (0.194) (0.261) (0.290)
Herfindahl index 0.153** 0.095 0.225*** 0.050 Market share 0.035 0.045 0.043 0.001
(0.065) (0.070) (0.063) (0.067) (0.064) (0.064) (0.071) (0.080)
NPL 0.012*** 0.013*** 0.011*** 0.011*** Herfindahl 0.130 0.125 0.115 0.128
(0.000) (0.000) (0.000) (0.000) index
FX loans 0.024*** 0.031*** (0.202) (0.202) (0.189) (0.211)
(0.004) (0.004) NPL 0.007*** 0.007*** 0.010*** 0.008***
Short-term loans 0.002 0.005 (0.001) (0.001) (0.001) (0.001)
(0.004) (0.004) FX loans 0.027*** 0.027*** 0.028*** 0.030***
Long-term loans 0.044*** 0.056*** (0.005) (0.005) (0.005) (0.006)
(0.005) (0.005) Short-term 0.006 0.006 0.012** 0.019***
Share private 0.017*** 0.016*** loans
(0.005) (0.005) (0.004) (0.004) (0.005) (0.006)
Share entrepreneur 0.032*** 0.016** Long-term loans 0.010* 0.008 0.034*** 0.031***
(0.008) (0.008) (0.006) (0.006) (0.008) (0.009)
Observations 2271 2271 2271 2271 Observations 2233 2233 1903 1903
Number of banks 110 110 110 110 Number of 108 108 102 102
Sargan–Hansen (p- 0.419 0.524 0.135 0.156 banks
value) Sargan–Hansen 0.135 0.157 0.456 0.463
(p-value)
Table presents the results of Eq. (2). The dependent variable is the bank-specific
average lending rate. The table lists coefficients and standard errors (in parenthe- Table presents the results of Eq. (2). The dependent variable is the bank-specific
ses) from Hausman and Taylor (1981) regression. All independent variables except interest rate on loans to private firms (columns 1 and 2) and entrepreneurs (col-
for ownership dummies are lagged by one quarter. Regressions include seasonal umns 3 and 4). The table lists coefficients and standard errors (in parentheses) from
dummies. Definitions of variables are provided in Table 1. Hausman and Taylor (1981) regression. Herfindahl Index, Market Share, FX loans,
*
Significance at 10% level. Short-term loans and Long-term loans are calculated for each borrower type. All
**
Significance at 5% level. independent variables except for ownership dummies are lagged by one quarter.
***
Significance at 1% level. Regressions include seasonal dummies. Definitions of variables are provided in
Table 1.
*
Significance at 10% level.
**
Significance at 5% level.
***
Significance at 1% level.
Even though the above theoretical model does not account for
the mode of foreign bank entry, it implies that the above segrega-
tion effect should be present in the wake of greenfield bank entry.  
Pit
The results for takeover banks are more ambiguous and depend on ln ¼ d0 þ d1 Stateit1 þ d2 Sharegreenfieldt1
the fact whether inheriting existing loan portfolios and personnel 1  Pit
allows them to enjoy information advantages similar to domestic þ d3 Sharetakeov ert1
banks or whether their willingness to rely on ‘‘soft’’ information þ d4 Bankcharacteristicsit1 þ d5 Macrot1
decreases. In the first case, takeover and domestic private banks
become direct competitors, which might have negative effects þ Seasont þ eit ; ð3Þ
(since they would compete in the same markets but foreign banks
Then we look at the nonperforming loans:
would have a cost advantage) but also positive spillover effects. In
the second case, acquisition of banks by foreign investors should NPLit ¼ c0 þ c1 stateit1 þ c2 Share greenfieldt1
have a segregation effect similar to the entry of greenfield banks
and induce the remaining domestic banks to lend more to opaque þ c3 Share takeov er t1 þ c4 Bank characteristicst1
borrowers. þ c5 Macrot1 þ Seasont þ eit ; ð4Þ
To test these hypotheses, we analyze the impact of foreign bank
entry on the supply of loans and portfolio quality of domestic where in all regressions Share greenfieldt is the share of greenfield
banks. To this end, we estimate two models on the subsample of and Share takeovert is the share of takeover banks in the total bank-
domestic banks. First, we analyze the portfolio shares: ing loans at quarter t, and the other variables are as defined above.
2958 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959

Table 8
Domestic banks: share of loans to private firms and entrepreneurs.

Share private (1) Share entrepreneurs (2) NPL private (3) NPL entrepreneurs (4)
*** **
Share greenfield 4.808 3.189 13.672 11.223***
(3.425) (1.207) (5.771) (2.185)
Share takeover 0.617*** 0.635*** 0.693** 0.155
(0.169) (0.191) (0.289) (0.351)
State-owned 0.142 0.283** 0.809*** 0.648**
(0.164) (0.143) (0.279) (0.262)
Costs 16.474** 30.464*** 46.719*** 7.637
(7.728) (5.982) (12.391) (11.007)
Capitalization 1.721*** 0.551** 0.450 0.194
(0.253) (0.223) (0.427) (0.410)
GDP growth 1.122 3.723*** 4.813** 1.194
(1.346) (1.133) (2.250) (2.085)
Inflation 14.122*** 6.997*** 10.642** 0.750
(2.706) (2.343) (4.623) (4.309)
Real interest rate 14.760*** 10.077*** 9.643* 3.167
(3.106) (2.535) (5.282) (4.665)
Market share 13.594*** 12.913*** 1.486 21.063***
(2.331) (3.879) (3.989) (7.091)
Herfindahl index 11.913* 1.651 20.771* 1.215
(6.726) (2.913) (11.374) (5.360)
NPL 0.145*** 0.077***
(0.022) (0.021)
Observations 837 837 837 837
Number of banks 56 56 56 56
R-squared 0.20 0.18 0.06 0.06

Table presents the results of Eq. (3).The dependent variable is the log-odds ratio of the share of loans to private firms (column 1) and entrepreneurs (column 2) in domestic
banks’ portfolios, and the ratio of nonperforming loans to total loans for private firms (column 3) and entrepreneurs (column 4), calculated as a deviation from the median.
The table lists coefficients and standard errors (in parentheses) from regressions estimated with bank fixed effects with robust standard errors clustered on banks. All
independent variables except for ownership dummy are lagged by one quarter. Regressions include seasonal dummies. Definitions of variables are provided in Table 1.
*
Significance at 10% level.
**
Significance at 5% level.
***
Significance at 1% level.

These regressions are estimated relying on fixed effect estima-
tion which becomes possible because all explanatory variables
are time-variant. Results presented in Table 8 show that the entry
of greenfield foreign banks encourages domestic banks to increase
the share of entrepreneurs in their portfolios. Moreover, their port-
folios become riskier for both transparent and opaque borrowers.
Such results suggest that greenfield banks are able to attract safest
borrowers, forcing domestic banks to refocus their lending to risk-
ier firms, which were probably turned down by foreign lenders
who possess superior screening techniques (Claeys and Hainz,
2007). This is consistent with a segregation effect of Dell’Ariccia
and Marquez (2004).
The effect of foreign bank entry via acquisition of domestic
banks has a distinctly different effect. First, they induce the
remaining domestic banks to lend less to both large firms and
entrepreneurs. Having shown earlier that takeover banks them-
selves devote a smaller share their portfolios to these borrowers,
we can conclude that we observe a certain convergence between
these banks. At the same time, the entry of foreign takeover banks
decreases the riskiness of domestic banks’ portfolios, which can be
interpreted as positive spillovers with respect to lending tech-
niques, loan monitoring or risk management. The fact that the ob-
served improvement concerns only loans to private firms suggests
that domestic banks lending to entrepreneurs might rely mostly on
relationship lending.
6. Conclusions
Using a unique dataset with detailed information on banks’
portfolios, we explore how foreign bank entry determines credit
allocation and lending rates in emerging markets. In particular,
using a dataset from Poland, we investigate the impact of the mode
of foreign entry—greenfield and takeover—on banks’ lending rates
controlling for the loan allocation to borrowers with different de-
grees of informational transparency, as well as by maturities and
currencies. Earlier studies have documented lower lending rates
and spreads of foreign entrants, which might stem from their supe-
rior efficiency (‘‘performance hypothesis’’), but also reflect borrower
informational capture (‘‘portfolio composition hypothesis’’). Our re-
sults are in line with the theoretical models underpinning the port-
folio composition hypothesis, showing that informational capture
determines bank credit allocation and lending rates (see e.g.,
Dell’Ariccia and Marquez, 2004; Sengupta, 2007).
Our main results can be summarized as follows. First, we show
that the mode of entry is a very important determinant of foreign
banks’ portfolio composition. Greenfield banks devote a smaller
share of their loan portfolio to opaque firms than domestic banks,
whereas this effect is not found for foreign banks that have ac-
quired existing domestic banks. We further find that over time
greenfield banks shifted towards more opaque borrowers. These
results are consistent with theories arguing that greenfield banks
have comparative disadvantages in lending to opaque borrowers
using soft information initially, but over time they also become pri-
vately informed and start servicing small enterprises.
Second, greenfield banks extend more loans in foreign currency,
even though this effect disappears after controlling for their better
access to foreign currency funding in international capital markets.
Moreover, greenfield banks extend less loans at longer maturities,
which may reflect their short-term commitment to host econo-
mies. Furthermore, we provide some evidence of convergence be-
tween greenfield and private banks in terms of currency
composition but not in terms of loan maturity. There is no robust
evidence that foreign banks that have entered via acquisitions pre-
fer to lend short-term or in foreign currency.
Having controlled for these portfolio composition effects, lend-
ing rates charged by foreign entrants – greenfield and takeover – to
 H. Degryse et al. / Journal of Banking & Finance 36 (2012) 2949–2959
different groups of borrowers (transparent and opaque) do not dif-
fer compared to those of other banks. We argue that the result
established in the existing literature indicating that the average
lending rate of greenfield banks is lower than that of domestic pri-
vate banks has to be attributed to a different portfolio composition:
greenfield banks offer less loans to opaque borrowers that exhibit a
higher cost of credit.
Finally, a higher participation of greenfield banks has encour-
aged domestic banks to shift the allocation of their portfolios to
more opaque borrowers with negative impacts on riskiness of their
loan portfolios, attesting to the segregation effect. Interestingly, the
entry of takeover banks has led to an improvement in the loan
quality of domestic banks’ loans to transparent firms. This can be
interpreted a spillover effects as domestic institutions learn to
use transaction-based lending techniques.
Acknowledgements
This paper’s findings, interpretations, and conclusions are en-
tirely those of the authors and do not necessarily represent the
views of the National Bank of Poland, the International Monetary
Fund, its Executive Directors, or the countries they represent. Part
of this paper circulated previously as ‘‘The effect of foreign bank
entry on the cost of credit in transition economies: which borrow-
ers benefit most?’’. The authors would like to thank anonymous
referees, Jochen R. Andritzky, Emanuel Barnea, Thorsten Beck, Mar-
tin Brown, Stijn Claessens, Paul De Grauwe, Adam Glogowski, Mar-
ta Golajewska, Christa Hainz, Kaoru Hosono, Vasso Ioannidou,
Matthias Koehler, Alina Luca, Daniel E. Nolle, Steven Ongena, Bruno
Parigi, Damiaan Persyn, José-Luis Peydró, Viorel Roscovan, Rajdeep
Sengupta, Dobromil Serwa, Greg Udell, Patrick Van Cayseele, Rudi
Vander Vennet, Delia Velculescu, Paul Wachtel, as well as partici-
pants at the 2010 Conference on ‘‘Banking in the light of the Global
Crisis’’ (Akko, Israel), the 2010 CESIfo Applied Micro conference,
the Carefin Conference on ‘‘Business Models in Banking: Is There
a Best Practice?’’, the 9th CAED conference, the XVI Tor Vergata
Conference on Banking and Finance, FIRS 2009 meetings, 2008
FMA European and International Conferences, 6th INFINITI Confer-
ence on Banking and Finance, and seminars at the IMF, the National
Bank of Poland, CEPII, VU Amsterdam, Universitat Pompeu Fabra,
University of Mannheim, and Tilburg University for many helpful
comments and suggestions. The paper was completed while Olena
Havrylchyk and Emilia Jurzyk were visiting the National Bank of
Poland, whose hospitality is gratefully appreciated. Emilia Jurzyk
gratefully acknowledges financial support from the Research Coun-
cil of the KU Leuven, in the framework of Central and Eastern Euro-
pean Initiatives. The remaining errors are sole responsibility of the
authors.
References
Althammer, W., Haselmann, R., 2011. Explaining foreign bank entrance in emerging
markets. Journal of Comparative Economics 39 (4), 486–498.
Barth, J., Caprio, G., Nolle, D., 2004. Comparative International Characteristics of
Banking. Office of the Comptroller of the Currency Economic and Policy Analysis
Working Paper 2004-1.
Beer, C., Ongena, S., Peter, M., 2010. Borrowing in foreign currency: Austrian
households as carry traders. Journal of Banking and Finance 34 (9), 2198–2211.
Berger, A., Udell, G., 2006. A more complete conceptual framework for SME finance.
Journal of Banking and Finance 30 (11), 2945–2966.
Berger, A., Klapper, L., Udell, G., 2001. The ability of banks to lend to informationally
opaque small businesses. Journal of Banking and Finance 25 (12), 2127–2167.
Berger, A., DeYoung, R., Genay, H., Udell, G., 2000. Globalization of financial
institutions: evidence from cross-border banking performance. Brookings-
Wharton Papers on Financial Services 3, 23–120.
Berger, A., Miller, N., Petersen, M., Rajan, R., Stein, J., 2005. Does function follow
organizational form? Evidence from the lending practices of large and small
banks. Journal of Financial Economics 76, 237–269.
2959
Berger, A., Saunders, A., Scalise, J., Udell, G., 1998. The effects of bank mergers and
acquisitions on small business lending. Journal of Financial Economics 50, 187–
229.
Bonin, J., Hasan, I., Wachtel, P., 2005. Bank performance, efficiency and ownership in
transition countries. Journal of Banking and Finance 29 (1), 31–53.
Brown, M., Ongena, S., Yesin, P., 2008. Currency Denomination of Bank Loans:
Evidence from Small Firms in Transition Countries. Center for Economic
Research Discussion Paper No. 16, Tilburg University.
Chen, S., Liao, Ch., 2011. Are foreign banks more profitable than domestic banks?
Home- and host-country effects of banking market structure, governance, and
supervision. Journal of Banking and Finance 35 (4), 819–839.
Claeys, S., Hainz, C., 2007. Acquisition versus Greenfield: The Impact of the Mode of
Foreign Bank Entry on Information and Bank Lending Rates. Sveriges Riksbank
Working Paper Series No. 210.
Clarke, G., Cull, R., Martinez Peria, M., 2006. Foreign bank participation and access to
credit across firms in developing countries. Journal of Comparative Economics
34 (4), 774–795.
Clarke, G., Cull, R., Martinez Peria, M., Sanchez, S., 2005. Bank lending to small
businesses in Latin America: does bank origin matter? Journal of Money, Credit
and Banking 37 (1), 83–118.
De Haas, R., Naaborg, I., 2006. Foreign banks in transition countries: to whom do
they lend and how are they financed? Financial Markets, Institutions &
Instruments 15 (4), 159–199.
de la Torre, A., Martinez Pera, M., Schmukler, S., 2010. Bank involvement with SMEs:
beyond relationship lending. Journal of Banking and Finance 34 (9), 2280–2293.
Dell’Ariccia, G., Marquez, R., 2004. Information and bank credit allocation. Journal of
Financial Economics 72, 185–214.
Detragiache, E., Tressel, T., Gupta, P., 2008. Foreign banks in poor countries: theory
and evidence. Journal of Finance 63 (5), 2123–2160.
Dooley, M., Shin, I., 2000. Private Inflows when Crises are Anticipated: A Case Study
of Korea. NBER Working Paper No. 7992.
European Central Bank, 2006. EU Banking Sector Stability. November 2006.
Farnoux, M., Lanteri, M., Schmidt, J., 2004. Foreign Direct Investment in the Polish
Financial Sector. Case Study Prepared for the CGFS Working Group on Financial
Sector FDI.
Giannetti, M., Ongena, S., 2009. Financial integration and firm performance:
evidence from foreign bank entry in emerging markets. Review of Finance 13,
181–223.
Giannetti, M., Ongena, S., 2012. ‘‘Lending by example’’: Direct and indirect effects of
foreign banks in emerging markets. Journal of International Economics 86 (1),
167–180.
Gormley, T., 2010. The impact of foreign bank entry in emerging markets: evidence
from India. Journal of Financial Intermediation 19 (1), 26–51.
Gormley, T., 2011. Costly Information, Foreign Entry, and Credit Access. American
Finance Association 2008 New Orleans Meetings Paper.
Gottesman, A., Roberts, G., 2004. Maturity and corporate loan pricing. The Financial
Review 39 (1), 55–77.
Hainz, C., 2005. Effects of bank insolvency on corporate incentives in transition
economies. Economics of Transition 13 (2), 261–286.
Havrylchyk, O., Jurzyk, E., 2011. Profitability of foreign banks in Central and Eastern
Europe: does the mode of entry matter? Economics of Transition 19 (3), 443–
472.
Hausman, J., Taylor, W., 1981. Panel data and unobservable individual effects.
Journal of Econometrics 16 (1), 155.
Helwege, J., Turner, C., 1999. The slope of the credit yield curve for speculative-
grade issuers. Journal of Finance 54, 1869–1884.
Koyama, T., Golub, S., 2006. OECD’s FDI Regulatory Restrictiveness Index: Revision
and Extension to More Economies. OECD Economics Department Working
Papers No. 525.
Martinez Peria, M., Mody, A., 2004. How foreign participation and market
concentration impact bank spreads: evidence from Latin America. Journal of
Money, Credit and Banking 36 (3), 511–537.
Mester, L., 1997. What’s the point of credit scoring? Federal Reserve Bank of
Philadelphia Business Review, September/October 1997, 3–16.
Mian, A., 2006. Distance constraints: the limits of foreign lending in poor
economies. Journal of Finance 61 (3), 1465–1505.
Petersen, M., Rajan, R., 2002. Does distance still matter? The information revolution
in small business lending. Journal of Finance 57, 2533–2570.
Pluemper, T., Troeger, V., 2007. Efficient estimation of time-invariant and rarely
changing variables in finite sample panel analyses with unit fixed effects.
Political Analysis 15 (2), 124–139.
Popov, A., Udell, G., 2010. Cross-Border Banking and the International Transmission
of Financial Distress. Working Paper No. 1203, European Central Bank.
Rodrik, D., Velasco, A., 2000. Short-term capital flows. In: Pleskovic, B., Stiglitz, J.
(Eds), Annual World Bank Conference on Development Economics 2000. World
Bank.
Sengupta, R., 2007. Foreign entry and bank competition. Journal of Financial
Economics 84, 502–528.
Sorsa, P., Bakker, B., Duenwald, C., Maechler, A., Tiffin, A., 2007. Vulnerabilities in
Emerging Southeastern Europe—How much cause for concern? Working Paper
No. 07/236, International Monetary Fund.
Stein, J., 2002. Information production and capital allocation: decentralized vs.
hierarchical firms. Journal of Finance 57, 1891–1921.
Weill, L., 2003. Banking efficiency in transition economies: the role of foreign
ownership. Economics of Transition 11 (3), 569–592.