ISSUES IN BANKING

Do Foreign Banks in India Indulge in

Cream Skimming?
Mandira Sarma, Anjali Prashad

Foreign banks in developing countries are often found

to indulge in cream skimming, a lending strategy that
targets only wealthy segments of the credit market and
excludes small and marginal borrowers from the general
pool of borrowers. This paper attempts to investigate
whether lending patterns of foreign banks in urban
regions of Indian states are indicative of such practices.
Using credit data on urban regions of 21 states of India
for 19992011, this paper finds empirical evidence of
cream skimming by foreign banks in India.

We thank the participants of the Workshop on Development Strategies:

Country Studies and International Comparisons, organised by HTW,
Berlin and East China Normal University in Shanghai, November 2012
and IGCISI Summer School (Delhi), July 2014, for helpful comments on
earlier drafts of this paper. We also thank two anonymous reviewers for
their insightful comments on previous versions of this paper. All errors
are ours.
Mandira Sarma (msarma.ms@gmail.com) teaches at the Centre for
International Trade and Development, School of International
Studies, Jawaharlal Nehru University, New Delhi. Anjali Prashad
(prasad.anjali3@gmail.com) is at the Development Planning Centre,
Institute of Economic Growth, University of Delhi.

120

1 Introducing Cream Skimming

iberalisation of entry norms for foreign banks is an important component of financial sector development
strategies in developing economies. As participation of
foreign banks in developing countries increases, their lending
strategies significantly affect credit availability in these economies (Berger et al 2001). The impact of foreign banks presence in host economies has been analysed from different perspectives. One perception is that foreign banks help in real
economic growth by mitigating credit constraints in developing host countries (Bruno and Hauswald 2009; Giannetti and
Ongena 2012). In this context, it is argued that lending patterns of foreign banks are more stable than domestic banks
because they have better access to international resources and
are supported by parent banks in times of financial distress
(Crdenas et al 2003). While in the context of developing
countries, many studies have found that presence of foreign
banks does not necessarily enhance overall credit availability
and may actually aggravate the conditions of credit constraints
rather than alleviating such constraints (Detriagache et al
2008; Beck and Peria 2010; Gormley 2010). This is because foreign banks often indulge in cream skimming, a lending strategy that involves extending credit to only wealthy and transparent segments of the credit market, while excluding segments that comprise poor and marginal borrowers (Berger
and Udell 1996).
Such cream skimming may induce serious distortion in the
credit market and may adversely affect the supply of aggregate
bank credit. Detragiache et al (2008) establish, both theoretically and empirically, a negative association between the presence of foreign banks and credit availability. In particular, they
found that in the context of developing countries, countries
with larger foreign bank presence have shallower credit markets (p 2125). Similarly, many studies have demonstrated that
foreign banks target best credit risks due to which their presence does not benefit opaque borrowers in host countries1
(see, for example, Mian 2006 for Pakistan; Beck and Peria
2010 for Mexico; Lin 2011 for China). Also, as foreign banks
cream skim the best credit risks, the quality of local borrowers
pool falls and domestic banks cut short their lending activities,
thus leading to decline in firms access to credit and fall in growth
of their business (Giannetti and Ongena 2007 for Eastern
European economies; Gormley 2010 for India).
Berger and Udell (1996), DellAriccia et al (1999), Berger
et al (2001), Stein (2002), Mian (2006) and Detragiache
MARCH 19, 2016

vol lI no 12

EPW

Economic & Political Weekly

ISSUES IN BANKING

et al (2008) provide explanations, based on asymmetry of

information, as to what drives foreign banks in developing
countries into cream skimming. As per this strand of literature,
foreign banks with their advanced monitoring and credit rating technologies tend to target those clients, who can provide
hard information, mainly the large and wealthy segments of
the credit market. The poor and small borrowers, who may not
be able to provide hard information, are excluded by foreign
banks as they face a comparative disadvantage in obtaining
soft information on these clients.
The consequences of cream skimming on the host countrys
banking system have been well documented. Lending market
segmentation is the immediate consequence of cream skimming. While foreign banks target the segments of the credit
market with minimal information asymmetry, the segments
with greater information asymmetry are left for domestic
lenders. Segmentation of this kind may create an imbalance in
the flow of bank credit across different borrower categories,
due to which a particular segment receives a steady flow of
bank credit in normal as well as in tight economic conditions
while others may be vulnerable to liquidity shortages. Also, if
such segmentation leaves domestic banks with a pool of highrisk borrowers, these banks may respond by restricting supply
of credit, leading to a fall in overall credit supply (Schmidt
2008; Detragiache et al 2008; Gormley 2010).
In light of the above literature, this paper attempts to empirically examine whether foreign banks lending behaviour in
India conforms to cream skimming. We attempt this by carrying out a state-level analysis of credit data on foreign and
domestic banks in urban regions of India. We cover only urban
regions because foreign banks have a negligible presence in
rural India. Unlike many studies that analyse firm-level data
on bank lending to investigate cream skimming behaviour, in
this paper we attempt to investigate the issue at an aggregate
level of urban regions of Indian states. The basic question that
we address is whether foreign banks in India target urban
regions in rich states; and if so, whether such targeting has led
to a negative impact on overall supply of credit in urban India.
The empirical methodology involves estimating a set of panel
regressions to investigate how average credit availability in
urban India is affected by the presence of foreign banks. The
panel regressions are carried out using relevant data for 21
Indian states during the years 19992011.
Our empirical results do seem to indicate evidence of cream
skimming by foreign banks in India. After accounting for per
capita income levels, foreign banks presence is found to be
negatively associated with overall credit availability of urban
regions of Indian states, while their presence in higher income
states is positively associated with credit availability. Further,
domestic banks contribution to total urban credit in states
where foreign banks are present is found to be negative. Thus,
while foreign banks presence seems to contribute negatively
towards overall urban credit availability, their presence in rich
states improves credit availability, indicating that they target
urban regions of rich states. The domestic banks seem to
squeeze credit supply in states where foreign banks are present.
Economic & Political Weekly

EPW

MARCH 19, 2016

vol lI no 12

These results substantiate an earlier study by Gormley (2010)

in the Indian context. Gormley (2010), using firm-level data for
the period 19912002, found that foreign banks benefited only
a small set of very profitable firms, while an average firms
likelihood of receiving credit dropped by 8 percentage points
upon entry of foreign banks in India. In our state-level study of
urban India, we find evidence of foreign banks benefiting only
the rich states while the overall impact of foreign banks on the
credit availability of all states being negative.
The remainder of the paper is as follows. Section 2 presents
some data on foreign banks lending pattern in India. Section 3
discusses the empirical model employed for investigation of
our research question. Results and implications are discussed
in Section 4, followed by conclusions in Section 5.
2 Foreign Banks in India: A Descriptive Analysis

As part of its phased liberalisation of the banking sector, the

Indian government agreed to the World Trade Organization
(WTO) (General Agreement on Trade in Services or GATS)
agreement in 1994 to issue five branch licences to existing and
new foreign banks every year. This limit was later increased to
12 branch licences a year with the signing of another WTO
agreement in 1997. Thereafter in 2005, the Reserve Bank of
India (RBI) released a Road Map for further liberalising the
entry of foreign banks in the Indian banking system. However,
this road map was put on hold, on account of the unprecedented
global financial crisis that began in 2007. In RBI (2011), the RBI
laid out a detailed framework for foreign banks entry in India.
In a more recent discussion paper, RBI has finalised a framework for wholly-owned subsidiary (WOS) mode of entry (RBI
2013). Thus, while all foreign banks in India currently have
Table 1: List of States Included in the Study
State/UTs

High-income states
1 Delhi
2 Chandigarh
3 Haryana
4 Puducherry
5 Maharashtra
6 Gujarat
Middle-income states
7 Tamil Nadu
8 Kerala
9 Punjab
10 Uttarakhand
11 Andhra Pradesh
12 Karnataka
13 West Bengal
Low-income states
14 Rajasthan
15 Chhattisgarh
16 Odisha
17 Jammu & Kashmir
18 Madhya Pradesh
19 Jharkhand
20 Uttar Pradesh
21 Bihar

Location (Region)

Per Capita Income

(Rs million, 2011)

Share in Total Income

Per Capita of 21
States, 2011

Northern
Northern
Northern
Southern
Western
Western

1,75,812
1,40,066
1,09,227
95,759
94,121
88,787

11.5
9.2
7.1
6.3
6.1
6.0

Southern
Southern
Northern
Central
Southern
Southern
Eastern

84,058
83,725
78,171
75,604
71,540
69,493
55,864

5.5
5.5
5.1
4.9
4.7
4.5
4.0

Northern
Central
Eastern
Northern
Central
Eastern
Central
Eastern

51,474
46,573
46,150
41,833
36,345
31,982
29,417
24,681

3.4
3.0
3.0
2.7
2.4
2.1
1.9
1.6

Source: For per capita incomeHandbook of Statistics on Indian Economy, RBI.

121

ISSUES IN BANKING

their presence in the branch mode, RBI has recently displayed

clear preference for the WOS mode.2
In this section, we present some analysis of data on foreign
banks participation in urban regions of Indian states for the
years 19992011. This analysis covers only urban and metropolitan regions (hereafter referred to as urban regions) of
Indian states.3 This is because foreign banks have negligible
presence in rural and semi-urban regions of India.4 We exclude the eight north-eastern states from our analysis, since
foreign banks face a stricter set of guidelines to set up presence
in these states than domestic banks. In addition, Himachal
Pradesh is also excluded from the study due to lack of availability of credit data for some years.
Thus, our analysis covers urban regions of 21 states listed in
Table 1 (p 121).5 In terms of per capita net state domestic product
(NSDP), we categorise them into three categories, namely, high-,
middle- and low-income states. In terms of economic structure,
the high-, middle-income states have a smaller share of informal (unorganised) sector than the all-India level, while the
share of informal (unorganised) sector in poor states is above
the all-India level.6 This distinction in the level of formalness in
the economic structure can be used as an indirect proxy for information asymmetry. Thus, while it is not possible to measure
information asymmetry at the regional level, we assume that
the urban regions of high-, middle-income states have less information asymmetry than in the low-income states.
The data used in this paper are from the following sources:
(i) Quarterly Statistics on Deposits and Credit of Scheduled
Commercial Banks published by RBI is used to collect data on
total bank credit, total bank deposits and number of bank
branches; (ii) data on state incomes are retrieved from the
Central Statistics Office database; and (iii) Basic Statistical
Returns of Scheduled Commercial Banks in India published
annually by RBI is used to assemble data on total number of
credit accounts per state in India.

banks was 10% in 1999. Thereafter, these shares declined to 4%

in 2008 and rose to 6% in 2012 (Figure 1). Thus, the share of
foreign banks in urban regions in different aspects of banking
activities is not very high and it is the domestic banks that are
much more dominant in the Indian banking system.
Foreign banks business activities in India are found to be
mostly concentrated in high-income states. As observed from
Figure 2, about 85% of total urban credit by foreign banks is
allocated to high-income states, about 14% to middle-income
states, and only 0.4% of their credit to low-income states.
According to Figure 3, 62% of foreign bank branches were
located in urban regions of high-income states, 30% in middleincome states, and 8% in low-income states. Thus, Figures 2
and 3 clearly indicate that foreign banks have chosen urban
regions of rich states to establish presence in India, and thus
have channelised a major proportion of their resources to
high-income segments. This observation provides tentative
evidence of foreign banks practising cream skimming in India.

Figure 1: Indicators of Foreign Banks' Participation (Urban Region Only,

21 States)

30

12

25

Credit share

Figure 2: State-wise Allocation of Foreign Banks' Credit with Regard to Per

Capita Income of States (Urban Region Only, 2011)
45

0.20
0.18

40

Share of credit (right axis)

0.16

35

0.14

30

0.12

25

Per capita NSDP (Rs mn, left axis)

0.10

20

0.08

Bihar

Jharkhand

Uttar Pradesh

Madhya Pradesh

Odisha

Jammu & Kashmir

Rajasthan

Chhattisgarh

Karnataka

West Bengal

Uttarakhand

Andhra Pradesh

Kerala

Punjab

Gujarat

Tamil Nadu

Maharashtra

0.00

Haryana

0.02

Puducherry

0.04

Delhi

0.06

10

Chandigarh

15

Figure 3: State-wise Distribution of Foreign Banks' Branches vis--vis

Per Capita Income of States (Urban Region Only, 2011)
Share of branches (right axis)

Asset share

10

20

15

Per capita NSDP (Rs mn, left axis)

10

6
Deposit share

Figure 1 presents the trend of indicators of foreign banks participation in urban regions of the 21 states during 19992012. As
seen from this figure, the share of foreign bank branches during
this period continues to be just about 1% of total bank branches
in urban regions in these states. Over this period of 13 years,
foreign banks share in total commercial bank assets remained
around 7%8% with a slight increase to 9% in 2008 and 2009.
Foreign banks share in total deposits and credit of commercial
122

Bihar

Uttar Pradesh

Jharkhand

Madhya Pradesh

Odisha

Jammu & Kashmir

Rajasthan

Chhattisgarh

Karnataka

West Bengal

Uttarakhand

Source: Quarterly Statistics on Deposits and Credit of Scheduled Commercial Banks (RBI).

Andhra Pradesh

Kerala

Punjab

Tamil Nadu

Gujarat

Maharashtra

Haryana

Puducherry

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

Delhi

Chandigarh

Share of branches

0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00

Source: Quarterly Statistics on Deposits and Credit of Scheduled Commercial Banks (RBI) and
Central Statistics Office.

3 Empirical Methodology

In this section, we turn to an empirical investigation of

whether foreign banks in India indulge in cream skimming
and if so, whether such lending behaviour has any impact on
the credit availability in the urban regions. The empirical
methodology employed here involves estimating a panel
MARCH 19, 2016

vol lI no 12

EPW

Economic & Political Weekly

ISSUES IN BANKING

regression equation using relevant data for urban regions of 21

Indian states during the years 19992011 (as described in the
previous section). Our dependent variable here is average
credit availability measured by credit to state domestic product. The basic form of the specific regression equation that we
estimate is as follows:
ln(C/SDP)i, t = i+1ln(PCI)i, t + 2(IFB)i, t
+ 3(PFB* ln PCI) i, t + 4 (PFB*s.DB)i, t + i, t
where,
C = total volume of urban bank credit, SDP = sate domestic
product and (C/GDP)i, t is ratio of total urban bank credit to
state domestic product for state i in time t.
Explanatory variables are discussed below:
ln(PCI)i, t: The logarithm of per capital income of state i at time
t, and captures the level of economic activity in the ith state and
thus indicates the demand for credit.
(IFB)i, t: It captures the intensity of foreign banks presence and
is measured by the ratio of foreign banks branches to total
bank branches in the urban regions of state i at time t.
(PFB* ln PCI)i, t: It is an interaction term, where, presence of
foreign bank (PFB) in a state is interacted with per capita income
(PCI) of the state. PFB here is a binary variable taking value 1 if
foreign banks have any urban office in the state in time t and
zero otherwise. This interaction term captures the differential
impact of the income level of a state on credit availability when
foreign banks are present (PFB=1) with relation to when foreign
banks are absent (PFB=0). Primarily, this interaction term
examines whether foreign banks presence in urban regions of
high-income/rich states leads to greater availability of credit
in these regions or vice versa. In other words, if foreign banks
target only rich urban regions then 3 should be positive.
(PFB*s.DB)i, t: This is another interaction term indicating interaction between presence of foreign banks (PFB) with domestic
banks credit share (s.DB). Again, presence of foreign banks
(PFB) in this interaction is a binary variable as defined above.
This term captures the differential impact of domestic bank
credit on average availability of urban credit when foreign
banks are present in relation to when foreign banks are absent.
Thus, if due to cream skimming of foreign banks, domestic
banks resort to cutting short on lending, then the coefficient 4
would have a negative sign.
This basic model is also estimated with time dummies to
control for time-specific effects. The time dummies account for
a host of unobserved demand-side factors not all of which can
be represented by per capital SDP alone. Some of these factors
that vary over time could be composition of industrial sector,
esinstitutional environment, changes in policy and so on.

Table 2 presents some summary statistics of the variables

discussed above for the sample at hand.
4 Results and Discussions

Table 3 presents the results of the econometric estimation of our

regression model. Column (1) presents the results of the basic
model without any time dummy, while column (2) presents
the results with a set of time dummies included as explanatory
variables in addition to those in the basic model.
Table 3: Regression Results
Explanatory Variables

Dep var: ln (Credit to SDP Ratio)

(1)
(2)

Constant

-6.092***
(0.00)
Ln(PCI)
0.407***
(0.00)
IFB
-25.58***
(0.00)
PFB x Ln(PCI)
0.267***
(0.00)
PFB x s.DB
-2.379***
(0.001)
Time dummies (a set of year dummies. Absent
The reference year is 1999)
No of observations
267
States/UTs
21
Hausman test
FIXED
F test
106.64
(0.000)
R squared
0.638

-0.259
(0.834)
-0.203
(0.111)
-2.972
(0.569)
0.230***
(0.00)
-2.172***
(0.00)
All highly significant
except 2000
263
21
FIXED
29.23
(0.000)
0.857

P-values within parentheses. *** =1% level of significance.

In column (1), per capita income is found to be positively

and significantly associated with overall availability of urban
credit. However, when we include time dummies (column
(2)), the significance of per capita income seems to diminish
in explaining credit availability. Thus, income levels do
impact overall credit availability, but its impact gets diluted
as we allow time-specific factors clubbed in times dummies in
our regression.
After accounting for income level, our first variable of interest,
namely, intensity of foreign banks presence, IFB, is found to be
negatively significant in the absence of time dummies and not
significant when time dummies are included as explanatory
variables. Thus foreign banks presence, in urban India, seems
to have a negative association with overall credit availability
although this impact does not show up significantly when
time dummies are included. Detragiache et al (2008), in their
cross-country study involving 89 developing countries, found
a negative association between foreign banks presence
(measured by asset share of foreign banks) and credit-togross domestic product (GDP) ratio, indicating an inverse relationship between penetration of foreign banks and credit

Table 2 : Summary Statistics

Variable

Obs

Mean

Standard Deviation

Min

Max

ln (credit to SDP ratio)

267

-1.668

0.962

-3.367

1.019

IFBIntensity of foreign banks presence (share of foreign bank branches in total)

267

0.007

0.007

0.000

0.025

ln (PCI)logarithm of per capita SDP

273

10.318

0.683

8.663

12.077

PFBbinary variable denoting presence (=1) or absence (=0) of foreign bank

267

0.776

0.418

(PFB LnPCI)foreign banks presence as a binary variable interacted with per capita SDP

267

8.118

4.415

12.077

(PFB s.DB)foreign banks presence as a binary variable interacted with domestic banks credit share 267

0.746

0.405

1.000

Economic & Political Weekly

EPW

MARCH 19, 2016

vol lI no 12

123

ISSUES IN BANKING

124

Figure 5a: Share of Credit Outstanding to Small Borrowers by Domestic and

Foreign Banks (Urban Regions Only)
120

Domestic banks

100
80
60
40

Foreign banks

20
2011

2010

2009

2008

2007

2006

2005

2004

0
2003

Source: Basic Statistical Returns of Scheduled Commercial Banks in India.

Figure 5b: Share of Credit Accounts of Small Borrowers in Domestic and

Foreign Banks (Urban Regions Only)
90
80
70
60
50
40
30
20
10
0

Domestic banks

2011

2010

2009

2008

2007

2006

2005

Foreign banks

2004

availability. Our result here seems to be in line with the finding

of Detragiache et al (2008).
The coefficient of the first interaction term (PFB*ln PCI) is
found to be significant and positive in both variants of our
regression equation. This coefficient can be interpreted as the
differential impact of per capita income on the credit availability
in the presence of the foreign banks in relation with its impact
in absence of foreign banks. The estimates of the coefficient of
this interaction term in both versions of the regression model
indicate that this differential impact is significant and positive.
In the other words, urban regions of rich states in India benefit
from foreign banks presence in terms of credit availability.
Thus, while foreign banks presence, in general, deteriorates
overall availability of credit for urban India (as given by the
negative sign of the second explanatory variable), their presence benefits urban regions of rich states. This supports the
premise that foreign banks capture the wealthy segments of
urban credit markets in India, indicating their cream skimming behaviour. The evidence of such cream skimming behaviour is also observed elsewhere, for example, in Pakistan,
Mian (2006) reported foreign banks lending to be biased towards large and hard information firms, and Beck and Peria
(2010) found that only rich and urban municipalities benefited
from foreign banks in Mexico.
The second interaction term of our regressions attempts to
capture how domestic banks react in the presence of foreign
banks. This interaction term measures the differential impact
of domestic banks credit supply on overall credit when
foreign banks are present as opposed to when foreign banks
are not present. The coefficient for this variable is found to be
negative and significant in both the specifications of our
model. This indicates that when foreign banks are present in
an urban region, the domestic banks tend to lend less on an
average as compared to when foreign banks are not present.

2003

As on March 2011, India was host to 41 foreign banks affiliating from 16 high-income
countries, five middle-income countries and three low-income countries. Foreign banks
from high-income source countries held 99% share in total foreign banks assets of India
and foreign banks from middle- and low-income contributed to only 1% share in total
foreign banks assets of India in 2012. Source countries are categorised as high-, middleand low-income according to World Banks classification of economies July 2012.
High-income countries are the United Kingdom, the US, Hong Kong, Singapore, Germany,
France, Japan, Canada, Switzerland, Australia, Korea Republic, UAE, Bahrain, Belgium,
the Netherlands, and Oman. Middle-income countries are Mauritius, South Africa, China,
Russia and Thailand. Low-income countries are Sri Lanka, Bangladesh, and Indonesia.
Source: RBI, Statistical Tables Related to Banks of India (Table B1).

2002

The Netherlands

Russia

Indonesia

Thailand

Bangladesh

Australia

China

Sri Lanka

Oman

Bahrain

South Africa

UAE

Mauritius

Switzerland

South Korea

Japan

Canada

France

Singapore

US

Germany

UK

Belgium

2002

15
10

2001

20

2001

25

2000

35
30

1999

45
40

This may seem somewhat contradictory when we look at the

trend of foreign banks credit share as depicted in Figure 1.
Figure 1 indicates a declining trend in the share of foreign
banks and hence a rise in the corresponding share of domestic
banks in aggregate urban credit. Thus, while domestic banks
share in total urban credit in India is rising in general (the
aggregate picture), the econometric analysis using state-level
data indicates that on an average, the differential impact
of domestic banks in urban credit supply in the presence of
foreign banks is negative as opposed to that in their absence
(micro picture). This indicates that perhaps the domestic
banks squeeze credit supply in the presence of foreign banks.
Thus, after accounting for income levels, we find that
foreign banks presence enhances urban credit availability of
richer states, although the overall impact of foreign banks
presence on aggregate urban credit is negative or insignificant.
Further, in the presence of foreign banks, domestic banks tend
to cut short on their lending, leading to an adverse impact
on overall credit availability of urban regions of India. These
results do seem to indicate cream skimming behaviour of
foreign banks in India.
While it is beyond the scope of this paper to analyse the
factors that might be driving foreign banks in India to indulge
in cream skimming, we attempt a tentative explanation as
follows. Most foreign banks in India belong to high-income
Organisation for Economic Co-operation and Development
(OECD) countries. As depicted in Figure 4, out of the 41 foreign
banks (as of March 2012), 30 banks belonged to 16 highincome countries, seven banks are from five middle-income
countries and four banks from three low-income countries.7
Foreign banks from high-income source countries together
held 99% share in total foreign banks assets of India while

2000

50

1999

Figure 4: Share of Foreign Banks Assets by Source Country, 2011 (March)

Source: Basic Statistical Returns of Scheduled Commercial Banks in India.

MARCH 19, 2016

vol lI no 12

EPW

Economic & Political Weekly

ISSUES IN BANKING

foreign banks from middle- and low-income countries contributed to only 1% share in total foreign banks assets in India
as of 2012. As banks from high-income countries are more
specialised in hard information-based lending, we therefore
expect much of the foreign banks to target urban regions of
richer Indian states. As mentioned earlier, high-income states
of India are also characterised by smaller share of unorganised
sector economy than the low-income states, and hence we
expect the credit markets of the urban regions of the highincome states to have less information asymmetry than in the
low-income states.
A look at the data on small borrowers in urban India reveals
that foreign banks do not cater to the credit needs of small borrowers.8 The data on share of credit outstanding and share of
credit accounts of small borrowers of urban India for the years
19992011 are depicted in Figures 5a and 5b (p 124) respectively. As shown by these figures, foreign banks share in credit
notes
1

2
3

See, for example, Mian (2006) for Pakistan;

Beck and Peria (2010) for Mexico; Lin (2011) for
China.
For details, see RBI (2011) and RBI (2013).
The Reserve Bank of India classifies regions of
India as rural, semi-urban, urban and metropolitan regions, based on the criterion of population size of the region. In this study, urban
and metropolitan centres are clubbed as urban
region.
Foreign banks comprise only about 0.03% of
total bank branches in rural and semi-urban
areas during 19992012. The low presence of
foreign banks in rural and semi-urban India is
due to the branch authorisation guidelines of
RBI, 201011, according to which foreign banks
require special permission to set up branches
in rural and semi-urban centres, while for urban
and metropolitan centres, they are treated on
par with the domestic banks. See Master Circular DBOD No BL BC 8/22 01 001/201011 dated
1 July 1 2010, RBI, Mumbai, for details on Indian
branch authorisation policy.
Although some of the listed states are actually
union territories (UTs), we refer to them as
states in this paper.
See, for example, Table 2 of Ghani et al (2013)
that reports share of unorganised sector employment (manufacturing as well as services
sector) for various states of India for different
years. The high-/middle-income states have a
lower share of unorganised sector employment
than all-India level while the share of unorganised employment in low-income states are
much above the all-India level.
Income classification of World Bank (2012) is
used to categorise source countries into high-,
middle- and low-income.
According to RBIs definition, small borrowers
are those with a credit limit up to Rs 0.2 million.

References
Beck, Thorsten and Maria S M Peria (2010): Foreign
Bank Participation and Outreach: Evidence
from Mexico, Journal of Financial Intermediation, Vol 19, pp 5273.
Berger, Allen N and Gregory F Udell (1996):
Universal Banking and the Future of Small
Business Lending, Financial System Design: The
Case for Universal Banking, Anthony Saunders
and Ingo Walter (eds), Burr Ridge, IL: Irwin
Publishing.
Economic & Political Weekly

EPW

MARCH 19, 2016

to the small borrowers is negligible, owing perhaps to the inability of these borrowers to provide hard information.
5 Conclusions

In this paper, using data on urban credit in 21 states of India for

the period 19992011, we investigated the impact of foreign
banks presence on availability of bank credit in India. Our empirical results indicate evidence of cream skimming behaviour
by foreign banks, despite their limited presence in India. While
foreign banks presence is found to contribute negatively towards
overall credit availability of urban regions, their presence in rich
states is found to improve urban credit availability, indicating that
they target urban regions of rich states. The domestic banks seem
to respond to this cream skimming by lending less in regions
where the foreign banks are present. These findings provide
interesting insights into foreign banks lending behaviour in
India in particular and in developing economies in general.

Berger, Allen N, Leora F Klapper and Gregory F

Udell (2001): The Ability of Banks to Lend to
Informationally Opaque Small Businesses,
Journal of Banking and Finance, Vol 25,
pp 212767.
Bruno, Valentina and Robert Hauswald (2009):
The Real Effect of Foreign Banks, Review of
Finance, Vol 18, pp 16831716.
Crdenas, Juan, Juan Pablo Graf and Pascual
ODogherty (2003): Foreign Bank Entry in
Emerging Market Economies: A Host Country
Perspective, CGFS Working Group on FDI in
the Financial Sector.
DellAriccia, Giovanni, Ezra Friedman and Robert
Marquez (1999): Adverse Selection as a Barrier
to Entry in the Banking Industry, RAND Journal
of Economics, Vol 30, No 3, 51534.
Detragiache, Enrica and Poonam Gupta (2006):
Foreign Banks in Emerging Market Crises:
Evidence from Malaysia, Journal of Financial
Stability, Vol 2, No 3, 21742.
Detragiache, Enrica, Thierry Tressel and Poonam
Gupta (2008): Foreign Banks in Poor Countries: Theory and Evidence, Journal of Finance,
Vol 33, No 5, pp 212359.
Ghani, Ejaz, William R Kerr and Stephen D
OConnell (2013): The Exceptional Persistence
of Indias Unorganized Sector, World Bank
Policy Research Working Paper No 6454.
Giannetti, Mariassunta and Steven Ongena (2007):
Financial Integration and Firm Performance:
Evidence from Foreign Bank Entry in Emerging
Markets, Review of Finance, Vol 13, pp 18123.

Giannetti, Mariassunta and Steven Ongena (2012):

Lending by Example: Direct and Indirect
Effects of Foreign Banks in Emerging Markets,
Journal of International Economics, Vol 86,
No 1, pp 16780.
Gormley, Todd A (2010): The Impact of Foreign
Bank Entry in Emerging Markets: Evidence
from India, Journal of Financial Intermediation, Vol 19, No 1, pp 2151.
Lin, Huidan (2011): Foreign Bank Entry and Firms
Access to Bank Credit: Evidence from China,
Journal of Banking & Finance, Vol 35, No 4,
pp100010.
Mian, Atif (2006): Distance Constraints: The Limit
of Foreign Lending in Poor Economies, Journal
of Finance, Vol 61, No 3, pp 14651505.
RBI (2011): Presence of Foreign Banks in India, Reserve Bank of India Discussion Paper, January.
(2013): Scheme of Setting up Wholly Owned
Subsidiaries (WSO) by Foreign Banks in India,
Reserve Bank of India, November.
Schmidt, Nikolaj (2008): Foreign Bank Entry: A
Liquidity Based Theory of Entry and Credit
Market Segmentation, Financial Markets
Group, London School of Economics and Political Science, Discussion Paper No 622, London,
UK.
Stein, Jeremy C (2002): Information Production
and Capital Allocation: Decentralized vs
Hierarchical Firms, Journal of Finance, Vol 7,
891921.
World Bank (2012): World Bank List of Economies,
World Bank, July.

Complete Annual EPW Sets Available

At Nominal Rates

L
Pr ow
ice
s

EPW has a few complete sets of the journal for 2004 and 2011 that are available at
nominal rates.
The entire set for each year is available for just Rs 100 plus postage and packing charges.
(The cost of postage for each setweighing around 10 kgby registered parcel will be
around Rs 400 to Rs 500. Packing charges will be Rs 100).
The total payable amount is Rs 700. Interested buyers can also call and visit our office in
Mumbai and collect the volumes by paying just Rs 100 each.
There are only a limited number of these unbound sets available. Institutions and
individuals interested in buying any of the sets can call the Circulation Department for
further details. Phone: 022-40638282
vol lI no 12

125