1 Foreign Entry and Banking Efficiency in Asia Bayu Kariastanto, GRIPS Wako Watanabe, Keio University May 24, 2012 International Conference, International.

1

Foreign Entry and Banking Efficiency in Asia

Bayu Kariastanto, GRIPSWako Watanabe, Keio University

May 24, 2012International Conference, “International Competition in

Banking: Theory and Practice”Ukrainian Academy of Banking, Sumy, Ukraine

Outline of the Presentation

Introduction to foreign banks in Asia Background theories and hypotheses Data Empirical methodology Results and interpretations

2

Overview of the Results

In Asia,Foreign banks are more cost efficient than private domestic banks but not more profitable.State owned banks are less profitable than private banks.Mode of entry (greenfield/JV/acquiring) does not affect profitability but greenfield banks are less cost efficient than local bank acquirers.

3

44

The Presence of Foreign Banks by RegionThe share of assets held by foreign banks within bank assets by

region （ % ）

Note: the number of countries inside the blaket

Source: Domanski (2005)

1995 2005 World (105) 15 23 North America (2) 10 21 Western Europe (19) 23 29 Eastern Europe (17) 25 58 Latin America (14) 18 38 Africa (25) 8 8 Middle East (9) 14 17 Central Asia (4) 2 2 East Asia and Oceania (13) 5 6

5

Foreign banks vs. domestic banks Foreign banks bring new lending technology and their

governance is more disciplined. Private banks vs. state owned banks State owned banks engage in political (“connected”)

lending. Minority foreign/state shares Minority shares may influence bank performance as

majority shares do?

Ownership Type and Bank Performance

6

Foreign banks vs. domestic banks Foreign banks more profitable in developing/transitional

countries (Bonin et al., 2004; Claessent, et al., 2000; Micco et al., 2004)

Private banks vs. state owned banks State owned banks as efficient as private banks (Bonin et

al., 2004) Minority shares (foreign/state) Minority foreign ownership improves state-owened

banks’ efficiency (Berger et al., 2008)

Ownership Type and Bank PerformanceEvidence

7

Mode of Entry and Bank Performance

Greenfield Open branches only in regional major cities Specialized on wholesale lending Utilize “hard” information ( = private & verifiable) Local bank Acquirers Acquire a local branch network. Utilize “soft information” ( = unverifiable) Joint Ventures Hybrids of greenfield and retail banks

8

De Novo Banks, Retail Banks or Joint Ventures

Theoretically, it is hard to predict which mode of foreign entry is more profitable/efficient.

Empirically… Lending rates are higher when foreign lenders are

greenfield than they are local bank acquirers (Claeys & Hainz, 2007)

99

Soft Information and Relationship Lending

Relationship lender

(a loan officer)SMEs

Other banks

Relationship

Soft (qualitative) information

Hard (quantitative) informationAsymmetri

c informatio

n

1010

Backgrounds: SME Lending and Soft Information

Local domestic lenders may be more profitable. A relationship lender who has exclusive access to certain

information about a borrower’s credit quality is able to earn a rent so that relationship lenders are more profitable than hard information based transaction lenders. (ex. Rajan, 1992)

Local lenders are able to select better credit risks whereas entrants are forced to lend to poor credit risks that local lenders choose not to lend to (Dell'Ariccia et al, 1999)

But more technologically advanced foreign lenders may be more cost efficient and more profitable.

1111

Soft Information and A Bank’s Organizational Structure

Loan review function of the headquarters

Branch manager

Loan officerSMEsRelationship

Soft info

Hard info

Soft info

Hard info

Hard

info

Documenting soft information is hard so it is difficult to deliver soft information from a bank’s branch to its headquarters.

The more delegated a bank’s branch is, the more incentive to gather soft information the branch has (because soft information is utilized for decision making).

1212

Organizational Structure and Bank Lending

Hypothesis

Loan officers at a bank with a decentralized decision making structure has more incentive to collect soft information than those at a bank with a centralized structure.

Evidence: Liberti & Mian (2009), Agarwal & Hauswald (2010)

1313

Operation Type and Foreign Bank Performance

Our question

Is a presumably more independent foreign subsidiary more efficient/profitable than a branch?

Evidence? Very few papers look at foreign banks’ operation types. A branch/subsidiary choice is made based on institutional

environments (Cerutti et al., 2007).

Data

The primary source of our data is Bankscope (Fitch IBCA & Bureau van Dijk) as well as banks’ homepages

Our data are the unbalanced panel of 980 banks from 33 countries in Asia (East, Southeast, South and Central Asia/Caucasus.) over the period of 1996 through 2011.

14

Trends of Ownership Types in Asia, Shares in terms of Total Assets

15

Descriptive Statistics: Ownership Types and Operation Types

Sample banks are divided into four ownership types, majority STATE (23%), majority private DOMESTIC (40%), majority FOREIGN (30%) and “no majority” (7%).

State, domestic, foreign and “no majority” represent 48%, 27%, 11% and 14% of total bank assets in Asia, respectively.

Only 6% of foreign banks operate as branches. This may be due to underepresentation of branches

in Bankscope.

16

Empirical Methodology: SFA Approach

The two step estimation approach is employed. At the first step, running the regression of the profit/cost function, each sample bank’s efficiency is estimated. At the second step, the estimated efficiency (level/rank) is regressed on variables to measure a bank’s attributes such as ownership types as well as control variables. The profit efficiency is more complete concept than the cost efficiency (Berger & Mester, 1997)

17

The Cost/Profit Function

The profit/cost function follows the specification employed by Bonin et al. (2004). Assumed to be a function of four outputs (total loans, total deposits, liquid assets, other earning assets) and two input prices (ratios of interest expenses to total deposits and non-interest expenses to fixed assets). Year and country dummies are included.

18

Ownership Type RegressionsProfit Efficiency

19

VariablesEfficiency

LevelRank

-0.021*** -0.072***(0.006) (0.024)

-0.043*** -0.149***(0.008) (0.026)

-0.015* -0.056(0.009) (0.036)

0.003 0.014(0.008) (0.035)

-0.010 -0.042(0.007) (0.027)

N 3521 3521Clusters 536 536

R-square 0 0

Minority state

Minorityforeign

Majorityforeign

Majority state

No majority

Ownership Type RegressionsCost Efficiency

20

VariablesEfficiency

LevelRank

0.058*** 0.058***(0.016) (0.016)

-0.023 -0.023(0.015) (0.015)

0.027 0.027(0.021) (0.021)

-0.001 -0.001(0.018) (0.018)

-0.014 -0.014(0.013) (0.013)

N 3521 3521Clusters 536 536

R-square 0 0

Majorityforeign

Majority state

No majority

Minority state

Minorityforeign

The Ownership Type Regressions

Banks are profit efficient in the following order,

Domestic > Foreign > State Domestic banks have better access to the soft information

(Mian, 2003; Claeys and Heinz, 2003) State-owned banks are often engaged in policy oriented

loans or politically motivated loans (Sapienza, 2004; Dobson and Kashyap, 2006; Berger, et al, 2008)

Foreign banks are more cost efficient than domestic banks.

Foreign banks are more technologically advanced.

21

Foreign Bank RegressionsProfit Efficiency

22

VariablesEfficiency

LevelRank

-0.013 -0.062(0.011) (0.043)

0.007 0.035(0.018) (0.075)

0.025 0.096*(0.016) (0.053)

-0.033 -0.151*(0.021) (0.093)

-0.160*** -0.374***(0.009) (0.035)

0.021 0.078(0.017) (0.072)

N 979 979Clusters 165 165

R-square 0 0

Greenfield

Joint venture

Acquisition,Minority stateGreenfield,Minority stateJoint venture,Minority state

Branch

Foreign Bank RegressionsCost Efficiency

23

VariablesEfficiency

LevelRank

-0.085*** -0.157***(0.029) (0.046)

-0.069 -0.107(0.050) (0.083)

0.078** 0.173**(0.038) (0.068)

-0.086 -0.117(0.054) (0.100)

-0.016 -0.007(0.023) (0.043)

-0.024 -0.134(0.048) (0.082)

N 979 979Clusters 165 165

R-square 0 0

Joint venture,Minority state

Greenfield

Joint venture

Branch

Acquisition,Minority stateGreenfield,Minority state

Foreign Bank Regressions

Mode of entry does not affect profit efficiency. Minority state ownership decreases greenfield

banks’ profit efficiency. The ex-post government’s stake in a bank established

as a greenfield bank reduces the bank’s profitability. Greenfield banks are less cost efficient than local bank

acquirers. Branches are generally more efficient than subsidiaries. Only successful branches may be in the data…

24

THANK YOU!!

25