Asia-Pacific Journal of Financial Studies (2013) 42, 5675

doi:10.1111/ajfs.12005

Testing for Competition in the South

Korean and Chinese Commercial Banking
Markets
Kang H. Park*
Harrison College of Business, Southeast Missouri State University
Received 11 October 2012; Accepted 10 December 2012

Abstract
This paper examines market concentration and competition in the South Korean and Chinese
commercial banking markets for the period of 19922008. This study empirically investigates
whether changes in bank concentration have affected the degree of competition in the Korean
and Chinese commercial banking industries by estimating the H statistic of the Panzar-Rosse
model. We also used the Boone Indicator model to confirm the results from the PanzarRosse model. The Korean banking industry has been monopolistically competitive for the
entire sample period while the Wald test of the H statistic shows competition in the Korean
banking actually increased to the level of perfect competition during the crisis period temporarily. Compared to the banking industry of Korea and other countries, the Chinese banking
industry is still highly concentrated and its level of competition is closer to oligopoly.
Keywords Bank consolidation; Bank competition; Korean banks; Chinese banks
JEL Classification: G21, L10

1. Introduction
The last two decades witnessed a surge in firm mergers. Although firm mergers have
been an established phenomenon over a long period, those occurring over the past
20 years, the so-called the fifth merger wave have been the most remarkable. The
banking industry has not been excluded from the merger wave with banking industries all over the world experiencing a fundamental change in market structure
through rapid consolidation. Financial deregulation and financial globalization have
triggered fierce competition among banks and necessitated consolidation to reduce
risk through business diversification and to take advantage of scale economies.
In the United States, lifting geographical restrictions on acquisitions and
branching in the 1980s accelerated mergers and acquisitions of banks. The number
of United States banks fell from 14 404 in 1980 to about 12 000 in 1990 and then
*Corresponding author: Kang H. Park, Southeast Missouri State University, Cape Girardeau,
MO, USA 63701. Tel: 73-651-2942, Fax: 573-651-2947, email: khpark@semo.edu.
56

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

to about 7100 in 2008, an almost 50% decline. Although the number of banks in
the United States declined and their average asset and deposit sizes became large,
there was no noticeable change in the concentration of the United States local
banking markets measured by the Herfindahl-Hirschman Index (hereafter HHI), a
standard measure of market concentration. This is because much bank consolidation in the United States is typically characterized by market extension, that is,
acquisitions involving two banks in different geographical markets. The United
States banking industry is still much less concentrated than those of other countries.
Over the last 20 years the South Korean (hereafter Korean) banking system has
undergone many changes: financial deregulation, financial crisis, and restructuring.
In this process, market concentration decreased as the number of banks increased
due to financial deregulation prior to the Asian financial crisis of 1997. However,
after the crisis, the concentration ratio increased because of a decline in the number
of banks due to bank closures and the creation of mega banks through bank consolidation. The Korean experience of bank mergers is different from the United States
experience. Bank mergers in Korea were the result of horizontal mergers among
banks with overlapping geographical markets contrary to market extension of interstate banking in the United States. Thus consolidation of banks in Korea has raised
a concern over the possibility of a decrease in the degree of competition in the
Korean commercial banking market.
The Chinese banking system experienced a quite different path from that of the
Korean banking system. Although there have been many bank foreclosures, takeovers and mergers in recent years, the number of new banks entering the Chinese
banking market far exceeded the number exiting. Until 1978, there was one bank in
China, the Peoples Bank of China. With economic reform, the Chinese Government authorized four state-owned commercial banks between 1979 and 1984 with
limited competition among them. Since then the Chinese Government has allowed
many joint equity banks and private banks in order to mobilize needed financial
resources for economic development. Furthermore, it authorized several policy
banks and city banks in the 1990s as a measure of financial liberalization in preparation for entry to the World Trade Organization. All these have contributed to a
continuous decrease in market concentration in the Chinese banking industry.
The purpose of this paper is to investigate empirically whether changes in bank
concentration are associated with the degree of competition in the Korean and Chinese commercial banking markets by estimating the H statistic of the Panzar-Rosse
model. This paper is organized as follows. Section 2 describes the trend of market
concentration in the Korean banking industry and the Chinese banking industry for
the period of 19922008. Section 3 reviews the related literature on the methodology of measuring the level of competition in the banking sector. Section 4 discusses
the methodology used to test for the degree of competition in empirical analysis.
Section 5 describes the data and interprets the estimates of the model. Section 6
provides a summary and conclusion.
2013 Korean Securities Association

57

K. H. Park et al.

Table 1 Market concentration of Korean and Chinese banks

1. HHI is the Herfindahl-Hirschman index and CR3 is the concentration ratio measured by the market
share of the largest k banks. k = 3 for Korea and 4 for China because CR3 is used by the Korean Financial Supervisory Authority while there are four dominant banks in China. Total assets are used to calculate HHI and CR. Total assets include assets in both banking accounts and trust accounts. 2. The total
number of the Korean commercial banks including both nationwide and regional banks was 24 until
1994 and increased to 26 in 1997, but declined since then to 13 in 2006. Even though China has more
than 200 banks, only 15 major banks are included in the sample because data availability is limited.
However, these 15 banks account for more than 70% of the total bank assets.

Korea

China

Year

HHI

CR3

HHI

CR4

1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008

876
827
788
715
707
664
808
937
983
1441
1481
1407
1325
1286
1466
1498
1491

36.42
34.69
33.15
30.15
29.78
28.32
34.46
38.67
40.99
52.23
54.38
53.18
51.09
51.17
52.27
58.66
58.54

2743
2625
2589
2534
2453
2257
2194
2101
2054
2004
1967
1822
1789
1726
1686
1669
1642

94.23
93.25
93.11
91.49
90.38
86.47
83.14
81.93
78.72
75.58
71.75
67.44
65.62
62.71
61.32
60.74
60.67

2. Bank Concentration
Market concentration can be measured in a number of ways. The most widely used
index is HHI, which is applied by the United States Department of Justice in implementing its anti-trust policy. Another straightforward method is to calculate what
share of the industrys output or assets is owned by a few dominant firms. This top
k-firm concentration ratio (CRk) is used by some governments including the
Korean Government to determine the degree of anti-competition of a proposed
merger.1 Table 1 presents HHI and CRk of total assets for Korea and China. We
obtained HHI and CRk of three variables, that is, total assets, total loans and total
deposits, and we found that the correlation coefficients of the HHI or CRk among
1

The Fair Trade Commission in Korea regards a market with CR1 greater than 50% or CR3
greater than 70% as a highly concentrated market.
58

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

the three variables are higher than 0.99. So, only HHI and CRk of total assets are
shown in Table 1.
In the case of Korea, HHI decreased from 876 in 1992 to 664 in 1997, but gradually increased to 1491 in 2006. There was a decreasing trend of market concentration until the Asian financial crisis of 1997. This trend actually began as early as
1982 when financial deregulation induced entry of new banks and caused fierce
competition among the existing banks. Until the Korean Government introduced a
series of financial reforms in 1982, the number of national commercial banks was
limited to five while ten much smaller regional banks were allowed in order to
stimulate regional economic development. With financial liberalization, the number
of nationwide commercial banks increased from five to 16 in 1997, leading to a
decrease in market concentration. Financial liberalization also allowed regional
banks to open their branches outside of their own regions. Just before the crisis,
there was concern about overbanking in Korea with 17 nationwide banks and ten
regional banks. However, closures of insolvent banks, and mergers with blue-chip
banks after the crisis, resulted in a drastic increase in these indices. The change in
market concentration after the crisis reflects the structural reform in the banking
sector carried out by the Korean Government.2
While a few nationwide mega banks were established through mergers and
acquisitions and creation of a financial holding company, about one half of regional
banks survived with the number declining from ten to six. Now there are seven
nationwide banks and six regional banks. The Korean banking industry experienced
polarization in bank size, leading to greater market concentration. Today, the three
largest banks, Kookmin Bank, Woori Bank and Shinhan Bank, hold about 60% of
total industry assets as a result of the government policy, and some observers are
concerned that this policy may have a negative effect on competition in the Korean
banking industry. The post-crisis period is also characterized by increasing market
share by foreign banks and increasing foreign ownership share of domestic banks.3
2

The Korean government began a two-phase financial restructuring. In its first-phase restructuring from 1998 to 2000, three types of merger occurred. First, five insolvent banks were
merged into five sound banks in the form of purchase of assets and assumption of liabilities
(P&As). Second, involuntary mergers between three groups of relatively sound banks were
initiated by the Korean government. Third, there was one voluntary merger of two banks.
The second and third types of mergers were in the form of mergers and acquisitions
(M&As). The second-phase restructuring, which began in 2001, focused on restoring bank
profitability. This structural reform also introduced the financial holding company system
and allowed mergers among larger banks, resulting in a few super-size banks. The number of
banks was reduced and the average asset size of banks increased.
3
Three domestic banks are controlled by foreign shareholders as of the end of 2004, and the
asset share of banks under foreign ownership, including branches of foreign banks and
domestic banks under foreign ownership, increased from 8.5% in 1997 to 22.4% in 2004.
However, the increase in the asset share is due mainly to foreign ownership of domestic
banks, not branch expansion of foreign banks.
2013 Korean Securities Association

59

K. H. Park et al.

In summary, the Korean banking sector prior to the crisis could be regarded as a
non-concentrated market with HHI less than 1000 in total assets, total loans and
total deposits, according to the horizontal merger guideline of the United States
Department of Justice. After the crisis, particularly after the second-phase restructuring, the Korean commercial banking market became a moderately concentrated
market with HHI ranging between 1000 and 1800. Although not reported in the
table, HHI is in excess of 1800 in some specific sub-markets, such as loans to
households and deposits in foreign currency. HHI figures for Korea were higher
than those for OECD countries of a similar population size. For example, the HHIs
of Spanish banks and Italian banks were below 1000.
Contrary to the increasing trend of Korean banks market concentration, the
Chinese banking system has experienced continually decreasing market concentration from 2743 in 1994 to 1642 in 2008. This change is clearly attributable to a
change in Chinese Government policy on banking, which allowed establishment of
more banks and promoted competition among them. In spite of some mergers of
banks occurring in recent years, the number of new banks created far exceeded the
number of banks foreclosed and merged. Before Deng Xiao Pings 1978 reform,
China had a mono bank, the Peoples Bank of China, playing both roles of central
and commercial banking. With reform, four specialized state banks were split from
the Peoples Bank of China between 1979 and 1984, leaving the Peoples Bank of
China functioning as Chinas central bank.4 Even though restrictions of these specialized banks to do business in only their designated territories were removed in
1985, competition among them was very limited until the mid-1990s. There was a
boost to competition when the Chinese government authorized establishment of
three policy banks.5 Since 1986, 14 joint-equity banks have been established, where
shares are held by the government, cooperatives and the private sector.6 During the
mid 1990s, the central government allowed local governments to establish local (or
city) banks. Currently the total number of banks exceeds 200, excluding foreign
banks, and the number increases year by year. Four Chinese banks are included in
4

Four state commercial banks split from the Peoples Bank of China are the Bank of China
(BOC), the Peoples Construction Bank of China (PCBC), the Industrial and Commerce
Bank of China (ICBC) and the Agricultural Bank of China (ABC). These specialized banks
were to provide banking services to a designated sector of the economy,
5
Three policy banks are the China Development Bank (CDB), the Export Import Bank of
China (EIBC) and the Agricultural and Development Bank of China (ADBC). CDB is chartered to provide long-term lending to finance construction projects for infrastructure and pillar industries. EIBCis established to provide loans for exports and imports of capital goods.
ADBC is to provide agricultural lending.
6
These are the Bank of Communication, China Merchants Bank, Shenzhen Development
Bank, Guangdong Development Bank, Pudong Development Bank, China Everbright Bank,
China Minsheng Banking Corporation, Hua Xia Bank, Fuijin Industrial Bank, Hainan Development Bank, China Investment Bank, Yantai Housing Saving Bank and Bengbu Housing
Saving Bank.
60

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

the worlds top 10 in 2012, as the global financial crisis of 20072008 reduced the
market values of the United States and Western banks.
3. Survey of the Literature
This section briefly reviews the theoretical models and empirical findings on bank
competition. The structure-conduct-performance (SCP) model suggests that
increasing market concentration leads to less competitive conduct in terms of
higher prices and fewer output and results in higher profits at the expense of lower
consumer welfare. Although there is a theoretical basis for these linkages, other
equilibrium conditions can lead to different relationships between market concentration and conduct.7
Two empirical methods have been developed to remedy shortcomings of the
SCP model by testing the conduct directly, without regard to industry structure.
One method is the Bresnahan (1982, 1989) and Lau (1982) model (B-L model),
which estimates the markup of price over marginal cost as a measure of market
power. Thus, this method is also called the markup test. This model is based on
two structural equations, an inverse demand equation and a supply equation
derived from the first order condition of profit maximization.8 The other method is
the Panzar and Rosse (1982, 1987) model (P-R model). This model measures the
extent to which a change in a vector of input prices is reflected in gross revenue.
Thus, this method is also called the revenue test. If the market is perfectly competitive, then the change will be fully reflected in revenue. Shaffer (2004) contrasted
both methods in detail and discussed their advantages and disadvantages.
This study applies the P-R model to the data of Korean and Chinese banks.
Numerous studies apply the P-R model empirically, beginning with Shaffer (1982)
who found monopolistic competition behaviors in a sample of New York banks in
1979. Nathan and Neave (1989) rejected the hypothesis of monopoly power of
7

As long as there are no sunk costs and hit-and-run entry is possible, then market contestability can yield competitive pricing regardless of the number of firms (Baumol et al., 1982).
The efficient structure hypothesis advances that efficient banks obtain higher profitability and
greater market share because of their efficiency, which will lead to a more concentrated market. Therefore, the association between structure and performance might be spurious unless
efficiency is controlled in the model (Smirlock, 1985). Adverse borrower selection may result
in spurious empirical SCP linkages too (Shaffer, 2002).
8
The following studies apply the B-L model empirically. Shaffer (1989) rejected the collusive
conduct hypothesis with a sample of United States banks, and Shaffer (1993) found that the
Canadian banks were competitive for the period 19651989 even with a relatively concentrated market. Berg and Kim (1996) showed that Cournot behavior is rejected in the Norwegian banking system. Fuentes and Sartre (1998) found that bank consolidation in Spain did
not weaken the competition level. Gruben and McComb (2003) found for Mexican banks
before 1995 that marginal prices were set below marginal costs and conclude that the Mexican market is super-competitive.
2013 Korean Securities Association

61

K. H. Park et al.

Canadian banks. Country-specific empirical studies include Vesala (1995) for Finland, Molyneux et al. (1996) for Japan, Coccorese (1998) for Italy, Hondroyiannis
et al. (1999) for Greece, and Hempell (2002) for Germany. Molyneux et al. (1994)
and Bikker and Groeneveld (2000) found monopolistic competition in several European countries. On the other hand, De Bandt and Davis (2000) found monopolistic
competition for large banks and monopoly for small banks in Germany and France.
Bikker and Haaf (2002) found that the banking industries in 23 OECD countries
for the period 19981999 are generally characterized by monopolistic competition
with the exception of Australia and Greece. Gelos and Roldos (2002) compared
eight European and Latin American countries and found that the bank consolidation process in its early stage has not lowered competition. Uchida and Tsutsui
(2005), from long-term Japanese panel data from 1974 to 2000, concluded that
market competition improved during the 1970s and 1980s, but worsened since
1997. Lee and Nagano (2008) reported that market concentration that brings about
the bank mergers does not necessarily result in low competition.
Compared to empirical studies on Western banking competition, there are few
studies on Korean or Chinese banking competition. Two different results are
reported for Korea. Kim (2003) measured competition using the Bresnahan-Lau
method based on aggregate monthly data from 1996 to 2002 and found that the
pricing behavior of Korean banks during this period was consistent with perfect
competition and that they behaved more competitively even after the increase in
concentration ratio. On the other hand, applying the P-R method, Lee and Lee
(2005) found that the Korean banking market showed monopolistically competitive
behavior for both the period of 19921997 and the period of 19982002, while market competition weakened significantly over the latter period. Park (2009) resolved
the conflicting results by dividing the sample period of 19922004 into three separate periods. Studies on bank competition in China are scanty and mostly descriptive rather than analytical, simply reporting the trend of bank concentration
without further investigative analysis.
4. Model
In this study we use the P-R model to assess the competitive nature of the Korean
and Chinese banking industries because this model is robust to the extent that market and bank level data are available. Let a banks revenue function be R = R (x, y1)
where x = a vector of products and y1 = a vector of exogenous variables shifting
the revenue function and let a banks cost function be C = C (x, w, y2) where w is
a vector of input prices and y2 = a vector of exogenous variables shifting the cost
function. y1 and y2 may have common variables.
Profit maximization by the bank requires that marginal revenue equal marginal
cost as R (x, y1) = C (x, w, y2). Panzar and Rosse (1987) calculate the sum of the
elasticities of the revenue with respect to input prices from the reduced-form revenue equation and define it as the H-statistics.
62

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

H R@R=@wi wi =R

where wi is ith input price. Panzar and Rosse show from the profit maximization
condition that the H-statistic is equal to unity (H = 1) in a perfectly competitive
market, and less than or equal to zero (H  0) under monopoly. Although the PR article also shows that 0 < H < 1 could be consistent with oligopolistic behavior,
it is common to regard 0 < H < 1 as the condition of Camberlinian monopolistic
competition. This interpretation is valid under the assumption that the observations
are in the long-run equilibrium (Nathan and Neave, 1989).
Following Park (2009) we specify the reduced-form revenue equation of a bank
as follows.
lnRit a b1 lnw1;it b2 lnw2;it b3 lnw3;it ck Rzk eit

where Rit is bank is revenue at time t, w1 is the input price of labor, w2 is the input
price of capital, w3 is the input price of funds, and zk is a vector of control variables
affecting the banks revenue function. The H-statistic is the sum of b1, b2 and b3.
In order to eliminate manual calculation of b1 + b2 + b3 and its standard error,
Equation (2) can be rearranged as follows.
lnRit a b1 lnw1;it  lnw3;it  b2 lnw2;it  lnw3;it 
b1 b2 b3 lnw3;it ck Rzk eit

The coefficient of ln(w3,it) can be regarded as the estimated H-statistics and its standard error can be used to test the significance of this estimate.
The P-R model is valid only if the market is in equilibrium. Following Shaffer
(1982), Molyneux et al. (1996), Claessens and Laeven (2004), and Park (2009),
Equation (4) is used to test the equilibrium conditions.
lnROAit a b1 lnw1;it b2 lnw2;it b3 lnw3;it ck Rzk eit

In equilibrium, rates of return on assets should not be statistically correlated with factor prices (H = 0). On the other hand, if the market is in disequilibrium, an increase
in factor prices would result in a temporary decline in the rates of return (H < 0).
5. Empirical Analysis
Traditionally the revenue (Rit) is typically measured by interest revenue or its ratio
to total assets, presuming that the main function of banks is financial intermediation. However, with weakening of financial intermediation in recent years and
diversification of bank assets, total revenue or its ratio to total assets is used in
some studies. We use both interest revenue (IR) and total revenue (TR) in this
study. ROA is the ratio of net after-tax income to total assets in percentage. The
unit labor cost (w1,it) is measured by the ratio of personnel expenses to the number
of employees, the unit capital cost (w2,it) is measured by the ratio of depreciation
allowance and other maintenance costs to total fixed assets, and the unit funding
2013 Korean Securities Association

63

K. H. Park et al.

cost (w3, it) is measured by the ratio of interest expenses to the sum of total deposits and borrowings.9
We also include several control variables in the model. Total assets (ASSET)
are included to see the size effect while the number of branches (BRANCH) is
included to account for the effect of bank networks. The ratio of non-performing
loans to total loans (NPL) is included to control for the credit risk effect. The
BIS risk-adjusted capital ratio (BIS) is alternatively used as a control variable for
credit market and operational risk.10 The ratio of non-interest revenue to total
revenue (NINT) is included to reflect the effect of changing financial intermediation or diversification. The variable BRANCH, representing bank network, was
eventually deleted from the regression estimation because of its high correlation
with ASSET.
Bikker et al. (2006) stated that inclusion of a scale explanatory variable such as
ASSET in the P-R model may cause overestimation of the level of competition and
may distort the tests on monopoly and perfect competition. So, we estimated competitive conditions in both ways, with and without the scale explanatory variable,
ASSET. However, the H values, regardless of inclusion or exclusion of ASSET in the
model, show similar test results with no indication that inclusion of a scale explanatory variable causes overestimation of the level of competition. Thus, in the sections
below we only report the estimation results with inclusion of ASSET in the model.
Even though the fixed effects model is usually regarded as more appropriate than
the random effects model, when population data instead of sample data are used,
as in the case of Korea, we use both fixed and random effects models for comparison purposes.11
Data used for Korea are from the Bank Management Statistics by the Bank of
Korea and financial statements of individual banks. Data used for China are from
Bank Scope and the Almanac of Chinese Banking and Finance. For Korea, we use
all domestic commercial banks in operation in any year during the sample period
of 19922008. The total number of Korean banks reached 26 banks in 1997 at its
peak just before the Asian financial crisis, but since then continuously declined to
16 banks in 2006.12 However, only 15 major Chinese domestic commercial banks
9

Unavailability of personnel expenses in further breakdown does not make it possible to

allow differing levels of human capital.
10
For Korea, the BIS risk-adjusted capital ratio is calculated according to the Bank of International Settlements guidelines, which assign varying risk weights to different types of assets.
However, for China, simply the equity ratio is used.
11
See Green (1993) and Hsiao (1986).
12
Seven nationwide banks are Kookmin Bank, Woori Bank, Shinhan Bank, Hana Bank, Korea
Exchange Bank, SC First Bank, and Korea Citi Bank, while six regional banks are Kyungnam
Bank, Jeonbuk Bank, Daegu Bank, Pusan Bank, Kwangju Bank and Cheju Bank. All seven
nationwide banks have their branches throughout the country and all six regional banks have
a branch in Seoul, the major financial center in Korea.
64

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

are included in the sample due to limitation of data availability.13 Fifteen Chinese
banks as a group account for more than 70% of the total bank assets, or total bank
loans or total bank deposits. Many foreign banks operate their branches in Korea
and Chinese, but they are not included in this study, not because their market share
is small, but because their data, which come from a different source, do not have
some variables used in this study.14
5.1. Korea
The estimation results of Equation 2 with the dependent variables of lnIR and lnTR
are reported in Table 2. The H-statistic, which is the sum of b1, b2 and b3, and the
competition condition test results are also given in Table 2. The Wald test rejects
the hypothesis of monopolistic market structure (H = 0) at the 1% level. It also
rejects the hypothesis of perfectly competitive market structure (H = 1) at the 1%
level.
The estimation results of the H values with two different dependent variables,
lnIR and lnTR, are robust as shown in Table 2. The empirical results lead us to
infer that the Korean commercial banking market was monopolistically competitive
during the sample period. However, when we re-estimate Equation 2 for the crisis
period (19972000), its level of competition increased to the level of perfect competition. This might be due to a breakdown of previous connections among banks,
making collusion more difficult.
All the regressions in Table 2 show goodness of fit, as indicated by very high
adjusted R2 values. The unit labor cost (w1,it) and the unit funding cost (w3,it) are
positive and statistically significant, indicating that an increase in unit costs of labor
or funds leads to higher revenue. The unit capital cost (w2,it) is not statistically significant. However, when the scale variable, ASSET, is excluded, the unit capital cost
exerts significant positive effect on revenue.
All control variables have expected signs. The positive sign of ASSET indicates
the presence of the size effect of bank assets. As expected, NINT (the ratio of
non-interest revenue to total revenue) has a significant negative effect on interest
revenue, but significant positive effect on total revenue. These results confirm the
weakening of the financial intermediary function of Korean banks during the
sample period. The ratio of interest revenue to total revenue over time has declined
over time. While this ratio has not changed much for regional banks, it declined
13

Fifteen Chinese banks include the Bank of China, the China Construction Bank, the Industrial and Commercial Bank of China, the Agricultural Bank of China, the Agricultural Development Bank of China (ADBC), China Development Bank (CDB), the Bank of
Communications, China CITIC Bank, China Everbright Bank, China Minsheng Bank, Guangdong Development Bank, Shenzhen Development Bank, China Merchants Bank, Shanghai
Pudong Development Bank and Industrial Bank.
14
Financial statements of foreign banks in Korea and China are available from the Bankscope
database, but they do not have some variables used in this study.
2013 Korean Securities Association

65

K. H. Park et al.

Table 2 Test of competition condition: Estimation Results of Equation 2 for Korean banks
1. lnIR is the natural logarithm of interest revenue while lnTR is the natural logarithm of total revenue.
2. The coefficients of the constant under the fixed effects model are not reported here.because there are
as many as the number of banks. 3. t values are shown in parentheses. *, ** and *** indicate significance
at the 10%, 5% and 1% levels respectively. 4. H statistic is the sum of lnW1, lnW2 and lnW3, and its t
value is obtained by estimating Equation 3.

Fixed effects model

lnIR

Random effects model

lnTR

lnIR

lnTR

0.127***
(7.053)
0.005
(0.313)
0.648***
(28.121)
0.942***
(43.151)
1.276***
(21.737)
0.001
(0.848)
0.012***
(8.231)
0.997
0.753***
(24.828)
551.35***
(0.000)
47.71***
(0.000)

0.256**
(1.917)
0.130***
(8.736)
0.022
(0.978)
0.654***
(29.562)
0.931***
(47.001)
0.389***
(6.519)
0.001
(0.775)
0.012***
(8.783)
0.997
0.812***
(26.553)
665.89***
(0.000)
42.63***
(0.000)

0.277**
(1.895)
0.123***
(8.462)
0.028
(1.531)
0.656***
(28.934)
0.929***
(46.082)
1.205***
(20.542)
0.000
(0.412)
0.014***
(8.812)
0.997
0.822***
(26.678)
636.29***
(0.000)
36.47***
(0.000)

Constant
lnW1
lnW2
lnW3
lnASSET
NINT
NPL SHARE
BIS
ADJ. R2
H statistic
Wald test: H = 0
(q-value)
Wald test: H = 1
(q-value)

0.126***
(7.325)
0.003
(0.083)
0.647***
(29.969)
0.947***
(44.757)
0.365***
(5.769)
0.002
(1.306)
0.011***
(7.915)
0.997
0.776***
(24.192)
537.27***
(0.000)
62.34***
(0.000)

continuously for national banks. This ratio for national banks stood at 0.8 in 1992,
but dropped to 0.5 in 2004. This coincides with the consolidation of banks and creation of mega banks. Mega banks have expanded their business into non-loanrelated activities such as security investments.
NPL had a significant negative effect on lnIR or lnTR while BIS had a significant
positive effect on them, indicating a stronger signaling effect of risk-adjusted equity
ratio on profitability. According to the signal theory (Berger, 1995), banks that expect
to have better performance in terms of profitability credibly transmit this information through a higher equity ratio. The equity ratio became even more important
after the crisis because government bail-out of banks was no longer guaranteed.
66

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

Table 3 Test of competition condition: Estimation results of Equation (2) for Chinese
banks
1. lnIR is the natural logarithm of interest revenue while lnTR is the natural logarithm of total revenue.
2. The coefficients of the constant under the fixed effects model are not reported here.because there are
as many as the number of banks. 3. t values are shown in parentheses. *, ** and *** indicate significance
at the 10%, 5% and 1% levels respectively. 4. H statistic is the sum of lnW1, lnW2 and lnW3, and its
t value is obtained by estimating Equation 3.

Fixed effects model

Constant
lnW1
lnW2
lnW3
lnASSET
NINT
NPL SHARE
BIS
ADJ. R2
H statistic
Wald test: H = 0
(q-value)
Wald test: H = 1
(q-value)

Random effects model

lnIR

lnTR

lnIR

lnTR

0.116**
(2.394)
0.063
(1.431)
0.018**
(2.645)
1.066***
(22.417)
0.270
(1.462)
0.004***
(4.167)
0.013
(1.480)
0.592
0.213***
(9.046)
23.92***
(0.000)
512.93***
(0.000)

0.131**
(2.607)
0.035
(0.459)
0.081**
(2.170)
1.009***
(21.383)
0.942
(1.271)
0.006***
(4.427)
0.014
(1.610)
0.623
0.245***
(10.277)
29.32***
(0.000)
575.62***
(0.000)

0.595*
(1.712)
0.161***
(3.697)
0.035*
(1.885)
0.093**
(9.251)
0.962***
(25.014)
0.347
(1.545)
0.004***
(5.426)
0.014
(1.607)
0.724
0.243***
(12.752)
28.34***
(0.000)
714.22***
(0.000)

0.654
(1.585)
0.171***
(3.491)
0.041
(1.026)
0.047**
(10.462)
0.932***
(28.936)
0.819
(1.231)
0.005***
(4.857)
0.016
(1.433)
0.768
0.268***
(11.744)
20.29***
(0.000)
790.43***
(0.000)

5.2. China
Table 3 presents the estimation results of Equation 2 with the dependent variables
of lnIR and lnTR along with the H-statistic, which is sum of b1, b2 and b3. The
Wald test is applied to test both the hypothesis of a monopolistic market structure
(H = 0) and the hypothesis of a perfectly competitive market structure (H = 1),
and both hypotheses are rejected at the 1% level. However, when we re-estimate
Equation 2 for the first half of the sample period (19922000), the hypothesis of
H = 1 is rejected, but the hypothesis of H = 0 cannot be rejected. This result indicates that there was a dramatic change in the competition level of the Chinese
banking industry over the time.
2013 Korean Securities Association

67

K. H. Park et al.

Compared to the Korean banking industry, the H values for the Chinese banking industry are very small regardless of which revenue (lnIR or lnTR) is used as a
dependent variable. This result indicates that the Chinese banking market is still far
from being a competitive market. Relatively high adjusted R2 values indicate the
goodness of fit for all the regressions in Table 3. All coefficients of the input costs,
that is, the unit labor cost (w1,it), the unit capital cost (w2,it) and the unit funding
cost (w3,it), have the positive sign as expected. However, their coefficient size is of
small value and some coefficients are not statistically significant. It can be inferred
from the empirical results that the Chinese banking market is far from a competitive market, rather closer to monopoly or oligopoly.
The significant and positive sign of ASSET indicates the strong presence of the
size effect. Contrary to the Korean case, NINT (the ratio of non-interest revenue to
total revenue) has no significant effect on both interest revenue and total revenue.
While there has been weakening of the financial intermediary function of Korean
banks, the dominant source of Chinese banks revenue is still interest revenue. Only
recently some Chinese banks have expanded their business into non-loan-related
activities. While NPL has a significant negative effect on lnIR or lnTR as expected,
equity ratio does not have a significant positive effect on them. There seems to be
no stronger signaling effect of the equity ratio on profitability in China.
5.3. Equilibrium Test
Table 4 gives the estimation results of equation 4 with the dependent variable,
lnROA, and the equilibrium tests for both Korea and China. Because the rate of
return on assets of some banks was of negative value, the dependent variable is
actually computed as ln (1 + ROA) where ROA is the ratio of net after-tax income
to total assets.15 Several troubled Chinese state and joint-equity banks in their
earlier years had negative rates of return on assets, and many Korean banks during
the Asian financial crisis had negative rates of return on assets.
For Korea, the estimates yield the H values that are consistent with long-run
equilibrium (H = 0). However, when we re-estimate Equation 4 only for the Asian
crisis period of 19972000, the hypothesis of H = 0 is rejected, even though the
value of H is close to zero. It may be inferred from the findings that the Korean
commercial banking market was in long-run equilibrium before the crisis, but fell
into disequilibrium during the crisis period. However, it made a rapid adjustment
to a new equilibrium after the crisis, which is similar to the findings of Molyneux
et al. (1996) for Japanese commercial banking. For China, the hypothesis that the
market is in equilibrium, that is H = 0, is rejected at the 2% level of significance.

15

Adding 1 to ROA before taking the logarithm is arbitrary, but is a common method used
to handle non-positive numbers in the logarithmic transformation. Bos and Koetter (2006)
pointed out that adding 1 affects composition of total error, but it does not affect coefficient
estimates, which are our ain concern.

68

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

Table 4 Test of equilibrium condition for Korea and China: Estimation results of Equation
(4) with the dependent variable, lnROA
1. Estimation results of the fixed effects model. The coefficients of the constant under the fixed effects
model are not reported here. 2. t values are shown in parentheses. *, ** and *** indicate significance at
the 10%, 5% and 1% levels respectively.

lnW1
lnW2
lnW3
lnASSET
NINT
NPL
BIS
ADJ. R2
H statistic
Wald test: H = 0
(q-value)

Korea

China

0.004
(0.511)
0.005
(1.316)
0.008***
(2.866)
0.009
(1.328)
0.018
(1.502)
0.003*
(1.978)
0.006***
(4.082)
0.634
0.004
(0.146)
2.53
(0.115)

0.009
(0.944)
0.007
(0.655)
0.058**
(2.175)
0.022*
(1.984)
0.004
(0.145)
0.012**
(2.158)
0.002
(1.343)
0.478
0.059***
(2.372)
5.712**
(0.017)

Continuous influx of banks over time and rapid changes in the structure of the
Chinese banking industry might result in disequilibrium conditions.
5.4. Trend of H Value Over Time in Korea and China
As some concerns have been raised in regard to the P-R models shortcomings,
including possible spurious indication of market power, we supplement our results
by employing two additional methods. One method is the Boone indicator to estimate market power for the entire sample period and for moving 3 year time
periods, which will be discussed in the next sub-section. Another attempt is to supplement the static estimates of the H-statistic by dynamic changes of the H-statistic
over time. To see how the H values changed over time, this statistic is estimated for
moving 3 year time periods, that is, 19921994, 19931995, 19941996 and so on.
We used 3 year moving averages partly to solve the degrees of freedom problems.
The estimation results of the H values for both Korea and China are reported
together with HHI in Table 5.
For Korea, market concentration measured by HHI has gradually decreased
until the Asian financial crisis, but increased after the crisis. The H-statistic jumped
2013 Korean Securities Association

69

K. H. Park et al.

to the level of 0.80.9 during the crisis period and then gradually decreased. For the
entire sample period, the correlation coefficient between HHI and the H statistic
with lnIR is only 0.002 while the correlation coefficient between HHI and the H
statistic with lnTR is 0.228 for the entire sample period. The results seem to indicate that increased market concentration over the last 10 years did not affect the
competition level in the Korean banking industry. However, the correlation coefficients between HHI and H-statistic with lnTR and H-statistic with lnTR for a
sub-period of 19972008 are 0.946 and 0.949 respectively. There was a structural

Table 5 Market concentration and competition level over time

Year
1. Korea
19921994
19931995
19941996
19951997
19961998
19971999
19982000
19992001
20002002
20012003
20022004
20032005
20042006
20052007
20062008
2. China
19921994
19931995
19941996
19951997
19961998
19971999
19982000
19992001
20002002
20012003
20022004
20032005
20042006
20052007
20062008

70

HHI - Total assets

H-statistic with lnIR

H-statistic with lnTR

830
777
736
695
726
802
909
1120
1301
1443
1404
1339
1359
1416
1485

0.520
0.609
0.525
0.410
0.877
0.884
0.751
0.690
0.672
0.636
0.638
0.627
0.641
0.623
0.653

0.543
0.623
0.554
0.461
0.924
0.944
0.780
0.675
0.664
0.642
0.598
0.608
0.613
0.592
0.614

2652
2583
2525
2415
2301
2184
2116
2053
2008
1931
1859
1779
1734
1694
1666

0.136
0.164
0.182
0.207
0.215
0.233
0.259
0.271
0.298
0.302
0.342
0.365
0.389
0.405
0.426

0.149
0.174
0.192
0.228
0.247
0.269
0.278
0.299
0.321
0.343
0.338
0.376
0.402
0.427
0.449

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

change between the pre-crisis and post-crisis periods (see Park and Weber, 2006a).
A strong negative correlation exists between market concentration and the level
of competition. Dynamic changes in the H values indicate less competition in the
Korean banking industry over time since the crisis.
Quite contrasting results are obtained for China. Both HHI and the H-statistic
in China show an opposite trend to that of Korea. Market concentration measured
by HHI has continuously declined from 2743 in 1992 to 1642 in 2008. The H statistic with both lnIR and lnTR has gradually increased over the same time period,
from less than 0.15 in the earlier period to more than 0.4 in the later period. So,
these two variables exhibit a high negative correlation. The correlation coefficient
between HHI and the H statistic with lnIR is 0.974 while the correlation coefficient between HHI and the H statistic with lnTR is 0.976. Decreased market concentration in the Chinese banking sector definitely contributed to improving the
level of competition in the banking industry in China, even though the effect may
be mild.
5.5. The Boone Indicator Model
The Boone model is based on the assumption that firms with lower marginal costs,
thus more efficient firms, gain higher market profits and market shares. Following
Boone (2004) and Van Leuvensteijin et al. (2007), the model for market share is
defined as follows.
ln Sit a b ln mcit uit

The sign of b is expected to be negative because the market shares of banks with
lower marginal costs are expected to increase. The magnitude of b is expected to be
larger when competition is strong. This b parameter is called the Boone indicator.
In order to estimate marginal cost, we use a translog cost function based on a
multi-product production function as follows.
ln Cit a Rbj ln qj;it Rck ln wk;it RRdjk ln qj;it ln qk;it
RRkjk ln qj;it ln wk;it RRujk ln wj;it ln wk;it eit

where Cit is the production cost of bank i at time t, qj,it is output of bank i at time
t, consisting of loans, securities and other services, and wk,it is input price of bank i
at time t, consisting of labor price, capital costs, and input price of funds.
Table 6 presents estimates of b parameter for the entire sample period and for
moving 3 year time periods. For the entire sample period, the Boone indicator for
Korea with a negative b value (1.12) indicates a relatively competitive banking
market while the Chinese banking market with a positive b value (1.97) is less competitive. Estimates of b value for moving 3 year time periods show two distinctive
trends, diminishing competition over time in Korea and improving competition in
China.
Table 7 reports correlation coefficients among the four variables measuring market concentration, degree of competition and market power: HHI, H-statistic with
2013 Korean Securities Association

71

K. H. Park et al.

Table 6 Boone Indicator estimates: Estimation results of Equation (5)

*,** and *** indicate significance at the 10%, 5% and 1% levels respectively.

Year

Korea
b value

China
b value

19922008
19921994
19931995
19941996
19951997
19961998
19971999
19982000
19992001
20002002
20012003
20022004
20032005
20042006
20052007
20062008

1.12***
1.23**
0.97
1.14*
0.86
2.99***
3.22***
2.61***
1.67**
1.52**
1.27*
1.03*
0.91*
0.76
0.83
0.79

1.97***
7.34**
5.97**
4.73**
3.87**
4.65**
2.38*
1.43*
0.24
0.19
0.17
0.34**
0.44**
0.29*
0.75
0.68

Table 7 Correlation coefficients

Numbers in parentheses are correlation coefficients when the pre-crisis period is excluded.

1. Korea
HHI
H value
H value
Boones
2. China
HHI
H value
H value
Boones

HHI

H value (lnIR)

H value (lnTR)

Boones b

(lnIR)
(lnTR)
b

1.0

0.002 (0.946)
1.0

0.228 (.949)
0.968 (0.994)
1.0

0.471 (0.963)
0.821 (0.957)
0.924 (0.973)
1.0

(lnIR)
(lnTR)
b

1.0

0.974
1.0

0.976
0.992
1.0

0.956
0.917
0.924
1.0

lnIR, H-statistic with lnTR, and Boones b value. Park and Weber (2006b) found a
structural change in the Korean banking market during the Asian financial crisis,
using the Chow breakpoint test. So, for Korea, two correlation coefficients, one for
the entire sample period and another for the post-crisis period are reported. All
four variables are highly correlated for both Korea and China.
72

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

6. Summary and Conclusions

In this study we examined the association between market concentration and bank
competition in Korea and China. For Korea, even though market concentration
decreased during the pre-crisis period due to financial deregulation, it has markedly
increased since the crisis because of reduction in the number of banks, bank consolidation, and creation of mega banks. With this change, there has been a growing
concern over market power in the Korean banking industry. Even though the Korean commercial banking market has become increasingly concentrated in the process of restructuring since the crisis, our study shows that bank competition has
not been greatly affected by bank consolidation.
The Korean banking industry has been monopolistically competitive for the
sample period of 19922008 while the competition condition test shows competition in the Korean banking actually increased to the level of perfect competition
during the crisis period (19972000). The Korean banking system may have
remained relatively competitive despite its consolidation partly due to the entry of
foreign banks and increased foreign ownership of domestic banks. Although an
increase in market concentration has not changed the overall level of competition
for the entire sample period, that is, still monopolistically competitive, market concentration has resulted in less competition over time evidenced by the gradual
decline of the H-statistic over time, diminishing Boones b values and an increase
in the average interest margin. Therefore, a growing concern over market power in
the Korean banking industry is worth noting. Future mergers that lead to a highly
concentrated market might very well reduce banking competition.
For China, the competitive conditions of the Chinese banking industry have
definitely improved over time. The Chinese banking system progressed from a one
bank system to a system with four state banks in the 1980s to more than 20 banks,
including joint-equity commercial banks in the 1990s, to more than 200 banks at
the present time. This study finds that in spite of a drastic decrease in market concentration in the Chinese banking industry, its competition conditions are far from
a competitive market, as evidenced by the small H statistic and positive or small
negative values of the Boone indicator. The sheer number of banks does not guarantee a competitive market. Lowering entry barriers for private banks and foreign
banks would further facilitate competition. Institutional changes and lifting government regulation on banking are also necessary to speed up competitive behaviors in
the Chinese banking market.
References
Baumol, W. J., J. C. Panzar, and R. D. Willig, 1982, Contestable markets and the theory of
industry structure, Harcourt Brace Jovanovich (San Diego).
Berg, S. A., and M. Kim, 1996, Banks as multioutput oligopolies: an empirical evaluation of
the retail and corporate banking markets, Proceedings of a Conference on Bank Structure
and Competition, Federal Reserve Bank of Chicago, pp. 183201.
2013 Korean Securities Association

73

K. H. Park et al.

Berger, A. N., 1995, The profit-structure relationship in banking tests of market-power and
efficient-structure hypotheses, Journal of Money, Credit and Banking, 27(2), pp. 404431.
Bikker, J. A., and J. M. Groeneveld, 2000, Competition and concentration in the EU banking
industry, Kredit und Kapital, 33 (1), pp. 6298.
Bikker, J. A., and K. Haaf, 2002, Competition, concentration and their relationships: An
empirical analysis of the banking industry, Journal of Banking & Finance, 26, pp. 2191
2214.
Bikker, J. A., L. Spierdijk, and P. Finnie, 2006, Misspecification in the Panzar-Rosse model:
Assessing competition in the banking industry, DNB Working Paper 114 (www.dnb.nl).
Boone, J., 2004, A new way to measure competition, CEPR Discussion Paper, No. 4330.
Bos, J. W., and M. Koetter, 2006, Handling losses in translog profit models. Mimeo, Utrecht
School of Economics, Utrecht University (The Netherlands).
Bresnahan, T., 1982, The oligopoly solution concept is identified, Economics Letters, 10, 8792.
Bresnahan, T., 1989, Empirical studies of industries with market power. In R. Schmalensee,
and R. D. Willig eds: Handbook of Industrial Organization. Vol. II, Chapter 17,
pp. 10111057. North-Holland Publishing (New York).
Claessens, S., and L. Laeven, 2004, What drives bank competition? Some international
evidence, Journal of Money, Credit and Banking, 36 (3), pp. 563583.
Coccorese, P., 1998, Assessing the competitive conditions in the Italian banking system: Some
empirical evidence, BNL Quarterly Review, 205, pp. 171191.
De Bandt, O., and E. P. Davis, 2000, Competition, contestability and market structure in
European banking sectors on the eve of EMU, Journal of Banking & Finance, 24 (6),
pp. 10451066.
Fuentes, I., and T. Sartre, 1998, Implications of restructuring in the banking industry: The
case of Spain, BIS Conference Papers, 7, 98120.
Gelos, R. G., and J. Roldos, 2002, Consolidation and market structure in emerging market
banking systems, IMF Working Paper, No. 02/186.
Green, W., 1993, Econometric analysis. Macmillan Publishing Co (New York).
Gruben, W. C., and R. McComb, 2003, Privatization, competition, and supercompetition in
the Mexican commercial banking system, Journal of Banking & Finance, 27 (2), pp. 229
249.
Hempell, H., 2002, Testing for competition among German Banks. Discussion Paper,
Economic Research Center for the Deutsche Bundesbank.
Hondroyiannis, G., S. Lolos, and E. Papapetrou, 1999, Assessing competitive conditions in
the Greek banking system, Journal of International Financial Markets, Institutions and
Money, 9, pp. 377391.
Hsiao, C., 1986, Analysis of panel data. Cambridge University Press (Cambridge).
Kim, H., 2003, Assessing concentration and competition in the Korean banking industry
(in Korean), KDI Policy Studies, 25, pp. 5798.
Lau, L., 1982, On identifying the degree of competitiveness from industry price and output
data, Economics Letters, 10, pp. 9399.
Lee, S., and J. Lee, 2005, Bank consolidation and bank competition: An empirical analysis of
the Korean banking industry (in Korean), Economic Papers, 8, pp. 102144.
Lee, M., and M. Nagano, 2008, Market competition before and after bank merger wave: a
comparative study of Korea and Japan, Pacific Economic Review, 13 (5), pp. 604619.

74

2013 Korean Securities Association

South Korean and Chinese Commercial Banking Markets

Molyneux, P., D. M. Lloyd-Williams, and J. Thornton, 1994, Competitive conditions in

European banking, Journal of Banking & Finance, 18, pp. 445459.
Molyneux, P., J. Thornton, and D. Lloyd-Williams, 1996, Competition and market contestability
in Japanese commercial banking, Journal of Economics and Business, 48 (1), pp. 3345.
Nathan, A., and E. H. Neave, 1989, Competition and contestability in Canadas financial
system: Empirical results, Canadian Journal of Economics, 22 (3), pp. 576594.
Panzar, J. C., and J. N. Rosse, 1982, Structure, conduct and comparative statistics, Bell
Laboratories Economic Discussion Paper.
Panzar, J. C., and J. N. Rosse, 1987, Testing for monopoly equilibrium, Journal of Industrial
Economics, 35, pp. 443456.
Park, K., 2009, Has bank consolidation in Korea lessened competition? Quarterly Review of
Economics and Finance, 49, pp. 651667.
Park, K., and W. Weber, 2006a, Profitability of Korean banks: Test of market structure
versus efficient structure, Journal of Economics and Business, 58, pp. 222239.
Park, K., and W. Weber, 2006b, A note on efficiency and productivity growth in the Korean
Banking industry, 19922002, Journal of Banking & Finance, 30, pp. 23712386.
Shaffer, S., 1982, A nonstructural test for competition in financial markets, Proceedings of
a Conference on Bank Structure and Competition, Federal Reserve Bank of Chicago,
pp. 225243.
Shaffer, S., 1989, Competition in the US banking industry, Economics Letters, 29 (4), pp. 323323.
Shaffer, S., 1993, A test of competition in Canadian banking, Journal of Money, Credit, and
Banking, 25 (1), pp. 4961.
Shaffer, S., 2002, Competitive bank pricing and adverse selection, with implications for testing
the SCP hypothesis, Quarterly Review of Economics and Finance, 42 (3), pp. 633647.
Shaffer, S., 2004, Patterns of competition in banking, Journal of Economics & Business, 56,
pp. 287313.
Smirlock, M., 1985, Evidence on the (non) relationship between concentration and
profitability in banking, Journal of Money, Credit, and Banking, 17, pp. 6983.
Uchida, H., and Y. Tsutsui, 2005, Has competition in the Japanese banking sector improved?,
Journal of Banking and Finance 29 (2), pp. 419439.
Van Leuvensteijin, M., J. Bikker, A. Van Rixtel, and C. Sorensen, 2007, A new approach to
measuring competition in the loan markets of the Euro area, European Central Bank
Working Paper, No. 768.
Vesala, J., 1995, Testing for competition in banking behavioral evidence from Finland. Bank
of Finland (Helsinki).

2013 Korean Securities Association

75