The Quarterly Review of Economics and Finance

49 (2009) 693710

Competitive behavior in Middle East and

North Africa banking systems
Rima Turk-Ariss
Lebanese American University, Business School, P.O. Box 13-5053, Beirut, Lebanon
Received 15 October 2007; received in revised form 5 March 2008; accepted 13 March 2008
Available online 21 March 2008

Abstract
This paper investigates the degree of market power in Middle East and North Africa (MENA) banking
systems where research on competitive conditions is scant. The banking sectors of MENA countries are
highly concentrated and they present unique characteristics in terms of ownership, structure and growth
potential. The degree of competitiveness is assessed based on the revenue elasticity to input prices approach,
and is related to a set of market and contestability indicators. The results show that, except for countries in
North Africa where monopolistic conditions are found, the prevailing market structure in MENA banking
is mostly monopolistically competitive. In line with the finding on other emerging and developed countries,
assuring greater market contestability by allowing more foreign bank participation and reducing activity
restrictions on banks is most important to guaranteeing competitiveness in the highly concentrated banking
systems of the MENA region.
2008 The Board of Trustees of the University of Illinois. Published by Elsevier B.V. All rights reserved.
JEL classication: D4; G21; L11; N25; O16
Keywords: MENA countries; Market structure; Competition; Contestability; Panzar and Rosse methodology

1. Introduction
Recent advances in information technology, financial innovations, globalization and deregulation have reduced margins in traditional banking activities worldwide. This set off a wave of
mergers between banks and other financial institutions causing drastic changes in the structure of
the banking industry. In turn, greater consolidation raises concerns regarding issues of competition in the local retail markets. The policy implications of such concerns are particularly relevant

Tel.: +961 1 786456x1644.

E-mail address: rima.turk@lau.edu.lb.

1062-9769/$ see front matter 2008 The Board of Trustees of the University of Illinois. Published by Elsevier B.V. All rights reserved.

doi:10.1016/j.qref.2008.03.002

694

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

Table 1
Three-bank total asset concentration average in MENA banking, 20002006
Country

Concentration

Algeria
Bahrain
Jordan
Kuwait
Lebanon
Morocco
Oman
Qatar
Saudi Arabia
Tunisia
Turkey
United Arab Emirates

85.92
80.47
87.15
68.51
37.86
49.50
79.14
90.99
56.11
45.09
44.60
49.62

Source: Authors calculations based on data retrieved from the BankScope database.

because regulators have traditionally used banking market structure as a policy variable to recommend measures aimed at fueling competition, promoting financial liberalization and removing
barriers to entry.1
The Middle East and North Africa (MENA) region is strategically located between Asian
economies and the Western world. Except for Turkey, MENA countries were colonized by the
French or the British until mid of the past century. Major institutions including financial intermediaries were established following the Western style, but they present some interesting features
that make them a challenging fieldwork. Financial sectors in MENA countries are generally still in
the early phases of economic development. Capital markets are weak or almost non-existent, and
financial markets are dominated by bank-financed credit mechanisms. In this framework, banks
are the main suppliers of credit to private and public investment projects and they also finance
government deficits. While this feature is common to many other emerging economies, banking
sectors in the MENA region are unique in three aspects.
First, the recent oil price hike marks the beginning of a new era that was last witnessed a quarter
of a century ago. Investment opportunities in the United States using petrodollars coming from
the Middle East became more restricted after 11 September, so that oil surplus funds have to be
channeled to productive uses elsewhere in the world. Monetary authorities in MENA countries
generally require banks to adopt international accounting standards and to comply with international regulatory requirements, including Basel II and anti-money laundering recommendations.
Still, a major concern is raised regarding the absorptive capacity of banks in the MENA region to
recycle oil surplus funds, justifying the concerns of policymakers with appropriately designing
policies for more efficient and stable banking systems.
Second, banking systems in the MENA region have traditionally been very highly concentrated
markets. Table 1 shows the three-bank average concentration ratios over the period 20002006
based on total assets. While Lebanon, Morocco, Saudi Arabia, Tunisia, Turkey and UAE have
a three-bank concentration ratio ranging between 37 and 57%, bank concentration in Algeria,
Bahrain, Jordan, Oman and Qatar exceeds 80%. In spite of historical high concentration ratios,
1 In the US, examples of regulatory changes include the relaxation of interstate branching in 1992, the Riegle-Neal
Interstate Banking and Branching Efficiency Act in 1994, and the Financial Services Modernization Act allowing the
operation of commercial banking, investment banking and insurance underwriting within the same holding company.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

695

intense merger activity is also taking place within and across several MENA countries. In some
countries where a large number of banks operate, like Lebanon, Turkey and the UAE, large banks
have sought to consolidate their position domestically before expanding in the region. In other
countries like Jordan and Kuwait, leading banks are strategically investing across borders in order
to enhance their growth potential. Ongoing consolidation of financial institutions within each
country and regionally justly intensifies public policy debates on issues of concentration and
competition in the banking industry.
Third, the governance structure of MENA banks is evolving, following accession to the World
Trade Organization (WTO) and greater commitment to financial liberalization. Traditionally,
banking institutions in the MENA region were either mostly family-owned businesses managed
by major shareholders who cater for their own personal interests, or dominated by state authorities, thus making it more difficult for new firms to compete in the industry. The past decades,
however, have altered the ownership of large shares of the banking systems from government
to private control and from domestic to foreign control. Such changes occurred as governments
privatized many of their state-owned banks and reduced barriers to foreign entry, in line with the
WTO accession requirements. Leading international financial intermediaries such as Citigroup,
HSBC, BMP Paribas, ABN Amro and Standard Chartered have generally set up a wide presence
in the MENA region through subsidiaries and are competing with domestic banks in different
segments of the industry. Foreign investors bring in state-of-the-art technology, sophisticated risk
management techniques and qualified human capital, forcing domestic banks to undergo major
structural reforms in order to compete on an equal platform with their peers. Together, financial
liberalization measures and incentives to attract foreign banks lead to better disclosure requirements which, together with a better regulatory environment, ultimately have a positive effect on
a countrys growth performance.
It is worthwhile noting that, in addition to intense competition from foreign banks, commercial banks in the MENA region are also facing increased competition from Islamic banks.
Recently, banks operating in accordance with the Islamic legal code have proliferated in the
MENA region and have become established as viable modes of financial intermediation, thereby
increasing competitive pressures in the industry. Table 2 shows the distribution of total assets in
the MENA region across commercial and Islamic banks in the years 2000 and 2006. The figures
indicate that the banking industry has more than doubled in size from $934,477 to $2,478,841
million between 2000 and 2006. However, this was accompanied by a decrease in conventional
banks share of the market which shrank from 94.12% in 2000 to 88.82% in 2006. During the
same period, the share of Islamic banks almost doubled from 5.88 to 11.18% of the banking
industry.
In view of these developments, interest in competitive conditions in banking in the MENA
region started to emerge. Issues of concentration and competition have been studied in the literature
Table 2
Distribution of total assets across commercial and Islamic banks in the MENA region
Specialization

Commercial banks
Islamic banks
All

2000

2006

Million US$

% of total

879,554
54,923
934,477

94.12
5.88
100.00

Million US$

Source: Authors calculations based on data retrieved from the BankScope database.

2,201,603
277,238
2,478,841

% of total
88.82
11.18
100.00

696

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

on the international scale (Claessens & Laeven, 2004), for developed countries (Bikker & Haaf,
2002) and for emerging economies (Gelos & Roldos, 2004), but not for MENA banking sectors.2
A recent paper by Al-Muharrami, Matthews, and Khabari (2006) assesses market structure and the
degree of competitiveness in the six Gulf Cooperation Council (GCC) countries using traditional
measures of concentration and the Panzar and Rosse (PR 1987) H-statistic. Their findings indicate
that banks operating in the GCC region operate under either monopolistically competitive or
perfectly competitive conditions, notwithstanding the relatively high degree of concentration
levels in the industry.
This study contributes to the literature on market structure in Arab banking systems by using a
larger sample to include non-GCC Arab countries of the Middle East and North Africa. It estimates
a measure of competitiveness in 12 very highly concentrated banking sectors in the MENA region,
and relates it to a set of industry and contestability indicators to explain differences in the degree
of competition across countries.
The findings reveal that, except for North Africa where the tests of hypotheses fail to reject
monopoly, monopolistic competition best describes market structure in other banking sectors
in the MENA region and that high concentration does not adversely affect competitive conditions in the region. In line with the results reported in the literature for most mature and
emerging markets (Bikker & Haaf, 2002; Claessens & Laeven, 2004; Gelos & Roldos, 2004),
banking systems with few activity restrictions and greater foreign bank participation are found
to be highly competitive. Further, the general level of economic development is found to be
a significant factor that explains differences in the degree of competitiveness in the MENA
region.
The rest of the paper is organized as follows. Section 2 surveys the research literature on
competitive conditions using the PR methodology. Section 3 presents the PR estimation to assess
the degree of competitiveness, based on the relationship between revenues and marginal costs.
Section 4 presents the data and the estimated competitiveness scores. Section 5 investigates the
factors that account for differences in competitive behavior in MENA banking. The conclusion is
offered in Section 6.
2. Literature review
The literature on the assessment of competitive behavior in banking is divided into two main
streams. The structural method has its roots in the theory of industrial organization that measures competitiveness following the Structure-Conduct-Performance (SCP) paradigm (fewer and
larger firms are more likely to engage in anticompetitive conduct)3 and the alternative efficiency
hypothesis (high concentration endogenously reflects the market share gains of efficient firms).4
In testing these theories, the empirical literature primarily uses the number of banks or the degree
of concentration in the banking sector as a proxy for market power.5 To cite a few, Berger and
Hannan (1989) report that deposit rates are significantly lower (and loan rates significantly higher)
in most concentrated markets, suggesting that greater concentration in banks implies a likelihood

2 The only MENA banking sectors included in the Claessens and Laeven (2004) study are Turkey and Lebanon, and
the only MENA country considered in the sample of Gelos and Roldos (2004) is Turkey.
3 See for example Berger and Hannan (1989), Hannan and Berger (1991) and Neumark and Sharpe (1992).
4 See for example Smirlock, Gilligan, and Marshall (1984), Smirlock (1985) and Shepherd (1986).
5 It is beyond the scope of this paper to provide a review of the deep SCP literature.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

697

for a price-setting behavior. Hannan and Berger (1991) and Neumark and Sharpe (1992) also
document deposit rate rigidity on the upward side.6
Other research, however, suggests that the number of banks, the Herfindahl and other concentration indexes are not, in themselves, sufficient indicators of the degree of competitive behavior.
In the context of studies of the so-called New Economic Industrial Organisation (NEIO), two
non-structural empirical tests for the degree of competitive behavior have been widely applied to
the banking sector. The model by Bresnahan (1982) and Lau (1982) allows for the estimation of
the degree of competition using aggregate industry data, and the alternative Panzar and Rosse (PR
1987) methodology employs bank-level data. While the first technique estimates demand, supply
and price equations simultaneously, the second allows for bank-specific differences in production
functions, provided that banks are examined under long-run equilibrium. Shaffer (2004b) presents
a detailed analysis concerning the advantages and disadvantages of the two NEIO tests for market
power, and concludes that the relative importance of each advantage and disadvantage is likely
to be specific to each sample (p. 297). Of the two methodologies, data availability at the firm
level presents an advantage to the PR test of competition, since revenues are more likely to be
observable than output prices and quantities. The idea is that the degree of bank competition
should not only be related to the industrys prevailing market structure, but rather to the actual
behavior of a marginal bank by estimating deviation from competitive pricing.
Empirical evidence on the degree of competition using the PR methodology is diverse. Apart
from Japan and, according to some work from Italy, a monopolistic competition structure is
reported for developed countries. In a pioneering research, Shaffer (1982) reveals monopolistic
competition among banks in New York. Nathan and Neave (1989) report perfect competition for
1982 and monopolistic competition for 19831984 for Canadian banks. The work of Molyneux,
Thornton, and Lloyd-Williams (1996) reports perfect collusion in the Japanese banking market.
An earlier study by Molyneux, Lloyd-Williams, and Thornton (1994) finds that monopolistic competition best describes the market structures in France, Germany, Spain and the UK. They could
not, however, reject monopoly for the case of Italy, a result that was contradicted by Coccorese
(1998) and Trivieri (2007). Different results are revealed by De Bandt and Davis (2000) who
report monopoly power in France and Germany and monopolistic competition in Italy. Still other
research by Rime (1999) on Swiss banks and Bikker and Groeneveld (2000) on 15 European
Union (EU) countries indicate that monopolistic competition is the prevailing market structure.
This finding is strongly confirmed by Bikker and Haaf (2002) based on a study covering 23 developed countries including Europe, the US, Canada, Japan and Australia. Bikker and Haaf report
that the banking industry in most countries exhibits monopolistic competition, but this finding is
less uniform when the market is split into local, national and international banks. Small banks are
found to exercise a degree of market power by catering more to local markets. Other research by
Weill (2004) finds that monopolistic competition best describes market conditions for a sample of
12 EU countries over the period 19941999, and Utero-Gonzales (2004) use the PR methodology
to compare the Spanish and UK banking systems.
More recently, Casu and Girardone (2006) extend the standard specification of the PR Hstatistic to account for bank efficiency for the former EU 15 member countries and find that the
degree of concentration is not related to the level of competition. In an extensive paper, Staikouras
and Koutsomanoli-Fillipaki (2006) find evidence of monopolistic competition for the EU 25

6 For a review of research on the effects of competition in general and on access to financing in particular, see Berger,
Demsetz, and Strahan (1999).

698

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

Member States over the 19982002 period and with larger banks behaving more competitively
than smaller banks, and new members showing higher levels of competition than former members.
Matthews, Murinde, and Zhao (2007) report that competitive conditions among the major British
banks are characterized by monopolistic competition. They find evidence that the intensity of
competition in the core market for bank lending remained approximately unchanged, although
competition appears to have become less intense in off-balance sheet business of British banks.
Recent research builds on the previous literature by attempting to explain why market concentration may not reliably predict pricing behavior. Beck, Demirguc-Kunt, and Levine (2003) and
Martinez, Soledad, and Mody (2004) show that competitive pricing may occur even in concentrated markets. Alternatively, cooperation can still be maintained in a market with a large number
of firms (Friedman, 1971). In any case, competition is a multifaceted issue, and factors other than
concentration determine the degree of competition in banking, called contestability. To assess the
competitive environment, broader measures of competition are considered, such as indicators for
regulation, entry restrictions into the banking system, and other legal impediments to bank competition. Claessens and Laeven (2004) relate the competitiveness of a countrys banking sector
with structural and regulatory indicators of the financial system. They use panel data (19942001)
to investigate the degree of competition in banking for 50 countries and attempt to identify factors
that explain the contestability of banking sectors across countries. They find that competitive
behavior in banking is positively related to foreign bank presence, less-severe entry restrictions,
and few activity restrictions. Studies by DeYoung and Nolle (1996) and Berger, DeYoung, Genay,
and Udell (2000) find that foreign-owned banks may compete in different ways from domestically
owned institutions. Barth, Caprio, and Levine (2004) further report that tighter entry requirements
negatively affect bank efficiency, and restricting foreign bank participation increases bank fragility.
With respect to research focusing on emerging economies, studies on competitive market
structure using the PR methodology have recently surged. Burdisso, Catena, and DAmato (2001)
find that competitive conditions in Argentina were very close to perfect competition during the
period 19971999. Gelos and Roldos (2004) examine the evolution of market structure in emerging
banking systems during the 1990s. They report that, although concentration increased during the
period of study, the banking markets in eight Central European and Latin American countries
have not become less competitive. By lowering barriers to entry and encouraging foreign banks
participation, greater consolidation did not lead to a decline in competition. Other research finds
evidence for a noncompetitive market structure in the Ghanaian (Buchs & Mathisen, 2005) and
Brazilian (Belaisch, 2003) banking systems. Hauner and Peiris (2005) investigate the banking
sector in Uganda and find that the level of competitiveness has significantly increased over time
and that foreign banks have responded efficiently to the increased competition. Yuan (2006)
examines the Chinese banking industry and reports that it was near a state of perfect competition
before foreign banks began to enter Chinas financial market. Finally, Yeyati and Micco (2007)
examine the impact of concentration and foreign penetration on competition and find that, while
concentration did not reduce competition in the industry, foreign penetration appears to have led
to less competitive banking sectors.
Research on Arab banking markets is scant to date. Al-Muharrami et al. (2006) examine the
degree of competitiveness in the GCC region using a variety of traditional measures of concentration and the PR H-statistic. Their results show that, while concentration in banking is high, most
Arab GCC markets operate under either monopolistically or perfectly competitive conditions.
This research extends the literature on MENA banking on two fronts. First, a broader sample of
countries in the MENA region is considered and second, factors that explain differences in the
degree of competitiveness in banking are investigated.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

699

3. Methodology
Panzar and Rosse (PR, 1987) conduct a test based on reduced-form revenue functions and
identify the market structure of the industry where the firm operates. The PR methodology allows
for the calculation of a measure of market structure, the H-statistic, as the sum of the elasticities
of total revenues of the bank with respect to its input prices. It is estimated from the following
reduced-form revenue equation on pooled samples for each country:
ln(TRit ) = + 1 ln(WL,it ) + 2 ln(WF,it ) + 3 ln(WK,it ) + 1 ln(Y1,it ) + 2 ln(Y2,it )
+ 3 ln(Y3,it ) + it

(1)

The dependent variable TRit indicates total revenues measured by the ratio of interest and
non-interest revenues to total assets, following Shaffer (1982), Nathan and Neave (1989), Casu
and Girardone (2006). Both traditional lending and fee-generating activities represent the major
sources of revenues for banks. Also, in globalized financial markets, the distinction between
traditional and non-interest income is becoming less relevant as banks compete on both fronts.
Three input prices are included in Eq. (1), WL,it as the cost of labor proxied by the ratio of personnel
expenses to total assets,7 WF,it as the cost of funds represented by the ratio of interest expenses
to total deposits and WK,it as the cost of fixed capital calculated as the ratio of other operating
and administrative expenses to total assets. Consistent with Bikker and Haaf (2002), Claessens
and Laeven (2004), Gelos and Roldos (2004) and Molyneux et al. (1996) and other bank-specific
control variables are included in the analysis. Y1,it and Y2,it represent the ratio of equity to total
assets and net loans to total assets, respectively, and control for business and portfolio mix of
the bank. These two explanatory variables are likely to reflect differences in risk and the capital
structure of banks. It is expected that better capitalization levels and a higher allocation of assets to
loans will generate more revenues and are therefore positively associated to the dependent variable.
To control for potential size effects across banks, Y3,it indicates total assets. While there is no
expectation about the sign on total assets, the results of the estimation would provide information
whether banks face economies or diseconomies of scale. The subscripts i and t refer to bank i
operating at time t. Because the data used concern institutions operating in the same industry and in
the same country, bank effects are most appropriate to capture differences in individual data. The
choice between a fixed and a random effect specification is based on the results of the Hausman test.
The PR H-statistic is then computed as the sum of the input price elasticities of total revenues.
Panzar and Rosse show that the H-statistic can reflect the structure and conduct of the market to
which the firm belongs, and it is interpreted as follows. Under long-run competitive equilibrium,
any increase in input prices should lead to an equivalent increase in total revenues, and firms that
cannot cover the increase in input prices will be forced to exit the market; therefore the H-statistic
would be equal to one. By contrast, if the firm operates as a monopoly, the H-statistic will be
negative since an upward shift in the marginal cost curve will be associated with a reduction in
revenue as a result of the optimal condition for the monopolist. Finally, if the market structure is
characterized by monopolistic competition, the H-statistic will lie between 0 and 1.
While Panzar and Rosse have shown that the H-statistic is a decreasing function of market power
for a pure monopoly, they point out that the interpretation of its value is less straightforward. In
7 The most precise proxy for the price of labor is the ratio between labor costs and the number of employees. However,
the BankScope database lacks suitable data on detailed personnel number for the majority of banks, and the next best
proxy used for the unit price of labor is the ratio of personnel expenses to total assets.

700

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

their illustration, they maintain that the magnitude as well as the sign of the statistic may be of
interest (PR, p. 446). Similarly, Shaffer (2004a, 2004b) casts doubts on the use of the H-statistic
as a continuous measure of competition, and concurs that the interpretation of the specific value
of H may be ambiguous (2004b, p. 297). Generalizing the result of Panzar and Rosse, Vesala
(1995) proves that, under certain conditions, the H-statistic is an inverse measure of market power
exercised on the part of firms. Along the same lines, Bikker and Haaf (2002) and Claessens
and Laeven (2004) argue that values of H between 0 and 1 can be interpreted as a continuous
measure of level of competition, in the sense that higher values of the statistic indicate stronger
competition than lower values. Several assumptions are needed to apply this framework in the
context of the banking industry. Banks are treated as single product firms, acting exclusively as
financial intermediaries that produce interest revenues using labor, capital and intermediated funds
(mainly deposits) as inputs (De Bandt & Davis, 2000). Moreover, it is assumed that higher factor
prices are not correlated with higher revenues generated by higher quality services, since this
may bias the computed H-statistic. Other assumptions include profit maximization and normally
shaped cost and revenue functions (Gelos & Roldos, 2004).
Most importantly in the PR framework, banks should be observed from a long-run equilibrium
perspective. In line with previous research, the problem of volatile economic environment in
the countries of study is overcome by considering a panel data specification and testing the
observations for long-run equilibrium using the following model:
ln(ROAit ) = + 1 ln(WL,it ) + 2 ln(WF,it ) + 3 ln(WK,it ), +1 ln(Y1,it ) + 2 ln(Y2,it )
+ 3 ln(Y3,it ) + it

(2)

where ROA is the pre-tax return on assets.8 The equilibrium statistic E is calculated as the sum
of the input price elasticities, and the hypothesis that its value is 0 is tested where, if rejected, the
market is not in equilibrium. The intuition behind the test is that, in the long-run, return on assets
is not related to input prices.9
4. Data and results for the degree of competition
Bank-level financial statements are retrieved from the BankScope database provided by FitchIBCA (International Bank Credit Analysis Ltd.). The sample represents panel data on commercial
banks in the MENA region for the years 20002006. Focusing on commercial banks enhances
the comparability of banks across MENA countries due to varying degrees of non-commercial
banks participation in banking activities.10 The BankScope database has comprehensive coverage
in most countries, accounting for over 90% of all banking assets. The original sample covered
16 MENA countries; however, some filtering rules were applied. Banks, for which data on the
main variables are not available, such as personnel expense or the cost of fixed capital, were
deleted11 ; and banks with less than three consecutive yearly observations were omitted. This
reduced the sample to an unbalanced panel for 12 countries consisting of 205 banks and a total
of 1117 observations over the period 20002006. Descriptive statistics on the variables included
are provided in Table 3. In terms of the number of banks, two countries dominate the sample,
8
9
10
11

A constant is added to ROA avoid taking the natural logarithm of a negative number.
See for example Claessens and Laeven (2004), Molyneux et al. (1996) and Shaffer (1982).
Islamic banks are not included in the analysis.
As such, Egypt, Lybia, Yemen and Iran are excluded from the sample.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

701

Table 3
Descriptive statistics for MENA banking systems over the period 20002006
Country

Number
of banks

Number of
observations

Statistic

TA

Loans

Equity

Net income

Algeria

16

57

Mean
S.D.

2713.88
3579.71

1190.14
1670.15

155.35
183.67

7.64
12.58

Bahrain

13

63

Mean
S.D.

5947.25
7171.65

2242.84
2891.34

679.65
669.24

76.93
110.53

Jordan

11

73

Mean
S.D.

5640.77
9119.49

2236.58
3653.60

656.62
1188.02

69.36
122.11

Kuwait

43

Mean
S.D.

8093.02
5517.89

3859.12
2938.01

972.79
660.05

193.07
180.11

Lebanon

53

258

Mean
S.D.

1624.92
2403.78

367.50
511.90

119.15
205.11

12.84
24.87

Morocco

10

67

Mean
S.D.

5604.30
3845.43

2460.94
1833.71

490.54
366.29

46.15
66.40

Oman

38

Mean
S.D.

2067.32
1530.41

1429.26
1000.61

248.97
158.15

31.76
44.22

Qatar

41

Mean
S.D.

3322.61
4305.12

1980.68
2787.85

519.88
650.28

83.37
120.07

Saudi Arabia

63

Mean
S.D.

15333.00
9476.37

7267.03
4887.69

1737.06
1240.20

417.63
346.68

Tunisia

16

97

Mean
S.D.

1392.41
1037.02

1006.25
770.22

134.80
100.57

9.63
13.48

Turkey

40

206

Mean
S.D.

8281.33
12701.70

2878.80
4385.23

967.67
1607.44

117.69
391.98

UAE

17

111

Mean
S.D.

5019.93
5881.73

2955.14
3751.53

702.87
695.10

121.73
148.68

Figures in US$ million. Source: Authors calculations based on data retrieved from the BankScope database

Lebanon and Turkey. Saudi Arabia has the largest average total assets over the period under study
in the region, followed by Turkey and Kuwait. Except for Bahrain, all banking sectors in the
Gulf Cooperation Council region have achieved the highest average profitability over the period
20002006 (with an average ROA exceeding 2.40%), followed by Turkey and Jordan, with the
lowest record of average ROA close to 0.30% in Algeria.
The estimation results of Eqs. (1) and (2) are reported in Tables 4 and 5, respectively. Table 4
shows the H-statistic for each country derived from individual regressions using bank effects
(fixed or random depending on the results of the Hausman test), and the H-statistic calculated
from a pooled regression on all countries estimated with country and bank fixed effects. The
results of equilibrium tests for all markets shown in Table 5 indicate that all countries are in an
equilibrium state. The calculated E-statistic is not significantly different from 0 at the 95% level
for all countries considered and for the pooled regression as well. The banking markets are thus in
long-run equilibrium and the PR H-statistic indicates that banks in the MENA region are operating
as a whole under conditions of monopolistic competition.
In Table 4, the coefficients on input prices WF and WL are generally positive and highly
significant, except for WFC which is only significant for Turkey and for the pooled regression. As

702

Table 4
H-statistics of banking systems in the MENA region
Algeriaa

Bahrainb

Jordana

Kuwaita

Lebanona

Moroccoa

Omanb

ln(WF )
ln(WL )
ln(WFC )
ln(equity/assets)
ln(loans/assets)
ln(total assets)
Constant

0.068 (0.061)
0.065 (0.192)
0.039 (0.082)
0.114 (0.129)
0.351 (0.179)*
0.031 (0.048)
0.972 (0.600)

0.241 (0.102)**
0.518 (0.093)***
0.019 (0.034)
0.153 (0.070)*
0.081 (0.044)*
0.06 (0.039)
0.997 (0.473)*

0.165 (0.035)***
0.107 (0.079)
0.017 (0.044)
0.067 (0.058)
0.19 (0.116)
0.032 (0.022)
0.712 (0.524)

0.439 (0.075)***
0.133 (0.149)
0.032 (0.040)
0.518 (0.126)**
0.403 (0.155)*
0.169 (0.070)*
0.285 (1.079)

0.421 (0.061)***
0.118 (0.039)***
0.001 (0.025)
0.087 (0.034)**
0.008 (0.035)
0.015 (0.013)
0.216 (0.217)

0.019 (0.091)
0.352 (0.146)**
0.062 (0.053)
0.139 (0.074)*
0.077 (0.054)
0.022 (0.040)
0.258 (0.753)

0.136 (0.025)***
0.158 (0.091)
0.027 (0.049)
0.202 (0.058)**
0.001 (0.129)
0.089 (0.031)**
0.723 (0.363)*

Number of obsvervations
R2

22
0.60

44
0.77

73
0.53

34
0.81

226
0.62

63
0.74

38
0.61

PR H-statistic

0.172 (0.161)

0.778 (0.125)

0.289 (0.110)

0.604 (0.143)

0.540 (0.086)

0.270 (0.170)

0.267 (0.086)

Monopoly H = 0
Perf. Comp. H = 1

Fail to reject
Reject***

Reject***
Fail to reject

Reject**
Reject***

Reject**
Reject**

Reject***
Reject***

Fail to reject
Reject***

Reject**
Reject***

Ln(TRTA)

Qatara

Saudi Arabiaa

Tunisiab

Turkeya

UAEb

Pooled sample

ln(WF )
ln(WL )
ln(WFC )
ln(equity/assets)
ln(loans/assets)
ln(total assets)
Constant

0.166 (0.087)
0.376 (0.068)***
0.003 (0.064)
0.08 (0.137)
0.336 (0.280)
0.018 (0.038)
0.255 (0.725)

0.238 (0.020)***
0.383 (0.069)***
0.02 (0.024)
0.263 (0.110)**
0.142 (0.075)*
0.088 (0.028)**
0.191 (0.412)

0.566 (0.108)***
0.383 (0.153)**
0.052 (0.078)
0.298 (0.088)***
0.153 (0.098)
0.066 (0.076)
5.39 (1.315)***

0.51 (0.096)***
0.246 (0.087)***
0.072 (0.034)**
0.062 (0.067)
0.052 (0.056)
0.018 (0.021)
0.566 (0.337)

0.24 (0.030)***
0.299 (0.059)***
0.021 (0.030)
0.174 (0.068)**
0.224 (0.069)***
0.01 (0.027)
0.463 (0.219)*

0.313 (0.0292)***
0.310 (0.0399)***
0.037 (0.0124)**
0.024 (0.038)
0.076 (0.060)
0.012 (0.022)
0.308 (0.267)

Number of obsverations
R2

32
0.67

63
0.81

76
0.51

156
0.68

103
0.74

930
0.55

PR H-statistic

0.539 (0.106)

0.602 (0.089)

0.897 (0.202)

0.829 (0.112)

0.561 (0.072)

0.659 (0.027)

Monopoly H = 0
Perf. Comp. H = 1

Reject***
Reject**

Reject***
Reject***

Fail to reject
Reject***

Reject***
Fail to reject

Reject***
Reject***

Reject***
Reject***

*, ** and *** significant at 10%, 5% and 1%, respectively.

a Country regressions are run using bank fixed effects and bank random effects, respectively based on the results of the Hausman test. Pooled regression is run with country
and bank fixed effects based on the results of the Hausman test. Robust standard errors are reported in parentheses.
b Country regressions are run using bank fixed effects and bank random effects, respectively based on the results of the Hausman test. Pooled regression is run with country
and bank fixed effects based on the results of the Hausman test. Robust standard errors are reported in parentheses.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

Ln(TRTA)

Table 5
Long run E-statistics of banking systems in the MENA region
Algeriaa

Bahrainb

Jordana

Kuwaita

Lebanona

Moroccoa

Omanb

ln(WF )
ln(WL )
ln(WFC )
ln(equity/assets)
ln(loans/assets)
ln(total assets)
Constant

0.005 (0.00)
0.014 (0.007)*
0.000 (0.00)
0.004 (0.00)
0.020 (0.009)*
0.002 (0.01)
0.106 (0.047)*

0.001 (0.00)
0.008 (0.01)
0.005 (0.002)*
0.072 (0.016)***
0.006 (0.01)
0.015 (0.01)
0.002 (0.05)

0.004 (0.002)*
0.005 (0.002)*
0.003 (0.001)***
0.014 (0.002)***
0.013 (0.006)*
0.003 (0.00)
0.040 (0.019)*

0.001 (0.00)
0.010 (0.01)
0.000 (0.00)
0.023 (0.009)*
0.005 (0.01)
0.013 (0.004)**
0.015 (0.05)

0.002 (0.00)
0.001 (0.00)
0.000 (0.00)
0.002 (0.00)
0.006 (0.001)***
0.005 (0.001)***
0.014 (0.01)

0.007 (0.002)***
0.032 (0.005)***
0.005 (0.00)
0.026 (0.010)**
0.002 (0.01)
0.002 (0.00)
0.059 (0.023)**

0.011 (0.005)*
0.048 (0.03)
0.008 (0.02)
0.078 (0.025)**
0.117 (0.11)
0.013 (0.03)
0.132 (0.15)

Number of observations
R2

22
0.71

49
0.71

73
0.77

34
0.74

249
0.14

90
0.61

38
0.46

LR E-statistic

0.019 (0.01)

LR equilibrium E = 0

Fail to

rejectc

0.019 (0.01)
Fail to

0.004 (0.00)

rejectd

Fail to

0.011 (0.01)

rejectd

Fail to

0.001 (0.00)

rejectd

Fail to

rejectd

0.020 (0.02)

0.051 (0.03)

rejectd

Fail to rejectd

Fail to

ROA

Qatara

Saudi Arabiaa

Tunisiab

Turkeya

UAEb

Pooled sample

ln(WF )
ln(WL )
ln(WFC )
ln(equity/assets)
ln(loans/assets)
ln(total assets)
Constant

0.002 (0.00)
0.044 (0.019)*
0.001 (0.00)
0.044 (0.020)*
0.022 (0.02)
0.005 (0.01)
0.292 (0.144)*

0.002 (0.00)
0.075 (0.022)**
0.000 (0.00)
0.017 (0.002)***
0.005 (0.01)
0.034 (0.00 7)***
0.110 (0.08)

0.024 (0.02)
0.042 (0.016)**
0.004 (0.00)
0.035 (0.017)*
0.023 (0.03)
0.008 (0.003)**
0.090 (0.08)

0.016 (0.01)
0.054 (0.03)
0.001 (0.00)
0.043 (0.015)**
0.009 (0.01)
0.009 (0.01)
0.214 (0.15)

0.000 (0.00)
0.014 (0.01)
0.003 (0.00)
0.020 (0.008)**
0.004 (0.01)
0.009 (0.01)
0.051 (0.03)

0.007 (0.01)
0.026 (0.01)
0.001 (0.00)
0.031 (0.008)***
0.013 (0.007)*
0.003 (0.00)
0.066 (0.05)

Number of observations
R2

32
0.57

63
0.81

97
0.4

206
0.31

103
0.29

1056
0.21

LR E-statistic

0.041 (0.02)

LR equilibrium E = 0

Fail to

rejectc

0.077 (0.03)
Fail to

rejectc

0.062 (0.03)
Fail to

rejectc

0.069 (0.04)
Fail to

rejectd

0.016 (0.01)
Fail to

rejectd

0.034 (0.02)
Fail to rejectd

703

*, ** and *** significant at 10%, 5% and 1% levels, respectively.

a Country regressions are run using bank fixed effects and bank random effects, respectively based on the results of the Hausman test. Pooled regression is run with country
and bank fixed effects based on the results of the Hausman test. Robust standard errors are reported in parentheses.
b Country regressions are run using bank fixed effects and bank random effects, respectively based on the results of the Hausman test. Pooled regression is run with country
and bank fixed effects based on the results of the Hausman test. Robust standard errors are reported in parentheses.
c Refers to the 5% level.
d Refers to the 10% level.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

ROA

704

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

expected, the coefficients on the loans to assets and equity to assets ratios are generally positive
in most countries. The sign on total assets, however, varies by country. To illustrate, while it is
positive and significant in the case of Kuwait and Saudi Arabia indicating positive scale effects,
it is significantly negative in Oman suggesting that banks face diseconomies of scale.
The H-statistic calculated from the pooled regression (H = 0.659) indicates that banks in the
MENA region generally operate under conditions of monopolistic competition; however, a closer
look at individual countries shows different patterns among them. Specifically, tests of hypotheses
for all countries of North Africa (Algeria, Morocco and Tunisia) fail to reject a monopolistic environment in the banking industry, while the test results fail to reject perfectly competitive markets
in Bahrain and Turkey.12 All other banks in the GCC and Middle East operate under conditions
of monopolistic competition. A monopolistic competition structure is, a priori, consistent with
the existence of product differentiation in banking and with the fact that banks tend to differ with
respect to product quality and advertising, although their core business is fairly homogeneous. It is
interesting to highlight the fact that, while banking sectors in the MENA region are traditionally
highly concentrated markets, their markets structure is generally monopolistically competitive
and not characterized as a monopoly except for countries in North Africa. The empirical findings
reveal that market power resulting from high concentration levels does not exclude competitive
behavior. This suggests that other factors may account for differences in the degree of competition
in banking.
5. Factors explaining differences in the degree of competition
This section uses second stage regression models to examine the factors that affect differences
in the degree of competition across the banking sectors in the MENA region. The sample consists
of panel data for the individual profiled countries over the period 20002006 and includes 84
observations. Different models are estimated following Eq. (3):
Hi = a + bCi + ei

(3)

The dependent variable is the H-statistics or the degree of competition derived from the stage
one regressions in each of the 12 MENA countries for the period 20002006. Alternatively, since
the H-statistic is a long-term variable that should not vary in time, the regressions are also run
the country-level estimated H-statistic for robustness. Ci is a vector of variables that are classified
into two categories, including industry control factors and indicators of banking structures. The
variables are obtained from Barth, Caprio, and Levine (2007) and the Economist Intelligence Unit
Country Profiles. A definition and description of the variables can be found in Table 6.
Industry control variables include three variables, all of which are calculated from the
BankScope database; efficiency, profitability and capitalization levels. Each industry control variable is calculated at the bank level and then averaged per year (for use in the year level regressions)
and per country (for use in the country level regressions). Efficiency is measured as the ratio of
non-interest expense to the sum of net interest income and non-interest revenues, with higher
values implying less efficiency levels; profitability is calculated as the return on assets and return
on equity; and capitalization is proxied by the ratio of equity to total assets. In line with classical industrial organization economics, banks not exposed to competition are able to exercise
12 Over the past two decades, Bahrain has attracted many foreign banks to operate as off-shore banking units. As a result,
several regional and international banks started operating in the country, thus putting more pressure on the prevailing
competitive conditions.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

705

Table 6
Description and definition of variables
Variable name

Description

H-statistic

Panzar and Rosse (1987) H-statistic calculated for every country over the sample
period 20002006
Source: Authors calculations using bank-level financial data from Fitch-IBCA
BankScope database for 20002006
Ratio of non-interest expense to the sum of net interest income and non-interest
revenues. A higher ratio indicates lower efficiency
Source: Authors calculations using bank-level financial data from Fitch-IBCA
BankScope database for 20002006.
Return on assets (the ratio of net income to total assets) and return on equity (the ratio
of net income to total equity)
Source: Authors calculations using bank-level financial data from Fitch-IBCA
BankScope database for 20002006.
Ratio of equity to total assets
Source: Authors calculations using bank-level financial data from Fitch-IBCA
BankScope database for 20002006
A measure of the degree of concentration in the banking industry, calculated as the
fraction of deposits held by the largest three commercial banks every year in each
country. Authors calculations using bank-level financial data from Fitch-IBCA
BankScope database
A measure of the degree of foreign ownership of banks, calculated as the fraction of
the banking systems assets that is in banks that are 50% or more foreign owned as of
year-end 2006. Source: Barth et al. (2007)
A measure of the limits imposed on commercial banks to engage in securities markets,
insurance and real estate business and owning non-financial institutions. A higher
score indicates more restrictions on banks to engage in such activities. Source: Barth
et al. (2007)
Logarithm of per capita GDP expressed in US dollars for the period 20002006.
Source: Economist Intelligence Unit Country Profiles

Efficiency

Profitability

Capitalization

Concentration

Foreign ownershipa

Activity restrictionsa

Per capita GDP

a As in other studies, data on market structure in banking is assumed to be relatively stable and it typically refers to
year-end 2006.

monopoly power and tend to be less efficient than banks subject to more competition (Berger &
Hannan, 1998), so that a negative relationship can be expected between the efficiency and the
H-statistic. With respect to bank profitability, it can be argued that banks are able to command
higher profitability ratios when operating under more monopolistic conditions and a negative sign
is expected on the coefficient of profitability in Eq. (3). Whether better capitalized banks operate
in more competitive environments is not clear cut, and there is no expectation for the sign of the
relationship between bank capital and the degree of competition.
Indicators of banking structures include bank concentration, and two proxies of market contestability or foreign ownership and activity restrictions. Bank concentration refers the total assets
and total deposits three-bank concentration levels. Is the traditional measure of bank concentration
a good indicator of the degree of competition in MENA banking? Is there evidence to suggest
that concentration in banking encourages monopoly practices and price-setting behavior? It is
expected that more bank concentration is associated with a lower degree of competition in the
industry. With respect to foreign ownership, it is calculated as the fraction of the banking systems
assets that are 50% or more foreign owned. Foreign banks can play an important role in improving the functioning of the payments system, mobilizing domestic savings and improving resource

706

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

allocation, thus modernizing and increasing the efficiency of the financial system. Levine (1997)
suggests that foreign banks help promote capital inflows and competition. It is hypothesized that
foreign presence in the banking sector improves competition and a positive sign is expected in Eq.
(3). The last explanatory variable, activity restrictions, is included to reflect the limits imposed
on commercial banks for engaging in non-traditional banking activities. Banks may or may not
be allowed to own and control non-financial firms, or to engage in securities or insurance underwriting and selling, or real estate investments. The variable is taken from the survey conducted
by Barth et al. (2007) with a higher score indicating more restrictions on banks to engage in such
activities. Its parameter estimate is expected to be negative since more open financial sectors with
limited restrictions are favorable to a more competitive environment. Similar to other studies,
foreign ownership and activity restrictions are assumed to be relatively stable over the period
under study (e.g. Beck et al., 2003; Claessens & Laeven, 2004). All regression models include the
natural logarithm of per capita GDP where a positive association is predicted between the level
of economic development and the degree of competition in banking. The estimation results are
reported in Table 7.
Panel A considers the yearly derived H-statistic as dependent variable, and panel B uses the
country level H-statistic instead for robustness. Models 1 and 2 examine the impact of each of
industry controls and banking structures separately on the degree of competitiveness and model
3 includes all explanatory variables. The findings generally show a pattern of consistency across
all estimated models.
The sign of the industry control variables is negative across all models. As predicted, banks
that operate in more monopolistic environments are also less efficient, consistent with the theory
of classical industrial organization. This finding is opposed to Casu and Girardone (2006) who
do not find evidence that more competitive markets are also more efficient. The negative sign on
profitability suggests that banks which are able to achieve high records of profitability operate
in more monopolistic environments. As for bank capitalization, its coefficient is significantly
negative in panel A, indicating that better capitalized banks who probably have high franchise
value operate in less competitive environments.
In line with the results obtained by Bikker and Haaf (2002) but opposite to the findings of
Claessens and Laeven (2004) who report a positive association and Casu and Girardone (2006)
who find a negative but not significant relationship between concentration and competition
the sign of the estimated parameter of concentration is significantly negative across all models.
This indicates that the degree of competition decreases with increases in market concentration
across MENA banking sectors. It seems that highly concentrated banking systems in the MENA
region result in more market power for leading firms. Caution, however, should be exercised in
interpreting the results since prior research did conclude that concentration may be a less robust
measure of competition compared to other contestability factors (Beck et al., 2003; Martinez et al.,
2004). According to Shaffer (2004a), if a factor is reported to be positively associated with larger
values of the H-statistic, one cannot reliably conclude that banks are behaving more competitively.
He argues that more robust tests of association are necessary to establish the relationship between
concentration and competition.
The two proxies of market contestability, foreign ownership and activity restrictions, have
the predicted positive and negative signs, respectively across all specifications. The theory of
contestable markets assumes that firms can freely enter and exit any market, and that potential
competitors have the same cost function as existing market participants (Baumol, Panzar, &
Willing, 1982). If these conditions prevail, the threat of entry enforces marginal cost pricing
and guarantees non-collusive behavior within a market. The empirical results indicate that cross-

Variable

Panel A: yearly H-statistic

Model 1

Efficiencya
Profitabilitya
Capitalizationa
Concentration
Foreign ownership
Activity restriction
Economic developmenta
Constant

1.5516 (0.5217)**
0.5653 (0.2439)*
4.0948 (1.7958)*

0.2403 (0.1110)*
0.9864 (0.8287)

Panel B: country level H-statistic

Model 2

Model 3

Model 1
2.2042 (0.6544)**
0.3067 (0.1455)*
0.6167 (0.6672)

0.7763 (0.3060)**
0.6566 (0.3566)*
0.3794 (0.2135)*
0.0995 (0.0361)**
0.4201 (0.5621)

0.1169 (0.1537)
0.3043 (0.2722)
5.6302 (2.6829)*
0.4575 (0.2183)*
0.7081 (0.2699)**
0.5767 (0.2304)**
0.2354 (0.0869)**
0.4079 (0.7371)

0.0863 (0.0628)
0.1575 (0.6004)

Model 2

Model 3

0.7763 (0.4631)*
0.5013 (0.0044)***
0.3292 (0.0082)***
0.1169 (0.0042)***
0.0017 (0.0610)

1.8983 (0.5085)***
0.2242 (0.1559)
0.3817 (0.7465)
0.1893 (0.09)**
0.435 (0.0891)***
0.2427 (0.0505)***
0.0668 (0.0438)
0.7163 (0.3801)

The dependent variable is the PR H-statistic derived for every country on a yearly basis (panel A) or on a pooled basis (panel B). Efficiency (higher values imply less efficiency),
profitability, capitalization and concentration are calculated from the BankScope database; foreign ownership and activity restrictions are obtained from Barth, Caprio, and
Levine (2007 Database) with higher figures implying more foreign ownership and more activity restrictions; economic development is obtained from the International Financial
Statistics database and is measured by the natural logarithm of GDP per capita. Robust errors are reported in parentheses. The sample consists of panel data for the 12 individual
profiled countries over the period 20002006 and includes 84 observations. *, ** and *** significant at 10%, 5% and 1% levels, respectively.
a Figures are averaged per year for regressions of panel A, and by country for regressions for panel B.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

Table 7
Regression results

707

708

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

country variations in the degree of bank competition are significantly positively related to the
level of foreign bank ownership in the market. This suggests that financial liberalization and more
openness of the local markets act to reduce and counterbalance the monopolistic pressures that
could result from the ongoing wave of bank consolidation. This finding is in line with the results
reported in the literature for most mature and emerging markets (Bikker & Haaf, 2002; Claessens
& Laeven, 2004; Gelos & Roldos, 2004), but it contrasts to Yeyati and Micco (2007) who report
that foreign penetration appears to have led to less competitive banking sectors in Latin America.
Similarly, the level of activity restrictions imposed on banks is significantly negatively related to
the degree of competition in banking across the MENA region. The greater the relative openness
of the banking sector and the fewer the limits imposed on banks to engage in non-traditional
banking activities, the more competitive the banking market is.
Finally, the results reveal a difference in the degree of competition in banking across the MENA
region for different levels of economic development. As predicted, a higher per capita GDP is
found to be positively associated with greater competitiveness scores.
6. Conclusions
This paper examines competitive structures in 12 highly concentrated MENA banking sectors
over the period 20002006 and investigates factors that can explain differences in the degree of
competitiveness. The PR methodology is used to account for the competitive nature of banking and
the estimated measure of competition, the H-statistic, is related to a number of industry controls
and prevailing banking structures. To my knowledge, no econometric analysis of the degree of
competitiveness in MENA banking was conducted before.
The results show that, except for countries of North Africa where monopolistic conditions
cannot be ruled out, the market structure in MENA banking systems can best be described as
monopolistically competitive. While concentration levels have historically been high in banking
in the region, there is cross-country evidence that MENA banking markets are contestable. Foreign
competition (or the threat thereof) moderates any eventual decline in competition resulting from
consolidation. Fewer activity restrictions reflected in more open banking systems with few
restrictions on entry and branching add to the competitive pressures in banking. This suggests that
assuring greater market contestability by allowing more foreign bank participation and reducing
activity restriction on banks is most important to guaranteeing competitiveness in MENA banking
systems.
Most MENA countries are committed to financial liberalization. Banks are increasingly
required to adopt international accounting standards and prudential guidelines of capitalization
and governance, while foreign banks are encouraged to become active players in domestic markets. The results convey an important message to regulators concerned with the traditional high
concentration levels observed in MENA banking sectors. Specifically, the analysis lends support
to policy decisions to develop contestable markets to assure competitiveness in banking.
References
Al-Muharrami, S., Matthews, K., & Khabari, Y. (2006). Market structure and competitive conditions in the Arab GCC
banking system. Journal of Banking & Finance, 30(12), 34873501.
Barth, J. R., Caprio, G., Jr., & Levine, R. (2004). Bank regulation and supervision: What works best? Journal of Financial
Intermediation, 13, 205248.
Barth, J. R., Caprio, G. Jr., & Levine, R. (2007). The Regulation and Supervision of Banks Around the World: A New
Database. World Bank Policy Research Working Paper No. WPS 2588.

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

709

Baumol, W. J., Panzar, J. C., & Willing, R. D. (1982). Contestable markets and the theory of industry structure. San
Diego: Harcourt Brace Jovanovich.
Beck, T., Demirguc-Kunt, A., & Levine, R. (2003). Bank Concentration and Crises. World Bank Working Paper.
Belaisch, A. (2003). Do Brazilian Banks Compete?, IMF Working Paper No. 03/113 (Washington: International Monetary
Fund).
Berger, A. N., & Hannan, T. H. (1989). The price-concentration relationship in banking. Review of Economics and
Statistics, 71, 291299.
Berger, A. N., Demsetz, R. S., & Strahan, P. E. (1999). The consolidation of the financial services industry: Causes,
consequences, and implications for the future. Journal of Banking and Finance, 23, 135194.
Berger, A. N., DeYoung, R., Genay, H., & Udell, G. F. (2000). The globalization of financial institutions: Evidence from
cross-border banking performance. Brookings-Wharton Papers on Financial Services, 3, 23158.
Bikker, J. A., & Groeneveld, J. M. (2000). Competition and concentration in the EU banking industry. Kredit and Kapital,
33, 6298.
Bikker, J. A., & Haaf, K. (2002). Competition, concentration and their relationship: An empirical analysis of the banking
industry. Journal of Banking and Finance, 26, 21912214.
Bresnahan, T. F. (1982). The oligopoly solution concept is identified. Economic Letters, 10, 8792.
Buchs, T., & Mathisen J. (2005). Competition and Efciency in Banking: Behavioral Evidence from Ghana, IMF Working
Paper No. 05/17 (Washington: International Monetary Fund).
Burdisso, T., Catena, M., DAmato, L. (2001). Bank Competition in Argentina. Central Bank of Argentina Working Paper
No. 20.
Casu, B., & Girardone, C. (2006). Bank competition and efficiency in the single European market. The Manchester School,
74(4), 441468.
Claessens, S., & Laeven, L. (2004). What drives bank competition? Some international evidence. Journal of Money,
Credit, and Banking, 36, 563583.
Coccorese, P. (1998). Assessing the competitive conditions in the Italian banking system: some empirical evidence. BNL
Quarterly Revue, 205, 171191.
De Bandt, O., & Davis, E. P. (2000). Competition, contestability and market structure in European banking sectors on the
eve of EMU. Journal of Banking and Finance, 24, 10451066.
DeYoung, R., & Nolle, D. (1996). Foreign-owned banks in the U.S.: Earning market share or buying it? Journal of Money,
Credit and Banking, 28(4), 622636.
Economist Intelligence Unit (various years), Country Profiles for MENA countries.
Friedman, J. W. (1971). A Non-cooperative equilibrium for supergames. Review of Economic Studies, 38, 112.
Gelos, R. G., & Roldos, J. (2004). Consolidation and market structure in emerging market banking systems. Emerging
Markets Review, 5, 3959.
Hannan, T. H., & Berger, A. N. (1991). The rigidity of prices: Evidence from the banking industry. American Economic
Review, 81, 938945.
Hauner, D. Peiris, S. (December, 2005). Bank Efciency and Competition in Low-Income Countries: The Case of Uganda,
IMF Working Paper No. 05/240, International Monetary Fund.
Lau, L. J. (1982). On identifying the degree of competitiveness from industry price and output data. Economics Letters,
10, 9399.
Levine, R. (1997). Financial development and economic growth: views and agenda. Journal of Economic Literature, 35,
688726.
Matthews, Murinde, K. V., & Zhao, T. (2007). Competitive conditions among the major British banks. Journal of Banking
& Finance, 31(7), 20252042.
Martinez, P., Soledad, M., & Mody, A. (2004, June). How foreign participation and market concentration impact bank
spreads: Evidence from Latin America. Journal of Money, Credit and Banking, 36-3, 511537.
Molyneux, P., Lloyd-Williams, D. M., & Thornton, J. (1994). Competitive conditions in European banking. Journal of
Banking and Finance, 18, 445459.
Molyneux, P., Thornton, J., & Lloyd-Williams, D. M. (1996). Competition and market contestability in Japanese commercial banking. Journal of Economics and Business, 48, 3345.
Nathan, A., & Neave, E. H. (1989). Competition and contestability in Canadas financial system: Empirical results.
Canadian Journal of Economics, 22, 576594.
Neumark, D., & Sharpe, S. A. (1992). Market structure and the nature of price rigidity: Evidence from the market for
consumer deposits. Quarterly Journal of Economics, 107, 657680.
Panzar, J. C., & Rosse, J. N. (1987). Testing for Monopoly equilibrium. Journal of Industrial Economics, 35,
443456.

710

R. Turk-Ariss / The Quarterly Review of Economics and Finance 49 (2009) 693710

Rime, B. (1999). Mesure du degre de concurrence dans le syst`eme bancaire Suisse a` laide du mod`ele de Panzar et Rosse.
Revue Suisse DEconomie Politique et de Statistique, 135(1), 2140.
Shaffer, S. (1982). A non-structural test for competition in financial markets. In Bank structure and competition Conference
Proceedings, Federal Reserve Bank of Chicago (pp. 225243).
Shaffer, S. (2004a). Comment on What drives bank competition? Some international evidence by Stijn Claessens and
Luc Laeven. Journal of Money, Credit, and Banking, 2, 36(3), 585592.
Shaffer, S. (2004b). Patterns of competition in banking. Journal of Economics and Business, 56(4), 287313.
Shepherd, W. G. (1986). Tobins q and the structureperformance relationship: Comment. American Economic Review,
76, 12051210.
Smirlock, M., Gilligan, T., & Marshall, W. (1984). Tobins q and the structureperformance relationship. American
Economic Review, 74, 10511060.
Smirlock, M. (1985). Evidence on the (non) relationship between concentration and profitability in banking. Journal of
Money, Credit, and Banking, 17, 6983.
Staikouras, C., & Koutsomanoli-Fillipaki, A. (2006). Competition and concentration in the new European banking
landscape. European Financial Management, 12(3), 443482.
Trivieri, F. (2007). Does cross-ownership affect competition? Evidence from the Italian banking industry. Journal of
International Financial Markets, Institutions and Money, 17(1), 79101.
Vesala, J. (1995). Testing for competition in banking: Behavioural evidence from Finland. Helsinki: Bank of Finland
Studies.
Utero-Gonzales, N. (2004). Banking competition in UK and Spain: Is there a Euro effect? In FMA European Conference.
Weill, L. (2004). On the relationship between competition and efficiency in the EU banking sector. Kredit und Kapital,
37(3), 329352.
Yeyati, E. L., & Micco, A. (2007). Concentration and foreign penetration in Latin American banking sector: Impact on
competition and risk. Journal of Banking and Finance, 31(6), 16331647.
Yuan, Y. (2006). The state of competition of the Chinese banking industry. Journal of Asian Economics, 17(3), 519534.