بررسی وضعیت رقابت در صنعت بانکداری یونانی

The present paper uses three different New Empirical Industrial Organization (NEIO) approaches (Panzar–Rosse, Bresnahan–Lau and Hall–Roeger models) to investigate competitive conditions in the Greek banking sector over the period 1995–2004. One important event which has taken place in the Greek banking industry, especially after 1998, is a wave of mergers and acquisitions. This study also measures and compares the degree of banking competition in two sub-periods, 1995–1998 and 1999–2004, in order to investigate the effect of mergers and acquisitions on the competitive structure of Greek banking. The empirical results of the three models indicate a shift from competitive to non-competitive conditions when moving from the first to the second sub-period. This finding suggests that mergers and acquisitions have affected the level of completion of Greek banking, rendering the industry less competitive. Furthermore, taking into consideration the negative effects of mergers and acquisitions on technical efficiency and the total factor productivity of Greek banking (Rezitis, 2008), the empirical findings of the present study provide some indications that one of the outcomes of the Greek banks’ merging activities, at least in the short run, might be to attain market power and thus higher profits, rather than higher efficiency and lower costs.

The Greek banking industry operated in an environment heavily controlled and regulated by the Bank of Greece until the mid-1980s. The Bank of Greece and two major state-owned banks, namely the National Bank of Greece and the Commercial Bank of Greece, almost completely dominated the banking industry. Towards the end of the 1980s, the industry gradually moved towards a more deregulated system due to international developments and the need to participate in the Single European Market and European Monetary Union (EMU). Since the beginning of the 1990s, the Greek banking industry has undergone unprecedented changes, caused by the deregulation of international financial markets, the establishment of the economic and monetary union (EMU), and the introduction and advancement of information technology. Some of the main consequences of these changes include the continuously decreasing market share of public banks accompanied by the increasing share of private banks and the recent wave of mergers and acquisitions (M&As) which have taken place in the Greek banking industry (Rezitis, 2008). These events may have vast effects on the competitive structure of the Greek banking industry. Thus, in order to evaluate the implications of these events regarding the competitive structure of the Greek banking industry, the industry's market structure needs to be investigated in order to determine the degree of competition and examine the impact the M&As are likely to have on the market structure and the behavior of banks. The purpose of this paper is to examine the degree of competition within the Greek banking industry during the period 1995–2004. The dataset used covers almost all Greek banks of the period 1995–2004 and has been compiled from the annual bank reports. It should be stated that two important events took place in the Greek banking industry during the study period. The first was the continuously decreasing market share of the state-controlled banks, while the second was the wave of M&As, with the most significant ones occurring after 1998. Thus, the present study aims to shed light on the effect of the M&As activity on the degree of competition in the Greek banking industry by investigating and comparing the degree of competition within it in the two sub-periods: 1995–1998 and 1999–2005. The results of this study will provide some indications to policymakers about the effects of M&As regarding the competitive structure of Greek banking. This information is particularly important in case additional measures are needed to ensure sufficient competition and financial stability in the Greek banking industry, especially in the current financial crisis, which has been affecting the Greek economy since 2008. Early empirical studies on the measurement of banking competition are based on the identification of an inverse relationship between market concentration and competition. Two paradigms, i.e. the structure–contact–performance (SCP) paradigm and the relative efficiency (RE) paradigm, are usually provided to support this relationship (Bain, 1951 and Gilbert, 1984). The SCP paradigm examines whether a highly concentrated market causes collusive behavior among larger banks resulting in market power increase and therefore higher market performance (higher prices and bank profitability), whereas the RE paradigm investigates whether it is the efficiency of larger banks that enables them to earn relatively higher profits because of lower costs and therefore increase their market share in the process. A large number of recent empirical studies still use banking concentration to evaluate the degree of banking competition (see Cetorelli, 1999). Other studies, e.g. Claessens and Laeven (2004), however, indicate that empirical evidence does not support the expected increasing monotonic relationship between market concentration and market power. Furthermore, according to Bikker (2004), concentration indices are increasingly unreliable when the number of banks is small and tend to exaggerate the level of competition in small countries. As a result, the reliance on concentration indices to measure bank competition can lead to measurement problems and misleading inferences. A more recent empirical framework for measuring bank competition is the New Empirical Industrial Organization (NEIO) methodology, which utilizes profit-maximizing comparative static conditions (see the survey of Bresnahan, 1989). Some basic features of NEIO are: (i) firms’ price-cost margins and marginal costs (MC) cannot be directly observed. Consequently, MC is either inferred from firm behavior or it is not measured at all; (ii) individual industries are taken to have important idiosyncrasies. Institutional detail at the industry level is deemed important and comparative statics of variation across industries or markets is seen as immaterial, unless the markets are closely related; (iii) firm and industry conduct are considered as unknown parameters to be estimated and then directly linked to analytical perceptions of firm and industry conduct; and finally (iv) the nature of market power is determined among a set of alternative hypotheses which are considered explicitly, thus the perfect competitive hypothesis is one of the alternatives between which the data can choose (Bresnahan, 1989: 1012). Two alternative NEIO procedures are usually employed in identifying the degree of banking competitiveness. The first uses the Panzar–Rosse H-statistic proposed by Rosse and Panzar (1977) and Panzar and Rosse, 1982 and Panzar and Rosse, 1987 while the second uses the Bresnahan–Lau approach proposed by Lau (1982) and Bresnahan, 1982 and Bresnahan, 1989. More specifically, the Panzar–Rosse approach provides a way to discriminate between different market structures based on the specification of the reduced-form revenue function at the individual bank level, while the Bresnahan–Lau approach obtains an indicator of the market power of banks from the estimation of a simultaneous equations system, e.g. market demand or supply function and a price setting equation. Shaffer (2004) presents a detailed analysis concerning the advantages and disadvantages of the two aforementioned approaches. The present paper applies three different NEIO approaches to empirically evaluate the degree of competition of the Greek banking industry. The first two approaches are the Panzar–Rosse and the Bresnahan–Lau approaches and the third is the Hall–Roeger approach. Note that the Hall–Roeger approach as originally proposed by Hall (1988) and subsequently modified by Roeger (1995) is based on the realization that the traditional Solow residuals should be independent of variation in the log-change of output in the absence of monopoly power. Based on Hindriks, 1999 and Hindriks, 2005, the Hall–Roeger model could be considered to be an NEIO approach because the equation that is estimated can be regarded as a supply relation, since in a way it relates price to marginal cost. The results of the present study provide evidence in support of either the SCP or the RE paradigm for Greek banking. This is because, after taking into consideration the empirical results of Rezitis (2008), which indicate that the effects of M&As on the technical efficiency and total factor productivity of Greek banks are somewhat negative, the present paper's evaluation of the degree of competition in Greek banking provides an indirect test of the SCP and RE paradigms. More specifically, if the empirical results of the present study indicated a lower degree of competition of Greek banking after merging then this would be an indication that the SCP paradigm should be supported over the RE paradigm for the Greek banking sector. It is worth stating that this is, to our knowledge, the first study applying the Hall–Roeger approach to the banking industry. This approach has been widely used in the literature for testing market power in various sectors of the economy and especially the manufacturing sector. In particular, the list of studies using the Hall–Roeger approach includes the studies by Wilhelmsson (2006) for the Swedish food industry; Aldaba (2005) for the Philippine manufacturing industry; Crespi and Gao (2005) for the US rice milling industry; Boyle (2004) for the Irish manufacturing industry; Dobrinsky et al. (2004) for Bulgarian and Hungarian manufacturing firms; Badinger (2004) for 17 sectors (including five service sectors) of a sample of 10 European countries; Ceritoglu (2002) for the Turkish industrial sector; Hindriks (1999) for the Dutch manufacturing industry; Abayasiri-Silva (1999) for the Australian manufacturing industry; and Martins et al. (1996) for the manufacturing sectors of 14 OECD countries. The other two approaches have been widely used in the literature in the area of measuring the degree of competition in banking industries around the world. Many studies have used the Panzar–Rosse approach to investigate competitive conditions and market structure in banking, for example the studies by Casu and Girardone (2006), Staikouras and Koutsomanoli-Fillipaki (2006), Bikker and Haaf, 2002 and Bikker and Haaf, 2000, De Bandt and Davis (2000), and Molyneux et al. (1994) for European banking; Allen and Liu (2007) for Canadian banking; Gutierrez de Rozas (2007) for Spanish banking; Pererera et al. (2006) for South Asian banking; Yuan (2006) for Chinese banking; Prasad and Ghosh (2005) for Italian banking; Buchs and Mathisen (2005) for Ghanaian banking; Claessens and Laeven (2004) for banking in 50 counties around the world; and Hondroyiannis et al. (1999) for Greek banking. There are fewer studies in the literature investigating banking competition using the Bresnahan–Lau approach. An indicative selection of works includes the studies by Brissimis et al. (2008) for the banking sectors of ten newly acceded EU countries; Kubo (2006) for Thai banking; Canhoto (2004) for Portuguese banking; Uchida and Tsutsui (2005) for Japanese banking; Coccorese (2009) and Angelini and Cetorelli (2003) for Italian banking; Bikker and Haaf (2000) for European banking; and Ribbon and Yosha (1999) for Israeli banking. Besides the study by Hondroyiannis et al. (1999), other studies providing empirical evidence for the competitive structure of the Greek banking industry are the studies undertaken by Bikker and Haaf (2002), Claessens and Laeven (2004) and Casu and Girardone (2006). Note, however, that all these studies apply the Panzar–Rosse approach in examining competitiveness in Greek banking. Thus, the present study is the first applying the Bresnahan–Lau and the Hall–Roeger approaches to Greek banking. The remainder of the paper is organized as follows: Section 2 outlines the methodological framework by theoretically presenting each one of the three empirical approaches. The data sample is discussed in Section 3. Section 4 presents model formulation and data variables for each one of the three empirical models, while Section 5 discusses the empirical results obtained. Finally, Section 6 offers a conclusion.

The present paper uses three different NEIO approaches (the Panzar–Rosse, Bresnahan–Lau, and Hall–Roeger models) to assess competitive conditions in the Greek banking sector over the period 1995–2004. Two important events took place in the Greek banking industry during the period under consideration. The first was the continuously decreasing market share of the state-controlled banks and the second was the wave of mergers and acquisitions which occurred especially after 1998. The present study also measures and compares the degree of banking competition between the two sub-periods, i.e. 1995–1998 and 1999–2004, in order to investigate the effect of mergers and acquisitions on the competitive structure of Greek banking. The empirical results of the Panzar–Rosse model indicate that during the whole period (1995–2004) the Greek banks appeared to operate in conditions of perfect competition. When the two sub-periods (1995–1998 and 1999–2004) are considered, the empirical results suggest that Greek banks operated in conditions of perfect competition in the first sub-period (1995–1998) and in conditions of monopolistic completion in the second sub-period (1999–2004). The results of the Bresnahan–Lau and the Hall–Roeger approaches suggest the presence of a non-competitive market structure in the Greek banking industry for the whole time period and a shift from a competitive to a non-competitive market structure from the first to the second sub-period. A possible explanation of the differences in the empirical results obtained in terms of the whole time period between the Panzar–Rosse model and the other two models, is that in the case of the Panzar–Rosse model the competitive conditions of the first sub-period might prevail while the opposite happens in the case of the other two models, i.e. the non-competitive conditions of the second sub-period could dominate. The empirical results of the three models, however, are similar when the two sub-periods are considered and suggest a shift from competitive to non-competitive conditions when moving from the first to the second sub-period. This finding is important in understanding the evolution and development of the market structure of Greek banking when mergers and acquisitions were underway, especially after 1998. In other words, mergers and acquisitions have affected the level of completion of Greek banking, rendering the industry less competitive. Furthermore, taking into consideration the negative effects of mergers and acquisitions on the technical efficiency and total factor productivity of Greek banking (Rezitis, 2008), the empirical findings of the present study supports the SPC paradigm, providing some indications that some of the main aims behind the merging activities of Greek banks, at least in the short run, might be to achieve market power and thus higher profits rather than higher efficiency and lower costs. These findings are particularly important to policymakers, especially under the current financial crisis, in which among governmental actions taken to halt and reverse the current financial distress is the financial support of the banking sector.