Aaltodoc publication archive (Aalto University institutional repository)
School of Business | Department of Economics | Economics | 2009
Thesis number: 12160
Competition in the Finnish banking sector - the case of cooperative banks
|Title:||Competition in the Finnish banking sector - the case of cooperative banks|
|Year:||2009 Language: eng|
|Department:||Department of Economics|
|Index terms:||kansantaloustiede; economics; kansantalous; national economy; pankit; banks; kilpailu; competition; osuuspankit; cooperative banks|
|Key terms:||competition, banking sector, Panzar-Rosse model, cooperative banks|
This study has two research questions which it aimed to answer. The first question is under which competitive conditions did member cooperative banks of the OP-Pohjola Group operate in 2000-2005? The second question is how does the existence of competitor branches in the same area affect the competitive conditions of the cooperative banks? Competition was measured with the Panzar-Rosse test which estimates an H-statistic that serves as a measure of competition. The test is derived from a general banking model that maximises profits at the bank and the industry level. The H-statistic is defined as the sum of a bank's elasticities of the reduced form revenues with respect to factor prices. The possible forms of competition in the model are monopoly, oligopolistic competition, perfect competition and monopolistic competition which are indicated by different values of H.
The data set used in this study was provided by the Finland's largest cooperative banking group OP-Pohjola Group and it includes information on income, costs and profits of 240 regional cooperative banks in the end of respective years 2000-2005. The test equation used total revenues to total assets as dependent variable and the estimation was performed both with fixed and random effects estimation methods since the Hausman test was not entirely conclusive. The results of both estimations suggested that the banks were operating under monopolistic competition in 2000-2005.
There are two reasons by which we can at least partly explain the relatively high degree of market power of cooperative banks indicated by the H-statistic. First, the Finnish banking sector has been dominated be three banking groups since 1999. Second, cooperative banks have disproportionately large branch networks compared to commercial banks which is one of the reasons for cooperative banks’ strong retail market positions. The large branch networks of cooperatives are due to the fact that cooperative banks maximise their members’ consumer surplus instead of profits. Therefore, cooperative banks are satisfied with smaller return on equity rates and can pursue a resource-intensive operating model. Consequently, cooperative banks often operate in areas where there are no competitors or only a few other banks' branches present. Previous studies also show that cooperative banks enjoy higher levels of market power than their commercial peers.
In order to answer the second research question, the sample was divided into three groups both by number of competitors' branches and by number of competitors' branches per 1000 inhabitants in the same area. When doing the estimations by number of competitor branches, the data set was divided into three groups with Group 1 consisting of cooperatives with zero competitor branches in the same area, Group 2 including cooperative banks with 1-3 competitor branches and Group 3 including cooperatives with 4 or more competitor branches. Similar estimations were run by number of competitor branches per 1000 inhabitants in the same area. Group 1 again consisted of banks with zero competitor branches. Groups 2 and 3 were formed by dividing the remaining banks to approximately similarly sized groups.
The results of the estimations by group were very similar in both types of group divisions. The results were consistent with monopolistic competition in all groups. Group 2 had the highest H-statistic in both estimations indicating the highest level of competition among the groups. Group 3 has the lowest H-statistic in both fixed effects estimations indicating lowest competitions among the groups. At the same time, tests of no difference between the H-statistics between groups could not be rejected. The only exception was the fixed effects estimation by number of competitor branches where the hypothesis of no difference between the H-statistics was rejected for Groups 2 and 3. Therefore, the tests do not support a hypothesis of a higher (relative) number of competitor branches leading to higher levels of competition. Also, the results suggest that the presence and number of competitor branches does not have an effect on the competitive stance of the cooperative banks. This result may relate to the fact that the number of competitor branches is relatively low in all areas in Finland.
Master's theses are stored at Learning Centre in Otaniemi.