THE IMPACT OF INDUSTRY CONCENTRATION ON PERFORMANCE, EXPLORING A COMPREHENSIVE BANK PERFORMANCE MODEL: THE CASE OF THE ETHIOPIAN BANKING SECTOR

摘要

There are wide empirical studies and literatures that advocate for a strong link between financial sector development and economy growth. Nevertheless, similar to the situation in Sub-Saharan Africa, the share of the financial sector in the overall economy of Ethiopia is at lower level and can be referred as a shallow financial market. In Ethiopia, reform measures to enhance the performance of this sector through liberalization measures as well as economic plans, Growth and Transformation Plan I & II (GTP I& II) that set bank growth strategies were enacted in an attempt to augment their contribution to the economy. Despite such reforms and policy attentions the banking system in Ethiopia is characterized by high concentration with an observed dominance of the state owned banks. Nonetheless, the lack of indisputably conclusive and substantial disagreement in empirical results, the Structure- Conduct- Performance (SCP) theorists claim that such kind of industry structure influences the conduct and performance of firms. The basic objective of this research is to test the validity of this view. A SCP analysis has been undertaken so as to examine whether the level of concentration influences performance of Ethiopian Banks. In addition, it aims at investigating the impact of other non-structure factors on performances with an overarching purpose of exploring a comprehensive model of bank performances. The study employed an explanatory sequential mixed method approach that combined both quantitative and qualitative studies to explore the central and subsidiary questions. A quantitative approach is adopted to form a causal link among different variables with bank performance measures. A panel data set from 1999 to 2015 for all (eighteen) commercial banks is used for the quantitative study. The quantitative study employs a two-stage estimation procedure to evaluate the impact of bank concentration on performance. In the first stage of the estimation process, a Data Envelopment Analysis (DEA) is employed to produce an efficiency estimates. The output of the DEA is later used as an input in the second stage of the estimation procedure, where a panel data regression model was employed to investigate the relationship between efficiency, concentration and other factors on profit and price performances. A qualitative approach that gathers data through semi-structured interviews with bank managers and regulatory staff is also used to justify the established relationship in the quantitative study as well as investigate the conduct of banks under given industry structure and banking environment. The result of the integrated approaches rejects the structure performance hypothesis and supports the scale efficiency version of the efficiency hypothesis. Therefore, better bank performance is derived from efficient operation rather than from collusive power of the banks in the industry. Theoretically as well it finds that bank collusion is not an easily pursued strategy among banks as long as there is a sizeable difference in bank size and ownership structure. Besides, there is a notable variation in their mission guiding their business motive. The test for efficiency variation through both parametric and non-parametric tests confirms that there is widely noted efficiency variation among banks operating in the country of which the state owned banks are consistently on the top of the frontier. The quantitative study whose result witnessed a different behavior of banks than the one suggested by the SCP hypotheses also finds that bank conduct is not necessarily a derivate of the industry structure and is shaped by several factors from internal and external environment. Most importantly, regulation is found to have a significant role in shaping the behavior of banks in the market. The study also finds that bank specific external and regulatory factors also explain the variation in performance of banks. Variables like ensuring income diversification, building resilient capital and liquidity base, maintaining asset quality, narrowing trade deficit, ensuring market and economy growth among others remained pertinent policy variables that impact bank performances. A separate view on the policy direction under the GTP II shows a mixed result where requirements for branch /agents growth and bill purchases have negative effect on performance while others like loan growth, deposit growth, capital increase and asset quality control have positive implications. With the backdrop of such findings, the study recommends that bank regulators and policy makers should have considerations to the multi-variables in the model in their attempt to design regulatory directives and macroeconomic policies intended to improve bank performance. In addition, the regulatory organ should limit measures that aggravate the concentration of the sector. The recently introduced actions such as merger between state-owned banks need to be carefully considered as it potentially affects the performances of other banks in the system. Banks and regulatory moves should be directed towards improving bank efficiencies and regulatory rigidities in some fronts like bill purchases and branch growth requirements need to be flexed. The study extends the research on industry concentration and performance employing the structural models: (SCP) or Efficiency Hypothesis (ESH), and applying a direct measure of efficiency with extensive panel data set to examine the Ethiopian banking system. Contributing to theory, the study showed how the non-structure factors results in banks to behave in different way and sometimes in different way than the one presumed by the structural theorists. Hence, it rejects the approach to determine the conduct of banks solely from the industry structure as it could lead to wrong generalization. The study has introduced valuable but neglected factors from the previous structure- performance researches. For instance, it has examined the effect of regulation on bank performances and finds that regulatory factors are more powerful influencers of market structure, banks conduct and internal management decisions. Besides, the above stated theoretical contributions, the study makes an addition to the development of mixed methods research in the study of industry concentration through integrating quantitative and qualitative approaches to investigate a research question. The modified performance model developed in the study contributes to forthcoming research works related to industry structure, efficiency and bank performance.