Abstract

Significant parts of the financial sector need to have sufficient capital to fulfill the basic functions of the banks. Regardless of the capital source, all banks have international standards that they must comply with in terms of capital adequacy. Provision of these standards leads to similar characteristics according to the capital adequacy of the banks. It may not always be possible to group banks with similar characteristics with exact divisions. At this point, the fuzzy clustering method, which allows banks to be assigned to different groups with specific membership grades, is at the forefront. Classical and fuzzy clustering approaches are used comparatively in this study which aims to group banks by capital adequacy ratios. In the study, 46 banks were grouped by Ward, K-Average and Fuzzy C-Average methods according to capital adequacy ratios for 2015. As a result, three sets of similar structures were obtained for each method. When the cluster structures were examined, it was observed that the clusters were heterogeneous in terms of capital resources. However, since the capital adequacy is taken into consideration instead of the capital resources while the grouping is carried out, it is a common result that the obtained clusters contain different types of banks. When membership ratings are examined, it is observed that the rest of the few except a few banks generally have a high membership level for a cluster. However, the membership grades for all the clusters of some banks are close to each other. Therefore, it is concluded that the clustering situation of these banks is more blurred than the other banks.