Font Size: 

Financial ratings are used to show the financial health and reliability of banks, companies and individuals. In this research, it is aimed to rank the banks of the Republic of Turkey by using their financial ratios. For this purpose, statistical learning methods were used. It can be difficult to tell which bank is close to going bankrupt by looking at their annual balance sheets. The aim of this study is to propose a rating system that allows to rank banks according to their level of success. The data used in this study were obtained from the website of the Banks Association of Turkey (TBB). Statistical learning methods were used to create an alternative rating model. We accepted the failure of a bank as a failure criterion for 3 consecutive years. Based on this hypothesis, we created a data set that lists the financial data of the banks for the last 3 years, as well as the labels showing whether a bank has failed or not. Using this data, we created classification models with different algorithms and compared them with each other. Since the Linear Discriminant Classification algorithm gave us the best performance, we chose it as the model for our study. Hierarchical clustering was applied to determine the rating of each bank using the scores obtained with the LDA algorithm.