Cluster Analysis and Discriminatory in Applied Study on Some Iraqi banks

Abstract

The bank fields is considered an important field in country ,this is becomes it's role in economy and it's effect on investment in this country .So in this study multivariate statistical methods were used, represented by cluster analysis and discriminate analysis for classifying and discerning the extent of homogeneity among banks. these methods were applied on (20) banks , and these are the most important bank in Iraq enrolled in Iraq Stock Exchange . Through the use of cluster analysis we classify banks in two homogenous groups, The first group include (16) banks (Commercial, Iraqi Islamic, Investment, Dijlah & Furat, National, Credit, Dar es salam, Sumer, Babylon, Economy, Gulf, Mosul, Union, Ashur, Mansour, Elaf Islamic ), regarding the second group include (4) banks (Baghdad , North , Kurdistan International , united ).Regarding discernment analysis, it applied to discern most important variables that leading to non homogeneity among concluded groups in cluster analysis. Through this application we predicted that each of variables (Traded share, Total revenues, Turnover ratio, Trading volume, Total expenses, Subscribed shares, Total sources of finance) has the most important value in classification. Further more discernment function was formulated which is used in prediction the importance of variables. And through discernment function we concluded that twenty banks classification was correct with out any error.