Applying factor, cluster, multidiscriminant, and path analyses for classification of the Lebanese commercial banks

Abstract

The banking sector in Lebanon has survived bravely despite the fact that it has been seriously and unexpectedly shaken several times in the past few years. This study identifies and clusters 53 out of 72 Lebanese Commercial Banks into groups containing similar and homogeneous characteristics, and measures the direct and indirect effects of capital and assets/equity ratios on profit for 100 Arab banks. Its purpose is to analyze, design, implement, and establish a consistent framework of analysis of bank data. To reach this purpose, the researcher uses cross-sectional secondary data to answer five major questions: What are the major characteristics of the selected commercial banks in Lebanon? -- How many factors underlie the Lebanese commercial banks? -- How many groups or clusters underlie the Lebanese commercial banks? -- What are the discriminating ratios that discriminate between the commercial banks in Lebanon? -- What are the direct and indirect effects of capital and assets/equity ratios on profit for 100 Arab Banks? Finally, this research is applicable in three different areas: Methodology, with its multivariate ratio analysis and its design of a recursive system -- Analysis, with its simplified graphical representation -- Conclusion. These mentioned areas will ultimately have three potential users: Policy makers in commercial banks who can use it as a guideline for their bank management -- Bank Control Commission to whom this study presents an outline for further research in the BDL statistics department -- The academic circles that may study an empirical situation rather than a rigid basic research. Undertaking the task of writing about the Lebanese and Arab banks is beset with many difficulties. The main difficulty is the collection and selection of reliable and adequate data which sometimes are not available and up-to-date or cannot be communicated for reasons of banking secrecy. In consequence of the lack of accurate and comprehensive statistics, the researcher relies on secondary information gathered from local and foreign bankers. Personal contacts with some professionals and experts in the field of banking proves to be a great help also. In this study, a model was developed to classify the Lebanese commercial banks and the Arab banks as well, based on their financial characteristics revealed by the financial ratios. Factor analysis was used as a data reduction technique to arrive at a limited number of crucial ratios. Five groups of commercial banks were arrived at by means of cluster analysis in which the financial ratios were the variables. A multidiscriminate model was using financial ratios as independent variables and clusters of 53 banks out of 72 as dependent variables. A recursive system was developed for 100 Arab banks to measure the direct and indirect effects of capital and assets/equity ratios on profit. Results of testing the model were positive for discriminating power, statistical significance, and predictive ability.