Comprehensive Assessment of Insurance Company’s Competitiveness by Cluster and Discriminative Analysis Methods

Abstract

The article considers the application of cluster analysis to the study of the dynamics of insurance companies, with its help it is found to which cluster the researched company belongs. Due to the structure of the insurance company’s liabilities to the insured, and hence the assets they manage, insurance companies are highly valued as long-term and stable investors. This has certain advantages not only for the insured, but also for society and the economy as a whole due to the role of financial intermediaries on the part of insurance companies. With the help of cluster analysis, companies were divided into three groups – strong, medium, weak in rating. However, for the most accurate assessment, discriminant analysis was used based on the already obtained results of cluster analysis. The affiliation of the studied object (insurance company) to one of the selected classes is determined on the basis of the constructed discriminant model. Classification tables for each observation, tables of squares of Mahalanobis distances and tables of a posteriori probabilities of affiliation are used. Discriminant analysis based on the already obtained results of cluster analysis was used. Discriminant analysis gave the final assessment and rating of insurance companies in Ukraine. Discriminant analysis is devoid of these shortcomings and includes statistical methods for classifying multidimensional observations in a situation where the researcher has so-called training samples. This type of analysis is multidimensional, as it uses several features of the object, the number of which can be large. The purpose of discriminant analysis is to, based on the measurement of various characteristics of the object to classify it, that is to refer to one of several given groups in some optimal way. It is assumed that the source data, along with the characteristics of the objects contain a variable of class, which determines the affiliation of the object to a particular group. Therefore, this analysis provides for the verification of the consistency of the classification. Discriminant analysis made it possible to make a final assessment and rating of insurance companies in Ukraine, and allowed to assess the quality of the model and the importance of the studied indicators. The initial data were formed with the help of qualitative and quantitative indicators, which made it possible to more accurately and broadly assess the state of insurance companies.