Hierarchical Fuzzy Method of Comparing Bank Products with Complex Tariff Packages

Abstract

The article considers a method for solving an information asymmetry problem that is relevant for the banking sector. The authors compare bank products and services in which tariff packages are consist of different informational components. The principal feature of the proposed method is to focus on obtaining a subjective evaluation of the attractiveness of the Bank’s offers for specific target groups. This makes the proposed method relevant for Central and commercial banks, as well as for consumers of banking products. Since complex tariff packages of different banks contain not only many indicators but also different sets, it is impossible to directly compare them. The proposed method allows solving the main informational problems of traditional methods for comparing banking products. Using the Fuzzy Inference System (FIS) allows a flexible approach to the aggregation of singular indicators of tariff packages and simplifies the expert procedures. To make rules for fuzzy inference, linguistic constructions close to the natural language can be used. This makes it easier to fill the knowledge base. To solve the set tasks, a methodology for constructing hierarchical FIS is used. In comparison with non-hierarchical FIS it can reduce the requirements to the number of rules in fuzzy inference rule base and system implementation time. The experiments with the fuzzy inference system, which implements a corporate credit cards tariff packages comparison of four Ukrainian banks, showed the effectiveness of the proposed approach. The obtained numerical results are adequately interpreted for the highlighted groups of banking product consumers.