Biloshyts'kyj O. The system of financial stability management models in insurance.

The dissertation contains the research results of modeling the financial stability in insurance in conditions of information asymmetry. Today research processes of financial stability of Ukrainian insurers are limited by appropriate analytical data as well as by models applicable for performing such analysis in conditions of uncertainty, scantiness and data falsity. It is proven, that scientific valid classification of insurers in correspondence to the level of financial indicators can be performed when these insurers will be analyzed in terms of their market similarity, and their quasi-homogeneity within certain sampling. There is multi-factor fuzzy-logic financial-results-based model for analysing financial stability of insurers in conditions of information asymmetry developed in the dissertation. There is also two-level neural-network model for complex analysis of insurer's financial stability developed in the dissertation. Having fuzzy data, it allows to get the integrated indicator of insurer's financial stability as well as to perform additional verification of results got while fuzzy-logic modeling process. The results of modeling financial stability in insurance when applying fuzzy-logic and neural-networks approaches has shown that the quality of modeling process is much depends on the set of qualitative and quantitative factors. First of all, it depends on the groups of factors being analyzed, data quality, and certain model parameters - meaning the quality of wages for neural network, definition of factors' threshold values, neural network architecture, transformation functions, quality of neurons, wages etc. The developed technology allows expert to perform decomposition of results to the level of groups of factors, to analyze gradations for each of the groups relatively to certain level, and make conclusions about the level of each factor. It allows expert to find correlations among different factors and also analyze impact of each factor on the integral value of insurer.