Pysanets K. Economic and mathematical modelling of scoring systems for borrower estimation.

In the thesis is settled and proposed solution of an actual problem of borrowers' credit risks dynamics estimation via scoring systems, in scope of which the complex of economic-mathematical models which allow to increase credit decisions quality are developed. It is proposed the methodological approach of risk management using cut-off curve, which allows to form effective lending strategies: profit maximization, risk reduction, credit portfolio size increase. The thesis deals with decision making methodological approach improvement in credit risk-management using LTV (life time value) indicator and credit risk factors identification procedures by means of information value (IV) dynamic estimation. Asymmetry of credit risk appearance time is found. The level of integral influence of the set of risk factors is estimated with consideration of their interconnection in consumer segment of credit market. Implementation of developed models of borrowers' credit risk dynamics estimation is done for the stage of loans issuance and stage of overdue loans recovery. Collection management processes optimization problem is solved.