КРЕДИТТІК ӨНІМДЕРДІҢ ӘРТҮРЛІ ТҮРЛЕРІ БОЙЫНША СКОРИНГТІК КАРТАЛАР

Abstract

Abstract. One of the main topics of discussion in credit risk management in financial corporations is the evolution of credit scoring. But since credit products are different depending on the risk and financing period, and there is not enough information about borrowers, a single method of issuing rating Cards is often useless. Paperwork takes into account the possibilities of score cards for consumer lending, refinancing, lending to small and medium-sized companies, car loans, mortgages, fintech and P2P. Therefore, this work, as mentioned above, can be considered as a comparative analysis of the most important elements that affect the likelihood of default by the borrower when calculating by segments., as well as the consideration of machine learning methods and the use of alternative data sources that allow improving the accuracy of the forecast. Depending on the typical loan product, research will allow you to give recommendations on choosing the best way to create scoring cards, thereby improving the accuracy of predicting the borrower's creditworthiness and thus reducing the degree of default risk.

Key words: credit scoring, scoring cards, credit products, refinancing, machine learning, risk management, creditworthiness, alternative data.