Artificial Intelligence System for Financial Risk Prediction in the Banking Sector

Abstract

AbstractThe aim of the study was to put forward and prove the hypothesis that using the AI ​​system it is possible to obtain a forecast of the share of overdue loans in the bank's portfolio. Based on the study, the following results were obtained. Theoretical foundations of the analysis and forecasting of financial risk in the banking sector under conditions of market uncertainty have been studied. The novelty of the study lies in the fact that the share of overdue loans in the bank's portfolio can be predicted based on the use of the developed artificial intelligence system - perceptron. The practical significance is that the perceptron AI system can be recommended for practical use. The study showed that the total loan portfolio of Russian banks in 2020 grew by 13.8% to 63.2 trillion rubles. The share of overdue debt on loans decreased over the past year from 11.9 to 11%. However, the problem with the quality of loans remains relevant. As a result of this project, the perseptron program was developed to predict the dynamics of the share of overdue loans in the portfolio of a commercial bank, which was formed on the Deductor platform. It was revealed that the share of overdue loans of commercial banks is influenced by many factors, including the factors included in the AI ​​system. The perseptron program has been developed to predict the dynamics of the share of overdue loans in the portfolio of a commercial bank.KeywordsFinancial riskMarket uncertaintyLoan portfolioPerceptronBad loansMoney