Comparison of Machine Learning Algorithms for Predicting Financial Risk in Cash Flow Statements

Abstract

Nowadays, making financial decisions and evaluating loan applications is a complex and sensitive process. Cash flow data, which shows the financial risk status of businesses, plays a key role in the evaluation of loan applications. A detailed analysis with machine learning algorithms evaluates the performance of different models in the loan classification process and highlights the role of cash flow data in the process. The study includes data from 282 companies for the quarterly periods between 2018 and 2022. It is observed that there are limited studies on loan classification with cash flow statement in the literature. Considering the suitability of the data used in the study to the data structure, the creation of effective algorithms and the evaluation of these algorithms with information criteria aimed to provide a unique approach in the field. The model for the 2nd quarter of 2019 was selected as the best model with 99% accuracy and 99% F1 value. It is also determined that variable selection with high accuracy rates in the models established for each quarter is important for predicting financial risk.