Prediction of financial failure in revolving fund enterprises and an application in the health sector

Revolving fund enterprises, which have a crucial place in the Turkish public financial management system, operate within the body of public administrations in fields such as education, agriculture, forestry and especially health. Although the initial purpose of the establishment was to use the idle capacity in the public sector, that has been changed over time and turned into the ability of the government to sell priceable public goods and services to the society with an understanding of business administration. It is seen that the MoH revolving fund enterprises, which are used as a very important financing tool in the execution of health services in our country, have experienced financial difficulties in recent years. With this research, the prediction of financial failure of the revolving fund enterprises operating under the Ministry of Health and the underlying main factors causing financial failure are aimed to be determined via data mining methods like CHAID decision trees algorithm and artificial neural network model applied to both financial and non-financial data for the years 2008-2020. As a result of the analysis, the most important variables affecting financial failure were determined as the ratio of operating cost to sales in the decision tree model, while liquidity ratios and the ratio of total hospital invoice to the global budget in artificial neural networks.