Assesment the influence of debt capital on the bankruptcy of enterprises in the agricultural sector

Purpose. The purpose of this study is to improve the approach to assessing the influence of loan capital on the bankruptcy of enterprises in the agricultural sector in the pre-war period. Methodology / approach. The article analyzes the significance of Altman, Lees, Springate, Duran, Tereshchenko and Matviychuk models. For the statistical base, open data for 2020 of 500 agricultural enterprises of Ukraine were selected, of which 12 became bankrupt according to the Unified Register of Enterprises, in respect of which bankruptcy proceedings were initiated according to the data of 2021–2022. Selected models of bankruptcy diagnosis, including loan capital (components of loan capital), were tested for the purpose of analyzing their effectiveness. Python Programming Language was used to test selected models for predicting the bankruptcy of agricultural enterprises. To evaluate the effectiveness of the models, such metrics as accuracy and the matrix of inconsistencies were calculated. Results. The Altman, Lis, Springate, Duran, Tereshchenko models, except the Matviychuk model, performed well on a statistical sample for determining bankrupt agricultural enterprises, which later really became bankrupt. But those that were subsequently stable were also classified as bankrupt. Duran's model, in which loan capital plays a significant role, showed the best results. Therefore, the amount of loan capital is important in predicting the bankruptcy of agricultural enterprises. The results of the paper refer to the pre-war period, however, the approach to processing data and forming conclusions is universal and can be applied to more recent data if available. Originality / scientific novelty. The novelty is the improvement of the approach to assessing the impact of loan capital on the bankruptcy of agricultural enterprises, which is based (1) on the analysis of the effectiveness of selected models for assessing the probability of bankruptcy and (2) on the basis of a significant sample of data from the financial reports of agricultural enterprises, which emphasizes the objectivity of the results obtained, as well as (3) on the use of Python for testing bankruptcy prediction models. Practical value / implications. Assessing the relationship between loan capital and bankruptcy can provide insight into the financial condition of agricultural enterprises and provide an opportunity to determine strategies to prevent or reduce bankruptcy risks. For agricultural enterprises, the results of this assessment can serve as a basis for providing recommendations for managing debt capital and other financial resources in order to avoid bankruptcy.