Predicting Financial Distress Using DEA and Multivariate Discriminant for Tourism, Restaurant, and Hotel Sector in Indonesia

Abstract

Financial distress is a condition in which a company experiences decline and difficulty to fulfil its financial obligations. Financial distress prediction aims to identify early warning indicators of impending financial disaster so businesses can begin financial reconstruction at the appropriate moment. This study aims to compare the statistically significant difference results of various financial prediction models and the level of accuracy of each model in 25 companies engaged in the hotel, restaurant, and tourism sectors in 2018-2021. By using Kruskal-Wallis Test, Mann-Whitney Post Hoc Test, and accuracy test to compare each model. The results of this study are based on the Kruskal-Wallis test, all models used are statistically significant differences. Meanwhile, when paired using the Mann-Whitney Post-Hoc Test, it was found that the Springate and Grover models did not have a statistically significant difference. In addition, the results of the test accuracy show that the DEA accuracy rate of 79%, and the Springate model with the lowest accuracy of 33%. The results interpret that each model has its own indicator for predicting financial distress. It is recommended in examining the financial distress of hotel, restaurant, and tourism companies using DEA model since it results in the highest accuracy rate. Keywords: DEA, financial distress, multivariate discriminant analysis