Financial Distress Prediction Model in the Construction Industry in Indonesia

Abstract

This study aims to evaluate the effectiveness of several financial distress prediction models in construction companies in Indonesia, especially in responding to changes in government infrastructure budget allocation policies. The models tested included Altman Z-Score, Zmijewski, Springate, Michal Karas, Grover G-Score, as well as Principal Component Analysis (PCA)-based models with logistic regression. The data used includes 33 construction companies listed on the Indonesia Stock Exchange (IDX) during the period 2017–2024 with a total of 258 annual company observations. The confusion matrix and logistic regression methods were used to assess the performance of each model. The results of the analysis showed that the PCA model provided the best performance with an accuracy rate of 94% for the T-1 prediction and 93% for the T-2. Logistic regression also showed that the PCA model had strong predictive clarity (Nagelkerke R² of 65.2% for T-1 and 44.9% for T-2). Profitability proved to be a significant predictor, and the government's focus on infrastructure spending strengthened the accuracy of the predictions. This study recommends the PCA model as the main tool for early detection of financial distress in construction companies.