The Performance of Financial Distress Prediction Models: Evidence from ASEAN Countries

Abstract

The objective of this study is to scrutinise selected financial distress prediction models across several categories, such as country, sector, COVID-19 era, utilising binomial probability test, confusion matrix, and ROC curve. The sample of this study included 21,250 observations from 6 countries and 21 sectors. In addition, it was collected on an annual basis from 2012 to 2021, which included both the pre- and during COVID-19 eras. According to the results of the binomial probability test,  the observed proportion are statistically different from the assumed probability based on accuracy and total positives. Meanwhile, based on the confusion matrix and ROC curve results, it was indicated that the Grover model performed best in predicting financial distress for all COVID-19 era, country, and sector observations. This suggests that the Grover model can be applied as a practical tool to predict financial distress in ASEAN. This study will contribute significantly to the literature on financial distress prediction since there are no studies that scrutinise and compare the performance of financial distress prediction models in ASEAN countries and different sectors.