Bankruptcy Prediction Model for Public Companies in Indonesia Stock Exchange Mochamad Nabil Faindra Putra
Faculty of Economics, Universitas Indonesia
Jalan Salemba Raya No. 4, Jakarta Pusat 10430, Indonesia
mochamad.nabil[at]ui.ac.id
Abstract
This study aims to create a new predictive model for bankruptcy in responses to public companies^ financial distress due to COVID-19 pandemic. The research method used is logistic regression to examine the relationship between bankruptcy and independent variables such as financial ratio and stock market ratio. The result shows that leverage, solvability, and profitability ratio affect more significant than others ratio. Since financial distress are not occurred suddenly, this study divides its model into 2, 1 year before distress (M1) and 2 years before distress (M2). The results show that M1 has better result, with classification accuracy in 93,5% (on default cut off point = 0,5). We also re-estimate others accounting-based model and compare our model into it. We found that our model performs better at holdout samples than others (+2,21% difference), but performs slightly behind the re-estimated Ohlson models at simulation samples (-0,61% difference)
Keywords: Bankruptcies, logistic regression, financial distress, public companies
