Forecasting Financial Distress of PN17 Companies in Malaysia Using Logistic Regression

Abstract

This study aims to predict financial distress among Malaysian companies included in Practice Note 17 (PN17). The analysis examines a sample of 35 companies classified as PN17 from 2014 to 2023. To create a balanced comparison, these companies are matched with 35 non-PN17 companies, resulting in a total sample size of 70 firms. Logistic regression was employed in this study because the financial ratios do not need to be normally distributed. The findings indicate that the model is most effective in the near term, specifically during the year of financial difficulty and up to one year prior to the occurrence, when it achieves the highest level of prediction accuracy. Financial metrics, including working capital, retained earnings, and earnings before interest and taxes (EBIT) are crucial in assessing a company's probability of experiencing financial difficulties, underscoring the vital need of liquidity, profitability, and financial soundness. However, when the prediction timeframe extends beyond two years, the model's precision decreases, highlighting the limitations of using conventional financial ratios for long-term forecasts. This suggests that while logistic regression is a valuable method for predicting short-term distress, its effectiveness diminishes in early-stage forecasts, when distress indicators are less prominent.