AN EMPIRICAL INVESTIGATION OF THE CHAOS BASED BANKRUPTCY MODEL FOR PREDICTING COVID-19 RELATED BANKRUPTCIES

Abstract

The COVID-19 pandemic has caused unprecedented damage to businesses across various industries, leading to a surge in bankruptcy filings worldwide. To predict COVID-19-related bankruptcies, this study examines the effectiveness of the Chaos Based Bankruptcy Model in forecasting bankruptcy using daily stock market returns. The Chaos Based Bankruptcy Model measures the degree of chaos in a system using the Lyapunov exponent calculated from the time series of daily stock prices. The model is based on the hypothesis that unhealthy systems exhibit less chaos than healthy systems.

Using data from before the pandemic, the binary logistic regression model successfully predicted the bankruptcy status of a company 70.8% of the time. The model compares the Lyapunov exponents of firms approaching bankruptcy with those of non-bankrupt pair match firms, based on the newer NAICS code, to identify firms that are more likely to declare bankruptcy. The coefficient of the model's Lyapunov exponent variable was -8.918, significant at 0.024 levels, indicating that the fewer the levels of chaos observed in the stock price trends of a firm, the higher the probability of bankruptcy.

This paper provides a novel method of bankruptcy prediction, which has not been explored in previous studies. Furthermore, this method has the potential to identify high-risk businesses in advance of financial stress, allowing companies to take preventive measures and avoid filing for bankruptcy. Future research will focus on applying this model to specific industries to predict bankruptcy risk in those industries, which will enable policymakers and investors to make informed decisions