Application of artificial neural networks in predicting financial distress in the JSE financial services and manufacturing companies

ABSTRACT

This study explored the role of artificial intelligence (AI) in predicting companies’ financial distress. We used Artificial Neural Networks (ANN) to develop and test financial distress prediction models for the financial services and manufacturing companies listed on the Johannesburg Stock Exchange (JSE) for the period 2000–2019. Our constructed ANN Models achieved classification accuracy rates of 81.03 and 96.6 percent for the financial services and manufacturing industries, respectively. Both models could also predict financial distress up to five years prior to the firm being classified as distressed. This study provided key theoretical and practical contributions to the current literature by highlighting the potential role of AI models in solving financial problems. Creditors can use the models built in this study as a default prediction tool, investors as an investment decision-making tool, and for business managers a performance guidance tool to ensure long term financial sustainability.

KEYWORDS:

• Artificial neural networks
• bankruptcy prediction
• financial distress
• financial ratios
• macroeconomic factors

JEL:

• Code: C63
• G17
• G33

Disclosure statement

No potential conflict of interest was reported by the authors.