Predicting corporate failure using discriminant analysis
Abstract
During its existence, a company inevitably faces economic and financial difficulties that can potentially lead to its failure. However, these problems do not arise suddenly, but usually result from the progressive accumulation of serious challenges faced by the company. To anticipate these failures, the use of failure prediction tools is essential as they enable timely implementation of necessary corrective measures. Therefore, the prediction of corporate difficulties is of crucial importance, while also allowing for classification between failing and non-failing companies.
The objective of this article is to predict corporate failures in Morocco. To achieve this, a methodology was implemented, involving the modeling of the inherent risk of failure using a sample of 30 Moroccan companies. For this purpose, a set of 6 financial ratios was used, calculated over the period from 2017 to 2019. These ratios provide key insights into the financial health of companies, their ability to meet obligations, and their profitability. Among these ratios, two were found to be particularly significant in explaining failure: the debt ratio and the financial profitability ratio.
To achieve the prediction of corporate failures, the technique of discriminant analysis was employed due to its proven robustness and predictive power. This model allows for the establishment of a boundary between failing and non-failing companies based on the selected financial ratios. In this study, the model achieved a 90% rate of correct classification, demonstrating its predictive effectiveness. This means that the model correctly classified 90% of the companies in the sample as failing or non-failing.
Keywords: Business failure, struggling companies, prediction of failure risk, financial ratios, discriminant analysis.
JEL Classification: G33
Paper type: Empirical research
Downloads
Copyright (c) 2024 Chaimae SALIH, Khaddouj KARIM
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.