Prediction of Bankruptcy: Evidence from Moroccan Agricultural Companies

Abstract

The aim of this research is to build a bankruptcy prediction model specific to the Moroccan agricultural firms through two methods: multidisciminant analysis and log it model. The second purpose is to choice the most suitable model for this specific sector. This paper is based on a sample of 75 healthy and 75 companies declared bankrupt. The model aims to test the predictive power of financial ratios reflecting the profitability, the liquidity, the solvency and the activity of companies.  The results indicate comfortable prediction accuracy for both models. However, the logistic regression outperforms MDA and has higher classification accuracy prediction. Beside academic contributions, this study may help credit risk managers and investors to provide them useful tools in the assessment of credit risk. The overall difference of this study from previous one is it can be considered the first paper which investigates the bankruptcy prediction in agricultural firms in Morocco.