Companies Bankruptcy Prediction by Using Altman Models and Comparing Them

Abstract

Bankruptcy prediction of economic institutions is considered a necessary matter at the present time in order to avoid the risks that may drive such institutions out of business. Given such fact, the current study was made to highlight the intellectual aspects of the subject of bankruptcy prediction and means of measuring it. There are five main types of models for predicting companies bankruptcy: one-way analysis of variance, multiple discriminant analysis, logarithmic analysis, recurrent algorithm analysis, and finally neural networks analysis, which is the most recent bankruptcy prediction method. These methods do not produce similar results. Most bankruptcy prediction studies used multiple discriminant analysis (MDA) and statistical methods for models development. These studies covered both large and small companies as well as private and public companies. MDA is the essence of this research paper which deals with Altman Model in detail and describes the changes that the original Z-Score equation has gone through. The study problem lies in arranging Altman Models for bankruptcy prediction of commercial companies in Iraq in accordance with the importance of each model.