Neural Network Vs Discriminant Analysis in Detecting Financial Distress Among Major Australian Companies

Neural Networks have seen an explosion of interest over the last few years, and are being successfully applied as statistical tools across an extraordinary range of problem domains, in areas as diverse as finance, medicine, engineering, geology, psychology and physics. Indeed, anywhere that there are problems of prediction, discrimination (classification), or control neural networks are being introduced. On the other hand under some distributional assumptions of the measured variables (eg: multivariate normal distribution) the traditional discriminant analysis can be employed for classification and for prediction purposes. the question arise now is "Why then Neural Network"? Neural Networks are adaptive statistical models based on analogy with the structure of brain. They do not differ essentially from standard statistical models. For example, one can find neural network architectures akin to the traditional discriminant analysis, principle component analysis, logistic regression and other techniques.

In this paper, numerous rations of a combination of successful and failed firms from the Australian Stock Exchange have been analysed using the neural network and the traditional discriminant analysis techniques. The overall result was good with an accuracy of over 80% correct prediction. Neural Networks approach appears to have an edge in performance over the traditional discriminant analysis.