Life testing of a firm: Some recent developments with application artificial neural networks ANN

Financial experts are interested in the detection of a firm’s slide toward failure at an early stage. In this paper we have briefly reviewed the various techniques for the detection of "survival" or"failure" of the firms. These techniques consist of statistical techniques like legit and probit models,multiple discriminant analysis, multivariate CUSUM methods etc. Artificial Neural Network (ANN) techniques can play a crucial role in survival analysis and quite often they are found better than statistical techniques. We have also developed a new hybrid model. A case study dealing with real data will be presented. The models discussed in this paper can be used in bankruptcy prediction of the firms in Taiwan.