Feedforward neural networks in the classiﬁcation of ﬁnancial information CARLOS SERRANO-CINCA Departamento de Contabilidad y Finanzas, Facultad de Ciencias Econ ´ omicas y Empresariales, Universidad de Zaragoza, Gran V´ıa 2, 50005 Zaragoza, Spain E-mail: serrano@posta.unizar.es Financial research has given rise to numerous studies in which, on the basis of the information provided by ﬁnancial statements, companies are classiﬁed into different groups. An example is that of the classiﬁcation of companies into those that are solvent and those that are insolvent. Linear discriminant analysis (LDA) and logistic regression have been the most commonly used statistical models in this type of work. One feedforward neural network, known as the multilayer perceptron (MLP), performs the same task as LDA and logistic regression which, a priori, makes it appropriate for the treatment of ﬁnancial information. In this paper, a practical case based on data from Spanish companies, shows, in an empirical form, the strengths and weaknesses of feedforward neural networks. The desirability of carrying out an exploratory data analysis of the ﬁnancial ratios in order to study their statistical properties, with the aim of achieving an appropriate model selection, is made clear. Keywords: neural networks, multilayer perceptron, bankruptcy prediction, Spanish banking system 1. INTRODUCTION Since the pioneering work of Altman (1968), linear discriminant analysis (LDA) has been the most commonly used statistical model in the prediction of corporate failure. However, its application is hampered by a series of restrictive assumptions and it suffers from a limited discriminatory power. Neural networks have been proposed to complement or substitute for traditional statistical models. White (1989), Cheng and Titterington (1994), Sarle (1994) and Ripley (1994) provide much insight into the statistical components of neural networks. The multilayer perceptron (MLP) is one of the most well known and widely used models of artiﬁcial neural networks. Gallinari et al. (1991) have demon- strated the relationship between LDA and MLP. Bell et al. (1990), Hart (1992), Yoon et al. (1993), Curram and Mingers (1994), Wilson and Sharda (1994) and Altman et al. (1994) have compared the classifying power of different statistical tools and of MLP. Feldman and Kingdon (1995) have surveyed some of the research issues used in applying neural networks to real-world problems and reviewed a number of neural network ﬁnancial applications. 1351–847X © 1997 Chapman & Hall The European Journal of Finance 3, 183–202 (1997)