Predicting Sovereign Debt Crises Using Artificial Neural Networks: A Comparative Approach

2 The Series " Documenti di Lavoro " of the Istituto di Studi e Analisi Economica – Institute for Studies and Economic Analyses (ISAE) hosts the preliminary results of the research projects carried out within ISAE. The diffusion of the papers is subject to the favourable opinion of an anonymous referee, whom we would like to thank. The opinions expressed are merely the Authors' own and in no way involve the ISAE responsability. The series is meant for experts and policy-makers with the aim of submitting proposals and raising suggestions and criticism. ABSTRACT Recent episodes of financial crises have revived the interest in developing models that are able to timely signal their occurrence. The literature has developed both parametric and non parametric models to predict these crises, the so called Early Warning Systems. Using data related to sovereign debt crises occurred in developing countries from 1980 to 2004, this paper shows that a further progress can be done applying a less developed non-parametric method, i.e. Artificial Neural Networks (ANN). Thanks to the high flexibility of neural networks and to the Universal Approximation Theorem an ANN based early warning system can, under certain conditions, outperform more consolidated methods. Financial crises occurred in emerging countries in the last decade of 20th century have revived theoretical and empirical interest in the topic in order to understand their causes and consequences as well as to develop statistic and econometric models that can timely signal their occurrence. Economic theory has developed three generation of models explaining financial crises: the " first " and " second generation " models focus on currency crises and public imbalances, while " third generation " models include a wider variety of crises and are better suitable at explaining episodes occurred in the late '90s which were caused, principally, by private imbalances. In the last decade, many empirical studies have concentrated their attention in developing models able to timely signal the occurrence of a financial crisis, the so-called early warning system (EWS). Using statistical and econometric techniques these models are applied to predict the likelihood of financial crises using a wide number of indicators related to internal and external factors, as well as social and political condition. The aim of this paper is to further develop the Early Warning System literature related to financial crisis. In particular, a less explored non-parametric method, i.e. Artificial Neural Network (ANN), is carried to test …