An early warning system approach for the identification of currency crises with data mining techniques

Abstract

It is a well-known fact that currency crises can be extremely costly to the countries, institutions and businesses with its damaging long-term effects. If the history of currency crises is investigated, it would be seen that crises cause many hazarding effects on economies, business policies and assets. It would generally have short-term and long-term effects and often spill over, via a variety of channels to other countries and companies. In this study, it is aimed to investigate the dominant factors that lead to currency crises. Discovering the nature and characteristics of currency crises and predicting possible currency crises in an early phase would save managers some time in better crisis management policies and corrective actions. We employed data mining techniques to construct an early warning system for the knowledge discovery and early prediction purposes of currency crises with the aim of providing an in-depth understanding of the topic for the researchers and managers. © 2012 Springer-Verlag London. © 2013 Elsevier B.V., All rights reserved.