Introducing an innovative mathematical method to predict the bankruptcy risk. Measures for the financial markets stability

摘要

This paper advances innovative measurement models for bankruptcy risk and develops a framework for further research in the use of latent multilevel static and dynamic models in modeling bankruptcy decision making. A brief presentation of literature is given, with an emphasis on the most recent developments in this field. The econometric model underlying the fuzzy sets theory applied to the measurement of bankruptcy risk is particularly examined here. The innovative part of the paper is the application of a new methodology to the analysis of bankruptcy. This model, called Item Response Theory (IRT), is a traditional technique in psychometrics and brings a set of advantages when being applied to the analysis of latent measures by a set of observable binary indicators, as is the bankruptcy risk. At the end, the paper concludes on the need for more collaboration among researchers in business, computer science and low, in order to develop new interdisciplinary measurement tools in finance.