La previsione dell’insolvenza ex art. 13 co. 3 C.c.i.: efficacia del test di classificazione binario e dell’analisi discriminante lineare

We argue that a multivariate dichotomous classification test is the best tool when forecasting bankruptcy in the context of the upcoming reform of the Italian law on insolvency and restructuring (d.lgs. 14/2019 – Codice della crisi d’impresa e dell’insolvenza).

To back our claim, we show an empirical comparison between the ability of the univariate dichotomous classification test and linear discriminant analysis to predict bankruptcy, and discuss the results with respect to the limitations of the two approaches.

The tested models are designed based on their features, considering the scope of application of the new legislation, and the evolution over time of the many bankruptcy prediction techniques.