No. 909 - Forecasting corporate default probabilities: a local logit approach for scenario analysis

Vai alla versione italiana Site Search

by Giuseppe Cascarino, Federica Ciocchetta, Stefano Pietrosanti and Ivan QuagliaFebruary 2025

The paper proposes a new model for forecasting Italian firms' default probability and for scenario analysis, using both firm-level data and macroeconomic series. The forecasting accuracy is assessed by estimating defaults during 2018-20 from the data available at the end of 2017, assuming a 'baseline scenario' using the values of the macroeconomic series actually recorded during this period. Additionally, an 'adverse scenario' is hypothesized, involving a deterioration in the macroeconomic environment analogous to the one witnessed during the sovereign debt crisis.

The model replicates the historical trend of the aggregate default rate of firms and ranks them appropriately according to their default risk. The forecast for the default rate of firms for 2018-20 under the baseline scenario demonstrates a high degree of accuracy. In the adverse scenario, the increase in the default rate is consistent with that experienced in Italy during the sovereign debt crisis.