The aim of this paper is to evaluate the contribution of several aspects of the (hard) bank-firm relationship in anticipating the corporate default event in a model including firms’ specific balance sheets characteristics. Using a unique dataset on a sample of 113 co-operative credit banks and about 12,000 firms operating in Italy between 2012 and 2014, this paper documents that private-hard information, coming from checking accounts activities and long-term loan performances, provides an additional explanatory power to public-hard information, gathered from balance sheets, in predicting the likelihood of default. Estimating a pooled probit model, it finds that usage of credit lines and credit lines limits violations on checking accounts, as well as credit lines limit violations and payments overdue on long-term loans, increase the accuracy prediction of default events by about ten percent one year before and by six percent two years before.
Estimating the additional predictive power of bank-firm hard information in default models. An analysis on a sample of Italian firms
Michele Modina;Filomena Pietrovito
2017-01-01
Abstract
The aim of this paper is to evaluate the contribution of several aspects of the (hard) bank-firm relationship in anticipating the corporate default event in a model including firms’ specific balance sheets characteristics. Using a unique dataset on a sample of 113 co-operative credit banks and about 12,000 firms operating in Italy between 2012 and 2014, this paper documents that private-hard information, coming from checking accounts activities and long-term loan performances, provides an additional explanatory power to public-hard information, gathered from balance sheets, in predicting the likelihood of default. Estimating a pooled probit model, it finds that usage of credit lines and credit lines limits violations on checking accounts, as well as credit lines limit violations and payments overdue on long-term loans, increase the accuracy prediction of default events by about ten percent one year before and by six percent two years before.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.