Explaining and forecasting bank loans. Good times and crisis

This article aims to develop a parsimonious model to explain and forecast bank loans to nonfinancial companies during calm periods as well as in situations of financial turmoil. It focuses on the French context, over a period including financial, banking and sovereign debt crises. Theoretical views and intuitions led us to gauge the marginal informational content of a large set of leading indicators in VAR and VECM models, and to investigate potential nonlinearity in credit dynamics. In accordance with firms and banks' balance sheet effects, the growth rate of equity prices appears to be one of the most interesting leading indicator as well as a significant threshold variable for explaining regime switching. However, it appears difficult to accurately predict the right credit dynamics regimes. A simple VAR model finally performs better.