Credit funding and banking fragility: A forecasting model for emerging economies

Abstract

Our paper proposes an empirical model to forecast banking fragility episodes using information from the credit funding sources. We predict the probability of occurrence of such events 3 and 6months ahead, employing a Bayesian Model Averaging on logistic regressions. We perform prediction exercises for nine emerging economies under a broad set of prior specifications, whose results are evaluated using predictive ability tests and the signaling analysis approach. Our findings indicate that the increasing use of wholesale funds provide signals of banking frailness. Moreover, pseudo out-of-sample predictions show that our warning tool is able to forecast financial fragility events.