| Summary: | In 2003 European Central Bank issued a qualitative survey of financing condi- tions. Called bank lending survey (BLS) and including 22 questions on credit standards, credit terms and conditions, and loan demand, the BLS was supposed to act as an additional information stream for the European Central Bank to be able to differentiate between supply and demand effects. This thesis gathers available BLS answers and evaluates their potential use in an early warning system. According to AUROC analysis of various logit models, based partially on traditional early warning indicators (EWIs) and partially on the BLS data, the study picks 27 BLS variables that significantly improve performance of the current EWIs over the period of 2003-2017. JEL Classification C23, C40, F47, G01, G21, G28 Keywords EWIs, bank lending survey, ROC, financial cycle Author's e-mail mayr.samuel@gmail.com Supervisor's e-mail adam.gersl@gmail.com
|
|---|
