Differentiated impact of spread determinants by personal loan category: Evidence from the Brazilian banking sector

Abstract

The empirical literature on the determinants of banking spreads has considered banks as pro-viders of one single product. According to this approach, the study of the effect of bank spreads determinants has assumed that this effect is uniform across the different types of loans offered by banks. The present study assesses the hypothesis that banking spreads attrib-utes have a differentiated impact on spreads, according to the loan category. To this end, we adopt a dynamic model, estimated through System GMM on the basis of a dataset of interest spreads charged on three categories of personal loans in Brazil. Our results support the hy-pothesis of differentiated impacts according to loan category, and are corroborated by a ro-bustness check, carried out through Difference GMM estimation of the adopted model. Over-all, these findings also suggest that regulators should observe the composition of banks’ loans portfolios when designing and implementing policies aiming at banking spread reduction.