Decomposing Banking Spreads of Multiple Types of Credit: a Structural Model Approach

Orientador(a): Carlos Viana de Carvalho

We propose a DSGE model with multiple types of credit and two financial frictions to address two objectives. First, we decompose Brazilian banking spreads into four components: administrative costs,  financial margins, taxes, and default risk. Second, we examine the effects of technology and monetary policy shocks on these spreads. The model features an imperfectly competitive banking sector that lends to heterogeneous households and to an entrepreneur subject to default risk, following Bernanke et al. (1999). Within this framework, each type of credit is characterized by different levels of impatience, costs, default rates, and market power. The model also incorporates a tax structure and financial costs consistent with the Brazilian economy. Then, we calibrate the model to match key determinants of banking spreads in Brazil. The main results are as follows: (i) the spread decomposition reported by the Central Bank of Brazil is closely replicated; (ii) reductions in costs and loan taxes -policies implemented in Brazil- lower all spreads; (iii) productivity and monetary policy shocks generate responses consistent with the financial frictions literature; and (iv) these results are robust when long-term credit is considered.