The banking sector plays a central role in economy and it is often highly concentrated: in many countries, the share of assets held by the five largest banks is around 80%. Moreover, in the last decade the degree of market concentration has increased. Despite the importance of banks and of bank competition, understanding of the effects of bank competition on the real economy is still limited. A key reason is that the degree of bank competition depends on market conditions.
In this paper, we assess the effects of bank competition on the cost of credit and economic activity. To this end, we look at how the merger and acquisition (M&A) episodes of large banks in Brazil affect local competition. Using detailed administrative data on loans and firms, we explore such M&A episodes across municipalities. Specifically, we employ a difference-in-differences strategy that compares changes in outcomes for markets that are affected or unaffected by M&A episodes.
First, we show that a reduction in bank competition increases lending spreads (ie the difference between lending and deposit rates) and shrinks credit volume. This decrease arises entirely through the extensive margin – ie fewer loans are made, rather than a reduction in the size of loans. Second, we show that these effects on credit markets feed through to the real economy. M&A affects firm output in both the tradable and non-tradable sector, indicating that firm financing is relevant to real outcomes. Among other counterfactuals, we show that if bank competition increased and Brazilian spreads fell to world levels, output would increase by approximately 5%.
We use heterogeneous exposure to large bank mergers to estimate the effect of bank competition on both financial and real variables in local Brazilian markets. Using detailed administrative data on loans and firms, we employ a difference-in- differences empirical strategy to identify the causal effect of bank competition. Following M&A episodes, spreads increase and there is persistently less lending in exposed markets. We also find that bank competition has real effects: a 1% increase in spreads leads to a 0.2% decline in employment. We develop a tractable model of heterogeneous firms and concentration in the banking sector. In our model, the semi-elasticity of credit to lending rates is a sufficient statistic for the effect of concentration on credit and output. We estimate this elasticity and show that the observed effects in the data and predicted by the model are consistent. Among other counterfactuals, we show that if the Brazilian lending spread were to fall to the world level, output would increase by approximately 5%.
JEL classification: G21, G34, E44
Keywords : bank competition, mergers and acquisitions, lending, spreads, output