An empirical investigation of the effects of concentration on profitability among US banks

This paper analyses the effects of concentration on profitability in the US banking sector from 1994-2005, using bank-level panel data. A new index of concentration is proposed, which reflects the depth and intensity of concentration. The econometric specification facilitates the simultaneous testing of the four main hypotheses in the literature concerning the relationship between concentration and profitability. Strong support is found for the Structure-Conduct-Performance hypothesis, as well as some support for the Relative Market Power hypothesis. The results are robust to alternative econometric techniques and specifications, and to various measures of profitability and of concentration. Further analysis sheds light on the nature and possible channels of the concentration-profitability relationship. A positive relationship is found between concentration and profitability even when the largest banks are excluded from the sample, suggesting that the relationship between concentration and profitability may act in a generalised structural way. In addition to very large banks, large banks and small banks also appear to benefit from concentration, but with no clear advantages to lower-middle-sized banks. Analysis of the effects of concentration on the components of profitability suggests that concentration may raise both interest and non-interest revenue, and reduce both interest and non-interest costs. Furthermore, concentration appears to depress bank deposit interest rates and raise both lending rates and the interest rate spread. This suggests that bank concentration might have negative effects on savings, investment, and growth.