Bewley Banks

How do movements in the distributions of bank size and income affect the macroeconomy? To answer this question we develop a dynamic general equilibrium model with heterogeneous financial intermediaries, incomplete markets, and aggregate uncertainty. We find that market incompleteness and uninsured idiosyncratic bank rate of return risk generate minimal concentration in the bank net worth distribution, leading to an "as-if" result, whereby the economy behaves as if it had a representative bank. However, introducing ex-ante heterogeneity in the banks' rates of return significantly raises concentration and amplifies real and financial fluctuations relative to the representative-bank case, as this increases a key sufficient statistic, the average marginal propensity to lend. We then extend the model with two empirically-validated features of the banking sector - countercyclical return risk and deposit market power - and show that these amplify and dampen aggregate fluctuations, respectively. Finally, because in the  model with ex-ante heterogeneity the distribution of bank size is highly concentrated, shocks to the largest banks can account for almost all of the aggregate variation that is due to idiosyncratic risk, leading to granular banking and economic cycles. The failure of granular banks ("too big to fail") produces sizeable macroeconomic crises.