Diamond–Dybvig and beyond: On the instability of banking

Abstract

Are financial intermediaries – in particular, banks – inherently unstable or fragile, and if so, why? We address this theoretically by analyzing whether model economies with financial intermediation are more prone than those without it to multiple, cyclic, or stochastic equilibria. Several formalizations are considered: Insurance-based banking; models with reputational considerations; those with fixed costs and delegated investment; and those where bank liabilities serve as payment instruments. Importantly for the issue at hand, in each case banking arrangements arise endogenously. While the economics and mathematics differ across specifications, they all predict that financial intermediation engenders instability in a precise sense.