On the instability of banking and other financial intermediation

BIS Working Papers  |  No 862  | 
08 May 2020


Are banks, or financial intermediaries more generally, inherently unstable and prone to volatility? Historically, even some of the staunchest proponents of laissez-faire such as Milton Friedman have viewed banking as inherently unstable and hence requiring government intervention.


In this paper, we investigate the logical possibility that financial intermediation might be inherently unstable. We remain agnostic and do not suppose a priori that financial intermediation engenders instability - or even if it does, we do not know if this is inherent in such activities or induced by policymakers' interventions. There is no generally accepted and all-purpose model of financial intermediaries to study this question, because these institutions perform a myriad of functions that are difficult to capture in a single setup. They serve as middlemen between savers and borrowers or asset sellers and buyers; they find, screen and monitor investment opportunities on behalf of depositors; and they provide liquidity insurance or maturity transformation, etc. So we analyse several formal models capturing the different facets of intermediation, as we want to know which ones, if any, lead to instability.


We find in each case that financial intermediaries can indeed engender instability: an economy with these institutions is more likely to have volatile dynamics than the same economy without them. In some cases, without intermediation the economy is stable, but with it becomes volatile; in others, intermediation expands the set of parameters for which the economy displays volatility. While the logic differs across models, in each case instability is directly related to the raison d'être for intermediation. Yet, while financial intermediation may be fragile in this sense, we emphasise that it still tends to increase welfare.


Are financial intermediaries inherently unstable and, if so, why? To address this, we analyse whether model economies with financial intermediation are particularly prone to multiple, cyclic or stochastic equilibria. Several formalisations are considered: a dynamic version of Diamond-Dybvig banking incorporating reputational considerations; a model with fixed costs and delegated investment as in Diamond; one with bank liabilities serving as payment instruments similar to currency in Lagos-Wright; and one with intermediaries as dealers in decentralised asset markets, similar to Duffie et al. Although the economics and mathematics differ across specifications, in each case financial intermediation engenders instability in a precise sense.

JEL classification: D02, E02, E44, G21

Keywords: banking, financial intermediation, instability, volatility