Introduction

In general the assets of the banking system (the liabilities of the borrowers) are less liquid than its liabilities (the assets of the depositors). This is an obvious claim, and almost every textbook on the matter will include the transformation of illiquid assets into more liquid liabilities amongst the reasons for the very existence of a banking system. Some headway has recently been made in the modelling of this process (Diamond and Dybvig (1983), Smith (1984)). There are two key features which are common to all theoretical models of liquidity transformation. The depositors are unsure as to when they will want to withdraw, and production technology is such that early disinvestment is penalized; production takes time.

This paper also makes use of these two features. Its aim is precisely that of gaining further insight into the liquidity transformation process and its consequences. The specification of consumers which I use is more complex than that used in previous models in this line. The consequences are far reaching and I contrast the findings of this paper with results previously obtained in the concluding section.

The combination of an illiquid technology with consumers' ‘desire for flexibility’ is enough to create the need for a financial intermediary with illiquid assets and more liquid liabilities. It is interesting that this is so without the introduction of any uncertainty in the technology. The banking system, transforming illiquid assets into more liquid ones for its depositors, provides a Pareto–superior equilibrium for the economy.