Financial Intermediation and Monetary Policy in a General Equilibrium Banking Model

Article excerpt

THE PROCESS by which monetary policy affects the level of financial intermediation, or the "credit channel," has been the focus of a great deal of research recently; see, for example, the survey paper by Hubbard (1994). An important issue, which is a focus of this paper, is when does expansionary monetary policy result in an increase in lending and when does the policy simply lead to more inflation. While there are many types of financial intermediaries, I focus on intermediation through a traditional banking system in which banks are subject to reserve requirements. Banks act as an intermediary between households and the government and between savers and producers (borrowers). Banks are able to raise funds by holding demand deposits and by issuing equity capital. Monetary policy is implemented by open-market operations and changes in nominal reserves. The policies are linked through the government's budget constraint.

By developing a model with a traditional banking sector, the effects of monetary policy on the balance sheet of the bank can be explored. Monetary policy can lead to a substitution among a bank's assets, which affects nominal lending and possibly real output. It can also lead to a change in the size of the bank's balance sheet by changing demand deposits and equity capital (the model has the feature that there is an optimal deposit--equity capital ratio). The real value of deposits and equity capital is equal to real savings in the economy. Prices adjust so that the real savings just equals the real value of the assets in the banking sector. Monetary policy may have real effects because it determines the amount of real wealth transferred between savers and dissavers. The banking model is related to the banking model developed by Taggart and Greenbaum (1978). Some of the same issues studied here have been raised in Bernanke and Gertler (1987).

On the lending side, borrowers are ex ante identical but, because of moral hazard and costly state verification, there is credit rationing in equilibrium; the lending side of the model is based on the Gale and Hellwig (1985) model and the Williamson model (1987). Banks have access to a monitoring technology, which gives them a comparative advantage in lending, that motivates the existence of a financial intermediary; this is similar to the approach taken by Boyd and Prescott (1986). The distributions of returns to producers depends on a state variable that is not observable at the time the loans are made. Hence, the actual return on loans received by banks is random and depends on aggregate shocks, although bank lending averages out idiosyncratic shocks.

When banks are able to write state-contingent standard debt contracts and monitoring costs are indexed for inflation, the real return to lending is unaffected by inflation. While government policy affects the size and composition of nominal assets available to the banking sector, prices adjust so that real transfers between savers and dissavers are unaffected by the policy. Hence, real lending and consumption allocations are unaffected by policy changes. Banks adjust their equity capital and depository liabilities to maintain their desired deposit-equity ratio. When real returns are affected by actual inflation, which is modeled here by assuming that monitoring costs rise (fall) by a smaller proportion when prices rise (fall), prices and real lending increase in response to an increase in nominal transfers (expansionary monetary policy).


There are four types of agents in this model: households, producers, banks, and the government. Households live for only two periods; each period there are two types of households--young and old (overtapping generations). Young agents receive an endowment y while old agents must provide for consumption by saving. Young agents decide how much to consume and how much to save of the perishable endowment, selling the portion saved in the goods market at a unit price of p. …