Portfolio Theory, Transaction Costs, and the Demand for Time Deposits

Abstract

Households do not rebalance their deposit portfolios in response to 200-300 basis point changes in relative yields. Is it because the deposits are poor substitutes or because transaction costs make it nonoptimal to rebalance? This study uses efficient frontier techniques from portfolio theory and a transaction-cost model to address these questions. The major findings are that the transaction-cost model explains deposit shares but the portfolio model does not and the gains from rebalancing are minuscule because banks made large changes in relative yields on poor substitutes while maintaining fairly constant spreads on close substitutes. Copyright 1995 by Ohio State University Press.