Deposit size, deposit mix and commercial bank asset portfolio composition



Journal Title

Journal ISSN

Volume Title



This dissertation is a study of the aggregate portfolio behavior of the United States commercial banking system during the period 1956-1965. Specifically, the study examines the roles played by deposit size-total net deposits held by the banking system-and deposit mix-the ratio of time deposits to total deposits-in determining the composition of the aggregate asset portfolio of the commercial banking system. The study also examines the effect of relative increases in public deposits, both federal and state and local, on the composition of the asset portfolio. The model used in the study treats commercial bank management as a risk averting investor attempting to maximize a utility function whose only arguments are expected return and risk. The model abstracts from all risk except that associated with unexpected deposit losses. As the risk of unexpected deposit losses rises bank, management must hold an increased portion of the asset portfolio in highly liquid forms in order to meet cash drains or run the risk of capital losses when relatively illiquid assets are sold under adverse market conditions. Evidence presented in the study indicates that deposit stability increases-i.e., the risk of unexpected deposit losses falls-as deposit size increases and as the ratio of time deposits to total deposit rises. Thus, as deposit size increases and the ratio of time deposits to total deposits rises, bank management can, and should, shift the composition of their asset portfolio toward longer-term, higher-yield assets. The empirical results of the study clearly indicate that deposit mix affects asset portfolio composition in the expected manner. The effect of deposit size is somewhat ambiguous, but it is generally in agreement with a priori expectations. Public deposits differ from regular deposits in that the bank holding such deposits must pledge collateral against the portion of the deposit in excess of FDIC insurance ceilings; the pledged assets are generally transferred to the depositing agency or its designated custodian. Although many securities can be used as collateral, U.S. Government securities are most often used. This study seeks to discover if such pledging requirements affect portfolio composition. Empirical findings indicate that as public deposits rise relative to total deposits asset portfolio composition is affected in two ways. First, the portion of U. S. Government securities maturing within one year falls as banks switch to longer term securities for pledging purposes. Second, the portion of the portfolio consisting of mortgages falls. The unique features of the model used are perhaps more interesting than the results obtained. First, the model does not deal with dollar holdings of each asset, rather, it deals with the percentage of the portfolio held in each asset. Second, it treats the entire asset portfolio and not just the portion remaining after reserve and liquidity requirements are met. Finally, the model does not assume the distributed lag form usually found in studies of the aggregate portfolio behavior of financial institutions. A detailed discussion of the rejection of the distributed lag model is included in the study.