This study tests whether a “portfolio effect” exists in a given branch bank; i.e., does the addition of branches reduce the variability of demand deposits for the bank? The approach taken is narrower than that in the usual portfolio selection model. The study measures risk by the intrayear coefficient of variation of average monthly demand deposits. The term “portfolio effect” was defined operationally as a reduction in the overall coefficient of variation of demand deposits through the addition of sets of branches. Relatively few portfolio effects were realized from the chronological addition of sets of branches. The absence of portfolio effects is largely attributable to (1) generally high positive correlations between deposits of the various sets of branches and the defined deposit base and (2) the small size of the sets of branches relative to the base deposits. Further, while the correlations were generally high and positive, they were also serially unstable which suggests they would be poor predictors of future correlations. Based on the experience of this bank, it does not appear that the reduction in demand deposit variability by adding branches is general or consistent enough to facilitate improved management of reserves or selection of branch locations. However, this conclusion does not necessarily imply that the bank should not have undertaken branch expansion. For example, this study does not include an analysis of branch profitability. While the results of this study based on a single branch bank cannot be generalized, they do suggest the need for a more comprehensive analysis of the impact of branching on the variability of demand deposits.
