A spatial interaction model for branch bank accounts

Abstract The Rabobank Group is a Dutch cooperative banking organization consisting of a nationwide network of 935 local banks with 2358 branches. Rabobank Nederland is developing a Decision Support System (DSS) for planning the branch network. One of the models in the DSS regards the distribution of Rabobank customers over the branches. We developed a spatial interaction model for the distribution of current accounts over 41 branches. In this type of model two factors play a key role, the attraction of branches and a distance decay function. To estimate these two factors we divided the region under study into zones and we determined the distance between the centers of the zones. In doing so, we paid attention to barriers crossing the zones. In this paper we present the structure of the DSS, the spatial interaction model, data gathering, determination of distances, parameter estimation and results.