Beau Dunker
Spatial Variation in Bank Branch Performance © 2004
The purpose of this paper is to identify the key variables associated with the gross revenue of a subset of Canadian bank branches from a major financial institution. This study can be used to rationalize the current branch network, or used in forecasting the potential revenues at new branches. The performance of a branch is a difficult measure without a benchmark, in which individual branch performance can be compared. Three research hypotheses were investigated. Each hypothesis was tested using multivariate step-wise regression models to identify the drivers of gross revenue within a specific market. A model was created for all branches within the cities of Vancouver, BC, Edmonton, AB, Mississauga, ON, Toronto, ON, and Montreal, QC. The second model was created using a subset market of branches in each of the 5 cities. A third model was created for suburban and urban areas in Montreal and Toronto. Finally, a model was created using a database without branch attributes to quantify the explanatory power these variables possess with gross revenue. The key variables identified in each model varied, as did the explanatory strengths. Each model was used to quantify the variance in performance in various spatially segregated markets.