ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS", Bangkok, May 23-25, 2001 
ANALYSING BRANCH BANK CLOSURES USING GIS AND THE SMART MODEL 
Lihua ZHAO, Barry J. GARMER 2 
School of Geography, The University of New South Wales 
Sydney 2052 NSW Australia 12 
Tel: 0061-2-9385 5919 Fax: 0061-2-9313-7878 Email: lihua zhao@hotmail.com 1 
Tel: 0061-2-9385 5583 Fax: 0061-2-9313-7878 Email: B.Garner@unsw.edu.au 2 
Keywords: Branch bank closure, GIS business application, MCDM, SDSS, SMART model 
Abstract: 
Branch bank network planning requires the analysis and evaluation of alternative branch locations taking into account a number of 
criteria. Although these traditionally relate to financial performance indicators, the spatial reorganisation of networks implies that an 
increasing number of spatial variables must be included in the analysis. While GIS provide the analytical facilities to support spatial 
information, a multi-criteria approach is required for the evaluation of alternatives in order to arrive at rational and justifiable decisions, 
since branch bank closure problems involve multiple and conflicting factors. This underlines the necessity of analysing branch closures 
based on the integration of GIS and Multiple Criteria Evaluation models. 
The specific aim of this paper is the identification of branch banks that are potential candidates for closure based on the analysis of 
spatial and demographic variables. GIS and SMART (the Simple Multi Attribute Rating Technique) are integrated to develop a Spatial 
Decision Support System (SDSS) to carry out the modelling process. The visual and spatial components of the system are based on the 
use of the ArcView GIS software for managing, generating and integrating data, and for the production of graphic displays. The SMART 
model in the Criterium DecisionPlus software suite is used to generate a ranking of branch banks in order of decreasing viability. 
Dynamic Data Exchange (DDE) is used to loosely couple the two software packages in a Windows environment. Preliminary results of 
the application of the system to identify candidates for closure using the Commonwealth Bank network in the Sydney metropolitan area 
are presented to demonstrate the application of the GIS-based SDSS. 
1. INTRODUCTION 
Since the 1980s, competition in the retail banking market has 
intensified, and recently the single branch channel system that 
has dominated the delivery of banking products and services in 
the past is increasingly being replaced by new multi-channel 
systems (Mois, 1999; Leyshon and Pollard, 2000). The spatial 
distribution of retail banking in Australia has changed 
dramatically in the recent past as a result of deregulation and 
innovations in information technology. Today virtually all retailing 
banks are seeking ways for reducing their costly branch banks, 
the identification of branches for closure can have an important 
influence on the overall profitability of the bank's operation. 
As the presence of branches in particular places can have 
external benefit as individuals and businesses depend on them 
for their personal and business banking need, the focus of 
management decisions is on how many and which branches 
should be closed or, alternatively, how many and which 
branches should be retained to adequately serve customers. 
This process, which is one of searching for the best location or 
pattern of locations, is generally considered to be a discrete 
multiple criteria location problem (Malczewski and Ogryczak, 
1995). Usually, only a fixed number of alternative locations is 
available for consideration, the choice among which is based on 
a systematic evaluation of multiple criteria relating to the 
performance and potentials of individual branches. Multi- 
Criteria Decision Making (MCDM) is the formal approach that 
has been developed and widely used as the basis for solving 
such locational choice problems. 
Branch bank network planning requires the analysis and 
evaluation of alternative branch locations taking into account a 
number of criteria. Although these traditionally relate to financial 
performance indicators, the spatial reorganisation of networks 
implies that an increasing number of spatial variables must be 
included in the analysis. An alternative approach based on a 
consideration of demographic criteria and spatial variables can 
guide decision-makers in examining different alternative 
solutions before arriving at a decision, and help decision-makers 
define the problem and consider possible factors when deciding 
whether or not a branch should be closed. The conflict that 
arises between banks and their customers regarding branch 
closure stresses the importance for decision makers to consider 
these criteria (SMH, 1999; The Age, 1999a; The Age, 1999b). 
The fact that there is a spatial dimension to the problem means 
that GIS are an appropriate tool for analysis. The advantage of 
applying GIS in this context is that they enable links to be 
established and spatial relations to be explored between data 
derived from different sources. Moreover, the powerful GIS 
display functions enable the results of analysis to be presented 
visually in a variety of useful ways (Zhao and Garner, 2000). 
While GIS provide the analytical facilities to support the spatial 
information, a multi-criteria approach is required for the 
evaluation of alternatives in order to arrive at a rational and 
justifiable decision since branch bank closure problems involve 
multiple and conflicting factors. This underlines the necessity of 
analysing branch closures based on the integration of GIS and 
MCDM models. 
The specific aim of this paper is the identification of branch 
banks that are potential candidates for closure based on the 
analysis of spatial and demographic variables. GIS and SMART 
(the Simple Multi Attribute Rating Technique) are integrated to 
develop a Spatial Decision Support System (SDSS) to carry out 
the modelling process. Preliminary results of the application of 
the system to identify candidates for closure using the 
Commonwealth Bank network in the Sydney metropolitan area 
are presented to demonstrate the application of the GIS-based 
SDSS. 
2. SPATIAL DECISION SUPPORT SYSTEMS 
SDSS are based on the integration of GIS functionalities with 
decision-making models to create what have been called 
“intelligent GIS” (Birkin et. al., 1996). The definition and history 
of the development of SDSS have been described in the 
literature (see Armstrong et al., 1986; Densham and Rushton 
1988; Densham, 1991). The goal of a SDSS is to provide an 
interactive visual tool to help decision-makers understand the 
spatial component of both the problem and the solution, as well 
as the interrelationships between these. Such systems are 
capable of handling a wide variety of ill-defined or semi- 
structured business problems in a user-friendly environment. 
Clarke and Clarke (1995) have reviewed some successful 
business applications of SDSS and documented a number of 
key issues that SDSS have successfully addressed, such as, 
adequately meeting market needs, helping predict what might