The simultaneous identification of strategic/performance groups and underlying dimensions for assessing an industry's competitive structure

The purpose of this paper is to integrate aspects of the literature on strategic and performance groups and explicitly derive strategic/performance groups which exhibit differences with respect to both strategy and performance, as well as display associations and potential interrelationships between the two sets of variables.

A two‐way clusterwise bilinear spatial model was formulated (e.g. a scalar products or vector multidimensional scaling model (MDS)) for the analysis of two‐way strategic and performance data which simultaneously performs MDS and cluster analysis. An efficient alternating least‐squares procedure was devised that estimates conditionally globally optimum estimates of the model parameters within each iterate in analytic, closed‐form expressions.

This bilinear MDS methodology was deployed in the context of strategic/performance group estimation using archival data for public banks in the NY‐NJ‐PA tri‐state area. For this illustration, four strategic/performance groups and two underlying dimensions were found.

Consideration of both strategy and performance data should be employed in describing the heterogeneity amongst firms competing in the same industry.

The paper provides a new spatial methodology to derive strategic/performance groups in any given industry to more completely summarize intra‐industry heterogeneity.