Profitability Analysis Using Data Envelopment Analysis-Discriminant Analysis: an Empirical Study

This paper addresses the profitability analysis of 31 chain stores belonging to a hairdressing company in central Taiwan. Four factors are adopted in the analysis with data of year 2010. The 31 stores are classified to four groups according to their levels of return on investment (ROI). The 31 stores are discriminated with apparent rate of 93.5% by a four-group data envelopment analysis-discriminant analysis (DEA-DA) model. The discriminant results reveal that two factors have positive effect for classifying a store to the group of high ROI level, while two factors have negative effect. This empirical study shows that DEA-DA is a suitable tool for the study company to discriminate the ROI attained in last year and can reveal the improvement aspects. Besides, the discriminant function obtained can be used to predict the ROI level of the oncoming year.