Performance evaluation using bootstrapping DEA techniques: Evidence from industry ratio analysis

In Data Envelopment Analysis (DEA) context financial data/ ratios have been used in order to produce a unified measure of performance metric. However, several scholars have indicated that the inclusion of financial ratios create biased efficiency estimates with implications on firms’ and industries’ performance evaluation. There have been several DEA formulations and techniques dealing with this problem including sensitivity analysis, Prior-Ratio-Analysis and DEA/ output–input ratio analysis for the assessment of the efficiency and ranking of the examined units. In addition to these computational approaches this paper in order to overcome these problems applies bootstrap techniques. Moreover it provides an application evaluating the performance of 23 Greek manufacturing sectors with the use of financial data. The results reveal that in the first stage of our sensitivity analysis the efficiencies obtained are biased. However, after applying the bootstrap techniques the sensitivity analysis reveals that the efficiency scores have been significantly improved.