Investigation of the relative efficiency for the Greek listed firms of the construction sector based on two DEA approaches for the period 2006–2012

This paper applies two different approaches of data envelopment analysis (DEA) in order to investigate the relative efficiency for the Greek listed firms of the construction sector before and during the recession (before recession: 2006–2008; in recession: 2009–2012). In the first stage of this study an output DEA version based on financial ratios is introduced. The main contribution of this approach concerns the use of the Recursive Partitioning Algorithm in order to establish two different groups of variables, one for each sub-period. This technique enhances the estimated power of the first version of DEA model, since it provides for each period the ratios with the highest estimated value and the proposed model adjusted to the current economic circumstances. Although the results show that the number of the inefficient firms remains the same for both sub periods, there is an adjustment in the classification of non efficient and efficient firms. The second model is based on an input–output version of DEA with the use of accounting data as variables. This model classifies the majority of the firms as inefficient for the first sub-period, while the percentage of the efficient firms increases during the second sub-period, according to their average efficiency scores. In order to estimate the reliability of the DEA model a bootstrap method is also applied. The results are significant for discussion especially in relation to the assets’ management policy of firms before and during the recession.