Efficiency evaluation under uncertainty: a stochastic DEA approach

In conventional data envelopment analysis (DEA) models, the efficiency measurement is carried out by using deterministic data typically referring to past observations. However, in many operative contexts, decision makers are called to predict the future performance for planning and control purposes. In these situations, ignoring the stochastic nature of data might lead to misleading results. The paper proposes a stochastic DEA approach based on the chance constrained paradigm and accounts for risk measured in terms of tail $$\gamma $$ γ -mean. A deterministic equivalent reformulation is presented under the assumption of discrete distributions. The computational experiments are carried out on an empirical case study related to the evaluation of the credit risk. The results demonstrate the validity of the proposed approach as proactive evaluation technique.