A new super-efficiency model in the presence of negative data

In a recent paper by Vencheh and Esmaeilzadeh, a super-efficiency model based on RDM model is formulated where some input and output data are negative. However, we show that the formulation in the paper is not always feasible or bounded for its purpose and cannot produce the numerical results claimed. Moreover, we demonstrate that the super-efficiency model based on RDM model is unable to provide a super-efficiency measure in some cases. To deal with this drawback we propose a new model, which is always feasible, to rank the DMUs in the presence of negative data.