STOCHASTIC FRONTIER MODELS BY COPULAS AND AN APPLICATION

The main endeavor in this study is to use copulas in the stochastic frontier analysis (SFA) to estimate production frontier and also technical efficiency of DMUs. After presenting theorical discussion using copulas in SFA, we finalize our study by providing an application on real data set. We compare the yielded efficiency scores by using copulas with results of the standard SFA and also CCR and BCC models. We concentrate on several Archimedean families and specially three families which were newly presented to the literature. These new families have trigonometric and hyperbolic generators and they are more flexible in modeling dependence structures.