This diploma thesis deals with the stochastic frontier approach -- a tool for the analysis of technical efficiency. First, general outline, basic terms and approaches to efficiency analysis are overviewed. Also the theoretical background of production functions is involved, focused on the Cobb-Douglas production function. Furthermore, the frontier approach and the stochastic frontier approach for cross-sections data is covered. Based on this, stochastic frontier model with time-varying technical... show full abstractThis diploma thesis deals with the stochastic frontier approach -- a tool for the analysis of technical efficiency. First, general outline, basic terms and approaches to efficiency analysis are overviewed. Also the theoretical background of production functions is involved, focused on the Cobb-Douglas production function. Furthermore, the frontier approach and the stochastic frontier approach for cross-sections data is covered. Based on this, stochastic frontier model with time-varying technical efficiency for panel data is introduced with respect to maximum likelihood estimation. The second part offers an application of the model to assess technical efficiency of ice hockey teams in the NHL for the 1997-2013 period. Output of the production is measured by the winning percentage, inputs are salaries of three categories -- goaltender, defenseman and centers/wingers. |