Estimating Efficiency Performance of Decision-Making Unit by using SFA and DEA Method: A Cross-Sectional Data Approach
In this paper, a cross-sectional samples data of 115 Malaysian stocks have been employed to compare both Data Envelopment Analysis (DEA) method and Stochastic Frontier Analysis (SFA) method. These approaches are used to provide a review of frontier conceptual measurement, strength and limitation of the parametric and non-parametric models. Stochastic frontier production function of Cobb-Douglas type was utilized for the estimation. The function was estimated using the maximum likelihood estimation technique. Two models in DEA, DEA-CCR and DEA-BCC are applied in this study and the ranking correlation between SFA method and both models DEA are determined by using the Spearman rank method. The result revealed using SFA, the mean technical efficiency of sample consumer product companies is 37.5% and implies that companies operating at means level of technical efficiency could produce 80.1% more output for given level of inputs if they become technically more efficient. From empirical results of the SFA method, we determined that the deviations from the efficient frontiers of production functions are largely attributed to inefficiency effects (technical inefficiency). Finally, the findings also showed that the difference in ranking stocks performance using DEA-CCR, DEA-BCC and SFA methods. The main contribution of the paper is showing the comparative performance based on both model, DEA and SFA method using financial ratio.