Maximum Likelihood Estimation of Variance Components in a Competitive Electricity Market

Competitive electricity markets are based on a set of hypotheses which sustains a vertical and horizontal segmentation of the generation, transmission and distribution sectors in different functions. In Argentina, this theoretical model was implemented in a wholesale electricity market (MEM in Spanish) and its operation for more than a decade produced a great amount of data which are analyzed by the Public Authorities to take further decisions. Proposals regarding reforms in the design of the Argentinean MEM must be based on a deep knowledge of the influence exerted by different factors connected with the main technological and economic MEM parameters. Consequently, it is important to have methodologies and statistical techniques which help Public Authorities to take decisions. In this paper, we present the application of the methodology of analysis of variance of data (ANOVA) to detect the influence that Large Users (GUMA in Spanish) has on the Contract (Forward) Market in the Argentinean MEM. A random-effect model of two factors is used: a ten-level factor for Large Users and a four-factor for months. The random variable observed correspond to the Contract Market price in Argentinean MEM during 2.003, 2.004, 2005 and 2.006.