Análise da persistência dos indicadores de continuidade do fornecimento de energia elétrica da Celg-D

• Autor(a) principal: Ribeiro, Viviane Pires
• Data de Publicação: 2018
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFG
• dARK ID: ark:/38995/0013000007ntx
• Texto Completo: http://repositorio.bc.ufg.br/tede/handle/tede/8650
• topic: Persistência aos choques; Conjuntos elétricos; DEC; FEC; Persistence to shocks; Electric sets; SAID; SAIF; CIENCIAS SOCIAIS APLICADAS::ECONOMIA

With the merger of Celg by the Enel group, one of the largest private companies in the Brazilian electric power sector, Enel Distribuição Goias has been facing serious problems related to its indicators of continued power supply, due to the fact that Celg-D has been in recent years, among the worst positions in the national ranking of service continuity and as one of the concessionaires that most compensates its consumers, directly affecting the company's results and generating losses to its shareholders. The requirement of the purchase agreement, defined by Aneel, is that, from 2019, the worsening trends in the SAID and SAIF indicators of Enel-GO will be reversed, under penalty of being intervened by the regulator and even losing the concession. In this scenario is the following question: Does the trend in the indicators of continuity of the Celg have the characteristic of persistence to the shocks, that is, it has long memory? The objective of this study is to analyze the persistence of the series of duration (SAID) and frequency (SAIF) of interruptions of the power supply of the company's electric sets during the periods prior to its privatization, from 2014 to 2016. Specifically, it seeks to test whether the series have long memory, to verify if the presence of structural breaks in the data changes the results and to identify if there is heterogeneity in the behavior of the series in different sets. For this, the methodology used is the method proposed by Robinson (1995) to analyze the persistence of the series and the Andrews and Ploberger (1994) test, with p-value using the approximations of Hansen (1997). The results suggest that there is a difference in the order of integration, depending on the type of series, duration or frequency and the electric set involved. Since the frequency series, in general, present more persistence than those of duration. When considering the potential presence of data breaks, there is a change in the order of integration of some series and in most of the sets the order found is less than 1. Therefore, both fractional integration analysis that considers the presence of break, do not consider, they point out that there are few sets that present persistent series and the majority of those who present are those that receive the highest compensations. However, it is not only the sets that receive large monetary amounts, in the form of compensations, that have series of persistent duration and / or frequency. Therefore, the hypothesis that part of the trend in the indicators has long-term components is partially rejected, in approximately 24% of the sets that received the highest DIC indicator compensations and in 40% of the sets that received the highest FIC compensation. In this way, the present work differs from the previous literature both in the sense of analyzing the behavior of the series of continuity indicators, with a methodology considered more robust, as well as by the database, not yet explored by the academy. Finally, the analysis allows both the direction of public and business policies and can be used by the company in making ex-ante decisions on the adoption of policies to reduce its indicators.