Submitted by Sara Ribeiro (sara.ribeiro@ucb.br) on 2017-06-09T16:58:04Z No. of bitstreams: 1 WashingtonMartinsdaSilvaTese2016.pdf: 1011628 bytes, checksum: 2748e189905b2f6bfb5c040f50fff9b3 (MD5) / Approved for entry into archive by Sara Ribeiro (sara.ribeiro@ucb.br) on 2017-06-09T16:58:20Z (GMT) No. of bitstreams: 1 WashingtonMartinsdaSilvaTese2016.pdf: 1011628 bytes, checksum: 2748e189905b2f6bfb5c040f50fff9b3 (MD5) / Made available in DSpace on 2017-06-09T16:58:20Z (GMT). No. of bitstreams: 1 WashingtonMartinsdaSilvaTese2016.pdf: 1011628 bytes, checksum: 2748e189905b2f6bfb5c040f50fff9b3 (MD5) Previous issue date: 2016-03-31 / This paper aims to examine two topics in the context of growth and modernization of the electricity sector: the short-term energy price and the realization of auctions for construction and operation of electricity transmission assets. In the first analysis was applied a state space model with GARCH volatility to estimate the short-term electricity price explained by the conventional thermal power generation and energy stored seasonally adjusted. The data analyzed refer to the subsystem South East / Central West of the Sistema Interligado Nacional (SIN) between 2001:7-2014:10. The analysis results found evidences that the estimated coefficient for variable thermal generation varies over the sample analyzed and has a statistically significant relationship with the shortterm energy price. On the other hand, although there is statistically significant inverse relationship between the ratio of stored energy with seasonal adjustments and the spot price of electricity, there was no evidence indicating changes in the magnitude of the coefficient over the sample. The period analyzed was disturbed considering that there was a water crisis due to the cooling of the impact of rainfall in reservoir levels, with consequent increase in the volume of thermal power generation. In addition, this paper also analyzes the transmission auctions held in the Brazilian electricity market in the period 1999 to 2015. We used the Endogenouos Switching Regression Model, known as a Roy model, or Tobit type 5 model with copula approach, to analyze the discounts offered in the bids, winners and losers, made by bidders who participated in the auction. This methodological approach is to identify the difference between the proponents and check for evidence of the presence of selectivity, ie influence of unobservable characteristics in the discount offered by tenderers. The results indicated that there is a dependent relationship between the errors of the result function and errors in estimating the selection equation and the best functional form found to the problem was a combination of the copula function "Joe" and "Plackett". / Este trabalho tem como objetivo analisar dois t??picos inseridos no contexto de crescimento e moderniza????o do setor de energia el??trica: o pre??o de energia de curto prazo e a realiza????o de leil??es para constru????o e opera????o de ativos de transmiss??o de energia el??trica. Na primeira an??lise foi aplicado um modelo em Espa??o de Estados, com modelagem de volatilidade GARCH, para estimar o pre??o de energia el??trica de curto prazo sendo explicado pela gera????o de energia t??rmica convencional e pela energia armazenada com ajuste sazonal. Os dados analisados referem-se ao subsistema Sudeste/Centro-Oeste, do Sistema Interligado Nacional (SIN), para o per??odo de 2001:7 a 2014:10. Nos resultados da an??lise foram encontradas evid??ncias de que o coeficiente estimado para a vari??vel gera????o t??rmica varia ao longo da amostra analisada e possui rela????o direta e estatisticamente significativa com o pre??o de Energia de curto prazo. Por outro lado, embora exista rela????o inversa e estatisticamente significativa entre o coeficiente da Energia Armazenada com ajustes sazonais e o pre??o spot de energia el??trica, n??o foram encontradas evid??ncias indicando altera????es na magnitude do coeficiente ao longo da amostra. O per??odo analisado foi conturbado considerando que houve uma crise h??drica decorrente do impacto do arrefecimento das precipita????es pluviom??tricas nos n??veis dos reservat??rios, com consequente aumento no volume da gera????o t??rmica de energia. Al??m disso, este trabalho tamb??m analisa os leil??es de transmiss??o realizados no mercado de energia el??trica brasileiro no per??odo de 1999 a 2015. Foi utilizado o Endogenouos Switching Regression Model, conhecido como modelo de Roy, ou modelo Tobit tipo 5, com abordagem de c??pulas, para analisar os des??gios oferecidos nos lances, vencedores e perdedores, realizados pelos proponentes que participaram dos leil??es. Esta abordagem metodol??gica visa identificar o diferencial entre os proponentes, e verificar se existem evidencias de presen??a de seletividade, ou seja, influ??ncia de caracter??sticas n??o-observ??veis no des??gio oferecidos pelos proponentes. Os resultados indicaram que existe rela????o de depend??ncia entre os erros das fun????es de resultado e os erros na estima????o da equa????o de sele????o e que a melhor forma 6 funcional encontrada para o problema foi uma combina????o entre a fun????o c??pula ???Joe??? e ???Plackett???.