Regressão espacial e avaliação de terrenos: um estudo de caso para a Cidade de Fortaleza-CE

• Ano de defesa: 2019
• Autor(a) principal: Bandeira, Sandro Ricardo Vasconcelos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Regressão espacial; Avaliação de imóveis; Tributação Imobiliária Municipal
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/53258

The real estate taxes have in common the fact that they have market value as their basis of calculation, that can be described as the most probable amount for which a good would be voluntarily and consciously negotiated, on a reference date, within the conditions of the current market. In order to estimate the market value of a property, Brazilian standards elect several equally valid methods, among which the direct comparative market method stands out with statistical inference by adopting the Method of Least Squares (MLS). When we insert the first law of geography by Waldo Tobler (1970) in statistical study it can be noticed that the location of a property by itself plays an extremely important role in the definition of this value, assuming, therefore, a notable contribution to form property prices. Two macromodels were proposed where the first included only classical analysis, using the Method of Least Squares (MLS), considering safe parameters adjustments by using Newey-West procedures and finally by proposing an analysis through spatial econometric tools, also appreciating the spatial behaviors of price as a dependent variable (SAR) and the MLS model residuals (SER). Therefore, the performance of each one of the approaches will be measured through a sample test - which was not included in the initial modeling - while comparing the model precision by using Akaike and Schwarz statistics to define the one that presents the most accurate and consistent results. In this work, it’s been developed a study case to estimate venal values of land in Fortaleza (capitol of Ceará State, Brazil) in order to exemplify the theoretical studies presented as well as to solidify a better understanding about this matter. In conclusion, we may say that the use of the spatial model, based on the Spatial Contribution of the Price produced the best results among all those tested.