Uso da regressão linear múltipla na avaliação de imóveis urbanos no oeste potiguar: estudo de caso em Assú/RN

Abstract

This paper uses multiple linear regression in the evaluation of urban properties in Assú/RN. For this, it is based on NBR 14653-2 in order to obtain an equation that estimate it, verifying if the implementation of this method would make tax collection in the city, and estimate if it would increase or decrease. For this, it is based on NBR 14653-2 that suggests the use of the Direct Comparison Approach Method for the identification the value of an urban property, using linear regression in the determination of the equation to calculate the market value of a property. Through the Municipal Tax Department, which uses the venal value as a basis for calculating the Real Property Transfer Tax (Imposto de Transmissão de Bens Imóveis - ITBI) and the Property Tax (Imposto Predial e Territorial Urbano - IPTU), data were collected from urban residential properties traded January 2017 to December 2018, obtained from the (Boletim de Cadastro Imobiliário - BCI), the Registration Certificate and the tax slip of the ITBI. The initial model presented six independent variables, which were reduced to three after the treatment of the data through the software SisDEA, and the venal values calculated from the model obtained in the study were compared both to the transfer prices informed for the ITBI generation and to the venal values that the municipal system allocates for each of these properties. The proposed technique obtained representative and approximate values of the transactions, at the same time that they were much higher than those practiced by the system that the Secretariat has. The work showed that the municipality loses revenue, especially in the collection of the IPTU, and the implementation of the new evaluation method is reliable and would make it fairer to the taxation of citizens, including increasing municipal revenue, making it possible to be a study base for the Secretariat and replicated to the various types of properties belonging to your profile.