Valoração de imóveis no Rio Grande do Sul: uma análise a partir de regressão espacial

Abstract

This work aims to determine the influence that the social economic factors together with the usual variable constructive cause in the valuation of property, in Rio Grande do Sul State. Property is a material good with distinct characteristics from the other goods. Many researchers seek different elements to explain this situation. In this context one factor stands out with great relevance, the neighborhood. The traditional inferential methods difficult the adequate modeling because of the multiplicity of the factors that influence the value of the properties in a given region. Spatial regression models were used to estimate their unit value (UV). The regression model LAG was used to standardize the data sample of the properties, it was heterogeneous. Kriging showed the estimated value of a property for a given region. The application of the methods was performed for a database obtained from Caixa Econômica Federal, containing properties transacted in Rio Grande do Sul State, from 2006 to 2008. The applied regression methods confirmed strongly significant indices on the obtained models for all the properties in the region. The spatial structure of the estimated indices minimized the autocorrelation existing in the residuals of the regression models, improving the reliability of assessment.