VALUATION OF PROPERTIES IN RIO GRANDE DO SUL: AN ANALYSIS FROM SPATIAL REGRESSION

Resumo

This work aims to determine the influence which the socioeconomic factors, along with the usual constructive variables, cause in the valuation of property in the state of Rio Grande do Sul. The real estate is an asset with distinctive features, different from others, and many researchers are pursuing different elements to explain this different behavior. In this regard, stands a very important factor, the neighborhood. The traditional inferential methods hinder the suitable modeling of the multiple factors that influence the value of the properties of a particular region. Thus, the spatial regression models were used to estimate the value of the same unit (VU). But the LAG regression model was used to standardize the sample data that showed heterogeneous properties. Nevertheless, the estimate of the kriging has demonstrated the value of a property to a particular region. The implementation of the methods was carried out to a database obtained from the Federal Savings Bank, containing properties traded in the state of Rio Grande do Sul, from 2005 to 2008. Moreover, applied regression methods indexes have confirmed significant indexes in the models obtained for all properties of a region. Thus, the spatial structure of indexes minimized the autocorrelation existing in the residuals of the regression model, improving the reliability of the assessment.