Implicit Prices Associated To The Main Causal Attributes In Real Estate Valuation, Obtencion De Precios Implicitos Para Atributos Determinantes En La Valoracion De Una Vivienda

An important question in many real estate markets is how to effectively identify property fair values using internal or external variables. In this paper, there are two main objectives. First, we estimate the value of a dwelling using the econometric models Artificial Neural Networks (ANN) and Classical Hedonic pricing models. Second, we obtain implicit prices of the main variables associated with the valuation process, comparing both in a case study. The ANN approach is preferred for two reasons. First because the degree of fit is better than for hedonic models and forecasted values are closer to observed transaction prices. Second because implicit prices for the main causal variables are closer to the buyer’s valuation. The ANN models are thus closer to the real behavior of the agents involved, than hedonic models. A case study with 2888 transactions is presented, corresponding to a medium size urban area in the South of Spain.