Using Artificial Neural Networks to Uncover Real Estate Market Transparency: The Market Value

In real estate property valuation, the availability of comparables is crucial. The reliability of the valuation of the market value depends on the number and on the accuracy of data that a professional can rely on. International standards suggest using historical prices as comparable since they are real transactions of sale/rent of a property that actually happened in a specific market. However, in the Italian real estate market, historical transaction prices are not available for professionals, and they have to base their valuations, primarily, on the asking prices enclosed in the selling advertisements. Asking prices can change in the future as they are subject to negotiation. Besides, sell ads always contain incomplete data or even wrong information. In this research, we employ Artificial Neural Networks to estimate how much offer prices and selling advertisements are misleading in property valuation in Italy. We, in a way, assess the opacity of the Italian real estate market, and we designate the major sources of error. The present work is a first step towards developing a model fitted for estimating data accuracy used generally in real estate estimates, namely, asking prices.