MULTIFACETED ANALYSIS OF THE PROCESS OF PRICE DEVELOPMENTS ON THE LOCAL REAL ESTATE MARKET BY MEANS OF THE REGRESSION TREES METHOD (C&RT)

Attention of rational purchaser or seller of real estate is focused on the prices and properties of similar real estates. In the light of the theory of valuation, differentiation in prices on the local market is explained by variation in a few most important properties of marketed real estate. Thus, in the practice of valuation, a important issue is to decide which of many possible attributes are significant and what is their individual contribution to the market price that can be achieved. In this paper, this problem has been solved using the model of regression trees (C&RT). The study covered undeveloped land designed in local plans for single-family housing development, situated on the northern outskirts of Krakow. A regression tree permitted a simple assessment of the impact of the different attributes on real estate prices. This method enables prediction of the market value, but it requires knowledge of at least several tens, or preferably many hundreds, of transactions relating to similar real estates. For that reason, it is predestined for mass valuations.