Three techniques of determination of the real estate price increase index, which are based on the analysis of pairs of similar estates, are compared. The indexes are calculated for a data base consisting of more than one hundred re-sale transactions of undeveloped real estates. The techniques analyzed differ in the weights of the re-sales transactions. The comparison reveals that the method preferring long time intervals between the subsequent sale transactions gives the best estimate of the price increase index, which is close to that calculated form linear regression.

5-13