Computational Valuation Model of Housing Price Using Pseudo Self Comparison Method

Abstract

Hedonic pricing method (HPM), which is commonly used for estimating real estate property values, considers the property’s internal and external characteristics for its valuation. Despite its popularity, however, the method lacks the mechanism that directly reflects the target property’s price fluctuation and the real estate market’s volatility over time. To overcome these limitations, we propose Pseudo Self Comparison Method (PSCM), which reduces the real estate valuation problem to finding a pseudo self, which is defined as a housing property that can most closely approximate the characteristics of the target housing property, and adjusting its previous transaction price to be in sync with the real estate market change. The proposed PSCM is tested for two scenarios in which the volatility of the real estate market varies greatly, using the transaction data compiled from Seoul, the capital of South Korea, and its surrounding region, Gyeonggi. The study results show almost five times lower estimation errors when predicting housing transaction prices using the PSCM compared to the HPM in both scenarios and in both areas. The proposed method is particularly useful for mass valuation of apartments or densely located housing units.