Auction price index model for residential properties

Abstract

The purpose of this study is to develop an auction price index model for residential properties. The model is needed to solve the problem of the unavailability of an auction property price index in Malaysia which is important in monitoring market price movement. The thin nature of auction property market has prompted this study to explore the applicability of rank transformation regression model for the auction property market data instead of using the conventional hedonic price model as employed in the Malaysian House Price Index. Data of high-rise auction property sold from year 2010 to 2014 in Kuala Lumpur was collected from Lelongtips.com. Review of journals, newspapers, legal acts as well as interview with valuers, auctioneers and lawyers were conducted in order to determine the factors to be considered in developing the auction property price model. The rank transformation regression model was analysed, its predictive power and precision performance were validated. This is followed by the application of hedonic Laspeyres Index method to construct the auction property price index. The result shows that rank transformation regression price model is applicable for traded auction property data. The quarterly Auction Price Index built from 2010Q1 to 2014Q4 revealed that the residential auction sale price is on an increasing trend, averaging at 2.4%, with some fluctuations. The construction of auction property price index through the integration of quarterly rank transformation regression and hedonic Laspeyres Index gives the benefit of a more accurate model prediction without the problem of geographical or temporal aggregation.