Investigating The Characteristics of Problematic Residential Properties at

INVESTIGATING THE CHARACTERISTICS OF PROBLEMATIC RESIDENTIAL PROPERTIES AT PUBLIC AUCTION SALES USING MULTIPLE REGRESSION ANALYSIS (MRA)

Zahiriah Yahya1

Chan Wai Seen2

Md Nasir Daud1

1 Faculty of Built Environment, University of Malaya

2 JS Valuers Research & Consultancy Sdn Bhd

ABSTRACT

In line with the increase non-performing loan for the purchased of residential properties, an increasing number of residential properties go on auction in the market. The properties which are held as collaterals, are being auctioned off by financial institutions to recover the loans given to the purchasers. Typically, prices of properties offered for auction by financial institutions will be reduced by 10% after unsuccessful auction. However, many properties remained unsold despite being auctioned at a substantial discount from their market value. Unsuccessful auction sales of properties even after substantial price reductions are attributed to poor demand for the properties. The question is why. Poor location, poor neighbourhood condition and poor building condition have been cited as possible explanations. Nonetheless, the actual causes have remained poorly understood. For the purpose of this study, properties that remained unsold after a number of auctions are known as problematic properties. This study aims to improve the understanding on this issue. Based on the literature review and focus group meeting with the stakeholders and experts, the characteristics of problematic properties had been identified and thereafter adopted as variables for these studies. Data on problematic properties within the study area from September 2005 and December 2007 have been collected and analysed. The characteristics of these problematic properties have been identified through site inspections and title searches. These data were then used to create the Multiple Regression Model (MRA) and Geographical Information System (GIS) Analysis. The analyses done in this study have proven that the phenomenon of the problematic residential properties at auction can be statistically modeled and estimated. In line with the scope of works of this research projects, four regression models have been generated and are deemed to be the most appropriate and fit to explain and estimate the problematic residential properties in auction sale.

Keywords: problematic residential properties, public auction, multiple regression analysis