Determinants of Residential Property Value in New Zealand: A Neural Network Approach

24 Pages Posted: 8 Nov 2016

Date Written: November 7, 2016

Abstract

Identifying the New Zealand housing market characteristics, the study examined the determinants of residential property price, and showed how different the marginal impact of these factors was. A semi-long linear model was estimated to assess the impact of demand and supply related factors on house prices based on prior studies. The study also employed the multi-layer neural network to evaluate the normalized importance of demand and supply determinants of the residential housing market. On average, both the semi-log-linear and the multi-layer neural network models supported the predictions made in the study that population, real GDP, immigration and housing credit on the demand side, and gross fixed capital formation (GDP) and residential investment on the supply side are the major determinants of residential property prices in New Zealand. Although both the models had shown promising results, the multi-layer neural network approach produced better results compared to the semi-log-linear model.

Suggested Citation

Abraham, Mathew Puravady, Determinants of Residential Property Value in New Zealand: A Neural Network Approach (November 7, 2016). Available at SSRN: https://ssrn.com/abstract=2866115 or http://dx.doi.org/10.2139/ssrn.2866115

Mathew Puravady Abraham (Contact Author)

Whitireia Business School ( email )

Level 2, 450 Queen Street
PO Box 106 219
Auckland, Auckland 1143
New Zealand

Register to save articles to
your library

Register

Paper statistics

Downloads
95
Abstract Views
522
rank
270,879
PlumX Metrics