A real-time mix-adjusted median property price index enabled by an efficient nearest neighbour approximation data structure

Abstract

Homeowners, first-time buyers, banks, governments and construction companies are highly interested in following the state of the property market. Currently, property price indexes are published several months out of date and hence do not offer the up-to-date information which housing market stakeholders need in order to make informed decisions. In this article, we present an enhanced version of a mix-adjusted median based property price index which uses geospatial property data and stratification in order to compare similar houses sold in different trading periods. The expansion of the algorithm to include additional parameters, enabled by both a richer dataset and the introduction of a new, efficient data structure for nearest neighbour approximation, allows for the construction of a far smoother and more robust index than the original algorithm produced.