The Rent Service
Property value predictor proves its worth in the rental sector
Geofutures’ property valuation space-time engine model was put to work in a pilot study for The Rent Service (TRS), an executive agency of the Department for Work and Pensions.
The model mimics the ‘human’ assessment technique of predicting rental values based on similar transactions close together in space and time.
Developed in collaboration with ESRI (Environmental Systems Research Institute), the tool is designed to provide market-based property rental values online at the touch of a button.
The model runs in parallel with a hedonic model using geographically weighted regression to assess the influences of multiple factors driving the rental market, including fine-scale property classifications, income levels and local schools.
The latter is used to refine and calibrate the outputs of the space-time engine. Taking Devon as a test area to ensure its usefulness across rural and urban localities, the pilot study produced rental value predictions with an impressively high standard of accuracy against actual rental data.
The space-time engine was first developed for a project for Transport for London and others, to model the response of the residential property market to infrastructure investment. In this study the techniques used proved readily adaptable to the rental market.
