Develop a predictive model that drives business strategy to determine property sales within the real estate industry based in KwaZulu Natal.

Abstract

The study focuses on property owner’s attitudes, behavior and personal drivers when deciding to sell their real estate property. There have been limited studies performed around types of behavior that drives property owners to sell their property. Very little information around this topic exists in South Africa and this poses a risk for property buyers and estate agents of a residential property. Determining the drivers that influence a property owner to sell their property will generate property stock for estate agents in terms of identifying when an individual in their life cycle will sell their property. Apart from estate agents obtaining property stock from property sellers, they will also have the opportunity to sell the property seller another property. This quantitative study seeks to examine three suburbs within KwaZulu Natal residential property market and formulate a regression model to best predict what motivates an individual to sell their real estate property. The research included a seven-year sample period of residential property sales from 2010 to 2017; along with South African citizen data sourced from Home Affairs and public domain information and adopts a regression analysis to interpret the data at the relevant significance level. The 80:20 rule based on the Pareto principal was used to split the data in a test and train dataset, with the train subset being used to the build the predictive model and the test dataset to evaluate the model accuracy. The results from the analysis applied on the three suburbs within KwaZulu Natal indicates a good fit with an accuracy of 73.4% prediction of properties that are highly likely to go on sale. Variables applied to the study that are found to be statistically significant include: 1. The price of the property; 2. Age of the property; 3. Property owner’s age, gender, lifestyle indicator (LSM) and loan finance credit risk score; 4. Historical property sales data and 5. Population suburb density. The relevant results were then interpreted and recommendations provided to property estate agents that indicated why an individual would sell their property.