| Summary: | This study analyses the housing price volatility in metropolitan areas in South Africa, particularly Port Elizabeth and East London residential housing markets. This study uses secondary statistical data, obtained from secondary sources. The study uses quarterly time series data for the period 1981:1 to 2015:3 giving 139 observations. The data will be collected from different sources. The main sources of data are real estate agencies (Trafalgar, Harcourts and Property24), the South African Department of Trade and Industry (dti) and supplemented by the South African Reserve Bank (SARB) and Statistics South Africa (Stats SA). The study shall use the ordinary least squares (OLS) method to estimate its results. Ordinarily, this is a generalised linear modelling technique that may be used to model a single response variable which has been recorded on at least an interval scale. This method requires that the underlying stochastic processes of the variables are stationary. That is, explanatory variables should exhibit constant means and variances over time. If the stochastic processes are not stationary, OLS produces unreliably significant coefficients. Results showed that household savings, household income and total growth in household buildings (TGH) are statistically significant in explaining changes in house prices. Jointly, all the explanatory variables can account for almost 52% of the changes in the dependent variable. The Durbin Watson statistic showed that there is no autocorrelation in the model. This shows that the model is good. Results from the regression show that there is a negative relationship between house prices and household savings. A one-unit increase in household savings leads to a 0.407 decrease in house prices. This relationship makes economic sense because when households save, there is less income available to buy houses. When there is less income available to buy houses, it would mean there is less demand for houses.
|
|---|
