Abstract:
Time series of the International Urban Housing Market, whether stationary or not, look highly irregular and yet they are often summarily fitted in linear models for forecasting purposes. This paper investigates the issue of nonlinearity by first extending the power of Range Rescaling analysis to uncover and to identify any type of nonlinearity that might be present in temporal sequences and next applies it to monthly supply and prices data (from 1949-01 to 1995-01) of the Canadian residential housing market.