Using Non-Linear Tests to Examine Integration Between Real Estate and Equity Markets

In this study we present an alterntive approach to test whether the real estate and equity markets are cointegrated. We develop a nonlinear test which allows for a stochastic trend term as opposed to a deterministic drift term. This is a reasonable approach, because if the real estate market is related to the equity market then it is desirable to incorporate the stochastic nature of the equity market into the model. We compare the results of the nonlinear model to the results obtained using conventional cointegration tests. The cointegration results supports the view that the real estate and equity markets are segmented, whereas the nonlinear model support the view that the markets are fractionally intgrated. A possible reason for this apparent dicrepancy between the results could be due to the underlying assumption of a linear relationship between the variables for the cointegration. It is possible that the tests of cointegration will reject that the two variables are related even though the relationship may be nonlinear.