Abstract
This study investigates the long-run stochastic properties of real estate assets by geographical breakdown. We also study their linkages with financial assets. The initial tests find that almost all property types exhibit the presence of nonstationarity. Thus, cointegrated methodologies are used. Structural breakpoints identified in the literature are used as a guide to divide the data into two windows, 1983–1989 and 1990–1996. The results show that real estate in the different regions exhibit a closer relationship with each other in the second period, compared with the first. Also, strong linkages between real estate regions and financial assets are noted in the second period. The South is the only region to exhibit segmentation in both periods. Overall, the information derived from our analysis sheds light on linkages among real estate assets and between real estate and financial assets and also provides a framework for creating diversified portfolios.