“…If the PVM is correct, then land rents and land prices should have the same time series properties, and the spread, defined as the stationary linear relationship between land rents and prices, should add useful information in forecasting future changes in rents given past changes in rents. A number of studies tested for two main reasons of rejection of the PVM: time‐varying discount rates (Falk, 1992; Hanson and Myers, 1995) and the presence of speculative bubbles (Featherstone and Baker, 1988; Tegene and Kuchler, 1993; Falk and Lee, 1998). Following this literature, co‐integration testing has become a routine in many studies, as a preliminary step before the empirical application of the PVM.…”