ESTIMATION AND INFERENCE WITH NEAR UNIT ROOTS

New methods are developed for identifying, estimating, and performing inference with nonstationary time series that have autoregressive roots near unity. The approach subsumes unit-root (UR), local (LUR), mildly integrated (MI), explosive (ME) specifications in the new model formulation. It is shown how a parameterization involving localizing rate sequence characterizes departures from unity can be consistently estimated all cases. Simple pivotal limit distributions enable valid about form degree of nonstationarity apply MI ME theory holds UR LUR Normalizing variance stabilizing properties explored. Simulations reported reveal some advantages this alternative formulation series. A housing market application conducted distinguishes differing forms house price behavior Australian state capital cities over past decade.