Alternative methods of detrending and the power of unit root tests

This paper suggests unit root tests based on detrending the series by a GLS regression, using an empirically plausible value of the autoregressive root. These tests are related to the point optimal tests of Dufour and King. Monte Carlo experiments show a clear gain in power, relative to other unit root tests such as the Dickey-Fuller tests, over a large and empirically relevant range of the parameter space.