SOME PROPERTIES OF EXACT TESTS FOR UNIT ROOTS

This paper is concerned with the null hypothesis that errors in a regr ession equation for time series data follow a random walk. We examine the power properties of most powerful invariant tests for the unit roo t null hypotheses against exact stationary and nonstationary first ord er autoregressive models. The analysis shows the importance of a const ant term and a linear trend variable in certain cases. The implication s of the results for models estimated using seasonal data are briefly discussed.