Sources of nonmonotonic power when testing for a shift in mean of a dynamic time series

摘要

Sources of nonmonotonic power are uncovered for a wide variety of tests for a shift in the mean of a dynamic time series. Two main sources of nonmonotonic power are found. The first source is the behavior of the estimate of error variance under the alternative hypothesis of a shift in mean. In particular if the error variance is estimated under the null hypothesis, nonmonotonic power can result. The second source is the presence of a lagged dependent variable in the estimated regression.