Blockwise bootstrap testing for stationarity

This paper proposes a bootstrap test for the null hypothesis that a stochastic process is stationary against the alternative hypothesis that it is integrated of order 1. The test is constructed by using a stationary bootstrap scheme, which involves resampling blocks of consecutive observations of random length. The first-order asymptotic correctness of the stationary bootstrap test is established for a large class of weakly dependent processes. The small-sample properties of the method are also investigated by means of Monte Carlo experiments.