Unit‐root testing against the alternative hypothesis of up to m structural breaks

In this paper we provide tests for the unit‐root hypothesis against the occurrence of an unspecified number of breaks which may be larger than 2 but smaller that the maximum number of breaks allowed, m , in univariate time‐series models. The advocated procedure is considerably less computationally intensive than those widely used in the literature. We provide critical values for the test and examine its small sample properties through Monte Carlo experiments.