In this paper, a new procedure to test for stationarity is proposed. The new test has three features. First, it is a point optimal test because its derivation is based on the Neyman-Pearson principles. Second, it achieves what is called a near ideal asymptotic rejection (NIAR) profile in the sense of Müller (2005). And third, it provides information about the potential relationship between the empirical size and the previously adopted significance level. The simulations presented at the end of the paper corroborate the asymptotic results and show that its performance improves that of other tests.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados