Testing the unit root hypothesis against the logistic smooth transition autoregressive model

摘要: In this paper two simple tests to distinguish between unit root processes and stationary nonlinear are proposed. New limit distribution results provided, together with F type test statistics for the joint linearity hypothesis against a specific alternative. Nonlinearity is defined through smooth transition autoregressive model. Due occasional size distortion in small samples, bootstrap method proposed estimating p-values of tests. Power simulations show that have at least same or higher power than corresponding Dickey-Fuller Finally, as an example, applied on seasonally adjusted U.S. monthly unemployment rate. The linear strongly rejected, showing considerable evidence series better described by process random walk.