Testing for smooth transition nonlinearity in the presence of outliers

Regime-switching models, like the smooth transition autoregressive [STAR] m odel, are typically applied to time series of moderate length. Hence, the n onlinear features that these models intend to describe may be reflected in only a few observations. Conversely neglected outliers in a linear time ser ies of moderate length may incorrectly suggest STAR (or other) type(s of) n onlinearity. In this article we propose outlier robust tests for STAR-type nonlinearity. These tests are designed such that they have a better level a nd power behavior than standard nonrobust tests in situations with outliers . We formally derive local and global robustness properties of the new test s. Extensive Monte Carlo simulations show the practical usefulness of the r obust tests. An application to several quarterly industrial production inde xes illustrates that apparent nonlinearity in time series sometimes seems d ue to only a few outliers.