We introduce a test for strict stationarity based on the fluctuations of the quantiles of the data, and we show that this test has power against the alternative hypothesis of unconditional heteroskedasticity while other tests for first order (level) stationarity as the KPSS test (Kwiatkowski et al., 1992) and, its robust version, the IKPSS test (de Jong et al., 2007) have low power against this alternative of time-varying variance. Moreover, our test has power against the alternative hypothesis of time-varying kurtosis, while the test for second order (covariance) stationarity introduced by Xiao and Lima (2007) has power close to size against this alternative. © 2013 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Lima, L. R., & Neri, B. (2013). A test for strict stationarity. In Advances in Intelligent Systems and Computing (Vol. 200 AISC, pp. 17–30). Springer Verlag. https://doi.org/10.1007/978-3-642-35443-4_2
Mendeley helps you to discover research relevant for your work.