Title: Unit root test in a semiparametric model
Authors: Sarah Bernadette Aracid - University of the Philippines Los Banos (Philippines) [presenting]
Erniel Barrios - University of the Philippines (Philippines)
Joseph Ryan Lansangan - University of the Philippines (Philippines)
Abstract: Presence of unit root in time series data is implicated in the persistent effect of random shocks in the behavior of a model, leading most unit root tests to be incorrectly-sized or have low power or both. A nonparametric test for the presence of unit root is proposed. To better understand the instance where unit root occurs, hence, mitigate the possible problem of present unit root tests, it is assumed that another time series $x_t$ possibly affect the target time series $y_t$ in addition to the autocorrelation dynamics. A nonparametric effect of $x_t$ can spare the autocorrelation structure from further contaminations, hence, the test can characterize presence of unit roots in $y_t$ easily. Simulation study showed that the proposed test yields better size and power compared to some popular tests for unit root.