.However, the IPS test suffers froma substantial decrease in power when heterogeneous trends are included in the specification as a resultof the small-sample bias correction applied to thet-statistics (Baltagi and Kao, 2000a,b; Breitung, 2000).To address thisissue,Breitung (2000)proposes a panel unit root test which employs unbiasedt-statistics. By allowing for heterogeneousdeterministic trends and short-run dynamics across countries without the need of bias adjustment, the Breitung test hasmore power to reject a false null and is not sensitive to the degree of augmentation of the ADF specifications.
đang được dịch, vui lòng đợi..