Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/75680
Authors: 
Kunst, Robert M.
Reutter, Michael
Year of Publication: 
2000
Series/Report no.: 
CESifo Working Paper 286
Abstract: 
Decisions on the presence of seasonal unit roots in economic time series are commonly taken on the basis of statistical hypothesis tests. Some of these tests have absence of unit roots as the null hypothesis, while others use unit roots as their null. Following a suggestion by Hylleberg (1995) to combine such tests in order to reach a clearer conclusion, we evaluate the merits of such test combinations on the basis of a Bayesian decision setup. We find that the potential gains over a pure application of the most common test due to Hylleberg et al. (1990) are small.
Document Type: 
Working Paper

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.