摘要
中国出口月度数据是一个很重要的外贸数据,它存在明显的季节性。通过季节性单位根检验———B&M检验及CH检验,可知其并不是在所有的季节频率均存在单位根,而仅在零频率和季节频率π存在单位根。同时进行过度差分分析,发现在对中国出口月度数据进行正确地建模或特征分析前,应该运用(1-L2)而不能用(1-L12)对其进行过滤,否则会过度差分。
China's monthly export data are very important. There is seasonality in the monthly data. By employing B&M and CH test on the data, it is found that the data didn't have unit roots at all seasonal frequency but frequency zero and π. Therefore, it should be applied to filter the China's export data before modeling and analyzing to prevent over difference.
出处
《厦门理工学院学报》
2006年第1期51-54,共4页
Journal of Xiamen University of Technology