中文摘要：

Groen和Kleibergen（2003）基于综列数据的误差纠正模型（PVECM）和极大似然估计方法，提出了与时间序列中Johanson（1991）协整检验类似的综列协整检验。其似然比检验统计量的极限分布是维纳过程的泛函，所以其临界值需要通过仿真试验来计算，但Groen和Kleibergen（2003）既没有给出具体的临界值也没有给出具体的程序。本文通过对维纳过程的随机积分进行仿真，估计基于PVECM的综列协整检验的临界值。与现有的部分临界值的比较结果显示，我们的计算程序是正确的，计算结果是可靠的。

英文摘要：

Similar to cointegration test in time series of Johansen （1991）, Groen and Kleibergen （2003） proposed likelihood ratio statistics to test the cointegration rank in panels based on the panel vector error correction model （PVECM） and its maximum likelihood estimators. The limit distributions of the test statistics are the functional of wiener process, so the critical values can only be obtained by simulation. Neither the critical values nor its estimation procedure were proposed in paper of Groen and Kleibergen （2003） . In this paper, we estimated the critical values of PVECM - based panel cointegration by a set of simulation. Comparing our result with the existing critical values, it is showed that our estimation procedure is correct and the proposed result is accurate.