New Simple Tests for Panel Cointegration
提出了两种新的基于残差的面板数据检验方法，用于检验无协整的原假设，无需修正数据的时间依赖性，且能容纳个体特定的短期动态、截距、趋势和斜率参数，蒙特卡洛模拟表明小样本下渐近性质良好。
In this paper, two new simple residual-based panel data tests are proposed for the null of no cointegration. The tests are simple because they do not require any correction for the temporal dependencies of the data. Yet they are able to accommodate individual specific short-run dynamics, individual specific intercept and trend terms, and individual specific slope parameters. The limiting distributions of the tests are derived and are shown to be free of nuisance parameters. The Monte Carlo results in this paper suggest that the asymptotic results are borne out well even in very small samples.