EViews.com

I have estimated the cointegration between economic growth and differents measures of finanical development using Pedroni’s Panel Cointegration Test. the results turn to be significant, however, the cofficients are negative. I checked differents studies that have used Pedroni’s Panel Cointegration Test and found that their results are also negative.

Can anyine explain why the results are negative (cofficients). here is my resuts:


SMC VT TR
Panel v-Statistic -3.84*** -3.86*** -3.86***
Panel rho-Statistic -9.24*** -8.37*** -8.84***
Panel PP-Statistic -34.25*** -36.67*** -36.05***
Panel ADF-Statistic -19.11*** -19.24*** -17.79***
Group rho-Statistic -6.88*** -6.21*** -6.62***
Group PP-Statistic -37.63*** -38.20*** -39.83***
Group ADF-Statistic -20.81*** -19.33*** -18.69***

While I'd have to look at the data to be certain that I understand what is going on in your particular case, you should note that most of the Pedroni statistics follow a normal distribution and diverge to negative infinity under the alternative hypothesis (Pedroni, 1999, "Critical Values for Cointegration Tests in Heterogeneous Panels with Multiple Regressors," Oxford Bulletin of Economics and Statistics, 653-670, see in particular p. 668).

thanks