Statistic | Value | Z-value | P-value | Bootstrapped p-value |
---|
G
Ï„
| −4.464 | −8.164 | 0.0000 | 0.003 |
G
α
| −26.554 | −6.762 | 0.0000 | 0.003 |
P
Ï„
| −14.398 | −8.825 | 0.0000 | 0.013 |
P
α
| −34.35 | −13.109 | 0.0000 | 0.003 |
- Note: All tests are implemented with the constant and trend in the test regression. For semi-parametric corrections, Newey and West (1987) developed the Bartlett kernel is employed. The lags and leads in the error-correction test are chosen according to the Akaike information criterion developed by Akaike (1973). All other bandwidth and lag orders are set according to the rule 4(T/100) 2/9 ≈ 1. The null hypothesis of the cointegration tests are no-cointegration. The p-values are for a one-sided test based on the normal distribution. The number of bootstraps to obtain bootstrap p-values, which are robust against cross-sectional dependences, is set to 400