Testing Panel Cointegration with Unobservable Dynamic Common Factors

The paper proposes statistics to test the null hypothesis of no cointegration in panel data when common factors drive the cross sectional dependence. We consider both the case in which regressors are independent of the common factors and the case in which regressors are correlated with the common factors. The proposed test statistics have limiting distributions that are independent of the common factors, making it possible to pool the individual statistics. Simulations show that the proposed procedures have good finite sample properties.