A Comparison of Johansen's, Bierens and the Subspace Algorithm Method for Cointegration Analysis

The methods listed in the title are compared by means of a simulation study and a real world application. The aspects compared in the simulations are: The performance of the tests of the different methods for the dimension of the cointegrating space and the quality of the estimated cointegrating space. It turns out that the subspace algorithm method, formulated in the state space framework and thus applicable for ARMA processes, performs at least comparable to the Johansen procedure and both perform significantly better than Bierens' method. The real world application is an investigation of the long-run properties of the neoclassical growth model for Austria. It turns out that the results do not fully support the theoretical predictions and that they are very versatile across the employed methods. The degree of versatility depends strongly upon the number of variables. For the case of 6 variables and about 100 observations huge differences occur, which lead us to conclude that the results of this typical situation in the applied literature should be interpreted with more caution than is commonly done.