Analyzing I(2) Systems by Transformed Vector Autoregressions

We characterize the restrictions imposed by the minimal I(2)-to-I(1) transformation that underlies much applied work, e.g. on money demand relationships or open-economy pricing relationships. The relationship between the parameters of the original I(2) vector autoregression, including the coefficients of polynomially cointegrating relationships, and the transformed I(1) model is characterized. We discuss estimation of the transformed model subject to restrictions as well as the more commonly used approach of unrestricted reduced rank regression. Only a minor loss of efficiency is incurred by ignoring the restrictions in the empirical example and a simulation study. A properly transformed vector autoregression thus provides a practical and effective means for inference on the parameters of the I(2) model.