AUTOREGRESSIVE TRANSFORMATIONS IN COINTEGRATED REGRESSIONS

Standard autocorrelation corrections applied to cointegrating regressi ons can lead to erroneous first-differencing. Such outcomes are shown to be possible under a range of environments, including cases with aut ocorrelation coefficients substantially less than 1. First-differencin g of a cointegrating regression results in estimates that may bear lit tle relation to the parameters in the original untransformed relation, resulting in misinterpretation of the parameter estimates. These resu lts are proved analytically and demonstrated with simulations and empi rical examples.