Hi All, I have a time series y_t and 2 other time series x1_t and x2t as regressors. I know that these 3 series are cointegrated via the Johansen tests. Hence I want to implement an error correction model with 1 lag for each variable (i.e. Lag y, lag x1 and lag x2) for projection purposes (suppose, I have future values for the regressors). Is there an R function I can use for this ECM model usage? Note that manually it is a bit challenging to pull off because the RHS of this model would have a lagged residual term, for which we have no future observations. Any help here would be appreciated. Regards, Preetam [[alternative HTML version deleted]]