John
2018-Aug-21 08:48 UTC
[R] Time series analysis: Granger causality with error-correction term
Hi, Which package/function do you recommend for Granger causality between x and y with an error correction term? In my problem, economic theory maintains that x~ I(1); y~I(1), x-y ~I(0) \begin{eqnarray} \Delta x_t = g_0 + \lambda_{x}(x_{t-1}-y_{t-1})+\sum_{k=1}^{n}g_{1k}\Delta x_{t-s}+\sum_{k=1}^{n}g_{2k}\Delta y_{t-s}+\epsilon_{1t} \\ \Delta y_t = g_0 + \lambda_{y}(x_{t-1}-y_{t-1})+\sum_{k=1}^{n}g_{3k}\Delta x_{t-s}+\sum_{k=1}^{n}g_{4k}\Delta y_{t-s}+\epsilon_{2t} \end{eqnarray} I am not sure if it would work if I run Granger causality between \Delta x and \Delta y, treating the error correction term (x_{t-1}-y_{t-1}) and exogenous variable. Thank you very much!! John [[alternative HTML version deleted]]