Econometrics help in modelling money demand?

I am current working on modelling money demand, showing that it depends on interest rates, gdp and the GDP defaltor. i worked out the short run model. My question is regarding the static long run solution. When i do a unit root test for cointigration, i reject the null and a unit root is not found, so it also mean that i cannot add an ECM term to my regression. Is that a problem, Is an ECM term needed for the Long run model. (PC Give used.)