I estimated an ARDL model and found evidence of cointegration . Altough the Bounds statistic is exceptionally high and the cointegrating equation gave meaningful cefficients, all the coefficients except the constant are insignificant as their p-values are much higher than 0.10 above 0.7.

Granger causalitys test using the Toda Yamamoto model with the same variables are also meaningful as expected according to economic theory.

Any explanations or suggestions? Thanks